diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,188333 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 12.505811250581125, + "eval_steps": 500, + "global_step": 26900, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00046490004649000463, + "grad_norm": Infinity, + "learning_rate": 1e-05, + "loss": 4.0094, + "step": 1 + }, + { + "epoch": 0.0009298000929800093, + "grad_norm": 35.45524215698242, + "learning_rate": 9.9999989999999e-06, + "loss": 4.1648, + "step": 2 + }, + { + "epoch": 0.001394700139470014, + "grad_norm": 24.217166900634766, + "learning_rate": 9.9999979999998e-06, + "loss": 3.6658, + "step": 3 + }, + { + "epoch": 0.0018596001859600185, + "grad_norm": 14.564384460449219, + "learning_rate": 9.999996999999701e-06, + "loss": 2.0655, + "step": 4 + }, + { + "epoch": 0.0023245002324500234, + "grad_norm": 22.69341468811035, + "learning_rate": 9.999995999999602e-06, + "loss": 3.0669, + "step": 5 + }, + { + "epoch": 0.002789400278940028, + "grad_norm": 14.560563087463379, + "learning_rate": 9.9999949999995e-06, + "loss": 1.8129, + "step": 6 + }, + { + "epoch": 0.0032543003254300326, + "grad_norm": 19.024784088134766, + "learning_rate": 9.999993999999401e-06, + "loss": 3.2103, + "step": 7 + }, + { + "epoch": 0.003719200371920037, + "grad_norm": 19.95561408996582, + "learning_rate": 9.9999929999993e-06, + "loss": 3.3367, + "step": 8 + }, + { + "epoch": 0.0041841004184100415, + "grad_norm": 12.347963333129883, + "learning_rate": 9.999991999999201e-06, + "loss": 1.6132, + "step": 9 + }, + { + "epoch": 0.004649000464900047, + "grad_norm": 15.344433784484863, + "learning_rate": 9.9999909999991e-06, + "loss": 2.253, + "step": 10 + }, + { + "epoch": 0.005113900511390051, + "grad_norm": 12.263849258422852, + "learning_rate": 9.999989999999e-06, + "loss": 1.8346, + "step": 11 + }, + { + "epoch": 0.005578800557880056, + "grad_norm": 18.816858291625977, + "learning_rate": 9.999988999998901e-06, + "loss": 2.9528, + "step": 12 + }, + { + "epoch": 0.00604370060437006, + "grad_norm": 19.531362533569336, + "learning_rate": 9.9999879999988e-06, + "loss": 2.0103, + "step": 13 + }, + { + "epoch": 0.006508600650860065, + "grad_norm": 12.620562553405762, + "learning_rate": 9.999986999998701e-06, + "loss": 1.2916, + "step": 14 + }, + { + "epoch": 0.00697350069735007, + "grad_norm": 10.75390911102295, + "learning_rate": 9.9999859999986e-06, + "loss": 1.2119, + "step": 15 + }, + { + "epoch": 0.007438400743840074, + "grad_norm": 13.622138977050781, + "learning_rate": 9.9999849999985e-06, + "loss": 1.4273, + "step": 16 + }, + { + "epoch": 0.00790330079033008, + "grad_norm": 9.720497131347656, + "learning_rate": 9.999983999998401e-06, + "loss": 1.1622, + "step": 17 + }, + { + "epoch": 0.008368200836820083, + "grad_norm": 12.709471702575684, + "learning_rate": 9.999982999998302e-06, + "loss": 1.9274, + "step": 18 + }, + { + "epoch": 0.008833100883310088, + "grad_norm": 12.005471229553223, + "learning_rate": 9.9999819999982e-06, + "loss": 1.5394, + "step": 19 + }, + { + "epoch": 0.009298000929800094, + "grad_norm": 14.935007095336914, + "learning_rate": 9.999980999998101e-06, + "loss": 1.5807, + "step": 20 + }, + { + "epoch": 0.009762900976290097, + "grad_norm": 10.8558931350708, + "learning_rate": 9.999979999998e-06, + "loss": 1.5113, + "step": 21 + }, + { + "epoch": 0.010227801022780102, + "grad_norm": 10.687922477722168, + "learning_rate": 9.999978999997901e-06, + "loss": 1.3766, + "step": 22 + }, + { + "epoch": 0.010692701069270108, + "grad_norm": 13.388243675231934, + "learning_rate": 9.9999779999978e-06, + "loss": 1.6063, + "step": 23 + }, + { + "epoch": 0.011157601115760111, + "grad_norm": 16.584619522094727, + "learning_rate": 9.9999769999977e-06, + "loss": 1.2262, + "step": 24 + }, + { + "epoch": 0.011622501162250116, + "grad_norm": 10.998594284057617, + "learning_rate": 9.999975999997601e-06, + "loss": 1.4316, + "step": 25 + }, + { + "epoch": 0.01208740120874012, + "grad_norm": 15.171038627624512, + "learning_rate": 9.9999749999975e-06, + "loss": 2.251, + "step": 26 + }, + { + "epoch": 0.012552301255230125, + "grad_norm": 14.543960571289062, + "learning_rate": 9.999973999997401e-06, + "loss": 2.0242, + "step": 27 + }, + { + "epoch": 0.01301720130172013, + "grad_norm": 9.565631866455078, + "learning_rate": 9.9999729999973e-06, + "loss": 0.9914, + "step": 28 + }, + { + "epoch": 0.013482101348210134, + "grad_norm": 10.753483772277832, + "learning_rate": 9.9999719999972e-06, + "loss": 1.1623, + "step": 29 + }, + { + "epoch": 0.01394700139470014, + "grad_norm": 13.203991889953613, + "learning_rate": 9.999970999997101e-06, + "loss": 1.1862, + "step": 30 + }, + { + "epoch": 0.014411901441190145, + "grad_norm": 10.397066116333008, + "learning_rate": 9.999969999997002e-06, + "loss": 1.0728, + "step": 31 + }, + { + "epoch": 0.014876801487680148, + "grad_norm": 8.681368827819824, + "learning_rate": 9.999968999996901e-06, + "loss": 0.9491, + "step": 32 + }, + { + "epoch": 0.015341701534170154, + "grad_norm": 13.763985633850098, + "learning_rate": 9.999967999996802e-06, + "loss": 1.5084, + "step": 33 + }, + { + "epoch": 0.01580660158066016, + "grad_norm": 14.60959529876709, + "learning_rate": 9.9999669999967e-06, + "loss": 1.6239, + "step": 34 + }, + { + "epoch": 0.016271501627150162, + "grad_norm": 10.305540084838867, + "learning_rate": 9.999965999996601e-06, + "loss": 1.1146, + "step": 35 + }, + { + "epoch": 0.016736401673640166, + "grad_norm": 10.438726425170898, + "learning_rate": 9.9999649999965e-06, + "loss": 1.0339, + "step": 36 + }, + { + "epoch": 0.017201301720130173, + "grad_norm": 7.2965803146362305, + "learning_rate": 9.9999639999964e-06, + "loss": 0.8246, + "step": 37 + }, + { + "epoch": 0.017666201766620176, + "grad_norm": 9.115209579467773, + "learning_rate": 9.999962999996301e-06, + "loss": 0.8741, + "step": 38 + }, + { + "epoch": 0.01813110181311018, + "grad_norm": 14.418665885925293, + "learning_rate": 9.9999619999962e-06, + "loss": 1.4621, + "step": 39 + }, + { + "epoch": 0.018596001859600187, + "grad_norm": 13.113110542297363, + "learning_rate": 9.999960999996101e-06, + "loss": 1.3376, + "step": 40 + }, + { + "epoch": 0.01906090190609019, + "grad_norm": 9.022974014282227, + "learning_rate": 9.999959999996e-06, + "loss": 1.2214, + "step": 41 + }, + { + "epoch": 0.019525801952580194, + "grad_norm": 11.54273796081543, + "learning_rate": 9.9999589999959e-06, + "loss": 0.8604, + "step": 42 + }, + { + "epoch": 0.0199907019990702, + "grad_norm": 11.838022232055664, + "learning_rate": 9.999957999995801e-06, + "loss": 1.2483, + "step": 43 + }, + { + "epoch": 0.020455602045560205, + "grad_norm": 11.613022804260254, + "learning_rate": 9.999956999995702e-06, + "loss": 1.2737, + "step": 44 + }, + { + "epoch": 0.02092050209205021, + "grad_norm": 15.26848030090332, + "learning_rate": 9.999955999995601e-06, + "loss": 1.7873, + "step": 45 + }, + { + "epoch": 0.021385402138540215, + "grad_norm": 7.977863788604736, + "learning_rate": 9.9999549999955e-06, + "loss": 0.8655, + "step": 46 + }, + { + "epoch": 0.02185030218503022, + "grad_norm": 15.820752143859863, + "learning_rate": 9.9999539999954e-06, + "loss": 1.2751, + "step": 47 + }, + { + "epoch": 0.022315202231520222, + "grad_norm": 10.579505920410156, + "learning_rate": 9.999952999995301e-06, + "loss": 1.2907, + "step": 48 + }, + { + "epoch": 0.02278010227801023, + "grad_norm": 10.590607643127441, + "learning_rate": 9.999951999995202e-06, + "loss": 1.174, + "step": 49 + }, + { + "epoch": 0.023245002324500233, + "grad_norm": 12.608613014221191, + "learning_rate": 9.999950999995101e-06, + "loss": 1.1209, + "step": 50 + }, + { + "epoch": 0.023709902370990237, + "grad_norm": 13.882878303527832, + "learning_rate": 9.999949999995002e-06, + "loss": 1.4178, + "step": 51 + }, + { + "epoch": 0.02417480241748024, + "grad_norm": 11.052614212036133, + "learning_rate": 9.9999489999949e-06, + "loss": 1.1266, + "step": 52 + }, + { + "epoch": 0.024639702463970247, + "grad_norm": 9.191442489624023, + "learning_rate": 9.999947999994801e-06, + "loss": 0.9098, + "step": 53 + }, + { + "epoch": 0.02510460251046025, + "grad_norm": 12.569636344909668, + "learning_rate": 9.9999469999947e-06, + "loss": 1.4742, + "step": 54 + }, + { + "epoch": 0.025569502556950254, + "grad_norm": 9.876986503601074, + "learning_rate": 9.9999459999946e-06, + "loss": 1.3735, + "step": 55 + }, + { + "epoch": 0.02603440260344026, + "grad_norm": 9.052992820739746, + "learning_rate": 9.999944999994501e-06, + "loss": 1.1767, + "step": 56 + }, + { + "epoch": 0.026499302649930265, + "grad_norm": 10.942220687866211, + "learning_rate": 9.9999439999944e-06, + "loss": 1.0153, + "step": 57 + }, + { + "epoch": 0.02696420269642027, + "grad_norm": 9.708091735839844, + "learning_rate": 9.999942999994301e-06, + "loss": 1.0748, + "step": 58 + }, + { + "epoch": 0.027429102742910275, + "grad_norm": 13.740374565124512, + "learning_rate": 9.9999419999942e-06, + "loss": 1.5147, + "step": 59 + }, + { + "epoch": 0.02789400278940028, + "grad_norm": 10.680349349975586, + "learning_rate": 9.9999409999941e-06, + "loss": 1.3974, + "step": 60 + }, + { + "epoch": 0.028358902835890282, + "grad_norm": 7.206218242645264, + "learning_rate": 9.999939999994001e-06, + "loss": 0.7813, + "step": 61 + }, + { + "epoch": 0.02882380288238029, + "grad_norm": 8.180681228637695, + "learning_rate": 9.999938999993902e-06, + "loss": 1.0207, + "step": 62 + }, + { + "epoch": 0.029288702928870293, + "grad_norm": 12.700065612792969, + "learning_rate": 9.999937999993801e-06, + "loss": 1.3614, + "step": 63 + }, + { + "epoch": 0.029753602975360297, + "grad_norm": 11.540399551391602, + "learning_rate": 9.999936999993702e-06, + "loss": 1.1768, + "step": 64 + }, + { + "epoch": 0.030218503021850304, + "grad_norm": 12.388829231262207, + "learning_rate": 9.9999359999936e-06, + "loss": 0.748, + "step": 65 + }, + { + "epoch": 0.030683403068340307, + "grad_norm": 10.886833190917969, + "learning_rate": 9.999934999993501e-06, + "loss": 1.2569, + "step": 66 + }, + { + "epoch": 0.03114830311483031, + "grad_norm": 12.890376091003418, + "learning_rate": 9.9999339999934e-06, + "loss": 1.1991, + "step": 67 + }, + { + "epoch": 0.03161320316132032, + "grad_norm": 10.96619987487793, + "learning_rate": 9.9999329999933e-06, + "loss": 1.1761, + "step": 68 + }, + { + "epoch": 0.03207810320781032, + "grad_norm": 10.481689453125, + "learning_rate": 9.999931999993202e-06, + "loss": 1.1553, + "step": 69 + }, + { + "epoch": 0.032543003254300325, + "grad_norm": 17.071203231811523, + "learning_rate": 9.9999309999931e-06, + "loss": 1.9027, + "step": 70 + }, + { + "epoch": 0.03300790330079033, + "grad_norm": 12.654949188232422, + "learning_rate": 9.999929999993001e-06, + "loss": 1.868, + "step": 71 + }, + { + "epoch": 0.03347280334728033, + "grad_norm": 10.395791053771973, + "learning_rate": 9.9999289999929e-06, + "loss": 1.5862, + "step": 72 + }, + { + "epoch": 0.03393770339377034, + "grad_norm": 16.348926544189453, + "learning_rate": 9.9999279999928e-06, + "loss": 1.5928, + "step": 73 + }, + { + "epoch": 0.034402603440260346, + "grad_norm": 13.96761703491211, + "learning_rate": 9.999926999992701e-06, + "loss": 0.9602, + "step": 74 + }, + { + "epoch": 0.03486750348675035, + "grad_norm": 10.813905715942383, + "learning_rate": 9.999925999992602e-06, + "loss": 1.1324, + "step": 75 + }, + { + "epoch": 0.03533240353324035, + "grad_norm": 10.727300643920898, + "learning_rate": 9.999924999992501e-06, + "loss": 0.9213, + "step": 76 + }, + { + "epoch": 0.03579730357973036, + "grad_norm": 8.690640449523926, + "learning_rate": 9.9999239999924e-06, + "loss": 1.0409, + "step": 77 + }, + { + "epoch": 0.03626220362622036, + "grad_norm": 8.851780891418457, + "learning_rate": 9.9999229999923e-06, + "loss": 0.7145, + "step": 78 + }, + { + "epoch": 0.03672710367271037, + "grad_norm": 11.941566467285156, + "learning_rate": 9.9999219999922e-06, + "loss": 1.0015, + "step": 79 + }, + { + "epoch": 0.037192003719200374, + "grad_norm": 13.13437271118164, + "learning_rate": 9.9999209999921e-06, + "loss": 1.1631, + "step": 80 + }, + { + "epoch": 0.03765690376569038, + "grad_norm": 7.366654872894287, + "learning_rate": 9.999919999992001e-06, + "loss": 0.5604, + "step": 81 + }, + { + "epoch": 0.03812180381218038, + "grad_norm": 12.891462326049805, + "learning_rate": 9.999918999991902e-06, + "loss": 1.256, + "step": 82 + }, + { + "epoch": 0.038586703858670385, + "grad_norm": 15.46192455291748, + "learning_rate": 9.9999179999918e-06, + "loss": 1.5319, + "step": 83 + }, + { + "epoch": 0.03905160390516039, + "grad_norm": 8.693193435668945, + "learning_rate": 9.999916999991701e-06, + "loss": 0.7045, + "step": 84 + }, + { + "epoch": 0.03951650395165039, + "grad_norm": 9.856704711914062, + "learning_rate": 9.9999159999916e-06, + "loss": 1.0581, + "step": 85 + }, + { + "epoch": 0.0399814039981404, + "grad_norm": 11.445311546325684, + "learning_rate": 9.9999149999915e-06, + "loss": 0.9341, + "step": 86 + }, + { + "epoch": 0.040446304044630406, + "grad_norm": 13.168042182922363, + "learning_rate": 9.999913999991401e-06, + "loss": 1.7574, + "step": 87 + }, + { + "epoch": 0.04091120409112041, + "grad_norm": 10.24945068359375, + "learning_rate": 9.9999129999913e-06, + "loss": 0.8866, + "step": 88 + }, + { + "epoch": 0.04137610413761041, + "grad_norm": 10.857465744018555, + "learning_rate": 9.999911999991201e-06, + "loss": 1.5692, + "step": 89 + }, + { + "epoch": 0.04184100418410042, + "grad_norm": 8.043157577514648, + "learning_rate": 9.9999109999911e-06, + "loss": 0.8113, + "step": 90 + }, + { + "epoch": 0.04230590423059042, + "grad_norm": 8.79730224609375, + "learning_rate": 9.999909999991e-06, + "loss": 0.8612, + "step": 91 + }, + { + "epoch": 0.04277080427708043, + "grad_norm": 10.052811622619629, + "learning_rate": 9.9999089999909e-06, + "loss": 1.2395, + "step": 92 + }, + { + "epoch": 0.043235704323570434, + "grad_norm": 14.225953102111816, + "learning_rate": 9.9999079999908e-06, + "loss": 1.356, + "step": 93 + }, + { + "epoch": 0.04370060437006044, + "grad_norm": 10.357001304626465, + "learning_rate": 9.999906999990701e-06, + "loss": 0.7332, + "step": 94 + }, + { + "epoch": 0.04416550441655044, + "grad_norm": 12.130985260009766, + "learning_rate": 9.999905999990602e-06, + "loss": 1.0922, + "step": 95 + }, + { + "epoch": 0.044630404463040445, + "grad_norm": 11.641082763671875, + "learning_rate": 9.9999049999905e-06, + "loss": 0.7133, + "step": 96 + }, + { + "epoch": 0.04509530450953045, + "grad_norm": 7.882996082305908, + "learning_rate": 9.999903999990401e-06, + "loss": 0.7548, + "step": 97 + }, + { + "epoch": 0.04556020455602046, + "grad_norm": 11.883736610412598, + "learning_rate": 9.9999029999903e-06, + "loss": 0.8384, + "step": 98 + }, + { + "epoch": 0.04602510460251046, + "grad_norm": 10.262872695922852, + "learning_rate": 9.999901999990201e-06, + "loss": 0.6792, + "step": 99 + }, + { + "epoch": 0.046490004649000466, + "grad_norm": 9.975241661071777, + "learning_rate": 9.999900999990102e-06, + "loss": 1.3343, + "step": 100 + }, + { + "epoch": 0.04695490469549047, + "grad_norm": 15.257561683654785, + "learning_rate": 9.99989999999e-06, + "loss": 1.2874, + "step": 101 + }, + { + "epoch": 0.04741980474198047, + "grad_norm": 13.679219245910645, + "learning_rate": 9.999898999989901e-06, + "loss": 1.4908, + "step": 102 + }, + { + "epoch": 0.04788470478847048, + "grad_norm": 13.38089370727539, + "learning_rate": 9.9998979999898e-06, + "loss": 1.3084, + "step": 103 + }, + { + "epoch": 0.04834960483496048, + "grad_norm": 9.568513870239258, + "learning_rate": 9.9998969999897e-06, + "loss": 1.0463, + "step": 104 + }, + { + "epoch": 0.04881450488145049, + "grad_norm": 10.715933799743652, + "learning_rate": 9.9998959999896e-06, + "loss": 0.9424, + "step": 105 + }, + { + "epoch": 0.049279404927940494, + "grad_norm": 7.810553550720215, + "learning_rate": 9.9998949999895e-06, + "loss": 0.7621, + "step": 106 + }, + { + "epoch": 0.0497443049744305, + "grad_norm": 10.991146087646484, + "learning_rate": 9.999893999989401e-06, + "loss": 0.9051, + "step": 107 + }, + { + "epoch": 0.0502092050209205, + "grad_norm": 7.645435810089111, + "learning_rate": 9.999892999989302e-06, + "loss": 0.6761, + "step": 108 + }, + { + "epoch": 0.050674105067410505, + "grad_norm": 8.872024536132812, + "learning_rate": 9.9998919999892e-06, + "loss": 0.6606, + "step": 109 + }, + { + "epoch": 0.05113900511390051, + "grad_norm": 7.798163414001465, + "learning_rate": 9.9998909999891e-06, + "loss": 0.8765, + "step": 110 + }, + { + "epoch": 0.05160390516039052, + "grad_norm": 9.771446228027344, + "learning_rate": 9.999889999989e-06, + "loss": 1.1118, + "step": 111 + }, + { + "epoch": 0.05206880520688052, + "grad_norm": 8.470248222351074, + "learning_rate": 9.999888999988901e-06, + "loss": 0.6983, + "step": 112 + }, + { + "epoch": 0.052533705253370526, + "grad_norm": 10.249551773071289, + "learning_rate": 9.999887999988802e-06, + "loss": 1.3984, + "step": 113 + }, + { + "epoch": 0.05299860529986053, + "grad_norm": 11.51451587677002, + "learning_rate": 9.9998869999887e-06, + "loss": 1.1831, + "step": 114 + }, + { + "epoch": 0.05346350534635053, + "grad_norm": 9.098623275756836, + "learning_rate": 9.999885999988601e-06, + "loss": 0.7726, + "step": 115 + }, + { + "epoch": 0.05392840539284054, + "grad_norm": 14.46179485321045, + "learning_rate": 9.9998849999885e-06, + "loss": 1.1952, + "step": 116 + }, + { + "epoch": 0.05439330543933055, + "grad_norm": 7.850644588470459, + "learning_rate": 9.999883999988401e-06, + "loss": 0.9409, + "step": 117 + }, + { + "epoch": 0.05485820548582055, + "grad_norm": 5.6950578689575195, + "learning_rate": 9.999882999988302e-06, + "loss": 0.494, + "step": 118 + }, + { + "epoch": 0.055323105532310554, + "grad_norm": 16.656362533569336, + "learning_rate": 9.9998819999882e-06, + "loss": 1.1345, + "step": 119 + }, + { + "epoch": 0.05578800557880056, + "grad_norm": 11.794100761413574, + "learning_rate": 9.999880999988101e-06, + "loss": 1.1309, + "step": 120 + }, + { + "epoch": 0.05625290562529056, + "grad_norm": 8.87034797668457, + "learning_rate": 9.999879999988e-06, + "loss": 0.7594, + "step": 121 + }, + { + "epoch": 0.056717805671780565, + "grad_norm": 12.082301139831543, + "learning_rate": 9.9998789999879e-06, + "loss": 1.2411, + "step": 122 + }, + { + "epoch": 0.05718270571827057, + "grad_norm": 8.33902645111084, + "learning_rate": 9.9998779999878e-06, + "loss": 1.0694, + "step": 123 + }, + { + "epoch": 0.05764760576476058, + "grad_norm": 7.467475891113281, + "learning_rate": 9.9998769999877e-06, + "loss": 0.5728, + "step": 124 + }, + { + "epoch": 0.05811250581125058, + "grad_norm": 11.401518821716309, + "learning_rate": 9.999875999987601e-06, + "loss": 1.0002, + "step": 125 + }, + { + "epoch": 0.058577405857740586, + "grad_norm": 15.05356216430664, + "learning_rate": 9.999874999987502e-06, + "loss": 1.5181, + "step": 126 + }, + { + "epoch": 0.05904230590423059, + "grad_norm": 8.834050178527832, + "learning_rate": 9.9998739999874e-06, + "loss": 0.5875, + "step": 127 + }, + { + "epoch": 0.05950720595072059, + "grad_norm": 13.168805122375488, + "learning_rate": 9.999872999987301e-06, + "loss": 0.9663, + "step": 128 + }, + { + "epoch": 0.0599721059972106, + "grad_norm": 15.627230644226074, + "learning_rate": 9.9998719999872e-06, + "loss": 1.3999, + "step": 129 + }, + { + "epoch": 0.06043700604370061, + "grad_norm": 15.754984855651855, + "learning_rate": 9.999870999987101e-06, + "loss": 1.8077, + "step": 130 + }, + { + "epoch": 0.06090190609019061, + "grad_norm": 8.332152366638184, + "learning_rate": 9.999869999987002e-06, + "loss": 0.7806, + "step": 131 + }, + { + "epoch": 0.061366806136680614, + "grad_norm": 9.220961570739746, + "learning_rate": 9.9998689999869e-06, + "loss": 1.3568, + "step": 132 + }, + { + "epoch": 0.06183170618317062, + "grad_norm": 5.494656085968018, + "learning_rate": 9.999867999986801e-06, + "loss": 0.5627, + "step": 133 + }, + { + "epoch": 0.06229660622966062, + "grad_norm": 10.039477348327637, + "learning_rate": 9.9998669999867e-06, + "loss": 1.2287, + "step": 134 + }, + { + "epoch": 0.06276150627615062, + "grad_norm": 9.080599784851074, + "learning_rate": 9.999865999986601e-06, + "loss": 0.7894, + "step": 135 + }, + { + "epoch": 0.06322640632264064, + "grad_norm": 12.161409378051758, + "learning_rate": 9.9998649999865e-06, + "loss": 1.6079, + "step": 136 + }, + { + "epoch": 0.06369130636913063, + "grad_norm": 7.488203525543213, + "learning_rate": 9.9998639999864e-06, + "loss": 0.9778, + "step": 137 + }, + { + "epoch": 0.06415620641562064, + "grad_norm": 10.444978713989258, + "learning_rate": 9.999862999986301e-06, + "loss": 1.0903, + "step": 138 + }, + { + "epoch": 0.06462110646211065, + "grad_norm": 9.332356452941895, + "learning_rate": 9.999861999986202e-06, + "loss": 0.8013, + "step": 139 + }, + { + "epoch": 0.06508600650860065, + "grad_norm": 8.156571388244629, + "learning_rate": 9.9998609999861e-06, + "loss": 0.6535, + "step": 140 + }, + { + "epoch": 0.06555090655509066, + "grad_norm": 10.224624633789062, + "learning_rate": 9.999859999986001e-06, + "loss": 1.1352, + "step": 141 + }, + { + "epoch": 0.06601580660158066, + "grad_norm": 13.111985206604004, + "learning_rate": 9.9998589999859e-06, + "loss": 1.1107, + "step": 142 + }, + { + "epoch": 0.06648070664807067, + "grad_norm": 10.90305233001709, + "learning_rate": 9.999857999985801e-06, + "loss": 1.0072, + "step": 143 + }, + { + "epoch": 0.06694560669456066, + "grad_norm": 11.907021522521973, + "learning_rate": 9.999856999985702e-06, + "loss": 0.8235, + "step": 144 + }, + { + "epoch": 0.06741050674105067, + "grad_norm": 11.470433235168457, + "learning_rate": 9.9998559999856e-06, + "loss": 1.4135, + "step": 145 + }, + { + "epoch": 0.06787540678754068, + "grad_norm": 7.618493556976318, + "learning_rate": 9.999854999985501e-06, + "loss": 0.8411, + "step": 146 + }, + { + "epoch": 0.06834030683403068, + "grad_norm": 9.48288631439209, + "learning_rate": 9.9998539999854e-06, + "loss": 1.0285, + "step": 147 + }, + { + "epoch": 0.06880520688052069, + "grad_norm": 10.986167907714844, + "learning_rate": 9.999852999985301e-06, + "loss": 0.7382, + "step": 148 + }, + { + "epoch": 0.06927010692701069, + "grad_norm": 10.433220863342285, + "learning_rate": 9.9998519999852e-06, + "loss": 1.0261, + "step": 149 + }, + { + "epoch": 0.0697350069735007, + "grad_norm": 9.618155479431152, + "learning_rate": 9.9998509999851e-06, + "loss": 1.3492, + "step": 150 + }, + { + "epoch": 0.0701999070199907, + "grad_norm": 11.387544631958008, + "learning_rate": 9.999849999985001e-06, + "loss": 1.1918, + "step": 151 + }, + { + "epoch": 0.0706648070664807, + "grad_norm": 5.403172016143799, + "learning_rate": 9.999848999984902e-06, + "loss": 0.4569, + "step": 152 + }, + { + "epoch": 0.07112970711297072, + "grad_norm": 9.911691665649414, + "learning_rate": 9.999847999984801e-06, + "loss": 0.8186, + "step": 153 + }, + { + "epoch": 0.07159460715946071, + "grad_norm": 8.923487663269043, + "learning_rate": 9.9998469999847e-06, + "loss": 0.9135, + "step": 154 + }, + { + "epoch": 0.07205950720595072, + "grad_norm": 13.906689643859863, + "learning_rate": 9.9998459999846e-06, + "loss": 1.9202, + "step": 155 + }, + { + "epoch": 0.07252440725244072, + "grad_norm": 7.307963848114014, + "learning_rate": 9.999844999984501e-06, + "loss": 0.5988, + "step": 156 + }, + { + "epoch": 0.07298930729893073, + "grad_norm": 11.176424980163574, + "learning_rate": 9.999843999984402e-06, + "loss": 1.183, + "step": 157 + }, + { + "epoch": 0.07345420734542074, + "grad_norm": 9.57575798034668, + "learning_rate": 9.9998429999843e-06, + "loss": 0.989, + "step": 158 + }, + { + "epoch": 0.07391910739191074, + "grad_norm": 11.591389656066895, + "learning_rate": 9.999841999984201e-06, + "loss": 1.3599, + "step": 159 + }, + { + "epoch": 0.07438400743840075, + "grad_norm": 14.514925956726074, + "learning_rate": 9.9998409999841e-06, + "loss": 1.2583, + "step": 160 + }, + { + "epoch": 0.07484890748489074, + "grad_norm": 8.144255638122559, + "learning_rate": 9.999839999984001e-06, + "loss": 0.8153, + "step": 161 + }, + { + "epoch": 0.07531380753138076, + "grad_norm": 12.214879035949707, + "learning_rate": 9.9998389999839e-06, + "loss": 1.4715, + "step": 162 + }, + { + "epoch": 0.07577870757787075, + "grad_norm": 11.226256370544434, + "learning_rate": 9.9998379999838e-06, + "loss": 1.0901, + "step": 163 + }, + { + "epoch": 0.07624360762436076, + "grad_norm": 11.073488235473633, + "learning_rate": 9.999836999983701e-06, + "loss": 1.0421, + "step": 164 + }, + { + "epoch": 0.07670850767085077, + "grad_norm": 7.693215847015381, + "learning_rate": 9.9998359999836e-06, + "loss": 1.1481, + "step": 165 + }, + { + "epoch": 0.07717340771734077, + "grad_norm": 8.723431587219238, + "learning_rate": 9.999834999983501e-06, + "loss": 1.1096, + "step": 166 + }, + { + "epoch": 0.07763830776383078, + "grad_norm": 6.891931533813477, + "learning_rate": 9.9998339999834e-06, + "loss": 0.649, + "step": 167 + }, + { + "epoch": 0.07810320781032078, + "grad_norm": 12.294376373291016, + "learning_rate": 9.9998329999833e-06, + "loss": 1.1025, + "step": 168 + }, + { + "epoch": 0.07856810785681079, + "grad_norm": 8.875847816467285, + "learning_rate": 9.999831999983201e-06, + "loss": 0.8409, + "step": 169 + }, + { + "epoch": 0.07903300790330078, + "grad_norm": 6.122175693511963, + "learning_rate": 9.999830999983102e-06, + "loss": 0.4917, + "step": 170 + }, + { + "epoch": 0.0794979079497908, + "grad_norm": 8.687474250793457, + "learning_rate": 9.999829999983e-06, + "loss": 0.7894, + "step": 171 + }, + { + "epoch": 0.0799628079962808, + "grad_norm": 8.625711441040039, + "learning_rate": 9.999828999982901e-06, + "loss": 0.9049, + "step": 172 + }, + { + "epoch": 0.0804277080427708, + "grad_norm": 9.215112686157227, + "learning_rate": 9.9998279999828e-06, + "loss": 1.0442, + "step": 173 + }, + { + "epoch": 0.08089260808926081, + "grad_norm": 11.410584449768066, + "learning_rate": 9.999826999982701e-06, + "loss": 1.2614, + "step": 174 + }, + { + "epoch": 0.08135750813575081, + "grad_norm": 9.599403381347656, + "learning_rate": 9.9998259999826e-06, + "loss": 0.8212, + "step": 175 + }, + { + "epoch": 0.08182240818224082, + "grad_norm": 10.258275032043457, + "learning_rate": 9.9998249999825e-06, + "loss": 1.0763, + "step": 176 + }, + { + "epoch": 0.08228730822873083, + "grad_norm": 9.602012634277344, + "learning_rate": 9.999823999982401e-06, + "loss": 0.8021, + "step": 177 + }, + { + "epoch": 0.08275220827522083, + "grad_norm": 7.627445220947266, + "learning_rate": 9.9998229999823e-06, + "loss": 0.8711, + "step": 178 + }, + { + "epoch": 0.08321710832171084, + "grad_norm": 6.402169227600098, + "learning_rate": 9.999821999982201e-06, + "loss": 0.5919, + "step": 179 + }, + { + "epoch": 0.08368200836820083, + "grad_norm": 13.741555213928223, + "learning_rate": 9.9998209999821e-06, + "loss": 1.9837, + "step": 180 + }, + { + "epoch": 0.08414690841469084, + "grad_norm": 7.391029357910156, + "learning_rate": 9.999819999982e-06, + "loss": 0.6778, + "step": 181 + }, + { + "epoch": 0.08461180846118084, + "grad_norm": 5.6059393882751465, + "learning_rate": 9.999818999981901e-06, + "loss": 0.5095, + "step": 182 + }, + { + "epoch": 0.08507670850767085, + "grad_norm": 11.671865463256836, + "learning_rate": 9.999817999981802e-06, + "loss": 1.2688, + "step": 183 + }, + { + "epoch": 0.08554160855416086, + "grad_norm": 17.802217483520508, + "learning_rate": 9.999816999981701e-06, + "loss": 1.7291, + "step": 184 + }, + { + "epoch": 0.08600650860065086, + "grad_norm": 11.164816856384277, + "learning_rate": 9.999815999981602e-06, + "loss": 1.3927, + "step": 185 + }, + { + "epoch": 0.08647140864714087, + "grad_norm": 11.890761375427246, + "learning_rate": 9.9998149999815e-06, + "loss": 1.4884, + "step": 186 + }, + { + "epoch": 0.08693630869363086, + "grad_norm": 15.062190055847168, + "learning_rate": 9.9998139999814e-06, + "loss": 1.1063, + "step": 187 + }, + { + "epoch": 0.08740120874012088, + "grad_norm": 9.281411170959473, + "learning_rate": 9.9998129999813e-06, + "loss": 1.0628, + "step": 188 + }, + { + "epoch": 0.08786610878661087, + "grad_norm": 11.374409675598145, + "learning_rate": 9.9998119999812e-06, + "loss": 1.2439, + "step": 189 + }, + { + "epoch": 0.08833100883310088, + "grad_norm": 8.981118202209473, + "learning_rate": 9.999810999981101e-06, + "loss": 0.6509, + "step": 190 + }, + { + "epoch": 0.08879590887959089, + "grad_norm": 12.481189727783203, + "learning_rate": 9.999809999981e-06, + "loss": 0.9519, + "step": 191 + }, + { + "epoch": 0.08926080892608089, + "grad_norm": 10.093208312988281, + "learning_rate": 9.999808999980901e-06, + "loss": 0.8573, + "step": 192 + }, + { + "epoch": 0.0897257089725709, + "grad_norm": 9.614958763122559, + "learning_rate": 9.9998079999808e-06, + "loss": 1.0506, + "step": 193 + }, + { + "epoch": 0.0901906090190609, + "grad_norm": 7.704128265380859, + "learning_rate": 9.9998069999807e-06, + "loss": 1.0679, + "step": 194 + }, + { + "epoch": 0.09065550906555091, + "grad_norm": 6.863243579864502, + "learning_rate": 9.999805999980601e-06, + "loss": 0.8844, + "step": 195 + }, + { + "epoch": 0.09112040911204092, + "grad_norm": 8.582168579101562, + "learning_rate": 9.999804999980502e-06, + "loss": 0.7516, + "step": 196 + }, + { + "epoch": 0.09158530915853091, + "grad_norm": 8.084257125854492, + "learning_rate": 9.999803999980401e-06, + "loss": 0.945, + "step": 197 + }, + { + "epoch": 0.09205020920502092, + "grad_norm": 9.292618751525879, + "learning_rate": 9.9998029999803e-06, + "loss": 1.0231, + "step": 198 + }, + { + "epoch": 0.09251510925151092, + "grad_norm": 15.87205696105957, + "learning_rate": 9.9998019999802e-06, + "loss": 1.3823, + "step": 199 + }, + { + "epoch": 0.09298000929800093, + "grad_norm": 10.551438331604004, + "learning_rate": 9.999800999980101e-06, + "loss": 1.052, + "step": 200 + }, + { + "epoch": 0.09344490934449093, + "grad_norm": 7.624856948852539, + "learning_rate": 9.99979999998e-06, + "loss": 0.7598, + "step": 201 + }, + { + "epoch": 0.09390980939098094, + "grad_norm": 6.1972432136535645, + "learning_rate": 9.999798999979901e-06, + "loss": 0.5272, + "step": 202 + }, + { + "epoch": 0.09437470943747095, + "grad_norm": 11.139930725097656, + "learning_rate": 9.999797999979802e-06, + "loss": 1.5578, + "step": 203 + }, + { + "epoch": 0.09483960948396095, + "grad_norm": 9.089067459106445, + "learning_rate": 9.9997969999797e-06, + "loss": 1.3616, + "step": 204 + }, + { + "epoch": 0.09530450953045096, + "grad_norm": 9.065717697143555, + "learning_rate": 9.999795999979601e-06, + "loss": 0.7426, + "step": 205 + }, + { + "epoch": 0.09576940957694095, + "grad_norm": 12.454169273376465, + "learning_rate": 9.9997949999795e-06, + "loss": 1.1155, + "step": 206 + }, + { + "epoch": 0.09623430962343096, + "grad_norm": 7.735940933227539, + "learning_rate": 9.9997939999794e-06, + "loss": 0.8382, + "step": 207 + }, + { + "epoch": 0.09669920966992096, + "grad_norm": 5.378530502319336, + "learning_rate": 9.999792999979301e-06, + "loss": 0.4092, + "step": 208 + }, + { + "epoch": 0.09716410971641097, + "grad_norm": 10.171887397766113, + "learning_rate": 9.9997919999792e-06, + "loss": 1.3748, + "step": 209 + }, + { + "epoch": 0.09762900976290098, + "grad_norm": 9.12307357788086, + "learning_rate": 9.999790999979101e-06, + "loss": 1.1907, + "step": 210 + }, + { + "epoch": 0.09809390980939098, + "grad_norm": 6.43080997467041, + "learning_rate": 9.999789999979e-06, + "loss": 0.5774, + "step": 211 + }, + { + "epoch": 0.09855880985588099, + "grad_norm": 9.757946014404297, + "learning_rate": 9.9997889999789e-06, + "loss": 1.1577, + "step": 212 + }, + { + "epoch": 0.09902370990237098, + "grad_norm": 6.781042575836182, + "learning_rate": 9.999787999978801e-06, + "loss": 0.5113, + "step": 213 + }, + { + "epoch": 0.099488609948861, + "grad_norm": 11.27275562286377, + "learning_rate": 9.999786999978702e-06, + "loss": 1.104, + "step": 214 + }, + { + "epoch": 0.099953509995351, + "grad_norm": 8.956888198852539, + "learning_rate": 9.999785999978601e-06, + "loss": 0.9176, + "step": 215 + }, + { + "epoch": 0.100418410041841, + "grad_norm": 7.434859275817871, + "learning_rate": 9.999784999978502e-06, + "loss": 0.7391, + "step": 216 + }, + { + "epoch": 0.10088331008833101, + "grad_norm": 6.644069194793701, + "learning_rate": 9.9997839999784e-06, + "loss": 0.6278, + "step": 217 + }, + { + "epoch": 0.10134821013482101, + "grad_norm": 8.018479347229004, + "learning_rate": 9.9997829999783e-06, + "loss": 0.6513, + "step": 218 + }, + { + "epoch": 0.10181311018131102, + "grad_norm": 6.4164910316467285, + "learning_rate": 9.9997819999782e-06, + "loss": 0.7401, + "step": 219 + }, + { + "epoch": 0.10227801022780102, + "grad_norm": 8.048410415649414, + "learning_rate": 9.999780999978101e-06, + "loss": 0.9064, + "step": 220 + }, + { + "epoch": 0.10274291027429103, + "grad_norm": 7.503805160522461, + "learning_rate": 9.999779999978002e-06, + "loss": 0.6988, + "step": 221 + }, + { + "epoch": 0.10320781032078104, + "grad_norm": 7.8786301612854, + "learning_rate": 9.9997789999779e-06, + "loss": 1.0381, + "step": 222 + }, + { + "epoch": 0.10367271036727103, + "grad_norm": 6.937041282653809, + "learning_rate": 9.999777999977801e-06, + "loss": 0.7753, + "step": 223 + }, + { + "epoch": 0.10413761041376104, + "grad_norm": 6.3568267822265625, + "learning_rate": 9.9997769999777e-06, + "loss": 0.4327, + "step": 224 + }, + { + "epoch": 0.10460251046025104, + "grad_norm": 18.106199264526367, + "learning_rate": 9.9997759999776e-06, + "loss": 1.8372, + "step": 225 + }, + { + "epoch": 0.10506741050674105, + "grad_norm": 8.308500289916992, + "learning_rate": 9.999774999977501e-06, + "loss": 0.8701, + "step": 226 + }, + { + "epoch": 0.10553231055323105, + "grad_norm": 10.403562545776367, + "learning_rate": 9.999773999977402e-06, + "loss": 1.318, + "step": 227 + }, + { + "epoch": 0.10599721059972106, + "grad_norm": 8.53259563446045, + "learning_rate": 9.999772999977301e-06, + "loss": 1.0504, + "step": 228 + }, + { + "epoch": 0.10646211064621107, + "grad_norm": 10.569559097290039, + "learning_rate": 9.9997719999772e-06, + "loss": 0.9306, + "step": 229 + }, + { + "epoch": 0.10692701069270107, + "grad_norm": 5.538365364074707, + "learning_rate": 9.9997709999771e-06, + "loss": 0.3862, + "step": 230 + }, + { + "epoch": 0.10739191073919108, + "grad_norm": 7.76457405090332, + "learning_rate": 9.999769999977e-06, + "loss": 0.7084, + "step": 231 + }, + { + "epoch": 0.10785681078568107, + "grad_norm": 12.011214256286621, + "learning_rate": 9.9997689999769e-06, + "loss": 1.1267, + "step": 232 + }, + { + "epoch": 0.10832171083217108, + "grad_norm": 9.098481178283691, + "learning_rate": 9.999767999976801e-06, + "loss": 0.6691, + "step": 233 + }, + { + "epoch": 0.1087866108786611, + "grad_norm": 8.762250900268555, + "learning_rate": 9.999766999976702e-06, + "loss": 0.9535, + "step": 234 + }, + { + "epoch": 0.10925151092515109, + "grad_norm": 11.982100486755371, + "learning_rate": 9.9997659999766e-06, + "loss": 0.917, + "step": 235 + }, + { + "epoch": 0.1097164109716411, + "grad_norm": 9.141411781311035, + "learning_rate": 9.999764999976501e-06, + "loss": 1.1823, + "step": 236 + }, + { + "epoch": 0.1101813110181311, + "grad_norm": 8.919357299804688, + "learning_rate": 9.9997639999764e-06, + "loss": 0.7707, + "step": 237 + }, + { + "epoch": 0.11064621106462111, + "grad_norm": 10.468092918395996, + "learning_rate": 9.999762999976301e-06, + "loss": 1.1858, + "step": 238 + }, + { + "epoch": 0.1111111111111111, + "grad_norm": 11.267462730407715, + "learning_rate": 9.999761999976202e-06, + "loss": 1.3644, + "step": 239 + }, + { + "epoch": 0.11157601115760112, + "grad_norm": 10.946197509765625, + "learning_rate": 9.9997609999761e-06, + "loss": 0.8251, + "step": 240 + }, + { + "epoch": 0.11204091120409113, + "grad_norm": 7.892282962799072, + "learning_rate": 9.999759999976001e-06, + "loss": 0.8013, + "step": 241 + }, + { + "epoch": 0.11250581125058112, + "grad_norm": 9.89450740814209, + "learning_rate": 9.9997589999759e-06, + "loss": 0.9003, + "step": 242 + }, + { + "epoch": 0.11297071129707113, + "grad_norm": 13.196460723876953, + "learning_rate": 9.9997579999758e-06, + "loss": 1.5021, + "step": 243 + }, + { + "epoch": 0.11343561134356113, + "grad_norm": 12.834817886352539, + "learning_rate": 9.9997569999757e-06, + "loss": 1.2474, + "step": 244 + }, + { + "epoch": 0.11390051139005114, + "grad_norm": 10.237266540527344, + "learning_rate": 9.9997559999756e-06, + "loss": 1.1973, + "step": 245 + }, + { + "epoch": 0.11436541143654114, + "grad_norm": 9.91955852508545, + "learning_rate": 9.999754999975501e-06, + "loss": 1.1511, + "step": 246 + }, + { + "epoch": 0.11483031148303115, + "grad_norm": 6.993465423583984, + "learning_rate": 9.999753999975402e-06, + "loss": 0.7146, + "step": 247 + }, + { + "epoch": 0.11529521152952116, + "grad_norm": 6.914819240570068, + "learning_rate": 9.9997529999753e-06, + "loss": 1.0926, + "step": 248 + }, + { + "epoch": 0.11576011157601115, + "grad_norm": 9.380569458007812, + "learning_rate": 9.999751999975201e-06, + "loss": 1.1108, + "step": 249 + }, + { + "epoch": 0.11622501162250116, + "grad_norm": 10.373343467712402, + "learning_rate": 9.9997509999751e-06, + "loss": 1.1679, + "step": 250 + }, + { + "epoch": 0.11668991166899116, + "grad_norm": 6.066802978515625, + "learning_rate": 9.999749999975001e-06, + "loss": 0.5798, + "step": 251 + }, + { + "epoch": 0.11715481171548117, + "grad_norm": 8.414020538330078, + "learning_rate": 9.999748999974902e-06, + "loss": 0.851, + "step": 252 + }, + { + "epoch": 0.11761971176197118, + "grad_norm": 9.911054611206055, + "learning_rate": 9.9997479999748e-06, + "loss": 1.1336, + "step": 253 + }, + { + "epoch": 0.11808461180846118, + "grad_norm": 7.923717021942139, + "learning_rate": 9.999746999974701e-06, + "loss": 0.8495, + "step": 254 + }, + { + "epoch": 0.11854951185495119, + "grad_norm": 8.026178359985352, + "learning_rate": 9.9997459999746e-06, + "loss": 0.7159, + "step": 255 + }, + { + "epoch": 0.11901441190144119, + "grad_norm": 7.211568355560303, + "learning_rate": 9.9997449999745e-06, + "loss": 0.8794, + "step": 256 + }, + { + "epoch": 0.1194793119479312, + "grad_norm": 5.9995880126953125, + "learning_rate": 9.9997439999744e-06, + "loss": 0.49, + "step": 257 + }, + { + "epoch": 0.1199442119944212, + "grad_norm": 7.7433695793151855, + "learning_rate": 9.9997429999743e-06, + "loss": 0.7973, + "step": 258 + }, + { + "epoch": 0.1204091120409112, + "grad_norm": 15.104406356811523, + "learning_rate": 9.999741999974201e-06, + "loss": 1.3159, + "step": 259 + }, + { + "epoch": 0.12087401208740121, + "grad_norm": 9.563790321350098, + "learning_rate": 9.999740999974102e-06, + "loss": 1.0211, + "step": 260 + }, + { + "epoch": 0.12133891213389121, + "grad_norm": 11.435441970825195, + "learning_rate": 9.999739999974e-06, + "loss": 1.3678, + "step": 261 + }, + { + "epoch": 0.12180381218038122, + "grad_norm": 4.941492080688477, + "learning_rate": 9.9997389999739e-06, + "loss": 0.402, + "step": 262 + }, + { + "epoch": 0.12226871222687122, + "grad_norm": 10.62851619720459, + "learning_rate": 9.9997379999738e-06, + "loss": 1.0565, + "step": 263 + }, + { + "epoch": 0.12273361227336123, + "grad_norm": 11.709250450134277, + "learning_rate": 9.999736999973701e-06, + "loss": 0.702, + "step": 264 + }, + { + "epoch": 0.12319851231985124, + "grad_norm": 11.895252227783203, + "learning_rate": 9.999735999973602e-06, + "loss": 0.9878, + "step": 265 + }, + { + "epoch": 0.12366341236634124, + "grad_norm": 5.623959541320801, + "learning_rate": 9.9997349999735e-06, + "loss": 0.566, + "step": 266 + }, + { + "epoch": 0.12412831241283125, + "grad_norm": 11.229691505432129, + "learning_rate": 9.999733999973401e-06, + "loss": 0.689, + "step": 267 + }, + { + "epoch": 0.12459321245932124, + "grad_norm": 8.1890287399292, + "learning_rate": 9.9997329999733e-06, + "loss": 0.8812, + "step": 268 + }, + { + "epoch": 0.12505811250581125, + "grad_norm": 6.76049280166626, + "learning_rate": 9.999731999973201e-06, + "loss": 0.7061, + "step": 269 + }, + { + "epoch": 0.12552301255230125, + "grad_norm": 8.10275936126709, + "learning_rate": 9.9997309999731e-06, + "loss": 0.9413, + "step": 270 + }, + { + "epoch": 0.12598791259879125, + "grad_norm": 12.06290340423584, + "learning_rate": 9.999729999973e-06, + "loss": 0.9154, + "step": 271 + }, + { + "epoch": 0.12645281264528127, + "grad_norm": 12.641207695007324, + "learning_rate": 9.999728999972901e-06, + "loss": 1.5829, + "step": 272 + }, + { + "epoch": 0.12691771269177127, + "grad_norm": 6.974831581115723, + "learning_rate": 9.9997279999728e-06, + "loss": 1.0644, + "step": 273 + }, + { + "epoch": 0.12738261273826126, + "grad_norm": 9.40944766998291, + "learning_rate": 9.9997269999727e-06, + "loss": 0.6973, + "step": 274 + }, + { + "epoch": 0.1278475127847513, + "grad_norm": 7.871720790863037, + "learning_rate": 9.9997259999726e-06, + "loss": 0.7061, + "step": 275 + }, + { + "epoch": 0.12831241283124128, + "grad_norm": 12.039237022399902, + "learning_rate": 9.9997249999725e-06, + "loss": 1.4316, + "step": 276 + }, + { + "epoch": 0.12877731287773128, + "grad_norm": 10.586935043334961, + "learning_rate": 9.999723999972401e-06, + "loss": 0.7804, + "step": 277 + }, + { + "epoch": 0.1292422129242213, + "grad_norm": 11.915858268737793, + "learning_rate": 9.999722999972302e-06, + "loss": 1.1441, + "step": 278 + }, + { + "epoch": 0.1297071129707113, + "grad_norm": 7.242831230163574, + "learning_rate": 9.9997219999722e-06, + "loss": 0.7883, + "step": 279 + }, + { + "epoch": 0.1301720130172013, + "grad_norm": 12.292387008666992, + "learning_rate": 9.999720999972101e-06, + "loss": 1.5032, + "step": 280 + }, + { + "epoch": 0.1306369130636913, + "grad_norm": 9.370741844177246, + "learning_rate": 9.999719999972e-06, + "loss": 1.12, + "step": 281 + }, + { + "epoch": 0.13110181311018132, + "grad_norm": 9.412545204162598, + "learning_rate": 9.999718999971901e-06, + "loss": 0.7944, + "step": 282 + }, + { + "epoch": 0.13156671315667132, + "grad_norm": 5.701841354370117, + "learning_rate": 9.999717999971802e-06, + "loss": 0.6152, + "step": 283 + }, + { + "epoch": 0.1320316132031613, + "grad_norm": 10.617661476135254, + "learning_rate": 9.9997169999717e-06, + "loss": 1.0023, + "step": 284 + }, + { + "epoch": 0.13249651324965134, + "grad_norm": 9.593035697937012, + "learning_rate": 9.999715999971601e-06, + "loss": 1.0563, + "step": 285 + }, + { + "epoch": 0.13296141329614133, + "grad_norm": 7.4250030517578125, + "learning_rate": 9.9997149999715e-06, + "loss": 0.8223, + "step": 286 + }, + { + "epoch": 0.13342631334263133, + "grad_norm": 11.55223560333252, + "learning_rate": 9.999713999971401e-06, + "loss": 1.0257, + "step": 287 + }, + { + "epoch": 0.13389121338912133, + "grad_norm": 6.441282749176025, + "learning_rate": 9.9997129999713e-06, + "loss": 0.5345, + "step": 288 + }, + { + "epoch": 0.13435611343561135, + "grad_norm": 6.219207763671875, + "learning_rate": 9.9997119999712e-06, + "loss": 0.529, + "step": 289 + }, + { + "epoch": 0.13482101348210135, + "grad_norm": 9.252127647399902, + "learning_rate": 9.999710999971101e-06, + "loss": 0.9256, + "step": 290 + }, + { + "epoch": 0.13528591352859135, + "grad_norm": 7.892247200012207, + "learning_rate": 9.999709999971002e-06, + "loss": 1.039, + "step": 291 + }, + { + "epoch": 0.13575081357508137, + "grad_norm": 8.784099578857422, + "learning_rate": 9.9997089999709e-06, + "loss": 1.0464, + "step": 292 + }, + { + "epoch": 0.13621571362157137, + "grad_norm": 9.945881843566895, + "learning_rate": 9.999707999970801e-06, + "loss": 1.1701, + "step": 293 + }, + { + "epoch": 0.13668061366806136, + "grad_norm": 8.645524024963379, + "learning_rate": 9.9997069999707e-06, + "loss": 1.0741, + "step": 294 + }, + { + "epoch": 0.13714551371455136, + "grad_norm": 11.112211227416992, + "learning_rate": 9.999705999970601e-06, + "loss": 1.1889, + "step": 295 + }, + { + "epoch": 0.13761041376104138, + "grad_norm": 7.312314987182617, + "learning_rate": 9.999704999970502e-06, + "loss": 0.69, + "step": 296 + }, + { + "epoch": 0.13807531380753138, + "grad_norm": 7.862229347229004, + "learning_rate": 9.9997039999704e-06, + "loss": 0.8151, + "step": 297 + }, + { + "epoch": 0.13854021385402138, + "grad_norm": 6.953213214874268, + "learning_rate": 9.999702999970301e-06, + "loss": 0.7467, + "step": 298 + }, + { + "epoch": 0.1390051139005114, + "grad_norm": 13.456071853637695, + "learning_rate": 9.9997019999702e-06, + "loss": 0.8261, + "step": 299 + }, + { + "epoch": 0.1394700139470014, + "grad_norm": 8.059375762939453, + "learning_rate": 9.999700999970101e-06, + "loss": 1.1772, + "step": 300 + }, + { + "epoch": 0.1399349139934914, + "grad_norm": 12.619104385375977, + "learning_rate": 9.99969999997e-06, + "loss": 1.1054, + "step": 301 + }, + { + "epoch": 0.1403998140399814, + "grad_norm": 9.00656795501709, + "learning_rate": 9.9996989999699e-06, + "loss": 1.196, + "step": 302 + }, + { + "epoch": 0.14086471408647142, + "grad_norm": 9.230240821838379, + "learning_rate": 9.999697999969801e-06, + "loss": 1.0729, + "step": 303 + }, + { + "epoch": 0.1413296141329614, + "grad_norm": 8.453560829162598, + "learning_rate": 9.999696999969702e-06, + "loss": 0.9889, + "step": 304 + }, + { + "epoch": 0.1417945141794514, + "grad_norm": 7.742706775665283, + "learning_rate": 9.999695999969601e-06, + "loss": 0.674, + "step": 305 + }, + { + "epoch": 0.14225941422594143, + "grad_norm": 6.092482089996338, + "learning_rate": 9.9996949999695e-06, + "loss": 0.5281, + "step": 306 + }, + { + "epoch": 0.14272431427243143, + "grad_norm": 12.05079174041748, + "learning_rate": 9.9996939999694e-06, + "loss": 0.9217, + "step": 307 + }, + { + "epoch": 0.14318921431892143, + "grad_norm": 11.561873435974121, + "learning_rate": 9.999692999969301e-06, + "loss": 1.0365, + "step": 308 + }, + { + "epoch": 0.14365411436541142, + "grad_norm": 13.429388999938965, + "learning_rate": 9.999691999969202e-06, + "loss": 1.3968, + "step": 309 + }, + { + "epoch": 0.14411901441190145, + "grad_norm": 7.858463287353516, + "learning_rate": 9.9996909999691e-06, + "loss": 0.7293, + "step": 310 + }, + { + "epoch": 0.14458391445839144, + "grad_norm": 9.144617080688477, + "learning_rate": 9.999689999969001e-06, + "loss": 0.7688, + "step": 311 + }, + { + "epoch": 0.14504881450488144, + "grad_norm": 9.854680061340332, + "learning_rate": 9.9996889999689e-06, + "loss": 0.999, + "step": 312 + }, + { + "epoch": 0.14551371455137146, + "grad_norm": 19.348724365234375, + "learning_rate": 9.999687999968801e-06, + "loss": 1.2113, + "step": 313 + }, + { + "epoch": 0.14597861459786146, + "grad_norm": 13.300654411315918, + "learning_rate": 9.9996869999687e-06, + "loss": 0.9613, + "step": 314 + }, + { + "epoch": 0.14644351464435146, + "grad_norm": 9.462535858154297, + "learning_rate": 9.9996859999686e-06, + "loss": 0.9527, + "step": 315 + }, + { + "epoch": 0.14690841469084148, + "grad_norm": 11.3388090133667, + "learning_rate": 9.999684999968501e-06, + "loss": 0.9808, + "step": 316 + }, + { + "epoch": 0.14737331473733148, + "grad_norm": 12.701608657836914, + "learning_rate": 9.9996839999684e-06, + "loss": 0.9674, + "step": 317 + }, + { + "epoch": 0.14783821478382148, + "grad_norm": 11.985180854797363, + "learning_rate": 9.999682999968301e-06, + "loss": 0.7445, + "step": 318 + }, + { + "epoch": 0.14830311483031147, + "grad_norm": 6.6830363273620605, + "learning_rate": 9.9996819999682e-06, + "loss": 0.5676, + "step": 319 + }, + { + "epoch": 0.1487680148768015, + "grad_norm": 17.123056411743164, + "learning_rate": 9.9996809999681e-06, + "loss": 0.8858, + "step": 320 + }, + { + "epoch": 0.1492329149232915, + "grad_norm": 9.98444938659668, + "learning_rate": 9.999679999968001e-06, + "loss": 0.6621, + "step": 321 + }, + { + "epoch": 0.1496978149697815, + "grad_norm": 8.065140724182129, + "learning_rate": 9.999678999967902e-06, + "loss": 0.5673, + "step": 322 + }, + { + "epoch": 0.15016271501627151, + "grad_norm": 7.101948261260986, + "learning_rate": 9.999677999967801e-06, + "loss": 0.9337, + "step": 323 + }, + { + "epoch": 0.1506276150627615, + "grad_norm": 8.5210599899292, + "learning_rate": 9.999676999967702e-06, + "loss": 1.1879, + "step": 324 + }, + { + "epoch": 0.1510925151092515, + "grad_norm": 13.893704414367676, + "learning_rate": 9.9996759999676e-06, + "loss": 1.3883, + "step": 325 + }, + { + "epoch": 0.1515574151557415, + "grad_norm": 9.152368545532227, + "learning_rate": 9.999674999967501e-06, + "loss": 1.1952, + "step": 326 + }, + { + "epoch": 0.15202231520223153, + "grad_norm": 9.124520301818848, + "learning_rate": 9.9996739999674e-06, + "loss": 0.9439, + "step": 327 + }, + { + "epoch": 0.15248721524872152, + "grad_norm": 8.915891647338867, + "learning_rate": 9.9996729999673e-06, + "loss": 0.9467, + "step": 328 + }, + { + "epoch": 0.15295211529521152, + "grad_norm": 11.785049438476562, + "learning_rate": 9.999671999967201e-06, + "loss": 1.239, + "step": 329 + }, + { + "epoch": 0.15341701534170155, + "grad_norm": 7.638711452484131, + "learning_rate": 9.9996709999671e-06, + "loss": 0.6521, + "step": 330 + }, + { + "epoch": 0.15388191538819154, + "grad_norm": 13.12578010559082, + "learning_rate": 9.999669999967001e-06, + "loss": 1.2717, + "step": 331 + }, + { + "epoch": 0.15434681543468154, + "grad_norm": 10.207740783691406, + "learning_rate": 9.9996689999669e-06, + "loss": 0.7813, + "step": 332 + }, + { + "epoch": 0.15481171548117154, + "grad_norm": 10.358567237854004, + "learning_rate": 9.9996679999668e-06, + "loss": 0.8681, + "step": 333 + }, + { + "epoch": 0.15527661552766156, + "grad_norm": 4.920478820800781, + "learning_rate": 9.999666999966701e-06, + "loss": 0.3851, + "step": 334 + }, + { + "epoch": 0.15574151557415156, + "grad_norm": 7.195193767547607, + "learning_rate": 9.999665999966602e-06, + "loss": 0.8943, + "step": 335 + }, + { + "epoch": 0.15620641562064155, + "grad_norm": 10.269038200378418, + "learning_rate": 9.999664999966501e-06, + "loss": 1.2482, + "step": 336 + }, + { + "epoch": 0.15667131566713158, + "grad_norm": 11.979458808898926, + "learning_rate": 9.999663999966402e-06, + "loss": 1.1506, + "step": 337 + }, + { + "epoch": 0.15713621571362157, + "grad_norm": 7.4199748039245605, + "learning_rate": 9.9996629999663e-06, + "loss": 0.523, + "step": 338 + }, + { + "epoch": 0.15760111576011157, + "grad_norm": 8.933382034301758, + "learning_rate": 9.9996619999662e-06, + "loss": 0.6206, + "step": 339 + }, + { + "epoch": 0.15806601580660157, + "grad_norm": 12.299149513244629, + "learning_rate": 9.9996609999661e-06, + "loss": 1.2337, + "step": 340 + }, + { + "epoch": 0.1585309158530916, + "grad_norm": 11.12102222442627, + "learning_rate": 9.999659999966001e-06, + "loss": 1.0561, + "step": 341 + }, + { + "epoch": 0.1589958158995816, + "grad_norm": 11.38724136352539, + "learning_rate": 9.999658999965902e-06, + "loss": 1.0401, + "step": 342 + }, + { + "epoch": 0.15946071594607159, + "grad_norm": 8.496471405029297, + "learning_rate": 9.9996579999658e-06, + "loss": 1.2376, + "step": 343 + }, + { + "epoch": 0.1599256159925616, + "grad_norm": 9.59240436553955, + "learning_rate": 9.999656999965701e-06, + "loss": 0.9605, + "step": 344 + }, + { + "epoch": 0.1603905160390516, + "grad_norm": 8.462102890014648, + "learning_rate": 9.9996559999656e-06, + "loss": 0.7632, + "step": 345 + }, + { + "epoch": 0.1608554160855416, + "grad_norm": 8.93777847290039, + "learning_rate": 9.9996549999655e-06, + "loss": 0.7141, + "step": 346 + }, + { + "epoch": 0.16132031613203163, + "grad_norm": 8.670785903930664, + "learning_rate": 9.999653999965401e-06, + "loss": 0.8479, + "step": 347 + }, + { + "epoch": 0.16178521617852162, + "grad_norm": 8.92241096496582, + "learning_rate": 9.999652999965302e-06, + "loss": 1.2104, + "step": 348 + }, + { + "epoch": 0.16225011622501162, + "grad_norm": 11.54997444152832, + "learning_rate": 9.999651999965201e-06, + "loss": 1.0483, + "step": 349 + }, + { + "epoch": 0.16271501627150162, + "grad_norm": 7.185827255249023, + "learning_rate": 9.9996509999651e-06, + "loss": 0.487, + "step": 350 + }, + { + "epoch": 0.16317991631799164, + "grad_norm": 8.953634262084961, + "learning_rate": 9.999649999965e-06, + "loss": 1.0037, + "step": 351 + }, + { + "epoch": 0.16364481636448164, + "grad_norm": 9.15186595916748, + "learning_rate": 9.999648999964901e-06, + "loss": 0.7094, + "step": 352 + }, + { + "epoch": 0.16410971641097163, + "grad_norm": 10.119549751281738, + "learning_rate": 9.9996479999648e-06, + "loss": 1.1838, + "step": 353 + }, + { + "epoch": 0.16457461645746166, + "grad_norm": 9.315022468566895, + "learning_rate": 9.999646999964701e-06, + "loss": 0.8905, + "step": 354 + }, + { + "epoch": 0.16503951650395166, + "grad_norm": 10.435300827026367, + "learning_rate": 9.999645999964602e-06, + "loss": 1.1032, + "step": 355 + }, + { + "epoch": 0.16550441655044165, + "grad_norm": 6.094074726104736, + "learning_rate": 9.9996449999645e-06, + "loss": 0.6248, + "step": 356 + }, + { + "epoch": 0.16596931659693165, + "grad_norm": 9.23153018951416, + "learning_rate": 9.999643999964401e-06, + "loss": 0.6708, + "step": 357 + }, + { + "epoch": 0.16643421664342167, + "grad_norm": 6.390391826629639, + "learning_rate": 9.9996429999643e-06, + "loss": 0.838, + "step": 358 + }, + { + "epoch": 0.16689911668991167, + "grad_norm": 9.152975082397461, + "learning_rate": 9.9996419999642e-06, + "loss": 0.6869, + "step": 359 + }, + { + "epoch": 0.16736401673640167, + "grad_norm": 11.0722074508667, + "learning_rate": 9.999640999964101e-06, + "loss": 1.4135, + "step": 360 + }, + { + "epoch": 0.1678289167828917, + "grad_norm": 6.525270938873291, + "learning_rate": 9.999639999964e-06, + "loss": 0.6298, + "step": 361 + }, + { + "epoch": 0.1682938168293817, + "grad_norm": 8.108715057373047, + "learning_rate": 9.999638999963901e-06, + "loss": 1.0779, + "step": 362 + }, + { + "epoch": 0.16875871687587168, + "grad_norm": 9.540878295898438, + "learning_rate": 9.9996379999638e-06, + "loss": 0.9028, + "step": 363 + }, + { + "epoch": 0.16922361692236168, + "grad_norm": 9.609642028808594, + "learning_rate": 9.9996369999637e-06, + "loss": 1.181, + "step": 364 + }, + { + "epoch": 0.1696885169688517, + "grad_norm": 10.981318473815918, + "learning_rate": 9.999635999963601e-06, + "loss": 1.3474, + "step": 365 + }, + { + "epoch": 0.1701534170153417, + "grad_norm": 9.830684661865234, + "learning_rate": 9.9996349999635e-06, + "loss": 0.7052, + "step": 366 + }, + { + "epoch": 0.1706183170618317, + "grad_norm": 14.448360443115234, + "learning_rate": 9.999633999963401e-06, + "loss": 1.4121, + "step": 367 + }, + { + "epoch": 0.17108321710832172, + "grad_norm": 6.7998762130737305, + "learning_rate": 9.999632999963302e-06, + "loss": 0.8969, + "step": 368 + }, + { + "epoch": 0.17154811715481172, + "grad_norm": 8.06762981414795, + "learning_rate": 9.9996319999632e-06, + "loss": 0.9821, + "step": 369 + }, + { + "epoch": 0.17201301720130172, + "grad_norm": 10.103517532348633, + "learning_rate": 9.9996309999631e-06, + "loss": 0.6505, + "step": 370 + }, + { + "epoch": 0.1724779172477917, + "grad_norm": 9.847838401794434, + "learning_rate": 9.999629999963e-06, + "loss": 0.9565, + "step": 371 + }, + { + "epoch": 0.17294281729428174, + "grad_norm": 12.264676094055176, + "learning_rate": 9.999628999962901e-06, + "loss": 1.016, + "step": 372 + }, + { + "epoch": 0.17340771734077173, + "grad_norm": 14.056889533996582, + "learning_rate": 9.999627999962802e-06, + "loss": 1.1764, + "step": 373 + }, + { + "epoch": 0.17387261738726173, + "grad_norm": 10.973730087280273, + "learning_rate": 9.9996269999627e-06, + "loss": 1.4014, + "step": 374 + }, + { + "epoch": 0.17433751743375175, + "grad_norm": 9.78699016571045, + "learning_rate": 9.999625999962601e-06, + "loss": 1.025, + "step": 375 + }, + { + "epoch": 0.17480241748024175, + "grad_norm": 6.848630905151367, + "learning_rate": 9.9996249999625e-06, + "loss": 0.6679, + "step": 376 + }, + { + "epoch": 0.17526731752673175, + "grad_norm": 8.655278205871582, + "learning_rate": 9.9996239999624e-06, + "loss": 0.753, + "step": 377 + }, + { + "epoch": 0.17573221757322174, + "grad_norm": 10.006077766418457, + "learning_rate": 9.999622999962301e-06, + "loss": 1.1677, + "step": 378 + }, + { + "epoch": 0.17619711761971177, + "grad_norm": 8.873841285705566, + "learning_rate": 9.999621999962202e-06, + "loss": 1.1956, + "step": 379 + }, + { + "epoch": 0.17666201766620176, + "grad_norm": 11.022208213806152, + "learning_rate": 9.999620999962101e-06, + "loss": 0.8935, + "step": 380 + }, + { + "epoch": 0.17712691771269176, + "grad_norm": 4.7020463943481445, + "learning_rate": 9.999619999962e-06, + "loss": 0.4221, + "step": 381 + }, + { + "epoch": 0.17759181775918179, + "grad_norm": 8.167107582092285, + "learning_rate": 9.9996189999619e-06, + "loss": 0.9364, + "step": 382 + }, + { + "epoch": 0.17805671780567178, + "grad_norm": 8.398082733154297, + "learning_rate": 9.9996179999618e-06, + "loss": 0.7412, + "step": 383 + }, + { + "epoch": 0.17852161785216178, + "grad_norm": 11.760035514831543, + "learning_rate": 9.9996169999617e-06, + "loss": 1.0779, + "step": 384 + }, + { + "epoch": 0.1789865178986518, + "grad_norm": 5.843050003051758, + "learning_rate": 9.999615999961601e-06, + "loss": 0.651, + "step": 385 + }, + { + "epoch": 0.1794514179451418, + "grad_norm": 7.996475696563721, + "learning_rate": 9.999614999961502e-06, + "loss": 1.0146, + "step": 386 + }, + { + "epoch": 0.1799163179916318, + "grad_norm": 8.653067588806152, + "learning_rate": 9.9996139999614e-06, + "loss": 0.7164, + "step": 387 + }, + { + "epoch": 0.1803812180381218, + "grad_norm": 6.381991386413574, + "learning_rate": 9.999612999961301e-06, + "loss": 0.3696, + "step": 388 + }, + { + "epoch": 0.18084611808461182, + "grad_norm": 9.563328742980957, + "learning_rate": 9.9996119999612e-06, + "loss": 1.0129, + "step": 389 + }, + { + "epoch": 0.18131101813110181, + "grad_norm": 11.251789093017578, + "learning_rate": 9.999610999961101e-06, + "loss": 1.0601, + "step": 390 + }, + { + "epoch": 0.1817759181775918, + "grad_norm": 8.412424087524414, + "learning_rate": 9.999609999961002e-06, + "loss": 1.137, + "step": 391 + }, + { + "epoch": 0.18224081822408184, + "grad_norm": 5.0486907958984375, + "learning_rate": 9.9996089999609e-06, + "loss": 0.5562, + "step": 392 + }, + { + "epoch": 0.18270571827057183, + "grad_norm": 5.968600273132324, + "learning_rate": 9.999607999960801e-06, + "loss": 0.583, + "step": 393 + }, + { + "epoch": 0.18317061831706183, + "grad_norm": 11.43770694732666, + "learning_rate": 9.9996069999607e-06, + "loss": 1.137, + "step": 394 + }, + { + "epoch": 0.18363551836355183, + "grad_norm": 6.772749423980713, + "learning_rate": 9.9996059999606e-06, + "loss": 0.7023, + "step": 395 + }, + { + "epoch": 0.18410041841004185, + "grad_norm": 6.41163444519043, + "learning_rate": 9.9996049999605e-06, + "loss": 0.5686, + "step": 396 + }, + { + "epoch": 0.18456531845653185, + "grad_norm": 6.309353351593018, + "learning_rate": 9.9996039999604e-06, + "loss": 0.4522, + "step": 397 + }, + { + "epoch": 0.18503021850302184, + "grad_norm": 12.959024429321289, + "learning_rate": 9.999602999960301e-06, + "loss": 0.5484, + "step": 398 + }, + { + "epoch": 0.18549511854951187, + "grad_norm": 10.195695877075195, + "learning_rate": 9.999601999960202e-06, + "loss": 1.1236, + "step": 399 + }, + { + "epoch": 0.18596001859600186, + "grad_norm": 10.67815113067627, + "learning_rate": 9.9996009999601e-06, + "loss": 1.0199, + "step": 400 + }, + { + "epoch": 0.18642491864249186, + "grad_norm": 9.70882797241211, + "learning_rate": 9.999599999960001e-06, + "loss": 0.9709, + "step": 401 + }, + { + "epoch": 0.18688981868898186, + "grad_norm": 8.720227241516113, + "learning_rate": 9.9995989999599e-06, + "loss": 0.7228, + "step": 402 + }, + { + "epoch": 0.18735471873547188, + "grad_norm": 6.728839874267578, + "learning_rate": 9.999597999959801e-06, + "loss": 0.7777, + "step": 403 + }, + { + "epoch": 0.18781961878196188, + "grad_norm": 8.135492324829102, + "learning_rate": 9.999596999959702e-06, + "loss": 0.3976, + "step": 404 + }, + { + "epoch": 0.18828451882845187, + "grad_norm": 10.888492584228516, + "learning_rate": 9.9995959999596e-06, + "loss": 1.1992, + "step": 405 + }, + { + "epoch": 0.1887494188749419, + "grad_norm": 8.427240371704102, + "learning_rate": 9.999594999959501e-06, + "loss": 0.8209, + "step": 406 + }, + { + "epoch": 0.1892143189214319, + "grad_norm": 6.078670024871826, + "learning_rate": 9.9995939999594e-06, + "loss": 0.6371, + "step": 407 + }, + { + "epoch": 0.1896792189679219, + "grad_norm": 8.472917556762695, + "learning_rate": 9.999592999959301e-06, + "loss": 0.8888, + "step": 408 + }, + { + "epoch": 0.1901441190144119, + "grad_norm": 9.414752006530762, + "learning_rate": 9.9995919999592e-06, + "loss": 1.1614, + "step": 409 + }, + { + "epoch": 0.1906090190609019, + "grad_norm": 6.906506538391113, + "learning_rate": 9.9995909999591e-06, + "loss": 0.6992, + "step": 410 + }, + { + "epoch": 0.1910739191073919, + "grad_norm": 14.92821979522705, + "learning_rate": 9.999589999959001e-06, + "loss": 0.7815, + "step": 411 + }, + { + "epoch": 0.1915388191538819, + "grad_norm": 7.877196311950684, + "learning_rate": 9.999588999958902e-06, + "loss": 0.781, + "step": 412 + }, + { + "epoch": 0.19200371920037193, + "grad_norm": 9.668245315551758, + "learning_rate": 9.9995879999588e-06, + "loss": 0.9955, + "step": 413 + }, + { + "epoch": 0.19246861924686193, + "grad_norm": 7.207817554473877, + "learning_rate": 9.9995869999587e-06, + "loss": 0.7135, + "step": 414 + }, + { + "epoch": 0.19293351929335192, + "grad_norm": 8.58276081085205, + "learning_rate": 9.9995859999586e-06, + "loss": 1.128, + "step": 415 + }, + { + "epoch": 0.19339841933984192, + "grad_norm": 7.06909704208374, + "learning_rate": 9.999584999958501e-06, + "loss": 0.6395, + "step": 416 + }, + { + "epoch": 0.19386331938633194, + "grad_norm": 7.785849571228027, + "learning_rate": 9.999583999958402e-06, + "loss": 0.7514, + "step": 417 + }, + { + "epoch": 0.19432821943282194, + "grad_norm": 10.765631675720215, + "learning_rate": 9.9995829999583e-06, + "loss": 0.8706, + "step": 418 + }, + { + "epoch": 0.19479311947931194, + "grad_norm": 7.276439666748047, + "learning_rate": 9.999581999958201e-06, + "loss": 0.7903, + "step": 419 + }, + { + "epoch": 0.19525801952580196, + "grad_norm": 10.225470542907715, + "learning_rate": 9.9995809999581e-06, + "loss": 1.0569, + "step": 420 + }, + { + "epoch": 0.19572291957229196, + "grad_norm": 7.663893699645996, + "learning_rate": 9.999579999958001e-06, + "loss": 0.5474, + "step": 421 + }, + { + "epoch": 0.19618781961878196, + "grad_norm": 7.587120056152344, + "learning_rate": 9.9995789999579e-06, + "loss": 0.8314, + "step": 422 + }, + { + "epoch": 0.19665271966527198, + "grad_norm": 8.469603538513184, + "learning_rate": 9.9995779999578e-06, + "loss": 0.9705, + "step": 423 + }, + { + "epoch": 0.19711761971176198, + "grad_norm": 6.758122444152832, + "learning_rate": 9.999576999957701e-06, + "loss": 0.8181, + "step": 424 + }, + { + "epoch": 0.19758251975825197, + "grad_norm": 6.603053569793701, + "learning_rate": 9.9995759999576e-06, + "loss": 0.8779, + "step": 425 + }, + { + "epoch": 0.19804741980474197, + "grad_norm": 6.95527458190918, + "learning_rate": 9.999574999957501e-06, + "loss": 0.5607, + "step": 426 + }, + { + "epoch": 0.198512319851232, + "grad_norm": 14.598949432373047, + "learning_rate": 9.9995739999574e-06, + "loss": 1.7187, + "step": 427 + }, + { + "epoch": 0.198977219897722, + "grad_norm": 7.673601150512695, + "learning_rate": 9.9995729999573e-06, + "loss": 0.6847, + "step": 428 + }, + { + "epoch": 0.199442119944212, + "grad_norm": 9.08953857421875, + "learning_rate": 9.999571999957201e-06, + "loss": 1.2583, + "step": 429 + }, + { + "epoch": 0.199907019990702, + "grad_norm": 8.250752449035645, + "learning_rate": 9.999570999957102e-06, + "loss": 0.5402, + "step": 430 + }, + { + "epoch": 0.200371920037192, + "grad_norm": 11.522260665893555, + "learning_rate": 9.999569999957e-06, + "loss": 0.9404, + "step": 431 + }, + { + "epoch": 0.200836820083682, + "grad_norm": 6.919488906860352, + "learning_rate": 9.999568999956901e-06, + "loss": 0.8471, + "step": 432 + }, + { + "epoch": 0.201301720130172, + "grad_norm": 15.133489608764648, + "learning_rate": 9.9995679999568e-06, + "loss": 0.6648, + "step": 433 + }, + { + "epoch": 0.20176662017666203, + "grad_norm": 7.824070453643799, + "learning_rate": 9.999566999956701e-06, + "loss": 0.9713, + "step": 434 + }, + { + "epoch": 0.20223152022315202, + "grad_norm": 9.405685424804688, + "learning_rate": 9.9995659999566e-06, + "loss": 1.317, + "step": 435 + }, + { + "epoch": 0.20269642026964202, + "grad_norm": 7.080969333648682, + "learning_rate": 9.9995649999565e-06, + "loss": 0.9341, + "step": 436 + }, + { + "epoch": 0.20316132031613204, + "grad_norm": 9.221504211425781, + "learning_rate": 9.999563999956401e-06, + "loss": 1.1097, + "step": 437 + }, + { + "epoch": 0.20362622036262204, + "grad_norm": 5.357058525085449, + "learning_rate": 9.9995629999563e-06, + "loss": 0.5194, + "step": 438 + }, + { + "epoch": 0.20409112040911204, + "grad_norm": 6.210283279418945, + "learning_rate": 9.999561999956201e-06, + "loss": 0.5186, + "step": 439 + }, + { + "epoch": 0.20455602045560203, + "grad_norm": 8.414419174194336, + "learning_rate": 9.9995609999561e-06, + "loss": 1.0243, + "step": 440 + }, + { + "epoch": 0.20502092050209206, + "grad_norm": 9.082574844360352, + "learning_rate": 9.999559999956e-06, + "loss": 1.1182, + "step": 441 + }, + { + "epoch": 0.20548582054858205, + "grad_norm": 10.086358070373535, + "learning_rate": 9.999558999955901e-06, + "loss": 1.1393, + "step": 442 + }, + { + "epoch": 0.20595072059507205, + "grad_norm": 8.429665565490723, + "learning_rate": 9.999557999955802e-06, + "loss": 1.0179, + "step": 443 + }, + { + "epoch": 0.20641562064156208, + "grad_norm": 6.505364894866943, + "learning_rate": 9.9995569999557e-06, + "loss": 0.834, + "step": 444 + }, + { + "epoch": 0.20688052068805207, + "grad_norm": 7.7085466384887695, + "learning_rate": 9.999555999955601e-06, + "loss": 0.9891, + "step": 445 + }, + { + "epoch": 0.20734542073454207, + "grad_norm": 16.61835479736328, + "learning_rate": 9.9995549999555e-06, + "loss": 0.6081, + "step": 446 + }, + { + "epoch": 0.20781032078103207, + "grad_norm": 5.938403129577637, + "learning_rate": 9.999553999955401e-06, + "loss": 0.4228, + "step": 447 + }, + { + "epoch": 0.2082752208275221, + "grad_norm": 6.174816131591797, + "learning_rate": 9.999552999955302e-06, + "loss": 0.5757, + "step": 448 + }, + { + "epoch": 0.2087401208740121, + "grad_norm": 8.406527519226074, + "learning_rate": 9.9995519999552e-06, + "loss": 0.7935, + "step": 449 + }, + { + "epoch": 0.20920502092050208, + "grad_norm": 9.443435668945312, + "learning_rate": 9.999550999955101e-06, + "loss": 0.9466, + "step": 450 + }, + { + "epoch": 0.2096699209669921, + "grad_norm": 6.9016218185424805, + "learning_rate": 9.999549999955e-06, + "loss": 0.6954, + "step": 451 + }, + { + "epoch": 0.2101348210134821, + "grad_norm": 8.143157958984375, + "learning_rate": 9.999548999954901e-06, + "loss": 0.798, + "step": 452 + }, + { + "epoch": 0.2105997210599721, + "grad_norm": 7.225801467895508, + "learning_rate": 9.9995479999548e-06, + "loss": 0.5397, + "step": 453 + }, + { + "epoch": 0.2110646211064621, + "grad_norm": 7.873186111450195, + "learning_rate": 9.9995469999547e-06, + "loss": 0.9154, + "step": 454 + }, + { + "epoch": 0.21152952115295212, + "grad_norm": 6.243428707122803, + "learning_rate": 9.999545999954601e-06, + "loss": 0.5773, + "step": 455 + }, + { + "epoch": 0.21199442119944212, + "grad_norm": 11.937853813171387, + "learning_rate": 9.999544999954502e-06, + "loss": 0.7871, + "step": 456 + }, + { + "epoch": 0.21245932124593211, + "grad_norm": 7.818880558013916, + "learning_rate": 9.999543999954401e-06, + "loss": 0.8021, + "step": 457 + }, + { + "epoch": 0.21292422129242214, + "grad_norm": 8.968914031982422, + "learning_rate": 9.9995429999543e-06, + "loss": 0.7897, + "step": 458 + }, + { + "epoch": 0.21338912133891214, + "grad_norm": 7.5700788497924805, + "learning_rate": 9.9995419999542e-06, + "loss": 0.7496, + "step": 459 + }, + { + "epoch": 0.21385402138540213, + "grad_norm": 5.7492852210998535, + "learning_rate": 9.999540999954101e-06, + "loss": 0.628, + "step": 460 + }, + { + "epoch": 0.21431892143189216, + "grad_norm": 7.608848571777344, + "learning_rate": 9.999539999954002e-06, + "loss": 0.7443, + "step": 461 + }, + { + "epoch": 0.21478382147838215, + "grad_norm": 8.77096176147461, + "learning_rate": 9.9995389999539e-06, + "loss": 0.6331, + "step": 462 + }, + { + "epoch": 0.21524872152487215, + "grad_norm": 7.307488918304443, + "learning_rate": 9.999537999953801e-06, + "loss": 0.5387, + "step": 463 + }, + { + "epoch": 0.21571362157136215, + "grad_norm": 10.195259094238281, + "learning_rate": 9.9995369999537e-06, + "loss": 1.0431, + "step": 464 + }, + { + "epoch": 0.21617852161785217, + "grad_norm": 9.572640419006348, + "learning_rate": 9.999535999953601e-06, + "loss": 1.0802, + "step": 465 + }, + { + "epoch": 0.21664342166434217, + "grad_norm": 9.975297927856445, + "learning_rate": 9.9995349999535e-06, + "loss": 0.5431, + "step": 466 + }, + { + "epoch": 0.21710832171083216, + "grad_norm": 9.473767280578613, + "learning_rate": 9.9995339999534e-06, + "loss": 1.0349, + "step": 467 + }, + { + "epoch": 0.2175732217573222, + "grad_norm": 10.577099800109863, + "learning_rate": 9.999532999953301e-06, + "loss": 0.6105, + "step": 468 + }, + { + "epoch": 0.21803812180381218, + "grad_norm": 16.08302879333496, + "learning_rate": 9.9995319999532e-06, + "loss": 1.0745, + "step": 469 + }, + { + "epoch": 0.21850302185030218, + "grad_norm": 5.1359944343566895, + "learning_rate": 9.999530999953101e-06, + "loss": 0.4641, + "step": 470 + }, + { + "epoch": 0.21896792189679218, + "grad_norm": 10.35550308227539, + "learning_rate": 9.999529999953e-06, + "loss": 1.1089, + "step": 471 + }, + { + "epoch": 0.2194328219432822, + "grad_norm": 9.087845802307129, + "learning_rate": 9.9995289999529e-06, + "loss": 0.969, + "step": 472 + }, + { + "epoch": 0.2198977219897722, + "grad_norm": 10.588335037231445, + "learning_rate": 9.999527999952801e-06, + "loss": 0.8416, + "step": 473 + }, + { + "epoch": 0.2203626220362622, + "grad_norm": 8.068855285644531, + "learning_rate": 9.999526999952702e-06, + "loss": 0.7444, + "step": 474 + }, + { + "epoch": 0.22082752208275222, + "grad_norm": 8.360237121582031, + "learning_rate": 9.999525999952601e-06, + "loss": 0.8588, + "step": 475 + }, + { + "epoch": 0.22129242212924222, + "grad_norm": 4.928906440734863, + "learning_rate": 9.999524999952502e-06, + "loss": 0.4938, + "step": 476 + }, + { + "epoch": 0.2217573221757322, + "grad_norm": 7.238918781280518, + "learning_rate": 9.9995239999524e-06, + "loss": 0.8268, + "step": 477 + }, + { + "epoch": 0.2222222222222222, + "grad_norm": 7.499417304992676, + "learning_rate": 9.999522999952301e-06, + "loss": 0.85, + "step": 478 + }, + { + "epoch": 0.22268712226871223, + "grad_norm": 7.6138458251953125, + "learning_rate": 9.9995219999522e-06, + "loss": 0.8774, + "step": 479 + }, + { + "epoch": 0.22315202231520223, + "grad_norm": 8.667255401611328, + "learning_rate": 9.9995209999521e-06, + "loss": 0.8724, + "step": 480 + }, + { + "epoch": 0.22361692236169223, + "grad_norm": 6.555504322052002, + "learning_rate": 9.999519999952001e-06, + "loss": 0.7641, + "step": 481 + }, + { + "epoch": 0.22408182240818225, + "grad_norm": 7.702280521392822, + "learning_rate": 9.9995189999519e-06, + "loss": 1.0179, + "step": 482 + }, + { + "epoch": 0.22454672245467225, + "grad_norm": 10.255595207214355, + "learning_rate": 9.999517999951801e-06, + "loss": 1.0248, + "step": 483 + }, + { + "epoch": 0.22501162250116225, + "grad_norm": 10.408937454223633, + "learning_rate": 9.9995169999517e-06, + "loss": 1.0517, + "step": 484 + }, + { + "epoch": 0.22547652254765224, + "grad_norm": 7.790832996368408, + "learning_rate": 9.9995159999516e-06, + "loss": 0.8333, + "step": 485 + }, + { + "epoch": 0.22594142259414227, + "grad_norm": 5.1031670570373535, + "learning_rate": 9.999514999951501e-06, + "loss": 0.3627, + "step": 486 + }, + { + "epoch": 0.22640632264063226, + "grad_norm": 6.928451061248779, + "learning_rate": 9.999513999951402e-06, + "loss": 0.6859, + "step": 487 + }, + { + "epoch": 0.22687122268712226, + "grad_norm": 6.315357685089111, + "learning_rate": 9.999512999951301e-06, + "loss": 0.6028, + "step": 488 + }, + { + "epoch": 0.22733612273361228, + "grad_norm": 7.588940143585205, + "learning_rate": 9.999511999951202e-06, + "loss": 1.0034, + "step": 489 + }, + { + "epoch": 0.22780102278010228, + "grad_norm": 6.882167816162109, + "learning_rate": 9.9995109999511e-06, + "loss": 0.8942, + "step": 490 + }, + { + "epoch": 0.22826592282659228, + "grad_norm": 8.056483268737793, + "learning_rate": 9.999509999951e-06, + "loss": 0.741, + "step": 491 + }, + { + "epoch": 0.22873082287308227, + "grad_norm": 11.170844078063965, + "learning_rate": 9.9995089999509e-06, + "loss": 1.01, + "step": 492 + }, + { + "epoch": 0.2291957229195723, + "grad_norm": 7.661205768585205, + "learning_rate": 9.999507999950801e-06, + "loss": 0.7144, + "step": 493 + }, + { + "epoch": 0.2296606229660623, + "grad_norm": 5.905482769012451, + "learning_rate": 9.999506999950702e-06, + "loss": 0.4721, + "step": 494 + }, + { + "epoch": 0.2301255230125523, + "grad_norm": 7.769119739532471, + "learning_rate": 9.9995059999506e-06, + "loss": 0.6789, + "step": 495 + }, + { + "epoch": 0.23059042305904232, + "grad_norm": 8.539901733398438, + "learning_rate": 9.999504999950501e-06, + "loss": 0.6226, + "step": 496 + }, + { + "epoch": 0.2310553231055323, + "grad_norm": 11.434021949768066, + "learning_rate": 9.9995039999504e-06, + "loss": 1.1829, + "step": 497 + }, + { + "epoch": 0.2315202231520223, + "grad_norm": 8.325376510620117, + "learning_rate": 9.9995029999503e-06, + "loss": 0.7998, + "step": 498 + }, + { + "epoch": 0.23198512319851233, + "grad_norm": 9.647318840026855, + "learning_rate": 9.999501999950201e-06, + "loss": 1.1004, + "step": 499 + }, + { + "epoch": 0.23245002324500233, + "grad_norm": 6.5285468101501465, + "learning_rate": 9.9995009999501e-06, + "loss": 0.6185, + "step": 500 + }, + { + "epoch": 0.23291492329149233, + "grad_norm": 5.297163963317871, + "learning_rate": 9.999499999950001e-06, + "loss": 0.4656, + "step": 501 + }, + { + "epoch": 0.23337982333798232, + "grad_norm": 6.268032073974609, + "learning_rate": 9.9994989999499e-06, + "loss": 0.667, + "step": 502 + }, + { + "epoch": 0.23384472338447235, + "grad_norm": 7.185512542724609, + "learning_rate": 9.9994979999498e-06, + "loss": 0.8648, + "step": 503 + }, + { + "epoch": 0.23430962343096234, + "grad_norm": 7.655709266662598, + "learning_rate": 9.9994969999497e-06, + "loss": 0.8568, + "step": 504 + }, + { + "epoch": 0.23477452347745234, + "grad_norm": 10.160104751586914, + "learning_rate": 9.9994959999496e-06, + "loss": 1.1547, + "step": 505 + }, + { + "epoch": 0.23523942352394236, + "grad_norm": 4.271467685699463, + "learning_rate": 9.999494999949501e-06, + "loss": 0.2968, + "step": 506 + }, + { + "epoch": 0.23570432357043236, + "grad_norm": 6.41612434387207, + "learning_rate": 9.999493999949402e-06, + "loss": 0.7374, + "step": 507 + }, + { + "epoch": 0.23616922361692236, + "grad_norm": 12.275501251220703, + "learning_rate": 9.9994929999493e-06, + "loss": 1.1117, + "step": 508 + }, + { + "epoch": 0.23663412366341235, + "grad_norm": 8.228565216064453, + "learning_rate": 9.999491999949201e-06, + "loss": 0.929, + "step": 509 + }, + { + "epoch": 0.23709902370990238, + "grad_norm": 5.35228967666626, + "learning_rate": 9.9994909999491e-06, + "loss": 0.5329, + "step": 510 + }, + { + "epoch": 0.23756392375639238, + "grad_norm": 8.3446626663208, + "learning_rate": 9.999489999949001e-06, + "loss": 0.6703, + "step": 511 + }, + { + "epoch": 0.23802882380288237, + "grad_norm": 8.50833511352539, + "learning_rate": 9.999488999948902e-06, + "loss": 0.8182, + "step": 512 + }, + { + "epoch": 0.2384937238493724, + "grad_norm": 10.962355613708496, + "learning_rate": 9.9994879999488e-06, + "loss": 0.8974, + "step": 513 + }, + { + "epoch": 0.2389586238958624, + "grad_norm": 6.595521450042725, + "learning_rate": 9.999486999948701e-06, + "loss": 0.7281, + "step": 514 + }, + { + "epoch": 0.2394235239423524, + "grad_norm": 6.450929641723633, + "learning_rate": 9.9994859999486e-06, + "loss": 0.5888, + "step": 515 + }, + { + "epoch": 0.2398884239888424, + "grad_norm": 6.523677349090576, + "learning_rate": 9.9994849999485e-06, + "loss": 0.5276, + "step": 516 + }, + { + "epoch": 0.2403533240353324, + "grad_norm": 5.93682861328125, + "learning_rate": 9.999483999948401e-06, + "loss": 0.4664, + "step": 517 + }, + { + "epoch": 0.2408182240818224, + "grad_norm": 6.629948139190674, + "learning_rate": 9.9994829999483e-06, + "loss": 0.8562, + "step": 518 + }, + { + "epoch": 0.2412831241283124, + "grad_norm": 8.135575294494629, + "learning_rate": 9.999481999948201e-06, + "loss": 0.9798, + "step": 519 + }, + { + "epoch": 0.24174802417480243, + "grad_norm": 11.763782501220703, + "learning_rate": 9.999480999948102e-06, + "loss": 1.0396, + "step": 520 + }, + { + "epoch": 0.24221292422129242, + "grad_norm": 12.140315055847168, + "learning_rate": 9.999479999948e-06, + "loss": 0.8135, + "step": 521 + }, + { + "epoch": 0.24267782426778242, + "grad_norm": 7.540204048156738, + "learning_rate": 9.9994789999479e-06, + "loss": 0.7969, + "step": 522 + }, + { + "epoch": 0.24314272431427242, + "grad_norm": 7.701476097106934, + "learning_rate": 9.9994779999478e-06, + "loss": 0.888, + "step": 523 + }, + { + "epoch": 0.24360762436076244, + "grad_norm": 7.562258243560791, + "learning_rate": 9.999476999947701e-06, + "loss": 0.7791, + "step": 524 + }, + { + "epoch": 0.24407252440725244, + "grad_norm": 5.975600719451904, + "learning_rate": 9.999475999947602e-06, + "loss": 0.528, + "step": 525 + }, + { + "epoch": 0.24453742445374244, + "grad_norm": 9.480173110961914, + "learning_rate": 9.9994749999475e-06, + "loss": 0.6133, + "step": 526 + }, + { + "epoch": 0.24500232450023246, + "grad_norm": 9.845833778381348, + "learning_rate": 9.999473999947401e-06, + "loss": 0.9246, + "step": 527 + }, + { + "epoch": 0.24546722454672246, + "grad_norm": 7.719487190246582, + "learning_rate": 9.9994729999473e-06, + "loss": 0.8678, + "step": 528 + }, + { + "epoch": 0.24593212459321245, + "grad_norm": 17.737228393554688, + "learning_rate": 9.9994719999472e-06, + "loss": 0.9902, + "step": 529 + }, + { + "epoch": 0.24639702463970248, + "grad_norm": 9.731313705444336, + "learning_rate": 9.999470999947102e-06, + "loss": 0.6769, + "step": 530 + }, + { + "epoch": 0.24686192468619247, + "grad_norm": 8.44924259185791, + "learning_rate": 9.999469999947e-06, + "loss": 0.4666, + "step": 531 + }, + { + "epoch": 0.24732682473268247, + "grad_norm": 9.33395004272461, + "learning_rate": 9.999468999946901e-06, + "loss": 0.9531, + "step": 532 + }, + { + "epoch": 0.24779172477917247, + "grad_norm": 5.13940954208374, + "learning_rate": 9.9994679999468e-06, + "loss": 0.4778, + "step": 533 + }, + { + "epoch": 0.2482566248256625, + "grad_norm": 7.518404483795166, + "learning_rate": 9.9994669999467e-06, + "loss": 0.5717, + "step": 534 + }, + { + "epoch": 0.2487215248721525, + "grad_norm": 11.078776359558105, + "learning_rate": 9.9994659999466e-06, + "loss": 1.2404, + "step": 535 + }, + { + "epoch": 0.24918642491864249, + "grad_norm": 10.304607391357422, + "learning_rate": 9.9994649999465e-06, + "loss": 1.1034, + "step": 536 + }, + { + "epoch": 0.2496513249651325, + "grad_norm": 7.984467506408691, + "learning_rate": 9.999463999946401e-06, + "loss": 0.8253, + "step": 537 + }, + { + "epoch": 0.2501162250116225, + "grad_norm": 6.842625141143799, + "learning_rate": 9.999462999946302e-06, + "loss": 0.6571, + "step": 538 + }, + { + "epoch": 0.25058112505811253, + "grad_norm": 7.96988582611084, + "learning_rate": 9.9994619999462e-06, + "loss": 0.6682, + "step": 539 + }, + { + "epoch": 0.2510460251046025, + "grad_norm": 7.837594032287598, + "learning_rate": 9.999460999946101e-06, + "loss": 0.4803, + "step": 540 + }, + { + "epoch": 0.2515109251510925, + "grad_norm": 13.91159725189209, + "learning_rate": 9.999459999946e-06, + "loss": 1.1755, + "step": 541 + }, + { + "epoch": 0.2519758251975825, + "grad_norm": 6.0708231925964355, + "learning_rate": 9.999458999945901e-06, + "loss": 0.7012, + "step": 542 + }, + { + "epoch": 0.2524407252440725, + "grad_norm": 5.535235404968262, + "learning_rate": 9.999457999945802e-06, + "loss": 0.6488, + "step": 543 + }, + { + "epoch": 0.25290562529056254, + "grad_norm": 11.400635719299316, + "learning_rate": 9.9994569999457e-06, + "loss": 0.6437, + "step": 544 + }, + { + "epoch": 0.2533705253370525, + "grad_norm": 6.541198253631592, + "learning_rate": 9.999455999945601e-06, + "loss": 0.7113, + "step": 545 + }, + { + "epoch": 0.25383542538354253, + "grad_norm": 9.176311492919922, + "learning_rate": 9.9994549999455e-06, + "loss": 0.8503, + "step": 546 + }, + { + "epoch": 0.25430032543003256, + "grad_norm": 10.958780288696289, + "learning_rate": 9.9994539999454e-06, + "loss": 0.7153, + "step": 547 + }, + { + "epoch": 0.2547652254765225, + "grad_norm": 8.376851081848145, + "learning_rate": 9.9994529999453e-06, + "loss": 0.6369, + "step": 548 + }, + { + "epoch": 0.25523012552301255, + "grad_norm": 7.384718894958496, + "learning_rate": 9.9994519999452e-06, + "loss": 0.6324, + "step": 549 + }, + { + "epoch": 0.2556950255695026, + "grad_norm": 7.221340179443359, + "learning_rate": 9.999450999945101e-06, + "loss": 0.7283, + "step": 550 + }, + { + "epoch": 0.25615992561599255, + "grad_norm": 13.819881439208984, + "learning_rate": 9.999449999945002e-06, + "loss": 1.124, + "step": 551 + }, + { + "epoch": 0.25662482566248257, + "grad_norm": 6.655117511749268, + "learning_rate": 9.9994489999449e-06, + "loss": 0.6765, + "step": 552 + }, + { + "epoch": 0.2570897257089726, + "grad_norm": 10.409466743469238, + "learning_rate": 9.999447999944801e-06, + "loss": 1.159, + "step": 553 + }, + { + "epoch": 0.25755462575546256, + "grad_norm": 9.057459831237793, + "learning_rate": 9.9994469999447e-06, + "loss": 1.0038, + "step": 554 + }, + { + "epoch": 0.2580195258019526, + "grad_norm": 5.022355079650879, + "learning_rate": 9.999445999944601e-06, + "loss": 0.533, + "step": 555 + }, + { + "epoch": 0.2584844258484426, + "grad_norm": 12.428532600402832, + "learning_rate": 9.999444999944502e-06, + "loss": 1.0651, + "step": 556 + }, + { + "epoch": 0.2589493258949326, + "grad_norm": 5.006133556365967, + "learning_rate": 9.9994439999444e-06, + "loss": 0.404, + "step": 557 + }, + { + "epoch": 0.2594142259414226, + "grad_norm": 11.792752265930176, + "learning_rate": 9.999442999944301e-06, + "loss": 1.3443, + "step": 558 + }, + { + "epoch": 0.2598791259879126, + "grad_norm": 7.92299222946167, + "learning_rate": 9.9994419999442e-06, + "loss": 0.8035, + "step": 559 + }, + { + "epoch": 0.2603440260344026, + "grad_norm": 6.7516188621521, + "learning_rate": 9.999440999944101e-06, + "loss": 0.6359, + "step": 560 + }, + { + "epoch": 0.2608089260808926, + "grad_norm": 7.262951850891113, + "learning_rate": 9.999439999944e-06, + "loss": 0.8408, + "step": 561 + }, + { + "epoch": 0.2612738261273826, + "grad_norm": 13.230724334716797, + "learning_rate": 9.9994389999439e-06, + "loss": 1.1236, + "step": 562 + }, + { + "epoch": 0.2617387261738726, + "grad_norm": 19.633588790893555, + "learning_rate": 9.999437999943801e-06, + "loss": 1.0223, + "step": 563 + }, + { + "epoch": 0.26220362622036264, + "grad_norm": 8.23546314239502, + "learning_rate": 9.999436999943702e-06, + "loss": 0.8883, + "step": 564 + }, + { + "epoch": 0.2626685262668526, + "grad_norm": 6.832429885864258, + "learning_rate": 9.9994359999436e-06, + "loss": 0.7021, + "step": 565 + }, + { + "epoch": 0.26313342631334263, + "grad_norm": 9.22818660736084, + "learning_rate": 9.9994349999435e-06, + "loss": 1.0896, + "step": 566 + }, + { + "epoch": 0.26359832635983266, + "grad_norm": 11.163102149963379, + "learning_rate": 9.9994339999434e-06, + "loss": 1.1858, + "step": 567 + }, + { + "epoch": 0.2640632264063226, + "grad_norm": 6.470144748687744, + "learning_rate": 9.999432999943301e-06, + "loss": 0.6756, + "step": 568 + }, + { + "epoch": 0.26452812645281265, + "grad_norm": 9.18777084350586, + "learning_rate": 9.999431999943202e-06, + "loss": 1.2223, + "step": 569 + }, + { + "epoch": 0.2649930264993027, + "grad_norm": 9.353349685668945, + "learning_rate": 9.9994309999431e-06, + "loss": 0.8075, + "step": 570 + }, + { + "epoch": 0.26545792654579264, + "grad_norm": 7.36619758605957, + "learning_rate": 9.999429999943001e-06, + "loss": 1.1098, + "step": 571 + }, + { + "epoch": 0.26592282659228267, + "grad_norm": 11.240906715393066, + "learning_rate": 9.9994289999429e-06, + "loss": 1.0663, + "step": 572 + }, + { + "epoch": 0.26638772663877264, + "grad_norm": 8.558723449707031, + "learning_rate": 9.999427999942801e-06, + "loss": 1.1503, + "step": 573 + }, + { + "epoch": 0.26685262668526266, + "grad_norm": 7.396143436431885, + "learning_rate": 9.9994269999427e-06, + "loss": 0.6374, + "step": 574 + }, + { + "epoch": 0.2673175267317527, + "grad_norm": 6.5769267082214355, + "learning_rate": 9.9994259999426e-06, + "loss": 0.6555, + "step": 575 + }, + { + "epoch": 0.26778242677824265, + "grad_norm": 6.414391040802002, + "learning_rate": 9.999424999942501e-06, + "loss": 0.4803, + "step": 576 + }, + { + "epoch": 0.2682473268247327, + "grad_norm": 7.157084941864014, + "learning_rate": 9.9994239999424e-06, + "loss": 0.5573, + "step": 577 + }, + { + "epoch": 0.2687122268712227, + "grad_norm": 7.034717082977295, + "learning_rate": 9.999422999942301e-06, + "loss": 0.6346, + "step": 578 + }, + { + "epoch": 0.2691771269177127, + "grad_norm": 12.007482528686523, + "learning_rate": 9.9994219999422e-06, + "loss": 0.8655, + "step": 579 + }, + { + "epoch": 0.2696420269642027, + "grad_norm": 5.176408767700195, + "learning_rate": 9.9994209999421e-06, + "loss": 0.4615, + "step": 580 + }, + { + "epoch": 0.2701069270106927, + "grad_norm": 11.960644721984863, + "learning_rate": 9.999419999942001e-06, + "loss": 1.1239, + "step": 581 + }, + { + "epoch": 0.2705718270571827, + "grad_norm": 7.49454402923584, + "learning_rate": 9.999418999941902e-06, + "loss": 0.9724, + "step": 582 + }, + { + "epoch": 0.2710367271036727, + "grad_norm": 6.400639533996582, + "learning_rate": 9.9994179999418e-06, + "loss": 0.4446, + "step": 583 + }, + { + "epoch": 0.27150162715016274, + "grad_norm": 6.722039222717285, + "learning_rate": 9.999416999941701e-06, + "loss": 0.6653, + "step": 584 + }, + { + "epoch": 0.2719665271966527, + "grad_norm": 7.034249782562256, + "learning_rate": 9.9994159999416e-06, + "loss": 0.9461, + "step": 585 + }, + { + "epoch": 0.27243142724314273, + "grad_norm": 11.097445487976074, + "learning_rate": 9.999414999941501e-06, + "loss": 1.0197, + "step": 586 + }, + { + "epoch": 0.27289632728963276, + "grad_norm": 5.337111473083496, + "learning_rate": 9.9994139999414e-06, + "loss": 0.4843, + "step": 587 + }, + { + "epoch": 0.2733612273361227, + "grad_norm": 9.682230949401855, + "learning_rate": 9.9994129999413e-06, + "loss": 1.0421, + "step": 588 + }, + { + "epoch": 0.27382612738261275, + "grad_norm": 10.357223510742188, + "learning_rate": 9.999411999941201e-06, + "loss": 1.4936, + "step": 589 + }, + { + "epoch": 0.2742910274291027, + "grad_norm": 6.416812419891357, + "learning_rate": 9.9994109999411e-06, + "loss": 0.4515, + "step": 590 + }, + { + "epoch": 0.27475592747559274, + "grad_norm": 5.606784343719482, + "learning_rate": 9.999409999941001e-06, + "loss": 0.4146, + "step": 591 + }, + { + "epoch": 0.27522082752208277, + "grad_norm": 8.358246803283691, + "learning_rate": 9.9994089999409e-06, + "loss": 0.601, + "step": 592 + }, + { + "epoch": 0.27568572756857274, + "grad_norm": 6.6344313621521, + "learning_rate": 9.9994079999408e-06, + "loss": 0.6428, + "step": 593 + }, + { + "epoch": 0.27615062761506276, + "grad_norm": 4.615457057952881, + "learning_rate": 9.999406999940701e-06, + "loss": 0.4201, + "step": 594 + }, + { + "epoch": 0.2766155276615528, + "grad_norm": 6.483162879943848, + "learning_rate": 9.999405999940602e-06, + "loss": 0.7018, + "step": 595 + }, + { + "epoch": 0.27708042770804275, + "grad_norm": 5.983182430267334, + "learning_rate": 9.999404999940501e-06, + "loss": 0.6627, + "step": 596 + }, + { + "epoch": 0.2775453277545328, + "grad_norm": 5.813223361968994, + "learning_rate": 9.999403999940402e-06, + "loss": 0.7276, + "step": 597 + }, + { + "epoch": 0.2780102278010228, + "grad_norm": 6.928962707519531, + "learning_rate": 9.9994029999403e-06, + "loss": 0.5048, + "step": 598 + }, + { + "epoch": 0.27847512784751277, + "grad_norm": 5.225326061248779, + "learning_rate": 9.999401999940201e-06, + "loss": 0.3381, + "step": 599 + }, + { + "epoch": 0.2789400278940028, + "grad_norm": 9.919087409973145, + "learning_rate": 9.9994009999401e-06, + "loss": 0.9094, + "step": 600 + }, + { + "epoch": 0.2794049279404928, + "grad_norm": 11.632762908935547, + "learning_rate": 9.99939999994e-06, + "loss": 1.2108, + "step": 601 + }, + { + "epoch": 0.2798698279869828, + "grad_norm": 9.164794921875, + "learning_rate": 9.999398999939901e-06, + "loss": 0.9887, + "step": 602 + }, + { + "epoch": 0.2803347280334728, + "grad_norm": 8.147980690002441, + "learning_rate": 9.9993979999398e-06, + "loss": 0.9041, + "step": 603 + }, + { + "epoch": 0.2807996280799628, + "grad_norm": 7.805147647857666, + "learning_rate": 9.999396999939701e-06, + "loss": 0.817, + "step": 604 + }, + { + "epoch": 0.2812645281264528, + "grad_norm": 5.515103816986084, + "learning_rate": 9.9993959999396e-06, + "loss": 0.3795, + "step": 605 + }, + { + "epoch": 0.28172942817294283, + "grad_norm": 7.834560871124268, + "learning_rate": 9.9993949999395e-06, + "loss": 0.6609, + "step": 606 + }, + { + "epoch": 0.2821943282194328, + "grad_norm": 9.948634147644043, + "learning_rate": 9.999393999939401e-06, + "loss": 1.2672, + "step": 607 + }, + { + "epoch": 0.2826592282659228, + "grad_norm": 6.774588584899902, + "learning_rate": 9.999392999939302e-06, + "loss": 0.8744, + "step": 608 + }, + { + "epoch": 0.28312412831241285, + "grad_norm": 8.75749683380127, + "learning_rate": 9.999391999939201e-06, + "loss": 0.8122, + "step": 609 + }, + { + "epoch": 0.2835890283589028, + "grad_norm": 7.093261241912842, + "learning_rate": 9.9993909999391e-06, + "loss": 0.4545, + "step": 610 + }, + { + "epoch": 0.28405392840539284, + "grad_norm": 7.011244297027588, + "learning_rate": 9.999389999939e-06, + "loss": 0.8453, + "step": 611 + }, + { + "epoch": 0.28451882845188287, + "grad_norm": 5.268610954284668, + "learning_rate": 9.999388999938901e-06, + "loss": 0.4128, + "step": 612 + }, + { + "epoch": 0.28498372849837283, + "grad_norm": 4.663233280181885, + "learning_rate": 9.999387999938802e-06, + "loss": 0.3806, + "step": 613 + }, + { + "epoch": 0.28544862854486286, + "grad_norm": 7.7916259765625, + "learning_rate": 9.999386999938701e-06, + "loss": 0.5695, + "step": 614 + }, + { + "epoch": 0.2859135285913529, + "grad_norm": 11.315993309020996, + "learning_rate": 9.999385999938602e-06, + "loss": 0.9487, + "step": 615 + }, + { + "epoch": 0.28637842863784285, + "grad_norm": 12.5484037399292, + "learning_rate": 9.9993849999385e-06, + "loss": 0.5892, + "step": 616 + }, + { + "epoch": 0.2868433286843329, + "grad_norm": 9.621021270751953, + "learning_rate": 9.999383999938401e-06, + "loss": 1.1435, + "step": 617 + }, + { + "epoch": 0.28730822873082285, + "grad_norm": 8.90722370147705, + "learning_rate": 9.9993829999383e-06, + "loss": 0.6222, + "step": 618 + }, + { + "epoch": 0.28777312877731287, + "grad_norm": 11.631423950195312, + "learning_rate": 9.9993819999382e-06, + "loss": 1.0125, + "step": 619 + }, + { + "epoch": 0.2882380288238029, + "grad_norm": 8.280855178833008, + "learning_rate": 9.999380999938101e-06, + "loss": 0.7922, + "step": 620 + }, + { + "epoch": 0.28870292887029286, + "grad_norm": 7.233405590057373, + "learning_rate": 9.999379999938e-06, + "loss": 0.7464, + "step": 621 + }, + { + "epoch": 0.2891678289167829, + "grad_norm": 9.543344497680664, + "learning_rate": 9.999378999937901e-06, + "loss": 0.8782, + "step": 622 + }, + { + "epoch": 0.2896327289632729, + "grad_norm": 12.929618835449219, + "learning_rate": 9.9993779999378e-06, + "loss": 0.9245, + "step": 623 + }, + { + "epoch": 0.2900976290097629, + "grad_norm": 6.330466270446777, + "learning_rate": 9.9993769999377e-06, + "loss": 0.7135, + "step": 624 + }, + { + "epoch": 0.2905625290562529, + "grad_norm": 8.574357032775879, + "learning_rate": 9.999375999937601e-06, + "loss": 1.0064, + "step": 625 + }, + { + "epoch": 0.29102742910274293, + "grad_norm": 8.032795906066895, + "learning_rate": 9.999374999937502e-06, + "loss": 0.6664, + "step": 626 + }, + { + "epoch": 0.2914923291492329, + "grad_norm": 9.6295166015625, + "learning_rate": 9.999373999937401e-06, + "loss": 1.0698, + "step": 627 + }, + { + "epoch": 0.2919572291957229, + "grad_norm": 9.942059516906738, + "learning_rate": 9.999372999937302e-06, + "loss": 1.0926, + "step": 628 + }, + { + "epoch": 0.29242212924221295, + "grad_norm": 8.44109058380127, + "learning_rate": 9.9993719999372e-06, + "loss": 0.9444, + "step": 629 + }, + { + "epoch": 0.2928870292887029, + "grad_norm": 4.669463157653809, + "learning_rate": 9.999370999937101e-06, + "loss": 0.4578, + "step": 630 + }, + { + "epoch": 0.29335192933519294, + "grad_norm": 9.1060791015625, + "learning_rate": 9.999369999937e-06, + "loss": 0.6016, + "step": 631 + }, + { + "epoch": 0.29381682938168296, + "grad_norm": 8.133634567260742, + "learning_rate": 9.9993689999369e-06, + "loss": 0.8066, + "step": 632 + }, + { + "epoch": 0.29428172942817293, + "grad_norm": 7.093649864196777, + "learning_rate": 9.999367999936801e-06, + "loss": 1.103, + "step": 633 + }, + { + "epoch": 0.29474662947466296, + "grad_norm": 5.031795024871826, + "learning_rate": 9.9993669999367e-06, + "loss": 0.4046, + "step": 634 + }, + { + "epoch": 0.2952115295211529, + "grad_norm": 5.678933143615723, + "learning_rate": 9.999365999936601e-06, + "loss": 0.5507, + "step": 635 + }, + { + "epoch": 0.29567642956764295, + "grad_norm": 6.549428462982178, + "learning_rate": 9.9993649999365e-06, + "loss": 0.5474, + "step": 636 + }, + { + "epoch": 0.296141329614133, + "grad_norm": 6.7951130867004395, + "learning_rate": 9.9993639999364e-06, + "loss": 0.7366, + "step": 637 + }, + { + "epoch": 0.29660622966062294, + "grad_norm": 10.207186698913574, + "learning_rate": 9.999362999936301e-06, + "loss": 1.0272, + "step": 638 + }, + { + "epoch": 0.29707112970711297, + "grad_norm": 9.056384086608887, + "learning_rate": 9.999361999936202e-06, + "loss": 1.0269, + "step": 639 + }, + { + "epoch": 0.297536029753603, + "grad_norm": 8.569476127624512, + "learning_rate": 9.999360999936101e-06, + "loss": 0.7606, + "step": 640 + }, + { + "epoch": 0.29800092980009296, + "grad_norm": 7.5767083168029785, + "learning_rate": 9.999359999936e-06, + "loss": 0.645, + "step": 641 + }, + { + "epoch": 0.298465829846583, + "grad_norm": 6.882831573486328, + "learning_rate": 9.9993589999359e-06, + "loss": 0.7988, + "step": 642 + }, + { + "epoch": 0.298930729893073, + "grad_norm": 5.96028470993042, + "learning_rate": 9.9993579999358e-06, + "loss": 0.6608, + "step": 643 + }, + { + "epoch": 0.299395629939563, + "grad_norm": 10.06063175201416, + "learning_rate": 9.9993569999357e-06, + "loss": 1.0138, + "step": 644 + }, + { + "epoch": 0.299860529986053, + "grad_norm": 8.534432411193848, + "learning_rate": 9.999355999935601e-06, + "loss": 1.1652, + "step": 645 + }, + { + "epoch": 0.30032543003254303, + "grad_norm": 5.39007568359375, + "learning_rate": 9.999354999935502e-06, + "loss": 0.5707, + "step": 646 + }, + { + "epoch": 0.300790330079033, + "grad_norm": 6.288309097290039, + "learning_rate": 9.9993539999354e-06, + "loss": 0.7335, + "step": 647 + }, + { + "epoch": 0.301255230125523, + "grad_norm": 5.435885429382324, + "learning_rate": 9.999352999935301e-06, + "loss": 0.4094, + "step": 648 + }, + { + "epoch": 0.301720130172013, + "grad_norm": 10.231627464294434, + "learning_rate": 9.9993519999352e-06, + "loss": 1.1625, + "step": 649 + }, + { + "epoch": 0.302185030218503, + "grad_norm": 7.468729019165039, + "learning_rate": 9.9993509999351e-06, + "loss": 0.842, + "step": 650 + }, + { + "epoch": 0.30264993026499304, + "grad_norm": 8.094898223876953, + "learning_rate": 9.999349999935001e-06, + "loss": 0.7753, + "step": 651 + }, + { + "epoch": 0.303114830311483, + "grad_norm": 5.494748592376709, + "learning_rate": 9.9993489999349e-06, + "loss": 0.5405, + "step": 652 + }, + { + "epoch": 0.30357973035797303, + "grad_norm": 9.052905082702637, + "learning_rate": 9.999347999934801e-06, + "loss": 0.7942, + "step": 653 + }, + { + "epoch": 0.30404463040446306, + "grad_norm": 5.8653244972229, + "learning_rate": 9.9993469999347e-06, + "loss": 0.5852, + "step": 654 + }, + { + "epoch": 0.304509530450953, + "grad_norm": 9.74548625946045, + "learning_rate": 9.9993459999346e-06, + "loss": 0.9696, + "step": 655 + }, + { + "epoch": 0.30497443049744305, + "grad_norm": 10.153861999511719, + "learning_rate": 9.9993449999345e-06, + "loss": 0.8301, + "step": 656 + }, + { + "epoch": 0.3054393305439331, + "grad_norm": 8.55717945098877, + "learning_rate": 9.9993439999344e-06, + "loss": 0.909, + "step": 657 + }, + { + "epoch": 0.30590423059042304, + "grad_norm": 7.638063907623291, + "learning_rate": 9.999342999934301e-06, + "loss": 0.697, + "step": 658 + }, + { + "epoch": 0.30636913063691307, + "grad_norm": 9.875725746154785, + "learning_rate": 9.999341999934202e-06, + "loss": 0.597, + "step": 659 + }, + { + "epoch": 0.3068340306834031, + "grad_norm": 6.808213233947754, + "learning_rate": 9.9993409999341e-06, + "loss": 0.5179, + "step": 660 + }, + { + "epoch": 0.30729893072989306, + "grad_norm": 6.881759166717529, + "learning_rate": 9.999339999934001e-06, + "loss": 0.7329, + "step": 661 + }, + { + "epoch": 0.3077638307763831, + "grad_norm": 9.727898597717285, + "learning_rate": 9.9993389999339e-06, + "loss": 0.7006, + "step": 662 + }, + { + "epoch": 0.3082287308228731, + "grad_norm": 7.647222518920898, + "learning_rate": 9.999337999933801e-06, + "loss": 0.8937, + "step": 663 + }, + { + "epoch": 0.3086936308693631, + "grad_norm": 7.3665032386779785, + "learning_rate": 9.999336999933702e-06, + "loss": 0.9192, + "step": 664 + }, + { + "epoch": 0.3091585309158531, + "grad_norm": 9.146772384643555, + "learning_rate": 9.9993359999336e-06, + "loss": 0.7732, + "step": 665 + }, + { + "epoch": 0.30962343096234307, + "grad_norm": 9.692351341247559, + "learning_rate": 9.999334999933501e-06, + "loss": 0.7911, + "step": 666 + }, + { + "epoch": 0.3100883310088331, + "grad_norm": 6.084076404571533, + "learning_rate": 9.9993339999334e-06, + "loss": 0.6815, + "step": 667 + }, + { + "epoch": 0.3105532310553231, + "grad_norm": 6.3625593185424805, + "learning_rate": 9.9993329999333e-06, + "loss": 0.8136, + "step": 668 + }, + { + "epoch": 0.3110181311018131, + "grad_norm": 11.481460571289062, + "learning_rate": 9.9993319999332e-06, + "loss": 1.3649, + "step": 669 + }, + { + "epoch": 0.3114830311483031, + "grad_norm": 6.079823017120361, + "learning_rate": 9.9993309999331e-06, + "loss": 0.747, + "step": 670 + }, + { + "epoch": 0.31194793119479314, + "grad_norm": 9.31806468963623, + "learning_rate": 9.999329999933001e-06, + "loss": 0.765, + "step": 671 + }, + { + "epoch": 0.3124128312412831, + "grad_norm": 6.420828342437744, + "learning_rate": 9.999328999932902e-06, + "loss": 0.7776, + "step": 672 + }, + { + "epoch": 0.31287773128777313, + "grad_norm": 6.701288223266602, + "learning_rate": 9.9993279999328e-06, + "loss": 0.7142, + "step": 673 + }, + { + "epoch": 0.31334263133426316, + "grad_norm": 6.985331058502197, + "learning_rate": 9.9993269999327e-06, + "loss": 0.5423, + "step": 674 + }, + { + "epoch": 0.3138075313807531, + "grad_norm": 9.100723266601562, + "learning_rate": 9.9993259999326e-06, + "loss": 1.0876, + "step": 675 + }, + { + "epoch": 0.31427243142724315, + "grad_norm": 12.6342134475708, + "learning_rate": 9.999324999932501e-06, + "loss": 1.3338, + "step": 676 + }, + { + "epoch": 0.3147373314737332, + "grad_norm": 7.172208786010742, + "learning_rate": 9.999323999932402e-06, + "loss": 1.0658, + "step": 677 + }, + { + "epoch": 0.31520223152022314, + "grad_norm": 6.986496448516846, + "learning_rate": 9.9993229999323e-06, + "loss": 0.4958, + "step": 678 + }, + { + "epoch": 0.31566713156671317, + "grad_norm": 8.342713356018066, + "learning_rate": 9.999321999932201e-06, + "loss": 0.8286, + "step": 679 + }, + { + "epoch": 0.31613203161320313, + "grad_norm": 5.781044006347656, + "learning_rate": 9.9993209999321e-06, + "loss": 0.4903, + "step": 680 + }, + { + "epoch": 0.31659693165969316, + "grad_norm": 8.446235656738281, + "learning_rate": 9.999319999932001e-06, + "loss": 0.8266, + "step": 681 + }, + { + "epoch": 0.3170618317061832, + "grad_norm": 6.402684211730957, + "learning_rate": 9.999318999931902e-06, + "loss": 0.5619, + "step": 682 + }, + { + "epoch": 0.31752673175267315, + "grad_norm": 6.729738235473633, + "learning_rate": 9.9993179999318e-06, + "loss": 0.8917, + "step": 683 + }, + { + "epoch": 0.3179916317991632, + "grad_norm": 6.066112518310547, + "learning_rate": 9.999316999931701e-06, + "loss": 0.7721, + "step": 684 + }, + { + "epoch": 0.3184565318456532, + "grad_norm": 7.651355743408203, + "learning_rate": 9.9993159999316e-06, + "loss": 0.7294, + "step": 685 + }, + { + "epoch": 0.31892143189214317, + "grad_norm": 9.900164604187012, + "learning_rate": 9.9993149999315e-06, + "loss": 1.0122, + "step": 686 + }, + { + "epoch": 0.3193863319386332, + "grad_norm": 10.292511940002441, + "learning_rate": 9.9993139999314e-06, + "loss": 1.0124, + "step": 687 + }, + { + "epoch": 0.3198512319851232, + "grad_norm": 4.302212715148926, + "learning_rate": 9.9993129999313e-06, + "loss": 0.5144, + "step": 688 + }, + { + "epoch": 0.3203161320316132, + "grad_norm": 8.767983436584473, + "learning_rate": 9.999311999931201e-06, + "loss": 0.8215, + "step": 689 + }, + { + "epoch": 0.3207810320781032, + "grad_norm": 5.1155853271484375, + "learning_rate": 9.999310999931102e-06, + "loss": 0.5228, + "step": 690 + }, + { + "epoch": 0.32124593212459324, + "grad_norm": 9.273005485534668, + "learning_rate": 9.999309999931e-06, + "loss": 0.6815, + "step": 691 + }, + { + "epoch": 0.3217108321710832, + "grad_norm": 10.411373138427734, + "learning_rate": 9.999308999930901e-06, + "loss": 1.2132, + "step": 692 + }, + { + "epoch": 0.32217573221757323, + "grad_norm": 6.937475204467773, + "learning_rate": 9.9993079999308e-06, + "loss": 0.8121, + "step": 693 + }, + { + "epoch": 0.32264063226406325, + "grad_norm": 7.716465473175049, + "learning_rate": 9.999306999930701e-06, + "loss": 0.6451, + "step": 694 + }, + { + "epoch": 0.3231055323105532, + "grad_norm": 8.266775131225586, + "learning_rate": 9.999305999930602e-06, + "loss": 0.9765, + "step": 695 + }, + { + "epoch": 0.32357043235704325, + "grad_norm": 14.949325561523438, + "learning_rate": 9.9993049999305e-06, + "loss": 1.0075, + "step": 696 + }, + { + "epoch": 0.3240353324035332, + "grad_norm": 10.170968055725098, + "learning_rate": 9.999303999930401e-06, + "loss": 0.9796, + "step": 697 + }, + { + "epoch": 0.32450023245002324, + "grad_norm": 8.349474906921387, + "learning_rate": 9.9993029999303e-06, + "loss": 0.8872, + "step": 698 + }, + { + "epoch": 0.32496513249651326, + "grad_norm": 11.549878120422363, + "learning_rate": 9.999301999930201e-06, + "loss": 1.0818, + "step": 699 + }, + { + "epoch": 0.32543003254300323, + "grad_norm": 7.878251075744629, + "learning_rate": 9.9993009999301e-06, + "loss": 0.8011, + "step": 700 + }, + { + "epoch": 0.32589493258949326, + "grad_norm": 7.70474910736084, + "learning_rate": 9.99929999993e-06, + "loss": 0.6236, + "step": 701 + }, + { + "epoch": 0.3263598326359833, + "grad_norm": 6.497116565704346, + "learning_rate": 9.999298999929901e-06, + "loss": 0.6365, + "step": 702 + }, + { + "epoch": 0.32682473268247325, + "grad_norm": 7.746426582336426, + "learning_rate": 9.999297999929802e-06, + "loss": 0.9236, + "step": 703 + }, + { + "epoch": 0.3272896327289633, + "grad_norm": 3.8815536499023438, + "learning_rate": 9.9992969999297e-06, + "loss": 0.3354, + "step": 704 + }, + { + "epoch": 0.3277545327754533, + "grad_norm": 6.74808406829834, + "learning_rate": 9.999295999929601e-06, + "loss": 0.8496, + "step": 705 + }, + { + "epoch": 0.32821943282194327, + "grad_norm": 9.633684158325195, + "learning_rate": 9.9992949999295e-06, + "loss": 1.18, + "step": 706 + }, + { + "epoch": 0.3286843328684333, + "grad_norm": 8.208608627319336, + "learning_rate": 9.999293999929401e-06, + "loss": 0.7683, + "step": 707 + }, + { + "epoch": 0.3291492329149233, + "grad_norm": 9.090843200683594, + "learning_rate": 9.999292999929302e-06, + "loss": 0.8905, + "step": 708 + }, + { + "epoch": 0.3296141329614133, + "grad_norm": 6.549350261688232, + "learning_rate": 9.9992919999292e-06, + "loss": 0.4455, + "step": 709 + }, + { + "epoch": 0.3300790330079033, + "grad_norm": 7.522887706756592, + "learning_rate": 9.999290999929101e-06, + "loss": 0.5887, + "step": 710 + }, + { + "epoch": 0.3305439330543933, + "grad_norm": 9.751531600952148, + "learning_rate": 9.999289999929e-06, + "loss": 0.9195, + "step": 711 + }, + { + "epoch": 0.3310088331008833, + "grad_norm": 6.60086727142334, + "learning_rate": 9.999288999928901e-06, + "loss": 0.8108, + "step": 712 + }, + { + "epoch": 0.33147373314737333, + "grad_norm": 9.040903091430664, + "learning_rate": 9.9992879999288e-06, + "loss": 1.2369, + "step": 713 + }, + { + "epoch": 0.3319386331938633, + "grad_norm": 11.704545974731445, + "learning_rate": 9.9992869999287e-06, + "loss": 1.2378, + "step": 714 + }, + { + "epoch": 0.3324035332403533, + "grad_norm": 6.062154769897461, + "learning_rate": 9.999285999928601e-06, + "loss": 0.5675, + "step": 715 + }, + { + "epoch": 0.33286843328684335, + "grad_norm": 8.876365661621094, + "learning_rate": 9.999284999928502e-06, + "loss": 0.9172, + "step": 716 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 9.96982479095459, + "learning_rate": 9.9992839999284e-06, + "loss": 1.149, + "step": 717 + }, + { + "epoch": 0.33379823337982334, + "grad_norm": 6.419450283050537, + "learning_rate": 9.9992829999283e-06, + "loss": 0.7291, + "step": 718 + }, + { + "epoch": 0.33426313342631336, + "grad_norm": 9.75366497039795, + "learning_rate": 9.9992819999282e-06, + "loss": 1.083, + "step": 719 + }, + { + "epoch": 0.33472803347280333, + "grad_norm": 8.09605598449707, + "learning_rate": 9.999280999928101e-06, + "loss": 0.8197, + "step": 720 + }, + { + "epoch": 0.33519293351929336, + "grad_norm": 7.144834041595459, + "learning_rate": 9.999279999928002e-06, + "loss": 0.886, + "step": 721 + }, + { + "epoch": 0.3356578335657834, + "grad_norm": 9.010927200317383, + "learning_rate": 9.9992789999279e-06, + "loss": 1.0524, + "step": 722 + }, + { + "epoch": 0.33612273361227335, + "grad_norm": 9.0764741897583, + "learning_rate": 9.999277999927801e-06, + "loss": 1.2744, + "step": 723 + }, + { + "epoch": 0.3365876336587634, + "grad_norm": 7.260586261749268, + "learning_rate": 9.9992769999277e-06, + "loss": 0.8086, + "step": 724 + }, + { + "epoch": 0.33705253370525334, + "grad_norm": 8.807286262512207, + "learning_rate": 9.999275999927601e-06, + "loss": 0.8795, + "step": 725 + }, + { + "epoch": 0.33751743375174337, + "grad_norm": 4.646021366119385, + "learning_rate": 9.9992749999275e-06, + "loss": 0.3275, + "step": 726 + }, + { + "epoch": 0.3379823337982334, + "grad_norm": 7.317996501922607, + "learning_rate": 9.9992739999274e-06, + "loss": 0.764, + "step": 727 + }, + { + "epoch": 0.33844723384472336, + "grad_norm": 8.594866752624512, + "learning_rate": 9.999272999927301e-06, + "loss": 0.8645, + "step": 728 + }, + { + "epoch": 0.3389121338912134, + "grad_norm": 10.238399505615234, + "learning_rate": 9.9992719999272e-06, + "loss": 0.9305, + "step": 729 + }, + { + "epoch": 0.3393770339377034, + "grad_norm": 8.378079414367676, + "learning_rate": 9.999270999927101e-06, + "loss": 0.7333, + "step": 730 + }, + { + "epoch": 0.3398419339841934, + "grad_norm": 6.922446250915527, + "learning_rate": 9.999269999927e-06, + "loss": 0.8925, + "step": 731 + }, + { + "epoch": 0.3403068340306834, + "grad_norm": 8.58515453338623, + "learning_rate": 9.9992689999269e-06, + "loss": 0.7266, + "step": 732 + }, + { + "epoch": 0.3407717340771734, + "grad_norm": 7.034302711486816, + "learning_rate": 9.999267999926801e-06, + "loss": 1.1224, + "step": 733 + }, + { + "epoch": 0.3412366341236634, + "grad_norm": 4.696310997009277, + "learning_rate": 9.999266999926702e-06, + "loss": 0.2539, + "step": 734 + }, + { + "epoch": 0.3417015341701534, + "grad_norm": 6.231134414672852, + "learning_rate": 9.9992659999266e-06, + "loss": 0.7339, + "step": 735 + }, + { + "epoch": 0.34216643421664344, + "grad_norm": 8.309358596801758, + "learning_rate": 9.999264999926501e-06, + "loss": 0.8395, + "step": 736 + }, + { + "epoch": 0.3426313342631334, + "grad_norm": 5.856487274169922, + "learning_rate": 9.9992639999264e-06, + "loss": 0.6175, + "step": 737 + }, + { + "epoch": 0.34309623430962344, + "grad_norm": 8.423868179321289, + "learning_rate": 9.999262999926301e-06, + "loss": 1.0918, + "step": 738 + }, + { + "epoch": 0.34356113435611346, + "grad_norm": 6.200765609741211, + "learning_rate": 9.9992619999262e-06, + "loss": 0.6279, + "step": 739 + }, + { + "epoch": 0.34402603440260343, + "grad_norm": 6.159182548522949, + "learning_rate": 9.9992609999261e-06, + "loss": 0.5848, + "step": 740 + }, + { + "epoch": 0.34449093444909346, + "grad_norm": 11.884621620178223, + "learning_rate": 9.999259999926001e-06, + "loss": 1.3165, + "step": 741 + }, + { + "epoch": 0.3449558344955834, + "grad_norm": 9.143893241882324, + "learning_rate": 9.9992589999259e-06, + "loss": 0.7127, + "step": 742 + }, + { + "epoch": 0.34542073454207345, + "grad_norm": 8.780732154846191, + "learning_rate": 9.999257999925801e-06, + "loss": 0.7969, + "step": 743 + }, + { + "epoch": 0.3458856345885635, + "grad_norm": 11.326672554016113, + "learning_rate": 9.9992569999257e-06, + "loss": 0.8123, + "step": 744 + }, + { + "epoch": 0.34635053463505344, + "grad_norm": 7.559329509735107, + "learning_rate": 9.9992559999256e-06, + "loss": 0.9679, + "step": 745 + }, + { + "epoch": 0.34681543468154347, + "grad_norm": 8.388748168945312, + "learning_rate": 9.999254999925501e-06, + "loss": 0.9887, + "step": 746 + }, + { + "epoch": 0.3472803347280335, + "grad_norm": 17.457496643066406, + "learning_rate": 9.999253999925402e-06, + "loss": 1.4443, + "step": 747 + }, + { + "epoch": 0.34774523477452346, + "grad_norm": 7.361182689666748, + "learning_rate": 9.999252999925301e-06, + "loss": 0.7739, + "step": 748 + }, + { + "epoch": 0.3482101348210135, + "grad_norm": 8.572721481323242, + "learning_rate": 9.999251999925202e-06, + "loss": 0.9423, + "step": 749 + }, + { + "epoch": 0.3486750348675035, + "grad_norm": 9.235068321228027, + "learning_rate": 9.9992509999251e-06, + "loss": 0.5725, + "step": 750 + }, + { + "epoch": 0.3491399349139935, + "grad_norm": 8.21972370147705, + "learning_rate": 9.999249999925001e-06, + "loss": 0.7857, + "step": 751 + }, + { + "epoch": 0.3496048349604835, + "grad_norm": 7.796825885772705, + "learning_rate": 9.9992489999249e-06, + "loss": 0.6856, + "step": 752 + }, + { + "epoch": 0.3500697350069735, + "grad_norm": 9.95273208618164, + "learning_rate": 9.9992479999248e-06, + "loss": 1.5018, + "step": 753 + }, + { + "epoch": 0.3505346350534635, + "grad_norm": 8.9869966506958, + "learning_rate": 9.999246999924701e-06, + "loss": 0.8875, + "step": 754 + }, + { + "epoch": 0.3509995350999535, + "grad_norm": 7.487219333648682, + "learning_rate": 9.9992459999246e-06, + "loss": 0.5417, + "step": 755 + }, + { + "epoch": 0.3514644351464435, + "grad_norm": 4.667087078094482, + "learning_rate": 9.999244999924501e-06, + "loss": 0.3381, + "step": 756 + }, + { + "epoch": 0.3519293351929335, + "grad_norm": 9.1000394821167, + "learning_rate": 9.9992439999244e-06, + "loss": 0.9371, + "step": 757 + }, + { + "epoch": 0.35239423523942354, + "grad_norm": 7.29518461227417, + "learning_rate": 9.9992429999243e-06, + "loss": 0.8096, + "step": 758 + }, + { + "epoch": 0.3528591352859135, + "grad_norm": 6.404296398162842, + "learning_rate": 9.999241999924201e-06, + "loss": 0.634, + "step": 759 + }, + { + "epoch": 0.35332403533240353, + "grad_norm": 6.637044429779053, + "learning_rate": 9.999240999924102e-06, + "loss": 0.4532, + "step": 760 + }, + { + "epoch": 0.35378893537889355, + "grad_norm": 7.208938121795654, + "learning_rate": 9.999239999924001e-06, + "loss": 0.879, + "step": 761 + }, + { + "epoch": 0.3542538354253835, + "grad_norm": 8.90187931060791, + "learning_rate": 9.9992389999239e-06, + "loss": 0.8539, + "step": 762 + }, + { + "epoch": 0.35471873547187355, + "grad_norm": 11.424521446228027, + "learning_rate": 9.9992379999238e-06, + "loss": 1.453, + "step": 763 + }, + { + "epoch": 0.35518363551836357, + "grad_norm": 6.313692569732666, + "learning_rate": 9.999236999923701e-06, + "loss": 0.7419, + "step": 764 + }, + { + "epoch": 0.35564853556485354, + "grad_norm": 7.748394012451172, + "learning_rate": 9.9992359999236e-06, + "loss": 0.6041, + "step": 765 + }, + { + "epoch": 0.35611343561134357, + "grad_norm": 5.7579874992370605, + "learning_rate": 9.999234999923501e-06, + "loss": 0.3747, + "step": 766 + }, + { + "epoch": 0.3565783356578336, + "grad_norm": 9.530736923217773, + "learning_rate": 9.999233999923402e-06, + "loss": 0.759, + "step": 767 + }, + { + "epoch": 0.35704323570432356, + "grad_norm": 6.463360786437988, + "learning_rate": 9.9992329999233e-06, + "loss": 0.6593, + "step": 768 + }, + { + "epoch": 0.3575081357508136, + "grad_norm": 9.083394050598145, + "learning_rate": 9.999231999923201e-06, + "loss": 1.0065, + "step": 769 + }, + { + "epoch": 0.3579730357973036, + "grad_norm": 7.982889652252197, + "learning_rate": 9.9992309999231e-06, + "loss": 0.5432, + "step": 770 + }, + { + "epoch": 0.3584379358437936, + "grad_norm": 10.948837280273438, + "learning_rate": 9.999229999923e-06, + "loss": 1.1049, + "step": 771 + }, + { + "epoch": 0.3589028358902836, + "grad_norm": 8.75843620300293, + "learning_rate": 9.999228999922901e-06, + "loss": 0.6796, + "step": 772 + }, + { + "epoch": 0.35936773593677357, + "grad_norm": 7.015105724334717, + "learning_rate": 9.9992279999228e-06, + "loss": 0.5231, + "step": 773 + }, + { + "epoch": 0.3598326359832636, + "grad_norm": 6.255554676055908, + "learning_rate": 9.999226999922701e-06, + "loss": 0.8084, + "step": 774 + }, + { + "epoch": 0.3602975360297536, + "grad_norm": 7.806128978729248, + "learning_rate": 9.9992259999226e-06, + "loss": 0.9347, + "step": 775 + }, + { + "epoch": 0.3607624360762436, + "grad_norm": 6.597543716430664, + "learning_rate": 9.9992249999225e-06, + "loss": 0.7482, + "step": 776 + }, + { + "epoch": 0.3612273361227336, + "grad_norm": 6.406163215637207, + "learning_rate": 9.999223999922401e-06, + "loss": 0.472, + "step": 777 + }, + { + "epoch": 0.36169223616922364, + "grad_norm": 5.887228012084961, + "learning_rate": 9.999222999922302e-06, + "loss": 0.6974, + "step": 778 + }, + { + "epoch": 0.3621571362157136, + "grad_norm": 11.7066068649292, + "learning_rate": 9.999221999922201e-06, + "loss": 0.906, + "step": 779 + }, + { + "epoch": 0.36262203626220363, + "grad_norm": 5.383739948272705, + "learning_rate": 9.999220999922102e-06, + "loss": 0.5188, + "step": 780 + }, + { + "epoch": 0.36308693630869365, + "grad_norm": 6.491984844207764, + "learning_rate": 9.999219999922e-06, + "loss": 0.5361, + "step": 781 + }, + { + "epoch": 0.3635518363551836, + "grad_norm": 8.329986572265625, + "learning_rate": 9.9992189999219e-06, + "loss": 0.6696, + "step": 782 + }, + { + "epoch": 0.36401673640167365, + "grad_norm": 8.831902503967285, + "learning_rate": 9.9992179999218e-06, + "loss": 0.6799, + "step": 783 + }, + { + "epoch": 0.36448163644816367, + "grad_norm": 5.972829818725586, + "learning_rate": 9.999216999921701e-06, + "loss": 0.5922, + "step": 784 + }, + { + "epoch": 0.36494653649465364, + "grad_norm": 16.160491943359375, + "learning_rate": 9.999215999921602e-06, + "loss": 1.4477, + "step": 785 + }, + { + "epoch": 0.36541143654114366, + "grad_norm": 7.93678617477417, + "learning_rate": 9.9992149999215e-06, + "loss": 0.6983, + "step": 786 + }, + { + "epoch": 0.36587633658763363, + "grad_norm": 5.831811428070068, + "learning_rate": 9.999213999921401e-06, + "loss": 0.4545, + "step": 787 + }, + { + "epoch": 0.36634123663412366, + "grad_norm": 5.272959232330322, + "learning_rate": 9.9992129999213e-06, + "loss": 0.506, + "step": 788 + }, + { + "epoch": 0.3668061366806137, + "grad_norm": 7.811851501464844, + "learning_rate": 9.9992119999212e-06, + "loss": 0.904, + "step": 789 + }, + { + "epoch": 0.36727103672710365, + "grad_norm": 5.326284408569336, + "learning_rate": 9.999210999921101e-06, + "loss": 0.4603, + "step": 790 + }, + { + "epoch": 0.3677359367735937, + "grad_norm": 5.619653701782227, + "learning_rate": 9.999209999921002e-06, + "loss": 0.431, + "step": 791 + }, + { + "epoch": 0.3682008368200837, + "grad_norm": 9.569487571716309, + "learning_rate": 9.999208999920901e-06, + "loss": 0.6451, + "step": 792 + }, + { + "epoch": 0.36866573686657367, + "grad_norm": 8.0109224319458, + "learning_rate": 9.9992079999208e-06, + "loss": 0.747, + "step": 793 + }, + { + "epoch": 0.3691306369130637, + "grad_norm": 5.6519999504089355, + "learning_rate": 9.9992069999207e-06, + "loss": 0.6654, + "step": 794 + }, + { + "epoch": 0.3695955369595537, + "grad_norm": 7.302247047424316, + "learning_rate": 9.9992059999206e-06, + "loss": 0.5344, + "step": 795 + }, + { + "epoch": 0.3700604370060437, + "grad_norm": 5.041618824005127, + "learning_rate": 9.9992049999205e-06, + "loss": 0.4189, + "step": 796 + }, + { + "epoch": 0.3705253370525337, + "grad_norm": 11.222954750061035, + "learning_rate": 9.999203999920401e-06, + "loss": 1.1672, + "step": 797 + }, + { + "epoch": 0.37099023709902373, + "grad_norm": 6.815182209014893, + "learning_rate": 9.999202999920302e-06, + "loss": 0.7242, + "step": 798 + }, + { + "epoch": 0.3714551371455137, + "grad_norm": 7.181813716888428, + "learning_rate": 9.9992019999202e-06, + "loss": 0.6185, + "step": 799 + }, + { + "epoch": 0.3719200371920037, + "grad_norm": 6.474527359008789, + "learning_rate": 9.999200999920101e-06, + "loss": 0.4889, + "step": 800 + }, + { + "epoch": 0.3723849372384937, + "grad_norm": 8.41978931427002, + "learning_rate": 9.99919999992e-06, + "loss": 0.6049, + "step": 801 + }, + { + "epoch": 0.3728498372849837, + "grad_norm": 7.1072678565979, + "learning_rate": 9.999198999919901e-06, + "loss": 0.8686, + "step": 802 + }, + { + "epoch": 0.37331473733147374, + "grad_norm": 13.258295059204102, + "learning_rate": 9.999197999919802e-06, + "loss": 1.7923, + "step": 803 + }, + { + "epoch": 0.3737796373779637, + "grad_norm": 7.993046283721924, + "learning_rate": 9.9991969999197e-06, + "loss": 1.1072, + "step": 804 + }, + { + "epoch": 0.37424453742445374, + "grad_norm": 11.521987915039062, + "learning_rate": 9.999195999919601e-06, + "loss": 0.862, + "step": 805 + }, + { + "epoch": 0.37470943747094376, + "grad_norm": 8.727118492126465, + "learning_rate": 9.9991949999195e-06, + "loss": 1.0848, + "step": 806 + }, + { + "epoch": 0.37517433751743373, + "grad_norm": 8.784647941589355, + "learning_rate": 9.9991939999194e-06, + "loss": 0.9128, + "step": 807 + }, + { + "epoch": 0.37563923756392376, + "grad_norm": 8.82391357421875, + "learning_rate": 9.9991929999193e-06, + "loss": 1.1019, + "step": 808 + }, + { + "epoch": 0.3761041376104138, + "grad_norm": 11.026349067687988, + "learning_rate": 9.9991919999192e-06, + "loss": 0.6306, + "step": 809 + }, + { + "epoch": 0.37656903765690375, + "grad_norm": 5.4052276611328125, + "learning_rate": 9.999190999919101e-06, + "loss": 0.5974, + "step": 810 + }, + { + "epoch": 0.3770339377033938, + "grad_norm": 9.46250057220459, + "learning_rate": 9.999189999919002e-06, + "loss": 1.1768, + "step": 811 + }, + { + "epoch": 0.3774988377498838, + "grad_norm": 6.950920104980469, + "learning_rate": 9.9991889999189e-06, + "loss": 0.4523, + "step": 812 + }, + { + "epoch": 0.37796373779637377, + "grad_norm": 11.489983558654785, + "learning_rate": 9.999187999918801e-06, + "loss": 0.4488, + "step": 813 + }, + { + "epoch": 0.3784286378428638, + "grad_norm": 7.808286666870117, + "learning_rate": 9.9991869999187e-06, + "loss": 0.8457, + "step": 814 + }, + { + "epoch": 0.3788935378893538, + "grad_norm": 7.63928747177124, + "learning_rate": 9.999185999918601e-06, + "loss": 0.7752, + "step": 815 + }, + { + "epoch": 0.3793584379358438, + "grad_norm": 12.34257984161377, + "learning_rate": 9.999184999918502e-06, + "loss": 0.8517, + "step": 816 + }, + { + "epoch": 0.3798233379823338, + "grad_norm": 5.517758846282959, + "learning_rate": 9.9991839999184e-06, + "loss": 0.7328, + "step": 817 + }, + { + "epoch": 0.3802882380288238, + "grad_norm": 8.388490676879883, + "learning_rate": 9.999182999918301e-06, + "loss": 0.8368, + "step": 818 + }, + { + "epoch": 0.3807531380753138, + "grad_norm": 11.529109001159668, + "learning_rate": 9.9991819999182e-06, + "loss": 1.5869, + "step": 819 + }, + { + "epoch": 0.3812180381218038, + "grad_norm": 9.217801094055176, + "learning_rate": 9.9991809999181e-06, + "loss": 0.732, + "step": 820 + }, + { + "epoch": 0.3816829381682938, + "grad_norm": 7.2455854415893555, + "learning_rate": 9.999179999918e-06, + "loss": 0.7905, + "step": 821 + }, + { + "epoch": 0.3821478382147838, + "grad_norm": 12.127432823181152, + "learning_rate": 9.9991789999179e-06, + "loss": 0.7662, + "step": 822 + }, + { + "epoch": 0.38261273826127384, + "grad_norm": 5.092895030975342, + "learning_rate": 9.999177999917801e-06, + "loss": 0.3332, + "step": 823 + }, + { + "epoch": 0.3830776383077638, + "grad_norm": 9.73701000213623, + "learning_rate": 9.999176999917702e-06, + "loss": 0.6974, + "step": 824 + }, + { + "epoch": 0.38354253835425384, + "grad_norm": 6.644396781921387, + "learning_rate": 9.9991759999176e-06, + "loss": 0.8032, + "step": 825 + }, + { + "epoch": 0.38400743840074386, + "grad_norm": 9.466633796691895, + "learning_rate": 9.9991749999175e-06, + "loss": 0.967, + "step": 826 + }, + { + "epoch": 0.38447233844723383, + "grad_norm": 10.396162033081055, + "learning_rate": 9.9991739999174e-06, + "loss": 0.9569, + "step": 827 + }, + { + "epoch": 0.38493723849372385, + "grad_norm": 12.631900787353516, + "learning_rate": 9.999172999917301e-06, + "loss": 1.1241, + "step": 828 + }, + { + "epoch": 0.3854021385402139, + "grad_norm": 7.108704090118408, + "learning_rate": 9.999171999917202e-06, + "loss": 0.8536, + "step": 829 + }, + { + "epoch": 0.38586703858670385, + "grad_norm": 9.185564041137695, + "learning_rate": 9.9991709999171e-06, + "loss": 1.1285, + "step": 830 + }, + { + "epoch": 0.38633193863319387, + "grad_norm": 7.34623908996582, + "learning_rate": 9.999169999917001e-06, + "loss": 0.8178, + "step": 831 + }, + { + "epoch": 0.38679683867968384, + "grad_norm": 5.920477867126465, + "learning_rate": 9.9991689999169e-06, + "loss": 0.5097, + "step": 832 + }, + { + "epoch": 0.38726173872617387, + "grad_norm": 6.950290679931641, + "learning_rate": 9.999167999916801e-06, + "loss": 0.6122, + "step": 833 + }, + { + "epoch": 0.3877266387726639, + "grad_norm": 5.31902551651001, + "learning_rate": 9.9991669999167e-06, + "loss": 0.6205, + "step": 834 + }, + { + "epoch": 0.38819153881915386, + "grad_norm": 12.910463333129883, + "learning_rate": 9.9991659999166e-06, + "loss": 1.1839, + "step": 835 + }, + { + "epoch": 0.3886564388656439, + "grad_norm": 5.920276165008545, + "learning_rate": 9.999164999916501e-06, + "loss": 0.5932, + "step": 836 + }, + { + "epoch": 0.3891213389121339, + "grad_norm": 8.114962577819824, + "learning_rate": 9.9991639999164e-06, + "loss": 0.7129, + "step": 837 + }, + { + "epoch": 0.3895862389586239, + "grad_norm": 6.562771320343018, + "learning_rate": 9.9991629999163e-06, + "loss": 0.5059, + "step": 838 + }, + { + "epoch": 0.3900511390051139, + "grad_norm": 5.4476470947265625, + "learning_rate": 9.9991619999162e-06, + "loss": 0.473, + "step": 839 + }, + { + "epoch": 0.3905160390516039, + "grad_norm": 6.995213031768799, + "learning_rate": 9.9991609999161e-06, + "loss": 0.7863, + "step": 840 + }, + { + "epoch": 0.3909809390980939, + "grad_norm": 11.817872047424316, + "learning_rate": 9.999159999916001e-06, + "loss": 1.1661, + "step": 841 + }, + { + "epoch": 0.3914458391445839, + "grad_norm": 8.006256103515625, + "learning_rate": 9.999158999915902e-06, + "loss": 0.9857, + "step": 842 + }, + { + "epoch": 0.39191073919107394, + "grad_norm": 8.517629623413086, + "learning_rate": 9.9991579999158e-06, + "loss": 0.6969, + "step": 843 + }, + { + "epoch": 0.3923756392375639, + "grad_norm": 17.66554832458496, + "learning_rate": 9.999156999915701e-06, + "loss": 0.6601, + "step": 844 + }, + { + "epoch": 0.39284053928405394, + "grad_norm": 9.122300148010254, + "learning_rate": 9.9991559999156e-06, + "loss": 0.8894, + "step": 845 + }, + { + "epoch": 0.39330543933054396, + "grad_norm": 10.8937406539917, + "learning_rate": 9.999154999915501e-06, + "loss": 1.0954, + "step": 846 + }, + { + "epoch": 0.39377033937703393, + "grad_norm": 5.9544358253479, + "learning_rate": 9.999153999915402e-06, + "loss": 0.6986, + "step": 847 + }, + { + "epoch": 0.39423523942352395, + "grad_norm": 4.907937049865723, + "learning_rate": 9.9991529999153e-06, + "loss": 0.4081, + "step": 848 + }, + { + "epoch": 0.3947001394700139, + "grad_norm": 8.26550006866455, + "learning_rate": 9.999151999915201e-06, + "loss": 0.8091, + "step": 849 + }, + { + "epoch": 0.39516503951650395, + "grad_norm": 7.496113300323486, + "learning_rate": 9.9991509999151e-06, + "loss": 0.7573, + "step": 850 + }, + { + "epoch": 0.39562993956299397, + "grad_norm": 13.209342002868652, + "learning_rate": 9.999149999915001e-06, + "loss": 1.2114, + "step": 851 + }, + { + "epoch": 0.39609483960948394, + "grad_norm": 8.345544815063477, + "learning_rate": 9.9991489999149e-06, + "loss": 0.3967, + "step": 852 + }, + { + "epoch": 0.39655973965597396, + "grad_norm": 6.96198034286499, + "learning_rate": 9.9991479999148e-06, + "loss": 0.7509, + "step": 853 + }, + { + "epoch": 0.397024639702464, + "grad_norm": 4.82371187210083, + "learning_rate": 9.999146999914701e-06, + "loss": 0.4655, + "step": 854 + }, + { + "epoch": 0.39748953974895396, + "grad_norm": 6.932863235473633, + "learning_rate": 9.999145999914602e-06, + "loss": 0.906, + "step": 855 + }, + { + "epoch": 0.397954439795444, + "grad_norm": 8.067989349365234, + "learning_rate": 9.9991449999145e-06, + "loss": 0.8757, + "step": 856 + }, + { + "epoch": 0.398419339841934, + "grad_norm": 12.401895523071289, + "learning_rate": 9.999143999914401e-06, + "loss": 1.8046, + "step": 857 + }, + { + "epoch": 0.398884239888424, + "grad_norm": 8.108248710632324, + "learning_rate": 9.9991429999143e-06, + "loss": 0.7144, + "step": 858 + }, + { + "epoch": 0.399349139934914, + "grad_norm": 8.92480754852295, + "learning_rate": 9.999141999914201e-06, + "loss": 1.0187, + "step": 859 + }, + { + "epoch": 0.399814039981404, + "grad_norm": 11.739278793334961, + "learning_rate": 9.999140999914102e-06, + "loss": 0.696, + "step": 860 + }, + { + "epoch": 0.400278940027894, + "grad_norm": 9.423574447631836, + "learning_rate": 9.999139999914e-06, + "loss": 0.8861, + "step": 861 + }, + { + "epoch": 0.400743840074384, + "grad_norm": 5.569077968597412, + "learning_rate": 9.999138999913901e-06, + "loss": 0.4593, + "step": 862 + }, + { + "epoch": 0.401208740120874, + "grad_norm": 6.269590854644775, + "learning_rate": 9.9991379999138e-06, + "loss": 0.5315, + "step": 863 + }, + { + "epoch": 0.401673640167364, + "grad_norm": 10.623257637023926, + "learning_rate": 9.999136999913701e-06, + "loss": 0.6844, + "step": 864 + }, + { + "epoch": 0.40213854021385403, + "grad_norm": 9.49766731262207, + "learning_rate": 9.9991359999136e-06, + "loss": 1.0143, + "step": 865 + }, + { + "epoch": 0.402603440260344, + "grad_norm": 10.06883716583252, + "learning_rate": 9.9991349999135e-06, + "loss": 1.0506, + "step": 866 + }, + { + "epoch": 0.403068340306834, + "grad_norm": 9.333463668823242, + "learning_rate": 9.999133999913401e-06, + "loss": 0.7765, + "step": 867 + }, + { + "epoch": 0.40353324035332405, + "grad_norm": 5.42322301864624, + "learning_rate": 9.999132999913302e-06, + "loss": 0.5346, + "step": 868 + }, + { + "epoch": 0.403998140399814, + "grad_norm": 9.060174942016602, + "learning_rate": 9.999131999913201e-06, + "loss": 0.8697, + "step": 869 + }, + { + "epoch": 0.40446304044630405, + "grad_norm": 6.127751350402832, + "learning_rate": 9.9991309999131e-06, + "loss": 0.5183, + "step": 870 + }, + { + "epoch": 0.40492794049279407, + "grad_norm": 7.7811126708984375, + "learning_rate": 9.999129999913e-06, + "loss": 0.805, + "step": 871 + }, + { + "epoch": 0.40539284053928404, + "grad_norm": 8.832974433898926, + "learning_rate": 9.999128999912901e-06, + "loss": 0.9695, + "step": 872 + }, + { + "epoch": 0.40585774058577406, + "grad_norm": 6.70252799987793, + "learning_rate": 9.999127999912802e-06, + "loss": 0.9897, + "step": 873 + }, + { + "epoch": 0.4063226406322641, + "grad_norm": 8.64156436920166, + "learning_rate": 9.9991269999127e-06, + "loss": 1.078, + "step": 874 + }, + { + "epoch": 0.40678754067875406, + "grad_norm": 6.713927268981934, + "learning_rate": 9.999125999912601e-06, + "loss": 0.8151, + "step": 875 + }, + { + "epoch": 0.4072524407252441, + "grad_norm": 11.628863334655762, + "learning_rate": 9.9991249999125e-06, + "loss": 1.4475, + "step": 876 + }, + { + "epoch": 0.4077173407717341, + "grad_norm": 10.5166015625, + "learning_rate": 9.999123999912401e-06, + "loss": 0.9501, + "step": 877 + }, + { + "epoch": 0.4081822408182241, + "grad_norm": 11.467411994934082, + "learning_rate": 9.9991229999123e-06, + "loss": 1.4651, + "step": 878 + }, + { + "epoch": 0.4086471408647141, + "grad_norm": 6.269761085510254, + "learning_rate": 9.9991219999122e-06, + "loss": 0.5941, + "step": 879 + }, + { + "epoch": 0.40911204091120407, + "grad_norm": 6.977835178375244, + "learning_rate": 9.999120999912101e-06, + "loss": 0.7304, + "step": 880 + }, + { + "epoch": 0.4095769409576941, + "grad_norm": 4.400878429412842, + "learning_rate": 9.999119999912e-06, + "loss": 0.3837, + "step": 881 + }, + { + "epoch": 0.4100418410041841, + "grad_norm": 6.610584259033203, + "learning_rate": 9.999118999911901e-06, + "loss": 0.7659, + "step": 882 + }, + { + "epoch": 0.4105067410506741, + "grad_norm": 9.919295310974121, + "learning_rate": 9.9991179999118e-06, + "loss": 1.2338, + "step": 883 + }, + { + "epoch": 0.4109716410971641, + "grad_norm": 8.274896621704102, + "learning_rate": 9.9991169999117e-06, + "loss": 0.8293, + "step": 884 + }, + { + "epoch": 0.41143654114365413, + "grad_norm": 8.052297592163086, + "learning_rate": 9.999115999911601e-06, + "loss": 0.7058, + "step": 885 + }, + { + "epoch": 0.4119014411901441, + "grad_norm": 4.349321365356445, + "learning_rate": 9.999114999911502e-06, + "loss": 0.3123, + "step": 886 + }, + { + "epoch": 0.4123663412366341, + "grad_norm": 5.143773555755615, + "learning_rate": 9.999113999911401e-06, + "loss": 0.3934, + "step": 887 + }, + { + "epoch": 0.41283124128312415, + "grad_norm": 15.171216011047363, + "learning_rate": 9.999112999911302e-06, + "loss": 1.1736, + "step": 888 + }, + { + "epoch": 0.4132961413296141, + "grad_norm": 7.153528690338135, + "learning_rate": 9.9991119999112e-06, + "loss": 0.8198, + "step": 889 + }, + { + "epoch": 0.41376104137610414, + "grad_norm": 5.6346588134765625, + "learning_rate": 9.999110999911101e-06, + "loss": 0.3928, + "step": 890 + }, + { + "epoch": 0.41422594142259417, + "grad_norm": 6.53761625289917, + "learning_rate": 9.999109999911e-06, + "loss": 0.5048, + "step": 891 + }, + { + "epoch": 0.41469084146908414, + "grad_norm": 4.6781158447265625, + "learning_rate": 9.9991089999109e-06, + "loss": 0.6285, + "step": 892 + }, + { + "epoch": 0.41515574151557416, + "grad_norm": 9.894698143005371, + "learning_rate": 9.999107999910801e-06, + "loss": 0.5449, + "step": 893 + }, + { + "epoch": 0.41562064156206413, + "grad_norm": 4.112925052642822, + "learning_rate": 9.9991069999107e-06, + "loss": 0.3049, + "step": 894 + }, + { + "epoch": 0.41608554160855415, + "grad_norm": 6.994103908538818, + "learning_rate": 9.999105999910601e-06, + "loss": 0.5403, + "step": 895 + }, + { + "epoch": 0.4165504416550442, + "grad_norm": 6.99237060546875, + "learning_rate": 9.9991049999105e-06, + "loss": 1.2386, + "step": 896 + }, + { + "epoch": 0.41701534170153415, + "grad_norm": 7.004312515258789, + "learning_rate": 9.9991039999104e-06, + "loss": 0.7929, + "step": 897 + }, + { + "epoch": 0.4174802417480242, + "grad_norm": 7.550278186798096, + "learning_rate": 9.999102999910301e-06, + "loss": 0.7552, + "step": 898 + }, + { + "epoch": 0.4179451417945142, + "grad_norm": 6.272104263305664, + "learning_rate": 9.999101999910202e-06, + "loss": 0.6494, + "step": 899 + }, + { + "epoch": 0.41841004184100417, + "grad_norm": 8.515576362609863, + "learning_rate": 9.999100999910101e-06, + "loss": 0.9199, + "step": 900 + }, + { + "epoch": 0.4188749418874942, + "grad_norm": 8.391157150268555, + "learning_rate": 9.999099999910002e-06, + "loss": 1.0303, + "step": 901 + }, + { + "epoch": 0.4193398419339842, + "grad_norm": 13.058817863464355, + "learning_rate": 9.9990989999099e-06, + "loss": 0.6591, + "step": 902 + }, + { + "epoch": 0.4198047419804742, + "grad_norm": 7.379490375518799, + "learning_rate": 9.9990979999098e-06, + "loss": 0.7589, + "step": 903 + }, + { + "epoch": 0.4202696420269642, + "grad_norm": 8.819229125976562, + "learning_rate": 9.9990969999097e-06, + "loss": 0.6622, + "step": 904 + }, + { + "epoch": 0.42073454207345423, + "grad_norm": 5.3865647315979, + "learning_rate": 9.9990959999096e-06, + "loss": 0.468, + "step": 905 + }, + { + "epoch": 0.4211994421199442, + "grad_norm": 8.333093643188477, + "learning_rate": 9.999094999909501e-06, + "loss": 0.8584, + "step": 906 + }, + { + "epoch": 0.4216643421664342, + "grad_norm": 7.87498664855957, + "learning_rate": 9.9990939999094e-06, + "loss": 0.64, + "step": 907 + }, + { + "epoch": 0.4221292422129242, + "grad_norm": 7.442760944366455, + "learning_rate": 9.999092999909301e-06, + "loss": 0.472, + "step": 908 + }, + { + "epoch": 0.4225941422594142, + "grad_norm": 8.212023735046387, + "learning_rate": 9.9990919999092e-06, + "loss": 0.9305, + "step": 909 + }, + { + "epoch": 0.42305904230590424, + "grad_norm": 10.599408149719238, + "learning_rate": 9.9990909999091e-06, + "loss": 1.3524, + "step": 910 + }, + { + "epoch": 0.4235239423523942, + "grad_norm": 10.76672649383545, + "learning_rate": 9.999089999909001e-06, + "loss": 1.474, + "step": 911 + }, + { + "epoch": 0.42398884239888424, + "grad_norm": 9.313421249389648, + "learning_rate": 9.999088999908902e-06, + "loss": 0.7873, + "step": 912 + }, + { + "epoch": 0.42445374244537426, + "grad_norm": 7.256422996520996, + "learning_rate": 9.999087999908801e-06, + "loss": 0.7661, + "step": 913 + }, + { + "epoch": 0.42491864249186423, + "grad_norm": 5.221542835235596, + "learning_rate": 9.9990869999087e-06, + "loss": 0.525, + "step": 914 + }, + { + "epoch": 0.42538354253835425, + "grad_norm": 10.634737014770508, + "learning_rate": 9.9990859999086e-06, + "loss": 1.402, + "step": 915 + }, + { + "epoch": 0.4258484425848443, + "grad_norm": 12.82314682006836, + "learning_rate": 9.999084999908501e-06, + "loss": 1.0587, + "step": 916 + }, + { + "epoch": 0.42631334263133425, + "grad_norm": 6.47551965713501, + "learning_rate": 9.9990839999084e-06, + "loss": 0.6867, + "step": 917 + }, + { + "epoch": 0.42677824267782427, + "grad_norm": 7.47194242477417, + "learning_rate": 9.999082999908301e-06, + "loss": 0.6697, + "step": 918 + }, + { + "epoch": 0.4272431427243143, + "grad_norm": 7.119606018066406, + "learning_rate": 9.999081999908202e-06, + "loss": 0.5997, + "step": 919 + }, + { + "epoch": 0.42770804277080426, + "grad_norm": 5.1243062019348145, + "learning_rate": 9.9990809999081e-06, + "loss": 0.619, + "step": 920 + }, + { + "epoch": 0.4281729428172943, + "grad_norm": 5.870535373687744, + "learning_rate": 9.999079999908001e-06, + "loss": 0.5307, + "step": 921 + }, + { + "epoch": 0.4286378428637843, + "grad_norm": 5.810249328613281, + "learning_rate": 9.9990789999079e-06, + "loss": 0.4722, + "step": 922 + }, + { + "epoch": 0.4291027429102743, + "grad_norm": 6.710216999053955, + "learning_rate": 9.9990779999078e-06, + "loss": 0.7019, + "step": 923 + }, + { + "epoch": 0.4295676429567643, + "grad_norm": 9.604846000671387, + "learning_rate": 9.999076999907701e-06, + "loss": 1.4211, + "step": 924 + }, + { + "epoch": 0.4300325430032543, + "grad_norm": 8.920656204223633, + "learning_rate": 9.9990759999076e-06, + "loss": 0.6873, + "step": 925 + }, + { + "epoch": 0.4304974430497443, + "grad_norm": 6.45170783996582, + "learning_rate": 9.999074999907501e-06, + "loss": 0.6982, + "step": 926 + }, + { + "epoch": 0.4309623430962343, + "grad_norm": 8.143094062805176, + "learning_rate": 9.9990739999074e-06, + "loss": 0.8845, + "step": 927 + }, + { + "epoch": 0.4314272431427243, + "grad_norm": 10.619364738464355, + "learning_rate": 9.9990729999073e-06, + "loss": 0.9608, + "step": 928 + }, + { + "epoch": 0.4318921431892143, + "grad_norm": 10.383740425109863, + "learning_rate": 9.999071999907201e-06, + "loss": 1.1155, + "step": 929 + }, + { + "epoch": 0.43235704323570434, + "grad_norm": 9.993014335632324, + "learning_rate": 9.9990709999071e-06, + "loss": 0.9789, + "step": 930 + }, + { + "epoch": 0.4328219432821943, + "grad_norm": 11.42905044555664, + "learning_rate": 9.999069999907001e-06, + "loss": 1.3062, + "step": 931 + }, + { + "epoch": 0.43328684332868433, + "grad_norm": 8.149091720581055, + "learning_rate": 9.999068999906902e-06, + "loss": 0.9225, + "step": 932 + }, + { + "epoch": 0.43375174337517436, + "grad_norm": 7.182446479797363, + "learning_rate": 9.9990679999068e-06, + "loss": 0.679, + "step": 933 + }, + { + "epoch": 0.43421664342166433, + "grad_norm": 6.386519908905029, + "learning_rate": 9.9990669999067e-06, + "loss": 0.6643, + "step": 934 + }, + { + "epoch": 0.43468154346815435, + "grad_norm": 7.783380508422852, + "learning_rate": 9.9990659999066e-06, + "loss": 0.5367, + "step": 935 + }, + { + "epoch": 0.4351464435146444, + "grad_norm": 6.7544941902160645, + "learning_rate": 9.999064999906501e-06, + "loss": 0.6318, + "step": 936 + }, + { + "epoch": 0.43561134356113435, + "grad_norm": 7.725551128387451, + "learning_rate": 9.999063999906402e-06, + "loss": 0.6823, + "step": 937 + }, + { + "epoch": 0.43607624360762437, + "grad_norm": 7.017218112945557, + "learning_rate": 9.9990629999063e-06, + "loss": 0.5116, + "step": 938 + }, + { + "epoch": 0.43654114365411434, + "grad_norm": 5.589216232299805, + "learning_rate": 9.999061999906201e-06, + "loss": 0.6756, + "step": 939 + }, + { + "epoch": 0.43700604370060436, + "grad_norm": 10.63843822479248, + "learning_rate": 9.9990609999061e-06, + "loss": 1.0423, + "step": 940 + }, + { + "epoch": 0.4374709437470944, + "grad_norm": 16.093523025512695, + "learning_rate": 9.999059999906e-06, + "loss": 1.76, + "step": 941 + }, + { + "epoch": 0.43793584379358436, + "grad_norm": 6.633858680725098, + "learning_rate": 9.999058999905901e-06, + "loss": 0.4238, + "step": 942 + }, + { + "epoch": 0.4384007438400744, + "grad_norm": 8.105208396911621, + "learning_rate": 9.999057999905802e-06, + "loss": 1.0925, + "step": 943 + }, + { + "epoch": 0.4388656438865644, + "grad_norm": 12.805190086364746, + "learning_rate": 9.999056999905701e-06, + "loss": 1.1832, + "step": 944 + }, + { + "epoch": 0.4393305439330544, + "grad_norm": 7.489113807678223, + "learning_rate": 9.9990559999056e-06, + "loss": 0.8976, + "step": 945 + }, + { + "epoch": 0.4397954439795444, + "grad_norm": 9.066057205200195, + "learning_rate": 9.9990549999055e-06, + "loss": 0.4295, + "step": 946 + }, + { + "epoch": 0.4402603440260344, + "grad_norm": 9.869705200195312, + "learning_rate": 9.9990539999054e-06, + "loss": 1.1814, + "step": 947 + }, + { + "epoch": 0.4407252440725244, + "grad_norm": 13.662261962890625, + "learning_rate": 9.9990529999053e-06, + "loss": 1.2591, + "step": 948 + }, + { + "epoch": 0.4411901441190144, + "grad_norm": 5.440181732177734, + "learning_rate": 9.999051999905201e-06, + "loss": 0.575, + "step": 949 + }, + { + "epoch": 0.44165504416550444, + "grad_norm": 7.088125705718994, + "learning_rate": 9.999050999905102e-06, + "loss": 0.5754, + "step": 950 + }, + { + "epoch": 0.4421199442119944, + "grad_norm": 11.611734390258789, + "learning_rate": 9.999049999905e-06, + "loss": 1.1355, + "step": 951 + }, + { + "epoch": 0.44258484425848443, + "grad_norm": 5.906369686126709, + "learning_rate": 9.999048999904901e-06, + "loss": 0.4178, + "step": 952 + }, + { + "epoch": 0.44304974430497446, + "grad_norm": 5.410999298095703, + "learning_rate": 9.9990479999048e-06, + "loss": 0.5339, + "step": 953 + }, + { + "epoch": 0.4435146443514644, + "grad_norm": 8.200979232788086, + "learning_rate": 9.999046999904701e-06, + "loss": 1.332, + "step": 954 + }, + { + "epoch": 0.44397954439795445, + "grad_norm": 8.507502555847168, + "learning_rate": 9.999045999904602e-06, + "loss": 0.8913, + "step": 955 + }, + { + "epoch": 0.4444444444444444, + "grad_norm": 7.402044773101807, + "learning_rate": 9.9990449999045e-06, + "loss": 0.7313, + "step": 956 + }, + { + "epoch": 0.44490934449093444, + "grad_norm": 7.3146796226501465, + "learning_rate": 9.999043999904401e-06, + "loss": 0.6142, + "step": 957 + }, + { + "epoch": 0.44537424453742447, + "grad_norm": 9.160687446594238, + "learning_rate": 9.9990429999043e-06, + "loss": 0.8667, + "step": 958 + }, + { + "epoch": 0.44583914458391444, + "grad_norm": 6.630360126495361, + "learning_rate": 9.9990419999042e-06, + "loss": 0.5622, + "step": 959 + }, + { + "epoch": 0.44630404463040446, + "grad_norm": 8.89023494720459, + "learning_rate": 9.9990409999041e-06, + "loss": 0.9983, + "step": 960 + }, + { + "epoch": 0.4467689446768945, + "grad_norm": 7.165299415588379, + "learning_rate": 9.999039999904e-06, + "loss": 1.0156, + "step": 961 + }, + { + "epoch": 0.44723384472338445, + "grad_norm": 9.901755332946777, + "learning_rate": 9.999038999903901e-06, + "loss": 1.3582, + "step": 962 + }, + { + "epoch": 0.4476987447698745, + "grad_norm": 6.208670616149902, + "learning_rate": 9.999037999903802e-06, + "loss": 0.6693, + "step": 963 + }, + { + "epoch": 0.4481636448163645, + "grad_norm": 5.910864353179932, + "learning_rate": 9.9990369999037e-06, + "loss": 0.6031, + "step": 964 + }, + { + "epoch": 0.4486285448628545, + "grad_norm": 10.877969741821289, + "learning_rate": 9.999035999903601e-06, + "loss": 1.5033, + "step": 965 + }, + { + "epoch": 0.4490934449093445, + "grad_norm": 6.021311283111572, + "learning_rate": 9.9990349999035e-06, + "loss": 0.6908, + "step": 966 + }, + { + "epoch": 0.4495583449558345, + "grad_norm": 7.971142768859863, + "learning_rate": 9.999033999903401e-06, + "loss": 0.9471, + "step": 967 + }, + { + "epoch": 0.4500232450023245, + "grad_norm": 5.242122173309326, + "learning_rate": 9.999032999903302e-06, + "loss": 0.4884, + "step": 968 + }, + { + "epoch": 0.4504881450488145, + "grad_norm": 7.733642578125, + "learning_rate": 9.9990319999032e-06, + "loss": 0.7738, + "step": 969 + }, + { + "epoch": 0.4509530450953045, + "grad_norm": 6.091546058654785, + "learning_rate": 9.999030999903101e-06, + "loss": 0.5953, + "step": 970 + }, + { + "epoch": 0.4514179451417945, + "grad_norm": 6.905654430389404, + "learning_rate": 9.999029999903e-06, + "loss": 0.4975, + "step": 971 + }, + { + "epoch": 0.45188284518828453, + "grad_norm": 5.6894354820251465, + "learning_rate": 9.999028999902901e-06, + "loss": 0.679, + "step": 972 + }, + { + "epoch": 0.4523477452347745, + "grad_norm": 5.8789544105529785, + "learning_rate": 9.9990279999028e-06, + "loss": 0.5562, + "step": 973 + }, + { + "epoch": 0.4528126452812645, + "grad_norm": 6.301591873168945, + "learning_rate": 9.9990269999027e-06, + "loss": 0.535, + "step": 974 + }, + { + "epoch": 0.45327754532775455, + "grad_norm": 11.641485214233398, + "learning_rate": 9.999025999902601e-06, + "loss": 1.1352, + "step": 975 + }, + { + "epoch": 0.4537424453742445, + "grad_norm": 7.17354679107666, + "learning_rate": 9.999024999902502e-06, + "loss": 0.5443, + "step": 976 + }, + { + "epoch": 0.45420734542073454, + "grad_norm": 4.6022629737854, + "learning_rate": 9.9990239999024e-06, + "loss": 0.4232, + "step": 977 + }, + { + "epoch": 0.45467224546722457, + "grad_norm": 8.57508373260498, + "learning_rate": 9.9990229999023e-06, + "loss": 0.862, + "step": 978 + }, + { + "epoch": 0.45513714551371454, + "grad_norm": 8.534709930419922, + "learning_rate": 9.9990219999022e-06, + "loss": 0.6639, + "step": 979 + }, + { + "epoch": 0.45560204556020456, + "grad_norm": 7.1432719230651855, + "learning_rate": 9.999020999902101e-06, + "loss": 0.6453, + "step": 980 + }, + { + "epoch": 0.4560669456066946, + "grad_norm": 4.8909525871276855, + "learning_rate": 9.999019999902002e-06, + "loss": 0.5173, + "step": 981 + }, + { + "epoch": 0.45653184565318455, + "grad_norm": 8.899125099182129, + "learning_rate": 9.9990189999019e-06, + "loss": 1.1315, + "step": 982 + }, + { + "epoch": 0.4569967456996746, + "grad_norm": 9.179483413696289, + "learning_rate": 9.999017999901801e-06, + "loss": 1.9182, + "step": 983 + }, + { + "epoch": 0.45746164574616455, + "grad_norm": 6.445417881011963, + "learning_rate": 9.9990169999017e-06, + "loss": 0.4487, + "step": 984 + }, + { + "epoch": 0.45792654579265457, + "grad_norm": 7.437139511108398, + "learning_rate": 9.999015999901601e-06, + "loss": 0.6599, + "step": 985 + }, + { + "epoch": 0.4583914458391446, + "grad_norm": 6.5848517417907715, + "learning_rate": 9.9990149999015e-06, + "loss": 0.5309, + "step": 986 + }, + { + "epoch": 0.45885634588563456, + "grad_norm": 7.267918109893799, + "learning_rate": 9.9990139999014e-06, + "loss": 0.7765, + "step": 987 + }, + { + "epoch": 0.4593212459321246, + "grad_norm": 8.215215682983398, + "learning_rate": 9.999012999901301e-06, + "loss": 0.8364, + "step": 988 + }, + { + "epoch": 0.4597861459786146, + "grad_norm": 8.794707298278809, + "learning_rate": 9.9990119999012e-06, + "loss": 1.3581, + "step": 989 + }, + { + "epoch": 0.4602510460251046, + "grad_norm": 12.117571830749512, + "learning_rate": 9.9990109999011e-06, + "loss": 1.1001, + "step": 990 + }, + { + "epoch": 0.4607159460715946, + "grad_norm": 9.459122657775879, + "learning_rate": 9.999009999901e-06, + "loss": 0.6675, + "step": 991 + }, + { + "epoch": 0.46118084611808463, + "grad_norm": 7.28818941116333, + "learning_rate": 9.9990089999009e-06, + "loss": 0.7394, + "step": 992 + }, + { + "epoch": 0.4616457461645746, + "grad_norm": 9.7479887008667, + "learning_rate": 9.999007999900801e-06, + "loss": 0.7369, + "step": 993 + }, + { + "epoch": 0.4621106462110646, + "grad_norm": 7.793990612030029, + "learning_rate": 9.999006999900702e-06, + "loss": 0.5611, + "step": 994 + }, + { + "epoch": 0.46257554625755465, + "grad_norm": 9.005208015441895, + "learning_rate": 9.9990059999006e-06, + "loss": 0.6072, + "step": 995 + }, + { + "epoch": 0.4630404463040446, + "grad_norm": 7.6074538230896, + "learning_rate": 9.999004999900501e-06, + "loss": 0.8434, + "step": 996 + }, + { + "epoch": 0.46350534635053464, + "grad_norm": 9.262664794921875, + "learning_rate": 9.9990039999004e-06, + "loss": 0.8949, + "step": 997 + }, + { + "epoch": 0.46397024639702467, + "grad_norm": 8.38783073425293, + "learning_rate": 9.999002999900301e-06, + "loss": 0.8553, + "step": 998 + }, + { + "epoch": 0.46443514644351463, + "grad_norm": 8.565309524536133, + "learning_rate": 9.9990019999002e-06, + "loss": 0.7459, + "step": 999 + }, + { + "epoch": 0.46490004649000466, + "grad_norm": 24.209789276123047, + "learning_rate": 9.9990009999001e-06, + "loss": 1.0281, + "step": 1000 + }, + { + "epoch": 0.46536494653649463, + "grad_norm": 7.019515514373779, + "learning_rate": 9.998999999900001e-06, + "loss": 0.5502, + "step": 1001 + }, + { + "epoch": 0.46582984658298465, + "grad_norm": 6.642327308654785, + "learning_rate": 9.9989989998999e-06, + "loss": 0.5628, + "step": 1002 + }, + { + "epoch": 0.4662947466294747, + "grad_norm": 12.422435760498047, + "learning_rate": 9.998997999899801e-06, + "loss": 1.4055, + "step": 1003 + }, + { + "epoch": 0.46675964667596465, + "grad_norm": 6.617071628570557, + "learning_rate": 9.9989969998997e-06, + "loss": 0.5407, + "step": 1004 + }, + { + "epoch": 0.46722454672245467, + "grad_norm": 6.208919525146484, + "learning_rate": 9.9989959998996e-06, + "loss": 0.5101, + "step": 1005 + }, + { + "epoch": 0.4676894467689447, + "grad_norm": 8.929449081420898, + "learning_rate": 9.998994999899501e-06, + "loss": 0.9718, + "step": 1006 + }, + { + "epoch": 0.46815434681543466, + "grad_norm": 7.079061985015869, + "learning_rate": 9.998993999899402e-06, + "loss": 0.7642, + "step": 1007 + }, + { + "epoch": 0.4686192468619247, + "grad_norm": 7.230600357055664, + "learning_rate": 9.9989929998993e-06, + "loss": 0.8734, + "step": 1008 + }, + { + "epoch": 0.4690841469084147, + "grad_norm": 5.949636459350586, + "learning_rate": 9.998991999899201e-06, + "loss": 0.5031, + "step": 1009 + }, + { + "epoch": 0.4695490469549047, + "grad_norm": 7.293687343597412, + "learning_rate": 9.9989909998991e-06, + "loss": 0.9145, + "step": 1010 + }, + { + "epoch": 0.4700139470013947, + "grad_norm": 7.455058574676514, + "learning_rate": 9.998989999899001e-06, + "loss": 0.7559, + "step": 1011 + }, + { + "epoch": 0.47047884704788473, + "grad_norm": 5.817276954650879, + "learning_rate": 9.998988999898902e-06, + "loss": 0.6296, + "step": 1012 + }, + { + "epoch": 0.4709437470943747, + "grad_norm": 9.70932674407959, + "learning_rate": 9.9989879998988e-06, + "loss": 0.8464, + "step": 1013 + }, + { + "epoch": 0.4714086471408647, + "grad_norm": 8.223123550415039, + "learning_rate": 9.998986999898701e-06, + "loss": 1.3747, + "step": 1014 + }, + { + "epoch": 0.4718735471873547, + "grad_norm": 5.871592998504639, + "learning_rate": 9.9989859998986e-06, + "loss": 0.5415, + "step": 1015 + }, + { + "epoch": 0.4723384472338447, + "grad_norm": 9.297321319580078, + "learning_rate": 9.998984999898501e-06, + "loss": 0.9638, + "step": 1016 + }, + { + "epoch": 0.47280334728033474, + "grad_norm": 6.350240230560303, + "learning_rate": 9.9989839998984e-06, + "loss": 0.6216, + "step": 1017 + }, + { + "epoch": 0.4732682473268247, + "grad_norm": 7.308672904968262, + "learning_rate": 9.9989829998983e-06, + "loss": 0.8024, + "step": 1018 + }, + { + "epoch": 0.47373314737331473, + "grad_norm": 5.593734264373779, + "learning_rate": 9.998981999898201e-06, + "loss": 0.6941, + "step": 1019 + }, + { + "epoch": 0.47419804741980476, + "grad_norm": 7.232341766357422, + "learning_rate": 9.998980999898102e-06, + "loss": 0.7007, + "step": 1020 + }, + { + "epoch": 0.4746629474662947, + "grad_norm": 5.669683933258057, + "learning_rate": 9.998979999898001e-06, + "loss": 0.4855, + "step": 1021 + }, + { + "epoch": 0.47512784751278475, + "grad_norm": 7.367640972137451, + "learning_rate": 9.9989789998979e-06, + "loss": 0.9144, + "step": 1022 + }, + { + "epoch": 0.4755927475592748, + "grad_norm": 11.108624458312988, + "learning_rate": 9.9989779998978e-06, + "loss": 1.2512, + "step": 1023 + }, + { + "epoch": 0.47605764760576474, + "grad_norm": 7.994923114776611, + "learning_rate": 9.998976999897701e-06, + "loss": 0.6083, + "step": 1024 + }, + { + "epoch": 0.47652254765225477, + "grad_norm": 7.731326103210449, + "learning_rate": 9.998975999897602e-06, + "loss": 0.8522, + "step": 1025 + }, + { + "epoch": 0.4769874476987448, + "grad_norm": 7.688309192657471, + "learning_rate": 9.9989749998975e-06, + "loss": 0.4182, + "step": 1026 + }, + { + "epoch": 0.47745234774523476, + "grad_norm": 7.767177104949951, + "learning_rate": 9.998973999897401e-06, + "loss": 0.9537, + "step": 1027 + }, + { + "epoch": 0.4779172477917248, + "grad_norm": 6.574499130249023, + "learning_rate": 9.9989729998973e-06, + "loss": 0.7218, + "step": 1028 + }, + { + "epoch": 0.4783821478382148, + "grad_norm": 10.167267799377441, + "learning_rate": 9.998971999897201e-06, + "loss": 0.6755, + "step": 1029 + }, + { + "epoch": 0.4788470478847048, + "grad_norm": 6.795566558837891, + "learning_rate": 9.9989709998971e-06, + "loss": 0.6867, + "step": 1030 + }, + { + "epoch": 0.4793119479311948, + "grad_norm": 9.833226203918457, + "learning_rate": 9.998969999897e-06, + "loss": 0.7092, + "step": 1031 + }, + { + "epoch": 0.4797768479776848, + "grad_norm": 11.001030921936035, + "learning_rate": 9.998968999896901e-06, + "loss": 1.1113, + "step": 1032 + }, + { + "epoch": 0.4802417480241748, + "grad_norm": 5.950009822845459, + "learning_rate": 9.9989679998968e-06, + "loss": 0.3619, + "step": 1033 + }, + { + "epoch": 0.4807066480706648, + "grad_norm": 6.7932353019714355, + "learning_rate": 9.998966999896701e-06, + "loss": 0.5962, + "step": 1034 + }, + { + "epoch": 0.4811715481171548, + "grad_norm": 11.298108100891113, + "learning_rate": 9.9989659998966e-06, + "loss": 0.9078, + "step": 1035 + }, + { + "epoch": 0.4816364481636448, + "grad_norm": 6.049661159515381, + "learning_rate": 9.9989649998965e-06, + "loss": 0.5832, + "step": 1036 + }, + { + "epoch": 0.48210134821013484, + "grad_norm": 6.894578456878662, + "learning_rate": 9.998963999896401e-06, + "loss": 0.8146, + "step": 1037 + }, + { + "epoch": 0.4825662482566248, + "grad_norm": 13.074256896972656, + "learning_rate": 9.998962999896302e-06, + "loss": 0.9323, + "step": 1038 + }, + { + "epoch": 0.48303114830311483, + "grad_norm": 6.259931564331055, + "learning_rate": 9.998961999896201e-06, + "loss": 0.5772, + "step": 1039 + }, + { + "epoch": 0.48349604834960486, + "grad_norm": 7.0895867347717285, + "learning_rate": 9.998960999896102e-06, + "loss": 0.4563, + "step": 1040 + }, + { + "epoch": 0.4839609483960948, + "grad_norm": 7.831277847290039, + "learning_rate": 9.998959999896e-06, + "loss": 0.9301, + "step": 1041 + }, + { + "epoch": 0.48442584844258485, + "grad_norm": 5.495150089263916, + "learning_rate": 9.998958999895901e-06, + "loss": 0.5579, + "step": 1042 + }, + { + "epoch": 0.4848907484890749, + "grad_norm": 7.630045413970947, + "learning_rate": 9.9989579998958e-06, + "loss": 0.9217, + "step": 1043 + }, + { + "epoch": 0.48535564853556484, + "grad_norm": 5.966329574584961, + "learning_rate": 9.9989569998957e-06, + "loss": 0.5499, + "step": 1044 + }, + { + "epoch": 0.48582054858205487, + "grad_norm": 6.676249980926514, + "learning_rate": 9.998955999895601e-06, + "loss": 0.6902, + "step": 1045 + }, + { + "epoch": 0.48628544862854484, + "grad_norm": 7.37178373336792, + "learning_rate": 9.9989549998955e-06, + "loss": 0.7364, + "step": 1046 + }, + { + "epoch": 0.48675034867503486, + "grad_norm": 4.797945499420166, + "learning_rate": 9.998953999895401e-06, + "loss": 0.5416, + "step": 1047 + }, + { + "epoch": 0.4872152487215249, + "grad_norm": 6.450428485870361, + "learning_rate": 9.9989529998953e-06, + "loss": 0.7921, + "step": 1048 + }, + { + "epoch": 0.48768014876801485, + "grad_norm": 8.062814712524414, + "learning_rate": 9.9989519998952e-06, + "loss": 0.3543, + "step": 1049 + }, + { + "epoch": 0.4881450488145049, + "grad_norm": 5.4737067222595215, + "learning_rate": 9.998950999895101e-06, + "loss": 0.6311, + "step": 1050 + }, + { + "epoch": 0.4886099488609949, + "grad_norm": 5.52265739440918, + "learning_rate": 9.998949999895002e-06, + "loss": 0.7124, + "step": 1051 + }, + { + "epoch": 0.48907484890748487, + "grad_norm": 6.946440696716309, + "learning_rate": 9.998948999894901e-06, + "loss": 0.9008, + "step": 1052 + }, + { + "epoch": 0.4895397489539749, + "grad_norm": 5.441209316253662, + "learning_rate": 9.998947999894802e-06, + "loss": 0.4475, + "step": 1053 + }, + { + "epoch": 0.4900046490004649, + "grad_norm": 7.4508442878723145, + "learning_rate": 9.9989469998947e-06, + "loss": 0.5732, + "step": 1054 + }, + { + "epoch": 0.4904695490469549, + "grad_norm": 5.159357070922852, + "learning_rate": 9.9989459998946e-06, + "loss": 0.5226, + "step": 1055 + }, + { + "epoch": 0.4909344490934449, + "grad_norm": 7.480625629425049, + "learning_rate": 9.9989449998945e-06, + "loss": 0.8072, + "step": 1056 + }, + { + "epoch": 0.49139934913993494, + "grad_norm": 9.4108304977417, + "learning_rate": 9.998943999894401e-06, + "loss": 0.9371, + "step": 1057 + }, + { + "epoch": 0.4918642491864249, + "grad_norm": 9.409568786621094, + "learning_rate": 9.998942999894302e-06, + "loss": 1.1948, + "step": 1058 + }, + { + "epoch": 0.49232914923291493, + "grad_norm": 8.296445846557617, + "learning_rate": 9.9989419998942e-06, + "loss": 0.7333, + "step": 1059 + }, + { + "epoch": 0.49279404927940496, + "grad_norm": 8.532403945922852, + "learning_rate": 9.998940999894101e-06, + "loss": 0.7629, + "step": 1060 + }, + { + "epoch": 0.4932589493258949, + "grad_norm": 6.353894233703613, + "learning_rate": 9.998939999894e-06, + "loss": 0.8201, + "step": 1061 + }, + { + "epoch": 0.49372384937238495, + "grad_norm": 11.502490043640137, + "learning_rate": 9.9989389998939e-06, + "loss": 0.8103, + "step": 1062 + }, + { + "epoch": 0.4941887494188749, + "grad_norm": 7.684313774108887, + "learning_rate": 9.998937999893801e-06, + "loss": 1.029, + "step": 1063 + }, + { + "epoch": 0.49465364946536494, + "grad_norm": 5.602656364440918, + "learning_rate": 9.9989369998937e-06, + "loss": 0.5061, + "step": 1064 + }, + { + "epoch": 0.49511854951185497, + "grad_norm": 5.207149505615234, + "learning_rate": 9.998935999893601e-06, + "loss": 0.5242, + "step": 1065 + }, + { + "epoch": 0.49558344955834494, + "grad_norm": 5.575476169586182, + "learning_rate": 9.9989349998935e-06, + "loss": 0.4229, + "step": 1066 + }, + { + "epoch": 0.49604834960483496, + "grad_norm": 8.066596031188965, + "learning_rate": 9.9989339998934e-06, + "loss": 0.7979, + "step": 1067 + }, + { + "epoch": 0.496513249651325, + "grad_norm": 7.651096343994141, + "learning_rate": 9.9989329998933e-06, + "loss": 0.6793, + "step": 1068 + }, + { + "epoch": 0.49697814969781495, + "grad_norm": 6.922614097595215, + "learning_rate": 9.9989319998932e-06, + "loss": 0.665, + "step": 1069 + }, + { + "epoch": 0.497443049744305, + "grad_norm": 7.139037132263184, + "learning_rate": 9.998930999893101e-06, + "loss": 0.794, + "step": 1070 + }, + { + "epoch": 0.497907949790795, + "grad_norm": 6.871804237365723, + "learning_rate": 9.998929999893002e-06, + "loss": 0.4858, + "step": 1071 + }, + { + "epoch": 0.49837284983728497, + "grad_norm": 15.687426567077637, + "learning_rate": 9.9989289998929e-06, + "loss": 1.8083, + "step": 1072 + }, + { + "epoch": 0.498837749883775, + "grad_norm": 9.675277709960938, + "learning_rate": 9.998927999892801e-06, + "loss": 1.0744, + "step": 1073 + }, + { + "epoch": 0.499302649930265, + "grad_norm": 7.048737525939941, + "learning_rate": 9.9989269998927e-06, + "loss": 0.4764, + "step": 1074 + }, + { + "epoch": 0.499767549976755, + "grad_norm": 6.963862419128418, + "learning_rate": 9.998925999892601e-06, + "loss": 0.713, + "step": 1075 + }, + { + "epoch": 0.500232450023245, + "grad_norm": 10.212060928344727, + "learning_rate": 9.998924999892502e-06, + "loss": 0.6995, + "step": 1076 + }, + { + "epoch": 0.500697350069735, + "grad_norm": 9.998250007629395, + "learning_rate": 9.9989239998924e-06, + "loss": 1.5655, + "step": 1077 + }, + { + "epoch": 0.5011622501162251, + "grad_norm": 9.531150817871094, + "learning_rate": 9.998922999892301e-06, + "loss": 0.639, + "step": 1078 + }, + { + "epoch": 0.501627150162715, + "grad_norm": 11.354978561401367, + "learning_rate": 9.9989219998922e-06, + "loss": 0.6956, + "step": 1079 + }, + { + "epoch": 0.502092050209205, + "grad_norm": 7.213747024536133, + "learning_rate": 9.9989209998921e-06, + "loss": 0.4877, + "step": 1080 + }, + { + "epoch": 0.502556950255695, + "grad_norm": 8.41362190246582, + "learning_rate": 9.998919999892001e-06, + "loss": 0.8087, + "step": 1081 + }, + { + "epoch": 0.503021850302185, + "grad_norm": 7.607968330383301, + "learning_rate": 9.9989189998919e-06, + "loss": 0.442, + "step": 1082 + }, + { + "epoch": 0.5034867503486751, + "grad_norm": 5.780048847198486, + "learning_rate": 9.998917999891801e-06, + "loss": 0.7222, + "step": 1083 + }, + { + "epoch": 0.503951650395165, + "grad_norm": 8.365687370300293, + "learning_rate": 9.998916999891702e-06, + "loss": 1.0587, + "step": 1084 + }, + { + "epoch": 0.504416550441655, + "grad_norm": 5.802889823913574, + "learning_rate": 9.9989159998916e-06, + "loss": 0.5816, + "step": 1085 + }, + { + "epoch": 0.504881450488145, + "grad_norm": 6.988004207611084, + "learning_rate": 9.9989149998915e-06, + "loss": 0.8973, + "step": 1086 + }, + { + "epoch": 0.5053463505346351, + "grad_norm": 7.25809907913208, + "learning_rate": 9.9989139998914e-06, + "loss": 0.6595, + "step": 1087 + }, + { + "epoch": 0.5058112505811251, + "grad_norm": 8.18140983581543, + "learning_rate": 9.998912999891301e-06, + "loss": 1.2756, + "step": 1088 + }, + { + "epoch": 0.5062761506276151, + "grad_norm": 8.08736515045166, + "learning_rate": 9.998911999891202e-06, + "loss": 0.7555, + "step": 1089 + }, + { + "epoch": 0.506741050674105, + "grad_norm": 9.218304634094238, + "learning_rate": 9.9989109998911e-06, + "loss": 0.8757, + "step": 1090 + }, + { + "epoch": 0.507205950720595, + "grad_norm": 4.836251258850098, + "learning_rate": 9.998909999891001e-06, + "loss": 0.4226, + "step": 1091 + }, + { + "epoch": 0.5076708507670851, + "grad_norm": 6.847232818603516, + "learning_rate": 9.9989089998909e-06, + "loss": 0.4373, + "step": 1092 + }, + { + "epoch": 0.5081357508135751, + "grad_norm": 7.614401340484619, + "learning_rate": 9.9989079998908e-06, + "loss": 0.6371, + "step": 1093 + }, + { + "epoch": 0.5086006508600651, + "grad_norm": 7.118531703948975, + "learning_rate": 9.998906999890701e-06, + "loss": 0.588, + "step": 1094 + }, + { + "epoch": 0.5090655509065551, + "grad_norm": 5.9433417320251465, + "learning_rate": 9.9989059998906e-06, + "loss": 0.4109, + "step": 1095 + }, + { + "epoch": 0.509530450953045, + "grad_norm": 6.926442623138428, + "learning_rate": 9.998904999890501e-06, + "loss": 0.9267, + "step": 1096 + }, + { + "epoch": 0.5099953509995351, + "grad_norm": 10.534729957580566, + "learning_rate": 9.9989039998904e-06, + "loss": 0.9463, + "step": 1097 + }, + { + "epoch": 0.5104602510460251, + "grad_norm": 6.4423675537109375, + "learning_rate": 9.9989029998903e-06, + "loss": 0.5125, + "step": 1098 + }, + { + "epoch": 0.5109251510925151, + "grad_norm": 10.573360443115234, + "learning_rate": 9.9989019998902e-06, + "loss": 0.8593, + "step": 1099 + }, + { + "epoch": 0.5113900511390052, + "grad_norm": 6.084214210510254, + "learning_rate": 9.9989009998901e-06, + "loss": 0.9552, + "step": 1100 + }, + { + "epoch": 0.5118549511854951, + "grad_norm": 6.368248462677002, + "learning_rate": 9.998899999890001e-06, + "loss": 0.421, + "step": 1101 + }, + { + "epoch": 0.5123198512319851, + "grad_norm": 8.092704772949219, + "learning_rate": 9.998898999889902e-06, + "loss": 0.7279, + "step": 1102 + }, + { + "epoch": 0.5127847512784751, + "grad_norm": 6.600542068481445, + "learning_rate": 9.9988979998898e-06, + "loss": 0.6141, + "step": 1103 + }, + { + "epoch": 0.5132496513249651, + "grad_norm": 9.915974617004395, + "learning_rate": 9.998896999889701e-06, + "loss": 0.8306, + "step": 1104 + }, + { + "epoch": 0.5137145513714552, + "grad_norm": 10.132452011108398, + "learning_rate": 9.9988959998896e-06, + "loss": 1.0878, + "step": 1105 + }, + { + "epoch": 0.5141794514179452, + "grad_norm": 6.194248199462891, + "learning_rate": 9.998894999889501e-06, + "loss": 0.6184, + "step": 1106 + }, + { + "epoch": 0.5146443514644351, + "grad_norm": 9.72970199584961, + "learning_rate": 9.998893999889402e-06, + "loss": 0.7888, + "step": 1107 + }, + { + "epoch": 0.5151092515109251, + "grad_norm": 7.672325134277344, + "learning_rate": 9.9988929998893e-06, + "loss": 0.7245, + "step": 1108 + }, + { + "epoch": 0.5155741515574152, + "grad_norm": 5.238547325134277, + "learning_rate": 9.998891999889201e-06, + "loss": 0.5183, + "step": 1109 + }, + { + "epoch": 0.5160390516039052, + "grad_norm": 6.327530860900879, + "learning_rate": 9.9988909998891e-06, + "loss": 0.5136, + "step": 1110 + }, + { + "epoch": 0.5165039516503952, + "grad_norm": 5.548627853393555, + "learning_rate": 9.998889999889e-06, + "loss": 0.5466, + "step": 1111 + }, + { + "epoch": 0.5169688516968852, + "grad_norm": 5.467564582824707, + "learning_rate": 9.9988889998889e-06, + "loss": 0.5379, + "step": 1112 + }, + { + "epoch": 0.5174337517433751, + "grad_norm": 6.649033546447754, + "learning_rate": 9.9988879998888e-06, + "loss": 0.8509, + "step": 1113 + }, + { + "epoch": 0.5178986517898652, + "grad_norm": 17.829862594604492, + "learning_rate": 9.998886999888701e-06, + "loss": 1.6377, + "step": 1114 + }, + { + "epoch": 0.5183635518363552, + "grad_norm": 6.653289794921875, + "learning_rate": 9.998885999888602e-06, + "loss": 0.531, + "step": 1115 + }, + { + "epoch": 0.5188284518828452, + "grad_norm": 5.650300025939941, + "learning_rate": 9.9988849998885e-06, + "loss": 0.5052, + "step": 1116 + }, + { + "epoch": 0.5192933519293352, + "grad_norm": 12.294291496276855, + "learning_rate": 9.998883999888401e-06, + "loss": 1.2082, + "step": 1117 + }, + { + "epoch": 0.5197582519758251, + "grad_norm": 7.83322811126709, + "learning_rate": 9.9988829998883e-06, + "loss": 0.8653, + "step": 1118 + }, + { + "epoch": 0.5202231520223152, + "grad_norm": 8.767287254333496, + "learning_rate": 9.998881999888201e-06, + "loss": 1.3369, + "step": 1119 + }, + { + "epoch": 0.5206880520688052, + "grad_norm": 7.5623698234558105, + "learning_rate": 9.998880999888102e-06, + "loss": 0.8274, + "step": 1120 + }, + { + "epoch": 0.5211529521152952, + "grad_norm": 8.252159118652344, + "learning_rate": 9.998879999888e-06, + "loss": 0.9545, + "step": 1121 + }, + { + "epoch": 0.5216178521617852, + "grad_norm": 6.657195091247559, + "learning_rate": 9.998878999887901e-06, + "loss": 0.8339, + "step": 1122 + }, + { + "epoch": 0.5220827522082753, + "grad_norm": 7.515926361083984, + "learning_rate": 9.9988779998878e-06, + "loss": 0.6762, + "step": 1123 + }, + { + "epoch": 0.5225476522547652, + "grad_norm": 7.711641788482666, + "learning_rate": 9.998876999887701e-06, + "loss": 0.6152, + "step": 1124 + }, + { + "epoch": 0.5230125523012552, + "grad_norm": 9.214947700500488, + "learning_rate": 9.9988759998876e-06, + "loss": 0.8778, + "step": 1125 + }, + { + "epoch": 0.5234774523477452, + "grad_norm": 8.756752967834473, + "learning_rate": 9.9988749998875e-06, + "loss": 1.1081, + "step": 1126 + }, + { + "epoch": 0.5239423523942353, + "grad_norm": 9.123619079589844, + "learning_rate": 9.998873999887401e-06, + "loss": 1.3348, + "step": 1127 + }, + { + "epoch": 0.5244072524407253, + "grad_norm": 7.568092346191406, + "learning_rate": 9.998872999887302e-06, + "loss": 0.8521, + "step": 1128 + }, + { + "epoch": 0.5248721524872152, + "grad_norm": 9.72684383392334, + "learning_rate": 9.9988719998872e-06, + "loss": 0.8623, + "step": 1129 + }, + { + "epoch": 0.5253370525337052, + "grad_norm": 7.299685001373291, + "learning_rate": 9.9988709998871e-06, + "loss": 0.898, + "step": 1130 + }, + { + "epoch": 0.5258019525801952, + "grad_norm": 6.098969459533691, + "learning_rate": 9.998869999887e-06, + "loss": 0.4433, + "step": 1131 + }, + { + "epoch": 0.5262668526266853, + "grad_norm": 5.6247944831848145, + "learning_rate": 9.998868999886901e-06, + "loss": 0.5988, + "step": 1132 + }, + { + "epoch": 0.5267317526731753, + "grad_norm": 7.2255377769470215, + "learning_rate": 9.998867999886802e-06, + "loss": 0.6564, + "step": 1133 + }, + { + "epoch": 0.5271966527196653, + "grad_norm": 4.771019458770752, + "learning_rate": 9.9988669998867e-06, + "loss": 0.4594, + "step": 1134 + }, + { + "epoch": 0.5276615527661552, + "grad_norm": 7.554280757904053, + "learning_rate": 9.998865999886601e-06, + "loss": 0.8949, + "step": 1135 + }, + { + "epoch": 0.5281264528126453, + "grad_norm": 7.238678455352783, + "learning_rate": 9.9988649998865e-06, + "loss": 0.6597, + "step": 1136 + }, + { + "epoch": 0.5285913528591353, + "grad_norm": 7.304091453552246, + "learning_rate": 9.998863999886401e-06, + "loss": 0.9494, + "step": 1137 + }, + { + "epoch": 0.5290562529056253, + "grad_norm": 6.153872489929199, + "learning_rate": 9.9988629998863e-06, + "loss": 0.4631, + "step": 1138 + }, + { + "epoch": 0.5295211529521153, + "grad_norm": 10.157283782958984, + "learning_rate": 9.9988619998862e-06, + "loss": 1.2481, + "step": 1139 + }, + { + "epoch": 0.5299860529986054, + "grad_norm": 6.109690189361572, + "learning_rate": 9.998860999886101e-06, + "loss": 0.6111, + "step": 1140 + }, + { + "epoch": 0.5304509530450953, + "grad_norm": 5.83418607711792, + "learning_rate": 9.998859999886e-06, + "loss": 0.5874, + "step": 1141 + }, + { + "epoch": 0.5309158530915853, + "grad_norm": 7.006343364715576, + "learning_rate": 9.998858999885901e-06, + "loss": 0.5079, + "step": 1142 + }, + { + "epoch": 0.5313807531380753, + "grad_norm": 7.400877952575684, + "learning_rate": 9.9988579998858e-06, + "loss": 0.6314, + "step": 1143 + }, + { + "epoch": 0.5318456531845653, + "grad_norm": 8.856679916381836, + "learning_rate": 9.9988569998857e-06, + "loss": 0.7828, + "step": 1144 + }, + { + "epoch": 0.5323105532310554, + "grad_norm": 4.028034687042236, + "learning_rate": 9.998855999885601e-06, + "loss": 0.3323, + "step": 1145 + }, + { + "epoch": 0.5327754532775453, + "grad_norm": 8.19267749786377, + "learning_rate": 9.998854999885502e-06, + "loss": 0.8177, + "step": 1146 + }, + { + "epoch": 0.5332403533240353, + "grad_norm": 5.59384298324585, + "learning_rate": 9.9988539998854e-06, + "loss": 0.7285, + "step": 1147 + }, + { + "epoch": 0.5337052533705253, + "grad_norm": 6.911167144775391, + "learning_rate": 9.998852999885301e-06, + "loss": 0.7368, + "step": 1148 + }, + { + "epoch": 0.5341701534170153, + "grad_norm": 7.524102687835693, + "learning_rate": 9.9988519998852e-06, + "loss": 0.7057, + "step": 1149 + }, + { + "epoch": 0.5346350534635054, + "grad_norm": 10.3854341506958, + "learning_rate": 9.998850999885101e-06, + "loss": 1.1222, + "step": 1150 + }, + { + "epoch": 0.5350999535099954, + "grad_norm": 7.09688663482666, + "learning_rate": 9.998849999885e-06, + "loss": 0.8652, + "step": 1151 + }, + { + "epoch": 0.5355648535564853, + "grad_norm": 6.810689926147461, + "learning_rate": 9.9988489998849e-06, + "loss": 0.9242, + "step": 1152 + }, + { + "epoch": 0.5360297536029753, + "grad_norm": 6.674005508422852, + "learning_rate": 9.998847999884801e-06, + "loss": 0.6053, + "step": 1153 + }, + { + "epoch": 0.5364946536494654, + "grad_norm": 6.340108871459961, + "learning_rate": 9.9988469998847e-06, + "loss": 0.8287, + "step": 1154 + }, + { + "epoch": 0.5369595536959554, + "grad_norm": 13.675806045532227, + "learning_rate": 9.998845999884601e-06, + "loss": 0.8687, + "step": 1155 + }, + { + "epoch": 0.5374244537424454, + "grad_norm": 5.292993068695068, + "learning_rate": 9.9988449998845e-06, + "loss": 0.3108, + "step": 1156 + }, + { + "epoch": 0.5378893537889354, + "grad_norm": 8.915117263793945, + "learning_rate": 9.9988439998844e-06, + "loss": 0.9074, + "step": 1157 + }, + { + "epoch": 0.5383542538354253, + "grad_norm": 9.513984680175781, + "learning_rate": 9.998842999884301e-06, + "loss": 0.5283, + "step": 1158 + }, + { + "epoch": 0.5388191538819154, + "grad_norm": 7.2305755615234375, + "learning_rate": 9.998841999884202e-06, + "loss": 0.6814, + "step": 1159 + }, + { + "epoch": 0.5392840539284054, + "grad_norm": 8.482539176940918, + "learning_rate": 9.998840999884101e-06, + "loss": 0.8955, + "step": 1160 + }, + { + "epoch": 0.5397489539748954, + "grad_norm": 6.1795878410339355, + "learning_rate": 9.998839999884002e-06, + "loss": 0.4001, + "step": 1161 + }, + { + "epoch": 0.5402138540213854, + "grad_norm": 8.002968788146973, + "learning_rate": 9.9988389998839e-06, + "loss": 0.6249, + "step": 1162 + }, + { + "epoch": 0.5406787540678754, + "grad_norm": 5.46239709854126, + "learning_rate": 9.998837999883801e-06, + "loss": 0.4779, + "step": 1163 + }, + { + "epoch": 0.5411436541143654, + "grad_norm": 5.252678871154785, + "learning_rate": 9.9988369998837e-06, + "loss": 0.5003, + "step": 1164 + }, + { + "epoch": 0.5416085541608554, + "grad_norm": 6.753422260284424, + "learning_rate": 9.9988359998836e-06, + "loss": 0.7681, + "step": 1165 + }, + { + "epoch": 0.5420734542073454, + "grad_norm": 11.336237907409668, + "learning_rate": 9.998834999883501e-06, + "loss": 1.082, + "step": 1166 + }, + { + "epoch": 0.5425383542538355, + "grad_norm": 11.999167442321777, + "learning_rate": 9.9988339998834e-06, + "loss": 1.1167, + "step": 1167 + }, + { + "epoch": 0.5430032543003255, + "grad_norm": 12.010167121887207, + "learning_rate": 9.998832999883301e-06, + "loss": 0.8278, + "step": 1168 + }, + { + "epoch": 0.5434681543468154, + "grad_norm": 4.863112926483154, + "learning_rate": 9.9988319998832e-06, + "loss": 0.5013, + "step": 1169 + }, + { + "epoch": 0.5439330543933054, + "grad_norm": 6.103987693786621, + "learning_rate": 9.9988309998831e-06, + "loss": 0.4849, + "step": 1170 + }, + { + "epoch": 0.5443979544397954, + "grad_norm": 5.750401496887207, + "learning_rate": 9.998829999883001e-06, + "loss": 0.4879, + "step": 1171 + }, + { + "epoch": 0.5448628544862855, + "grad_norm": 10.24885082244873, + "learning_rate": 9.998828999882902e-06, + "loss": 0.7727, + "step": 1172 + }, + { + "epoch": 0.5453277545327755, + "grad_norm": 8.99259090423584, + "learning_rate": 9.998827999882801e-06, + "loss": 0.7348, + "step": 1173 + }, + { + "epoch": 0.5457926545792655, + "grad_norm": 7.780148029327393, + "learning_rate": 9.9988269998827e-06, + "loss": 0.5728, + "step": 1174 + }, + { + "epoch": 0.5462575546257554, + "grad_norm": 7.659729957580566, + "learning_rate": 9.9988259998826e-06, + "loss": 0.6233, + "step": 1175 + }, + { + "epoch": 0.5467224546722455, + "grad_norm": 7.934966564178467, + "learning_rate": 9.998824999882501e-06, + "loss": 0.8193, + "step": 1176 + }, + { + "epoch": 0.5471873547187355, + "grad_norm": 10.553998947143555, + "learning_rate": 9.998823999882402e-06, + "loss": 0.9653, + "step": 1177 + }, + { + "epoch": 0.5476522547652255, + "grad_norm": 6.4479594230651855, + "learning_rate": 9.9988229998823e-06, + "loss": 0.6261, + "step": 1178 + }, + { + "epoch": 0.5481171548117155, + "grad_norm": 8.497949600219727, + "learning_rate": 9.998821999882201e-06, + "loss": 0.6477, + "step": 1179 + }, + { + "epoch": 0.5485820548582054, + "grad_norm": 9.80571460723877, + "learning_rate": 9.9988209998821e-06, + "loss": 0.9608, + "step": 1180 + }, + { + "epoch": 0.5490469549046955, + "grad_norm": 14.84951114654541, + "learning_rate": 9.998819999882001e-06, + "loss": 1.3394, + "step": 1181 + }, + { + "epoch": 0.5495118549511855, + "grad_norm": 6.981801986694336, + "learning_rate": 9.9988189998819e-06, + "loss": 0.7772, + "step": 1182 + }, + { + "epoch": 0.5499767549976755, + "grad_norm": 5.281968116760254, + "learning_rate": 9.9988179998818e-06, + "loss": 0.4684, + "step": 1183 + }, + { + "epoch": 0.5504416550441655, + "grad_norm": 7.11374044418335, + "learning_rate": 9.998816999881701e-06, + "loss": 0.758, + "step": 1184 + }, + { + "epoch": 0.5509065550906556, + "grad_norm": 5.632142066955566, + "learning_rate": 9.9988159998816e-06, + "loss": 0.4589, + "step": 1185 + }, + { + "epoch": 0.5513714551371455, + "grad_norm": 8.167473793029785, + "learning_rate": 9.998814999881501e-06, + "loss": 0.6838, + "step": 1186 + }, + { + "epoch": 0.5518363551836355, + "grad_norm": 8.272212028503418, + "learning_rate": 9.9988139998814e-06, + "loss": 0.7224, + "step": 1187 + }, + { + "epoch": 0.5523012552301255, + "grad_norm": 3.7549638748168945, + "learning_rate": 9.9988129998813e-06, + "loss": 0.2447, + "step": 1188 + }, + { + "epoch": 0.5527661552766155, + "grad_norm": 8.699580192565918, + "learning_rate": 9.998811999881201e-06, + "loss": 1.015, + "step": 1189 + }, + { + "epoch": 0.5532310553231056, + "grad_norm": 11.238309860229492, + "learning_rate": 9.998810999881102e-06, + "loss": 0.9591, + "step": 1190 + }, + { + "epoch": 0.5536959553695955, + "grad_norm": 9.989469528198242, + "learning_rate": 9.998809999881001e-06, + "loss": 0.8406, + "step": 1191 + }, + { + "epoch": 0.5541608554160855, + "grad_norm": 5.343141555786133, + "learning_rate": 9.998808999880902e-06, + "loss": 0.3478, + "step": 1192 + }, + { + "epoch": 0.5546257554625755, + "grad_norm": 10.454519271850586, + "learning_rate": 9.9988079998808e-06, + "loss": 1.1532, + "step": 1193 + }, + { + "epoch": 0.5550906555090656, + "grad_norm": 7.539825439453125, + "learning_rate": 9.998806999880701e-06, + "loss": 1.0556, + "step": 1194 + }, + { + "epoch": 0.5555555555555556, + "grad_norm": 7.068440914154053, + "learning_rate": 9.9988059998806e-06, + "loss": 0.5104, + "step": 1195 + }, + { + "epoch": 0.5560204556020456, + "grad_norm": 7.301443099975586, + "learning_rate": 9.9988049998805e-06, + "loss": 0.7774, + "step": 1196 + }, + { + "epoch": 0.5564853556485355, + "grad_norm": 7.32213020324707, + "learning_rate": 9.998803999880401e-06, + "loss": 0.703, + "step": 1197 + }, + { + "epoch": 0.5569502556950255, + "grad_norm": 3.7514960765838623, + "learning_rate": 9.9988029998803e-06, + "loss": 0.1627, + "step": 1198 + }, + { + "epoch": 0.5574151557415156, + "grad_norm": 6.855757236480713, + "learning_rate": 9.998801999880201e-06, + "loss": 0.7124, + "step": 1199 + }, + { + "epoch": 0.5578800557880056, + "grad_norm": 7.1989054679870605, + "learning_rate": 9.9988009998801e-06, + "loss": 0.8365, + "step": 1200 + }, + { + "epoch": 0.5583449558344956, + "grad_norm": 7.623052597045898, + "learning_rate": 9.99879999988e-06, + "loss": 0.8905, + "step": 1201 + }, + { + "epoch": 0.5588098558809856, + "grad_norm": 5.2378082275390625, + "learning_rate": 9.998798999879901e-06, + "loss": 0.5574, + "step": 1202 + }, + { + "epoch": 0.5592747559274756, + "grad_norm": 6.8349289894104, + "learning_rate": 9.998797999879802e-06, + "loss": 0.8494, + "step": 1203 + }, + { + "epoch": 0.5597396559739656, + "grad_norm": 10.801729202270508, + "learning_rate": 9.998796999879701e-06, + "loss": 0.6413, + "step": 1204 + }, + { + "epoch": 0.5602045560204556, + "grad_norm": 10.353248596191406, + "learning_rate": 9.9987959998796e-06, + "loss": 0.7206, + "step": 1205 + }, + { + "epoch": 0.5606694560669456, + "grad_norm": 10.118453025817871, + "learning_rate": 9.9987949998795e-06, + "loss": 1.0842, + "step": 1206 + }, + { + "epoch": 0.5611343561134357, + "grad_norm": 6.048868656158447, + "learning_rate": 9.9987939998794e-06, + "loss": 0.8619, + "step": 1207 + }, + { + "epoch": 0.5615992561599256, + "grad_norm": 4.61629056930542, + "learning_rate": 9.9987929998793e-06, + "loss": 0.3242, + "step": 1208 + }, + { + "epoch": 0.5620641562064156, + "grad_norm": 6.17100715637207, + "learning_rate": 9.998791999879201e-06, + "loss": 0.7863, + "step": 1209 + }, + { + "epoch": 0.5625290562529056, + "grad_norm": 7.1041951179504395, + "learning_rate": 9.998790999879102e-06, + "loss": 0.7134, + "step": 1210 + }, + { + "epoch": 0.5629939562993956, + "grad_norm": 8.462076187133789, + "learning_rate": 9.998789999879e-06, + "loss": 1.0219, + "step": 1211 + }, + { + "epoch": 0.5634588563458857, + "grad_norm": 6.326632976531982, + "learning_rate": 9.998788999878901e-06, + "loss": 0.8215, + "step": 1212 + }, + { + "epoch": 0.5639237563923757, + "grad_norm": 6.709658622741699, + "learning_rate": 9.9987879998788e-06, + "loss": 0.4904, + "step": 1213 + }, + { + "epoch": 0.5643886564388656, + "grad_norm": 6.0241498947143555, + "learning_rate": 9.9987869998787e-06, + "loss": 0.5675, + "step": 1214 + }, + { + "epoch": 0.5648535564853556, + "grad_norm": 6.321414947509766, + "learning_rate": 9.998785999878601e-06, + "loss": 0.6735, + "step": 1215 + }, + { + "epoch": 0.5653184565318456, + "grad_norm": 16.37961196899414, + "learning_rate": 9.9987849998785e-06, + "loss": 1.1412, + "step": 1216 + }, + { + "epoch": 0.5657833565783357, + "grad_norm": 6.6295318603515625, + "learning_rate": 9.998783999878401e-06, + "loss": 0.6496, + "step": 1217 + }, + { + "epoch": 0.5662482566248257, + "grad_norm": 8.519168853759766, + "learning_rate": 9.9987829998783e-06, + "loss": 0.7471, + "step": 1218 + }, + { + "epoch": 0.5667131566713157, + "grad_norm": 5.951240062713623, + "learning_rate": 9.9987819998782e-06, + "loss": 0.6029, + "step": 1219 + }, + { + "epoch": 0.5671780567178056, + "grad_norm": 5.203462600708008, + "learning_rate": 9.9987809998781e-06, + "loss": 0.4991, + "step": 1220 + }, + { + "epoch": 0.5676429567642957, + "grad_norm": 9.097908973693848, + "learning_rate": 9.998779999878e-06, + "loss": 0.6814, + "step": 1221 + }, + { + "epoch": 0.5681078568107857, + "grad_norm": 10.489225387573242, + "learning_rate": 9.998778999877901e-06, + "loss": 0.8764, + "step": 1222 + }, + { + "epoch": 0.5685727568572757, + "grad_norm": 5.412479400634766, + "learning_rate": 9.998777999877802e-06, + "loss": 0.5426, + "step": 1223 + }, + { + "epoch": 0.5690376569037657, + "grad_norm": 7.247320652008057, + "learning_rate": 9.9987769998777e-06, + "loss": 0.671, + "step": 1224 + }, + { + "epoch": 0.5695025569502556, + "grad_norm": 6.907779693603516, + "learning_rate": 9.998775999877601e-06, + "loss": 0.8564, + "step": 1225 + }, + { + "epoch": 0.5699674569967457, + "grad_norm": 7.670416831970215, + "learning_rate": 9.9987749998775e-06, + "loss": 0.485, + "step": 1226 + }, + { + "epoch": 0.5704323570432357, + "grad_norm": 6.793957233428955, + "learning_rate": 9.998773999877401e-06, + "loss": 0.6226, + "step": 1227 + }, + { + "epoch": 0.5708972570897257, + "grad_norm": 7.165489196777344, + "learning_rate": 9.998772999877302e-06, + "loss": 0.8166, + "step": 1228 + }, + { + "epoch": 0.5713621571362157, + "grad_norm": 6.65006685256958, + "learning_rate": 9.9987719998772e-06, + "loss": 0.7468, + "step": 1229 + }, + { + "epoch": 0.5718270571827058, + "grad_norm": 6.237794399261475, + "learning_rate": 9.998770999877101e-06, + "loss": 0.6713, + "step": 1230 + }, + { + "epoch": 0.5722919572291957, + "grad_norm": 6.662072658538818, + "learning_rate": 9.998769999877e-06, + "loss": 0.4829, + "step": 1231 + }, + { + "epoch": 0.5727568572756857, + "grad_norm": 6.238709449768066, + "learning_rate": 9.9987689998769e-06, + "loss": 0.7572, + "step": 1232 + }, + { + "epoch": 0.5732217573221757, + "grad_norm": 6.051784992218018, + "learning_rate": 9.9987679998768e-06, + "loss": 0.5634, + "step": 1233 + }, + { + "epoch": 0.5736866573686658, + "grad_norm": 10.435773849487305, + "learning_rate": 9.9987669998767e-06, + "loss": 1.1471, + "step": 1234 + }, + { + "epoch": 0.5741515574151558, + "grad_norm": 9.941508293151855, + "learning_rate": 9.998765999876601e-06, + "loss": 0.4376, + "step": 1235 + }, + { + "epoch": 0.5746164574616457, + "grad_norm": 5.699219226837158, + "learning_rate": 9.998764999876502e-06, + "loss": 0.4748, + "step": 1236 + }, + { + "epoch": 0.5750813575081357, + "grad_norm": 6.406047344207764, + "learning_rate": 9.9987639998764e-06, + "loss": 0.6281, + "step": 1237 + }, + { + "epoch": 0.5755462575546257, + "grad_norm": 7.726429462432861, + "learning_rate": 9.9987629998763e-06, + "loss": 1.0905, + "step": 1238 + }, + { + "epoch": 0.5760111576011158, + "grad_norm": 9.63988971710205, + "learning_rate": 9.9987619998762e-06, + "loss": 1.0665, + "step": 1239 + }, + { + "epoch": 0.5764760576476058, + "grad_norm": 11.740028381347656, + "learning_rate": 9.998760999876101e-06, + "loss": 1.0781, + "step": 1240 + }, + { + "epoch": 0.5769409576940958, + "grad_norm": 12.851069450378418, + "learning_rate": 9.998759999876002e-06, + "loss": 1.1465, + "step": 1241 + }, + { + "epoch": 0.5774058577405857, + "grad_norm": 8.853347778320312, + "learning_rate": 9.9987589998759e-06, + "loss": 1.1419, + "step": 1242 + }, + { + "epoch": 0.5778707577870758, + "grad_norm": 10.25047779083252, + "learning_rate": 9.998757999875801e-06, + "loss": 1.0882, + "step": 1243 + }, + { + "epoch": 0.5783356578335658, + "grad_norm": 7.005180358886719, + "learning_rate": 9.9987569998757e-06, + "loss": 0.8854, + "step": 1244 + }, + { + "epoch": 0.5788005578800558, + "grad_norm": 5.884579658508301, + "learning_rate": 9.998755999875601e-06, + "loss": 0.5258, + "step": 1245 + }, + { + "epoch": 0.5792654579265458, + "grad_norm": 8.895803451538086, + "learning_rate": 9.998754999875502e-06, + "loss": 1.048, + "step": 1246 + }, + { + "epoch": 0.5797303579730358, + "grad_norm": 7.50189733505249, + "learning_rate": 9.9987539998754e-06, + "loss": 0.7725, + "step": 1247 + }, + { + "epoch": 0.5801952580195258, + "grad_norm": 6.165489673614502, + "learning_rate": 9.998752999875301e-06, + "loss": 0.5593, + "step": 1248 + }, + { + "epoch": 0.5806601580660158, + "grad_norm": 6.937504768371582, + "learning_rate": 9.9987519998752e-06, + "loss": 0.6547, + "step": 1249 + }, + { + "epoch": 0.5811250581125058, + "grad_norm": 5.12592887878418, + "learning_rate": 9.9987509998751e-06, + "loss": 0.4515, + "step": 1250 + }, + { + "epoch": 0.5815899581589958, + "grad_norm": 7.1295318603515625, + "learning_rate": 9.998749999875e-06, + "loss": 0.3968, + "step": 1251 + }, + { + "epoch": 0.5820548582054859, + "grad_norm": 6.528859615325928, + "learning_rate": 9.9987489998749e-06, + "loss": 0.7501, + "step": 1252 + }, + { + "epoch": 0.5825197582519758, + "grad_norm": 7.334307670593262, + "learning_rate": 9.998747999874801e-06, + "loss": 0.7498, + "step": 1253 + }, + { + "epoch": 0.5829846582984658, + "grad_norm": 5.430944442749023, + "learning_rate": 9.998746999874702e-06, + "loss": 0.6747, + "step": 1254 + }, + { + "epoch": 0.5834495583449558, + "grad_norm": 8.684850692749023, + "learning_rate": 9.9987459998746e-06, + "loss": 0.8592, + "step": 1255 + }, + { + "epoch": 0.5839144583914458, + "grad_norm": 8.935995101928711, + "learning_rate": 9.998744999874501e-06, + "loss": 0.7484, + "step": 1256 + }, + { + "epoch": 0.5843793584379359, + "grad_norm": 4.859222888946533, + "learning_rate": 9.9987439998744e-06, + "loss": 0.369, + "step": 1257 + }, + { + "epoch": 0.5848442584844259, + "grad_norm": 9.138697624206543, + "learning_rate": 9.998742999874301e-06, + "loss": 0.5202, + "step": 1258 + }, + { + "epoch": 0.5853091585309158, + "grad_norm": 7.615625858306885, + "learning_rate": 9.998741999874202e-06, + "loss": 0.8021, + "step": 1259 + }, + { + "epoch": 0.5857740585774058, + "grad_norm": 8.281704902648926, + "learning_rate": 9.9987409998741e-06, + "loss": 0.6333, + "step": 1260 + }, + { + "epoch": 0.5862389586238959, + "grad_norm": 8.878538131713867, + "learning_rate": 9.998739999874001e-06, + "loss": 0.8999, + "step": 1261 + }, + { + "epoch": 0.5867038586703859, + "grad_norm": 8.095192909240723, + "learning_rate": 9.9987389998739e-06, + "loss": 0.8841, + "step": 1262 + }, + { + "epoch": 0.5871687587168759, + "grad_norm": 4.948946475982666, + "learning_rate": 9.998737999873801e-06, + "loss": 0.6455, + "step": 1263 + }, + { + "epoch": 0.5876336587633659, + "grad_norm": 11.662871360778809, + "learning_rate": 9.9987369998737e-06, + "loss": 0.6701, + "step": 1264 + }, + { + "epoch": 0.5880985588098558, + "grad_norm": 9.504828453063965, + "learning_rate": 9.9987359998736e-06, + "loss": 0.8849, + "step": 1265 + }, + { + "epoch": 0.5885634588563459, + "grad_norm": 10.189004898071289, + "learning_rate": 9.998734999873501e-06, + "loss": 0.9194, + "step": 1266 + }, + { + "epoch": 0.5890283589028359, + "grad_norm": 8.31252670288086, + "learning_rate": 9.998733999873402e-06, + "loss": 1.0281, + "step": 1267 + }, + { + "epoch": 0.5894932589493259, + "grad_norm": 5.270479202270508, + "learning_rate": 9.9987329998733e-06, + "loss": 0.4713, + "step": 1268 + }, + { + "epoch": 0.5899581589958159, + "grad_norm": 8.101851463317871, + "learning_rate": 9.998731999873201e-06, + "loss": 1.0257, + "step": 1269 + }, + { + "epoch": 0.5904230590423059, + "grad_norm": 6.352687835693359, + "learning_rate": 9.9987309998731e-06, + "loss": 0.4663, + "step": 1270 + }, + { + "epoch": 0.5908879590887959, + "grad_norm": 6.9872050285339355, + "learning_rate": 9.998729999873001e-06, + "loss": 0.6939, + "step": 1271 + }, + { + "epoch": 0.5913528591352859, + "grad_norm": 7.856936931610107, + "learning_rate": 9.998728999872902e-06, + "loss": 1.0611, + "step": 1272 + }, + { + "epoch": 0.5918177591817759, + "grad_norm": 6.583176136016846, + "learning_rate": 9.9987279998728e-06, + "loss": 0.5963, + "step": 1273 + }, + { + "epoch": 0.592282659228266, + "grad_norm": 11.044299125671387, + "learning_rate": 9.998726999872701e-06, + "loss": 0.6389, + "step": 1274 + }, + { + "epoch": 0.592747559274756, + "grad_norm": 5.3225178718566895, + "learning_rate": 9.9987259998726e-06, + "loss": 0.7171, + "step": 1275 + }, + { + "epoch": 0.5932124593212459, + "grad_norm": 10.62732219696045, + "learning_rate": 9.998724999872501e-06, + "loss": 0.9811, + "step": 1276 + }, + { + "epoch": 0.5936773593677359, + "grad_norm": 8.610602378845215, + "learning_rate": 9.9987239998724e-06, + "loss": 0.9382, + "step": 1277 + }, + { + "epoch": 0.5941422594142259, + "grad_norm": 7.383626461029053, + "learning_rate": 9.9987229998723e-06, + "loss": 0.6264, + "step": 1278 + }, + { + "epoch": 0.594607159460716, + "grad_norm": 6.374850749969482, + "learning_rate": 9.998721999872201e-06, + "loss": 0.6624, + "step": 1279 + }, + { + "epoch": 0.595072059507206, + "grad_norm": 6.161890029907227, + "learning_rate": 9.998720999872102e-06, + "loss": 0.5218, + "step": 1280 + }, + { + "epoch": 0.595536959553696, + "grad_norm": 12.816766738891602, + "learning_rate": 9.998719999872e-06, + "loss": 0.9589, + "step": 1281 + }, + { + "epoch": 0.5960018596001859, + "grad_norm": 7.554964542388916, + "learning_rate": 9.9987189998719e-06, + "loss": 0.8957, + "step": 1282 + }, + { + "epoch": 0.596466759646676, + "grad_norm": 10.320991516113281, + "learning_rate": 9.9987179998718e-06, + "loss": 0.578, + "step": 1283 + }, + { + "epoch": 0.596931659693166, + "grad_norm": 6.703699111938477, + "learning_rate": 9.998716999871701e-06, + "loss": 0.477, + "step": 1284 + }, + { + "epoch": 0.597396559739656, + "grad_norm": 7.518505096435547, + "learning_rate": 9.998715999871602e-06, + "loss": 0.7338, + "step": 1285 + }, + { + "epoch": 0.597861459786146, + "grad_norm": 6.030684471130371, + "learning_rate": 9.9987149998715e-06, + "loss": 0.6, + "step": 1286 + }, + { + "epoch": 0.5983263598326359, + "grad_norm": 5.840198993682861, + "learning_rate": 9.998713999871401e-06, + "loss": 0.5768, + "step": 1287 + }, + { + "epoch": 0.598791259879126, + "grad_norm": 7.274750232696533, + "learning_rate": 9.9987129998713e-06, + "loss": 0.6734, + "step": 1288 + }, + { + "epoch": 0.599256159925616, + "grad_norm": 5.450262069702148, + "learning_rate": 9.998711999871201e-06, + "loss": 0.507, + "step": 1289 + }, + { + "epoch": 0.599721059972106, + "grad_norm": 9.735919952392578, + "learning_rate": 9.9987109998711e-06, + "loss": 1.0079, + "step": 1290 + }, + { + "epoch": 0.600185960018596, + "grad_norm": 7.162991523742676, + "learning_rate": 9.998709999871e-06, + "loss": 1.0361, + "step": 1291 + }, + { + "epoch": 0.6006508600650861, + "grad_norm": 7.0911760330200195, + "learning_rate": 9.998708999870901e-06, + "loss": 0.7398, + "step": 1292 + }, + { + "epoch": 0.601115760111576, + "grad_norm": 5.475833892822266, + "learning_rate": 9.9987079998708e-06, + "loss": 0.4902, + "step": 1293 + }, + { + "epoch": 0.601580660158066, + "grad_norm": 7.434250354766846, + "learning_rate": 9.998706999870701e-06, + "loss": 0.9992, + "step": 1294 + }, + { + "epoch": 0.602045560204556, + "grad_norm": 8.854272842407227, + "learning_rate": 9.9987059998706e-06, + "loss": 0.4028, + "step": 1295 + }, + { + "epoch": 0.602510460251046, + "grad_norm": 10.611333847045898, + "learning_rate": 9.9987049998705e-06, + "loss": 1.3369, + "step": 1296 + }, + { + "epoch": 0.6029753602975361, + "grad_norm": 6.56153678894043, + "learning_rate": 9.998703999870401e-06, + "loss": 0.5944, + "step": 1297 + }, + { + "epoch": 0.603440260344026, + "grad_norm": 6.96737813949585, + "learning_rate": 9.998702999870302e-06, + "loss": 0.4666, + "step": 1298 + }, + { + "epoch": 0.603905160390516, + "grad_norm": 7.832687854766846, + "learning_rate": 9.9987019998702e-06, + "loss": 0.537, + "step": 1299 + }, + { + "epoch": 0.604370060437006, + "grad_norm": 10.14950180053711, + "learning_rate": 9.998700999870101e-06, + "loss": 1.2397, + "step": 1300 + }, + { + "epoch": 0.604834960483496, + "grad_norm": 6.05402946472168, + "learning_rate": 9.99869999987e-06, + "loss": 0.7286, + "step": 1301 + }, + { + "epoch": 0.6052998605299861, + "grad_norm": 5.909971714019775, + "learning_rate": 9.998698999869901e-06, + "loss": 0.4136, + "step": 1302 + }, + { + "epoch": 0.6057647605764761, + "grad_norm": 7.279001712799072, + "learning_rate": 9.9986979998698e-06, + "loss": 0.8693, + "step": 1303 + }, + { + "epoch": 0.606229660622966, + "grad_norm": 6.7187299728393555, + "learning_rate": 9.9986969998697e-06, + "loss": 0.7424, + "step": 1304 + }, + { + "epoch": 0.606694560669456, + "grad_norm": 6.216047286987305, + "learning_rate": 9.998695999869601e-06, + "loss": 0.6413, + "step": 1305 + }, + { + "epoch": 0.6071594607159461, + "grad_norm": 12.672630310058594, + "learning_rate": 9.9986949998695e-06, + "loss": 0.7502, + "step": 1306 + }, + { + "epoch": 0.6076243607624361, + "grad_norm": 5.532203674316406, + "learning_rate": 9.998693999869401e-06, + "loss": 0.4439, + "step": 1307 + }, + { + "epoch": 0.6080892608089261, + "grad_norm": 6.356549263000488, + "learning_rate": 9.9986929998693e-06, + "loss": 0.6255, + "step": 1308 + }, + { + "epoch": 0.6085541608554161, + "grad_norm": 4.935951232910156, + "learning_rate": 9.9986919998692e-06, + "loss": 0.3174, + "step": 1309 + }, + { + "epoch": 0.609019060901906, + "grad_norm": 5.144252300262451, + "learning_rate": 9.998690999869101e-06, + "loss": 0.4819, + "step": 1310 + }, + { + "epoch": 0.6094839609483961, + "grad_norm": 7.370270252227783, + "learning_rate": 9.998689999869002e-06, + "loss": 0.7622, + "step": 1311 + }, + { + "epoch": 0.6099488609948861, + "grad_norm": 5.741710186004639, + "learning_rate": 9.998688999868901e-06, + "loss": 0.5608, + "step": 1312 + }, + { + "epoch": 0.6104137610413761, + "grad_norm": 5.967807292938232, + "learning_rate": 9.998687999868802e-06, + "loss": 0.4155, + "step": 1313 + }, + { + "epoch": 0.6108786610878661, + "grad_norm": 10.78149700164795, + "learning_rate": 9.9986869998687e-06, + "loss": 0.6304, + "step": 1314 + }, + { + "epoch": 0.6113435611343561, + "grad_norm": 5.935286521911621, + "learning_rate": 9.998685999868601e-06, + "loss": 0.6312, + "step": 1315 + }, + { + "epoch": 0.6118084611808461, + "grad_norm": 9.760269165039062, + "learning_rate": 9.9986849998685e-06, + "loss": 1.3718, + "step": 1316 + }, + { + "epoch": 0.6122733612273361, + "grad_norm": 10.472648620605469, + "learning_rate": 9.9986839998684e-06, + "loss": 0.7055, + "step": 1317 + }, + { + "epoch": 0.6127382612738261, + "grad_norm": 9.658890724182129, + "learning_rate": 9.998682999868301e-06, + "loss": 1.2013, + "step": 1318 + }, + { + "epoch": 0.6132031613203162, + "grad_norm": 9.485122680664062, + "learning_rate": 9.9986819998682e-06, + "loss": 0.6031, + "step": 1319 + }, + { + "epoch": 0.6136680613668062, + "grad_norm": 7.716256141662598, + "learning_rate": 9.998680999868101e-06, + "loss": 0.319, + "step": 1320 + }, + { + "epoch": 0.6141329614132961, + "grad_norm": 7.167984485626221, + "learning_rate": 9.998679999868e-06, + "loss": 0.788, + "step": 1321 + }, + { + "epoch": 0.6145978614597861, + "grad_norm": 8.37767505645752, + "learning_rate": 9.9986789998679e-06, + "loss": 0.4226, + "step": 1322 + }, + { + "epoch": 0.6150627615062761, + "grad_norm": 8.91722583770752, + "learning_rate": 9.998677999867801e-06, + "loss": 0.9359, + "step": 1323 + }, + { + "epoch": 0.6155276615527662, + "grad_norm": 8.422332763671875, + "learning_rate": 9.998676999867702e-06, + "loss": 0.5939, + "step": 1324 + }, + { + "epoch": 0.6159925615992562, + "grad_norm": 7.777310371398926, + "learning_rate": 9.998675999867601e-06, + "loss": 0.6049, + "step": 1325 + }, + { + "epoch": 0.6164574616457462, + "grad_norm": 6.9589009284973145, + "learning_rate": 9.9986749998675e-06, + "loss": 1.0234, + "step": 1326 + }, + { + "epoch": 0.6169223616922361, + "grad_norm": 7.545405387878418, + "learning_rate": 9.9986739998674e-06, + "loss": 0.6593, + "step": 1327 + }, + { + "epoch": 0.6173872617387262, + "grad_norm": 9.4163236618042, + "learning_rate": 9.998672999867301e-06, + "loss": 1.3969, + "step": 1328 + }, + { + "epoch": 0.6178521617852162, + "grad_norm": 9.072214126586914, + "learning_rate": 9.9986719998672e-06, + "loss": 0.9113, + "step": 1329 + }, + { + "epoch": 0.6183170618317062, + "grad_norm": 15.076622009277344, + "learning_rate": 9.998670999867101e-06, + "loss": 1.0701, + "step": 1330 + }, + { + "epoch": 0.6187819618781962, + "grad_norm": 6.515867233276367, + "learning_rate": 9.998669999867002e-06, + "loss": 0.6636, + "step": 1331 + }, + { + "epoch": 0.6192468619246861, + "grad_norm": 5.250950813293457, + "learning_rate": 9.9986689998669e-06, + "loss": 0.4553, + "step": 1332 + }, + { + "epoch": 0.6197117619711762, + "grad_norm": 6.208817481994629, + "learning_rate": 9.998667999866801e-06, + "loss": 0.6012, + "step": 1333 + }, + { + "epoch": 0.6201766620176662, + "grad_norm": 8.055994033813477, + "learning_rate": 9.9986669998667e-06, + "loss": 0.7665, + "step": 1334 + }, + { + "epoch": 0.6206415620641562, + "grad_norm": 8.463043212890625, + "learning_rate": 9.9986659998666e-06, + "loss": 0.6594, + "step": 1335 + }, + { + "epoch": 0.6211064621106462, + "grad_norm": 7.638784408569336, + "learning_rate": 9.998664999866501e-06, + "loss": 0.9354, + "step": 1336 + }, + { + "epoch": 0.6215713621571363, + "grad_norm": 6.52336311340332, + "learning_rate": 9.9986639998664e-06, + "loss": 0.785, + "step": 1337 + }, + { + "epoch": 0.6220362622036262, + "grad_norm": 13.312156677246094, + "learning_rate": 9.998662999866301e-06, + "loss": 1.3122, + "step": 1338 + }, + { + "epoch": 0.6225011622501162, + "grad_norm": 10.179048538208008, + "learning_rate": 9.9986619998662e-06, + "loss": 0.3186, + "step": 1339 + }, + { + "epoch": 0.6229660622966062, + "grad_norm": 8.535512924194336, + "learning_rate": 9.9986609998661e-06, + "loss": 1.0695, + "step": 1340 + }, + { + "epoch": 0.6234309623430963, + "grad_norm": 6.04429292678833, + "learning_rate": 9.998659999866001e-06, + "loss": 0.4478, + "step": 1341 + }, + { + "epoch": 0.6238958623895863, + "grad_norm": 8.700727462768555, + "learning_rate": 9.998658999865902e-06, + "loss": 0.6236, + "step": 1342 + }, + { + "epoch": 0.6243607624360762, + "grad_norm": 8.552109718322754, + "learning_rate": 9.998657999865801e-06, + "loss": 0.8378, + "step": 1343 + }, + { + "epoch": 0.6248256624825662, + "grad_norm": 6.839399814605713, + "learning_rate": 9.998656999865702e-06, + "loss": 0.9138, + "step": 1344 + }, + { + "epoch": 0.6252905625290562, + "grad_norm": 6.754815578460693, + "learning_rate": 9.9986559998656e-06, + "loss": 0.5625, + "step": 1345 + }, + { + "epoch": 0.6257554625755463, + "grad_norm": 6.672272205352783, + "learning_rate": 9.9986549998655e-06, + "loss": 0.5867, + "step": 1346 + }, + { + "epoch": 0.6262203626220363, + "grad_norm": 6.959560394287109, + "learning_rate": 9.9986539998654e-06, + "loss": 0.7598, + "step": 1347 + }, + { + "epoch": 0.6266852626685263, + "grad_norm": 6.904024600982666, + "learning_rate": 9.998652999865301e-06, + "loss": 0.6896, + "step": 1348 + }, + { + "epoch": 0.6271501627150162, + "grad_norm": 9.013762474060059, + "learning_rate": 9.998651999865202e-06, + "loss": 0.6477, + "step": 1349 + }, + { + "epoch": 0.6276150627615062, + "grad_norm": 7.811771869659424, + "learning_rate": 9.9986509998651e-06, + "loss": 0.8784, + "step": 1350 + }, + { + "epoch": 0.6280799628079963, + "grad_norm": 6.3183465003967285, + "learning_rate": 9.998649999865001e-06, + "loss": 0.4898, + "step": 1351 + }, + { + "epoch": 0.6285448628544863, + "grad_norm": 8.596487045288086, + "learning_rate": 9.9986489998649e-06, + "loss": 0.9963, + "step": 1352 + }, + { + "epoch": 0.6290097629009763, + "grad_norm": 7.773998260498047, + "learning_rate": 9.9986479998648e-06, + "loss": 0.6862, + "step": 1353 + }, + { + "epoch": 0.6294746629474663, + "grad_norm": 7.092819690704346, + "learning_rate": 9.998646999864701e-06, + "loss": 0.8182, + "step": 1354 + }, + { + "epoch": 0.6299395629939563, + "grad_norm": 9.255434036254883, + "learning_rate": 9.998645999864602e-06, + "loss": 0.8595, + "step": 1355 + }, + { + "epoch": 0.6304044630404463, + "grad_norm": 10.310294151306152, + "learning_rate": 9.998644999864501e-06, + "loss": 1.9956, + "step": 1356 + }, + { + "epoch": 0.6308693630869363, + "grad_norm": 6.350977897644043, + "learning_rate": 9.9986439998644e-06, + "loss": 0.4802, + "step": 1357 + }, + { + "epoch": 0.6313342631334263, + "grad_norm": 7.502711772918701, + "learning_rate": 9.9986429998643e-06, + "loss": 0.814, + "step": 1358 + }, + { + "epoch": 0.6317991631799164, + "grad_norm": 7.511549949645996, + "learning_rate": 9.9986419998642e-06, + "loss": 0.8653, + "step": 1359 + }, + { + "epoch": 0.6322640632264063, + "grad_norm": 7.939019680023193, + "learning_rate": 9.9986409998641e-06, + "loss": 0.6478, + "step": 1360 + }, + { + "epoch": 0.6327289632728963, + "grad_norm": 9.730749130249023, + "learning_rate": 9.998639999864001e-06, + "loss": 1.0707, + "step": 1361 + }, + { + "epoch": 0.6331938633193863, + "grad_norm": 6.097006797790527, + "learning_rate": 9.998638999863902e-06, + "loss": 0.5338, + "step": 1362 + }, + { + "epoch": 0.6336587633658763, + "grad_norm": 6.911073684692383, + "learning_rate": 9.9986379998638e-06, + "loss": 0.7511, + "step": 1363 + }, + { + "epoch": 0.6341236634123664, + "grad_norm": 10.397120475769043, + "learning_rate": 9.998636999863701e-06, + "loss": 0.8151, + "step": 1364 + }, + { + "epoch": 0.6345885634588564, + "grad_norm": 9.272366523742676, + "learning_rate": 9.9986359998636e-06, + "loss": 1.0082, + "step": 1365 + }, + { + "epoch": 0.6350534635053463, + "grad_norm": 5.5689616203308105, + "learning_rate": 9.9986349998635e-06, + "loss": 0.5428, + "step": 1366 + }, + { + "epoch": 0.6355183635518363, + "grad_norm": 8.248492240905762, + "learning_rate": 9.998633999863401e-06, + "loss": 1.0633, + "step": 1367 + }, + { + "epoch": 0.6359832635983264, + "grad_norm": 3.995124101638794, + "learning_rate": 9.9986329998633e-06, + "loss": 0.2417, + "step": 1368 + }, + { + "epoch": 0.6364481636448164, + "grad_norm": 4.613018035888672, + "learning_rate": 9.998631999863201e-06, + "loss": 0.4789, + "step": 1369 + }, + { + "epoch": 0.6369130636913064, + "grad_norm": 9.513778686523438, + "learning_rate": 9.9986309998631e-06, + "loss": 0.6987, + "step": 1370 + }, + { + "epoch": 0.6373779637377964, + "grad_norm": 6.865842342376709, + "learning_rate": 9.998629999863e-06, + "loss": 0.6982, + "step": 1371 + }, + { + "epoch": 0.6378428637842863, + "grad_norm": 7.714282035827637, + "learning_rate": 9.9986289998629e-06, + "loss": 0.7679, + "step": 1372 + }, + { + "epoch": 0.6383077638307764, + "grad_norm": 6.850600242614746, + "learning_rate": 9.9986279998628e-06, + "loss": 0.7289, + "step": 1373 + }, + { + "epoch": 0.6387726638772664, + "grad_norm": 5.010479927062988, + "learning_rate": 9.998626999862701e-06, + "loss": 0.4684, + "step": 1374 + }, + { + "epoch": 0.6392375639237564, + "grad_norm": 7.782680988311768, + "learning_rate": 9.998625999862602e-06, + "loss": 0.9767, + "step": 1375 + }, + { + "epoch": 0.6397024639702464, + "grad_norm": 4.411804676055908, + "learning_rate": 9.9986249998625e-06, + "loss": 0.2687, + "step": 1376 + }, + { + "epoch": 0.6401673640167364, + "grad_norm": 8.475860595703125, + "learning_rate": 9.998623999862401e-06, + "loss": 1.153, + "step": 1377 + }, + { + "epoch": 0.6406322640632264, + "grad_norm": 5.261316299438477, + "learning_rate": 9.9986229998623e-06, + "loss": 0.4339, + "step": 1378 + }, + { + "epoch": 0.6410971641097164, + "grad_norm": 12.558639526367188, + "learning_rate": 9.998621999862201e-06, + "loss": 0.9774, + "step": 1379 + }, + { + "epoch": 0.6415620641562064, + "grad_norm": 8.975348472595215, + "learning_rate": 9.998620999862102e-06, + "loss": 1.3178, + "step": 1380 + }, + { + "epoch": 0.6420269642026964, + "grad_norm": 4.752361297607422, + "learning_rate": 9.998619999862e-06, + "loss": 0.4722, + "step": 1381 + }, + { + "epoch": 0.6424918642491865, + "grad_norm": 8.365156173706055, + "learning_rate": 9.998618999861901e-06, + "loss": 0.6105, + "step": 1382 + }, + { + "epoch": 0.6429567642956764, + "grad_norm": 8.569055557250977, + "learning_rate": 9.9986179998618e-06, + "loss": 0.9854, + "step": 1383 + }, + { + "epoch": 0.6434216643421664, + "grad_norm": 7.768671035766602, + "learning_rate": 9.9986169998617e-06, + "loss": 1.1138, + "step": 1384 + }, + { + "epoch": 0.6438865643886564, + "grad_norm": 8.682705879211426, + "learning_rate": 9.9986159998616e-06, + "loss": 0.8145, + "step": 1385 + }, + { + "epoch": 0.6443514644351465, + "grad_norm": 9.04330825805664, + "learning_rate": 9.9986149998615e-06, + "loss": 0.9585, + "step": 1386 + }, + { + "epoch": 0.6448163644816365, + "grad_norm": 4.9751482009887695, + "learning_rate": 9.998613999861401e-06, + "loss": 0.4003, + "step": 1387 + }, + { + "epoch": 0.6452812645281265, + "grad_norm": 6.60760498046875, + "learning_rate": 9.998612999861302e-06, + "loss": 0.8314, + "step": 1388 + }, + { + "epoch": 0.6457461645746164, + "grad_norm": 11.551517486572266, + "learning_rate": 9.9986119998612e-06, + "loss": 1.2988, + "step": 1389 + }, + { + "epoch": 0.6462110646211064, + "grad_norm": 8.102108001708984, + "learning_rate": 9.9986109998611e-06, + "loss": 0.8993, + "step": 1390 + }, + { + "epoch": 0.6466759646675965, + "grad_norm": 8.185206413269043, + "learning_rate": 9.998609999861e-06, + "loss": 0.8338, + "step": 1391 + }, + { + "epoch": 0.6471408647140865, + "grad_norm": 5.403822422027588, + "learning_rate": 9.998608999860901e-06, + "loss": 0.5043, + "step": 1392 + }, + { + "epoch": 0.6476057647605765, + "grad_norm": 17.98576545715332, + "learning_rate": 9.998607999860802e-06, + "loss": 0.6733, + "step": 1393 + }, + { + "epoch": 0.6480706648070664, + "grad_norm": 7.0294342041015625, + "learning_rate": 9.9986069998607e-06, + "loss": 0.6943, + "step": 1394 + }, + { + "epoch": 0.6485355648535565, + "grad_norm": 6.983304500579834, + "learning_rate": 9.998605999860601e-06, + "loss": 0.7092, + "step": 1395 + }, + { + "epoch": 0.6490004649000465, + "grad_norm": 7.262372970581055, + "learning_rate": 9.9986049998605e-06, + "loss": 0.7994, + "step": 1396 + }, + { + "epoch": 0.6494653649465365, + "grad_norm": 10.416267395019531, + "learning_rate": 9.998603999860401e-06, + "loss": 0.8, + "step": 1397 + }, + { + "epoch": 0.6499302649930265, + "grad_norm": 4.967037677764893, + "learning_rate": 9.9986029998603e-06, + "loss": 0.4336, + "step": 1398 + }, + { + "epoch": 0.6503951650395166, + "grad_norm": 5.580690860748291, + "learning_rate": 9.9986019998602e-06, + "loss": 0.6675, + "step": 1399 + }, + { + "epoch": 0.6508600650860065, + "grad_norm": 7.61266565322876, + "learning_rate": 9.998600999860101e-06, + "loss": 0.9137, + "step": 1400 + }, + { + "epoch": 0.6513249651324965, + "grad_norm": 9.062201499938965, + "learning_rate": 9.99859999986e-06, + "loss": 0.934, + "step": 1401 + }, + { + "epoch": 0.6517898651789865, + "grad_norm": 7.559817790985107, + "learning_rate": 9.9985989998599e-06, + "loss": 1.0283, + "step": 1402 + }, + { + "epoch": 0.6522547652254765, + "grad_norm": 11.701279640197754, + "learning_rate": 9.9985979998598e-06, + "loss": 1.5473, + "step": 1403 + }, + { + "epoch": 0.6527196652719666, + "grad_norm": 4.1497907638549805, + "learning_rate": 9.9985969998597e-06, + "loss": 0.3646, + "step": 1404 + }, + { + "epoch": 0.6531845653184565, + "grad_norm": 11.094557762145996, + "learning_rate": 9.998595999859601e-06, + "loss": 0.3202, + "step": 1405 + }, + { + "epoch": 0.6536494653649465, + "grad_norm": 6.801319599151611, + "learning_rate": 9.998594999859502e-06, + "loss": 0.5358, + "step": 1406 + }, + { + "epoch": 0.6541143654114365, + "grad_norm": 7.102147579193115, + "learning_rate": 9.9985939998594e-06, + "loss": 0.7865, + "step": 1407 + }, + { + "epoch": 0.6545792654579266, + "grad_norm": 5.193842887878418, + "learning_rate": 9.998592999859301e-06, + "loss": 0.4595, + "step": 1408 + }, + { + "epoch": 0.6550441655044166, + "grad_norm": 8.791281700134277, + "learning_rate": 9.9985919998592e-06, + "loss": 1.1154, + "step": 1409 + }, + { + "epoch": 0.6555090655509066, + "grad_norm": 3.4129974842071533, + "learning_rate": 9.998590999859101e-06, + "loss": 0.2787, + "step": 1410 + }, + { + "epoch": 0.6559739655973965, + "grad_norm": 8.319037437438965, + "learning_rate": 9.998589999859002e-06, + "loss": 0.6508, + "step": 1411 + }, + { + "epoch": 0.6564388656438865, + "grad_norm": 7.641232967376709, + "learning_rate": 9.9985889998589e-06, + "loss": 0.9108, + "step": 1412 + }, + { + "epoch": 0.6569037656903766, + "grad_norm": 6.270259380340576, + "learning_rate": 9.998587999858801e-06, + "loss": 0.3937, + "step": 1413 + }, + { + "epoch": 0.6573686657368666, + "grad_norm": 4.727664470672607, + "learning_rate": 9.9985869998587e-06, + "loss": 0.3766, + "step": 1414 + }, + { + "epoch": 0.6578335657833566, + "grad_norm": 6.76374626159668, + "learning_rate": 9.998585999858601e-06, + "loss": 0.6742, + "step": 1415 + }, + { + "epoch": 0.6582984658298466, + "grad_norm": 6.445496559143066, + "learning_rate": 9.9985849998585e-06, + "loss": 0.5289, + "step": 1416 + }, + { + "epoch": 0.6587633658763365, + "grad_norm": 6.294702053070068, + "learning_rate": 9.9985839998584e-06, + "loss": 0.7758, + "step": 1417 + }, + { + "epoch": 0.6592282659228266, + "grad_norm": 7.191268444061279, + "learning_rate": 9.998582999858301e-06, + "loss": 0.489, + "step": 1418 + }, + { + "epoch": 0.6596931659693166, + "grad_norm": 6.695955276489258, + "learning_rate": 9.998581999858202e-06, + "loss": 0.8324, + "step": 1419 + }, + { + "epoch": 0.6601580660158066, + "grad_norm": 6.459774494171143, + "learning_rate": 9.9985809998581e-06, + "loss": 0.7164, + "step": 1420 + }, + { + "epoch": 0.6606229660622966, + "grad_norm": 8.091227531433105, + "learning_rate": 9.998579999858001e-06, + "loss": 1.0813, + "step": 1421 + }, + { + "epoch": 0.6610878661087866, + "grad_norm": 7.248233318328857, + "learning_rate": 9.9985789998579e-06, + "loss": 0.6626, + "step": 1422 + }, + { + "epoch": 0.6615527661552766, + "grad_norm": 4.970542907714844, + "learning_rate": 9.998577999857801e-06, + "loss": 0.5622, + "step": 1423 + }, + { + "epoch": 0.6620176662017666, + "grad_norm": 7.024615287780762, + "learning_rate": 9.998576999857702e-06, + "loss": 0.5329, + "step": 1424 + }, + { + "epoch": 0.6624825662482566, + "grad_norm": 7.281092166900635, + "learning_rate": 9.9985759998576e-06, + "loss": 0.9953, + "step": 1425 + }, + { + "epoch": 0.6629474662947467, + "grad_norm": 7.081650257110596, + "learning_rate": 9.998574999857501e-06, + "loss": 0.6994, + "step": 1426 + }, + { + "epoch": 0.6634123663412367, + "grad_norm": 7.167159557342529, + "learning_rate": 9.9985739998574e-06, + "loss": 0.8711, + "step": 1427 + }, + { + "epoch": 0.6638772663877266, + "grad_norm": 4.242079257965088, + "learning_rate": 9.998572999857301e-06, + "loss": 0.5606, + "step": 1428 + }, + { + "epoch": 0.6643421664342166, + "grad_norm": 12.135812759399414, + "learning_rate": 9.9985719998572e-06, + "loss": 1.0879, + "step": 1429 + }, + { + "epoch": 0.6648070664807066, + "grad_norm": 7.6369733810424805, + "learning_rate": 9.9985709998571e-06, + "loss": 0.6969, + "step": 1430 + }, + { + "epoch": 0.6652719665271967, + "grad_norm": 8.912035942077637, + "learning_rate": 9.998569999857001e-06, + "loss": 0.9021, + "step": 1431 + }, + { + "epoch": 0.6657368665736867, + "grad_norm": 9.961666107177734, + "learning_rate": 9.998568999856902e-06, + "loss": 0.7369, + "step": 1432 + }, + { + "epoch": 0.6662017666201767, + "grad_norm": 6.519305229187012, + "learning_rate": 9.998567999856801e-06, + "loss": 0.4612, + "step": 1433 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 8.547667503356934, + "learning_rate": 9.9985669998567e-06, + "loss": 0.9408, + "step": 1434 + }, + { + "epoch": 0.6671315667131567, + "grad_norm": 9.497273445129395, + "learning_rate": 9.9985659998566e-06, + "loss": 0.4962, + "step": 1435 + }, + { + "epoch": 0.6675964667596467, + "grad_norm": 8.109962463378906, + "learning_rate": 9.998564999856501e-06, + "loss": 0.9962, + "step": 1436 + }, + { + "epoch": 0.6680613668061367, + "grad_norm": 8.565103530883789, + "learning_rate": 9.998563999856402e-06, + "loss": 1.5004, + "step": 1437 + }, + { + "epoch": 0.6685262668526267, + "grad_norm": 7.046835899353027, + "learning_rate": 9.9985629998563e-06, + "loss": 0.5261, + "step": 1438 + }, + { + "epoch": 0.6689911668991166, + "grad_norm": 7.00200891494751, + "learning_rate": 9.998561999856201e-06, + "loss": 0.6849, + "step": 1439 + }, + { + "epoch": 0.6694560669456067, + "grad_norm": 12.187256813049316, + "learning_rate": 9.9985609998561e-06, + "loss": 0.6207, + "step": 1440 + }, + { + "epoch": 0.6699209669920967, + "grad_norm": 7.086055755615234, + "learning_rate": 9.998559999856001e-06, + "loss": 0.5676, + "step": 1441 + }, + { + "epoch": 0.6703858670385867, + "grad_norm": 11.299099922180176, + "learning_rate": 9.9985589998559e-06, + "loss": 0.6952, + "step": 1442 + }, + { + "epoch": 0.6708507670850767, + "grad_norm": 6.74464225769043, + "learning_rate": 9.9985579998558e-06, + "loss": 0.6572, + "step": 1443 + }, + { + "epoch": 0.6713156671315668, + "grad_norm": 6.789754867553711, + "learning_rate": 9.998556999855701e-06, + "loss": 0.8385, + "step": 1444 + }, + { + "epoch": 0.6717805671780567, + "grad_norm": 7.989667892456055, + "learning_rate": 9.9985559998556e-06, + "loss": 0.5938, + "step": 1445 + }, + { + "epoch": 0.6722454672245467, + "grad_norm": 6.487307071685791, + "learning_rate": 9.998554999855501e-06, + "loss": 0.6674, + "step": 1446 + }, + { + "epoch": 0.6727103672710367, + "grad_norm": 8.224594116210938, + "learning_rate": 9.9985539998554e-06, + "loss": 1.0405, + "step": 1447 + }, + { + "epoch": 0.6731752673175267, + "grad_norm": 6.343496322631836, + "learning_rate": 9.9985529998553e-06, + "loss": 0.7159, + "step": 1448 + }, + { + "epoch": 0.6736401673640168, + "grad_norm": 5.695250988006592, + "learning_rate": 9.998551999855201e-06, + "loss": 0.6155, + "step": 1449 + }, + { + "epoch": 0.6741050674105067, + "grad_norm": 9.216282844543457, + "learning_rate": 9.998550999855102e-06, + "loss": 0.6169, + "step": 1450 + }, + { + "epoch": 0.6745699674569967, + "grad_norm": 7.762765407562256, + "learning_rate": 9.998549999855e-06, + "loss": 0.6577, + "step": 1451 + }, + { + "epoch": 0.6750348675034867, + "grad_norm": 6.0613226890563965, + "learning_rate": 9.998548999854902e-06, + "loss": 0.7735, + "step": 1452 + }, + { + "epoch": 0.6754997675499768, + "grad_norm": 7.048237323760986, + "learning_rate": 9.9985479998548e-06, + "loss": 0.6416, + "step": 1453 + }, + { + "epoch": 0.6759646675964668, + "grad_norm": 9.686210632324219, + "learning_rate": 9.998546999854701e-06, + "loss": 0.9014, + "step": 1454 + }, + { + "epoch": 0.6764295676429568, + "grad_norm": 4.761448383331299, + "learning_rate": 9.9985459998546e-06, + "loss": 0.3807, + "step": 1455 + }, + { + "epoch": 0.6768944676894467, + "grad_norm": 7.560929775238037, + "learning_rate": 9.9985449998545e-06, + "loss": 0.6067, + "step": 1456 + }, + { + "epoch": 0.6773593677359367, + "grad_norm": 12.798995971679688, + "learning_rate": 9.998543999854401e-06, + "loss": 1.3074, + "step": 1457 + }, + { + "epoch": 0.6778242677824268, + "grad_norm": 7.084683895111084, + "learning_rate": 9.9985429998543e-06, + "loss": 0.5563, + "step": 1458 + }, + { + "epoch": 0.6782891678289168, + "grad_norm": 7.520930290222168, + "learning_rate": 9.998541999854201e-06, + "loss": 0.6117, + "step": 1459 + }, + { + "epoch": 0.6787540678754068, + "grad_norm": 8.478391647338867, + "learning_rate": 9.9985409998541e-06, + "loss": 1.2262, + "step": 1460 + }, + { + "epoch": 0.6792189679218968, + "grad_norm": 11.823731422424316, + "learning_rate": 9.998539999854e-06, + "loss": 1.0672, + "step": 1461 + }, + { + "epoch": 0.6796838679683868, + "grad_norm": 6.9932756423950195, + "learning_rate": 9.998538999853901e-06, + "loss": 0.6895, + "step": 1462 + }, + { + "epoch": 0.6801487680148768, + "grad_norm": 8.66246509552002, + "learning_rate": 9.998537999853802e-06, + "loss": 0.8485, + "step": 1463 + }, + { + "epoch": 0.6806136680613668, + "grad_norm": 9.082144737243652, + "learning_rate": 9.998536999853701e-06, + "loss": 1.0162, + "step": 1464 + }, + { + "epoch": 0.6810785681078568, + "grad_norm": 10.550289154052734, + "learning_rate": 9.998535999853602e-06, + "loss": 1.2757, + "step": 1465 + }, + { + "epoch": 0.6815434681543469, + "grad_norm": 7.006915092468262, + "learning_rate": 9.9985349998535e-06, + "loss": 0.5954, + "step": 1466 + }, + { + "epoch": 0.6820083682008368, + "grad_norm": 7.203117847442627, + "learning_rate": 9.9985339998534e-06, + "loss": 0.952, + "step": 1467 + }, + { + "epoch": 0.6824732682473268, + "grad_norm": 4.692962646484375, + "learning_rate": 9.9985329998533e-06, + "loss": 0.2669, + "step": 1468 + }, + { + "epoch": 0.6829381682938168, + "grad_norm": 6.058074951171875, + "learning_rate": 9.9985319998532e-06, + "loss": 0.5362, + "step": 1469 + }, + { + "epoch": 0.6834030683403068, + "grad_norm": 7.698408603668213, + "learning_rate": 9.998530999853101e-06, + "loss": 0.7043, + "step": 1470 + }, + { + "epoch": 0.6838679683867969, + "grad_norm": 8.860274314880371, + "learning_rate": 9.998529999853e-06, + "loss": 0.7877, + "step": 1471 + }, + { + "epoch": 0.6843328684332869, + "grad_norm": 7.542948246002197, + "learning_rate": 9.998528999852901e-06, + "loss": 0.5328, + "step": 1472 + }, + { + "epoch": 0.6847977684797768, + "grad_norm": 6.659553527832031, + "learning_rate": 9.9985279998528e-06, + "loss": 0.8648, + "step": 1473 + }, + { + "epoch": 0.6852626685262668, + "grad_norm": 8.960122108459473, + "learning_rate": 9.9985269998527e-06, + "loss": 0.7811, + "step": 1474 + }, + { + "epoch": 0.6857275685727569, + "grad_norm": 8.812795639038086, + "learning_rate": 9.998525999852601e-06, + "loss": 1.0936, + "step": 1475 + }, + { + "epoch": 0.6861924686192469, + "grad_norm": 7.670197486877441, + "learning_rate": 9.998524999852502e-06, + "loss": 0.7769, + "step": 1476 + }, + { + "epoch": 0.6866573686657369, + "grad_norm": 6.505554676055908, + "learning_rate": 9.998523999852401e-06, + "loss": 0.572, + "step": 1477 + }, + { + "epoch": 0.6871222687122269, + "grad_norm": 9.232853889465332, + "learning_rate": 9.9985229998523e-06, + "loss": 1.1119, + "step": 1478 + }, + { + "epoch": 0.6875871687587168, + "grad_norm": 9.978343963623047, + "learning_rate": 9.9985219998522e-06, + "loss": 0.7873, + "step": 1479 + }, + { + "epoch": 0.6880520688052069, + "grad_norm": 6.910945415496826, + "learning_rate": 9.998520999852101e-06, + "loss": 0.9856, + "step": 1480 + }, + { + "epoch": 0.6885169688516969, + "grad_norm": 7.287930011749268, + "learning_rate": 9.998519999852e-06, + "loss": 0.7585, + "step": 1481 + }, + { + "epoch": 0.6889818688981869, + "grad_norm": 7.031306743621826, + "learning_rate": 9.998518999851901e-06, + "loss": 0.523, + "step": 1482 + }, + { + "epoch": 0.6894467689446769, + "grad_norm": 5.139354228973389, + "learning_rate": 9.998517999851802e-06, + "loss": 0.4449, + "step": 1483 + }, + { + "epoch": 0.6899116689911668, + "grad_norm": 5.298123359680176, + "learning_rate": 9.9985169998517e-06, + "loss": 0.4604, + "step": 1484 + }, + { + "epoch": 0.6903765690376569, + "grad_norm": 7.401618957519531, + "learning_rate": 9.998515999851601e-06, + "loss": 0.7459, + "step": 1485 + }, + { + "epoch": 0.6908414690841469, + "grad_norm": 5.316839218139648, + "learning_rate": 9.9985149998515e-06, + "loss": 0.3559, + "step": 1486 + }, + { + "epoch": 0.6913063691306369, + "grad_norm": 7.539909839630127, + "learning_rate": 9.9985139998514e-06, + "loss": 0.5278, + "step": 1487 + }, + { + "epoch": 0.691771269177127, + "grad_norm": 6.137341499328613, + "learning_rate": 9.998512999851301e-06, + "loss": 0.4433, + "step": 1488 + }, + { + "epoch": 0.692236169223617, + "grad_norm": 5.681013107299805, + "learning_rate": 9.9985119998512e-06, + "loss": 0.5665, + "step": 1489 + }, + { + "epoch": 0.6927010692701069, + "grad_norm": 5.849409103393555, + "learning_rate": 9.998510999851101e-06, + "loss": 0.5087, + "step": 1490 + }, + { + "epoch": 0.6931659693165969, + "grad_norm": 14.251544952392578, + "learning_rate": 9.998509999851e-06, + "loss": 1.3124, + "step": 1491 + }, + { + "epoch": 0.6936308693630869, + "grad_norm": 5.877683639526367, + "learning_rate": 9.9985089998509e-06, + "loss": 0.4908, + "step": 1492 + }, + { + "epoch": 0.694095769409577, + "grad_norm": 7.8239874839782715, + "learning_rate": 9.998507999850801e-06, + "loss": 1.2002, + "step": 1493 + }, + { + "epoch": 0.694560669456067, + "grad_norm": 5.932570934295654, + "learning_rate": 9.9985069998507e-06, + "loss": 0.6003, + "step": 1494 + }, + { + "epoch": 0.695025569502557, + "grad_norm": 4.372016906738281, + "learning_rate": 9.998505999850601e-06, + "loss": 0.3586, + "step": 1495 + }, + { + "epoch": 0.6954904695490469, + "grad_norm": 8.993486404418945, + "learning_rate": 9.998504999850502e-06, + "loss": 0.6613, + "step": 1496 + }, + { + "epoch": 0.6959553695955369, + "grad_norm": 7.73392391204834, + "learning_rate": 9.9985039998504e-06, + "loss": 0.5772, + "step": 1497 + }, + { + "epoch": 0.696420269642027, + "grad_norm": 7.268515110015869, + "learning_rate": 9.9985029998503e-06, + "loss": 0.5273, + "step": 1498 + }, + { + "epoch": 0.696885169688517, + "grad_norm": 6.379449367523193, + "learning_rate": 9.9985019998502e-06, + "loss": 0.6213, + "step": 1499 + }, + { + "epoch": 0.697350069735007, + "grad_norm": 6.79326868057251, + "learning_rate": 9.998500999850101e-06, + "loss": 0.3969, + "step": 1500 + }, + { + "epoch": 0.6978149697814969, + "grad_norm": 7.443460941314697, + "learning_rate": 9.998499999850002e-06, + "loss": 0.7099, + "step": 1501 + }, + { + "epoch": 0.698279869827987, + "grad_norm": 9.08292007446289, + "learning_rate": 9.9984989998499e-06, + "loss": 0.8672, + "step": 1502 + }, + { + "epoch": 0.698744769874477, + "grad_norm": 6.722146034240723, + "learning_rate": 9.998497999849801e-06, + "loss": 0.8813, + "step": 1503 + }, + { + "epoch": 0.699209669920967, + "grad_norm": 5.9666314125061035, + "learning_rate": 9.9984969998497e-06, + "loss": 0.5112, + "step": 1504 + }, + { + "epoch": 0.699674569967457, + "grad_norm": 4.684502124786377, + "learning_rate": 9.9984959998496e-06, + "loss": 0.463, + "step": 1505 + }, + { + "epoch": 0.700139470013947, + "grad_norm": 7.14704704284668, + "learning_rate": 9.998494999849501e-06, + "loss": 0.7228, + "step": 1506 + }, + { + "epoch": 0.700604370060437, + "grad_norm": 5.571743488311768, + "learning_rate": 9.998493999849402e-06, + "loss": 0.5258, + "step": 1507 + }, + { + "epoch": 0.701069270106927, + "grad_norm": 5.839298725128174, + "learning_rate": 9.998492999849301e-06, + "loss": 0.8668, + "step": 1508 + }, + { + "epoch": 0.701534170153417, + "grad_norm": 7.851373195648193, + "learning_rate": 9.9984919998492e-06, + "loss": 0.9257, + "step": 1509 + }, + { + "epoch": 0.701999070199907, + "grad_norm": 7.202126979827881, + "learning_rate": 9.9984909998491e-06, + "loss": 0.6092, + "step": 1510 + }, + { + "epoch": 0.7024639702463971, + "grad_norm": 5.9331512451171875, + "learning_rate": 9.998489999849e-06, + "loss": 0.4156, + "step": 1511 + }, + { + "epoch": 0.702928870292887, + "grad_norm": 6.843191623687744, + "learning_rate": 9.9984889998489e-06, + "loss": 0.6062, + "step": 1512 + }, + { + "epoch": 0.703393770339377, + "grad_norm": 5.297097682952881, + "learning_rate": 9.998487999848801e-06, + "loss": 0.3166, + "step": 1513 + }, + { + "epoch": 0.703858670385867, + "grad_norm": 7.3613152503967285, + "learning_rate": 9.998486999848702e-06, + "loss": 0.5987, + "step": 1514 + }, + { + "epoch": 0.704323570432357, + "grad_norm": 8.264558792114258, + "learning_rate": 9.9984859998486e-06, + "loss": 0.6728, + "step": 1515 + }, + { + "epoch": 0.7047884704788471, + "grad_norm": 6.760379791259766, + "learning_rate": 9.998484999848501e-06, + "loss": 0.72, + "step": 1516 + }, + { + "epoch": 0.7052533705253371, + "grad_norm": 8.472474098205566, + "learning_rate": 9.9984839998484e-06, + "loss": 0.7506, + "step": 1517 + }, + { + "epoch": 0.705718270571827, + "grad_norm": 7.920993328094482, + "learning_rate": 9.998482999848301e-06, + "loss": 0.5991, + "step": 1518 + }, + { + "epoch": 0.706183170618317, + "grad_norm": 7.889115810394287, + "learning_rate": 9.998481999848202e-06, + "loss": 0.7916, + "step": 1519 + }, + { + "epoch": 0.7066480706648071, + "grad_norm": 6.877645015716553, + "learning_rate": 9.9984809998481e-06, + "loss": 0.773, + "step": 1520 + }, + { + "epoch": 0.7071129707112971, + "grad_norm": 6.183680534362793, + "learning_rate": 9.998479999848001e-06, + "loss": 0.6939, + "step": 1521 + }, + { + "epoch": 0.7075778707577871, + "grad_norm": 6.600738525390625, + "learning_rate": 9.9984789998479e-06, + "loss": 0.7418, + "step": 1522 + }, + { + "epoch": 0.7080427708042771, + "grad_norm": 5.261500835418701, + "learning_rate": 9.9984779998478e-06, + "loss": 0.4076, + "step": 1523 + }, + { + "epoch": 0.708507670850767, + "grad_norm": 12.29023551940918, + "learning_rate": 9.9984769998477e-06, + "loss": 1.0139, + "step": 1524 + }, + { + "epoch": 0.7089725708972571, + "grad_norm": 7.8997673988342285, + "learning_rate": 9.9984759998476e-06, + "loss": 0.5327, + "step": 1525 + }, + { + "epoch": 0.7094374709437471, + "grad_norm": 10.107629776000977, + "learning_rate": 9.998474999847501e-06, + "loss": 1.5078, + "step": 1526 + }, + { + "epoch": 0.7099023709902371, + "grad_norm": 4.108511447906494, + "learning_rate": 9.998473999847402e-06, + "loss": 0.3518, + "step": 1527 + }, + { + "epoch": 0.7103672710367271, + "grad_norm": 6.493999481201172, + "learning_rate": 9.9984729998473e-06, + "loss": 0.7192, + "step": 1528 + }, + { + "epoch": 0.7108321710832171, + "grad_norm": 5.969768047332764, + "learning_rate": 9.998471999847201e-06, + "loss": 0.5928, + "step": 1529 + }, + { + "epoch": 0.7112970711297071, + "grad_norm": 8.836640357971191, + "learning_rate": 9.9984709998471e-06, + "loss": 0.7596, + "step": 1530 + }, + { + "epoch": 0.7117619711761971, + "grad_norm": 8.287631034851074, + "learning_rate": 9.998469999847001e-06, + "loss": 0.9075, + "step": 1531 + }, + { + "epoch": 0.7122268712226871, + "grad_norm": 11.08590030670166, + "learning_rate": 9.998468999846902e-06, + "loss": 0.8935, + "step": 1532 + }, + { + "epoch": 0.7126917712691772, + "grad_norm": 4.978476524353027, + "learning_rate": 9.9984679998468e-06, + "loss": 0.5287, + "step": 1533 + }, + { + "epoch": 0.7131566713156672, + "grad_norm": 6.880387306213379, + "learning_rate": 9.998466999846701e-06, + "loss": 0.7125, + "step": 1534 + }, + { + "epoch": 0.7136215713621571, + "grad_norm": 7.89290714263916, + "learning_rate": 9.9984659998466e-06, + "loss": 0.6826, + "step": 1535 + }, + { + "epoch": 0.7140864714086471, + "grad_norm": 11.400527000427246, + "learning_rate": 9.998464999846501e-06, + "loss": 0.4848, + "step": 1536 + }, + { + "epoch": 0.7145513714551371, + "grad_norm": 5.6953864097595215, + "learning_rate": 9.9984639998464e-06, + "loss": 0.5002, + "step": 1537 + }, + { + "epoch": 0.7150162715016272, + "grad_norm": 11.20948600769043, + "learning_rate": 9.9984629998463e-06, + "loss": 0.8593, + "step": 1538 + }, + { + "epoch": 0.7154811715481172, + "grad_norm": 10.47797679901123, + "learning_rate": 9.998461999846201e-06, + "loss": 1.0394, + "step": 1539 + }, + { + "epoch": 0.7159460715946072, + "grad_norm": 6.341732501983643, + "learning_rate": 9.998460999846102e-06, + "loss": 0.7239, + "step": 1540 + }, + { + "epoch": 0.7164109716410971, + "grad_norm": 8.772225379943848, + "learning_rate": 9.998459999846e-06, + "loss": 1.0696, + "step": 1541 + }, + { + "epoch": 0.7168758716875872, + "grad_norm": 6.752238750457764, + "learning_rate": 9.9984589998459e-06, + "loss": 0.8437, + "step": 1542 + }, + { + "epoch": 0.7173407717340772, + "grad_norm": 8.406490325927734, + "learning_rate": 9.9984579998458e-06, + "loss": 0.5185, + "step": 1543 + }, + { + "epoch": 0.7178056717805672, + "grad_norm": 6.447621822357178, + "learning_rate": 9.998456999845701e-06, + "loss": 0.5569, + "step": 1544 + }, + { + "epoch": 0.7182705718270572, + "grad_norm": 6.471087455749512, + "learning_rate": 9.998455999845602e-06, + "loss": 0.485, + "step": 1545 + }, + { + "epoch": 0.7187354718735471, + "grad_norm": 6.579780578613281, + "learning_rate": 9.9984549998455e-06, + "loss": 0.573, + "step": 1546 + }, + { + "epoch": 0.7192003719200372, + "grad_norm": 7.466217041015625, + "learning_rate": 9.998453999845401e-06, + "loss": 0.7356, + "step": 1547 + }, + { + "epoch": 0.7196652719665272, + "grad_norm": 5.970634460449219, + "learning_rate": 9.9984529998453e-06, + "loss": 0.5761, + "step": 1548 + }, + { + "epoch": 0.7201301720130172, + "grad_norm": 5.740000247955322, + "learning_rate": 9.998451999845201e-06, + "loss": 0.4384, + "step": 1549 + }, + { + "epoch": 0.7205950720595072, + "grad_norm": 8.785966873168945, + "learning_rate": 9.9984509998451e-06, + "loss": 1.1286, + "step": 1550 + }, + { + "epoch": 0.7210599721059973, + "grad_norm": 9.399487495422363, + "learning_rate": 9.998449999845e-06, + "loss": 1.0909, + "step": 1551 + }, + { + "epoch": 0.7215248721524872, + "grad_norm": 7.337953567504883, + "learning_rate": 9.998448999844901e-06, + "loss": 0.2996, + "step": 1552 + }, + { + "epoch": 0.7219897721989772, + "grad_norm": 6.911343574523926, + "learning_rate": 9.9984479998448e-06, + "loss": 0.8681, + "step": 1553 + }, + { + "epoch": 0.7224546722454672, + "grad_norm": 5.642986297607422, + "learning_rate": 9.9984469998447e-06, + "loss": 0.5574, + "step": 1554 + }, + { + "epoch": 0.7229195722919572, + "grad_norm": 9.419486045837402, + "learning_rate": 9.9984459998446e-06, + "loss": 0.778, + "step": 1555 + }, + { + "epoch": 0.7233844723384473, + "grad_norm": 6.331673622131348, + "learning_rate": 9.9984449998445e-06, + "loss": 0.4681, + "step": 1556 + }, + { + "epoch": 0.7238493723849372, + "grad_norm": 6.472997665405273, + "learning_rate": 9.998443999844401e-06, + "loss": 0.6478, + "step": 1557 + }, + { + "epoch": 0.7243142724314272, + "grad_norm": 7.285573959350586, + "learning_rate": 9.998442999844302e-06, + "loss": 0.9404, + "step": 1558 + }, + { + "epoch": 0.7247791724779172, + "grad_norm": 5.827949523925781, + "learning_rate": 9.9984419998442e-06, + "loss": 0.4592, + "step": 1559 + }, + { + "epoch": 0.7252440725244073, + "grad_norm": 7.357102394104004, + "learning_rate": 9.998440999844101e-06, + "loss": 0.4166, + "step": 1560 + }, + { + "epoch": 0.7257089725708973, + "grad_norm": 6.029745101928711, + "learning_rate": 9.998439999844e-06, + "loss": 0.4367, + "step": 1561 + }, + { + "epoch": 0.7261738726173873, + "grad_norm": 7.287923812866211, + "learning_rate": 9.998438999843901e-06, + "loss": 1.036, + "step": 1562 + }, + { + "epoch": 0.7266387726638772, + "grad_norm": 6.682716369628906, + "learning_rate": 9.9984379998438e-06, + "loss": 0.5103, + "step": 1563 + }, + { + "epoch": 0.7271036727103672, + "grad_norm": 10.391717910766602, + "learning_rate": 9.9984369998437e-06, + "loss": 0.7424, + "step": 1564 + }, + { + "epoch": 0.7275685727568573, + "grad_norm": 7.527161598205566, + "learning_rate": 9.998435999843601e-06, + "loss": 0.6807, + "step": 1565 + }, + { + "epoch": 0.7280334728033473, + "grad_norm": 7.777228355407715, + "learning_rate": 9.9984349998435e-06, + "loss": 0.8872, + "step": 1566 + }, + { + "epoch": 0.7284983728498373, + "grad_norm": 7.197262763977051, + "learning_rate": 9.998433999843401e-06, + "loss": 0.97, + "step": 1567 + }, + { + "epoch": 0.7289632728963273, + "grad_norm": 8.546728134155273, + "learning_rate": 9.9984329998433e-06, + "loss": 0.8155, + "step": 1568 + }, + { + "epoch": 0.7294281729428173, + "grad_norm": 6.705996990203857, + "learning_rate": 9.9984319998432e-06, + "loss": 0.4201, + "step": 1569 + }, + { + "epoch": 0.7298930729893073, + "grad_norm": 6.315218925476074, + "learning_rate": 9.998430999843101e-06, + "loss": 0.5687, + "step": 1570 + }, + { + "epoch": 0.7303579730357973, + "grad_norm": 7.848804473876953, + "learning_rate": 9.998429999843002e-06, + "loss": 0.507, + "step": 1571 + }, + { + "epoch": 0.7308228730822873, + "grad_norm": 6.893216133117676, + "learning_rate": 9.9984289998429e-06, + "loss": 0.876, + "step": 1572 + }, + { + "epoch": 0.7312877731287774, + "grad_norm": 7.644804000854492, + "learning_rate": 9.998427999842801e-06, + "loss": 0.5431, + "step": 1573 + }, + { + "epoch": 0.7317526731752673, + "grad_norm": 7.929372310638428, + "learning_rate": 9.9984269998427e-06, + "loss": 0.7191, + "step": 1574 + }, + { + "epoch": 0.7322175732217573, + "grad_norm": 8.645465850830078, + "learning_rate": 9.998425999842601e-06, + "loss": 0.6059, + "step": 1575 + }, + { + "epoch": 0.7326824732682473, + "grad_norm": 6.039076328277588, + "learning_rate": 9.998424999842502e-06, + "loss": 0.2675, + "step": 1576 + }, + { + "epoch": 0.7331473733147373, + "grad_norm": 4.969431400299072, + "learning_rate": 9.9984239998424e-06, + "loss": 0.443, + "step": 1577 + }, + { + "epoch": 0.7336122733612274, + "grad_norm": 10.162643432617188, + "learning_rate": 9.998422999842301e-06, + "loss": 0.4591, + "step": 1578 + }, + { + "epoch": 0.7340771734077174, + "grad_norm": 10.379923820495605, + "learning_rate": 9.9984219998422e-06, + "loss": 0.7472, + "step": 1579 + }, + { + "epoch": 0.7345420734542073, + "grad_norm": 7.531208515167236, + "learning_rate": 9.998420999842101e-06, + "loss": 0.8199, + "step": 1580 + }, + { + "epoch": 0.7350069735006973, + "grad_norm": 8.29791259765625, + "learning_rate": 9.998419999842e-06, + "loss": 0.8903, + "step": 1581 + }, + { + "epoch": 0.7354718735471873, + "grad_norm": 5.357228755950928, + "learning_rate": 9.9984189998419e-06, + "loss": 0.3552, + "step": 1582 + }, + { + "epoch": 0.7359367735936774, + "grad_norm": 7.788773059844971, + "learning_rate": 9.998417999841801e-06, + "loss": 0.7765, + "step": 1583 + }, + { + "epoch": 0.7364016736401674, + "grad_norm": 5.469699859619141, + "learning_rate": 9.998416999841702e-06, + "loss": 0.5632, + "step": 1584 + }, + { + "epoch": 0.7368665736866574, + "grad_norm": 5.900763511657715, + "learning_rate": 9.998415999841601e-06, + "loss": 0.4877, + "step": 1585 + }, + { + "epoch": 0.7373314737331473, + "grad_norm": 7.7364325523376465, + "learning_rate": 9.9984149998415e-06, + "loss": 0.5335, + "step": 1586 + }, + { + "epoch": 0.7377963737796374, + "grad_norm": 6.858334541320801, + "learning_rate": 9.9984139998414e-06, + "loss": 0.8478, + "step": 1587 + }, + { + "epoch": 0.7382612738261274, + "grad_norm": 8.623242378234863, + "learning_rate": 9.998412999841301e-06, + "loss": 0.701, + "step": 1588 + }, + { + "epoch": 0.7387261738726174, + "grad_norm": 5.280101776123047, + "learning_rate": 9.998411999841202e-06, + "loss": 0.4924, + "step": 1589 + }, + { + "epoch": 0.7391910739191074, + "grad_norm": 5.789385795593262, + "learning_rate": 9.9984109998411e-06, + "loss": 0.59, + "step": 1590 + }, + { + "epoch": 0.7396559739655973, + "grad_norm": 4.961820125579834, + "learning_rate": 9.998409999841001e-06, + "loss": 0.2873, + "step": 1591 + }, + { + "epoch": 0.7401208740120874, + "grad_norm": 7.2772979736328125, + "learning_rate": 9.9984089998409e-06, + "loss": 0.5057, + "step": 1592 + }, + { + "epoch": 0.7405857740585774, + "grad_norm": 6.567513942718506, + "learning_rate": 9.998407999840801e-06, + "loss": 0.6206, + "step": 1593 + }, + { + "epoch": 0.7410506741050674, + "grad_norm": 8.95686149597168, + "learning_rate": 9.9984069998407e-06, + "loss": 0.5221, + "step": 1594 + }, + { + "epoch": 0.7415155741515574, + "grad_norm": 4.556466102600098, + "learning_rate": 9.9984059998406e-06, + "loss": 0.2746, + "step": 1595 + }, + { + "epoch": 0.7419804741980475, + "grad_norm": 12.926765441894531, + "learning_rate": 9.998404999840501e-06, + "loss": 0.9526, + "step": 1596 + }, + { + "epoch": 0.7424453742445374, + "grad_norm": 8.901631355285645, + "learning_rate": 9.9984039998404e-06, + "loss": 0.3299, + "step": 1597 + }, + { + "epoch": 0.7429102742910274, + "grad_norm": 7.630202770233154, + "learning_rate": 9.998402999840301e-06, + "loss": 0.7641, + "step": 1598 + }, + { + "epoch": 0.7433751743375174, + "grad_norm": 5.945196151733398, + "learning_rate": 9.9984019998402e-06, + "loss": 0.6691, + "step": 1599 + }, + { + "epoch": 0.7438400743840075, + "grad_norm": 6.815633296966553, + "learning_rate": 9.9984009998401e-06, + "loss": 0.5469, + "step": 1600 + }, + { + "epoch": 0.7443049744304975, + "grad_norm": 7.843914031982422, + "learning_rate": 9.998399999840001e-06, + "loss": 0.7185, + "step": 1601 + }, + { + "epoch": 0.7447698744769874, + "grad_norm": 6.745875835418701, + "learning_rate": 9.998398999839902e-06, + "loss": 0.5109, + "step": 1602 + }, + { + "epoch": 0.7452347745234774, + "grad_norm": 7.282474994659424, + "learning_rate": 9.998397999839801e-06, + "loss": 0.7961, + "step": 1603 + }, + { + "epoch": 0.7456996745699674, + "grad_norm": 13.131388664245605, + "learning_rate": 9.998396999839702e-06, + "loss": 0.771, + "step": 1604 + }, + { + "epoch": 0.7461645746164575, + "grad_norm": 11.053718566894531, + "learning_rate": 9.9983959998396e-06, + "loss": 1.2378, + "step": 1605 + }, + { + "epoch": 0.7466294746629475, + "grad_norm": 8.99249267578125, + "learning_rate": 9.998394999839501e-06, + "loss": 0.5726, + "step": 1606 + }, + { + "epoch": 0.7470943747094375, + "grad_norm": 6.141333103179932, + "learning_rate": 9.9983939998394e-06, + "loss": 0.4953, + "step": 1607 + }, + { + "epoch": 0.7475592747559274, + "grad_norm": 5.819798469543457, + "learning_rate": 9.9983929998393e-06, + "loss": 0.6044, + "step": 1608 + }, + { + "epoch": 0.7480241748024175, + "grad_norm": 9.4017915725708, + "learning_rate": 9.998391999839201e-06, + "loss": 0.7458, + "step": 1609 + }, + { + "epoch": 0.7484890748489075, + "grad_norm": 8.094745635986328, + "learning_rate": 9.9983909998391e-06, + "loss": 1.149, + "step": 1610 + }, + { + "epoch": 0.7489539748953975, + "grad_norm": 5.543362617492676, + "learning_rate": 9.998389999839001e-06, + "loss": 0.2483, + "step": 1611 + }, + { + "epoch": 0.7494188749418875, + "grad_norm": 7.706406116485596, + "learning_rate": 9.9983889998389e-06, + "loss": 0.7138, + "step": 1612 + }, + { + "epoch": 0.7498837749883775, + "grad_norm": 7.952761173248291, + "learning_rate": 9.9983879998388e-06, + "loss": 0.7832, + "step": 1613 + }, + { + "epoch": 0.7503486750348675, + "grad_norm": 9.29058837890625, + "learning_rate": 9.998386999838701e-06, + "loss": 0.596, + "step": 1614 + }, + { + "epoch": 0.7508135750813575, + "grad_norm": 8.9371976852417, + "learning_rate": 9.998385999838602e-06, + "loss": 1.4383, + "step": 1615 + }, + { + "epoch": 0.7512784751278475, + "grad_norm": 6.009920597076416, + "learning_rate": 9.998384999838501e-06, + "loss": 0.6004, + "step": 1616 + }, + { + "epoch": 0.7517433751743375, + "grad_norm": 6.175505638122559, + "learning_rate": 9.998383999838402e-06, + "loss": 0.7071, + "step": 1617 + }, + { + "epoch": 0.7522082752208276, + "grad_norm": 6.100338935852051, + "learning_rate": 9.9983829998383e-06, + "loss": 0.4478, + "step": 1618 + }, + { + "epoch": 0.7526731752673175, + "grad_norm": 7.368991851806641, + "learning_rate": 9.9983819998382e-06, + "loss": 0.7377, + "step": 1619 + }, + { + "epoch": 0.7531380753138075, + "grad_norm": 5.966150283813477, + "learning_rate": 9.9983809998381e-06, + "loss": 0.5136, + "step": 1620 + }, + { + "epoch": 0.7536029753602975, + "grad_norm": 5.961118221282959, + "learning_rate": 9.998379999838001e-06, + "loss": 0.687, + "step": 1621 + }, + { + "epoch": 0.7540678754067875, + "grad_norm": 6.021643161773682, + "learning_rate": 9.998378999837902e-06, + "loss": 0.3651, + "step": 1622 + }, + { + "epoch": 0.7545327754532776, + "grad_norm": 5.33844518661499, + "learning_rate": 9.9983779998378e-06, + "loss": 0.5003, + "step": 1623 + }, + { + "epoch": 0.7549976754997676, + "grad_norm": 7.901539325714111, + "learning_rate": 9.998376999837701e-06, + "loss": 0.5274, + "step": 1624 + }, + { + "epoch": 0.7554625755462575, + "grad_norm": 7.537463665008545, + "learning_rate": 9.9983759998376e-06, + "loss": 0.9298, + "step": 1625 + }, + { + "epoch": 0.7559274755927475, + "grad_norm": 10.160776138305664, + "learning_rate": 9.9983749998375e-06, + "loss": 0.7493, + "step": 1626 + }, + { + "epoch": 0.7563923756392376, + "grad_norm": 6.955471038818359, + "learning_rate": 9.998373999837401e-06, + "loss": 0.7566, + "step": 1627 + }, + { + "epoch": 0.7568572756857276, + "grad_norm": 8.492168426513672, + "learning_rate": 9.9983729998373e-06, + "loss": 0.8878, + "step": 1628 + }, + { + "epoch": 0.7573221757322176, + "grad_norm": 4.876123905181885, + "learning_rate": 9.998371999837201e-06, + "loss": 0.3731, + "step": 1629 + }, + { + "epoch": 0.7577870757787076, + "grad_norm": 12.625633239746094, + "learning_rate": 9.9983709998371e-06, + "loss": 1.1688, + "step": 1630 + }, + { + "epoch": 0.7582519758251975, + "grad_norm": 5.402047634124756, + "learning_rate": 9.998369999837e-06, + "loss": 0.6087, + "step": 1631 + }, + { + "epoch": 0.7587168758716876, + "grad_norm": 6.270816802978516, + "learning_rate": 9.9983689998369e-06, + "loss": 0.5831, + "step": 1632 + }, + { + "epoch": 0.7591817759181776, + "grad_norm": 6.3493146896362305, + "learning_rate": 9.9983679998368e-06, + "loss": 0.5499, + "step": 1633 + }, + { + "epoch": 0.7596466759646676, + "grad_norm": 7.0903120040893555, + "learning_rate": 9.998366999836701e-06, + "loss": 0.5481, + "step": 1634 + }, + { + "epoch": 0.7601115760111576, + "grad_norm": 8.29167652130127, + "learning_rate": 9.998365999836602e-06, + "loss": 1.1382, + "step": 1635 + }, + { + "epoch": 0.7605764760576476, + "grad_norm": 5.431929588317871, + "learning_rate": 9.9983649998365e-06, + "loss": 0.4332, + "step": 1636 + }, + { + "epoch": 0.7610413761041376, + "grad_norm": 5.993124485015869, + "learning_rate": 9.998363999836401e-06, + "loss": 0.7115, + "step": 1637 + }, + { + "epoch": 0.7615062761506276, + "grad_norm": 6.174988269805908, + "learning_rate": 9.9983629998363e-06, + "loss": 0.7751, + "step": 1638 + }, + { + "epoch": 0.7619711761971176, + "grad_norm": 9.603148460388184, + "learning_rate": 9.9983619998362e-06, + "loss": 1.0618, + "step": 1639 + }, + { + "epoch": 0.7624360762436077, + "grad_norm": 6.411449432373047, + "learning_rate": 9.998360999836101e-06, + "loss": 0.6142, + "step": 1640 + }, + { + "epoch": 0.7629009762900977, + "grad_norm": 6.596259593963623, + "learning_rate": 9.998359999836e-06, + "loss": 0.6687, + "step": 1641 + }, + { + "epoch": 0.7633658763365876, + "grad_norm": 5.171804428100586, + "learning_rate": 9.998358999835901e-06, + "loss": 0.5431, + "step": 1642 + }, + { + "epoch": 0.7638307763830776, + "grad_norm": 8.291741371154785, + "learning_rate": 9.9983579998358e-06, + "loss": 0.6289, + "step": 1643 + }, + { + "epoch": 0.7642956764295676, + "grad_norm": 7.272487163543701, + "learning_rate": 9.9983569998357e-06, + "loss": 0.484, + "step": 1644 + }, + { + "epoch": 0.7647605764760577, + "grad_norm": 8.07724666595459, + "learning_rate": 9.998355999835601e-06, + "loss": 0.6619, + "step": 1645 + }, + { + "epoch": 0.7652254765225477, + "grad_norm": 5.526209831237793, + "learning_rate": 9.9983549998355e-06, + "loss": 0.5867, + "step": 1646 + }, + { + "epoch": 0.7656903765690377, + "grad_norm": 5.3393874168396, + "learning_rate": 9.998353999835401e-06, + "loss": 0.5433, + "step": 1647 + }, + { + "epoch": 0.7661552766155276, + "grad_norm": 5.577486038208008, + "learning_rate": 9.998352999835302e-06, + "loss": 0.4217, + "step": 1648 + }, + { + "epoch": 0.7666201766620176, + "grad_norm": 9.468759536743164, + "learning_rate": 9.9983519998352e-06, + "loss": 1.2196, + "step": 1649 + }, + { + "epoch": 0.7670850767085077, + "grad_norm": 7.812718868255615, + "learning_rate": 9.9983509998351e-06, + "loss": 0.832, + "step": 1650 + }, + { + "epoch": 0.7675499767549977, + "grad_norm": 8.39531421661377, + "learning_rate": 9.998349999835e-06, + "loss": 0.9698, + "step": 1651 + }, + { + "epoch": 0.7680148768014877, + "grad_norm": 5.172719478607178, + "learning_rate": 9.998348999834901e-06, + "loss": 0.9384, + "step": 1652 + }, + { + "epoch": 0.7684797768479776, + "grad_norm": 10.306553840637207, + "learning_rate": 9.998347999834802e-06, + "loss": 0.9171, + "step": 1653 + }, + { + "epoch": 0.7689446768944677, + "grad_norm": 9.751111030578613, + "learning_rate": 9.9983469998347e-06, + "loss": 0.9116, + "step": 1654 + }, + { + "epoch": 0.7694095769409577, + "grad_norm": 10.726200103759766, + "learning_rate": 9.998345999834601e-06, + "loss": 1.2489, + "step": 1655 + }, + { + "epoch": 0.7698744769874477, + "grad_norm": 5.685938358306885, + "learning_rate": 9.9983449998345e-06, + "loss": 0.8216, + "step": 1656 + }, + { + "epoch": 0.7703393770339377, + "grad_norm": 10.60875129699707, + "learning_rate": 9.9983439998344e-06, + "loss": 1.1112, + "step": 1657 + }, + { + "epoch": 0.7708042770804278, + "grad_norm": 6.9257121086120605, + "learning_rate": 9.998342999834301e-06, + "loss": 0.583, + "step": 1658 + }, + { + "epoch": 0.7712691771269177, + "grad_norm": 5.076784133911133, + "learning_rate": 9.9983419998342e-06, + "loss": 0.5616, + "step": 1659 + }, + { + "epoch": 0.7717340771734077, + "grad_norm": 4.796905040740967, + "learning_rate": 9.998340999834101e-06, + "loss": 0.5204, + "step": 1660 + }, + { + "epoch": 0.7721989772198977, + "grad_norm": 7.1346845626831055, + "learning_rate": 9.998339999834e-06, + "loss": 0.6916, + "step": 1661 + }, + { + "epoch": 0.7726638772663877, + "grad_norm": 7.252756118774414, + "learning_rate": 9.9983389998339e-06, + "loss": 0.4155, + "step": 1662 + }, + { + "epoch": 0.7731287773128778, + "grad_norm": 5.151723861694336, + "learning_rate": 9.9983379998338e-06, + "loss": 0.467, + "step": 1663 + }, + { + "epoch": 0.7735936773593677, + "grad_norm": 12.17117691040039, + "learning_rate": 9.9983369998337e-06, + "loss": 1.1957, + "step": 1664 + }, + { + "epoch": 0.7740585774058577, + "grad_norm": 6.783204078674316, + "learning_rate": 9.998335999833601e-06, + "loss": 0.7178, + "step": 1665 + }, + { + "epoch": 0.7745234774523477, + "grad_norm": 12.81383228302002, + "learning_rate": 9.998334999833502e-06, + "loss": 1.2, + "step": 1666 + }, + { + "epoch": 0.7749883774988378, + "grad_norm": 6.429178237915039, + "learning_rate": 9.9983339998334e-06, + "loss": 0.4475, + "step": 1667 + }, + { + "epoch": 0.7754532775453278, + "grad_norm": 6.748467922210693, + "learning_rate": 9.998332999833301e-06, + "loss": 0.8903, + "step": 1668 + }, + { + "epoch": 0.7759181775918178, + "grad_norm": 9.411212921142578, + "learning_rate": 9.9983319998332e-06, + "loss": 0.7943, + "step": 1669 + }, + { + "epoch": 0.7763830776383077, + "grad_norm": 7.449085712432861, + "learning_rate": 9.998330999833101e-06, + "loss": 0.7804, + "step": 1670 + }, + { + "epoch": 0.7768479776847977, + "grad_norm": 7.878346920013428, + "learning_rate": 9.998329999833002e-06, + "loss": 0.5704, + "step": 1671 + }, + { + "epoch": 0.7773128777312878, + "grad_norm": 11.244363784790039, + "learning_rate": 9.9983289998329e-06, + "loss": 0.8319, + "step": 1672 + }, + { + "epoch": 0.7777777777777778, + "grad_norm": 9.509714126586914, + "learning_rate": 9.998327999832801e-06, + "loss": 0.5848, + "step": 1673 + }, + { + "epoch": 0.7782426778242678, + "grad_norm": 5.955776691436768, + "learning_rate": 9.9983269998327e-06, + "loss": 0.7352, + "step": 1674 + }, + { + "epoch": 0.7787075778707578, + "grad_norm": 6.127941608428955, + "learning_rate": 9.9983259998326e-06, + "loss": 0.4232, + "step": 1675 + }, + { + "epoch": 0.7791724779172478, + "grad_norm": 6.314253330230713, + "learning_rate": 9.9983249998325e-06, + "loss": 0.6669, + "step": 1676 + }, + { + "epoch": 0.7796373779637378, + "grad_norm": 6.90627384185791, + "learning_rate": 9.9983239998324e-06, + "loss": 0.6603, + "step": 1677 + }, + { + "epoch": 0.7801022780102278, + "grad_norm": 9.063176155090332, + "learning_rate": 9.998322999832301e-06, + "loss": 0.6499, + "step": 1678 + }, + { + "epoch": 0.7805671780567178, + "grad_norm": 5.712819576263428, + "learning_rate": 9.998321999832202e-06, + "loss": 0.5798, + "step": 1679 + }, + { + "epoch": 0.7810320781032078, + "grad_norm": 18.506881713867188, + "learning_rate": 9.9983209998321e-06, + "loss": 1.0615, + "step": 1680 + }, + { + "epoch": 0.7814969781496978, + "grad_norm": 6.313080310821533, + "learning_rate": 9.998319999832001e-06, + "loss": 0.5847, + "step": 1681 + }, + { + "epoch": 0.7819618781961878, + "grad_norm": 6.557002067565918, + "learning_rate": 9.9983189998319e-06, + "loss": 0.4695, + "step": 1682 + }, + { + "epoch": 0.7824267782426778, + "grad_norm": 8.186263084411621, + "learning_rate": 9.998317999831801e-06, + "loss": 0.761, + "step": 1683 + }, + { + "epoch": 0.7828916782891678, + "grad_norm": 6.0825066566467285, + "learning_rate": 9.998316999831702e-06, + "loss": 0.6058, + "step": 1684 + }, + { + "epoch": 0.7833565783356579, + "grad_norm": 5.701247215270996, + "learning_rate": 9.9983159998316e-06, + "loss": 0.5204, + "step": 1685 + }, + { + "epoch": 0.7838214783821479, + "grad_norm": 5.794501781463623, + "learning_rate": 9.998314999831501e-06, + "loss": 0.4064, + "step": 1686 + }, + { + "epoch": 0.7842863784286378, + "grad_norm": 6.808113098144531, + "learning_rate": 9.9983139998314e-06, + "loss": 0.6165, + "step": 1687 + }, + { + "epoch": 0.7847512784751278, + "grad_norm": 5.825628280639648, + "learning_rate": 9.998312999831301e-06, + "loss": 0.5699, + "step": 1688 + }, + { + "epoch": 0.7852161785216178, + "grad_norm": 9.419376373291016, + "learning_rate": 9.9983119998312e-06, + "loss": 0.7755, + "step": 1689 + }, + { + "epoch": 0.7856810785681079, + "grad_norm": 7.642823219299316, + "learning_rate": 9.9983109998311e-06, + "loss": 0.7454, + "step": 1690 + }, + { + "epoch": 0.7861459786145979, + "grad_norm": 7.966249465942383, + "learning_rate": 9.998309999831001e-06, + "loss": 1.0338, + "step": 1691 + }, + { + "epoch": 0.7866108786610879, + "grad_norm": 8.465629577636719, + "learning_rate": 9.998308999830902e-06, + "loss": 0.5535, + "step": 1692 + }, + { + "epoch": 0.7870757787075778, + "grad_norm": 7.84459114074707, + "learning_rate": 9.9983079998308e-06, + "loss": 0.6677, + "step": 1693 + }, + { + "epoch": 0.7875406787540679, + "grad_norm": 6.063119888305664, + "learning_rate": 9.9983069998307e-06, + "loss": 0.5728, + "step": 1694 + }, + { + "epoch": 0.7880055788005579, + "grad_norm": 7.81863260269165, + "learning_rate": 9.9983059998306e-06, + "loss": 0.6363, + "step": 1695 + }, + { + "epoch": 0.7884704788470479, + "grad_norm": 5.924655914306641, + "learning_rate": 9.998304999830501e-06, + "loss": 0.4781, + "step": 1696 + }, + { + "epoch": 0.7889353788935379, + "grad_norm": 4.122714996337891, + "learning_rate": 9.998303999830402e-06, + "loss": 0.3431, + "step": 1697 + }, + { + "epoch": 0.7894002789400278, + "grad_norm": 8.981864929199219, + "learning_rate": 9.9983029998303e-06, + "loss": 0.8199, + "step": 1698 + }, + { + "epoch": 0.7898651789865179, + "grad_norm": 5.718377590179443, + "learning_rate": 9.998301999830201e-06, + "loss": 0.7866, + "step": 1699 + }, + { + "epoch": 0.7903300790330079, + "grad_norm": 7.5303144454956055, + "learning_rate": 9.9983009998301e-06, + "loss": 0.6618, + "step": 1700 + }, + { + "epoch": 0.7907949790794979, + "grad_norm": 4.619516372680664, + "learning_rate": 9.998299999830001e-06, + "loss": 0.3468, + "step": 1701 + }, + { + "epoch": 0.7912598791259879, + "grad_norm": 5.141032695770264, + "learning_rate": 9.9982989998299e-06, + "loss": 0.3837, + "step": 1702 + }, + { + "epoch": 0.791724779172478, + "grad_norm": 9.08244514465332, + "learning_rate": 9.9982979998298e-06, + "loss": 0.9096, + "step": 1703 + }, + { + "epoch": 0.7921896792189679, + "grad_norm": 5.101423263549805, + "learning_rate": 9.998296999829701e-06, + "loss": 0.7161, + "step": 1704 + }, + { + "epoch": 0.7926545792654579, + "grad_norm": 7.639341354370117, + "learning_rate": 9.9982959998296e-06, + "loss": 1.0466, + "step": 1705 + }, + { + "epoch": 0.7931194793119479, + "grad_norm": 5.812538146972656, + "learning_rate": 9.998294999829501e-06, + "loss": 0.4627, + "step": 1706 + }, + { + "epoch": 0.793584379358438, + "grad_norm": 6.300144195556641, + "learning_rate": 9.9982939998294e-06, + "loss": 0.4363, + "step": 1707 + }, + { + "epoch": 0.794049279404928, + "grad_norm": 7.396739959716797, + "learning_rate": 9.9982929998293e-06, + "loss": 0.7334, + "step": 1708 + }, + { + "epoch": 0.7945141794514179, + "grad_norm": 6.356086254119873, + "learning_rate": 9.998291999829201e-06, + "loss": 0.4904, + "step": 1709 + }, + { + "epoch": 0.7949790794979079, + "grad_norm": 9.790695190429688, + "learning_rate": 9.998290999829102e-06, + "loss": 1.0118, + "step": 1710 + }, + { + "epoch": 0.7954439795443979, + "grad_norm": 6.143927097320557, + "learning_rate": 9.998289999829e-06, + "loss": 0.6724, + "step": 1711 + }, + { + "epoch": 0.795908879590888, + "grad_norm": 8.34294319152832, + "learning_rate": 9.998288999828901e-06, + "loss": 0.7512, + "step": 1712 + }, + { + "epoch": 0.796373779637378, + "grad_norm": 4.58821964263916, + "learning_rate": 9.9982879998288e-06, + "loss": 0.4937, + "step": 1713 + }, + { + "epoch": 0.796838679683868, + "grad_norm": 8.523152351379395, + "learning_rate": 9.998286999828701e-06, + "loss": 0.5175, + "step": 1714 + }, + { + "epoch": 0.7973035797303579, + "grad_norm": 9.677740097045898, + "learning_rate": 9.9982859998286e-06, + "loss": 1.2575, + "step": 1715 + }, + { + "epoch": 0.797768479776848, + "grad_norm": 7.932932376861572, + "learning_rate": 9.9982849998285e-06, + "loss": 0.9251, + "step": 1716 + }, + { + "epoch": 0.798233379823338, + "grad_norm": 8.906527519226074, + "learning_rate": 9.998283999828401e-06, + "loss": 1.0814, + "step": 1717 + }, + { + "epoch": 0.798698279869828, + "grad_norm": 6.1788811683654785, + "learning_rate": 9.9982829998283e-06, + "loss": 0.9935, + "step": 1718 + }, + { + "epoch": 0.799163179916318, + "grad_norm": 5.328824996948242, + "learning_rate": 9.998281999828201e-06, + "loss": 0.485, + "step": 1719 + }, + { + "epoch": 0.799628079962808, + "grad_norm": 10.686690330505371, + "learning_rate": 9.9982809998281e-06, + "loss": 1.0887, + "step": 1720 + }, + { + "epoch": 0.800092980009298, + "grad_norm": 10.024508476257324, + "learning_rate": 9.998279999828e-06, + "loss": 1.0594, + "step": 1721 + }, + { + "epoch": 0.800557880055788, + "grad_norm": 5.6274800300598145, + "learning_rate": 9.998278999827901e-06, + "loss": 0.6934, + "step": 1722 + }, + { + "epoch": 0.801022780102278, + "grad_norm": 9.403301239013672, + "learning_rate": 9.998277999827802e-06, + "loss": 1.1587, + "step": 1723 + }, + { + "epoch": 0.801487680148768, + "grad_norm": 9.243465423583984, + "learning_rate": 9.998276999827701e-06, + "loss": 0.7051, + "step": 1724 + }, + { + "epoch": 0.8019525801952581, + "grad_norm": 6.40681791305542, + "learning_rate": 9.998275999827602e-06, + "loss": 0.4574, + "step": 1725 + }, + { + "epoch": 0.802417480241748, + "grad_norm": 6.527696132659912, + "learning_rate": 9.9982749998275e-06, + "loss": 0.6783, + "step": 1726 + }, + { + "epoch": 0.802882380288238, + "grad_norm": 5.337404251098633, + "learning_rate": 9.998273999827401e-06, + "loss": 0.3938, + "step": 1727 + }, + { + "epoch": 0.803347280334728, + "grad_norm": 4.8317108154296875, + "learning_rate": 9.9982729998273e-06, + "loss": 0.3449, + "step": 1728 + }, + { + "epoch": 0.803812180381218, + "grad_norm": 9.638764381408691, + "learning_rate": 9.9982719998272e-06, + "loss": 0.811, + "step": 1729 + }, + { + "epoch": 0.8042770804277081, + "grad_norm": 6.105317115783691, + "learning_rate": 9.998270999827101e-06, + "loss": 0.6953, + "step": 1730 + }, + { + "epoch": 0.8047419804741981, + "grad_norm": 8.312463760375977, + "learning_rate": 9.998269999827e-06, + "loss": 0.6747, + "step": 1731 + }, + { + "epoch": 0.805206880520688, + "grad_norm": 4.725527286529541, + "learning_rate": 9.998268999826901e-06, + "loss": 0.3607, + "step": 1732 + }, + { + "epoch": 0.805671780567178, + "grad_norm": 8.103647232055664, + "learning_rate": 9.9982679998268e-06, + "loss": 0.6159, + "step": 1733 + }, + { + "epoch": 0.806136680613668, + "grad_norm": 8.315047264099121, + "learning_rate": 9.9982669998267e-06, + "loss": 0.6332, + "step": 1734 + }, + { + "epoch": 0.8066015806601581, + "grad_norm": 6.1559529304504395, + "learning_rate": 9.998265999826601e-06, + "loss": 0.7481, + "step": 1735 + }, + { + "epoch": 0.8070664807066481, + "grad_norm": 4.345332622528076, + "learning_rate": 9.998264999826502e-06, + "loss": 0.3711, + "step": 1736 + }, + { + "epoch": 0.8075313807531381, + "grad_norm": 6.357636451721191, + "learning_rate": 9.998263999826401e-06, + "loss": 0.6485, + "step": 1737 + }, + { + "epoch": 0.807996280799628, + "grad_norm": 7.7893195152282715, + "learning_rate": 9.9982629998263e-06, + "loss": 0.7652, + "step": 1738 + }, + { + "epoch": 0.8084611808461181, + "grad_norm": 5.586349010467529, + "learning_rate": 9.9982619998262e-06, + "loss": 0.6019, + "step": 1739 + }, + { + "epoch": 0.8089260808926081, + "grad_norm": 8.822538375854492, + "learning_rate": 9.998260999826101e-06, + "loss": 0.9044, + "step": 1740 + }, + { + "epoch": 0.8093909809390981, + "grad_norm": 7.367069244384766, + "learning_rate": 9.998259999826002e-06, + "loss": 0.9608, + "step": 1741 + }, + { + "epoch": 0.8098558809855881, + "grad_norm": 8.291058540344238, + "learning_rate": 9.9982589998259e-06, + "loss": 0.6095, + "step": 1742 + }, + { + "epoch": 0.810320781032078, + "grad_norm": 6.586578845977783, + "learning_rate": 9.998257999825801e-06, + "loss": 0.5935, + "step": 1743 + }, + { + "epoch": 0.8107856810785681, + "grad_norm": 6.680955410003662, + "learning_rate": 9.9982569998257e-06, + "loss": 0.7514, + "step": 1744 + }, + { + "epoch": 0.8112505811250581, + "grad_norm": 9.273926734924316, + "learning_rate": 9.998255999825601e-06, + "loss": 0.6301, + "step": 1745 + }, + { + "epoch": 0.8117154811715481, + "grad_norm": 12.677363395690918, + "learning_rate": 9.9982549998255e-06, + "loss": 0.7388, + "step": 1746 + }, + { + "epoch": 0.8121803812180381, + "grad_norm": 6.135332107543945, + "learning_rate": 9.9982539998254e-06, + "loss": 0.556, + "step": 1747 + }, + { + "epoch": 0.8126452812645282, + "grad_norm": 9.569008827209473, + "learning_rate": 9.998252999825301e-06, + "loss": 0.8359, + "step": 1748 + }, + { + "epoch": 0.8131101813110181, + "grad_norm": 15.688093185424805, + "learning_rate": 9.9982519998252e-06, + "loss": 2.9017, + "step": 1749 + }, + { + "epoch": 0.8135750813575081, + "grad_norm": 11.105977058410645, + "learning_rate": 9.998250999825101e-06, + "loss": 0.982, + "step": 1750 + }, + { + "epoch": 0.8140399814039981, + "grad_norm": 8.220104217529297, + "learning_rate": 9.998249999825e-06, + "loss": 0.9758, + "step": 1751 + }, + { + "epoch": 0.8145048814504882, + "grad_norm": 6.047811508178711, + "learning_rate": 9.9982489998249e-06, + "loss": 0.5099, + "step": 1752 + }, + { + "epoch": 0.8149697814969782, + "grad_norm": 6.43534517288208, + "learning_rate": 9.998247999824801e-06, + "loss": 0.8315, + "step": 1753 + }, + { + "epoch": 0.8154346815434682, + "grad_norm": 12.415273666381836, + "learning_rate": 9.998246999824702e-06, + "loss": 1.0697, + "step": 1754 + }, + { + "epoch": 0.8158995815899581, + "grad_norm": 6.746450424194336, + "learning_rate": 9.998245999824601e-06, + "loss": 0.7997, + "step": 1755 + }, + { + "epoch": 0.8163644816364481, + "grad_norm": 5.4247355461120605, + "learning_rate": 9.998244999824502e-06, + "loss": 0.5146, + "step": 1756 + }, + { + "epoch": 0.8168293816829382, + "grad_norm": 8.113396644592285, + "learning_rate": 9.9982439998244e-06, + "loss": 0.8811, + "step": 1757 + }, + { + "epoch": 0.8172942817294282, + "grad_norm": 7.026947021484375, + "learning_rate": 9.998242999824301e-06, + "loss": 0.9432, + "step": 1758 + }, + { + "epoch": 0.8177591817759182, + "grad_norm": 6.7126383781433105, + "learning_rate": 9.9982419998242e-06, + "loss": 0.6968, + "step": 1759 + }, + { + "epoch": 0.8182240818224081, + "grad_norm": 6.390560150146484, + "learning_rate": 9.9982409998241e-06, + "loss": 0.4867, + "step": 1760 + }, + { + "epoch": 0.8186889818688982, + "grad_norm": 5.323058128356934, + "learning_rate": 9.998239999824001e-06, + "loss": 0.3621, + "step": 1761 + }, + { + "epoch": 0.8191538819153882, + "grad_norm": 6.952043533325195, + "learning_rate": 9.9982389998239e-06, + "loss": 0.735, + "step": 1762 + }, + { + "epoch": 0.8196187819618782, + "grad_norm": 6.8238420486450195, + "learning_rate": 9.998237999823801e-06, + "loss": 1.0431, + "step": 1763 + }, + { + "epoch": 0.8200836820083682, + "grad_norm": 6.2494378089904785, + "learning_rate": 9.9982369998237e-06, + "loss": 0.4693, + "step": 1764 + }, + { + "epoch": 0.8205485820548583, + "grad_norm": 5.56302547454834, + "learning_rate": 9.9982359998236e-06, + "loss": 0.6693, + "step": 1765 + }, + { + "epoch": 0.8210134821013482, + "grad_norm": 8.486534118652344, + "learning_rate": 9.998234999823501e-06, + "loss": 0.7642, + "step": 1766 + }, + { + "epoch": 0.8214783821478382, + "grad_norm": 9.979066848754883, + "learning_rate": 9.998233999823402e-06, + "loss": 0.8459, + "step": 1767 + }, + { + "epoch": 0.8219432821943282, + "grad_norm": 5.991272926330566, + "learning_rate": 9.998232999823301e-06, + "loss": 0.7338, + "step": 1768 + }, + { + "epoch": 0.8224081822408182, + "grad_norm": 5.773449897766113, + "learning_rate": 9.9982319998232e-06, + "loss": 0.6044, + "step": 1769 + }, + { + "epoch": 0.8228730822873083, + "grad_norm": 5.798567295074463, + "learning_rate": 9.9982309998231e-06, + "loss": 0.6402, + "step": 1770 + }, + { + "epoch": 0.8233379823337982, + "grad_norm": 5.278077602386475, + "learning_rate": 9.998229999823e-06, + "loss": 0.6119, + "step": 1771 + }, + { + "epoch": 0.8238028823802882, + "grad_norm": 9.010823249816895, + "learning_rate": 9.9982289998229e-06, + "loss": 0.8575, + "step": 1772 + }, + { + "epoch": 0.8242677824267782, + "grad_norm": 7.139113903045654, + "learning_rate": 9.998227999822801e-06, + "loss": 0.6447, + "step": 1773 + }, + { + "epoch": 0.8247326824732683, + "grad_norm": 12.210677146911621, + "learning_rate": 9.998226999822702e-06, + "loss": 0.8075, + "step": 1774 + }, + { + "epoch": 0.8251975825197583, + "grad_norm": 5.7334465980529785, + "learning_rate": 9.9982259998226e-06, + "loss": 0.762, + "step": 1775 + }, + { + "epoch": 0.8256624825662483, + "grad_norm": 10.03715991973877, + "learning_rate": 9.998224999822501e-06, + "loss": 0.815, + "step": 1776 + }, + { + "epoch": 0.8261273826127382, + "grad_norm": 7.065068244934082, + "learning_rate": 9.9982239998224e-06, + "loss": 0.6338, + "step": 1777 + }, + { + "epoch": 0.8265922826592282, + "grad_norm": 9.442242622375488, + "learning_rate": 9.9982229998223e-06, + "loss": 1.0578, + "step": 1778 + }, + { + "epoch": 0.8270571827057183, + "grad_norm": 5.225984573364258, + "learning_rate": 9.998221999822201e-06, + "loss": 0.3825, + "step": 1779 + }, + { + "epoch": 0.8275220827522083, + "grad_norm": 5.49752140045166, + "learning_rate": 9.9982209998221e-06, + "loss": 0.5396, + "step": 1780 + }, + { + "epoch": 0.8279869827986983, + "grad_norm": 7.091917991638184, + "learning_rate": 9.998219999822001e-06, + "loss": 0.4761, + "step": 1781 + }, + { + "epoch": 0.8284518828451883, + "grad_norm": 6.162967205047607, + "learning_rate": 9.9982189998219e-06, + "loss": 0.5777, + "step": 1782 + }, + { + "epoch": 0.8289167828916782, + "grad_norm": 6.594886302947998, + "learning_rate": 9.9982179998218e-06, + "loss": 0.7224, + "step": 1783 + }, + { + "epoch": 0.8293816829381683, + "grad_norm": 5.772677421569824, + "learning_rate": 9.9982169998217e-06, + "loss": 0.667, + "step": 1784 + }, + { + "epoch": 0.8298465829846583, + "grad_norm": 6.471079349517822, + "learning_rate": 9.9982159998216e-06, + "loss": 0.5401, + "step": 1785 + }, + { + "epoch": 0.8303114830311483, + "grad_norm": 11.50944995880127, + "learning_rate": 9.998214999821501e-06, + "loss": 0.677, + "step": 1786 + }, + { + "epoch": 0.8307763830776383, + "grad_norm": 6.743424415588379, + "learning_rate": 9.998213999821402e-06, + "loss": 0.665, + "step": 1787 + }, + { + "epoch": 0.8312412831241283, + "grad_norm": 8.384634017944336, + "learning_rate": 9.9982129998213e-06, + "loss": 1.0937, + "step": 1788 + }, + { + "epoch": 0.8317061831706183, + "grad_norm": 10.73773193359375, + "learning_rate": 9.998211999821201e-06, + "loss": 1.1763, + "step": 1789 + }, + { + "epoch": 0.8321710832171083, + "grad_norm": 11.768834114074707, + "learning_rate": 9.9982109998211e-06, + "loss": 0.8803, + "step": 1790 + }, + { + "epoch": 0.8326359832635983, + "grad_norm": 6.2612104415893555, + "learning_rate": 9.998209999821001e-06, + "loss": 0.9059, + "step": 1791 + }, + { + "epoch": 0.8331008833100884, + "grad_norm": 12.732711791992188, + "learning_rate": 9.998208999820902e-06, + "loss": 0.8896, + "step": 1792 + }, + { + "epoch": 0.8335657833565784, + "grad_norm": 4.326085090637207, + "learning_rate": 9.9982079998208e-06, + "loss": 0.3882, + "step": 1793 + }, + { + "epoch": 0.8340306834030683, + "grad_norm": 7.129255294799805, + "learning_rate": 9.998206999820701e-06, + "loss": 0.7892, + "step": 1794 + }, + { + "epoch": 0.8344955834495583, + "grad_norm": 6.024131774902344, + "learning_rate": 9.9982059998206e-06, + "loss": 0.5261, + "step": 1795 + }, + { + "epoch": 0.8349604834960483, + "grad_norm": 6.3001227378845215, + "learning_rate": 9.9982049998205e-06, + "loss": 0.7836, + "step": 1796 + }, + { + "epoch": 0.8354253835425384, + "grad_norm": 6.359413146972656, + "learning_rate": 9.9982039998204e-06, + "loss": 0.6019, + "step": 1797 + }, + { + "epoch": 0.8358902835890284, + "grad_norm": 6.2938232421875, + "learning_rate": 9.9982029998203e-06, + "loss": 0.6799, + "step": 1798 + }, + { + "epoch": 0.8363551836355184, + "grad_norm": 4.673009872436523, + "learning_rate": 9.998201999820201e-06, + "loss": 0.4152, + "step": 1799 + }, + { + "epoch": 0.8368200836820083, + "grad_norm": 8.948838233947754, + "learning_rate": 9.998200999820102e-06, + "loss": 0.88, + "step": 1800 + }, + { + "epoch": 0.8372849837284984, + "grad_norm": 7.4146728515625, + "learning_rate": 9.99819999982e-06, + "loss": 0.7725, + "step": 1801 + }, + { + "epoch": 0.8377498837749884, + "grad_norm": 6.176004409790039, + "learning_rate": 9.9981989998199e-06, + "loss": 0.6288, + "step": 1802 + }, + { + "epoch": 0.8382147838214784, + "grad_norm": 7.287527561187744, + "learning_rate": 9.9981979998198e-06, + "loss": 0.7085, + "step": 1803 + }, + { + "epoch": 0.8386796838679684, + "grad_norm": 19.327533721923828, + "learning_rate": 9.998196999819701e-06, + "loss": 0.7913, + "step": 1804 + }, + { + "epoch": 0.8391445839144583, + "grad_norm": 8.015626907348633, + "learning_rate": 9.998195999819602e-06, + "loss": 0.6444, + "step": 1805 + }, + { + "epoch": 0.8396094839609484, + "grad_norm": 5.290071964263916, + "learning_rate": 9.9981949998195e-06, + "loss": 0.6469, + "step": 1806 + }, + { + "epoch": 0.8400743840074384, + "grad_norm": 6.683074951171875, + "learning_rate": 9.998193999819401e-06, + "loss": 0.4445, + "step": 1807 + }, + { + "epoch": 0.8405392840539284, + "grad_norm": 7.136590480804443, + "learning_rate": 9.9981929998193e-06, + "loss": 0.7148, + "step": 1808 + }, + { + "epoch": 0.8410041841004184, + "grad_norm": 4.228488922119141, + "learning_rate": 9.998191999819201e-06, + "loss": 0.3308, + "step": 1809 + }, + { + "epoch": 0.8414690841469085, + "grad_norm": 6.73750114440918, + "learning_rate": 9.998190999819102e-06, + "loss": 0.713, + "step": 1810 + }, + { + "epoch": 0.8419339841933984, + "grad_norm": 9.554896354675293, + "learning_rate": 9.998189999819e-06, + "loss": 0.8676, + "step": 1811 + }, + { + "epoch": 0.8423988842398884, + "grad_norm": 9.927472114562988, + "learning_rate": 9.998188999818901e-06, + "loss": 1.1004, + "step": 1812 + }, + { + "epoch": 0.8428637842863784, + "grad_norm": 8.308180809020996, + "learning_rate": 9.9981879998188e-06, + "loss": 0.5172, + "step": 1813 + }, + { + "epoch": 0.8433286843328685, + "grad_norm": 6.6081767082214355, + "learning_rate": 9.9981869998187e-06, + "loss": 0.4387, + "step": 1814 + }, + { + "epoch": 0.8437935843793585, + "grad_norm": 6.141027927398682, + "learning_rate": 9.9981859998186e-06, + "loss": 0.7863, + "step": 1815 + }, + { + "epoch": 0.8442584844258484, + "grad_norm": 11.76423454284668, + "learning_rate": 9.9981849998185e-06, + "loss": 1.0281, + "step": 1816 + }, + { + "epoch": 0.8447233844723384, + "grad_norm": 5.0612592697143555, + "learning_rate": 9.998183999818401e-06, + "loss": 0.6408, + "step": 1817 + }, + { + "epoch": 0.8451882845188284, + "grad_norm": 5.351144313812256, + "learning_rate": 9.998182999818302e-06, + "loss": 0.6095, + "step": 1818 + }, + { + "epoch": 0.8456531845653185, + "grad_norm": 7.60306453704834, + "learning_rate": 9.9981819998182e-06, + "loss": 0.5881, + "step": 1819 + }, + { + "epoch": 0.8461180846118085, + "grad_norm": 8.472026824951172, + "learning_rate": 9.998180999818101e-06, + "loss": 0.3478, + "step": 1820 + }, + { + "epoch": 0.8465829846582985, + "grad_norm": 8.706148147583008, + "learning_rate": 9.998179999818e-06, + "loss": 0.9717, + "step": 1821 + }, + { + "epoch": 0.8470478847047884, + "grad_norm": 6.306098937988281, + "learning_rate": 9.998178999817901e-06, + "loss": 0.7021, + "step": 1822 + }, + { + "epoch": 0.8475127847512784, + "grad_norm": 8.230513572692871, + "learning_rate": 9.998177999817802e-06, + "loss": 0.6692, + "step": 1823 + }, + { + "epoch": 0.8479776847977685, + "grad_norm": 6.976614475250244, + "learning_rate": 9.9981769998177e-06, + "loss": 0.8926, + "step": 1824 + }, + { + "epoch": 0.8484425848442585, + "grad_norm": 7.439591407775879, + "learning_rate": 9.998175999817601e-06, + "loss": 0.7588, + "step": 1825 + }, + { + "epoch": 0.8489074848907485, + "grad_norm": 7.09647798538208, + "learning_rate": 9.9981749998175e-06, + "loss": 0.9588, + "step": 1826 + }, + { + "epoch": 0.8493723849372385, + "grad_norm": 6.3226470947265625, + "learning_rate": 9.9981739998174e-06, + "loss": 1.0663, + "step": 1827 + }, + { + "epoch": 0.8498372849837285, + "grad_norm": 5.30750036239624, + "learning_rate": 9.9981729998173e-06, + "loss": 0.4091, + "step": 1828 + }, + { + "epoch": 0.8503021850302185, + "grad_norm": 6.069753170013428, + "learning_rate": 9.9981719998172e-06, + "loss": 0.5697, + "step": 1829 + }, + { + "epoch": 0.8507670850767085, + "grad_norm": 10.382354736328125, + "learning_rate": 9.998170999817101e-06, + "loss": 1.2399, + "step": 1830 + }, + { + "epoch": 0.8512319851231985, + "grad_norm": 8.606959342956543, + "learning_rate": 9.998169999817002e-06, + "loss": 0.6513, + "step": 1831 + }, + { + "epoch": 0.8516968851696886, + "grad_norm": 10.443204879760742, + "learning_rate": 9.9981689998169e-06, + "loss": 1.171, + "step": 1832 + }, + { + "epoch": 0.8521617852161785, + "grad_norm": 5.074190139770508, + "learning_rate": 9.998167999816801e-06, + "loss": 0.2981, + "step": 1833 + }, + { + "epoch": 0.8526266852626685, + "grad_norm": 6.61204195022583, + "learning_rate": 9.9981669998167e-06, + "loss": 0.82, + "step": 1834 + }, + { + "epoch": 0.8530915853091585, + "grad_norm": 8.033239364624023, + "learning_rate": 9.998165999816601e-06, + "loss": 0.738, + "step": 1835 + }, + { + "epoch": 0.8535564853556485, + "grad_norm": 12.455591201782227, + "learning_rate": 9.998164999816502e-06, + "loss": 1.0636, + "step": 1836 + }, + { + "epoch": 0.8540213854021386, + "grad_norm": 6.663708209991455, + "learning_rate": 9.9981639998164e-06, + "loss": 0.5012, + "step": 1837 + }, + { + "epoch": 0.8544862854486286, + "grad_norm": 16.7215518951416, + "learning_rate": 9.998162999816301e-06, + "loss": 3.3687, + "step": 1838 + }, + { + "epoch": 0.8549511854951185, + "grad_norm": 9.075592041015625, + "learning_rate": 9.9981619998162e-06, + "loss": 0.7244, + "step": 1839 + }, + { + "epoch": 0.8554160855416085, + "grad_norm": 6.882232666015625, + "learning_rate": 9.998160999816101e-06, + "loss": 0.4366, + "step": 1840 + }, + { + "epoch": 0.8558809855880986, + "grad_norm": 5.00037956237793, + "learning_rate": 9.998159999816e-06, + "loss": 0.5305, + "step": 1841 + }, + { + "epoch": 0.8563458856345886, + "grad_norm": 5.977245330810547, + "learning_rate": 9.9981589998159e-06, + "loss": 0.5565, + "step": 1842 + }, + { + "epoch": 0.8568107856810786, + "grad_norm": 6.809998512268066, + "learning_rate": 9.998157999815801e-06, + "loss": 0.5534, + "step": 1843 + }, + { + "epoch": 0.8572756857275686, + "grad_norm": 5.786665916442871, + "learning_rate": 9.998156999815702e-06, + "loss": 0.6786, + "step": 1844 + }, + { + "epoch": 0.8577405857740585, + "grad_norm": 5.274286270141602, + "learning_rate": 9.9981559998156e-06, + "loss": 0.3573, + "step": 1845 + }, + { + "epoch": 0.8582054858205486, + "grad_norm": 9.238567352294922, + "learning_rate": 9.9981549998155e-06, + "loss": 0.8189, + "step": 1846 + }, + { + "epoch": 0.8586703858670386, + "grad_norm": 9.70542049407959, + "learning_rate": 9.9981539998154e-06, + "loss": 1.1012, + "step": 1847 + }, + { + "epoch": 0.8591352859135286, + "grad_norm": 5.891972064971924, + "learning_rate": 9.998152999815301e-06, + "loss": 0.509, + "step": 1848 + }, + { + "epoch": 0.8596001859600186, + "grad_norm": 7.622063159942627, + "learning_rate": 9.998151999815202e-06, + "loss": 0.9793, + "step": 1849 + }, + { + "epoch": 0.8600650860065086, + "grad_norm": 7.932720184326172, + "learning_rate": 9.9981509998151e-06, + "loss": 0.89, + "step": 1850 + }, + { + "epoch": 0.8605299860529986, + "grad_norm": 11.006072044372559, + "learning_rate": 9.998149999815001e-06, + "loss": 0.9296, + "step": 1851 + }, + { + "epoch": 0.8609948860994886, + "grad_norm": 6.373332500457764, + "learning_rate": 9.9981489998149e-06, + "loss": 0.7444, + "step": 1852 + }, + { + "epoch": 0.8614597861459786, + "grad_norm": 11.86872673034668, + "learning_rate": 9.998147999814801e-06, + "loss": 0.7226, + "step": 1853 + }, + { + "epoch": 0.8619246861924686, + "grad_norm": 10.855386734008789, + "learning_rate": 9.9981469998147e-06, + "loss": 1.3247, + "step": 1854 + }, + { + "epoch": 0.8623895862389587, + "grad_norm": 10.554044723510742, + "learning_rate": 9.9981459998146e-06, + "loss": 1.0991, + "step": 1855 + }, + { + "epoch": 0.8628544862854486, + "grad_norm": 6.017852783203125, + "learning_rate": 9.998144999814501e-06, + "loss": 0.5857, + "step": 1856 + }, + { + "epoch": 0.8633193863319386, + "grad_norm": 11.718999862670898, + "learning_rate": 9.9981439998144e-06, + "loss": 1.0611, + "step": 1857 + }, + { + "epoch": 0.8637842863784286, + "grad_norm": 7.990134239196777, + "learning_rate": 9.998142999814301e-06, + "loss": 0.8956, + "step": 1858 + }, + { + "epoch": 0.8642491864249187, + "grad_norm": 6.208372592926025, + "learning_rate": 9.9981419998142e-06, + "loss": 0.6235, + "step": 1859 + }, + { + "epoch": 0.8647140864714087, + "grad_norm": 7.075347900390625, + "learning_rate": 9.9981409998141e-06, + "loss": 0.9449, + "step": 1860 + }, + { + "epoch": 0.8651789865178987, + "grad_norm": 6.880540370941162, + "learning_rate": 9.998139999814001e-06, + "loss": 0.9243, + "step": 1861 + }, + { + "epoch": 0.8656438865643886, + "grad_norm": 7.302027702331543, + "learning_rate": 9.998138999813902e-06, + "loss": 0.8877, + "step": 1862 + }, + { + "epoch": 0.8661087866108786, + "grad_norm": 7.142313003540039, + "learning_rate": 9.9981379998138e-06, + "loss": 0.7017, + "step": 1863 + }, + { + "epoch": 0.8665736866573687, + "grad_norm": 6.301131248474121, + "learning_rate": 9.998136999813701e-06, + "loss": 0.5636, + "step": 1864 + }, + { + "epoch": 0.8670385867038587, + "grad_norm": 8.41162395477295, + "learning_rate": 9.9981359998136e-06, + "loss": 0.5153, + "step": 1865 + }, + { + "epoch": 0.8675034867503487, + "grad_norm": 4.800826549530029, + "learning_rate": 9.998134999813501e-06, + "loss": 0.4384, + "step": 1866 + }, + { + "epoch": 0.8679683867968386, + "grad_norm": 7.323402404785156, + "learning_rate": 9.9981339998134e-06, + "loss": 0.6155, + "step": 1867 + }, + { + "epoch": 0.8684332868433287, + "grad_norm": 4.839722633361816, + "learning_rate": 9.9981329998133e-06, + "loss": 0.5549, + "step": 1868 + }, + { + "epoch": 0.8688981868898187, + "grad_norm": 6.816176414489746, + "learning_rate": 9.998131999813201e-06, + "loss": 1.0448, + "step": 1869 + }, + { + "epoch": 0.8693630869363087, + "grad_norm": 7.96941614151001, + "learning_rate": 9.9981309998131e-06, + "loss": 0.8749, + "step": 1870 + }, + { + "epoch": 0.8698279869827987, + "grad_norm": 5.364948272705078, + "learning_rate": 9.998129999813001e-06, + "loss": 0.4476, + "step": 1871 + }, + { + "epoch": 0.8702928870292888, + "grad_norm": 4.82069730758667, + "learning_rate": 9.9981289998129e-06, + "loss": 0.56, + "step": 1872 + }, + { + "epoch": 0.8707577870757787, + "grad_norm": 9.707784652709961, + "learning_rate": 9.9981279998128e-06, + "loss": 1.3657, + "step": 1873 + }, + { + "epoch": 0.8712226871222687, + "grad_norm": 6.953049659729004, + "learning_rate": 9.998126999812701e-06, + "loss": 0.6835, + "step": 1874 + }, + { + "epoch": 0.8716875871687587, + "grad_norm": 8.098422050476074, + "learning_rate": 9.998125999812602e-06, + "loss": 0.8633, + "step": 1875 + }, + { + "epoch": 0.8721524872152487, + "grad_norm": 6.583449840545654, + "learning_rate": 9.998124999812501e-06, + "loss": 0.6153, + "step": 1876 + }, + { + "epoch": 0.8726173872617388, + "grad_norm": 8.539508819580078, + "learning_rate": 9.998123999812402e-06, + "loss": 0.5264, + "step": 1877 + }, + { + "epoch": 0.8730822873082287, + "grad_norm": 11.405159950256348, + "learning_rate": 9.9981229998123e-06, + "loss": 0.6705, + "step": 1878 + }, + { + "epoch": 0.8735471873547187, + "grad_norm": 7.369908332824707, + "learning_rate": 9.998121999812201e-06, + "loss": 0.5235, + "step": 1879 + }, + { + "epoch": 0.8740120874012087, + "grad_norm": 6.48301362991333, + "learning_rate": 9.9981209998121e-06, + "loss": 0.6239, + "step": 1880 + }, + { + "epoch": 0.8744769874476988, + "grad_norm": 8.763883590698242, + "learning_rate": 9.998119999812e-06, + "loss": 0.8087, + "step": 1881 + }, + { + "epoch": 0.8749418874941888, + "grad_norm": 6.8775224685668945, + "learning_rate": 9.998118999811901e-06, + "loss": 0.5379, + "step": 1882 + }, + { + "epoch": 0.8754067875406788, + "grad_norm": 7.240195274353027, + "learning_rate": 9.9981179998118e-06, + "loss": 0.949, + "step": 1883 + }, + { + "epoch": 0.8758716875871687, + "grad_norm": 8.116570472717285, + "learning_rate": 9.998116999811701e-06, + "loss": 0.8662, + "step": 1884 + }, + { + "epoch": 0.8763365876336587, + "grad_norm": 6.552999019622803, + "learning_rate": 9.9981159998116e-06, + "loss": 0.7186, + "step": 1885 + }, + { + "epoch": 0.8768014876801488, + "grad_norm": 4.506673812866211, + "learning_rate": 9.9981149998115e-06, + "loss": 0.4402, + "step": 1886 + }, + { + "epoch": 0.8772663877266388, + "grad_norm": 9.638957023620605, + "learning_rate": 9.998113999811401e-06, + "loss": 1.3625, + "step": 1887 + }, + { + "epoch": 0.8777312877731288, + "grad_norm": 8.6482572555542, + "learning_rate": 9.998112999811302e-06, + "loss": 0.839, + "step": 1888 + }, + { + "epoch": 0.8781961878196188, + "grad_norm": 7.147324085235596, + "learning_rate": 9.998111999811201e-06, + "loss": 0.6279, + "step": 1889 + }, + { + "epoch": 0.8786610878661087, + "grad_norm": 5.351322650909424, + "learning_rate": 9.9981109998111e-06, + "loss": 0.3725, + "step": 1890 + }, + { + "epoch": 0.8791259879125988, + "grad_norm": 4.183652400970459, + "learning_rate": 9.998109999811e-06, + "loss": 0.3023, + "step": 1891 + }, + { + "epoch": 0.8795908879590888, + "grad_norm": 9.31440544128418, + "learning_rate": 9.998108999810901e-06, + "loss": 1.1098, + "step": 1892 + }, + { + "epoch": 0.8800557880055788, + "grad_norm": 4.189338207244873, + "learning_rate": 9.9981079998108e-06, + "loss": 0.4192, + "step": 1893 + }, + { + "epoch": 0.8805206880520688, + "grad_norm": 2.870708465576172, + "learning_rate": 9.998106999810701e-06, + "loss": 0.2843, + "step": 1894 + }, + { + "epoch": 0.8809855880985588, + "grad_norm": 11.8063325881958, + "learning_rate": 9.998105999810602e-06, + "loss": 0.928, + "step": 1895 + }, + { + "epoch": 0.8814504881450488, + "grad_norm": 9.081511497497559, + "learning_rate": 9.9981049998105e-06, + "loss": 1.1424, + "step": 1896 + }, + { + "epoch": 0.8819153881915388, + "grad_norm": 6.1090192794799805, + "learning_rate": 9.998103999810401e-06, + "loss": 0.5776, + "step": 1897 + }, + { + "epoch": 0.8823802882380288, + "grad_norm": 8.847550392150879, + "learning_rate": 9.9981029998103e-06, + "loss": 0.782, + "step": 1898 + }, + { + "epoch": 0.8828451882845189, + "grad_norm": 7.0014424324035645, + "learning_rate": 9.9981019998102e-06, + "loss": 0.7268, + "step": 1899 + }, + { + "epoch": 0.8833100883310089, + "grad_norm": 10.251333236694336, + "learning_rate": 9.998100999810101e-06, + "loss": 1.0734, + "step": 1900 + }, + { + "epoch": 0.8837749883774988, + "grad_norm": 8.99980640411377, + "learning_rate": 9.99809999981e-06, + "loss": 1.1857, + "step": 1901 + }, + { + "epoch": 0.8842398884239888, + "grad_norm": 17.38448715209961, + "learning_rate": 9.998098999809901e-06, + "loss": 0.9322, + "step": 1902 + }, + { + "epoch": 0.8847047884704788, + "grad_norm": 7.382213592529297, + "learning_rate": 9.9980979998098e-06, + "loss": 0.8586, + "step": 1903 + }, + { + "epoch": 0.8851696885169689, + "grad_norm": 5.46842098236084, + "learning_rate": 9.9980969998097e-06, + "loss": 0.7217, + "step": 1904 + }, + { + "epoch": 0.8856345885634589, + "grad_norm": 6.970139503479004, + "learning_rate": 9.998095999809601e-06, + "loss": 0.8183, + "step": 1905 + }, + { + "epoch": 0.8860994886099489, + "grad_norm": 5.840770721435547, + "learning_rate": 9.998094999809502e-06, + "loss": 0.4639, + "step": 1906 + }, + { + "epoch": 0.8865643886564388, + "grad_norm": 7.8703413009643555, + "learning_rate": 9.998093999809401e-06, + "loss": 1.1739, + "step": 1907 + }, + { + "epoch": 0.8870292887029289, + "grad_norm": 7.1035356521606445, + "learning_rate": 9.998092999809302e-06, + "loss": 0.7627, + "step": 1908 + }, + { + "epoch": 0.8874941887494189, + "grad_norm": 3.846522569656372, + "learning_rate": 9.9980919998092e-06, + "loss": 0.3205, + "step": 1909 + }, + { + "epoch": 0.8879590887959089, + "grad_norm": 6.265820503234863, + "learning_rate": 9.9980909998091e-06, + "loss": 0.4939, + "step": 1910 + }, + { + "epoch": 0.8884239888423989, + "grad_norm": 9.085920333862305, + "learning_rate": 9.998089999809e-06, + "loss": 0.7295, + "step": 1911 + }, + { + "epoch": 0.8888888888888888, + "grad_norm": 5.760479927062988, + "learning_rate": 9.9980889998089e-06, + "loss": 0.4692, + "step": 1912 + }, + { + "epoch": 0.8893537889353789, + "grad_norm": 7.557728290557861, + "learning_rate": 9.998087999808802e-06, + "loss": 0.4448, + "step": 1913 + }, + { + "epoch": 0.8898186889818689, + "grad_norm": 8.720195770263672, + "learning_rate": 9.9980869998087e-06, + "loss": 0.5169, + "step": 1914 + }, + { + "epoch": 0.8902835890283589, + "grad_norm": 6.295958995819092, + "learning_rate": 9.998085999808601e-06, + "loss": 0.5146, + "step": 1915 + }, + { + "epoch": 0.8907484890748489, + "grad_norm": 7.038425922393799, + "learning_rate": 9.9980849998085e-06, + "loss": 0.8515, + "step": 1916 + }, + { + "epoch": 0.891213389121339, + "grad_norm": 7.059398651123047, + "learning_rate": 9.9980839998084e-06, + "loss": 0.7036, + "step": 1917 + }, + { + "epoch": 0.8916782891678289, + "grad_norm": 10.349127769470215, + "learning_rate": 9.998082999808301e-06, + "loss": 1.198, + "step": 1918 + }, + { + "epoch": 0.8921431892143189, + "grad_norm": 6.803155422210693, + "learning_rate": 9.998081999808202e-06, + "loss": 0.6337, + "step": 1919 + }, + { + "epoch": 0.8926080892608089, + "grad_norm": 6.109830856323242, + "learning_rate": 9.998080999808101e-06, + "loss": 0.6908, + "step": 1920 + }, + { + "epoch": 0.893072989307299, + "grad_norm": 5.925183296203613, + "learning_rate": 9.998079999808e-06, + "loss": 0.6654, + "step": 1921 + }, + { + "epoch": 0.893537889353789, + "grad_norm": 14.726906776428223, + "learning_rate": 9.9980789998079e-06, + "loss": 1.2916, + "step": 1922 + }, + { + "epoch": 0.8940027894002789, + "grad_norm": 8.004148483276367, + "learning_rate": 9.9980779998078e-06, + "loss": 0.7784, + "step": 1923 + }, + { + "epoch": 0.8944676894467689, + "grad_norm": 6.965612888336182, + "learning_rate": 9.9980769998077e-06, + "loss": 0.9041, + "step": 1924 + }, + { + "epoch": 0.8949325894932589, + "grad_norm": 4.831741809844971, + "learning_rate": 9.998075999807601e-06, + "loss": 0.3999, + "step": 1925 + }, + { + "epoch": 0.895397489539749, + "grad_norm": 5.202110290527344, + "learning_rate": 9.998074999807502e-06, + "loss": 0.6205, + "step": 1926 + }, + { + "epoch": 0.895862389586239, + "grad_norm": 6.872384071350098, + "learning_rate": 9.9980739998074e-06, + "loss": 0.618, + "step": 1927 + }, + { + "epoch": 0.896327289632729, + "grad_norm": 11.750117301940918, + "learning_rate": 9.998072999807301e-06, + "loss": 0.4718, + "step": 1928 + }, + { + "epoch": 0.8967921896792189, + "grad_norm": 5.555392265319824, + "learning_rate": 9.9980719998072e-06, + "loss": 0.6573, + "step": 1929 + }, + { + "epoch": 0.897257089725709, + "grad_norm": 5.481790065765381, + "learning_rate": 9.9980709998071e-06, + "loss": 0.5323, + "step": 1930 + }, + { + "epoch": 0.897721989772199, + "grad_norm": 8.21962833404541, + "learning_rate": 9.998069999807001e-06, + "loss": 1.0715, + "step": 1931 + }, + { + "epoch": 0.898186889818689, + "grad_norm": 7.443788528442383, + "learning_rate": 9.9980689998069e-06, + "loss": 0.6263, + "step": 1932 + }, + { + "epoch": 0.898651789865179, + "grad_norm": 8.97794246673584, + "learning_rate": 9.998067999806801e-06, + "loss": 0.7803, + "step": 1933 + }, + { + "epoch": 0.899116689911669, + "grad_norm": 9.306828498840332, + "learning_rate": 9.9980669998067e-06, + "loss": 0.9411, + "step": 1934 + }, + { + "epoch": 0.899581589958159, + "grad_norm": 12.560994148254395, + "learning_rate": 9.9980659998066e-06, + "loss": 0.5201, + "step": 1935 + }, + { + "epoch": 0.900046490004649, + "grad_norm": 7.80911111831665, + "learning_rate": 9.9980649998065e-06, + "loss": 0.7559, + "step": 1936 + }, + { + "epoch": 0.900511390051139, + "grad_norm": 6.1473164558410645, + "learning_rate": 9.9980639998064e-06, + "loss": 0.6578, + "step": 1937 + }, + { + "epoch": 0.900976290097629, + "grad_norm": 7.868904113769531, + "learning_rate": 9.998062999806301e-06, + "loss": 0.9784, + "step": 1938 + }, + { + "epoch": 0.901441190144119, + "grad_norm": 4.736620903015137, + "learning_rate": 9.998061999806202e-06, + "loss": 0.52, + "step": 1939 + }, + { + "epoch": 0.901906090190609, + "grad_norm": 5.796806335449219, + "learning_rate": 9.9980609998061e-06, + "loss": 0.6372, + "step": 1940 + }, + { + "epoch": 0.902370990237099, + "grad_norm": 4.638171672821045, + "learning_rate": 9.998059999806001e-06, + "loss": 0.4565, + "step": 1941 + }, + { + "epoch": 0.902835890283589, + "grad_norm": 5.392114639282227, + "learning_rate": 9.9980589998059e-06, + "loss": 0.4409, + "step": 1942 + }, + { + "epoch": 0.903300790330079, + "grad_norm": 4.595608234405518, + "learning_rate": 9.998057999805801e-06, + "loss": 0.4231, + "step": 1943 + }, + { + "epoch": 0.9037656903765691, + "grad_norm": 7.544233322143555, + "learning_rate": 9.998056999805702e-06, + "loss": 0.7563, + "step": 1944 + }, + { + "epoch": 0.9042305904230591, + "grad_norm": 7.4274725914001465, + "learning_rate": 9.9980559998056e-06, + "loss": 0.7455, + "step": 1945 + }, + { + "epoch": 0.904695490469549, + "grad_norm": 5.968679428100586, + "learning_rate": 9.998054999805501e-06, + "loss": 0.5479, + "step": 1946 + }, + { + "epoch": 0.905160390516039, + "grad_norm": 7.013175964355469, + "learning_rate": 9.9980539998054e-06, + "loss": 0.7021, + "step": 1947 + }, + { + "epoch": 0.905625290562529, + "grad_norm": 5.2601118087768555, + "learning_rate": 9.9980529998053e-06, + "loss": 0.7136, + "step": 1948 + }, + { + "epoch": 0.9060901906090191, + "grad_norm": 7.475469589233398, + "learning_rate": 9.9980519998052e-06, + "loss": 0.4627, + "step": 1949 + }, + { + "epoch": 0.9065550906555091, + "grad_norm": 7.0341973304748535, + "learning_rate": 9.9980509998051e-06, + "loss": 1.0782, + "step": 1950 + }, + { + "epoch": 0.9070199907019991, + "grad_norm": 5.816357612609863, + "learning_rate": 9.998049999805001e-06, + "loss": 0.8015, + "step": 1951 + }, + { + "epoch": 0.907484890748489, + "grad_norm": 7.409904479980469, + "learning_rate": 9.998048999804902e-06, + "loss": 0.5705, + "step": 1952 + }, + { + "epoch": 0.9079497907949791, + "grad_norm": 10.423155784606934, + "learning_rate": 9.9980479998048e-06, + "loss": 1.4992, + "step": 1953 + }, + { + "epoch": 0.9084146908414691, + "grad_norm": 5.664531707763672, + "learning_rate": 9.9980469998047e-06, + "loss": 0.3362, + "step": 1954 + }, + { + "epoch": 0.9088795908879591, + "grad_norm": 9.039563179016113, + "learning_rate": 9.9980459998046e-06, + "loss": 0.7579, + "step": 1955 + }, + { + "epoch": 0.9093444909344491, + "grad_norm": 7.758370876312256, + "learning_rate": 9.998044999804501e-06, + "loss": 0.9802, + "step": 1956 + }, + { + "epoch": 0.909809390980939, + "grad_norm": 7.565301418304443, + "learning_rate": 9.998043999804402e-06, + "loss": 0.5121, + "step": 1957 + }, + { + "epoch": 0.9102742910274291, + "grad_norm": 6.068579196929932, + "learning_rate": 9.9980429998043e-06, + "loss": 0.6714, + "step": 1958 + }, + { + "epoch": 0.9107391910739191, + "grad_norm": 4.802227020263672, + "learning_rate": 9.998041999804201e-06, + "loss": 0.4638, + "step": 1959 + }, + { + "epoch": 0.9112040911204091, + "grad_norm": 4.146444320678711, + "learning_rate": 9.9980409998041e-06, + "loss": 0.4352, + "step": 1960 + }, + { + "epoch": 0.9116689911668991, + "grad_norm": 5.856989860534668, + "learning_rate": 9.998039999804001e-06, + "loss": 0.5579, + "step": 1961 + }, + { + "epoch": 0.9121338912133892, + "grad_norm": 7.302022933959961, + "learning_rate": 9.9980389998039e-06, + "loss": 0.5702, + "step": 1962 + }, + { + "epoch": 0.9125987912598791, + "grad_norm": 7.059780597686768, + "learning_rate": 9.9980379998038e-06, + "loss": 0.7216, + "step": 1963 + }, + { + "epoch": 0.9130636913063691, + "grad_norm": 6.659173965454102, + "learning_rate": 9.998036999803701e-06, + "loss": 0.6032, + "step": 1964 + }, + { + "epoch": 0.9135285913528591, + "grad_norm": 4.85455846786499, + "learning_rate": 9.9980359998036e-06, + "loss": 0.3869, + "step": 1965 + }, + { + "epoch": 0.9139934913993492, + "grad_norm": 8.195603370666504, + "learning_rate": 9.9980349998035e-06, + "loss": 0.9981, + "step": 1966 + }, + { + "epoch": 0.9144583914458392, + "grad_norm": 8.06721019744873, + "learning_rate": 9.9980339998034e-06, + "loss": 0.7921, + "step": 1967 + }, + { + "epoch": 0.9149232914923291, + "grad_norm": 6.604793071746826, + "learning_rate": 9.9980329998033e-06, + "loss": 0.5284, + "step": 1968 + }, + { + "epoch": 0.9153881915388191, + "grad_norm": 6.847445011138916, + "learning_rate": 9.998031999803201e-06, + "loss": 0.8199, + "step": 1969 + }, + { + "epoch": 0.9158530915853091, + "grad_norm": 11.531997680664062, + "learning_rate": 9.998030999803102e-06, + "loss": 1.1321, + "step": 1970 + }, + { + "epoch": 0.9163179916317992, + "grad_norm": 6.201798439025879, + "learning_rate": 9.998029999803e-06, + "loss": 0.7137, + "step": 1971 + }, + { + "epoch": 0.9167828916782892, + "grad_norm": 5.264861106872559, + "learning_rate": 9.998028999802901e-06, + "loss": 0.6562, + "step": 1972 + }, + { + "epoch": 0.9172477917247792, + "grad_norm": 6.861499309539795, + "learning_rate": 9.9980279998028e-06, + "loss": 0.5801, + "step": 1973 + }, + { + "epoch": 0.9177126917712691, + "grad_norm": 6.857534885406494, + "learning_rate": 9.998026999802701e-06, + "loss": 0.7001, + "step": 1974 + }, + { + "epoch": 0.9181775918177592, + "grad_norm": 6.399560451507568, + "learning_rate": 9.998025999802602e-06, + "loss": 0.5979, + "step": 1975 + }, + { + "epoch": 0.9186424918642492, + "grad_norm": 6.352320671081543, + "learning_rate": 9.9980249998025e-06, + "loss": 0.6732, + "step": 1976 + }, + { + "epoch": 0.9191073919107392, + "grad_norm": 6.773539066314697, + "learning_rate": 9.998023999802401e-06, + "loss": 0.7021, + "step": 1977 + }, + { + "epoch": 0.9195722919572292, + "grad_norm": 9.680203437805176, + "learning_rate": 9.9980229998023e-06, + "loss": 0.6636, + "step": 1978 + }, + { + "epoch": 0.9200371920037193, + "grad_norm": 5.725888252258301, + "learning_rate": 9.998021999802201e-06, + "loss": 0.7961, + "step": 1979 + }, + { + "epoch": 0.9205020920502092, + "grad_norm": 4.942446231842041, + "learning_rate": 9.9980209998021e-06, + "loss": 0.5154, + "step": 1980 + }, + { + "epoch": 0.9209669920966992, + "grad_norm": 9.453999519348145, + "learning_rate": 9.998019999802e-06, + "loss": 0.8421, + "step": 1981 + }, + { + "epoch": 0.9214318921431892, + "grad_norm": 6.511327266693115, + "learning_rate": 9.998018999801901e-06, + "loss": 0.511, + "step": 1982 + }, + { + "epoch": 0.9218967921896792, + "grad_norm": 6.069790363311768, + "learning_rate": 9.998017999801802e-06, + "loss": 0.5521, + "step": 1983 + }, + { + "epoch": 0.9223616922361693, + "grad_norm": 6.180865287780762, + "learning_rate": 9.9980169998017e-06, + "loss": 0.4127, + "step": 1984 + }, + { + "epoch": 0.9228265922826592, + "grad_norm": 7.908862113952637, + "learning_rate": 9.998015999801601e-06, + "loss": 0.563, + "step": 1985 + }, + { + "epoch": 0.9232914923291492, + "grad_norm": 7.56239652633667, + "learning_rate": 9.9980149998015e-06, + "loss": 0.5163, + "step": 1986 + }, + { + "epoch": 0.9237563923756392, + "grad_norm": 8.487849235534668, + "learning_rate": 9.998013999801401e-06, + "loss": 0.5375, + "step": 1987 + }, + { + "epoch": 0.9242212924221292, + "grad_norm": 7.6108808517456055, + "learning_rate": 9.998012999801302e-06, + "loss": 0.3629, + "step": 1988 + }, + { + "epoch": 0.9246861924686193, + "grad_norm": 6.54315185546875, + "learning_rate": 9.9980119998012e-06, + "loss": 0.4262, + "step": 1989 + }, + { + "epoch": 0.9251510925151093, + "grad_norm": 5.446584701538086, + "learning_rate": 9.998010999801101e-06, + "loss": 0.6411, + "step": 1990 + }, + { + "epoch": 0.9256159925615992, + "grad_norm": 7.336849689483643, + "learning_rate": 9.998009999801e-06, + "loss": 0.7715, + "step": 1991 + }, + { + "epoch": 0.9260808926080892, + "grad_norm": 7.117382526397705, + "learning_rate": 9.998008999800901e-06, + "loss": 0.8998, + "step": 1992 + }, + { + "epoch": 0.9265457926545793, + "grad_norm": 9.497781753540039, + "learning_rate": 9.9980079998008e-06, + "loss": 0.5575, + "step": 1993 + }, + { + "epoch": 0.9270106927010693, + "grad_norm": 9.159769058227539, + "learning_rate": 9.9980069998007e-06, + "loss": 0.787, + "step": 1994 + }, + { + "epoch": 0.9274755927475593, + "grad_norm": 8.30476188659668, + "learning_rate": 9.998005999800601e-06, + "loss": 0.7392, + "step": 1995 + }, + { + "epoch": 0.9279404927940493, + "grad_norm": 6.413250923156738, + "learning_rate": 9.998004999800502e-06, + "loss": 0.604, + "step": 1996 + }, + { + "epoch": 0.9284053928405392, + "grad_norm": 5.179686546325684, + "learning_rate": 9.998003999800401e-06, + "loss": 0.458, + "step": 1997 + }, + { + "epoch": 0.9288702928870293, + "grad_norm": 8.256731986999512, + "learning_rate": 9.9980029998003e-06, + "loss": 0.7629, + "step": 1998 + }, + { + "epoch": 0.9293351929335193, + "grad_norm": 9.574298858642578, + "learning_rate": 9.9980019998002e-06, + "loss": 1.1696, + "step": 1999 + }, + { + "epoch": 0.9298000929800093, + "grad_norm": 4.935337543487549, + "learning_rate": 9.998000999800101e-06, + "loss": 0.5978, + "step": 2000 + }, + { + "epoch": 0.9302649930264993, + "grad_norm": 12.197901725769043, + "learning_rate": 9.997999999800002e-06, + "loss": 0.872, + "step": 2001 + }, + { + "epoch": 0.9307298930729893, + "grad_norm": 5.4915361404418945, + "learning_rate": 9.9979989997999e-06, + "loss": 0.4488, + "step": 2002 + }, + { + "epoch": 0.9311947931194793, + "grad_norm": 10.043405532836914, + "learning_rate": 9.997997999799801e-06, + "loss": 0.6908, + "step": 2003 + }, + { + "epoch": 0.9316596931659693, + "grad_norm": 9.672263145446777, + "learning_rate": 9.9979969997997e-06, + "loss": 0.7914, + "step": 2004 + }, + { + "epoch": 0.9321245932124593, + "grad_norm": 4.903146266937256, + "learning_rate": 9.997995999799601e-06, + "loss": 0.42, + "step": 2005 + }, + { + "epoch": 0.9325894932589494, + "grad_norm": 10.434723854064941, + "learning_rate": 9.9979949997995e-06, + "loss": 0.9622, + "step": 2006 + }, + { + "epoch": 0.9330543933054394, + "grad_norm": 10.919610023498535, + "learning_rate": 9.9979939997994e-06, + "loss": 0.9534, + "step": 2007 + }, + { + "epoch": 0.9335192933519293, + "grad_norm": 5.49722957611084, + "learning_rate": 9.997992999799301e-06, + "loss": 0.4971, + "step": 2008 + }, + { + "epoch": 0.9339841933984193, + "grad_norm": 7.6359944343566895, + "learning_rate": 9.9979919997992e-06, + "loss": 0.6721, + "step": 2009 + }, + { + "epoch": 0.9344490934449093, + "grad_norm": 4.077872276306152, + "learning_rate": 9.997990999799101e-06, + "loss": 0.2669, + "step": 2010 + }, + { + "epoch": 0.9349139934913994, + "grad_norm": 7.23444128036499, + "learning_rate": 9.997989999799e-06, + "loss": 0.5765, + "step": 2011 + }, + { + "epoch": 0.9353788935378894, + "grad_norm": 5.050586223602295, + "learning_rate": 9.9979889997989e-06, + "loss": 0.5308, + "step": 2012 + }, + { + "epoch": 0.9358437935843794, + "grad_norm": 8.193659782409668, + "learning_rate": 9.997987999798801e-06, + "loss": 0.7139, + "step": 2013 + }, + { + "epoch": 0.9363086936308693, + "grad_norm": 6.534438610076904, + "learning_rate": 9.997986999798702e-06, + "loss": 0.7633, + "step": 2014 + }, + { + "epoch": 0.9367735936773594, + "grad_norm": 5.7243170738220215, + "learning_rate": 9.9979859997986e-06, + "loss": 0.5166, + "step": 2015 + }, + { + "epoch": 0.9372384937238494, + "grad_norm": 4.841761589050293, + "learning_rate": 9.997984999798501e-06, + "loss": 0.3595, + "step": 2016 + }, + { + "epoch": 0.9377033937703394, + "grad_norm": 5.3186564445495605, + "learning_rate": 9.9979839997984e-06, + "loss": 0.4807, + "step": 2017 + }, + { + "epoch": 0.9381682938168294, + "grad_norm": 9.547323226928711, + "learning_rate": 9.997982999798301e-06, + "loss": 0.5587, + "step": 2018 + }, + { + "epoch": 0.9386331938633193, + "grad_norm": Infinity, + "learning_rate": 9.997982999798301e-06, + "loss": 0.5053, + "step": 2019 + }, + { + "epoch": 0.9390980939098094, + "grad_norm": 7.007751941680908, + "learning_rate": 9.9979819997982e-06, + "loss": 0.5571, + "step": 2020 + }, + { + "epoch": 0.9395629939562994, + "grad_norm": 6.0856242179870605, + "learning_rate": 9.9979809997981e-06, + "loss": 0.5827, + "step": 2021 + }, + { + "epoch": 0.9400278940027894, + "grad_norm": 6.738037586212158, + "learning_rate": 9.997979999798001e-06, + "loss": 0.6278, + "step": 2022 + }, + { + "epoch": 0.9404927940492794, + "grad_norm": 6.904698371887207, + "learning_rate": 9.9979789997979e-06, + "loss": 0.8361, + "step": 2023 + }, + { + "epoch": 0.9409576940957695, + "grad_norm": 5.530892848968506, + "learning_rate": 9.997977999797801e-06, + "loss": 0.4719, + "step": 2024 + }, + { + "epoch": 0.9414225941422594, + "grad_norm": 7.500560760498047, + "learning_rate": 9.9979769997977e-06, + "loss": 0.7671, + "step": 2025 + }, + { + "epoch": 0.9418874941887494, + "grad_norm": 7.641007423400879, + "learning_rate": 9.9979759997976e-06, + "loss": 0.841, + "step": 2026 + }, + { + "epoch": 0.9423523942352394, + "grad_norm": 5.354994773864746, + "learning_rate": 9.997974999797501e-06, + "loss": 0.5047, + "step": 2027 + }, + { + "epoch": 0.9428172942817294, + "grad_norm": 5.354684352874756, + "learning_rate": 9.997973999797402e-06, + "loss": 0.5972, + "step": 2028 + }, + { + "epoch": 0.9432821943282195, + "grad_norm": 6.05451774597168, + "learning_rate": 9.997972999797301e-06, + "loss": 0.5963, + "step": 2029 + }, + { + "epoch": 0.9437470943747094, + "grad_norm": 5.877597808837891, + "learning_rate": 9.997971999797202e-06, + "loss": 0.5961, + "step": 2030 + }, + { + "epoch": 0.9442119944211994, + "grad_norm": 6.589680194854736, + "learning_rate": 9.9979709997971e-06, + "loss": 0.4341, + "step": 2031 + }, + { + "epoch": 0.9446768944676894, + "grad_norm": 5.684527397155762, + "learning_rate": 9.997969999797e-06, + "loss": 0.5529, + "step": 2032 + }, + { + "epoch": 0.9451417945141795, + "grad_norm": 7.02030086517334, + "learning_rate": 9.9979689997969e-06, + "loss": 0.6508, + "step": 2033 + }, + { + "epoch": 0.9456066945606695, + "grad_norm": 7.9410624504089355, + "learning_rate": 9.9979679997968e-06, + "loss": 0.3307, + "step": 2034 + }, + { + "epoch": 0.9460715946071595, + "grad_norm": 9.72636604309082, + "learning_rate": 9.997966999796701e-06, + "loss": 0.9682, + "step": 2035 + }, + { + "epoch": 0.9465364946536494, + "grad_norm": 4.712926387786865, + "learning_rate": 9.9979659997966e-06, + "loss": 0.3772, + "step": 2036 + }, + { + "epoch": 0.9470013947001394, + "grad_norm": 5.219086647033691, + "learning_rate": 9.997964999796501e-06, + "loss": 0.5328, + "step": 2037 + }, + { + "epoch": 0.9474662947466295, + "grad_norm": 6.731646537780762, + "learning_rate": 9.9979639997964e-06, + "loss": 0.7561, + "step": 2038 + }, + { + "epoch": 0.9479311947931195, + "grad_norm": 9.596319198608398, + "learning_rate": 9.9979629997963e-06, + "loss": 0.7675, + "step": 2039 + }, + { + "epoch": 0.9483960948396095, + "grad_norm": 7.252112865447998, + "learning_rate": 9.997961999796201e-06, + "loss": 0.6647, + "step": 2040 + }, + { + "epoch": 0.9488609948860995, + "grad_norm": 8.437779426574707, + "learning_rate": 9.997960999796102e-06, + "loss": 0.8016, + "step": 2041 + }, + { + "epoch": 0.9493258949325895, + "grad_norm": 7.371647834777832, + "learning_rate": 9.997959999796001e-06, + "loss": 0.7652, + "step": 2042 + }, + { + "epoch": 0.9497907949790795, + "grad_norm": 12.383209228515625, + "learning_rate": 9.9979589997959e-06, + "loss": 0.9312, + "step": 2043 + }, + { + "epoch": 0.9502556950255695, + "grad_norm": 6.1172709465026855, + "learning_rate": 9.9979579997958e-06, + "loss": 0.5391, + "step": 2044 + }, + { + "epoch": 0.9507205950720595, + "grad_norm": 4.59469747543335, + "learning_rate": 9.997956999795701e-06, + "loss": 0.3729, + "step": 2045 + }, + { + "epoch": 0.9511854951185496, + "grad_norm": 5.602328777313232, + "learning_rate": 9.9979559997956e-06, + "loss": 0.4381, + "step": 2046 + }, + { + "epoch": 0.9516503951650395, + "grad_norm": 5.906530857086182, + "learning_rate": 9.997954999795501e-06, + "loss": 0.5827, + "step": 2047 + }, + { + "epoch": 0.9521152952115295, + "grad_norm": 5.49058723449707, + "learning_rate": 9.997953999795402e-06, + "loss": 0.3333, + "step": 2048 + }, + { + "epoch": 0.9525801952580195, + "grad_norm": 4.6076741218566895, + "learning_rate": 9.9979529997953e-06, + "loss": 0.2932, + "step": 2049 + }, + { + "epoch": 0.9530450953045095, + "grad_norm": 7.643979549407959, + "learning_rate": 9.997951999795201e-06, + "loss": 0.5641, + "step": 2050 + }, + { + "epoch": 0.9535099953509996, + "grad_norm": 7.4816508293151855, + "learning_rate": 9.9979509997951e-06, + "loss": 0.7562, + "step": 2051 + }, + { + "epoch": 0.9539748953974896, + "grad_norm": 4.874110221862793, + "learning_rate": 9.997949999795e-06, + "loss": 0.4729, + "step": 2052 + }, + { + "epoch": 0.9544397954439795, + "grad_norm": 8.057563781738281, + "learning_rate": 9.997948999794901e-06, + "loss": 1.032, + "step": 2053 + }, + { + "epoch": 0.9549046954904695, + "grad_norm": 4.937479496002197, + "learning_rate": 9.9979479997948e-06, + "loss": 0.5271, + "step": 2054 + }, + { + "epoch": 0.9553695955369595, + "grad_norm": 5.544114112854004, + "learning_rate": 9.997946999794701e-06, + "loss": 0.6828, + "step": 2055 + }, + { + "epoch": 0.9558344955834496, + "grad_norm": 9.312363624572754, + "learning_rate": 9.9979459997946e-06, + "loss": 0.9264, + "step": 2056 + }, + { + "epoch": 0.9562993956299396, + "grad_norm": 5.377913951873779, + "learning_rate": 9.9979449997945e-06, + "loss": 0.5272, + "step": 2057 + }, + { + "epoch": 0.9567642956764296, + "grad_norm": 6.578389644622803, + "learning_rate": 9.997943999794401e-06, + "loss": 0.6292, + "step": 2058 + }, + { + "epoch": 0.9572291957229195, + "grad_norm": 7.574567794799805, + "learning_rate": 9.9979429997943e-06, + "loss": 0.7687, + "step": 2059 + }, + { + "epoch": 0.9576940957694096, + "grad_norm": 8.263895034790039, + "learning_rate": 9.997941999794201e-06, + "loss": 0.9538, + "step": 2060 + }, + { + "epoch": 0.9581589958158996, + "grad_norm": 4.704774856567383, + "learning_rate": 9.997940999794102e-06, + "loss": 0.526, + "step": 2061 + }, + { + "epoch": 0.9586238958623896, + "grad_norm": 5.3664045333862305, + "learning_rate": 9.997939999794e-06, + "loss": 0.6036, + "step": 2062 + }, + { + "epoch": 0.9590887959088796, + "grad_norm": 6.030858516693115, + "learning_rate": 9.9979389997939e-06, + "loss": 0.6379, + "step": 2063 + }, + { + "epoch": 0.9595536959553695, + "grad_norm": 6.745009422302246, + "learning_rate": 9.9979379997938e-06, + "loss": 0.8101, + "step": 2064 + }, + { + "epoch": 0.9600185960018596, + "grad_norm": 5.627882957458496, + "learning_rate": 9.997936999793701e-06, + "loss": 0.6545, + "step": 2065 + }, + { + "epoch": 0.9604834960483496, + "grad_norm": 8.466575622558594, + "learning_rate": 9.997935999793602e-06, + "loss": 0.9762, + "step": 2066 + }, + { + "epoch": 0.9609483960948396, + "grad_norm": 5.905485153198242, + "learning_rate": 9.9979349997935e-06, + "loss": 0.462, + "step": 2067 + }, + { + "epoch": 0.9614132961413296, + "grad_norm": 6.150757789611816, + "learning_rate": 9.997933999793401e-06, + "loss": 0.7599, + "step": 2068 + }, + { + "epoch": 0.9618781961878197, + "grad_norm": 8.943175315856934, + "learning_rate": 9.9979329997933e-06, + "loss": 0.612, + "step": 2069 + }, + { + "epoch": 0.9623430962343096, + "grad_norm": 5.288191318511963, + "learning_rate": 9.9979319997932e-06, + "loss": 0.6353, + "step": 2070 + }, + { + "epoch": 0.9628079962807996, + "grad_norm": 7.4047980308532715, + "learning_rate": 9.997930999793101e-06, + "loss": 0.6921, + "step": 2071 + }, + { + "epoch": 0.9632728963272896, + "grad_norm": 6.997310161590576, + "learning_rate": 9.997929999793002e-06, + "loss": 0.6787, + "step": 2072 + }, + { + "epoch": 0.9637377963737797, + "grad_norm": 6.654079437255859, + "learning_rate": 9.997928999792901e-06, + "loss": 0.658, + "step": 2073 + }, + { + "epoch": 0.9642026964202697, + "grad_norm": 4.701853275299072, + "learning_rate": 9.9979279997928e-06, + "loss": 0.3668, + "step": 2074 + }, + { + "epoch": 0.9646675964667596, + "grad_norm": 11.473533630371094, + "learning_rate": 9.9979269997927e-06, + "loss": 1.169, + "step": 2075 + }, + { + "epoch": 0.9651324965132496, + "grad_norm": 5.619405269622803, + "learning_rate": 9.9979259997926e-06, + "loss": 0.525, + "step": 2076 + }, + { + "epoch": 0.9655973965597396, + "grad_norm": 8.055761337280273, + "learning_rate": 9.9979249997925e-06, + "loss": 0.8679, + "step": 2077 + }, + { + "epoch": 0.9660622966062297, + "grad_norm": 8.404850959777832, + "learning_rate": 9.997923999792401e-06, + "loss": 0.7089, + "step": 2078 + }, + { + "epoch": 0.9665271966527197, + "grad_norm": 6.816584587097168, + "learning_rate": 9.997922999792302e-06, + "loss": 0.836, + "step": 2079 + }, + { + "epoch": 0.9669920966992097, + "grad_norm": 3.5605084896087646, + "learning_rate": 9.9979219997922e-06, + "loss": 0.2384, + "step": 2080 + }, + { + "epoch": 0.9674569967456996, + "grad_norm": 12.788117408752441, + "learning_rate": 9.997920999792101e-06, + "loss": 1.1426, + "step": 2081 + }, + { + "epoch": 0.9679218967921897, + "grad_norm": 7.560024261474609, + "learning_rate": 9.997919999792e-06, + "loss": 0.8146, + "step": 2082 + }, + { + "epoch": 0.9683867968386797, + "grad_norm": 5.990693092346191, + "learning_rate": 9.997918999791901e-06, + "loss": 0.6514, + "step": 2083 + }, + { + "epoch": 0.9688516968851697, + "grad_norm": 8.610835075378418, + "learning_rate": 9.997917999791802e-06, + "loss": 0.7361, + "step": 2084 + }, + { + "epoch": 0.9693165969316597, + "grad_norm": 9.36592960357666, + "learning_rate": 9.9979169997917e-06, + "loss": 0.7166, + "step": 2085 + }, + { + "epoch": 0.9697814969781497, + "grad_norm": 5.146718978881836, + "learning_rate": 9.997915999791601e-06, + "loss": 0.74, + "step": 2086 + }, + { + "epoch": 0.9702463970246397, + "grad_norm": 4.923673152923584, + "learning_rate": 9.9979149997915e-06, + "loss": 0.6336, + "step": 2087 + }, + { + "epoch": 0.9707112970711297, + "grad_norm": 5.904451847076416, + "learning_rate": 9.9979139997914e-06, + "loss": 0.8298, + "step": 2088 + }, + { + "epoch": 0.9711761971176197, + "grad_norm": 4.555729866027832, + "learning_rate": 9.9979129997913e-06, + "loss": 0.4875, + "step": 2089 + }, + { + "epoch": 0.9716410971641097, + "grad_norm": 6.435046672821045, + "learning_rate": 9.9979119997912e-06, + "loss": 0.6175, + "step": 2090 + }, + { + "epoch": 0.9721059972105998, + "grad_norm": 5.150032997131348, + "learning_rate": 9.997910999791101e-06, + "loss": 0.5428, + "step": 2091 + }, + { + "epoch": 0.9725708972570897, + "grad_norm": 7.824830532073975, + "learning_rate": 9.997909999791002e-06, + "loss": 0.8167, + "step": 2092 + }, + { + "epoch": 0.9730357973035797, + "grad_norm": 7.743847846984863, + "learning_rate": 9.9979089997909e-06, + "loss": 0.7975, + "step": 2093 + }, + { + "epoch": 0.9735006973500697, + "grad_norm": 7.598583698272705, + "learning_rate": 9.997907999790801e-06, + "loss": 0.9491, + "step": 2094 + }, + { + "epoch": 0.9739655973965597, + "grad_norm": 4.582650184631348, + "learning_rate": 9.9979069997907e-06, + "loss": 0.4194, + "step": 2095 + }, + { + "epoch": 0.9744304974430498, + "grad_norm": 7.318029880523682, + "learning_rate": 9.997905999790601e-06, + "loss": 0.6409, + "step": 2096 + }, + { + "epoch": 0.9748953974895398, + "grad_norm": 7.07444953918457, + "learning_rate": 9.997904999790502e-06, + "loss": 0.8506, + "step": 2097 + }, + { + "epoch": 0.9753602975360297, + "grad_norm": 9.920256614685059, + "learning_rate": 9.9979039997904e-06, + "loss": 0.8545, + "step": 2098 + }, + { + "epoch": 0.9758251975825197, + "grad_norm": 3.2221672534942627, + "learning_rate": 9.997902999790301e-06, + "loss": 0.2617, + "step": 2099 + }, + { + "epoch": 0.9762900976290098, + "grad_norm": 4.643921852111816, + "learning_rate": 9.9979019997902e-06, + "loss": 0.3064, + "step": 2100 + }, + { + "epoch": 0.9767549976754998, + "grad_norm": 6.41953706741333, + "learning_rate": 9.9979009997901e-06, + "loss": 0.4171, + "step": 2101 + }, + { + "epoch": 0.9772198977219898, + "grad_norm": 9.536880493164062, + "learning_rate": 9.99789999979e-06, + "loss": 1.2337, + "step": 2102 + }, + { + "epoch": 0.9776847977684798, + "grad_norm": 6.889066219329834, + "learning_rate": 9.9978989997899e-06, + "loss": 0.4719, + "step": 2103 + }, + { + "epoch": 0.9781496978149697, + "grad_norm": 5.0573883056640625, + "learning_rate": 9.997897999789801e-06, + "loss": 0.6112, + "step": 2104 + }, + { + "epoch": 0.9786145978614598, + "grad_norm": 6.680283546447754, + "learning_rate": 9.997896999789702e-06, + "loss": 0.7792, + "step": 2105 + }, + { + "epoch": 0.9790794979079498, + "grad_norm": 6.474630355834961, + "learning_rate": 9.9978959997896e-06, + "loss": 0.4743, + "step": 2106 + }, + { + "epoch": 0.9795443979544398, + "grad_norm": 8.453758239746094, + "learning_rate": 9.9978949997895e-06, + "loss": 0.9196, + "step": 2107 + }, + { + "epoch": 0.9800092980009298, + "grad_norm": 6.733182430267334, + "learning_rate": 9.9978939997894e-06, + "loss": 0.7812, + "step": 2108 + }, + { + "epoch": 0.9804741980474198, + "grad_norm": 7.609188556671143, + "learning_rate": 9.997892999789301e-06, + "loss": 0.5282, + "step": 2109 + }, + { + "epoch": 0.9809390980939098, + "grad_norm": 8.26236629486084, + "learning_rate": 9.997891999789202e-06, + "loss": 0.6725, + "step": 2110 + }, + { + "epoch": 0.9814039981403998, + "grad_norm": 7.625654220581055, + "learning_rate": 9.9978909997891e-06, + "loss": 0.7323, + "step": 2111 + }, + { + "epoch": 0.9818688981868898, + "grad_norm": 5.952244281768799, + "learning_rate": 9.997889999789001e-06, + "loss": 0.4972, + "step": 2112 + }, + { + "epoch": 0.9823337982333799, + "grad_norm": 5.8398261070251465, + "learning_rate": 9.9978889997889e-06, + "loss": 0.5783, + "step": 2113 + }, + { + "epoch": 0.9827986982798699, + "grad_norm": 8.597421646118164, + "learning_rate": 9.997887999788801e-06, + "loss": 0.6533, + "step": 2114 + }, + { + "epoch": 0.9832635983263598, + "grad_norm": 9.784174919128418, + "learning_rate": 9.9978869997887e-06, + "loss": 1.0318, + "step": 2115 + }, + { + "epoch": 0.9837284983728498, + "grad_norm": 4.937501430511475, + "learning_rate": 9.9978859997886e-06, + "loss": 0.6368, + "step": 2116 + }, + { + "epoch": 0.9841933984193398, + "grad_norm": 8.724623680114746, + "learning_rate": 9.997884999788501e-06, + "loss": 0.7439, + "step": 2117 + }, + { + "epoch": 0.9846582984658299, + "grad_norm": 10.041658401489258, + "learning_rate": 9.9978839997884e-06, + "loss": 2.2638, + "step": 2118 + }, + { + "epoch": 0.9851231985123199, + "grad_norm": 5.903381824493408, + "learning_rate": 9.9978829997883e-06, + "loss": 0.6018, + "step": 2119 + }, + { + "epoch": 0.9855880985588099, + "grad_norm": 7.5454864501953125, + "learning_rate": 9.9978819997882e-06, + "loss": 0.6415, + "step": 2120 + }, + { + "epoch": 0.9860529986052998, + "grad_norm": 6.3772501945495605, + "learning_rate": 9.9978809997881e-06, + "loss": 0.6419, + "step": 2121 + }, + { + "epoch": 0.9865178986517898, + "grad_norm": 5.946975231170654, + "learning_rate": 9.997879999788001e-06, + "loss": 0.6665, + "step": 2122 + }, + { + "epoch": 0.9869827986982799, + "grad_norm": 12.249963760375977, + "learning_rate": 9.997878999787902e-06, + "loss": 0.7363, + "step": 2123 + }, + { + "epoch": 0.9874476987447699, + "grad_norm": 6.159754276275635, + "learning_rate": 9.9978779997878e-06, + "loss": 0.5898, + "step": 2124 + }, + { + "epoch": 0.9879125987912599, + "grad_norm": 4.877467632293701, + "learning_rate": 9.997876999787701e-06, + "loss": 0.3936, + "step": 2125 + }, + { + "epoch": 0.9883774988377498, + "grad_norm": 8.09626293182373, + "learning_rate": 9.9978759997876e-06, + "loss": 0.3863, + "step": 2126 + }, + { + "epoch": 0.9888423988842399, + "grad_norm": 5.988336563110352, + "learning_rate": 9.997874999787501e-06, + "loss": 0.6667, + "step": 2127 + }, + { + "epoch": 0.9893072989307299, + "grad_norm": 7.53723669052124, + "learning_rate": 9.9978739997874e-06, + "loss": 0.6196, + "step": 2128 + }, + { + "epoch": 0.9897721989772199, + "grad_norm": 8.824223518371582, + "learning_rate": 9.9978729997873e-06, + "loss": 0.8509, + "step": 2129 + }, + { + "epoch": 0.9902370990237099, + "grad_norm": 8.785229682922363, + "learning_rate": 9.997871999787201e-06, + "loss": 0.796, + "step": 2130 + }, + { + "epoch": 0.9907019990702, + "grad_norm": 4.904662132263184, + "learning_rate": 9.9978709997871e-06, + "loss": 0.4405, + "step": 2131 + }, + { + "epoch": 0.9911668991166899, + "grad_norm": 9.963423728942871, + "learning_rate": 9.997869999787001e-06, + "loss": 0.8185, + "step": 2132 + }, + { + "epoch": 0.9916317991631799, + "grad_norm": 5.94773530960083, + "learning_rate": 9.9978689997869e-06, + "loss": 0.5407, + "step": 2133 + }, + { + "epoch": 0.9920966992096699, + "grad_norm": 7.069963455200195, + "learning_rate": 9.9978679997868e-06, + "loss": 0.6216, + "step": 2134 + }, + { + "epoch": 0.9925615992561599, + "grad_norm": 11.746253967285156, + "learning_rate": 9.997866999786701e-06, + "loss": 1.0265, + "step": 2135 + }, + { + "epoch": 0.99302649930265, + "grad_norm": 5.648895740509033, + "learning_rate": 9.997865999786602e-06, + "loss": 0.4472, + "step": 2136 + }, + { + "epoch": 0.9934913993491399, + "grad_norm": 5.03840970993042, + "learning_rate": 9.9978649997865e-06, + "loss": 0.7008, + "step": 2137 + }, + { + "epoch": 0.9939562993956299, + "grad_norm": 7.756815433502197, + "learning_rate": 9.997863999786401e-06, + "loss": 0.8233, + "step": 2138 + }, + { + "epoch": 0.9944211994421199, + "grad_norm": 6.88035774230957, + "learning_rate": 9.9978629997863e-06, + "loss": 0.818, + "step": 2139 + }, + { + "epoch": 0.99488609948861, + "grad_norm": 5.054443359375, + "learning_rate": 9.997861999786201e-06, + "loss": 0.4823, + "step": 2140 + }, + { + "epoch": 0.9953509995351, + "grad_norm": 6.110260486602783, + "learning_rate": 9.997860999786102e-06, + "loss": 0.5988, + "step": 2141 + }, + { + "epoch": 0.99581589958159, + "grad_norm": 5.910305023193359, + "learning_rate": 9.997859999786e-06, + "loss": 0.4699, + "step": 2142 + }, + { + "epoch": 0.9962807996280799, + "grad_norm": 8.240856170654297, + "learning_rate": 9.997858999785901e-06, + "loss": 0.9363, + "step": 2143 + }, + { + "epoch": 0.9967456996745699, + "grad_norm": 4.635610580444336, + "learning_rate": 9.9978579997858e-06, + "loss": 0.434, + "step": 2144 + }, + { + "epoch": 0.99721059972106, + "grad_norm": 6.201440811157227, + "learning_rate": 9.997856999785701e-06, + "loss": 0.7266, + "step": 2145 + }, + { + "epoch": 0.99767549976755, + "grad_norm": 7.59931755065918, + "learning_rate": 9.9978559997856e-06, + "loss": 0.6327, + "step": 2146 + }, + { + "epoch": 0.99814039981404, + "grad_norm": 4.813102722167969, + "learning_rate": 9.9978549997855e-06, + "loss": 0.5466, + "step": 2147 + }, + { + "epoch": 0.99860529986053, + "grad_norm": 9.08767032623291, + "learning_rate": 9.997853999785401e-06, + "loss": 0.8082, + "step": 2148 + }, + { + "epoch": 0.99907019990702, + "grad_norm": 5.842503547668457, + "learning_rate": 9.997852999785302e-06, + "loss": 0.6071, + "step": 2149 + }, + { + "epoch": 0.99953509995351, + "grad_norm": 5.305007457733154, + "learning_rate": 9.997851999785201e-06, + "loss": 0.4524, + "step": 2150 + }, + { + "epoch": 1.0, + "grad_norm": 7.052679538726807, + "learning_rate": 9.9978509997851e-06, + "loss": 0.5589, + "step": 2151 + }, + { + "epoch": 1.00046490004649, + "grad_norm": 5.186915874481201, + "learning_rate": 9.997849999785e-06, + "loss": 0.6173, + "step": 2152 + }, + { + "epoch": 1.00092980009298, + "grad_norm": 4.275538444519043, + "learning_rate": 9.997848999784901e-06, + "loss": 0.4525, + "step": 2153 + }, + { + "epoch": 1.00139470013947, + "grad_norm": 5.872393608093262, + "learning_rate": 9.997847999784802e-06, + "loss": 0.6802, + "step": 2154 + }, + { + "epoch": 1.00185960018596, + "grad_norm": 5.586630821228027, + "learning_rate": 9.9978469997847e-06, + "loss": 0.5654, + "step": 2155 + }, + { + "epoch": 1.0023245002324501, + "grad_norm": 4.997726917266846, + "learning_rate": 9.997845999784601e-06, + "loss": 0.5322, + "step": 2156 + }, + { + "epoch": 1.00278940027894, + "grad_norm": 7.881160736083984, + "learning_rate": 9.9978449997845e-06, + "loss": 0.41, + "step": 2157 + }, + { + "epoch": 1.00325430032543, + "grad_norm": 5.584551811218262, + "learning_rate": 9.997843999784401e-06, + "loss": 0.5211, + "step": 2158 + }, + { + "epoch": 1.00371920037192, + "grad_norm": 4.925804615020752, + "learning_rate": 9.9978429997843e-06, + "loss": 0.5609, + "step": 2159 + }, + { + "epoch": 1.00418410041841, + "grad_norm": 12.069150924682617, + "learning_rate": 9.9978419997842e-06, + "loss": 0.4487, + "step": 2160 + }, + { + "epoch": 1.0046490004649, + "grad_norm": 6.653909206390381, + "learning_rate": 9.997840999784101e-06, + "loss": 0.7402, + "step": 2161 + }, + { + "epoch": 1.00511390051139, + "grad_norm": 8.425396919250488, + "learning_rate": 9.997839999784e-06, + "loss": 0.6274, + "step": 2162 + }, + { + "epoch": 1.00557880055788, + "grad_norm": 3.8707244396209717, + "learning_rate": 9.997838999783901e-06, + "loss": 0.3488, + "step": 2163 + }, + { + "epoch": 1.00604370060437, + "grad_norm": 6.142043590545654, + "learning_rate": 9.9978379997838e-06, + "loss": 0.6317, + "step": 2164 + }, + { + "epoch": 1.0065086006508601, + "grad_norm": 5.608409881591797, + "learning_rate": 9.9978369997837e-06, + "loss": 0.6897, + "step": 2165 + }, + { + "epoch": 1.0069735006973501, + "grad_norm": 6.403521537780762, + "learning_rate": 9.997835999783601e-06, + "loss": 0.5351, + "step": 2166 + }, + { + "epoch": 1.0074384007438402, + "grad_norm": 8.58456039428711, + "learning_rate": 9.997834999783502e-06, + "loss": 0.87, + "step": 2167 + }, + { + "epoch": 1.00790330079033, + "grad_norm": 7.016386032104492, + "learning_rate": 9.997833999783401e-06, + "loss": 0.5238, + "step": 2168 + }, + { + "epoch": 1.00836820083682, + "grad_norm": 5.250643730163574, + "learning_rate": 9.997832999783302e-06, + "loss": 0.4963, + "step": 2169 + }, + { + "epoch": 1.00883310088331, + "grad_norm": 4.470435619354248, + "learning_rate": 9.9978319997832e-06, + "loss": 0.6522, + "step": 2170 + }, + { + "epoch": 1.0092980009298, + "grad_norm": 4.227511882781982, + "learning_rate": 9.997830999783101e-06, + "loss": 0.3267, + "step": 2171 + }, + { + "epoch": 1.00976290097629, + "grad_norm": 9.263947486877441, + "learning_rate": 9.997829999783e-06, + "loss": 0.9124, + "step": 2172 + }, + { + "epoch": 1.01022780102278, + "grad_norm": 4.661016941070557, + "learning_rate": 9.9978289997829e-06, + "loss": 0.4057, + "step": 2173 + }, + { + "epoch": 1.0106927010692701, + "grad_norm": 6.7056121826171875, + "learning_rate": 9.997827999782801e-06, + "loss": 0.73, + "step": 2174 + }, + { + "epoch": 1.0111576011157601, + "grad_norm": 5.626008987426758, + "learning_rate": 9.9978269997827e-06, + "loss": 0.3801, + "step": 2175 + }, + { + "epoch": 1.0116225011622502, + "grad_norm": 6.346704959869385, + "learning_rate": 9.997825999782601e-06, + "loss": 0.6638, + "step": 2176 + }, + { + "epoch": 1.0120874012087402, + "grad_norm": 5.11220121383667, + "learning_rate": 9.9978249997825e-06, + "loss": 0.489, + "step": 2177 + }, + { + "epoch": 1.0125523012552302, + "grad_norm": 3.307112216949463, + "learning_rate": 9.9978239997824e-06, + "loss": 0.2674, + "step": 2178 + }, + { + "epoch": 1.0130172013017202, + "grad_norm": 7.293837547302246, + "learning_rate": 9.997822999782301e-06, + "loss": 0.7327, + "step": 2179 + }, + { + "epoch": 1.01348210134821, + "grad_norm": 6.325375080108643, + "learning_rate": 9.997821999782202e-06, + "loss": 0.5542, + "step": 2180 + }, + { + "epoch": 1.0139470013947, + "grad_norm": 5.508871078491211, + "learning_rate": 9.997820999782101e-06, + "loss": 0.5561, + "step": 2181 + }, + { + "epoch": 1.01441190144119, + "grad_norm": 6.485572338104248, + "learning_rate": 9.997819999782002e-06, + "loss": 0.524, + "step": 2182 + }, + { + "epoch": 1.0148768014876801, + "grad_norm": 6.053825378417969, + "learning_rate": 9.9978189997819e-06, + "loss": 0.496, + "step": 2183 + }, + { + "epoch": 1.0153417015341701, + "grad_norm": 4.631051063537598, + "learning_rate": 9.9978179997818e-06, + "loss": 0.4991, + "step": 2184 + }, + { + "epoch": 1.0158066015806602, + "grad_norm": 3.665053606033325, + "learning_rate": 9.9978169997817e-06, + "loss": 0.2947, + "step": 2185 + }, + { + "epoch": 1.0162715016271502, + "grad_norm": 3.7250192165374756, + "learning_rate": 9.997815999781601e-06, + "loss": 0.3512, + "step": 2186 + }, + { + "epoch": 1.0167364016736402, + "grad_norm": 4.022524356842041, + "learning_rate": 9.997814999781502e-06, + "loss": 0.3831, + "step": 2187 + }, + { + "epoch": 1.0172013017201302, + "grad_norm": 8.176545143127441, + "learning_rate": 9.9978139997814e-06, + "loss": 0.7769, + "step": 2188 + }, + { + "epoch": 1.0176662017666203, + "grad_norm": 6.526941299438477, + "learning_rate": 9.997812999781301e-06, + "loss": 0.5176, + "step": 2189 + }, + { + "epoch": 1.0181311018131103, + "grad_norm": 7.33513879776001, + "learning_rate": 9.9978119997812e-06, + "loss": 0.5665, + "step": 2190 + }, + { + "epoch": 1.0185960018596, + "grad_norm": 6.181335926055908, + "learning_rate": 9.9978109997811e-06, + "loss": 0.7226, + "step": 2191 + }, + { + "epoch": 1.01906090190609, + "grad_norm": 6.835940361022949, + "learning_rate": 9.997809999781001e-06, + "loss": 0.7464, + "step": 2192 + }, + { + "epoch": 1.0195258019525801, + "grad_norm": 4.405381202697754, + "learning_rate": 9.9978089997809e-06, + "loss": 0.5713, + "step": 2193 + }, + { + "epoch": 1.0199907019990702, + "grad_norm": 7.204245567321777, + "learning_rate": 9.997807999780801e-06, + "loss": 0.5859, + "step": 2194 + }, + { + "epoch": 1.0204556020455602, + "grad_norm": 5.357060432434082, + "learning_rate": 9.9978069997807e-06, + "loss": 0.5145, + "step": 2195 + }, + { + "epoch": 1.0209205020920502, + "grad_norm": 4.497661113739014, + "learning_rate": 9.9978059997806e-06, + "loss": 0.2409, + "step": 2196 + }, + { + "epoch": 1.0213854021385402, + "grad_norm": 6.06287145614624, + "learning_rate": 9.9978049997805e-06, + "loss": 0.4924, + "step": 2197 + }, + { + "epoch": 1.0218503021850303, + "grad_norm": 7.005344867706299, + "learning_rate": 9.9978039997804e-06, + "loss": 0.8394, + "step": 2198 + }, + { + "epoch": 1.0223152022315203, + "grad_norm": 6.115183353424072, + "learning_rate": 9.997802999780301e-06, + "loss": 0.6139, + "step": 2199 + }, + { + "epoch": 1.0227801022780103, + "grad_norm": 4.044814586639404, + "learning_rate": 9.997801999780202e-06, + "loss": 0.3409, + "step": 2200 + }, + { + "epoch": 1.0232450023245003, + "grad_norm": 3.616730213165283, + "learning_rate": 9.9978009997801e-06, + "loss": 0.1932, + "step": 2201 + }, + { + "epoch": 1.0237099023709901, + "grad_norm": 5.886814594268799, + "learning_rate": 9.997799999780001e-06, + "loss": 0.3934, + "step": 2202 + }, + { + "epoch": 1.0241748024174802, + "grad_norm": 4.333410739898682, + "learning_rate": 9.9977989997799e-06, + "loss": 0.3912, + "step": 2203 + }, + { + "epoch": 1.0246397024639702, + "grad_norm": 7.240286827087402, + "learning_rate": 9.9977979997798e-06, + "loss": 0.6581, + "step": 2204 + }, + { + "epoch": 1.0251046025104602, + "grad_norm": 5.6075310707092285, + "learning_rate": 9.997796999779701e-06, + "loss": 0.5381, + "step": 2205 + }, + { + "epoch": 1.0255695025569502, + "grad_norm": 6.54600715637207, + "learning_rate": 9.9977959997796e-06, + "loss": 0.5901, + "step": 2206 + }, + { + "epoch": 1.0260344026034403, + "grad_norm": 7.897469997406006, + "learning_rate": 9.997794999779501e-06, + "loss": 0.8173, + "step": 2207 + }, + { + "epoch": 1.0264993026499303, + "grad_norm": 7.370254993438721, + "learning_rate": 9.9977939997794e-06, + "loss": 0.4772, + "step": 2208 + }, + { + "epoch": 1.0269642026964203, + "grad_norm": 8.602293968200684, + "learning_rate": 9.9977929997793e-06, + "loss": 0.6242, + "step": 2209 + }, + { + "epoch": 1.0274291027429103, + "grad_norm": 5.367918491363525, + "learning_rate": 9.997791999779201e-06, + "loss": 0.7409, + "step": 2210 + }, + { + "epoch": 1.0278940027894004, + "grad_norm": 5.529574871063232, + "learning_rate": 9.9977909997791e-06, + "loss": 0.5336, + "step": 2211 + }, + { + "epoch": 1.0283589028358904, + "grad_norm": 6.171641826629639, + "learning_rate": 9.997789999779001e-06, + "loss": 0.4108, + "step": 2212 + }, + { + "epoch": 1.0288238028823802, + "grad_norm": 8.571629524230957, + "learning_rate": 9.997788999778902e-06, + "loss": 0.9111, + "step": 2213 + }, + { + "epoch": 1.0292887029288702, + "grad_norm": 11.971243858337402, + "learning_rate": 9.9977879997788e-06, + "loss": 1.2434, + "step": 2214 + }, + { + "epoch": 1.0297536029753602, + "grad_norm": 4.733326435089111, + "learning_rate": 9.9977869997787e-06, + "loss": 0.4849, + "step": 2215 + }, + { + "epoch": 1.0302185030218503, + "grad_norm": 9.91348648071289, + "learning_rate": 9.9977859997786e-06, + "loss": 0.8465, + "step": 2216 + }, + { + "epoch": 1.0306834030683403, + "grad_norm": 5.205984592437744, + "learning_rate": 9.997784999778501e-06, + "loss": 0.6143, + "step": 2217 + }, + { + "epoch": 1.0311483031148303, + "grad_norm": 5.461762428283691, + "learning_rate": 9.997783999778402e-06, + "loss": 0.5961, + "step": 2218 + }, + { + "epoch": 1.0316132031613203, + "grad_norm": 7.506382942199707, + "learning_rate": 9.9977829997783e-06, + "loss": 0.7592, + "step": 2219 + }, + { + "epoch": 1.0320781032078103, + "grad_norm": 6.706875801086426, + "learning_rate": 9.997781999778201e-06, + "loss": 0.5067, + "step": 2220 + }, + { + "epoch": 1.0325430032543004, + "grad_norm": 8.536949157714844, + "learning_rate": 9.9977809997781e-06, + "loss": 0.781, + "step": 2221 + }, + { + "epoch": 1.0330079033007904, + "grad_norm": 4.548848628997803, + "learning_rate": 9.997779999778e-06, + "loss": 0.4266, + "step": 2222 + }, + { + "epoch": 1.0334728033472804, + "grad_norm": 7.228769302368164, + "learning_rate": 9.997778999777901e-06, + "loss": 0.584, + "step": 2223 + }, + { + "epoch": 1.0339377033937704, + "grad_norm": 4.739604949951172, + "learning_rate": 9.9977779997778e-06, + "loss": 0.3379, + "step": 2224 + }, + { + "epoch": 1.0344026034402602, + "grad_norm": 5.044796466827393, + "learning_rate": 9.997776999777701e-06, + "loss": 0.4661, + "step": 2225 + }, + { + "epoch": 1.0348675034867503, + "grad_norm": 4.373696804046631, + "learning_rate": 9.9977759997776e-06, + "loss": 0.2413, + "step": 2226 + }, + { + "epoch": 1.0353324035332403, + "grad_norm": 6.8540825843811035, + "learning_rate": 9.9977749997775e-06, + "loss": 0.652, + "step": 2227 + }, + { + "epoch": 1.0357973035797303, + "grad_norm": 8.029919624328613, + "learning_rate": 9.9977739997774e-06, + "loss": 0.7922, + "step": 2228 + }, + { + "epoch": 1.0362622036262203, + "grad_norm": 4.1693644523620605, + "learning_rate": 9.9977729997773e-06, + "loss": 0.2955, + "step": 2229 + }, + { + "epoch": 1.0367271036727104, + "grad_norm": 6.092689514160156, + "learning_rate": 9.997771999777201e-06, + "loss": 0.5561, + "step": 2230 + }, + { + "epoch": 1.0371920037192004, + "grad_norm": 3.9274704456329346, + "learning_rate": 9.997770999777102e-06, + "loss": 0.3254, + "step": 2231 + }, + { + "epoch": 1.0376569037656904, + "grad_norm": 9.249741554260254, + "learning_rate": 9.997769999777e-06, + "loss": 0.7874, + "step": 2232 + }, + { + "epoch": 1.0381218038121804, + "grad_norm": 8.203621864318848, + "learning_rate": 9.997768999776901e-06, + "loss": 0.5484, + "step": 2233 + }, + { + "epoch": 1.0385867038586705, + "grad_norm": 8.543249130249023, + "learning_rate": 9.9977679997768e-06, + "loss": 0.5705, + "step": 2234 + }, + { + "epoch": 1.0390516039051605, + "grad_norm": 3.985335111618042, + "learning_rate": 9.997766999776701e-06, + "loss": 0.3187, + "step": 2235 + }, + { + "epoch": 1.0395165039516503, + "grad_norm": 4.449018955230713, + "learning_rate": 9.997765999776602e-06, + "loss": 0.4654, + "step": 2236 + }, + { + "epoch": 1.0399814039981403, + "grad_norm": 7.415290355682373, + "learning_rate": 9.9977649997765e-06, + "loss": 0.6227, + "step": 2237 + }, + { + "epoch": 1.0404463040446303, + "grad_norm": 7.643729209899902, + "learning_rate": 9.997763999776401e-06, + "loss": 0.4822, + "step": 2238 + }, + { + "epoch": 1.0409112040911204, + "grad_norm": 5.471475124359131, + "learning_rate": 9.9977629997763e-06, + "loss": 0.5163, + "step": 2239 + }, + { + "epoch": 1.0413761041376104, + "grad_norm": 5.770743370056152, + "learning_rate": 9.9977619997762e-06, + "loss": 0.5762, + "step": 2240 + }, + { + "epoch": 1.0418410041841004, + "grad_norm": 4.892403602600098, + "learning_rate": 9.9977609997761e-06, + "loss": 0.4577, + "step": 2241 + }, + { + "epoch": 1.0423059042305904, + "grad_norm": 5.398508548736572, + "learning_rate": 9.997759999776e-06, + "loss": 0.3126, + "step": 2242 + }, + { + "epoch": 1.0427708042770805, + "grad_norm": 7.33344841003418, + "learning_rate": 9.997758999775901e-06, + "loss": 0.5308, + "step": 2243 + }, + { + "epoch": 1.0432357043235705, + "grad_norm": 4.286957263946533, + "learning_rate": 9.997757999775802e-06, + "loss": 0.3221, + "step": 2244 + }, + { + "epoch": 1.0437006043700605, + "grad_norm": 4.584007740020752, + "learning_rate": 9.9977569997757e-06, + "loss": 0.4208, + "step": 2245 + }, + { + "epoch": 1.0441655044165505, + "grad_norm": 5.038762092590332, + "learning_rate": 9.997755999775601e-06, + "loss": 0.2997, + "step": 2246 + }, + { + "epoch": 1.0446304044630403, + "grad_norm": 5.932769298553467, + "learning_rate": 9.9977549997755e-06, + "loss": 0.4828, + "step": 2247 + }, + { + "epoch": 1.0450953045095304, + "grad_norm": 4.91826057434082, + "learning_rate": 9.997753999775401e-06, + "loss": 0.4664, + "step": 2248 + }, + { + "epoch": 1.0455602045560204, + "grad_norm": 6.987888813018799, + "learning_rate": 9.997752999775302e-06, + "loss": 0.8756, + "step": 2249 + }, + { + "epoch": 1.0460251046025104, + "grad_norm": 7.171616077423096, + "learning_rate": 9.9977519997752e-06, + "loss": 0.8037, + "step": 2250 + }, + { + "epoch": 1.0464900046490004, + "grad_norm": 5.801848888397217, + "learning_rate": 9.997750999775101e-06, + "loss": 0.4451, + "step": 2251 + }, + { + "epoch": 1.0469549046954905, + "grad_norm": 10.593640327453613, + "learning_rate": 9.997749999775e-06, + "loss": 0.9863, + "step": 2252 + }, + { + "epoch": 1.0474198047419805, + "grad_norm": 6.550298690795898, + "learning_rate": 9.997748999774901e-06, + "loss": 0.3527, + "step": 2253 + }, + { + "epoch": 1.0478847047884705, + "grad_norm": 11.999054908752441, + "learning_rate": 9.9977479997748e-06, + "loss": 0.3593, + "step": 2254 + }, + { + "epoch": 1.0483496048349605, + "grad_norm": 8.272343635559082, + "learning_rate": 9.9977469997747e-06, + "loss": 0.5757, + "step": 2255 + }, + { + "epoch": 1.0488145048814506, + "grad_norm": 5.403914451599121, + "learning_rate": 9.997745999774601e-06, + "loss": 0.4543, + "step": 2256 + }, + { + "epoch": 1.0492794049279406, + "grad_norm": 5.212682723999023, + "learning_rate": 9.997744999774502e-06, + "loss": 0.434, + "step": 2257 + }, + { + "epoch": 1.0497443049744306, + "grad_norm": 5.710812091827393, + "learning_rate": 9.9977439997744e-06, + "loss": 0.4982, + "step": 2258 + }, + { + "epoch": 1.0502092050209204, + "grad_norm": 5.348522663116455, + "learning_rate": 9.9977429997743e-06, + "loss": 0.4828, + "step": 2259 + }, + { + "epoch": 1.0506741050674104, + "grad_norm": 4.580734729766846, + "learning_rate": 9.9977419997742e-06, + "loss": 0.5099, + "step": 2260 + }, + { + "epoch": 1.0511390051139005, + "grad_norm": 6.070713996887207, + "learning_rate": 9.997740999774101e-06, + "loss": 0.5049, + "step": 2261 + }, + { + "epoch": 1.0516039051603905, + "grad_norm": 4.8380937576293945, + "learning_rate": 9.997739999774002e-06, + "loss": 0.4304, + "step": 2262 + }, + { + "epoch": 1.0520688052068805, + "grad_norm": 6.4368157386779785, + "learning_rate": 9.9977389997739e-06, + "loss": 0.5909, + "step": 2263 + }, + { + "epoch": 1.0525337052533705, + "grad_norm": 6.630300521850586, + "learning_rate": 9.997737999773801e-06, + "loss": 0.7272, + "step": 2264 + }, + { + "epoch": 1.0529986052998606, + "grad_norm": 7.029788494110107, + "learning_rate": 9.9977369997737e-06, + "loss": 1.0295, + "step": 2265 + }, + { + "epoch": 1.0534635053463506, + "grad_norm": 8.29638671875, + "learning_rate": 9.997735999773601e-06, + "loss": 0.8073, + "step": 2266 + }, + { + "epoch": 1.0539284053928406, + "grad_norm": 4.884097099304199, + "learning_rate": 9.9977349997735e-06, + "loss": 0.3173, + "step": 2267 + }, + { + "epoch": 1.0543933054393306, + "grad_norm": 4.829146862030029, + "learning_rate": 9.9977339997734e-06, + "loss": 0.4857, + "step": 2268 + }, + { + "epoch": 1.0548582054858207, + "grad_norm": 4.2324442863464355, + "learning_rate": 9.997732999773301e-06, + "loss": 0.2874, + "step": 2269 + }, + { + "epoch": 1.0553231055323105, + "grad_norm": 8.683405876159668, + "learning_rate": 9.9977319997732e-06, + "loss": 0.6518, + "step": 2270 + }, + { + "epoch": 1.0557880055788005, + "grad_norm": 7.322556018829346, + "learning_rate": 9.997730999773101e-06, + "loss": 0.6266, + "step": 2271 + }, + { + "epoch": 1.0562529056252905, + "grad_norm": 5.382328510284424, + "learning_rate": 9.997729999773e-06, + "loss": 0.5073, + "step": 2272 + }, + { + "epoch": 1.0567178056717805, + "grad_norm": 3.953864097595215, + "learning_rate": 9.9977289997729e-06, + "loss": 0.3029, + "step": 2273 + }, + { + "epoch": 1.0571827057182706, + "grad_norm": 5.8605217933654785, + "learning_rate": 9.997727999772801e-06, + "loss": 0.5845, + "step": 2274 + }, + { + "epoch": 1.0576476057647606, + "grad_norm": 7.972323417663574, + "learning_rate": 9.997726999772702e-06, + "loss": 0.6474, + "step": 2275 + }, + { + "epoch": 1.0581125058112506, + "grad_norm": 14.964778900146484, + "learning_rate": 9.9977259997726e-06, + "loss": 0.6398, + "step": 2276 + }, + { + "epoch": 1.0585774058577406, + "grad_norm": 3.568136215209961, + "learning_rate": 9.997724999772501e-06, + "loss": 0.2387, + "step": 2277 + }, + { + "epoch": 1.0590423059042307, + "grad_norm": 6.595362663269043, + "learning_rate": 9.9977239997724e-06, + "loss": 0.577, + "step": 2278 + }, + { + "epoch": 1.0595072059507207, + "grad_norm": 4.277549743652344, + "learning_rate": 9.997722999772301e-06, + "loss": 0.2959, + "step": 2279 + }, + { + "epoch": 1.0599721059972107, + "grad_norm": 6.539231777191162, + "learning_rate": 9.9977219997722e-06, + "loss": 0.4908, + "step": 2280 + }, + { + "epoch": 1.0604370060437005, + "grad_norm": 4.379379749298096, + "learning_rate": 9.9977209997721e-06, + "loss": 0.362, + "step": 2281 + }, + { + "epoch": 1.0609019060901905, + "grad_norm": 5.793797969818115, + "learning_rate": 9.997719999772001e-06, + "loss": 0.7637, + "step": 2282 + }, + { + "epoch": 1.0613668061366806, + "grad_norm": 4.091028213500977, + "learning_rate": 9.9977189997719e-06, + "loss": 0.3525, + "step": 2283 + }, + { + "epoch": 1.0618317061831706, + "grad_norm": 8.173943519592285, + "learning_rate": 9.997717999771801e-06, + "loss": 0.6136, + "step": 2284 + }, + { + "epoch": 1.0622966062296606, + "grad_norm": 6.104858875274658, + "learning_rate": 9.9977169997717e-06, + "loss": 0.8017, + "step": 2285 + }, + { + "epoch": 1.0627615062761506, + "grad_norm": 7.699484825134277, + "learning_rate": 9.9977159997716e-06, + "loss": 0.6826, + "step": 2286 + }, + { + "epoch": 1.0632264063226406, + "grad_norm": 4.417400360107422, + "learning_rate": 9.997714999771501e-06, + "loss": 0.4694, + "step": 2287 + }, + { + "epoch": 1.0636913063691307, + "grad_norm": 8.530783653259277, + "learning_rate": 9.997713999771402e-06, + "loss": 0.8366, + "step": 2288 + }, + { + "epoch": 1.0641562064156207, + "grad_norm": 4.765913963317871, + "learning_rate": 9.9977129997713e-06, + "loss": 0.2695, + "step": 2289 + }, + { + "epoch": 1.0646211064621107, + "grad_norm": 5.041229248046875, + "learning_rate": 9.997711999771201e-06, + "loss": 0.5338, + "step": 2290 + }, + { + "epoch": 1.0650860065086007, + "grad_norm": 5.887575149536133, + "learning_rate": 9.9977109997711e-06, + "loss": 0.733, + "step": 2291 + }, + { + "epoch": 1.0655509065550905, + "grad_norm": 6.35178279876709, + "learning_rate": 9.997709999771001e-06, + "loss": 0.627, + "step": 2292 + }, + { + "epoch": 1.0660158066015806, + "grad_norm": 5.635557651519775, + "learning_rate": 9.9977089997709e-06, + "loss": 0.6633, + "step": 2293 + }, + { + "epoch": 1.0664807066480706, + "grad_norm": 6.361604690551758, + "learning_rate": 9.9977079997708e-06, + "loss": 0.5483, + "step": 2294 + }, + { + "epoch": 1.0669456066945606, + "grad_norm": 6.01836633682251, + "learning_rate": 9.997706999770701e-06, + "loss": 0.6285, + "step": 2295 + }, + { + "epoch": 1.0674105067410506, + "grad_norm": 7.9159464836120605, + "learning_rate": 9.9977059997706e-06, + "loss": 0.6753, + "step": 2296 + }, + { + "epoch": 1.0678754067875407, + "grad_norm": 6.844576835632324, + "learning_rate": 9.997704999770501e-06, + "loss": 0.9334, + "step": 2297 + }, + { + "epoch": 1.0683403068340307, + "grad_norm": 6.470669269561768, + "learning_rate": 9.9977039997704e-06, + "loss": 0.3741, + "step": 2298 + }, + { + "epoch": 1.0688052068805207, + "grad_norm": 8.145240783691406, + "learning_rate": 9.9977029997703e-06, + "loss": 0.4335, + "step": 2299 + }, + { + "epoch": 1.0692701069270107, + "grad_norm": 5.874666690826416, + "learning_rate": 9.997701999770201e-06, + "loss": 0.6107, + "step": 2300 + }, + { + "epoch": 1.0697350069735008, + "grad_norm": 3.4745543003082275, + "learning_rate": 9.997700999770102e-06, + "loss": 0.4367, + "step": 2301 + }, + { + "epoch": 1.0701999070199908, + "grad_norm": 7.097092628479004, + "learning_rate": 9.997699999770001e-06, + "loss": 0.4653, + "step": 2302 + }, + { + "epoch": 1.0706648070664806, + "grad_norm": 5.692465782165527, + "learning_rate": 9.9976989997699e-06, + "loss": 0.405, + "step": 2303 + }, + { + "epoch": 1.0711297071129706, + "grad_norm": 4.269026756286621, + "learning_rate": 9.9976979997698e-06, + "loss": 0.3544, + "step": 2304 + }, + { + "epoch": 1.0715946071594606, + "grad_norm": 6.475755214691162, + "learning_rate": 9.997696999769701e-06, + "loss": 0.7498, + "step": 2305 + }, + { + "epoch": 1.0720595072059507, + "grad_norm": 6.278639793395996, + "learning_rate": 9.997695999769602e-06, + "loss": 0.5717, + "step": 2306 + }, + { + "epoch": 1.0725244072524407, + "grad_norm": 6.173007011413574, + "learning_rate": 9.9976949997695e-06, + "loss": 0.7751, + "step": 2307 + }, + { + "epoch": 1.0729893072989307, + "grad_norm": 5.830672740936279, + "learning_rate": 9.997693999769401e-06, + "loss": 0.7061, + "step": 2308 + }, + { + "epoch": 1.0734542073454207, + "grad_norm": 5.3299784660339355, + "learning_rate": 9.9976929997693e-06, + "loss": 0.3091, + "step": 2309 + }, + { + "epoch": 1.0739191073919108, + "grad_norm": 5.703367710113525, + "learning_rate": 9.997691999769201e-06, + "loss": 0.3596, + "step": 2310 + }, + { + "epoch": 1.0743840074384008, + "grad_norm": 6.301151275634766, + "learning_rate": 9.9976909997691e-06, + "loss": 0.6819, + "step": 2311 + }, + { + "epoch": 1.0748489074848908, + "grad_norm": 5.101972579956055, + "learning_rate": 9.997689999769e-06, + "loss": 0.3805, + "step": 2312 + }, + { + "epoch": 1.0753138075313808, + "grad_norm": 5.48438835144043, + "learning_rate": 9.997688999768901e-06, + "loss": 0.41, + "step": 2313 + }, + { + "epoch": 1.0757787075778706, + "grad_norm": 5.47978401184082, + "learning_rate": 9.9976879997688e-06, + "loss": 0.5973, + "step": 2314 + }, + { + "epoch": 1.0762436076243607, + "grad_norm": 4.578532695770264, + "learning_rate": 9.997686999768701e-06, + "loss": 0.3668, + "step": 2315 + }, + { + "epoch": 1.0767085076708507, + "grad_norm": 4.626312732696533, + "learning_rate": 9.9976859997686e-06, + "loss": 0.4048, + "step": 2316 + }, + { + "epoch": 1.0771734077173407, + "grad_norm": 5.047887325286865, + "learning_rate": 9.9976849997685e-06, + "loss": 0.4781, + "step": 2317 + }, + { + "epoch": 1.0776383077638307, + "grad_norm": 4.496944427490234, + "learning_rate": 9.997683999768401e-06, + "loss": 0.4223, + "step": 2318 + }, + { + "epoch": 1.0781032078103208, + "grad_norm": 7.993547439575195, + "learning_rate": 9.997682999768302e-06, + "loss": 0.704, + "step": 2319 + }, + { + "epoch": 1.0785681078568108, + "grad_norm": 5.688601016998291, + "learning_rate": 9.997681999768201e-06, + "loss": 0.3406, + "step": 2320 + }, + { + "epoch": 1.0790330079033008, + "grad_norm": 7.450418949127197, + "learning_rate": 9.997680999768102e-06, + "loss": 0.5588, + "step": 2321 + }, + { + "epoch": 1.0794979079497908, + "grad_norm": 4.981571674346924, + "learning_rate": 9.997679999768e-06, + "loss": 0.3262, + "step": 2322 + }, + { + "epoch": 1.0799628079962809, + "grad_norm": 6.094048500061035, + "learning_rate": 9.997678999767901e-06, + "loss": 0.6622, + "step": 2323 + }, + { + "epoch": 1.0804277080427709, + "grad_norm": 3.2390596866607666, + "learning_rate": 9.9976779997678e-06, + "loss": 0.2189, + "step": 2324 + }, + { + "epoch": 1.080892608089261, + "grad_norm": 7.533057689666748, + "learning_rate": 9.9976769997677e-06, + "loss": 1.0155, + "step": 2325 + }, + { + "epoch": 1.0813575081357507, + "grad_norm": 7.094898700714111, + "learning_rate": 9.997675999767601e-06, + "loss": 0.5331, + "step": 2326 + }, + { + "epoch": 1.0818224081822407, + "grad_norm": 7.812102317810059, + "learning_rate": 9.9976749997675e-06, + "loss": 0.8827, + "step": 2327 + }, + { + "epoch": 1.0822873082287308, + "grad_norm": 3.734771728515625, + "learning_rate": 9.997673999767401e-06, + "loss": 0.2689, + "step": 2328 + }, + { + "epoch": 1.0827522082752208, + "grad_norm": 13.061522483825684, + "learning_rate": 9.9976729997673e-06, + "loss": 0.5026, + "step": 2329 + }, + { + "epoch": 1.0832171083217108, + "grad_norm": 6.734863758087158, + "learning_rate": 9.9976719997672e-06, + "loss": 0.5361, + "step": 2330 + }, + { + "epoch": 1.0836820083682008, + "grad_norm": 7.291978359222412, + "learning_rate": 9.997670999767101e-06, + "loss": 0.7191, + "step": 2331 + }, + { + "epoch": 1.0841469084146909, + "grad_norm": 7.714723110198975, + "learning_rate": 9.997669999767002e-06, + "loss": 0.8604, + "step": 2332 + }, + { + "epoch": 1.0846118084611809, + "grad_norm": 4.330498218536377, + "learning_rate": 9.997668999766901e-06, + "loss": 0.3815, + "step": 2333 + }, + { + "epoch": 1.085076708507671, + "grad_norm": 5.0927839279174805, + "learning_rate": 9.9976679997668e-06, + "loss": 0.5101, + "step": 2334 + }, + { + "epoch": 1.085541608554161, + "grad_norm": 6.492030143737793, + "learning_rate": 9.9976669997667e-06, + "loss": 0.5098, + "step": 2335 + }, + { + "epoch": 1.086006508600651, + "grad_norm": 3.889798641204834, + "learning_rate": 9.9976659997666e-06, + "loss": 0.3334, + "step": 2336 + }, + { + "epoch": 1.086471408647141, + "grad_norm": 5.928637981414795, + "learning_rate": 9.9976649997665e-06, + "loss": 0.5941, + "step": 2337 + }, + { + "epoch": 1.0869363086936308, + "grad_norm": 2.892026424407959, + "learning_rate": 9.997663999766401e-06, + "loss": 0.1866, + "step": 2338 + }, + { + "epoch": 1.0874012087401208, + "grad_norm": 5.175277233123779, + "learning_rate": 9.997662999766302e-06, + "loss": 0.7561, + "step": 2339 + }, + { + "epoch": 1.0878661087866108, + "grad_norm": 6.264210224151611, + "learning_rate": 9.9976619997662e-06, + "loss": 0.4566, + "step": 2340 + }, + { + "epoch": 1.0883310088331009, + "grad_norm": 8.886448860168457, + "learning_rate": 9.997660999766101e-06, + "loss": 0.6213, + "step": 2341 + }, + { + "epoch": 1.0887959088795909, + "grad_norm": 6.125986099243164, + "learning_rate": 9.997659999766e-06, + "loss": 0.5351, + "step": 2342 + }, + { + "epoch": 1.089260808926081, + "grad_norm": 7.788107872009277, + "learning_rate": 9.9976589997659e-06, + "loss": 0.5902, + "step": 2343 + }, + { + "epoch": 1.089725708972571, + "grad_norm": 8.316930770874023, + "learning_rate": 9.997657999765801e-06, + "loss": 0.8615, + "step": 2344 + }, + { + "epoch": 1.090190609019061, + "grad_norm": 6.544632434844971, + "learning_rate": 9.9976569997657e-06, + "loss": 0.6002, + "step": 2345 + }, + { + "epoch": 1.090655509065551, + "grad_norm": 5.561204433441162, + "learning_rate": 9.997655999765601e-06, + "loss": 0.4765, + "step": 2346 + }, + { + "epoch": 1.091120409112041, + "grad_norm": 5.441018581390381, + "learning_rate": 9.9976549997655e-06, + "loss": 0.6644, + "step": 2347 + }, + { + "epoch": 1.091585309158531, + "grad_norm": 7.338642120361328, + "learning_rate": 9.9976539997654e-06, + "loss": 0.8373, + "step": 2348 + }, + { + "epoch": 1.0920502092050208, + "grad_norm": 5.757531642913818, + "learning_rate": 9.9976529997653e-06, + "loss": 0.6333, + "step": 2349 + }, + { + "epoch": 1.0925151092515109, + "grad_norm": 5.592671871185303, + "learning_rate": 9.9976519997652e-06, + "loss": 0.7216, + "step": 2350 + }, + { + "epoch": 1.0929800092980009, + "grad_norm": 8.644911766052246, + "learning_rate": 9.997650999765101e-06, + "loss": 0.6512, + "step": 2351 + }, + { + "epoch": 1.093444909344491, + "grad_norm": 10.899642944335938, + "learning_rate": 9.997649999765002e-06, + "loss": 1.0386, + "step": 2352 + }, + { + "epoch": 1.093909809390981, + "grad_norm": 7.8922505378723145, + "learning_rate": 9.9976489997649e-06, + "loss": 0.5525, + "step": 2353 + }, + { + "epoch": 1.094374709437471, + "grad_norm": 5.284404277801514, + "learning_rate": 9.997647999764801e-06, + "loss": 0.5433, + "step": 2354 + }, + { + "epoch": 1.094839609483961, + "grad_norm": 9.05160903930664, + "learning_rate": 9.9976469997647e-06, + "loss": 0.5232, + "step": 2355 + }, + { + "epoch": 1.095304509530451, + "grad_norm": 6.916476726531982, + "learning_rate": 9.997645999764601e-06, + "loss": 0.5992, + "step": 2356 + }, + { + "epoch": 1.095769409576941, + "grad_norm": 5.003836154937744, + "learning_rate": 9.997644999764502e-06, + "loss": 0.4588, + "step": 2357 + }, + { + "epoch": 1.096234309623431, + "grad_norm": 8.09193229675293, + "learning_rate": 9.9976439997644e-06, + "loss": 0.9172, + "step": 2358 + }, + { + "epoch": 1.096699209669921, + "grad_norm": 9.929859161376953, + "learning_rate": 9.997642999764301e-06, + "loss": 0.5277, + "step": 2359 + }, + { + "epoch": 1.0971641097164109, + "grad_norm": 7.427032470703125, + "learning_rate": 9.9976419997642e-06, + "loss": 0.7006, + "step": 2360 + }, + { + "epoch": 1.097629009762901, + "grad_norm": 16.31114387512207, + "learning_rate": 9.9976409997641e-06, + "loss": 0.8418, + "step": 2361 + }, + { + "epoch": 1.098093909809391, + "grad_norm": 6.62791109085083, + "learning_rate": 9.997639999764e-06, + "loss": 0.766, + "step": 2362 + }, + { + "epoch": 1.098558809855881, + "grad_norm": 5.81619119644165, + "learning_rate": 9.9976389997639e-06, + "loss": 0.6349, + "step": 2363 + }, + { + "epoch": 1.099023709902371, + "grad_norm": 6.840479373931885, + "learning_rate": 9.997637999763801e-06, + "loss": 0.7036, + "step": 2364 + }, + { + "epoch": 1.099488609948861, + "grad_norm": 6.7824625968933105, + "learning_rate": 9.997636999763702e-06, + "loss": 0.5563, + "step": 2365 + }, + { + "epoch": 1.099953509995351, + "grad_norm": 3.0745582580566406, + "learning_rate": 9.9976359997636e-06, + "loss": 0.2268, + "step": 2366 + }, + { + "epoch": 1.100418410041841, + "grad_norm": 7.302048683166504, + "learning_rate": 9.9976349997635e-06, + "loss": 0.3728, + "step": 2367 + }, + { + "epoch": 1.100883310088331, + "grad_norm": 5.961839199066162, + "learning_rate": 9.9976339997634e-06, + "loss": 0.5676, + "step": 2368 + }, + { + "epoch": 1.101348210134821, + "grad_norm": 6.042909622192383, + "learning_rate": 9.997632999763301e-06, + "loss": 0.6296, + "step": 2369 + }, + { + "epoch": 1.1018131101813111, + "grad_norm": 15.187084197998047, + "learning_rate": 9.997631999763202e-06, + "loss": 0.4442, + "step": 2370 + }, + { + "epoch": 1.102278010227801, + "grad_norm": 10.77792739868164, + "learning_rate": 9.9976309997631e-06, + "loss": 0.5531, + "step": 2371 + }, + { + "epoch": 1.102742910274291, + "grad_norm": 5.503527641296387, + "learning_rate": 9.997629999763001e-06, + "loss": 0.5807, + "step": 2372 + }, + { + "epoch": 1.103207810320781, + "grad_norm": 7.572511672973633, + "learning_rate": 9.9976289997629e-06, + "loss": 0.6575, + "step": 2373 + }, + { + "epoch": 1.103672710367271, + "grad_norm": 6.279473781585693, + "learning_rate": 9.9976279997628e-06, + "loss": 0.6393, + "step": 2374 + }, + { + "epoch": 1.104137610413761, + "grad_norm": 4.920038223266602, + "learning_rate": 9.997626999762702e-06, + "loss": 0.4856, + "step": 2375 + }, + { + "epoch": 1.104602510460251, + "grad_norm": 6.005365371704102, + "learning_rate": 9.9976259997626e-06, + "loss": 0.6243, + "step": 2376 + }, + { + "epoch": 1.105067410506741, + "grad_norm": 7.314875602722168, + "learning_rate": 9.997624999762501e-06, + "loss": 0.956, + "step": 2377 + }, + { + "epoch": 1.105532310553231, + "grad_norm": 5.971256732940674, + "learning_rate": 9.9976239997624e-06, + "loss": 0.5948, + "step": 2378 + }, + { + "epoch": 1.1059972105997211, + "grad_norm": 4.619717597961426, + "learning_rate": 9.9976229997623e-06, + "loss": 0.3372, + "step": 2379 + }, + { + "epoch": 1.1064621106462111, + "grad_norm": 5.050589561462402, + "learning_rate": 9.9976219997622e-06, + "loss": 0.5125, + "step": 2380 + }, + { + "epoch": 1.1069270106927012, + "grad_norm": 7.593270778656006, + "learning_rate": 9.9976209997621e-06, + "loss": 1.1726, + "step": 2381 + }, + { + "epoch": 1.107391910739191, + "grad_norm": 4.727348804473877, + "learning_rate": 9.997619999762001e-06, + "loss": 0.432, + "step": 2382 + }, + { + "epoch": 1.107856810785681, + "grad_norm": 5.188518524169922, + "learning_rate": 9.997618999761902e-06, + "loss": 0.3218, + "step": 2383 + }, + { + "epoch": 1.108321710832171, + "grad_norm": 5.24407958984375, + "learning_rate": 9.9976179997618e-06, + "loss": 0.5395, + "step": 2384 + }, + { + "epoch": 1.108786610878661, + "grad_norm": 5.395861625671387, + "learning_rate": 9.997616999761701e-06, + "loss": 0.5683, + "step": 2385 + }, + { + "epoch": 1.109251510925151, + "grad_norm": 5.336498737335205, + "learning_rate": 9.9976159997616e-06, + "loss": 0.6137, + "step": 2386 + }, + { + "epoch": 1.109716410971641, + "grad_norm": 9.492446899414062, + "learning_rate": 9.997614999761501e-06, + "loss": 0.4125, + "step": 2387 + }, + { + "epoch": 1.1101813110181311, + "grad_norm": 9.941079139709473, + "learning_rate": 9.997613999761402e-06, + "loss": 0.9642, + "step": 2388 + }, + { + "epoch": 1.1106462110646211, + "grad_norm": 7.675682067871094, + "learning_rate": 9.9976129997613e-06, + "loss": 0.659, + "step": 2389 + }, + { + "epoch": 1.1111111111111112, + "grad_norm": 7.950647830963135, + "learning_rate": 9.997611999761201e-06, + "loss": 0.7493, + "step": 2390 + }, + { + "epoch": 1.1115760111576012, + "grad_norm": 6.656844615936279, + "learning_rate": 9.9976109997611e-06, + "loss": 0.6096, + "step": 2391 + }, + { + "epoch": 1.1120409112040912, + "grad_norm": 9.552692413330078, + "learning_rate": 9.997609999761e-06, + "loss": 0.6952, + "step": 2392 + }, + { + "epoch": 1.112505811250581, + "grad_norm": 6.339468002319336, + "learning_rate": 9.9976089997609e-06, + "loss": 0.6298, + "step": 2393 + }, + { + "epoch": 1.112970711297071, + "grad_norm": 7.95170783996582, + "learning_rate": 9.9976079997608e-06, + "loss": 0.5884, + "step": 2394 + }, + { + "epoch": 1.113435611343561, + "grad_norm": 5.963256359100342, + "learning_rate": 9.997606999760701e-06, + "loss": 0.5483, + "step": 2395 + }, + { + "epoch": 1.113900511390051, + "grad_norm": 8.001626014709473, + "learning_rate": 9.997605999760602e-06, + "loss": 0.517, + "step": 2396 + }, + { + "epoch": 1.114365411436541, + "grad_norm": 4.469066619873047, + "learning_rate": 9.9976049997605e-06, + "loss": 0.3751, + "step": 2397 + }, + { + "epoch": 1.1148303114830311, + "grad_norm": 5.572115421295166, + "learning_rate": 9.997603999760401e-06, + "loss": 0.6878, + "step": 2398 + }, + { + "epoch": 1.1152952115295212, + "grad_norm": 7.23897647857666, + "learning_rate": 9.9976029997603e-06, + "loss": 0.8158, + "step": 2399 + }, + { + "epoch": 1.1157601115760112, + "grad_norm": 12.079607009887695, + "learning_rate": 9.997601999760201e-06, + "loss": 0.2505, + "step": 2400 + }, + { + "epoch": 1.1162250116225012, + "grad_norm": 9.248923301696777, + "learning_rate": 9.997600999760102e-06, + "loss": 0.3705, + "step": 2401 + }, + { + "epoch": 1.1166899116689912, + "grad_norm": 3.400501012802124, + "learning_rate": 9.99759999976e-06, + "loss": 0.3358, + "step": 2402 + }, + { + "epoch": 1.1171548117154813, + "grad_norm": 4.259141445159912, + "learning_rate": 9.997598999759901e-06, + "loss": 0.4663, + "step": 2403 + }, + { + "epoch": 1.1176197117619713, + "grad_norm": 3.5217442512512207, + "learning_rate": 9.9975979997598e-06, + "loss": 0.1911, + "step": 2404 + }, + { + "epoch": 1.118084611808461, + "grad_norm": 5.526280403137207, + "learning_rate": 9.997596999759701e-06, + "loss": 0.4514, + "step": 2405 + }, + { + "epoch": 1.118549511854951, + "grad_norm": 6.3075947761535645, + "learning_rate": 9.9975959997596e-06, + "loss": 0.5822, + "step": 2406 + }, + { + "epoch": 1.1190144119014411, + "grad_norm": 5.2327985763549805, + "learning_rate": 9.9975949997595e-06, + "loss": 0.3492, + "step": 2407 + }, + { + "epoch": 1.1194793119479312, + "grad_norm": 5.294242858886719, + "learning_rate": 9.997593999759401e-06, + "loss": 0.5117, + "step": 2408 + }, + { + "epoch": 1.1199442119944212, + "grad_norm": 6.59063196182251, + "learning_rate": 9.997592999759302e-06, + "loss": 0.5791, + "step": 2409 + }, + { + "epoch": 1.1204091120409112, + "grad_norm": 5.626022815704346, + "learning_rate": 9.9975919997592e-06, + "loss": 0.8827, + "step": 2410 + }, + { + "epoch": 1.1208740120874012, + "grad_norm": 3.4629194736480713, + "learning_rate": 9.9975909997591e-06, + "loss": 0.2369, + "step": 2411 + }, + { + "epoch": 1.1213389121338913, + "grad_norm": 5.828598976135254, + "learning_rate": 9.997589999759e-06, + "loss": 0.4761, + "step": 2412 + }, + { + "epoch": 1.1218038121803813, + "grad_norm": 7.097200870513916, + "learning_rate": 9.997588999758901e-06, + "loss": 0.791, + "step": 2413 + }, + { + "epoch": 1.1222687122268713, + "grad_norm": 8.267175674438477, + "learning_rate": 9.997587999758802e-06, + "loss": 0.9895, + "step": 2414 + }, + { + "epoch": 1.1227336122733613, + "grad_norm": 5.854556083679199, + "learning_rate": 9.9975869997587e-06, + "loss": 0.4794, + "step": 2415 + }, + { + "epoch": 1.1231985123198513, + "grad_norm": 4.559449195861816, + "learning_rate": 9.997585999758601e-06, + "loss": 0.4363, + "step": 2416 + }, + { + "epoch": 1.1236634123663412, + "grad_norm": 5.598221302032471, + "learning_rate": 9.9975849997585e-06, + "loss": 0.4089, + "step": 2417 + }, + { + "epoch": 1.1241283124128312, + "grad_norm": 4.680405616760254, + "learning_rate": 9.997583999758401e-06, + "loss": 0.4706, + "step": 2418 + }, + { + "epoch": 1.1245932124593212, + "grad_norm": 10.449075698852539, + "learning_rate": 9.9975829997583e-06, + "loss": 0.9998, + "step": 2419 + }, + { + "epoch": 1.1250581125058112, + "grad_norm": 10.452759742736816, + "learning_rate": 9.9975819997582e-06, + "loss": 0.6626, + "step": 2420 + }, + { + "epoch": 1.1255230125523012, + "grad_norm": 6.471254825592041, + "learning_rate": 9.997580999758101e-06, + "loss": 0.4675, + "step": 2421 + }, + { + "epoch": 1.1259879125987913, + "grad_norm": 5.976332187652588, + "learning_rate": 9.997579999758e-06, + "loss": 0.6157, + "step": 2422 + }, + { + "epoch": 1.1264528126452813, + "grad_norm": 8.749670028686523, + "learning_rate": 9.997578999757901e-06, + "loss": 0.9315, + "step": 2423 + }, + { + "epoch": 1.1269177126917713, + "grad_norm": 6.718826770782471, + "learning_rate": 9.9975779997578e-06, + "loss": 0.852, + "step": 2424 + }, + { + "epoch": 1.1273826127382613, + "grad_norm": 12.042798042297363, + "learning_rate": 9.9975769997577e-06, + "loss": 0.5293, + "step": 2425 + }, + { + "epoch": 1.1278475127847514, + "grad_norm": 6.306645393371582, + "learning_rate": 9.997575999757601e-06, + "loss": 0.4311, + "step": 2426 + }, + { + "epoch": 1.1283124128312414, + "grad_norm": 4.935123920440674, + "learning_rate": 9.997574999757502e-06, + "loss": 0.3028, + "step": 2427 + }, + { + "epoch": 1.1287773128777312, + "grad_norm": 6.784600734710693, + "learning_rate": 9.9975739997574e-06, + "loss": 0.6833, + "step": 2428 + }, + { + "epoch": 1.1292422129242212, + "grad_norm": 5.996852397918701, + "learning_rate": 9.997572999757301e-06, + "loss": 0.5909, + "step": 2429 + }, + { + "epoch": 1.1297071129707112, + "grad_norm": 6.1175947189331055, + "learning_rate": 9.9975719997572e-06, + "loss": 0.5384, + "step": 2430 + }, + { + "epoch": 1.1301720130172013, + "grad_norm": 4.49419641494751, + "learning_rate": 9.997570999757101e-06, + "loss": 0.5822, + "step": 2431 + }, + { + "epoch": 1.1306369130636913, + "grad_norm": 8.505579948425293, + "learning_rate": 9.997569999757e-06, + "loss": 0.7652, + "step": 2432 + }, + { + "epoch": 1.1311018131101813, + "grad_norm": 5.98047399520874, + "learning_rate": 9.9975689997569e-06, + "loss": 0.4717, + "step": 2433 + }, + { + "epoch": 1.1315667131566713, + "grad_norm": 5.392478942871094, + "learning_rate": 9.997567999756801e-06, + "loss": 0.6675, + "step": 2434 + }, + { + "epoch": 1.1320316132031614, + "grad_norm": 10.09037971496582, + "learning_rate": 9.9975669997567e-06, + "loss": 0.9708, + "step": 2435 + }, + { + "epoch": 1.1324965132496514, + "grad_norm": 3.4246058464050293, + "learning_rate": 9.997565999756601e-06, + "loss": 0.196, + "step": 2436 + }, + { + "epoch": 1.1329614132961414, + "grad_norm": 6.357741355895996, + "learning_rate": 9.9975649997565e-06, + "loss": 0.4467, + "step": 2437 + }, + { + "epoch": 1.1334263133426314, + "grad_norm": 7.053163051605225, + "learning_rate": 9.9975639997564e-06, + "loss": 0.5856, + "step": 2438 + }, + { + "epoch": 1.1338912133891212, + "grad_norm": 3.69844126701355, + "learning_rate": 9.997562999756301e-06, + "loss": 0.3054, + "step": 2439 + }, + { + "epoch": 1.1343561134356113, + "grad_norm": 4.546398639678955, + "learning_rate": 9.997561999756202e-06, + "loss": 0.2472, + "step": 2440 + }, + { + "epoch": 1.1348210134821013, + "grad_norm": 7.560128211975098, + "learning_rate": 9.997560999756101e-06, + "loss": 0.8764, + "step": 2441 + }, + { + "epoch": 1.1352859135285913, + "grad_norm": 2.8547205924987793, + "learning_rate": 9.997559999756002e-06, + "loss": 0.1828, + "step": 2442 + }, + { + "epoch": 1.1357508135750813, + "grad_norm": 7.076803207397461, + "learning_rate": 9.9975589997559e-06, + "loss": 0.6682, + "step": 2443 + }, + { + "epoch": 1.1362157136215714, + "grad_norm": 8.992269515991211, + "learning_rate": 9.997557999755801e-06, + "loss": 1.2475, + "step": 2444 + }, + { + "epoch": 1.1366806136680614, + "grad_norm": 7.488985061645508, + "learning_rate": 9.9975569997557e-06, + "loss": 0.6162, + "step": 2445 + }, + { + "epoch": 1.1371455137145514, + "grad_norm": 7.7775750160217285, + "learning_rate": 9.9975559997556e-06, + "loss": 0.8968, + "step": 2446 + }, + { + "epoch": 1.1376104137610414, + "grad_norm": 7.40713357925415, + "learning_rate": 9.997554999755501e-06, + "loss": 0.8984, + "step": 2447 + }, + { + "epoch": 1.1380753138075315, + "grad_norm": 6.582225799560547, + "learning_rate": 9.9975539997554e-06, + "loss": 0.682, + "step": 2448 + }, + { + "epoch": 1.1385402138540215, + "grad_norm": 9.653417587280273, + "learning_rate": 9.997552999755301e-06, + "loss": 0.7756, + "step": 2449 + }, + { + "epoch": 1.1390051139005113, + "grad_norm": 5.065432548522949, + "learning_rate": 9.9975519997552e-06, + "loss": 0.4486, + "step": 2450 + }, + { + "epoch": 1.1394700139470013, + "grad_norm": 5.023778915405273, + "learning_rate": 9.9975509997551e-06, + "loss": 0.5368, + "step": 2451 + }, + { + "epoch": 1.1399349139934913, + "grad_norm": 5.1837477684021, + "learning_rate": 9.997549999755001e-06, + "loss": 0.4511, + "step": 2452 + }, + { + "epoch": 1.1403998140399814, + "grad_norm": 8.20516586303711, + "learning_rate": 9.997548999754902e-06, + "loss": 0.5091, + "step": 2453 + }, + { + "epoch": 1.1408647140864714, + "grad_norm": 11.929564476013184, + "learning_rate": 9.997547999754801e-06, + "loss": 0.535, + "step": 2454 + }, + { + "epoch": 1.1413296141329614, + "grad_norm": 6.418271541595459, + "learning_rate": 9.9975469997547e-06, + "loss": 0.732, + "step": 2455 + }, + { + "epoch": 1.1417945141794514, + "grad_norm": 7.4072265625, + "learning_rate": 9.9975459997546e-06, + "loss": 0.7672, + "step": 2456 + }, + { + "epoch": 1.1422594142259415, + "grad_norm": 6.101611137390137, + "learning_rate": 9.997544999754501e-06, + "loss": 0.4136, + "step": 2457 + }, + { + "epoch": 1.1427243142724315, + "grad_norm": 6.036932945251465, + "learning_rate": 9.9975439997544e-06, + "loss": 0.7251, + "step": 2458 + }, + { + "epoch": 1.1431892143189215, + "grad_norm": 4.998466491699219, + "learning_rate": 9.997542999754301e-06, + "loss": 0.4438, + "step": 2459 + }, + { + "epoch": 1.1436541143654115, + "grad_norm": 5.829908847808838, + "learning_rate": 9.997541999754202e-06, + "loss": 0.4434, + "step": 2460 + }, + { + "epoch": 1.1441190144119013, + "grad_norm": 4.8318328857421875, + "learning_rate": 9.9975409997541e-06, + "loss": 0.5749, + "step": 2461 + }, + { + "epoch": 1.1445839144583914, + "grad_norm": 4.405141353607178, + "learning_rate": 9.997539999754001e-06, + "loss": 0.5042, + "step": 2462 + }, + { + "epoch": 1.1450488145048814, + "grad_norm": 4.527180194854736, + "learning_rate": 9.9975389997539e-06, + "loss": 0.5063, + "step": 2463 + }, + { + "epoch": 1.1455137145513714, + "grad_norm": 5.941801071166992, + "learning_rate": 9.9975379997538e-06, + "loss": 0.4991, + "step": 2464 + }, + { + "epoch": 1.1459786145978614, + "grad_norm": 5.852427005767822, + "learning_rate": 9.997536999753701e-06, + "loss": 0.5576, + "step": 2465 + }, + { + "epoch": 1.1464435146443515, + "grad_norm": 7.8806633949279785, + "learning_rate": 9.9975359997536e-06, + "loss": 0.7815, + "step": 2466 + }, + { + "epoch": 1.1469084146908415, + "grad_norm": 7.294925212860107, + "learning_rate": 9.997534999753501e-06, + "loss": 0.7954, + "step": 2467 + }, + { + "epoch": 1.1473733147373315, + "grad_norm": 6.835493087768555, + "learning_rate": 9.9975339997534e-06, + "loss": 0.6325, + "step": 2468 + }, + { + "epoch": 1.1478382147838215, + "grad_norm": 6.414812088012695, + "learning_rate": 9.9975329997533e-06, + "loss": 0.3439, + "step": 2469 + }, + { + "epoch": 1.1483031148303116, + "grad_norm": 10.408889770507812, + "learning_rate": 9.997531999753201e-06, + "loss": 1.4004, + "step": 2470 + }, + { + "epoch": 1.1487680148768016, + "grad_norm": 5.441980361938477, + "learning_rate": 9.997530999753102e-06, + "loss": 0.6933, + "step": 2471 + }, + { + "epoch": 1.1492329149232914, + "grad_norm": 8.871638298034668, + "learning_rate": 9.997529999753001e-06, + "loss": 1.0385, + "step": 2472 + }, + { + "epoch": 1.1496978149697814, + "grad_norm": 5.108837127685547, + "learning_rate": 9.997528999752902e-06, + "loss": 0.6647, + "step": 2473 + }, + { + "epoch": 1.1501627150162714, + "grad_norm": 6.604003429412842, + "learning_rate": 9.9975279997528e-06, + "loss": 0.538, + "step": 2474 + }, + { + "epoch": 1.1506276150627615, + "grad_norm": 6.250598907470703, + "learning_rate": 9.9975269997527e-06, + "loss": 0.4529, + "step": 2475 + }, + { + "epoch": 1.1510925151092515, + "grad_norm": 5.69655704498291, + "learning_rate": 9.9975259997526e-06, + "loss": 0.6872, + "step": 2476 + }, + { + "epoch": 1.1515574151557415, + "grad_norm": 5.741939067840576, + "learning_rate": 9.9975249997525e-06, + "loss": 0.7107, + "step": 2477 + }, + { + "epoch": 1.1520223152022315, + "grad_norm": 5.01854133605957, + "learning_rate": 9.997523999752401e-06, + "loss": 0.527, + "step": 2478 + }, + { + "epoch": 1.1524872152487216, + "grad_norm": 5.74107551574707, + "learning_rate": 9.9975229997523e-06, + "loss": 0.7779, + "step": 2479 + }, + { + "epoch": 1.1529521152952116, + "grad_norm": 6.985475540161133, + "learning_rate": 9.997521999752201e-06, + "loss": 0.9365, + "step": 2480 + }, + { + "epoch": 1.1534170153417016, + "grad_norm": 4.875875949859619, + "learning_rate": 9.9975209997521e-06, + "loss": 0.5433, + "step": 2481 + }, + { + "epoch": 1.1538819153881916, + "grad_norm": 5.910841464996338, + "learning_rate": 9.997519999752e-06, + "loss": 0.5435, + "step": 2482 + }, + { + "epoch": 1.1543468154346814, + "grad_norm": 6.33705472946167, + "learning_rate": 9.997518999751901e-06, + "loss": 0.7199, + "step": 2483 + }, + { + "epoch": 1.1548117154811715, + "grad_norm": 7.556943893432617, + "learning_rate": 9.997517999751802e-06, + "loss": 0.6744, + "step": 2484 + }, + { + "epoch": 1.1552766155276615, + "grad_norm": 6.1311726570129395, + "learning_rate": 9.997516999751701e-06, + "loss": 0.7422, + "step": 2485 + }, + { + "epoch": 1.1557415155741515, + "grad_norm": 5.400205612182617, + "learning_rate": 9.9975159997516e-06, + "loss": 0.4244, + "step": 2486 + }, + { + "epoch": 1.1562064156206415, + "grad_norm": 5.301734924316406, + "learning_rate": 9.9975149997515e-06, + "loss": 0.4363, + "step": 2487 + }, + { + "epoch": 1.1566713156671315, + "grad_norm": 5.185722827911377, + "learning_rate": 9.9975139997514e-06, + "loss": 0.483, + "step": 2488 + }, + { + "epoch": 1.1571362157136216, + "grad_norm": 5.667693138122559, + "learning_rate": 9.9975129997513e-06, + "loss": 0.5751, + "step": 2489 + }, + { + "epoch": 1.1576011157601116, + "grad_norm": 6.2976789474487305, + "learning_rate": 9.997511999751201e-06, + "loss": 0.4664, + "step": 2490 + }, + { + "epoch": 1.1580660158066016, + "grad_norm": 5.397450923919678, + "learning_rate": 9.997510999751102e-06, + "loss": 0.5615, + "step": 2491 + }, + { + "epoch": 1.1585309158530916, + "grad_norm": 4.270632743835449, + "learning_rate": 9.997509999751e-06, + "loss": 0.2137, + "step": 2492 + }, + { + "epoch": 1.1589958158995817, + "grad_norm": 5.1881103515625, + "learning_rate": 9.997508999750901e-06, + "loss": 0.5511, + "step": 2493 + }, + { + "epoch": 1.1594607159460715, + "grad_norm": 5.629428863525391, + "learning_rate": 9.9975079997508e-06, + "loss": 0.4684, + "step": 2494 + }, + { + "epoch": 1.1599256159925617, + "grad_norm": 6.957577705383301, + "learning_rate": 9.9975069997507e-06, + "loss": 0.7164, + "step": 2495 + }, + { + "epoch": 1.1603905160390515, + "grad_norm": 9.167976379394531, + "learning_rate": 9.997505999750601e-06, + "loss": 1.1266, + "step": 2496 + }, + { + "epoch": 1.1608554160855415, + "grad_norm": 5.251193523406982, + "learning_rate": 9.9975049997505e-06, + "loss": 0.4568, + "step": 2497 + }, + { + "epoch": 1.1613203161320316, + "grad_norm": 7.733233451843262, + "learning_rate": 9.997503999750401e-06, + "loss": 0.7735, + "step": 2498 + }, + { + "epoch": 1.1617852161785216, + "grad_norm": 5.27536678314209, + "learning_rate": 9.9975029997503e-06, + "loss": 0.5238, + "step": 2499 + }, + { + "epoch": 1.1622501162250116, + "grad_norm": 9.75834846496582, + "learning_rate": 9.9975019997502e-06, + "loss": 0.9721, + "step": 2500 + }, + { + "epoch": 1.1627150162715016, + "grad_norm": 8.460372924804688, + "learning_rate": 9.9975009997501e-06, + "loss": 0.8038, + "step": 2501 + }, + { + "epoch": 1.1631799163179917, + "grad_norm": 6.8149800300598145, + "learning_rate": 9.99749999975e-06, + "loss": 0.6766, + "step": 2502 + }, + { + "epoch": 1.1636448163644817, + "grad_norm": 7.829797744750977, + "learning_rate": 9.997498999749901e-06, + "loss": 0.7652, + "step": 2503 + }, + { + "epoch": 1.1641097164109717, + "grad_norm": 5.247959613800049, + "learning_rate": 9.997497999749802e-06, + "loss": 0.4493, + "step": 2504 + }, + { + "epoch": 1.1645746164574617, + "grad_norm": 4.699702739715576, + "learning_rate": 9.9974969997497e-06, + "loss": 0.573, + "step": 2505 + }, + { + "epoch": 1.1650395165039518, + "grad_norm": 7.606330394744873, + "learning_rate": 9.997495999749601e-06, + "loss": 0.507, + "step": 2506 + }, + { + "epoch": 1.1655044165504416, + "grad_norm": 10.089770317077637, + "learning_rate": 9.9974949997495e-06, + "loss": 0.7816, + "step": 2507 + }, + { + "epoch": 1.1659693165969316, + "grad_norm": 5.09708309173584, + "learning_rate": 9.997493999749401e-06, + "loss": 0.4506, + "step": 2508 + }, + { + "epoch": 1.1664342166434216, + "grad_norm": 6.209015369415283, + "learning_rate": 9.997492999749302e-06, + "loss": 0.4871, + "step": 2509 + }, + { + "epoch": 1.1668991166899116, + "grad_norm": 6.702918529510498, + "learning_rate": 9.9974919997492e-06, + "loss": 0.7025, + "step": 2510 + }, + { + "epoch": 1.1673640167364017, + "grad_norm": 18.880878448486328, + "learning_rate": 9.997490999749101e-06, + "loss": 2.735, + "step": 2511 + }, + { + "epoch": 1.1678289167828917, + "grad_norm": 10.945382118225098, + "learning_rate": 9.997489999749e-06, + "loss": 1.0161, + "step": 2512 + }, + { + "epoch": 1.1682938168293817, + "grad_norm": 4.653984069824219, + "learning_rate": 9.9974889997489e-06, + "loss": 0.4916, + "step": 2513 + }, + { + "epoch": 1.1687587168758717, + "grad_norm": 4.64724063873291, + "learning_rate": 9.9974879997488e-06, + "loss": 0.5517, + "step": 2514 + }, + { + "epoch": 1.1692236169223618, + "grad_norm": 11.449668884277344, + "learning_rate": 9.9974869997487e-06, + "loss": 1.2336, + "step": 2515 + }, + { + "epoch": 1.1696885169688518, + "grad_norm": 6.010657787322998, + "learning_rate": 9.997485999748601e-06, + "loss": 0.607, + "step": 2516 + }, + { + "epoch": 1.1701534170153418, + "grad_norm": 7.569952011108398, + "learning_rate": 9.997484999748502e-06, + "loss": 0.8893, + "step": 2517 + }, + { + "epoch": 1.1706183170618316, + "grad_norm": 4.272054195404053, + "learning_rate": 9.9974839997484e-06, + "loss": 0.5421, + "step": 2518 + }, + { + "epoch": 1.1710832171083216, + "grad_norm": 5.260537624359131, + "learning_rate": 9.9974829997483e-06, + "loss": 0.4736, + "step": 2519 + }, + { + "epoch": 1.1715481171548117, + "grad_norm": 7.341501712799072, + "learning_rate": 9.9974819997482e-06, + "loss": 0.6358, + "step": 2520 + }, + { + "epoch": 1.1720130172013017, + "grad_norm": 8.913859367370605, + "learning_rate": 9.997480999748101e-06, + "loss": 1.2795, + "step": 2521 + }, + { + "epoch": 1.1724779172477917, + "grad_norm": 6.590238571166992, + "learning_rate": 9.997479999748002e-06, + "loss": 0.5359, + "step": 2522 + }, + { + "epoch": 1.1729428172942817, + "grad_norm": 5.907708168029785, + "learning_rate": 9.9974789997479e-06, + "loss": 0.8589, + "step": 2523 + }, + { + "epoch": 1.1734077173407718, + "grad_norm": 6.299588203430176, + "learning_rate": 9.997477999747801e-06, + "loss": 0.7086, + "step": 2524 + }, + { + "epoch": 1.1738726173872618, + "grad_norm": 5.790817737579346, + "learning_rate": 9.9974769997477e-06, + "loss": 0.3825, + "step": 2525 + }, + { + "epoch": 1.1743375174337518, + "grad_norm": 6.868861675262451, + "learning_rate": 9.997475999747601e-06, + "loss": 0.5051, + "step": 2526 + }, + { + "epoch": 1.1748024174802418, + "grad_norm": 4.813432693481445, + "learning_rate": 9.9974749997475e-06, + "loss": 0.3577, + "step": 2527 + }, + { + "epoch": 1.1752673175267319, + "grad_norm": 5.788732528686523, + "learning_rate": 9.9974739997474e-06, + "loss": 0.4737, + "step": 2528 + }, + { + "epoch": 1.1757322175732217, + "grad_norm": 5.266735076904297, + "learning_rate": 9.997472999747301e-06, + "loss": 0.4223, + "step": 2529 + }, + { + "epoch": 1.1761971176197117, + "grad_norm": 5.3130292892456055, + "learning_rate": 9.9974719997472e-06, + "loss": 0.5043, + "step": 2530 + }, + { + "epoch": 1.1766620176662017, + "grad_norm": 8.204442024230957, + "learning_rate": 9.9974709997471e-06, + "loss": 0.7362, + "step": 2531 + }, + { + "epoch": 1.1771269177126917, + "grad_norm": 5.370755672454834, + "learning_rate": 9.997469999747e-06, + "loss": 0.5231, + "step": 2532 + }, + { + "epoch": 1.1775918177591818, + "grad_norm": 6.4374213218688965, + "learning_rate": 9.9974689997469e-06, + "loss": 0.6893, + "step": 2533 + }, + { + "epoch": 1.1780567178056718, + "grad_norm": 5.476044178009033, + "learning_rate": 9.997467999746801e-06, + "loss": 0.5137, + "step": 2534 + }, + { + "epoch": 1.1785216178521618, + "grad_norm": 3.9351894855499268, + "learning_rate": 9.997466999746702e-06, + "loss": 0.3557, + "step": 2535 + }, + { + "epoch": 1.1789865178986518, + "grad_norm": 6.224011421203613, + "learning_rate": 9.9974659997466e-06, + "loss": 0.6602, + "step": 2536 + }, + { + "epoch": 1.1794514179451419, + "grad_norm": 6.849653720855713, + "learning_rate": 9.997464999746501e-06, + "loss": 0.7028, + "step": 2537 + }, + { + "epoch": 1.1799163179916319, + "grad_norm": 5.88706636428833, + "learning_rate": 9.9974639997464e-06, + "loss": 0.6351, + "step": 2538 + }, + { + "epoch": 1.180381218038122, + "grad_norm": 6.124752044677734, + "learning_rate": 9.997462999746301e-06, + "loss": 0.2856, + "step": 2539 + }, + { + "epoch": 1.1808461180846117, + "grad_norm": 5.328833103179932, + "learning_rate": 9.997461999746202e-06, + "loss": 0.3388, + "step": 2540 + }, + { + "epoch": 1.1813110181311017, + "grad_norm": 6.624303340911865, + "learning_rate": 9.9974609997461e-06, + "loss": 0.485, + "step": 2541 + }, + { + "epoch": 1.1817759181775918, + "grad_norm": 9.255170822143555, + "learning_rate": 9.997459999746001e-06, + "loss": 0.5515, + "step": 2542 + }, + { + "epoch": 1.1822408182240818, + "grad_norm": 5.105090618133545, + "learning_rate": 9.9974589997459e-06, + "loss": 0.4213, + "step": 2543 + }, + { + "epoch": 1.1827057182705718, + "grad_norm": 6.985569953918457, + "learning_rate": 9.997457999745801e-06, + "loss": 0.5781, + "step": 2544 + }, + { + "epoch": 1.1831706183170618, + "grad_norm": 6.330123424530029, + "learning_rate": 9.9974569997457e-06, + "loss": 0.7874, + "step": 2545 + }, + { + "epoch": 1.1836355183635519, + "grad_norm": 6.578890800476074, + "learning_rate": 9.9974559997456e-06, + "loss": 0.8012, + "step": 2546 + }, + { + "epoch": 1.1841004184100419, + "grad_norm": 7.221005439758301, + "learning_rate": 9.997454999745501e-06, + "loss": 0.9477, + "step": 2547 + }, + { + "epoch": 1.184565318456532, + "grad_norm": 7.803560256958008, + "learning_rate": 9.997453999745402e-06, + "loss": 1.0769, + "step": 2548 + }, + { + "epoch": 1.185030218503022, + "grad_norm": 3.718223810195923, + "learning_rate": 9.9974529997453e-06, + "loss": 0.2845, + "step": 2549 + }, + { + "epoch": 1.185495118549512, + "grad_norm": 6.060736179351807, + "learning_rate": 9.997451999745201e-06, + "loss": 0.6427, + "step": 2550 + }, + { + "epoch": 1.1859600185960018, + "grad_norm": 5.421525955200195, + "learning_rate": 9.9974509997451e-06, + "loss": 0.342, + "step": 2551 + }, + { + "epoch": 1.1864249186424918, + "grad_norm": 5.861371994018555, + "learning_rate": 9.997449999745001e-06, + "loss": 0.2681, + "step": 2552 + }, + { + "epoch": 1.1868898186889818, + "grad_norm": 5.629678249359131, + "learning_rate": 9.997448999744902e-06, + "loss": 0.5752, + "step": 2553 + }, + { + "epoch": 1.1873547187354718, + "grad_norm": 7.672821998596191, + "learning_rate": 9.9974479997448e-06, + "loss": 0.5744, + "step": 2554 + }, + { + "epoch": 1.1878196187819619, + "grad_norm": 7.870491981506348, + "learning_rate": 9.997446999744701e-06, + "loss": 0.6583, + "step": 2555 + }, + { + "epoch": 1.1882845188284519, + "grad_norm": 4.881421089172363, + "learning_rate": 9.9974459997446e-06, + "loss": 0.5653, + "step": 2556 + }, + { + "epoch": 1.188749418874942, + "grad_norm": 8.948473930358887, + "learning_rate": 9.997444999744501e-06, + "loss": 0.7927, + "step": 2557 + }, + { + "epoch": 1.189214318921432, + "grad_norm": 6.6802802085876465, + "learning_rate": 9.9974439997444e-06, + "loss": 0.6829, + "step": 2558 + }, + { + "epoch": 1.189679218967922, + "grad_norm": 4.381592273712158, + "learning_rate": 9.9974429997443e-06, + "loss": 0.4767, + "step": 2559 + }, + { + "epoch": 1.190144119014412, + "grad_norm": 8.602169036865234, + "learning_rate": 9.997441999744201e-06, + "loss": 0.2991, + "step": 2560 + }, + { + "epoch": 1.190609019060902, + "grad_norm": 3.9259040355682373, + "learning_rate": 9.997440999744102e-06, + "loss": 0.2603, + "step": 2561 + }, + { + "epoch": 1.1910739191073918, + "grad_norm": 5.620212078094482, + "learning_rate": 9.997439999744e-06, + "loss": 0.6689, + "step": 2562 + }, + { + "epoch": 1.1915388191538818, + "grad_norm": 18.180763244628906, + "learning_rate": 9.9974389997439e-06, + "loss": 0.8283, + "step": 2563 + }, + { + "epoch": 1.1920037192003718, + "grad_norm": 5.380748748779297, + "learning_rate": 9.9974379997438e-06, + "loss": 0.5643, + "step": 2564 + }, + { + "epoch": 1.1924686192468619, + "grad_norm": 3.4938406944274902, + "learning_rate": 9.997436999743701e-06, + "loss": 0.3099, + "step": 2565 + }, + { + "epoch": 1.192933519293352, + "grad_norm": 7.058894157409668, + "learning_rate": 9.997435999743602e-06, + "loss": 0.6574, + "step": 2566 + }, + { + "epoch": 1.193398419339842, + "grad_norm": 9.508307456970215, + "learning_rate": 9.9974349997435e-06, + "loss": 1.3128, + "step": 2567 + }, + { + "epoch": 1.193863319386332, + "grad_norm": 5.534472942352295, + "learning_rate": 9.997433999743401e-06, + "loss": 0.4537, + "step": 2568 + }, + { + "epoch": 1.194328219432822, + "grad_norm": 4.720019817352295, + "learning_rate": 9.9974329997433e-06, + "loss": 0.3794, + "step": 2569 + }, + { + "epoch": 1.194793119479312, + "grad_norm": 4.477630138397217, + "learning_rate": 9.997431999743201e-06, + "loss": 0.3271, + "step": 2570 + }, + { + "epoch": 1.195258019525802, + "grad_norm": 9.052522659301758, + "learning_rate": 9.9974309997431e-06, + "loss": 1.099, + "step": 2571 + }, + { + "epoch": 1.195722919572292, + "grad_norm": 5.496120929718018, + "learning_rate": 9.997429999743e-06, + "loss": 0.5262, + "step": 2572 + }, + { + "epoch": 1.1961878196187818, + "grad_norm": 5.838382244110107, + "learning_rate": 9.997428999742901e-06, + "loss": 0.4608, + "step": 2573 + }, + { + "epoch": 1.196652719665272, + "grad_norm": 5.598231792449951, + "learning_rate": 9.9974279997428e-06, + "loss": 0.3478, + "step": 2574 + }, + { + "epoch": 1.197117619711762, + "grad_norm": 7.206724643707275, + "learning_rate": 9.997426999742701e-06, + "loss": 0.6449, + "step": 2575 + }, + { + "epoch": 1.197582519758252, + "grad_norm": 4.778626441955566, + "learning_rate": 9.9974259997426e-06, + "loss": 0.2756, + "step": 2576 + }, + { + "epoch": 1.198047419804742, + "grad_norm": 7.516315460205078, + "learning_rate": 9.9974249997425e-06, + "loss": 0.8354, + "step": 2577 + }, + { + "epoch": 1.198512319851232, + "grad_norm": 4.1554412841796875, + "learning_rate": 9.997423999742401e-06, + "loss": 0.4191, + "step": 2578 + }, + { + "epoch": 1.198977219897722, + "grad_norm": 6.126855850219727, + "learning_rate": 9.997422999742302e-06, + "loss": 0.5258, + "step": 2579 + }, + { + "epoch": 1.199442119944212, + "grad_norm": 3.6037158966064453, + "learning_rate": 9.9974219997422e-06, + "loss": 0.1796, + "step": 2580 + }, + { + "epoch": 1.199907019990702, + "grad_norm": 10.42006778717041, + "learning_rate": 9.997420999742101e-06, + "loss": 0.9072, + "step": 2581 + }, + { + "epoch": 1.200371920037192, + "grad_norm": 5.1546549797058105, + "learning_rate": 9.997419999742e-06, + "loss": 0.4952, + "step": 2582 + }, + { + "epoch": 1.200836820083682, + "grad_norm": 6.587945938110352, + "learning_rate": 9.997418999741901e-06, + "loss": 0.5267, + "step": 2583 + }, + { + "epoch": 1.201301720130172, + "grad_norm": 4.709897041320801, + "learning_rate": 9.9974179997418e-06, + "loss": 0.3094, + "step": 2584 + }, + { + "epoch": 1.2017666201766621, + "grad_norm": 7.134087562561035, + "learning_rate": 9.9974169997417e-06, + "loss": 0.3712, + "step": 2585 + }, + { + "epoch": 1.202231520223152, + "grad_norm": 8.970295906066895, + "learning_rate": 9.997415999741601e-06, + "loss": 0.8399, + "step": 2586 + }, + { + "epoch": 1.202696420269642, + "grad_norm": 12.547086715698242, + "learning_rate": 9.9974149997415e-06, + "loss": 0.3296, + "step": 2587 + }, + { + "epoch": 1.203161320316132, + "grad_norm": 4.005743026733398, + "learning_rate": 9.997413999741401e-06, + "loss": 0.2127, + "step": 2588 + }, + { + "epoch": 1.203626220362622, + "grad_norm": 5.025252819061279, + "learning_rate": 9.9974129997413e-06, + "loss": 0.5104, + "step": 2589 + }, + { + "epoch": 1.204091120409112, + "grad_norm": 4.081263065338135, + "learning_rate": 9.9974119997412e-06, + "loss": 0.2177, + "step": 2590 + }, + { + "epoch": 1.204556020455602, + "grad_norm": 5.5428619384765625, + "learning_rate": 9.997410999741101e-06, + "loss": 0.8016, + "step": 2591 + }, + { + "epoch": 1.205020920502092, + "grad_norm": 5.991549491882324, + "learning_rate": 9.997409999741002e-06, + "loss": 0.5013, + "step": 2592 + }, + { + "epoch": 1.205485820548582, + "grad_norm": 10.563076972961426, + "learning_rate": 9.997408999740901e-06, + "loss": 0.6367, + "step": 2593 + }, + { + "epoch": 1.2059507205950721, + "grad_norm": 7.977787494659424, + "learning_rate": 9.997407999740802e-06, + "loss": 0.7399, + "step": 2594 + }, + { + "epoch": 1.2064156206415622, + "grad_norm": 5.0684494972229, + "learning_rate": 9.9974069997407e-06, + "loss": 0.4175, + "step": 2595 + }, + { + "epoch": 1.2068805206880522, + "grad_norm": 6.166652679443359, + "learning_rate": 9.9974059997406e-06, + "loss": 0.392, + "step": 2596 + }, + { + "epoch": 1.207345420734542, + "grad_norm": 5.640847206115723, + "learning_rate": 9.9974049997405e-06, + "loss": 0.5719, + "step": 2597 + }, + { + "epoch": 1.207810320781032, + "grad_norm": 4.022738456726074, + "learning_rate": 9.9974039997404e-06, + "loss": 0.2055, + "step": 2598 + }, + { + "epoch": 1.208275220827522, + "grad_norm": 5.478492736816406, + "learning_rate": 9.997402999740301e-06, + "loss": 0.5141, + "step": 2599 + }, + { + "epoch": 1.208740120874012, + "grad_norm": 4.581672191619873, + "learning_rate": 9.9974019997402e-06, + "loss": 0.496, + "step": 2600 + }, + { + "epoch": 1.209205020920502, + "grad_norm": 5.804956436157227, + "learning_rate": 9.997400999740101e-06, + "loss": 0.4994, + "step": 2601 + }, + { + "epoch": 1.209669920966992, + "grad_norm": 4.794339656829834, + "learning_rate": 9.99739999974e-06, + "loss": 0.3494, + "step": 2602 + }, + { + "epoch": 1.2101348210134821, + "grad_norm": 8.11595630645752, + "learning_rate": 9.9973989997399e-06, + "loss": 0.8176, + "step": 2603 + }, + { + "epoch": 1.2105997210599722, + "grad_norm": 4.580057144165039, + "learning_rate": 9.997397999739801e-06, + "loss": 0.338, + "step": 2604 + }, + { + "epoch": 1.2110646211064622, + "grad_norm": 5.728760242462158, + "learning_rate": 9.9973969997397e-06, + "loss": 0.318, + "step": 2605 + }, + { + "epoch": 1.2115295211529522, + "grad_norm": 5.151236057281494, + "learning_rate": 9.997395999739601e-06, + "loss": 0.5825, + "step": 2606 + }, + { + "epoch": 1.2119944211994422, + "grad_norm": 4.562431812286377, + "learning_rate": 9.9973949997395e-06, + "loss": 0.4798, + "step": 2607 + }, + { + "epoch": 1.212459321245932, + "grad_norm": 4.434548854827881, + "learning_rate": 9.9973939997394e-06, + "loss": 0.4489, + "step": 2608 + }, + { + "epoch": 1.212924221292422, + "grad_norm": 4.537761211395264, + "learning_rate": 9.997392999739301e-06, + "loss": 0.3824, + "step": 2609 + }, + { + "epoch": 1.213389121338912, + "grad_norm": 6.001027584075928, + "learning_rate": 9.9973919997392e-06, + "loss": 0.567, + "step": 2610 + }, + { + "epoch": 1.213854021385402, + "grad_norm": 5.87844181060791, + "learning_rate": 9.997390999739101e-06, + "loss": 0.4634, + "step": 2611 + }, + { + "epoch": 1.2143189214318921, + "grad_norm": 8.921910285949707, + "learning_rate": 9.997389999739002e-06, + "loss": 0.6176, + "step": 2612 + }, + { + "epoch": 1.2147838214783822, + "grad_norm": 8.359898567199707, + "learning_rate": 9.9973889997389e-06, + "loss": 0.7205, + "step": 2613 + }, + { + "epoch": 1.2152487215248722, + "grad_norm": 6.297598838806152, + "learning_rate": 9.997387999738801e-06, + "loss": 0.5691, + "step": 2614 + }, + { + "epoch": 1.2157136215713622, + "grad_norm": 6.36012601852417, + "learning_rate": 9.9973869997387e-06, + "loss": 0.444, + "step": 2615 + }, + { + "epoch": 1.2161785216178522, + "grad_norm": 5.298403263092041, + "learning_rate": 9.9973859997386e-06, + "loss": 0.4491, + "step": 2616 + }, + { + "epoch": 1.2166434216643423, + "grad_norm": 5.4406046867370605, + "learning_rate": 9.997384999738501e-06, + "loss": 0.4593, + "step": 2617 + }, + { + "epoch": 1.2171083217108323, + "grad_norm": 5.1051716804504395, + "learning_rate": 9.9973839997384e-06, + "loss": 0.3856, + "step": 2618 + }, + { + "epoch": 1.217573221757322, + "grad_norm": 8.18179702758789, + "learning_rate": 9.997382999738301e-06, + "loss": 0.5649, + "step": 2619 + }, + { + "epoch": 1.218038121803812, + "grad_norm": 4.346715450286865, + "learning_rate": 9.9973819997382e-06, + "loss": 0.2737, + "step": 2620 + }, + { + "epoch": 1.2185030218503021, + "grad_norm": 6.097378253936768, + "learning_rate": 9.9973809997381e-06, + "loss": 0.3193, + "step": 2621 + }, + { + "epoch": 1.2189679218967922, + "grad_norm": 5.98035192489624, + "learning_rate": 9.997379999738001e-06, + "loss": 0.4644, + "step": 2622 + }, + { + "epoch": 1.2194328219432822, + "grad_norm": 6.956916809082031, + "learning_rate": 9.997378999737902e-06, + "loss": 0.6593, + "step": 2623 + }, + { + "epoch": 1.2198977219897722, + "grad_norm": 6.900369167327881, + "learning_rate": 9.997377999737801e-06, + "loss": 0.5737, + "step": 2624 + }, + { + "epoch": 1.2203626220362622, + "grad_norm": 5.383298873901367, + "learning_rate": 9.997376999737702e-06, + "loss": 0.6631, + "step": 2625 + }, + { + "epoch": 1.2208275220827522, + "grad_norm": 8.027989387512207, + "learning_rate": 9.9973759997376e-06, + "loss": 0.6114, + "step": 2626 + }, + { + "epoch": 1.2212924221292423, + "grad_norm": 10.238443374633789, + "learning_rate": 9.9973749997375e-06, + "loss": 0.6644, + "step": 2627 + }, + { + "epoch": 1.2217573221757323, + "grad_norm": 7.217168807983398, + "learning_rate": 9.9973739997374e-06, + "loss": 0.7317, + "step": 2628 + }, + { + "epoch": 1.2222222222222223, + "grad_norm": 5.408633232116699, + "learning_rate": 9.997372999737301e-06, + "loss": 0.6105, + "step": 2629 + }, + { + "epoch": 1.2226871222687121, + "grad_norm": 5.763835906982422, + "learning_rate": 9.997371999737202e-06, + "loss": 0.468, + "step": 2630 + }, + { + "epoch": 1.2231520223152021, + "grad_norm": 5.4705305099487305, + "learning_rate": 9.9973709997371e-06, + "loss": 0.3583, + "step": 2631 + }, + { + "epoch": 1.2236169223616922, + "grad_norm": 8.649785041809082, + "learning_rate": 9.997369999737001e-06, + "loss": 1.1172, + "step": 2632 + }, + { + "epoch": 1.2240818224081822, + "grad_norm": 6.414024353027344, + "learning_rate": 9.9973689997369e-06, + "loss": 0.5695, + "step": 2633 + }, + { + "epoch": 1.2245467224546722, + "grad_norm": 5.758132457733154, + "learning_rate": 9.9973679997368e-06, + "loss": 0.7185, + "step": 2634 + }, + { + "epoch": 1.2250116225011622, + "grad_norm": 7.000990867614746, + "learning_rate": 9.997366999736701e-06, + "loss": 0.4122, + "step": 2635 + }, + { + "epoch": 1.2254765225476523, + "grad_norm": 3.4056577682495117, + "learning_rate": 9.997365999736602e-06, + "loss": 0.3361, + "step": 2636 + }, + { + "epoch": 1.2259414225941423, + "grad_norm": 4.988608360290527, + "learning_rate": 9.997364999736501e-06, + "loss": 0.3054, + "step": 2637 + }, + { + "epoch": 1.2264063226406323, + "grad_norm": 5.4448089599609375, + "learning_rate": 9.9973639997364e-06, + "loss": 0.5333, + "step": 2638 + }, + { + "epoch": 1.2268712226871223, + "grad_norm": 4.18998908996582, + "learning_rate": 9.9973629997363e-06, + "loss": 0.3297, + "step": 2639 + }, + { + "epoch": 1.2273361227336124, + "grad_norm": 7.210175514221191, + "learning_rate": 9.9973619997362e-06, + "loss": 0.7289, + "step": 2640 + }, + { + "epoch": 1.2278010227801022, + "grad_norm": 5.323444366455078, + "learning_rate": 9.9973609997361e-06, + "loss": 0.5092, + "step": 2641 + }, + { + "epoch": 1.2282659228265922, + "grad_norm": 5.589160919189453, + "learning_rate": 9.997359999736001e-06, + "loss": 0.6467, + "step": 2642 + }, + { + "epoch": 1.2287308228730822, + "grad_norm": 6.115922927856445, + "learning_rate": 9.997358999735902e-06, + "loss": 0.7874, + "step": 2643 + }, + { + "epoch": 1.2291957229195722, + "grad_norm": 8.449370384216309, + "learning_rate": 9.9973579997358e-06, + "loss": 0.698, + "step": 2644 + }, + { + "epoch": 1.2296606229660623, + "grad_norm": 4.602459907531738, + "learning_rate": 9.997356999735701e-06, + "loss": 0.3831, + "step": 2645 + }, + { + "epoch": 1.2301255230125523, + "grad_norm": 7.293886661529541, + "learning_rate": 9.9973559997356e-06, + "loss": 0.8705, + "step": 2646 + }, + { + "epoch": 1.2305904230590423, + "grad_norm": 5.019686698913574, + "learning_rate": 9.997354999735501e-06, + "loss": 0.4681, + "step": 2647 + }, + { + "epoch": 1.2310553231055323, + "grad_norm": 4.217807292938232, + "learning_rate": 9.997353999735402e-06, + "loss": 0.4949, + "step": 2648 + }, + { + "epoch": 1.2315202231520224, + "grad_norm": 2.548393726348877, + "learning_rate": 9.9973529997353e-06, + "loss": 0.1302, + "step": 2649 + }, + { + "epoch": 1.2319851231985124, + "grad_norm": 4.8713836669921875, + "learning_rate": 9.997351999735201e-06, + "loss": 0.4937, + "step": 2650 + }, + { + "epoch": 1.2324500232450024, + "grad_norm": 3.9713382720947266, + "learning_rate": 9.9973509997351e-06, + "loss": 0.2487, + "step": 2651 + }, + { + "epoch": 1.2329149232914922, + "grad_norm": 6.904343605041504, + "learning_rate": 9.997349999735e-06, + "loss": 0.4992, + "step": 2652 + }, + { + "epoch": 1.2333798233379822, + "grad_norm": 8.136007308959961, + "learning_rate": 9.9973489997349e-06, + "loss": 0.7639, + "step": 2653 + }, + { + "epoch": 1.2338447233844723, + "grad_norm": 5.9682745933532715, + "learning_rate": 9.9973479997348e-06, + "loss": 0.546, + "step": 2654 + }, + { + "epoch": 1.2343096234309623, + "grad_norm": 7.244222640991211, + "learning_rate": 9.997346999734701e-06, + "loss": 0.6445, + "step": 2655 + }, + { + "epoch": 1.2347745234774523, + "grad_norm": 6.470603942871094, + "learning_rate": 9.997345999734602e-06, + "loss": 0.6006, + "step": 2656 + }, + { + "epoch": 1.2352394235239423, + "grad_norm": 6.9305267333984375, + "learning_rate": 9.9973449997345e-06, + "loss": 0.6336, + "step": 2657 + }, + { + "epoch": 1.2357043235704324, + "grad_norm": 6.800852298736572, + "learning_rate": 9.997343999734401e-06, + "loss": 0.7656, + "step": 2658 + }, + { + "epoch": 1.2361692236169224, + "grad_norm": 5.400609970092773, + "learning_rate": 9.9973429997343e-06, + "loss": 0.4008, + "step": 2659 + }, + { + "epoch": 1.2366341236634124, + "grad_norm": 6.35512113571167, + "learning_rate": 9.997341999734201e-06, + "loss": 0.5213, + "step": 2660 + }, + { + "epoch": 1.2370990237099024, + "grad_norm": 8.271468162536621, + "learning_rate": 9.997340999734102e-06, + "loss": 0.5167, + "step": 2661 + }, + { + "epoch": 1.2375639237563925, + "grad_norm": 4.049558639526367, + "learning_rate": 9.997339999734e-06, + "loss": 0.2892, + "step": 2662 + }, + { + "epoch": 1.2380288238028823, + "grad_norm": 6.425408363342285, + "learning_rate": 9.997338999733901e-06, + "loss": 0.6973, + "step": 2663 + }, + { + "epoch": 1.2384937238493725, + "grad_norm": 7.057345390319824, + "learning_rate": 9.9973379997338e-06, + "loss": 0.5073, + "step": 2664 + }, + { + "epoch": 1.2389586238958623, + "grad_norm": 3.599252700805664, + "learning_rate": 9.9973369997337e-06, + "loss": 0.2247, + "step": 2665 + }, + { + "epoch": 1.2394235239423523, + "grad_norm": 4.48404598236084, + "learning_rate": 9.9973359997336e-06, + "loss": 0.5096, + "step": 2666 + }, + { + "epoch": 1.2398884239888424, + "grad_norm": 5.984477519989014, + "learning_rate": 9.9973349997335e-06, + "loss": 0.5013, + "step": 2667 + }, + { + "epoch": 1.2403533240353324, + "grad_norm": 10.021673202514648, + "learning_rate": 9.997333999733401e-06, + "loss": 1.1547, + "step": 2668 + }, + { + "epoch": 1.2408182240818224, + "grad_norm": 4.812919616699219, + "learning_rate": 9.997332999733302e-06, + "loss": 0.4045, + "step": 2669 + }, + { + "epoch": 1.2412831241283124, + "grad_norm": 6.616232395172119, + "learning_rate": 9.9973319997332e-06, + "loss": 0.4408, + "step": 2670 + }, + { + "epoch": 1.2417480241748025, + "grad_norm": 5.5959882736206055, + "learning_rate": 9.9973309997331e-06, + "loss": 0.3665, + "step": 2671 + }, + { + "epoch": 1.2422129242212925, + "grad_norm": 4.435755729675293, + "learning_rate": 9.997329999733e-06, + "loss": 0.3474, + "step": 2672 + }, + { + "epoch": 1.2426778242677825, + "grad_norm": 8.655323028564453, + "learning_rate": 9.997328999732901e-06, + "loss": 0.6226, + "step": 2673 + }, + { + "epoch": 1.2431427243142723, + "grad_norm": 4.596333026885986, + "learning_rate": 9.997327999732802e-06, + "loss": 0.4754, + "step": 2674 + }, + { + "epoch": 1.2436076243607626, + "grad_norm": 4.764501571655273, + "learning_rate": 9.9973269997327e-06, + "loss": 0.4587, + "step": 2675 + }, + { + "epoch": 1.2440725244072524, + "grad_norm": 5.73415470123291, + "learning_rate": 9.997325999732601e-06, + "loss": 0.5131, + "step": 2676 + }, + { + "epoch": 1.2445374244537424, + "grad_norm": 5.599452495574951, + "learning_rate": 9.9973249997325e-06, + "loss": 0.3807, + "step": 2677 + }, + { + "epoch": 1.2450023245002324, + "grad_norm": 8.409965515136719, + "learning_rate": 9.997323999732401e-06, + "loss": 0.5359, + "step": 2678 + }, + { + "epoch": 1.2454672245467224, + "grad_norm": 6.280808448791504, + "learning_rate": 9.9973229997323e-06, + "loss": 0.4484, + "step": 2679 + }, + { + "epoch": 1.2459321245932125, + "grad_norm": 5.698644638061523, + "learning_rate": 9.9973219997322e-06, + "loss": 0.6027, + "step": 2680 + }, + { + "epoch": 1.2463970246397025, + "grad_norm": 5.583163261413574, + "learning_rate": 9.997320999732101e-06, + "loss": 0.3635, + "step": 2681 + }, + { + "epoch": 1.2468619246861925, + "grad_norm": 4.992555618286133, + "learning_rate": 9.997319999732e-06, + "loss": 0.3301, + "step": 2682 + }, + { + "epoch": 1.2473268247326825, + "grad_norm": 8.726938247680664, + "learning_rate": 9.9973189997319e-06, + "loss": 0.6864, + "step": 2683 + }, + { + "epoch": 1.2477917247791726, + "grad_norm": 6.850297927856445, + "learning_rate": 9.9973179997318e-06, + "loss": 0.6339, + "step": 2684 + }, + { + "epoch": 1.2482566248256626, + "grad_norm": 7.709141254425049, + "learning_rate": 9.9973169997317e-06, + "loss": 0.4523, + "step": 2685 + }, + { + "epoch": 1.2487215248721526, + "grad_norm": 4.250770568847656, + "learning_rate": 9.997315999731601e-06, + "loss": 0.4163, + "step": 2686 + }, + { + "epoch": 1.2491864249186424, + "grad_norm": 5.103882789611816, + "learning_rate": 9.997314999731502e-06, + "loss": 0.4031, + "step": 2687 + }, + { + "epoch": 1.2496513249651324, + "grad_norm": 6.153959274291992, + "learning_rate": 9.9973139997314e-06, + "loss": 0.3053, + "step": 2688 + }, + { + "epoch": 1.2501162250116225, + "grad_norm": 7.554106712341309, + "learning_rate": 9.997312999731301e-06, + "loss": 0.8749, + "step": 2689 + }, + { + "epoch": 1.2505811250581125, + "grad_norm": 4.049548625946045, + "learning_rate": 9.9973119997312e-06, + "loss": 0.3065, + "step": 2690 + }, + { + "epoch": 1.2510460251046025, + "grad_norm": 8.239001274108887, + "learning_rate": 9.997310999731101e-06, + "loss": 0.2481, + "step": 2691 + }, + { + "epoch": 1.2515109251510925, + "grad_norm": 5.514925956726074, + "learning_rate": 9.997309999731002e-06, + "loss": 0.4475, + "step": 2692 + }, + { + "epoch": 1.2519758251975825, + "grad_norm": 9.2302885055542, + "learning_rate": 9.9973089997309e-06, + "loss": 0.7242, + "step": 2693 + }, + { + "epoch": 1.2524407252440726, + "grad_norm": 7.509133338928223, + "learning_rate": 9.997307999730801e-06, + "loss": 0.6389, + "step": 2694 + }, + { + "epoch": 1.2529056252905626, + "grad_norm": 6.5458245277404785, + "learning_rate": 9.9973069997307e-06, + "loss": 0.4806, + "step": 2695 + }, + { + "epoch": 1.2533705253370524, + "grad_norm": 5.108046531677246, + "learning_rate": 9.997305999730601e-06, + "loss": 0.3918, + "step": 2696 + }, + { + "epoch": 1.2538354253835426, + "grad_norm": 9.708495140075684, + "learning_rate": 9.9973049997305e-06, + "loss": 0.5244, + "step": 2697 + }, + { + "epoch": 1.2543003254300324, + "grad_norm": 5.701374053955078, + "learning_rate": 9.9973039997304e-06, + "loss": 0.3388, + "step": 2698 + }, + { + "epoch": 1.2547652254765225, + "grad_norm": 9.973392486572266, + "learning_rate": 9.997302999730301e-06, + "loss": 0.2991, + "step": 2699 + }, + { + "epoch": 1.2552301255230125, + "grad_norm": 5.985474586486816, + "learning_rate": 9.997301999730202e-06, + "loss": 0.6768, + "step": 2700 + }, + { + "epoch": 1.2556950255695025, + "grad_norm": 6.591020584106445, + "learning_rate": 9.9973009997301e-06, + "loss": 0.4584, + "step": 2701 + }, + { + "epoch": 1.2561599256159925, + "grad_norm": 7.542978763580322, + "learning_rate": 9.997299999730001e-06, + "loss": 0.8087, + "step": 2702 + }, + { + "epoch": 1.2566248256624826, + "grad_norm": 7.283514022827148, + "learning_rate": 9.9972989997299e-06, + "loss": 0.7109, + "step": 2703 + }, + { + "epoch": 1.2570897257089726, + "grad_norm": 7.046607971191406, + "learning_rate": 9.997297999729801e-06, + "loss": 0.7872, + "step": 2704 + }, + { + "epoch": 1.2575546257554626, + "grad_norm": 6.396254539489746, + "learning_rate": 9.997296999729702e-06, + "loss": 0.6735, + "step": 2705 + }, + { + "epoch": 1.2580195258019526, + "grad_norm": 5.282597064971924, + "learning_rate": 9.9972959997296e-06, + "loss": 0.5974, + "step": 2706 + }, + { + "epoch": 1.2584844258484427, + "grad_norm": 6.035130023956299, + "learning_rate": 9.997294999729501e-06, + "loss": 0.4445, + "step": 2707 + }, + { + "epoch": 1.2589493258949327, + "grad_norm": 5.0162200927734375, + "learning_rate": 9.9972939997294e-06, + "loss": 0.58, + "step": 2708 + }, + { + "epoch": 1.2594142259414225, + "grad_norm": 6.221617221832275, + "learning_rate": 9.997292999729301e-06, + "loss": 0.6769, + "step": 2709 + }, + { + "epoch": 1.2598791259879125, + "grad_norm": 6.693264961242676, + "learning_rate": 9.9972919997292e-06, + "loss": 0.725, + "step": 2710 + }, + { + "epoch": 1.2603440260344025, + "grad_norm": 5.368199348449707, + "learning_rate": 9.9972909997291e-06, + "loss": 0.5511, + "step": 2711 + }, + { + "epoch": 1.2608089260808926, + "grad_norm": 5.722194671630859, + "learning_rate": 9.997289999729001e-06, + "loss": 0.5908, + "step": 2712 + }, + { + "epoch": 1.2612738261273826, + "grad_norm": 5.18423318862915, + "learning_rate": 9.997288999728902e-06, + "loss": 0.5036, + "step": 2713 + }, + { + "epoch": 1.2617387261738726, + "grad_norm": 7.5571112632751465, + "learning_rate": 9.997287999728801e-06, + "loss": 0.5346, + "step": 2714 + }, + { + "epoch": 1.2622036262203626, + "grad_norm": 7.339358329772949, + "learning_rate": 9.9972869997287e-06, + "loss": 0.3908, + "step": 2715 + }, + { + "epoch": 1.2626685262668527, + "grad_norm": 7.364192008972168, + "learning_rate": 9.9972859997286e-06, + "loss": 0.7916, + "step": 2716 + }, + { + "epoch": 1.2631334263133427, + "grad_norm": 3.764864444732666, + "learning_rate": 9.997284999728501e-06, + "loss": 0.3179, + "step": 2717 + }, + { + "epoch": 1.2635983263598327, + "grad_norm": 8.114385604858398, + "learning_rate": 9.997283999728402e-06, + "loss": 0.8479, + "step": 2718 + }, + { + "epoch": 1.2640632264063227, + "grad_norm": 12.81151008605957, + "learning_rate": 9.9972829997283e-06, + "loss": 0.9808, + "step": 2719 + }, + { + "epoch": 1.2645281264528125, + "grad_norm": 7.532785892486572, + "learning_rate": 9.997281999728201e-06, + "loss": 0.7437, + "step": 2720 + }, + { + "epoch": 1.2649930264993028, + "grad_norm": 6.243711471557617, + "learning_rate": 9.9972809997281e-06, + "loss": 0.6045, + "step": 2721 + }, + { + "epoch": 1.2654579265457926, + "grad_norm": 5.9096760749816895, + "learning_rate": 9.997279999728001e-06, + "loss": 0.7318, + "step": 2722 + }, + { + "epoch": 1.2659228265922826, + "grad_norm": 6.0962443351745605, + "learning_rate": 9.9972789997279e-06, + "loss": 0.7559, + "step": 2723 + }, + { + "epoch": 1.2663877266387726, + "grad_norm": 4.512932300567627, + "learning_rate": 9.9972779997278e-06, + "loss": 0.3651, + "step": 2724 + }, + { + "epoch": 1.2668526266852627, + "grad_norm": 6.074883460998535, + "learning_rate": 9.997276999727701e-06, + "loss": 0.4957, + "step": 2725 + }, + { + "epoch": 1.2673175267317527, + "grad_norm": 5.501577854156494, + "learning_rate": 9.9972759997276e-06, + "loss": 0.4122, + "step": 2726 + }, + { + "epoch": 1.2677824267782427, + "grad_norm": 4.703200817108154, + "learning_rate": 9.997274999727501e-06, + "loss": 0.4342, + "step": 2727 + }, + { + "epoch": 1.2682473268247327, + "grad_norm": 4.391674041748047, + "learning_rate": 9.9972739997274e-06, + "loss": 0.42, + "step": 2728 + }, + { + "epoch": 1.2687122268712228, + "grad_norm": 5.8041181564331055, + "learning_rate": 9.9972729997273e-06, + "loss": 0.5381, + "step": 2729 + }, + { + "epoch": 1.2691771269177128, + "grad_norm": 10.218130111694336, + "learning_rate": 9.997271999727201e-06, + "loss": 0.716, + "step": 2730 + }, + { + "epoch": 1.2696420269642026, + "grad_norm": 4.804999828338623, + "learning_rate": 9.997270999727102e-06, + "loss": 0.3779, + "step": 2731 + }, + { + "epoch": 1.2701069270106928, + "grad_norm": 7.058648586273193, + "learning_rate": 9.997269999727001e-06, + "loss": 0.4555, + "step": 2732 + }, + { + "epoch": 1.2705718270571826, + "grad_norm": 5.409944534301758, + "learning_rate": 9.997268999726902e-06, + "loss": 0.643, + "step": 2733 + }, + { + "epoch": 1.2710367271036727, + "grad_norm": 4.2153544425964355, + "learning_rate": 9.9972679997268e-06, + "loss": 0.4005, + "step": 2734 + }, + { + "epoch": 1.2715016271501627, + "grad_norm": 7.597991943359375, + "learning_rate": 9.997266999726701e-06, + "loss": 0.4677, + "step": 2735 + }, + { + "epoch": 1.2719665271966527, + "grad_norm": 6.30781364440918, + "learning_rate": 9.9972659997266e-06, + "loss": 0.7205, + "step": 2736 + }, + { + "epoch": 1.2724314272431427, + "grad_norm": 5.692300796508789, + "learning_rate": 9.9972649997265e-06, + "loss": 0.6514, + "step": 2737 + }, + { + "epoch": 1.2728963272896328, + "grad_norm": 6.324886798858643, + "learning_rate": 9.997263999726401e-06, + "loss": 0.527, + "step": 2738 + }, + { + "epoch": 1.2733612273361228, + "grad_norm": 3.492736577987671, + "learning_rate": 9.9972629997263e-06, + "loss": 0.3289, + "step": 2739 + }, + { + "epoch": 1.2738261273826128, + "grad_norm": 7.048805236816406, + "learning_rate": 9.997261999726201e-06, + "loss": 0.5917, + "step": 2740 + }, + { + "epoch": 1.2742910274291028, + "grad_norm": 7.304534435272217, + "learning_rate": 9.9972609997261e-06, + "loss": 0.5683, + "step": 2741 + }, + { + "epoch": 1.2747559274755926, + "grad_norm": 7.269115924835205, + "learning_rate": 9.997259999726e-06, + "loss": 0.5852, + "step": 2742 + }, + { + "epoch": 1.2752208275220829, + "grad_norm": 4.185292720794678, + "learning_rate": 9.997258999725901e-06, + "loss": 0.3349, + "step": 2743 + }, + { + "epoch": 1.2756857275685727, + "grad_norm": 4.922520637512207, + "learning_rate": 9.997257999725802e-06, + "loss": 0.4605, + "step": 2744 + }, + { + "epoch": 1.2761506276150627, + "grad_norm": 4.998137474060059, + "learning_rate": 9.997256999725701e-06, + "loss": 0.3155, + "step": 2745 + }, + { + "epoch": 1.2766155276615527, + "grad_norm": 5.260077476501465, + "learning_rate": 9.9972559997256e-06, + "loss": 0.5768, + "step": 2746 + }, + { + "epoch": 1.2770804277080428, + "grad_norm": 5.192551136016846, + "learning_rate": 9.9972549997255e-06, + "loss": 0.5107, + "step": 2747 + }, + { + "epoch": 1.2775453277545328, + "grad_norm": 5.021882057189941, + "learning_rate": 9.9972539997254e-06, + "loss": 0.2962, + "step": 2748 + }, + { + "epoch": 1.2780102278010228, + "grad_norm": 7.304816722869873, + "learning_rate": 9.9972529997253e-06, + "loss": 0.64, + "step": 2749 + }, + { + "epoch": 1.2784751278475128, + "grad_norm": 7.39225959777832, + "learning_rate": 9.9972519997252e-06, + "loss": 0.6342, + "step": 2750 + }, + { + "epoch": 1.2789400278940029, + "grad_norm": 6.420746326446533, + "learning_rate": 9.997250999725101e-06, + "loss": 0.6206, + "step": 2751 + }, + { + "epoch": 1.2794049279404929, + "grad_norm": 5.9164252281188965, + "learning_rate": 9.997249999725e-06, + "loss": 0.698, + "step": 2752 + }, + { + "epoch": 1.2798698279869827, + "grad_norm": 6.251963138580322, + "learning_rate": 9.997248999724901e-06, + "loss": 0.6025, + "step": 2753 + }, + { + "epoch": 1.280334728033473, + "grad_norm": 5.601487636566162, + "learning_rate": 9.9972479997248e-06, + "loss": 0.4104, + "step": 2754 + }, + { + "epoch": 1.2807996280799627, + "grad_norm": 5.821593761444092, + "learning_rate": 9.9972469997247e-06, + "loss": 0.76, + "step": 2755 + }, + { + "epoch": 1.2812645281264528, + "grad_norm": 3.50291109085083, + "learning_rate": 9.997245999724601e-06, + "loss": 0.2987, + "step": 2756 + }, + { + "epoch": 1.2817294281729428, + "grad_norm": 7.924383163452148, + "learning_rate": 9.9972449997245e-06, + "loss": 0.5634, + "step": 2757 + }, + { + "epoch": 1.2821943282194328, + "grad_norm": 4.14771842956543, + "learning_rate": 9.997243999724401e-06, + "loss": 0.373, + "step": 2758 + }, + { + "epoch": 1.2826592282659228, + "grad_norm": 12.124897003173828, + "learning_rate": 9.9972429997243e-06, + "loss": 1.0872, + "step": 2759 + }, + { + "epoch": 1.2831241283124128, + "grad_norm": 7.810494422912598, + "learning_rate": 9.9972419997242e-06, + "loss": 1.132, + "step": 2760 + }, + { + "epoch": 1.2835890283589029, + "grad_norm": 6.312707424163818, + "learning_rate": 9.9972409997241e-06, + "loss": 0.5067, + "step": 2761 + }, + { + "epoch": 1.284053928405393, + "grad_norm": 6.640231609344482, + "learning_rate": 9.997239999724e-06, + "loss": 0.5841, + "step": 2762 + }, + { + "epoch": 1.284518828451883, + "grad_norm": 5.587459564208984, + "learning_rate": 9.997238999723901e-06, + "loss": 0.5271, + "step": 2763 + }, + { + "epoch": 1.2849837284983727, + "grad_norm": 7.432864665985107, + "learning_rate": 9.997237999723802e-06, + "loss": 0.6442, + "step": 2764 + }, + { + "epoch": 1.285448628544863, + "grad_norm": 5.5885539054870605, + "learning_rate": 9.9972369997237e-06, + "loss": 0.1985, + "step": 2765 + }, + { + "epoch": 1.2859135285913528, + "grad_norm": 8.635177612304688, + "learning_rate": 9.997235999723601e-06, + "loss": 0.5442, + "step": 2766 + }, + { + "epoch": 1.2863784286378428, + "grad_norm": 7.403300762176514, + "learning_rate": 9.9972349997235e-06, + "loss": 0.9816, + "step": 2767 + }, + { + "epoch": 1.2868433286843328, + "grad_norm": 5.006382465362549, + "learning_rate": 9.9972339997234e-06, + "loss": 0.4105, + "step": 2768 + }, + { + "epoch": 1.2873082287308228, + "grad_norm": 8.68249225616455, + "learning_rate": 9.997232999723301e-06, + "loss": 0.9612, + "step": 2769 + }, + { + "epoch": 1.2877731287773129, + "grad_norm": 4.657000541687012, + "learning_rate": 9.9972319997232e-06, + "loss": 0.4367, + "step": 2770 + }, + { + "epoch": 1.288238028823803, + "grad_norm": 5.066672325134277, + "learning_rate": 9.997230999723101e-06, + "loss": 0.4355, + "step": 2771 + }, + { + "epoch": 1.288702928870293, + "grad_norm": 4.600392818450928, + "learning_rate": 9.997229999723e-06, + "loss": 0.2979, + "step": 2772 + }, + { + "epoch": 1.289167828916783, + "grad_norm": 5.180381774902344, + "learning_rate": 9.9972289997229e-06, + "loss": 0.2988, + "step": 2773 + }, + { + "epoch": 1.289632728963273, + "grad_norm": 5.875922203063965, + "learning_rate": 9.997227999722801e-06, + "loss": 0.6207, + "step": 2774 + }, + { + "epoch": 1.2900976290097628, + "grad_norm": 4.668985843658447, + "learning_rate": 9.9972269997227e-06, + "loss": 0.3236, + "step": 2775 + }, + { + "epoch": 1.290562529056253, + "grad_norm": 6.2828474044799805, + "learning_rate": 9.997225999722601e-06, + "loss": 0.818, + "step": 2776 + }, + { + "epoch": 1.2910274291027428, + "grad_norm": 6.983347415924072, + "learning_rate": 9.997224999722502e-06, + "loss": 0.8267, + "step": 2777 + }, + { + "epoch": 1.2914923291492328, + "grad_norm": 5.623066425323486, + "learning_rate": 9.9972239997224e-06, + "loss": 0.6857, + "step": 2778 + }, + { + "epoch": 1.2919572291957229, + "grad_norm": 7.228401184082031, + "learning_rate": 9.9972229997223e-06, + "loss": 0.5399, + "step": 2779 + }, + { + "epoch": 1.292422129242213, + "grad_norm": 5.215794086456299, + "learning_rate": 9.9972219997222e-06, + "loss": 0.3861, + "step": 2780 + }, + { + "epoch": 1.292887029288703, + "grad_norm": 5.716991901397705, + "learning_rate": 9.997220999722101e-06, + "loss": 0.3965, + "step": 2781 + }, + { + "epoch": 1.293351929335193, + "grad_norm": 11.921241760253906, + "learning_rate": 9.997219999722002e-06, + "loss": 1.123, + "step": 2782 + }, + { + "epoch": 1.293816829381683, + "grad_norm": 5.711712837219238, + "learning_rate": 9.9972189997219e-06, + "loss": 0.3751, + "step": 2783 + }, + { + "epoch": 1.294281729428173, + "grad_norm": 6.731318950653076, + "learning_rate": 9.997217999721801e-06, + "loss": 0.5773, + "step": 2784 + }, + { + "epoch": 1.294746629474663, + "grad_norm": 7.864459037780762, + "learning_rate": 9.9972169997217e-06, + "loss": 0.6295, + "step": 2785 + }, + { + "epoch": 1.2952115295211528, + "grad_norm": 6.187872409820557, + "learning_rate": 9.9972159997216e-06, + "loss": 0.5059, + "step": 2786 + }, + { + "epoch": 1.295676429567643, + "grad_norm": 5.924149990081787, + "learning_rate": 9.997214999721501e-06, + "loss": 0.3772, + "step": 2787 + }, + { + "epoch": 1.2961413296141329, + "grad_norm": 5.532605171203613, + "learning_rate": 9.997213999721402e-06, + "loss": 0.5264, + "step": 2788 + }, + { + "epoch": 1.2966062296606229, + "grad_norm": 5.3132219314575195, + "learning_rate": 9.997212999721301e-06, + "loss": 0.4474, + "step": 2789 + }, + { + "epoch": 1.297071129707113, + "grad_norm": 8.898503303527832, + "learning_rate": 9.9972119997212e-06, + "loss": 0.8467, + "step": 2790 + }, + { + "epoch": 1.297536029753603, + "grad_norm": 4.846219062805176, + "learning_rate": 9.9972109997211e-06, + "loss": 0.5145, + "step": 2791 + }, + { + "epoch": 1.298000929800093, + "grad_norm": 10.542691230773926, + "learning_rate": 9.997209999721e-06, + "loss": 0.9658, + "step": 2792 + }, + { + "epoch": 1.298465829846583, + "grad_norm": 6.79809045791626, + "learning_rate": 9.9972089997209e-06, + "loss": 0.4856, + "step": 2793 + }, + { + "epoch": 1.298930729893073, + "grad_norm": 6.961491584777832, + "learning_rate": 9.997207999720801e-06, + "loss": 0.6033, + "step": 2794 + }, + { + "epoch": 1.299395629939563, + "grad_norm": 9.439806938171387, + "learning_rate": 9.997206999720702e-06, + "loss": 0.7159, + "step": 2795 + }, + { + "epoch": 1.299860529986053, + "grad_norm": 9.773975372314453, + "learning_rate": 9.9972059997206e-06, + "loss": 1.0012, + "step": 2796 + }, + { + "epoch": 1.300325430032543, + "grad_norm": 5.281250953674316, + "learning_rate": 9.997204999720501e-06, + "loss": 0.434, + "step": 2797 + }, + { + "epoch": 1.300790330079033, + "grad_norm": 12.246696472167969, + "learning_rate": 9.9972039997204e-06, + "loss": 0.9519, + "step": 2798 + }, + { + "epoch": 1.301255230125523, + "grad_norm": 5.57805871963501, + "learning_rate": 9.997202999720301e-06, + "loss": 0.5769, + "step": 2799 + }, + { + "epoch": 1.301720130172013, + "grad_norm": 6.086439609527588, + "learning_rate": 9.997201999720202e-06, + "loss": 0.6333, + "step": 2800 + }, + { + "epoch": 1.302185030218503, + "grad_norm": 6.100707054138184, + "learning_rate": 9.9972009997201e-06, + "loss": 0.6334, + "step": 2801 + }, + { + "epoch": 1.302649930264993, + "grad_norm": 7.301699638366699, + "learning_rate": 9.997199999720001e-06, + "loss": 0.4747, + "step": 2802 + }, + { + "epoch": 1.303114830311483, + "grad_norm": 4.9622416496276855, + "learning_rate": 9.9971989997199e-06, + "loss": 0.3891, + "step": 2803 + }, + { + "epoch": 1.303579730357973, + "grad_norm": 4.5188188552856445, + "learning_rate": 9.9971979997198e-06, + "loss": 0.4456, + "step": 2804 + }, + { + "epoch": 1.304044630404463, + "grad_norm": 7.025173664093018, + "learning_rate": 9.9971969997197e-06, + "loss": 0.7057, + "step": 2805 + }, + { + "epoch": 1.304509530450953, + "grad_norm": 12.450777053833008, + "learning_rate": 9.9971959997196e-06, + "loss": 0.9234, + "step": 2806 + }, + { + "epoch": 1.304974430497443, + "grad_norm": 5.84674072265625, + "learning_rate": 9.997194999719501e-06, + "loss": 0.6451, + "step": 2807 + }, + { + "epoch": 1.3054393305439331, + "grad_norm": 4.7732319831848145, + "learning_rate": 9.997193999719402e-06, + "loss": 0.2971, + "step": 2808 + }, + { + "epoch": 1.3059042305904232, + "grad_norm": 4.925719738006592, + "learning_rate": 9.9971929997193e-06, + "loss": 0.4589, + "step": 2809 + }, + { + "epoch": 1.306369130636913, + "grad_norm": 6.15333366394043, + "learning_rate": 9.997191999719201e-06, + "loss": 0.4707, + "step": 2810 + }, + { + "epoch": 1.3068340306834032, + "grad_norm": 3.5904548168182373, + "learning_rate": 9.9971909997191e-06, + "loss": 0.3344, + "step": 2811 + }, + { + "epoch": 1.307298930729893, + "grad_norm": 7.275781631469727, + "learning_rate": 9.997189999719001e-06, + "loss": 0.5645, + "step": 2812 + }, + { + "epoch": 1.307763830776383, + "grad_norm": 5.507754802703857, + "learning_rate": 9.997188999718902e-06, + "loss": 0.3936, + "step": 2813 + }, + { + "epoch": 1.308228730822873, + "grad_norm": 5.871829032897949, + "learning_rate": 9.9971879997188e-06, + "loss": 0.6211, + "step": 2814 + }, + { + "epoch": 1.308693630869363, + "grad_norm": 9.161715507507324, + "learning_rate": 9.997186999718701e-06, + "loss": 0.8231, + "step": 2815 + }, + { + "epoch": 1.309158530915853, + "grad_norm": 4.650484085083008, + "learning_rate": 9.9971859997186e-06, + "loss": 0.2817, + "step": 2816 + }, + { + "epoch": 1.3096234309623431, + "grad_norm": 4.688835620880127, + "learning_rate": 9.997184999718501e-06, + "loss": 0.4959, + "step": 2817 + }, + { + "epoch": 1.3100883310088332, + "grad_norm": 6.077690124511719, + "learning_rate": 9.9971839997184e-06, + "loss": 0.5702, + "step": 2818 + }, + { + "epoch": 1.3105532310553232, + "grad_norm": 5.661258220672607, + "learning_rate": 9.9971829997183e-06, + "loss": 0.6282, + "step": 2819 + }, + { + "epoch": 1.3110181311018132, + "grad_norm": 7.270230293273926, + "learning_rate": 9.997181999718201e-06, + "loss": 0.6966, + "step": 2820 + }, + { + "epoch": 1.311483031148303, + "grad_norm": 7.12741231918335, + "learning_rate": 9.997180999718102e-06, + "loss": 0.7839, + "step": 2821 + }, + { + "epoch": 1.3119479311947932, + "grad_norm": 7.819937229156494, + "learning_rate": 9.997179999718e-06, + "loss": 0.5498, + "step": 2822 + }, + { + "epoch": 1.312412831241283, + "grad_norm": 5.202445030212402, + "learning_rate": 9.9971789997179e-06, + "loss": 0.6607, + "step": 2823 + }, + { + "epoch": 1.312877731287773, + "grad_norm": 6.950046062469482, + "learning_rate": 9.9971779997178e-06, + "loss": 0.6046, + "step": 2824 + }, + { + "epoch": 1.313342631334263, + "grad_norm": 11.502656936645508, + "learning_rate": 9.997176999717701e-06, + "loss": 0.553, + "step": 2825 + }, + { + "epoch": 1.3138075313807531, + "grad_norm": 4.934115886688232, + "learning_rate": 9.997175999717602e-06, + "loss": 0.4518, + "step": 2826 + }, + { + "epoch": 1.3142724314272431, + "grad_norm": 6.304600715637207, + "learning_rate": 9.9971749997175e-06, + "loss": 0.7142, + "step": 2827 + }, + { + "epoch": 1.3147373314737332, + "grad_norm": 6.039148807525635, + "learning_rate": 9.997173999717401e-06, + "loss": 0.5063, + "step": 2828 + }, + { + "epoch": 1.3152022315202232, + "grad_norm": 5.078033447265625, + "learning_rate": 9.9971729997173e-06, + "loss": 0.3574, + "step": 2829 + }, + { + "epoch": 1.3156671315667132, + "grad_norm": 5.620978355407715, + "learning_rate": 9.997171999717201e-06, + "loss": 0.5718, + "step": 2830 + }, + { + "epoch": 1.3161320316132032, + "grad_norm": 5.605066299438477, + "learning_rate": 9.9971709997171e-06, + "loss": 0.5296, + "step": 2831 + }, + { + "epoch": 1.316596931659693, + "grad_norm": 4.331998825073242, + "learning_rate": 9.997169999717e-06, + "loss": 0.4796, + "step": 2832 + }, + { + "epoch": 1.3170618317061833, + "grad_norm": 4.9565229415893555, + "learning_rate": 9.997168999716901e-06, + "loss": 0.4021, + "step": 2833 + }, + { + "epoch": 1.317526731752673, + "grad_norm": 7.631538391113281, + "learning_rate": 9.9971679997168e-06, + "loss": 0.5151, + "step": 2834 + }, + { + "epoch": 1.3179916317991631, + "grad_norm": 5.8691935539245605, + "learning_rate": 9.9971669997167e-06, + "loss": 0.6619, + "step": 2835 + }, + { + "epoch": 1.3184565318456531, + "grad_norm": 6.644981384277344, + "learning_rate": 9.9971659997166e-06, + "loss": 0.5804, + "step": 2836 + }, + { + "epoch": 1.3189214318921432, + "grad_norm": 5.008767127990723, + "learning_rate": 9.9971649997165e-06, + "loss": 0.3486, + "step": 2837 + }, + { + "epoch": 1.3193863319386332, + "grad_norm": 5.279478549957275, + "learning_rate": 9.997163999716401e-06, + "loss": 0.2848, + "step": 2838 + }, + { + "epoch": 1.3198512319851232, + "grad_norm": 7.3319315910339355, + "learning_rate": 9.997162999716302e-06, + "loss": 0.5612, + "step": 2839 + }, + { + "epoch": 1.3203161320316132, + "grad_norm": 8.333239555358887, + "learning_rate": 9.9971619997162e-06, + "loss": 0.5831, + "step": 2840 + }, + { + "epoch": 1.3207810320781033, + "grad_norm": 6.965899467468262, + "learning_rate": 9.997160999716101e-06, + "loss": 0.7353, + "step": 2841 + }, + { + "epoch": 1.3212459321245933, + "grad_norm": 5.028989315032959, + "learning_rate": 9.997159999716e-06, + "loss": 0.3733, + "step": 2842 + }, + { + "epoch": 1.321710832171083, + "grad_norm": 5.9558024406433105, + "learning_rate": 9.997158999715901e-06, + "loss": 0.1944, + "step": 2843 + }, + { + "epoch": 1.3221757322175733, + "grad_norm": 8.883451461791992, + "learning_rate": 9.9971579997158e-06, + "loss": 0.7994, + "step": 2844 + }, + { + "epoch": 1.3226406322640631, + "grad_norm": 5.733861446380615, + "learning_rate": 9.9971569997157e-06, + "loss": 0.4878, + "step": 2845 + }, + { + "epoch": 1.3231055323105532, + "grad_norm": 6.148090839385986, + "learning_rate": 9.997155999715601e-06, + "loss": 0.6029, + "step": 2846 + }, + { + "epoch": 1.3235704323570432, + "grad_norm": 6.382185935974121, + "learning_rate": 9.9971549997155e-06, + "loss": 0.8188, + "step": 2847 + }, + { + "epoch": 1.3240353324035332, + "grad_norm": 4.796387195587158, + "learning_rate": 9.997153999715401e-06, + "loss": 0.6054, + "step": 2848 + }, + { + "epoch": 1.3245002324500232, + "grad_norm": 4.800085067749023, + "learning_rate": 9.9971529997153e-06, + "loss": 0.3557, + "step": 2849 + }, + { + "epoch": 1.3249651324965133, + "grad_norm": 9.271683692932129, + "learning_rate": 9.9971519997152e-06, + "loss": 0.7987, + "step": 2850 + }, + { + "epoch": 1.3254300325430033, + "grad_norm": 4.239929676055908, + "learning_rate": 9.997150999715101e-06, + "loss": 0.3672, + "step": 2851 + }, + { + "epoch": 1.3258949325894933, + "grad_norm": 4.741929531097412, + "learning_rate": 9.997149999715002e-06, + "loss": 0.4106, + "step": 2852 + }, + { + "epoch": 1.3263598326359833, + "grad_norm": 6.321047782897949, + "learning_rate": 9.9971489997149e-06, + "loss": 0.7923, + "step": 2853 + }, + { + "epoch": 1.3268247326824731, + "grad_norm": 8.913509368896484, + "learning_rate": 9.997147999714801e-06, + "loss": 0.9953, + "step": 2854 + }, + { + "epoch": 1.3272896327289634, + "grad_norm": 8.749062538146973, + "learning_rate": 9.9971469997147e-06, + "loss": 0.9124, + "step": 2855 + }, + { + "epoch": 1.3277545327754532, + "grad_norm": 4.440212249755859, + "learning_rate": 9.997145999714601e-06, + "loss": 0.4069, + "step": 2856 + }, + { + "epoch": 1.3282194328219432, + "grad_norm": 5.5567307472229, + "learning_rate": 9.997144999714502e-06, + "loss": 0.3252, + "step": 2857 + }, + { + "epoch": 1.3286843328684332, + "grad_norm": 4.567773342132568, + "learning_rate": 9.9971439997144e-06, + "loss": 0.4409, + "step": 2858 + }, + { + "epoch": 1.3291492329149233, + "grad_norm": 7.067738056182861, + "learning_rate": 9.997142999714301e-06, + "loss": 0.3451, + "step": 2859 + }, + { + "epoch": 1.3296141329614133, + "grad_norm": 6.889719009399414, + "learning_rate": 9.9971419997142e-06, + "loss": 0.4711, + "step": 2860 + }, + { + "epoch": 1.3300790330079033, + "grad_norm": 5.509834289550781, + "learning_rate": 9.997140999714101e-06, + "loss": 0.5084, + "step": 2861 + }, + { + "epoch": 1.3305439330543933, + "grad_norm": 5.134122371673584, + "learning_rate": 9.997139999714e-06, + "loss": 0.3846, + "step": 2862 + }, + { + "epoch": 1.3310088331008834, + "grad_norm": 9.445636749267578, + "learning_rate": 9.9971389997139e-06, + "loss": 1.109, + "step": 2863 + }, + { + "epoch": 1.3314737331473734, + "grad_norm": 7.827271938323975, + "learning_rate": 9.997137999713801e-06, + "loss": 0.7681, + "step": 2864 + }, + { + "epoch": 1.3319386331938632, + "grad_norm": 3.9228086471557617, + "learning_rate": 9.997136999713702e-06, + "loss": 0.4447, + "step": 2865 + }, + { + "epoch": 1.3324035332403534, + "grad_norm": 5.022489547729492, + "learning_rate": 9.997135999713601e-06, + "loss": 0.4578, + "step": 2866 + }, + { + "epoch": 1.3328684332868432, + "grad_norm": 6.936535835266113, + "learning_rate": 9.9971349997135e-06, + "loss": 0.6753, + "step": 2867 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 5.1979498863220215, + "learning_rate": 9.9971339997134e-06, + "loss": 0.5104, + "step": 2868 + }, + { + "epoch": 1.3337982333798233, + "grad_norm": 6.7310028076171875, + "learning_rate": 9.997132999713301e-06, + "loss": 0.4746, + "step": 2869 + }, + { + "epoch": 1.3342631334263133, + "grad_norm": 7.616934776306152, + "learning_rate": 9.997131999713202e-06, + "loss": 0.9362, + "step": 2870 + }, + { + "epoch": 1.3347280334728033, + "grad_norm": 6.915678977966309, + "learning_rate": 9.9971309997131e-06, + "loss": 0.7124, + "step": 2871 + }, + { + "epoch": 1.3351929335192934, + "grad_norm": 3.891177177429199, + "learning_rate": 9.997129999713001e-06, + "loss": 0.2967, + "step": 2872 + }, + { + "epoch": 1.3356578335657834, + "grad_norm": 6.438318252563477, + "learning_rate": 9.9971289997129e-06, + "loss": 0.5948, + "step": 2873 + }, + { + "epoch": 1.3361227336122734, + "grad_norm": 9.398539543151855, + "learning_rate": 9.997127999712801e-06, + "loss": 0.8592, + "step": 2874 + }, + { + "epoch": 1.3365876336587634, + "grad_norm": 6.4921345710754395, + "learning_rate": 9.9971269997127e-06, + "loss": 0.8238, + "step": 2875 + }, + { + "epoch": 1.3370525337052532, + "grad_norm": 5.062838554382324, + "learning_rate": 9.9971259997126e-06, + "loss": 0.5742, + "step": 2876 + }, + { + "epoch": 1.3375174337517435, + "grad_norm": 5.353534698486328, + "learning_rate": 9.997124999712501e-06, + "loss": 0.3528, + "step": 2877 + }, + { + "epoch": 1.3379823337982333, + "grad_norm": 8.999159812927246, + "learning_rate": 9.9971239997124e-06, + "loss": 1.1727, + "step": 2878 + }, + { + "epoch": 1.3384472338447233, + "grad_norm": 5.90580415725708, + "learning_rate": 9.997122999712301e-06, + "loss": 0.8439, + "step": 2879 + }, + { + "epoch": 1.3389121338912133, + "grad_norm": 3.5811450481414795, + "learning_rate": 9.9971219997122e-06, + "loss": 0.2704, + "step": 2880 + }, + { + "epoch": 1.3393770339377034, + "grad_norm": 4.677485942840576, + "learning_rate": 9.9971209997121e-06, + "loss": 0.5215, + "step": 2881 + }, + { + "epoch": 1.3398419339841934, + "grad_norm": 5.923542499542236, + "learning_rate": 9.997119999712001e-06, + "loss": 0.9105, + "step": 2882 + }, + { + "epoch": 1.3403068340306834, + "grad_norm": 7.059866428375244, + "learning_rate": 9.997118999711902e-06, + "loss": 0.5531, + "step": 2883 + }, + { + "epoch": 1.3407717340771734, + "grad_norm": 5.641503810882568, + "learning_rate": 9.997117999711801e-06, + "loss": 0.492, + "step": 2884 + }, + { + "epoch": 1.3412366341236635, + "grad_norm": 9.17851448059082, + "learning_rate": 9.997116999711702e-06, + "loss": 0.9328, + "step": 2885 + }, + { + "epoch": 1.3417015341701535, + "grad_norm": 3.451347589492798, + "learning_rate": 9.9971159997116e-06, + "loss": 0.2729, + "step": 2886 + }, + { + "epoch": 1.3421664342166435, + "grad_norm": 6.959483623504639, + "learning_rate": 9.9971149997115e-06, + "loss": 0.7705, + "step": 2887 + }, + { + "epoch": 1.3426313342631335, + "grad_norm": 6.36794376373291, + "learning_rate": 9.9971139997114e-06, + "loss": 0.4048, + "step": 2888 + }, + { + "epoch": 1.3430962343096233, + "grad_norm": 6.148439884185791, + "learning_rate": 9.9971129997113e-06, + "loss": 0.8068, + "step": 2889 + }, + { + "epoch": 1.3435611343561136, + "grad_norm": 6.155517101287842, + "learning_rate": 9.997111999711201e-06, + "loss": 0.3918, + "step": 2890 + }, + { + "epoch": 1.3440260344026034, + "grad_norm": 4.250600814819336, + "learning_rate": 9.9971109997111e-06, + "loss": 0.401, + "step": 2891 + }, + { + "epoch": 1.3444909344490934, + "grad_norm": 4.061110019683838, + "learning_rate": 9.997109999711001e-06, + "loss": 0.3127, + "step": 2892 + }, + { + "epoch": 1.3449558344955834, + "grad_norm": 3.824622631072998, + "learning_rate": 9.9971089997109e-06, + "loss": 0.3961, + "step": 2893 + }, + { + "epoch": 1.3454207345420734, + "grad_norm": 4.946393966674805, + "learning_rate": 9.9971079997108e-06, + "loss": 0.6136, + "step": 2894 + }, + { + "epoch": 1.3458856345885635, + "grad_norm": 6.347789764404297, + "learning_rate": 9.997106999710701e-06, + "loss": 0.56, + "step": 2895 + }, + { + "epoch": 1.3463505346350535, + "grad_norm": 4.690427303314209, + "learning_rate": 9.997105999710602e-06, + "loss": 0.564, + "step": 2896 + }, + { + "epoch": 1.3468154346815435, + "grad_norm": 8.876750946044922, + "learning_rate": 9.997104999710501e-06, + "loss": 0.9238, + "step": 2897 + }, + { + "epoch": 1.3472803347280335, + "grad_norm": 4.256500720977783, + "learning_rate": 9.9971039997104e-06, + "loss": 0.3957, + "step": 2898 + }, + { + "epoch": 1.3477452347745236, + "grad_norm": 5.302395343780518, + "learning_rate": 9.9971029997103e-06, + "loss": 0.5847, + "step": 2899 + }, + { + "epoch": 1.3482101348210134, + "grad_norm": 5.075392723083496, + "learning_rate": 9.9971019997102e-06, + "loss": 0.307, + "step": 2900 + }, + { + "epoch": 1.3486750348675036, + "grad_norm": 6.554705619812012, + "learning_rate": 9.9971009997101e-06, + "loss": 0.7053, + "step": 2901 + }, + { + "epoch": 1.3491399349139934, + "grad_norm": 4.960315227508545, + "learning_rate": 9.997099999710001e-06, + "loss": 0.445, + "step": 2902 + }, + { + "epoch": 1.3496048349604834, + "grad_norm": 5.414975166320801, + "learning_rate": 9.997098999709902e-06, + "loss": 0.61, + "step": 2903 + }, + { + "epoch": 1.3500697350069735, + "grad_norm": 3.8710172176361084, + "learning_rate": 9.9970979997098e-06, + "loss": 0.3256, + "step": 2904 + }, + { + "epoch": 1.3505346350534635, + "grad_norm": 6.215335369110107, + "learning_rate": 9.997096999709701e-06, + "loss": 0.5043, + "step": 2905 + }, + { + "epoch": 1.3509995350999535, + "grad_norm": 5.418049335479736, + "learning_rate": 9.9970959997096e-06, + "loss": 0.4403, + "step": 2906 + }, + { + "epoch": 1.3514644351464435, + "grad_norm": 4.667141437530518, + "learning_rate": 9.9970949997095e-06, + "loss": 0.4269, + "step": 2907 + }, + { + "epoch": 1.3519293351929336, + "grad_norm": 16.66932487487793, + "learning_rate": 9.997093999709401e-06, + "loss": 1.2335, + "step": 2908 + }, + { + "epoch": 1.3523942352394236, + "grad_norm": 6.498439788818359, + "learning_rate": 9.9970929997093e-06, + "loss": 0.83, + "step": 2909 + }, + { + "epoch": 1.3528591352859136, + "grad_norm": 8.196728706359863, + "learning_rate": 9.997091999709201e-06, + "loss": 0.8042, + "step": 2910 + }, + { + "epoch": 1.3533240353324034, + "grad_norm": 6.904376029968262, + "learning_rate": 9.9970909997091e-06, + "loss": 0.4887, + "step": 2911 + }, + { + "epoch": 1.3537889353788937, + "grad_norm": 5.84160852432251, + "learning_rate": 9.997089999709e-06, + "loss": 0.3097, + "step": 2912 + }, + { + "epoch": 1.3542538354253835, + "grad_norm": 6.565975189208984, + "learning_rate": 9.9970889997089e-06, + "loss": 0.5793, + "step": 2913 + }, + { + "epoch": 1.3547187354718735, + "grad_norm": 6.599172115325928, + "learning_rate": 9.9970879997088e-06, + "loss": 0.481, + "step": 2914 + }, + { + "epoch": 1.3551836355183635, + "grad_norm": 7.624661922454834, + "learning_rate": 9.997086999708701e-06, + "loss": 1.1301, + "step": 2915 + }, + { + "epoch": 1.3556485355648535, + "grad_norm": 5.091336727142334, + "learning_rate": 9.997085999708602e-06, + "loss": 0.5399, + "step": 2916 + }, + { + "epoch": 1.3561134356113436, + "grad_norm": 8.01740837097168, + "learning_rate": 9.9970849997085e-06, + "loss": 0.5041, + "step": 2917 + }, + { + "epoch": 1.3565783356578336, + "grad_norm": 6.679240703582764, + "learning_rate": 9.997083999708401e-06, + "loss": 0.8251, + "step": 2918 + }, + { + "epoch": 1.3570432357043236, + "grad_norm": 7.060033321380615, + "learning_rate": 9.9970829997083e-06, + "loss": 0.38, + "step": 2919 + }, + { + "epoch": 1.3575081357508136, + "grad_norm": 8.89455795288086, + "learning_rate": 9.997081999708201e-06, + "loss": 0.8759, + "step": 2920 + }, + { + "epoch": 1.3579730357973037, + "grad_norm": 6.222494602203369, + "learning_rate": 9.997080999708102e-06, + "loss": 0.3175, + "step": 2921 + }, + { + "epoch": 1.3584379358437935, + "grad_norm": 5.505598545074463, + "learning_rate": 9.997079999708e-06, + "loss": 0.4429, + "step": 2922 + }, + { + "epoch": 1.3589028358902837, + "grad_norm": 6.09587287902832, + "learning_rate": 9.997078999707901e-06, + "loss": 0.7554, + "step": 2923 + }, + { + "epoch": 1.3593677359367735, + "grad_norm": 5.226953029632568, + "learning_rate": 9.9970779997078e-06, + "loss": 0.4637, + "step": 2924 + }, + { + "epoch": 1.3598326359832635, + "grad_norm": 7.531311511993408, + "learning_rate": 9.9970769997077e-06, + "loss": 0.3984, + "step": 2925 + }, + { + "epoch": 1.3602975360297536, + "grad_norm": 10.681822776794434, + "learning_rate": 9.997075999707601e-06, + "loss": 0.3776, + "step": 2926 + }, + { + "epoch": 1.3607624360762436, + "grad_norm": 6.5604987144470215, + "learning_rate": 9.9970749997075e-06, + "loss": 0.9958, + "step": 2927 + }, + { + "epoch": 1.3612273361227336, + "grad_norm": 5.929450035095215, + "learning_rate": 9.997073999707401e-06, + "loss": 0.7121, + "step": 2928 + }, + { + "epoch": 1.3616922361692236, + "grad_norm": 5.736649513244629, + "learning_rate": 9.997072999707302e-06, + "loss": 0.5959, + "step": 2929 + }, + { + "epoch": 1.3621571362157137, + "grad_norm": 7.162381649017334, + "learning_rate": 9.9970719997072e-06, + "loss": 0.5967, + "step": 2930 + }, + { + "epoch": 1.3626220362622037, + "grad_norm": 7.123595714569092, + "learning_rate": 9.9970709997071e-06, + "loss": 0.7322, + "step": 2931 + }, + { + "epoch": 1.3630869363086937, + "grad_norm": 6.184546947479248, + "learning_rate": 9.997069999707e-06, + "loss": 0.6037, + "step": 2932 + }, + { + "epoch": 1.3635518363551835, + "grad_norm": 5.349216938018799, + "learning_rate": 9.997068999706901e-06, + "loss": 0.4284, + "step": 2933 + }, + { + "epoch": 1.3640167364016738, + "grad_norm": 6.1142191886901855, + "learning_rate": 9.997067999706802e-06, + "loss": 0.7459, + "step": 2934 + }, + { + "epoch": 1.3644816364481636, + "grad_norm": 5.828237056732178, + "learning_rate": 9.9970669997067e-06, + "loss": 0.779, + "step": 2935 + }, + { + "epoch": 1.3649465364946536, + "grad_norm": 4.141225814819336, + "learning_rate": 9.997065999706601e-06, + "loss": 0.2568, + "step": 2936 + }, + { + "epoch": 1.3654114365411436, + "grad_norm": 5.644075870513916, + "learning_rate": 9.9970649997065e-06, + "loss": 0.619, + "step": 2937 + }, + { + "epoch": 1.3658763365876336, + "grad_norm": 4.846052169799805, + "learning_rate": 9.9970639997064e-06, + "loss": 0.5639, + "step": 2938 + }, + { + "epoch": 1.3663412366341237, + "grad_norm": 5.013890266418457, + "learning_rate": 9.997062999706301e-06, + "loss": 0.3208, + "step": 2939 + }, + { + "epoch": 1.3668061366806137, + "grad_norm": 4.630516052246094, + "learning_rate": 9.9970619997062e-06, + "loss": 0.2414, + "step": 2940 + }, + { + "epoch": 1.3672710367271037, + "grad_norm": 5.041431903839111, + "learning_rate": 9.997060999706101e-06, + "loss": 0.4882, + "step": 2941 + }, + { + "epoch": 1.3677359367735937, + "grad_norm": 8.56820011138916, + "learning_rate": 9.997059999706e-06, + "loss": 0.6311, + "step": 2942 + }, + { + "epoch": 1.3682008368200838, + "grad_norm": 7.691431045532227, + "learning_rate": 9.9970589997059e-06, + "loss": 0.6735, + "step": 2943 + }, + { + "epoch": 1.3686657368665736, + "grad_norm": 5.065143585205078, + "learning_rate": 9.9970579997058e-06, + "loss": 0.4236, + "step": 2944 + }, + { + "epoch": 1.3691306369130638, + "grad_norm": 5.693542003631592, + "learning_rate": 9.9970569997057e-06, + "loss": 0.5319, + "step": 2945 + }, + { + "epoch": 1.3695955369595536, + "grad_norm": 4.872269630432129, + "learning_rate": 9.997055999705601e-06, + "loss": 0.3044, + "step": 2946 + }, + { + "epoch": 1.3700604370060436, + "grad_norm": 5.495224475860596, + "learning_rate": 9.997054999705502e-06, + "loss": 0.6363, + "step": 2947 + }, + { + "epoch": 1.3705253370525337, + "grad_norm": 5.168615341186523, + "learning_rate": 9.9970539997054e-06, + "loss": 0.5531, + "step": 2948 + }, + { + "epoch": 1.3709902370990237, + "grad_norm": 5.227132320404053, + "learning_rate": 9.997052999705301e-06, + "loss": 0.4797, + "step": 2949 + }, + { + "epoch": 1.3714551371455137, + "grad_norm": 6.4675984382629395, + "learning_rate": 9.9970519997052e-06, + "loss": 0.6035, + "step": 2950 + }, + { + "epoch": 1.3719200371920037, + "grad_norm": 8.578699111938477, + "learning_rate": 9.997050999705101e-06, + "loss": 0.5406, + "step": 2951 + }, + { + "epoch": 1.3723849372384938, + "grad_norm": 5.978692531585693, + "learning_rate": 9.997049999705002e-06, + "loss": 0.4652, + "step": 2952 + }, + { + "epoch": 1.3728498372849838, + "grad_norm": 6.923114776611328, + "learning_rate": 9.9970489997049e-06, + "loss": 0.6451, + "step": 2953 + }, + { + "epoch": 1.3733147373314738, + "grad_norm": 6.059421539306641, + "learning_rate": 9.997047999704801e-06, + "loss": 0.4945, + "step": 2954 + }, + { + "epoch": 1.3737796373779636, + "grad_norm": 8.0183744430542, + "learning_rate": 9.9970469997047e-06, + "loss": 0.7017, + "step": 2955 + }, + { + "epoch": 1.3742445374244538, + "grad_norm": 4.7838287353515625, + "learning_rate": 9.9970459997046e-06, + "loss": 0.426, + "step": 2956 + }, + { + "epoch": 1.3747094374709437, + "grad_norm": 3.7528460025787354, + "learning_rate": 9.9970449997045e-06, + "loss": 0.2104, + "step": 2957 + }, + { + "epoch": 1.3751743375174337, + "grad_norm": 5.965926170349121, + "learning_rate": 9.9970439997044e-06, + "loss": 0.2746, + "step": 2958 + }, + { + "epoch": 1.3756392375639237, + "grad_norm": 5.246718883514404, + "learning_rate": 9.997042999704301e-06, + "loss": 0.4701, + "step": 2959 + }, + { + "epoch": 1.3761041376104137, + "grad_norm": 6.68441915512085, + "learning_rate": 9.997041999704202e-06, + "loss": 0.5868, + "step": 2960 + }, + { + "epoch": 1.3765690376569037, + "grad_norm": 4.6266045570373535, + "learning_rate": 9.9970409997041e-06, + "loss": 0.5237, + "step": 2961 + }, + { + "epoch": 1.3770339377033938, + "grad_norm": 5.4320549964904785, + "learning_rate": 9.997039999704001e-06, + "loss": 0.4197, + "step": 2962 + }, + { + "epoch": 1.3774988377498838, + "grad_norm": 3.966773748397827, + "learning_rate": 9.9970389997039e-06, + "loss": 0.3935, + "step": 2963 + }, + { + "epoch": 1.3779637377963738, + "grad_norm": 6.076874256134033, + "learning_rate": 9.997037999703801e-06, + "loss": 0.689, + "step": 2964 + }, + { + "epoch": 1.3784286378428638, + "grad_norm": 5.186761856079102, + "learning_rate": 9.997036999703702e-06, + "loss": 0.5105, + "step": 2965 + }, + { + "epoch": 1.3788935378893539, + "grad_norm": 7.362705230712891, + "learning_rate": 9.9970359997036e-06, + "loss": 0.6645, + "step": 2966 + }, + { + "epoch": 1.379358437935844, + "grad_norm": 7.129617691040039, + "learning_rate": 9.997034999703501e-06, + "loss": 0.8908, + "step": 2967 + }, + { + "epoch": 1.3798233379823337, + "grad_norm": 5.900271415710449, + "learning_rate": 9.9970339997034e-06, + "loss": 0.6458, + "step": 2968 + }, + { + "epoch": 1.3802882380288237, + "grad_norm": 5.099453449249268, + "learning_rate": 9.997032999703301e-06, + "loss": 0.3694, + "step": 2969 + }, + { + "epoch": 1.3807531380753137, + "grad_norm": 5.407529354095459, + "learning_rate": 9.9970319997032e-06, + "loss": 0.4825, + "step": 2970 + }, + { + "epoch": 1.3812180381218038, + "grad_norm": 10.908854484558105, + "learning_rate": 9.9970309997031e-06, + "loss": 0.4117, + "step": 2971 + }, + { + "epoch": 1.3816829381682938, + "grad_norm": 4.936193466186523, + "learning_rate": 9.997029999703001e-06, + "loss": 0.5531, + "step": 2972 + }, + { + "epoch": 1.3821478382147838, + "grad_norm": 4.698526859283447, + "learning_rate": 9.997028999702902e-06, + "loss": 0.314, + "step": 2973 + }, + { + "epoch": 1.3826127382612738, + "grad_norm": 4.579212665557861, + "learning_rate": 9.9970279997028e-06, + "loss": 0.4462, + "step": 2974 + }, + { + "epoch": 1.3830776383077639, + "grad_norm": 5.7176194190979, + "learning_rate": 9.9970269997027e-06, + "loss": 0.4701, + "step": 2975 + }, + { + "epoch": 1.383542538354254, + "grad_norm": 9.064122200012207, + "learning_rate": 9.9970259997026e-06, + "loss": 0.6887, + "step": 2976 + }, + { + "epoch": 1.384007438400744, + "grad_norm": 6.988226413726807, + "learning_rate": 9.997024999702501e-06, + "loss": 0.3955, + "step": 2977 + }, + { + "epoch": 1.384472338447234, + "grad_norm": 6.640456199645996, + "learning_rate": 9.997023999702402e-06, + "loss": 0.6988, + "step": 2978 + }, + { + "epoch": 1.3849372384937237, + "grad_norm": 6.984334468841553, + "learning_rate": 9.9970229997023e-06, + "loss": 0.8867, + "step": 2979 + }, + { + "epoch": 1.385402138540214, + "grad_norm": 5.207383632659912, + "learning_rate": 9.997021999702201e-06, + "loss": 0.4883, + "step": 2980 + }, + { + "epoch": 1.3858670385867038, + "grad_norm": 7.188784122467041, + "learning_rate": 9.9970209997021e-06, + "loss": 0.3651, + "step": 2981 + }, + { + "epoch": 1.3863319386331938, + "grad_norm": 6.588418006896973, + "learning_rate": 9.997019999702001e-06, + "loss": 0.4548, + "step": 2982 + }, + { + "epoch": 1.3867968386796838, + "grad_norm": 4.918613910675049, + "learning_rate": 9.9970189997019e-06, + "loss": 0.5226, + "step": 2983 + }, + { + "epoch": 1.3872617387261739, + "grad_norm": 6.07687520980835, + "learning_rate": 9.9970179997018e-06, + "loss": 0.7042, + "step": 2984 + }, + { + "epoch": 1.387726638772664, + "grad_norm": 5.204805850982666, + "learning_rate": 9.997016999701701e-06, + "loss": 0.3967, + "step": 2985 + }, + { + "epoch": 1.388191538819154, + "grad_norm": 5.608415603637695, + "learning_rate": 9.9970159997016e-06, + "loss": 0.73, + "step": 2986 + }, + { + "epoch": 1.388656438865644, + "grad_norm": 5.0341010093688965, + "learning_rate": 9.997014999701501e-06, + "loss": 0.385, + "step": 2987 + }, + { + "epoch": 1.389121338912134, + "grad_norm": 4.687176704406738, + "learning_rate": 9.9970139997014e-06, + "loss": 0.5096, + "step": 2988 + }, + { + "epoch": 1.389586238958624, + "grad_norm": 4.128174304962158, + "learning_rate": 9.9970129997013e-06, + "loss": 0.3319, + "step": 2989 + }, + { + "epoch": 1.3900511390051138, + "grad_norm": 5.160621643066406, + "learning_rate": 9.997011999701201e-06, + "loss": 0.3673, + "step": 2990 + }, + { + "epoch": 1.390516039051604, + "grad_norm": 10.557696342468262, + "learning_rate": 9.997010999701102e-06, + "loss": 0.7124, + "step": 2991 + }, + { + "epoch": 1.3909809390980938, + "grad_norm": 4.039104461669922, + "learning_rate": 9.997009999701e-06, + "loss": 0.3586, + "step": 2992 + }, + { + "epoch": 1.3914458391445839, + "grad_norm": 5.320417404174805, + "learning_rate": 9.997008999700901e-06, + "loss": 0.4817, + "step": 2993 + }, + { + "epoch": 1.3919107391910739, + "grad_norm": 4.962473392486572, + "learning_rate": 9.9970079997008e-06, + "loss": 0.2827, + "step": 2994 + }, + { + "epoch": 1.392375639237564, + "grad_norm": 7.616919040679932, + "learning_rate": 9.997006999700701e-06, + "loss": 0.4772, + "step": 2995 + }, + { + "epoch": 1.392840539284054, + "grad_norm": 7.033681392669678, + "learning_rate": 9.9970059997006e-06, + "loss": 0.396, + "step": 2996 + }, + { + "epoch": 1.393305439330544, + "grad_norm": 6.669976711273193, + "learning_rate": 9.9970049997005e-06, + "loss": 0.8093, + "step": 2997 + }, + { + "epoch": 1.393770339377034, + "grad_norm": 9.116025924682617, + "learning_rate": 9.997003999700401e-06, + "loss": 0.8857, + "step": 2998 + }, + { + "epoch": 1.394235239423524, + "grad_norm": 7.126028537750244, + "learning_rate": 9.9970029997003e-06, + "loss": 0.6284, + "step": 2999 + }, + { + "epoch": 1.394700139470014, + "grad_norm": 7.008157253265381, + "learning_rate": 9.997001999700201e-06, + "loss": 0.724, + "step": 3000 + }, + { + "epoch": 1.3951650395165038, + "grad_norm": 6.476599216461182, + "learning_rate": 9.9970009997001e-06, + "loss": 0.6094, + "step": 3001 + }, + { + "epoch": 1.395629939562994, + "grad_norm": 7.572150230407715, + "learning_rate": 9.9969999997e-06, + "loss": 0.7889, + "step": 3002 + }, + { + "epoch": 1.3960948396094839, + "grad_norm": 4.941886901855469, + "learning_rate": 9.996998999699901e-06, + "loss": 0.2992, + "step": 3003 + }, + { + "epoch": 1.396559739655974, + "grad_norm": 4.205989837646484, + "learning_rate": 9.996997999699802e-06, + "loss": 0.4397, + "step": 3004 + }, + { + "epoch": 1.397024639702464, + "grad_norm": 5.814023494720459, + "learning_rate": 9.996996999699701e-06, + "loss": 0.4705, + "step": 3005 + }, + { + "epoch": 1.397489539748954, + "grad_norm": 5.887749195098877, + "learning_rate": 9.996995999699602e-06, + "loss": 0.496, + "step": 3006 + }, + { + "epoch": 1.397954439795444, + "grad_norm": 7.994316577911377, + "learning_rate": 9.9969949996995e-06, + "loss": 0.6376, + "step": 3007 + }, + { + "epoch": 1.398419339841934, + "grad_norm": 7.074978351593018, + "learning_rate": 9.996993999699401e-06, + "loss": 1.0073, + "step": 3008 + }, + { + "epoch": 1.398884239888424, + "grad_norm": 10.06895637512207, + "learning_rate": 9.9969929996993e-06, + "loss": 0.7473, + "step": 3009 + }, + { + "epoch": 1.399349139934914, + "grad_norm": 4.645018100738525, + "learning_rate": 9.9969919996992e-06, + "loss": 0.3341, + "step": 3010 + }, + { + "epoch": 1.399814039981404, + "grad_norm": 5.550559043884277, + "learning_rate": 9.996990999699101e-06, + "loss": 0.4503, + "step": 3011 + }, + { + "epoch": 1.4002789400278939, + "grad_norm": 5.719535827636719, + "learning_rate": 9.996989999699e-06, + "loss": 0.3666, + "step": 3012 + }, + { + "epoch": 1.4007438400743841, + "grad_norm": 7.837536811828613, + "learning_rate": 9.996988999698901e-06, + "loss": 0.8706, + "step": 3013 + }, + { + "epoch": 1.401208740120874, + "grad_norm": 5.347418308258057, + "learning_rate": 9.9969879996988e-06, + "loss": 0.525, + "step": 3014 + }, + { + "epoch": 1.401673640167364, + "grad_norm": 5.331034183502197, + "learning_rate": 9.9969869996987e-06, + "loss": 0.4727, + "step": 3015 + }, + { + "epoch": 1.402138540213854, + "grad_norm": 9.33145809173584, + "learning_rate": 9.996985999698601e-06, + "loss": 1.2549, + "step": 3016 + }, + { + "epoch": 1.402603440260344, + "grad_norm": 9.147745132446289, + "learning_rate": 9.996984999698502e-06, + "loss": 0.9418, + "step": 3017 + }, + { + "epoch": 1.403068340306834, + "grad_norm": 5.0116729736328125, + "learning_rate": 9.996983999698401e-06, + "loss": 0.5632, + "step": 3018 + }, + { + "epoch": 1.403533240353324, + "grad_norm": 4.7877607345581055, + "learning_rate": 9.9969829996983e-06, + "loss": 0.6224, + "step": 3019 + }, + { + "epoch": 1.403998140399814, + "grad_norm": 4.374802112579346, + "learning_rate": 9.9969819996982e-06, + "loss": 0.3093, + "step": 3020 + }, + { + "epoch": 1.404463040446304, + "grad_norm": 5.2151312828063965, + "learning_rate": 9.996980999698101e-06, + "loss": 0.3735, + "step": 3021 + }, + { + "epoch": 1.4049279404927941, + "grad_norm": 5.87799072265625, + "learning_rate": 9.996979999698002e-06, + "loss": 0.6005, + "step": 3022 + }, + { + "epoch": 1.405392840539284, + "grad_norm": 6.435915946960449, + "learning_rate": 9.9969789996979e-06, + "loss": 0.7108, + "step": 3023 + }, + { + "epoch": 1.4058577405857742, + "grad_norm": 6.305276870727539, + "learning_rate": 9.996977999697801e-06, + "loss": 0.5191, + "step": 3024 + }, + { + "epoch": 1.406322640632264, + "grad_norm": 7.509685039520264, + "learning_rate": 9.9969769996977e-06, + "loss": 0.6686, + "step": 3025 + }, + { + "epoch": 1.406787540678754, + "grad_norm": 4.634374618530273, + "learning_rate": 9.996975999697601e-06, + "loss": 0.427, + "step": 3026 + }, + { + "epoch": 1.407252440725244, + "grad_norm": 6.904226303100586, + "learning_rate": 9.9969749996975e-06, + "loss": 0.8465, + "step": 3027 + }, + { + "epoch": 1.407717340771734, + "grad_norm": 7.0079121589660645, + "learning_rate": 9.9969739996974e-06, + "loss": 0.3517, + "step": 3028 + }, + { + "epoch": 1.408182240818224, + "grad_norm": 4.280802249908447, + "learning_rate": 9.996972999697301e-06, + "loss": 0.4353, + "step": 3029 + }, + { + "epoch": 1.408647140864714, + "grad_norm": 6.517807960510254, + "learning_rate": 9.9969719996972e-06, + "loss": 0.5408, + "step": 3030 + }, + { + "epoch": 1.4091120409112041, + "grad_norm": 5.478367805480957, + "learning_rate": 9.996970999697101e-06, + "loss": 0.3263, + "step": 3031 + }, + { + "epoch": 1.4095769409576941, + "grad_norm": 9.002290725708008, + "learning_rate": 9.996969999697e-06, + "loss": 0.6106, + "step": 3032 + }, + { + "epoch": 1.4100418410041842, + "grad_norm": 6.021955966949463, + "learning_rate": 9.9969689996969e-06, + "loss": 0.6125, + "step": 3033 + }, + { + "epoch": 1.410506741050674, + "grad_norm": 4.647383689880371, + "learning_rate": 9.996967999696801e-06, + "loss": 0.4483, + "step": 3034 + }, + { + "epoch": 1.4109716410971642, + "grad_norm": 6.496819019317627, + "learning_rate": 9.996966999696702e-06, + "loss": 0.4889, + "step": 3035 + }, + { + "epoch": 1.411436541143654, + "grad_norm": 6.353121757507324, + "learning_rate": 9.996965999696601e-06, + "loss": 0.2932, + "step": 3036 + }, + { + "epoch": 1.411901441190144, + "grad_norm": 6.328413963317871, + "learning_rate": 9.996964999696502e-06, + "loss": 0.5051, + "step": 3037 + }, + { + "epoch": 1.412366341236634, + "grad_norm": 5.529501914978027, + "learning_rate": 9.9969639996964e-06, + "loss": 0.4986, + "step": 3038 + }, + { + "epoch": 1.412831241283124, + "grad_norm": 6.127214431762695, + "learning_rate": 9.9969629996963e-06, + "loss": 0.2786, + "step": 3039 + }, + { + "epoch": 1.4132961413296141, + "grad_norm": 10.556620597839355, + "learning_rate": 9.9969619996962e-06, + "loss": 0.5854, + "step": 3040 + }, + { + "epoch": 1.4137610413761041, + "grad_norm": 10.319358825683594, + "learning_rate": 9.9969609996961e-06, + "loss": 0.9483, + "step": 3041 + }, + { + "epoch": 1.4142259414225942, + "grad_norm": 2.1524486541748047, + "learning_rate": 9.996959999696001e-06, + "loss": 0.1114, + "step": 3042 + }, + { + "epoch": 1.4146908414690842, + "grad_norm": 6.284522533416748, + "learning_rate": 9.9969589996959e-06, + "loss": 0.4774, + "step": 3043 + }, + { + "epoch": 1.4151557415155742, + "grad_norm": 5.817793846130371, + "learning_rate": 9.996957999695801e-06, + "loss": 0.7355, + "step": 3044 + }, + { + "epoch": 1.415620641562064, + "grad_norm": 8.99681568145752, + "learning_rate": 9.9969569996957e-06, + "loss": 1.0188, + "step": 3045 + }, + { + "epoch": 1.4160855416085543, + "grad_norm": 11.639630317687988, + "learning_rate": 9.9969559996956e-06, + "loss": 0.7315, + "step": 3046 + }, + { + "epoch": 1.416550441655044, + "grad_norm": 4.7814788818359375, + "learning_rate": 9.996954999695501e-06, + "loss": 0.2929, + "step": 3047 + }, + { + "epoch": 1.417015341701534, + "grad_norm": 13.751479148864746, + "learning_rate": 9.996953999695402e-06, + "loss": 1.3575, + "step": 3048 + }, + { + "epoch": 1.4174802417480241, + "grad_norm": 10.366204261779785, + "learning_rate": 9.996952999695301e-06, + "loss": 0.5142, + "step": 3049 + }, + { + "epoch": 1.4179451417945141, + "grad_norm": 5.739081859588623, + "learning_rate": 9.9969519996952e-06, + "loss": 0.3881, + "step": 3050 + }, + { + "epoch": 1.4184100418410042, + "grad_norm": 4.85914945602417, + "learning_rate": 9.9969509996951e-06, + "loss": 0.5216, + "step": 3051 + }, + { + "epoch": 1.4188749418874942, + "grad_norm": 4.759453773498535, + "learning_rate": 9.996949999695e-06, + "loss": 0.5499, + "step": 3052 + }, + { + "epoch": 1.4193398419339842, + "grad_norm": 5.446372985839844, + "learning_rate": 9.9969489996949e-06, + "loss": 0.4324, + "step": 3053 + }, + { + "epoch": 1.4198047419804742, + "grad_norm": 6.069986820220947, + "learning_rate": 9.996947999694801e-06, + "loss": 0.6745, + "step": 3054 + }, + { + "epoch": 1.4202696420269643, + "grad_norm": 7.710259914398193, + "learning_rate": 9.996946999694702e-06, + "loss": 0.5959, + "step": 3055 + }, + { + "epoch": 1.4207345420734543, + "grad_norm": 5.092002868652344, + "learning_rate": 9.9969459996946e-06, + "loss": 0.4138, + "step": 3056 + }, + { + "epoch": 1.4211994421199443, + "grad_norm": 3.5107309818267822, + "learning_rate": 9.996944999694501e-06, + "loss": 0.2914, + "step": 3057 + }, + { + "epoch": 1.4216643421664341, + "grad_norm": 4.288033962249756, + "learning_rate": 9.9969439996944e-06, + "loss": 0.3011, + "step": 3058 + }, + { + "epoch": 1.4221292422129241, + "grad_norm": 6.569981098175049, + "learning_rate": 9.9969429996943e-06, + "loss": 0.6514, + "step": 3059 + }, + { + "epoch": 1.4225941422594142, + "grad_norm": 5.499393463134766, + "learning_rate": 9.996941999694201e-06, + "loss": 0.452, + "step": 3060 + }, + { + "epoch": 1.4230590423059042, + "grad_norm": 5.799331188201904, + "learning_rate": 9.9969409996941e-06, + "loss": 0.6166, + "step": 3061 + }, + { + "epoch": 1.4235239423523942, + "grad_norm": 14.25661563873291, + "learning_rate": 9.996939999694001e-06, + "loss": 1.0427, + "step": 3062 + }, + { + "epoch": 1.4239888423988842, + "grad_norm": 4.377359867095947, + "learning_rate": 9.9969389996939e-06, + "loss": 0.2546, + "step": 3063 + }, + { + "epoch": 1.4244537424453743, + "grad_norm": 8.351799011230469, + "learning_rate": 9.9969379996938e-06, + "loss": 0.4187, + "step": 3064 + }, + { + "epoch": 1.4249186424918643, + "grad_norm": 8.312785148620605, + "learning_rate": 9.9969369996937e-06, + "loss": 0.8174, + "step": 3065 + }, + { + "epoch": 1.4253835425383543, + "grad_norm": 7.428474426269531, + "learning_rate": 9.9969359996936e-06, + "loss": 0.7004, + "step": 3066 + }, + { + "epoch": 1.4258484425848443, + "grad_norm": 4.858312606811523, + "learning_rate": 9.996934999693501e-06, + "loss": 0.449, + "step": 3067 + }, + { + "epoch": 1.4263133426313344, + "grad_norm": 5.22831916809082, + "learning_rate": 9.996933999693402e-06, + "loss": 0.308, + "step": 3068 + }, + { + "epoch": 1.4267782426778242, + "grad_norm": 6.406515121459961, + "learning_rate": 9.9969329996933e-06, + "loss": 0.5744, + "step": 3069 + }, + { + "epoch": 1.4272431427243144, + "grad_norm": 8.19558334350586, + "learning_rate": 9.996931999693201e-06, + "loss": 0.6416, + "step": 3070 + }, + { + "epoch": 1.4277080427708042, + "grad_norm": 8.23897647857666, + "learning_rate": 9.9969309996931e-06, + "loss": 0.401, + "step": 3071 + }, + { + "epoch": 1.4281729428172942, + "grad_norm": 7.364901065826416, + "learning_rate": 9.996929999693001e-06, + "loss": 0.5962, + "step": 3072 + }, + { + "epoch": 1.4286378428637843, + "grad_norm": 5.9245076179504395, + "learning_rate": 9.996928999692902e-06, + "loss": 0.402, + "step": 3073 + }, + { + "epoch": 1.4291027429102743, + "grad_norm": 4.167618751525879, + "learning_rate": 9.9969279996928e-06, + "loss": 0.3357, + "step": 3074 + }, + { + "epoch": 1.4295676429567643, + "grad_norm": 6.241053581237793, + "learning_rate": 9.996926999692701e-06, + "loss": 0.5423, + "step": 3075 + }, + { + "epoch": 1.4300325430032543, + "grad_norm": 3.7199504375457764, + "learning_rate": 9.9969259996926e-06, + "loss": 0.3708, + "step": 3076 + }, + { + "epoch": 1.4304974430497444, + "grad_norm": 5.430951118469238, + "learning_rate": 9.9969249996925e-06, + "loss": 0.5029, + "step": 3077 + }, + { + "epoch": 1.4309623430962344, + "grad_norm": 5.818283557891846, + "learning_rate": 9.9969239996924e-06, + "loss": 0.7256, + "step": 3078 + }, + { + "epoch": 1.4314272431427244, + "grad_norm": 6.438594341278076, + "learning_rate": 9.9969229996923e-06, + "loss": 0.8406, + "step": 3079 + }, + { + "epoch": 1.4318921431892142, + "grad_norm": 7.591070175170898, + "learning_rate": 9.996921999692201e-06, + "loss": 0.7088, + "step": 3080 + }, + { + "epoch": 1.4323570432357045, + "grad_norm": 9.91330337524414, + "learning_rate": 9.996920999692102e-06, + "loss": 0.708, + "step": 3081 + }, + { + "epoch": 1.4328219432821943, + "grad_norm": 6.381015777587891, + "learning_rate": 9.996919999692e-06, + "loss": 0.6836, + "step": 3082 + }, + { + "epoch": 1.4332868433286843, + "grad_norm": 3.7927377223968506, + "learning_rate": 9.9969189996919e-06, + "loss": 0.5039, + "step": 3083 + }, + { + "epoch": 1.4337517433751743, + "grad_norm": 5.495319366455078, + "learning_rate": 9.9969179996918e-06, + "loss": 0.5185, + "step": 3084 + }, + { + "epoch": 1.4342166434216643, + "grad_norm": 5.472263336181641, + "learning_rate": 9.996916999691701e-06, + "loss": 0.6552, + "step": 3085 + }, + { + "epoch": 1.4346815434681544, + "grad_norm": 7.455924987792969, + "learning_rate": 9.996915999691602e-06, + "loss": 0.7219, + "step": 3086 + }, + { + "epoch": 1.4351464435146444, + "grad_norm": 5.828441619873047, + "learning_rate": 9.9969149996915e-06, + "loss": 0.6915, + "step": 3087 + }, + { + "epoch": 1.4356113435611344, + "grad_norm": 5.468839645385742, + "learning_rate": 9.996913999691401e-06, + "loss": 0.4446, + "step": 3088 + }, + { + "epoch": 1.4360762436076244, + "grad_norm": 5.381757736206055, + "learning_rate": 9.9969129996913e-06, + "loss": 0.4651, + "step": 3089 + }, + { + "epoch": 1.4365411436541144, + "grad_norm": 5.328107833862305, + "learning_rate": 9.996911999691201e-06, + "loss": 0.407, + "step": 3090 + }, + { + "epoch": 1.4370060437006043, + "grad_norm": 5.911623477935791, + "learning_rate": 9.996910999691102e-06, + "loss": 0.4854, + "step": 3091 + }, + { + "epoch": 1.4374709437470945, + "grad_norm": 5.490938663482666, + "learning_rate": 9.996909999691e-06, + "loss": 0.3539, + "step": 3092 + }, + { + "epoch": 1.4379358437935843, + "grad_norm": 5.19872522354126, + "learning_rate": 9.996908999690901e-06, + "loss": 0.2684, + "step": 3093 + }, + { + "epoch": 1.4384007438400743, + "grad_norm": 6.556756496429443, + "learning_rate": 9.9969079996908e-06, + "loss": 0.7003, + "step": 3094 + }, + { + "epoch": 1.4388656438865643, + "grad_norm": 3.5463547706604004, + "learning_rate": 9.9969069996907e-06, + "loss": 0.1867, + "step": 3095 + }, + { + "epoch": 1.4393305439330544, + "grad_norm": 10.19078540802002, + "learning_rate": 9.9969059996906e-06, + "loss": 1.1988, + "step": 3096 + }, + { + "epoch": 1.4397954439795444, + "grad_norm": 8.180513381958008, + "learning_rate": 9.9969049996905e-06, + "loss": 0.6164, + "step": 3097 + }, + { + "epoch": 1.4402603440260344, + "grad_norm": 5.57110071182251, + "learning_rate": 9.996903999690401e-06, + "loss": 0.4657, + "step": 3098 + }, + { + "epoch": 1.4407252440725244, + "grad_norm": 10.56284236907959, + "learning_rate": 9.996902999690302e-06, + "loss": 0.781, + "step": 3099 + }, + { + "epoch": 1.4411901441190145, + "grad_norm": 7.252917289733887, + "learning_rate": 9.9969019996902e-06, + "loss": 0.7386, + "step": 3100 + }, + { + "epoch": 1.4416550441655045, + "grad_norm": 4.3960442543029785, + "learning_rate": 9.996900999690101e-06, + "loss": 0.3062, + "step": 3101 + }, + { + "epoch": 1.4421199442119943, + "grad_norm": 5.745699405670166, + "learning_rate": 9.99689999969e-06, + "loss": 0.5463, + "step": 3102 + }, + { + "epoch": 1.4425848442584845, + "grad_norm": 4.950019836425781, + "learning_rate": 9.996898999689901e-06, + "loss": 0.4201, + "step": 3103 + }, + { + "epoch": 1.4430497443049743, + "grad_norm": 6.431091785430908, + "learning_rate": 9.996897999689802e-06, + "loss": 0.6818, + "step": 3104 + }, + { + "epoch": 1.4435146443514644, + "grad_norm": 8.308435440063477, + "learning_rate": 9.9968969996897e-06, + "loss": 0.7948, + "step": 3105 + }, + { + "epoch": 1.4439795443979544, + "grad_norm": 7.073814392089844, + "learning_rate": 9.996895999689601e-06, + "loss": 0.4994, + "step": 3106 + }, + { + "epoch": 1.4444444444444444, + "grad_norm": 7.532236099243164, + "learning_rate": 9.9968949996895e-06, + "loss": 0.6126, + "step": 3107 + }, + { + "epoch": 1.4449093444909344, + "grad_norm": 9.196242332458496, + "learning_rate": 9.996893999689401e-06, + "loss": 1.3988, + "step": 3108 + }, + { + "epoch": 1.4453742445374245, + "grad_norm": 4.079174518585205, + "learning_rate": 9.9968929996893e-06, + "loss": 0.3137, + "step": 3109 + }, + { + "epoch": 1.4458391445839145, + "grad_norm": 6.456685543060303, + "learning_rate": 9.9968919996892e-06, + "loss": 0.485, + "step": 3110 + }, + { + "epoch": 1.4463040446304045, + "grad_norm": 6.236655235290527, + "learning_rate": 9.996890999689101e-06, + "loss": 0.3004, + "step": 3111 + }, + { + "epoch": 1.4467689446768945, + "grad_norm": 5.212365627288818, + "learning_rate": 9.996889999689002e-06, + "loss": 0.5097, + "step": 3112 + }, + { + "epoch": 1.4472338447233843, + "grad_norm": 6.925333499908447, + "learning_rate": 9.9968889996889e-06, + "loss": 0.952, + "step": 3113 + }, + { + "epoch": 1.4476987447698746, + "grad_norm": 4.776047706604004, + "learning_rate": 9.996887999688801e-06, + "loss": 0.4792, + "step": 3114 + }, + { + "epoch": 1.4481636448163644, + "grad_norm": 8.162449836730957, + "learning_rate": 9.9968869996887e-06, + "loss": 0.6733, + "step": 3115 + }, + { + "epoch": 1.4486285448628544, + "grad_norm": 6.992868423461914, + "learning_rate": 9.996885999688601e-06, + "loss": 0.4165, + "step": 3116 + }, + { + "epoch": 1.4490934449093444, + "grad_norm": 6.5172014236450195, + "learning_rate": 9.996884999688502e-06, + "loss": 0.4943, + "step": 3117 + }, + { + "epoch": 1.4495583449558345, + "grad_norm": 7.439804553985596, + "learning_rate": 9.9968839996884e-06, + "loss": 0.5598, + "step": 3118 + }, + { + "epoch": 1.4500232450023245, + "grad_norm": 5.951997756958008, + "learning_rate": 9.996882999688301e-06, + "loss": 0.5029, + "step": 3119 + }, + { + "epoch": 1.4504881450488145, + "grad_norm": 7.169592380523682, + "learning_rate": 9.9968819996882e-06, + "loss": 0.5124, + "step": 3120 + }, + { + "epoch": 1.4509530450953045, + "grad_norm": 5.007530212402344, + "learning_rate": 9.996880999688101e-06, + "loss": 0.6062, + "step": 3121 + }, + { + "epoch": 1.4514179451417946, + "grad_norm": 7.663766384124756, + "learning_rate": 9.996879999688e-06, + "loss": 0.7885, + "step": 3122 + }, + { + "epoch": 1.4518828451882846, + "grad_norm": 4.1581220626831055, + "learning_rate": 9.9968789996879e-06, + "loss": 0.3792, + "step": 3123 + }, + { + "epoch": 1.4523477452347744, + "grad_norm": 4.665257453918457, + "learning_rate": 9.996877999687801e-06, + "loss": 0.3876, + "step": 3124 + }, + { + "epoch": 1.4528126452812646, + "grad_norm": 6.130786895751953, + "learning_rate": 9.996876999687702e-06, + "loss": 0.8664, + "step": 3125 + }, + { + "epoch": 1.4532775453277544, + "grad_norm": 6.705748081207275, + "learning_rate": 9.9968759996876e-06, + "loss": 0.3392, + "step": 3126 + }, + { + "epoch": 1.4537424453742445, + "grad_norm": 5.550660133361816, + "learning_rate": 9.9968749996875e-06, + "loss": 0.5064, + "step": 3127 + }, + { + "epoch": 1.4542073454207345, + "grad_norm": 11.295355796813965, + "learning_rate": 9.9968739996874e-06, + "loss": 1.2764, + "step": 3128 + }, + { + "epoch": 1.4546722454672245, + "grad_norm": 6.210816860198975, + "learning_rate": 9.996872999687301e-06, + "loss": 0.4956, + "step": 3129 + }, + { + "epoch": 1.4551371455137145, + "grad_norm": 13.104247093200684, + "learning_rate": 9.996871999687202e-06, + "loss": 1.9309, + "step": 3130 + }, + { + "epoch": 1.4556020455602046, + "grad_norm": 4.28470516204834, + "learning_rate": 9.9968709996871e-06, + "loss": 0.2309, + "step": 3131 + }, + { + "epoch": 1.4560669456066946, + "grad_norm": 5.524116039276123, + "learning_rate": 9.996869999687001e-06, + "loss": 0.5618, + "step": 3132 + }, + { + "epoch": 1.4565318456531846, + "grad_norm": 8.925786018371582, + "learning_rate": 9.9968689996869e-06, + "loss": 1.1421, + "step": 3133 + }, + { + "epoch": 1.4569967456996746, + "grad_norm": 4.815215110778809, + "learning_rate": 9.996867999686801e-06, + "loss": 0.4721, + "step": 3134 + }, + { + "epoch": 1.4574616457461644, + "grad_norm": 5.927204608917236, + "learning_rate": 9.9968669996867e-06, + "loss": 0.7505, + "step": 3135 + }, + { + "epoch": 1.4579265457926547, + "grad_norm": 6.644270420074463, + "learning_rate": 9.9968659996866e-06, + "loss": 0.4491, + "step": 3136 + }, + { + "epoch": 1.4583914458391445, + "grad_norm": 5.216222286224365, + "learning_rate": 9.996864999686501e-06, + "loss": 0.7552, + "step": 3137 + }, + { + "epoch": 1.4588563458856345, + "grad_norm": 5.60206413269043, + "learning_rate": 9.9968639996864e-06, + "loss": 0.6782, + "step": 3138 + }, + { + "epoch": 1.4593212459321245, + "grad_norm": 4.354546070098877, + "learning_rate": 9.996862999686301e-06, + "loss": 0.3004, + "step": 3139 + }, + { + "epoch": 1.4597861459786146, + "grad_norm": 6.719423294067383, + "learning_rate": 9.9968619996862e-06, + "loss": 0.7717, + "step": 3140 + }, + { + "epoch": 1.4602510460251046, + "grad_norm": 7.990881443023682, + "learning_rate": 9.9968609996861e-06, + "loss": 0.7671, + "step": 3141 + }, + { + "epoch": 1.4607159460715946, + "grad_norm": 7.320895671844482, + "learning_rate": 9.996859999686001e-06, + "loss": 0.7181, + "step": 3142 + }, + { + "epoch": 1.4611808461180846, + "grad_norm": 5.962873458862305, + "learning_rate": 9.996858999685902e-06, + "loss": 0.74, + "step": 3143 + }, + { + "epoch": 1.4616457461645747, + "grad_norm": 4.749288082122803, + "learning_rate": 9.9968579996858e-06, + "loss": 0.3981, + "step": 3144 + }, + { + "epoch": 1.4621106462110647, + "grad_norm": 5.228191375732422, + "learning_rate": 9.996856999685701e-06, + "loss": 0.5448, + "step": 3145 + }, + { + "epoch": 1.4625755462575547, + "grad_norm": 4.624213695526123, + "learning_rate": 9.9968559996856e-06, + "loss": 0.3608, + "step": 3146 + }, + { + "epoch": 1.4630404463040447, + "grad_norm": 5.432234287261963, + "learning_rate": 9.996854999685501e-06, + "loss": 0.7393, + "step": 3147 + }, + { + "epoch": 1.4635053463505345, + "grad_norm": 8.630559921264648, + "learning_rate": 9.9968539996854e-06, + "loss": 0.8946, + "step": 3148 + }, + { + "epoch": 1.4639702463970248, + "grad_norm": 6.079057216644287, + "learning_rate": 9.9968529996853e-06, + "loss": 0.6679, + "step": 3149 + }, + { + "epoch": 1.4644351464435146, + "grad_norm": 4.880687236785889, + "learning_rate": 9.996851999685201e-06, + "loss": 0.3816, + "step": 3150 + }, + { + "epoch": 1.4649000464900046, + "grad_norm": 6.755713939666748, + "learning_rate": 9.9968509996851e-06, + "loss": 0.8175, + "step": 3151 + }, + { + "epoch": 1.4653649465364946, + "grad_norm": 7.210269451141357, + "learning_rate": 9.996849999685001e-06, + "loss": 0.5283, + "step": 3152 + }, + { + "epoch": 1.4658298465829847, + "grad_norm": 8.100931167602539, + "learning_rate": 9.9968489996849e-06, + "loss": 0.7214, + "step": 3153 + }, + { + "epoch": 1.4662947466294747, + "grad_norm": 5.861679553985596, + "learning_rate": 9.9968479996848e-06, + "loss": 0.5263, + "step": 3154 + }, + { + "epoch": 1.4667596466759647, + "grad_norm": 5.349968910217285, + "learning_rate": 9.996846999684701e-06, + "loss": 0.3979, + "step": 3155 + }, + { + "epoch": 1.4672245467224547, + "grad_norm": 9.66298770904541, + "learning_rate": 9.996845999684602e-06, + "loss": 1.0768, + "step": 3156 + }, + { + "epoch": 1.4676894467689447, + "grad_norm": 5.976162910461426, + "learning_rate": 9.996844999684501e-06, + "loss": 0.5465, + "step": 3157 + }, + { + "epoch": 1.4681543468154348, + "grad_norm": 8.612092018127441, + "learning_rate": 9.996843999684402e-06, + "loss": 0.746, + "step": 3158 + }, + { + "epoch": 1.4686192468619246, + "grad_norm": 6.823472023010254, + "learning_rate": 9.9968429996843e-06, + "loss": 0.7005, + "step": 3159 + }, + { + "epoch": 1.4690841469084148, + "grad_norm": 4.501873970031738, + "learning_rate": 9.996841999684201e-06, + "loss": 0.4364, + "step": 3160 + }, + { + "epoch": 1.4695490469549046, + "grad_norm": 3.624150037765503, + "learning_rate": 9.9968409996841e-06, + "loss": 0.2233, + "step": 3161 + }, + { + "epoch": 1.4700139470013946, + "grad_norm": 6.067880630493164, + "learning_rate": 9.996839999684e-06, + "loss": 0.5518, + "step": 3162 + }, + { + "epoch": 1.4704788470478847, + "grad_norm": 8.157172203063965, + "learning_rate": 9.996838999683901e-06, + "loss": 0.7727, + "step": 3163 + }, + { + "epoch": 1.4709437470943747, + "grad_norm": 5.253787517547607, + "learning_rate": 9.9968379996838e-06, + "loss": 0.7214, + "step": 3164 + }, + { + "epoch": 1.4714086471408647, + "grad_norm": 4.614781379699707, + "learning_rate": 9.996836999683701e-06, + "loss": 0.5965, + "step": 3165 + }, + { + "epoch": 1.4718735471873547, + "grad_norm": 7.223054885864258, + "learning_rate": 9.9968359996836e-06, + "loss": 1.0265, + "step": 3166 + }, + { + "epoch": 1.4723384472338448, + "grad_norm": 6.8492584228515625, + "learning_rate": 9.9968349996835e-06, + "loss": 0.6251, + "step": 3167 + }, + { + "epoch": 1.4728033472803348, + "grad_norm": 4.215879440307617, + "learning_rate": 9.996833999683401e-06, + "loss": 0.5267, + "step": 3168 + }, + { + "epoch": 1.4732682473268248, + "grad_norm": 5.915916442871094, + "learning_rate": 9.9968329996833e-06, + "loss": 0.5772, + "step": 3169 + }, + { + "epoch": 1.4737331473733146, + "grad_norm": 6.209413528442383, + "learning_rate": 9.996831999683201e-06, + "loss": 0.6616, + "step": 3170 + }, + { + "epoch": 1.4741980474198049, + "grad_norm": 8.544997215270996, + "learning_rate": 9.9968309996831e-06, + "loss": 0.674, + "step": 3171 + }, + { + "epoch": 1.4746629474662947, + "grad_norm": 5.790303707122803, + "learning_rate": 9.996829999683e-06, + "loss": 0.5372, + "step": 3172 + }, + { + "epoch": 1.4751278475127847, + "grad_norm": 4.724123477935791, + "learning_rate": 9.996828999682901e-06, + "loss": 0.3897, + "step": 3173 + }, + { + "epoch": 1.4755927475592747, + "grad_norm": 6.590796947479248, + "learning_rate": 9.9968279996828e-06, + "loss": 0.9764, + "step": 3174 + }, + { + "epoch": 1.4760576476057647, + "grad_norm": 6.497898578643799, + "learning_rate": 9.996826999682701e-06, + "loss": 0.6778, + "step": 3175 + }, + { + "epoch": 1.4765225476522548, + "grad_norm": 5.239213466644287, + "learning_rate": 9.996825999682602e-06, + "loss": 0.4564, + "step": 3176 + }, + { + "epoch": 1.4769874476987448, + "grad_norm": 4.637018203735352, + "learning_rate": 9.9968249996825e-06, + "loss": 0.3032, + "step": 3177 + }, + { + "epoch": 1.4774523477452348, + "grad_norm": 8.371675491333008, + "learning_rate": 9.996823999682401e-06, + "loss": 0.5792, + "step": 3178 + }, + { + "epoch": 1.4779172477917248, + "grad_norm": 8.896050453186035, + "learning_rate": 9.9968229996823e-06, + "loss": 0.6386, + "step": 3179 + }, + { + "epoch": 1.4783821478382149, + "grad_norm": 7.205750942230225, + "learning_rate": 9.9968219996822e-06, + "loss": 0.6032, + "step": 3180 + }, + { + "epoch": 1.4788470478847047, + "grad_norm": 5.583934307098389, + "learning_rate": 9.996820999682101e-06, + "loss": 0.4188, + "step": 3181 + }, + { + "epoch": 1.479311947931195, + "grad_norm": 8.791656494140625, + "learning_rate": 9.996819999682e-06, + "loss": 0.7898, + "step": 3182 + }, + { + "epoch": 1.4797768479776847, + "grad_norm": 6.605480670928955, + "learning_rate": 9.996818999681901e-06, + "loss": 0.4782, + "step": 3183 + }, + { + "epoch": 1.4802417480241747, + "grad_norm": 5.39771842956543, + "learning_rate": 9.9968179996818e-06, + "loss": 0.352, + "step": 3184 + }, + { + "epoch": 1.4807066480706648, + "grad_norm": 5.103763103485107, + "learning_rate": 9.9968169996817e-06, + "loss": 0.5221, + "step": 3185 + }, + { + "epoch": 1.4811715481171548, + "grad_norm": 5.205180644989014, + "learning_rate": 9.996815999681601e-06, + "loss": 0.4084, + "step": 3186 + }, + { + "epoch": 1.4816364481636448, + "grad_norm": 4.946037769317627, + "learning_rate": 9.996814999681502e-06, + "loss": 0.4844, + "step": 3187 + }, + { + "epoch": 1.4821013482101348, + "grad_norm": 5.075271129608154, + "learning_rate": 9.996813999681401e-06, + "loss": 0.6997, + "step": 3188 + }, + { + "epoch": 1.4825662482566249, + "grad_norm": 4.9508819580078125, + "learning_rate": 9.996812999681302e-06, + "loss": 0.4938, + "step": 3189 + }, + { + "epoch": 1.4830311483031149, + "grad_norm": 6.443787574768066, + "learning_rate": 9.9968119996812e-06, + "loss": 0.5876, + "step": 3190 + }, + { + "epoch": 1.483496048349605, + "grad_norm": 8.78442668914795, + "learning_rate": 9.9968109996811e-06, + "loss": 0.8398, + "step": 3191 + }, + { + "epoch": 1.4839609483960947, + "grad_norm": 5.09363317489624, + "learning_rate": 9.996809999681e-06, + "loss": 0.3611, + "step": 3192 + }, + { + "epoch": 1.484425848442585, + "grad_norm": 9.370162010192871, + "learning_rate": 9.996808999680901e-06, + "loss": 0.9153, + "step": 3193 + }, + { + "epoch": 1.4848907484890748, + "grad_norm": 4.425540447235107, + "learning_rate": 9.996807999680802e-06, + "loss": 0.397, + "step": 3194 + }, + { + "epoch": 1.4853556485355648, + "grad_norm": 4.387710094451904, + "learning_rate": 9.9968069996807e-06, + "loss": 0.2584, + "step": 3195 + }, + { + "epoch": 1.4858205485820548, + "grad_norm": 6.5157623291015625, + "learning_rate": 9.996805999680601e-06, + "loss": 0.5295, + "step": 3196 + }, + { + "epoch": 1.4862854486285448, + "grad_norm": 5.583672046661377, + "learning_rate": 9.9968049996805e-06, + "loss": 0.3954, + "step": 3197 + }, + { + "epoch": 1.4867503486750349, + "grad_norm": 4.372612476348877, + "learning_rate": 9.9968039996804e-06, + "loss": 0.3351, + "step": 3198 + }, + { + "epoch": 1.4872152487215249, + "grad_norm": 5.193045139312744, + "learning_rate": 9.996802999680301e-06, + "loss": 0.4397, + "step": 3199 + }, + { + "epoch": 1.487680148768015, + "grad_norm": 6.196756362915039, + "learning_rate": 9.996801999680202e-06, + "loss": 0.5326, + "step": 3200 + }, + { + "epoch": 1.488145048814505, + "grad_norm": 6.854388236999512, + "learning_rate": 9.996800999680101e-06, + "loss": 0.5929, + "step": 3201 + }, + { + "epoch": 1.488609948860995, + "grad_norm": 6.756391525268555, + "learning_rate": 9.99679999968e-06, + "loss": 0.7481, + "step": 3202 + }, + { + "epoch": 1.4890748489074848, + "grad_norm": 8.010823249816895, + "learning_rate": 9.9967989996799e-06, + "loss": 0.4191, + "step": 3203 + }, + { + "epoch": 1.489539748953975, + "grad_norm": 4.477957248687744, + "learning_rate": 9.9967979996798e-06, + "loss": 0.3664, + "step": 3204 + }, + { + "epoch": 1.4900046490004648, + "grad_norm": 5.354471206665039, + "learning_rate": 9.9967969996797e-06, + "loss": 0.5344, + "step": 3205 + }, + { + "epoch": 1.4904695490469548, + "grad_norm": 5.61762809753418, + "learning_rate": 9.996795999679601e-06, + "loss": 0.6849, + "step": 3206 + }, + { + "epoch": 1.4909344490934449, + "grad_norm": 4.594954013824463, + "learning_rate": 9.996794999679502e-06, + "loss": 0.4421, + "step": 3207 + }, + { + "epoch": 1.4913993491399349, + "grad_norm": 5.566478729248047, + "learning_rate": 9.9967939996794e-06, + "loss": 0.7613, + "step": 3208 + }, + { + "epoch": 1.491864249186425, + "grad_norm": 4.466547966003418, + "learning_rate": 9.996792999679301e-06, + "loss": 0.3151, + "step": 3209 + }, + { + "epoch": 1.492329149232915, + "grad_norm": 6.828961372375488, + "learning_rate": 9.9967919996792e-06, + "loss": 0.4897, + "step": 3210 + }, + { + "epoch": 1.492794049279405, + "grad_norm": 7.974672794342041, + "learning_rate": 9.9967909996791e-06, + "loss": 0.657, + "step": 3211 + }, + { + "epoch": 1.493258949325895, + "grad_norm": 6.610179424285889, + "learning_rate": 9.996789999679001e-06, + "loss": 0.5869, + "step": 3212 + }, + { + "epoch": 1.493723849372385, + "grad_norm": 5.571815013885498, + "learning_rate": 9.9967889996789e-06, + "loss": 0.4477, + "step": 3213 + }, + { + "epoch": 1.4941887494188748, + "grad_norm": 10.070664405822754, + "learning_rate": 9.996787999678801e-06, + "loss": 0.3608, + "step": 3214 + }, + { + "epoch": 1.494653649465365, + "grad_norm": 6.251002311706543, + "learning_rate": 9.9967869996787e-06, + "loss": 0.5961, + "step": 3215 + }, + { + "epoch": 1.4951185495118549, + "grad_norm": 8.049627304077148, + "learning_rate": 9.9967859996786e-06, + "loss": 0.5379, + "step": 3216 + }, + { + "epoch": 1.4955834495583449, + "grad_norm": 6.4481682777404785, + "learning_rate": 9.9967849996785e-06, + "loss": 0.7184, + "step": 3217 + }, + { + "epoch": 1.496048349604835, + "grad_norm": 12.761979103088379, + "learning_rate": 9.9967839996784e-06, + "loss": 0.7301, + "step": 3218 + }, + { + "epoch": 1.496513249651325, + "grad_norm": 6.921463489532471, + "learning_rate": 9.996782999678301e-06, + "loss": 0.5251, + "step": 3219 + }, + { + "epoch": 1.496978149697815, + "grad_norm": 7.051331043243408, + "learning_rate": 9.996781999678202e-06, + "loss": 0.6952, + "step": 3220 + }, + { + "epoch": 1.497443049744305, + "grad_norm": 6.954566478729248, + "learning_rate": 9.9967809996781e-06, + "loss": 0.6513, + "step": 3221 + }, + { + "epoch": 1.497907949790795, + "grad_norm": 8.219438552856445, + "learning_rate": 9.996779999678001e-06, + "loss": 0.627, + "step": 3222 + }, + { + "epoch": 1.498372849837285, + "grad_norm": 6.358550548553467, + "learning_rate": 9.9967789996779e-06, + "loss": 0.539, + "step": 3223 + }, + { + "epoch": 1.498837749883775, + "grad_norm": 8.642419815063477, + "learning_rate": 9.996777999677801e-06, + "loss": 0.3427, + "step": 3224 + }, + { + "epoch": 1.499302649930265, + "grad_norm": 10.670586585998535, + "learning_rate": 9.996776999677702e-06, + "loss": 0.5211, + "step": 3225 + }, + { + "epoch": 1.499767549976755, + "grad_norm": 6.8987555503845215, + "learning_rate": 9.9967759996776e-06, + "loss": 0.6353, + "step": 3226 + }, + { + "epoch": 1.500232450023245, + "grad_norm": 4.263932228088379, + "learning_rate": 9.996774999677501e-06, + "loss": 0.2865, + "step": 3227 + }, + { + "epoch": 1.5006973500697351, + "grad_norm": 7.477247714996338, + "learning_rate": 9.9967739996774e-06, + "loss": 0.9821, + "step": 3228 + }, + { + "epoch": 1.501162250116225, + "grad_norm": 4.5664448738098145, + "learning_rate": 9.9967729996773e-06, + "loss": 0.3912, + "step": 3229 + }, + { + "epoch": 1.501627150162715, + "grad_norm": 3.971313953399658, + "learning_rate": 9.9967719996772e-06, + "loss": 0.2442, + "step": 3230 + }, + { + "epoch": 1.502092050209205, + "grad_norm": 3.2479946613311768, + "learning_rate": 9.9967709996771e-06, + "loss": 0.2362, + "step": 3231 + }, + { + "epoch": 1.502556950255695, + "grad_norm": 9.072308540344238, + "learning_rate": 9.996769999677001e-06, + "loss": 0.8358, + "step": 3232 + }, + { + "epoch": 1.503021850302185, + "grad_norm": 5.006146430969238, + "learning_rate": 9.996768999676902e-06, + "loss": 0.2572, + "step": 3233 + }, + { + "epoch": 1.503486750348675, + "grad_norm": 5.590768814086914, + "learning_rate": 9.9967679996768e-06, + "loss": 0.5523, + "step": 3234 + }, + { + "epoch": 1.503951650395165, + "grad_norm": 6.77153205871582, + "learning_rate": 9.9967669996767e-06, + "loss": 0.5999, + "step": 3235 + }, + { + "epoch": 1.504416550441655, + "grad_norm": 7.126492500305176, + "learning_rate": 9.9967659996766e-06, + "loss": 0.6559, + "step": 3236 + }, + { + "epoch": 1.5048814504881451, + "grad_norm": 4.1793742179870605, + "learning_rate": 9.996764999676501e-06, + "loss": 0.3058, + "step": 3237 + }, + { + "epoch": 1.505346350534635, + "grad_norm": 4.024850368499756, + "learning_rate": 9.996763999676402e-06, + "loss": 0.2847, + "step": 3238 + }, + { + "epoch": 1.5058112505811252, + "grad_norm": 5.166171550750732, + "learning_rate": 9.9967629996763e-06, + "loss": 0.5909, + "step": 3239 + }, + { + "epoch": 1.506276150627615, + "grad_norm": 7.761171340942383, + "learning_rate": 9.996761999676201e-06, + "loss": 0.5359, + "step": 3240 + }, + { + "epoch": 1.506741050674105, + "grad_norm": 7.622218132019043, + "learning_rate": 9.9967609996761e-06, + "loss": 0.7059, + "step": 3241 + }, + { + "epoch": 1.507205950720595, + "grad_norm": 9.272522926330566, + "learning_rate": 9.996759999676001e-06, + "loss": 0.916, + "step": 3242 + }, + { + "epoch": 1.507670850767085, + "grad_norm": 6.549522876739502, + "learning_rate": 9.9967589996759e-06, + "loss": 0.5528, + "step": 3243 + }, + { + "epoch": 1.508135750813575, + "grad_norm": 10.21436882019043, + "learning_rate": 9.9967579996758e-06, + "loss": 1.0336, + "step": 3244 + }, + { + "epoch": 1.5086006508600651, + "grad_norm": 5.104884624481201, + "learning_rate": 9.996756999675701e-06, + "loss": 0.4633, + "step": 3245 + }, + { + "epoch": 1.5090655509065551, + "grad_norm": 8.183640480041504, + "learning_rate": 9.9967559996756e-06, + "loss": 0.6557, + "step": 3246 + }, + { + "epoch": 1.509530450953045, + "grad_norm": 5.565601348876953, + "learning_rate": 9.9967549996755e-06, + "loss": 0.6066, + "step": 3247 + }, + { + "epoch": 1.5099953509995352, + "grad_norm": 8.908255577087402, + "learning_rate": 9.9967539996754e-06, + "loss": 0.5904, + "step": 3248 + }, + { + "epoch": 1.510460251046025, + "grad_norm": 7.0414581298828125, + "learning_rate": 9.9967529996753e-06, + "loss": 0.4946, + "step": 3249 + }, + { + "epoch": 1.5109251510925152, + "grad_norm": 6.48568868637085, + "learning_rate": 9.996751999675201e-06, + "loss": 0.8222, + "step": 3250 + }, + { + "epoch": 1.511390051139005, + "grad_norm": 3.7924699783325195, + "learning_rate": 9.996750999675102e-06, + "loss": 0.2764, + "step": 3251 + }, + { + "epoch": 1.511854951185495, + "grad_norm": 5.776539325714111, + "learning_rate": 9.996749999675e-06, + "loss": 0.7667, + "step": 3252 + }, + { + "epoch": 1.512319851231985, + "grad_norm": 6.097172737121582, + "learning_rate": 9.996748999674901e-06, + "loss": 0.4222, + "step": 3253 + }, + { + "epoch": 1.5127847512784751, + "grad_norm": 5.071048259735107, + "learning_rate": 9.9967479996748e-06, + "loss": 0.4859, + "step": 3254 + }, + { + "epoch": 1.5132496513249651, + "grad_norm": 4.1703948974609375, + "learning_rate": 9.996746999674701e-06, + "loss": 0.3543, + "step": 3255 + }, + { + "epoch": 1.5137145513714552, + "grad_norm": 4.215035915374756, + "learning_rate": 9.996745999674602e-06, + "loss": 0.34, + "step": 3256 + }, + { + "epoch": 1.5141794514179452, + "grad_norm": 4.912278652191162, + "learning_rate": 9.9967449996745e-06, + "loss": 0.4679, + "step": 3257 + }, + { + "epoch": 1.514644351464435, + "grad_norm": 4.588719844818115, + "learning_rate": 9.996743999674401e-06, + "loss": 0.5784, + "step": 3258 + }, + { + "epoch": 1.5151092515109252, + "grad_norm": 5.816802978515625, + "learning_rate": 9.9967429996743e-06, + "loss": 0.5533, + "step": 3259 + }, + { + "epoch": 1.515574151557415, + "grad_norm": 7.150609970092773, + "learning_rate": 9.996741999674201e-06, + "loss": 0.7633, + "step": 3260 + }, + { + "epoch": 1.5160390516039053, + "grad_norm": 4.784478187561035, + "learning_rate": 9.9967409996741e-06, + "loss": 0.4882, + "step": 3261 + }, + { + "epoch": 1.516503951650395, + "grad_norm": 5.363261699676514, + "learning_rate": 9.996739999674e-06, + "loss": 0.6699, + "step": 3262 + }, + { + "epoch": 1.5169688516968853, + "grad_norm": 6.985927104949951, + "learning_rate": 9.996738999673901e-06, + "loss": 0.7743, + "step": 3263 + }, + { + "epoch": 1.5174337517433751, + "grad_norm": 5.223776817321777, + "learning_rate": 9.996737999673802e-06, + "loss": 0.5173, + "step": 3264 + }, + { + "epoch": 1.5178986517898652, + "grad_norm": 6.065086364746094, + "learning_rate": 9.9967369996737e-06, + "loss": 0.5084, + "step": 3265 + }, + { + "epoch": 1.5183635518363552, + "grad_norm": 4.967368125915527, + "learning_rate": 9.996735999673601e-06, + "loss": 0.4487, + "step": 3266 + }, + { + "epoch": 1.5188284518828452, + "grad_norm": 7.925069332122803, + "learning_rate": 9.9967349996735e-06, + "loss": 0.5779, + "step": 3267 + }, + { + "epoch": 1.5192933519293352, + "grad_norm": 10.088817596435547, + "learning_rate": 9.996733999673401e-06, + "loss": 0.6478, + "step": 3268 + }, + { + "epoch": 1.519758251975825, + "grad_norm": 8.13489818572998, + "learning_rate": 9.996732999673302e-06, + "loss": 0.6426, + "step": 3269 + }, + { + "epoch": 1.5202231520223153, + "grad_norm": 4.362918853759766, + "learning_rate": 9.9967319996732e-06, + "loss": 0.2894, + "step": 3270 + }, + { + "epoch": 1.520688052068805, + "grad_norm": 4.3630757331848145, + "learning_rate": 9.996730999673101e-06, + "loss": 0.3466, + "step": 3271 + }, + { + "epoch": 1.5211529521152953, + "grad_norm": 9.412307739257812, + "learning_rate": 9.996729999673e-06, + "loss": 0.9311, + "step": 3272 + }, + { + "epoch": 1.5216178521617851, + "grad_norm": 7.164676189422607, + "learning_rate": 9.996728999672901e-06, + "loss": 0.5045, + "step": 3273 + }, + { + "epoch": 1.5220827522082754, + "grad_norm": 6.527884483337402, + "learning_rate": 9.9967279996728e-06, + "loss": 0.4899, + "step": 3274 + }, + { + "epoch": 1.5225476522547652, + "grad_norm": 5.225215911865234, + "learning_rate": 9.9967269996727e-06, + "loss": 0.4689, + "step": 3275 + }, + { + "epoch": 1.5230125523012552, + "grad_norm": 6.976796627044678, + "learning_rate": 9.996725999672601e-06, + "loss": 0.7314, + "step": 3276 + }, + { + "epoch": 1.5234774523477452, + "grad_norm": 4.319855213165283, + "learning_rate": 9.996724999672502e-06, + "loss": 0.3187, + "step": 3277 + }, + { + "epoch": 1.5239423523942353, + "grad_norm": 4.590564250946045, + "learning_rate": 9.996723999672401e-06, + "loss": 0.505, + "step": 3278 + }, + { + "epoch": 1.5244072524407253, + "grad_norm": 4.826691150665283, + "learning_rate": 9.9967229996723e-06, + "loss": 0.4844, + "step": 3279 + }, + { + "epoch": 1.524872152487215, + "grad_norm": 3.939751386642456, + "learning_rate": 9.9967219996722e-06, + "loss": 0.1698, + "step": 3280 + }, + { + "epoch": 1.5253370525337053, + "grad_norm": 5.0786895751953125, + "learning_rate": 9.996720999672101e-06, + "loss": 0.4684, + "step": 3281 + }, + { + "epoch": 1.5258019525801951, + "grad_norm": 5.976478099822998, + "learning_rate": 9.996719999672002e-06, + "loss": 0.5724, + "step": 3282 + }, + { + "epoch": 1.5262668526266854, + "grad_norm": 7.112088680267334, + "learning_rate": 9.9967189996719e-06, + "loss": 0.8633, + "step": 3283 + }, + { + "epoch": 1.5267317526731752, + "grad_norm": 8.540013313293457, + "learning_rate": 9.996717999671801e-06, + "loss": 1.024, + "step": 3284 + }, + { + "epoch": 1.5271966527196654, + "grad_norm": 8.847336769104004, + "learning_rate": 9.9967169996717e-06, + "loss": 0.2572, + "step": 3285 + }, + { + "epoch": 1.5276615527661552, + "grad_norm": 8.75747013092041, + "learning_rate": 9.996715999671601e-06, + "loss": 0.9145, + "step": 3286 + }, + { + "epoch": 1.5281264528126453, + "grad_norm": 5.360870361328125, + "learning_rate": 9.9967149996715e-06, + "loss": 0.3716, + "step": 3287 + }, + { + "epoch": 1.5285913528591353, + "grad_norm": 5.4536638259887695, + "learning_rate": 9.9967139996714e-06, + "loss": 0.4115, + "step": 3288 + }, + { + "epoch": 1.5290562529056253, + "grad_norm": 7.026416778564453, + "learning_rate": 9.996712999671301e-06, + "loss": 0.4503, + "step": 3289 + }, + { + "epoch": 1.5295211529521153, + "grad_norm": 5.092644214630127, + "learning_rate": 9.9967119996712e-06, + "loss": 0.5029, + "step": 3290 + }, + { + "epoch": 1.5299860529986054, + "grad_norm": 4.618101119995117, + "learning_rate": 9.996710999671101e-06, + "loss": 0.524, + "step": 3291 + }, + { + "epoch": 1.5304509530450954, + "grad_norm": 4.123658180236816, + "learning_rate": 9.996709999671e-06, + "loss": 0.2771, + "step": 3292 + }, + { + "epoch": 1.5309158530915852, + "grad_norm": 4.59637451171875, + "learning_rate": 9.9967089996709e-06, + "loss": 0.3096, + "step": 3293 + }, + { + "epoch": 1.5313807531380754, + "grad_norm": 5.348351955413818, + "learning_rate": 9.996707999670801e-06, + "loss": 0.6408, + "step": 3294 + }, + { + "epoch": 1.5318456531845652, + "grad_norm": 5.580860137939453, + "learning_rate": 9.996706999670702e-06, + "loss": 0.6376, + "step": 3295 + }, + { + "epoch": 1.5323105532310555, + "grad_norm": 8.568695068359375, + "learning_rate": 9.9967059996706e-06, + "loss": 0.9832, + "step": 3296 + }, + { + "epoch": 1.5327754532775453, + "grad_norm": 5.116918563842773, + "learning_rate": 9.996704999670502e-06, + "loss": 0.6418, + "step": 3297 + }, + { + "epoch": 1.5332403533240353, + "grad_norm": 5.5918660163879395, + "learning_rate": 9.9967039996704e-06, + "loss": 0.4189, + "step": 3298 + }, + { + "epoch": 1.5337052533705253, + "grad_norm": 13.76392936706543, + "learning_rate": 9.996702999670301e-06, + "loss": 1.1724, + "step": 3299 + }, + { + "epoch": 1.5341701534170153, + "grad_norm": 4.4357218742370605, + "learning_rate": 9.9967019996702e-06, + "loss": 0.33, + "step": 3300 + }, + { + "epoch": 1.5346350534635054, + "grad_norm": 7.07509708404541, + "learning_rate": 9.9967009996701e-06, + "loss": 0.5027, + "step": 3301 + }, + { + "epoch": 1.5350999535099954, + "grad_norm": 7.135797500610352, + "learning_rate": 9.996699999670001e-06, + "loss": 0.7018, + "step": 3302 + }, + { + "epoch": 1.5355648535564854, + "grad_norm": 6.773794651031494, + "learning_rate": 9.9966989996699e-06, + "loss": 0.6565, + "step": 3303 + }, + { + "epoch": 1.5360297536029752, + "grad_norm": 5.062862873077393, + "learning_rate": 9.996697999669801e-06, + "loss": 0.4375, + "step": 3304 + }, + { + "epoch": 1.5364946536494655, + "grad_norm": 5.538901329040527, + "learning_rate": 9.9966969996697e-06, + "loss": 0.6244, + "step": 3305 + }, + { + "epoch": 1.5369595536959553, + "grad_norm": 6.095108985900879, + "learning_rate": 9.9966959996696e-06, + "loss": 0.5565, + "step": 3306 + }, + { + "epoch": 1.5374244537424455, + "grad_norm": 6.8282670974731445, + "learning_rate": 9.996694999669501e-06, + "loss": 0.824, + "step": 3307 + }, + { + "epoch": 1.5378893537889353, + "grad_norm": 7.446592807769775, + "learning_rate": 9.996693999669402e-06, + "loss": 0.682, + "step": 3308 + }, + { + "epoch": 1.5383542538354253, + "grad_norm": 5.923242568969727, + "learning_rate": 9.996692999669301e-06, + "loss": 0.8299, + "step": 3309 + }, + { + "epoch": 1.5388191538819154, + "grad_norm": 6.90551233291626, + "learning_rate": 9.9966919996692e-06, + "loss": 0.5968, + "step": 3310 + }, + { + "epoch": 1.5392840539284054, + "grad_norm": 5.648077011108398, + "learning_rate": 9.9966909996691e-06, + "loss": 0.5556, + "step": 3311 + }, + { + "epoch": 1.5397489539748954, + "grad_norm": 6.771881103515625, + "learning_rate": 9.996689999669e-06, + "loss": 0.6887, + "step": 3312 + }, + { + "epoch": 1.5402138540213854, + "grad_norm": 9.854801177978516, + "learning_rate": 9.9966889996689e-06, + "loss": 0.8543, + "step": 3313 + }, + { + "epoch": 1.5406787540678755, + "grad_norm": 6.462681293487549, + "learning_rate": 9.9966879996688e-06, + "loss": 0.796, + "step": 3314 + }, + { + "epoch": 1.5411436541143653, + "grad_norm": 4.384909629821777, + "learning_rate": 9.996686999668701e-06, + "loss": 0.3088, + "step": 3315 + }, + { + "epoch": 1.5416085541608555, + "grad_norm": 4.6994123458862305, + "learning_rate": 9.9966859996686e-06, + "loss": 0.4438, + "step": 3316 + }, + { + "epoch": 1.5420734542073453, + "grad_norm": 5.243156433105469, + "learning_rate": 9.996684999668501e-06, + "loss": 0.3279, + "step": 3317 + }, + { + "epoch": 1.5425383542538356, + "grad_norm": 9.425540924072266, + "learning_rate": 9.9966839996684e-06, + "loss": 0.5118, + "step": 3318 + }, + { + "epoch": 1.5430032543003254, + "grad_norm": 4.336418628692627, + "learning_rate": 9.9966829996683e-06, + "loss": 0.4465, + "step": 3319 + }, + { + "epoch": 1.5434681543468154, + "grad_norm": 8.952005386352539, + "learning_rate": 9.996681999668201e-06, + "loss": 0.4163, + "step": 3320 + }, + { + "epoch": 1.5439330543933054, + "grad_norm": 4.949583530426025, + "learning_rate": 9.9966809996681e-06, + "loss": 0.6166, + "step": 3321 + }, + { + "epoch": 1.5443979544397954, + "grad_norm": 12.75858211517334, + "learning_rate": 9.996679999668001e-06, + "loss": 2.8892, + "step": 3322 + }, + { + "epoch": 1.5448628544862855, + "grad_norm": 4.265242576599121, + "learning_rate": 9.9966789996679e-06, + "loss": 0.3393, + "step": 3323 + }, + { + "epoch": 1.5453277545327755, + "grad_norm": 6.238925933837891, + "learning_rate": 9.9966779996678e-06, + "loss": 0.6353, + "step": 3324 + }, + { + "epoch": 1.5457926545792655, + "grad_norm": 7.1703996658325195, + "learning_rate": 9.996676999667701e-06, + "loss": 0.7685, + "step": 3325 + }, + { + "epoch": 1.5462575546257553, + "grad_norm": 6.917335033416748, + "learning_rate": 9.9966759996676e-06, + "loss": 0.6958, + "step": 3326 + }, + { + "epoch": 1.5467224546722456, + "grad_norm": 4.607722759246826, + "learning_rate": 9.996674999667501e-06, + "loss": 0.3212, + "step": 3327 + }, + { + "epoch": 1.5471873547187354, + "grad_norm": 4.51751708984375, + "learning_rate": 9.996673999667402e-06, + "loss": 0.3711, + "step": 3328 + }, + { + "epoch": 1.5476522547652256, + "grad_norm": 5.351672649383545, + "learning_rate": 9.9966729996673e-06, + "loss": 0.3607, + "step": 3329 + }, + { + "epoch": 1.5481171548117154, + "grad_norm": 7.5239338874816895, + "learning_rate": 9.996671999667201e-06, + "loss": 0.7877, + "step": 3330 + }, + { + "epoch": 1.5485820548582054, + "grad_norm": 5.704563140869141, + "learning_rate": 9.9966709996671e-06, + "loss": 0.4933, + "step": 3331 + }, + { + "epoch": 1.5490469549046955, + "grad_norm": 8.74876880645752, + "learning_rate": 9.996669999667e-06, + "loss": 0.6763, + "step": 3332 + }, + { + "epoch": 1.5495118549511855, + "grad_norm": 7.489928722381592, + "learning_rate": 9.996668999666901e-06, + "loss": 1.0384, + "step": 3333 + }, + { + "epoch": 1.5499767549976755, + "grad_norm": 3.99367094039917, + "learning_rate": 9.9966679996668e-06, + "loss": 0.3219, + "step": 3334 + }, + { + "epoch": 1.5504416550441655, + "grad_norm": 6.687647342681885, + "learning_rate": 9.996666999666701e-06, + "loss": 0.4957, + "step": 3335 + }, + { + "epoch": 1.5509065550906556, + "grad_norm": 4.276191234588623, + "learning_rate": 9.9966659996666e-06, + "loss": 0.4058, + "step": 3336 + }, + { + "epoch": 1.5513714551371454, + "grad_norm": 6.0826592445373535, + "learning_rate": 9.9966649996665e-06, + "loss": 0.88, + "step": 3337 + }, + { + "epoch": 1.5518363551836356, + "grad_norm": 8.507759094238281, + "learning_rate": 9.996663999666401e-06, + "loss": 0.7227, + "step": 3338 + }, + { + "epoch": 1.5523012552301254, + "grad_norm": 5.456730365753174, + "learning_rate": 9.9966629996663e-06, + "loss": 0.4209, + "step": 3339 + }, + { + "epoch": 1.5527661552766157, + "grad_norm": 5.092073917388916, + "learning_rate": 9.996661999666201e-06, + "loss": 0.4205, + "step": 3340 + }, + { + "epoch": 1.5532310553231055, + "grad_norm": 6.718460559844971, + "learning_rate": 9.996660999666102e-06, + "loss": 0.9128, + "step": 3341 + }, + { + "epoch": 1.5536959553695955, + "grad_norm": 5.908188343048096, + "learning_rate": 9.996659999666e-06, + "loss": 0.4408, + "step": 3342 + }, + { + "epoch": 1.5541608554160855, + "grad_norm": 5.546911239624023, + "learning_rate": 9.9966589996659e-06, + "loss": 0.4108, + "step": 3343 + }, + { + "epoch": 1.5546257554625755, + "grad_norm": 8.816423416137695, + "learning_rate": 9.9966579996658e-06, + "loss": 0.9292, + "step": 3344 + }, + { + "epoch": 1.5550906555090656, + "grad_norm": 5.323498725891113, + "learning_rate": 9.996656999665701e-06, + "loss": 0.4823, + "step": 3345 + }, + { + "epoch": 1.5555555555555556, + "grad_norm": 8.943218231201172, + "learning_rate": 9.996655999665602e-06, + "loss": 0.6407, + "step": 3346 + }, + { + "epoch": 1.5560204556020456, + "grad_norm": 5.508942127227783, + "learning_rate": 9.9966549996655e-06, + "loss": 0.5111, + "step": 3347 + }, + { + "epoch": 1.5564853556485354, + "grad_norm": 6.876700401306152, + "learning_rate": 9.996653999665401e-06, + "loss": 0.8197, + "step": 3348 + }, + { + "epoch": 1.5569502556950257, + "grad_norm": 4.199520111083984, + "learning_rate": 9.9966529996653e-06, + "loss": 0.3276, + "step": 3349 + }, + { + "epoch": 1.5574151557415155, + "grad_norm": 5.272153377532959, + "learning_rate": 9.9966519996652e-06, + "loss": 0.6554, + "step": 3350 + }, + { + "epoch": 1.5578800557880057, + "grad_norm": 6.533261775970459, + "learning_rate": 9.996650999665101e-06, + "loss": 0.6469, + "step": 3351 + }, + { + "epoch": 1.5583449558344955, + "grad_norm": 8.158873558044434, + "learning_rate": 9.996649999665002e-06, + "loss": 0.5422, + "step": 3352 + }, + { + "epoch": 1.5588098558809858, + "grad_norm": 3.9577295780181885, + "learning_rate": 9.996648999664901e-06, + "loss": 0.461, + "step": 3353 + }, + { + "epoch": 1.5592747559274756, + "grad_norm": 4.876617908477783, + "learning_rate": 9.9966479996648e-06, + "loss": 0.3908, + "step": 3354 + }, + { + "epoch": 1.5597396559739656, + "grad_norm": 6.469640254974365, + "learning_rate": 9.9966469996647e-06, + "loss": 0.683, + "step": 3355 + }, + { + "epoch": 1.5602045560204556, + "grad_norm": 5.645909786224365, + "learning_rate": 9.9966459996646e-06, + "loss": 0.5529, + "step": 3356 + }, + { + "epoch": 1.5606694560669456, + "grad_norm": 7.724992275238037, + "learning_rate": 9.9966449996645e-06, + "loss": 0.7014, + "step": 3357 + }, + { + "epoch": 1.5611343561134357, + "grad_norm": 8.298425674438477, + "learning_rate": 9.996643999664401e-06, + "loss": 0.8113, + "step": 3358 + }, + { + "epoch": 1.5615992561599255, + "grad_norm": 8.420448303222656, + "learning_rate": 9.996642999664302e-06, + "loss": 0.8662, + "step": 3359 + }, + { + "epoch": 1.5620641562064157, + "grad_norm": 8.356173515319824, + "learning_rate": 9.9966419996642e-06, + "loss": 1.0151, + "step": 3360 + }, + { + "epoch": 1.5625290562529055, + "grad_norm": 7.5408220291137695, + "learning_rate": 9.996640999664101e-06, + "loss": 0.72, + "step": 3361 + }, + { + "epoch": 1.5629939562993957, + "grad_norm": 7.340937614440918, + "learning_rate": 9.996639999664e-06, + "loss": 0.6822, + "step": 3362 + }, + { + "epoch": 1.5634588563458856, + "grad_norm": 11.60794734954834, + "learning_rate": 9.996638999663901e-06, + "loss": 1.0941, + "step": 3363 + }, + { + "epoch": 1.5639237563923758, + "grad_norm": 12.510573387145996, + "learning_rate": 9.996637999663802e-06, + "loss": 0.7837, + "step": 3364 + }, + { + "epoch": 1.5643886564388656, + "grad_norm": 7.204160213470459, + "learning_rate": 9.9966369996637e-06, + "loss": 0.763, + "step": 3365 + }, + { + "epoch": 1.5648535564853556, + "grad_norm": 6.0938544273376465, + "learning_rate": 9.996635999663601e-06, + "loss": 0.6269, + "step": 3366 + }, + { + "epoch": 1.5653184565318456, + "grad_norm": 10.121529579162598, + "learning_rate": 9.9966349996635e-06, + "loss": 0.6321, + "step": 3367 + }, + { + "epoch": 1.5657833565783357, + "grad_norm": 5.722692012786865, + "learning_rate": 9.9966339996634e-06, + "loss": 0.4478, + "step": 3368 + }, + { + "epoch": 1.5662482566248257, + "grad_norm": 4.534372329711914, + "learning_rate": 9.9966329996633e-06, + "loss": 0.3428, + "step": 3369 + }, + { + "epoch": 1.5667131566713157, + "grad_norm": 3.946101188659668, + "learning_rate": 9.9966319996632e-06, + "loss": 0.3186, + "step": 3370 + }, + { + "epoch": 1.5671780567178057, + "grad_norm": 5.611434459686279, + "learning_rate": 9.996630999663101e-06, + "loss": 0.5318, + "step": 3371 + }, + { + "epoch": 1.5676429567642955, + "grad_norm": 5.5816168785095215, + "learning_rate": 9.996629999663002e-06, + "loss": 0.3411, + "step": 3372 + }, + { + "epoch": 1.5681078568107858, + "grad_norm": 6.712675094604492, + "learning_rate": 9.9966289996629e-06, + "loss": 0.5479, + "step": 3373 + }, + { + "epoch": 1.5685727568572756, + "grad_norm": 6.188445091247559, + "learning_rate": 9.996627999662801e-06, + "loss": 0.3897, + "step": 3374 + }, + { + "epoch": 1.5690376569037658, + "grad_norm": 4.631490707397461, + "learning_rate": 9.9966269996627e-06, + "loss": 0.3115, + "step": 3375 + }, + { + "epoch": 1.5695025569502556, + "grad_norm": 8.815617561340332, + "learning_rate": 9.996625999662601e-06, + "loss": 0.5772, + "step": 3376 + }, + { + "epoch": 1.5699674569967457, + "grad_norm": 10.276517868041992, + "learning_rate": 9.996624999662502e-06, + "loss": 0.8879, + "step": 3377 + }, + { + "epoch": 1.5704323570432357, + "grad_norm": 9.898716926574707, + "learning_rate": 9.9966239996624e-06, + "loss": 1.1083, + "step": 3378 + }, + { + "epoch": 1.5708972570897257, + "grad_norm": 4.007884979248047, + "learning_rate": 9.996622999662301e-06, + "loss": 0.331, + "step": 3379 + }, + { + "epoch": 1.5713621571362157, + "grad_norm": 5.625577926635742, + "learning_rate": 9.9966219996622e-06, + "loss": 0.5188, + "step": 3380 + }, + { + "epoch": 1.5718270571827058, + "grad_norm": 7.170387268066406, + "learning_rate": 9.996620999662101e-06, + "loss": 0.6247, + "step": 3381 + }, + { + "epoch": 1.5722919572291958, + "grad_norm": 6.122554779052734, + "learning_rate": 9.996619999662e-06, + "loss": 0.6108, + "step": 3382 + }, + { + "epoch": 1.5727568572756856, + "grad_norm": 6.879653453826904, + "learning_rate": 9.9966189996619e-06, + "loss": 0.6003, + "step": 3383 + }, + { + "epoch": 1.5732217573221758, + "grad_norm": 5.614171981811523, + "learning_rate": 9.996617999661801e-06, + "loss": 0.4729, + "step": 3384 + }, + { + "epoch": 1.5736866573686656, + "grad_norm": 7.206139087677002, + "learning_rate": 9.996616999661702e-06, + "loss": 0.5135, + "step": 3385 + }, + { + "epoch": 1.5741515574151559, + "grad_norm": 7.65886926651001, + "learning_rate": 9.9966159996616e-06, + "loss": 0.6937, + "step": 3386 + }, + { + "epoch": 1.5746164574616457, + "grad_norm": 3.9348814487457275, + "learning_rate": 9.9966149996615e-06, + "loss": 0.479, + "step": 3387 + }, + { + "epoch": 1.5750813575081357, + "grad_norm": 7.467320919036865, + "learning_rate": 9.9966139996614e-06, + "loss": 0.6522, + "step": 3388 + }, + { + "epoch": 1.5755462575546257, + "grad_norm": 5.932638168334961, + "learning_rate": 9.996612999661301e-06, + "loss": 0.6802, + "step": 3389 + }, + { + "epoch": 1.5760111576011158, + "grad_norm": 7.070004463195801, + "learning_rate": 9.996611999661202e-06, + "loss": 0.72, + "step": 3390 + }, + { + "epoch": 1.5764760576476058, + "grad_norm": 5.205458164215088, + "learning_rate": 9.9966109996611e-06, + "loss": 0.6223, + "step": 3391 + }, + { + "epoch": 1.5769409576940958, + "grad_norm": 3.693929672241211, + "learning_rate": 9.996609999661001e-06, + "loss": 0.3031, + "step": 3392 + }, + { + "epoch": 1.5774058577405858, + "grad_norm": 4.631025314331055, + "learning_rate": 9.9966089996609e-06, + "loss": 0.5127, + "step": 3393 + }, + { + "epoch": 1.5778707577870756, + "grad_norm": 6.707803726196289, + "learning_rate": 9.996607999660801e-06, + "loss": 0.7505, + "step": 3394 + }, + { + "epoch": 1.5783356578335659, + "grad_norm": 6.388518333435059, + "learning_rate": 9.9966069996607e-06, + "loss": 0.586, + "step": 3395 + }, + { + "epoch": 1.5788005578800557, + "grad_norm": 4.258230209350586, + "learning_rate": 9.9966059996606e-06, + "loss": 0.2269, + "step": 3396 + }, + { + "epoch": 1.579265457926546, + "grad_norm": 7.381649017333984, + "learning_rate": 9.996604999660501e-06, + "loss": 0.8259, + "step": 3397 + }, + { + "epoch": 1.5797303579730357, + "grad_norm": 4.3709635734558105, + "learning_rate": 9.9966039996604e-06, + "loss": 0.387, + "step": 3398 + }, + { + "epoch": 1.5801952580195258, + "grad_norm": 6.483020782470703, + "learning_rate": 9.9966029996603e-06, + "loss": 0.7279, + "step": 3399 + }, + { + "epoch": 1.5806601580660158, + "grad_norm": 5.160656929016113, + "learning_rate": 9.9966019996602e-06, + "loss": 0.4913, + "step": 3400 + }, + { + "epoch": 1.5811250581125058, + "grad_norm": 5.024487018585205, + "learning_rate": 9.9966009996601e-06, + "loss": 0.5099, + "step": 3401 + }, + { + "epoch": 1.5815899581589958, + "grad_norm": 7.293968677520752, + "learning_rate": 9.996599999660001e-06, + "loss": 0.8257, + "step": 3402 + }, + { + "epoch": 1.5820548582054859, + "grad_norm": 6.756013870239258, + "learning_rate": 9.996598999659902e-06, + "loss": 0.7344, + "step": 3403 + }, + { + "epoch": 1.5825197582519759, + "grad_norm": 5.266984462738037, + "learning_rate": 9.9965979996598e-06, + "loss": 0.6639, + "step": 3404 + }, + { + "epoch": 1.5829846582984657, + "grad_norm": 5.262167453765869, + "learning_rate": 9.996596999659701e-06, + "loss": 0.5144, + "step": 3405 + }, + { + "epoch": 1.583449558344956, + "grad_norm": 5.678562164306641, + "learning_rate": 9.9965959996596e-06, + "loss": 0.4506, + "step": 3406 + }, + { + "epoch": 1.5839144583914457, + "grad_norm": 8.263038635253906, + "learning_rate": 9.996594999659501e-06, + "loss": 1.0178, + "step": 3407 + }, + { + "epoch": 1.584379358437936, + "grad_norm": 6.508664131164551, + "learning_rate": 9.9965939996594e-06, + "loss": 0.6437, + "step": 3408 + }, + { + "epoch": 1.5848442584844258, + "grad_norm": 7.456343173980713, + "learning_rate": 9.9965929996593e-06, + "loss": 0.991, + "step": 3409 + }, + { + "epoch": 1.5853091585309158, + "grad_norm": 5.842319488525391, + "learning_rate": 9.996591999659201e-06, + "loss": 0.6379, + "step": 3410 + }, + { + "epoch": 1.5857740585774058, + "grad_norm": 4.624527931213379, + "learning_rate": 9.9965909996591e-06, + "loss": 0.329, + "step": 3411 + }, + { + "epoch": 1.5862389586238959, + "grad_norm": 7.40773344039917, + "learning_rate": 9.996589999659001e-06, + "loss": 0.5116, + "step": 3412 + }, + { + "epoch": 1.5867038586703859, + "grad_norm": 7.311116695404053, + "learning_rate": 9.9965889996589e-06, + "loss": 0.6498, + "step": 3413 + }, + { + "epoch": 1.587168758716876, + "grad_norm": 5.558883190155029, + "learning_rate": 9.9965879996588e-06, + "loss": 0.5997, + "step": 3414 + }, + { + "epoch": 1.587633658763366, + "grad_norm": 6.626638412475586, + "learning_rate": 9.996586999658701e-06, + "loss": 0.5927, + "step": 3415 + }, + { + "epoch": 1.5880985588098557, + "grad_norm": 6.16867208480835, + "learning_rate": 9.996585999658602e-06, + "loss": 0.5899, + "step": 3416 + }, + { + "epoch": 1.588563458856346, + "grad_norm": 5.019034385681152, + "learning_rate": 9.9965849996585e-06, + "loss": 0.3038, + "step": 3417 + }, + { + "epoch": 1.5890283589028358, + "grad_norm": 5.612858295440674, + "learning_rate": 9.996583999658401e-06, + "loss": 0.3409, + "step": 3418 + }, + { + "epoch": 1.589493258949326, + "grad_norm": 7.875426292419434, + "learning_rate": 9.9965829996583e-06, + "loss": 0.4831, + "step": 3419 + }, + { + "epoch": 1.5899581589958158, + "grad_norm": 3.9803242683410645, + "learning_rate": 9.996581999658201e-06, + "loss": 0.4429, + "step": 3420 + }, + { + "epoch": 1.5904230590423059, + "grad_norm": 7.800559043884277, + "learning_rate": 9.996580999658102e-06, + "loss": 0.4301, + "step": 3421 + }, + { + "epoch": 1.5908879590887959, + "grad_norm": 7.185056209564209, + "learning_rate": 9.996579999658e-06, + "loss": 0.7116, + "step": 3422 + }, + { + "epoch": 1.591352859135286, + "grad_norm": 5.0339508056640625, + "learning_rate": 9.996578999657901e-06, + "loss": 0.4124, + "step": 3423 + }, + { + "epoch": 1.591817759181776, + "grad_norm": 6.425482749938965, + "learning_rate": 9.9965779996578e-06, + "loss": 0.6254, + "step": 3424 + }, + { + "epoch": 1.592282659228266, + "grad_norm": 8.443370819091797, + "learning_rate": 9.996576999657701e-06, + "loss": 0.6574, + "step": 3425 + }, + { + "epoch": 1.592747559274756, + "grad_norm": 4.713174819946289, + "learning_rate": 9.9965759996576e-06, + "loss": 0.3537, + "step": 3426 + }, + { + "epoch": 1.5932124593212458, + "grad_norm": 5.9546217918396, + "learning_rate": 9.9965749996575e-06, + "loss": 0.419, + "step": 3427 + }, + { + "epoch": 1.593677359367736, + "grad_norm": 6.994750499725342, + "learning_rate": 9.996573999657401e-06, + "loss": 0.6589, + "step": 3428 + }, + { + "epoch": 1.5941422594142258, + "grad_norm": 4.974900245666504, + "learning_rate": 9.996572999657302e-06, + "loss": 0.3083, + "step": 3429 + }, + { + "epoch": 1.594607159460716, + "grad_norm": 9.240700721740723, + "learning_rate": 9.996571999657201e-06, + "loss": 0.99, + "step": 3430 + }, + { + "epoch": 1.5950720595072059, + "grad_norm": 4.559171199798584, + "learning_rate": 9.9965709996571e-06, + "loss": 0.3815, + "step": 3431 + }, + { + "epoch": 1.5955369595536961, + "grad_norm": 3.4001352787017822, + "learning_rate": 9.996569999657e-06, + "loss": 0.2768, + "step": 3432 + }, + { + "epoch": 1.596001859600186, + "grad_norm": 3.9652140140533447, + "learning_rate": 9.996568999656901e-06, + "loss": 0.416, + "step": 3433 + }, + { + "epoch": 1.596466759646676, + "grad_norm": 4.813519477844238, + "learning_rate": 9.996567999656802e-06, + "loss": 0.5289, + "step": 3434 + }, + { + "epoch": 1.596931659693166, + "grad_norm": 3.8269736766815186, + "learning_rate": 9.9965669996567e-06, + "loss": 0.2894, + "step": 3435 + }, + { + "epoch": 1.597396559739656, + "grad_norm": 5.229704856872559, + "learning_rate": 9.996565999656601e-06, + "loss": 0.6977, + "step": 3436 + }, + { + "epoch": 1.597861459786146, + "grad_norm": 4.67594051361084, + "learning_rate": 9.9965649996565e-06, + "loss": 0.3452, + "step": 3437 + }, + { + "epoch": 1.5983263598326358, + "grad_norm": 4.760647296905518, + "learning_rate": 9.996563999656401e-06, + "loss": 0.328, + "step": 3438 + }, + { + "epoch": 1.598791259879126, + "grad_norm": 5.959355354309082, + "learning_rate": 9.9965629996563e-06, + "loss": 0.5969, + "step": 3439 + }, + { + "epoch": 1.5992561599256159, + "grad_norm": 4.411102294921875, + "learning_rate": 9.9965619996562e-06, + "loss": 0.3291, + "step": 3440 + }, + { + "epoch": 1.5997210599721061, + "grad_norm": 5.2322564125061035, + "learning_rate": 9.996560999656101e-06, + "loss": 0.7078, + "step": 3441 + }, + { + "epoch": 1.600185960018596, + "grad_norm": 15.361722946166992, + "learning_rate": 9.996559999656e-06, + "loss": 1.4656, + "step": 3442 + }, + { + "epoch": 1.6006508600650862, + "grad_norm": 7.743661403656006, + "learning_rate": 9.996558999655901e-06, + "loss": 0.8122, + "step": 3443 + }, + { + "epoch": 1.601115760111576, + "grad_norm": 7.372518539428711, + "learning_rate": 9.9965579996558e-06, + "loss": 0.87, + "step": 3444 + }, + { + "epoch": 1.601580660158066, + "grad_norm": 5.901437282562256, + "learning_rate": 9.9965569996557e-06, + "loss": 0.4344, + "step": 3445 + }, + { + "epoch": 1.602045560204556, + "grad_norm": 4.512515068054199, + "learning_rate": 9.996555999655601e-06, + "loss": 0.4914, + "step": 3446 + }, + { + "epoch": 1.602510460251046, + "grad_norm": 6.311244010925293, + "learning_rate": 9.996554999655502e-06, + "loss": 0.6187, + "step": 3447 + }, + { + "epoch": 1.602975360297536, + "grad_norm": 4.342381477355957, + "learning_rate": 9.996553999655401e-06, + "loss": 0.3808, + "step": 3448 + }, + { + "epoch": 1.6034402603440259, + "grad_norm": 5.626481056213379, + "learning_rate": 9.996552999655302e-06, + "loss": 0.5287, + "step": 3449 + }, + { + "epoch": 1.6039051603905161, + "grad_norm": 4.867796421051025, + "learning_rate": 9.9965519996552e-06, + "loss": 0.5355, + "step": 3450 + }, + { + "epoch": 1.604370060437006, + "grad_norm": 5.366731643676758, + "learning_rate": 9.9965509996551e-06, + "loss": 0.6982, + "step": 3451 + }, + { + "epoch": 1.6048349604834962, + "grad_norm": 6.0948166847229, + "learning_rate": 9.996549999655e-06, + "loss": 0.5218, + "step": 3452 + }, + { + "epoch": 1.605299860529986, + "grad_norm": 5.801759243011475, + "learning_rate": 9.9965489996549e-06, + "loss": 0.5924, + "step": 3453 + }, + { + "epoch": 1.6057647605764762, + "grad_norm": 4.022035598754883, + "learning_rate": 9.996547999654801e-06, + "loss": 0.233, + "step": 3454 + }, + { + "epoch": 1.606229660622966, + "grad_norm": 5.53837776184082, + "learning_rate": 9.9965469996547e-06, + "loss": 0.5963, + "step": 3455 + }, + { + "epoch": 1.606694560669456, + "grad_norm": 5.282047748565674, + "learning_rate": 9.996545999654601e-06, + "loss": 0.6094, + "step": 3456 + }, + { + "epoch": 1.607159460715946, + "grad_norm": 6.173508167266846, + "learning_rate": 9.9965449996545e-06, + "loss": 0.7184, + "step": 3457 + }, + { + "epoch": 1.607624360762436, + "grad_norm": 6.801712989807129, + "learning_rate": 9.9965439996544e-06, + "loss": 0.648, + "step": 3458 + }, + { + "epoch": 1.6080892608089261, + "grad_norm": 5.313076972961426, + "learning_rate": 9.996542999654301e-06, + "loss": 0.5035, + "step": 3459 + }, + { + "epoch": 1.6085541608554161, + "grad_norm": 5.983987808227539, + "learning_rate": 9.996541999654202e-06, + "loss": 0.4127, + "step": 3460 + }, + { + "epoch": 1.6090190609019062, + "grad_norm": 3.658863067626953, + "learning_rate": 9.996540999654101e-06, + "loss": 0.2078, + "step": 3461 + }, + { + "epoch": 1.609483960948396, + "grad_norm": 3.9470083713531494, + "learning_rate": 9.996539999654e-06, + "loss": 0.3416, + "step": 3462 + }, + { + "epoch": 1.6099488609948862, + "grad_norm": 5.285759449005127, + "learning_rate": 9.9965389996539e-06, + "loss": 0.4881, + "step": 3463 + }, + { + "epoch": 1.610413761041376, + "grad_norm": 6.229647159576416, + "learning_rate": 9.9965379996538e-06, + "loss": 0.6186, + "step": 3464 + }, + { + "epoch": 1.6108786610878663, + "grad_norm": 9.852486610412598, + "learning_rate": 9.9965369996537e-06, + "loss": 1.047, + "step": 3465 + }, + { + "epoch": 1.611343561134356, + "grad_norm": 5.427558422088623, + "learning_rate": 9.996535999653601e-06, + "loss": 0.5578, + "step": 3466 + }, + { + "epoch": 1.611808461180846, + "grad_norm": 6.542568683624268, + "learning_rate": 9.996534999653502e-06, + "loss": 0.8138, + "step": 3467 + }, + { + "epoch": 1.612273361227336, + "grad_norm": 5.957124710083008, + "learning_rate": 9.9965339996534e-06, + "loss": 0.2052, + "step": 3468 + }, + { + "epoch": 1.6127382612738261, + "grad_norm": 6.960590362548828, + "learning_rate": 9.996532999653301e-06, + "loss": 0.4484, + "step": 3469 + }, + { + "epoch": 1.6132031613203162, + "grad_norm": 3.9993691444396973, + "learning_rate": 9.9965319996532e-06, + "loss": 0.2195, + "step": 3470 + }, + { + "epoch": 1.6136680613668062, + "grad_norm": 5.396333694458008, + "learning_rate": 9.9965309996531e-06, + "loss": 0.6066, + "step": 3471 + }, + { + "epoch": 1.6141329614132962, + "grad_norm": 6.002114772796631, + "learning_rate": 9.996529999653001e-06, + "loss": 0.5823, + "step": 3472 + }, + { + "epoch": 1.614597861459786, + "grad_norm": 13.650059700012207, + "learning_rate": 9.9965289996529e-06, + "loss": 0.3473, + "step": 3473 + }, + { + "epoch": 1.6150627615062763, + "grad_norm": 7.264416694641113, + "learning_rate": 9.996527999652801e-06, + "loss": 0.7198, + "step": 3474 + }, + { + "epoch": 1.615527661552766, + "grad_norm": 10.012381553649902, + "learning_rate": 9.9965269996527e-06, + "loss": 0.8048, + "step": 3475 + }, + { + "epoch": 1.6159925615992563, + "grad_norm": 5.321561813354492, + "learning_rate": 9.9965259996526e-06, + "loss": 0.5188, + "step": 3476 + }, + { + "epoch": 1.616457461645746, + "grad_norm": 7.345298767089844, + "learning_rate": 9.9965249996525e-06, + "loss": 0.7448, + "step": 3477 + }, + { + "epoch": 1.6169223616922361, + "grad_norm": 12.077078819274902, + "learning_rate": 9.9965239996524e-06, + "loss": 1.2647, + "step": 3478 + }, + { + "epoch": 1.6173872617387262, + "grad_norm": 5.6606316566467285, + "learning_rate": 9.996522999652301e-06, + "loss": 0.5498, + "step": 3479 + }, + { + "epoch": 1.6178521617852162, + "grad_norm": 6.593053817749023, + "learning_rate": 9.996521999652202e-06, + "loss": 0.4082, + "step": 3480 + }, + { + "epoch": 1.6183170618317062, + "grad_norm": 7.060173511505127, + "learning_rate": 9.9965209996521e-06, + "loss": 0.5263, + "step": 3481 + }, + { + "epoch": 1.6187819618781962, + "grad_norm": 6.553162574768066, + "learning_rate": 9.996519999652001e-06, + "loss": 0.6738, + "step": 3482 + }, + { + "epoch": 1.6192468619246863, + "grad_norm": 9.152692794799805, + "learning_rate": 9.9965189996519e-06, + "loss": 0.8277, + "step": 3483 + }, + { + "epoch": 1.619711761971176, + "grad_norm": 8.372237205505371, + "learning_rate": 9.9965179996518e-06, + "loss": 0.6287, + "step": 3484 + }, + { + "epoch": 1.6201766620176663, + "grad_norm": 5.508787631988525, + "learning_rate": 9.996516999651701e-06, + "loss": 0.6431, + "step": 3485 + }, + { + "epoch": 1.620641562064156, + "grad_norm": 6.575632572174072, + "learning_rate": 9.9965159996516e-06, + "loss": 0.7211, + "step": 3486 + }, + { + "epoch": 1.6211064621106464, + "grad_norm": 7.62727689743042, + "learning_rate": 9.996514999651501e-06, + "loss": 0.7315, + "step": 3487 + }, + { + "epoch": 1.6215713621571362, + "grad_norm": 8.32397747039795, + "learning_rate": 9.9965139996514e-06, + "loss": 0.9656, + "step": 3488 + }, + { + "epoch": 1.6220362622036262, + "grad_norm": 4.459382057189941, + "learning_rate": 9.9965129996513e-06, + "loss": 0.4645, + "step": 3489 + }, + { + "epoch": 1.6225011622501162, + "grad_norm": 5.285238742828369, + "learning_rate": 9.996511999651201e-06, + "loss": 0.454, + "step": 3490 + }, + { + "epoch": 1.6229660622966062, + "grad_norm": 7.442856311798096, + "learning_rate": 9.9965109996511e-06, + "loss": 0.7574, + "step": 3491 + }, + { + "epoch": 1.6234309623430963, + "grad_norm": 3.770479440689087, + "learning_rate": 9.996509999651001e-06, + "loss": 0.3671, + "step": 3492 + }, + { + "epoch": 1.6238958623895863, + "grad_norm": 8.735837936401367, + "learning_rate": 9.996508999650902e-06, + "loss": 0.9258, + "step": 3493 + }, + { + "epoch": 1.6243607624360763, + "grad_norm": 13.479876518249512, + "learning_rate": 9.9965079996508e-06, + "loss": 2.3073, + "step": 3494 + }, + { + "epoch": 1.624825662482566, + "grad_norm": 8.078025817871094, + "learning_rate": 9.9965069996507e-06, + "loss": 0.7524, + "step": 3495 + }, + { + "epoch": 1.6252905625290563, + "grad_norm": 5.681537628173828, + "learning_rate": 9.9965059996506e-06, + "loss": 0.4673, + "step": 3496 + }, + { + "epoch": 1.6257554625755462, + "grad_norm": 4.7437543869018555, + "learning_rate": 9.996504999650501e-06, + "loss": 0.3222, + "step": 3497 + }, + { + "epoch": 1.6262203626220364, + "grad_norm": 4.844328880310059, + "learning_rate": 9.996503999650402e-06, + "loss": 0.3283, + "step": 3498 + }, + { + "epoch": 1.6266852626685262, + "grad_norm": 5.610156536102295, + "learning_rate": 9.9965029996503e-06, + "loss": 0.7991, + "step": 3499 + }, + { + "epoch": 1.6271501627150162, + "grad_norm": 6.0347371101379395, + "learning_rate": 9.996501999650201e-06, + "loss": 0.6551, + "step": 3500 + }, + { + "epoch": 1.6276150627615062, + "grad_norm": 9.47046184539795, + "learning_rate": 9.9965009996501e-06, + "loss": 1.0503, + "step": 3501 + }, + { + "epoch": 1.6280799628079963, + "grad_norm": 6.159233570098877, + "learning_rate": 9.99649999965e-06, + "loss": 0.4355, + "step": 3502 + }, + { + "epoch": 1.6285448628544863, + "grad_norm": 5.11868143081665, + "learning_rate": 9.996498999649901e-06, + "loss": 0.3495, + "step": 3503 + }, + { + "epoch": 1.6290097629009763, + "grad_norm": 4.778929233551025, + "learning_rate": 9.9964979996498e-06, + "loss": 0.3943, + "step": 3504 + }, + { + "epoch": 1.6294746629474663, + "grad_norm": 5.583984851837158, + "learning_rate": 9.996496999649701e-06, + "loss": 0.5022, + "step": 3505 + }, + { + "epoch": 1.6299395629939561, + "grad_norm": 10.956315040588379, + "learning_rate": 9.9964959996496e-06, + "loss": 0.7386, + "step": 3506 + }, + { + "epoch": 1.6304044630404464, + "grad_norm": 6.578444957733154, + "learning_rate": 9.9964949996495e-06, + "loss": 0.5273, + "step": 3507 + }, + { + "epoch": 1.6308693630869362, + "grad_norm": 6.483483791351318, + "learning_rate": 9.9964939996494e-06, + "loss": 0.5158, + "step": 3508 + }, + { + "epoch": 1.6313342631334264, + "grad_norm": 4.566122531890869, + "learning_rate": 9.9964929996493e-06, + "loss": 0.4866, + "step": 3509 + }, + { + "epoch": 1.6317991631799162, + "grad_norm": 5.519860744476318, + "learning_rate": 9.996491999649201e-06, + "loss": 0.5198, + "step": 3510 + }, + { + "epoch": 1.6322640632264063, + "grad_norm": 6.089426040649414, + "learning_rate": 9.996490999649102e-06, + "loss": 0.5468, + "step": 3511 + }, + { + "epoch": 1.6327289632728963, + "grad_norm": 7.374702453613281, + "learning_rate": 9.996489999649e-06, + "loss": 0.5641, + "step": 3512 + }, + { + "epoch": 1.6331938633193863, + "grad_norm": 5.526103496551514, + "learning_rate": 9.996488999648901e-06, + "loss": 0.7343, + "step": 3513 + }, + { + "epoch": 1.6336587633658763, + "grad_norm": 11.051496505737305, + "learning_rate": 9.9964879996488e-06, + "loss": 0.8346, + "step": 3514 + }, + { + "epoch": 1.6341236634123664, + "grad_norm": 7.060535907745361, + "learning_rate": 9.996486999648701e-06, + "loss": 0.5322, + "step": 3515 + }, + { + "epoch": 1.6345885634588564, + "grad_norm": 5.829392910003662, + "learning_rate": 9.996485999648602e-06, + "loss": 0.5919, + "step": 3516 + }, + { + "epoch": 1.6350534635053462, + "grad_norm": 4.187569618225098, + "learning_rate": 9.9964849996485e-06, + "loss": 0.5267, + "step": 3517 + }, + { + "epoch": 1.6355183635518364, + "grad_norm": 6.400564193725586, + "learning_rate": 9.996483999648401e-06, + "loss": 0.7187, + "step": 3518 + }, + { + "epoch": 1.6359832635983262, + "grad_norm": 6.1430182456970215, + "learning_rate": 9.9964829996483e-06, + "loss": 0.6207, + "step": 3519 + }, + { + "epoch": 1.6364481636448165, + "grad_norm": 5.199199676513672, + "learning_rate": 9.9964819996482e-06, + "loss": 0.508, + "step": 3520 + }, + { + "epoch": 1.6369130636913063, + "grad_norm": 4.459421634674072, + "learning_rate": 9.9964809996481e-06, + "loss": 0.3575, + "step": 3521 + }, + { + "epoch": 1.6373779637377965, + "grad_norm": 5.14977502822876, + "learning_rate": 9.996479999648e-06, + "loss": 0.4333, + "step": 3522 + }, + { + "epoch": 1.6378428637842863, + "grad_norm": 6.94900369644165, + "learning_rate": 9.996478999647901e-06, + "loss": 0.8197, + "step": 3523 + }, + { + "epoch": 1.6383077638307764, + "grad_norm": 4.305420875549316, + "learning_rate": 9.996477999647802e-06, + "loss": 0.5675, + "step": 3524 + }, + { + "epoch": 1.6387726638772664, + "grad_norm": 29.704341888427734, + "learning_rate": 9.9964769996477e-06, + "loss": 0.9894, + "step": 3525 + }, + { + "epoch": 1.6392375639237564, + "grad_norm": 5.112705707550049, + "learning_rate": 9.996475999647601e-06, + "loss": 0.5391, + "step": 3526 + }, + { + "epoch": 1.6397024639702464, + "grad_norm": 6.287240982055664, + "learning_rate": 9.9964749996475e-06, + "loss": 0.5124, + "step": 3527 + }, + { + "epoch": 1.6401673640167362, + "grad_norm": 6.207174301147461, + "learning_rate": 9.996473999647401e-06, + "loss": 0.6597, + "step": 3528 + }, + { + "epoch": 1.6406322640632265, + "grad_norm": 5.468986511230469, + "learning_rate": 9.996472999647302e-06, + "loss": 0.6635, + "step": 3529 + }, + { + "epoch": 1.6410971641097163, + "grad_norm": 5.581027030944824, + "learning_rate": 9.9964719996472e-06, + "loss": 0.5139, + "step": 3530 + }, + { + "epoch": 1.6415620641562065, + "grad_norm": 4.731346607208252, + "learning_rate": 9.996470999647101e-06, + "loss": 0.464, + "step": 3531 + }, + { + "epoch": 1.6420269642026963, + "grad_norm": 5.324331760406494, + "learning_rate": 9.996469999647e-06, + "loss": 0.5269, + "step": 3532 + }, + { + "epoch": 1.6424918642491866, + "grad_norm": 6.125245571136475, + "learning_rate": 9.996468999646901e-06, + "loss": 0.654, + "step": 3533 + }, + { + "epoch": 1.6429567642956764, + "grad_norm": 7.1643805503845215, + "learning_rate": 9.9964679996468e-06, + "loss": 0.7443, + "step": 3534 + }, + { + "epoch": 1.6434216643421664, + "grad_norm": 5.784786224365234, + "learning_rate": 9.9964669996467e-06, + "loss": 0.5733, + "step": 3535 + }, + { + "epoch": 1.6438865643886564, + "grad_norm": 9.118682861328125, + "learning_rate": 9.996465999646601e-06, + "loss": 0.9854, + "step": 3536 + }, + { + "epoch": 1.6443514644351465, + "grad_norm": 5.051069259643555, + "learning_rate": 9.996464999646502e-06, + "loss": 0.386, + "step": 3537 + }, + { + "epoch": 1.6448163644816365, + "grad_norm": 8.697630882263184, + "learning_rate": 9.9964639996464e-06, + "loss": 0.7492, + "step": 3538 + }, + { + "epoch": 1.6452812645281265, + "grad_norm": 5.136373043060303, + "learning_rate": 9.9964629996463e-06, + "loss": 0.4407, + "step": 3539 + }, + { + "epoch": 1.6457461645746165, + "grad_norm": 4.894552707672119, + "learning_rate": 9.9964619996462e-06, + "loss": 0.2512, + "step": 3540 + }, + { + "epoch": 1.6462110646211063, + "grad_norm": 5.883354187011719, + "learning_rate": 9.996460999646101e-06, + "loss": 0.523, + "step": 3541 + }, + { + "epoch": 1.6466759646675966, + "grad_norm": 6.201507568359375, + "learning_rate": 9.996459999646002e-06, + "loss": 0.5302, + "step": 3542 + }, + { + "epoch": 1.6471408647140864, + "grad_norm": 6.907440662384033, + "learning_rate": 9.9964589996459e-06, + "loss": 0.8344, + "step": 3543 + }, + { + "epoch": 1.6476057647605766, + "grad_norm": 9.083303451538086, + "learning_rate": 9.996457999645801e-06, + "loss": 0.4169, + "step": 3544 + }, + { + "epoch": 1.6480706648070664, + "grad_norm": 5.520013809204102, + "learning_rate": 9.9964569996457e-06, + "loss": 0.6056, + "step": 3545 + }, + { + "epoch": 1.6485355648535565, + "grad_norm": 5.565127849578857, + "learning_rate": 9.996455999645601e-06, + "loss": 0.6463, + "step": 3546 + }, + { + "epoch": 1.6490004649000465, + "grad_norm": 11.218369483947754, + "learning_rate": 9.9964549996455e-06, + "loss": 0.8135, + "step": 3547 + }, + { + "epoch": 1.6494653649465365, + "grad_norm": 4.804922580718994, + "learning_rate": 9.9964539996454e-06, + "loss": 0.4901, + "step": 3548 + }, + { + "epoch": 1.6499302649930265, + "grad_norm": 5.084923267364502, + "learning_rate": 9.996452999645301e-06, + "loss": 0.3434, + "step": 3549 + }, + { + "epoch": 1.6503951650395166, + "grad_norm": 3.336388349533081, + "learning_rate": 9.9964519996452e-06, + "loss": 0.346, + "step": 3550 + }, + { + "epoch": 1.6508600650860066, + "grad_norm": 4.780605316162109, + "learning_rate": 9.996450999645101e-06, + "loss": 0.4917, + "step": 3551 + }, + { + "epoch": 1.6513249651324964, + "grad_norm": 4.388759613037109, + "learning_rate": 9.996449999645e-06, + "loss": 0.3795, + "step": 3552 + }, + { + "epoch": 1.6517898651789866, + "grad_norm": 6.162712574005127, + "learning_rate": 9.9964489996449e-06, + "loss": 0.624, + "step": 3553 + }, + { + "epoch": 1.6522547652254764, + "grad_norm": 6.8191070556640625, + "learning_rate": 9.996447999644801e-06, + "loss": 0.582, + "step": 3554 + }, + { + "epoch": 1.6527196652719667, + "grad_norm": 8.679901123046875, + "learning_rate": 9.996446999644702e-06, + "loss": 0.7122, + "step": 3555 + }, + { + "epoch": 1.6531845653184565, + "grad_norm": 11.036418914794922, + "learning_rate": 9.9964459996446e-06, + "loss": 0.7715, + "step": 3556 + }, + { + "epoch": 1.6536494653649465, + "grad_norm": 9.319889068603516, + "learning_rate": 9.996444999644501e-06, + "loss": 1.6952, + "step": 3557 + }, + { + "epoch": 1.6541143654114365, + "grad_norm": 8.501663208007812, + "learning_rate": 9.9964439996444e-06, + "loss": 0.7584, + "step": 3558 + }, + { + "epoch": 1.6545792654579266, + "grad_norm": 5.776938438415527, + "learning_rate": 9.996442999644301e-06, + "loss": 0.6914, + "step": 3559 + }, + { + "epoch": 1.6550441655044166, + "grad_norm": 5.168510913848877, + "learning_rate": 9.9964419996442e-06, + "loss": 0.4716, + "step": 3560 + }, + { + "epoch": 1.6555090655509066, + "grad_norm": 8.658303260803223, + "learning_rate": 9.9964409996441e-06, + "loss": 0.4194, + "step": 3561 + }, + { + "epoch": 1.6559739655973966, + "grad_norm": 7.324795722961426, + "learning_rate": 9.996439999644001e-06, + "loss": 0.5676, + "step": 3562 + }, + { + "epoch": 1.6564388656438864, + "grad_norm": 6.174038410186768, + "learning_rate": 9.9964389996439e-06, + "loss": 0.5626, + "step": 3563 + }, + { + "epoch": 1.6569037656903767, + "grad_norm": 5.972703456878662, + "learning_rate": 9.996437999643801e-06, + "loss": 0.6534, + "step": 3564 + }, + { + "epoch": 1.6573686657368665, + "grad_norm": 7.276310920715332, + "learning_rate": 9.9964369996437e-06, + "loss": 0.6502, + "step": 3565 + }, + { + "epoch": 1.6578335657833567, + "grad_norm": 6.540353775024414, + "learning_rate": 9.9964359996436e-06, + "loss": 0.5266, + "step": 3566 + }, + { + "epoch": 1.6582984658298465, + "grad_norm": 6.979139804840088, + "learning_rate": 9.996434999643501e-06, + "loss": 0.5399, + "step": 3567 + }, + { + "epoch": 1.6587633658763365, + "grad_norm": 4.456345558166504, + "learning_rate": 9.996433999643402e-06, + "loss": 0.3257, + "step": 3568 + }, + { + "epoch": 1.6592282659228266, + "grad_norm": 4.566025733947754, + "learning_rate": 9.996432999643301e-06, + "loss": 0.4454, + "step": 3569 + }, + { + "epoch": 1.6596931659693166, + "grad_norm": 8.395252227783203, + "learning_rate": 9.996431999643202e-06, + "loss": 0.5524, + "step": 3570 + }, + { + "epoch": 1.6601580660158066, + "grad_norm": 5.829723358154297, + "learning_rate": 9.9964309996431e-06, + "loss": 0.5785, + "step": 3571 + }, + { + "epoch": 1.6606229660622966, + "grad_norm": 4.0736236572265625, + "learning_rate": 9.996429999643001e-06, + "loss": 0.3834, + "step": 3572 + }, + { + "epoch": 1.6610878661087867, + "grad_norm": 8.824352264404297, + "learning_rate": 9.9964289996429e-06, + "loss": 0.7771, + "step": 3573 + }, + { + "epoch": 1.6615527661552765, + "grad_norm": 10.612982749938965, + "learning_rate": 9.9964279996428e-06, + "loss": 0.5227, + "step": 3574 + }, + { + "epoch": 1.6620176662017667, + "grad_norm": 8.976003646850586, + "learning_rate": 9.996426999642701e-06, + "loss": 0.383, + "step": 3575 + }, + { + "epoch": 1.6624825662482565, + "grad_norm": 7.740998268127441, + "learning_rate": 9.9964259996426e-06, + "loss": 0.7674, + "step": 3576 + }, + { + "epoch": 1.6629474662947468, + "grad_norm": 5.191046237945557, + "learning_rate": 9.996424999642501e-06, + "loss": 0.3971, + "step": 3577 + }, + { + "epoch": 1.6634123663412366, + "grad_norm": 7.02493143081665, + "learning_rate": 9.9964239996424e-06, + "loss": 0.6452, + "step": 3578 + }, + { + "epoch": 1.6638772663877266, + "grad_norm": 4.3740925788879395, + "learning_rate": 9.9964229996423e-06, + "loss": 0.5011, + "step": 3579 + }, + { + "epoch": 1.6643421664342166, + "grad_norm": 7.812779903411865, + "learning_rate": 9.996421999642201e-06, + "loss": 0.8625, + "step": 3580 + }, + { + "epoch": 1.6648070664807066, + "grad_norm": 7.826124668121338, + "learning_rate": 9.996420999642102e-06, + "loss": 0.7766, + "step": 3581 + }, + { + "epoch": 1.6652719665271967, + "grad_norm": 6.472758769989014, + "learning_rate": 9.996419999642001e-06, + "loss": 0.7936, + "step": 3582 + }, + { + "epoch": 1.6657368665736867, + "grad_norm": 3.8113837242126465, + "learning_rate": 9.9964189996419e-06, + "loss": 0.2631, + "step": 3583 + }, + { + "epoch": 1.6662017666201767, + "grad_norm": 6.02735710144043, + "learning_rate": 9.9964179996418e-06, + "loss": 0.4767, + "step": 3584 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 7.807922840118408, + "learning_rate": 9.996416999641701e-06, + "loss": 0.6118, + "step": 3585 + }, + { + "epoch": 1.6671315667131568, + "grad_norm": 7.72080659866333, + "learning_rate": 9.996415999641602e-06, + "loss": 0.5402, + "step": 3586 + }, + { + "epoch": 1.6675964667596466, + "grad_norm": 6.548558235168457, + "learning_rate": 9.9964149996415e-06, + "loss": 0.7357, + "step": 3587 + }, + { + "epoch": 1.6680613668061368, + "grad_norm": 4.867877960205078, + "learning_rate": 9.996413999641401e-06, + "loss": 0.4088, + "step": 3588 + }, + { + "epoch": 1.6685262668526266, + "grad_norm": 5.734102725982666, + "learning_rate": 9.9964129996413e-06, + "loss": 0.4761, + "step": 3589 + }, + { + "epoch": 1.6689911668991166, + "grad_norm": 6.807816505432129, + "learning_rate": 9.996411999641201e-06, + "loss": 0.719, + "step": 3590 + }, + { + "epoch": 1.6694560669456067, + "grad_norm": 6.377374172210693, + "learning_rate": 9.9964109996411e-06, + "loss": 0.3867, + "step": 3591 + }, + { + "epoch": 1.6699209669920967, + "grad_norm": 4.36317253112793, + "learning_rate": 9.996409999641e-06, + "loss": 0.3083, + "step": 3592 + }, + { + "epoch": 1.6703858670385867, + "grad_norm": 7.180074691772461, + "learning_rate": 9.996408999640901e-06, + "loss": 0.903, + "step": 3593 + }, + { + "epoch": 1.6708507670850767, + "grad_norm": 10.902867317199707, + "learning_rate": 9.9964079996408e-06, + "loss": 0.4582, + "step": 3594 + }, + { + "epoch": 1.6713156671315668, + "grad_norm": 5.8044610023498535, + "learning_rate": 9.996406999640701e-06, + "loss": 0.5704, + "step": 3595 + }, + { + "epoch": 1.6717805671780566, + "grad_norm": 5.41924524307251, + "learning_rate": 9.9964059996406e-06, + "loss": 0.4725, + "step": 3596 + }, + { + "epoch": 1.6722454672245468, + "grad_norm": 7.154657363891602, + "learning_rate": 9.9964049996405e-06, + "loss": 0.6025, + "step": 3597 + }, + { + "epoch": 1.6727103672710366, + "grad_norm": 5.821597576141357, + "learning_rate": 9.996403999640401e-06, + "loss": 0.5906, + "step": 3598 + }, + { + "epoch": 1.6731752673175269, + "grad_norm": 6.515263080596924, + "learning_rate": 9.996402999640302e-06, + "loss": 0.5258, + "step": 3599 + }, + { + "epoch": 1.6736401673640167, + "grad_norm": 8.081713676452637, + "learning_rate": 9.996401999640201e-06, + "loss": 0.8221, + "step": 3600 + }, + { + "epoch": 1.6741050674105067, + "grad_norm": 5.997825622558594, + "learning_rate": 9.996400999640102e-06, + "loss": 0.7748, + "step": 3601 + }, + { + "epoch": 1.6745699674569967, + "grad_norm": 6.624133586883545, + "learning_rate": 9.99639999964e-06, + "loss": 0.689, + "step": 3602 + }, + { + "epoch": 1.6750348675034867, + "grad_norm": 5.012633800506592, + "learning_rate": 9.9963989996399e-06, + "loss": 0.3766, + "step": 3603 + }, + { + "epoch": 1.6754997675499768, + "grad_norm": 5.87786340713501, + "learning_rate": 9.9963979996398e-06, + "loss": 0.5883, + "step": 3604 + }, + { + "epoch": 1.6759646675964668, + "grad_norm": 4.627636432647705, + "learning_rate": 9.9963969996397e-06, + "loss": 0.3745, + "step": 3605 + }, + { + "epoch": 1.6764295676429568, + "grad_norm": 10.600364685058594, + "learning_rate": 9.996395999639601e-06, + "loss": 0.6698, + "step": 3606 + }, + { + "epoch": 1.6768944676894466, + "grad_norm": 5.956223011016846, + "learning_rate": 9.9963949996395e-06, + "loss": 0.8867, + "step": 3607 + }, + { + "epoch": 1.6773593677359369, + "grad_norm": 7.148153781890869, + "learning_rate": 9.996393999639401e-06, + "loss": 0.7835, + "step": 3608 + }, + { + "epoch": 1.6778242677824267, + "grad_norm": 7.447465419769287, + "learning_rate": 9.9963929996393e-06, + "loss": 0.7758, + "step": 3609 + }, + { + "epoch": 1.678289167828917, + "grad_norm": 7.939570903778076, + "learning_rate": 9.9963919996392e-06, + "loss": 0.9014, + "step": 3610 + }, + { + "epoch": 1.6787540678754067, + "grad_norm": 4.4137797355651855, + "learning_rate": 9.996390999639101e-06, + "loss": 0.3882, + "step": 3611 + }, + { + "epoch": 1.679218967921897, + "grad_norm": 6.682091236114502, + "learning_rate": 9.996389999639002e-06, + "loss": 0.5349, + "step": 3612 + }, + { + "epoch": 1.6796838679683868, + "grad_norm": 13.66723918914795, + "learning_rate": 9.996388999638901e-06, + "loss": 0.5908, + "step": 3613 + }, + { + "epoch": 1.6801487680148768, + "grad_norm": 5.565954208374023, + "learning_rate": 9.9963879996388e-06, + "loss": 0.7042, + "step": 3614 + }, + { + "epoch": 1.6806136680613668, + "grad_norm": 4.932942867279053, + "learning_rate": 9.9963869996387e-06, + "loss": 0.3522, + "step": 3615 + }, + { + "epoch": 1.6810785681078568, + "grad_norm": 4.669673442840576, + "learning_rate": 9.9963859996386e-06, + "loss": 0.4648, + "step": 3616 + }, + { + "epoch": 1.6815434681543469, + "grad_norm": 4.912144660949707, + "learning_rate": 9.9963849996385e-06, + "loss": 0.4867, + "step": 3617 + }, + { + "epoch": 1.6820083682008367, + "grad_norm": 7.003923416137695, + "learning_rate": 9.996383999638401e-06, + "loss": 0.8123, + "step": 3618 + }, + { + "epoch": 1.682473268247327, + "grad_norm": 6.069211483001709, + "learning_rate": 9.996382999638302e-06, + "loss": 0.8433, + "step": 3619 + }, + { + "epoch": 1.6829381682938167, + "grad_norm": 6.4041242599487305, + "learning_rate": 9.9963819996382e-06, + "loss": 0.432, + "step": 3620 + }, + { + "epoch": 1.683403068340307, + "grad_norm": 3.097606658935547, + "learning_rate": 9.996380999638101e-06, + "loss": 0.1175, + "step": 3621 + }, + { + "epoch": 1.6838679683867968, + "grad_norm": 5.881203651428223, + "learning_rate": 9.996379999638e-06, + "loss": 0.6552, + "step": 3622 + }, + { + "epoch": 1.684332868433287, + "grad_norm": 6.553452014923096, + "learning_rate": 9.9963789996379e-06, + "loss": 0.9456, + "step": 3623 + }, + { + "epoch": 1.6847977684797768, + "grad_norm": 5.263757705688477, + "learning_rate": 9.996377999637801e-06, + "loss": 0.4055, + "step": 3624 + }, + { + "epoch": 1.6852626685262668, + "grad_norm": 4.307650566101074, + "learning_rate": 9.9963769996377e-06, + "loss": 0.2656, + "step": 3625 + }, + { + "epoch": 1.6857275685727569, + "grad_norm": 6.438689708709717, + "learning_rate": 9.996375999637601e-06, + "loss": 0.5639, + "step": 3626 + }, + { + "epoch": 1.6861924686192469, + "grad_norm": 8.885516166687012, + "learning_rate": 9.9963749996375e-06, + "loss": 0.4707, + "step": 3627 + }, + { + "epoch": 1.686657368665737, + "grad_norm": 3.869417190551758, + "learning_rate": 9.9963739996374e-06, + "loss": 0.3897, + "step": 3628 + }, + { + "epoch": 1.687122268712227, + "grad_norm": 5.12237548828125, + "learning_rate": 9.9963729996373e-06, + "loss": 0.5575, + "step": 3629 + }, + { + "epoch": 1.687587168758717, + "grad_norm": 4.430319786071777, + "learning_rate": 9.9963719996372e-06, + "loss": 0.387, + "step": 3630 + }, + { + "epoch": 1.6880520688052068, + "grad_norm": 6.688879013061523, + "learning_rate": 9.996370999637101e-06, + "loss": 0.6096, + "step": 3631 + }, + { + "epoch": 1.688516968851697, + "grad_norm": 5.549559116363525, + "learning_rate": 9.996369999637002e-06, + "loss": 0.3668, + "step": 3632 + }, + { + "epoch": 1.6889818688981868, + "grad_norm": 5.198909282684326, + "learning_rate": 9.9963689996369e-06, + "loss": 0.4998, + "step": 3633 + }, + { + "epoch": 1.689446768944677, + "grad_norm": 5.625499725341797, + "learning_rate": 9.996367999636801e-06, + "loss": 0.3872, + "step": 3634 + }, + { + "epoch": 1.6899116689911668, + "grad_norm": 6.20835018157959, + "learning_rate": 9.9963669996367e-06, + "loss": 0.6709, + "step": 3635 + }, + { + "epoch": 1.6903765690376569, + "grad_norm": 4.718818187713623, + "learning_rate": 9.996365999636601e-06, + "loss": 0.3475, + "step": 3636 + }, + { + "epoch": 1.690841469084147, + "grad_norm": 6.238049507141113, + "learning_rate": 9.996364999636502e-06, + "loss": 0.7707, + "step": 3637 + }, + { + "epoch": 1.691306369130637, + "grad_norm": 5.248319625854492, + "learning_rate": 9.9963639996364e-06, + "loss": 0.4268, + "step": 3638 + }, + { + "epoch": 1.691771269177127, + "grad_norm": 10.82319164276123, + "learning_rate": 9.996362999636301e-06, + "loss": 0.4851, + "step": 3639 + }, + { + "epoch": 1.692236169223617, + "grad_norm": 5.985126495361328, + "learning_rate": 9.9963619996362e-06, + "loss": 0.4591, + "step": 3640 + }, + { + "epoch": 1.692701069270107, + "grad_norm": 9.577142715454102, + "learning_rate": 9.9963609996361e-06, + "loss": 0.7171, + "step": 3641 + }, + { + "epoch": 1.6931659693165968, + "grad_norm": 6.597506999969482, + "learning_rate": 9.996359999636e-06, + "loss": 0.6083, + "step": 3642 + }, + { + "epoch": 1.693630869363087, + "grad_norm": 7.128410339355469, + "learning_rate": 9.9963589996359e-06, + "loss": 0.8264, + "step": 3643 + }, + { + "epoch": 1.6940957694095768, + "grad_norm": 6.3196611404418945, + "learning_rate": 9.996357999635801e-06, + "loss": 0.3474, + "step": 3644 + }, + { + "epoch": 1.694560669456067, + "grad_norm": 9.286418914794922, + "learning_rate": 9.996356999635702e-06, + "loss": 1.0141, + "step": 3645 + }, + { + "epoch": 1.695025569502557, + "grad_norm": 5.543634414672852, + "learning_rate": 9.9963559996356e-06, + "loss": 0.6253, + "step": 3646 + }, + { + "epoch": 1.695490469549047, + "grad_norm": 7.591969013214111, + "learning_rate": 9.9963549996355e-06, + "loss": 0.5121, + "step": 3647 + }, + { + "epoch": 1.695955369595537, + "grad_norm": 5.993334770202637, + "learning_rate": 9.9963539996354e-06, + "loss": 0.4508, + "step": 3648 + }, + { + "epoch": 1.696420269642027, + "grad_norm": 7.368037223815918, + "learning_rate": 9.996352999635301e-06, + "loss": 0.4394, + "step": 3649 + }, + { + "epoch": 1.696885169688517, + "grad_norm": 11.310816764831543, + "learning_rate": 9.996351999635202e-06, + "loss": 0.9988, + "step": 3650 + }, + { + "epoch": 1.697350069735007, + "grad_norm": 9.765625953674316, + "learning_rate": 9.9963509996351e-06, + "loss": 0.6481, + "step": 3651 + }, + { + "epoch": 1.697814969781497, + "grad_norm": 4.441245079040527, + "learning_rate": 9.996349999635001e-06, + "loss": 0.3139, + "step": 3652 + }, + { + "epoch": 1.6982798698279868, + "grad_norm": 3.529480218887329, + "learning_rate": 9.9963489996349e-06, + "loss": 0.3138, + "step": 3653 + }, + { + "epoch": 1.698744769874477, + "grad_norm": 7.146917343139648, + "learning_rate": 9.996347999634801e-06, + "loss": 0.6718, + "step": 3654 + }, + { + "epoch": 1.699209669920967, + "grad_norm": 6.132967472076416, + "learning_rate": 9.996346999634702e-06, + "loss": 0.6843, + "step": 3655 + }, + { + "epoch": 1.6996745699674571, + "grad_norm": 4.586550712585449, + "learning_rate": 9.9963459996346e-06, + "loss": 0.4212, + "step": 3656 + }, + { + "epoch": 1.700139470013947, + "grad_norm": 5.573400974273682, + "learning_rate": 9.996344999634501e-06, + "loss": 0.462, + "step": 3657 + }, + { + "epoch": 1.700604370060437, + "grad_norm": 4.420722961425781, + "learning_rate": 9.9963439996344e-06, + "loss": 0.2796, + "step": 3658 + }, + { + "epoch": 1.701069270106927, + "grad_norm": 6.202597141265869, + "learning_rate": 9.9963429996343e-06, + "loss": 0.96, + "step": 3659 + }, + { + "epoch": 1.701534170153417, + "grad_norm": 4.551113128662109, + "learning_rate": 9.9963419996342e-06, + "loss": 0.4451, + "step": 3660 + }, + { + "epoch": 1.701999070199907, + "grad_norm": 5.561594009399414, + "learning_rate": 9.9963409996341e-06, + "loss": 0.546, + "step": 3661 + }, + { + "epoch": 1.702463970246397, + "grad_norm": 4.983760356903076, + "learning_rate": 9.996339999634001e-06, + "loss": 0.5082, + "step": 3662 + }, + { + "epoch": 1.702928870292887, + "grad_norm": 7.670078754425049, + "learning_rate": 9.996338999633902e-06, + "loss": 0.5977, + "step": 3663 + }, + { + "epoch": 1.703393770339377, + "grad_norm": 5.383225440979004, + "learning_rate": 9.9963379996338e-06, + "loss": 0.4032, + "step": 3664 + }, + { + "epoch": 1.7038586703858671, + "grad_norm": 3.22456955909729, + "learning_rate": 9.996336999633701e-06, + "loss": 0.3195, + "step": 3665 + }, + { + "epoch": 1.704323570432357, + "grad_norm": 4.579484462738037, + "learning_rate": 9.9963359996336e-06, + "loss": 0.4243, + "step": 3666 + }, + { + "epoch": 1.7047884704788472, + "grad_norm": 5.0451741218566895, + "learning_rate": 9.996334999633501e-06, + "loss": 0.4333, + "step": 3667 + }, + { + "epoch": 1.705253370525337, + "grad_norm": 9.194003105163574, + "learning_rate": 9.996333999633402e-06, + "loss": 0.7939, + "step": 3668 + }, + { + "epoch": 1.705718270571827, + "grad_norm": 3.7402186393737793, + "learning_rate": 9.9963329996333e-06, + "loss": 0.4184, + "step": 3669 + }, + { + "epoch": 1.706183170618317, + "grad_norm": 6.048062324523926, + "learning_rate": 9.996331999633201e-06, + "loss": 0.5797, + "step": 3670 + }, + { + "epoch": 1.706648070664807, + "grad_norm": 10.783827781677246, + "learning_rate": 9.9963309996331e-06, + "loss": 0.755, + "step": 3671 + }, + { + "epoch": 1.707112970711297, + "grad_norm": 4.443427085876465, + "learning_rate": 9.996329999633e-06, + "loss": 0.354, + "step": 3672 + }, + { + "epoch": 1.707577870757787, + "grad_norm": 5.730403423309326, + "learning_rate": 9.9963289996329e-06, + "loss": 0.5754, + "step": 3673 + }, + { + "epoch": 1.7080427708042771, + "grad_norm": 9.441051483154297, + "learning_rate": 9.9963279996328e-06, + "loss": 0.6923, + "step": 3674 + }, + { + "epoch": 1.708507670850767, + "grad_norm": 6.227530002593994, + "learning_rate": 9.996326999632701e-06, + "loss": 0.3684, + "step": 3675 + }, + { + "epoch": 1.7089725708972572, + "grad_norm": 6.492858409881592, + "learning_rate": 9.996325999632602e-06, + "loss": 0.556, + "step": 3676 + }, + { + "epoch": 1.709437470943747, + "grad_norm": 3.991198778152466, + "learning_rate": 9.9963249996325e-06, + "loss": 0.342, + "step": 3677 + }, + { + "epoch": 1.7099023709902372, + "grad_norm": 8.996414184570312, + "learning_rate": 9.996323999632401e-06, + "loss": 0.7164, + "step": 3678 + }, + { + "epoch": 1.710367271036727, + "grad_norm": 3.3251116275787354, + "learning_rate": 9.9963229996323e-06, + "loss": 0.2385, + "step": 3679 + }, + { + "epoch": 1.710832171083217, + "grad_norm": 6.146736145019531, + "learning_rate": 9.996321999632201e-06, + "loss": 0.5648, + "step": 3680 + }, + { + "epoch": 1.711297071129707, + "grad_norm": 4.924281597137451, + "learning_rate": 9.996320999632102e-06, + "loss": 0.3753, + "step": 3681 + }, + { + "epoch": 1.711761971176197, + "grad_norm": 4.99364709854126, + "learning_rate": 9.996319999632e-06, + "loss": 0.4467, + "step": 3682 + }, + { + "epoch": 1.7122268712226871, + "grad_norm": 5.212062358856201, + "learning_rate": 9.996318999631901e-06, + "loss": 0.4511, + "step": 3683 + }, + { + "epoch": 1.7126917712691772, + "grad_norm": 5.6988701820373535, + "learning_rate": 9.9963179996318e-06, + "loss": 0.5275, + "step": 3684 + }, + { + "epoch": 1.7131566713156672, + "grad_norm": 5.211921691894531, + "learning_rate": 9.996316999631701e-06, + "loss": 0.48, + "step": 3685 + }, + { + "epoch": 1.713621571362157, + "grad_norm": 6.309988021850586, + "learning_rate": 9.9963159996316e-06, + "loss": 0.4839, + "step": 3686 + }, + { + "epoch": 1.7140864714086472, + "grad_norm": 3.893841505050659, + "learning_rate": 9.9963149996315e-06, + "loss": 0.2581, + "step": 3687 + }, + { + "epoch": 1.714551371455137, + "grad_norm": 5.6922287940979, + "learning_rate": 9.996313999631401e-06, + "loss": 0.6185, + "step": 3688 + }, + { + "epoch": 1.7150162715016273, + "grad_norm": 5.561637878417969, + "learning_rate": 9.996312999631302e-06, + "loss": 0.4753, + "step": 3689 + }, + { + "epoch": 1.715481171548117, + "grad_norm": 5.058505535125732, + "learning_rate": 9.9963119996312e-06, + "loss": 0.4591, + "step": 3690 + }, + { + "epoch": 1.7159460715946073, + "grad_norm": 4.606479167938232, + "learning_rate": 9.9963109996311e-06, + "loss": 0.3072, + "step": 3691 + }, + { + "epoch": 1.7164109716410971, + "grad_norm": 4.5242133140563965, + "learning_rate": 9.996309999631e-06, + "loss": 0.4126, + "step": 3692 + }, + { + "epoch": 1.7168758716875872, + "grad_norm": 7.237139701843262, + "learning_rate": 9.996308999630901e-06, + "loss": 0.3625, + "step": 3693 + }, + { + "epoch": 1.7173407717340772, + "grad_norm": 7.231119632720947, + "learning_rate": 9.996307999630802e-06, + "loss": 0.801, + "step": 3694 + }, + { + "epoch": 1.7178056717805672, + "grad_norm": 3.0699522495269775, + "learning_rate": 9.9963069996307e-06, + "loss": 0.2204, + "step": 3695 + }, + { + "epoch": 1.7182705718270572, + "grad_norm": 6.401881694793701, + "learning_rate": 9.996305999630601e-06, + "loss": 0.6409, + "step": 3696 + }, + { + "epoch": 1.718735471873547, + "grad_norm": 5.721397399902344, + "learning_rate": 9.9963049996305e-06, + "loss": 0.5542, + "step": 3697 + }, + { + "epoch": 1.7192003719200373, + "grad_norm": 5.509790420532227, + "learning_rate": 9.996303999630401e-06, + "loss": 0.788, + "step": 3698 + }, + { + "epoch": 1.719665271966527, + "grad_norm": 4.984347820281982, + "learning_rate": 9.9963029996303e-06, + "loss": 0.4426, + "step": 3699 + }, + { + "epoch": 1.7201301720130173, + "grad_norm": 7.076423168182373, + "learning_rate": 9.9963019996302e-06, + "loss": 0.6234, + "step": 3700 + }, + { + "epoch": 1.7205950720595071, + "grad_norm": 7.312623977661133, + "learning_rate": 9.996300999630101e-06, + "loss": 0.6908, + "step": 3701 + }, + { + "epoch": 1.7210599721059974, + "grad_norm": 4.867172718048096, + "learning_rate": 9.99629999963e-06, + "loss": 0.5502, + "step": 3702 + }, + { + "epoch": 1.7215248721524872, + "grad_norm": 6.052154064178467, + "learning_rate": 9.996298999629901e-06, + "loss": 0.4793, + "step": 3703 + }, + { + "epoch": 1.7219897721989772, + "grad_norm": 6.715153694152832, + "learning_rate": 9.9962979996298e-06, + "loss": 0.6342, + "step": 3704 + }, + { + "epoch": 1.7224546722454672, + "grad_norm": 5.830573081970215, + "learning_rate": 9.9962969996297e-06, + "loss": 0.543, + "step": 3705 + }, + { + "epoch": 1.7229195722919572, + "grad_norm": 5.960351467132568, + "learning_rate": 9.996295999629601e-06, + "loss": 1.1082, + "step": 3706 + }, + { + "epoch": 1.7233844723384473, + "grad_norm": 11.494726181030273, + "learning_rate": 9.996294999629502e-06, + "loss": 0.5706, + "step": 3707 + }, + { + "epoch": 1.723849372384937, + "grad_norm": 4.035188674926758, + "learning_rate": 9.9962939996294e-06, + "loss": 0.4263, + "step": 3708 + }, + { + "epoch": 1.7243142724314273, + "grad_norm": 4.800931453704834, + "learning_rate": 9.996292999629301e-06, + "loss": 0.5101, + "step": 3709 + }, + { + "epoch": 1.7247791724779171, + "grad_norm": 7.388068675994873, + "learning_rate": 9.9962919996292e-06, + "loss": 0.4556, + "step": 3710 + }, + { + "epoch": 1.7252440725244074, + "grad_norm": 7.035270690917969, + "learning_rate": 9.996290999629101e-06, + "loss": 0.8719, + "step": 3711 + }, + { + "epoch": 1.7257089725708972, + "grad_norm": 3.4778692722320557, + "learning_rate": 9.996289999629e-06, + "loss": 0.2185, + "step": 3712 + }, + { + "epoch": 1.7261738726173874, + "grad_norm": 4.864955902099609, + "learning_rate": 9.9962889996289e-06, + "loss": 0.552, + "step": 3713 + }, + { + "epoch": 1.7266387726638772, + "grad_norm": 5.081823348999023, + "learning_rate": 9.996287999628801e-06, + "loss": 0.4271, + "step": 3714 + }, + { + "epoch": 1.7271036727103672, + "grad_norm": 8.586714744567871, + "learning_rate": 9.9962869996287e-06, + "loss": 0.8122, + "step": 3715 + }, + { + "epoch": 1.7275685727568573, + "grad_norm": 6.178071022033691, + "learning_rate": 9.996285999628601e-06, + "loss": 0.5319, + "step": 3716 + }, + { + "epoch": 1.7280334728033473, + "grad_norm": 5.1879682540893555, + "learning_rate": 9.9962849996285e-06, + "loss": 0.4496, + "step": 3717 + }, + { + "epoch": 1.7284983728498373, + "grad_norm": 8.30210018157959, + "learning_rate": 9.9962839996284e-06, + "loss": 0.8064, + "step": 3718 + }, + { + "epoch": 1.7289632728963273, + "grad_norm": 4.9429707527160645, + "learning_rate": 9.996282999628301e-06, + "loss": 0.6435, + "step": 3719 + }, + { + "epoch": 1.7294281729428174, + "grad_norm": 5.612203121185303, + "learning_rate": 9.996281999628202e-06, + "loss": 0.5376, + "step": 3720 + }, + { + "epoch": 1.7298930729893072, + "grad_norm": 5.065670490264893, + "learning_rate": 9.996280999628101e-06, + "loss": 0.5592, + "step": 3721 + }, + { + "epoch": 1.7303579730357974, + "grad_norm": 8.218449592590332, + "learning_rate": 9.996279999628002e-06, + "loss": 0.4711, + "step": 3722 + }, + { + "epoch": 1.7308228730822872, + "grad_norm": 4.745869159698486, + "learning_rate": 9.9962789996279e-06, + "loss": 0.2713, + "step": 3723 + }, + { + "epoch": 1.7312877731287775, + "grad_norm": 4.181095123291016, + "learning_rate": 9.996277999627801e-06, + "loss": 0.3642, + "step": 3724 + }, + { + "epoch": 1.7317526731752673, + "grad_norm": 5.918643951416016, + "learning_rate": 9.9962769996277e-06, + "loss": 0.5024, + "step": 3725 + }, + { + "epoch": 1.7322175732217573, + "grad_norm": 7.877264976501465, + "learning_rate": 9.9962759996276e-06, + "loss": 0.7836, + "step": 3726 + }, + { + "epoch": 1.7326824732682473, + "grad_norm": 5.000230312347412, + "learning_rate": 9.996274999627501e-06, + "loss": 0.5311, + "step": 3727 + }, + { + "epoch": 1.7331473733147373, + "grad_norm": 4.668985843658447, + "learning_rate": 9.9962739996274e-06, + "loss": 0.6072, + "step": 3728 + }, + { + "epoch": 1.7336122733612274, + "grad_norm": 4.909816741943359, + "learning_rate": 9.996272999627301e-06, + "loss": 0.4262, + "step": 3729 + }, + { + "epoch": 1.7340771734077174, + "grad_norm": 5.000997066497803, + "learning_rate": 9.9962719996272e-06, + "loss": 0.4001, + "step": 3730 + }, + { + "epoch": 1.7345420734542074, + "grad_norm": 5.705283164978027, + "learning_rate": 9.9962709996271e-06, + "loss": 0.5677, + "step": 3731 + }, + { + "epoch": 1.7350069735006972, + "grad_norm": 6.999468803405762, + "learning_rate": 9.996269999627001e-06, + "loss": 0.4986, + "step": 3732 + }, + { + "epoch": 1.7354718735471875, + "grad_norm": 9.045291900634766, + "learning_rate": 9.9962689996269e-06, + "loss": 1.0997, + "step": 3733 + }, + { + "epoch": 1.7359367735936773, + "grad_norm": 5.892764091491699, + "learning_rate": 9.996267999626801e-06, + "loss": 0.6799, + "step": 3734 + }, + { + "epoch": 1.7364016736401675, + "grad_norm": 5.648662567138672, + "learning_rate": 9.9962669996267e-06, + "loss": 0.4547, + "step": 3735 + }, + { + "epoch": 1.7368665736866573, + "grad_norm": 5.574146747589111, + "learning_rate": 9.9962659996266e-06, + "loss": 0.5468, + "step": 3736 + }, + { + "epoch": 1.7373314737331473, + "grad_norm": 11.932151794433594, + "learning_rate": 9.996264999626501e-06, + "loss": 0.4211, + "step": 3737 + }, + { + "epoch": 1.7377963737796374, + "grad_norm": 4.796853542327881, + "learning_rate": 9.9962639996264e-06, + "loss": 0.4322, + "step": 3738 + }, + { + "epoch": 1.7382612738261274, + "grad_norm": 4.558944225311279, + "learning_rate": 9.996262999626301e-06, + "loss": 0.4462, + "step": 3739 + }, + { + "epoch": 1.7387261738726174, + "grad_norm": 5.110750675201416, + "learning_rate": 9.996261999626202e-06, + "loss": 0.5511, + "step": 3740 + }, + { + "epoch": 1.7391910739191074, + "grad_norm": 5.489506721496582, + "learning_rate": 9.9962609996261e-06, + "loss": 0.3555, + "step": 3741 + }, + { + "epoch": 1.7396559739655975, + "grad_norm": 7.7366132736206055, + "learning_rate": 9.996259999626001e-06, + "loss": 0.821, + "step": 3742 + }, + { + "epoch": 1.7401208740120873, + "grad_norm": 4.2294697761535645, + "learning_rate": 9.9962589996259e-06, + "loss": 0.3291, + "step": 3743 + }, + { + "epoch": 1.7405857740585775, + "grad_norm": 4.7394819259643555, + "learning_rate": 9.9962579996258e-06, + "loss": 0.4933, + "step": 3744 + }, + { + "epoch": 1.7410506741050673, + "grad_norm": 6.109631538391113, + "learning_rate": 9.996256999625701e-06, + "loss": 0.7701, + "step": 3745 + }, + { + "epoch": 1.7415155741515576, + "grad_norm": 4.433706760406494, + "learning_rate": 9.9962559996256e-06, + "loss": 0.334, + "step": 3746 + }, + { + "epoch": 1.7419804741980474, + "grad_norm": 5.076189041137695, + "learning_rate": 9.996254999625501e-06, + "loss": 0.3725, + "step": 3747 + }, + { + "epoch": 1.7424453742445374, + "grad_norm": 5.862693786621094, + "learning_rate": 9.9962539996254e-06, + "loss": 0.6942, + "step": 3748 + }, + { + "epoch": 1.7429102742910274, + "grad_norm": 6.768830299377441, + "learning_rate": 9.9962529996253e-06, + "loss": 0.6904, + "step": 3749 + }, + { + "epoch": 1.7433751743375174, + "grad_norm": 4.013789176940918, + "learning_rate": 9.996251999625201e-06, + "loss": 0.3653, + "step": 3750 + }, + { + "epoch": 1.7438400743840075, + "grad_norm": 6.352783203125, + "learning_rate": 9.996250999625102e-06, + "loss": 0.3639, + "step": 3751 + }, + { + "epoch": 1.7443049744304975, + "grad_norm": 5.730754852294922, + "learning_rate": 9.996249999625001e-06, + "loss": 0.5112, + "step": 3752 + }, + { + "epoch": 1.7447698744769875, + "grad_norm": 4.296408176422119, + "learning_rate": 9.996248999624902e-06, + "loss": 0.2986, + "step": 3753 + }, + { + "epoch": 1.7452347745234773, + "grad_norm": 5.599689483642578, + "learning_rate": 9.9962479996248e-06, + "loss": 0.5637, + "step": 3754 + }, + { + "epoch": 1.7456996745699676, + "grad_norm": 8.777532577514648, + "learning_rate": 9.9962469996247e-06, + "loss": 0.7193, + "step": 3755 + }, + { + "epoch": 1.7461645746164574, + "grad_norm": 4.099140167236328, + "learning_rate": 9.9962459996246e-06, + "loss": 0.1773, + "step": 3756 + }, + { + "epoch": 1.7466294746629476, + "grad_norm": 8.938737869262695, + "learning_rate": 9.9962449996245e-06, + "loss": 0.7953, + "step": 3757 + }, + { + "epoch": 1.7470943747094374, + "grad_norm": 8.726188659667969, + "learning_rate": 9.996243999624402e-06, + "loss": 1.0278, + "step": 3758 + }, + { + "epoch": 1.7475592747559274, + "grad_norm": 4.4052734375, + "learning_rate": 9.9962429996243e-06, + "loss": 0.3507, + "step": 3759 + }, + { + "epoch": 1.7480241748024175, + "grad_norm": 8.698685646057129, + "learning_rate": 9.996241999624201e-06, + "loss": 0.846, + "step": 3760 + }, + { + "epoch": 1.7484890748489075, + "grad_norm": 6.941092014312744, + "learning_rate": 9.9962409996241e-06, + "loss": 0.617, + "step": 3761 + }, + { + "epoch": 1.7489539748953975, + "grad_norm": 10.402631759643555, + "learning_rate": 9.996239999624e-06, + "loss": 0.2895, + "step": 3762 + }, + { + "epoch": 1.7494188749418875, + "grad_norm": 6.535046577453613, + "learning_rate": 9.996238999623901e-06, + "loss": 0.5594, + "step": 3763 + }, + { + "epoch": 1.7498837749883775, + "grad_norm": 5.34915018081665, + "learning_rate": 9.996237999623802e-06, + "loss": 0.5105, + "step": 3764 + }, + { + "epoch": 1.7503486750348674, + "grad_norm": 7.394558429718018, + "learning_rate": 9.996236999623701e-06, + "loss": 0.707, + "step": 3765 + }, + { + "epoch": 1.7508135750813576, + "grad_norm": 7.475854396820068, + "learning_rate": 9.9962359996236e-06, + "loss": 0.8663, + "step": 3766 + }, + { + "epoch": 1.7512784751278474, + "grad_norm": 6.16146993637085, + "learning_rate": 9.9962349996235e-06, + "loss": 0.5124, + "step": 3767 + }, + { + "epoch": 1.7517433751743376, + "grad_norm": 5.840900421142578, + "learning_rate": 9.9962339996234e-06, + "loss": 0.5634, + "step": 3768 + }, + { + "epoch": 1.7522082752208274, + "grad_norm": 5.828770160675049, + "learning_rate": 9.9962329996233e-06, + "loss": 0.4851, + "step": 3769 + }, + { + "epoch": 1.7526731752673175, + "grad_norm": 4.677632808685303, + "learning_rate": 9.996231999623201e-06, + "loss": 0.309, + "step": 3770 + }, + { + "epoch": 1.7531380753138075, + "grad_norm": 7.170637607574463, + "learning_rate": 9.996230999623102e-06, + "loss": 0.6578, + "step": 3771 + }, + { + "epoch": 1.7536029753602975, + "grad_norm": 6.091549873352051, + "learning_rate": 9.996229999623e-06, + "loss": 0.4425, + "step": 3772 + }, + { + "epoch": 1.7540678754067875, + "grad_norm": 5.487090587615967, + "learning_rate": 9.996228999622901e-06, + "loss": 0.5368, + "step": 3773 + }, + { + "epoch": 1.7545327754532776, + "grad_norm": 5.80232048034668, + "learning_rate": 9.9962279996228e-06, + "loss": 0.561, + "step": 3774 + }, + { + "epoch": 1.7549976754997676, + "grad_norm": 6.225417137145996, + "learning_rate": 9.9962269996227e-06, + "loss": 0.6457, + "step": 3775 + }, + { + "epoch": 1.7554625755462574, + "grad_norm": 5.240749835968018, + "learning_rate": 9.996225999622601e-06, + "loss": 0.3719, + "step": 3776 + }, + { + "epoch": 1.7559274755927476, + "grad_norm": 7.530819892883301, + "learning_rate": 9.9962249996225e-06, + "loss": 0.6008, + "step": 3777 + }, + { + "epoch": 1.7563923756392374, + "grad_norm": 6.635265827178955, + "learning_rate": 9.996223999622401e-06, + "loss": 0.3729, + "step": 3778 + }, + { + "epoch": 1.7568572756857277, + "grad_norm": 5.736187934875488, + "learning_rate": 9.9962229996223e-06, + "loss": 0.6357, + "step": 3779 + }, + { + "epoch": 1.7573221757322175, + "grad_norm": 6.061929225921631, + "learning_rate": 9.9962219996222e-06, + "loss": 0.7389, + "step": 3780 + }, + { + "epoch": 1.7577870757787077, + "grad_norm": 5.011348247528076, + "learning_rate": 9.9962209996221e-06, + "loss": 0.5439, + "step": 3781 + }, + { + "epoch": 1.7582519758251975, + "grad_norm": 5.976624011993408, + "learning_rate": 9.996219999622e-06, + "loss": 0.5494, + "step": 3782 + }, + { + "epoch": 1.7587168758716876, + "grad_norm": 5.023672580718994, + "learning_rate": 9.996218999621901e-06, + "loss": 0.3584, + "step": 3783 + }, + { + "epoch": 1.7591817759181776, + "grad_norm": 6.137037754058838, + "learning_rate": 9.996217999621802e-06, + "loss": 0.7133, + "step": 3784 + }, + { + "epoch": 1.7596466759646676, + "grad_norm": 6.122052192687988, + "learning_rate": 9.9962169996217e-06, + "loss": 0.6768, + "step": 3785 + }, + { + "epoch": 1.7601115760111576, + "grad_norm": 5.504772186279297, + "learning_rate": 9.996215999621601e-06, + "loss": 0.5342, + "step": 3786 + }, + { + "epoch": 1.7605764760576474, + "grad_norm": 5.930167198181152, + "learning_rate": 9.9962149996215e-06, + "loss": 0.5776, + "step": 3787 + }, + { + "epoch": 1.7610413761041377, + "grad_norm": 4.702578544616699, + "learning_rate": 9.996213999621401e-06, + "loss": 0.3873, + "step": 3788 + }, + { + "epoch": 1.7615062761506275, + "grad_norm": 4.24972677230835, + "learning_rate": 9.996212999621302e-06, + "loss": 0.4052, + "step": 3789 + }, + { + "epoch": 1.7619711761971177, + "grad_norm": 6.615088939666748, + "learning_rate": 9.9962119996212e-06, + "loss": 0.6641, + "step": 3790 + }, + { + "epoch": 1.7624360762436075, + "grad_norm": 5.3468852043151855, + "learning_rate": 9.996210999621101e-06, + "loss": 0.3892, + "step": 3791 + }, + { + "epoch": 1.7629009762900978, + "grad_norm": 6.6802978515625, + "learning_rate": 9.996209999621e-06, + "loss": 0.5614, + "step": 3792 + }, + { + "epoch": 1.7633658763365876, + "grad_norm": 4.700436592102051, + "learning_rate": 9.9962089996209e-06, + "loss": 0.2061, + "step": 3793 + }, + { + "epoch": 1.7638307763830776, + "grad_norm": 4.9558329582214355, + "learning_rate": 9.9962079996208e-06, + "loss": 0.3904, + "step": 3794 + }, + { + "epoch": 1.7642956764295676, + "grad_norm": 5.434839725494385, + "learning_rate": 9.9962069996207e-06, + "loss": 0.4155, + "step": 3795 + }, + { + "epoch": 1.7647605764760577, + "grad_norm": 5.2188568115234375, + "learning_rate": 9.996205999620601e-06, + "loss": 0.613, + "step": 3796 + }, + { + "epoch": 1.7652254765225477, + "grad_norm": 5.660783290863037, + "learning_rate": 9.996204999620502e-06, + "loss": 0.4763, + "step": 3797 + }, + { + "epoch": 1.7656903765690377, + "grad_norm": 5.195202350616455, + "learning_rate": 9.9962039996204e-06, + "loss": 0.4702, + "step": 3798 + }, + { + "epoch": 1.7661552766155277, + "grad_norm": 10.579172134399414, + "learning_rate": 9.9962029996203e-06, + "loss": 0.7561, + "step": 3799 + }, + { + "epoch": 1.7666201766620175, + "grad_norm": 5.965630531311035, + "learning_rate": 9.9962019996202e-06, + "loss": 0.6419, + "step": 3800 + }, + { + "epoch": 1.7670850767085078, + "grad_norm": 6.162230014801025, + "learning_rate": 9.996200999620101e-06, + "loss": 0.6063, + "step": 3801 + }, + { + "epoch": 1.7675499767549976, + "grad_norm": 6.776439189910889, + "learning_rate": 9.996199999620002e-06, + "loss": 0.8216, + "step": 3802 + }, + { + "epoch": 1.7680148768014878, + "grad_norm": 5.364141941070557, + "learning_rate": 9.9961989996199e-06, + "loss": 0.5674, + "step": 3803 + }, + { + "epoch": 1.7684797768479776, + "grad_norm": 6.398838043212891, + "learning_rate": 9.996197999619801e-06, + "loss": 0.6961, + "step": 3804 + }, + { + "epoch": 1.7689446768944677, + "grad_norm": 4.205951690673828, + "learning_rate": 9.9961969996197e-06, + "loss": 0.2722, + "step": 3805 + }, + { + "epoch": 1.7694095769409577, + "grad_norm": 5.836526393890381, + "learning_rate": 9.996195999619601e-06, + "loss": 0.6371, + "step": 3806 + }, + { + "epoch": 1.7698744769874477, + "grad_norm": 3.6670122146606445, + "learning_rate": 9.9961949996195e-06, + "loss": 0.2415, + "step": 3807 + }, + { + "epoch": 1.7703393770339377, + "grad_norm": 5.95225715637207, + "learning_rate": 9.9961939996194e-06, + "loss": 0.5444, + "step": 3808 + }, + { + "epoch": 1.7708042770804278, + "grad_norm": 6.879260063171387, + "learning_rate": 9.996192999619301e-06, + "loss": 0.8728, + "step": 3809 + }, + { + "epoch": 1.7712691771269178, + "grad_norm": 4.534753322601318, + "learning_rate": 9.9961919996192e-06, + "loss": 0.2887, + "step": 3810 + }, + { + "epoch": 1.7717340771734076, + "grad_norm": 6.279197692871094, + "learning_rate": 9.9961909996191e-06, + "loss": 0.6394, + "step": 3811 + }, + { + "epoch": 1.7721989772198978, + "grad_norm": 6.178213596343994, + "learning_rate": 9.996189999619e-06, + "loss": 0.5941, + "step": 3812 + }, + { + "epoch": 1.7726638772663876, + "grad_norm": 6.077298164367676, + "learning_rate": 9.9961889996189e-06, + "loss": 0.6442, + "step": 3813 + }, + { + "epoch": 1.7731287773128779, + "grad_norm": 8.640789985656738, + "learning_rate": 9.996187999618801e-06, + "loss": 0.9711, + "step": 3814 + }, + { + "epoch": 1.7735936773593677, + "grad_norm": 5.741477966308594, + "learning_rate": 9.996186999618702e-06, + "loss": 0.9572, + "step": 3815 + }, + { + "epoch": 1.7740585774058577, + "grad_norm": 9.741394996643066, + "learning_rate": 9.9961859996186e-06, + "loss": 0.7478, + "step": 3816 + }, + { + "epoch": 1.7745234774523477, + "grad_norm": 4.5894975662231445, + "learning_rate": 9.996184999618501e-06, + "loss": 0.3872, + "step": 3817 + }, + { + "epoch": 1.7749883774988378, + "grad_norm": 5.093929767608643, + "learning_rate": 9.9961839996184e-06, + "loss": 0.4624, + "step": 3818 + }, + { + "epoch": 1.7754532775453278, + "grad_norm": 8.058117866516113, + "learning_rate": 9.996182999618301e-06, + "loss": 0.4081, + "step": 3819 + }, + { + "epoch": 1.7759181775918178, + "grad_norm": 4.143045425415039, + "learning_rate": 9.996181999618202e-06, + "loss": 0.2438, + "step": 3820 + }, + { + "epoch": 1.7763830776383078, + "grad_norm": 5.0545430183410645, + "learning_rate": 9.9961809996181e-06, + "loss": 0.453, + "step": 3821 + }, + { + "epoch": 1.7768479776847976, + "grad_norm": 5.594204425811768, + "learning_rate": 9.996179999618001e-06, + "loss": 0.5674, + "step": 3822 + }, + { + "epoch": 1.7773128777312879, + "grad_norm": 5.852578639984131, + "learning_rate": 9.9961789996179e-06, + "loss": 0.4502, + "step": 3823 + }, + { + "epoch": 1.7777777777777777, + "grad_norm": 5.0113725662231445, + "learning_rate": 9.996177999617801e-06, + "loss": 0.321, + "step": 3824 + }, + { + "epoch": 1.778242677824268, + "grad_norm": 3.312018394470215, + "learning_rate": 9.9961769996177e-06, + "loss": 0.3319, + "step": 3825 + }, + { + "epoch": 1.7787075778707577, + "grad_norm": 7.8759765625, + "learning_rate": 9.9961759996176e-06, + "loss": 0.7886, + "step": 3826 + }, + { + "epoch": 1.7791724779172478, + "grad_norm": 6.024354934692383, + "learning_rate": 9.996174999617501e-06, + "loss": 0.4814, + "step": 3827 + }, + { + "epoch": 1.7796373779637378, + "grad_norm": 7.74758243560791, + "learning_rate": 9.996173999617402e-06, + "loss": 0.3589, + "step": 3828 + }, + { + "epoch": 1.7801022780102278, + "grad_norm": 4.8964643478393555, + "learning_rate": 9.9961729996173e-06, + "loss": 0.578, + "step": 3829 + }, + { + "epoch": 1.7805671780567178, + "grad_norm": 5.828339099884033, + "learning_rate": 9.996171999617201e-06, + "loss": 0.4363, + "step": 3830 + }, + { + "epoch": 1.7810320781032078, + "grad_norm": 4.112491130828857, + "learning_rate": 9.9961709996171e-06, + "loss": 0.3793, + "step": 3831 + }, + { + "epoch": 1.7814969781496979, + "grad_norm": 4.804193019866943, + "learning_rate": 9.996169999617001e-06, + "loss": 0.449, + "step": 3832 + }, + { + "epoch": 1.7819618781961877, + "grad_norm": 6.619786739349365, + "learning_rate": 9.996168999616902e-06, + "loss": 0.8065, + "step": 3833 + }, + { + "epoch": 1.782426778242678, + "grad_norm": 8.545326232910156, + "learning_rate": 9.9961679996168e-06, + "loss": 0.55, + "step": 3834 + }, + { + "epoch": 1.7828916782891677, + "grad_norm": 6.5814385414123535, + "learning_rate": 9.996166999616701e-06, + "loss": 0.4376, + "step": 3835 + }, + { + "epoch": 1.783356578335658, + "grad_norm": 3.20603346824646, + "learning_rate": 9.9961659996166e-06, + "loss": 0.15, + "step": 3836 + }, + { + "epoch": 1.7838214783821478, + "grad_norm": 6.26145601272583, + "learning_rate": 9.996164999616501e-06, + "loss": 0.7886, + "step": 3837 + }, + { + "epoch": 1.7842863784286378, + "grad_norm": 3.814070224761963, + "learning_rate": 9.9961639996164e-06, + "loss": 0.2319, + "step": 3838 + }, + { + "epoch": 1.7847512784751278, + "grad_norm": 5.597795486450195, + "learning_rate": 9.9961629996163e-06, + "loss": 0.6501, + "step": 3839 + }, + { + "epoch": 1.7852161785216178, + "grad_norm": 5.15036153793335, + "learning_rate": 9.996161999616201e-06, + "loss": 0.3561, + "step": 3840 + }, + { + "epoch": 1.7856810785681079, + "grad_norm": 5.044939041137695, + "learning_rate": 9.996160999616102e-06, + "loss": 0.6266, + "step": 3841 + }, + { + "epoch": 1.786145978614598, + "grad_norm": 5.349740982055664, + "learning_rate": 9.996159999616001e-06, + "loss": 0.5074, + "step": 3842 + }, + { + "epoch": 1.786610878661088, + "grad_norm": 3.5445775985717773, + "learning_rate": 9.9961589996159e-06, + "loss": 0.2443, + "step": 3843 + }, + { + "epoch": 1.7870757787075777, + "grad_norm": 7.518460750579834, + "learning_rate": 9.9961579996158e-06, + "loss": 0.2647, + "step": 3844 + }, + { + "epoch": 1.787540678754068, + "grad_norm": 9.061639785766602, + "learning_rate": 9.996156999615701e-06, + "loss": 0.8243, + "step": 3845 + }, + { + "epoch": 1.7880055788005578, + "grad_norm": 6.0923357009887695, + "learning_rate": 9.996155999615602e-06, + "loss": 0.6194, + "step": 3846 + }, + { + "epoch": 1.788470478847048, + "grad_norm": 6.568515300750732, + "learning_rate": 9.9961549996155e-06, + "loss": 0.4064, + "step": 3847 + }, + { + "epoch": 1.7889353788935378, + "grad_norm": 3.2741971015930176, + "learning_rate": 9.996153999615401e-06, + "loss": 0.2599, + "step": 3848 + }, + { + "epoch": 1.7894002789400278, + "grad_norm": 6.809337615966797, + "learning_rate": 9.9961529996153e-06, + "loss": 0.7811, + "step": 3849 + }, + { + "epoch": 1.7898651789865179, + "grad_norm": 7.543499946594238, + "learning_rate": 9.996151999615201e-06, + "loss": 0.5472, + "step": 3850 + }, + { + "epoch": 1.790330079033008, + "grad_norm": 6.75208044052124, + "learning_rate": 9.9961509996151e-06, + "loss": 0.6173, + "step": 3851 + }, + { + "epoch": 1.790794979079498, + "grad_norm": 4.092970848083496, + "learning_rate": 9.996149999615e-06, + "loss": 0.4099, + "step": 3852 + }, + { + "epoch": 1.791259879125988, + "grad_norm": 4.651479721069336, + "learning_rate": 9.996148999614901e-06, + "loss": 0.4104, + "step": 3853 + }, + { + "epoch": 1.791724779172478, + "grad_norm": 6.019876003265381, + "learning_rate": 9.9961479996148e-06, + "loss": 0.4383, + "step": 3854 + }, + { + "epoch": 1.7921896792189678, + "grad_norm": 4.026945114135742, + "learning_rate": 9.996146999614701e-06, + "loss": 0.2738, + "step": 3855 + }, + { + "epoch": 1.792654579265458, + "grad_norm": 4.653628826141357, + "learning_rate": 9.9961459996146e-06, + "loss": 0.4425, + "step": 3856 + }, + { + "epoch": 1.7931194793119478, + "grad_norm": 8.48490047454834, + "learning_rate": 9.9961449996145e-06, + "loss": 1.0821, + "step": 3857 + }, + { + "epoch": 1.793584379358438, + "grad_norm": 4.240468502044678, + "learning_rate": 9.996143999614401e-06, + "loss": 0.3829, + "step": 3858 + }, + { + "epoch": 1.7940492794049279, + "grad_norm": 7.253324031829834, + "learning_rate": 9.996142999614302e-06, + "loss": 0.7621, + "step": 3859 + }, + { + "epoch": 1.794514179451418, + "grad_norm": 7.624653339385986, + "learning_rate": 9.9961419996142e-06, + "loss": 0.5762, + "step": 3860 + }, + { + "epoch": 1.794979079497908, + "grad_norm": 6.149784564971924, + "learning_rate": 9.996140999614101e-06, + "loss": 0.6682, + "step": 3861 + }, + { + "epoch": 1.795443979544398, + "grad_norm": 7.068849086761475, + "learning_rate": 9.996139999614e-06, + "loss": 0.514, + "step": 3862 + }, + { + "epoch": 1.795908879590888, + "grad_norm": 6.37371826171875, + "learning_rate": 9.996138999613901e-06, + "loss": 0.5734, + "step": 3863 + }, + { + "epoch": 1.796373779637378, + "grad_norm": 4.215679168701172, + "learning_rate": 9.9961379996138e-06, + "loss": 0.4023, + "step": 3864 + }, + { + "epoch": 1.796838679683868, + "grad_norm": 8.55172061920166, + "learning_rate": 9.9961369996137e-06, + "loss": 0.8247, + "step": 3865 + }, + { + "epoch": 1.7973035797303578, + "grad_norm": 9.685636520385742, + "learning_rate": 9.996135999613601e-06, + "loss": 0.7769, + "step": 3866 + }, + { + "epoch": 1.797768479776848, + "grad_norm": 6.197904109954834, + "learning_rate": 9.9961349996135e-06, + "loss": 0.3963, + "step": 3867 + }, + { + "epoch": 1.7982333798233379, + "grad_norm": 7.155196189880371, + "learning_rate": 9.996133999613401e-06, + "loss": 0.6297, + "step": 3868 + }, + { + "epoch": 1.798698279869828, + "grad_norm": 6.872702121734619, + "learning_rate": 9.9961329996133e-06, + "loss": 0.5385, + "step": 3869 + }, + { + "epoch": 1.799163179916318, + "grad_norm": 6.807054042816162, + "learning_rate": 9.9961319996132e-06, + "loss": 0.8248, + "step": 3870 + }, + { + "epoch": 1.7996280799628082, + "grad_norm": 3.877255916595459, + "learning_rate": 9.996130999613101e-06, + "loss": 0.3617, + "step": 3871 + }, + { + "epoch": 1.800092980009298, + "grad_norm": 4.11314582824707, + "learning_rate": 9.996129999613002e-06, + "loss": 0.4406, + "step": 3872 + }, + { + "epoch": 1.800557880055788, + "grad_norm": 6.437507629394531, + "learning_rate": 9.996128999612901e-06, + "loss": 0.6689, + "step": 3873 + }, + { + "epoch": 1.801022780102278, + "grad_norm": 4.048006057739258, + "learning_rate": 9.9961279996128e-06, + "loss": 0.2672, + "step": 3874 + }, + { + "epoch": 1.801487680148768, + "grad_norm": 8.594390869140625, + "learning_rate": 9.9961269996127e-06, + "loss": 0.7267, + "step": 3875 + }, + { + "epoch": 1.801952580195258, + "grad_norm": 8.36662483215332, + "learning_rate": 9.9961259996126e-06, + "loss": 0.4324, + "step": 3876 + }, + { + "epoch": 1.8024174802417479, + "grad_norm": 6.632443904876709, + "learning_rate": 9.9961249996125e-06, + "loss": 0.3981, + "step": 3877 + }, + { + "epoch": 1.802882380288238, + "grad_norm": 12.597007751464844, + "learning_rate": 9.9961239996124e-06, + "loss": 1.4103, + "step": 3878 + }, + { + "epoch": 1.803347280334728, + "grad_norm": 4.217723369598389, + "learning_rate": 9.996122999612301e-06, + "loss": 0.4301, + "step": 3879 + }, + { + "epoch": 1.8038121803812182, + "grad_norm": 4.329727649688721, + "learning_rate": 9.9961219996122e-06, + "loss": 0.5021, + "step": 3880 + }, + { + "epoch": 1.804277080427708, + "grad_norm": 7.705010890960693, + "learning_rate": 9.996120999612101e-06, + "loss": 0.7555, + "step": 3881 + }, + { + "epoch": 1.8047419804741982, + "grad_norm": 7.053459644317627, + "learning_rate": 9.996119999612e-06, + "loss": 0.8603, + "step": 3882 + }, + { + "epoch": 1.805206880520688, + "grad_norm": 4.607448577880859, + "learning_rate": 9.9961189996119e-06, + "loss": 0.3245, + "step": 3883 + }, + { + "epoch": 1.805671780567178, + "grad_norm": 4.685016632080078, + "learning_rate": 9.996117999611801e-06, + "loss": 0.4798, + "step": 3884 + }, + { + "epoch": 1.806136680613668, + "grad_norm": 4.925126552581787, + "learning_rate": 9.9961169996117e-06, + "loss": 0.3979, + "step": 3885 + }, + { + "epoch": 1.806601580660158, + "grad_norm": 6.1241559982299805, + "learning_rate": 9.996115999611601e-06, + "loss": 0.5251, + "step": 3886 + }, + { + "epoch": 1.807066480706648, + "grad_norm": 7.271735668182373, + "learning_rate": 9.9961149996115e-06, + "loss": 0.3899, + "step": 3887 + }, + { + "epoch": 1.8075313807531381, + "grad_norm": 4.573271751403809, + "learning_rate": 9.9961139996114e-06, + "loss": 0.3675, + "step": 3888 + }, + { + "epoch": 1.8079962807996282, + "grad_norm": 6.822360038757324, + "learning_rate": 9.996112999611301e-06, + "loss": 0.4795, + "step": 3889 + }, + { + "epoch": 1.808461180846118, + "grad_norm": 4.922654628753662, + "learning_rate": 9.9961119996112e-06, + "loss": 0.3754, + "step": 3890 + }, + { + "epoch": 1.8089260808926082, + "grad_norm": 4.85671854019165, + "learning_rate": 9.996110999611101e-06, + "loss": 0.3971, + "step": 3891 + }, + { + "epoch": 1.809390980939098, + "grad_norm": 8.138055801391602, + "learning_rate": 9.996109999611002e-06, + "loss": 1.1545, + "step": 3892 + }, + { + "epoch": 1.8098558809855883, + "grad_norm": 6.118337154388428, + "learning_rate": 9.9961089996109e-06, + "loss": 0.5445, + "step": 3893 + }, + { + "epoch": 1.810320781032078, + "grad_norm": 6.9169816970825195, + "learning_rate": 9.996107999610801e-06, + "loss": 0.629, + "step": 3894 + }, + { + "epoch": 1.810785681078568, + "grad_norm": 10.256726264953613, + "learning_rate": 9.9961069996107e-06, + "loss": 0.5383, + "step": 3895 + }, + { + "epoch": 1.811250581125058, + "grad_norm": 7.261306285858154, + "learning_rate": 9.9961059996106e-06, + "loss": 0.7389, + "step": 3896 + }, + { + "epoch": 1.8117154811715481, + "grad_norm": 7.501967430114746, + "learning_rate": 9.996104999610501e-06, + "loss": 0.7388, + "step": 3897 + }, + { + "epoch": 1.8121803812180381, + "grad_norm": 5.630184173583984, + "learning_rate": 9.9961039996104e-06, + "loss": 0.5282, + "step": 3898 + }, + { + "epoch": 1.8126452812645282, + "grad_norm": 5.724059104919434, + "learning_rate": 9.996102999610301e-06, + "loss": 0.631, + "step": 3899 + }, + { + "epoch": 1.8131101813110182, + "grad_norm": 3.977979898452759, + "learning_rate": 9.9961019996102e-06, + "loss": 0.2971, + "step": 3900 + }, + { + "epoch": 1.813575081357508, + "grad_norm": 7.7879486083984375, + "learning_rate": 9.9961009996101e-06, + "loss": 0.7603, + "step": 3901 + }, + { + "epoch": 1.8140399814039982, + "grad_norm": 5.1766581535339355, + "learning_rate": 9.996099999610001e-06, + "loss": 0.5586, + "step": 3902 + }, + { + "epoch": 1.814504881450488, + "grad_norm": 7.79730224609375, + "learning_rate": 9.9960989996099e-06, + "loss": 0.378, + "step": 3903 + }, + { + "epoch": 1.8149697814969783, + "grad_norm": 6.021176815032959, + "learning_rate": 9.996097999609801e-06, + "loss": 0.3383, + "step": 3904 + }, + { + "epoch": 1.815434681543468, + "grad_norm": 5.471567153930664, + "learning_rate": 9.996096999609702e-06, + "loss": 0.6001, + "step": 3905 + }, + { + "epoch": 1.8158995815899581, + "grad_norm": 8.010743141174316, + "learning_rate": 9.9960959996096e-06, + "loss": 0.621, + "step": 3906 + }, + { + "epoch": 1.8163644816364481, + "grad_norm": 4.894117832183838, + "learning_rate": 9.9960949996095e-06, + "loss": 0.2616, + "step": 3907 + }, + { + "epoch": 1.8168293816829382, + "grad_norm": 7.956266403198242, + "learning_rate": 9.9960939996094e-06, + "loss": 0.624, + "step": 3908 + }, + { + "epoch": 1.8172942817294282, + "grad_norm": 5.451071262359619, + "learning_rate": 9.996092999609301e-06, + "loss": 0.5619, + "step": 3909 + }, + { + "epoch": 1.8177591817759182, + "grad_norm": 4.8820929527282715, + "learning_rate": 9.996091999609202e-06, + "loss": 0.3854, + "step": 3910 + }, + { + "epoch": 1.8182240818224082, + "grad_norm": 6.20636510848999, + "learning_rate": 9.9960909996091e-06, + "loss": 0.4397, + "step": 3911 + }, + { + "epoch": 1.818688981868898, + "grad_norm": 6.008831977844238, + "learning_rate": 9.996089999609001e-06, + "loss": 0.7075, + "step": 3912 + }, + { + "epoch": 1.8191538819153883, + "grad_norm": 5.599728107452393, + "learning_rate": 9.9960889996089e-06, + "loss": 0.3873, + "step": 3913 + }, + { + "epoch": 1.819618781961878, + "grad_norm": 6.794156551361084, + "learning_rate": 9.9960879996088e-06, + "loss": 0.7284, + "step": 3914 + }, + { + "epoch": 1.8200836820083683, + "grad_norm": 5.196821212768555, + "learning_rate": 9.996086999608701e-06, + "loss": 0.4296, + "step": 3915 + }, + { + "epoch": 1.8205485820548581, + "grad_norm": 6.922805309295654, + "learning_rate": 9.996085999608602e-06, + "loss": 0.6292, + "step": 3916 + }, + { + "epoch": 1.8210134821013482, + "grad_norm": 14.409069061279297, + "learning_rate": 9.996084999608501e-06, + "loss": 0.5183, + "step": 3917 + }, + { + "epoch": 1.8214783821478382, + "grad_norm": 4.694238662719727, + "learning_rate": 9.9960839996084e-06, + "loss": 0.4065, + "step": 3918 + }, + { + "epoch": 1.8219432821943282, + "grad_norm": 5.377286434173584, + "learning_rate": 9.9960829996083e-06, + "loss": 0.4012, + "step": 3919 + }, + { + "epoch": 1.8224081822408182, + "grad_norm": 5.160375595092773, + "learning_rate": 9.9960819996082e-06, + "loss": 0.4637, + "step": 3920 + }, + { + "epoch": 1.8228730822873083, + "grad_norm": 4.927406311035156, + "learning_rate": 9.9960809996081e-06, + "loss": 0.556, + "step": 3921 + }, + { + "epoch": 1.8233379823337983, + "grad_norm": 5.653059482574463, + "learning_rate": 9.996079999608001e-06, + "loss": 0.5057, + "step": 3922 + }, + { + "epoch": 1.823802882380288, + "grad_norm": 6.456662654876709, + "learning_rate": 9.996078999607902e-06, + "loss": 0.9028, + "step": 3923 + }, + { + "epoch": 1.8242677824267783, + "grad_norm": 5.908863067626953, + "learning_rate": 9.9960779996078e-06, + "loss": 0.6666, + "step": 3924 + }, + { + "epoch": 1.8247326824732681, + "grad_norm": 7.234760284423828, + "learning_rate": 9.996076999607701e-06, + "loss": 0.7901, + "step": 3925 + }, + { + "epoch": 1.8251975825197584, + "grad_norm": 5.869619369506836, + "learning_rate": 9.9960759996076e-06, + "loss": 0.3521, + "step": 3926 + }, + { + "epoch": 1.8256624825662482, + "grad_norm": 7.642486572265625, + "learning_rate": 9.996074999607501e-06, + "loss": 0.6603, + "step": 3927 + }, + { + "epoch": 1.8261273826127382, + "grad_norm": 6.72847843170166, + "learning_rate": 9.996073999607402e-06, + "loss": 0.3534, + "step": 3928 + }, + { + "epoch": 1.8265922826592282, + "grad_norm": 4.069160461425781, + "learning_rate": 9.9960729996073e-06, + "loss": 0.42, + "step": 3929 + }, + { + "epoch": 1.8270571827057183, + "grad_norm": 5.939012050628662, + "learning_rate": 9.996071999607201e-06, + "loss": 0.7373, + "step": 3930 + }, + { + "epoch": 1.8275220827522083, + "grad_norm": 8.733787536621094, + "learning_rate": 9.9960709996071e-06, + "loss": 0.8397, + "step": 3931 + }, + { + "epoch": 1.8279869827986983, + "grad_norm": 3.591513156890869, + "learning_rate": 9.996069999607e-06, + "loss": 0.2216, + "step": 3932 + }, + { + "epoch": 1.8284518828451883, + "grad_norm": 7.037027835845947, + "learning_rate": 9.9960689996069e-06, + "loss": 0.5236, + "step": 3933 + }, + { + "epoch": 1.8289167828916781, + "grad_norm": 4.586752891540527, + "learning_rate": 9.9960679996068e-06, + "loss": 0.346, + "step": 3934 + }, + { + "epoch": 1.8293816829381684, + "grad_norm": 7.021925926208496, + "learning_rate": 9.996066999606701e-06, + "loss": 0.7104, + "step": 3935 + }, + { + "epoch": 1.8298465829846582, + "grad_norm": 6.8194661140441895, + "learning_rate": 9.996065999606602e-06, + "loss": 0.38, + "step": 3936 + }, + { + "epoch": 1.8303114830311484, + "grad_norm": 7.028560638427734, + "learning_rate": 9.9960649996065e-06, + "loss": 0.6886, + "step": 3937 + }, + { + "epoch": 1.8307763830776382, + "grad_norm": 5.755892276763916, + "learning_rate": 9.996063999606401e-06, + "loss": 0.5206, + "step": 3938 + }, + { + "epoch": 1.8312412831241283, + "grad_norm": 6.814227104187012, + "learning_rate": 9.9960629996063e-06, + "loss": 0.5076, + "step": 3939 + }, + { + "epoch": 1.8317061831706183, + "grad_norm": 6.179660320281982, + "learning_rate": 9.996061999606201e-06, + "loss": 0.6288, + "step": 3940 + }, + { + "epoch": 1.8321710832171083, + "grad_norm": 7.85093879699707, + "learning_rate": 9.996060999606102e-06, + "loss": 0.8629, + "step": 3941 + }, + { + "epoch": 1.8326359832635983, + "grad_norm": 6.013796329498291, + "learning_rate": 9.996059999606e-06, + "loss": 0.4922, + "step": 3942 + }, + { + "epoch": 1.8331008833100884, + "grad_norm": 7.886419773101807, + "learning_rate": 9.996058999605901e-06, + "loss": 0.8891, + "step": 3943 + }, + { + "epoch": 1.8335657833565784, + "grad_norm": 9.088862419128418, + "learning_rate": 9.9960579996058e-06, + "loss": 0.9256, + "step": 3944 + }, + { + "epoch": 1.8340306834030682, + "grad_norm": 6.056003570556641, + "learning_rate": 9.9960569996057e-06, + "loss": 0.63, + "step": 3945 + }, + { + "epoch": 1.8344955834495584, + "grad_norm": 6.741253852844238, + "learning_rate": 9.9960559996056e-06, + "loss": 0.6347, + "step": 3946 + }, + { + "epoch": 1.8349604834960482, + "grad_norm": 6.139286041259766, + "learning_rate": 9.9960549996055e-06, + "loss": 0.5391, + "step": 3947 + }, + { + "epoch": 1.8354253835425385, + "grad_norm": 5.562475204467773, + "learning_rate": 9.996053999605401e-06, + "loss": 0.4452, + "step": 3948 + }, + { + "epoch": 1.8358902835890283, + "grad_norm": 9.241254806518555, + "learning_rate": 9.996052999605302e-06, + "loss": 0.6451, + "step": 3949 + }, + { + "epoch": 1.8363551836355185, + "grad_norm": 6.071413516998291, + "learning_rate": 9.9960519996052e-06, + "loss": 0.5814, + "step": 3950 + }, + { + "epoch": 1.8368200836820083, + "grad_norm": 5.4992451667785645, + "learning_rate": 9.9960509996051e-06, + "loss": 0.7025, + "step": 3951 + }, + { + "epoch": 1.8372849837284984, + "grad_norm": 7.552049160003662, + "learning_rate": 9.996049999605e-06, + "loss": 0.6424, + "step": 3952 + }, + { + "epoch": 1.8377498837749884, + "grad_norm": 6.585880279541016, + "learning_rate": 9.996048999604901e-06, + "loss": 0.649, + "step": 3953 + }, + { + "epoch": 1.8382147838214784, + "grad_norm": 5.240715026855469, + "learning_rate": 9.996047999604802e-06, + "loss": 0.4432, + "step": 3954 + }, + { + "epoch": 1.8386796838679684, + "grad_norm": 5.204375267028809, + "learning_rate": 9.9960469996047e-06, + "loss": 0.4791, + "step": 3955 + }, + { + "epoch": 1.8391445839144582, + "grad_norm": 5.604049205780029, + "learning_rate": 9.996045999604601e-06, + "loss": 0.6411, + "step": 3956 + }, + { + "epoch": 1.8396094839609485, + "grad_norm": 7.659050941467285, + "learning_rate": 9.9960449996045e-06, + "loss": 0.5786, + "step": 3957 + }, + { + "epoch": 1.8400743840074383, + "grad_norm": 7.142995357513428, + "learning_rate": 9.996043999604401e-06, + "loss": 0.6606, + "step": 3958 + }, + { + "epoch": 1.8405392840539285, + "grad_norm": 4.961292743682861, + "learning_rate": 9.9960429996043e-06, + "loss": 0.4185, + "step": 3959 + }, + { + "epoch": 1.8410041841004183, + "grad_norm": 7.074472904205322, + "learning_rate": 9.9960419996042e-06, + "loss": 0.5525, + "step": 3960 + }, + { + "epoch": 1.8414690841469086, + "grad_norm": 5.724681854248047, + "learning_rate": 9.996040999604101e-06, + "loss": 0.3304, + "step": 3961 + }, + { + "epoch": 1.8419339841933984, + "grad_norm": 4.238550186157227, + "learning_rate": 9.996039999604e-06, + "loss": 0.3049, + "step": 3962 + }, + { + "epoch": 1.8423988842398884, + "grad_norm": 8.28498649597168, + "learning_rate": 9.9960389996039e-06, + "loss": 0.902, + "step": 3963 + }, + { + "epoch": 1.8428637842863784, + "grad_norm": 4.938450813293457, + "learning_rate": 9.9960379996038e-06, + "loss": 0.3625, + "step": 3964 + }, + { + "epoch": 1.8433286843328685, + "grad_norm": 6.003771781921387, + "learning_rate": 9.9960369996037e-06, + "loss": 0.6854, + "step": 3965 + }, + { + "epoch": 1.8437935843793585, + "grad_norm": 5.119253158569336, + "learning_rate": 9.996035999603601e-06, + "loss": 0.4169, + "step": 3966 + }, + { + "epoch": 1.8442584844258483, + "grad_norm": 5.382247447967529, + "learning_rate": 9.996034999603502e-06, + "loss": 0.4333, + "step": 3967 + }, + { + "epoch": 1.8447233844723385, + "grad_norm": 5.7179036140441895, + "learning_rate": 9.9960339996034e-06, + "loss": 0.4733, + "step": 3968 + }, + { + "epoch": 1.8451882845188283, + "grad_norm": 10.128433227539062, + "learning_rate": 9.996032999603301e-06, + "loss": 1.136, + "step": 3969 + }, + { + "epoch": 1.8456531845653186, + "grad_norm": 4.784890174865723, + "learning_rate": 9.9960319996032e-06, + "loss": 0.5659, + "step": 3970 + }, + { + "epoch": 1.8461180846118084, + "grad_norm": 5.622333526611328, + "learning_rate": 9.996030999603101e-06, + "loss": 0.4901, + "step": 3971 + }, + { + "epoch": 1.8465829846582986, + "grad_norm": 5.433602333068848, + "learning_rate": 9.996029999603e-06, + "loss": 0.6767, + "step": 3972 + }, + { + "epoch": 1.8470478847047884, + "grad_norm": 4.537357330322266, + "learning_rate": 9.9960289996029e-06, + "loss": 0.4067, + "step": 3973 + }, + { + "epoch": 1.8475127847512784, + "grad_norm": 4.76195764541626, + "learning_rate": 9.996027999602801e-06, + "loss": 0.3337, + "step": 3974 + }, + { + "epoch": 1.8479776847977685, + "grad_norm": 6.702915191650391, + "learning_rate": 9.9960269996027e-06, + "loss": 0.4312, + "step": 3975 + }, + { + "epoch": 1.8484425848442585, + "grad_norm": 6.715512275695801, + "learning_rate": 9.996025999602601e-06, + "loss": 0.6859, + "step": 3976 + }, + { + "epoch": 1.8489074848907485, + "grad_norm": 6.8779425621032715, + "learning_rate": 9.9960249996025e-06, + "loss": 0.7027, + "step": 3977 + }, + { + "epoch": 1.8493723849372385, + "grad_norm": 7.576364994049072, + "learning_rate": 9.9960239996024e-06, + "loss": 0.4037, + "step": 3978 + }, + { + "epoch": 1.8498372849837286, + "grad_norm": 4.778060436248779, + "learning_rate": 9.996022999602301e-06, + "loss": 0.438, + "step": 3979 + }, + { + "epoch": 1.8503021850302184, + "grad_norm": 10.108707427978516, + "learning_rate": 9.996021999602202e-06, + "loss": 0.2811, + "step": 3980 + }, + { + "epoch": 1.8507670850767086, + "grad_norm": 4.211582660675049, + "learning_rate": 9.9960209996021e-06, + "loss": 0.303, + "step": 3981 + }, + { + "epoch": 1.8512319851231984, + "grad_norm": 5.687553882598877, + "learning_rate": 9.996019999602001e-06, + "loss": 0.6102, + "step": 3982 + }, + { + "epoch": 1.8516968851696887, + "grad_norm": 9.356795310974121, + "learning_rate": 9.9960189996019e-06, + "loss": 0.5013, + "step": 3983 + }, + { + "epoch": 1.8521617852161785, + "grad_norm": 6.956449031829834, + "learning_rate": 9.996017999601801e-06, + "loss": 0.6754, + "step": 3984 + }, + { + "epoch": 1.8526266852626685, + "grad_norm": 6.736838340759277, + "learning_rate": 9.996016999601702e-06, + "loss": 0.86, + "step": 3985 + }, + { + "epoch": 1.8530915853091585, + "grad_norm": 5.469196796417236, + "learning_rate": 9.9960159996016e-06, + "loss": 0.7557, + "step": 3986 + }, + { + "epoch": 1.8535564853556485, + "grad_norm": 5.075472354888916, + "learning_rate": 9.996014999601501e-06, + "loss": 0.3769, + "step": 3987 + }, + { + "epoch": 1.8540213854021386, + "grad_norm": 6.220958709716797, + "learning_rate": 9.9960139996014e-06, + "loss": 0.3977, + "step": 3988 + }, + { + "epoch": 1.8544862854486286, + "grad_norm": 7.265904426574707, + "learning_rate": 9.996012999601301e-06, + "loss": 0.9077, + "step": 3989 + }, + { + "epoch": 1.8549511854951186, + "grad_norm": 7.7932820320129395, + "learning_rate": 9.9960119996012e-06, + "loss": 0.8412, + "step": 3990 + }, + { + "epoch": 1.8554160855416084, + "grad_norm": 4.671389102935791, + "learning_rate": 9.9960109996011e-06, + "loss": 0.411, + "step": 3991 + }, + { + "epoch": 1.8558809855880987, + "grad_norm": 5.928183555603027, + "learning_rate": 9.996009999601001e-06, + "loss": 0.5194, + "step": 3992 + }, + { + "epoch": 1.8563458856345885, + "grad_norm": 3.525024652481079, + "learning_rate": 9.996008999600902e-06, + "loss": 0.2711, + "step": 3993 + }, + { + "epoch": 1.8568107856810787, + "grad_norm": 5.2591047286987305, + "learning_rate": 9.996007999600801e-06, + "loss": 0.3942, + "step": 3994 + }, + { + "epoch": 1.8572756857275685, + "grad_norm": 4.49232816696167, + "learning_rate": 9.9960069996007e-06, + "loss": 0.4184, + "step": 3995 + }, + { + "epoch": 1.8577405857740585, + "grad_norm": 4.086883544921875, + "learning_rate": 9.9960059996006e-06, + "loss": 0.3012, + "step": 3996 + }, + { + "epoch": 1.8582054858205486, + "grad_norm": 7.733066082000732, + "learning_rate": 9.996004999600501e-06, + "loss": 0.5206, + "step": 3997 + }, + { + "epoch": 1.8586703858670386, + "grad_norm": 6.054320335388184, + "learning_rate": 9.996003999600402e-06, + "loss": 0.7134, + "step": 3998 + }, + { + "epoch": 1.8591352859135286, + "grad_norm": 5.136349678039551, + "learning_rate": 9.9960029996003e-06, + "loss": 0.3928, + "step": 3999 + }, + { + "epoch": 1.8596001859600186, + "grad_norm": 5.63085412979126, + "learning_rate": 9.996001999600201e-06, + "loss": 0.6435, + "step": 4000 + }, + { + "epoch": 1.8600650860065087, + "grad_norm": 3.981372594833374, + "learning_rate": 9.9960009996001e-06, + "loss": 0.4415, + "step": 4001 + }, + { + "epoch": 1.8605299860529985, + "grad_norm": 7.589320659637451, + "learning_rate": 9.995999999600001e-06, + "loss": 0.5758, + "step": 4002 + }, + { + "epoch": 1.8609948860994887, + "grad_norm": 9.60495662689209, + "learning_rate": 9.9959989995999e-06, + "loss": 0.4996, + "step": 4003 + }, + { + "epoch": 1.8614597861459785, + "grad_norm": 4.568313121795654, + "learning_rate": 9.9959979995998e-06, + "loss": 0.4186, + "step": 4004 + }, + { + "epoch": 1.8619246861924688, + "grad_norm": 6.978579521179199, + "learning_rate": 9.995996999599701e-06, + "loss": 0.8524, + "step": 4005 + }, + { + "epoch": 1.8623895862389586, + "grad_norm": 7.425955295562744, + "learning_rate": 9.9959959995996e-06, + "loss": 0.4834, + "step": 4006 + }, + { + "epoch": 1.8628544862854486, + "grad_norm": 4.16478157043457, + "learning_rate": 9.995994999599501e-06, + "loss": 0.4085, + "step": 4007 + }, + { + "epoch": 1.8633193863319386, + "grad_norm": 7.23736572265625, + "learning_rate": 9.9959939995994e-06, + "loss": 0.5764, + "step": 4008 + }, + { + "epoch": 1.8637842863784286, + "grad_norm": 6.041782855987549, + "learning_rate": 9.9959929995993e-06, + "loss": 0.7859, + "step": 4009 + }, + { + "epoch": 1.8642491864249187, + "grad_norm": 6.355955600738525, + "learning_rate": 9.995991999599201e-06, + "loss": 0.685, + "step": 4010 + }, + { + "epoch": 1.8647140864714087, + "grad_norm": 6.525869369506836, + "learning_rate": 9.995990999599102e-06, + "loss": 0.5211, + "step": 4011 + }, + { + "epoch": 1.8651789865178987, + "grad_norm": 8.033628463745117, + "learning_rate": 9.995989999599001e-06, + "loss": 0.8307, + "step": 4012 + }, + { + "epoch": 1.8656438865643885, + "grad_norm": 7.869875431060791, + "learning_rate": 9.995988999598902e-06, + "loss": 0.6525, + "step": 4013 + }, + { + "epoch": 1.8661087866108788, + "grad_norm": 5.643305778503418, + "learning_rate": 9.9959879995988e-06, + "loss": 0.6644, + "step": 4014 + }, + { + "epoch": 1.8665736866573686, + "grad_norm": 5.985032081604004, + "learning_rate": 9.9959869995987e-06, + "loss": 0.4261, + "step": 4015 + }, + { + "epoch": 1.8670385867038588, + "grad_norm": 6.071505546569824, + "learning_rate": 9.9959859995986e-06, + "loss": 0.6002, + "step": 4016 + }, + { + "epoch": 1.8675034867503486, + "grad_norm": 4.293980598449707, + "learning_rate": 9.9959849995985e-06, + "loss": 0.3876, + "step": 4017 + }, + { + "epoch": 1.8679683867968386, + "grad_norm": 4.699156761169434, + "learning_rate": 9.995983999598401e-06, + "loss": 0.4694, + "step": 4018 + }, + { + "epoch": 1.8684332868433287, + "grad_norm": 4.740266799926758, + "learning_rate": 9.9959829995983e-06, + "loss": 0.5197, + "step": 4019 + }, + { + "epoch": 1.8688981868898187, + "grad_norm": 6.9483256340026855, + "learning_rate": 9.995981999598201e-06, + "loss": 0.6458, + "step": 4020 + }, + { + "epoch": 1.8693630869363087, + "grad_norm": 6.334929466247559, + "learning_rate": 9.9959809995981e-06, + "loss": 0.7973, + "step": 4021 + }, + { + "epoch": 1.8698279869827987, + "grad_norm": 5.822216987609863, + "learning_rate": 9.995979999598e-06, + "loss": 0.5844, + "step": 4022 + }, + { + "epoch": 1.8702928870292888, + "grad_norm": 4.6156463623046875, + "learning_rate": 9.995978999597901e-06, + "loss": 0.4888, + "step": 4023 + }, + { + "epoch": 1.8707577870757786, + "grad_norm": 4.760200500488281, + "learning_rate": 9.995977999597802e-06, + "loss": 0.4223, + "step": 4024 + }, + { + "epoch": 1.8712226871222688, + "grad_norm": 6.338533401489258, + "learning_rate": 9.995976999597701e-06, + "loss": 0.4412, + "step": 4025 + }, + { + "epoch": 1.8716875871687586, + "grad_norm": 3.882080316543579, + "learning_rate": 9.9959759995976e-06, + "loss": 0.2344, + "step": 4026 + }, + { + "epoch": 1.8721524872152489, + "grad_norm": 7.679769515991211, + "learning_rate": 9.9959749995975e-06, + "loss": 0.5292, + "step": 4027 + }, + { + "epoch": 1.8726173872617387, + "grad_norm": 8.450347900390625, + "learning_rate": 9.9959739995974e-06, + "loss": 0.5697, + "step": 4028 + }, + { + "epoch": 1.8730822873082287, + "grad_norm": 3.8492209911346436, + "learning_rate": 9.9959729995973e-06, + "loss": 0.2957, + "step": 4029 + }, + { + "epoch": 1.8735471873547187, + "grad_norm": 6.676304817199707, + "learning_rate": 9.995971999597201e-06, + "loss": 0.7761, + "step": 4030 + }, + { + "epoch": 1.8740120874012087, + "grad_norm": 6.2436113357543945, + "learning_rate": 9.995970999597102e-06, + "loss": 0.4534, + "step": 4031 + }, + { + "epoch": 1.8744769874476988, + "grad_norm": 4.441768646240234, + "learning_rate": 9.995969999597e-06, + "loss": 0.4046, + "step": 4032 + }, + { + "epoch": 1.8749418874941888, + "grad_norm": 5.412728309631348, + "learning_rate": 9.995968999596901e-06, + "loss": 0.3817, + "step": 4033 + }, + { + "epoch": 1.8754067875406788, + "grad_norm": 5.42721700668335, + "learning_rate": 9.9959679995968e-06, + "loss": 0.5114, + "step": 4034 + }, + { + "epoch": 1.8758716875871686, + "grad_norm": 7.847777366638184, + "learning_rate": 9.9959669995967e-06, + "loss": 0.5774, + "step": 4035 + }, + { + "epoch": 1.8763365876336588, + "grad_norm": 9.258280754089355, + "learning_rate": 9.995965999596601e-06, + "loss": 0.5016, + "step": 4036 + }, + { + "epoch": 1.8768014876801487, + "grad_norm": 5.2107768058776855, + "learning_rate": 9.9959649995965e-06, + "loss": 0.4041, + "step": 4037 + }, + { + "epoch": 1.877266387726639, + "grad_norm": 4.412253379821777, + "learning_rate": 9.995963999596401e-06, + "loss": 0.3903, + "step": 4038 + }, + { + "epoch": 1.8777312877731287, + "grad_norm": 5.009997367858887, + "learning_rate": 9.9959629995963e-06, + "loss": 0.4435, + "step": 4039 + }, + { + "epoch": 1.878196187819619, + "grad_norm": 5.421339988708496, + "learning_rate": 9.9959619995962e-06, + "loss": 0.4604, + "step": 4040 + }, + { + "epoch": 1.8786610878661087, + "grad_norm": 5.017758369445801, + "learning_rate": 9.9959609995961e-06, + "loss": 0.6059, + "step": 4041 + }, + { + "epoch": 1.8791259879125988, + "grad_norm": 9.530436515808105, + "learning_rate": 9.995959999596e-06, + "loss": 1.0526, + "step": 4042 + }, + { + "epoch": 1.8795908879590888, + "grad_norm": 7.693924427032471, + "learning_rate": 9.995958999595901e-06, + "loss": 0.7224, + "step": 4043 + }, + { + "epoch": 1.8800557880055788, + "grad_norm": 6.942763328552246, + "learning_rate": 9.995957999595802e-06, + "loss": 0.7907, + "step": 4044 + }, + { + "epoch": 1.8805206880520688, + "grad_norm": 6.317469596862793, + "learning_rate": 9.9959569995957e-06, + "loss": 0.755, + "step": 4045 + }, + { + "epoch": 1.8809855880985586, + "grad_norm": 7.073362350463867, + "learning_rate": 9.995955999595601e-06, + "loss": 0.5301, + "step": 4046 + }, + { + "epoch": 1.881450488145049, + "grad_norm": 6.251765727996826, + "learning_rate": 9.9959549995955e-06, + "loss": 0.3272, + "step": 4047 + }, + { + "epoch": 1.8819153881915387, + "grad_norm": 6.392611026763916, + "learning_rate": 9.9959539995954e-06, + "loss": 0.4416, + "step": 4048 + }, + { + "epoch": 1.882380288238029, + "grad_norm": 8.273425102233887, + "learning_rate": 9.995952999595301e-06, + "loss": 0.4589, + "step": 4049 + }, + { + "epoch": 1.8828451882845187, + "grad_norm": 6.166413307189941, + "learning_rate": 9.9959519995952e-06, + "loss": 0.6694, + "step": 4050 + }, + { + "epoch": 1.883310088331009, + "grad_norm": 6.601210594177246, + "learning_rate": 9.995950999595101e-06, + "loss": 0.8455, + "step": 4051 + }, + { + "epoch": 1.8837749883774988, + "grad_norm": 6.6365251541137695, + "learning_rate": 9.995949999595e-06, + "loss": 0.7775, + "step": 4052 + }, + { + "epoch": 1.8842398884239888, + "grad_norm": 7.481632232666016, + "learning_rate": 9.9959489995949e-06, + "loss": 0.742, + "step": 4053 + }, + { + "epoch": 1.8847047884704788, + "grad_norm": 5.252692699432373, + "learning_rate": 9.995947999594801e-06, + "loss": 0.3709, + "step": 4054 + }, + { + "epoch": 1.8851696885169689, + "grad_norm": 5.171414375305176, + "learning_rate": 9.9959469995947e-06, + "loss": 0.6416, + "step": 4055 + }, + { + "epoch": 1.885634588563459, + "grad_norm": 4.865359783172607, + "learning_rate": 9.995945999594601e-06, + "loss": 0.433, + "step": 4056 + }, + { + "epoch": 1.886099488609949, + "grad_norm": 5.758243083953857, + "learning_rate": 9.995944999594502e-06, + "loss": 0.4666, + "step": 4057 + }, + { + "epoch": 1.886564388656439, + "grad_norm": 4.766274452209473, + "learning_rate": 9.9959439995944e-06, + "loss": 0.3805, + "step": 4058 + }, + { + "epoch": 1.8870292887029287, + "grad_norm": 4.137103080749512, + "learning_rate": 9.9959429995943e-06, + "loss": 0.3159, + "step": 4059 + }, + { + "epoch": 1.887494188749419, + "grad_norm": 4.915076732635498, + "learning_rate": 9.9959419995942e-06, + "loss": 0.581, + "step": 4060 + }, + { + "epoch": 1.8879590887959088, + "grad_norm": 11.48361587524414, + "learning_rate": 9.995940999594101e-06, + "loss": 1.1333, + "step": 4061 + }, + { + "epoch": 1.888423988842399, + "grad_norm": 6.797233581542969, + "learning_rate": 9.995939999594002e-06, + "loss": 0.5239, + "step": 4062 + }, + { + "epoch": 1.8888888888888888, + "grad_norm": 7.765871047973633, + "learning_rate": 9.9959389995939e-06, + "loss": 0.8059, + "step": 4063 + }, + { + "epoch": 1.8893537889353789, + "grad_norm": 7.067188262939453, + "learning_rate": 9.995937999593801e-06, + "loss": 0.5396, + "step": 4064 + }, + { + "epoch": 1.8898186889818689, + "grad_norm": 7.46940279006958, + "learning_rate": 9.9959369995937e-06, + "loss": 0.7809, + "step": 4065 + }, + { + "epoch": 1.890283589028359, + "grad_norm": Infinity, + "learning_rate": 9.9959369995937e-06, + "loss": 0.853, + "step": 4066 + }, + { + "epoch": 1.890748489074849, + "grad_norm": 4.590436935424805, + "learning_rate": 9.9959359995936e-06, + "loss": 0.4913, + "step": 4067 + }, + { + "epoch": 1.891213389121339, + "grad_norm": 5.327308654785156, + "learning_rate": 9.995934999593501e-06, + "loss": 0.4904, + "step": 4068 + }, + { + "epoch": 1.891678289167829, + "grad_norm": 3.5848991870880127, + "learning_rate": 9.9959339995934e-06, + "loss": 0.2437, + "step": 4069 + }, + { + "epoch": 1.8921431892143188, + "grad_norm": 5.999992847442627, + "learning_rate": 9.995932999593301e-06, + "loss": 0.2805, + "step": 4070 + }, + { + "epoch": 1.892608089260809, + "grad_norm": 4.863650321960449, + "learning_rate": 9.9959319995932e-06, + "loss": 0.4208, + "step": 4071 + }, + { + "epoch": 1.8930729893072988, + "grad_norm": 4.352815628051758, + "learning_rate": 9.9959309995931e-06, + "loss": 0.5423, + "step": 4072 + }, + { + "epoch": 1.893537889353789, + "grad_norm": 4.170658111572266, + "learning_rate": 9.995929999593e-06, + "loss": 0.3459, + "step": 4073 + }, + { + "epoch": 1.8940027894002789, + "grad_norm": 6.915919780731201, + "learning_rate": 9.9959289995929e-06, + "loss": 0.8592, + "step": 4074 + }, + { + "epoch": 1.894467689446769, + "grad_norm": 5.944794178009033, + "learning_rate": 9.995927999592801e-06, + "loss": 0.5507, + "step": 4075 + }, + { + "epoch": 1.894932589493259, + "grad_norm": 6.148126602172852, + "learning_rate": 9.995926999592702e-06, + "loss": 0.4815, + "step": 4076 + }, + { + "epoch": 1.895397489539749, + "grad_norm": 6.198424816131592, + "learning_rate": 9.9959259995926e-06, + "loss": 0.3834, + "step": 4077 + }, + { + "epoch": 1.895862389586239, + "grad_norm": 6.265298366546631, + "learning_rate": 9.995924999592501e-06, + "loss": 0.5431, + "step": 4078 + }, + { + "epoch": 1.896327289632729, + "grad_norm": 5.791954040527344, + "learning_rate": 9.9959239995924e-06, + "loss": 0.5716, + "step": 4079 + }, + { + "epoch": 1.896792189679219, + "grad_norm": 3.9306938648223877, + "learning_rate": 9.995922999592301e-06, + "loss": 0.419, + "step": 4080 + }, + { + "epoch": 1.8972570897257088, + "grad_norm": 4.499544620513916, + "learning_rate": 9.995921999592202e-06, + "loss": 0.3453, + "step": 4081 + }, + { + "epoch": 1.897721989772199, + "grad_norm": 8.280305862426758, + "learning_rate": 9.9959209995921e-06, + "loss": 0.7192, + "step": 4082 + }, + { + "epoch": 1.8981868898186889, + "grad_norm": 8.863299369812012, + "learning_rate": 9.995919999592001e-06, + "loss": 0.6969, + "step": 4083 + }, + { + "epoch": 1.8986517898651791, + "grad_norm": 8.314441680908203, + "learning_rate": 9.9959189995919e-06, + "loss": 0.6163, + "step": 4084 + }, + { + "epoch": 1.899116689911669, + "grad_norm": 4.29686164855957, + "learning_rate": 9.9959179995918e-06, + "loss": 0.4354, + "step": 4085 + }, + { + "epoch": 1.899581589958159, + "grad_norm": 6.065218925476074, + "learning_rate": 9.9959169995917e-06, + "loss": 0.6244, + "step": 4086 + }, + { + "epoch": 1.900046490004649, + "grad_norm": 4.031548976898193, + "learning_rate": 9.9959159995916e-06, + "loss": 0.3662, + "step": 4087 + }, + { + "epoch": 1.900511390051139, + "grad_norm": 4.47170352935791, + "learning_rate": 9.995914999591501e-06, + "loss": 0.3568, + "step": 4088 + }, + { + "epoch": 1.900976290097629, + "grad_norm": 4.585382461547852, + "learning_rate": 9.995913999591402e-06, + "loss": 0.4024, + "step": 4089 + }, + { + "epoch": 1.901441190144119, + "grad_norm": 5.918221950531006, + "learning_rate": 9.9959129995913e-06, + "loss": 0.6304, + "step": 4090 + }, + { + "epoch": 1.901906090190609, + "grad_norm": 4.6386942863464355, + "learning_rate": 9.995911999591201e-06, + "loss": 0.3843, + "step": 4091 + }, + { + "epoch": 1.9023709902370989, + "grad_norm": 10.189749717712402, + "learning_rate": 9.9959109995911e-06, + "loss": 0.4645, + "step": 4092 + }, + { + "epoch": 1.9028358902835891, + "grad_norm": 5.148841857910156, + "learning_rate": 9.995909999591001e-06, + "loss": 0.5243, + "step": 4093 + }, + { + "epoch": 1.903300790330079, + "grad_norm": 9.912589073181152, + "learning_rate": 9.995908999590902e-06, + "loss": 0.5195, + "step": 4094 + }, + { + "epoch": 1.9037656903765692, + "grad_norm": 7.3616414070129395, + "learning_rate": 9.9959079995908e-06, + "loss": 0.9273, + "step": 4095 + }, + { + "epoch": 1.904230590423059, + "grad_norm": 9.715949058532715, + "learning_rate": 9.995906999590701e-06, + "loss": 0.8705, + "step": 4096 + }, + { + "epoch": 1.904695490469549, + "grad_norm": 5.82841682434082, + "learning_rate": 9.9959059995906e-06, + "loss": 0.5715, + "step": 4097 + }, + { + "epoch": 1.905160390516039, + "grad_norm": 6.0427117347717285, + "learning_rate": 9.995904999590501e-06, + "loss": 0.4436, + "step": 4098 + }, + { + "epoch": 1.905625290562529, + "grad_norm": 4.354869842529297, + "learning_rate": 9.9959039995904e-06, + "loss": 0.4277, + "step": 4099 + }, + { + "epoch": 1.906090190609019, + "grad_norm": 4.3649516105651855, + "learning_rate": 9.9959029995903e-06, + "loss": 0.473, + "step": 4100 + }, + { + "epoch": 1.906555090655509, + "grad_norm": 7.016865253448486, + "learning_rate": 9.995901999590201e-06, + "loss": 0.4152, + "step": 4101 + }, + { + "epoch": 1.9070199907019991, + "grad_norm": 6.011719703674316, + "learning_rate": 9.995900999590102e-06, + "loss": 0.4317, + "step": 4102 + }, + { + "epoch": 1.907484890748489, + "grad_norm": 9.552576065063477, + "learning_rate": 9.99589999959e-06, + "loss": 1.1617, + "step": 4103 + }, + { + "epoch": 1.9079497907949792, + "grad_norm": 4.513777732849121, + "learning_rate": 9.9958989995899e-06, + "loss": 0.4501, + "step": 4104 + }, + { + "epoch": 1.908414690841469, + "grad_norm": 4.494040012359619, + "learning_rate": 9.9958979995898e-06, + "loss": 0.3924, + "step": 4105 + }, + { + "epoch": 1.9088795908879592, + "grad_norm": 5.822569370269775, + "learning_rate": 9.995896999589701e-06, + "loss": 0.6516, + "step": 4106 + }, + { + "epoch": 1.909344490934449, + "grad_norm": 8.447339057922363, + "learning_rate": 9.995895999589602e-06, + "loss": 0.7124, + "step": 4107 + }, + { + "epoch": 1.909809390980939, + "grad_norm": 4.768677234649658, + "learning_rate": 9.9958949995895e-06, + "loss": 0.5163, + "step": 4108 + }, + { + "epoch": 1.910274291027429, + "grad_norm": 5.76719856262207, + "learning_rate": 9.995893999589401e-06, + "loss": 0.5255, + "step": 4109 + }, + { + "epoch": 1.910739191073919, + "grad_norm": 5.216991901397705, + "learning_rate": 9.9958929995893e-06, + "loss": 0.2768, + "step": 4110 + }, + { + "epoch": 1.9112040911204091, + "grad_norm": 5.413909435272217, + "learning_rate": 9.995891999589201e-06, + "loss": 0.6209, + "step": 4111 + }, + { + "epoch": 1.9116689911668991, + "grad_norm": 5.1020660400390625, + "learning_rate": 9.9958909995891e-06, + "loss": 0.4956, + "step": 4112 + }, + { + "epoch": 1.9121338912133892, + "grad_norm": 5.06400728225708, + "learning_rate": 9.995889999589e-06, + "loss": 0.5145, + "step": 4113 + }, + { + "epoch": 1.912598791259879, + "grad_norm": 9.871618270874023, + "learning_rate": 9.995888999588901e-06, + "loss": 0.5298, + "step": 4114 + }, + { + "epoch": 1.9130636913063692, + "grad_norm": 4.9306721687316895, + "learning_rate": 9.9958879995888e-06, + "loss": 0.6285, + "step": 4115 + }, + { + "epoch": 1.913528591352859, + "grad_norm": 9.626954078674316, + "learning_rate": 9.995886999588701e-06, + "loss": 1.2585, + "step": 4116 + }, + { + "epoch": 1.9139934913993493, + "grad_norm": 6.616055011749268, + "learning_rate": 9.9958859995886e-06, + "loss": 0.4424, + "step": 4117 + }, + { + "epoch": 1.914458391445839, + "grad_norm": 6.997819423675537, + "learning_rate": 9.9958849995885e-06, + "loss": 0.5345, + "step": 4118 + }, + { + "epoch": 1.914923291492329, + "grad_norm": 5.886908531188965, + "learning_rate": 9.995883999588401e-06, + "loss": 0.4146, + "step": 4119 + }, + { + "epoch": 1.9153881915388191, + "grad_norm": 5.165041446685791, + "learning_rate": 9.995882999588302e-06, + "loss": 0.3914, + "step": 4120 + }, + { + "epoch": 1.9158530915853091, + "grad_norm": 4.869058132171631, + "learning_rate": 9.9958819995882e-06, + "loss": 0.3466, + "step": 4121 + }, + { + "epoch": 1.9163179916317992, + "grad_norm": 4.913190841674805, + "learning_rate": 9.995880999588101e-06, + "loss": 0.5361, + "step": 4122 + }, + { + "epoch": 1.9167828916782892, + "grad_norm": 7.06567907333374, + "learning_rate": 9.995879999588e-06, + "loss": 0.67, + "step": 4123 + }, + { + "epoch": 1.9172477917247792, + "grad_norm": 6.707674503326416, + "learning_rate": 9.995878999587901e-06, + "loss": 0.4013, + "step": 4124 + }, + { + "epoch": 1.917712691771269, + "grad_norm": 6.0370659828186035, + "learning_rate": 9.9958779995878e-06, + "loss": 0.6588, + "step": 4125 + }, + { + "epoch": 1.9181775918177593, + "grad_norm": 5.688657760620117, + "learning_rate": 9.9958769995877e-06, + "loss": 0.6674, + "step": 4126 + }, + { + "epoch": 1.918642491864249, + "grad_norm": 3.526567220687866, + "learning_rate": 9.995875999587601e-06, + "loss": 0.2118, + "step": 4127 + }, + { + "epoch": 1.9191073919107393, + "grad_norm": 7.596861362457275, + "learning_rate": 9.9958749995875e-06, + "loss": 0.5736, + "step": 4128 + }, + { + "epoch": 1.9195722919572291, + "grad_norm": 5.667518138885498, + "learning_rate": 9.995873999587401e-06, + "loss": 0.366, + "step": 4129 + }, + { + "epoch": 1.9200371920037194, + "grad_norm": 4.875471115112305, + "learning_rate": 9.9958729995873e-06, + "loss": 0.4879, + "step": 4130 + }, + { + "epoch": 1.9205020920502092, + "grad_norm": 5.3464789390563965, + "learning_rate": 9.9958719995872e-06, + "loss": 0.5239, + "step": 4131 + }, + { + "epoch": 1.9209669920966992, + "grad_norm": 9.458271026611328, + "learning_rate": 9.995870999587101e-06, + "loss": 0.5686, + "step": 4132 + }, + { + "epoch": 1.9214318921431892, + "grad_norm": 4.721645832061768, + "learning_rate": 9.995869999587002e-06, + "loss": 0.4648, + "step": 4133 + }, + { + "epoch": 1.9218967921896792, + "grad_norm": 5.733346939086914, + "learning_rate": 9.9958689995869e-06, + "loss": 0.4529, + "step": 4134 + }, + { + "epoch": 1.9223616922361693, + "grad_norm": 5.566469669342041, + "learning_rate": 9.995867999586801e-06, + "loss": 0.604, + "step": 4135 + }, + { + "epoch": 1.922826592282659, + "grad_norm": 5.981134414672852, + "learning_rate": 9.9958669995867e-06, + "loss": 0.3604, + "step": 4136 + }, + { + "epoch": 1.9232914923291493, + "grad_norm": 21.449222564697266, + "learning_rate": 9.995865999586601e-06, + "loss": 0.9555, + "step": 4137 + }, + { + "epoch": 1.9237563923756391, + "grad_norm": 4.326531887054443, + "learning_rate": 9.9958649995865e-06, + "loss": 0.32, + "step": 4138 + }, + { + "epoch": 1.9242212924221294, + "grad_norm": 4.4277448654174805, + "learning_rate": 9.9958639995864e-06, + "loss": 0.3911, + "step": 4139 + }, + { + "epoch": 1.9246861924686192, + "grad_norm": 6.067160606384277, + "learning_rate": 9.995862999586301e-06, + "loss": 0.6044, + "step": 4140 + }, + { + "epoch": 1.9251510925151094, + "grad_norm": 3.3782577514648438, + "learning_rate": 9.9958619995862e-06, + "loss": 0.2995, + "step": 4141 + }, + { + "epoch": 1.9256159925615992, + "grad_norm": 7.15701961517334, + "learning_rate": 9.995860999586101e-06, + "loss": 0.8655, + "step": 4142 + }, + { + "epoch": 1.9260808926080892, + "grad_norm": 9.248754501342773, + "learning_rate": 9.995859999586e-06, + "loss": 0.7956, + "step": 4143 + }, + { + "epoch": 1.9265457926545793, + "grad_norm": 6.66008186340332, + "learning_rate": 9.9958589995859e-06, + "loss": 0.521, + "step": 4144 + }, + { + "epoch": 1.9270106927010693, + "grad_norm": 5.8256144523620605, + "learning_rate": 9.995857999585801e-06, + "loss": 0.5032, + "step": 4145 + }, + { + "epoch": 1.9274755927475593, + "grad_norm": 8.384822845458984, + "learning_rate": 9.995856999585702e-06, + "loss": 1.058, + "step": 4146 + }, + { + "epoch": 1.9279404927940493, + "grad_norm": 7.669069290161133, + "learning_rate": 9.995855999585601e-06, + "loss": 0.3009, + "step": 4147 + }, + { + "epoch": 1.9284053928405394, + "grad_norm": 5.588844299316406, + "learning_rate": 9.9958549995855e-06, + "loss": 0.4585, + "step": 4148 + }, + { + "epoch": 1.9288702928870292, + "grad_norm": 6.281683444976807, + "learning_rate": 9.9958539995854e-06, + "loss": 0.5092, + "step": 4149 + }, + { + "epoch": 1.9293351929335194, + "grad_norm": 8.069167137145996, + "learning_rate": 9.995852999585301e-06, + "loss": 0.5769, + "step": 4150 + }, + { + "epoch": 1.9298000929800092, + "grad_norm": 6.4850172996521, + "learning_rate": 9.995851999585202e-06, + "loss": 0.6756, + "step": 4151 + }, + { + "epoch": 1.9302649930264995, + "grad_norm": 4.684398651123047, + "learning_rate": 9.9958509995851e-06, + "loss": 0.473, + "step": 4152 + }, + { + "epoch": 1.9307298930729893, + "grad_norm": 7.6515583992004395, + "learning_rate": 9.995849999585001e-06, + "loss": 0.4593, + "step": 4153 + }, + { + "epoch": 1.9311947931194793, + "grad_norm": 7.227930545806885, + "learning_rate": 9.9958489995849e-06, + "loss": 0.8807, + "step": 4154 + }, + { + "epoch": 1.9316596931659693, + "grad_norm": 4.527309417724609, + "learning_rate": 9.995847999584801e-06, + "loss": 0.2803, + "step": 4155 + }, + { + "epoch": 1.9321245932124593, + "grad_norm": 5.1004462242126465, + "learning_rate": 9.9958469995847e-06, + "loss": 0.3748, + "step": 4156 + }, + { + "epoch": 1.9325894932589494, + "grad_norm": 6.301362037658691, + "learning_rate": 9.9958459995846e-06, + "loss": 0.5257, + "step": 4157 + }, + { + "epoch": 1.9330543933054394, + "grad_norm": 5.988612651824951, + "learning_rate": 9.995844999584501e-06, + "loss": 0.5934, + "step": 4158 + }, + { + "epoch": 1.9335192933519294, + "grad_norm": 4.975613594055176, + "learning_rate": 9.9958439995844e-06, + "loss": 0.4196, + "step": 4159 + }, + { + "epoch": 1.9339841933984192, + "grad_norm": 5.682332992553711, + "learning_rate": 9.995842999584301e-06, + "loss": 0.7989, + "step": 4160 + }, + { + "epoch": 1.9344490934449095, + "grad_norm": 6.418298244476318, + "learning_rate": 9.9958419995842e-06, + "loss": 0.4349, + "step": 4161 + }, + { + "epoch": 1.9349139934913993, + "grad_norm": 6.164441108703613, + "learning_rate": 9.9958409995841e-06, + "loss": 0.523, + "step": 4162 + }, + { + "epoch": 1.9353788935378895, + "grad_norm": 10.351937294006348, + "learning_rate": 9.995839999584001e-06, + "loss": 0.9703, + "step": 4163 + }, + { + "epoch": 1.9358437935843793, + "grad_norm": 5.643045902252197, + "learning_rate": 9.995838999583902e-06, + "loss": 0.5378, + "step": 4164 + }, + { + "epoch": 1.9363086936308693, + "grad_norm": 5.663859844207764, + "learning_rate": 9.995837999583801e-06, + "loss": 0.4696, + "step": 4165 + }, + { + "epoch": 1.9367735936773594, + "grad_norm": 4.9103546142578125, + "learning_rate": 9.995836999583702e-06, + "loss": 0.4041, + "step": 4166 + }, + { + "epoch": 1.9372384937238494, + "grad_norm": 8.675719261169434, + "learning_rate": 9.9958359995836e-06, + "loss": 0.9532, + "step": 4167 + }, + { + "epoch": 1.9377033937703394, + "grad_norm": 6.7207770347595215, + "learning_rate": 9.9958349995835e-06, + "loss": 0.5182, + "step": 4168 + }, + { + "epoch": 1.9381682938168294, + "grad_norm": 5.9141974449157715, + "learning_rate": 9.9958339995834e-06, + "loss": 0.4289, + "step": 4169 + }, + { + "epoch": 1.9386331938633194, + "grad_norm": 3.3510537147521973, + "learning_rate": 9.9958329995833e-06, + "loss": 0.2145, + "step": 4170 + }, + { + "epoch": 1.9390980939098093, + "grad_norm": 5.943368911743164, + "learning_rate": 9.995831999583201e-06, + "loss": 0.6788, + "step": 4171 + }, + { + "epoch": 1.9395629939562995, + "grad_norm": 11.375713348388672, + "learning_rate": 9.9958309995831e-06, + "loss": 1.1614, + "step": 4172 + }, + { + "epoch": 1.9400278940027893, + "grad_norm": 6.209079265594482, + "learning_rate": 9.995829999583001e-06, + "loss": 0.4701, + "step": 4173 + }, + { + "epoch": 1.9404927940492795, + "grad_norm": 4.814525127410889, + "learning_rate": 9.9958289995829e-06, + "loss": 0.4035, + "step": 4174 + }, + { + "epoch": 1.9409576940957693, + "grad_norm": 6.084917068481445, + "learning_rate": 9.9958279995828e-06, + "loss": 0.4663, + "step": 4175 + }, + { + "epoch": 1.9414225941422594, + "grad_norm": 4.047932147979736, + "learning_rate": 9.995826999582701e-06, + "loss": 0.323, + "step": 4176 + }, + { + "epoch": 1.9418874941887494, + "grad_norm": 4.937535762786865, + "learning_rate": 9.995825999582602e-06, + "loss": 0.5037, + "step": 4177 + }, + { + "epoch": 1.9423523942352394, + "grad_norm": 4.473650932312012, + "learning_rate": 9.995824999582501e-06, + "loss": 0.3937, + "step": 4178 + }, + { + "epoch": 1.9428172942817294, + "grad_norm": 3.5388705730438232, + "learning_rate": 9.9958239995824e-06, + "loss": 0.166, + "step": 4179 + }, + { + "epoch": 1.9432821943282195, + "grad_norm": 3.9624643325805664, + "learning_rate": 9.9958229995823e-06, + "loss": 0.3022, + "step": 4180 + }, + { + "epoch": 1.9437470943747095, + "grad_norm": 5.795173645019531, + "learning_rate": 9.9958219995822e-06, + "loss": 0.7741, + "step": 4181 + }, + { + "epoch": 1.9442119944211993, + "grad_norm": 5.622011661529541, + "learning_rate": 9.9958209995821e-06, + "loss": 0.7087, + "step": 4182 + }, + { + "epoch": 1.9446768944676895, + "grad_norm": 6.795257091522217, + "learning_rate": 9.995819999582001e-06, + "loss": 0.4869, + "step": 4183 + }, + { + "epoch": 1.9451417945141793, + "grad_norm": 5.513619422912598, + "learning_rate": 9.995818999581902e-06, + "loss": 0.4812, + "step": 4184 + }, + { + "epoch": 1.9456066945606696, + "grad_norm": 5.005978107452393, + "learning_rate": 9.9958179995818e-06, + "loss": 0.4083, + "step": 4185 + }, + { + "epoch": 1.9460715946071594, + "grad_norm": 6.985101699829102, + "learning_rate": 9.995816999581701e-06, + "loss": 0.7904, + "step": 4186 + }, + { + "epoch": 1.9465364946536494, + "grad_norm": 4.406337261199951, + "learning_rate": 9.9958159995816e-06, + "loss": 0.4346, + "step": 4187 + }, + { + "epoch": 1.9470013947001394, + "grad_norm": 6.030765056610107, + "learning_rate": 9.9958149995815e-06, + "loss": 0.471, + "step": 4188 + }, + { + "epoch": 1.9474662947466295, + "grad_norm": 5.919139862060547, + "learning_rate": 9.995813999581401e-06, + "loss": 0.5462, + "step": 4189 + }, + { + "epoch": 1.9479311947931195, + "grad_norm": 6.589946269989014, + "learning_rate": 9.9958129995813e-06, + "loss": 0.4579, + "step": 4190 + }, + { + "epoch": 1.9483960948396095, + "grad_norm": 5.421365737915039, + "learning_rate": 9.995811999581201e-06, + "loss": 0.457, + "step": 4191 + }, + { + "epoch": 1.9488609948860995, + "grad_norm": 4.4790730476379395, + "learning_rate": 9.9958109995811e-06, + "loss": 0.5155, + "step": 4192 + }, + { + "epoch": 1.9493258949325893, + "grad_norm": 6.232719898223877, + "learning_rate": 9.995809999581e-06, + "loss": 0.7728, + "step": 4193 + }, + { + "epoch": 1.9497907949790796, + "grad_norm": 3.571298360824585, + "learning_rate": 9.9958089995809e-06, + "loss": 0.2581, + "step": 4194 + }, + { + "epoch": 1.9502556950255694, + "grad_norm": 5.401448726654053, + "learning_rate": 9.9958079995808e-06, + "loss": 0.4108, + "step": 4195 + }, + { + "epoch": 1.9507205950720596, + "grad_norm": 4.933201313018799, + "learning_rate": 9.995806999580701e-06, + "loss": 0.4934, + "step": 4196 + }, + { + "epoch": 1.9511854951185494, + "grad_norm": 4.67263126373291, + "learning_rate": 9.995805999580602e-06, + "loss": 0.4792, + "step": 4197 + }, + { + "epoch": 1.9516503951650395, + "grad_norm": 5.247239589691162, + "learning_rate": 9.9958049995805e-06, + "loss": 0.4515, + "step": 4198 + }, + { + "epoch": 1.9521152952115295, + "grad_norm": 5.4264068603515625, + "learning_rate": 9.995803999580401e-06, + "loss": 0.365, + "step": 4199 + }, + { + "epoch": 1.9525801952580195, + "grad_norm": 5.219094276428223, + "learning_rate": 9.9958029995803e-06, + "loss": 0.4531, + "step": 4200 + }, + { + "epoch": 1.9530450953045095, + "grad_norm": 5.7197699546813965, + "learning_rate": 9.995801999580201e-06, + "loss": 0.5849, + "step": 4201 + }, + { + "epoch": 1.9535099953509996, + "grad_norm": 9.379204750061035, + "learning_rate": 9.995800999580102e-06, + "loss": 0.6605, + "step": 4202 + }, + { + "epoch": 1.9539748953974896, + "grad_norm": 4.853122711181641, + "learning_rate": 9.99579999958e-06, + "loss": 0.2753, + "step": 4203 + }, + { + "epoch": 1.9544397954439794, + "grad_norm": 6.614706039428711, + "learning_rate": 9.995798999579901e-06, + "loss": 0.6355, + "step": 4204 + }, + { + "epoch": 1.9549046954904696, + "grad_norm": 6.497218132019043, + "learning_rate": 9.9957979995798e-06, + "loss": 0.4692, + "step": 4205 + }, + { + "epoch": 1.9553695955369594, + "grad_norm": 6.508086681365967, + "learning_rate": 9.9957969995797e-06, + "loss": 0.8176, + "step": 4206 + }, + { + "epoch": 1.9558344955834497, + "grad_norm": 4.450139045715332, + "learning_rate": 9.9957959995796e-06, + "loss": 0.4738, + "step": 4207 + }, + { + "epoch": 1.9562993956299395, + "grad_norm": 3.7311298847198486, + "learning_rate": 9.9957949995795e-06, + "loss": 0.3896, + "step": 4208 + }, + { + "epoch": 1.9567642956764297, + "grad_norm": 8.769356727600098, + "learning_rate": 9.995793999579401e-06, + "loss": 0.9035, + "step": 4209 + }, + { + "epoch": 1.9572291957229195, + "grad_norm": 5.985103130340576, + "learning_rate": 9.995792999579302e-06, + "loss": 0.4551, + "step": 4210 + }, + { + "epoch": 1.9576940957694096, + "grad_norm": 9.419586181640625, + "learning_rate": 9.9957919995792e-06, + "loss": 0.8039, + "step": 4211 + }, + { + "epoch": 1.9581589958158996, + "grad_norm": 6.310050964355469, + "learning_rate": 9.9957909995791e-06, + "loss": 0.5234, + "step": 4212 + }, + { + "epoch": 1.9586238958623896, + "grad_norm": 6.70223331451416, + "learning_rate": 9.995789999579e-06, + "loss": 0.8327, + "step": 4213 + }, + { + "epoch": 1.9590887959088796, + "grad_norm": 10.596755027770996, + "learning_rate": 9.995788999578901e-06, + "loss": 1.0019, + "step": 4214 + }, + { + "epoch": 1.9595536959553694, + "grad_norm": 6.705220699310303, + "learning_rate": 9.995787999578802e-06, + "loss": 0.4287, + "step": 4215 + }, + { + "epoch": 1.9600185960018597, + "grad_norm": 4.174105167388916, + "learning_rate": 9.9957869995787e-06, + "loss": 0.4439, + "step": 4216 + }, + { + "epoch": 1.9604834960483495, + "grad_norm": 4.658987522125244, + "learning_rate": 9.995785999578601e-06, + "loss": 0.3618, + "step": 4217 + }, + { + "epoch": 1.9609483960948397, + "grad_norm": 7.921549320220947, + "learning_rate": 9.9957849995785e-06, + "loss": 0.8006, + "step": 4218 + }, + { + "epoch": 1.9614132961413295, + "grad_norm": 6.4664764404296875, + "learning_rate": 9.9957839995784e-06, + "loss": 0.551, + "step": 4219 + }, + { + "epoch": 1.9618781961878198, + "grad_norm": 8.585976600646973, + "learning_rate": 9.995782999578302e-06, + "loss": 0.7118, + "step": 4220 + }, + { + "epoch": 1.9623430962343096, + "grad_norm": 5.620660781860352, + "learning_rate": 9.9957819995782e-06, + "loss": 0.5285, + "step": 4221 + }, + { + "epoch": 1.9628079962807996, + "grad_norm": 5.082192420959473, + "learning_rate": 9.995780999578101e-06, + "loss": 0.4962, + "step": 4222 + }, + { + "epoch": 1.9632728963272896, + "grad_norm": 6.153079032897949, + "learning_rate": 9.995779999578e-06, + "loss": 0.7881, + "step": 4223 + }, + { + "epoch": 1.9637377963737797, + "grad_norm": 7.308237075805664, + "learning_rate": 9.9957789995779e-06, + "loss": 0.6725, + "step": 4224 + }, + { + "epoch": 1.9642026964202697, + "grad_norm": 4.84792947769165, + "learning_rate": 9.9957779995778e-06, + "loss": 0.6492, + "step": 4225 + }, + { + "epoch": 1.9646675964667595, + "grad_norm": 5.779921531677246, + "learning_rate": 9.9957769995777e-06, + "loss": 0.7588, + "step": 4226 + }, + { + "epoch": 1.9651324965132497, + "grad_norm": 5.233814239501953, + "learning_rate": 9.995775999577601e-06, + "loss": 0.4871, + "step": 4227 + }, + { + "epoch": 1.9655973965597395, + "grad_norm": 10.40495777130127, + "learning_rate": 9.995774999577502e-06, + "loss": 0.95, + "step": 4228 + }, + { + "epoch": 1.9660622966062298, + "grad_norm": 6.7852349281311035, + "learning_rate": 9.9957739995774e-06, + "loss": 0.6179, + "step": 4229 + }, + { + "epoch": 1.9665271966527196, + "grad_norm": 3.5919835567474365, + "learning_rate": 9.995772999577301e-06, + "loss": 0.4307, + "step": 4230 + }, + { + "epoch": 1.9669920966992098, + "grad_norm": 5.207309246063232, + "learning_rate": 9.9957719995772e-06, + "loss": 0.4393, + "step": 4231 + }, + { + "epoch": 1.9674569967456996, + "grad_norm": 5.5177226066589355, + "learning_rate": 9.995770999577101e-06, + "loss": 0.2533, + "step": 4232 + }, + { + "epoch": 1.9679218967921897, + "grad_norm": 7.673786640167236, + "learning_rate": 9.995769999577002e-06, + "loss": 0.7531, + "step": 4233 + }, + { + "epoch": 1.9683867968386797, + "grad_norm": 6.278751850128174, + "learning_rate": 9.9957689995769e-06, + "loss": 0.4826, + "step": 4234 + }, + { + "epoch": 1.9688516968851697, + "grad_norm": 4.624988079071045, + "learning_rate": 9.995767999576801e-06, + "loss": 0.5166, + "step": 4235 + }, + { + "epoch": 1.9693165969316597, + "grad_norm": 6.578075885772705, + "learning_rate": 9.9957669995767e-06, + "loss": 0.4421, + "step": 4236 + }, + { + "epoch": 1.9697814969781497, + "grad_norm": 5.329639911651611, + "learning_rate": 9.9957659995766e-06, + "loss": 0.5349, + "step": 4237 + }, + { + "epoch": 1.9702463970246398, + "grad_norm": 5.766910552978516, + "learning_rate": 9.9957649995765e-06, + "loss": 0.5658, + "step": 4238 + }, + { + "epoch": 1.9707112970711296, + "grad_norm": 6.640281677246094, + "learning_rate": 9.9957639995764e-06, + "loss": 0.7269, + "step": 4239 + }, + { + "epoch": 1.9711761971176198, + "grad_norm": 4.376102924346924, + "learning_rate": 9.995762999576301e-06, + "loss": 0.4537, + "step": 4240 + }, + { + "epoch": 1.9716410971641096, + "grad_norm": 5.389187812805176, + "learning_rate": 9.995761999576202e-06, + "loss": 0.4758, + "step": 4241 + }, + { + "epoch": 1.9721059972105999, + "grad_norm": 4.992763042449951, + "learning_rate": 9.9957609995761e-06, + "loss": 0.4545, + "step": 4242 + }, + { + "epoch": 1.9725708972570897, + "grad_norm": 5.259459972381592, + "learning_rate": 9.995759999576001e-06, + "loss": 0.4069, + "step": 4243 + }, + { + "epoch": 1.9730357973035797, + "grad_norm": 5.693251609802246, + "learning_rate": 9.9957589995759e-06, + "loss": 0.3635, + "step": 4244 + }, + { + "epoch": 1.9735006973500697, + "grad_norm": 6.433508396148682, + "learning_rate": 9.995757999575801e-06, + "loss": 0.6269, + "step": 4245 + }, + { + "epoch": 1.9739655973965597, + "grad_norm": 3.4030044078826904, + "learning_rate": 9.995756999575702e-06, + "loss": 0.1582, + "step": 4246 + }, + { + "epoch": 1.9744304974430498, + "grad_norm": 6.4811482429504395, + "learning_rate": 9.9957559995756e-06, + "loss": 0.7532, + "step": 4247 + }, + { + "epoch": 1.9748953974895398, + "grad_norm": 6.559160232543945, + "learning_rate": 9.995754999575501e-06, + "loss": 0.8576, + "step": 4248 + }, + { + "epoch": 1.9753602975360298, + "grad_norm": 5.444293975830078, + "learning_rate": 9.9957539995754e-06, + "loss": 0.513, + "step": 4249 + }, + { + "epoch": 1.9758251975825196, + "grad_norm": 8.633881568908691, + "learning_rate": 9.995752999575301e-06, + "loss": 0.7231, + "step": 4250 + }, + { + "epoch": 1.9762900976290099, + "grad_norm": 3.6789393424987793, + "learning_rate": 9.9957519995752e-06, + "loss": 0.4081, + "step": 4251 + }, + { + "epoch": 1.9767549976754997, + "grad_norm": 7.424015998840332, + "learning_rate": 9.9957509995751e-06, + "loss": 0.8487, + "step": 4252 + }, + { + "epoch": 1.97721989772199, + "grad_norm": 9.98383903503418, + "learning_rate": 9.995749999575001e-06, + "loss": 0.6257, + "step": 4253 + }, + { + "epoch": 1.9776847977684797, + "grad_norm": 6.747807502746582, + "learning_rate": 9.995748999574902e-06, + "loss": 0.5983, + "step": 4254 + }, + { + "epoch": 1.9781496978149697, + "grad_norm": 5.487414836883545, + "learning_rate": 9.9957479995748e-06, + "loss": 0.6964, + "step": 4255 + }, + { + "epoch": 1.9786145978614598, + "grad_norm": 6.833139896392822, + "learning_rate": 9.9957469995747e-06, + "loss": 0.5656, + "step": 4256 + }, + { + "epoch": 1.9790794979079498, + "grad_norm": 5.339278697967529, + "learning_rate": 9.9957459995746e-06, + "loss": 0.5314, + "step": 4257 + }, + { + "epoch": 1.9795443979544398, + "grad_norm": 7.675800323486328, + "learning_rate": 9.995744999574501e-06, + "loss": 0.5499, + "step": 4258 + }, + { + "epoch": 1.9800092980009298, + "grad_norm": 7.239145278930664, + "learning_rate": 9.995743999574402e-06, + "loss": 0.7905, + "step": 4259 + }, + { + "epoch": 1.9804741980474199, + "grad_norm": 5.7163214683532715, + "learning_rate": 9.9957429995743e-06, + "loss": 0.4377, + "step": 4260 + }, + { + "epoch": 1.9809390980939097, + "grad_norm": 3.5204238891601562, + "learning_rate": 9.995741999574201e-06, + "loss": 0.4172, + "step": 4261 + }, + { + "epoch": 1.9814039981404, + "grad_norm": 5.834833145141602, + "learning_rate": 9.9957409995741e-06, + "loss": 0.4245, + "step": 4262 + }, + { + "epoch": 1.9818688981868897, + "grad_norm": 6.800993919372559, + "learning_rate": 9.995739999574001e-06, + "loss": 0.6847, + "step": 4263 + }, + { + "epoch": 1.98233379823338, + "grad_norm": 4.318042278289795, + "learning_rate": 9.9957389995739e-06, + "loss": 0.3791, + "step": 4264 + }, + { + "epoch": 1.9827986982798698, + "grad_norm": 5.688649654388428, + "learning_rate": 9.9957379995738e-06, + "loss": 0.4952, + "step": 4265 + }, + { + "epoch": 1.9832635983263598, + "grad_norm": 5.077238082885742, + "learning_rate": 9.995736999573701e-06, + "loss": 0.4108, + "step": 4266 + }, + { + "epoch": 1.9837284983728498, + "grad_norm": 7.060328483581543, + "learning_rate": 9.9957359995736e-06, + "loss": 0.8126, + "step": 4267 + }, + { + "epoch": 1.9841933984193398, + "grad_norm": 10.465643882751465, + "learning_rate": 9.995734999573501e-06, + "loss": 1.0573, + "step": 4268 + }, + { + "epoch": 1.9846582984658299, + "grad_norm": 4.84328556060791, + "learning_rate": 9.9957339995734e-06, + "loss": 0.4594, + "step": 4269 + }, + { + "epoch": 1.9851231985123199, + "grad_norm": 4.195311069488525, + "learning_rate": 9.9957329995733e-06, + "loss": 0.2407, + "step": 4270 + }, + { + "epoch": 1.98558809855881, + "grad_norm": 2.9313387870788574, + "learning_rate": 9.995731999573201e-06, + "loss": 0.2777, + "step": 4271 + }, + { + "epoch": 1.9860529986052997, + "grad_norm": 5.086322784423828, + "learning_rate": 9.995730999573102e-06, + "loss": 0.3977, + "step": 4272 + }, + { + "epoch": 1.98651789865179, + "grad_norm": 7.2705488204956055, + "learning_rate": 9.995729999573e-06, + "loss": 0.7332, + "step": 4273 + }, + { + "epoch": 1.9869827986982798, + "grad_norm": 5.631176471710205, + "learning_rate": 9.995728999572901e-06, + "loss": 0.421, + "step": 4274 + }, + { + "epoch": 1.98744769874477, + "grad_norm": 5.621024131774902, + "learning_rate": 9.9957279995728e-06, + "loss": 0.5131, + "step": 4275 + }, + { + "epoch": 1.9879125987912598, + "grad_norm": 6.499343395233154, + "learning_rate": 9.995726999572701e-06, + "loss": 0.7395, + "step": 4276 + }, + { + "epoch": 1.9883774988377498, + "grad_norm": 6.786025047302246, + "learning_rate": 9.9957259995726e-06, + "loss": 0.8303, + "step": 4277 + }, + { + "epoch": 1.9888423988842399, + "grad_norm": 5.976122856140137, + "learning_rate": 9.9957249995725e-06, + "loss": 0.6786, + "step": 4278 + }, + { + "epoch": 1.9893072989307299, + "grad_norm": 5.223127365112305, + "learning_rate": 9.995723999572401e-06, + "loss": 0.4623, + "step": 4279 + }, + { + "epoch": 1.98977219897722, + "grad_norm": 6.233969688415527, + "learning_rate": 9.9957229995723e-06, + "loss": 0.6986, + "step": 4280 + }, + { + "epoch": 1.99023709902371, + "grad_norm": 5.626959323883057, + "learning_rate": 9.995721999572201e-06, + "loss": 0.5615, + "step": 4281 + }, + { + "epoch": 1.9907019990702, + "grad_norm": 9.472993850708008, + "learning_rate": 9.9957209995721e-06, + "loss": 0.6057, + "step": 4282 + }, + { + "epoch": 1.9911668991166898, + "grad_norm": 6.535130977630615, + "learning_rate": 9.995719999572e-06, + "loss": 0.6118, + "step": 4283 + }, + { + "epoch": 1.99163179916318, + "grad_norm": 4.35960578918457, + "learning_rate": 9.995718999571901e-06, + "loss": 0.455, + "step": 4284 + }, + { + "epoch": 1.9920966992096698, + "grad_norm": 6.728945255279541, + "learning_rate": 9.995717999571802e-06, + "loss": 0.6399, + "step": 4285 + }, + { + "epoch": 1.99256159925616, + "grad_norm": 4.323879241943359, + "learning_rate": 9.995716999571701e-06, + "loss": 0.2799, + "step": 4286 + }, + { + "epoch": 1.9930264993026499, + "grad_norm": 8.99842357635498, + "learning_rate": 9.995715999571602e-06, + "loss": 0.8408, + "step": 4287 + }, + { + "epoch": 1.9934913993491399, + "grad_norm": 8.349610328674316, + "learning_rate": 9.9957149995715e-06, + "loss": 0.6574, + "step": 4288 + }, + { + "epoch": 1.99395629939563, + "grad_norm": 5.802945613861084, + "learning_rate": 9.995713999571401e-06, + "loss": 0.6078, + "step": 4289 + }, + { + "epoch": 1.99442119944212, + "grad_norm": 5.664584636688232, + "learning_rate": 9.9957129995713e-06, + "loss": 0.3794, + "step": 4290 + }, + { + "epoch": 1.99488609948861, + "grad_norm": 6.528872013092041, + "learning_rate": 9.9957119995712e-06, + "loss": 0.3018, + "step": 4291 + }, + { + "epoch": 1.9953509995351, + "grad_norm": 7.520498752593994, + "learning_rate": 9.995710999571101e-06, + "loss": 0.6026, + "step": 4292 + }, + { + "epoch": 1.99581589958159, + "grad_norm": 4.556714057922363, + "learning_rate": 9.995709999571e-06, + "loss": 0.2849, + "step": 4293 + }, + { + "epoch": 1.9962807996280798, + "grad_norm": 5.41218376159668, + "learning_rate": 9.995708999570901e-06, + "loss": 0.4244, + "step": 4294 + }, + { + "epoch": 1.99674569967457, + "grad_norm": 7.9949822425842285, + "learning_rate": 9.9957079995708e-06, + "loss": 0.9211, + "step": 4295 + }, + { + "epoch": 1.9972105997210599, + "grad_norm": 5.522622108459473, + "learning_rate": 9.9957069995707e-06, + "loss": 0.4014, + "step": 4296 + }, + { + "epoch": 1.99767549976755, + "grad_norm": 5.126392841339111, + "learning_rate": 9.995705999570601e-06, + "loss": 0.5204, + "step": 4297 + }, + { + "epoch": 1.99814039981404, + "grad_norm": 5.620323657989502, + "learning_rate": 9.9957049995705e-06, + "loss": 0.3668, + "step": 4298 + }, + { + "epoch": 1.9986052998605301, + "grad_norm": 8.034504890441895, + "learning_rate": 9.995703999570401e-06, + "loss": 0.7268, + "step": 4299 + }, + { + "epoch": 1.99907019990702, + "grad_norm": 7.373629093170166, + "learning_rate": 9.9957029995703e-06, + "loss": 0.7507, + "step": 4300 + }, + { + "epoch": 1.99953509995351, + "grad_norm": 10.420496940612793, + "learning_rate": 9.9957019995702e-06, + "loss": 0.9885, + "step": 4301 + }, + { + "epoch": 2.0, + "grad_norm": 8.71774673461914, + "learning_rate": 9.995700999570101e-06, + "loss": 0.6652, + "step": 4302 + }, + { + "epoch": 2.00046490004649, + "grad_norm": 6.100430965423584, + "learning_rate": 9.99569999957e-06, + "loss": 0.6398, + "step": 4303 + }, + { + "epoch": 2.00092980009298, + "grad_norm": 3.669694662094116, + "learning_rate": 9.995698999569901e-06, + "loss": 0.1928, + "step": 4304 + }, + { + "epoch": 2.00139470013947, + "grad_norm": 5.094183444976807, + "learning_rate": 9.995697999569802e-06, + "loss": 0.6142, + "step": 4305 + }, + { + "epoch": 2.00185960018596, + "grad_norm": 6.5204548835754395, + "learning_rate": 9.9956969995697e-06, + "loss": 0.9872, + "step": 4306 + }, + { + "epoch": 2.00232450023245, + "grad_norm": 5.095763206481934, + "learning_rate": 9.995695999569601e-06, + "loss": 0.7973, + "step": 4307 + }, + { + "epoch": 2.00278940027894, + "grad_norm": 4.687575817108154, + "learning_rate": 9.9956949995695e-06, + "loss": 0.4602, + "step": 4308 + }, + { + "epoch": 2.00325430032543, + "grad_norm": 4.110746383666992, + "learning_rate": 9.9956939995694e-06, + "loss": 0.3847, + "step": 4309 + }, + { + "epoch": 2.00371920037192, + "grad_norm": 5.662601470947266, + "learning_rate": 9.995692999569301e-06, + "loss": 0.5105, + "step": 4310 + }, + { + "epoch": 2.00418410041841, + "grad_norm": 4.382277011871338, + "learning_rate": 9.9956919995692e-06, + "loss": 0.4055, + "step": 4311 + }, + { + "epoch": 2.0046490004649002, + "grad_norm": 4.883382797241211, + "learning_rate": 9.995690999569101e-06, + "loss": 0.4811, + "step": 4312 + }, + { + "epoch": 2.00511390051139, + "grad_norm": 8.61436653137207, + "learning_rate": 9.995689999569e-06, + "loss": 0.2916, + "step": 4313 + }, + { + "epoch": 2.00557880055788, + "grad_norm": 5.3906121253967285, + "learning_rate": 9.9956889995689e-06, + "loss": 0.3767, + "step": 4314 + }, + { + "epoch": 2.00604370060437, + "grad_norm": 4.001357078552246, + "learning_rate": 9.995687999568801e-06, + "loss": 0.3049, + "step": 4315 + }, + { + "epoch": 2.00650860065086, + "grad_norm": 5.316244125366211, + "learning_rate": 9.995686999568702e-06, + "loss": 0.588, + "step": 4316 + }, + { + "epoch": 2.00697350069735, + "grad_norm": 3.558897018432617, + "learning_rate": 9.995685999568601e-06, + "loss": 0.169, + "step": 4317 + }, + { + "epoch": 2.00743840074384, + "grad_norm": 4.841050624847412, + "learning_rate": 9.995684999568502e-06, + "loss": 0.4006, + "step": 4318 + }, + { + "epoch": 2.00790330079033, + "grad_norm": 4.719240188598633, + "learning_rate": 9.9956839995684e-06, + "loss": 0.6871, + "step": 4319 + }, + { + "epoch": 2.00836820083682, + "grad_norm": 4.9895920753479, + "learning_rate": 9.9956829995683e-06, + "loss": 0.4639, + "step": 4320 + }, + { + "epoch": 2.0088331008833102, + "grad_norm": 4.166966438293457, + "learning_rate": 9.9956819995682e-06, + "loss": 0.2126, + "step": 4321 + }, + { + "epoch": 2.0092980009298, + "grad_norm": 7.166627407073975, + "learning_rate": 9.9956809995681e-06, + "loss": 0.8353, + "step": 4322 + }, + { + "epoch": 2.0097629009762903, + "grad_norm": 4.0901384353637695, + "learning_rate": 9.995679999568001e-06, + "loss": 0.1839, + "step": 4323 + }, + { + "epoch": 2.01022780102278, + "grad_norm": 5.781834125518799, + "learning_rate": 9.9956789995679e-06, + "loss": 0.3939, + "step": 4324 + }, + { + "epoch": 2.01069270106927, + "grad_norm": 5.563934803009033, + "learning_rate": 9.995677999567801e-06, + "loss": 0.4332, + "step": 4325 + }, + { + "epoch": 2.01115760111576, + "grad_norm": 8.521034240722656, + "learning_rate": 9.9956769995677e-06, + "loss": 0.5522, + "step": 4326 + }, + { + "epoch": 2.01162250116225, + "grad_norm": 3.7162959575653076, + "learning_rate": 9.9956759995676e-06, + "loss": 0.3473, + "step": 4327 + }, + { + "epoch": 2.01208740120874, + "grad_norm": 4.250004768371582, + "learning_rate": 9.995674999567501e-06, + "loss": 0.2958, + "step": 4328 + }, + { + "epoch": 2.01255230125523, + "grad_norm": 4.312964916229248, + "learning_rate": 9.995673999567402e-06, + "loss": 0.34, + "step": 4329 + }, + { + "epoch": 2.0130172013017202, + "grad_norm": 5.479302883148193, + "learning_rate": 9.995672999567301e-06, + "loss": 0.4668, + "step": 4330 + }, + { + "epoch": 2.01348210134821, + "grad_norm": 6.8368730545043945, + "learning_rate": 9.9956719995672e-06, + "loss": 0.6632, + "step": 4331 + }, + { + "epoch": 2.0139470013947003, + "grad_norm": 4.686013698577881, + "learning_rate": 9.9956709995671e-06, + "loss": 0.2484, + "step": 4332 + }, + { + "epoch": 2.01441190144119, + "grad_norm": 4.833444595336914, + "learning_rate": 9.995669999567e-06, + "loss": 0.5424, + "step": 4333 + }, + { + "epoch": 2.0148768014876803, + "grad_norm": 10.654145240783691, + "learning_rate": 9.9956689995669e-06, + "loss": 1.5767, + "step": 4334 + }, + { + "epoch": 2.01534170153417, + "grad_norm": 3.584796905517578, + "learning_rate": 9.995667999566801e-06, + "loss": 0.3265, + "step": 4335 + }, + { + "epoch": 2.01580660158066, + "grad_norm": 3.153062343597412, + "learning_rate": 9.995666999566702e-06, + "loss": 0.2421, + "step": 4336 + }, + { + "epoch": 2.01627150162715, + "grad_norm": 3.2670884132385254, + "learning_rate": 9.9956659995666e-06, + "loss": 0.2478, + "step": 4337 + }, + { + "epoch": 2.01673640167364, + "grad_norm": 7.517485618591309, + "learning_rate": 9.995664999566501e-06, + "loss": 0.736, + "step": 4338 + }, + { + "epoch": 2.0172013017201302, + "grad_norm": 6.70169734954834, + "learning_rate": 9.9956639995664e-06, + "loss": 0.6389, + "step": 4339 + }, + { + "epoch": 2.01766620176662, + "grad_norm": 3.594085216522217, + "learning_rate": 9.9956629995663e-06, + "loss": 0.2783, + "step": 4340 + }, + { + "epoch": 2.0181311018131103, + "grad_norm": 7.173691272735596, + "learning_rate": 9.995661999566201e-06, + "loss": 0.782, + "step": 4341 + }, + { + "epoch": 2.0185960018596, + "grad_norm": 4.549412727355957, + "learning_rate": 9.9956609995661e-06, + "loss": 0.2729, + "step": 4342 + }, + { + "epoch": 2.0190609019060903, + "grad_norm": 6.154842853546143, + "learning_rate": 9.995659999566001e-06, + "loss": 0.791, + "step": 4343 + }, + { + "epoch": 2.01952580195258, + "grad_norm": 5.854216575622559, + "learning_rate": 9.9956589995659e-06, + "loss": 0.4804, + "step": 4344 + }, + { + "epoch": 2.0199907019990704, + "grad_norm": 8.278240203857422, + "learning_rate": 9.9956579995658e-06, + "loss": 0.9967, + "step": 4345 + }, + { + "epoch": 2.02045560204556, + "grad_norm": 5.565300464630127, + "learning_rate": 9.9956569995657e-06, + "loss": 0.488, + "step": 4346 + }, + { + "epoch": 2.02092050209205, + "grad_norm": 8.066060066223145, + "learning_rate": 9.9956559995656e-06, + "loss": 0.7996, + "step": 4347 + }, + { + "epoch": 2.0213854021385402, + "grad_norm": 5.428238391876221, + "learning_rate": 9.995654999565501e-06, + "loss": 0.4858, + "step": 4348 + }, + { + "epoch": 2.02185030218503, + "grad_norm": 5.472867012023926, + "learning_rate": 9.995653999565402e-06, + "loss": 0.6256, + "step": 4349 + }, + { + "epoch": 2.0223152022315203, + "grad_norm": 6.797527313232422, + "learning_rate": 9.9956529995653e-06, + "loss": 0.7777, + "step": 4350 + }, + { + "epoch": 2.02278010227801, + "grad_norm": 5.655513763427734, + "learning_rate": 9.995651999565201e-06, + "loss": 0.5042, + "step": 4351 + }, + { + "epoch": 2.0232450023245003, + "grad_norm": 3.598375082015991, + "learning_rate": 9.9956509995651e-06, + "loss": 0.2264, + "step": 4352 + }, + { + "epoch": 2.02370990237099, + "grad_norm": 4.627654075622559, + "learning_rate": 9.995649999565001e-06, + "loss": 0.4531, + "step": 4353 + }, + { + "epoch": 2.0241748024174804, + "grad_norm": 2.658728837966919, + "learning_rate": 9.995648999564902e-06, + "loss": 0.1808, + "step": 4354 + }, + { + "epoch": 2.02463970246397, + "grad_norm": 8.505918502807617, + "learning_rate": 9.9956479995648e-06, + "loss": 0.5671, + "step": 4355 + }, + { + "epoch": 2.0251046025104604, + "grad_norm": 4.66569709777832, + "learning_rate": 9.995646999564701e-06, + "loss": 0.4244, + "step": 4356 + }, + { + "epoch": 2.0255695025569502, + "grad_norm": 4.697343826293945, + "learning_rate": 9.9956459995646e-06, + "loss": 0.3827, + "step": 4357 + }, + { + "epoch": 2.0260344026034405, + "grad_norm": 4.088134765625, + "learning_rate": 9.9956449995645e-06, + "loss": 0.2303, + "step": 4358 + }, + { + "epoch": 2.0264993026499303, + "grad_norm": 5.277453899383545, + "learning_rate": 9.9956439995644e-06, + "loss": 0.3786, + "step": 4359 + }, + { + "epoch": 2.02696420269642, + "grad_norm": 7.889645576477051, + "learning_rate": 9.9956429995643e-06, + "loss": 0.4387, + "step": 4360 + }, + { + "epoch": 2.0274291027429103, + "grad_norm": 4.434329032897949, + "learning_rate": 9.995641999564201e-06, + "loss": 0.3333, + "step": 4361 + }, + { + "epoch": 2.0278940027894, + "grad_norm": 5.011424541473389, + "learning_rate": 9.995640999564102e-06, + "loss": 0.4775, + "step": 4362 + }, + { + "epoch": 2.0283589028358904, + "grad_norm": 6.2434916496276855, + "learning_rate": 9.995639999564e-06, + "loss": 0.5956, + "step": 4363 + }, + { + "epoch": 2.02882380288238, + "grad_norm": 6.0461812019348145, + "learning_rate": 9.9956389995639e-06, + "loss": 0.226, + "step": 4364 + }, + { + "epoch": 2.0292887029288704, + "grad_norm": 6.212432384490967, + "learning_rate": 9.9956379995638e-06, + "loss": 0.7055, + "step": 4365 + }, + { + "epoch": 2.0297536029753602, + "grad_norm": 6.034689903259277, + "learning_rate": 9.995636999563701e-06, + "loss": 0.5628, + "step": 4366 + }, + { + "epoch": 2.0302185030218505, + "grad_norm": 3.99892258644104, + "learning_rate": 9.995635999563602e-06, + "loss": 0.2848, + "step": 4367 + }, + { + "epoch": 2.0306834030683403, + "grad_norm": 5.231868267059326, + "learning_rate": 9.9956349995635e-06, + "loss": 0.418, + "step": 4368 + }, + { + "epoch": 2.0311483031148305, + "grad_norm": 5.768941879272461, + "learning_rate": 9.995633999563401e-06, + "loss": 0.5575, + "step": 4369 + }, + { + "epoch": 2.0316132031613203, + "grad_norm": 5.288180828094482, + "learning_rate": 9.9956329995633e-06, + "loss": 0.4638, + "step": 4370 + }, + { + "epoch": 2.03207810320781, + "grad_norm": 5.316526889801025, + "learning_rate": 9.995631999563201e-06, + "loss": 0.4287, + "step": 4371 + }, + { + "epoch": 2.0325430032543004, + "grad_norm": 4.45741081237793, + "learning_rate": 9.9956309995631e-06, + "loss": 0.3299, + "step": 4372 + }, + { + "epoch": 2.03300790330079, + "grad_norm": 5.014162540435791, + "learning_rate": 9.995629999563e-06, + "loss": 0.4482, + "step": 4373 + }, + { + "epoch": 2.0334728033472804, + "grad_norm": 5.8032636642456055, + "learning_rate": 9.995628999562901e-06, + "loss": 0.5919, + "step": 4374 + }, + { + "epoch": 2.0339377033937702, + "grad_norm": 3.7330541610717773, + "learning_rate": 9.9956279995628e-06, + "loss": 0.4319, + "step": 4375 + }, + { + "epoch": 2.0344026034402605, + "grad_norm": 2.9090425968170166, + "learning_rate": 9.9956269995627e-06, + "loss": 0.2183, + "step": 4376 + }, + { + "epoch": 2.0348675034867503, + "grad_norm": 4.218691349029541, + "learning_rate": 9.9956259995626e-06, + "loss": 0.3609, + "step": 4377 + }, + { + "epoch": 2.0353324035332405, + "grad_norm": 7.700176239013672, + "learning_rate": 9.9956249995625e-06, + "loss": 0.5188, + "step": 4378 + }, + { + "epoch": 2.0357973035797303, + "grad_norm": 6.822635173797607, + "learning_rate": 9.995623999562401e-06, + "loss": 0.9185, + "step": 4379 + }, + { + "epoch": 2.0362622036262206, + "grad_norm": 5.971060752868652, + "learning_rate": 9.995622999562302e-06, + "loss": 0.5607, + "step": 4380 + }, + { + "epoch": 2.0367271036727104, + "grad_norm": 3.842524290084839, + "learning_rate": 9.9956219995622e-06, + "loss": 0.1961, + "step": 4381 + }, + { + "epoch": 2.0371920037192, + "grad_norm": 6.418776988983154, + "learning_rate": 9.995620999562101e-06, + "loss": 0.6462, + "step": 4382 + }, + { + "epoch": 2.0376569037656904, + "grad_norm": 5.239255428314209, + "learning_rate": 9.995619999562e-06, + "loss": 0.4547, + "step": 4383 + }, + { + "epoch": 2.03812180381218, + "grad_norm": 4.326357364654541, + "learning_rate": 9.995618999561901e-06, + "loss": 0.3552, + "step": 4384 + }, + { + "epoch": 2.0385867038586705, + "grad_norm": 5.593657970428467, + "learning_rate": 9.995617999561802e-06, + "loss": 0.4393, + "step": 4385 + }, + { + "epoch": 2.0390516039051603, + "grad_norm": 8.719419479370117, + "learning_rate": 9.9956169995617e-06, + "loss": 0.3275, + "step": 4386 + }, + { + "epoch": 2.0395165039516505, + "grad_norm": 5.297443389892578, + "learning_rate": 9.995615999561601e-06, + "loss": 0.3552, + "step": 4387 + }, + { + "epoch": 2.0399814039981403, + "grad_norm": 4.696147441864014, + "learning_rate": 9.9956149995615e-06, + "loss": 0.4119, + "step": 4388 + }, + { + "epoch": 2.0404463040446306, + "grad_norm": 3.2086164951324463, + "learning_rate": 9.995613999561401e-06, + "loss": 0.2651, + "step": 4389 + }, + { + "epoch": 2.0409112040911204, + "grad_norm": 6.151645660400391, + "learning_rate": 9.9956129995613e-06, + "loss": 0.4809, + "step": 4390 + }, + { + "epoch": 2.0413761041376106, + "grad_norm": 8.604133605957031, + "learning_rate": 9.9956119995612e-06, + "loss": 0.5447, + "step": 4391 + }, + { + "epoch": 2.0418410041841004, + "grad_norm": 3.9845468997955322, + "learning_rate": 9.995610999561101e-06, + "loss": 0.2295, + "step": 4392 + }, + { + "epoch": 2.04230590423059, + "grad_norm": 3.9006552696228027, + "learning_rate": 9.995609999561002e-06, + "loss": 0.2361, + "step": 4393 + }, + { + "epoch": 2.0427708042770805, + "grad_norm": 4.827364921569824, + "learning_rate": 9.9956089995609e-06, + "loss": 0.3709, + "step": 4394 + }, + { + "epoch": 2.0432357043235703, + "grad_norm": 5.229556083679199, + "learning_rate": 9.995607999560801e-06, + "loss": 0.558, + "step": 4395 + }, + { + "epoch": 2.0437006043700605, + "grad_norm": 4.65888786315918, + "learning_rate": 9.9956069995607e-06, + "loss": 0.2556, + "step": 4396 + }, + { + "epoch": 2.0441655044165503, + "grad_norm": 7.854198932647705, + "learning_rate": 9.995605999560601e-06, + "loss": 0.5627, + "step": 4397 + }, + { + "epoch": 2.0446304044630406, + "grad_norm": 6.135409355163574, + "learning_rate": 9.995604999560502e-06, + "loss": 0.5, + "step": 4398 + }, + { + "epoch": 2.0450953045095304, + "grad_norm": 6.340995788574219, + "learning_rate": 9.9956039995604e-06, + "loss": 0.3382, + "step": 4399 + }, + { + "epoch": 2.0455602045560206, + "grad_norm": 10.238054275512695, + "learning_rate": 9.995602999560301e-06, + "loss": 0.8088, + "step": 4400 + }, + { + "epoch": 2.0460251046025104, + "grad_norm": 6.947607517242432, + "learning_rate": 9.9956019995602e-06, + "loss": 0.638, + "step": 4401 + }, + { + "epoch": 2.0464900046490007, + "grad_norm": 8.51333999633789, + "learning_rate": 9.995600999560101e-06, + "loss": 0.8697, + "step": 4402 + }, + { + "epoch": 2.0469549046954905, + "grad_norm": 5.99129056930542, + "learning_rate": 9.99559999956e-06, + "loss": 0.517, + "step": 4403 + }, + { + "epoch": 2.0474198047419803, + "grad_norm": 4.13267707824707, + "learning_rate": 9.9955989995599e-06, + "loss": 0.2998, + "step": 4404 + }, + { + "epoch": 2.0478847047884705, + "grad_norm": 8.955910682678223, + "learning_rate": 9.995597999559801e-06, + "loss": 0.7886, + "step": 4405 + }, + { + "epoch": 2.0483496048349603, + "grad_norm": 4.838528156280518, + "learning_rate": 9.995596999559702e-06, + "loss": 0.4677, + "step": 4406 + }, + { + "epoch": 2.0488145048814506, + "grad_norm": 4.65894889831543, + "learning_rate": 9.9955959995596e-06, + "loss": 0.4073, + "step": 4407 + }, + { + "epoch": 2.0492794049279404, + "grad_norm": 7.286484718322754, + "learning_rate": 9.9955949995595e-06, + "loss": 0.5855, + "step": 4408 + }, + { + "epoch": 2.0497443049744306, + "grad_norm": 6.328083515167236, + "learning_rate": 9.9955939995594e-06, + "loss": 0.6875, + "step": 4409 + }, + { + "epoch": 2.0502092050209204, + "grad_norm": 4.965205669403076, + "learning_rate": 9.995592999559301e-06, + "loss": 0.3761, + "step": 4410 + }, + { + "epoch": 2.0506741050674107, + "grad_norm": 4.089475154876709, + "learning_rate": 9.995591999559202e-06, + "loss": 0.2101, + "step": 4411 + }, + { + "epoch": 2.0511390051139005, + "grad_norm": 4.7504119873046875, + "learning_rate": 9.9955909995591e-06, + "loss": 0.3835, + "step": 4412 + }, + { + "epoch": 2.0516039051603907, + "grad_norm": 3.45050048828125, + "learning_rate": 9.995589999559001e-06, + "loss": 0.2017, + "step": 4413 + }, + { + "epoch": 2.0520688052068805, + "grad_norm": 4.61947774887085, + "learning_rate": 9.9955889995589e-06, + "loss": 0.3526, + "step": 4414 + }, + { + "epoch": 2.0525337052533703, + "grad_norm": 7.2058892250061035, + "learning_rate": 9.995587999558801e-06, + "loss": 0.7388, + "step": 4415 + }, + { + "epoch": 2.0529986052998606, + "grad_norm": 7.462276458740234, + "learning_rate": 9.9955869995587e-06, + "loss": 0.5079, + "step": 4416 + }, + { + "epoch": 2.0534635053463504, + "grad_norm": 6.135153770446777, + "learning_rate": 9.9955859995586e-06, + "loss": 0.4706, + "step": 4417 + }, + { + "epoch": 2.0539284053928406, + "grad_norm": 3.984694004058838, + "learning_rate": 9.995584999558501e-06, + "loss": 0.3763, + "step": 4418 + }, + { + "epoch": 2.0543933054393304, + "grad_norm": 4.5922346115112305, + "learning_rate": 9.9955839995584e-06, + "loss": 0.3205, + "step": 4419 + }, + { + "epoch": 2.0548582054858207, + "grad_norm": 4.2664875984191895, + "learning_rate": 9.995582999558301e-06, + "loss": 0.2098, + "step": 4420 + }, + { + "epoch": 2.0553231055323105, + "grad_norm": 3.9844701290130615, + "learning_rate": 9.9955819995582e-06, + "loss": 0.3406, + "step": 4421 + }, + { + "epoch": 2.0557880055788007, + "grad_norm": 6.283055782318115, + "learning_rate": 9.9955809995581e-06, + "loss": 0.738, + "step": 4422 + }, + { + "epoch": 2.0562529056252905, + "grad_norm": 6.350402355194092, + "learning_rate": 9.995579999558001e-06, + "loss": 0.6828, + "step": 4423 + }, + { + "epoch": 2.0567178056717808, + "grad_norm": 5.68876314163208, + "learning_rate": 9.995578999557902e-06, + "loss": 0.3707, + "step": 4424 + }, + { + "epoch": 2.0571827057182706, + "grad_norm": 4.549171447753906, + "learning_rate": 9.9955779995578e-06, + "loss": 0.296, + "step": 4425 + }, + { + "epoch": 2.0576476057647604, + "grad_norm": 8.83993148803711, + "learning_rate": 9.995576999557701e-06, + "loss": 0.6791, + "step": 4426 + }, + { + "epoch": 2.0581125058112506, + "grad_norm": 3.348956346511841, + "learning_rate": 9.9955759995576e-06, + "loss": 0.2946, + "step": 4427 + }, + { + "epoch": 2.0585774058577404, + "grad_norm": 4.958040237426758, + "learning_rate": 9.995574999557501e-06, + "loss": 0.2564, + "step": 4428 + }, + { + "epoch": 2.0590423059042307, + "grad_norm": 3.24186635017395, + "learning_rate": 9.9955739995574e-06, + "loss": 0.2897, + "step": 4429 + }, + { + "epoch": 2.0595072059507205, + "grad_norm": 4.305703639984131, + "learning_rate": 9.9955729995573e-06, + "loss": 0.3443, + "step": 4430 + }, + { + "epoch": 2.0599721059972107, + "grad_norm": 7.6159749031066895, + "learning_rate": 9.995571999557201e-06, + "loss": 0.8762, + "step": 4431 + }, + { + "epoch": 2.0604370060437005, + "grad_norm": 4.456196308135986, + "learning_rate": 9.9955709995571e-06, + "loss": 0.3093, + "step": 4432 + }, + { + "epoch": 2.0609019060901907, + "grad_norm": 6.137641906738281, + "learning_rate": 9.995569999557001e-06, + "loss": 0.622, + "step": 4433 + }, + { + "epoch": 2.0613668061366806, + "grad_norm": 3.9072725772857666, + "learning_rate": 9.9955689995569e-06, + "loss": 0.3297, + "step": 4434 + }, + { + "epoch": 2.061831706183171, + "grad_norm": 5.701833724975586, + "learning_rate": 9.9955679995568e-06, + "loss": 0.4771, + "step": 4435 + }, + { + "epoch": 2.0622966062296606, + "grad_norm": 5.1864471435546875, + "learning_rate": 9.995566999556701e-06, + "loss": 0.3171, + "step": 4436 + }, + { + "epoch": 2.062761506276151, + "grad_norm": 4.138635635375977, + "learning_rate": 9.995565999556602e-06, + "loss": 0.3575, + "step": 4437 + }, + { + "epoch": 2.0632264063226406, + "grad_norm": 7.505455017089844, + "learning_rate": 9.995564999556501e-06, + "loss": 0.6761, + "step": 4438 + }, + { + "epoch": 2.0636913063691305, + "grad_norm": 4.110079765319824, + "learning_rate": 9.9955639995564e-06, + "loss": 0.2608, + "step": 4439 + }, + { + "epoch": 2.0641562064156207, + "grad_norm": 4.585425853729248, + "learning_rate": 9.9955629995563e-06, + "loss": 0.3431, + "step": 4440 + }, + { + "epoch": 2.0646211064621105, + "grad_norm": 4.588489532470703, + "learning_rate": 9.9955619995562e-06, + "loss": 0.2993, + "step": 4441 + }, + { + "epoch": 2.0650860065086007, + "grad_norm": 4.158510208129883, + "learning_rate": 9.9955609995561e-06, + "loss": 0.3566, + "step": 4442 + }, + { + "epoch": 2.0655509065550905, + "grad_norm": 5.062590599060059, + "learning_rate": 9.995559999556e-06, + "loss": 0.5296, + "step": 4443 + }, + { + "epoch": 2.066015806601581, + "grad_norm": 6.528486728668213, + "learning_rate": 9.995558999555901e-06, + "loss": 0.3688, + "step": 4444 + }, + { + "epoch": 2.0664807066480706, + "grad_norm": 3.3784115314483643, + "learning_rate": 9.9955579995558e-06, + "loss": 0.1512, + "step": 4445 + }, + { + "epoch": 2.066945606694561, + "grad_norm": 4.632044792175293, + "learning_rate": 9.995556999555701e-06, + "loss": 0.2746, + "step": 4446 + }, + { + "epoch": 2.0674105067410506, + "grad_norm": 3.362703323364258, + "learning_rate": 9.9955559995556e-06, + "loss": 0.1615, + "step": 4447 + }, + { + "epoch": 2.067875406787541, + "grad_norm": 2.98360276222229, + "learning_rate": 9.9955549995555e-06, + "loss": 0.2731, + "step": 4448 + }, + { + "epoch": 2.0683403068340307, + "grad_norm": 6.080516338348389, + "learning_rate": 9.995553999555401e-06, + "loss": 0.6297, + "step": 4449 + }, + { + "epoch": 2.0688052068805205, + "grad_norm": 3.736675262451172, + "learning_rate": 9.9955529995553e-06, + "loss": 0.255, + "step": 4450 + }, + { + "epoch": 2.0692701069270107, + "grad_norm": 5.072849750518799, + "learning_rate": 9.995551999555201e-06, + "loss": 0.5534, + "step": 4451 + }, + { + "epoch": 2.0697350069735005, + "grad_norm": 4.229574203491211, + "learning_rate": 9.9955509995551e-06, + "loss": 0.2999, + "step": 4452 + }, + { + "epoch": 2.070199907019991, + "grad_norm": 5.77674674987793, + "learning_rate": 9.995549999555e-06, + "loss": 0.5272, + "step": 4453 + }, + { + "epoch": 2.0706648070664806, + "grad_norm": 4.305337429046631, + "learning_rate": 9.995548999554901e-06, + "loss": 0.3512, + "step": 4454 + }, + { + "epoch": 2.071129707112971, + "grad_norm": 8.060033798217773, + "learning_rate": 9.9955479995548e-06, + "loss": 0.6401, + "step": 4455 + }, + { + "epoch": 2.0715946071594606, + "grad_norm": 9.337298393249512, + "learning_rate": 9.995546999554701e-06, + "loss": 0.6232, + "step": 4456 + }, + { + "epoch": 2.072059507205951, + "grad_norm": 3.7578608989715576, + "learning_rate": 9.995545999554602e-06, + "loss": 0.2918, + "step": 4457 + }, + { + "epoch": 2.0725244072524407, + "grad_norm": 6.999575614929199, + "learning_rate": 9.9955449995545e-06, + "loss": 0.7639, + "step": 4458 + }, + { + "epoch": 2.072989307298931, + "grad_norm": 6.393758296966553, + "learning_rate": 9.995543999554401e-06, + "loss": 0.4689, + "step": 4459 + }, + { + "epoch": 2.0734542073454207, + "grad_norm": 6.801157474517822, + "learning_rate": 9.9955429995543e-06, + "loss": 0.8224, + "step": 4460 + }, + { + "epoch": 2.0739191073919105, + "grad_norm": 5.318775177001953, + "learning_rate": 9.9955419995542e-06, + "loss": 0.3792, + "step": 4461 + }, + { + "epoch": 2.074384007438401, + "grad_norm": 6.2710137367248535, + "learning_rate": 9.995540999554101e-06, + "loss": 0.465, + "step": 4462 + }, + { + "epoch": 2.0748489074848906, + "grad_norm": 5.116504192352295, + "learning_rate": 9.995539999554e-06, + "loss": 0.4578, + "step": 4463 + }, + { + "epoch": 2.075313807531381, + "grad_norm": 4.164783477783203, + "learning_rate": 9.995538999553901e-06, + "loss": 0.3804, + "step": 4464 + }, + { + "epoch": 2.0757787075778706, + "grad_norm": 5.398975372314453, + "learning_rate": 9.9955379995538e-06, + "loss": 0.3741, + "step": 4465 + }, + { + "epoch": 2.076243607624361, + "grad_norm": 4.544362545013428, + "learning_rate": 9.9955369995537e-06, + "loss": 0.3392, + "step": 4466 + }, + { + "epoch": 2.0767085076708507, + "grad_norm": 5.445427894592285, + "learning_rate": 9.995535999553601e-06, + "loss": 0.4189, + "step": 4467 + }, + { + "epoch": 2.077173407717341, + "grad_norm": 5.377079963684082, + "learning_rate": 9.9955349995535e-06, + "loss": 0.5432, + "step": 4468 + }, + { + "epoch": 2.0776383077638307, + "grad_norm": 5.41078519821167, + "learning_rate": 9.995533999553401e-06, + "loss": 0.6549, + "step": 4469 + }, + { + "epoch": 2.078103207810321, + "grad_norm": 5.969748020172119, + "learning_rate": 9.995532999553302e-06, + "loss": 0.616, + "step": 4470 + }, + { + "epoch": 2.078568107856811, + "grad_norm": 6.971005916595459, + "learning_rate": 9.9955319995532e-06, + "loss": 0.499, + "step": 4471 + }, + { + "epoch": 2.0790330079033006, + "grad_norm": 6.383263111114502, + "learning_rate": 9.9955309995531e-06, + "loss": 0.5999, + "step": 4472 + }, + { + "epoch": 2.079497907949791, + "grad_norm": 5.150323390960693, + "learning_rate": 9.995529999553e-06, + "loss": 0.3951, + "step": 4473 + }, + { + "epoch": 2.0799628079962806, + "grad_norm": 5.359004497528076, + "learning_rate": 9.995528999552901e-06, + "loss": 0.3944, + "step": 4474 + }, + { + "epoch": 2.080427708042771, + "grad_norm": 4.072765350341797, + "learning_rate": 9.995527999552802e-06, + "loss": 0.2521, + "step": 4475 + }, + { + "epoch": 2.0808926080892607, + "grad_norm": 6.066372394561768, + "learning_rate": 9.9955269995527e-06, + "loss": 0.5854, + "step": 4476 + }, + { + "epoch": 2.081357508135751, + "grad_norm": 3.7731075286865234, + "learning_rate": 9.995525999552601e-06, + "loss": 0.312, + "step": 4477 + }, + { + "epoch": 2.0818224081822407, + "grad_norm": 4.634199142456055, + "learning_rate": 9.9955249995525e-06, + "loss": 0.4263, + "step": 4478 + }, + { + "epoch": 2.082287308228731, + "grad_norm": 8.189164161682129, + "learning_rate": 9.9955239995524e-06, + "loss": 0.7227, + "step": 4479 + }, + { + "epoch": 2.082752208275221, + "grad_norm": 5.2085347175598145, + "learning_rate": 9.995522999552301e-06, + "loss": 0.3738, + "step": 4480 + }, + { + "epoch": 2.083217108321711, + "grad_norm": 6.330471515655518, + "learning_rate": 9.995521999552202e-06, + "loss": 0.5794, + "step": 4481 + }, + { + "epoch": 2.083682008368201, + "grad_norm": 4.566965579986572, + "learning_rate": 9.995520999552101e-06, + "loss": 0.4076, + "step": 4482 + }, + { + "epoch": 2.0841469084146906, + "grad_norm": 4.9148783683776855, + "learning_rate": 9.995519999552e-06, + "loss": 0.5468, + "step": 4483 + }, + { + "epoch": 2.084611808461181, + "grad_norm": 4.126989364624023, + "learning_rate": 9.9955189995519e-06, + "loss": 0.4003, + "step": 4484 + }, + { + "epoch": 2.0850767085076707, + "grad_norm": 6.489853858947754, + "learning_rate": 9.9955179995518e-06, + "loss": 0.5989, + "step": 4485 + }, + { + "epoch": 2.085541608554161, + "grad_norm": 4.403002738952637, + "learning_rate": 9.9955169995517e-06, + "loss": 0.3497, + "step": 4486 + }, + { + "epoch": 2.0860065086006507, + "grad_norm": 6.007173538208008, + "learning_rate": 9.995515999551601e-06, + "loss": 0.6563, + "step": 4487 + }, + { + "epoch": 2.086471408647141, + "grad_norm": 5.2308807373046875, + "learning_rate": 9.995514999551502e-06, + "loss": 0.5786, + "step": 4488 + }, + { + "epoch": 2.086936308693631, + "grad_norm": 5.30234956741333, + "learning_rate": 9.9955139995514e-06, + "loss": 0.2574, + "step": 4489 + }, + { + "epoch": 2.087401208740121, + "grad_norm": 3.356689691543579, + "learning_rate": 9.995512999551301e-06, + "loss": 0.252, + "step": 4490 + }, + { + "epoch": 2.087866108786611, + "grad_norm": 8.895821571350098, + "learning_rate": 9.9955119995512e-06, + "loss": 0.8158, + "step": 4491 + }, + { + "epoch": 2.088331008833101, + "grad_norm": 4.165257453918457, + "learning_rate": 9.995510999551101e-06, + "loss": 0.3039, + "step": 4492 + }, + { + "epoch": 2.088795908879591, + "grad_norm": 5.214908599853516, + "learning_rate": 9.995509999551002e-06, + "loss": 0.4907, + "step": 4493 + }, + { + "epoch": 2.0892608089260807, + "grad_norm": 4.592205047607422, + "learning_rate": 9.9955089995509e-06, + "loss": 0.4865, + "step": 4494 + }, + { + "epoch": 2.089725708972571, + "grad_norm": 4.395826816558838, + "learning_rate": 9.995507999550801e-06, + "loss": 0.2009, + "step": 4495 + }, + { + "epoch": 2.0901906090190607, + "grad_norm": 5.153807163238525, + "learning_rate": 9.9955069995507e-06, + "loss": 0.5053, + "step": 4496 + }, + { + "epoch": 2.090655509065551, + "grad_norm": 5.32899284362793, + "learning_rate": 9.9955059995506e-06, + "loss": 0.514, + "step": 4497 + }, + { + "epoch": 2.091120409112041, + "grad_norm": 4.573617458343506, + "learning_rate": 9.9955049995505e-06, + "loss": 0.3287, + "step": 4498 + }, + { + "epoch": 2.091585309158531, + "grad_norm": 5.310750484466553, + "learning_rate": 9.9955039995504e-06, + "loss": 0.5379, + "step": 4499 + }, + { + "epoch": 2.092050209205021, + "grad_norm": 7.873195171356201, + "learning_rate": 9.995502999550301e-06, + "loss": 0.3049, + "step": 4500 + }, + { + "epoch": 2.092515109251511, + "grad_norm": 4.763248920440674, + "learning_rate": 9.995501999550202e-06, + "loss": 0.4889, + "step": 4501 + }, + { + "epoch": 2.092980009298001, + "grad_norm": 3.5237812995910645, + "learning_rate": 9.9955009995501e-06, + "loss": 0.1435, + "step": 4502 + }, + { + "epoch": 2.093444909344491, + "grad_norm": 7.735689163208008, + "learning_rate": 9.995499999550001e-06, + "loss": 0.5613, + "step": 4503 + }, + { + "epoch": 2.093909809390981, + "grad_norm": 3.868567943572998, + "learning_rate": 9.9954989995499e-06, + "loss": 0.2175, + "step": 4504 + }, + { + "epoch": 2.0943747094374707, + "grad_norm": 3.7745823860168457, + "learning_rate": 9.995497999549801e-06, + "loss": 0.2707, + "step": 4505 + }, + { + "epoch": 2.094839609483961, + "grad_norm": 3.4765141010284424, + "learning_rate": 9.995496999549702e-06, + "loss": 0.2017, + "step": 4506 + }, + { + "epoch": 2.0953045095304508, + "grad_norm": 5.987328052520752, + "learning_rate": 9.9954959995496e-06, + "loss": 0.4369, + "step": 4507 + }, + { + "epoch": 2.095769409576941, + "grad_norm": 2.8795788288116455, + "learning_rate": 9.995494999549501e-06, + "loss": 0.18, + "step": 4508 + }, + { + "epoch": 2.096234309623431, + "grad_norm": 4.063522815704346, + "learning_rate": 9.9954939995494e-06, + "loss": 0.2161, + "step": 4509 + }, + { + "epoch": 2.096699209669921, + "grad_norm": 4.575754165649414, + "learning_rate": 9.9954929995493e-06, + "loss": 0.4363, + "step": 4510 + }, + { + "epoch": 2.097164109716411, + "grad_norm": 5.305547714233398, + "learning_rate": 9.9954919995492e-06, + "loss": 0.5399, + "step": 4511 + }, + { + "epoch": 2.097629009762901, + "grad_norm": 6.658756732940674, + "learning_rate": 9.9954909995491e-06, + "loss": 0.4678, + "step": 4512 + }, + { + "epoch": 2.098093909809391, + "grad_norm": 4.010667324066162, + "learning_rate": 9.995489999549001e-06, + "loss": 0.3118, + "step": 4513 + }, + { + "epoch": 2.098558809855881, + "grad_norm": 5.4602203369140625, + "learning_rate": 9.995488999548902e-06, + "loss": 0.5742, + "step": 4514 + }, + { + "epoch": 2.099023709902371, + "grad_norm": 5.365551471710205, + "learning_rate": 9.9954879995488e-06, + "loss": 0.5068, + "step": 4515 + }, + { + "epoch": 2.099488609948861, + "grad_norm": 6.776236057281494, + "learning_rate": 9.9954869995487e-06, + "loss": 0.2561, + "step": 4516 + }, + { + "epoch": 2.099953509995351, + "grad_norm": 8.361666679382324, + "learning_rate": 9.9954859995486e-06, + "loss": 0.5557, + "step": 4517 + }, + { + "epoch": 2.100418410041841, + "grad_norm": 5.733133316040039, + "learning_rate": 9.995484999548501e-06, + "loss": 0.4835, + "step": 4518 + }, + { + "epoch": 2.100883310088331, + "grad_norm": 6.771964073181152, + "learning_rate": 9.995483999548402e-06, + "loss": 0.7423, + "step": 4519 + }, + { + "epoch": 2.101348210134821, + "grad_norm": 11.839274406433105, + "learning_rate": 9.9954829995483e-06, + "loss": 0.4419, + "step": 4520 + }, + { + "epoch": 2.101813110181311, + "grad_norm": 4.448545455932617, + "learning_rate": 9.995481999548201e-06, + "loss": 0.429, + "step": 4521 + }, + { + "epoch": 2.102278010227801, + "grad_norm": 4.411886215209961, + "learning_rate": 9.9954809995481e-06, + "loss": 0.3478, + "step": 4522 + }, + { + "epoch": 2.102742910274291, + "grad_norm": 11.613450050354004, + "learning_rate": 9.995479999548001e-06, + "loss": 1.0489, + "step": 4523 + }, + { + "epoch": 2.103207810320781, + "grad_norm": 7.882637977600098, + "learning_rate": 9.9954789995479e-06, + "loss": 0.6431, + "step": 4524 + }, + { + "epoch": 2.103672710367271, + "grad_norm": 4.715241432189941, + "learning_rate": 9.9954779995478e-06, + "loss": 0.303, + "step": 4525 + }, + { + "epoch": 2.104137610413761, + "grad_norm": 4.0706024169921875, + "learning_rate": 9.995476999547701e-06, + "loss": 0.2395, + "step": 4526 + }, + { + "epoch": 2.104602510460251, + "grad_norm": 5.69342565536499, + "learning_rate": 9.9954759995476e-06, + "loss": 0.4489, + "step": 4527 + }, + { + "epoch": 2.105067410506741, + "grad_norm": 4.881242752075195, + "learning_rate": 9.9954749995475e-06, + "loss": 0.3257, + "step": 4528 + }, + { + "epoch": 2.105532310553231, + "grad_norm": 5.3210554122924805, + "learning_rate": 9.9954739995474e-06, + "loss": 0.3792, + "step": 4529 + }, + { + "epoch": 2.105997210599721, + "grad_norm": 4.527942657470703, + "learning_rate": 9.9954729995473e-06, + "loss": 0.4432, + "step": 4530 + }, + { + "epoch": 2.106462110646211, + "grad_norm": 8.261435508728027, + "learning_rate": 9.995471999547201e-06, + "loss": 0.5893, + "step": 4531 + }, + { + "epoch": 2.106927010692701, + "grad_norm": 4.677630424499512, + "learning_rate": 9.995470999547102e-06, + "loss": 0.2803, + "step": 4532 + }, + { + "epoch": 2.107391910739191, + "grad_norm": 8.608001708984375, + "learning_rate": 9.995469999547e-06, + "loss": 0.7597, + "step": 4533 + }, + { + "epoch": 2.107856810785681, + "grad_norm": 4.312868118286133, + "learning_rate": 9.995468999546901e-06, + "loss": 0.2568, + "step": 4534 + }, + { + "epoch": 2.108321710832171, + "grad_norm": 4.265228271484375, + "learning_rate": 9.9954679995468e-06, + "loss": 0.4319, + "step": 4535 + }, + { + "epoch": 2.1087866108786613, + "grad_norm": 6.7259955406188965, + "learning_rate": 9.995466999546701e-06, + "loss": 0.3891, + "step": 4536 + }, + { + "epoch": 2.109251510925151, + "grad_norm": 5.179603099822998, + "learning_rate": 9.9954659995466e-06, + "loss": 0.2346, + "step": 4537 + }, + { + "epoch": 2.1097164109716413, + "grad_norm": 8.697959899902344, + "learning_rate": 9.9954649995465e-06, + "loss": 0.4661, + "step": 4538 + }, + { + "epoch": 2.110181311018131, + "grad_norm": 5.724148750305176, + "learning_rate": 9.995463999546401e-06, + "loss": 0.593, + "step": 4539 + }, + { + "epoch": 2.110646211064621, + "grad_norm": 6.162160873413086, + "learning_rate": 9.9954629995463e-06, + "loss": 0.4377, + "step": 4540 + }, + { + "epoch": 2.111111111111111, + "grad_norm": 7.337324619293213, + "learning_rate": 9.995461999546201e-06, + "loss": 0.5685, + "step": 4541 + }, + { + "epoch": 2.111576011157601, + "grad_norm": 7.004763126373291, + "learning_rate": 9.9954609995461e-06, + "loss": 0.5315, + "step": 4542 + }, + { + "epoch": 2.112040911204091, + "grad_norm": 3.9406282901763916, + "learning_rate": 9.995459999546e-06, + "loss": 0.3126, + "step": 4543 + }, + { + "epoch": 2.112505811250581, + "grad_norm": 5.474116802215576, + "learning_rate": 9.995458999545901e-06, + "loss": 0.4429, + "step": 4544 + }, + { + "epoch": 2.1129707112970713, + "grad_norm": 5.896844863891602, + "learning_rate": 9.995457999545802e-06, + "loss": 0.5244, + "step": 4545 + }, + { + "epoch": 2.113435611343561, + "grad_norm": 9.949884414672852, + "learning_rate": 9.9954569995457e-06, + "loss": 1.0591, + "step": 4546 + }, + { + "epoch": 2.1139005113900513, + "grad_norm": 6.013916492462158, + "learning_rate": 9.995455999545601e-06, + "loss": 0.6039, + "step": 4547 + }, + { + "epoch": 2.114365411436541, + "grad_norm": 5.972525119781494, + "learning_rate": 9.9954549995455e-06, + "loss": 0.3005, + "step": 4548 + }, + { + "epoch": 2.1148303114830314, + "grad_norm": 5.493051052093506, + "learning_rate": 9.995453999545401e-06, + "loss": 0.5275, + "step": 4549 + }, + { + "epoch": 2.115295211529521, + "grad_norm": 4.711793422698975, + "learning_rate": 9.995452999545302e-06, + "loss": 0.4131, + "step": 4550 + }, + { + "epoch": 2.115760111576011, + "grad_norm": 4.24843168258667, + "learning_rate": 9.9954519995452e-06, + "loss": 0.2294, + "step": 4551 + }, + { + "epoch": 2.116225011622501, + "grad_norm": 5.912128925323486, + "learning_rate": 9.995450999545101e-06, + "loss": 0.4711, + "step": 4552 + }, + { + "epoch": 2.116689911668991, + "grad_norm": 4.968247413635254, + "learning_rate": 9.995449999545e-06, + "loss": 0.4134, + "step": 4553 + }, + { + "epoch": 2.1171548117154813, + "grad_norm": 4.456722259521484, + "learning_rate": 9.995448999544901e-06, + "loss": 0.2907, + "step": 4554 + }, + { + "epoch": 2.117619711761971, + "grad_norm": 5.796450138092041, + "learning_rate": 9.9954479995448e-06, + "loss": 0.5939, + "step": 4555 + }, + { + "epoch": 2.1180846118084613, + "grad_norm": 5.221571445465088, + "learning_rate": 9.9954469995447e-06, + "loss": 0.4846, + "step": 4556 + }, + { + "epoch": 2.118549511854951, + "grad_norm": 7.8954386711120605, + "learning_rate": 9.995445999544601e-06, + "loss": 0.2876, + "step": 4557 + }, + { + "epoch": 2.1190144119014414, + "grad_norm": 4.0814595222473145, + "learning_rate": 9.995444999544502e-06, + "loss": 0.3115, + "step": 4558 + }, + { + "epoch": 2.119479311947931, + "grad_norm": 6.372552871704102, + "learning_rate": 9.995443999544401e-06, + "loss": 0.4023, + "step": 4559 + }, + { + "epoch": 2.1199442119944214, + "grad_norm": 7.5814008712768555, + "learning_rate": 9.9954429995443e-06, + "loss": 0.3336, + "step": 4560 + }, + { + "epoch": 2.120409112040911, + "grad_norm": 6.167369842529297, + "learning_rate": 9.9954419995442e-06, + "loss": 0.7265, + "step": 4561 + }, + { + "epoch": 2.120874012087401, + "grad_norm": 5.609127998352051, + "learning_rate": 9.995440999544101e-06, + "loss": 0.3544, + "step": 4562 + }, + { + "epoch": 2.1213389121338913, + "grad_norm": 6.021010398864746, + "learning_rate": 9.995439999544002e-06, + "loss": 0.5963, + "step": 4563 + }, + { + "epoch": 2.121803812180381, + "grad_norm": 5.336641788482666, + "learning_rate": 9.9954389995439e-06, + "loss": 0.6423, + "step": 4564 + }, + { + "epoch": 2.1222687122268713, + "grad_norm": 4.186577320098877, + "learning_rate": 9.995437999543801e-06, + "loss": 0.3469, + "step": 4565 + }, + { + "epoch": 2.122733612273361, + "grad_norm": 5.084349632263184, + "learning_rate": 9.9954369995437e-06, + "loss": 0.2641, + "step": 4566 + }, + { + "epoch": 2.1231985123198513, + "grad_norm": 5.134176254272461, + "learning_rate": 9.995435999543601e-06, + "loss": 0.4857, + "step": 4567 + }, + { + "epoch": 2.123663412366341, + "grad_norm": 6.512207984924316, + "learning_rate": 9.9954349995435e-06, + "loss": 0.3385, + "step": 4568 + }, + { + "epoch": 2.1241283124128314, + "grad_norm": 6.356027603149414, + "learning_rate": 9.9954339995434e-06, + "loss": 0.3502, + "step": 4569 + }, + { + "epoch": 2.124593212459321, + "grad_norm": 5.5209245681762695, + "learning_rate": 9.995432999543301e-06, + "loss": 0.363, + "step": 4570 + }, + { + "epoch": 2.1250581125058114, + "grad_norm": 6.20979118347168, + "learning_rate": 9.9954319995432e-06, + "loss": 0.4697, + "step": 4571 + }, + { + "epoch": 2.1255230125523012, + "grad_norm": 4.572920799255371, + "learning_rate": 9.995430999543101e-06, + "loss": 0.2772, + "step": 4572 + }, + { + "epoch": 2.125987912598791, + "grad_norm": 5.586182117462158, + "learning_rate": 9.995429999543e-06, + "loss": 0.4718, + "step": 4573 + }, + { + "epoch": 2.1264528126452813, + "grad_norm": 6.048141956329346, + "learning_rate": 9.9954289995429e-06, + "loss": 0.7563, + "step": 4574 + }, + { + "epoch": 2.126917712691771, + "grad_norm": 5.590456485748291, + "learning_rate": 9.995427999542801e-06, + "loss": 0.4878, + "step": 4575 + }, + { + "epoch": 2.1273826127382613, + "grad_norm": 7.631307601928711, + "learning_rate": 9.995426999542702e-06, + "loss": 0.4994, + "step": 4576 + }, + { + "epoch": 2.127847512784751, + "grad_norm": 4.975677013397217, + "learning_rate": 9.995425999542601e-06, + "loss": 0.3976, + "step": 4577 + }, + { + "epoch": 2.1283124128312414, + "grad_norm": 5.093850612640381, + "learning_rate": 9.995424999542502e-06, + "loss": 0.3796, + "step": 4578 + }, + { + "epoch": 2.128777312877731, + "grad_norm": 6.08567476272583, + "learning_rate": 9.9954239995424e-06, + "loss": 0.456, + "step": 4579 + }, + { + "epoch": 2.1292422129242214, + "grad_norm": 4.277512073516846, + "learning_rate": 9.9954229995423e-06, + "loss": 0.4012, + "step": 4580 + }, + { + "epoch": 2.1297071129707112, + "grad_norm": 6.019728183746338, + "learning_rate": 9.9954219995422e-06, + "loss": 0.5935, + "step": 4581 + }, + { + "epoch": 2.1301720130172015, + "grad_norm": 5.698948860168457, + "learning_rate": 9.9954209995421e-06, + "loss": 0.649, + "step": 4582 + }, + { + "epoch": 2.1306369130636913, + "grad_norm": 6.017704963684082, + "learning_rate": 9.995419999542001e-06, + "loss": 0.557, + "step": 4583 + }, + { + "epoch": 2.131101813110181, + "grad_norm": 9.263684272766113, + "learning_rate": 9.9954189995419e-06, + "loss": 0.8523, + "step": 4584 + }, + { + "epoch": 2.1315667131566713, + "grad_norm": 3.368750810623169, + "learning_rate": 9.995417999541801e-06, + "loss": 0.273, + "step": 4585 + }, + { + "epoch": 2.132031613203161, + "grad_norm": 4.290988922119141, + "learning_rate": 9.9954169995417e-06, + "loss": 0.2405, + "step": 4586 + }, + { + "epoch": 2.1324965132496514, + "grad_norm": 6.543356895446777, + "learning_rate": 9.9954159995416e-06, + "loss": 0.7159, + "step": 4587 + }, + { + "epoch": 2.132961413296141, + "grad_norm": 5.382047176361084, + "learning_rate": 9.995414999541501e-06, + "loss": 0.543, + "step": 4588 + }, + { + "epoch": 2.1334263133426314, + "grad_norm": 4.926426410675049, + "learning_rate": 9.995413999541402e-06, + "loss": 0.4866, + "step": 4589 + }, + { + "epoch": 2.1338912133891212, + "grad_norm": 12.18683910369873, + "learning_rate": 9.995412999541301e-06, + "loss": 1.3824, + "step": 4590 + }, + { + "epoch": 2.1343561134356115, + "grad_norm": 5.355868816375732, + "learning_rate": 9.9954119995412e-06, + "loss": 0.3881, + "step": 4591 + }, + { + "epoch": 2.1348210134821013, + "grad_norm": 4.589108467102051, + "learning_rate": 9.9954109995411e-06, + "loss": 0.4113, + "step": 4592 + }, + { + "epoch": 2.1352859135285915, + "grad_norm": 6.586643695831299, + "learning_rate": 9.995409999541e-06, + "loss": 0.3625, + "step": 4593 + }, + { + "epoch": 2.1357508135750813, + "grad_norm": 5.236112594604492, + "learning_rate": 9.9954089995409e-06, + "loss": 0.4572, + "step": 4594 + }, + { + "epoch": 2.1362157136215716, + "grad_norm": 4.505550861358643, + "learning_rate": 9.9954079995408e-06, + "loss": 0.5469, + "step": 4595 + }, + { + "epoch": 2.1366806136680614, + "grad_norm": 6.0099382400512695, + "learning_rate": 9.995406999540701e-06, + "loss": 0.5796, + "step": 4596 + }, + { + "epoch": 2.137145513714551, + "grad_norm": 7.066587924957275, + "learning_rate": 9.9954059995406e-06, + "loss": 0.3507, + "step": 4597 + }, + { + "epoch": 2.1376104137610414, + "grad_norm": 7.106479167938232, + "learning_rate": 9.995404999540501e-06, + "loss": 0.4737, + "step": 4598 + }, + { + "epoch": 2.1380753138075312, + "grad_norm": 2.8655142784118652, + "learning_rate": 9.9954039995404e-06, + "loss": 0.2456, + "step": 4599 + }, + { + "epoch": 2.1385402138540215, + "grad_norm": 4.8315935134887695, + "learning_rate": 9.9954029995403e-06, + "loss": 0.3927, + "step": 4600 + }, + { + "epoch": 2.1390051139005113, + "grad_norm": 4.2220048904418945, + "learning_rate": 9.995401999540201e-06, + "loss": 0.4307, + "step": 4601 + }, + { + "epoch": 2.1394700139470015, + "grad_norm": 4.457540035247803, + "learning_rate": 9.9954009995401e-06, + "loss": 0.2712, + "step": 4602 + }, + { + "epoch": 2.1399349139934913, + "grad_norm": 4.89407205581665, + "learning_rate": 9.995399999540001e-06, + "loss": 0.5115, + "step": 4603 + }, + { + "epoch": 2.1403998140399816, + "grad_norm": 5.4724016189575195, + "learning_rate": 9.9953989995399e-06, + "loss": 0.361, + "step": 4604 + }, + { + "epoch": 2.1408647140864714, + "grad_norm": 8.249471664428711, + "learning_rate": 9.9953979995398e-06, + "loss": 0.3761, + "step": 4605 + }, + { + "epoch": 2.141329614132961, + "grad_norm": 5.903730392456055, + "learning_rate": 9.9953969995397e-06, + "loss": 0.5458, + "step": 4606 + }, + { + "epoch": 2.1417945141794514, + "grad_norm": 4.160175323486328, + "learning_rate": 9.9953959995396e-06, + "loss": 0.2801, + "step": 4607 + }, + { + "epoch": 2.1422594142259412, + "grad_norm": 6.114616870880127, + "learning_rate": 9.995394999539501e-06, + "loss": 0.6096, + "step": 4608 + }, + { + "epoch": 2.1427243142724315, + "grad_norm": 5.814290523529053, + "learning_rate": 9.995393999539402e-06, + "loss": 0.4801, + "step": 4609 + }, + { + "epoch": 2.1431892143189213, + "grad_norm": 4.909616470336914, + "learning_rate": 9.9953929995393e-06, + "loss": 0.5775, + "step": 4610 + }, + { + "epoch": 2.1436541143654115, + "grad_norm": 6.813101291656494, + "learning_rate": 9.995391999539201e-06, + "loss": 0.3149, + "step": 4611 + }, + { + "epoch": 2.1441190144119013, + "grad_norm": 4.54240083694458, + "learning_rate": 9.9953909995391e-06, + "loss": 0.3614, + "step": 4612 + }, + { + "epoch": 2.1445839144583916, + "grad_norm": 6.315625190734863, + "learning_rate": 9.995389999539e-06, + "loss": 0.4616, + "step": 4613 + }, + { + "epoch": 2.1450488145048814, + "grad_norm": 6.93907356262207, + "learning_rate": 9.995388999538901e-06, + "loss": 0.958, + "step": 4614 + }, + { + "epoch": 2.1455137145513716, + "grad_norm": 5.6072516441345215, + "learning_rate": 9.9953879995388e-06, + "loss": 0.4536, + "step": 4615 + }, + { + "epoch": 2.1459786145978614, + "grad_norm": 6.536250591278076, + "learning_rate": 9.995386999538701e-06, + "loss": 0.685, + "step": 4616 + }, + { + "epoch": 2.1464435146443517, + "grad_norm": 6.941993713378906, + "learning_rate": 9.9953859995386e-06, + "loss": 0.6111, + "step": 4617 + }, + { + "epoch": 2.1469084146908415, + "grad_norm": 6.530501365661621, + "learning_rate": 9.9953849995385e-06, + "loss": 0.4215, + "step": 4618 + }, + { + "epoch": 2.1473733147373313, + "grad_norm": 5.384678840637207, + "learning_rate": 9.995383999538401e-06, + "loss": 0.5116, + "step": 4619 + }, + { + "epoch": 2.1478382147838215, + "grad_norm": 5.49621057510376, + "learning_rate": 9.9953829995383e-06, + "loss": 0.4605, + "step": 4620 + }, + { + "epoch": 2.1483031148303113, + "grad_norm": 5.151256084442139, + "learning_rate": 9.995381999538201e-06, + "loss": 0.3538, + "step": 4621 + }, + { + "epoch": 2.1487680148768016, + "grad_norm": 5.450331687927246, + "learning_rate": 9.995380999538102e-06, + "loss": 0.3549, + "step": 4622 + }, + { + "epoch": 2.1492329149232914, + "grad_norm": 4.616213798522949, + "learning_rate": 9.995379999538e-06, + "loss": 0.3781, + "step": 4623 + }, + { + "epoch": 2.1496978149697816, + "grad_norm": 4.195201873779297, + "learning_rate": 9.9953789995379e-06, + "loss": 0.26, + "step": 4624 + }, + { + "epoch": 2.1501627150162714, + "grad_norm": 4.711382865905762, + "learning_rate": 9.9953779995378e-06, + "loss": 0.3567, + "step": 4625 + }, + { + "epoch": 2.1506276150627617, + "grad_norm": 5.3123626708984375, + "learning_rate": 9.995376999537701e-06, + "loss": 0.4916, + "step": 4626 + }, + { + "epoch": 2.1510925151092515, + "grad_norm": 8.704558372497559, + "learning_rate": 9.995375999537602e-06, + "loss": 0.5222, + "step": 4627 + }, + { + "epoch": 2.1515574151557413, + "grad_norm": 4.64261531829834, + "learning_rate": 9.9953749995375e-06, + "loss": 0.3637, + "step": 4628 + }, + { + "epoch": 2.1520223152022315, + "grad_norm": 6.44559383392334, + "learning_rate": 9.995373999537401e-06, + "loss": 0.4942, + "step": 4629 + }, + { + "epoch": 2.1524872152487213, + "grad_norm": 3.565222978591919, + "learning_rate": 9.9953729995373e-06, + "loss": 0.4048, + "step": 4630 + }, + { + "epoch": 2.1529521152952116, + "grad_norm": 6.838956356048584, + "learning_rate": 9.9953719995372e-06, + "loss": 0.7977, + "step": 4631 + }, + { + "epoch": 2.1534170153417014, + "grad_norm": 3.721721887588501, + "learning_rate": 9.995370999537101e-06, + "loss": 0.2703, + "step": 4632 + }, + { + "epoch": 2.1538819153881916, + "grad_norm": 5.4065985679626465, + "learning_rate": 9.995369999537e-06, + "loss": 0.4767, + "step": 4633 + }, + { + "epoch": 2.1543468154346814, + "grad_norm": 3.9923527240753174, + "learning_rate": 9.995368999536901e-06, + "loss": 0.2099, + "step": 4634 + }, + { + "epoch": 2.1548117154811717, + "grad_norm": 4.487736225128174, + "learning_rate": 9.9953679995368e-06, + "loss": 0.3498, + "step": 4635 + }, + { + "epoch": 2.1552766155276615, + "grad_norm": 7.231865882873535, + "learning_rate": 9.9953669995367e-06, + "loss": 0.4588, + "step": 4636 + }, + { + "epoch": 2.1557415155741517, + "grad_norm": 5.083963871002197, + "learning_rate": 9.9953659995366e-06, + "loss": 0.5333, + "step": 4637 + }, + { + "epoch": 2.1562064156206415, + "grad_norm": 5.567556858062744, + "learning_rate": 9.9953649995365e-06, + "loss": 0.524, + "step": 4638 + }, + { + "epoch": 2.1566713156671318, + "grad_norm": 4.449225902557373, + "learning_rate": 9.995363999536401e-06, + "loss": 0.4818, + "step": 4639 + }, + { + "epoch": 2.1571362157136216, + "grad_norm": 4.958736896514893, + "learning_rate": 9.995362999536302e-06, + "loss": 0.6097, + "step": 4640 + }, + { + "epoch": 2.1576011157601114, + "grad_norm": 4.7312164306640625, + "learning_rate": 9.9953619995362e-06, + "loss": 0.2879, + "step": 4641 + }, + { + "epoch": 2.1580660158066016, + "grad_norm": 4.3443922996521, + "learning_rate": 9.995360999536101e-06, + "loss": 0.2481, + "step": 4642 + }, + { + "epoch": 2.1585309158530914, + "grad_norm": 7.005352973937988, + "learning_rate": 9.995359999536e-06, + "loss": 0.5458, + "step": 4643 + }, + { + "epoch": 2.1589958158995817, + "grad_norm": 6.340753078460693, + "learning_rate": 9.995358999535901e-06, + "loss": 0.4932, + "step": 4644 + }, + { + "epoch": 2.1594607159460715, + "grad_norm": 5.429219722747803, + "learning_rate": 9.995357999535802e-06, + "loss": 0.4435, + "step": 4645 + }, + { + "epoch": 2.1599256159925617, + "grad_norm": 8.042932510375977, + "learning_rate": 9.9953569995357e-06, + "loss": 0.5478, + "step": 4646 + }, + { + "epoch": 2.1603905160390515, + "grad_norm": 4.940004825592041, + "learning_rate": 9.995355999535601e-06, + "loss": 0.3295, + "step": 4647 + }, + { + "epoch": 2.1608554160855418, + "grad_norm": 4.5640764236450195, + "learning_rate": 9.9953549995355e-06, + "loss": 0.3151, + "step": 4648 + }, + { + "epoch": 2.1613203161320316, + "grad_norm": 6.976424217224121, + "learning_rate": 9.9953539995354e-06, + "loss": 0.6376, + "step": 4649 + }, + { + "epoch": 2.161785216178522, + "grad_norm": 4.40910005569458, + "learning_rate": 9.9953529995353e-06, + "loss": 0.3617, + "step": 4650 + }, + { + "epoch": 2.1622501162250116, + "grad_norm": 3.6116981506347656, + "learning_rate": 9.9953519995352e-06, + "loss": 0.2406, + "step": 4651 + }, + { + "epoch": 2.1627150162715014, + "grad_norm": 5.93511962890625, + "learning_rate": 9.995350999535101e-06, + "loss": 0.5769, + "step": 4652 + }, + { + "epoch": 2.1631799163179917, + "grad_norm": 4.076377868652344, + "learning_rate": 9.995349999535002e-06, + "loss": 0.3019, + "step": 4653 + }, + { + "epoch": 2.1636448163644815, + "grad_norm": 6.238147735595703, + "learning_rate": 9.9953489995349e-06, + "loss": 0.7753, + "step": 4654 + }, + { + "epoch": 2.1641097164109717, + "grad_norm": 4.472458839416504, + "learning_rate": 9.995347999534801e-06, + "loss": 0.4207, + "step": 4655 + }, + { + "epoch": 2.1645746164574615, + "grad_norm": 5.4903717041015625, + "learning_rate": 9.9953469995347e-06, + "loss": 0.5731, + "step": 4656 + }, + { + "epoch": 2.1650395165039518, + "grad_norm": 3.4855666160583496, + "learning_rate": 9.995345999534601e-06, + "loss": 0.1802, + "step": 4657 + }, + { + "epoch": 2.1655044165504416, + "grad_norm": 5.310297966003418, + "learning_rate": 9.995344999534502e-06, + "loss": 0.6076, + "step": 4658 + }, + { + "epoch": 2.165969316596932, + "grad_norm": 8.832290649414062, + "learning_rate": 9.9953439995344e-06, + "loss": 0.4044, + "step": 4659 + }, + { + "epoch": 2.1664342166434216, + "grad_norm": 7.357120037078857, + "learning_rate": 9.995342999534301e-06, + "loss": 0.2815, + "step": 4660 + }, + { + "epoch": 2.166899116689912, + "grad_norm": 9.800457000732422, + "learning_rate": 9.9953419995342e-06, + "loss": 0.6458, + "step": 4661 + }, + { + "epoch": 2.1673640167364017, + "grad_norm": 5.577800273895264, + "learning_rate": 9.995340999534101e-06, + "loss": 0.4007, + "step": 4662 + }, + { + "epoch": 2.1678289167828915, + "grad_norm": 7.135362148284912, + "learning_rate": 9.995339999534e-06, + "loss": 0.7449, + "step": 4663 + }, + { + "epoch": 2.1682938168293817, + "grad_norm": 5.164811134338379, + "learning_rate": 9.9953389995339e-06, + "loss": 0.5047, + "step": 4664 + }, + { + "epoch": 2.1687587168758715, + "grad_norm": 3.923771619796753, + "learning_rate": 9.995337999533801e-06, + "loss": 0.27, + "step": 4665 + }, + { + "epoch": 2.1692236169223618, + "grad_norm": 5.828920841217041, + "learning_rate": 9.995336999533702e-06, + "loss": 0.4054, + "step": 4666 + }, + { + "epoch": 2.1696885169688516, + "grad_norm": 5.406777858734131, + "learning_rate": 9.9953359995336e-06, + "loss": 0.5975, + "step": 4667 + }, + { + "epoch": 2.170153417015342, + "grad_norm": 4.112321853637695, + "learning_rate": 9.9953349995335e-06, + "loss": 0.3912, + "step": 4668 + }, + { + "epoch": 2.1706183170618316, + "grad_norm": 3.201089859008789, + "learning_rate": 9.9953339995334e-06, + "loss": 0.2541, + "step": 4669 + }, + { + "epoch": 2.171083217108322, + "grad_norm": 5.303885459899902, + "learning_rate": 9.995332999533301e-06, + "loss": 0.5246, + "step": 4670 + }, + { + "epoch": 2.1715481171548117, + "grad_norm": 3.3147356510162354, + "learning_rate": 9.995331999533202e-06, + "loss": 0.2008, + "step": 4671 + }, + { + "epoch": 2.172013017201302, + "grad_norm": 8.37025260925293, + "learning_rate": 9.9953309995331e-06, + "loss": 1.1097, + "step": 4672 + }, + { + "epoch": 2.1724779172477917, + "grad_norm": 4.656601428985596, + "learning_rate": 9.995329999533001e-06, + "loss": 0.3817, + "step": 4673 + }, + { + "epoch": 2.172942817294282, + "grad_norm": 4.510209560394287, + "learning_rate": 9.9953289995329e-06, + "loss": 0.2554, + "step": 4674 + }, + { + "epoch": 2.1734077173407718, + "grad_norm": 4.854275226593018, + "learning_rate": 9.995327999532801e-06, + "loss": 0.4306, + "step": 4675 + }, + { + "epoch": 2.1738726173872616, + "grad_norm": 7.832681179046631, + "learning_rate": 9.9953269995327e-06, + "loss": 0.8256, + "step": 4676 + }, + { + "epoch": 2.174337517433752, + "grad_norm": 5.169404029846191, + "learning_rate": 9.9953259995326e-06, + "loss": 0.3572, + "step": 4677 + }, + { + "epoch": 2.1748024174802416, + "grad_norm": 5.486791610717773, + "learning_rate": 9.995324999532501e-06, + "loss": 0.5839, + "step": 4678 + }, + { + "epoch": 2.175267317526732, + "grad_norm": 7.347311496734619, + "learning_rate": 9.9953239995324e-06, + "loss": 0.6899, + "step": 4679 + }, + { + "epoch": 2.1757322175732217, + "grad_norm": 5.2112956047058105, + "learning_rate": 9.9953229995323e-06, + "loss": 0.3132, + "step": 4680 + }, + { + "epoch": 2.176197117619712, + "grad_norm": 6.361586093902588, + "learning_rate": 9.9953219995322e-06, + "loss": 0.4706, + "step": 4681 + }, + { + "epoch": 2.1766620176662017, + "grad_norm": 9.567825317382812, + "learning_rate": 9.9953209995321e-06, + "loss": 0.3675, + "step": 4682 + }, + { + "epoch": 2.177126917712692, + "grad_norm": 4.820364475250244, + "learning_rate": 9.995319999532001e-06, + "loss": 0.5342, + "step": 4683 + }, + { + "epoch": 2.1775918177591818, + "grad_norm": 7.602067470550537, + "learning_rate": 9.995318999531902e-06, + "loss": 0.6236, + "step": 4684 + }, + { + "epoch": 2.1780567178056716, + "grad_norm": 3.27824330329895, + "learning_rate": 9.9953179995318e-06, + "loss": 0.1736, + "step": 4685 + }, + { + "epoch": 2.178521617852162, + "grad_norm": 4.495482921600342, + "learning_rate": 9.995316999531701e-06, + "loss": 0.4255, + "step": 4686 + }, + { + "epoch": 2.1789865178986516, + "grad_norm": 5.479061603546143, + "learning_rate": 9.9953159995316e-06, + "loss": 0.6091, + "step": 4687 + }, + { + "epoch": 2.179451417945142, + "grad_norm": 5.873513221740723, + "learning_rate": 9.995314999531501e-06, + "loss": 0.5266, + "step": 4688 + }, + { + "epoch": 2.1799163179916317, + "grad_norm": 4.636577606201172, + "learning_rate": 9.9953139995314e-06, + "loss": 0.4149, + "step": 4689 + }, + { + "epoch": 2.180381218038122, + "grad_norm": 6.931615352630615, + "learning_rate": 9.9953129995313e-06, + "loss": 0.9364, + "step": 4690 + }, + { + "epoch": 2.1808461180846117, + "grad_norm": 5.896842956542969, + "learning_rate": 9.995311999531201e-06, + "loss": 0.4916, + "step": 4691 + }, + { + "epoch": 2.181311018131102, + "grad_norm": 4.996251106262207, + "learning_rate": 9.9953109995311e-06, + "loss": 0.4165, + "step": 4692 + }, + { + "epoch": 2.1817759181775918, + "grad_norm": 6.051182746887207, + "learning_rate": 9.995309999531001e-06, + "loss": 0.6839, + "step": 4693 + }, + { + "epoch": 2.182240818224082, + "grad_norm": 5.308141231536865, + "learning_rate": 9.9953089995309e-06, + "loss": 0.4905, + "step": 4694 + }, + { + "epoch": 2.182705718270572, + "grad_norm": 8.122102737426758, + "learning_rate": 9.9953079995308e-06, + "loss": 0.4088, + "step": 4695 + }, + { + "epoch": 2.183170618317062, + "grad_norm": 5.190510272979736, + "learning_rate": 9.995306999530701e-06, + "loss": 0.451, + "step": 4696 + }, + { + "epoch": 2.183635518363552, + "grad_norm": 5.479786396026611, + "learning_rate": 9.995305999530602e-06, + "loss": 0.4375, + "step": 4697 + }, + { + "epoch": 2.1841004184100417, + "grad_norm": 10.74671459197998, + "learning_rate": 9.9953049995305e-06, + "loss": 0.6268, + "step": 4698 + }, + { + "epoch": 2.184565318456532, + "grad_norm": 9.532346725463867, + "learning_rate": 9.995303999530401e-06, + "loss": 0.846, + "step": 4699 + }, + { + "epoch": 2.1850302185030217, + "grad_norm": 7.589521408081055, + "learning_rate": 9.9953029995303e-06, + "loss": 0.7258, + "step": 4700 + }, + { + "epoch": 2.185495118549512, + "grad_norm": 5.223341941833496, + "learning_rate": 9.995301999530201e-06, + "loss": 0.6807, + "step": 4701 + }, + { + "epoch": 2.1859600185960018, + "grad_norm": 5.869809627532959, + "learning_rate": 9.9953009995301e-06, + "loss": 0.301, + "step": 4702 + }, + { + "epoch": 2.186424918642492, + "grad_norm": 3.986255407333374, + "learning_rate": 9.99529999953e-06, + "loss": 0.2765, + "step": 4703 + }, + { + "epoch": 2.186889818688982, + "grad_norm": 12.054126739501953, + "learning_rate": 9.995298999529901e-06, + "loss": 0.7728, + "step": 4704 + }, + { + "epoch": 2.187354718735472, + "grad_norm": 5.388518333435059, + "learning_rate": 9.9952979995298e-06, + "loss": 0.4256, + "step": 4705 + }, + { + "epoch": 2.187819618781962, + "grad_norm": 6.378868579864502, + "learning_rate": 9.995296999529701e-06, + "loss": 0.5159, + "step": 4706 + }, + { + "epoch": 2.1882845188284517, + "grad_norm": 3.4802162647247314, + "learning_rate": 9.9952959995296e-06, + "loss": 0.2792, + "step": 4707 + }, + { + "epoch": 2.188749418874942, + "grad_norm": 3.4714813232421875, + "learning_rate": 9.9952949995295e-06, + "loss": 0.1492, + "step": 4708 + }, + { + "epoch": 2.1892143189214317, + "grad_norm": 6.194065093994141, + "learning_rate": 9.995293999529401e-06, + "loss": 0.5048, + "step": 4709 + }, + { + "epoch": 2.189679218967922, + "grad_norm": 5.252769947052002, + "learning_rate": 9.995292999529302e-06, + "loss": 0.6341, + "step": 4710 + }, + { + "epoch": 2.1901441190144117, + "grad_norm": 4.678628921508789, + "learning_rate": 9.995291999529201e-06, + "loss": 0.3969, + "step": 4711 + }, + { + "epoch": 2.190609019060902, + "grad_norm": 3.5408871173858643, + "learning_rate": 9.9952909995291e-06, + "loss": 0.2385, + "step": 4712 + }, + { + "epoch": 2.191073919107392, + "grad_norm": 4.40771484375, + "learning_rate": 9.995289999529e-06, + "loss": 0.3433, + "step": 4713 + }, + { + "epoch": 2.191538819153882, + "grad_norm": 4.637880802154541, + "learning_rate": 9.995288999528901e-06, + "loss": 0.5437, + "step": 4714 + }, + { + "epoch": 2.192003719200372, + "grad_norm": 5.146994590759277, + "learning_rate": 9.995287999528802e-06, + "loss": 0.4889, + "step": 4715 + }, + { + "epoch": 2.192468619246862, + "grad_norm": 6.523789405822754, + "learning_rate": 9.9952869995287e-06, + "loss": 0.4369, + "step": 4716 + }, + { + "epoch": 2.192933519293352, + "grad_norm": 5.291518211364746, + "learning_rate": 9.995285999528601e-06, + "loss": 0.4015, + "step": 4717 + }, + { + "epoch": 2.193398419339842, + "grad_norm": 5.236917972564697, + "learning_rate": 9.9952849995285e-06, + "loss": 0.3931, + "step": 4718 + }, + { + "epoch": 2.193863319386332, + "grad_norm": 5.887327194213867, + "learning_rate": 9.995283999528401e-06, + "loss": 0.4398, + "step": 4719 + }, + { + "epoch": 2.1943282194328217, + "grad_norm": 4.925662517547607, + "learning_rate": 9.9952829995283e-06, + "loss": 0.239, + "step": 4720 + }, + { + "epoch": 2.194793119479312, + "grad_norm": 9.391730308532715, + "learning_rate": 9.9952819995282e-06, + "loss": 0.4372, + "step": 4721 + }, + { + "epoch": 2.195258019525802, + "grad_norm": 7.326934337615967, + "learning_rate": 9.995280999528101e-06, + "loss": 0.7084, + "step": 4722 + }, + { + "epoch": 2.195722919572292, + "grad_norm": 6.210968017578125, + "learning_rate": 9.995279999528e-06, + "loss": 0.7565, + "step": 4723 + }, + { + "epoch": 2.196187819618782, + "grad_norm": 5.368100166320801, + "learning_rate": 9.995278999527901e-06, + "loss": 0.5149, + "step": 4724 + }, + { + "epoch": 2.196652719665272, + "grad_norm": 7.722994804382324, + "learning_rate": 9.9952779995278e-06, + "loss": 0.5167, + "step": 4725 + }, + { + "epoch": 2.197117619711762, + "grad_norm": 5.456569194793701, + "learning_rate": 9.9952769995277e-06, + "loss": 0.7952, + "step": 4726 + }, + { + "epoch": 2.197582519758252, + "grad_norm": 5.373854637145996, + "learning_rate": 9.995275999527601e-06, + "loss": 0.4625, + "step": 4727 + }, + { + "epoch": 2.198047419804742, + "grad_norm": 7.402164459228516, + "learning_rate": 9.995274999527502e-06, + "loss": 0.7927, + "step": 4728 + }, + { + "epoch": 2.198512319851232, + "grad_norm": 3.8194878101348877, + "learning_rate": 9.995273999527401e-06, + "loss": 0.38, + "step": 4729 + }, + { + "epoch": 2.198977219897722, + "grad_norm": 6.6133317947387695, + "learning_rate": 9.995272999527302e-06, + "loss": 0.7481, + "step": 4730 + }, + { + "epoch": 2.199442119944212, + "grad_norm": 5.47524881362915, + "learning_rate": 9.9952719995272e-06, + "loss": 0.576, + "step": 4731 + }, + { + "epoch": 2.199907019990702, + "grad_norm": 6.528716087341309, + "learning_rate": 9.9952709995271e-06, + "loss": 0.5885, + "step": 4732 + }, + { + "epoch": 2.200371920037192, + "grad_norm": 6.098817348480225, + "learning_rate": 9.995269999527e-06, + "loss": 0.4612, + "step": 4733 + }, + { + "epoch": 2.200836820083682, + "grad_norm": 4.150088787078857, + "learning_rate": 9.9952689995269e-06, + "loss": 0.2828, + "step": 4734 + }, + { + "epoch": 2.201301720130172, + "grad_norm": 5.7722272872924805, + "learning_rate": 9.995267999526801e-06, + "loss": 0.4826, + "step": 4735 + }, + { + "epoch": 2.201766620176662, + "grad_norm": 4.883697032928467, + "learning_rate": 9.9952669995267e-06, + "loss": 0.3719, + "step": 4736 + }, + { + "epoch": 2.202231520223152, + "grad_norm": 6.498202800750732, + "learning_rate": 9.995265999526601e-06, + "loss": 0.4246, + "step": 4737 + }, + { + "epoch": 2.202696420269642, + "grad_norm": 4.401515960693359, + "learning_rate": 9.9952649995265e-06, + "loss": 0.3025, + "step": 4738 + }, + { + "epoch": 2.203161320316132, + "grad_norm": 6.925644397735596, + "learning_rate": 9.9952639995264e-06, + "loss": 0.9836, + "step": 4739 + }, + { + "epoch": 2.2036262203626222, + "grad_norm": 8.135836601257324, + "learning_rate": 9.995262999526301e-06, + "loss": 0.4661, + "step": 4740 + }, + { + "epoch": 2.204091120409112, + "grad_norm": 4.7579731941223145, + "learning_rate": 9.995261999526202e-06, + "loss": 0.2677, + "step": 4741 + }, + { + "epoch": 2.204556020455602, + "grad_norm": 3.526810646057129, + "learning_rate": 9.995260999526101e-06, + "loss": 0.3051, + "step": 4742 + }, + { + "epoch": 2.205020920502092, + "grad_norm": 4.862575054168701, + "learning_rate": 9.995259999526e-06, + "loss": 0.2947, + "step": 4743 + }, + { + "epoch": 2.205485820548582, + "grad_norm": 5.849588871002197, + "learning_rate": 9.9952589995259e-06, + "loss": 0.333, + "step": 4744 + }, + { + "epoch": 2.205950720595072, + "grad_norm": 5.267209053039551, + "learning_rate": 9.9952579995258e-06, + "loss": 0.5968, + "step": 4745 + }, + { + "epoch": 2.206415620641562, + "grad_norm": 7.925967216491699, + "learning_rate": 9.9952569995257e-06, + "loss": 0.4586, + "step": 4746 + }, + { + "epoch": 2.206880520688052, + "grad_norm": 5.556106090545654, + "learning_rate": 9.995255999525601e-06, + "loss": 0.378, + "step": 4747 + }, + { + "epoch": 2.207345420734542, + "grad_norm": 6.813588619232178, + "learning_rate": 9.995254999525502e-06, + "loss": 0.4865, + "step": 4748 + }, + { + "epoch": 2.2078103207810322, + "grad_norm": 6.45041561126709, + "learning_rate": 9.9952539995254e-06, + "loss": 0.4288, + "step": 4749 + }, + { + "epoch": 2.208275220827522, + "grad_norm": 3.706378936767578, + "learning_rate": 9.995252999525301e-06, + "loss": 0.2312, + "step": 4750 + }, + { + "epoch": 2.2087401208740123, + "grad_norm": 8.063298225402832, + "learning_rate": 9.9952519995252e-06, + "loss": 0.5998, + "step": 4751 + }, + { + "epoch": 2.209205020920502, + "grad_norm": 3.4566588401794434, + "learning_rate": 9.9952509995251e-06, + "loss": 0.2926, + "step": 4752 + }, + { + "epoch": 2.2096699209669923, + "grad_norm": 4.402578353881836, + "learning_rate": 9.995249999525001e-06, + "loss": 0.2143, + "step": 4753 + }, + { + "epoch": 2.210134821013482, + "grad_norm": 9.2678861618042, + "learning_rate": 9.9952489995249e-06, + "loss": 0.3796, + "step": 4754 + }, + { + "epoch": 2.210599721059972, + "grad_norm": 4.706636905670166, + "learning_rate": 9.995247999524801e-06, + "loss": 0.5175, + "step": 4755 + }, + { + "epoch": 2.211064621106462, + "grad_norm": 7.069522380828857, + "learning_rate": 9.9952469995247e-06, + "loss": 0.7204, + "step": 4756 + }, + { + "epoch": 2.211529521152952, + "grad_norm": 4.9172797203063965, + "learning_rate": 9.9952459995246e-06, + "loss": 0.3919, + "step": 4757 + }, + { + "epoch": 2.2119944211994422, + "grad_norm": 4.506686687469482, + "learning_rate": 9.9952449995245e-06, + "loss": 0.3302, + "step": 4758 + }, + { + "epoch": 2.212459321245932, + "grad_norm": 5.06798791885376, + "learning_rate": 9.9952439995244e-06, + "loss": 0.3628, + "step": 4759 + }, + { + "epoch": 2.2129242212924223, + "grad_norm": 3.648850917816162, + "learning_rate": 9.995242999524301e-06, + "loss": 0.3377, + "step": 4760 + }, + { + "epoch": 2.213389121338912, + "grad_norm": 5.674792289733887, + "learning_rate": 9.995241999524202e-06, + "loss": 0.3093, + "step": 4761 + }, + { + "epoch": 2.2138540213854023, + "grad_norm": 4.60029935836792, + "learning_rate": 9.9952409995241e-06, + "loss": 0.3116, + "step": 4762 + }, + { + "epoch": 2.214318921431892, + "grad_norm": 4.093806743621826, + "learning_rate": 9.995239999524001e-06, + "loss": 0.29, + "step": 4763 + }, + { + "epoch": 2.214783821478382, + "grad_norm": 4.695716857910156, + "learning_rate": 9.9952389995239e-06, + "loss": 0.2734, + "step": 4764 + }, + { + "epoch": 2.215248721524872, + "grad_norm": 6.220455646514893, + "learning_rate": 9.995237999523801e-06, + "loss": 0.57, + "step": 4765 + }, + { + "epoch": 2.215713621571362, + "grad_norm": 6.565047740936279, + "learning_rate": 9.995236999523702e-06, + "loss": 0.5457, + "step": 4766 + }, + { + "epoch": 2.2161785216178522, + "grad_norm": 4.100856304168701, + "learning_rate": 9.9952359995236e-06, + "loss": 0.2463, + "step": 4767 + }, + { + "epoch": 2.216643421664342, + "grad_norm": 4.732679843902588, + "learning_rate": 9.995234999523501e-06, + "loss": 0.4108, + "step": 4768 + }, + { + "epoch": 2.2171083217108323, + "grad_norm": 6.517456531524658, + "learning_rate": 9.9952339995234e-06, + "loss": 0.6161, + "step": 4769 + }, + { + "epoch": 2.217573221757322, + "grad_norm": 3.0325639247894287, + "learning_rate": 9.9952329995233e-06, + "loss": 0.1541, + "step": 4770 + }, + { + "epoch": 2.2180381218038123, + "grad_norm": 6.377007484436035, + "learning_rate": 9.9952319995232e-06, + "loss": 0.3551, + "step": 4771 + }, + { + "epoch": 2.218503021850302, + "grad_norm": 8.213379859924316, + "learning_rate": 9.9952309995231e-06, + "loss": 0.8438, + "step": 4772 + }, + { + "epoch": 2.2189679218967924, + "grad_norm": 7.12801456451416, + "learning_rate": 9.995229999523001e-06, + "loss": 0.7402, + "step": 4773 + }, + { + "epoch": 2.219432821943282, + "grad_norm": 5.606762409210205, + "learning_rate": 9.995228999522902e-06, + "loss": 0.6499, + "step": 4774 + }, + { + "epoch": 2.2198977219897724, + "grad_norm": 6.323250770568848, + "learning_rate": 9.9952279995228e-06, + "loss": 0.6406, + "step": 4775 + }, + { + "epoch": 2.2203626220362622, + "grad_norm": 4.826825141906738, + "learning_rate": 9.9952269995227e-06, + "loss": 0.3928, + "step": 4776 + }, + { + "epoch": 2.220827522082752, + "grad_norm": 5.0407328605651855, + "learning_rate": 9.9952259995226e-06, + "loss": 0.3655, + "step": 4777 + }, + { + "epoch": 2.2212924221292423, + "grad_norm": 3.3398382663726807, + "learning_rate": 9.995224999522501e-06, + "loss": 0.1562, + "step": 4778 + }, + { + "epoch": 2.221757322175732, + "grad_norm": 6.9389777183532715, + "learning_rate": 9.995223999522402e-06, + "loss": 0.9421, + "step": 4779 + }, + { + "epoch": 2.2222222222222223, + "grad_norm": 7.005780220031738, + "learning_rate": 9.9952229995223e-06, + "loss": 0.3808, + "step": 4780 + }, + { + "epoch": 2.222687122268712, + "grad_norm": 3.6752591133117676, + "learning_rate": 9.995221999522201e-06, + "loss": 0.2597, + "step": 4781 + }, + { + "epoch": 2.2231520223152024, + "grad_norm": 3.724236249923706, + "learning_rate": 9.9952209995221e-06, + "loss": 0.2145, + "step": 4782 + }, + { + "epoch": 2.223616922361692, + "grad_norm": 6.740995407104492, + "learning_rate": 9.995219999522e-06, + "loss": 0.5088, + "step": 4783 + }, + { + "epoch": 2.2240818224081824, + "grad_norm": 5.249281883239746, + "learning_rate": 9.995218999521901e-06, + "loss": 0.5029, + "step": 4784 + }, + { + "epoch": 2.224546722454672, + "grad_norm": 7.162242889404297, + "learning_rate": 9.9952179995218e-06, + "loss": 0.52, + "step": 4785 + }, + { + "epoch": 2.225011622501162, + "grad_norm": 12.878426551818848, + "learning_rate": 9.995216999521701e-06, + "loss": 0.5136, + "step": 4786 + }, + { + "epoch": 2.2254765225476523, + "grad_norm": 10.092291831970215, + "learning_rate": 9.9952159995216e-06, + "loss": 0.5083, + "step": 4787 + }, + { + "epoch": 2.225941422594142, + "grad_norm": 7.279494285583496, + "learning_rate": 9.9952149995215e-06, + "loss": 0.5433, + "step": 4788 + }, + { + "epoch": 2.2264063226406323, + "grad_norm": 5.5762505531311035, + "learning_rate": 9.9952139995214e-06, + "loss": 0.4863, + "step": 4789 + }, + { + "epoch": 2.226871222687122, + "grad_norm": 3.7529077529907227, + "learning_rate": 9.9952129995213e-06, + "loss": 0.3988, + "step": 4790 + }, + { + "epoch": 2.2273361227336124, + "grad_norm": 6.389484882354736, + "learning_rate": 9.995211999521201e-06, + "loss": 0.3873, + "step": 4791 + }, + { + "epoch": 2.227801022780102, + "grad_norm": 8.680354118347168, + "learning_rate": 9.995210999521102e-06, + "loss": 0.8477, + "step": 4792 + }, + { + "epoch": 2.2282659228265924, + "grad_norm": 7.271360397338867, + "learning_rate": 9.995209999521e-06, + "loss": 0.4449, + "step": 4793 + }, + { + "epoch": 2.228730822873082, + "grad_norm": 4.525271415710449, + "learning_rate": 9.995208999520901e-06, + "loss": 0.4465, + "step": 4794 + }, + { + "epoch": 2.2291957229195725, + "grad_norm": 6.6604485511779785, + "learning_rate": 9.9952079995208e-06, + "loss": 0.5052, + "step": 4795 + }, + { + "epoch": 2.2296606229660623, + "grad_norm": 4.5107316970825195, + "learning_rate": 9.995206999520701e-06, + "loss": 0.3739, + "step": 4796 + }, + { + "epoch": 2.2301255230125525, + "grad_norm": 5.2968244552612305, + "learning_rate": 9.995205999520602e-06, + "loss": 0.3025, + "step": 4797 + }, + { + "epoch": 2.2305904230590423, + "grad_norm": 5.187340259552002, + "learning_rate": 9.9952049995205e-06, + "loss": 0.3152, + "step": 4798 + }, + { + "epoch": 2.231055323105532, + "grad_norm": 13.481966018676758, + "learning_rate": 9.995203999520401e-06, + "loss": 0.5992, + "step": 4799 + }, + { + "epoch": 2.2315202231520224, + "grad_norm": 5.464190483093262, + "learning_rate": 9.9952029995203e-06, + "loss": 0.3821, + "step": 4800 + }, + { + "epoch": 2.231985123198512, + "grad_norm": 5.564152717590332, + "learning_rate": 9.9952019995202e-06, + "loss": 0.2949, + "step": 4801 + }, + { + "epoch": 2.2324500232450024, + "grad_norm": 5.094796180725098, + "learning_rate": 9.9952009995201e-06, + "loss": 0.3162, + "step": 4802 + }, + { + "epoch": 2.232914923291492, + "grad_norm": 4.548390865325928, + "learning_rate": 9.99519999952e-06, + "loss": 0.373, + "step": 4803 + }, + { + "epoch": 2.2333798233379825, + "grad_norm": 5.492648601531982, + "learning_rate": 9.995198999519901e-06, + "loss": 0.5996, + "step": 4804 + }, + { + "epoch": 2.2338447233844723, + "grad_norm": 5.262760639190674, + "learning_rate": 9.995197999519802e-06, + "loss": 0.3585, + "step": 4805 + }, + { + "epoch": 2.2343096234309625, + "grad_norm": 5.09044075012207, + "learning_rate": 9.9951969995197e-06, + "loss": 0.3679, + "step": 4806 + }, + { + "epoch": 2.2347745234774523, + "grad_norm": 4.634350776672363, + "learning_rate": 9.995195999519601e-06, + "loss": 0.4114, + "step": 4807 + }, + { + "epoch": 2.2352394235239426, + "grad_norm": 4.927405834197998, + "learning_rate": 9.9951949995195e-06, + "loss": 0.5814, + "step": 4808 + }, + { + "epoch": 2.2357043235704324, + "grad_norm": 5.366201877593994, + "learning_rate": 9.995193999519401e-06, + "loss": 0.501, + "step": 4809 + }, + { + "epoch": 2.236169223616922, + "grad_norm": 4.452559947967529, + "learning_rate": 9.995192999519302e-06, + "loss": 0.3768, + "step": 4810 + }, + { + "epoch": 2.2366341236634124, + "grad_norm": 3.3913793563842773, + "learning_rate": 9.9951919995192e-06, + "loss": 0.1624, + "step": 4811 + }, + { + "epoch": 2.237099023709902, + "grad_norm": 4.070425033569336, + "learning_rate": 9.995190999519101e-06, + "loss": 0.3702, + "step": 4812 + }, + { + "epoch": 2.2375639237563925, + "grad_norm": 5.832078456878662, + "learning_rate": 9.995189999519e-06, + "loss": 0.3898, + "step": 4813 + }, + { + "epoch": 2.2380288238028823, + "grad_norm": 5.2009711265563965, + "learning_rate": 9.995188999518901e-06, + "loss": 0.2155, + "step": 4814 + }, + { + "epoch": 2.2384937238493725, + "grad_norm": 5.490665912628174, + "learning_rate": 9.9951879995188e-06, + "loss": 0.382, + "step": 4815 + }, + { + "epoch": 2.2389586238958623, + "grad_norm": 4.03530216217041, + "learning_rate": 9.9951869995187e-06, + "loss": 0.2634, + "step": 4816 + }, + { + "epoch": 2.2394235239423526, + "grad_norm": 6.066982269287109, + "learning_rate": 9.995185999518601e-06, + "loss": 0.4726, + "step": 4817 + }, + { + "epoch": 2.2398884239888424, + "grad_norm": 5.375803470611572, + "learning_rate": 9.995184999518502e-06, + "loss": 0.3944, + "step": 4818 + }, + { + "epoch": 2.2403533240353326, + "grad_norm": 6.823390007019043, + "learning_rate": 9.9951839995184e-06, + "loss": 0.8272, + "step": 4819 + }, + { + "epoch": 2.2408182240818224, + "grad_norm": 4.096299648284912, + "learning_rate": 9.9951829995183e-06, + "loss": 0.3014, + "step": 4820 + }, + { + "epoch": 2.241283124128312, + "grad_norm": 6.513105869293213, + "learning_rate": 9.9951819995182e-06, + "loss": 0.5556, + "step": 4821 + }, + { + "epoch": 2.2417480241748025, + "grad_norm": 7.095858573913574, + "learning_rate": 9.995180999518101e-06, + "loss": 0.6031, + "step": 4822 + }, + { + "epoch": 2.2422129242212923, + "grad_norm": 4.02918004989624, + "learning_rate": 9.995179999518002e-06, + "loss": 0.4034, + "step": 4823 + }, + { + "epoch": 2.2426778242677825, + "grad_norm": 5.614934921264648, + "learning_rate": 9.9951789995179e-06, + "loss": 0.324, + "step": 4824 + }, + { + "epoch": 2.2431427243142723, + "grad_norm": 7.147292137145996, + "learning_rate": 9.995177999517801e-06, + "loss": 0.3691, + "step": 4825 + }, + { + "epoch": 2.2436076243607626, + "grad_norm": 4.565422058105469, + "learning_rate": 9.9951769995177e-06, + "loss": 0.3918, + "step": 4826 + }, + { + "epoch": 2.2440725244072524, + "grad_norm": 7.685978889465332, + "learning_rate": 9.995175999517601e-06, + "loss": 0.3486, + "step": 4827 + }, + { + "epoch": 2.2445374244537426, + "grad_norm": 4.896158695220947, + "learning_rate": 9.9951749995175e-06, + "loss": 0.2038, + "step": 4828 + }, + { + "epoch": 2.2450023245002324, + "grad_norm": 6.351507186889648, + "learning_rate": 9.9951739995174e-06, + "loss": 0.4549, + "step": 4829 + }, + { + "epoch": 2.2454672245467227, + "grad_norm": 6.007656574249268, + "learning_rate": 9.995172999517301e-06, + "loss": 0.2734, + "step": 4830 + }, + { + "epoch": 2.2459321245932125, + "grad_norm": 5.334366798400879, + "learning_rate": 9.9951719995172e-06, + "loss": 0.5232, + "step": 4831 + }, + { + "epoch": 2.2463970246397027, + "grad_norm": 13.071989059448242, + "learning_rate": 9.995170999517101e-06, + "loss": 0.9049, + "step": 4832 + }, + { + "epoch": 2.2468619246861925, + "grad_norm": 4.792056083679199, + "learning_rate": 9.995169999517e-06, + "loss": 0.3769, + "step": 4833 + }, + { + "epoch": 2.2473268247326823, + "grad_norm": 6.682924270629883, + "learning_rate": 9.9951689995169e-06, + "loss": 0.5833, + "step": 4834 + }, + { + "epoch": 2.2477917247791726, + "grad_norm": 6.8039655685424805, + "learning_rate": 9.995167999516801e-06, + "loss": 0.6447, + "step": 4835 + }, + { + "epoch": 2.2482566248256624, + "grad_norm": 6.856713771820068, + "learning_rate": 9.995166999516702e-06, + "loss": 0.5462, + "step": 4836 + }, + { + "epoch": 2.2487215248721526, + "grad_norm": 4.413552761077881, + "learning_rate": 9.9951659995166e-06, + "loss": 0.3347, + "step": 4837 + }, + { + "epoch": 2.2491864249186424, + "grad_norm": 5.283041000366211, + "learning_rate": 9.995164999516501e-06, + "loss": 0.3604, + "step": 4838 + }, + { + "epoch": 2.2496513249651326, + "grad_norm": 4.832571506500244, + "learning_rate": 9.9951639995164e-06, + "loss": 0.2889, + "step": 4839 + }, + { + "epoch": 2.2501162250116225, + "grad_norm": 6.085830211639404, + "learning_rate": 9.995162999516301e-06, + "loss": 0.7504, + "step": 4840 + }, + { + "epoch": 2.2505811250581127, + "grad_norm": 4.892759323120117, + "learning_rate": 9.9951619995162e-06, + "loss": 0.4089, + "step": 4841 + }, + { + "epoch": 2.2510460251046025, + "grad_norm": 4.958770275115967, + "learning_rate": 9.9951609995161e-06, + "loss": 0.6042, + "step": 4842 + }, + { + "epoch": 2.2515109251510923, + "grad_norm": 6.997579097747803, + "learning_rate": 9.995159999516001e-06, + "loss": 0.6133, + "step": 4843 + }, + { + "epoch": 2.2519758251975825, + "grad_norm": 5.444996356964111, + "learning_rate": 9.9951589995159e-06, + "loss": 0.3762, + "step": 4844 + }, + { + "epoch": 2.2524407252440724, + "grad_norm": 5.3052544593811035, + "learning_rate": 9.995157999515801e-06, + "loss": 0.5221, + "step": 4845 + }, + { + "epoch": 2.2529056252905626, + "grad_norm": 6.515061855316162, + "learning_rate": 9.9951569995157e-06, + "loss": 0.4058, + "step": 4846 + }, + { + "epoch": 2.2533705253370524, + "grad_norm": 4.329367637634277, + "learning_rate": 9.9951559995156e-06, + "loss": 0.3261, + "step": 4847 + }, + { + "epoch": 2.2538354253835426, + "grad_norm": 5.8770551681518555, + "learning_rate": 9.995154999515501e-06, + "loss": 0.4566, + "step": 4848 + }, + { + "epoch": 2.2543003254300324, + "grad_norm": 5.6770524978637695, + "learning_rate": 9.995153999515402e-06, + "loss": 0.8045, + "step": 4849 + }, + { + "epoch": 2.2547652254765227, + "grad_norm": 5.95387601852417, + "learning_rate": 9.995152999515301e-06, + "loss": 0.4771, + "step": 4850 + }, + { + "epoch": 2.2552301255230125, + "grad_norm": 5.760194301605225, + "learning_rate": 9.995151999515202e-06, + "loss": 0.4291, + "step": 4851 + }, + { + "epoch": 2.2556950255695027, + "grad_norm": 4.121948719024658, + "learning_rate": 9.9951509995151e-06, + "loss": 0.2099, + "step": 4852 + }, + { + "epoch": 2.2561599256159925, + "grad_norm": 5.672104358673096, + "learning_rate": 9.995149999515001e-06, + "loss": 0.5097, + "step": 4853 + }, + { + "epoch": 2.256624825662483, + "grad_norm": 5.067205905914307, + "learning_rate": 9.9951489995149e-06, + "loss": 0.385, + "step": 4854 + }, + { + "epoch": 2.2570897257089726, + "grad_norm": 7.463171005249023, + "learning_rate": 9.9951479995148e-06, + "loss": 0.4492, + "step": 4855 + }, + { + "epoch": 2.2575546257554624, + "grad_norm": 2.5146942138671875, + "learning_rate": 9.995146999514701e-06, + "loss": 0.1245, + "step": 4856 + }, + { + "epoch": 2.2580195258019526, + "grad_norm": 6.110436916351318, + "learning_rate": 9.9951459995146e-06, + "loss": 0.7697, + "step": 4857 + }, + { + "epoch": 2.2584844258484424, + "grad_norm": 6.6458001136779785, + "learning_rate": 9.995144999514501e-06, + "loss": 0.2505, + "step": 4858 + }, + { + "epoch": 2.2589493258949327, + "grad_norm": 4.5757036209106445, + "learning_rate": 9.9951439995144e-06, + "loss": 0.2444, + "step": 4859 + }, + { + "epoch": 2.2594142259414225, + "grad_norm": 4.452184200286865, + "learning_rate": 9.9951429995143e-06, + "loss": 0.3236, + "step": 4860 + }, + { + "epoch": 2.2598791259879127, + "grad_norm": 2.949408769607544, + "learning_rate": 9.995141999514201e-06, + "loss": 0.1373, + "step": 4861 + }, + { + "epoch": 2.2603440260344025, + "grad_norm": 6.372926712036133, + "learning_rate": 9.9951409995141e-06, + "loss": 0.371, + "step": 4862 + }, + { + "epoch": 2.260808926080893, + "grad_norm": 6.935194969177246, + "learning_rate": 9.995139999514001e-06, + "loss": 0.5335, + "step": 4863 + }, + { + "epoch": 2.2612738261273826, + "grad_norm": 4.289972305297852, + "learning_rate": 9.9951389995139e-06, + "loss": 0.3645, + "step": 4864 + }, + { + "epoch": 2.2617387261738724, + "grad_norm": 3.8105926513671875, + "learning_rate": 9.9951379995138e-06, + "loss": 0.2052, + "step": 4865 + }, + { + "epoch": 2.2622036262203626, + "grad_norm": 7.357482433319092, + "learning_rate": 9.995136999513701e-06, + "loss": 0.4809, + "step": 4866 + }, + { + "epoch": 2.2626685262668524, + "grad_norm": 4.5787248611450195, + "learning_rate": 9.9951359995136e-06, + "loss": 0.2595, + "step": 4867 + }, + { + "epoch": 2.2631334263133427, + "grad_norm": 6.3570356369018555, + "learning_rate": 9.9951349995135e-06, + "loss": 0.8529, + "step": 4868 + }, + { + "epoch": 2.2635983263598325, + "grad_norm": 5.119509696960449, + "learning_rate": 9.995133999513402e-06, + "loss": 0.2739, + "step": 4869 + }, + { + "epoch": 2.2640632264063227, + "grad_norm": 4.90076208114624, + "learning_rate": 9.9951329995133e-06, + "loss": 0.5028, + "step": 4870 + }, + { + "epoch": 2.2645281264528125, + "grad_norm": 3.910270929336548, + "learning_rate": 9.995131999513201e-06, + "loss": 0.2189, + "step": 4871 + }, + { + "epoch": 2.264993026499303, + "grad_norm": 5.327520370483398, + "learning_rate": 9.9951309995131e-06, + "loss": 0.3205, + "step": 4872 + }, + { + "epoch": 2.2654579265457926, + "grad_norm": 5.865079402923584, + "learning_rate": 9.995129999513e-06, + "loss": 0.3901, + "step": 4873 + }, + { + "epoch": 2.265922826592283, + "grad_norm": 6.660824298858643, + "learning_rate": 9.995128999512901e-06, + "loss": 0.4656, + "step": 4874 + }, + { + "epoch": 2.2663877266387726, + "grad_norm": 5.722811698913574, + "learning_rate": 9.9951279995128e-06, + "loss": 0.4211, + "step": 4875 + }, + { + "epoch": 2.266852626685263, + "grad_norm": 4.2776265144348145, + "learning_rate": 9.995126999512701e-06, + "loss": 0.382, + "step": 4876 + }, + { + "epoch": 2.2673175267317527, + "grad_norm": 4.2834858894348145, + "learning_rate": 9.9951259995126e-06, + "loss": 0.3673, + "step": 4877 + }, + { + "epoch": 2.2677824267782425, + "grad_norm": 5.2735748291015625, + "learning_rate": 9.9951249995125e-06, + "loss": 0.5543, + "step": 4878 + }, + { + "epoch": 2.2682473268247327, + "grad_norm": 6.850246906280518, + "learning_rate": 9.995123999512401e-06, + "loss": 0.59, + "step": 4879 + }, + { + "epoch": 2.2687122268712225, + "grad_norm": 4.746832370758057, + "learning_rate": 9.995122999512302e-06, + "loss": 0.3907, + "step": 4880 + }, + { + "epoch": 2.269177126917713, + "grad_norm": 5.283148765563965, + "learning_rate": 9.995121999512201e-06, + "loss": 0.4176, + "step": 4881 + }, + { + "epoch": 2.2696420269642026, + "grad_norm": 4.798195838928223, + "learning_rate": 9.995120999512102e-06, + "loss": 0.3846, + "step": 4882 + }, + { + "epoch": 2.270106927010693, + "grad_norm": 11.251285552978516, + "learning_rate": 9.995119999512e-06, + "loss": 0.4579, + "step": 4883 + }, + { + "epoch": 2.2705718270571826, + "grad_norm": 4.59340238571167, + "learning_rate": 9.9951189995119e-06, + "loss": 0.4702, + "step": 4884 + }, + { + "epoch": 2.271036727103673, + "grad_norm": 5.430858612060547, + "learning_rate": 9.9951179995118e-06, + "loss": 0.1192, + "step": 4885 + }, + { + "epoch": 2.2715016271501627, + "grad_norm": 5.211923599243164, + "learning_rate": 9.9951169995117e-06, + "loss": 0.2031, + "step": 4886 + }, + { + "epoch": 2.2719665271966525, + "grad_norm": 2.8279318809509277, + "learning_rate": 9.995115999511601e-06, + "loss": 0.1719, + "step": 4887 + }, + { + "epoch": 2.2724314272431427, + "grad_norm": 4.84147834777832, + "learning_rate": 9.9951149995115e-06, + "loss": 0.3313, + "step": 4888 + }, + { + "epoch": 2.272896327289633, + "grad_norm": 5.333591461181641, + "learning_rate": 9.995113999511401e-06, + "loss": 0.3986, + "step": 4889 + }, + { + "epoch": 2.273361227336123, + "grad_norm": 6.519214630126953, + "learning_rate": 9.9951129995113e-06, + "loss": 0.7745, + "step": 4890 + }, + { + "epoch": 2.2738261273826126, + "grad_norm": 6.665240287780762, + "learning_rate": 9.9951119995112e-06, + "loss": 0.6718, + "step": 4891 + }, + { + "epoch": 2.274291027429103, + "grad_norm": 5.170334339141846, + "learning_rate": 9.995110999511101e-06, + "loss": 0.3727, + "step": 4892 + }, + { + "epoch": 2.2747559274755926, + "grad_norm": 5.2417497634887695, + "learning_rate": 9.995109999511002e-06, + "loss": 0.4732, + "step": 4893 + }, + { + "epoch": 2.275220827522083, + "grad_norm": 5.548876762390137, + "learning_rate": 9.995108999510901e-06, + "loss": 0.6347, + "step": 4894 + }, + { + "epoch": 2.2756857275685727, + "grad_norm": 5.6656599044799805, + "learning_rate": 9.9951079995108e-06, + "loss": 0.6537, + "step": 4895 + }, + { + "epoch": 2.276150627615063, + "grad_norm": 3.3486411571502686, + "learning_rate": 9.9951069995107e-06, + "loss": 0.1654, + "step": 4896 + }, + { + "epoch": 2.2766155276615527, + "grad_norm": 3.8743743896484375, + "learning_rate": 9.9951059995106e-06, + "loss": 0.3386, + "step": 4897 + }, + { + "epoch": 2.277080427708043, + "grad_norm": 3.7226710319519043, + "learning_rate": 9.9951049995105e-06, + "loss": 0.3328, + "step": 4898 + }, + { + "epoch": 2.2775453277545328, + "grad_norm": 6.595566749572754, + "learning_rate": 9.995103999510401e-06, + "loss": 0.8321, + "step": 4899 + }, + { + "epoch": 2.2780102278010226, + "grad_norm": 5.534704685211182, + "learning_rate": 9.995102999510302e-06, + "loss": 0.3631, + "step": 4900 + }, + { + "epoch": 2.278475127847513, + "grad_norm": 5.047486782073975, + "learning_rate": 9.9951019995102e-06, + "loss": 0.3879, + "step": 4901 + }, + { + "epoch": 2.2789400278940026, + "grad_norm": 7.500573635101318, + "learning_rate": 9.995100999510101e-06, + "loss": 0.3687, + "step": 4902 + }, + { + "epoch": 2.279404927940493, + "grad_norm": 4.5807647705078125, + "learning_rate": 9.99509999951e-06, + "loss": 0.3121, + "step": 4903 + }, + { + "epoch": 2.2798698279869827, + "grad_norm": 4.527140140533447, + "learning_rate": 9.9950989995099e-06, + "loss": 0.4092, + "step": 4904 + }, + { + "epoch": 2.280334728033473, + "grad_norm": 4.467066764831543, + "learning_rate": 9.995097999509801e-06, + "loss": 0.4576, + "step": 4905 + }, + { + "epoch": 2.2807996280799627, + "grad_norm": 6.230132102966309, + "learning_rate": 9.9950969995097e-06, + "loss": 0.4477, + "step": 4906 + }, + { + "epoch": 2.281264528126453, + "grad_norm": 10.033646583557129, + "learning_rate": 9.995095999509601e-06, + "loss": 0.7249, + "step": 4907 + }, + { + "epoch": 2.2817294281729428, + "grad_norm": 3.7723612785339355, + "learning_rate": 9.9950949995095e-06, + "loss": 0.2417, + "step": 4908 + }, + { + "epoch": 2.2821943282194326, + "grad_norm": 4.2674994468688965, + "learning_rate": 9.9950939995094e-06, + "loss": 0.4998, + "step": 4909 + }, + { + "epoch": 2.282659228265923, + "grad_norm": 5.153303146362305, + "learning_rate": 9.9950929995093e-06, + "loss": 0.4478, + "step": 4910 + }, + { + "epoch": 2.283124128312413, + "grad_norm": 5.130427837371826, + "learning_rate": 9.9950919995092e-06, + "loss": 0.5898, + "step": 4911 + }, + { + "epoch": 2.283589028358903, + "grad_norm": 4.174790859222412, + "learning_rate": 9.995090999509101e-06, + "loss": 0.409, + "step": 4912 + }, + { + "epoch": 2.2840539284053927, + "grad_norm": 5.677173137664795, + "learning_rate": 9.995089999509002e-06, + "loss": 0.316, + "step": 4913 + }, + { + "epoch": 2.284518828451883, + "grad_norm": 4.10029411315918, + "learning_rate": 9.9950889995089e-06, + "loss": 0.3362, + "step": 4914 + }, + { + "epoch": 2.2849837284983727, + "grad_norm": 7.6214118003845215, + "learning_rate": 9.995087999508801e-06, + "loss": 0.5358, + "step": 4915 + }, + { + "epoch": 2.285448628544863, + "grad_norm": 5.245863914489746, + "learning_rate": 9.9950869995087e-06, + "loss": 0.4627, + "step": 4916 + }, + { + "epoch": 2.2859135285913528, + "grad_norm": 5.362049102783203, + "learning_rate": 9.995085999508601e-06, + "loss": 0.4456, + "step": 4917 + }, + { + "epoch": 2.286378428637843, + "grad_norm": 6.709080696105957, + "learning_rate": 9.995084999508502e-06, + "loss": 0.5712, + "step": 4918 + }, + { + "epoch": 2.286843328684333, + "grad_norm": 5.235257148742676, + "learning_rate": 9.9950839995084e-06, + "loss": 0.5454, + "step": 4919 + }, + { + "epoch": 2.287308228730823, + "grad_norm": 4.868673324584961, + "learning_rate": 9.995082999508301e-06, + "loss": 0.3342, + "step": 4920 + }, + { + "epoch": 2.287773128777313, + "grad_norm": 4.807116508483887, + "learning_rate": 9.9950819995082e-06, + "loss": 0.3468, + "step": 4921 + }, + { + "epoch": 2.2882380288238027, + "grad_norm": 9.181120872497559, + "learning_rate": 9.9950809995081e-06, + "loss": 0.8709, + "step": 4922 + }, + { + "epoch": 2.288702928870293, + "grad_norm": 4.504263877868652, + "learning_rate": 9.995079999508e-06, + "loss": 0.5018, + "step": 4923 + }, + { + "epoch": 2.2891678289167827, + "grad_norm": 7.605844974517822, + "learning_rate": 9.9950789995079e-06, + "loss": 0.9482, + "step": 4924 + }, + { + "epoch": 2.289632728963273, + "grad_norm": 5.298529148101807, + "learning_rate": 9.995077999507801e-06, + "loss": 0.3307, + "step": 4925 + }, + { + "epoch": 2.2900976290097628, + "grad_norm": 5.163459300994873, + "learning_rate": 9.995076999507702e-06, + "loss": 0.4937, + "step": 4926 + }, + { + "epoch": 2.290562529056253, + "grad_norm": 3.962120532989502, + "learning_rate": 9.9950759995076e-06, + "loss": 0.159, + "step": 4927 + }, + { + "epoch": 2.291027429102743, + "grad_norm": 5.859421253204346, + "learning_rate": 9.9950749995075e-06, + "loss": 0.3201, + "step": 4928 + }, + { + "epoch": 2.291492329149233, + "grad_norm": 7.0709428787231445, + "learning_rate": 9.9950739995074e-06, + "loss": 0.5704, + "step": 4929 + }, + { + "epoch": 2.291957229195723, + "grad_norm": 5.467860221862793, + "learning_rate": 9.995072999507301e-06, + "loss": 0.5673, + "step": 4930 + }, + { + "epoch": 2.292422129242213, + "grad_norm": 8.530523300170898, + "learning_rate": 9.995071999507202e-06, + "loss": 0.9096, + "step": 4931 + }, + { + "epoch": 2.292887029288703, + "grad_norm": 5.296736717224121, + "learning_rate": 9.9950709995071e-06, + "loss": 0.6698, + "step": 4932 + }, + { + "epoch": 2.293351929335193, + "grad_norm": 6.0523247718811035, + "learning_rate": 9.995069999507001e-06, + "loss": 0.5397, + "step": 4933 + }, + { + "epoch": 2.293816829381683, + "grad_norm": 4.819683074951172, + "learning_rate": 9.9950689995069e-06, + "loss": 0.4014, + "step": 4934 + }, + { + "epoch": 2.2942817294281728, + "grad_norm": 5.120641231536865, + "learning_rate": 9.995067999506801e-06, + "loss": 0.2135, + "step": 4935 + }, + { + "epoch": 2.294746629474663, + "grad_norm": 4.289938449859619, + "learning_rate": 9.9950669995067e-06, + "loss": 0.4437, + "step": 4936 + }, + { + "epoch": 2.295211529521153, + "grad_norm": 6.457924842834473, + "learning_rate": 9.9950659995066e-06, + "loss": 0.4062, + "step": 4937 + }, + { + "epoch": 2.295676429567643, + "grad_norm": 5.940118312835693, + "learning_rate": 9.995064999506501e-06, + "loss": 0.4626, + "step": 4938 + }, + { + "epoch": 2.296141329614133, + "grad_norm": 3.8558261394500732, + "learning_rate": 9.9950639995064e-06, + "loss": 0.2118, + "step": 4939 + }, + { + "epoch": 2.296606229660623, + "grad_norm": 3.7592902183532715, + "learning_rate": 9.9950629995063e-06, + "loss": 0.2368, + "step": 4940 + }, + { + "epoch": 2.297071129707113, + "grad_norm": 5.778518199920654, + "learning_rate": 9.9950619995062e-06, + "loss": 0.3239, + "step": 4941 + }, + { + "epoch": 2.297536029753603, + "grad_norm": 6.51664400100708, + "learning_rate": 9.9950609995061e-06, + "loss": 0.251, + "step": 4942 + }, + { + "epoch": 2.298000929800093, + "grad_norm": 5.426025867462158, + "learning_rate": 9.995059999506001e-06, + "loss": 0.4795, + "step": 4943 + }, + { + "epoch": 2.2984658298465828, + "grad_norm": 4.847101211547852, + "learning_rate": 9.995058999505902e-06, + "loss": 0.3492, + "step": 4944 + }, + { + "epoch": 2.298930729893073, + "grad_norm": 8.344991683959961, + "learning_rate": 9.9950579995058e-06, + "loss": 0.6087, + "step": 4945 + }, + { + "epoch": 2.299395629939563, + "grad_norm": 5.618790626525879, + "learning_rate": 9.995056999505701e-06, + "loss": 0.5757, + "step": 4946 + }, + { + "epoch": 2.299860529986053, + "grad_norm": 4.104610919952393, + "learning_rate": 9.9950559995056e-06, + "loss": 0.2908, + "step": 4947 + }, + { + "epoch": 2.300325430032543, + "grad_norm": 4.0479559898376465, + "learning_rate": 9.995054999505501e-06, + "loss": 0.3518, + "step": 4948 + }, + { + "epoch": 2.300790330079033, + "grad_norm": 4.598454475402832, + "learning_rate": 9.995053999505402e-06, + "loss": 0.3695, + "step": 4949 + }, + { + "epoch": 2.301255230125523, + "grad_norm": 5.529533863067627, + "learning_rate": 9.9950529995053e-06, + "loss": 0.4655, + "step": 4950 + }, + { + "epoch": 2.301720130172013, + "grad_norm": 11.528863906860352, + "learning_rate": 9.995051999505201e-06, + "loss": 0.2685, + "step": 4951 + }, + { + "epoch": 2.302185030218503, + "grad_norm": 7.729918956756592, + "learning_rate": 9.9950509995051e-06, + "loss": 0.5531, + "step": 4952 + }, + { + "epoch": 2.302649930264993, + "grad_norm": 5.34896993637085, + "learning_rate": 9.995049999505001e-06, + "loss": 0.5447, + "step": 4953 + }, + { + "epoch": 2.303114830311483, + "grad_norm": 4.954716682434082, + "learning_rate": 9.9950489995049e-06, + "loss": 0.3184, + "step": 4954 + }, + { + "epoch": 2.3035797303579733, + "grad_norm": 4.493119239807129, + "learning_rate": 9.9950479995048e-06, + "loss": 0.3791, + "step": 4955 + }, + { + "epoch": 2.304044630404463, + "grad_norm": 4.4055962562561035, + "learning_rate": 9.995046999504701e-06, + "loss": 0.243, + "step": 4956 + }, + { + "epoch": 2.304509530450953, + "grad_norm": 5.944924831390381, + "learning_rate": 9.995045999504602e-06, + "loss": 0.3357, + "step": 4957 + }, + { + "epoch": 2.304974430497443, + "grad_norm": 4.746670722961426, + "learning_rate": 9.9950449995045e-06, + "loss": 0.379, + "step": 4958 + }, + { + "epoch": 2.305439330543933, + "grad_norm": 6.812967777252197, + "learning_rate": 9.995043999504401e-06, + "loss": 0.8131, + "step": 4959 + }, + { + "epoch": 2.305904230590423, + "grad_norm": 10.168140411376953, + "learning_rate": 9.9950429995043e-06, + "loss": 0.9491, + "step": 4960 + }, + { + "epoch": 2.306369130636913, + "grad_norm": 6.819264888763428, + "learning_rate": 9.995041999504201e-06, + "loss": 0.3792, + "step": 4961 + }, + { + "epoch": 2.306834030683403, + "grad_norm": 3.830120801925659, + "learning_rate": 9.995040999504102e-06, + "loss": 0.4092, + "step": 4962 + }, + { + "epoch": 2.307298930729893, + "grad_norm": 6.457525730133057, + "learning_rate": 9.995039999504e-06, + "loss": 0.3214, + "step": 4963 + }, + { + "epoch": 2.3077638307763833, + "grad_norm": 3.6312248706817627, + "learning_rate": 9.995038999503901e-06, + "loss": 0.2612, + "step": 4964 + }, + { + "epoch": 2.308228730822873, + "grad_norm": 6.219481945037842, + "learning_rate": 9.9950379995038e-06, + "loss": 0.4191, + "step": 4965 + }, + { + "epoch": 2.308693630869363, + "grad_norm": 5.887006759643555, + "learning_rate": 9.995036999503701e-06, + "loss": 0.4798, + "step": 4966 + }, + { + "epoch": 2.309158530915853, + "grad_norm": 4.153655052185059, + "learning_rate": 9.9950359995036e-06, + "loss": 0.2579, + "step": 4967 + }, + { + "epoch": 2.309623430962343, + "grad_norm": 6.347054958343506, + "learning_rate": 9.9950349995035e-06, + "loss": 0.5121, + "step": 4968 + }, + { + "epoch": 2.310088331008833, + "grad_norm": 5.257510185241699, + "learning_rate": 9.995033999503401e-06, + "loss": 0.4356, + "step": 4969 + }, + { + "epoch": 2.310553231055323, + "grad_norm": 5.95878267288208, + "learning_rate": 9.995032999503302e-06, + "loss": 0.5851, + "step": 4970 + }, + { + "epoch": 2.311018131101813, + "grad_norm": 5.574553966522217, + "learning_rate": 9.9950319995032e-06, + "loss": 0.4637, + "step": 4971 + }, + { + "epoch": 2.311483031148303, + "grad_norm": 3.599973201751709, + "learning_rate": 9.9950309995031e-06, + "loss": 0.2551, + "step": 4972 + }, + { + "epoch": 2.3119479311947932, + "grad_norm": 7.084350109100342, + "learning_rate": 9.995029999503e-06, + "loss": 0.6909, + "step": 4973 + }, + { + "epoch": 2.312412831241283, + "grad_norm": 4.22507905960083, + "learning_rate": 9.995028999502901e-06, + "loss": 0.4614, + "step": 4974 + }, + { + "epoch": 2.3128777312877733, + "grad_norm": 6.289647579193115, + "learning_rate": 9.995027999502802e-06, + "loss": 0.5371, + "step": 4975 + }, + { + "epoch": 2.313342631334263, + "grad_norm": 10.13327407836914, + "learning_rate": 9.9950269995027e-06, + "loss": 0.7146, + "step": 4976 + }, + { + "epoch": 2.3138075313807533, + "grad_norm": 7.243377685546875, + "learning_rate": 9.995025999502601e-06, + "loss": 0.7458, + "step": 4977 + }, + { + "epoch": 2.314272431427243, + "grad_norm": 6.075034141540527, + "learning_rate": 9.9950249995025e-06, + "loss": 0.5078, + "step": 4978 + }, + { + "epoch": 2.314737331473733, + "grad_norm": 6.060837268829346, + "learning_rate": 9.995023999502401e-06, + "loss": 0.3808, + "step": 4979 + }, + { + "epoch": 2.315202231520223, + "grad_norm": 3.1713223457336426, + "learning_rate": 9.9950229995023e-06, + "loss": 0.1489, + "step": 4980 + }, + { + "epoch": 2.315667131566713, + "grad_norm": 5.606597423553467, + "learning_rate": 9.9950219995022e-06, + "loss": 0.3994, + "step": 4981 + }, + { + "epoch": 2.3161320316132032, + "grad_norm": 5.1020097732543945, + "learning_rate": 9.995020999502101e-06, + "loss": 0.4582, + "step": 4982 + }, + { + "epoch": 2.316596931659693, + "grad_norm": 5.620251178741455, + "learning_rate": 9.995019999502e-06, + "loss": 0.4974, + "step": 4983 + }, + { + "epoch": 2.3170618317061833, + "grad_norm": 3.7402405738830566, + "learning_rate": 9.995018999501901e-06, + "loss": 0.2552, + "step": 4984 + }, + { + "epoch": 2.317526731752673, + "grad_norm": 4.0455522537231445, + "learning_rate": 9.9950179995018e-06, + "loss": 0.2745, + "step": 4985 + }, + { + "epoch": 2.3179916317991633, + "grad_norm": 6.074113845825195, + "learning_rate": 9.9950169995017e-06, + "loss": 0.3521, + "step": 4986 + }, + { + "epoch": 2.318456531845653, + "grad_norm": 7.846375942230225, + "learning_rate": 9.995015999501601e-06, + "loss": 0.6504, + "step": 4987 + }, + { + "epoch": 2.318921431892143, + "grad_norm": 5.173695087432861, + "learning_rate": 9.995014999501502e-06, + "loss": 0.3248, + "step": 4988 + }, + { + "epoch": 2.319386331938633, + "grad_norm": 6.509139060974121, + "learning_rate": 9.9950139995014e-06, + "loss": 0.4162, + "step": 4989 + }, + { + "epoch": 2.3198512319851234, + "grad_norm": 4.3112101554870605, + "learning_rate": 9.995012999501301e-06, + "loss": 0.4915, + "step": 4990 + }, + { + "epoch": 2.3203161320316132, + "grad_norm": 5.52281379699707, + "learning_rate": 9.9950119995012e-06, + "loss": 0.3101, + "step": 4991 + }, + { + "epoch": 2.320781032078103, + "grad_norm": 5.1637983322143555, + "learning_rate": 9.995010999501101e-06, + "loss": 0.546, + "step": 4992 + }, + { + "epoch": 2.3212459321245933, + "grad_norm": 4.019276142120361, + "learning_rate": 9.995009999501e-06, + "loss": 0.2225, + "step": 4993 + }, + { + "epoch": 2.321710832171083, + "grad_norm": 5.684107780456543, + "learning_rate": 9.9950089995009e-06, + "loss": 0.4134, + "step": 4994 + }, + { + "epoch": 2.3221757322175733, + "grad_norm": 8.173792839050293, + "learning_rate": 9.995007999500801e-06, + "loss": 0.5878, + "step": 4995 + }, + { + "epoch": 2.322640632264063, + "grad_norm": 5.488272190093994, + "learning_rate": 9.9950069995007e-06, + "loss": 0.3845, + "step": 4996 + }, + { + "epoch": 2.3231055323105534, + "grad_norm": 4.0345025062561035, + "learning_rate": 9.995005999500601e-06, + "loss": 0.2302, + "step": 4997 + }, + { + "epoch": 2.323570432357043, + "grad_norm": 3.7282471656799316, + "learning_rate": 9.9950049995005e-06, + "loss": 0.2232, + "step": 4998 + }, + { + "epoch": 2.3240353324035334, + "grad_norm": 6.300872802734375, + "learning_rate": 9.9950039995004e-06, + "loss": 0.7967, + "step": 4999 + }, + { + "epoch": 2.3245002324500232, + "grad_norm": 5.45170783996582, + "learning_rate": 9.995002999500301e-06, + "loss": 0.4049, + "step": 5000 + }, + { + "epoch": 2.324965132496513, + "grad_norm": 12.892953872680664, + "learning_rate": 9.995001999500202e-06, + "loss": 0.4205, + "step": 5001 + }, + { + "epoch": 2.3254300325430033, + "grad_norm": 8.534778594970703, + "learning_rate": 9.995000999500101e-06, + "loss": 0.3997, + "step": 5002 + }, + { + "epoch": 2.325894932589493, + "grad_norm": 7.077780246734619, + "learning_rate": 9.9949999995e-06, + "loss": 0.2968, + "step": 5003 + }, + { + "epoch": 2.3263598326359833, + "grad_norm": 5.437408447265625, + "learning_rate": 9.9949989994999e-06, + "loss": 0.4451, + "step": 5004 + }, + { + "epoch": 2.326824732682473, + "grad_norm": 5.029237270355225, + "learning_rate": 9.9949979994998e-06, + "loss": 0.3662, + "step": 5005 + }, + { + "epoch": 2.3272896327289634, + "grad_norm": 3.6290738582611084, + "learning_rate": 9.9949969994997e-06, + "loss": 0.1422, + "step": 5006 + }, + { + "epoch": 2.327754532775453, + "grad_norm": 4.469902992248535, + "learning_rate": 9.9949959994996e-06, + "loss": 0.2617, + "step": 5007 + }, + { + "epoch": 2.3282194328219434, + "grad_norm": 6.45231819152832, + "learning_rate": 9.994994999499501e-06, + "loss": 0.6346, + "step": 5008 + }, + { + "epoch": 2.3286843328684332, + "grad_norm": 6.0898823738098145, + "learning_rate": 9.9949939994994e-06, + "loss": 0.4701, + "step": 5009 + }, + { + "epoch": 2.3291492329149235, + "grad_norm": 6.590389251708984, + "learning_rate": 9.994992999499301e-06, + "loss": 0.63, + "step": 5010 + }, + { + "epoch": 2.3296141329614133, + "grad_norm": 5.381944179534912, + "learning_rate": 9.9949919994992e-06, + "loss": 0.3787, + "step": 5011 + }, + { + "epoch": 2.3300790330079035, + "grad_norm": 4.4200968742370605, + "learning_rate": 9.9949909994991e-06, + "loss": 0.4499, + "step": 5012 + }, + { + "epoch": 2.3305439330543933, + "grad_norm": 3.4988176822662354, + "learning_rate": 9.994989999499001e-06, + "loss": 0.18, + "step": 5013 + }, + { + "epoch": 2.331008833100883, + "grad_norm": 5.286315441131592, + "learning_rate": 9.9949889994989e-06, + "loss": 0.2665, + "step": 5014 + }, + { + "epoch": 2.3314737331473734, + "grad_norm": 4.503076553344727, + "learning_rate": 9.994987999498801e-06, + "loss": 0.2548, + "step": 5015 + }, + { + "epoch": 2.331938633193863, + "grad_norm": 3.5230462551116943, + "learning_rate": 9.9949869994987e-06, + "loss": 0.299, + "step": 5016 + }, + { + "epoch": 2.3324035332403534, + "grad_norm": 5.7877631187438965, + "learning_rate": 9.9949859994986e-06, + "loss": 0.5416, + "step": 5017 + }, + { + "epoch": 2.3328684332868432, + "grad_norm": 5.868993759155273, + "learning_rate": 9.994984999498501e-06, + "loss": 0.5879, + "step": 5018 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 6.427070617675781, + "learning_rate": 9.9949839994984e-06, + "loss": 0.4378, + "step": 5019 + }, + { + "epoch": 2.3337982333798233, + "grad_norm": 6.937946319580078, + "learning_rate": 9.994982999498301e-06, + "loss": 0.5427, + "step": 5020 + }, + { + "epoch": 2.3342631334263135, + "grad_norm": 6.384261131286621, + "learning_rate": 9.994981999498202e-06, + "loss": 0.3204, + "step": 5021 + }, + { + "epoch": 2.3347280334728033, + "grad_norm": 4.606331825256348, + "learning_rate": 9.9949809994981e-06, + "loss": 0.2606, + "step": 5022 + }, + { + "epoch": 2.335192933519293, + "grad_norm": 4.238741874694824, + "learning_rate": 9.994979999498001e-06, + "loss": 0.2194, + "step": 5023 + }, + { + "epoch": 2.3356578335657834, + "grad_norm": 4.9829487800598145, + "learning_rate": 9.9949789994979e-06, + "loss": 0.4574, + "step": 5024 + }, + { + "epoch": 2.336122733612273, + "grad_norm": 5.186346054077148, + "learning_rate": 9.9949779994978e-06, + "loss": 0.4474, + "step": 5025 + }, + { + "epoch": 2.3365876336587634, + "grad_norm": 7.06009578704834, + "learning_rate": 9.994976999497701e-06, + "loss": 0.5541, + "step": 5026 + }, + { + "epoch": 2.3370525337052532, + "grad_norm": 7.1381516456604, + "learning_rate": 9.9949759994976e-06, + "loss": 0.4373, + "step": 5027 + }, + { + "epoch": 2.3375174337517435, + "grad_norm": 6.932162284851074, + "learning_rate": 9.994974999497501e-06, + "loss": 0.6086, + "step": 5028 + }, + { + "epoch": 2.3379823337982333, + "grad_norm": 5.932460784912109, + "learning_rate": 9.9949739994974e-06, + "loss": 0.6184, + "step": 5029 + }, + { + "epoch": 2.3384472338447235, + "grad_norm": 4.833393096923828, + "learning_rate": 9.9949729994973e-06, + "loss": 0.1996, + "step": 5030 + }, + { + "epoch": 2.3389121338912133, + "grad_norm": 4.272271156311035, + "learning_rate": 9.994971999497201e-06, + "loss": 0.4278, + "step": 5031 + }, + { + "epoch": 2.3393770339377036, + "grad_norm": 8.34500789642334, + "learning_rate": 9.9949709994971e-06, + "loss": 0.8177, + "step": 5032 + }, + { + "epoch": 2.3398419339841934, + "grad_norm": 9.190461158752441, + "learning_rate": 9.994969999497001e-06, + "loss": 0.4545, + "step": 5033 + }, + { + "epoch": 2.3403068340306836, + "grad_norm": 3.6861114501953125, + "learning_rate": 9.994968999496902e-06, + "loss": 0.2485, + "step": 5034 + }, + { + "epoch": 2.3407717340771734, + "grad_norm": 5.228905200958252, + "learning_rate": 9.9949679994968e-06, + "loss": 0.2269, + "step": 5035 + }, + { + "epoch": 2.3412366341236632, + "grad_norm": 6.4408369064331055, + "learning_rate": 9.9949669994967e-06, + "loss": 0.658, + "step": 5036 + }, + { + "epoch": 2.3417015341701535, + "grad_norm": 8.45034408569336, + "learning_rate": 9.9949659994966e-06, + "loss": 0.7019, + "step": 5037 + }, + { + "epoch": 2.3421664342166433, + "grad_norm": 9.579833030700684, + "learning_rate": 9.994964999496501e-06, + "loss": 0.8582, + "step": 5038 + }, + { + "epoch": 2.3426313342631335, + "grad_norm": 4.326962947845459, + "learning_rate": 9.994963999496402e-06, + "loss": 0.4387, + "step": 5039 + }, + { + "epoch": 2.3430962343096233, + "grad_norm": 6.197934150695801, + "learning_rate": 9.9949629994963e-06, + "loss": 0.5854, + "step": 5040 + }, + { + "epoch": 2.3435611343561136, + "grad_norm": 11.117565155029297, + "learning_rate": 9.994961999496201e-06, + "loss": 0.4677, + "step": 5041 + }, + { + "epoch": 2.3440260344026034, + "grad_norm": 4.456210136413574, + "learning_rate": 9.9949609994961e-06, + "loss": 0.33, + "step": 5042 + }, + { + "epoch": 2.3444909344490936, + "grad_norm": 4.512657165527344, + "learning_rate": 9.994959999496e-06, + "loss": 0.1933, + "step": 5043 + }, + { + "epoch": 2.3449558344955834, + "grad_norm": 8.61464786529541, + "learning_rate": 9.994958999495901e-06, + "loss": 0.8936, + "step": 5044 + }, + { + "epoch": 2.3454207345420732, + "grad_norm": 7.144443988800049, + "learning_rate": 9.994957999495802e-06, + "loss": 0.7609, + "step": 5045 + }, + { + "epoch": 2.3458856345885635, + "grad_norm": 5.8215413093566895, + "learning_rate": 9.994956999495701e-06, + "loss": 0.3825, + "step": 5046 + }, + { + "epoch": 2.3463505346350533, + "grad_norm": 4.22491455078125, + "learning_rate": 9.9949559994956e-06, + "loss": 0.3261, + "step": 5047 + }, + { + "epoch": 2.3468154346815435, + "grad_norm": 8.900324821472168, + "learning_rate": 9.9949549994955e-06, + "loss": 0.2476, + "step": 5048 + }, + { + "epoch": 2.3472803347280333, + "grad_norm": 5.992573261260986, + "learning_rate": 9.9949539994954e-06, + "loss": 0.439, + "step": 5049 + }, + { + "epoch": 2.3477452347745236, + "grad_norm": 6.011528968811035, + "learning_rate": 9.9949529994953e-06, + "loss": 0.5521, + "step": 5050 + }, + { + "epoch": 2.3482101348210134, + "grad_norm": 5.111826419830322, + "learning_rate": 9.994951999495201e-06, + "loss": 0.2409, + "step": 5051 + }, + { + "epoch": 2.3486750348675036, + "grad_norm": 4.9876017570495605, + "learning_rate": 9.994950999495102e-06, + "loss": 0.2808, + "step": 5052 + }, + { + "epoch": 2.3491399349139934, + "grad_norm": 7.419528961181641, + "learning_rate": 9.994949999495e-06, + "loss": 0.7578, + "step": 5053 + }, + { + "epoch": 2.3496048349604837, + "grad_norm": 6.559978008270264, + "learning_rate": 9.994948999494901e-06, + "loss": 0.4748, + "step": 5054 + }, + { + "epoch": 2.3500697350069735, + "grad_norm": 5.190243244171143, + "learning_rate": 9.9949479994948e-06, + "loss": 0.5314, + "step": 5055 + }, + { + "epoch": 2.3505346350534637, + "grad_norm": 6.263399124145508, + "learning_rate": 9.9949469994947e-06, + "loss": 0.8504, + "step": 5056 + }, + { + "epoch": 2.3509995350999535, + "grad_norm": 5.290328025817871, + "learning_rate": 9.994945999494601e-06, + "loss": 0.3855, + "step": 5057 + }, + { + "epoch": 2.3514644351464433, + "grad_norm": 4.052076816558838, + "learning_rate": 9.9949449994945e-06, + "loss": 0.3477, + "step": 5058 + }, + { + "epoch": 2.3519293351929336, + "grad_norm": 6.739137649536133, + "learning_rate": 9.994943999494401e-06, + "loss": 0.368, + "step": 5059 + }, + { + "epoch": 2.3523942352394234, + "grad_norm": 6.045091152191162, + "learning_rate": 9.9949429994943e-06, + "loss": 0.4212, + "step": 5060 + }, + { + "epoch": 2.3528591352859136, + "grad_norm": 4.068432331085205, + "learning_rate": 9.9949419994942e-06, + "loss": 0.2874, + "step": 5061 + }, + { + "epoch": 2.3533240353324034, + "grad_norm": 8.688581466674805, + "learning_rate": 9.9949409994941e-06, + "loss": 0.484, + "step": 5062 + }, + { + "epoch": 2.3537889353788937, + "grad_norm": 5.3468146324157715, + "learning_rate": 9.994939999494e-06, + "loss": 0.534, + "step": 5063 + }, + { + "epoch": 2.3542538354253835, + "grad_norm": 9.496050834655762, + "learning_rate": 9.994938999493901e-06, + "loss": 0.5956, + "step": 5064 + }, + { + "epoch": 2.3547187354718737, + "grad_norm": 6.174869537353516, + "learning_rate": 9.994937999493802e-06, + "loss": 0.5692, + "step": 5065 + }, + { + "epoch": 2.3551836355183635, + "grad_norm": 5.185247898101807, + "learning_rate": 9.9949369994937e-06, + "loss": 0.3734, + "step": 5066 + }, + { + "epoch": 2.3556485355648533, + "grad_norm": 5.7406325340271, + "learning_rate": 9.994935999493601e-06, + "loss": 0.4672, + "step": 5067 + }, + { + "epoch": 2.3561134356113436, + "grad_norm": 7.093161106109619, + "learning_rate": 9.9949349994935e-06, + "loss": 0.543, + "step": 5068 + }, + { + "epoch": 2.356578335657834, + "grad_norm": 5.405251502990723, + "learning_rate": 9.994933999493401e-06, + "loss": 0.5513, + "step": 5069 + }, + { + "epoch": 2.3570432357043236, + "grad_norm": 4.305744647979736, + "learning_rate": 9.994932999493302e-06, + "loss": 0.2937, + "step": 5070 + }, + { + "epoch": 2.3575081357508134, + "grad_norm": 5.564687728881836, + "learning_rate": 9.9949319994932e-06, + "loss": 0.3497, + "step": 5071 + }, + { + "epoch": 2.3579730357973037, + "grad_norm": 7.468056678771973, + "learning_rate": 9.994930999493101e-06, + "loss": 0.6557, + "step": 5072 + }, + { + "epoch": 2.3584379358437935, + "grad_norm": 5.99576997756958, + "learning_rate": 9.994929999493e-06, + "loss": 0.5136, + "step": 5073 + }, + { + "epoch": 2.3589028358902837, + "grad_norm": 14.288616180419922, + "learning_rate": 9.9949289994929e-06, + "loss": 0.9661, + "step": 5074 + }, + { + "epoch": 2.3593677359367735, + "grad_norm": 7.332887649536133, + "learning_rate": 9.9949279994928e-06, + "loss": 0.5154, + "step": 5075 + }, + { + "epoch": 2.3598326359832638, + "grad_norm": 5.80228853225708, + "learning_rate": 9.9949269994927e-06, + "loss": 0.3634, + "step": 5076 + }, + { + "epoch": 2.3602975360297536, + "grad_norm": 5.403787612915039, + "learning_rate": 9.994925999492601e-06, + "loss": 0.4207, + "step": 5077 + }, + { + "epoch": 2.360762436076244, + "grad_norm": 4.66761589050293, + "learning_rate": 9.994924999492502e-06, + "loss": 0.2311, + "step": 5078 + }, + { + "epoch": 2.3612273361227336, + "grad_norm": 5.646284103393555, + "learning_rate": 9.9949239994924e-06, + "loss": 0.4462, + "step": 5079 + }, + { + "epoch": 2.3616922361692234, + "grad_norm": 5.5342488288879395, + "learning_rate": 9.9949229994923e-06, + "loss": 0.4604, + "step": 5080 + }, + { + "epoch": 2.3621571362157137, + "grad_norm": 5.601319313049316, + "learning_rate": 9.9949219994922e-06, + "loss": 0.4458, + "step": 5081 + }, + { + "epoch": 2.3626220362622035, + "grad_norm": 3.5761334896087646, + "learning_rate": 9.994920999492101e-06, + "loss": 0.1971, + "step": 5082 + }, + { + "epoch": 2.3630869363086937, + "grad_norm": 7.087026119232178, + "learning_rate": 9.994919999492002e-06, + "loss": 0.631, + "step": 5083 + }, + { + "epoch": 2.3635518363551835, + "grad_norm": 7.337288856506348, + "learning_rate": 9.9949189994919e-06, + "loss": 0.7654, + "step": 5084 + }, + { + "epoch": 2.3640167364016738, + "grad_norm": 6.687684059143066, + "learning_rate": 9.994917999491801e-06, + "loss": 0.6195, + "step": 5085 + }, + { + "epoch": 2.3644816364481636, + "grad_norm": 6.9143781661987305, + "learning_rate": 9.9949169994917e-06, + "loss": 0.6649, + "step": 5086 + }, + { + "epoch": 2.364946536494654, + "grad_norm": 5.290834426879883, + "learning_rate": 9.994915999491601e-06, + "loss": 0.4938, + "step": 5087 + }, + { + "epoch": 2.3654114365411436, + "grad_norm": 5.4535698890686035, + "learning_rate": 9.9949149994915e-06, + "loss": 0.4226, + "step": 5088 + }, + { + "epoch": 2.3658763365876334, + "grad_norm": 7.049666404724121, + "learning_rate": 9.9949139994914e-06, + "loss": 0.7379, + "step": 5089 + }, + { + "epoch": 2.3663412366341237, + "grad_norm": 6.398994445800781, + "learning_rate": 9.994912999491301e-06, + "loss": 0.4511, + "step": 5090 + }, + { + "epoch": 2.366806136680614, + "grad_norm": 5.572786808013916, + "learning_rate": 9.9949119994912e-06, + "loss": 0.3965, + "step": 5091 + }, + { + "epoch": 2.3672710367271037, + "grad_norm": 6.701457977294922, + "learning_rate": 9.9949109994911e-06, + "loss": 0.4287, + "step": 5092 + }, + { + "epoch": 2.3677359367735935, + "grad_norm": 6.96864128112793, + "learning_rate": 9.994909999491e-06, + "loss": 0.3607, + "step": 5093 + }, + { + "epoch": 2.3682008368200838, + "grad_norm": 7.629732131958008, + "learning_rate": 9.9949089994909e-06, + "loss": 0.8593, + "step": 5094 + }, + { + "epoch": 2.3686657368665736, + "grad_norm": 6.438701152801514, + "learning_rate": 9.994907999490801e-06, + "loss": 0.5048, + "step": 5095 + }, + { + "epoch": 2.369130636913064, + "grad_norm": 4.002255439758301, + "learning_rate": 9.994906999490702e-06, + "loss": 0.2942, + "step": 5096 + }, + { + "epoch": 2.3695955369595536, + "grad_norm": 4.915142059326172, + "learning_rate": 9.9949059994906e-06, + "loss": 0.2976, + "step": 5097 + }, + { + "epoch": 2.370060437006044, + "grad_norm": 5.052490711212158, + "learning_rate": 9.994904999490501e-06, + "loss": 0.529, + "step": 5098 + }, + { + "epoch": 2.3705253370525337, + "grad_norm": 8.778657913208008, + "learning_rate": 9.9949039994904e-06, + "loss": 0.7482, + "step": 5099 + }, + { + "epoch": 2.370990237099024, + "grad_norm": 7.410974502563477, + "learning_rate": 9.994902999490301e-06, + "loss": 0.619, + "step": 5100 + }, + { + "epoch": 2.3714551371455137, + "grad_norm": 5.257546424865723, + "learning_rate": 9.9949019994902e-06, + "loss": 0.4491, + "step": 5101 + }, + { + "epoch": 2.3719200371920035, + "grad_norm": 5.274449825286865, + "learning_rate": 9.9949009994901e-06, + "loss": 0.366, + "step": 5102 + }, + { + "epoch": 2.3723849372384938, + "grad_norm": 4.35119104385376, + "learning_rate": 9.994899999490001e-06, + "loss": 0.501, + "step": 5103 + }, + { + "epoch": 2.3728498372849836, + "grad_norm": 10.065338134765625, + "learning_rate": 9.9948989994899e-06, + "loss": 0.3572, + "step": 5104 + }, + { + "epoch": 2.373314737331474, + "grad_norm": 5.082964897155762, + "learning_rate": 9.994897999489801e-06, + "loss": 0.4837, + "step": 5105 + }, + { + "epoch": 2.3737796373779636, + "grad_norm": 7.385058403015137, + "learning_rate": 9.9948969994897e-06, + "loss": 0.5708, + "step": 5106 + }, + { + "epoch": 2.374244537424454, + "grad_norm": 5.96180534362793, + "learning_rate": 9.9948959994896e-06, + "loss": 0.4211, + "step": 5107 + }, + { + "epoch": 2.3747094374709437, + "grad_norm": 6.7290120124816895, + "learning_rate": 9.994894999489501e-06, + "loss": 0.4718, + "step": 5108 + }, + { + "epoch": 2.375174337517434, + "grad_norm": 4.660354137420654, + "learning_rate": 9.994893999489402e-06, + "loss": 0.2964, + "step": 5109 + }, + { + "epoch": 2.3756392375639237, + "grad_norm": 3.817291259765625, + "learning_rate": 9.9948929994893e-06, + "loss": 0.3216, + "step": 5110 + }, + { + "epoch": 2.376104137610414, + "grad_norm": 6.417239665985107, + "learning_rate": 9.994891999489201e-06, + "loss": 0.6643, + "step": 5111 + }, + { + "epoch": 2.3765690376569037, + "grad_norm": 4.875197410583496, + "learning_rate": 9.9948909994891e-06, + "loss": 0.5815, + "step": 5112 + }, + { + "epoch": 2.377033937703394, + "grad_norm": 5.600948810577393, + "learning_rate": 9.994889999489001e-06, + "loss": 0.6697, + "step": 5113 + }, + { + "epoch": 2.377498837749884, + "grad_norm": 4.107156753540039, + "learning_rate": 9.994888999488902e-06, + "loss": 0.2257, + "step": 5114 + }, + { + "epoch": 2.3779637377963736, + "grad_norm": 4.390016078948975, + "learning_rate": 9.9948879994888e-06, + "loss": 0.2796, + "step": 5115 + }, + { + "epoch": 2.378428637842864, + "grad_norm": 7.054908752441406, + "learning_rate": 9.994886999488701e-06, + "loss": 0.634, + "step": 5116 + }, + { + "epoch": 2.3788935378893536, + "grad_norm": 5.747330188751221, + "learning_rate": 9.9948859994886e-06, + "loss": 0.5461, + "step": 5117 + }, + { + "epoch": 2.379358437935844, + "grad_norm": 8.151705741882324, + "learning_rate": 9.994884999488501e-06, + "loss": 0.5034, + "step": 5118 + }, + { + "epoch": 2.3798233379823337, + "grad_norm": 6.768886566162109, + "learning_rate": 9.9948839994884e-06, + "loss": 0.6168, + "step": 5119 + }, + { + "epoch": 2.380288238028824, + "grad_norm": 4.672508716583252, + "learning_rate": 9.9948829994883e-06, + "loss": 0.3372, + "step": 5120 + }, + { + "epoch": 2.3807531380753137, + "grad_norm": 5.47153902053833, + "learning_rate": 9.994881999488201e-06, + "loss": 0.4566, + "step": 5121 + }, + { + "epoch": 2.381218038121804, + "grad_norm": 5.613348007202148, + "learning_rate": 9.994880999488102e-06, + "loss": 0.3782, + "step": 5122 + }, + { + "epoch": 2.381682938168294, + "grad_norm": 6.045223712921143, + "learning_rate": 9.994879999488001e-06, + "loss": 0.5216, + "step": 5123 + }, + { + "epoch": 2.3821478382147836, + "grad_norm": 6.2340168952941895, + "learning_rate": 9.9948789994879e-06, + "loss": 0.2802, + "step": 5124 + }, + { + "epoch": 2.382612738261274, + "grad_norm": 5.543851375579834, + "learning_rate": 9.9948779994878e-06, + "loss": 0.2799, + "step": 5125 + }, + { + "epoch": 2.3830776383077636, + "grad_norm": 4.715404510498047, + "learning_rate": 9.994876999487701e-06, + "loss": 0.3915, + "step": 5126 + }, + { + "epoch": 2.383542538354254, + "grad_norm": 6.740653038024902, + "learning_rate": 9.994875999487602e-06, + "loss": 0.4159, + "step": 5127 + }, + { + "epoch": 2.3840074384007437, + "grad_norm": 3.830519199371338, + "learning_rate": 9.9948749994875e-06, + "loss": 0.2937, + "step": 5128 + }, + { + "epoch": 2.384472338447234, + "grad_norm": 4.738029956817627, + "learning_rate": 9.994873999487401e-06, + "loss": 0.4625, + "step": 5129 + }, + { + "epoch": 2.3849372384937237, + "grad_norm": 6.3989057540893555, + "learning_rate": 9.9948729994873e-06, + "loss": 0.588, + "step": 5130 + }, + { + "epoch": 2.385402138540214, + "grad_norm": 4.5789570808410645, + "learning_rate": 9.994871999487201e-06, + "loss": 0.2727, + "step": 5131 + }, + { + "epoch": 2.385867038586704, + "grad_norm": 6.977818965911865, + "learning_rate": 9.9948709994871e-06, + "loss": 0.4549, + "step": 5132 + }, + { + "epoch": 2.386331938633194, + "grad_norm": 5.188814163208008, + "learning_rate": 9.994869999487e-06, + "loss": 0.497, + "step": 5133 + }, + { + "epoch": 2.386796838679684, + "grad_norm": 6.148141384124756, + "learning_rate": 9.994868999486901e-06, + "loss": 0.2628, + "step": 5134 + }, + { + "epoch": 2.387261738726174, + "grad_norm": 5.03603458404541, + "learning_rate": 9.9948679994868e-06, + "loss": 0.3839, + "step": 5135 + }, + { + "epoch": 2.387726638772664, + "grad_norm": 18.963098526000977, + "learning_rate": 9.994866999486701e-06, + "loss": 0.2867, + "step": 5136 + }, + { + "epoch": 2.3881915388191537, + "grad_norm": 5.26957368850708, + "learning_rate": 9.9948659994866e-06, + "loss": 0.2493, + "step": 5137 + }, + { + "epoch": 2.388656438865644, + "grad_norm": 4.94774866104126, + "learning_rate": 9.9948649994865e-06, + "loss": 0.3265, + "step": 5138 + }, + { + "epoch": 2.3891213389121337, + "grad_norm": 5.316683769226074, + "learning_rate": 9.994863999486401e-06, + "loss": 0.6427, + "step": 5139 + }, + { + "epoch": 2.389586238958624, + "grad_norm": 4.1961870193481445, + "learning_rate": 9.994862999486302e-06, + "loss": 0.2812, + "step": 5140 + }, + { + "epoch": 2.390051139005114, + "grad_norm": 4.964874267578125, + "learning_rate": 9.9948619994862e-06, + "loss": 0.3894, + "step": 5141 + }, + { + "epoch": 2.390516039051604, + "grad_norm": 3.606333017349243, + "learning_rate": 9.994860999486102e-06, + "loss": 0.3379, + "step": 5142 + }, + { + "epoch": 2.390980939098094, + "grad_norm": 3.4350666999816895, + "learning_rate": 9.994859999486e-06, + "loss": 0.1653, + "step": 5143 + }, + { + "epoch": 2.391445839144584, + "grad_norm": 5.001581192016602, + "learning_rate": 9.9948589994859e-06, + "loss": 0.4169, + "step": 5144 + }, + { + "epoch": 2.391910739191074, + "grad_norm": 5.231367111206055, + "learning_rate": 9.9948579994858e-06, + "loss": 0.4687, + "step": 5145 + }, + { + "epoch": 2.3923756392375637, + "grad_norm": 5.792492866516113, + "learning_rate": 9.9948569994857e-06, + "loss": 0.5161, + "step": 5146 + }, + { + "epoch": 2.392840539284054, + "grad_norm": 3.7573862075805664, + "learning_rate": 9.994855999485601e-06, + "loss": 0.2994, + "step": 5147 + }, + { + "epoch": 2.393305439330544, + "grad_norm": 4.931885242462158, + "learning_rate": 9.9948549994855e-06, + "loss": 0.404, + "step": 5148 + }, + { + "epoch": 2.393770339377034, + "grad_norm": 7.641971588134766, + "learning_rate": 9.994853999485401e-06, + "loss": 0.4766, + "step": 5149 + }, + { + "epoch": 2.394235239423524, + "grad_norm": 5.825547695159912, + "learning_rate": 9.9948529994853e-06, + "loss": 0.5808, + "step": 5150 + }, + { + "epoch": 2.394700139470014, + "grad_norm": 3.158874273300171, + "learning_rate": 9.9948519994852e-06, + "loss": 0.1594, + "step": 5151 + }, + { + "epoch": 2.395165039516504, + "grad_norm": 4.607054233551025, + "learning_rate": 9.994850999485101e-06, + "loss": 0.3722, + "step": 5152 + }, + { + "epoch": 2.395629939562994, + "grad_norm": 5.046706199645996, + "learning_rate": 9.994849999485002e-06, + "loss": 0.6401, + "step": 5153 + }, + { + "epoch": 2.396094839609484, + "grad_norm": 4.162281036376953, + "learning_rate": 9.994848999484901e-06, + "loss": 0.2925, + "step": 5154 + }, + { + "epoch": 2.396559739655974, + "grad_norm": 6.493382930755615, + "learning_rate": 9.9948479994848e-06, + "loss": 0.5471, + "step": 5155 + }, + { + "epoch": 2.397024639702464, + "grad_norm": 6.762735366821289, + "learning_rate": 9.9948469994847e-06, + "loss": 0.3579, + "step": 5156 + }, + { + "epoch": 2.397489539748954, + "grad_norm": 5.069240093231201, + "learning_rate": 9.9948459994846e-06, + "loss": 0.3766, + "step": 5157 + }, + { + "epoch": 2.397954439795444, + "grad_norm": 6.3324360847473145, + "learning_rate": 9.9948449994845e-06, + "loss": 0.6511, + "step": 5158 + }, + { + "epoch": 2.398419339841934, + "grad_norm": 7.717749118804932, + "learning_rate": 9.9948439994844e-06, + "loss": 0.8267, + "step": 5159 + }, + { + "epoch": 2.398884239888424, + "grad_norm": 4.312648296356201, + "learning_rate": 9.994842999484301e-06, + "loss": 0.2478, + "step": 5160 + }, + { + "epoch": 2.399349139934914, + "grad_norm": 4.901788711547852, + "learning_rate": 9.9948419994842e-06, + "loss": 0.387, + "step": 5161 + }, + { + "epoch": 2.399814039981404, + "grad_norm": 4.999730110168457, + "learning_rate": 9.994840999484101e-06, + "loss": 0.5449, + "step": 5162 + }, + { + "epoch": 2.400278940027894, + "grad_norm": 5.1698102951049805, + "learning_rate": 9.994839999484e-06, + "loss": 0.6007, + "step": 5163 + }, + { + "epoch": 2.400743840074384, + "grad_norm": 5.731054306030273, + "learning_rate": 9.9948389994839e-06, + "loss": 0.1581, + "step": 5164 + }, + { + "epoch": 2.401208740120874, + "grad_norm": 9.859057426452637, + "learning_rate": 9.994837999483801e-06, + "loss": 0.4604, + "step": 5165 + }, + { + "epoch": 2.401673640167364, + "grad_norm": 6.995851993560791, + "learning_rate": 9.9948369994837e-06, + "loss": 0.6109, + "step": 5166 + }, + { + "epoch": 2.402138540213854, + "grad_norm": 4.262461185455322, + "learning_rate": 9.994835999483601e-06, + "loss": 0.361, + "step": 5167 + }, + { + "epoch": 2.402603440260344, + "grad_norm": 3.57582950592041, + "learning_rate": 9.9948349994835e-06, + "loss": 0.1538, + "step": 5168 + }, + { + "epoch": 2.403068340306834, + "grad_norm": 8.630762100219727, + "learning_rate": 9.9948339994834e-06, + "loss": 0.5756, + "step": 5169 + }, + { + "epoch": 2.4035332403533243, + "grad_norm": 5.279118537902832, + "learning_rate": 9.9948329994833e-06, + "loss": 0.3964, + "step": 5170 + }, + { + "epoch": 2.403998140399814, + "grad_norm": 6.242101669311523, + "learning_rate": 9.9948319994832e-06, + "loss": 0.4137, + "step": 5171 + }, + { + "epoch": 2.404463040446304, + "grad_norm": 5.223903179168701, + "learning_rate": 9.994830999483101e-06, + "loss": 0.4534, + "step": 5172 + }, + { + "epoch": 2.404927940492794, + "grad_norm": 7.589288711547852, + "learning_rate": 9.994829999483002e-06, + "loss": 0.5108, + "step": 5173 + }, + { + "epoch": 2.405392840539284, + "grad_norm": 5.377110958099365, + "learning_rate": 9.9948289994829e-06, + "loss": 0.3763, + "step": 5174 + }, + { + "epoch": 2.405857740585774, + "grad_norm": 4.170370578765869, + "learning_rate": 9.994827999482801e-06, + "loss": 0.3991, + "step": 5175 + }, + { + "epoch": 2.406322640632264, + "grad_norm": 6.004370212554932, + "learning_rate": 9.9948269994827e-06, + "loss": 0.6139, + "step": 5176 + }, + { + "epoch": 2.4067875406787542, + "grad_norm": 6.844902515411377, + "learning_rate": 9.9948259994826e-06, + "loss": 0.4895, + "step": 5177 + }, + { + "epoch": 2.407252440725244, + "grad_norm": 6.824481964111328, + "learning_rate": 9.994824999482501e-06, + "loss": 0.6379, + "step": 5178 + }, + { + "epoch": 2.4077173407717343, + "grad_norm": 4.563008785247803, + "learning_rate": 9.9948239994824e-06, + "loss": 0.1907, + "step": 5179 + }, + { + "epoch": 2.408182240818224, + "grad_norm": 5.012880802154541, + "learning_rate": 9.994822999482301e-06, + "loss": 0.3106, + "step": 5180 + }, + { + "epoch": 2.408647140864714, + "grad_norm": 4.226649761199951, + "learning_rate": 9.9948219994822e-06, + "loss": 0.3195, + "step": 5181 + }, + { + "epoch": 2.409112040911204, + "grad_norm": 6.017914772033691, + "learning_rate": 9.9948209994821e-06, + "loss": 0.582, + "step": 5182 + }, + { + "epoch": 2.409576940957694, + "grad_norm": 6.643052101135254, + "learning_rate": 9.994819999482001e-06, + "loss": 0.5198, + "step": 5183 + }, + { + "epoch": 2.410041841004184, + "grad_norm": 4.087279796600342, + "learning_rate": 9.9948189994819e-06, + "loss": 0.266, + "step": 5184 + }, + { + "epoch": 2.410506741050674, + "grad_norm": 4.686944961547852, + "learning_rate": 9.994817999481801e-06, + "loss": 0.4167, + "step": 5185 + }, + { + "epoch": 2.410971641097164, + "grad_norm": 9.541116714477539, + "learning_rate": 9.994816999481702e-06, + "loss": 0.5888, + "step": 5186 + }, + { + "epoch": 2.411436541143654, + "grad_norm": 8.573952674865723, + "learning_rate": 9.9948159994816e-06, + "loss": 0.5922, + "step": 5187 + }, + { + "epoch": 2.4119014411901443, + "grad_norm": 8.34106731414795, + "learning_rate": 9.9948149994815e-06, + "loss": 0.2486, + "step": 5188 + }, + { + "epoch": 2.412366341236634, + "grad_norm": 3.9778292179107666, + "learning_rate": 9.9948139994814e-06, + "loss": 0.2861, + "step": 5189 + }, + { + "epoch": 2.4128312412831243, + "grad_norm": 3.6343815326690674, + "learning_rate": 9.994812999481301e-06, + "loss": 0.2151, + "step": 5190 + }, + { + "epoch": 2.413296141329614, + "grad_norm": 5.135373115539551, + "learning_rate": 9.994811999481202e-06, + "loss": 0.516, + "step": 5191 + }, + { + "epoch": 2.4137610413761044, + "grad_norm": 6.028303146362305, + "learning_rate": 9.9948109994811e-06, + "loss": 0.4614, + "step": 5192 + }, + { + "epoch": 2.414225941422594, + "grad_norm": 7.323265552520752, + "learning_rate": 9.994809999481001e-06, + "loss": 0.7668, + "step": 5193 + }, + { + "epoch": 2.414690841469084, + "grad_norm": 5.3483500480651855, + "learning_rate": 9.9948089994809e-06, + "loss": 0.4011, + "step": 5194 + }, + { + "epoch": 2.415155741515574, + "grad_norm": 8.201322555541992, + "learning_rate": 9.9948079994808e-06, + "loss": 0.7739, + "step": 5195 + }, + { + "epoch": 2.415620641562064, + "grad_norm": 5.381821155548096, + "learning_rate": 9.994806999480701e-06, + "loss": 0.4638, + "step": 5196 + }, + { + "epoch": 2.4160855416085543, + "grad_norm": 5.359497547149658, + "learning_rate": 9.994805999480602e-06, + "loss": 0.3742, + "step": 5197 + }, + { + "epoch": 2.416550441655044, + "grad_norm": 9.926708221435547, + "learning_rate": 9.994804999480501e-06, + "loss": 0.6117, + "step": 5198 + }, + { + "epoch": 2.4170153417015343, + "grad_norm": 4.215640068054199, + "learning_rate": 9.9948039994804e-06, + "loss": 0.2661, + "step": 5199 + }, + { + "epoch": 2.417480241748024, + "grad_norm": 4.708570957183838, + "learning_rate": 9.9948029994803e-06, + "loss": 0.3535, + "step": 5200 + }, + { + "epoch": 2.4179451417945144, + "grad_norm": 3.9371135234832764, + "learning_rate": 9.9948019994802e-06, + "loss": 0.3261, + "step": 5201 + }, + { + "epoch": 2.418410041841004, + "grad_norm": 16.943683624267578, + "learning_rate": 9.9948009994801e-06, + "loss": 2.1199, + "step": 5202 + }, + { + "epoch": 2.418874941887494, + "grad_norm": 4.469849586486816, + "learning_rate": 9.994799999480001e-06, + "loss": 0.3406, + "step": 5203 + }, + { + "epoch": 2.419339841933984, + "grad_norm": 7.706470489501953, + "learning_rate": 9.994798999479902e-06, + "loss": 0.9715, + "step": 5204 + }, + { + "epoch": 2.419804741980474, + "grad_norm": 5.02799654006958, + "learning_rate": 9.9947979994798e-06, + "loss": 0.2949, + "step": 5205 + }, + { + "epoch": 2.4202696420269643, + "grad_norm": 2.8106184005737305, + "learning_rate": 9.994796999479701e-06, + "loss": 0.2204, + "step": 5206 + }, + { + "epoch": 2.420734542073454, + "grad_norm": 6.4220685958862305, + "learning_rate": 9.9947959994796e-06, + "loss": 0.4672, + "step": 5207 + }, + { + "epoch": 2.4211994421199443, + "grad_norm": 8.946244239807129, + "learning_rate": 9.994794999479501e-06, + "loss": 0.4856, + "step": 5208 + }, + { + "epoch": 2.421664342166434, + "grad_norm": 8.352355003356934, + "learning_rate": 9.994793999479402e-06, + "loss": 0.517, + "step": 5209 + }, + { + "epoch": 2.4221292422129244, + "grad_norm": 4.452331066131592, + "learning_rate": 9.9947929994793e-06, + "loss": 0.4125, + "step": 5210 + }, + { + "epoch": 2.422594142259414, + "grad_norm": 4.607297897338867, + "learning_rate": 9.994791999479201e-06, + "loss": 0.4245, + "step": 5211 + }, + { + "epoch": 2.4230590423059044, + "grad_norm": 7.628004550933838, + "learning_rate": 9.9947909994791e-06, + "loss": 0.3988, + "step": 5212 + }, + { + "epoch": 2.423523942352394, + "grad_norm": 4.799656391143799, + "learning_rate": 9.994789999479e-06, + "loss": 0.2994, + "step": 5213 + }, + { + "epoch": 2.4239888423988845, + "grad_norm": 6.999420642852783, + "learning_rate": 9.9947889994789e-06, + "loss": 0.4693, + "step": 5214 + }, + { + "epoch": 2.4244537424453743, + "grad_norm": 7.002630233764648, + "learning_rate": 9.9947879994788e-06, + "loss": 0.75, + "step": 5215 + }, + { + "epoch": 2.424918642491864, + "grad_norm": 4.208200454711914, + "learning_rate": 9.994786999478701e-06, + "loss": 0.2519, + "step": 5216 + }, + { + "epoch": 2.4253835425383543, + "grad_norm": 4.877662658691406, + "learning_rate": 9.994785999478602e-06, + "loss": 0.351, + "step": 5217 + }, + { + "epoch": 2.425848442584844, + "grad_norm": 4.994498252868652, + "learning_rate": 9.9947849994785e-06, + "loss": 0.1808, + "step": 5218 + }, + { + "epoch": 2.4263133426313344, + "grad_norm": 6.517800331115723, + "learning_rate": 9.994783999478401e-06, + "loss": 0.5353, + "step": 5219 + }, + { + "epoch": 2.426778242677824, + "grad_norm": 4.660003185272217, + "learning_rate": 9.9947829994783e-06, + "loss": 0.3318, + "step": 5220 + }, + { + "epoch": 2.4272431427243144, + "grad_norm": 7.063469886779785, + "learning_rate": 9.994781999478201e-06, + "loss": 0.6658, + "step": 5221 + }, + { + "epoch": 2.427708042770804, + "grad_norm": 6.865651607513428, + "learning_rate": 9.994780999478102e-06, + "loss": 0.6239, + "step": 5222 + }, + { + "epoch": 2.4281729428172945, + "grad_norm": 5.017030715942383, + "learning_rate": 9.994779999478e-06, + "loss": 0.5458, + "step": 5223 + }, + { + "epoch": 2.4286378428637843, + "grad_norm": 4.763394832611084, + "learning_rate": 9.994778999477901e-06, + "loss": 0.4708, + "step": 5224 + }, + { + "epoch": 2.429102742910274, + "grad_norm": 4.2668776512146, + "learning_rate": 9.9947779994778e-06, + "loss": 0.4078, + "step": 5225 + }, + { + "epoch": 2.4295676429567643, + "grad_norm": 3.3326168060302734, + "learning_rate": 9.994776999477701e-06, + "loss": 0.212, + "step": 5226 + }, + { + "epoch": 2.430032543003254, + "grad_norm": 4.546849727630615, + "learning_rate": 9.9947759994776e-06, + "loss": 0.4178, + "step": 5227 + }, + { + "epoch": 2.4304974430497444, + "grad_norm": 7.713714599609375, + "learning_rate": 9.9947749994775e-06, + "loss": 0.6354, + "step": 5228 + }, + { + "epoch": 2.430962343096234, + "grad_norm": 7.353805065155029, + "learning_rate": 9.994773999477401e-06, + "loss": 0.8711, + "step": 5229 + }, + { + "epoch": 2.4314272431427244, + "grad_norm": 5.69140100479126, + "learning_rate": 9.994772999477302e-06, + "loss": 0.4853, + "step": 5230 + }, + { + "epoch": 2.431892143189214, + "grad_norm": 5.77006196975708, + "learning_rate": 9.9947719994772e-06, + "loss": 0.5215, + "step": 5231 + }, + { + "epoch": 2.4323570432357045, + "grad_norm": 3.465768814086914, + "learning_rate": 9.9947709994771e-06, + "loss": 0.2559, + "step": 5232 + }, + { + "epoch": 2.4328219432821943, + "grad_norm": 6.018754482269287, + "learning_rate": 9.994769999477e-06, + "loss": 0.4852, + "step": 5233 + }, + { + "epoch": 2.4332868433286845, + "grad_norm": 4.793680667877197, + "learning_rate": 9.994768999476901e-06, + "loss": 0.3647, + "step": 5234 + }, + { + "epoch": 2.4337517433751743, + "grad_norm": 4.301246166229248, + "learning_rate": 9.994767999476802e-06, + "loss": 0.422, + "step": 5235 + }, + { + "epoch": 2.4342166434216645, + "grad_norm": 6.049656391143799, + "learning_rate": 9.9947669994767e-06, + "loss": 0.5409, + "step": 5236 + }, + { + "epoch": 2.4346815434681544, + "grad_norm": 4.371786594390869, + "learning_rate": 9.994765999476601e-06, + "loss": 0.4181, + "step": 5237 + }, + { + "epoch": 2.435146443514644, + "grad_norm": 3.9757721424102783, + "learning_rate": 9.9947649994765e-06, + "loss": 0.3598, + "step": 5238 + }, + { + "epoch": 2.4356113435611344, + "grad_norm": 6.478806972503662, + "learning_rate": 9.994763999476401e-06, + "loss": 0.4723, + "step": 5239 + }, + { + "epoch": 2.436076243607624, + "grad_norm": 6.911035060882568, + "learning_rate": 9.9947629994763e-06, + "loss": 0.6313, + "step": 5240 + }, + { + "epoch": 2.4365411436541144, + "grad_norm": 4.091418743133545, + "learning_rate": 9.9947619994762e-06, + "loss": 0.2494, + "step": 5241 + }, + { + "epoch": 2.4370060437006043, + "grad_norm": 3.4461565017700195, + "learning_rate": 9.994760999476101e-06, + "loss": 0.259, + "step": 5242 + }, + { + "epoch": 2.4374709437470945, + "grad_norm": 5.595682144165039, + "learning_rate": 9.994759999476e-06, + "loss": 0.6753, + "step": 5243 + }, + { + "epoch": 2.4379358437935843, + "grad_norm": 4.675457954406738, + "learning_rate": 9.9947589994759e-06, + "loss": 0.2728, + "step": 5244 + }, + { + "epoch": 2.4384007438400745, + "grad_norm": 3.812704086303711, + "learning_rate": 9.9947579994758e-06, + "loss": 0.2772, + "step": 5245 + }, + { + "epoch": 2.4388656438865643, + "grad_norm": 11.097214698791504, + "learning_rate": 9.9947569994757e-06, + "loss": 0.7632, + "step": 5246 + }, + { + "epoch": 2.439330543933054, + "grad_norm": 5.436697006225586, + "learning_rate": 9.994755999475601e-06, + "loss": 0.2213, + "step": 5247 + }, + { + "epoch": 2.4397954439795444, + "grad_norm": 10.578607559204102, + "learning_rate": 9.994754999475502e-06, + "loss": 0.8084, + "step": 5248 + }, + { + "epoch": 2.4402603440260346, + "grad_norm": 10.860505104064941, + "learning_rate": 9.9947539994754e-06, + "loss": 0.6276, + "step": 5249 + }, + { + "epoch": 2.4407252440725244, + "grad_norm": 4.6581807136535645, + "learning_rate": 9.994752999475301e-06, + "loss": 0.2925, + "step": 5250 + }, + { + "epoch": 2.4411901441190142, + "grad_norm": 4.497875690460205, + "learning_rate": 9.9947519994752e-06, + "loss": 0.3979, + "step": 5251 + }, + { + "epoch": 2.4416550441655045, + "grad_norm": 4.442148685455322, + "learning_rate": 9.994750999475101e-06, + "loss": 0.3548, + "step": 5252 + }, + { + "epoch": 2.4421199442119943, + "grad_norm": 5.821349143981934, + "learning_rate": 9.994749999475e-06, + "loss": 0.4362, + "step": 5253 + }, + { + "epoch": 2.4425848442584845, + "grad_norm": 6.582073211669922, + "learning_rate": 9.9947489994749e-06, + "loss": 0.3434, + "step": 5254 + }, + { + "epoch": 2.4430497443049743, + "grad_norm": 5.711733818054199, + "learning_rate": 9.994747999474801e-06, + "loss": 0.7919, + "step": 5255 + }, + { + "epoch": 2.4435146443514646, + "grad_norm": 5.081416130065918, + "learning_rate": 9.9947469994747e-06, + "loss": 0.4736, + "step": 5256 + }, + { + "epoch": 2.4439795443979544, + "grad_norm": 5.732506275177002, + "learning_rate": 9.994745999474601e-06, + "loss": 0.5436, + "step": 5257 + }, + { + "epoch": 2.4444444444444446, + "grad_norm": 3.6825671195983887, + "learning_rate": 9.9947449994745e-06, + "loss": 0.3058, + "step": 5258 + }, + { + "epoch": 2.4449093444909344, + "grad_norm": 4.588293075561523, + "learning_rate": 9.9947439994744e-06, + "loss": 0.3656, + "step": 5259 + }, + { + "epoch": 2.4453742445374242, + "grad_norm": 5.256502628326416, + "learning_rate": 9.994742999474301e-06, + "loss": 0.3255, + "step": 5260 + }, + { + "epoch": 2.4458391445839145, + "grad_norm": 6.124085426330566, + "learning_rate": 9.994741999474202e-06, + "loss": 0.4625, + "step": 5261 + }, + { + "epoch": 2.4463040446304043, + "grad_norm": 4.387022495269775, + "learning_rate": 9.9947409994741e-06, + "loss": 0.3366, + "step": 5262 + }, + { + "epoch": 2.4467689446768945, + "grad_norm": 5.703780174255371, + "learning_rate": 9.994739999474001e-06, + "loss": 0.3589, + "step": 5263 + }, + { + "epoch": 2.4472338447233843, + "grad_norm": 4.089512348175049, + "learning_rate": 9.9947389994739e-06, + "loss": 0.3759, + "step": 5264 + }, + { + "epoch": 2.4476987447698746, + "grad_norm": 6.605905055999756, + "learning_rate": 9.994737999473801e-06, + "loss": 0.5677, + "step": 5265 + }, + { + "epoch": 2.4481636448163644, + "grad_norm": 6.175969123840332, + "learning_rate": 9.9947369994737e-06, + "loss": 0.2969, + "step": 5266 + }, + { + "epoch": 2.4486285448628546, + "grad_norm": 5.34889554977417, + "learning_rate": 9.9947359994736e-06, + "loss": 0.3121, + "step": 5267 + }, + { + "epoch": 2.4490934449093444, + "grad_norm": 3.9868452548980713, + "learning_rate": 9.994734999473501e-06, + "loss": 0.3757, + "step": 5268 + }, + { + "epoch": 2.4495583449558347, + "grad_norm": 6.085297584533691, + "learning_rate": 9.9947339994734e-06, + "loss": 0.5591, + "step": 5269 + }, + { + "epoch": 2.4500232450023245, + "grad_norm": 5.5268096923828125, + "learning_rate": 9.994732999473301e-06, + "loss": 0.3159, + "step": 5270 + }, + { + "epoch": 2.4504881450488147, + "grad_norm": 8.182568550109863, + "learning_rate": 9.9947319994732e-06, + "loss": 0.6864, + "step": 5271 + }, + { + "epoch": 2.4509530450953045, + "grad_norm": 7.020611763000488, + "learning_rate": 9.9947309994731e-06, + "loss": 0.481, + "step": 5272 + }, + { + "epoch": 2.4514179451417943, + "grad_norm": 5.505092144012451, + "learning_rate": 9.994729999473001e-06, + "loss": 0.4216, + "step": 5273 + }, + { + "epoch": 2.4518828451882846, + "grad_norm": 6.49795389175415, + "learning_rate": 9.994728999472902e-06, + "loss": 0.4512, + "step": 5274 + }, + { + "epoch": 2.4523477452347744, + "grad_norm": 4.938880443572998, + "learning_rate": 9.994727999472801e-06, + "loss": 0.4138, + "step": 5275 + }, + { + "epoch": 2.4528126452812646, + "grad_norm": 5.042511940002441, + "learning_rate": 9.9947269994727e-06, + "loss": 0.4899, + "step": 5276 + }, + { + "epoch": 2.4532775453277544, + "grad_norm": 6.8693037033081055, + "learning_rate": 9.9947259994726e-06, + "loss": 0.3347, + "step": 5277 + }, + { + "epoch": 2.4537424453742447, + "grad_norm": 4.354583740234375, + "learning_rate": 9.994724999472501e-06, + "loss": 0.3454, + "step": 5278 + }, + { + "epoch": 2.4542073454207345, + "grad_norm": 7.888007164001465, + "learning_rate": 9.994723999472402e-06, + "loss": 0.4411, + "step": 5279 + }, + { + "epoch": 2.4546722454672247, + "grad_norm": 5.053161144256592, + "learning_rate": 9.9947229994723e-06, + "loss": 0.5604, + "step": 5280 + }, + { + "epoch": 2.4551371455137145, + "grad_norm": 8.824329376220703, + "learning_rate": 9.994721999472201e-06, + "loss": 0.2247, + "step": 5281 + }, + { + "epoch": 2.4556020455602043, + "grad_norm": 4.139098167419434, + "learning_rate": 9.9947209994721e-06, + "loss": 0.4173, + "step": 5282 + }, + { + "epoch": 2.4560669456066946, + "grad_norm": 9.702286720275879, + "learning_rate": 9.994719999472001e-06, + "loss": 0.5728, + "step": 5283 + }, + { + "epoch": 2.4565318456531844, + "grad_norm": 4.895423889160156, + "learning_rate": 9.9947189994719e-06, + "loss": 0.441, + "step": 5284 + }, + { + "epoch": 2.4569967456996746, + "grad_norm": 7.065365314483643, + "learning_rate": 9.9947179994718e-06, + "loss": 0.6001, + "step": 5285 + }, + { + "epoch": 2.4574616457461644, + "grad_norm": 6.090455532073975, + "learning_rate": 9.994716999471701e-06, + "loss": 0.4585, + "step": 5286 + }, + { + "epoch": 2.4579265457926547, + "grad_norm": 6.306117534637451, + "learning_rate": 9.9947159994716e-06, + "loss": 0.7635, + "step": 5287 + }, + { + "epoch": 2.4583914458391445, + "grad_norm": 3.2298686504364014, + "learning_rate": 9.994714999471501e-06, + "loss": 0.1627, + "step": 5288 + }, + { + "epoch": 2.4588563458856347, + "grad_norm": 3.4888288974761963, + "learning_rate": 9.9947139994714e-06, + "loss": 0.2605, + "step": 5289 + }, + { + "epoch": 2.4593212459321245, + "grad_norm": 5.425975799560547, + "learning_rate": 9.9947129994713e-06, + "loss": 0.3976, + "step": 5290 + }, + { + "epoch": 2.459786145978615, + "grad_norm": 5.587878704071045, + "learning_rate": 9.994711999471201e-06, + "loss": 0.6601, + "step": 5291 + }, + { + "epoch": 2.4602510460251046, + "grad_norm": 7.357375621795654, + "learning_rate": 9.994710999471102e-06, + "loss": 0.4711, + "step": 5292 + }, + { + "epoch": 2.460715946071595, + "grad_norm": 5.173671722412109, + "learning_rate": 9.994709999471001e-06, + "loss": 0.2965, + "step": 5293 + }, + { + "epoch": 2.4611808461180846, + "grad_norm": 3.916966676712036, + "learning_rate": 9.994708999470902e-06, + "loss": 0.2599, + "step": 5294 + }, + { + "epoch": 2.4616457461645744, + "grad_norm": 4.252391338348389, + "learning_rate": 9.9947079994708e-06, + "loss": 0.4078, + "step": 5295 + }, + { + "epoch": 2.4621106462110647, + "grad_norm": 4.953356742858887, + "learning_rate": 9.9947069994707e-06, + "loss": 0.3514, + "step": 5296 + }, + { + "epoch": 2.4625755462575545, + "grad_norm": 3.9054934978485107, + "learning_rate": 9.9947059994706e-06, + "loss": 0.311, + "step": 5297 + }, + { + "epoch": 2.4630404463040447, + "grad_norm": 6.321239948272705, + "learning_rate": 9.9947049994705e-06, + "loss": 0.564, + "step": 5298 + }, + { + "epoch": 2.4635053463505345, + "grad_norm": 4.978963851928711, + "learning_rate": 9.994703999470401e-06, + "loss": 0.452, + "step": 5299 + }, + { + "epoch": 2.4639702463970248, + "grad_norm": 4.978423118591309, + "learning_rate": 9.9947029994703e-06, + "loss": 0.3283, + "step": 5300 + }, + { + "epoch": 2.4644351464435146, + "grad_norm": 5.936254024505615, + "learning_rate": 9.994701999470201e-06, + "loss": 0.5677, + "step": 5301 + }, + { + "epoch": 2.464900046490005, + "grad_norm": 9.727485656738281, + "learning_rate": 9.9947009994701e-06, + "loss": 0.1504, + "step": 5302 + }, + { + "epoch": 2.4653649465364946, + "grad_norm": 4.295800685882568, + "learning_rate": 9.99469999947e-06, + "loss": 0.5011, + "step": 5303 + }, + { + "epoch": 2.4658298465829844, + "grad_norm": 8.245399475097656, + "learning_rate": 9.994698999469901e-06, + "loss": 0.3445, + "step": 5304 + }, + { + "epoch": 2.4662947466294747, + "grad_norm": 6.13284969329834, + "learning_rate": 9.994697999469802e-06, + "loss": 0.6446, + "step": 5305 + }, + { + "epoch": 2.4667596466759645, + "grad_norm": 5.561898708343506, + "learning_rate": 9.994696999469701e-06, + "loss": 0.4217, + "step": 5306 + }, + { + "epoch": 2.4672245467224547, + "grad_norm": 7.466893196105957, + "learning_rate": 9.9946959994696e-06, + "loss": 0.5553, + "step": 5307 + }, + { + "epoch": 2.4676894467689445, + "grad_norm": 5.626180171966553, + "learning_rate": 9.9946949994695e-06, + "loss": 0.3727, + "step": 5308 + }, + { + "epoch": 2.4681543468154348, + "grad_norm": 5.051806449890137, + "learning_rate": 9.9946939994694e-06, + "loss": 0.5431, + "step": 5309 + }, + { + "epoch": 2.4686192468619246, + "grad_norm": 10.380768775939941, + "learning_rate": 9.9946929994693e-06, + "loss": 0.8105, + "step": 5310 + }, + { + "epoch": 2.469084146908415, + "grad_norm": 8.56005573272705, + "learning_rate": 9.994691999469201e-06, + "loss": 0.6712, + "step": 5311 + }, + { + "epoch": 2.4695490469549046, + "grad_norm": 8.192479133605957, + "learning_rate": 9.994690999469102e-06, + "loss": 0.702, + "step": 5312 + }, + { + "epoch": 2.470013947001395, + "grad_norm": 5.681506633758545, + "learning_rate": 9.994689999469e-06, + "loss": 0.3143, + "step": 5313 + }, + { + "epoch": 2.4704788470478847, + "grad_norm": 6.0106120109558105, + "learning_rate": 9.994688999468901e-06, + "loss": 0.5901, + "step": 5314 + }, + { + "epoch": 2.470943747094375, + "grad_norm": 3.9780685901641846, + "learning_rate": 9.9946879994688e-06, + "loss": 0.2844, + "step": 5315 + }, + { + "epoch": 2.4714086471408647, + "grad_norm": 5.24560022354126, + "learning_rate": 9.9946869994687e-06, + "loss": 0.6176, + "step": 5316 + }, + { + "epoch": 2.4718735471873545, + "grad_norm": 5.7736310958862305, + "learning_rate": 9.994685999468601e-06, + "loss": 0.4588, + "step": 5317 + }, + { + "epoch": 2.4723384472338448, + "grad_norm": 8.623875617980957, + "learning_rate": 9.9946849994685e-06, + "loss": 0.865, + "step": 5318 + }, + { + "epoch": 2.4728033472803346, + "grad_norm": 4.774918079376221, + "learning_rate": 9.994683999468401e-06, + "loss": 0.2028, + "step": 5319 + }, + { + "epoch": 2.473268247326825, + "grad_norm": 3.2313311100006104, + "learning_rate": 9.9946829994683e-06, + "loss": 0.2046, + "step": 5320 + }, + { + "epoch": 2.4737331473733146, + "grad_norm": 5.737895488739014, + "learning_rate": 9.9946819994682e-06, + "loss": 0.5488, + "step": 5321 + }, + { + "epoch": 2.474198047419805, + "grad_norm": 5.313209533691406, + "learning_rate": 9.9946809994681e-06, + "loss": 0.5006, + "step": 5322 + }, + { + "epoch": 2.4746629474662947, + "grad_norm": 6.225863933563232, + "learning_rate": 9.994679999468e-06, + "loss": 0.4419, + "step": 5323 + }, + { + "epoch": 2.475127847512785, + "grad_norm": 4.823225021362305, + "learning_rate": 9.994678999467901e-06, + "loss": 0.219, + "step": 5324 + }, + { + "epoch": 2.4755927475592747, + "grad_norm": 5.787018299102783, + "learning_rate": 9.994677999467802e-06, + "loss": 0.4319, + "step": 5325 + }, + { + "epoch": 2.4760576476057645, + "grad_norm": 5.088578701019287, + "learning_rate": 9.9946769994677e-06, + "loss": 0.4089, + "step": 5326 + }, + { + "epoch": 2.4765225476522548, + "grad_norm": 4.808043956756592, + "learning_rate": 9.994675999467601e-06, + "loss": 0.2923, + "step": 5327 + }, + { + "epoch": 2.476987447698745, + "grad_norm": 7.8184051513671875, + "learning_rate": 9.9946749994675e-06, + "loss": 0.3786, + "step": 5328 + }, + { + "epoch": 2.477452347745235, + "grad_norm": 5.6823835372924805, + "learning_rate": 9.9946739994674e-06, + "loss": 0.4508, + "step": 5329 + }, + { + "epoch": 2.4779172477917246, + "grad_norm": 4.808340072631836, + "learning_rate": 9.994672999467302e-06, + "loss": 0.3578, + "step": 5330 + }, + { + "epoch": 2.478382147838215, + "grad_norm": 6.2468581199646, + "learning_rate": 9.9946719994672e-06, + "loss": 0.4906, + "step": 5331 + }, + { + "epoch": 2.4788470478847047, + "grad_norm": 5.57279634475708, + "learning_rate": 9.994670999467101e-06, + "loss": 0.4576, + "step": 5332 + }, + { + "epoch": 2.479311947931195, + "grad_norm": 4.69193696975708, + "learning_rate": 9.994669999467e-06, + "loss": 0.3126, + "step": 5333 + }, + { + "epoch": 2.4797768479776847, + "grad_norm": 5.802849769592285, + "learning_rate": 9.9946689994669e-06, + "loss": 0.4366, + "step": 5334 + }, + { + "epoch": 2.480241748024175, + "grad_norm": 3.854609489440918, + "learning_rate": 9.9946679994668e-06, + "loss": 0.2976, + "step": 5335 + }, + { + "epoch": 2.4807066480706648, + "grad_norm": 5.945973873138428, + "learning_rate": 9.9946669994667e-06, + "loss": 0.5861, + "step": 5336 + }, + { + "epoch": 2.481171548117155, + "grad_norm": 4.63891077041626, + "learning_rate": 9.994665999466601e-06, + "loss": 0.2726, + "step": 5337 + }, + { + "epoch": 2.481636448163645, + "grad_norm": 6.370110988616943, + "learning_rate": 9.994664999466502e-06, + "loss": 0.4949, + "step": 5338 + }, + { + "epoch": 2.4821013482101346, + "grad_norm": 6.823953628540039, + "learning_rate": 9.9946639994664e-06, + "loss": 0.5866, + "step": 5339 + }, + { + "epoch": 2.482566248256625, + "grad_norm": 4.536090850830078, + "learning_rate": 9.9946629994663e-06, + "loss": 0.4158, + "step": 5340 + }, + { + "epoch": 2.4830311483031147, + "grad_norm": 9.739404678344727, + "learning_rate": 9.9946619994662e-06, + "loss": 0.4268, + "step": 5341 + }, + { + "epoch": 2.483496048349605, + "grad_norm": 9.026057243347168, + "learning_rate": 9.994660999466101e-06, + "loss": 0.6345, + "step": 5342 + }, + { + "epoch": 2.4839609483960947, + "grad_norm": 5.59595251083374, + "learning_rate": 9.994659999466002e-06, + "loss": 0.5375, + "step": 5343 + }, + { + "epoch": 2.484425848442585, + "grad_norm": 9.662765502929688, + "learning_rate": 9.9946589994659e-06, + "loss": 0.4483, + "step": 5344 + }, + { + "epoch": 2.4848907484890748, + "grad_norm": 5.626204967498779, + "learning_rate": 9.994657999465801e-06, + "loss": 0.3451, + "step": 5345 + }, + { + "epoch": 2.485355648535565, + "grad_norm": 5.072089195251465, + "learning_rate": 9.9946569994657e-06, + "loss": 0.4213, + "step": 5346 + }, + { + "epoch": 2.485820548582055, + "grad_norm": 7.4724884033203125, + "learning_rate": 9.9946559994656e-06, + "loss": 0.7282, + "step": 5347 + }, + { + "epoch": 2.4862854486285446, + "grad_norm": 7.921525001525879, + "learning_rate": 9.994654999465501e-06, + "loss": 0.5647, + "step": 5348 + }, + { + "epoch": 2.486750348675035, + "grad_norm": 7.488558292388916, + "learning_rate": 9.9946539994654e-06, + "loss": 0.448, + "step": 5349 + }, + { + "epoch": 2.487215248721525, + "grad_norm": 5.996471405029297, + "learning_rate": 9.994652999465301e-06, + "loss": 0.5363, + "step": 5350 + }, + { + "epoch": 2.487680148768015, + "grad_norm": 5.962993144989014, + "learning_rate": 9.9946519994652e-06, + "loss": 0.6218, + "step": 5351 + }, + { + "epoch": 2.4881450488145047, + "grad_norm": 8.791666984558105, + "learning_rate": 9.9946509994651e-06, + "loss": 0.3601, + "step": 5352 + }, + { + "epoch": 2.488609948860995, + "grad_norm": 5.284543037414551, + "learning_rate": 9.994649999465e-06, + "loss": 0.4124, + "step": 5353 + }, + { + "epoch": 2.4890748489074848, + "grad_norm": 4.901197910308838, + "learning_rate": 9.9946489994649e-06, + "loss": 0.3558, + "step": 5354 + }, + { + "epoch": 2.489539748953975, + "grad_norm": 9.018011093139648, + "learning_rate": 9.994647999464801e-06, + "loss": 0.6152, + "step": 5355 + }, + { + "epoch": 2.490004649000465, + "grad_norm": 4.06335973739624, + "learning_rate": 9.994646999464702e-06, + "loss": 0.2668, + "step": 5356 + }, + { + "epoch": 2.490469549046955, + "grad_norm": 5.016957759857178, + "learning_rate": 9.9946459994646e-06, + "loss": 0.4999, + "step": 5357 + }, + { + "epoch": 2.490934449093445, + "grad_norm": 5.796869277954102, + "learning_rate": 9.994644999464501e-06, + "loss": 0.448, + "step": 5358 + }, + { + "epoch": 2.491399349139935, + "grad_norm": 5.230278968811035, + "learning_rate": 9.9946439994644e-06, + "loss": 0.4872, + "step": 5359 + }, + { + "epoch": 2.491864249186425, + "grad_norm": 7.723790645599365, + "learning_rate": 9.994642999464301e-06, + "loss": 0.9087, + "step": 5360 + }, + { + "epoch": 2.4923291492329147, + "grad_norm": 5.411709308624268, + "learning_rate": 9.994641999464202e-06, + "loss": 0.3754, + "step": 5361 + }, + { + "epoch": 2.492794049279405, + "grad_norm": 4.71145486831665, + "learning_rate": 9.9946409994641e-06, + "loss": 0.5409, + "step": 5362 + }, + { + "epoch": 2.4932589493258948, + "grad_norm": 3.760636329650879, + "learning_rate": 9.994639999464001e-06, + "loss": 0.2579, + "step": 5363 + }, + { + "epoch": 2.493723849372385, + "grad_norm": 6.07564640045166, + "learning_rate": 9.9946389994639e-06, + "loss": 0.5554, + "step": 5364 + }, + { + "epoch": 2.494188749418875, + "grad_norm": 5.86343240737915, + "learning_rate": 9.9946379994638e-06, + "loss": 0.4881, + "step": 5365 + }, + { + "epoch": 2.494653649465365, + "grad_norm": 4.629968166351318, + "learning_rate": 9.9946369994637e-06, + "loss": 0.4733, + "step": 5366 + }, + { + "epoch": 2.495118549511855, + "grad_norm": 5.545675754547119, + "learning_rate": 9.9946359994636e-06, + "loss": 0.6131, + "step": 5367 + }, + { + "epoch": 2.495583449558345, + "grad_norm": 5.994877338409424, + "learning_rate": 9.994634999463501e-06, + "loss": 0.3921, + "step": 5368 + }, + { + "epoch": 2.496048349604835, + "grad_norm": 8.490959167480469, + "learning_rate": 9.994633999463402e-06, + "loss": 0.7008, + "step": 5369 + }, + { + "epoch": 2.496513249651325, + "grad_norm": 8.199443817138672, + "learning_rate": 9.9946329994633e-06, + "loss": 0.7951, + "step": 5370 + }, + { + "epoch": 2.496978149697815, + "grad_norm": 3.3102786540985107, + "learning_rate": 9.994631999463201e-06, + "loss": 0.2585, + "step": 5371 + }, + { + "epoch": 2.497443049744305, + "grad_norm": 10.333353996276855, + "learning_rate": 9.9946309994631e-06, + "loss": 1.2911, + "step": 5372 + }, + { + "epoch": 2.497907949790795, + "grad_norm": 4.321412086486816, + "learning_rate": 9.994629999463001e-06, + "loss": 0.2607, + "step": 5373 + }, + { + "epoch": 2.498372849837285, + "grad_norm": 6.191905498504639, + "learning_rate": 9.994628999462902e-06, + "loss": 0.4563, + "step": 5374 + }, + { + "epoch": 2.498837749883775, + "grad_norm": 5.426208019256592, + "learning_rate": 9.9946279994628e-06, + "loss": 0.5653, + "step": 5375 + }, + { + "epoch": 2.499302649930265, + "grad_norm": 5.140883922576904, + "learning_rate": 9.994626999462701e-06, + "loss": 0.3959, + "step": 5376 + }, + { + "epoch": 2.499767549976755, + "grad_norm": 5.478837013244629, + "learning_rate": 9.9946259994626e-06, + "loss": 0.2945, + "step": 5377 + }, + { + "epoch": 2.500232450023245, + "grad_norm": 3.438326358795166, + "learning_rate": 9.994624999462501e-06, + "loss": 0.1883, + "step": 5378 + }, + { + "epoch": 2.500697350069735, + "grad_norm": 4.122995376586914, + "learning_rate": 9.9946239994624e-06, + "loss": 0.4108, + "step": 5379 + }, + { + "epoch": 2.501162250116225, + "grad_norm": 7.692357063293457, + "learning_rate": 9.9946229994623e-06, + "loss": 0.7831, + "step": 5380 + }, + { + "epoch": 2.501627150162715, + "grad_norm": 4.58071231842041, + "learning_rate": 9.994621999462201e-06, + "loss": 0.3376, + "step": 5381 + }, + { + "epoch": 2.502092050209205, + "grad_norm": 2.581460475921631, + "learning_rate": 9.994620999462102e-06, + "loss": 0.1505, + "step": 5382 + }, + { + "epoch": 2.502556950255695, + "grad_norm": 3.771118640899658, + "learning_rate": 9.994619999462e-06, + "loss": 0.2, + "step": 5383 + }, + { + "epoch": 2.503021850302185, + "grad_norm": 4.155485153198242, + "learning_rate": 9.9946189994619e-06, + "loss": 0.3769, + "step": 5384 + }, + { + "epoch": 2.5034867503486753, + "grad_norm": 3.8352463245391846, + "learning_rate": 9.9946179994618e-06, + "loss": 0.3794, + "step": 5385 + }, + { + "epoch": 2.503951650395165, + "grad_norm": 5.2496867179870605, + "learning_rate": 9.994616999461701e-06, + "loss": 0.3591, + "step": 5386 + }, + { + "epoch": 2.504416550441655, + "grad_norm": 6.631891250610352, + "learning_rate": 9.994615999461602e-06, + "loss": 0.712, + "step": 5387 + }, + { + "epoch": 2.504881450488145, + "grad_norm": 5.092591762542725, + "learning_rate": 9.9946149994615e-06, + "loss": 0.2335, + "step": 5388 + }, + { + "epoch": 2.505346350534635, + "grad_norm": 4.839165210723877, + "learning_rate": 9.994613999461401e-06, + "loss": 0.3902, + "step": 5389 + }, + { + "epoch": 2.505811250581125, + "grad_norm": 4.659469127655029, + "learning_rate": 9.9946129994613e-06, + "loss": 0.3089, + "step": 5390 + }, + { + "epoch": 2.506276150627615, + "grad_norm": 5.975346565246582, + "learning_rate": 9.994611999461201e-06, + "loss": 0.4612, + "step": 5391 + }, + { + "epoch": 2.506741050674105, + "grad_norm": 4.196157932281494, + "learning_rate": 9.9946109994611e-06, + "loss": 0.2778, + "step": 5392 + }, + { + "epoch": 2.507205950720595, + "grad_norm": 6.456628322601318, + "learning_rate": 9.994609999461e-06, + "loss": 0.5065, + "step": 5393 + }, + { + "epoch": 2.5076708507670853, + "grad_norm": 4.602474212646484, + "learning_rate": 9.994608999460901e-06, + "loss": 0.3939, + "step": 5394 + }, + { + "epoch": 2.508135750813575, + "grad_norm": 5.548401832580566, + "learning_rate": 9.9946079994608e-06, + "loss": 0.4897, + "step": 5395 + }, + { + "epoch": 2.508600650860065, + "grad_norm": 10.150665283203125, + "learning_rate": 9.994606999460701e-06, + "loss": 0.9934, + "step": 5396 + }, + { + "epoch": 2.509065550906555, + "grad_norm": 4.789935111999512, + "learning_rate": 9.9946059994606e-06, + "loss": 0.2596, + "step": 5397 + }, + { + "epoch": 2.509530450953045, + "grad_norm": 6.656611442565918, + "learning_rate": 9.9946049994605e-06, + "loss": 0.5548, + "step": 5398 + }, + { + "epoch": 2.509995350999535, + "grad_norm": 7.697559356689453, + "learning_rate": 9.994603999460401e-06, + "loss": 0.7883, + "step": 5399 + }, + { + "epoch": 2.510460251046025, + "grad_norm": 7.836308002471924, + "learning_rate": 9.994602999460302e-06, + "loss": 1.0623, + "step": 5400 + }, + { + "epoch": 2.5109251510925152, + "grad_norm": 8.500492095947266, + "learning_rate": 9.9946019994602e-06, + "loss": 1.1749, + "step": 5401 + }, + { + "epoch": 2.511390051139005, + "grad_norm": 4.52586030960083, + "learning_rate": 9.994600999460101e-06, + "loss": 0.3084, + "step": 5402 + }, + { + "epoch": 2.5118549511854953, + "grad_norm": 6.213005065917969, + "learning_rate": 9.99459999946e-06, + "loss": 0.5057, + "step": 5403 + }, + { + "epoch": 2.512319851231985, + "grad_norm": 7.190744876861572, + "learning_rate": 9.994598999459901e-06, + "loss": 0.6976, + "step": 5404 + }, + { + "epoch": 2.512784751278475, + "grad_norm": 7.690081596374512, + "learning_rate": 9.9945979994598e-06, + "loss": 0.532, + "step": 5405 + }, + { + "epoch": 2.513249651324965, + "grad_norm": 6.043677806854248, + "learning_rate": 9.9945969994597e-06, + "loss": 0.6484, + "step": 5406 + }, + { + "epoch": 2.5137145513714554, + "grad_norm": 4.803064823150635, + "learning_rate": 9.994595999459601e-06, + "loss": 0.3587, + "step": 5407 + }, + { + "epoch": 2.514179451417945, + "grad_norm": 4.746696472167969, + "learning_rate": 9.9945949994595e-06, + "loss": 0.2821, + "step": 5408 + }, + { + "epoch": 2.514644351464435, + "grad_norm": 5.982560157775879, + "learning_rate": 9.994593999459401e-06, + "loss": 0.5064, + "step": 5409 + }, + { + "epoch": 2.5151092515109252, + "grad_norm": 6.306202411651611, + "learning_rate": 9.9945929994593e-06, + "loss": 0.5063, + "step": 5410 + }, + { + "epoch": 2.515574151557415, + "grad_norm": 4.985077381134033, + "learning_rate": 9.9945919994592e-06, + "loss": 0.4596, + "step": 5411 + }, + { + "epoch": 2.5160390516039053, + "grad_norm": 6.4831366539001465, + "learning_rate": 9.994590999459101e-06, + "loss": 0.6085, + "step": 5412 + }, + { + "epoch": 2.516503951650395, + "grad_norm": 6.058990955352783, + "learning_rate": 9.994589999459002e-06, + "loss": 0.3502, + "step": 5413 + }, + { + "epoch": 2.5169688516968853, + "grad_norm": 4.731683254241943, + "learning_rate": 9.994588999458901e-06, + "loss": 0.4697, + "step": 5414 + }, + { + "epoch": 2.517433751743375, + "grad_norm": 4.892797470092773, + "learning_rate": 9.994587999458802e-06, + "loss": 0.444, + "step": 5415 + }, + { + "epoch": 2.5178986517898654, + "grad_norm": 8.090255737304688, + "learning_rate": 9.9945869994587e-06, + "loss": 0.7877, + "step": 5416 + }, + { + "epoch": 2.518363551836355, + "grad_norm": 6.22312068939209, + "learning_rate": 9.994585999458601e-06, + "loss": 0.6014, + "step": 5417 + }, + { + "epoch": 2.518828451882845, + "grad_norm": 9.931392669677734, + "learning_rate": 9.9945849994585e-06, + "loss": 0.853, + "step": 5418 + }, + { + "epoch": 2.5192933519293352, + "grad_norm": 5.241401195526123, + "learning_rate": 9.9945839994584e-06, + "loss": 0.4392, + "step": 5419 + }, + { + "epoch": 2.519758251975825, + "grad_norm": 6.418260097503662, + "learning_rate": 9.994582999458301e-06, + "loss": 0.5864, + "step": 5420 + }, + { + "epoch": 2.5202231520223153, + "grad_norm": 8.675437927246094, + "learning_rate": 9.9945819994582e-06, + "loss": 0.4227, + "step": 5421 + }, + { + "epoch": 2.520688052068805, + "grad_norm": 7.935267925262451, + "learning_rate": 9.994580999458101e-06, + "loss": 0.4696, + "step": 5422 + }, + { + "epoch": 2.5211529521152953, + "grad_norm": 5.639540195465088, + "learning_rate": 9.994579999458e-06, + "loss": 0.5727, + "step": 5423 + }, + { + "epoch": 2.521617852161785, + "grad_norm": 5.571815490722656, + "learning_rate": 9.9945789994579e-06, + "loss": 0.5411, + "step": 5424 + }, + { + "epoch": 2.5220827522082754, + "grad_norm": 4.261096477508545, + "learning_rate": 9.994577999457801e-06, + "loss": 0.1934, + "step": 5425 + }, + { + "epoch": 2.522547652254765, + "grad_norm": 7.774321556091309, + "learning_rate": 9.9945769994577e-06, + "loss": 0.3258, + "step": 5426 + }, + { + "epoch": 2.523012552301255, + "grad_norm": 9.480228424072266, + "learning_rate": 9.994575999457601e-06, + "loss": 0.7339, + "step": 5427 + }, + { + "epoch": 2.5234774523477452, + "grad_norm": 6.1617021560668945, + "learning_rate": 9.9945749994575e-06, + "loss": 0.4701, + "step": 5428 + }, + { + "epoch": 2.5239423523942355, + "grad_norm": 5.39612340927124, + "learning_rate": 9.9945739994574e-06, + "loss": 0.3232, + "step": 5429 + }, + { + "epoch": 2.5244072524407253, + "grad_norm": 4.7611870765686035, + "learning_rate": 9.994572999457301e-06, + "loss": 0.2526, + "step": 5430 + }, + { + "epoch": 2.524872152487215, + "grad_norm": 7.8655924797058105, + "learning_rate": 9.994571999457202e-06, + "loss": 0.621, + "step": 5431 + }, + { + "epoch": 2.5253370525337053, + "grad_norm": 6.12466287612915, + "learning_rate": 9.9945709994571e-06, + "loss": 0.4839, + "step": 5432 + }, + { + "epoch": 2.525801952580195, + "grad_norm": 5.804339408874512, + "learning_rate": 9.994569999457001e-06, + "loss": 0.4596, + "step": 5433 + }, + { + "epoch": 2.5262668526266854, + "grad_norm": 5.662735462188721, + "learning_rate": 9.9945689994569e-06, + "loss": 0.3297, + "step": 5434 + }, + { + "epoch": 2.526731752673175, + "grad_norm": 5.496204376220703, + "learning_rate": 9.994567999456801e-06, + "loss": 0.5898, + "step": 5435 + }, + { + "epoch": 2.5271966527196654, + "grad_norm": 6.449605941772461, + "learning_rate": 9.9945669994567e-06, + "loss": 0.4197, + "step": 5436 + }, + { + "epoch": 2.5276615527661552, + "grad_norm": 4.7642107009887695, + "learning_rate": 9.9945659994566e-06, + "loss": 0.295, + "step": 5437 + }, + { + "epoch": 2.5281264528126455, + "grad_norm": 6.51174259185791, + "learning_rate": 9.994564999456501e-06, + "loss": 0.5988, + "step": 5438 + }, + { + "epoch": 2.5285913528591353, + "grad_norm": 4.983196258544922, + "learning_rate": 9.9945639994564e-06, + "loss": 0.5346, + "step": 5439 + }, + { + "epoch": 2.529056252905625, + "grad_norm": 4.659725666046143, + "learning_rate": 9.994562999456301e-06, + "loss": 0.4945, + "step": 5440 + }, + { + "epoch": 2.5295211529521153, + "grad_norm": 6.339081764221191, + "learning_rate": 9.9945619994562e-06, + "loss": 0.6624, + "step": 5441 + }, + { + "epoch": 2.5299860529986056, + "grad_norm": 5.656674385070801, + "learning_rate": 9.9945609994561e-06, + "loss": 0.322, + "step": 5442 + }, + { + "epoch": 2.5304509530450954, + "grad_norm": 5.466838836669922, + "learning_rate": 9.994559999456001e-06, + "loss": 0.3809, + "step": 5443 + }, + { + "epoch": 2.530915853091585, + "grad_norm": 6.5073323249816895, + "learning_rate": 9.994558999455902e-06, + "loss": 0.5745, + "step": 5444 + }, + { + "epoch": 2.5313807531380754, + "grad_norm": 4.524012088775635, + "learning_rate": 9.994557999455801e-06, + "loss": 0.4301, + "step": 5445 + }, + { + "epoch": 2.5318456531845652, + "grad_norm": 3.6213579177856445, + "learning_rate": 9.994556999455702e-06, + "loss": 0.1604, + "step": 5446 + }, + { + "epoch": 2.5323105532310555, + "grad_norm": 5.750791549682617, + "learning_rate": 9.9945559994556e-06, + "loss": 0.3065, + "step": 5447 + }, + { + "epoch": 2.5327754532775453, + "grad_norm": 4.543478965759277, + "learning_rate": 9.9945549994555e-06, + "loss": 0.2639, + "step": 5448 + }, + { + "epoch": 2.533240353324035, + "grad_norm": 4.263202667236328, + "learning_rate": 9.9945539994554e-06, + "loss": 0.2952, + "step": 5449 + }, + { + "epoch": 2.5337052533705253, + "grad_norm": 9.024815559387207, + "learning_rate": 9.9945529994553e-06, + "loss": 0.6771, + "step": 5450 + }, + { + "epoch": 2.5341701534170156, + "grad_norm": 5.530869007110596, + "learning_rate": 9.994551999455201e-06, + "loss": 0.3817, + "step": 5451 + }, + { + "epoch": 2.5346350534635054, + "grad_norm": 8.703168869018555, + "learning_rate": 9.9945509994551e-06, + "loss": 1.035, + "step": 5452 + }, + { + "epoch": 2.535099953509995, + "grad_norm": 5.509153842926025, + "learning_rate": 9.994549999455001e-06, + "loss": 0.3954, + "step": 5453 + }, + { + "epoch": 2.5355648535564854, + "grad_norm": 5.017279148101807, + "learning_rate": 9.9945489994549e-06, + "loss": 0.2921, + "step": 5454 + }, + { + "epoch": 2.5360297536029752, + "grad_norm": 10.81110668182373, + "learning_rate": 9.9945479994548e-06, + "loss": 0.4227, + "step": 5455 + }, + { + "epoch": 2.5364946536494655, + "grad_norm": 4.9073028564453125, + "learning_rate": 9.994546999454701e-06, + "loss": 0.2813, + "step": 5456 + }, + { + "epoch": 2.5369595536959553, + "grad_norm": 6.104679584503174, + "learning_rate": 9.994545999454602e-06, + "loss": 0.3849, + "step": 5457 + }, + { + "epoch": 2.5374244537424455, + "grad_norm": 7.653542518615723, + "learning_rate": 9.994544999454501e-06, + "loss": 0.6301, + "step": 5458 + }, + { + "epoch": 2.5378893537889353, + "grad_norm": 5.649990081787109, + "learning_rate": 9.9945439994544e-06, + "loss": 0.5726, + "step": 5459 + }, + { + "epoch": 2.5383542538354256, + "grad_norm": 4.298945426940918, + "learning_rate": 9.9945429994543e-06, + "loss": 0.3844, + "step": 5460 + }, + { + "epoch": 2.5388191538819154, + "grad_norm": 8.787728309631348, + "learning_rate": 9.9945419994542e-06, + "loss": 0.6327, + "step": 5461 + }, + { + "epoch": 2.539284053928405, + "grad_norm": 6.191319942474365, + "learning_rate": 9.9945409994541e-06, + "loss": 0.442, + "step": 5462 + }, + { + "epoch": 2.5397489539748954, + "grad_norm": 4.134575366973877, + "learning_rate": 9.994539999454001e-06, + "loss": 0.3085, + "step": 5463 + }, + { + "epoch": 2.5402138540213857, + "grad_norm": 3.896096706390381, + "learning_rate": 9.994538999453902e-06, + "loss": 0.345, + "step": 5464 + }, + { + "epoch": 2.5406787540678755, + "grad_norm": 5.131866931915283, + "learning_rate": 9.9945379994538e-06, + "loss": 0.2569, + "step": 5465 + }, + { + "epoch": 2.5411436541143653, + "grad_norm": 6.378597259521484, + "learning_rate": 9.994536999453701e-06, + "loss": 0.5247, + "step": 5466 + }, + { + "epoch": 2.5416085541608555, + "grad_norm": 3.8579604625701904, + "learning_rate": 9.9945359994536e-06, + "loss": 0.3005, + "step": 5467 + }, + { + "epoch": 2.5420734542073453, + "grad_norm": 5.033864974975586, + "learning_rate": 9.9945349994535e-06, + "loss": 0.5418, + "step": 5468 + }, + { + "epoch": 2.5425383542538356, + "grad_norm": 3.867838144302368, + "learning_rate": 9.994533999453401e-06, + "loss": 0.2849, + "step": 5469 + }, + { + "epoch": 2.5430032543003254, + "grad_norm": 5.2056145668029785, + "learning_rate": 9.9945329994533e-06, + "loss": 0.3617, + "step": 5470 + }, + { + "epoch": 2.543468154346815, + "grad_norm": 2.8834621906280518, + "learning_rate": 9.994531999453201e-06, + "loss": 0.1696, + "step": 5471 + }, + { + "epoch": 2.5439330543933054, + "grad_norm": 6.933676242828369, + "learning_rate": 9.9945309994531e-06, + "loss": 0.6802, + "step": 5472 + }, + { + "epoch": 2.5443979544397957, + "grad_norm": 7.053461074829102, + "learning_rate": 9.994529999453e-06, + "loss": 0.5937, + "step": 5473 + }, + { + "epoch": 2.5448628544862855, + "grad_norm": 5.276704788208008, + "learning_rate": 9.9945289994529e-06, + "loss": 0.4099, + "step": 5474 + }, + { + "epoch": 2.5453277545327753, + "grad_norm": 4.859157085418701, + "learning_rate": 9.9945279994528e-06, + "loss": 0.4116, + "step": 5475 + }, + { + "epoch": 2.5457926545792655, + "grad_norm": 7.521435737609863, + "learning_rate": 9.994526999452701e-06, + "loss": 0.3316, + "step": 5476 + }, + { + "epoch": 2.5462575546257553, + "grad_norm": 3.8293888568878174, + "learning_rate": 9.994525999452602e-06, + "loss": 0.1988, + "step": 5477 + }, + { + "epoch": 2.5467224546722456, + "grad_norm": 9.669601440429688, + "learning_rate": 9.9945249994525e-06, + "loss": 0.4661, + "step": 5478 + }, + { + "epoch": 2.5471873547187354, + "grad_norm": 5.168359756469727, + "learning_rate": 9.994523999452401e-06, + "loss": 0.2036, + "step": 5479 + }, + { + "epoch": 2.5476522547652256, + "grad_norm": 5.4099297523498535, + "learning_rate": 9.9945229994523e-06, + "loss": 0.437, + "step": 5480 + }, + { + "epoch": 2.5481171548117154, + "grad_norm": 8.440621376037598, + "learning_rate": 9.994521999452201e-06, + "loss": 0.8031, + "step": 5481 + }, + { + "epoch": 2.5485820548582057, + "grad_norm": 5.164533615112305, + "learning_rate": 9.994520999452102e-06, + "loss": 0.3511, + "step": 5482 + }, + { + "epoch": 2.5490469549046955, + "grad_norm": 4.711756229400635, + "learning_rate": 9.994519999452e-06, + "loss": 0.3148, + "step": 5483 + }, + { + "epoch": 2.5495118549511853, + "grad_norm": 6.485785961151123, + "learning_rate": 9.994518999451901e-06, + "loss": 0.4421, + "step": 5484 + }, + { + "epoch": 2.5499767549976755, + "grad_norm": 8.957527160644531, + "learning_rate": 9.9945179994518e-06, + "loss": 0.4384, + "step": 5485 + }, + { + "epoch": 2.5504416550441658, + "grad_norm": 4.007526397705078, + "learning_rate": 9.9945169994517e-06, + "loss": 0.1863, + "step": 5486 + }, + { + "epoch": 2.5509065550906556, + "grad_norm": 8.242445945739746, + "learning_rate": 9.9945159994516e-06, + "loss": 0.8146, + "step": 5487 + }, + { + "epoch": 2.5513714551371454, + "grad_norm": 4.544416904449463, + "learning_rate": 9.9945149994515e-06, + "loss": 0.394, + "step": 5488 + }, + { + "epoch": 2.5518363551836356, + "grad_norm": 5.3691277503967285, + "learning_rate": 9.994513999451401e-06, + "loss": 0.5796, + "step": 5489 + }, + { + "epoch": 2.5523012552301254, + "grad_norm": 9.15218448638916, + "learning_rate": 9.994512999451302e-06, + "loss": 0.7445, + "step": 5490 + }, + { + "epoch": 2.5527661552766157, + "grad_norm": 6.075345516204834, + "learning_rate": 9.9945119994512e-06, + "loss": 0.5361, + "step": 5491 + }, + { + "epoch": 2.5532310553231055, + "grad_norm": 5.5956878662109375, + "learning_rate": 9.9945109994511e-06, + "loss": 0.5279, + "step": 5492 + }, + { + "epoch": 2.5536959553695953, + "grad_norm": 4.79944372177124, + "learning_rate": 9.994509999451e-06, + "loss": 0.4854, + "step": 5493 + }, + { + "epoch": 2.5541608554160855, + "grad_norm": 6.695560932159424, + "learning_rate": 9.994508999450901e-06, + "loss": 0.3543, + "step": 5494 + }, + { + "epoch": 2.5546257554625758, + "grad_norm": 6.081746578216553, + "learning_rate": 9.994507999450802e-06, + "loss": 0.4546, + "step": 5495 + }, + { + "epoch": 2.5550906555090656, + "grad_norm": 9.118782997131348, + "learning_rate": 9.9945069994507e-06, + "loss": 0.3831, + "step": 5496 + }, + { + "epoch": 2.5555555555555554, + "grad_norm": 5.3708176612854, + "learning_rate": 9.994505999450601e-06, + "loss": 0.3942, + "step": 5497 + }, + { + "epoch": 2.5560204556020456, + "grad_norm": 5.123554229736328, + "learning_rate": 9.9945049994505e-06, + "loss": 0.3908, + "step": 5498 + }, + { + "epoch": 2.5564853556485354, + "grad_norm": 7.519163608551025, + "learning_rate": 9.994503999450401e-06, + "loss": 0.4973, + "step": 5499 + }, + { + "epoch": 2.5569502556950257, + "grad_norm": 4.615323543548584, + "learning_rate": 9.994502999450302e-06, + "loss": 0.3964, + "step": 5500 + }, + { + "epoch": 2.5574151557415155, + "grad_norm": 7.1554718017578125, + "learning_rate": 9.9945019994502e-06, + "loss": 0.4771, + "step": 5501 + }, + { + "epoch": 2.5578800557880057, + "grad_norm": 3.784503221511841, + "learning_rate": 9.994500999450101e-06, + "loss": 0.3305, + "step": 5502 + }, + { + "epoch": 2.5583449558344955, + "grad_norm": 6.837099075317383, + "learning_rate": 9.99449999945e-06, + "loss": 0.5717, + "step": 5503 + }, + { + "epoch": 2.5588098558809858, + "grad_norm": 4.898493766784668, + "learning_rate": 9.9944989994499e-06, + "loss": 0.2932, + "step": 5504 + }, + { + "epoch": 2.5592747559274756, + "grad_norm": 5.6409783363342285, + "learning_rate": 9.9944979994498e-06, + "loss": 0.3471, + "step": 5505 + }, + { + "epoch": 2.5597396559739654, + "grad_norm": 5.7542548179626465, + "learning_rate": 9.9944969994497e-06, + "loss": 0.408, + "step": 5506 + }, + { + "epoch": 2.5602045560204556, + "grad_norm": 8.278038024902344, + "learning_rate": 9.994495999449601e-06, + "loss": 0.4585, + "step": 5507 + }, + { + "epoch": 2.560669456066946, + "grad_norm": 6.353991508483887, + "learning_rate": 9.994494999449502e-06, + "loss": 0.5671, + "step": 5508 + }, + { + "epoch": 2.5611343561134357, + "grad_norm": 4.077714443206787, + "learning_rate": 9.9944939994494e-06, + "loss": 0.4439, + "step": 5509 + }, + { + "epoch": 2.5615992561599255, + "grad_norm": 9.492223739624023, + "learning_rate": 9.994492999449301e-06, + "loss": 0.9305, + "step": 5510 + }, + { + "epoch": 2.5620641562064157, + "grad_norm": 5.040796756744385, + "learning_rate": 9.9944919994492e-06, + "loss": 0.4194, + "step": 5511 + }, + { + "epoch": 2.5625290562529055, + "grad_norm": 6.584272861480713, + "learning_rate": 9.994490999449101e-06, + "loss": 0.6169, + "step": 5512 + }, + { + "epoch": 2.5629939562993957, + "grad_norm": 4.812939167022705, + "learning_rate": 9.994489999449002e-06, + "loss": 0.3612, + "step": 5513 + }, + { + "epoch": 2.5634588563458856, + "grad_norm": 5.662762641906738, + "learning_rate": 9.9944889994489e-06, + "loss": 0.4972, + "step": 5514 + }, + { + "epoch": 2.563923756392376, + "grad_norm": 6.0910186767578125, + "learning_rate": 9.994487999448801e-06, + "loss": 0.6027, + "step": 5515 + }, + { + "epoch": 2.5643886564388656, + "grad_norm": 5.150008201599121, + "learning_rate": 9.9944869994487e-06, + "loss": 0.5927, + "step": 5516 + }, + { + "epoch": 2.564853556485356, + "grad_norm": 4.481881141662598, + "learning_rate": 9.9944859994486e-06, + "loss": 0.4318, + "step": 5517 + }, + { + "epoch": 2.5653184565318456, + "grad_norm": 11.201786994934082, + "learning_rate": 9.9944849994485e-06, + "loss": 0.4937, + "step": 5518 + }, + { + "epoch": 2.5657833565783355, + "grad_norm": 5.309040069580078, + "learning_rate": 9.9944839994484e-06, + "loss": 0.5132, + "step": 5519 + }, + { + "epoch": 2.5662482566248257, + "grad_norm": 2.495147705078125, + "learning_rate": 9.994482999448301e-06, + "loss": 0.1356, + "step": 5520 + }, + { + "epoch": 2.566713156671316, + "grad_norm": 4.935389518737793, + "learning_rate": 9.994481999448202e-06, + "loss": 0.3877, + "step": 5521 + }, + { + "epoch": 2.5671780567178057, + "grad_norm": 4.105488300323486, + "learning_rate": 9.9944809994481e-06, + "loss": 0.3897, + "step": 5522 + }, + { + "epoch": 2.5676429567642955, + "grad_norm": 3.724550247192383, + "learning_rate": 9.994479999448001e-06, + "loss": 0.3518, + "step": 5523 + }, + { + "epoch": 2.568107856810786, + "grad_norm": 4.648247718811035, + "learning_rate": 9.9944789994479e-06, + "loss": 0.4257, + "step": 5524 + }, + { + "epoch": 2.5685727568572756, + "grad_norm": 6.816096782684326, + "learning_rate": 9.994477999447801e-06, + "loss": 0.6775, + "step": 5525 + }, + { + "epoch": 2.569037656903766, + "grad_norm": 5.359671592712402, + "learning_rate": 9.994476999447702e-06, + "loss": 0.5116, + "step": 5526 + }, + { + "epoch": 2.5695025569502556, + "grad_norm": 5.25807523727417, + "learning_rate": 9.9944759994476e-06, + "loss": 0.5315, + "step": 5527 + }, + { + "epoch": 2.5699674569967454, + "grad_norm": 5.450571537017822, + "learning_rate": 9.994474999447501e-06, + "loss": 0.3996, + "step": 5528 + }, + { + "epoch": 2.5704323570432357, + "grad_norm": 4.964431285858154, + "learning_rate": 9.9944739994474e-06, + "loss": 0.5102, + "step": 5529 + }, + { + "epoch": 2.570897257089726, + "grad_norm": 7.643819332122803, + "learning_rate": 9.994472999447301e-06, + "loss": 0.462, + "step": 5530 + }, + { + "epoch": 2.5713621571362157, + "grad_norm": 8.995699882507324, + "learning_rate": 9.9944719994472e-06, + "loss": 0.6468, + "step": 5531 + }, + { + "epoch": 2.5718270571827055, + "grad_norm": 6.818390369415283, + "learning_rate": 9.9944709994471e-06, + "loss": 0.7094, + "step": 5532 + }, + { + "epoch": 2.572291957229196, + "grad_norm": 4.53121280670166, + "learning_rate": 9.994469999447001e-06, + "loss": 0.3489, + "step": 5533 + }, + { + "epoch": 2.5727568572756856, + "grad_norm": 3.7444937229156494, + "learning_rate": 9.994468999446902e-06, + "loss": 0.2394, + "step": 5534 + }, + { + "epoch": 2.573221757322176, + "grad_norm": 8.945982933044434, + "learning_rate": 9.9944679994468e-06, + "loss": 0.3778, + "step": 5535 + }, + { + "epoch": 2.5736866573686656, + "grad_norm": 7.051883220672607, + "learning_rate": 9.9944669994467e-06, + "loss": 0.4683, + "step": 5536 + }, + { + "epoch": 2.574151557415156, + "grad_norm": 5.86910343170166, + "learning_rate": 9.9944659994466e-06, + "loss": 0.3008, + "step": 5537 + }, + { + "epoch": 2.5746164574616457, + "grad_norm": 3.8125200271606445, + "learning_rate": 9.994464999446501e-06, + "loss": 0.253, + "step": 5538 + }, + { + "epoch": 2.575081357508136, + "grad_norm": 5.986684799194336, + "learning_rate": 9.994463999446402e-06, + "loss": 0.4378, + "step": 5539 + }, + { + "epoch": 2.5755462575546257, + "grad_norm": 3.1873233318328857, + "learning_rate": 9.9944629994463e-06, + "loss": 0.207, + "step": 5540 + }, + { + "epoch": 2.5760111576011155, + "grad_norm": 8.852529525756836, + "learning_rate": 9.994461999446201e-06, + "loss": 0.6008, + "step": 5541 + }, + { + "epoch": 2.576476057647606, + "grad_norm": 5.682873725891113, + "learning_rate": 9.9944609994461e-06, + "loss": 0.5566, + "step": 5542 + }, + { + "epoch": 2.576940957694096, + "grad_norm": 7.602056980133057, + "learning_rate": 9.994459999446001e-06, + "loss": 0.4901, + "step": 5543 + }, + { + "epoch": 2.577405857740586, + "grad_norm": 5.658641815185547, + "learning_rate": 9.9944589994459e-06, + "loss": 0.5082, + "step": 5544 + }, + { + "epoch": 2.5778707577870756, + "grad_norm": 6.403368949890137, + "learning_rate": 9.9944579994458e-06, + "loss": 0.5136, + "step": 5545 + }, + { + "epoch": 2.578335657833566, + "grad_norm": 5.032139301300049, + "learning_rate": 9.994456999445701e-06, + "loss": 0.5225, + "step": 5546 + }, + { + "epoch": 2.5788005578800557, + "grad_norm": 7.042543888092041, + "learning_rate": 9.9944559994456e-06, + "loss": 0.5133, + "step": 5547 + }, + { + "epoch": 2.579265457926546, + "grad_norm": 7.539360046386719, + "learning_rate": 9.994454999445501e-06, + "loss": 0.4496, + "step": 5548 + }, + { + "epoch": 2.5797303579730357, + "grad_norm": 4.283010959625244, + "learning_rate": 9.9944539994454e-06, + "loss": 0.335, + "step": 5549 + }, + { + "epoch": 2.5801952580195255, + "grad_norm": 5.714997291564941, + "learning_rate": 9.9944529994453e-06, + "loss": 0.6288, + "step": 5550 + }, + { + "epoch": 2.580660158066016, + "grad_norm": 4.936751842498779, + "learning_rate": 9.994451999445201e-06, + "loss": 0.4494, + "step": 5551 + }, + { + "epoch": 2.581125058112506, + "grad_norm": 5.100598335266113, + "learning_rate": 9.994450999445102e-06, + "loss": 0.379, + "step": 5552 + }, + { + "epoch": 2.581589958158996, + "grad_norm": 4.582352638244629, + "learning_rate": 9.994449999445e-06, + "loss": 0.2716, + "step": 5553 + }, + { + "epoch": 2.5820548582054856, + "grad_norm": 5.682868003845215, + "learning_rate": 9.994448999444901e-06, + "loss": 0.3656, + "step": 5554 + }, + { + "epoch": 2.582519758251976, + "grad_norm": 3.640592336654663, + "learning_rate": 9.9944479994448e-06, + "loss": 0.1877, + "step": 5555 + }, + { + "epoch": 2.5829846582984657, + "grad_norm": 7.5013322830200195, + "learning_rate": 9.994446999444701e-06, + "loss": 0.7331, + "step": 5556 + }, + { + "epoch": 2.583449558344956, + "grad_norm": 6.590694427490234, + "learning_rate": 9.9944459994446e-06, + "loss": 0.5877, + "step": 5557 + }, + { + "epoch": 2.5839144583914457, + "grad_norm": 8.131507873535156, + "learning_rate": 9.9944449994445e-06, + "loss": 0.6405, + "step": 5558 + }, + { + "epoch": 2.584379358437936, + "grad_norm": 5.792476177215576, + "learning_rate": 9.994443999444401e-06, + "loss": 0.4322, + "step": 5559 + }, + { + "epoch": 2.584844258484426, + "grad_norm": 5.657492160797119, + "learning_rate": 9.9944429994443e-06, + "loss": 0.4094, + "step": 5560 + }, + { + "epoch": 2.585309158530916, + "grad_norm": 5.215730667114258, + "learning_rate": 9.994441999444201e-06, + "loss": 0.3903, + "step": 5561 + }, + { + "epoch": 2.585774058577406, + "grad_norm": 3.9040069580078125, + "learning_rate": 9.9944409994441e-06, + "loss": 0.222, + "step": 5562 + }, + { + "epoch": 2.5862389586238956, + "grad_norm": 3.317333459854126, + "learning_rate": 9.994439999444e-06, + "loss": 0.238, + "step": 5563 + }, + { + "epoch": 2.586703858670386, + "grad_norm": 3.8548076152801514, + "learning_rate": 9.994438999443901e-06, + "loss": 0.2697, + "step": 5564 + }, + { + "epoch": 2.587168758716876, + "grad_norm": 7.593306541442871, + "learning_rate": 9.994437999443802e-06, + "loss": 0.5282, + "step": 5565 + }, + { + "epoch": 2.587633658763366, + "grad_norm": 7.118558406829834, + "learning_rate": 9.994436999443701e-06, + "loss": 0.3978, + "step": 5566 + }, + { + "epoch": 2.5880985588098557, + "grad_norm": 4.8605217933654785, + "learning_rate": 9.9944359994436e-06, + "loss": 0.248, + "step": 5567 + }, + { + "epoch": 2.588563458856346, + "grad_norm": 5.800318717956543, + "learning_rate": 9.9944349994435e-06, + "loss": 0.5525, + "step": 5568 + }, + { + "epoch": 2.589028358902836, + "grad_norm": 4.454716205596924, + "learning_rate": 9.9944339994434e-06, + "loss": 0.352, + "step": 5569 + }, + { + "epoch": 2.589493258949326, + "grad_norm": 10.811656951904297, + "learning_rate": 9.9944329994433e-06, + "loss": 0.2637, + "step": 5570 + }, + { + "epoch": 2.589958158995816, + "grad_norm": 6.738522052764893, + "learning_rate": 9.9944319994432e-06, + "loss": 0.4202, + "step": 5571 + }, + { + "epoch": 2.5904230590423056, + "grad_norm": 4.266721248626709, + "learning_rate": 9.994430999443101e-06, + "loss": 0.2604, + "step": 5572 + }, + { + "epoch": 2.590887959088796, + "grad_norm": 6.141563892364502, + "learning_rate": 9.994429999443e-06, + "loss": 0.358, + "step": 5573 + }, + { + "epoch": 2.591352859135286, + "grad_norm": 4.272834300994873, + "learning_rate": 9.994428999442901e-06, + "loss": 0.3944, + "step": 5574 + }, + { + "epoch": 2.591817759181776, + "grad_norm": 6.103463649749756, + "learning_rate": 9.9944279994428e-06, + "loss": 0.662, + "step": 5575 + }, + { + "epoch": 2.5922826592282657, + "grad_norm": 8.802579879760742, + "learning_rate": 9.9944269994427e-06, + "loss": 0.5427, + "step": 5576 + }, + { + "epoch": 2.592747559274756, + "grad_norm": 3.8700833320617676, + "learning_rate": 9.994425999442601e-06, + "loss": 0.3002, + "step": 5577 + }, + { + "epoch": 2.5932124593212458, + "grad_norm": 6.849687576293945, + "learning_rate": 9.9944249994425e-06, + "loss": 0.6974, + "step": 5578 + }, + { + "epoch": 2.593677359367736, + "grad_norm": 4.951376914978027, + "learning_rate": 9.994423999442401e-06, + "loss": 0.4385, + "step": 5579 + }, + { + "epoch": 2.594142259414226, + "grad_norm": 4.6672844886779785, + "learning_rate": 9.9944229994423e-06, + "loss": 0.4899, + "step": 5580 + }, + { + "epoch": 2.594607159460716, + "grad_norm": 7.455352783203125, + "learning_rate": 9.9944219994422e-06, + "loss": 0.5728, + "step": 5581 + }, + { + "epoch": 2.595072059507206, + "grad_norm": 4.0948405265808105, + "learning_rate": 9.994420999442101e-06, + "loss": 0.3558, + "step": 5582 + }, + { + "epoch": 2.595536959553696, + "grad_norm": 4.335521697998047, + "learning_rate": 9.994419999442e-06, + "loss": 0.1463, + "step": 5583 + }, + { + "epoch": 2.596001859600186, + "grad_norm": 6.092428207397461, + "learning_rate": 9.994418999441901e-06, + "loss": 0.7355, + "step": 5584 + }, + { + "epoch": 2.5964667596466757, + "grad_norm": 4.13913631439209, + "learning_rate": 9.994417999441802e-06, + "loss": 0.2691, + "step": 5585 + }, + { + "epoch": 2.596931659693166, + "grad_norm": 5.888411998748779, + "learning_rate": 9.9944169994417e-06, + "loss": 0.3278, + "step": 5586 + }, + { + "epoch": 2.597396559739656, + "grad_norm": 2.772752046585083, + "learning_rate": 9.994415999441601e-06, + "loss": 0.1654, + "step": 5587 + }, + { + "epoch": 2.597861459786146, + "grad_norm": 4.54464864730835, + "learning_rate": 9.9944149994415e-06, + "loss": 0.3898, + "step": 5588 + }, + { + "epoch": 2.598326359832636, + "grad_norm": 5.9221296310424805, + "learning_rate": 9.9944139994414e-06, + "loss": 0.2296, + "step": 5589 + }, + { + "epoch": 2.598791259879126, + "grad_norm": 5.196529388427734, + "learning_rate": 9.994412999441301e-06, + "loss": 0.4096, + "step": 5590 + }, + { + "epoch": 2.599256159925616, + "grad_norm": 5.453166484832764, + "learning_rate": 9.9944119994412e-06, + "loss": 0.6124, + "step": 5591 + }, + { + "epoch": 2.599721059972106, + "grad_norm": 7.189730644226074, + "learning_rate": 9.994410999441101e-06, + "loss": 0.4623, + "step": 5592 + }, + { + "epoch": 2.600185960018596, + "grad_norm": 5.971238136291504, + "learning_rate": 9.994409999441e-06, + "loss": 0.6389, + "step": 5593 + }, + { + "epoch": 2.600650860065086, + "grad_norm": 5.966085910797119, + "learning_rate": 9.9944089994409e-06, + "loss": 0.2566, + "step": 5594 + }, + { + "epoch": 2.601115760111576, + "grad_norm": 5.024973392486572, + "learning_rate": 9.994407999440801e-06, + "loss": 0.3892, + "step": 5595 + }, + { + "epoch": 2.601580660158066, + "grad_norm": 5.133455753326416, + "learning_rate": 9.994406999440702e-06, + "loss": 0.3385, + "step": 5596 + }, + { + "epoch": 2.602045560204556, + "grad_norm": 4.446136951446533, + "learning_rate": 9.994405999440601e-06, + "loss": 0.2523, + "step": 5597 + }, + { + "epoch": 2.602510460251046, + "grad_norm": 7.223288536071777, + "learning_rate": 9.994404999440502e-06, + "loss": 0.7346, + "step": 5598 + }, + { + "epoch": 2.602975360297536, + "grad_norm": 6.351736545562744, + "learning_rate": 9.9944039994404e-06, + "loss": 0.4761, + "step": 5599 + }, + { + "epoch": 2.603440260344026, + "grad_norm": 5.203135967254639, + "learning_rate": 9.9944029994403e-06, + "loss": 0.3744, + "step": 5600 + }, + { + "epoch": 2.603905160390516, + "grad_norm": 6.272673606872559, + "learning_rate": 9.9944019994402e-06, + "loss": 0.4288, + "step": 5601 + }, + { + "epoch": 2.604370060437006, + "grad_norm": 7.323306560516357, + "learning_rate": 9.994400999440101e-06, + "loss": 0.748, + "step": 5602 + }, + { + "epoch": 2.604834960483496, + "grad_norm": 5.894902229309082, + "learning_rate": 9.994399999440002e-06, + "loss": 0.4652, + "step": 5603 + }, + { + "epoch": 2.605299860529986, + "grad_norm": 4.948305130004883, + "learning_rate": 9.9943989994399e-06, + "loss": 0.2429, + "step": 5604 + }, + { + "epoch": 2.605764760576476, + "grad_norm": 4.670929908752441, + "learning_rate": 9.994397999439801e-06, + "loss": 0.4691, + "step": 5605 + }, + { + "epoch": 2.606229660622966, + "grad_norm": 5.0846662521362305, + "learning_rate": 9.9943969994397e-06, + "loss": 0.4122, + "step": 5606 + }, + { + "epoch": 2.606694560669456, + "grad_norm": 6.067317485809326, + "learning_rate": 9.9943959994396e-06, + "loss": 0.4758, + "step": 5607 + }, + { + "epoch": 2.607159460715946, + "grad_norm": 6.165361404418945, + "learning_rate": 9.994394999439501e-06, + "loss": 0.4983, + "step": 5608 + }, + { + "epoch": 2.6076243607624363, + "grad_norm": 6.810774326324463, + "learning_rate": 9.994393999439402e-06, + "loss": 0.6, + "step": 5609 + }, + { + "epoch": 2.608089260808926, + "grad_norm": 6.732766151428223, + "learning_rate": 9.994392999439301e-06, + "loss": 0.6733, + "step": 5610 + }, + { + "epoch": 2.608554160855416, + "grad_norm": 6.0650177001953125, + "learning_rate": 9.9943919994392e-06, + "loss": 0.5597, + "step": 5611 + }, + { + "epoch": 2.609019060901906, + "grad_norm": 4.959795951843262, + "learning_rate": 9.9943909994391e-06, + "loss": 0.3932, + "step": 5612 + }, + { + "epoch": 2.609483960948396, + "grad_norm": 4.917588233947754, + "learning_rate": 9.994389999439e-06, + "loss": 0.3316, + "step": 5613 + }, + { + "epoch": 2.609948860994886, + "grad_norm": 7.583770751953125, + "learning_rate": 9.9943889994389e-06, + "loss": 0.6542, + "step": 5614 + }, + { + "epoch": 2.610413761041376, + "grad_norm": 4.779882907867432, + "learning_rate": 9.994387999438801e-06, + "loss": 0.3489, + "step": 5615 + }, + { + "epoch": 2.6108786610878663, + "grad_norm": 5.156742572784424, + "learning_rate": 9.994386999438702e-06, + "loss": 0.5752, + "step": 5616 + }, + { + "epoch": 2.611343561134356, + "grad_norm": 6.167352199554443, + "learning_rate": 9.9943859994386e-06, + "loss": 0.5857, + "step": 5617 + }, + { + "epoch": 2.6118084611808463, + "grad_norm": 4.7601094245910645, + "learning_rate": 9.994384999438501e-06, + "loss": 0.2842, + "step": 5618 + }, + { + "epoch": 2.612273361227336, + "grad_norm": 5.377358436584473, + "learning_rate": 9.9943839994384e-06, + "loss": 0.2951, + "step": 5619 + }, + { + "epoch": 2.612738261273826, + "grad_norm": 3.771411895751953, + "learning_rate": 9.9943829994383e-06, + "loss": 0.2339, + "step": 5620 + }, + { + "epoch": 2.613203161320316, + "grad_norm": 10.928704261779785, + "learning_rate": 9.994381999438201e-06, + "loss": 0.5233, + "step": 5621 + }, + { + "epoch": 2.6136680613668064, + "grad_norm": 6.302888870239258, + "learning_rate": 9.9943809994381e-06, + "loss": 0.5815, + "step": 5622 + }, + { + "epoch": 2.614132961413296, + "grad_norm": 4.430446147918701, + "learning_rate": 9.994379999438001e-06, + "loss": 0.4529, + "step": 5623 + }, + { + "epoch": 2.614597861459786, + "grad_norm": 4.851295471191406, + "learning_rate": 9.9943789994379e-06, + "loss": 0.4528, + "step": 5624 + }, + { + "epoch": 2.6150627615062763, + "grad_norm": 5.747514724731445, + "learning_rate": 9.9943779994378e-06, + "loss": 0.5108, + "step": 5625 + }, + { + "epoch": 2.615527661552766, + "grad_norm": 4.789963722229004, + "learning_rate": 9.9943769994377e-06, + "loss": 0.3936, + "step": 5626 + }, + { + "epoch": 2.6159925615992563, + "grad_norm": 6.8414130210876465, + "learning_rate": 9.9943759994376e-06, + "loss": 0.447, + "step": 5627 + }, + { + "epoch": 2.616457461645746, + "grad_norm": 5.365880489349365, + "learning_rate": 9.994374999437501e-06, + "loss": 0.5056, + "step": 5628 + }, + { + "epoch": 2.616922361692236, + "grad_norm": 3.6509487628936768, + "learning_rate": 9.994373999437402e-06, + "loss": 0.214, + "step": 5629 + }, + { + "epoch": 2.617387261738726, + "grad_norm": 5.527547359466553, + "learning_rate": 9.9943729994373e-06, + "loss": 0.365, + "step": 5630 + }, + { + "epoch": 2.6178521617852164, + "grad_norm": 8.687482833862305, + "learning_rate": 9.994371999437201e-06, + "loss": 0.4817, + "step": 5631 + }, + { + "epoch": 2.618317061831706, + "grad_norm": 3.7697184085845947, + "learning_rate": 9.9943709994371e-06, + "loss": 0.2347, + "step": 5632 + }, + { + "epoch": 2.618781961878196, + "grad_norm": 6.327661991119385, + "learning_rate": 9.994369999437001e-06, + "loss": 0.6022, + "step": 5633 + }, + { + "epoch": 2.6192468619246863, + "grad_norm": 4.535174369812012, + "learning_rate": 9.994368999436902e-06, + "loss": 0.3558, + "step": 5634 + }, + { + "epoch": 2.619711761971176, + "grad_norm": 6.455977916717529, + "learning_rate": 9.9943679994368e-06, + "loss": 0.473, + "step": 5635 + }, + { + "epoch": 2.6201766620176663, + "grad_norm": 4.57008171081543, + "learning_rate": 9.994366999436701e-06, + "loss": 0.4083, + "step": 5636 + }, + { + "epoch": 2.620641562064156, + "grad_norm": 5.439036846160889, + "learning_rate": 9.9943659994366e-06, + "loss": 0.5095, + "step": 5637 + }, + { + "epoch": 2.6211064621106464, + "grad_norm": 3.9095680713653564, + "learning_rate": 9.9943649994365e-06, + "loss": 0.4529, + "step": 5638 + }, + { + "epoch": 2.621571362157136, + "grad_norm": 4.089998722076416, + "learning_rate": 9.9943639994364e-06, + "loss": 0.2115, + "step": 5639 + }, + { + "epoch": 2.6220362622036264, + "grad_norm": 10.51236343383789, + "learning_rate": 9.9943629994363e-06, + "loss": 0.7429, + "step": 5640 + }, + { + "epoch": 2.622501162250116, + "grad_norm": 7.856497287750244, + "learning_rate": 9.994361999436201e-06, + "loss": 0.5071, + "step": 5641 + }, + { + "epoch": 2.622966062296606, + "grad_norm": 8.387615203857422, + "learning_rate": 9.994360999436102e-06, + "loss": 0.5048, + "step": 5642 + }, + { + "epoch": 2.6234309623430963, + "grad_norm": 5.629804611206055, + "learning_rate": 9.994359999436e-06, + "loss": 0.4219, + "step": 5643 + }, + { + "epoch": 2.6238958623895865, + "grad_norm": 5.1036200523376465, + "learning_rate": 9.9943589994359e-06, + "loss": 0.3376, + "step": 5644 + }, + { + "epoch": 2.6243607624360763, + "grad_norm": 8.567352294921875, + "learning_rate": 9.9943579994358e-06, + "loss": 0.7775, + "step": 5645 + }, + { + "epoch": 2.624825662482566, + "grad_norm": 5.709506034851074, + "learning_rate": 9.994356999435701e-06, + "loss": 0.3895, + "step": 5646 + }, + { + "epoch": 2.6252905625290563, + "grad_norm": 7.047823429107666, + "learning_rate": 9.994355999435602e-06, + "loss": 0.6896, + "step": 5647 + }, + { + "epoch": 2.625755462575546, + "grad_norm": 5.275310516357422, + "learning_rate": 9.9943549994355e-06, + "loss": 0.6063, + "step": 5648 + }, + { + "epoch": 2.6262203626220364, + "grad_norm": 6.627861022949219, + "learning_rate": 9.994353999435401e-06, + "loss": 0.6583, + "step": 5649 + }, + { + "epoch": 2.626685262668526, + "grad_norm": 9.021591186523438, + "learning_rate": 9.9943529994353e-06, + "loss": 0.4022, + "step": 5650 + }, + { + "epoch": 2.627150162715016, + "grad_norm": 5.632351398468018, + "learning_rate": 9.994351999435201e-06, + "loss": 0.3025, + "step": 5651 + }, + { + "epoch": 2.6276150627615062, + "grad_norm": 5.084775447845459, + "learning_rate": 9.9943509994351e-06, + "loss": 0.3362, + "step": 5652 + }, + { + "epoch": 2.6280799628079965, + "grad_norm": 6.537253379821777, + "learning_rate": 9.994349999435e-06, + "loss": 0.411, + "step": 5653 + }, + { + "epoch": 2.6285448628544863, + "grad_norm": 5.803483963012695, + "learning_rate": 9.994348999434901e-06, + "loss": 0.4343, + "step": 5654 + }, + { + "epoch": 2.629009762900976, + "grad_norm": 6.317578315734863, + "learning_rate": 9.9943479994348e-06, + "loss": 0.5931, + "step": 5655 + }, + { + "epoch": 2.6294746629474663, + "grad_norm": 6.188268184661865, + "learning_rate": 9.9943469994347e-06, + "loss": 0.3819, + "step": 5656 + }, + { + "epoch": 2.629939562993956, + "grad_norm": 6.081116199493408, + "learning_rate": 9.9943459994346e-06, + "loss": 0.6077, + "step": 5657 + }, + { + "epoch": 2.6304044630404464, + "grad_norm": 5.749893665313721, + "learning_rate": 9.9943449994345e-06, + "loss": 0.5594, + "step": 5658 + }, + { + "epoch": 2.630869363086936, + "grad_norm": 6.281847953796387, + "learning_rate": 9.994343999434401e-06, + "loss": 0.3265, + "step": 5659 + }, + { + "epoch": 2.6313342631334264, + "grad_norm": 5.51943826675415, + "learning_rate": 9.994342999434302e-06, + "loss": 0.504, + "step": 5660 + }, + { + "epoch": 2.6317991631799162, + "grad_norm": 6.732715606689453, + "learning_rate": 9.9943419994342e-06, + "loss": 0.3207, + "step": 5661 + }, + { + "epoch": 2.6322640632264065, + "grad_norm": 10.069877624511719, + "learning_rate": 9.994340999434101e-06, + "loss": 0.6348, + "step": 5662 + }, + { + "epoch": 2.6327289632728963, + "grad_norm": 5.385249614715576, + "learning_rate": 9.994339999434e-06, + "loss": 0.2668, + "step": 5663 + }, + { + "epoch": 2.633193863319386, + "grad_norm": 5.021171569824219, + "learning_rate": 9.994338999433901e-06, + "loss": 0.5691, + "step": 5664 + }, + { + "epoch": 2.6336587633658763, + "grad_norm": 5.7259955406188965, + "learning_rate": 9.994337999433802e-06, + "loss": 0.4584, + "step": 5665 + }, + { + "epoch": 2.6341236634123666, + "grad_norm": 5.180727005004883, + "learning_rate": 9.9943369994337e-06, + "loss": 0.3688, + "step": 5666 + }, + { + "epoch": 2.6345885634588564, + "grad_norm": 8.542142868041992, + "learning_rate": 9.994335999433601e-06, + "loss": 0.4326, + "step": 5667 + }, + { + "epoch": 2.635053463505346, + "grad_norm": 4.879644393920898, + "learning_rate": 9.9943349994335e-06, + "loss": 0.4729, + "step": 5668 + }, + { + "epoch": 2.6355183635518364, + "grad_norm": 5.514164924621582, + "learning_rate": 9.994333999433401e-06, + "loss": 0.4491, + "step": 5669 + }, + { + "epoch": 2.6359832635983262, + "grad_norm": 7.922525882720947, + "learning_rate": 9.9943329994333e-06, + "loss": 0.7837, + "step": 5670 + }, + { + "epoch": 2.6364481636448165, + "grad_norm": 6.465142726898193, + "learning_rate": 9.9943319994332e-06, + "loss": 0.5769, + "step": 5671 + }, + { + "epoch": 2.6369130636913063, + "grad_norm": 11.010184288024902, + "learning_rate": 9.994330999433101e-06, + "loss": 0.7303, + "step": 5672 + }, + { + "epoch": 2.6373779637377965, + "grad_norm": 8.073168754577637, + "learning_rate": 9.994329999433002e-06, + "loss": 0.8005, + "step": 5673 + }, + { + "epoch": 2.6378428637842863, + "grad_norm": 5.13613748550415, + "learning_rate": 9.9943289994329e-06, + "loss": 0.3781, + "step": 5674 + }, + { + "epoch": 2.6383077638307766, + "grad_norm": 7.721363544464111, + "learning_rate": 9.994327999432801e-06, + "loss": 0.5034, + "step": 5675 + }, + { + "epoch": 2.6387726638772664, + "grad_norm": 6.818492889404297, + "learning_rate": 9.9943269994327e-06, + "loss": 0.5367, + "step": 5676 + }, + { + "epoch": 2.639237563923756, + "grad_norm": 3.577542304992676, + "learning_rate": 9.994325999432601e-06, + "loss": 0.3031, + "step": 5677 + }, + { + "epoch": 2.6397024639702464, + "grad_norm": 5.464631080627441, + "learning_rate": 9.994324999432502e-06, + "loss": 0.542, + "step": 5678 + }, + { + "epoch": 2.6401673640167362, + "grad_norm": 8.593255043029785, + "learning_rate": 9.9943239994324e-06, + "loss": 0.5773, + "step": 5679 + }, + { + "epoch": 2.6406322640632265, + "grad_norm": 5.77306604385376, + "learning_rate": 9.994322999432301e-06, + "loss": 0.4741, + "step": 5680 + }, + { + "epoch": 2.6410971641097163, + "grad_norm": 5.94674015045166, + "learning_rate": 9.9943219994322e-06, + "loss": 0.5027, + "step": 5681 + }, + { + "epoch": 2.6415620641562065, + "grad_norm": 4.371725082397461, + "learning_rate": 9.994320999432101e-06, + "loss": 0.2098, + "step": 5682 + }, + { + "epoch": 2.6420269642026963, + "grad_norm": 6.8127899169921875, + "learning_rate": 9.994319999432e-06, + "loss": 0.6518, + "step": 5683 + }, + { + "epoch": 2.6424918642491866, + "grad_norm": 6.3450727462768555, + "learning_rate": 9.9943189994319e-06, + "loss": 0.5506, + "step": 5684 + }, + { + "epoch": 2.6429567642956764, + "grad_norm": 4.218777179718018, + "learning_rate": 9.994317999431801e-06, + "loss": 0.267, + "step": 5685 + }, + { + "epoch": 2.643421664342166, + "grad_norm": 5.896463394165039, + "learning_rate": 9.994316999431702e-06, + "loss": 0.434, + "step": 5686 + }, + { + "epoch": 2.6438865643886564, + "grad_norm": 4.787370204925537, + "learning_rate": 9.994315999431601e-06, + "loss": 0.3178, + "step": 5687 + }, + { + "epoch": 2.6443514644351467, + "grad_norm": 5.396151542663574, + "learning_rate": 9.9943149994315e-06, + "loss": 0.2716, + "step": 5688 + }, + { + "epoch": 2.6448163644816365, + "grad_norm": 7.7830400466918945, + "learning_rate": 9.9943139994314e-06, + "loss": 0.6793, + "step": 5689 + }, + { + "epoch": 2.6452812645281263, + "grad_norm": 3.5873541831970215, + "learning_rate": 9.994312999431301e-06, + "loss": 0.2623, + "step": 5690 + }, + { + "epoch": 2.6457461645746165, + "grad_norm": 7.533114433288574, + "learning_rate": 9.994311999431202e-06, + "loss": 0.6806, + "step": 5691 + }, + { + "epoch": 2.6462110646211063, + "grad_norm": 5.688507556915283, + "learning_rate": 9.9943109994311e-06, + "loss": 0.3471, + "step": 5692 + }, + { + "epoch": 2.6466759646675966, + "grad_norm": 4.226088047027588, + "learning_rate": 9.994309999431001e-06, + "loss": 0.3348, + "step": 5693 + }, + { + "epoch": 2.6471408647140864, + "grad_norm": 11.327149391174316, + "learning_rate": 9.9943089994309e-06, + "loss": 0.6843, + "step": 5694 + }, + { + "epoch": 2.6476057647605766, + "grad_norm": 3.8060383796691895, + "learning_rate": 9.994307999430801e-06, + "loss": 0.386, + "step": 5695 + }, + { + "epoch": 2.6480706648070664, + "grad_norm": 3.540910243988037, + "learning_rate": 9.9943069994307e-06, + "loss": 0.2301, + "step": 5696 + }, + { + "epoch": 2.6485355648535567, + "grad_norm": 8.207377433776855, + "learning_rate": 9.9943059994306e-06, + "loss": 0.5635, + "step": 5697 + }, + { + "epoch": 2.6490004649000465, + "grad_norm": 6.0110907554626465, + "learning_rate": 9.994304999430501e-06, + "loss": 0.7789, + "step": 5698 + }, + { + "epoch": 2.6494653649465363, + "grad_norm": 4.279351234436035, + "learning_rate": 9.9943039994304e-06, + "loss": 0.2616, + "step": 5699 + }, + { + "epoch": 2.6499302649930265, + "grad_norm": 5.2989654541015625, + "learning_rate": 9.994302999430301e-06, + "loss": 0.5128, + "step": 5700 + }, + { + "epoch": 2.6503951650395168, + "grad_norm": 2.9206252098083496, + "learning_rate": 9.9943019994302e-06, + "loss": 0.1581, + "step": 5701 + }, + { + "epoch": 2.6508600650860066, + "grad_norm": 14.885785102844238, + "learning_rate": 9.9943009994301e-06, + "loss": 0.8592, + "step": 5702 + }, + { + "epoch": 2.6513249651324964, + "grad_norm": 8.271832466125488, + "learning_rate": 9.994299999430001e-06, + "loss": 0.5134, + "step": 5703 + }, + { + "epoch": 2.6517898651789866, + "grad_norm": 4.3880462646484375, + "learning_rate": 9.994298999429902e-06, + "loss": 0.4015, + "step": 5704 + }, + { + "epoch": 2.6522547652254764, + "grad_norm": 6.524986267089844, + "learning_rate": 9.9942979994298e-06, + "loss": 0.7913, + "step": 5705 + }, + { + "epoch": 2.6527196652719667, + "grad_norm": 6.228729248046875, + "learning_rate": 9.994296999429701e-06, + "loss": 0.4135, + "step": 5706 + }, + { + "epoch": 2.6531845653184565, + "grad_norm": 7.310221195220947, + "learning_rate": 9.9942959994296e-06, + "loss": 0.5829, + "step": 5707 + }, + { + "epoch": 2.6536494653649463, + "grad_norm": 5.204617500305176, + "learning_rate": 9.9942949994295e-06, + "loss": 0.3891, + "step": 5708 + }, + { + "epoch": 2.6541143654114365, + "grad_norm": 9.1510009765625, + "learning_rate": 9.9942939994294e-06, + "loss": 0.8482, + "step": 5709 + }, + { + "epoch": 2.6545792654579268, + "grad_norm": 8.673355102539062, + "learning_rate": 9.9942929994293e-06, + "loss": 0.8119, + "step": 5710 + }, + { + "epoch": 2.6550441655044166, + "grad_norm": 7.575273036956787, + "learning_rate": 9.994291999429201e-06, + "loss": 0.5384, + "step": 5711 + }, + { + "epoch": 2.6555090655509064, + "grad_norm": 3.9617347717285156, + "learning_rate": 9.9942909994291e-06, + "loss": 0.3794, + "step": 5712 + }, + { + "epoch": 2.6559739655973966, + "grad_norm": 6.387683391571045, + "learning_rate": 9.994289999429001e-06, + "loss": 0.9923, + "step": 5713 + }, + { + "epoch": 2.6564388656438864, + "grad_norm": 6.766337871551514, + "learning_rate": 9.9942889994289e-06, + "loss": 0.6489, + "step": 5714 + }, + { + "epoch": 2.6569037656903767, + "grad_norm": 6.589404582977295, + "learning_rate": 9.9942879994288e-06, + "loss": 0.5001, + "step": 5715 + }, + { + "epoch": 2.6573686657368665, + "grad_norm": 5.980020523071289, + "learning_rate": 9.994286999428701e-06, + "loss": 0.6572, + "step": 5716 + }, + { + "epoch": 2.6578335657833567, + "grad_norm": 5.218296051025391, + "learning_rate": 9.994285999428602e-06, + "loss": 0.4989, + "step": 5717 + }, + { + "epoch": 2.6582984658298465, + "grad_norm": 4.929196357727051, + "learning_rate": 9.994284999428501e-06, + "loss": 0.2809, + "step": 5718 + }, + { + "epoch": 2.6587633658763368, + "grad_norm": 4.856922149658203, + "learning_rate": 9.9942839994284e-06, + "loss": 0.4524, + "step": 5719 + }, + { + "epoch": 2.6592282659228266, + "grad_norm": 5.055176258087158, + "learning_rate": 9.9942829994283e-06, + "loss": 0.5326, + "step": 5720 + }, + { + "epoch": 2.6596931659693164, + "grad_norm": 4.733518123626709, + "learning_rate": 9.9942819994282e-06, + "loss": 0.2257, + "step": 5721 + }, + { + "epoch": 2.6601580660158066, + "grad_norm": 5.457587242126465, + "learning_rate": 9.9942809994281e-06, + "loss": 0.3592, + "step": 5722 + }, + { + "epoch": 2.660622966062297, + "grad_norm": 4.771544456481934, + "learning_rate": 9.994279999428e-06, + "loss": 0.2061, + "step": 5723 + }, + { + "epoch": 2.6610878661087867, + "grad_norm": 5.440240859985352, + "learning_rate": 9.994278999427901e-06, + "loss": 0.3883, + "step": 5724 + }, + { + "epoch": 2.6615527661552765, + "grad_norm": 4.109737396240234, + "learning_rate": 9.9942779994278e-06, + "loss": 0.2564, + "step": 5725 + }, + { + "epoch": 2.6620176662017667, + "grad_norm": 5.708200931549072, + "learning_rate": 9.994276999427701e-06, + "loss": 0.3492, + "step": 5726 + }, + { + "epoch": 2.6624825662482565, + "grad_norm": 4.036259174346924, + "learning_rate": 9.9942759994276e-06, + "loss": 0.2746, + "step": 5727 + }, + { + "epoch": 2.6629474662947468, + "grad_norm": 11.70553970336914, + "learning_rate": 9.9942749994275e-06, + "loss": 0.4186, + "step": 5728 + }, + { + "epoch": 2.6634123663412366, + "grad_norm": 6.771122455596924, + "learning_rate": 9.994273999427401e-06, + "loss": 0.6065, + "step": 5729 + }, + { + "epoch": 2.6638772663877264, + "grad_norm": 7.032674789428711, + "learning_rate": 9.9942729994273e-06, + "loss": 0.4219, + "step": 5730 + }, + { + "epoch": 2.6643421664342166, + "grad_norm": 5.920657157897949, + "learning_rate": 9.994271999427201e-06, + "loss": 0.4558, + "step": 5731 + }, + { + "epoch": 2.664807066480707, + "grad_norm": 7.498859405517578, + "learning_rate": 9.9942709994271e-06, + "loss": 0.39, + "step": 5732 + }, + { + "epoch": 2.6652719665271967, + "grad_norm": 4.771852970123291, + "learning_rate": 9.994269999427e-06, + "loss": 0.3539, + "step": 5733 + }, + { + "epoch": 2.6657368665736865, + "grad_norm": 5.563656806945801, + "learning_rate": 9.994268999426901e-06, + "loss": 0.5455, + "step": 5734 + }, + { + "epoch": 2.6662017666201767, + "grad_norm": 4.7830810546875, + "learning_rate": 9.9942679994268e-06, + "loss": 0.3785, + "step": 5735 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 8.470800399780273, + "learning_rate": 9.994266999426701e-06, + "loss": 0.9375, + "step": 5736 + }, + { + "epoch": 2.6671315667131568, + "grad_norm": 6.795756816864014, + "learning_rate": 9.994265999426602e-06, + "loss": 0.6788, + "step": 5737 + }, + { + "epoch": 2.6675964667596466, + "grad_norm": 6.473123073577881, + "learning_rate": 9.9942649994265e-06, + "loss": 0.3856, + "step": 5738 + }, + { + "epoch": 2.668061366806137, + "grad_norm": 7.0703911781311035, + "learning_rate": 9.994263999426401e-06, + "loss": 0.5837, + "step": 5739 + }, + { + "epoch": 2.6685262668526266, + "grad_norm": 6.011892795562744, + "learning_rate": 9.9942629994263e-06, + "loss": 0.7618, + "step": 5740 + }, + { + "epoch": 2.668991166899117, + "grad_norm": 6.047227382659912, + "learning_rate": 9.9942619994262e-06, + "loss": 0.3432, + "step": 5741 + }, + { + "epoch": 2.6694560669456067, + "grad_norm": 5.983806610107422, + "learning_rate": 9.994260999426101e-06, + "loss": 0.3739, + "step": 5742 + }, + { + "epoch": 2.6699209669920965, + "grad_norm": 6.489043235778809, + "learning_rate": 9.994259999426e-06, + "loss": 0.593, + "step": 5743 + }, + { + "epoch": 2.6703858670385867, + "grad_norm": 8.482552528381348, + "learning_rate": 9.994258999425901e-06, + "loss": 0.5311, + "step": 5744 + }, + { + "epoch": 2.670850767085077, + "grad_norm": 3.690697431564331, + "learning_rate": 9.9942579994258e-06, + "loss": 0.3602, + "step": 5745 + }, + { + "epoch": 2.6713156671315668, + "grad_norm": 6.961888790130615, + "learning_rate": 9.9942569994257e-06, + "loss": 0.6177, + "step": 5746 + }, + { + "epoch": 2.6717805671780566, + "grad_norm": 5.778287410736084, + "learning_rate": 9.994255999425601e-06, + "loss": 0.4286, + "step": 5747 + }, + { + "epoch": 2.672245467224547, + "grad_norm": 6.748373508453369, + "learning_rate": 9.9942549994255e-06, + "loss": 0.4721, + "step": 5748 + }, + { + "epoch": 2.6727103672710366, + "grad_norm": 4.404869556427002, + "learning_rate": 9.994253999425401e-06, + "loss": 0.3537, + "step": 5749 + }, + { + "epoch": 2.673175267317527, + "grad_norm": 4.332780361175537, + "learning_rate": 9.994252999425302e-06, + "loss": 0.2698, + "step": 5750 + }, + { + "epoch": 2.6736401673640167, + "grad_norm": 11.584376335144043, + "learning_rate": 9.9942519994252e-06, + "loss": 0.6931, + "step": 5751 + }, + { + "epoch": 2.6741050674105065, + "grad_norm": 5.5448174476623535, + "learning_rate": 9.9942509994251e-06, + "loss": 0.2739, + "step": 5752 + }, + { + "epoch": 2.6745699674569967, + "grad_norm": 6.473660469055176, + "learning_rate": 9.994249999425e-06, + "loss": 0.4101, + "step": 5753 + }, + { + "epoch": 2.675034867503487, + "grad_norm": 5.121837139129639, + "learning_rate": 9.994248999424901e-06, + "loss": 0.4758, + "step": 5754 + }, + { + "epoch": 2.6754997675499768, + "grad_norm": 3.2220380306243896, + "learning_rate": 9.994247999424802e-06, + "loss": 0.2042, + "step": 5755 + }, + { + "epoch": 2.6759646675964666, + "grad_norm": 12.167414665222168, + "learning_rate": 9.9942469994247e-06, + "loss": 0.8774, + "step": 5756 + }, + { + "epoch": 2.676429567642957, + "grad_norm": 6.20670223236084, + "learning_rate": 9.994245999424601e-06, + "loss": 0.4294, + "step": 5757 + }, + { + "epoch": 2.6768944676894466, + "grad_norm": 6.32527494430542, + "learning_rate": 9.9942449994245e-06, + "loss": 0.4171, + "step": 5758 + }, + { + "epoch": 2.677359367735937, + "grad_norm": 3.4173574447631836, + "learning_rate": 9.9942439994244e-06, + "loss": 0.1971, + "step": 5759 + }, + { + "epoch": 2.6778242677824267, + "grad_norm": 13.362882614135742, + "learning_rate": 9.994242999424301e-06, + "loss": 0.7195, + "step": 5760 + }, + { + "epoch": 2.678289167828917, + "grad_norm": 5.80898904800415, + "learning_rate": 9.994241999424202e-06, + "loss": 0.5116, + "step": 5761 + }, + { + "epoch": 2.6787540678754067, + "grad_norm": 6.528703689575195, + "learning_rate": 9.994240999424101e-06, + "loss": 0.7088, + "step": 5762 + }, + { + "epoch": 2.679218967921897, + "grad_norm": 8.999499320983887, + "learning_rate": 9.994239999424e-06, + "loss": 0.7204, + "step": 5763 + }, + { + "epoch": 2.6796838679683868, + "grad_norm": 4.258975505828857, + "learning_rate": 9.9942389994239e-06, + "loss": 0.497, + "step": 5764 + }, + { + "epoch": 2.6801487680148766, + "grad_norm": 18.30262565612793, + "learning_rate": 9.9942379994238e-06, + "loss": 0.7295, + "step": 5765 + }, + { + "epoch": 2.680613668061367, + "grad_norm": 3.538262128829956, + "learning_rate": 9.9942369994237e-06, + "loss": 0.2789, + "step": 5766 + }, + { + "epoch": 2.681078568107857, + "grad_norm": 6.802004814147949, + "learning_rate": 9.994235999423601e-06, + "loss": 0.3961, + "step": 5767 + }, + { + "epoch": 2.681543468154347, + "grad_norm": 6.78092622756958, + "learning_rate": 9.994234999423502e-06, + "loss": 0.5967, + "step": 5768 + }, + { + "epoch": 2.6820083682008367, + "grad_norm": 4.520244598388672, + "learning_rate": 9.9942339994234e-06, + "loss": 0.349, + "step": 5769 + }, + { + "epoch": 2.682473268247327, + "grad_norm": 7.974789142608643, + "learning_rate": 9.994232999423301e-06, + "loss": 0.934, + "step": 5770 + }, + { + "epoch": 2.6829381682938167, + "grad_norm": 6.283482551574707, + "learning_rate": 9.9942319994232e-06, + "loss": 0.329, + "step": 5771 + }, + { + "epoch": 2.683403068340307, + "grad_norm": 7.290154933929443, + "learning_rate": 9.994230999423101e-06, + "loss": 0.6764, + "step": 5772 + }, + { + "epoch": 2.6838679683867968, + "grad_norm": 5.159022331237793, + "learning_rate": 9.994229999423002e-06, + "loss": 0.3447, + "step": 5773 + }, + { + "epoch": 2.684332868433287, + "grad_norm": 7.347305774688721, + "learning_rate": 9.9942289994229e-06, + "loss": 0.3759, + "step": 5774 + }, + { + "epoch": 2.684797768479777, + "grad_norm": 6.251156330108643, + "learning_rate": 9.994227999422801e-06, + "loss": 0.6645, + "step": 5775 + }, + { + "epoch": 2.685262668526267, + "grad_norm": 7.0743584632873535, + "learning_rate": 9.9942269994227e-06, + "loss": 0.6758, + "step": 5776 + }, + { + "epoch": 2.685727568572757, + "grad_norm": 7.742031097412109, + "learning_rate": 9.9942259994226e-06, + "loss": 0.4764, + "step": 5777 + }, + { + "epoch": 2.6861924686192467, + "grad_norm": 5.000149726867676, + "learning_rate": 9.9942249994225e-06, + "loss": 0.6584, + "step": 5778 + }, + { + "epoch": 2.686657368665737, + "grad_norm": 4.981840133666992, + "learning_rate": 9.9942239994224e-06, + "loss": 0.4838, + "step": 5779 + }, + { + "epoch": 2.687122268712227, + "grad_norm": 5.31096887588501, + "learning_rate": 9.994222999422301e-06, + "loss": 0.4622, + "step": 5780 + }, + { + "epoch": 2.687587168758717, + "grad_norm": 7.974549770355225, + "learning_rate": 9.994221999422202e-06, + "loss": 0.8277, + "step": 5781 + }, + { + "epoch": 2.6880520688052068, + "grad_norm": 5.328675746917725, + "learning_rate": 9.9942209994221e-06, + "loss": 0.5603, + "step": 5782 + }, + { + "epoch": 2.688516968851697, + "grad_norm": 5.265301704406738, + "learning_rate": 9.994219999422001e-06, + "loss": 0.3372, + "step": 5783 + }, + { + "epoch": 2.688981868898187, + "grad_norm": 6.1014556884765625, + "learning_rate": 9.9942189994219e-06, + "loss": 0.7092, + "step": 5784 + }, + { + "epoch": 2.689446768944677, + "grad_norm": 7.80340051651001, + "learning_rate": 9.994217999421801e-06, + "loss": 0.549, + "step": 5785 + }, + { + "epoch": 2.689911668991167, + "grad_norm": 5.543661594390869, + "learning_rate": 9.994216999421702e-06, + "loss": 0.4124, + "step": 5786 + }, + { + "epoch": 2.6903765690376567, + "grad_norm": 6.1670942306518555, + "learning_rate": 9.9942159994216e-06, + "loss": 0.6052, + "step": 5787 + }, + { + "epoch": 2.690841469084147, + "grad_norm": 5.1447296142578125, + "learning_rate": 9.994214999421501e-06, + "loss": 0.37, + "step": 5788 + }, + { + "epoch": 2.691306369130637, + "grad_norm": 3.5740833282470703, + "learning_rate": 9.9942139994214e-06, + "loss": 0.3522, + "step": 5789 + }, + { + "epoch": 2.691771269177127, + "grad_norm": 5.374849319458008, + "learning_rate": 9.9942129994213e-06, + "loss": 0.447, + "step": 5790 + }, + { + "epoch": 2.6922361692236167, + "grad_norm": 6.329187393188477, + "learning_rate": 9.9942119994212e-06, + "loss": 0.2847, + "step": 5791 + }, + { + "epoch": 2.692701069270107, + "grad_norm": 5.677894592285156, + "learning_rate": 9.9942109994211e-06, + "loss": 0.4078, + "step": 5792 + }, + { + "epoch": 2.693165969316597, + "grad_norm": 4.476107597351074, + "learning_rate": 9.994209999421001e-06, + "loss": 0.2344, + "step": 5793 + }, + { + "epoch": 2.693630869363087, + "grad_norm": 8.900744438171387, + "learning_rate": 9.994208999420902e-06, + "loss": 0.5738, + "step": 5794 + }, + { + "epoch": 2.694095769409577, + "grad_norm": 5.506163597106934, + "learning_rate": 9.9942079994208e-06, + "loss": 0.424, + "step": 5795 + }, + { + "epoch": 2.694560669456067, + "grad_norm": 3.472215175628662, + "learning_rate": 9.9942069994207e-06, + "loss": 0.2069, + "step": 5796 + }, + { + "epoch": 2.695025569502557, + "grad_norm": 9.760252952575684, + "learning_rate": 9.9942059994206e-06, + "loss": 0.7898, + "step": 5797 + }, + { + "epoch": 2.695490469549047, + "grad_norm": 5.351093292236328, + "learning_rate": 9.994204999420501e-06, + "loss": 0.6306, + "step": 5798 + }, + { + "epoch": 2.695955369595537, + "grad_norm": 3.594207763671875, + "learning_rate": 9.994203999420402e-06, + "loss": 0.2864, + "step": 5799 + }, + { + "epoch": 2.6964202696420267, + "grad_norm": 3.650742769241333, + "learning_rate": 9.9942029994203e-06, + "loss": 0.1705, + "step": 5800 + }, + { + "epoch": 2.696885169688517, + "grad_norm": 4.416585445404053, + "learning_rate": 9.994201999420201e-06, + "loss": 0.4214, + "step": 5801 + }, + { + "epoch": 2.6973500697350072, + "grad_norm": 4.883419036865234, + "learning_rate": 9.9942009994201e-06, + "loss": 0.2853, + "step": 5802 + }, + { + "epoch": 2.697814969781497, + "grad_norm": 6.542292594909668, + "learning_rate": 9.994199999420001e-06, + "loss": 0.4573, + "step": 5803 + }, + { + "epoch": 2.698279869827987, + "grad_norm": 4.2676777839660645, + "learning_rate": 9.9941989994199e-06, + "loss": 0.3898, + "step": 5804 + }, + { + "epoch": 2.698744769874477, + "grad_norm": 5.511103630065918, + "learning_rate": 9.9941979994198e-06, + "loss": 0.5223, + "step": 5805 + }, + { + "epoch": 2.699209669920967, + "grad_norm": 4.291223049163818, + "learning_rate": 9.994196999419701e-06, + "loss": 0.3034, + "step": 5806 + }, + { + "epoch": 2.699674569967457, + "grad_norm": 4.292786121368408, + "learning_rate": 9.9941959994196e-06, + "loss": 0.2118, + "step": 5807 + }, + { + "epoch": 2.700139470013947, + "grad_norm": 4.829630374908447, + "learning_rate": 9.9941949994195e-06, + "loss": 0.426, + "step": 5808 + }, + { + "epoch": 2.7006043700604367, + "grad_norm": 3.822420120239258, + "learning_rate": 9.9941939994194e-06, + "loss": 0.252, + "step": 5809 + }, + { + "epoch": 2.701069270106927, + "grad_norm": 3.7516775131225586, + "learning_rate": 9.9941929994193e-06, + "loss": 0.2996, + "step": 5810 + }, + { + "epoch": 2.7015341701534172, + "grad_norm": 3.6433749198913574, + "learning_rate": 9.994191999419201e-06, + "loss": 0.3192, + "step": 5811 + }, + { + "epoch": 2.701999070199907, + "grad_norm": 5.008735179901123, + "learning_rate": 9.994190999419102e-06, + "loss": 0.3898, + "step": 5812 + }, + { + "epoch": 2.702463970246397, + "grad_norm": 6.784287929534912, + "learning_rate": 9.994189999419e-06, + "loss": 0.3548, + "step": 5813 + }, + { + "epoch": 2.702928870292887, + "grad_norm": 6.534629821777344, + "learning_rate": 9.994188999418901e-06, + "loss": 0.5956, + "step": 5814 + }, + { + "epoch": 2.703393770339377, + "grad_norm": 8.87839126586914, + "learning_rate": 9.9941879994188e-06, + "loss": 0.7201, + "step": 5815 + }, + { + "epoch": 2.703858670385867, + "grad_norm": 6.6680145263671875, + "learning_rate": 9.994186999418701e-06, + "loss": 0.2502, + "step": 5816 + }, + { + "epoch": 2.704323570432357, + "grad_norm": 4.385280132293701, + "learning_rate": 9.9941859994186e-06, + "loss": 0.2268, + "step": 5817 + }, + { + "epoch": 2.704788470478847, + "grad_norm": 6.844857692718506, + "learning_rate": 9.9941849994185e-06, + "loss": 0.6007, + "step": 5818 + }, + { + "epoch": 2.705253370525337, + "grad_norm": 3.810985803604126, + "learning_rate": 9.994183999418401e-06, + "loss": 0.1306, + "step": 5819 + }, + { + "epoch": 2.7057182705718272, + "grad_norm": 4.896864414215088, + "learning_rate": 9.9941829994183e-06, + "loss": 0.3946, + "step": 5820 + }, + { + "epoch": 2.706183170618317, + "grad_norm": 9.740348815917969, + "learning_rate": 9.994181999418201e-06, + "loss": 0.8167, + "step": 5821 + }, + { + "epoch": 2.706648070664807, + "grad_norm": 5.615018367767334, + "learning_rate": 9.9941809994181e-06, + "loss": 0.6207, + "step": 5822 + }, + { + "epoch": 2.707112970711297, + "grad_norm": 6.78955602645874, + "learning_rate": 9.994179999418e-06, + "loss": 0.3998, + "step": 5823 + }, + { + "epoch": 2.7075778707577873, + "grad_norm": 4.996095180511475, + "learning_rate": 9.994178999417901e-06, + "loss": 0.1955, + "step": 5824 + }, + { + "epoch": 2.708042770804277, + "grad_norm": 5.9418840408325195, + "learning_rate": 9.994177999417802e-06, + "loss": 0.4438, + "step": 5825 + }, + { + "epoch": 2.708507670850767, + "grad_norm": 7.261411666870117, + "learning_rate": 9.9941769994177e-06, + "loss": 0.6829, + "step": 5826 + }, + { + "epoch": 2.708972570897257, + "grad_norm": 9.123184204101562, + "learning_rate": 9.994175999417601e-06, + "loss": 0.8255, + "step": 5827 + }, + { + "epoch": 2.709437470943747, + "grad_norm": 4.55479097366333, + "learning_rate": 9.9941749994175e-06, + "loss": 0.2932, + "step": 5828 + }, + { + "epoch": 2.7099023709902372, + "grad_norm": 3.9926023483276367, + "learning_rate": 9.994173999417401e-06, + "loss": 0.3107, + "step": 5829 + }, + { + "epoch": 2.710367271036727, + "grad_norm": 4.5506415367126465, + "learning_rate": 9.994172999417302e-06, + "loss": 0.2498, + "step": 5830 + }, + { + "epoch": 2.710832171083217, + "grad_norm": 3.017362594604492, + "learning_rate": 9.9941719994172e-06, + "loss": 0.1476, + "step": 5831 + }, + { + "epoch": 2.711297071129707, + "grad_norm": 5.603129863739014, + "learning_rate": 9.994170999417101e-06, + "loss": 0.5155, + "step": 5832 + }, + { + "epoch": 2.7117619711761973, + "grad_norm": 4.602002143859863, + "learning_rate": 9.994169999417e-06, + "loss": 0.4789, + "step": 5833 + }, + { + "epoch": 2.712226871222687, + "grad_norm": 6.1082611083984375, + "learning_rate": 9.994168999416901e-06, + "loss": 0.1853, + "step": 5834 + }, + { + "epoch": 2.712691771269177, + "grad_norm": 4.941284656524658, + "learning_rate": 9.9941679994168e-06, + "loss": 0.4965, + "step": 5835 + }, + { + "epoch": 2.713156671315667, + "grad_norm": 4.089282989501953, + "learning_rate": 9.9941669994167e-06, + "loss": 0.294, + "step": 5836 + }, + { + "epoch": 2.713621571362157, + "grad_norm": 6.628998756408691, + "learning_rate": 9.994165999416601e-06, + "loss": 0.7467, + "step": 5837 + }, + { + "epoch": 2.7140864714086472, + "grad_norm": 4.2595672607421875, + "learning_rate": 9.994164999416502e-06, + "loss": 0.3352, + "step": 5838 + }, + { + "epoch": 2.714551371455137, + "grad_norm": 3.5889835357666016, + "learning_rate": 9.994163999416401e-06, + "loss": 0.1864, + "step": 5839 + }, + { + "epoch": 2.7150162715016273, + "grad_norm": 6.561220169067383, + "learning_rate": 9.9941629994163e-06, + "loss": 0.7238, + "step": 5840 + }, + { + "epoch": 2.715481171548117, + "grad_norm": 3.4543213844299316, + "learning_rate": 9.9941619994162e-06, + "loss": 0.2343, + "step": 5841 + }, + { + "epoch": 2.7159460715946073, + "grad_norm": 4.779330730438232, + "learning_rate": 9.994160999416101e-06, + "loss": 0.3662, + "step": 5842 + }, + { + "epoch": 2.716410971641097, + "grad_norm": 9.48723316192627, + "learning_rate": 9.994159999416002e-06, + "loss": 0.4248, + "step": 5843 + }, + { + "epoch": 2.716875871687587, + "grad_norm": 6.064364910125732, + "learning_rate": 9.9941589994159e-06, + "loss": 0.6507, + "step": 5844 + }, + { + "epoch": 2.717340771734077, + "grad_norm": 7.758840560913086, + "learning_rate": 9.994157999415801e-06, + "loss": 0.5982, + "step": 5845 + }, + { + "epoch": 2.7178056717805674, + "grad_norm": 7.425261974334717, + "learning_rate": 9.9941569994157e-06, + "loss": 0.5003, + "step": 5846 + }, + { + "epoch": 2.7182705718270572, + "grad_norm": 7.762746810913086, + "learning_rate": 9.994155999415601e-06, + "loss": 0.4408, + "step": 5847 + }, + { + "epoch": 2.718735471873547, + "grad_norm": 6.804959774017334, + "learning_rate": 9.9941549994155e-06, + "loss": 0.484, + "step": 5848 + }, + { + "epoch": 2.7192003719200373, + "grad_norm": 6.089260101318359, + "learning_rate": 9.9941539994154e-06, + "loss": 0.7276, + "step": 5849 + }, + { + "epoch": 2.719665271966527, + "grad_norm": 5.147066116333008, + "learning_rate": 9.994152999415301e-06, + "loss": 0.5324, + "step": 5850 + }, + { + "epoch": 2.7201301720130173, + "grad_norm": 5.999314785003662, + "learning_rate": 9.9941519994152e-06, + "loss": 0.5199, + "step": 5851 + }, + { + "epoch": 2.720595072059507, + "grad_norm": 5.269485950469971, + "learning_rate": 9.994150999415101e-06, + "loss": 0.4302, + "step": 5852 + }, + { + "epoch": 2.7210599721059974, + "grad_norm": 9.485448837280273, + "learning_rate": 9.994149999415e-06, + "loss": 0.4978, + "step": 5853 + }, + { + "epoch": 2.721524872152487, + "grad_norm": 7.336456775665283, + "learning_rate": 9.9941489994149e-06, + "loss": 0.5513, + "step": 5854 + }, + { + "epoch": 2.7219897721989774, + "grad_norm": 5.606080055236816, + "learning_rate": 9.994147999414801e-06, + "loss": 0.3595, + "step": 5855 + }, + { + "epoch": 2.7224546722454672, + "grad_norm": 4.939384937286377, + "learning_rate": 9.994146999414702e-06, + "loss": 0.2194, + "step": 5856 + }, + { + "epoch": 2.722919572291957, + "grad_norm": 8.767638206481934, + "learning_rate": 9.994145999414601e-06, + "loss": 0.6703, + "step": 5857 + }, + { + "epoch": 2.7233844723384473, + "grad_norm": 5.7984843254089355, + "learning_rate": 9.994144999414502e-06, + "loss": 0.5321, + "step": 5858 + }, + { + "epoch": 2.723849372384937, + "grad_norm": 7.845130443572998, + "learning_rate": 9.9941439994144e-06, + "loss": 0.4976, + "step": 5859 + }, + { + "epoch": 2.7243142724314273, + "grad_norm": 6.346533298492432, + "learning_rate": 9.9941429994143e-06, + "loss": 0.5211, + "step": 5860 + }, + { + "epoch": 2.724779172477917, + "grad_norm": 7.074356555938721, + "learning_rate": 9.9941419994142e-06, + "loss": 0.6079, + "step": 5861 + }, + { + "epoch": 2.7252440725244074, + "grad_norm": 5.931107997894287, + "learning_rate": 9.9941409994141e-06, + "loss": 0.5181, + "step": 5862 + }, + { + "epoch": 2.725708972570897, + "grad_norm": 8.804801940917969, + "learning_rate": 9.994139999414001e-06, + "loss": 0.4044, + "step": 5863 + }, + { + "epoch": 2.7261738726173874, + "grad_norm": 3.8510937690734863, + "learning_rate": 9.9941389994139e-06, + "loss": 0.2638, + "step": 5864 + }, + { + "epoch": 2.726638772663877, + "grad_norm": 5.948464870452881, + "learning_rate": 9.994137999413801e-06, + "loss": 0.3879, + "step": 5865 + }, + { + "epoch": 2.727103672710367, + "grad_norm": 4.448903560638428, + "learning_rate": 9.9941369994137e-06, + "loss": 0.3779, + "step": 5866 + }, + { + "epoch": 2.7275685727568573, + "grad_norm": 9.913300514221191, + "learning_rate": 9.9941359994136e-06, + "loss": 0.4678, + "step": 5867 + }, + { + "epoch": 2.7280334728033475, + "grad_norm": 5.367135047912598, + "learning_rate": 9.994134999413501e-06, + "loss": 0.4536, + "step": 5868 + }, + { + "epoch": 2.7284983728498373, + "grad_norm": 6.307028770446777, + "learning_rate": 9.994133999413402e-06, + "loss": 0.7147, + "step": 5869 + }, + { + "epoch": 2.728963272896327, + "grad_norm": 4.210731506347656, + "learning_rate": 9.994132999413301e-06, + "loss": 0.3083, + "step": 5870 + }, + { + "epoch": 2.7294281729428174, + "grad_norm": 3.6654646396636963, + "learning_rate": 9.9941319994132e-06, + "loss": 0.3022, + "step": 5871 + }, + { + "epoch": 2.729893072989307, + "grad_norm": 4.156524658203125, + "learning_rate": 9.9941309994131e-06, + "loss": 0.3199, + "step": 5872 + }, + { + "epoch": 2.7303579730357974, + "grad_norm": 5.649214267730713, + "learning_rate": 9.994129999413e-06, + "loss": 0.2563, + "step": 5873 + }, + { + "epoch": 2.730822873082287, + "grad_norm": 4.653870105743408, + "learning_rate": 9.9941289994129e-06, + "loss": 0.347, + "step": 5874 + }, + { + "epoch": 2.7312877731287775, + "grad_norm": 5.990279674530029, + "learning_rate": 9.994127999412801e-06, + "loss": 0.37, + "step": 5875 + }, + { + "epoch": 2.7317526731752673, + "grad_norm": 7.370069980621338, + "learning_rate": 9.994126999412702e-06, + "loss": 0.7413, + "step": 5876 + }, + { + "epoch": 2.7322175732217575, + "grad_norm": 5.5778679847717285, + "learning_rate": 9.9941259994126e-06, + "loss": 0.4523, + "step": 5877 + }, + { + "epoch": 2.7326824732682473, + "grad_norm": 5.690329551696777, + "learning_rate": 9.994124999412501e-06, + "loss": 0.4589, + "step": 5878 + }, + { + "epoch": 2.733147373314737, + "grad_norm": 7.177591800689697, + "learning_rate": 9.9941239994124e-06, + "loss": 1.1996, + "step": 5879 + }, + { + "epoch": 2.7336122733612274, + "grad_norm": 5.323742866516113, + "learning_rate": 9.9941229994123e-06, + "loss": 0.165, + "step": 5880 + }, + { + "epoch": 2.7340771734077176, + "grad_norm": 3.899819850921631, + "learning_rate": 9.994121999412201e-06, + "loss": 0.3523, + "step": 5881 + }, + { + "epoch": 2.7345420734542074, + "grad_norm": 4.895769119262695, + "learning_rate": 9.9941209994121e-06, + "loss": 0.3704, + "step": 5882 + }, + { + "epoch": 2.735006973500697, + "grad_norm": 5.550821781158447, + "learning_rate": 9.994119999412001e-06, + "loss": 0.4164, + "step": 5883 + }, + { + "epoch": 2.7354718735471875, + "grad_norm": 5.6692352294921875, + "learning_rate": 9.9941189994119e-06, + "loss": 0.3529, + "step": 5884 + }, + { + "epoch": 2.7359367735936773, + "grad_norm": 6.009722709655762, + "learning_rate": 9.9941179994118e-06, + "loss": 0.5636, + "step": 5885 + }, + { + "epoch": 2.7364016736401675, + "grad_norm": 4.156284809112549, + "learning_rate": 9.9941169994117e-06, + "loss": 0.3109, + "step": 5886 + }, + { + "epoch": 2.7368665736866573, + "grad_norm": 8.820524215698242, + "learning_rate": 9.9941159994116e-06, + "loss": 0.373, + "step": 5887 + }, + { + "epoch": 2.737331473733147, + "grad_norm": 4.9549407958984375, + "learning_rate": 9.994114999411501e-06, + "loss": 0.4153, + "step": 5888 + }, + { + "epoch": 2.7377963737796374, + "grad_norm": 4.285280227661133, + "learning_rate": 9.994113999411402e-06, + "loss": 0.2236, + "step": 5889 + }, + { + "epoch": 2.7382612738261276, + "grad_norm": 5.057969093322754, + "learning_rate": 9.9941129994113e-06, + "loss": 0.4381, + "step": 5890 + }, + { + "epoch": 2.7387261738726174, + "grad_norm": 5.4160027503967285, + "learning_rate": 9.994111999411201e-06, + "loss": 0.2873, + "step": 5891 + }, + { + "epoch": 2.739191073919107, + "grad_norm": 5.156164169311523, + "learning_rate": 9.9941109994111e-06, + "loss": 0.2693, + "step": 5892 + }, + { + "epoch": 2.7396559739655975, + "grad_norm": 6.140208721160889, + "learning_rate": 9.994109999411e-06, + "loss": 0.5394, + "step": 5893 + }, + { + "epoch": 2.7401208740120873, + "grad_norm": 7.3551344871521, + "learning_rate": 9.994108999410901e-06, + "loss": 0.5268, + "step": 5894 + }, + { + "epoch": 2.7405857740585775, + "grad_norm": 4.502669811248779, + "learning_rate": 9.9941079994108e-06, + "loss": 0.2051, + "step": 5895 + }, + { + "epoch": 2.7410506741050673, + "grad_norm": 6.747655868530273, + "learning_rate": 9.994106999410701e-06, + "loss": 0.688, + "step": 5896 + }, + { + "epoch": 2.7415155741515576, + "grad_norm": 2.710659980773926, + "learning_rate": 9.9941059994106e-06, + "loss": 0.1303, + "step": 5897 + }, + { + "epoch": 2.7419804741980474, + "grad_norm": 4.941000938415527, + "learning_rate": 9.9941049994105e-06, + "loss": 0.3072, + "step": 5898 + }, + { + "epoch": 2.7424453742445376, + "grad_norm": 6.9468793869018555, + "learning_rate": 9.994103999410401e-06, + "loss": 0.3023, + "step": 5899 + }, + { + "epoch": 2.7429102742910274, + "grad_norm": 5.229516506195068, + "learning_rate": 9.9941029994103e-06, + "loss": 0.5345, + "step": 5900 + }, + { + "epoch": 2.743375174337517, + "grad_norm": 5.909034729003906, + "learning_rate": 9.994101999410201e-06, + "loss": 0.5946, + "step": 5901 + }, + { + "epoch": 2.7438400743840075, + "grad_norm": 6.826345443725586, + "learning_rate": 9.994100999410102e-06, + "loss": 0.5454, + "step": 5902 + }, + { + "epoch": 2.7443049744304977, + "grad_norm": 5.477640151977539, + "learning_rate": 9.99409999941e-06, + "loss": 0.5357, + "step": 5903 + }, + { + "epoch": 2.7447698744769875, + "grad_norm": 5.970934867858887, + "learning_rate": 9.9940989994099e-06, + "loss": 0.5025, + "step": 5904 + }, + { + "epoch": 2.7452347745234773, + "grad_norm": 5.208261489868164, + "learning_rate": 9.9940979994098e-06, + "loss": 0.2832, + "step": 5905 + }, + { + "epoch": 2.7456996745699676, + "grad_norm": 4.947603702545166, + "learning_rate": 9.994096999409701e-06, + "loss": 0.3964, + "step": 5906 + }, + { + "epoch": 2.7461645746164574, + "grad_norm": 4.184919357299805, + "learning_rate": 9.994095999409602e-06, + "loss": 0.3165, + "step": 5907 + }, + { + "epoch": 2.7466294746629476, + "grad_norm": 8.785956382751465, + "learning_rate": 9.9940949994095e-06, + "loss": 0.4063, + "step": 5908 + }, + { + "epoch": 2.7470943747094374, + "grad_norm": 6.7527289390563965, + "learning_rate": 9.994093999409401e-06, + "loss": 0.3721, + "step": 5909 + }, + { + "epoch": 2.747559274755927, + "grad_norm": 5.412710666656494, + "learning_rate": 9.9940929994093e-06, + "loss": 0.3513, + "step": 5910 + }, + { + "epoch": 2.7480241748024175, + "grad_norm": 4.641419410705566, + "learning_rate": 9.9940919994092e-06, + "loss": 0.3119, + "step": 5911 + }, + { + "epoch": 2.7484890748489077, + "grad_norm": 7.117093563079834, + "learning_rate": 9.994090999409101e-06, + "loss": 0.6368, + "step": 5912 + }, + { + "epoch": 2.7489539748953975, + "grad_norm": 7.895333290100098, + "learning_rate": 9.994089999409e-06, + "loss": 0.6145, + "step": 5913 + }, + { + "epoch": 2.7494188749418873, + "grad_norm": 4.867049694061279, + "learning_rate": 9.994088999408901e-06, + "loss": 0.4542, + "step": 5914 + }, + { + "epoch": 2.7498837749883775, + "grad_norm": 6.7427473068237305, + "learning_rate": 9.9940879994088e-06, + "loss": 0.5807, + "step": 5915 + }, + { + "epoch": 2.7503486750348674, + "grad_norm": 3.754481792449951, + "learning_rate": 9.9940869994087e-06, + "loss": 0.1476, + "step": 5916 + }, + { + "epoch": 2.7508135750813576, + "grad_norm": 5.409637451171875, + "learning_rate": 9.9940859994086e-06, + "loss": 0.4085, + "step": 5917 + }, + { + "epoch": 2.7512784751278474, + "grad_norm": 6.959630966186523, + "learning_rate": 9.9940849994085e-06, + "loss": 0.4295, + "step": 5918 + }, + { + "epoch": 2.7517433751743376, + "grad_norm": 7.772634983062744, + "learning_rate": 9.994083999408401e-06, + "loss": 0.5794, + "step": 5919 + }, + { + "epoch": 2.7522082752208274, + "grad_norm": 4.509642124176025, + "learning_rate": 9.994082999408302e-06, + "loss": 0.3617, + "step": 5920 + }, + { + "epoch": 2.7526731752673177, + "grad_norm": 5.20396614074707, + "learning_rate": 9.9940819994082e-06, + "loss": 0.2802, + "step": 5921 + }, + { + "epoch": 2.7531380753138075, + "grad_norm": 4.412148475646973, + "learning_rate": 9.994080999408101e-06, + "loss": 0.4187, + "step": 5922 + }, + { + "epoch": 2.7536029753602973, + "grad_norm": 4.503915786743164, + "learning_rate": 9.994079999408e-06, + "loss": 0.2807, + "step": 5923 + }, + { + "epoch": 2.7540678754067875, + "grad_norm": 5.585047721862793, + "learning_rate": 9.994078999407901e-06, + "loss": 0.343, + "step": 5924 + }, + { + "epoch": 2.754532775453278, + "grad_norm": 5.056324005126953, + "learning_rate": 9.994077999407802e-06, + "loss": 0.4712, + "step": 5925 + }, + { + "epoch": 2.7549976754997676, + "grad_norm": 5.080580711364746, + "learning_rate": 9.9940769994077e-06, + "loss": 0.4115, + "step": 5926 + }, + { + "epoch": 2.7554625755462574, + "grad_norm": 5.513840198516846, + "learning_rate": 9.994075999407601e-06, + "loss": 0.5751, + "step": 5927 + }, + { + "epoch": 2.7559274755927476, + "grad_norm": 5.544318199157715, + "learning_rate": 9.9940749994075e-06, + "loss": 0.4629, + "step": 5928 + }, + { + "epoch": 2.7563923756392374, + "grad_norm": 6.91965913772583, + "learning_rate": 9.9940739994074e-06, + "loss": 0.4592, + "step": 5929 + }, + { + "epoch": 2.7568572756857277, + "grad_norm": 4.989914894104004, + "learning_rate": 9.9940729994073e-06, + "loss": 0.3876, + "step": 5930 + }, + { + "epoch": 2.7573221757322175, + "grad_norm": 6.632987976074219, + "learning_rate": 9.9940719994072e-06, + "loss": 0.6349, + "step": 5931 + }, + { + "epoch": 2.7577870757787077, + "grad_norm": 5.210072040557861, + "learning_rate": 9.994070999407101e-06, + "loss": 0.4489, + "step": 5932 + }, + { + "epoch": 2.7582519758251975, + "grad_norm": 6.107524871826172, + "learning_rate": 9.994069999407002e-06, + "loss": 0.3382, + "step": 5933 + }, + { + "epoch": 2.758716875871688, + "grad_norm": 6.42685604095459, + "learning_rate": 9.9940689994069e-06, + "loss": 0.4299, + "step": 5934 + }, + { + "epoch": 2.7591817759181776, + "grad_norm": 8.72636890411377, + "learning_rate": 9.994067999406801e-06, + "loss": 0.7181, + "step": 5935 + }, + { + "epoch": 2.7596466759646674, + "grad_norm": 6.617051601409912, + "learning_rate": 9.9940669994067e-06, + "loss": 0.3632, + "step": 5936 + }, + { + "epoch": 2.7601115760111576, + "grad_norm": 6.150485992431641, + "learning_rate": 9.994065999406601e-06, + "loss": 0.5728, + "step": 5937 + }, + { + "epoch": 2.7605764760576474, + "grad_norm": 6.229482173919678, + "learning_rate": 9.994064999406502e-06, + "loss": 0.403, + "step": 5938 + }, + { + "epoch": 2.7610413761041377, + "grad_norm": 4.358729362487793, + "learning_rate": 9.9940639994064e-06, + "loss": 0.279, + "step": 5939 + }, + { + "epoch": 2.7615062761506275, + "grad_norm": 4.762620449066162, + "learning_rate": 9.994062999406301e-06, + "loss": 0.4093, + "step": 5940 + }, + { + "epoch": 2.7619711761971177, + "grad_norm": 3.6771039962768555, + "learning_rate": 9.9940619994062e-06, + "loss": 0.2684, + "step": 5941 + }, + { + "epoch": 2.7624360762436075, + "grad_norm": 5.233631134033203, + "learning_rate": 9.994060999406101e-06, + "loss": 0.4478, + "step": 5942 + }, + { + "epoch": 2.762900976290098, + "grad_norm": 4.725604057312012, + "learning_rate": 9.994059999406e-06, + "loss": 0.3728, + "step": 5943 + }, + { + "epoch": 2.7633658763365876, + "grad_norm": 5.610170841217041, + "learning_rate": 9.9940589994059e-06, + "loss": 0.4661, + "step": 5944 + }, + { + "epoch": 2.7638307763830774, + "grad_norm": 5.702184677124023, + "learning_rate": 9.994057999405801e-06, + "loss": 0.3511, + "step": 5945 + }, + { + "epoch": 2.7642956764295676, + "grad_norm": 5.955216884613037, + "learning_rate": 9.994056999405702e-06, + "loss": 0.633, + "step": 5946 + }, + { + "epoch": 2.764760576476058, + "grad_norm": 5.9778313636779785, + "learning_rate": 9.9940559994056e-06, + "loss": 0.8856, + "step": 5947 + }, + { + "epoch": 2.7652254765225477, + "grad_norm": 4.983314514160156, + "learning_rate": 9.9940549994055e-06, + "loss": 0.3065, + "step": 5948 + }, + { + "epoch": 2.7656903765690375, + "grad_norm": 6.667550563812256, + "learning_rate": 9.9940539994054e-06, + "loss": 0.3546, + "step": 5949 + }, + { + "epoch": 2.7661552766155277, + "grad_norm": 7.094269752502441, + "learning_rate": 9.994052999405301e-06, + "loss": 1.0284, + "step": 5950 + }, + { + "epoch": 2.7666201766620175, + "grad_norm": 5.30837869644165, + "learning_rate": 9.994051999405202e-06, + "loss": 0.4252, + "step": 5951 + }, + { + "epoch": 2.767085076708508, + "grad_norm": 3.8155415058135986, + "learning_rate": 9.9940509994051e-06, + "loss": 0.2659, + "step": 5952 + }, + { + "epoch": 2.7675499767549976, + "grad_norm": 3.654851198196411, + "learning_rate": 9.994049999405001e-06, + "loss": 0.4171, + "step": 5953 + }, + { + "epoch": 2.768014876801488, + "grad_norm": 7.009414196014404, + "learning_rate": 9.9940489994049e-06, + "loss": 0.5154, + "step": 5954 + }, + { + "epoch": 2.7684797768479776, + "grad_norm": 4.906233787536621, + "learning_rate": 9.994047999404801e-06, + "loss": 0.435, + "step": 5955 + }, + { + "epoch": 2.768944676894468, + "grad_norm": 4.102152347564697, + "learning_rate": 9.9940469994047e-06, + "loss": 0.3496, + "step": 5956 + }, + { + "epoch": 2.7694095769409577, + "grad_norm": 5.154668807983398, + "learning_rate": 9.9940459994046e-06, + "loss": 0.4433, + "step": 5957 + }, + { + "epoch": 2.7698744769874475, + "grad_norm": 4.967031955718994, + "learning_rate": 9.994044999404501e-06, + "loss": 0.4257, + "step": 5958 + }, + { + "epoch": 2.7703393770339377, + "grad_norm": 6.083144664764404, + "learning_rate": 9.9940439994044e-06, + "loss": 0.5936, + "step": 5959 + }, + { + "epoch": 2.770804277080428, + "grad_norm": 5.452276229858398, + "learning_rate": 9.994042999404301e-06, + "loss": 0.4603, + "step": 5960 + }, + { + "epoch": 2.771269177126918, + "grad_norm": 6.37406063079834, + "learning_rate": 9.9940419994042e-06, + "loss": 0.4056, + "step": 5961 + }, + { + "epoch": 2.7717340771734076, + "grad_norm": 5.332402229309082, + "learning_rate": 9.9940409994041e-06, + "loss": 0.446, + "step": 5962 + }, + { + "epoch": 2.772198977219898, + "grad_norm": 5.28493595123291, + "learning_rate": 9.994039999404001e-06, + "loss": 0.4352, + "step": 5963 + }, + { + "epoch": 2.7726638772663876, + "grad_norm": 6.207212448120117, + "learning_rate": 9.994038999403902e-06, + "loss": 0.5699, + "step": 5964 + }, + { + "epoch": 2.773128777312878, + "grad_norm": 5.591403007507324, + "learning_rate": 9.9940379994038e-06, + "loss": 0.3897, + "step": 5965 + }, + { + "epoch": 2.7735936773593677, + "grad_norm": 5.089874267578125, + "learning_rate": 9.994036999403701e-06, + "loss": 0.3394, + "step": 5966 + }, + { + "epoch": 2.7740585774058575, + "grad_norm": 6.268100738525391, + "learning_rate": 9.9940359994036e-06, + "loss": 0.52, + "step": 5967 + }, + { + "epoch": 2.7745234774523477, + "grad_norm": 4.949416160583496, + "learning_rate": 9.994034999403501e-06, + "loss": 0.267, + "step": 5968 + }, + { + "epoch": 2.774988377498838, + "grad_norm": 5.006519794464111, + "learning_rate": 9.9940339994034e-06, + "loss": 0.5312, + "step": 5969 + }, + { + "epoch": 2.775453277545328, + "grad_norm": 4.3821940422058105, + "learning_rate": 9.9940329994033e-06, + "loss": 0.3318, + "step": 5970 + }, + { + "epoch": 2.7759181775918176, + "grad_norm": 7.444587707519531, + "learning_rate": 9.994031999403201e-06, + "loss": 0.5805, + "step": 5971 + }, + { + "epoch": 2.776383077638308, + "grad_norm": 4.435763359069824, + "learning_rate": 9.9940309994031e-06, + "loss": 0.3848, + "step": 5972 + }, + { + "epoch": 2.7768479776847976, + "grad_norm": 6.322258949279785, + "learning_rate": 9.994029999403001e-06, + "loss": 0.7338, + "step": 5973 + }, + { + "epoch": 2.777312877731288, + "grad_norm": 6.389178276062012, + "learning_rate": 9.9940289994029e-06, + "loss": 0.4847, + "step": 5974 + }, + { + "epoch": 2.7777777777777777, + "grad_norm": 6.2255048751831055, + "learning_rate": 9.9940279994028e-06, + "loss": 0.4476, + "step": 5975 + }, + { + "epoch": 2.778242677824268, + "grad_norm": 5.518285751342773, + "learning_rate": 9.994026999402701e-06, + "loss": 0.594, + "step": 5976 + }, + { + "epoch": 2.7787075778707577, + "grad_norm": 5.581151962280273, + "learning_rate": 9.994025999402602e-06, + "loss": 0.5809, + "step": 5977 + }, + { + "epoch": 2.779172477917248, + "grad_norm": 4.812685012817383, + "learning_rate": 9.9940249994025e-06, + "loss": 0.3975, + "step": 5978 + }, + { + "epoch": 2.7796373779637378, + "grad_norm": 4.486788749694824, + "learning_rate": 9.994023999402401e-06, + "loss": 0.3874, + "step": 5979 + }, + { + "epoch": 2.7801022780102276, + "grad_norm": 4.681423664093018, + "learning_rate": 9.9940229994023e-06, + "loss": 0.2735, + "step": 5980 + }, + { + "epoch": 2.780567178056718, + "grad_norm": 6.313925743103027, + "learning_rate": 9.994021999402201e-06, + "loss": 0.5019, + "step": 5981 + }, + { + "epoch": 2.781032078103208, + "grad_norm": 6.169017314910889, + "learning_rate": 9.9940209994021e-06, + "loss": 0.5827, + "step": 5982 + }, + { + "epoch": 2.781496978149698, + "grad_norm": 6.302913188934326, + "learning_rate": 9.994019999402e-06, + "loss": 0.5836, + "step": 5983 + }, + { + "epoch": 2.7819618781961877, + "grad_norm": 6.59501838684082, + "learning_rate": 9.994018999401901e-06, + "loss": 0.2681, + "step": 5984 + }, + { + "epoch": 2.782426778242678, + "grad_norm": 6.025727272033691, + "learning_rate": 9.9940179994018e-06, + "loss": 0.7319, + "step": 5985 + }, + { + "epoch": 2.7828916782891677, + "grad_norm": 4.913480758666992, + "learning_rate": 9.994016999401701e-06, + "loss": 0.5148, + "step": 5986 + }, + { + "epoch": 2.783356578335658, + "grad_norm": 4.83707332611084, + "learning_rate": 9.9940159994016e-06, + "loss": 0.4121, + "step": 5987 + }, + { + "epoch": 2.7838214783821478, + "grad_norm": 6.406055927276611, + "learning_rate": 9.9940149994015e-06, + "loss": 0.2425, + "step": 5988 + }, + { + "epoch": 2.7842863784286376, + "grad_norm": 6.734202861785889, + "learning_rate": 9.994013999401401e-06, + "loss": 0.5979, + "step": 5989 + }, + { + "epoch": 2.784751278475128, + "grad_norm": 5.124796390533447, + "learning_rate": 9.9940129994013e-06, + "loss": 0.4991, + "step": 5990 + }, + { + "epoch": 2.785216178521618, + "grad_norm": 5.3676533699035645, + "learning_rate": 9.994011999401201e-06, + "loss": 0.7186, + "step": 5991 + }, + { + "epoch": 2.785681078568108, + "grad_norm": 5.056426525115967, + "learning_rate": 9.9940109994011e-06, + "loss": 0.4191, + "step": 5992 + }, + { + "epoch": 2.7861459786145977, + "grad_norm": 4.333265781402588, + "learning_rate": 9.994009999401e-06, + "loss": 0.3317, + "step": 5993 + }, + { + "epoch": 2.786610878661088, + "grad_norm": 6.175198078155518, + "learning_rate": 9.994008999400901e-06, + "loss": 0.4717, + "step": 5994 + }, + { + "epoch": 2.7870757787075777, + "grad_norm": 5.499378681182861, + "learning_rate": 9.994007999400802e-06, + "loss": 0.3445, + "step": 5995 + }, + { + "epoch": 2.787540678754068, + "grad_norm": 4.276416301727295, + "learning_rate": 9.9940069994007e-06, + "loss": 0.3473, + "step": 5996 + }, + { + "epoch": 2.7880055788005578, + "grad_norm": 4.22672176361084, + "learning_rate": 9.994005999400601e-06, + "loss": 0.2006, + "step": 5997 + }, + { + "epoch": 2.788470478847048, + "grad_norm": 7.073930740356445, + "learning_rate": 9.9940049994005e-06, + "loss": 0.5883, + "step": 5998 + }, + { + "epoch": 2.788935378893538, + "grad_norm": 4.135695934295654, + "learning_rate": 9.994003999400401e-06, + "loss": 0.2435, + "step": 5999 + }, + { + "epoch": 2.789400278940028, + "grad_norm": 4.172452926635742, + "learning_rate": 9.9940029994003e-06, + "loss": 0.428, + "step": 6000 + }, + { + "epoch": 2.789865178986518, + "grad_norm": 4.359867572784424, + "learning_rate": 9.9940019994002e-06, + "loss": 0.2446, + "step": 6001 + }, + { + "epoch": 2.7903300790330077, + "grad_norm": 5.228458404541016, + "learning_rate": 9.994000999400101e-06, + "loss": 0.4986, + "step": 6002 + }, + { + "epoch": 2.790794979079498, + "grad_norm": 6.8362908363342285, + "learning_rate": 9.9939999994e-06, + "loss": 0.3159, + "step": 6003 + }, + { + "epoch": 2.791259879125988, + "grad_norm": 5.303159236907959, + "learning_rate": 9.993998999399901e-06, + "loss": 0.4093, + "step": 6004 + }, + { + "epoch": 2.791724779172478, + "grad_norm": 6.427446365356445, + "learning_rate": 9.9939979993998e-06, + "loss": 0.6949, + "step": 6005 + }, + { + "epoch": 2.7921896792189678, + "grad_norm": 5.819038391113281, + "learning_rate": 9.9939969993997e-06, + "loss": 0.513, + "step": 6006 + }, + { + "epoch": 2.792654579265458, + "grad_norm": 3.907478094100952, + "learning_rate": 9.993995999399601e-06, + "loss": 0.3072, + "step": 6007 + }, + { + "epoch": 2.793119479311948, + "grad_norm": 4.373895168304443, + "learning_rate": 9.993994999399502e-06, + "loss": 0.4931, + "step": 6008 + }, + { + "epoch": 2.793584379358438, + "grad_norm": 6.057862758636475, + "learning_rate": 9.993993999399401e-06, + "loss": 0.516, + "step": 6009 + }, + { + "epoch": 2.794049279404928, + "grad_norm": 6.4961018562316895, + "learning_rate": 9.993992999399302e-06, + "loss": 0.4773, + "step": 6010 + }, + { + "epoch": 2.7945141794514177, + "grad_norm": 4.708291530609131, + "learning_rate": 9.9939919993992e-06, + "loss": 0.5059, + "step": 6011 + }, + { + "epoch": 2.794979079497908, + "grad_norm": 4.807563304901123, + "learning_rate": 9.9939909993991e-06, + "loss": 0.328, + "step": 6012 + }, + { + "epoch": 2.795443979544398, + "grad_norm": 2.9572317600250244, + "learning_rate": 9.993989999399e-06, + "loss": 0.1537, + "step": 6013 + }, + { + "epoch": 2.795908879590888, + "grad_norm": 5.345416069030762, + "learning_rate": 9.9939889993989e-06, + "loss": 0.3809, + "step": 6014 + }, + { + "epoch": 2.7963737796373778, + "grad_norm": 6.894497394561768, + "learning_rate": 9.993987999398801e-06, + "loss": 0.5016, + "step": 6015 + }, + { + "epoch": 2.796838679683868, + "grad_norm": 5.720743656158447, + "learning_rate": 9.9939869993987e-06, + "loss": 0.4053, + "step": 6016 + }, + { + "epoch": 2.797303579730358, + "grad_norm": 4.584404468536377, + "learning_rate": 9.993985999398601e-06, + "loss": 0.3208, + "step": 6017 + }, + { + "epoch": 2.797768479776848, + "grad_norm": 4.777254104614258, + "learning_rate": 9.9939849993985e-06, + "loss": 0.4689, + "step": 6018 + }, + { + "epoch": 2.798233379823338, + "grad_norm": 4.9758219718933105, + "learning_rate": 9.9939839993984e-06, + "loss": 0.4206, + "step": 6019 + }, + { + "epoch": 2.798698279869828, + "grad_norm": 6.787013053894043, + "learning_rate": 9.993982999398301e-06, + "loss": 0.5157, + "step": 6020 + }, + { + "epoch": 2.799163179916318, + "grad_norm": 5.106445789337158, + "learning_rate": 9.993981999398202e-06, + "loss": 0.3866, + "step": 6021 + }, + { + "epoch": 2.799628079962808, + "grad_norm": 10.081835746765137, + "learning_rate": 9.993980999398101e-06, + "loss": 0.6283, + "step": 6022 + }, + { + "epoch": 2.800092980009298, + "grad_norm": 4.568618297576904, + "learning_rate": 9.993979999398e-06, + "loss": 0.2941, + "step": 6023 + }, + { + "epoch": 2.8005578800557878, + "grad_norm": 4.929654121398926, + "learning_rate": 9.9939789993979e-06, + "loss": 0.3462, + "step": 6024 + }, + { + "epoch": 2.801022780102278, + "grad_norm": 6.8580851554870605, + "learning_rate": 9.9939779993978e-06, + "loss": 0.4965, + "step": 6025 + }, + { + "epoch": 2.8014876801487683, + "grad_norm": 4.225681304931641, + "learning_rate": 9.9939769993977e-06, + "loss": 0.2504, + "step": 6026 + }, + { + "epoch": 2.801952580195258, + "grad_norm": 3.9875433444976807, + "learning_rate": 9.993975999397601e-06, + "loss": 0.3546, + "step": 6027 + }, + { + "epoch": 2.802417480241748, + "grad_norm": 6.836137294769287, + "learning_rate": 9.993974999397502e-06, + "loss": 0.6556, + "step": 6028 + }, + { + "epoch": 2.802882380288238, + "grad_norm": 7.078554630279541, + "learning_rate": 9.9939739993974e-06, + "loss": 0.4564, + "step": 6029 + }, + { + "epoch": 2.803347280334728, + "grad_norm": 5.053871154785156, + "learning_rate": 9.993972999397301e-06, + "loss": 0.5075, + "step": 6030 + }, + { + "epoch": 2.803812180381218, + "grad_norm": 4.376038551330566, + "learning_rate": 9.9939719993972e-06, + "loss": 0.353, + "step": 6031 + }, + { + "epoch": 2.804277080427708, + "grad_norm": 6.834493637084961, + "learning_rate": 9.9939709993971e-06, + "loss": 0.6437, + "step": 6032 + }, + { + "epoch": 2.804741980474198, + "grad_norm": 8.264509201049805, + "learning_rate": 9.993969999397001e-06, + "loss": 0.8622, + "step": 6033 + }, + { + "epoch": 2.805206880520688, + "grad_norm": 6.329646587371826, + "learning_rate": 9.9939689993969e-06, + "loss": 0.4305, + "step": 6034 + }, + { + "epoch": 2.8056717805671783, + "grad_norm": 4.235045433044434, + "learning_rate": 9.993967999396801e-06, + "loss": 0.4256, + "step": 6035 + }, + { + "epoch": 2.806136680613668, + "grad_norm": 5.8256378173828125, + "learning_rate": 9.9939669993967e-06, + "loss": 0.5401, + "step": 6036 + }, + { + "epoch": 2.806601580660158, + "grad_norm": 6.391780376434326, + "learning_rate": 9.9939659993966e-06, + "loss": 0.4662, + "step": 6037 + }, + { + "epoch": 2.807066480706648, + "grad_norm": 6.391071319580078, + "learning_rate": 9.9939649993965e-06, + "loss": 0.5989, + "step": 6038 + }, + { + "epoch": 2.8075313807531384, + "grad_norm": 3.711683988571167, + "learning_rate": 9.9939639993964e-06, + "loss": 0.3008, + "step": 6039 + }, + { + "epoch": 2.807996280799628, + "grad_norm": 6.461827754974365, + "learning_rate": 9.993962999396301e-06, + "loss": 0.6419, + "step": 6040 + }, + { + "epoch": 2.808461180846118, + "grad_norm": 3.638693332672119, + "learning_rate": 9.993961999396202e-06, + "loss": 0.226, + "step": 6041 + }, + { + "epoch": 2.808926080892608, + "grad_norm": 6.567357540130615, + "learning_rate": 9.9939609993961e-06, + "loss": 0.352, + "step": 6042 + }, + { + "epoch": 2.809390980939098, + "grad_norm": 4.6498823165893555, + "learning_rate": 9.993959999396001e-06, + "loss": 0.325, + "step": 6043 + }, + { + "epoch": 2.8098558809855883, + "grad_norm": 5.020832538604736, + "learning_rate": 9.9939589993959e-06, + "loss": 0.4299, + "step": 6044 + }, + { + "epoch": 2.810320781032078, + "grad_norm": 5.598250389099121, + "learning_rate": 9.993957999395801e-06, + "loss": 0.5903, + "step": 6045 + }, + { + "epoch": 2.810785681078568, + "grad_norm": 5.8697052001953125, + "learning_rate": 9.993956999395702e-06, + "loss": 0.4284, + "step": 6046 + }, + { + "epoch": 2.811250581125058, + "grad_norm": 6.919435501098633, + "learning_rate": 9.9939559993956e-06, + "loss": 0.4255, + "step": 6047 + }, + { + "epoch": 2.8117154811715483, + "grad_norm": 5.516733169555664, + "learning_rate": 9.993954999395501e-06, + "loss": 0.6922, + "step": 6048 + }, + { + "epoch": 2.812180381218038, + "grad_norm": 5.012948036193848, + "learning_rate": 9.9939539993954e-06, + "loss": 0.3954, + "step": 6049 + }, + { + "epoch": 2.812645281264528, + "grad_norm": 4.081091403961182, + "learning_rate": 9.9939529993953e-06, + "loss": 0.3312, + "step": 6050 + }, + { + "epoch": 2.813110181311018, + "grad_norm": 4.708012580871582, + "learning_rate": 9.9939519993952e-06, + "loss": 0.4104, + "step": 6051 + }, + { + "epoch": 2.813575081357508, + "grad_norm": 5.769990921020508, + "learning_rate": 9.9939509993951e-06, + "loss": 0.4421, + "step": 6052 + }, + { + "epoch": 2.8140399814039982, + "grad_norm": 4.801267147064209, + "learning_rate": 9.993949999395001e-06, + "loss": 0.3634, + "step": 6053 + }, + { + "epoch": 2.814504881450488, + "grad_norm": 7.989553451538086, + "learning_rate": 9.993948999394902e-06, + "loss": 0.4487, + "step": 6054 + }, + { + "epoch": 2.8149697814969783, + "grad_norm": 6.6078901290893555, + "learning_rate": 9.9939479993948e-06, + "loss": 0.6563, + "step": 6055 + }, + { + "epoch": 2.815434681543468, + "grad_norm": 4.863282203674316, + "learning_rate": 9.9939469993947e-06, + "loss": 0.2004, + "step": 6056 + }, + { + "epoch": 2.8158995815899583, + "grad_norm": 7.618734836578369, + "learning_rate": 9.9939459993946e-06, + "loss": 0.6417, + "step": 6057 + }, + { + "epoch": 2.816364481636448, + "grad_norm": 5.2107367515563965, + "learning_rate": 9.993944999394501e-06, + "loss": 0.2791, + "step": 6058 + }, + { + "epoch": 2.816829381682938, + "grad_norm": 9.135947227478027, + "learning_rate": 9.993943999394402e-06, + "loss": 0.8873, + "step": 6059 + }, + { + "epoch": 2.817294281729428, + "grad_norm": 5.723328590393066, + "learning_rate": 9.9939429993943e-06, + "loss": 0.4342, + "step": 6060 + }, + { + "epoch": 2.8177591817759184, + "grad_norm": 7.058468818664551, + "learning_rate": 9.993941999394201e-06, + "loss": 0.5009, + "step": 6061 + }, + { + "epoch": 2.8182240818224082, + "grad_norm": 8.714787483215332, + "learning_rate": 9.9939409993941e-06, + "loss": 0.5691, + "step": 6062 + }, + { + "epoch": 2.818688981868898, + "grad_norm": 4.979973316192627, + "learning_rate": 9.993939999394001e-06, + "loss": 0.4218, + "step": 6063 + }, + { + "epoch": 2.8191538819153883, + "grad_norm": 4.452539443969727, + "learning_rate": 9.993938999393902e-06, + "loss": 0.3087, + "step": 6064 + }, + { + "epoch": 2.819618781961878, + "grad_norm": 6.484631538391113, + "learning_rate": 9.9939379993938e-06, + "loss": 0.4263, + "step": 6065 + }, + { + "epoch": 2.8200836820083683, + "grad_norm": 5.585823059082031, + "learning_rate": 9.993936999393701e-06, + "loss": 0.4551, + "step": 6066 + }, + { + "epoch": 2.820548582054858, + "grad_norm": 4.775347709655762, + "learning_rate": 9.9939359993936e-06, + "loss": 0.3115, + "step": 6067 + }, + { + "epoch": 2.821013482101348, + "grad_norm": 4.885765075683594, + "learning_rate": 9.9939349993935e-06, + "loss": 0.3965, + "step": 6068 + }, + { + "epoch": 2.821478382147838, + "grad_norm": 3.913886785507202, + "learning_rate": 9.9939339993934e-06, + "loss": 0.3764, + "step": 6069 + }, + { + "epoch": 2.8219432821943284, + "grad_norm": 5.349390983581543, + "learning_rate": 9.9939329993933e-06, + "loss": 0.4855, + "step": 6070 + }, + { + "epoch": 2.8224081822408182, + "grad_norm": 5.175655364990234, + "learning_rate": 9.993931999393201e-06, + "loss": 0.4782, + "step": 6071 + }, + { + "epoch": 2.822873082287308, + "grad_norm": 5.074904918670654, + "learning_rate": 9.993930999393102e-06, + "loss": 0.5283, + "step": 6072 + }, + { + "epoch": 2.8233379823337983, + "grad_norm": 4.850214958190918, + "learning_rate": 9.993929999393e-06, + "loss": 0.381, + "step": 6073 + }, + { + "epoch": 2.823802882380288, + "grad_norm": 5.030768871307373, + "learning_rate": 9.993928999392901e-06, + "loss": 0.3564, + "step": 6074 + }, + { + "epoch": 2.8242677824267783, + "grad_norm": 4.681774139404297, + "learning_rate": 9.9939279993928e-06, + "loss": 0.2811, + "step": 6075 + }, + { + "epoch": 2.824732682473268, + "grad_norm": 5.4322919845581055, + "learning_rate": 9.993926999392701e-06, + "loss": 0.4741, + "step": 6076 + }, + { + "epoch": 2.8251975825197584, + "grad_norm": 6.600274562835693, + "learning_rate": 9.993925999392602e-06, + "loss": 0.6859, + "step": 6077 + }, + { + "epoch": 2.825662482566248, + "grad_norm": 8.119295120239258, + "learning_rate": 9.9939249993925e-06, + "loss": 0.1811, + "step": 6078 + }, + { + "epoch": 2.8261273826127384, + "grad_norm": 15.87875747680664, + "learning_rate": 9.993923999392401e-06, + "loss": 0.394, + "step": 6079 + }, + { + "epoch": 2.8265922826592282, + "grad_norm": 7.972775459289551, + "learning_rate": 9.9939229993923e-06, + "loss": 0.4542, + "step": 6080 + }, + { + "epoch": 2.827057182705718, + "grad_norm": 7.550102710723877, + "learning_rate": 9.9939219993922e-06, + "loss": 0.5667, + "step": 6081 + }, + { + "epoch": 2.8275220827522083, + "grad_norm": 6.9346795082092285, + "learning_rate": 9.9939209993921e-06, + "loss": 0.921, + "step": 6082 + }, + { + "epoch": 2.8279869827986985, + "grad_norm": 3.8629140853881836, + "learning_rate": 9.993919999392e-06, + "loss": 0.2466, + "step": 6083 + }, + { + "epoch": 2.8284518828451883, + "grad_norm": Infinity, + "learning_rate": 9.993919999392e-06, + "loss": 0.6134, + "step": 6084 + }, + { + "epoch": 2.828916782891678, + "grad_norm": 3.909471273422241, + "learning_rate": 9.993918999391901e-06, + "loss": 0.089, + "step": 6085 + }, + { + "epoch": 2.8293816829381684, + "grad_norm": 9.430512428283691, + "learning_rate": 9.993917999391802e-06, + "loss": 0.5998, + "step": 6086 + }, + { + "epoch": 2.829846582984658, + "grad_norm": 8.669350624084473, + "learning_rate": 9.9939169993917e-06, + "loss": 0.5285, + "step": 6087 + }, + { + "epoch": 2.8303114830311484, + "grad_norm": 4.404876232147217, + "learning_rate": 9.993915999391601e-06, + "loss": 0.197, + "step": 6088 + }, + { + "epoch": 2.8307763830776382, + "grad_norm": 6.605991363525391, + "learning_rate": 9.9939149993915e-06, + "loss": 0.341, + "step": 6089 + }, + { + "epoch": 2.831241283124128, + "grad_norm": 4.977310657501221, + "learning_rate": 9.993913999391401e-06, + "loss": 0.2401, + "step": 6090 + }, + { + "epoch": 2.8317061831706183, + "grad_norm": 5.63334321975708, + "learning_rate": 9.993912999391302e-06, + "loss": 0.4626, + "step": 6091 + }, + { + "epoch": 2.8321710832171085, + "grad_norm": 6.132293701171875, + "learning_rate": 9.9939119993912e-06, + "loss": 0.3875, + "step": 6092 + }, + { + "epoch": 2.8326359832635983, + "grad_norm": 4.991438388824463, + "learning_rate": 9.993910999391101e-06, + "loss": 0.3418, + "step": 6093 + }, + { + "epoch": 2.833100883310088, + "grad_norm": 5.403930187225342, + "learning_rate": 9.993909999391e-06, + "loss": 0.5131, + "step": 6094 + }, + { + "epoch": 2.8335657833565784, + "grad_norm": 6.281976222991943, + "learning_rate": 9.993908999390901e-06, + "loss": 0.6847, + "step": 6095 + }, + { + "epoch": 2.834030683403068, + "grad_norm": 6.7756147384643555, + "learning_rate": 9.9939079993908e-06, + "loss": 0.8094, + "step": 6096 + }, + { + "epoch": 2.8344955834495584, + "grad_norm": 7.209840774536133, + "learning_rate": 9.9939069993907e-06, + "loss": 0.7158, + "step": 6097 + }, + { + "epoch": 2.8349604834960482, + "grad_norm": 9.26366138458252, + "learning_rate": 9.993905999390601e-06, + "loss": 0.5644, + "step": 6098 + }, + { + "epoch": 2.8354253835425385, + "grad_norm": 3.362673044204712, + "learning_rate": 9.993904999390502e-06, + "loss": 0.1766, + "step": 6099 + }, + { + "epoch": 2.8358902835890283, + "grad_norm": 5.701622009277344, + "learning_rate": 9.9939039993904e-06, + "loss": 0.5164, + "step": 6100 + }, + { + "epoch": 2.8363551836355185, + "grad_norm": 6.188033580780029, + "learning_rate": 9.9939029993903e-06, + "loss": 0.6888, + "step": 6101 + }, + { + "epoch": 2.8368200836820083, + "grad_norm": 6.296255588531494, + "learning_rate": 9.9939019993902e-06, + "loss": 0.4475, + "step": 6102 + }, + { + "epoch": 2.837284983728498, + "grad_norm": 5.613093852996826, + "learning_rate": 9.993900999390101e-06, + "loss": 0.5214, + "step": 6103 + }, + { + "epoch": 2.8377498837749884, + "grad_norm": 5.637208461761475, + "learning_rate": 9.993899999390002e-06, + "loss": 0.4517, + "step": 6104 + }, + { + "epoch": 2.8382147838214786, + "grad_norm": 5.072276592254639, + "learning_rate": 9.9938989993899e-06, + "loss": 0.5681, + "step": 6105 + }, + { + "epoch": 2.8386796838679684, + "grad_norm": 6.738268852233887, + "learning_rate": 9.993897999389801e-06, + "loss": 0.5435, + "step": 6106 + }, + { + "epoch": 2.8391445839144582, + "grad_norm": 7.8566107749938965, + "learning_rate": 9.9938969993897e-06, + "loss": 0.6813, + "step": 6107 + }, + { + "epoch": 2.8396094839609485, + "grad_norm": 8.223309516906738, + "learning_rate": 9.993895999389601e-06, + "loss": 0.466, + "step": 6108 + }, + { + "epoch": 2.8400743840074383, + "grad_norm": 9.914838790893555, + "learning_rate": 9.9938949993895e-06, + "loss": 0.5283, + "step": 6109 + }, + { + "epoch": 2.8405392840539285, + "grad_norm": 4.984903812408447, + "learning_rate": 9.9938939993894e-06, + "loss": 0.3181, + "step": 6110 + }, + { + "epoch": 2.8410041841004183, + "grad_norm": 5.007750988006592, + "learning_rate": 9.993892999389301e-06, + "loss": 0.487, + "step": 6111 + }, + { + "epoch": 2.8414690841469086, + "grad_norm": 6.384644031524658, + "learning_rate": 9.9938919993892e-06, + "loss": 0.6007, + "step": 6112 + }, + { + "epoch": 2.8419339841933984, + "grad_norm": 5.101530075073242, + "learning_rate": 9.993890999389101e-06, + "loss": 0.4747, + "step": 6113 + }, + { + "epoch": 2.8423988842398886, + "grad_norm": 5.268805503845215, + "learning_rate": 9.993889999389e-06, + "loss": 0.3963, + "step": 6114 + }, + { + "epoch": 2.8428637842863784, + "grad_norm": 5.810764312744141, + "learning_rate": 9.9938889993889e-06, + "loss": 0.2982, + "step": 6115 + }, + { + "epoch": 2.8433286843328682, + "grad_norm": 3.7932207584381104, + "learning_rate": 9.993887999388801e-06, + "loss": 0.3225, + "step": 6116 + }, + { + "epoch": 2.8437935843793585, + "grad_norm": 3.2192301750183105, + "learning_rate": 9.993886999388702e-06, + "loss": 0.1917, + "step": 6117 + }, + { + "epoch": 2.8442584844258483, + "grad_norm": 7.127127647399902, + "learning_rate": 9.9938859993886e-06, + "loss": 0.6803, + "step": 6118 + }, + { + "epoch": 2.8447233844723385, + "grad_norm": 4.730709075927734, + "learning_rate": 9.993884999388501e-06, + "loss": 0.283, + "step": 6119 + }, + { + "epoch": 2.8451882845188283, + "grad_norm": 5.5086870193481445, + "learning_rate": 9.9938839993884e-06, + "loss": 0.5137, + "step": 6120 + }, + { + "epoch": 2.8456531845653186, + "grad_norm": 6.23792028427124, + "learning_rate": 9.993882999388301e-06, + "loss": 0.4357, + "step": 6121 + }, + { + "epoch": 2.8461180846118084, + "grad_norm": 5.879896640777588, + "learning_rate": 9.9938819993882e-06, + "loss": 0.5938, + "step": 6122 + }, + { + "epoch": 2.8465829846582986, + "grad_norm": 4.6238813400268555, + "learning_rate": 9.9938809993881e-06, + "loss": 0.3026, + "step": 6123 + }, + { + "epoch": 2.8470478847047884, + "grad_norm": 3.3456437587738037, + "learning_rate": 9.993879999388001e-06, + "loss": 0.1744, + "step": 6124 + }, + { + "epoch": 2.8475127847512782, + "grad_norm": 8.762755393981934, + "learning_rate": 9.9938789993879e-06, + "loss": 0.7157, + "step": 6125 + }, + { + "epoch": 2.8479776847977685, + "grad_norm": 4.35617733001709, + "learning_rate": 9.993877999387801e-06, + "loss": 0.3312, + "step": 6126 + }, + { + "epoch": 2.8484425848442587, + "grad_norm": 5.140875816345215, + "learning_rate": 9.9938769993877e-06, + "loss": 0.3846, + "step": 6127 + }, + { + "epoch": 2.8489074848907485, + "grad_norm": 4.36065149307251, + "learning_rate": 9.9938759993876e-06, + "loss": 0.263, + "step": 6128 + }, + { + "epoch": 2.8493723849372383, + "grad_norm": 3.4856936931610107, + "learning_rate": 9.993874999387501e-06, + "loss": 0.1224, + "step": 6129 + }, + { + "epoch": 2.8498372849837286, + "grad_norm": 5.062428951263428, + "learning_rate": 9.993873999387402e-06, + "loss": 0.3628, + "step": 6130 + }, + { + "epoch": 2.8503021850302184, + "grad_norm": 7.660738468170166, + "learning_rate": 9.993872999387301e-06, + "loss": 0.6005, + "step": 6131 + }, + { + "epoch": 2.8507670850767086, + "grad_norm": 6.510570526123047, + "learning_rate": 9.9938719993872e-06, + "loss": 0.6574, + "step": 6132 + }, + { + "epoch": 2.8512319851231984, + "grad_norm": 2.402975082397461, + "learning_rate": 9.9938709993871e-06, + "loss": 0.1212, + "step": 6133 + }, + { + "epoch": 2.8516968851696887, + "grad_norm": 5.346009254455566, + "learning_rate": 9.993869999387001e-06, + "loss": 0.3183, + "step": 6134 + }, + { + "epoch": 2.8521617852161785, + "grad_norm": 5.0023393630981445, + "learning_rate": 9.9938689993869e-06, + "loss": 0.5344, + "step": 6135 + }, + { + "epoch": 2.8526266852626687, + "grad_norm": 5.136613845825195, + "learning_rate": 9.9938679993868e-06, + "loss": 0.3789, + "step": 6136 + }, + { + "epoch": 2.8530915853091585, + "grad_norm": 5.3972272872924805, + "learning_rate": 9.993866999386701e-06, + "loss": 0.4624, + "step": 6137 + }, + { + "epoch": 2.8535564853556483, + "grad_norm": 7.341054916381836, + "learning_rate": 9.9938659993866e-06, + "loss": 0.6226, + "step": 6138 + }, + { + "epoch": 2.8540213854021386, + "grad_norm": 8.815167427062988, + "learning_rate": 9.993864999386501e-06, + "loss": 0.6307, + "step": 6139 + }, + { + "epoch": 2.854486285448629, + "grad_norm": 3.904153347015381, + "learning_rate": 9.9938639993864e-06, + "loss": 0.3655, + "step": 6140 + }, + { + "epoch": 2.8549511854951186, + "grad_norm": 4.101628303527832, + "learning_rate": 9.9938629993863e-06, + "loss": 0.2272, + "step": 6141 + }, + { + "epoch": 2.8554160855416084, + "grad_norm": 5.509921550750732, + "learning_rate": 9.993861999386201e-06, + "loss": 0.4334, + "step": 6142 + }, + { + "epoch": 2.8558809855880987, + "grad_norm": 6.004958629608154, + "learning_rate": 9.9938609993861e-06, + "loss": 0.3751, + "step": 6143 + }, + { + "epoch": 2.8563458856345885, + "grad_norm": 4.344845771789551, + "learning_rate": 9.993859999386001e-06, + "loss": 0.3374, + "step": 6144 + }, + { + "epoch": 2.8568107856810787, + "grad_norm": 5.274522304534912, + "learning_rate": 9.9938589993859e-06, + "loss": 0.3819, + "step": 6145 + }, + { + "epoch": 2.8572756857275685, + "grad_norm": 5.4383463859558105, + "learning_rate": 9.9938579993858e-06, + "loss": 0.3161, + "step": 6146 + }, + { + "epoch": 2.8577405857740583, + "grad_norm": 4.486917972564697, + "learning_rate": 9.993856999385701e-06, + "loss": 0.3119, + "step": 6147 + }, + { + "epoch": 2.8582054858205486, + "grad_norm": 15.87653636932373, + "learning_rate": 9.9938559993856e-06, + "loss": 0.313, + "step": 6148 + }, + { + "epoch": 2.858670385867039, + "grad_norm": 6.679238319396973, + "learning_rate": 9.993854999385501e-06, + "loss": 0.6426, + "step": 6149 + }, + { + "epoch": 2.8591352859135286, + "grad_norm": 5.543515682220459, + "learning_rate": 9.993853999385402e-06, + "loss": 0.2651, + "step": 6150 + }, + { + "epoch": 2.8596001859600184, + "grad_norm": 5.7297234535217285, + "learning_rate": 9.9938529993853e-06, + "loss": 0.582, + "step": 6151 + }, + { + "epoch": 2.8600650860065087, + "grad_norm": 7.0750226974487305, + "learning_rate": 9.993851999385201e-06, + "loss": 0.4325, + "step": 6152 + }, + { + "epoch": 2.8605299860529985, + "grad_norm": 5.646607398986816, + "learning_rate": 9.9938509993851e-06, + "loss": 0.3917, + "step": 6153 + }, + { + "epoch": 2.8609948860994887, + "grad_norm": 8.506218910217285, + "learning_rate": 9.993849999385e-06, + "loss": 0.5967, + "step": 6154 + }, + { + "epoch": 2.8614597861459785, + "grad_norm": 5.475658893585205, + "learning_rate": 9.993848999384901e-06, + "loss": 0.3453, + "step": 6155 + }, + { + "epoch": 2.8619246861924688, + "grad_norm": 6.5244059562683105, + "learning_rate": 9.9938479993848e-06, + "loss": 0.5467, + "step": 6156 + }, + { + "epoch": 2.8623895862389586, + "grad_norm": 6.497305870056152, + "learning_rate": 9.993846999384701e-06, + "loss": 0.4893, + "step": 6157 + }, + { + "epoch": 2.862854486285449, + "grad_norm": 4.843864917755127, + "learning_rate": 9.9938459993846e-06, + "loss": 0.3966, + "step": 6158 + }, + { + "epoch": 2.8633193863319386, + "grad_norm": 4.05419397354126, + "learning_rate": 9.9938449993845e-06, + "loss": 0.2378, + "step": 6159 + }, + { + "epoch": 2.8637842863784284, + "grad_norm": 4.282208442687988, + "learning_rate": 9.993843999384401e-06, + "loss": 0.342, + "step": 6160 + }, + { + "epoch": 2.8642491864249187, + "grad_norm": 7.4680938720703125, + "learning_rate": 9.993842999384302e-06, + "loss": 0.7928, + "step": 6161 + }, + { + "epoch": 2.864714086471409, + "grad_norm": 4.940791606903076, + "learning_rate": 9.993841999384201e-06, + "loss": 0.2823, + "step": 6162 + }, + { + "epoch": 2.8651789865178987, + "grad_norm": 4.051150798797607, + "learning_rate": 9.993840999384102e-06, + "loss": 0.3124, + "step": 6163 + }, + { + "epoch": 2.8656438865643885, + "grad_norm": 5.481119155883789, + "learning_rate": 9.993839999384e-06, + "loss": 0.5869, + "step": 6164 + }, + { + "epoch": 2.8661087866108788, + "grad_norm": 8.314779281616211, + "learning_rate": 9.9938389993839e-06, + "loss": 0.8149, + "step": 6165 + }, + { + "epoch": 2.8665736866573686, + "grad_norm": 6.7549662590026855, + "learning_rate": 9.9938379993838e-06, + "loss": 0.2946, + "step": 6166 + }, + { + "epoch": 2.867038586703859, + "grad_norm": 9.999578475952148, + "learning_rate": 9.9938369993837e-06, + "loss": 0.6694, + "step": 6167 + }, + { + "epoch": 2.8675034867503486, + "grad_norm": 6.2480788230896, + "learning_rate": 9.993835999383601e-06, + "loss": 0.6076, + "step": 6168 + }, + { + "epoch": 2.8679683867968384, + "grad_norm": 4.544184684753418, + "learning_rate": 9.9938349993835e-06, + "loss": 0.3849, + "step": 6169 + }, + { + "epoch": 2.8684332868433287, + "grad_norm": 5.317014694213867, + "learning_rate": 9.993833999383401e-06, + "loss": 0.3808, + "step": 6170 + }, + { + "epoch": 2.868898186889819, + "grad_norm": 6.7982892990112305, + "learning_rate": 9.9938329993833e-06, + "loss": 0.6, + "step": 6171 + }, + { + "epoch": 2.8693630869363087, + "grad_norm": 11.310079574584961, + "learning_rate": 9.9938319993832e-06, + "loss": 0.6388, + "step": 6172 + }, + { + "epoch": 2.8698279869827985, + "grad_norm": 4.913380146026611, + "learning_rate": 9.993830999383101e-06, + "loss": 0.4035, + "step": 6173 + }, + { + "epoch": 2.8702928870292888, + "grad_norm": 3.8810300827026367, + "learning_rate": 9.993829999383002e-06, + "loss": 0.2548, + "step": 6174 + }, + { + "epoch": 2.8707577870757786, + "grad_norm": 4.452574253082275, + "learning_rate": 9.993828999382901e-06, + "loss": 0.2612, + "step": 6175 + }, + { + "epoch": 2.871222687122269, + "grad_norm": 4.66523551940918, + "learning_rate": 9.9938279993828e-06, + "loss": 0.3713, + "step": 6176 + }, + { + "epoch": 2.8716875871687586, + "grad_norm": 6.587453842163086, + "learning_rate": 9.9938269993827e-06, + "loss": 0.6814, + "step": 6177 + }, + { + "epoch": 2.872152487215249, + "grad_norm": 3.1144425868988037, + "learning_rate": 9.9938259993826e-06, + "loss": 0.1486, + "step": 6178 + }, + { + "epoch": 2.8726173872617387, + "grad_norm": 6.8966803550720215, + "learning_rate": 9.9938249993825e-06, + "loss": 0.387, + "step": 6179 + }, + { + "epoch": 2.873082287308229, + "grad_norm": 9.511645317077637, + "learning_rate": 9.993823999382401e-06, + "loss": 2.2146, + "step": 6180 + }, + { + "epoch": 2.8735471873547187, + "grad_norm": 5.070169448852539, + "learning_rate": 9.993822999382302e-06, + "loss": 0.5042, + "step": 6181 + }, + { + "epoch": 2.8740120874012085, + "grad_norm": 5.706877708435059, + "learning_rate": 9.9938219993822e-06, + "loss": 0.3165, + "step": 6182 + }, + { + "epoch": 2.8744769874476988, + "grad_norm": 6.031602382659912, + "learning_rate": 9.993820999382101e-06, + "loss": 0.6506, + "step": 6183 + }, + { + "epoch": 2.874941887494189, + "grad_norm": 5.177016735076904, + "learning_rate": 9.993819999382e-06, + "loss": 0.2259, + "step": 6184 + }, + { + "epoch": 2.875406787540679, + "grad_norm": 5.204057216644287, + "learning_rate": 9.9938189993819e-06, + "loss": 0.553, + "step": 6185 + }, + { + "epoch": 2.8758716875871686, + "grad_norm": 4.260878562927246, + "learning_rate": 9.993817999381801e-06, + "loss": 0.3699, + "step": 6186 + }, + { + "epoch": 2.876336587633659, + "grad_norm": 4.610028266906738, + "learning_rate": 9.9938169993817e-06, + "loss": 0.3731, + "step": 6187 + }, + { + "epoch": 2.8768014876801487, + "grad_norm": 5.917640686035156, + "learning_rate": 9.993815999381601e-06, + "loss": 0.4383, + "step": 6188 + }, + { + "epoch": 2.877266387726639, + "grad_norm": 4.0324859619140625, + "learning_rate": 9.9938149993815e-06, + "loss": 0.2493, + "step": 6189 + }, + { + "epoch": 2.8777312877731287, + "grad_norm": 6.633136749267578, + "learning_rate": 9.9938139993814e-06, + "loss": 0.5619, + "step": 6190 + }, + { + "epoch": 2.878196187819619, + "grad_norm": 4.907777786254883, + "learning_rate": 9.9938129993813e-06, + "loss": 0.3186, + "step": 6191 + }, + { + "epoch": 2.8786610878661087, + "grad_norm": 7.362251281738281, + "learning_rate": 9.9938119993812e-06, + "loss": 0.5316, + "step": 6192 + }, + { + "epoch": 2.879125987912599, + "grad_norm": 5.1888251304626465, + "learning_rate": 9.993810999381101e-06, + "loss": 0.4681, + "step": 6193 + }, + { + "epoch": 2.879590887959089, + "grad_norm": 7.658495903015137, + "learning_rate": 9.993809999381002e-06, + "loss": 0.415, + "step": 6194 + }, + { + "epoch": 2.8800557880055786, + "grad_norm": 5.814704895019531, + "learning_rate": 9.9938089993809e-06, + "loss": 0.5455, + "step": 6195 + }, + { + "epoch": 2.880520688052069, + "grad_norm": 5.383439064025879, + "learning_rate": 9.993807999380801e-06, + "loss": 0.4784, + "step": 6196 + }, + { + "epoch": 2.8809855880985586, + "grad_norm": 5.886495590209961, + "learning_rate": 9.9938069993807e-06, + "loss": 0.4194, + "step": 6197 + }, + { + "epoch": 2.881450488145049, + "grad_norm": 5.0261359214782715, + "learning_rate": 9.993805999380601e-06, + "loss": 0.501, + "step": 6198 + }, + { + "epoch": 2.8819153881915387, + "grad_norm": 6.997060775756836, + "learning_rate": 9.993804999380502e-06, + "loss": 0.8229, + "step": 6199 + }, + { + "epoch": 2.882380288238029, + "grad_norm": 6.890752792358398, + "learning_rate": 9.9938039993804e-06, + "loss": 0.4908, + "step": 6200 + }, + { + "epoch": 2.8828451882845187, + "grad_norm": 3.5259671211242676, + "learning_rate": 9.993802999380301e-06, + "loss": 0.3406, + "step": 6201 + }, + { + "epoch": 2.883310088331009, + "grad_norm": 9.670546531677246, + "learning_rate": 9.9938019993802e-06, + "loss": 0.6785, + "step": 6202 + }, + { + "epoch": 2.883774988377499, + "grad_norm": 10.244342803955078, + "learning_rate": 9.9938009993801e-06, + "loss": 0.4148, + "step": 6203 + }, + { + "epoch": 2.8842398884239886, + "grad_norm": 6.130430221557617, + "learning_rate": 9.99379999938e-06, + "loss": 0.4426, + "step": 6204 + }, + { + "epoch": 2.884704788470479, + "grad_norm": 5.974501609802246, + "learning_rate": 9.9937989993799e-06, + "loss": 0.487, + "step": 6205 + }, + { + "epoch": 2.885169688516969, + "grad_norm": 4.617537975311279, + "learning_rate": 9.993797999379801e-06, + "loss": 0.3601, + "step": 6206 + }, + { + "epoch": 2.885634588563459, + "grad_norm": 4.801614284515381, + "learning_rate": 9.993796999379702e-06, + "loss": 0.4179, + "step": 6207 + }, + { + "epoch": 2.8860994886099487, + "grad_norm": 7.480029582977295, + "learning_rate": 9.9937959993796e-06, + "loss": 0.4011, + "step": 6208 + }, + { + "epoch": 2.886564388656439, + "grad_norm": 7.055490493774414, + "learning_rate": 9.9937949993795e-06, + "loss": 0.4948, + "step": 6209 + }, + { + "epoch": 2.8870292887029287, + "grad_norm": 4.633966445922852, + "learning_rate": 9.9937939993794e-06, + "loss": 0.3683, + "step": 6210 + }, + { + "epoch": 2.887494188749419, + "grad_norm": 2.6123273372650146, + "learning_rate": 9.993792999379301e-06, + "loss": 0.2319, + "step": 6211 + }, + { + "epoch": 2.887959088795909, + "grad_norm": 21.82329559326172, + "learning_rate": 9.993791999379202e-06, + "loss": 0.4834, + "step": 6212 + }, + { + "epoch": 2.888423988842399, + "grad_norm": 4.3406081199646, + "learning_rate": 9.9937909993791e-06, + "loss": 0.4199, + "step": 6213 + }, + { + "epoch": 2.888888888888889, + "grad_norm": 6.840439796447754, + "learning_rate": 9.993789999379001e-06, + "loss": 0.75, + "step": 6214 + }, + { + "epoch": 2.889353788935379, + "grad_norm": 5.442006587982178, + "learning_rate": 9.9937889993789e-06, + "loss": 0.3896, + "step": 6215 + }, + { + "epoch": 2.889818688981869, + "grad_norm": 7.860734939575195, + "learning_rate": 9.993787999378801e-06, + "loss": 0.5242, + "step": 6216 + }, + { + "epoch": 2.8902835890283587, + "grad_norm": 5.097373008728027, + "learning_rate": 9.9937869993787e-06, + "loss": 0.373, + "step": 6217 + }, + { + "epoch": 2.890748489074849, + "grad_norm": 4.201440811157227, + "learning_rate": 9.9937859993786e-06, + "loss": 0.2902, + "step": 6218 + }, + { + "epoch": 2.891213389121339, + "grad_norm": 6.553044319152832, + "learning_rate": 9.993784999378501e-06, + "loss": 0.3813, + "step": 6219 + }, + { + "epoch": 2.891678289167829, + "grad_norm": 5.226564884185791, + "learning_rate": 9.9937839993784e-06, + "loss": 0.2397, + "step": 6220 + }, + { + "epoch": 2.892143189214319, + "grad_norm": 5.1151299476623535, + "learning_rate": 9.9937829993783e-06, + "loss": 0.3158, + "step": 6221 + }, + { + "epoch": 2.892608089260809, + "grad_norm": 8.048579216003418, + "learning_rate": 9.9937819993782e-06, + "loss": 0.804, + "step": 6222 + }, + { + "epoch": 2.893072989307299, + "grad_norm": 4.716759204864502, + "learning_rate": 9.9937809993781e-06, + "loss": 0.2785, + "step": 6223 + }, + { + "epoch": 2.893537889353789, + "grad_norm": 7.319987773895264, + "learning_rate": 9.993779999378001e-06, + "loss": 0.8659, + "step": 6224 + }, + { + "epoch": 2.894002789400279, + "grad_norm": 5.379020690917969, + "learning_rate": 9.993778999377902e-06, + "loss": 0.4723, + "step": 6225 + }, + { + "epoch": 2.8944676894467687, + "grad_norm": 5.425897121429443, + "learning_rate": 9.9937779993778e-06, + "loss": 0.5024, + "step": 6226 + }, + { + "epoch": 2.894932589493259, + "grad_norm": 3.905726432800293, + "learning_rate": 9.993776999377701e-06, + "loss": 0.2731, + "step": 6227 + }, + { + "epoch": 2.895397489539749, + "grad_norm": 4.8174943923950195, + "learning_rate": 9.9937759993776e-06, + "loss": 0.3208, + "step": 6228 + }, + { + "epoch": 2.895862389586239, + "grad_norm": 4.689845085144043, + "learning_rate": 9.993774999377501e-06, + "loss": 0.4517, + "step": 6229 + }, + { + "epoch": 2.896327289632729, + "grad_norm": 6.571304798126221, + "learning_rate": 9.993773999377402e-06, + "loss": 0.513, + "step": 6230 + }, + { + "epoch": 2.896792189679219, + "grad_norm": 4.995659828186035, + "learning_rate": 9.9937729993773e-06, + "loss": 0.4178, + "step": 6231 + }, + { + "epoch": 2.897257089725709, + "grad_norm": 7.4959001541137695, + "learning_rate": 9.993771999377201e-06, + "loss": 0.599, + "step": 6232 + }, + { + "epoch": 2.897721989772199, + "grad_norm": 8.00810718536377, + "learning_rate": 9.9937709993771e-06, + "loss": 0.7828, + "step": 6233 + }, + { + "epoch": 2.898186889818689, + "grad_norm": 8.692723274230957, + "learning_rate": 9.993769999377001e-06, + "loss": 0.3096, + "step": 6234 + }, + { + "epoch": 2.898651789865179, + "grad_norm": 4.486476421356201, + "learning_rate": 9.9937689993769e-06, + "loss": 0.2673, + "step": 6235 + }, + { + "epoch": 2.899116689911669, + "grad_norm": 5.915162086486816, + "learning_rate": 9.9937679993768e-06, + "loss": 0.4953, + "step": 6236 + }, + { + "epoch": 2.899581589958159, + "grad_norm": 3.259812593460083, + "learning_rate": 9.993766999376701e-06, + "loss": 0.2483, + "step": 6237 + }, + { + "epoch": 2.900046490004649, + "grad_norm": 5.900418758392334, + "learning_rate": 9.993765999376602e-06, + "loss": 0.6259, + "step": 6238 + }, + { + "epoch": 2.900511390051139, + "grad_norm": 4.833314418792725, + "learning_rate": 9.9937649993765e-06, + "loss": 0.5471, + "step": 6239 + }, + { + "epoch": 2.900976290097629, + "grad_norm": 5.062804222106934, + "learning_rate": 9.993763999376401e-06, + "loss": 0.4535, + "step": 6240 + }, + { + "epoch": 2.9014411901441193, + "grad_norm": 5.650574207305908, + "learning_rate": 9.9937629993763e-06, + "loss": 0.4646, + "step": 6241 + }, + { + "epoch": 2.901906090190609, + "grad_norm": 4.566768646240234, + "learning_rate": 9.993761999376201e-06, + "loss": 0.4617, + "step": 6242 + }, + { + "epoch": 2.902370990237099, + "grad_norm": 6.037600517272949, + "learning_rate": 9.993760999376102e-06, + "loss": 0.6673, + "step": 6243 + }, + { + "epoch": 2.902835890283589, + "grad_norm": 8.577310562133789, + "learning_rate": 9.993759999376e-06, + "loss": 0.575, + "step": 6244 + }, + { + "epoch": 2.903300790330079, + "grad_norm": 5.753050327301025, + "learning_rate": 9.993758999375901e-06, + "loss": 0.5025, + "step": 6245 + }, + { + "epoch": 2.903765690376569, + "grad_norm": 5.762115001678467, + "learning_rate": 9.9937579993758e-06, + "loss": 0.3506, + "step": 6246 + }, + { + "epoch": 2.904230590423059, + "grad_norm": 5.178817272186279, + "learning_rate": 9.993756999375701e-06, + "loss": 0.2885, + "step": 6247 + }, + { + "epoch": 2.904695490469549, + "grad_norm": 4.636765956878662, + "learning_rate": 9.9937559993756e-06, + "loss": 0.2824, + "step": 6248 + }, + { + "epoch": 2.905160390516039, + "grad_norm": 5.011076927185059, + "learning_rate": 9.9937549993755e-06, + "loss": 0.6575, + "step": 6249 + }, + { + "epoch": 2.9056252905625293, + "grad_norm": 5.25462007522583, + "learning_rate": 9.993753999375401e-06, + "loss": 0.3854, + "step": 6250 + }, + { + "epoch": 2.906090190609019, + "grad_norm": 9.471476554870605, + "learning_rate": 9.993752999375302e-06, + "loss": 0.5866, + "step": 6251 + }, + { + "epoch": 2.906555090655509, + "grad_norm": 4.650920391082764, + "learning_rate": 9.9937519993752e-06, + "loss": 0.2978, + "step": 6252 + }, + { + "epoch": 2.907019990701999, + "grad_norm": 6.536296367645264, + "learning_rate": 9.9937509993751e-06, + "loss": 0.2221, + "step": 6253 + }, + { + "epoch": 2.907484890748489, + "grad_norm": 4.754899978637695, + "learning_rate": 9.993749999375e-06, + "loss": 0.2723, + "step": 6254 + }, + { + "epoch": 2.907949790794979, + "grad_norm": 4.746220588684082, + "learning_rate": 9.993748999374901e-06, + "loss": 0.3219, + "step": 6255 + }, + { + "epoch": 2.908414690841469, + "grad_norm": 5.541076183319092, + "learning_rate": 9.993747999374802e-06, + "loss": 0.2061, + "step": 6256 + }, + { + "epoch": 2.908879590887959, + "grad_norm": 5.814949035644531, + "learning_rate": 9.9937469993747e-06, + "loss": 0.5086, + "step": 6257 + }, + { + "epoch": 2.909344490934449, + "grad_norm": 4.221242427825928, + "learning_rate": 9.993745999374601e-06, + "loss": 0.2873, + "step": 6258 + }, + { + "epoch": 2.9098093909809393, + "grad_norm": 5.033374309539795, + "learning_rate": 9.9937449993745e-06, + "loss": 0.1515, + "step": 6259 + }, + { + "epoch": 2.910274291027429, + "grad_norm": 6.605466842651367, + "learning_rate": 9.993743999374401e-06, + "loss": 0.3852, + "step": 6260 + }, + { + "epoch": 2.910739191073919, + "grad_norm": 6.428306579589844, + "learning_rate": 9.9937429993743e-06, + "loss": 0.4473, + "step": 6261 + }, + { + "epoch": 2.911204091120409, + "grad_norm": 4.3988728523254395, + "learning_rate": 9.9937419993742e-06, + "loss": 0.1632, + "step": 6262 + }, + { + "epoch": 2.9116689911668994, + "grad_norm": 11.262743949890137, + "learning_rate": 9.993740999374101e-06, + "loss": 0.8517, + "step": 6263 + }, + { + "epoch": 2.912133891213389, + "grad_norm": 7.0118303298950195, + "learning_rate": 9.993739999374e-06, + "loss": 0.332, + "step": 6264 + }, + { + "epoch": 2.912598791259879, + "grad_norm": 3.3444230556488037, + "learning_rate": 9.993738999373901e-06, + "loss": 0.1126, + "step": 6265 + }, + { + "epoch": 2.913063691306369, + "grad_norm": 7.391544342041016, + "learning_rate": 9.9937379993738e-06, + "loss": 0.4589, + "step": 6266 + }, + { + "epoch": 2.913528591352859, + "grad_norm": 9.315759658813477, + "learning_rate": 9.9937369993737e-06, + "loss": 1.2822, + "step": 6267 + }, + { + "epoch": 2.9139934913993493, + "grad_norm": 6.062946319580078, + "learning_rate": 9.993735999373601e-06, + "loss": 0.2982, + "step": 6268 + }, + { + "epoch": 2.914458391445839, + "grad_norm": 4.372904300689697, + "learning_rate": 9.993734999373502e-06, + "loss": 0.2468, + "step": 6269 + }, + { + "epoch": 2.914923291492329, + "grad_norm": 6.446206092834473, + "learning_rate": 9.9937339993734e-06, + "loss": 0.3612, + "step": 6270 + }, + { + "epoch": 2.915388191538819, + "grad_norm": 3.7636117935180664, + "learning_rate": 9.993732999373301e-06, + "loss": 0.2669, + "step": 6271 + }, + { + "epoch": 2.9158530915853094, + "grad_norm": 6.037347793579102, + "learning_rate": 9.9937319993732e-06, + "loss": 0.4051, + "step": 6272 + }, + { + "epoch": 2.916317991631799, + "grad_norm": 5.460978031158447, + "learning_rate": 9.9937309993731e-06, + "loss": 0.4008, + "step": 6273 + }, + { + "epoch": 2.916782891678289, + "grad_norm": 8.624279022216797, + "learning_rate": 9.993729999373e-06, + "loss": 0.6807, + "step": 6274 + }, + { + "epoch": 2.917247791724779, + "grad_norm": 5.523046493530273, + "learning_rate": 9.9937289993729e-06, + "loss": 0.3261, + "step": 6275 + }, + { + "epoch": 2.917712691771269, + "grad_norm": 4.604807376861572, + "learning_rate": 9.993727999372801e-06, + "loss": 0.2745, + "step": 6276 + }, + { + "epoch": 2.9181775918177593, + "grad_norm": 5.036529541015625, + "learning_rate": 9.9937269993727e-06, + "loss": 0.4729, + "step": 6277 + }, + { + "epoch": 2.918642491864249, + "grad_norm": 11.035210609436035, + "learning_rate": 9.993725999372601e-06, + "loss": 0.9846, + "step": 6278 + }, + { + "epoch": 2.9191073919107393, + "grad_norm": 3.5786564350128174, + "learning_rate": 9.9937249993725e-06, + "loss": 0.291, + "step": 6279 + }, + { + "epoch": 2.919572291957229, + "grad_norm": 3.992973566055298, + "learning_rate": 9.9937239993724e-06, + "loss": 0.4108, + "step": 6280 + }, + { + "epoch": 2.9200371920037194, + "grad_norm": 5.42030668258667, + "learning_rate": 9.993722999372301e-06, + "loss": 0.5442, + "step": 6281 + }, + { + "epoch": 2.920502092050209, + "grad_norm": 7.280388832092285, + "learning_rate": 9.993721999372202e-06, + "loss": 0.4472, + "step": 6282 + }, + { + "epoch": 2.920966992096699, + "grad_norm": 3.246826648712158, + "learning_rate": 9.993720999372101e-06, + "loss": 0.2417, + "step": 6283 + }, + { + "epoch": 2.921431892143189, + "grad_norm": 3.8075461387634277, + "learning_rate": 9.993719999372e-06, + "loss": 0.326, + "step": 6284 + }, + { + "epoch": 2.9218967921896795, + "grad_norm": 16.185461044311523, + "learning_rate": 9.9937189993719e-06, + "loss": 2.2014, + "step": 6285 + }, + { + "epoch": 2.9223616922361693, + "grad_norm": 7.735650062561035, + "learning_rate": 9.9937179993718e-06, + "loss": 0.8493, + "step": 6286 + }, + { + "epoch": 2.922826592282659, + "grad_norm": 8.290908813476562, + "learning_rate": 9.9937169993717e-06, + "loss": 0.5617, + "step": 6287 + }, + { + "epoch": 2.9232914923291493, + "grad_norm": 6.963515758514404, + "learning_rate": 9.9937159993716e-06, + "loss": 0.4529, + "step": 6288 + }, + { + "epoch": 2.923756392375639, + "grad_norm": 3.7263643741607666, + "learning_rate": 9.993714999371501e-06, + "loss": 0.2473, + "step": 6289 + }, + { + "epoch": 2.9242212924221294, + "grad_norm": 5.059169292449951, + "learning_rate": 9.9937139993714e-06, + "loss": 0.452, + "step": 6290 + }, + { + "epoch": 2.924686192468619, + "grad_norm": 4.286563873291016, + "learning_rate": 9.993712999371301e-06, + "loss": 0.2524, + "step": 6291 + }, + { + "epoch": 2.9251510925151094, + "grad_norm": 5.0419921875, + "learning_rate": 9.9937119993712e-06, + "loss": 0.2977, + "step": 6292 + }, + { + "epoch": 2.925615992561599, + "grad_norm": 6.61260986328125, + "learning_rate": 9.9937109993711e-06, + "loss": 0.7671, + "step": 6293 + }, + { + "epoch": 2.9260808926080895, + "grad_norm": 8.497840881347656, + "learning_rate": 9.993709999371001e-06, + "loss": 0.6107, + "step": 6294 + }, + { + "epoch": 2.9265457926545793, + "grad_norm": 5.990030765533447, + "learning_rate": 9.9937089993709e-06, + "loss": 0.6529, + "step": 6295 + }, + { + "epoch": 2.927010692701069, + "grad_norm": 4.966625213623047, + "learning_rate": 9.993707999370801e-06, + "loss": 0.2275, + "step": 6296 + }, + { + "epoch": 2.9274755927475593, + "grad_norm": 5.415621757507324, + "learning_rate": 9.9937069993707e-06, + "loss": 0.2993, + "step": 6297 + }, + { + "epoch": 2.9279404927940496, + "grad_norm": 4.642990589141846, + "learning_rate": 9.9937059993706e-06, + "loss": 0.3381, + "step": 6298 + }, + { + "epoch": 2.9284053928405394, + "grad_norm": 3.4224727153778076, + "learning_rate": 9.993704999370501e-06, + "loss": 0.2215, + "step": 6299 + }, + { + "epoch": 2.928870292887029, + "grad_norm": 5.477099895477295, + "learning_rate": 9.9937039993704e-06, + "loss": 0.392, + "step": 6300 + }, + { + "epoch": 2.9293351929335194, + "grad_norm": 3.4411051273345947, + "learning_rate": 9.993702999370301e-06, + "loss": 0.1905, + "step": 6301 + }, + { + "epoch": 2.929800092980009, + "grad_norm": 8.522111892700195, + "learning_rate": 9.993701999370202e-06, + "loss": 0.4134, + "step": 6302 + }, + { + "epoch": 2.9302649930264995, + "grad_norm": 9.19692611694336, + "learning_rate": 9.9937009993701e-06, + "loss": 0.6562, + "step": 6303 + }, + { + "epoch": 2.9307298930729893, + "grad_norm": 6.336058139801025, + "learning_rate": 9.993699999370001e-06, + "loss": 0.7693, + "step": 6304 + }, + { + "epoch": 2.931194793119479, + "grad_norm": 5.75673246383667, + "learning_rate": 9.9936989993699e-06, + "loss": 0.6634, + "step": 6305 + }, + { + "epoch": 2.9316596931659693, + "grad_norm": 4.580772876739502, + "learning_rate": 9.9936979993698e-06, + "loss": 0.3178, + "step": 6306 + }, + { + "epoch": 2.9321245932124596, + "grad_norm": 6.39060640335083, + "learning_rate": 9.993696999369701e-06, + "loss": 0.7027, + "step": 6307 + }, + { + "epoch": 2.9325894932589494, + "grad_norm": 6.023180961608887, + "learning_rate": 9.9936959993696e-06, + "loss": 0.5172, + "step": 6308 + }, + { + "epoch": 2.933054393305439, + "grad_norm": 5.477637767791748, + "learning_rate": 9.993694999369501e-06, + "loss": 0.4321, + "step": 6309 + }, + { + "epoch": 2.9335192933519294, + "grad_norm": 6.0965681076049805, + "learning_rate": 9.9936939993694e-06, + "loss": 0.356, + "step": 6310 + }, + { + "epoch": 2.933984193398419, + "grad_norm": 5.163066387176514, + "learning_rate": 9.9936929993693e-06, + "loss": 0.4896, + "step": 6311 + }, + { + "epoch": 2.9344490934449095, + "grad_norm": 3.35683274269104, + "learning_rate": 9.993691999369201e-06, + "loss": 0.1669, + "step": 6312 + }, + { + "epoch": 2.9349139934913993, + "grad_norm": 6.842613697052002, + "learning_rate": 9.9936909993691e-06, + "loss": 0.3993, + "step": 6313 + }, + { + "epoch": 2.9353788935378895, + "grad_norm": 6.222872734069824, + "learning_rate": 9.993689999369001e-06, + "loss": 0.3061, + "step": 6314 + }, + { + "epoch": 2.9358437935843793, + "grad_norm": 6.839595317840576, + "learning_rate": 9.993688999368902e-06, + "loss": 0.7929, + "step": 6315 + }, + { + "epoch": 2.9363086936308695, + "grad_norm": 6.307823181152344, + "learning_rate": 9.9936879993688e-06, + "loss": 0.4528, + "step": 6316 + }, + { + "epoch": 2.9367735936773594, + "grad_norm": 4.735023021697998, + "learning_rate": 9.9936869993687e-06, + "loss": 0.3406, + "step": 6317 + }, + { + "epoch": 2.937238493723849, + "grad_norm": 10.579448699951172, + "learning_rate": 9.9936859993686e-06, + "loss": 0.2498, + "step": 6318 + }, + { + "epoch": 2.9377033937703394, + "grad_norm": 4.33495569229126, + "learning_rate": 9.993684999368501e-06, + "loss": 0.3385, + "step": 6319 + }, + { + "epoch": 2.9381682938168296, + "grad_norm": 5.321931838989258, + "learning_rate": 9.993683999368402e-06, + "loss": 0.469, + "step": 6320 + }, + { + "epoch": 2.9386331938633194, + "grad_norm": 5.194030284881592, + "learning_rate": 9.9936829993683e-06, + "loss": 0.4979, + "step": 6321 + }, + { + "epoch": 2.9390980939098093, + "grad_norm": 5.713241100311279, + "learning_rate": 9.993681999368201e-06, + "loss": 0.3776, + "step": 6322 + }, + { + "epoch": 2.9395629939562995, + "grad_norm": 5.890443801879883, + "learning_rate": 9.9936809993681e-06, + "loss": 0.5075, + "step": 6323 + }, + { + "epoch": 2.9400278940027893, + "grad_norm": 6.436061859130859, + "learning_rate": 9.993679999368e-06, + "loss": 0.5821, + "step": 6324 + }, + { + "epoch": 2.9404927940492795, + "grad_norm": 4.03553581237793, + "learning_rate": 9.993678999367901e-06, + "loss": 0.2763, + "step": 6325 + }, + { + "epoch": 2.9409576940957693, + "grad_norm": 5.986965656280518, + "learning_rate": 9.993677999367802e-06, + "loss": 0.3855, + "step": 6326 + }, + { + "epoch": 2.941422594142259, + "grad_norm": 10.530686378479004, + "learning_rate": 9.993676999367701e-06, + "loss": 0.6336, + "step": 6327 + }, + { + "epoch": 2.9418874941887494, + "grad_norm": 6.4302592277526855, + "learning_rate": 9.9936759993676e-06, + "loss": 0.4538, + "step": 6328 + }, + { + "epoch": 2.9423523942352396, + "grad_norm": 4.583762168884277, + "learning_rate": 9.9936749993675e-06, + "loss": 0.4751, + "step": 6329 + }, + { + "epoch": 2.9428172942817294, + "grad_norm": 5.607551097869873, + "learning_rate": 9.9936739993674e-06, + "loss": 0.2547, + "step": 6330 + }, + { + "epoch": 2.9432821943282192, + "grad_norm": 5.3372802734375, + "learning_rate": 9.9936729993673e-06, + "loss": 0.5655, + "step": 6331 + }, + { + "epoch": 2.9437470943747095, + "grad_norm": 6.1826958656311035, + "learning_rate": 9.993671999367201e-06, + "loss": 0.2372, + "step": 6332 + }, + { + "epoch": 2.9442119944211993, + "grad_norm": 6.862386226654053, + "learning_rate": 9.993670999367102e-06, + "loss": 0.489, + "step": 6333 + }, + { + "epoch": 2.9446768944676895, + "grad_norm": 4.778977394104004, + "learning_rate": 9.993669999367e-06, + "loss": 0.4496, + "step": 6334 + }, + { + "epoch": 2.9451417945141793, + "grad_norm": 5.3114776611328125, + "learning_rate": 9.993668999366901e-06, + "loss": 0.3333, + "step": 6335 + }, + { + "epoch": 2.9456066945606696, + "grad_norm": 5.500539302825928, + "learning_rate": 9.9936679993668e-06, + "loss": 0.5209, + "step": 6336 + }, + { + "epoch": 2.9460715946071594, + "grad_norm": 5.044684886932373, + "learning_rate": 9.993666999366701e-06, + "loss": 0.4523, + "step": 6337 + }, + { + "epoch": 2.9465364946536496, + "grad_norm": 6.08042573928833, + "learning_rate": 9.993665999366602e-06, + "loss": 0.3579, + "step": 6338 + }, + { + "epoch": 2.9470013947001394, + "grad_norm": 5.5607147216796875, + "learning_rate": 9.9936649993665e-06, + "loss": 0.6017, + "step": 6339 + }, + { + "epoch": 2.9474662947466292, + "grad_norm": 6.169000148773193, + "learning_rate": 9.993663999366401e-06, + "loss": 0.5961, + "step": 6340 + }, + { + "epoch": 2.9479311947931195, + "grad_norm": 4.737051963806152, + "learning_rate": 9.9936629993663e-06, + "loss": 0.4516, + "step": 6341 + }, + { + "epoch": 2.9483960948396097, + "grad_norm": 5.43651008605957, + "learning_rate": 9.9936619993662e-06, + "loss": 0.5111, + "step": 6342 + }, + { + "epoch": 2.9488609948860995, + "grad_norm": 19.69056510925293, + "learning_rate": 9.9936609993661e-06, + "loss": 0.7426, + "step": 6343 + }, + { + "epoch": 2.9493258949325893, + "grad_norm": 3.6386330127716064, + "learning_rate": 9.993659999366e-06, + "loss": 0.1618, + "step": 6344 + }, + { + "epoch": 2.9497907949790796, + "grad_norm": 4.346364498138428, + "learning_rate": 9.993658999365901e-06, + "loss": 0.3065, + "step": 6345 + }, + { + "epoch": 2.9502556950255694, + "grad_norm": 5.39323091506958, + "learning_rate": 9.993657999365802e-06, + "loss": 0.459, + "step": 6346 + }, + { + "epoch": 2.9507205950720596, + "grad_norm": 7.866148948669434, + "learning_rate": 9.9936569993657e-06, + "loss": 0.5969, + "step": 6347 + }, + { + "epoch": 2.9511854951185494, + "grad_norm": 4.92398738861084, + "learning_rate": 9.993655999365601e-06, + "loss": 0.4491, + "step": 6348 + }, + { + "epoch": 2.9516503951650392, + "grad_norm": 6.721884250640869, + "learning_rate": 9.9936549993655e-06, + "loss": 0.5072, + "step": 6349 + }, + { + "epoch": 2.9521152952115295, + "grad_norm": 4.6786370277404785, + "learning_rate": 9.993653999365401e-06, + "loss": 0.4548, + "step": 6350 + }, + { + "epoch": 2.9525801952580197, + "grad_norm": 10.095656394958496, + "learning_rate": 9.993652999365302e-06, + "loss": 0.3018, + "step": 6351 + }, + { + "epoch": 2.9530450953045095, + "grad_norm": 5.440873622894287, + "learning_rate": 9.9936519993652e-06, + "loss": 0.6133, + "step": 6352 + }, + { + "epoch": 2.9535099953509993, + "grad_norm": 6.775108337402344, + "learning_rate": 9.993650999365101e-06, + "loss": 0.5417, + "step": 6353 + }, + { + "epoch": 2.9539748953974896, + "grad_norm": 5.068181991577148, + "learning_rate": 9.993649999365e-06, + "loss": 0.3826, + "step": 6354 + }, + { + "epoch": 2.9544397954439794, + "grad_norm": 7.391268730163574, + "learning_rate": 9.9936489993649e-06, + "loss": 0.3577, + "step": 6355 + }, + { + "epoch": 2.9549046954904696, + "grad_norm": 11.523311614990234, + "learning_rate": 9.9936479993648e-06, + "loss": 0.5059, + "step": 6356 + }, + { + "epoch": 2.9553695955369594, + "grad_norm": 5.644241809844971, + "learning_rate": 9.9936469993647e-06, + "loss": 0.5391, + "step": 6357 + }, + { + "epoch": 2.9558344955834497, + "grad_norm": 9.088332176208496, + "learning_rate": 9.993645999364601e-06, + "loss": 0.679, + "step": 6358 + }, + { + "epoch": 2.9562993956299395, + "grad_norm": 5.700749397277832, + "learning_rate": 9.993644999364502e-06, + "loss": 0.4444, + "step": 6359 + }, + { + "epoch": 2.9567642956764297, + "grad_norm": 6.957569122314453, + "learning_rate": 9.9936439993644e-06, + "loss": 0.5927, + "step": 6360 + }, + { + "epoch": 2.9572291957229195, + "grad_norm": 5.845780372619629, + "learning_rate": 9.9936429993643e-06, + "loss": 0.4381, + "step": 6361 + }, + { + "epoch": 2.9576940957694093, + "grad_norm": 6.732683181762695, + "learning_rate": 9.9936419993642e-06, + "loss": 0.5196, + "step": 6362 + }, + { + "epoch": 2.9581589958158996, + "grad_norm": 3.760100841522217, + "learning_rate": 9.993640999364101e-06, + "loss": 0.2447, + "step": 6363 + }, + { + "epoch": 2.95862389586239, + "grad_norm": 3.183006763458252, + "learning_rate": 9.993639999364002e-06, + "loss": 0.1274, + "step": 6364 + }, + { + "epoch": 2.9590887959088796, + "grad_norm": 6.8279924392700195, + "learning_rate": 9.9936389993639e-06, + "loss": 0.6203, + "step": 6365 + }, + { + "epoch": 2.9595536959553694, + "grad_norm": 7.146036624908447, + "learning_rate": 9.993637999363801e-06, + "loss": 0.7108, + "step": 6366 + }, + { + "epoch": 2.9600185960018597, + "grad_norm": 5.627175331115723, + "learning_rate": 9.9936369993637e-06, + "loss": 0.3804, + "step": 6367 + }, + { + "epoch": 2.9604834960483495, + "grad_norm": 3.9589898586273193, + "learning_rate": 9.993635999363601e-06, + "loss": 0.2118, + "step": 6368 + }, + { + "epoch": 2.9609483960948397, + "grad_norm": 5.699357509613037, + "learning_rate": 9.9936349993635e-06, + "loss": 0.4378, + "step": 6369 + }, + { + "epoch": 2.9614132961413295, + "grad_norm": 6.594386100769043, + "learning_rate": 9.9936339993634e-06, + "loss": 0.7882, + "step": 6370 + }, + { + "epoch": 2.96187819618782, + "grad_norm": 5.370503902435303, + "learning_rate": 9.993632999363301e-06, + "loss": 0.3561, + "step": 6371 + }, + { + "epoch": 2.9623430962343096, + "grad_norm": 5.345922946929932, + "learning_rate": 9.9936319993632e-06, + "loss": 0.5612, + "step": 6372 + }, + { + "epoch": 2.9628079962808, + "grad_norm": 3.759234666824341, + "learning_rate": 9.9936309993631e-06, + "loss": 0.18, + "step": 6373 + }, + { + "epoch": 2.9632728963272896, + "grad_norm": 5.692526817321777, + "learning_rate": 9.993629999363e-06, + "loss": 0.3536, + "step": 6374 + }, + { + "epoch": 2.9637377963737794, + "grad_norm": 4.124083518981934, + "learning_rate": 9.9936289993629e-06, + "loss": 0.2933, + "step": 6375 + }, + { + "epoch": 2.9642026964202697, + "grad_norm": 4.817544937133789, + "learning_rate": 9.993627999362801e-06, + "loss": 0.4054, + "step": 6376 + }, + { + "epoch": 2.9646675964667595, + "grad_norm": 6.681967258453369, + "learning_rate": 9.993626999362702e-06, + "loss": 0.4243, + "step": 6377 + }, + { + "epoch": 2.9651324965132497, + "grad_norm": 8.201699256896973, + "learning_rate": 9.9936259993626e-06, + "loss": 0.7759, + "step": 6378 + }, + { + "epoch": 2.9655973965597395, + "grad_norm": 6.866189002990723, + "learning_rate": 9.993624999362501e-06, + "loss": 0.4239, + "step": 6379 + }, + { + "epoch": 2.9660622966062298, + "grad_norm": 8.285715103149414, + "learning_rate": 9.9936239993624e-06, + "loss": 0.7813, + "step": 6380 + }, + { + "epoch": 2.9665271966527196, + "grad_norm": 4.245107650756836, + "learning_rate": 9.993622999362301e-06, + "loss": 0.2672, + "step": 6381 + }, + { + "epoch": 2.96699209669921, + "grad_norm": 11.358372688293457, + "learning_rate": 9.9936219993622e-06, + "loss": 0.6847, + "step": 6382 + }, + { + "epoch": 2.9674569967456996, + "grad_norm": 7.030549049377441, + "learning_rate": 9.9936209993621e-06, + "loss": 1.1109, + "step": 6383 + }, + { + "epoch": 2.9679218967921894, + "grad_norm": 5.79581880569458, + "learning_rate": 9.993619999362001e-06, + "loss": 0.5852, + "step": 6384 + }, + { + "epoch": 2.9683867968386797, + "grad_norm": 7.493926525115967, + "learning_rate": 9.9936189993619e-06, + "loss": 0.9665, + "step": 6385 + }, + { + "epoch": 2.96885169688517, + "grad_norm": 4.904453277587891, + "learning_rate": 9.993617999361801e-06, + "loss": 0.4346, + "step": 6386 + }, + { + "epoch": 2.9693165969316597, + "grad_norm": 7.895668983459473, + "learning_rate": 9.9936169993617e-06, + "loss": 0.4066, + "step": 6387 + }, + { + "epoch": 2.9697814969781495, + "grad_norm": 6.922050952911377, + "learning_rate": 9.9936159993616e-06, + "loss": 0.5111, + "step": 6388 + }, + { + "epoch": 2.9702463970246398, + "grad_norm": 5.377651214599609, + "learning_rate": 9.993614999361501e-06, + "loss": 0.3937, + "step": 6389 + }, + { + "epoch": 2.9707112970711296, + "grad_norm": 6.133293151855469, + "learning_rate": 9.993613999361402e-06, + "loss": 0.7081, + "step": 6390 + }, + { + "epoch": 2.97117619711762, + "grad_norm": 7.744842052459717, + "learning_rate": 9.9936129993613e-06, + "loss": 0.4354, + "step": 6391 + }, + { + "epoch": 2.9716410971641096, + "grad_norm": 3.762486457824707, + "learning_rate": 9.993611999361201e-06, + "loss": 0.1565, + "step": 6392 + }, + { + "epoch": 2.9721059972106, + "grad_norm": 3.894503355026245, + "learning_rate": 9.9936109993611e-06, + "loss": 0.2028, + "step": 6393 + }, + { + "epoch": 2.9725708972570897, + "grad_norm": 5.400960922241211, + "learning_rate": 9.993609999361001e-06, + "loss": 0.2686, + "step": 6394 + }, + { + "epoch": 2.97303579730358, + "grad_norm": 6.161404132843018, + "learning_rate": 9.993608999360902e-06, + "loss": 0.3378, + "step": 6395 + }, + { + "epoch": 2.9735006973500697, + "grad_norm": 3.933742046356201, + "learning_rate": 9.9936079993608e-06, + "loss": 0.3696, + "step": 6396 + }, + { + "epoch": 2.9739655973965595, + "grad_norm": 7.0749287605285645, + "learning_rate": 9.993606999360701e-06, + "loss": 0.5568, + "step": 6397 + }, + { + "epoch": 2.9744304974430498, + "grad_norm": 8.317172050476074, + "learning_rate": 9.9936059993606e-06, + "loss": 0.542, + "step": 6398 + }, + { + "epoch": 2.97489539748954, + "grad_norm": 4.583962917327881, + "learning_rate": 9.993604999360501e-06, + "loss": 0.4859, + "step": 6399 + }, + { + "epoch": 2.97536029753603, + "grad_norm": 4.0428619384765625, + "learning_rate": 9.9936039993604e-06, + "loss": 0.2405, + "step": 6400 + }, + { + "epoch": 2.9758251975825196, + "grad_norm": 6.336739540100098, + "learning_rate": 9.9936029993603e-06, + "loss": 0.5182, + "step": 6401 + }, + { + "epoch": 2.97629009762901, + "grad_norm": 3.78452205657959, + "learning_rate": 9.993601999360201e-06, + "loss": 0.2177, + "step": 6402 + }, + { + "epoch": 2.9767549976754997, + "grad_norm": 5.006394386291504, + "learning_rate": 9.993600999360102e-06, + "loss": 0.3263, + "step": 6403 + }, + { + "epoch": 2.97721989772199, + "grad_norm": 6.543952465057373, + "learning_rate": 9.993599999360001e-06, + "loss": 0.5846, + "step": 6404 + }, + { + "epoch": 2.9776847977684797, + "grad_norm": 6.37080717086792, + "learning_rate": 9.9935989993599e-06, + "loss": 0.4715, + "step": 6405 + }, + { + "epoch": 2.9781496978149695, + "grad_norm": 5.760970115661621, + "learning_rate": 9.9935979993598e-06, + "loss": 0.5194, + "step": 6406 + }, + { + "epoch": 2.9786145978614598, + "grad_norm": 4.849841594696045, + "learning_rate": 9.993596999359701e-06, + "loss": 0.4913, + "step": 6407 + }, + { + "epoch": 2.97907949790795, + "grad_norm": 5.99839973449707, + "learning_rate": 9.993595999359602e-06, + "loss": 0.8366, + "step": 6408 + }, + { + "epoch": 2.97954439795444, + "grad_norm": 4.627472400665283, + "learning_rate": 9.9935949993595e-06, + "loss": 0.2541, + "step": 6409 + }, + { + "epoch": 2.9800092980009296, + "grad_norm": 6.2652363777160645, + "learning_rate": 9.993593999359401e-06, + "loss": 0.5966, + "step": 6410 + }, + { + "epoch": 2.98047419804742, + "grad_norm": 4.375805377960205, + "learning_rate": 9.9935929993593e-06, + "loss": 0.2761, + "step": 6411 + }, + { + "epoch": 2.9809390980939097, + "grad_norm": 6.872925281524658, + "learning_rate": 9.993591999359201e-06, + "loss": 0.5736, + "step": 6412 + }, + { + "epoch": 2.9814039981404, + "grad_norm": 6.144716262817383, + "learning_rate": 9.9935909993591e-06, + "loss": 0.4376, + "step": 6413 + }, + { + "epoch": 2.9818688981868897, + "grad_norm": 6.7622575759887695, + "learning_rate": 9.993589999359e-06, + "loss": 0.5421, + "step": 6414 + }, + { + "epoch": 2.98233379823338, + "grad_norm": 4.81020450592041, + "learning_rate": 9.993588999358901e-06, + "loss": 0.3398, + "step": 6415 + }, + { + "epoch": 2.9827986982798698, + "grad_norm": 6.148555755615234, + "learning_rate": 9.9935879993588e-06, + "loss": 0.6929, + "step": 6416 + }, + { + "epoch": 2.98326359832636, + "grad_norm": 6.174932479858398, + "learning_rate": 9.993586999358701e-06, + "loss": 0.5837, + "step": 6417 + }, + { + "epoch": 2.98372849837285, + "grad_norm": 4.854167938232422, + "learning_rate": 9.9935859993586e-06, + "loss": 0.3688, + "step": 6418 + }, + { + "epoch": 2.9841933984193396, + "grad_norm": 4.55521821975708, + "learning_rate": 9.9935849993585e-06, + "loss": 0.3255, + "step": 6419 + }, + { + "epoch": 2.98465829846583, + "grad_norm": 4.616926670074463, + "learning_rate": 9.993583999358401e-06, + "loss": 0.3765, + "step": 6420 + }, + { + "epoch": 2.98512319851232, + "grad_norm": 5.631952285766602, + "learning_rate": 9.993582999358302e-06, + "loss": 0.6119, + "step": 6421 + }, + { + "epoch": 2.98558809855881, + "grad_norm": 5.0046892166137695, + "learning_rate": 9.993581999358201e-06, + "loss": 0.2935, + "step": 6422 + }, + { + "epoch": 2.9860529986052997, + "grad_norm": 5.022066593170166, + "learning_rate": 9.993580999358102e-06, + "loss": 0.3258, + "step": 6423 + }, + { + "epoch": 2.98651789865179, + "grad_norm": 4.926651954650879, + "learning_rate": 9.993579999358e-06, + "loss": 0.3311, + "step": 6424 + }, + { + "epoch": 2.9869827986982798, + "grad_norm": 5.685019493103027, + "learning_rate": 9.9935789993579e-06, + "loss": 0.264, + "step": 6425 + }, + { + "epoch": 2.98744769874477, + "grad_norm": 4.477784633636475, + "learning_rate": 9.9935779993578e-06, + "loss": 0.3387, + "step": 6426 + }, + { + "epoch": 2.98791259879126, + "grad_norm": 6.358789443969727, + "learning_rate": 9.9935769993577e-06, + "loss": 0.468, + "step": 6427 + }, + { + "epoch": 2.9883774988377496, + "grad_norm": 4.237431049346924, + "learning_rate": 9.993575999357601e-06, + "loss": 0.3065, + "step": 6428 + }, + { + "epoch": 2.98884239888424, + "grad_norm": 4.875698089599609, + "learning_rate": 9.9935749993575e-06, + "loss": 0.3381, + "step": 6429 + }, + { + "epoch": 2.98930729893073, + "grad_norm": 4.157297611236572, + "learning_rate": 9.993573999357401e-06, + "loss": 0.279, + "step": 6430 + }, + { + "epoch": 2.98977219897722, + "grad_norm": 5.195772647857666, + "learning_rate": 9.9935729993573e-06, + "loss": 0.3971, + "step": 6431 + }, + { + "epoch": 2.9902370990237097, + "grad_norm": 5.635176181793213, + "learning_rate": 9.9935719993572e-06, + "loss": 0.4576, + "step": 6432 + }, + { + "epoch": 2.9907019990702, + "grad_norm": 7.9704742431640625, + "learning_rate": 9.993570999357101e-06, + "loss": 0.6741, + "step": 6433 + }, + { + "epoch": 2.9911668991166898, + "grad_norm": 17.93187713623047, + "learning_rate": 9.993569999357002e-06, + "loss": 0.2234, + "step": 6434 + }, + { + "epoch": 2.99163179916318, + "grad_norm": 6.806302070617676, + "learning_rate": 9.993568999356901e-06, + "loss": 0.3771, + "step": 6435 + }, + { + "epoch": 2.99209669920967, + "grad_norm": 9.434968948364258, + "learning_rate": 9.9935679993568e-06, + "loss": 0.3047, + "step": 6436 + }, + { + "epoch": 2.99256159925616, + "grad_norm": 5.115718364715576, + "learning_rate": 9.9935669993567e-06, + "loss": 0.3032, + "step": 6437 + }, + { + "epoch": 2.99302649930265, + "grad_norm": 4.358196258544922, + "learning_rate": 9.9935659993566e-06, + "loss": 0.3838, + "step": 6438 + }, + { + "epoch": 2.99349139934914, + "grad_norm": 5.736541271209717, + "learning_rate": 9.9935649993565e-06, + "loss": 0.2936, + "step": 6439 + }, + { + "epoch": 2.99395629939563, + "grad_norm": 5.561332702636719, + "learning_rate": 9.9935639993564e-06, + "loss": 0.5061, + "step": 6440 + }, + { + "epoch": 2.9944211994421197, + "grad_norm": 3.966151714324951, + "learning_rate": 9.993562999356301e-06, + "loss": 0.1792, + "step": 6441 + }, + { + "epoch": 2.99488609948861, + "grad_norm": 4.633851051330566, + "learning_rate": 9.9935619993562e-06, + "loss": 0.5124, + "step": 6442 + }, + { + "epoch": 2.9953509995351, + "grad_norm": 7.040566921234131, + "learning_rate": 9.993560999356101e-06, + "loss": 0.5487, + "step": 6443 + }, + { + "epoch": 2.99581589958159, + "grad_norm": 4.480220794677734, + "learning_rate": 9.993559999356e-06, + "loss": 0.319, + "step": 6444 + }, + { + "epoch": 2.99628079962808, + "grad_norm": 5.428402900695801, + "learning_rate": 9.9935589993559e-06, + "loss": 0.3965, + "step": 6445 + }, + { + "epoch": 2.99674569967457, + "grad_norm": 6.5423712730407715, + "learning_rate": 9.993557999355801e-06, + "loss": 0.654, + "step": 6446 + }, + { + "epoch": 2.99721059972106, + "grad_norm": 6.848749160766602, + "learning_rate": 9.9935569993557e-06, + "loss": 0.6927, + "step": 6447 + }, + { + "epoch": 2.99767549976755, + "grad_norm": 11.296600341796875, + "learning_rate": 9.993555999355601e-06, + "loss": 0.4038, + "step": 6448 + }, + { + "epoch": 2.99814039981404, + "grad_norm": 5.760401248931885, + "learning_rate": 9.9935549993555e-06, + "loss": 0.638, + "step": 6449 + }, + { + "epoch": 2.99860529986053, + "grad_norm": 5.914560794830322, + "learning_rate": 9.9935539993554e-06, + "loss": 0.4042, + "step": 6450 + }, + { + "epoch": 2.99907019990702, + "grad_norm": 5.855527877807617, + "learning_rate": 9.9935529993553e-06, + "loss": 0.4879, + "step": 6451 + }, + { + "epoch": 2.99953509995351, + "grad_norm": 11.870196342468262, + "learning_rate": 9.9935519993552e-06, + "loss": 0.9743, + "step": 6452 + }, + { + "epoch": 3.0, + "grad_norm": 9.029461860656738, + "learning_rate": 9.993550999355101e-06, + "loss": 0.7812, + "step": 6453 + }, + { + "epoch": 3.00046490004649, + "grad_norm": 5.730465888977051, + "learning_rate": 9.993549999355002e-06, + "loss": 0.473, + "step": 6454 + }, + { + "epoch": 3.00092980009298, + "grad_norm": 3.6704909801483154, + "learning_rate": 9.9935489993549e-06, + "loss": 0.1923, + "step": 6455 + }, + { + "epoch": 3.00139470013947, + "grad_norm": 5.631181716918945, + "learning_rate": 9.993547999354801e-06, + "loss": 0.4242, + "step": 6456 + }, + { + "epoch": 3.00185960018596, + "grad_norm": 7.594720363616943, + "learning_rate": 9.9935469993547e-06, + "loss": 0.3924, + "step": 6457 + }, + { + "epoch": 3.00232450023245, + "grad_norm": 5.269809722900391, + "learning_rate": 9.9935459993546e-06, + "loss": 0.2134, + "step": 6458 + }, + { + "epoch": 3.00278940027894, + "grad_norm": 4.438982009887695, + "learning_rate": 9.993544999354501e-06, + "loss": 0.1758, + "step": 6459 + }, + { + "epoch": 3.00325430032543, + "grad_norm": 3.860016345977783, + "learning_rate": 9.9935439993544e-06, + "loss": 0.255, + "step": 6460 + }, + { + "epoch": 3.00371920037192, + "grad_norm": 3.8967888355255127, + "learning_rate": 9.993542999354301e-06, + "loss": 0.2771, + "step": 6461 + }, + { + "epoch": 3.00418410041841, + "grad_norm": 4.123082160949707, + "learning_rate": 9.9935419993542e-06, + "loss": 0.2799, + "step": 6462 + }, + { + "epoch": 3.0046490004649002, + "grad_norm": 6.204929828643799, + "learning_rate": 9.9935409993541e-06, + "loss": 0.3764, + "step": 6463 + }, + { + "epoch": 3.00511390051139, + "grad_norm": 4.146798610687256, + "learning_rate": 9.993539999354001e-06, + "loss": 0.246, + "step": 6464 + }, + { + "epoch": 3.00557880055788, + "grad_norm": 4.54416561126709, + "learning_rate": 9.9935389993539e-06, + "loss": 0.2915, + "step": 6465 + }, + { + "epoch": 3.00604370060437, + "grad_norm": 3.812983512878418, + "learning_rate": 9.993537999353801e-06, + "loss": 0.3539, + "step": 6466 + }, + { + "epoch": 3.00650860065086, + "grad_norm": 3.804270029067993, + "learning_rate": 9.993536999353702e-06, + "loss": 0.3372, + "step": 6467 + }, + { + "epoch": 3.00697350069735, + "grad_norm": 4.940443515777588, + "learning_rate": 9.9935359993536e-06, + "loss": 0.3859, + "step": 6468 + }, + { + "epoch": 3.00743840074384, + "grad_norm": 4.163417339324951, + "learning_rate": 9.9935349993535e-06, + "loss": 0.3327, + "step": 6469 + }, + { + "epoch": 3.00790330079033, + "grad_norm": 4.748514175415039, + "learning_rate": 9.9935339993534e-06, + "loss": 0.297, + "step": 6470 + }, + { + "epoch": 3.00836820083682, + "grad_norm": 7.240175247192383, + "learning_rate": 9.993532999353301e-06, + "loss": 0.2639, + "step": 6471 + }, + { + "epoch": 3.0088331008833102, + "grad_norm": 6.074373245239258, + "learning_rate": 9.993531999353202e-06, + "loss": 0.3754, + "step": 6472 + }, + { + "epoch": 3.0092980009298, + "grad_norm": 5.303124904632568, + "learning_rate": 9.9935309993531e-06, + "loss": 0.2859, + "step": 6473 + }, + { + "epoch": 3.0097629009762903, + "grad_norm": 6.640782833099365, + "learning_rate": 9.993529999353001e-06, + "loss": 0.5, + "step": 6474 + }, + { + "epoch": 3.01022780102278, + "grad_norm": 4.08430814743042, + "learning_rate": 9.9935289993529e-06, + "loss": 0.2946, + "step": 6475 + }, + { + "epoch": 3.01069270106927, + "grad_norm": 5.482104778289795, + "learning_rate": 9.9935279993528e-06, + "loss": 0.4067, + "step": 6476 + }, + { + "epoch": 3.01115760111576, + "grad_norm": 4.338759422302246, + "learning_rate": 9.993526999352701e-06, + "loss": 0.3273, + "step": 6477 + }, + { + "epoch": 3.01162250116225, + "grad_norm": 5.1552934646606445, + "learning_rate": 9.9935259993526e-06, + "loss": 0.4063, + "step": 6478 + }, + { + "epoch": 3.01208740120874, + "grad_norm": 4.408321380615234, + "learning_rate": 9.993524999352501e-06, + "loss": 0.2323, + "step": 6479 + }, + { + "epoch": 3.01255230125523, + "grad_norm": 4.758731842041016, + "learning_rate": 9.9935239993524e-06, + "loss": 0.2998, + "step": 6480 + }, + { + "epoch": 3.0130172013017202, + "grad_norm": 6.4231085777282715, + "learning_rate": 9.9935229993523e-06, + "loss": 0.3284, + "step": 6481 + }, + { + "epoch": 3.01348210134821, + "grad_norm": 4.224413871765137, + "learning_rate": 9.9935219993522e-06, + "loss": 0.3444, + "step": 6482 + }, + { + "epoch": 3.0139470013947003, + "grad_norm": 5.92491340637207, + "learning_rate": 9.9935209993521e-06, + "loss": 0.1819, + "step": 6483 + }, + { + "epoch": 3.01441190144119, + "grad_norm": 4.655552864074707, + "learning_rate": 9.993519999352001e-06, + "loss": 0.3295, + "step": 6484 + }, + { + "epoch": 3.0148768014876803, + "grad_norm": 4.928391933441162, + "learning_rate": 9.993518999351902e-06, + "loss": 0.3463, + "step": 6485 + }, + { + "epoch": 3.01534170153417, + "grad_norm": 4.970902442932129, + "learning_rate": 9.9935179993518e-06, + "loss": 0.3915, + "step": 6486 + }, + { + "epoch": 3.01580660158066, + "grad_norm": 6.357214450836182, + "learning_rate": 9.993516999351701e-06, + "loss": 0.5155, + "step": 6487 + }, + { + "epoch": 3.01627150162715, + "grad_norm": 14.895002365112305, + "learning_rate": 9.9935159993516e-06, + "loss": 1.8611, + "step": 6488 + }, + { + "epoch": 3.01673640167364, + "grad_norm": 6.636816501617432, + "learning_rate": 9.993514999351501e-06, + "loss": 0.2256, + "step": 6489 + }, + { + "epoch": 3.0172013017201302, + "grad_norm": 9.115354537963867, + "learning_rate": 9.993513999351402e-06, + "loss": 0.4126, + "step": 6490 + }, + { + "epoch": 3.01766620176662, + "grad_norm": 6.668649196624756, + "learning_rate": 9.9935129993513e-06, + "loss": 0.4557, + "step": 6491 + }, + { + "epoch": 3.0181311018131103, + "grad_norm": 6.922341346740723, + "learning_rate": 9.993511999351201e-06, + "loss": 0.5573, + "step": 6492 + }, + { + "epoch": 3.0185960018596, + "grad_norm": 4.421820640563965, + "learning_rate": 9.9935109993511e-06, + "loss": 0.4337, + "step": 6493 + }, + { + "epoch": 3.0190609019060903, + "grad_norm": 5.0498046875, + "learning_rate": 9.993509999351e-06, + "loss": 0.353, + "step": 6494 + }, + { + "epoch": 3.01952580195258, + "grad_norm": 5.859799385070801, + "learning_rate": 9.9935089993509e-06, + "loss": 0.3841, + "step": 6495 + }, + { + "epoch": 3.0199907019990704, + "grad_norm": 5.439741611480713, + "learning_rate": 9.9935079993508e-06, + "loss": 0.4779, + "step": 6496 + }, + { + "epoch": 3.02045560204556, + "grad_norm": 5.294633865356445, + "learning_rate": 9.993506999350701e-06, + "loss": 0.4176, + "step": 6497 + }, + { + "epoch": 3.02092050209205, + "grad_norm": 7.970809459686279, + "learning_rate": 9.993505999350602e-06, + "loss": 0.7313, + "step": 6498 + }, + { + "epoch": 3.0213854021385402, + "grad_norm": 4.070308685302734, + "learning_rate": 9.9935049993505e-06, + "loss": 0.2726, + "step": 6499 + }, + { + "epoch": 3.02185030218503, + "grad_norm": 5.259549140930176, + "learning_rate": 9.993503999350401e-06, + "loss": 0.3135, + "step": 6500 + }, + { + "epoch": 3.0223152022315203, + "grad_norm": 4.2081685066223145, + "learning_rate": 9.9935029993503e-06, + "loss": 0.3759, + "step": 6501 + }, + { + "epoch": 3.02278010227801, + "grad_norm": 7.847158908843994, + "learning_rate": 9.993501999350201e-06, + "loss": 0.3303, + "step": 6502 + }, + { + "epoch": 3.0232450023245003, + "grad_norm": 4.751351833343506, + "learning_rate": 9.993500999350102e-06, + "loss": 0.2258, + "step": 6503 + }, + { + "epoch": 3.02370990237099, + "grad_norm": 7.029623508453369, + "learning_rate": 9.99349999935e-06, + "loss": 0.6702, + "step": 6504 + }, + { + "epoch": 3.0241748024174804, + "grad_norm": 6.276358604431152, + "learning_rate": 9.993498999349901e-06, + "loss": 0.3456, + "step": 6505 + }, + { + "epoch": 3.02463970246397, + "grad_norm": 5.562404632568359, + "learning_rate": 9.9934979993498e-06, + "loss": 0.4705, + "step": 6506 + }, + { + "epoch": 3.0251046025104604, + "grad_norm": 6.232768535614014, + "learning_rate": 9.993496999349701e-06, + "loss": 0.4021, + "step": 6507 + }, + { + "epoch": 3.0255695025569502, + "grad_norm": 5.284687042236328, + "learning_rate": 9.9934959993496e-06, + "loss": 0.2579, + "step": 6508 + }, + { + "epoch": 3.0260344026034405, + "grad_norm": 4.32225227355957, + "learning_rate": 9.9934949993495e-06, + "loss": 0.3226, + "step": 6509 + }, + { + "epoch": 3.0264993026499303, + "grad_norm": 5.710925579071045, + "learning_rate": 9.993493999349401e-06, + "loss": 0.2972, + "step": 6510 + }, + { + "epoch": 3.02696420269642, + "grad_norm": 4.3011274337768555, + "learning_rate": 9.993492999349302e-06, + "loss": 0.3411, + "step": 6511 + }, + { + "epoch": 3.0274291027429103, + "grad_norm": 9.091463088989258, + "learning_rate": 9.9934919993492e-06, + "loss": 0.651, + "step": 6512 + }, + { + "epoch": 3.0278940027894, + "grad_norm": 3.5887978076934814, + "learning_rate": 9.9934909993491e-06, + "loss": 0.1713, + "step": 6513 + }, + { + "epoch": 3.0283589028358904, + "grad_norm": 7.344145774841309, + "learning_rate": 9.993489999349e-06, + "loss": 0.3679, + "step": 6514 + }, + { + "epoch": 3.02882380288238, + "grad_norm": 5.560347080230713, + "learning_rate": 9.993488999348901e-06, + "loss": 0.5794, + "step": 6515 + }, + { + "epoch": 3.0292887029288704, + "grad_norm": 4.983603000640869, + "learning_rate": 9.993487999348802e-06, + "loss": 0.3326, + "step": 6516 + }, + { + "epoch": 3.0297536029753602, + "grad_norm": 7.387828350067139, + "learning_rate": 9.9934869993487e-06, + "loss": 0.6517, + "step": 6517 + }, + { + "epoch": 3.0302185030218505, + "grad_norm": 5.522238731384277, + "learning_rate": 9.993485999348601e-06, + "loss": 0.4111, + "step": 6518 + }, + { + "epoch": 3.0306834030683403, + "grad_norm": 7.305154323577881, + "learning_rate": 9.9934849993485e-06, + "loss": 0.5732, + "step": 6519 + }, + { + "epoch": 3.0311483031148305, + "grad_norm": 7.603208541870117, + "learning_rate": 9.993483999348401e-06, + "loss": 0.378, + "step": 6520 + }, + { + "epoch": 3.0316132031613203, + "grad_norm": 4.164745807647705, + "learning_rate": 9.9934829993483e-06, + "loss": 0.3131, + "step": 6521 + }, + { + "epoch": 3.03207810320781, + "grad_norm": 4.091364860534668, + "learning_rate": 9.9934819993482e-06, + "loss": 0.3174, + "step": 6522 + }, + { + "epoch": 3.0325430032543004, + "grad_norm": 5.257673740386963, + "learning_rate": 9.993480999348101e-06, + "loss": 0.347, + "step": 6523 + }, + { + "epoch": 3.03300790330079, + "grad_norm": 4.649690628051758, + "learning_rate": 9.993479999348e-06, + "loss": 0.2552, + "step": 6524 + }, + { + "epoch": 3.0334728033472804, + "grad_norm": 9.03918170928955, + "learning_rate": 9.993478999347901e-06, + "loss": 0.4536, + "step": 6525 + }, + { + "epoch": 3.0339377033937702, + "grad_norm": 5.3081560134887695, + "learning_rate": 9.9934779993478e-06, + "loss": 0.2724, + "step": 6526 + }, + { + "epoch": 3.0344026034402605, + "grad_norm": 6.7225565910339355, + "learning_rate": 9.9934769993477e-06, + "loss": 0.5679, + "step": 6527 + }, + { + "epoch": 3.0348675034867503, + "grad_norm": 6.98511266708374, + "learning_rate": 9.993475999347601e-06, + "loss": 0.6348, + "step": 6528 + }, + { + "epoch": 3.0353324035332405, + "grad_norm": 5.8139848709106445, + "learning_rate": 9.993474999347502e-06, + "loss": 0.5377, + "step": 6529 + }, + { + "epoch": 3.0357973035797303, + "grad_norm": 6.179758071899414, + "learning_rate": 9.9934739993474e-06, + "loss": 0.4014, + "step": 6530 + }, + { + "epoch": 3.0362622036262206, + "grad_norm": 2.9910755157470703, + "learning_rate": 9.993472999347301e-06, + "loss": 0.1676, + "step": 6531 + }, + { + "epoch": 3.0367271036727104, + "grad_norm": 4.6572723388671875, + "learning_rate": 9.9934719993472e-06, + "loss": 0.2444, + "step": 6532 + }, + { + "epoch": 3.0371920037192, + "grad_norm": 6.807001113891602, + "learning_rate": 9.993470999347101e-06, + "loss": 0.1919, + "step": 6533 + }, + { + "epoch": 3.0376569037656904, + "grad_norm": 4.998176097869873, + "learning_rate": 9.993469999347e-06, + "loss": 0.2779, + "step": 6534 + }, + { + "epoch": 3.03812180381218, + "grad_norm": 6.690260410308838, + "learning_rate": 9.9934689993469e-06, + "loss": 0.4646, + "step": 6535 + }, + { + "epoch": 3.0385867038586705, + "grad_norm": 5.482710838317871, + "learning_rate": 9.993467999346801e-06, + "loss": 0.4013, + "step": 6536 + }, + { + "epoch": 3.0390516039051603, + "grad_norm": 5.542501926422119, + "learning_rate": 9.9934669993467e-06, + "loss": 0.2642, + "step": 6537 + }, + { + "epoch": 3.0395165039516505, + "grad_norm": 7.398382186889648, + "learning_rate": 9.993465999346601e-06, + "loss": 0.3812, + "step": 6538 + }, + { + "epoch": 3.0399814039981403, + "grad_norm": 7.1309638023376465, + "learning_rate": 9.9934649993465e-06, + "loss": 0.341, + "step": 6539 + }, + { + "epoch": 3.0404463040446306, + "grad_norm": 5.004190444946289, + "learning_rate": 9.9934639993464e-06, + "loss": 0.3356, + "step": 6540 + }, + { + "epoch": 3.0409112040911204, + "grad_norm": 4.666360855102539, + "learning_rate": 9.993462999346301e-06, + "loss": 0.3294, + "step": 6541 + }, + { + "epoch": 3.0413761041376106, + "grad_norm": 4.334031105041504, + "learning_rate": 9.993461999346202e-06, + "loss": 0.3421, + "step": 6542 + }, + { + "epoch": 3.0418410041841004, + "grad_norm": 4.203700065612793, + "learning_rate": 9.9934609993461e-06, + "loss": 0.2996, + "step": 6543 + }, + { + "epoch": 3.04230590423059, + "grad_norm": 5.655267715454102, + "learning_rate": 9.993459999346001e-06, + "loss": 0.4806, + "step": 6544 + }, + { + "epoch": 3.0427708042770805, + "grad_norm": 4.544229984283447, + "learning_rate": 9.9934589993459e-06, + "loss": 0.317, + "step": 6545 + }, + { + "epoch": 3.0432357043235703, + "grad_norm": 5.128479480743408, + "learning_rate": 9.993457999345801e-06, + "loss": 0.2818, + "step": 6546 + }, + { + "epoch": 3.0437006043700605, + "grad_norm": 6.736198902130127, + "learning_rate": 9.9934569993457e-06, + "loss": 0.537, + "step": 6547 + }, + { + "epoch": 3.0441655044165503, + "grad_norm": 3.97683048248291, + "learning_rate": 9.9934559993456e-06, + "loss": 0.4294, + "step": 6548 + }, + { + "epoch": 3.0446304044630406, + "grad_norm": 4.809457302093506, + "learning_rate": 9.993454999345501e-06, + "loss": 0.2439, + "step": 6549 + }, + { + "epoch": 3.0450953045095304, + "grad_norm": 8.22905445098877, + "learning_rate": 9.9934539993454e-06, + "loss": 0.3582, + "step": 6550 + }, + { + "epoch": 3.0455602045560206, + "grad_norm": 8.177946090698242, + "learning_rate": 9.993452999345301e-06, + "loss": 0.576, + "step": 6551 + }, + { + "epoch": 3.0460251046025104, + "grad_norm": 4.581088066101074, + "learning_rate": 9.9934519993452e-06, + "loss": 0.3956, + "step": 6552 + }, + { + "epoch": 3.0464900046490007, + "grad_norm": 6.213138103485107, + "learning_rate": 9.9934509993451e-06, + "loss": 0.5595, + "step": 6553 + }, + { + "epoch": 3.0469549046954905, + "grad_norm": 5.3755621910095215, + "learning_rate": 9.993449999345001e-06, + "loss": 0.3349, + "step": 6554 + }, + { + "epoch": 3.0474198047419803, + "grad_norm": 6.604043483734131, + "learning_rate": 9.9934489993449e-06, + "loss": 0.5537, + "step": 6555 + }, + { + "epoch": 3.0478847047884705, + "grad_norm": 7.790536403656006, + "learning_rate": 9.993447999344801e-06, + "loss": 0.583, + "step": 6556 + }, + { + "epoch": 3.0483496048349603, + "grad_norm": 3.014172077178955, + "learning_rate": 9.9934469993447e-06, + "loss": 0.155, + "step": 6557 + }, + { + "epoch": 3.0488145048814506, + "grad_norm": 4.82693338394165, + "learning_rate": 9.9934459993446e-06, + "loss": 0.4339, + "step": 6558 + }, + { + "epoch": 3.0492794049279404, + "grad_norm": 3.7862799167633057, + "learning_rate": 9.993444999344501e-06, + "loss": 0.2321, + "step": 6559 + }, + { + "epoch": 3.0497443049744306, + "grad_norm": 4.538299560546875, + "learning_rate": 9.993443999344402e-06, + "loss": 0.3058, + "step": 6560 + }, + { + "epoch": 3.0502092050209204, + "grad_norm": 7.801648139953613, + "learning_rate": 9.9934429993443e-06, + "loss": 0.4542, + "step": 6561 + }, + { + "epoch": 3.0506741050674107, + "grad_norm": 3.2836856842041016, + "learning_rate": 9.993441999344201e-06, + "loss": 0.2117, + "step": 6562 + }, + { + "epoch": 3.0511390051139005, + "grad_norm": 8.337841033935547, + "learning_rate": 9.9934409993441e-06, + "loss": 0.4604, + "step": 6563 + }, + { + "epoch": 3.0516039051603907, + "grad_norm": 5.2921271324157715, + "learning_rate": 9.993439999344001e-06, + "loss": 0.2893, + "step": 6564 + }, + { + "epoch": 3.0520688052068805, + "grad_norm": 4.749012470245361, + "learning_rate": 9.9934389993439e-06, + "loss": 0.2972, + "step": 6565 + }, + { + "epoch": 3.0525337052533703, + "grad_norm": 5.5878825187683105, + "learning_rate": 9.9934379993438e-06, + "loss": 0.3197, + "step": 6566 + }, + { + "epoch": 3.0529986052998606, + "grad_norm": 3.9616522789001465, + "learning_rate": 9.993436999343701e-06, + "loss": 0.2302, + "step": 6567 + }, + { + "epoch": 3.0534635053463504, + "grad_norm": 4.582493305206299, + "learning_rate": 9.9934359993436e-06, + "loss": 0.3282, + "step": 6568 + }, + { + "epoch": 3.0539284053928406, + "grad_norm": 5.293577194213867, + "learning_rate": 9.993434999343501e-06, + "loss": 0.3408, + "step": 6569 + }, + { + "epoch": 3.0543933054393304, + "grad_norm": 7.858805179595947, + "learning_rate": 9.9934339993434e-06, + "loss": 0.5759, + "step": 6570 + }, + { + "epoch": 3.0548582054858207, + "grad_norm": 5.616227149963379, + "learning_rate": 9.9934329993433e-06, + "loss": 0.449, + "step": 6571 + }, + { + "epoch": 3.0553231055323105, + "grad_norm": 5.160716533660889, + "learning_rate": 9.993431999343201e-06, + "loss": 0.2706, + "step": 6572 + }, + { + "epoch": 3.0557880055788007, + "grad_norm": 4.890803813934326, + "learning_rate": 9.993430999343102e-06, + "loss": 0.202, + "step": 6573 + }, + { + "epoch": 3.0562529056252905, + "grad_norm": 11.1590576171875, + "learning_rate": 9.993429999343001e-06, + "loss": 0.5112, + "step": 6574 + }, + { + "epoch": 3.0567178056717808, + "grad_norm": 6.837064743041992, + "learning_rate": 9.993428999342902e-06, + "loss": 0.322, + "step": 6575 + }, + { + "epoch": 3.0571827057182706, + "grad_norm": 5.699185371398926, + "learning_rate": 9.9934279993428e-06, + "loss": 0.2077, + "step": 6576 + }, + { + "epoch": 3.0576476057647604, + "grad_norm": 5.295156955718994, + "learning_rate": 9.9934269993427e-06, + "loss": 0.4442, + "step": 6577 + }, + { + "epoch": 3.0581125058112506, + "grad_norm": 3.091982364654541, + "learning_rate": 9.9934259993426e-06, + "loss": 0.1353, + "step": 6578 + }, + { + "epoch": 3.0585774058577404, + "grad_norm": 5.283457279205322, + "learning_rate": 9.9934249993425e-06, + "loss": 0.3398, + "step": 6579 + }, + { + "epoch": 3.0590423059042307, + "grad_norm": 7.353675365447998, + "learning_rate": 9.993423999342401e-06, + "loss": 0.4604, + "step": 6580 + }, + { + "epoch": 3.0595072059507205, + "grad_norm": 3.8820395469665527, + "learning_rate": 9.9934229993423e-06, + "loss": 0.1981, + "step": 6581 + }, + { + "epoch": 3.0599721059972107, + "grad_norm": 5.867837429046631, + "learning_rate": 9.993421999342201e-06, + "loss": 0.4596, + "step": 6582 + }, + { + "epoch": 3.0604370060437005, + "grad_norm": 4.3733720779418945, + "learning_rate": 9.9934209993421e-06, + "loss": 0.2636, + "step": 6583 + }, + { + "epoch": 3.0609019060901907, + "grad_norm": 4.339534759521484, + "learning_rate": 9.993419999342e-06, + "loss": 0.1732, + "step": 6584 + }, + { + "epoch": 3.0613668061366806, + "grad_norm": 4.674377918243408, + "learning_rate": 9.993418999341901e-06, + "loss": 0.3036, + "step": 6585 + }, + { + "epoch": 3.061831706183171, + "grad_norm": 5.484984397888184, + "learning_rate": 9.993417999341802e-06, + "loss": 0.44, + "step": 6586 + }, + { + "epoch": 3.0622966062296606, + "grad_norm": 5.245548248291016, + "learning_rate": 9.993416999341701e-06, + "loss": 0.2217, + "step": 6587 + }, + { + "epoch": 3.062761506276151, + "grad_norm": 3.904958963394165, + "learning_rate": 9.9934159993416e-06, + "loss": 0.33, + "step": 6588 + }, + { + "epoch": 3.0632264063226406, + "grad_norm": 4.927145004272461, + "learning_rate": 9.9934149993415e-06, + "loss": 0.1185, + "step": 6589 + }, + { + "epoch": 3.0636913063691305, + "grad_norm": 7.039090633392334, + "learning_rate": 9.9934139993414e-06, + "loss": 0.4337, + "step": 6590 + }, + { + "epoch": 3.0641562064156207, + "grad_norm": 4.828031539916992, + "learning_rate": 9.9934129993413e-06, + "loss": 0.1975, + "step": 6591 + }, + { + "epoch": 3.0646211064621105, + "grad_norm": 3.7929959297180176, + "learning_rate": 9.993411999341201e-06, + "loss": 0.2402, + "step": 6592 + }, + { + "epoch": 3.0650860065086007, + "grad_norm": 5.363829135894775, + "learning_rate": 9.993410999341102e-06, + "loss": 0.3624, + "step": 6593 + }, + { + "epoch": 3.0655509065550905, + "grad_norm": 4.769661903381348, + "learning_rate": 9.993409999341e-06, + "loss": 0.3695, + "step": 6594 + }, + { + "epoch": 3.066015806601581, + "grad_norm": 5.5653300285339355, + "learning_rate": 9.993408999340901e-06, + "loss": 0.5059, + "step": 6595 + }, + { + "epoch": 3.0664807066480706, + "grad_norm": 5.341317176818848, + "learning_rate": 9.9934079993408e-06, + "loss": 0.4524, + "step": 6596 + }, + { + "epoch": 3.066945606694561, + "grad_norm": 5.929602146148682, + "learning_rate": 9.9934069993407e-06, + "loss": 0.3324, + "step": 6597 + }, + { + "epoch": 3.0674105067410506, + "grad_norm": 4.25147008895874, + "learning_rate": 9.993405999340601e-06, + "loss": 0.2802, + "step": 6598 + }, + { + "epoch": 3.067875406787541, + "grad_norm": 7.913113117218018, + "learning_rate": 9.9934049993405e-06, + "loss": 0.4122, + "step": 6599 + }, + { + "epoch": 3.0683403068340307, + "grad_norm": 4.45880126953125, + "learning_rate": 9.993403999340401e-06, + "loss": 0.3594, + "step": 6600 + }, + { + "epoch": 3.0688052068805205, + "grad_norm": 2.744159460067749, + "learning_rate": 9.9934029993403e-06, + "loss": 0.1245, + "step": 6601 + }, + { + "epoch": 3.0692701069270107, + "grad_norm": 6.715419769287109, + "learning_rate": 9.9934019993402e-06, + "loss": 0.2661, + "step": 6602 + }, + { + "epoch": 3.0697350069735005, + "grad_norm": 7.4110918045043945, + "learning_rate": 9.9934009993401e-06, + "loss": 0.6309, + "step": 6603 + }, + { + "epoch": 3.070199907019991, + "grad_norm": 6.185568809509277, + "learning_rate": 9.99339999934e-06, + "loss": 0.4382, + "step": 6604 + }, + { + "epoch": 3.0706648070664806, + "grad_norm": 7.066239833831787, + "learning_rate": 9.993398999339901e-06, + "loss": 0.4702, + "step": 6605 + }, + { + "epoch": 3.071129707112971, + "grad_norm": 7.650163173675537, + "learning_rate": 9.993397999339802e-06, + "loss": 0.7377, + "step": 6606 + }, + { + "epoch": 3.0715946071594606, + "grad_norm": 5.062954902648926, + "learning_rate": 9.9933969993397e-06, + "loss": 0.3163, + "step": 6607 + }, + { + "epoch": 3.072059507205951, + "grad_norm": 4.817061901092529, + "learning_rate": 9.993395999339601e-06, + "loss": 0.4102, + "step": 6608 + }, + { + "epoch": 3.0725244072524407, + "grad_norm": 9.250951766967773, + "learning_rate": 9.9933949993395e-06, + "loss": 0.369, + "step": 6609 + }, + { + "epoch": 3.072989307298931, + "grad_norm": 7.119334697723389, + "learning_rate": 9.993393999339401e-06, + "loss": 0.5931, + "step": 6610 + }, + { + "epoch": 3.0734542073454207, + "grad_norm": 5.7962775230407715, + "learning_rate": 9.993392999339302e-06, + "loss": 0.5785, + "step": 6611 + }, + { + "epoch": 3.0739191073919105, + "grad_norm": 4.506809234619141, + "learning_rate": 9.9933919993392e-06, + "loss": 0.3278, + "step": 6612 + }, + { + "epoch": 3.074384007438401, + "grad_norm": 5.277029037475586, + "learning_rate": 9.993390999339101e-06, + "loss": 0.4024, + "step": 6613 + }, + { + "epoch": 3.0748489074848906, + "grad_norm": 3.018601179122925, + "learning_rate": 9.993389999339e-06, + "loss": 0.1286, + "step": 6614 + }, + { + "epoch": 3.075313807531381, + "grad_norm": 4.088971138000488, + "learning_rate": 9.9933889993389e-06, + "loss": 0.2293, + "step": 6615 + }, + { + "epoch": 3.0757787075778706, + "grad_norm": 3.932277202606201, + "learning_rate": 9.9933879993388e-06, + "loss": 0.2343, + "step": 6616 + }, + { + "epoch": 3.076243607624361, + "grad_norm": 6.85985803604126, + "learning_rate": 9.9933869993387e-06, + "loss": 0.4706, + "step": 6617 + }, + { + "epoch": 3.0767085076708507, + "grad_norm": 5.954880237579346, + "learning_rate": 9.993385999338601e-06, + "loss": 0.3979, + "step": 6618 + }, + { + "epoch": 3.077173407717341, + "grad_norm": 3.255471706390381, + "learning_rate": 9.993384999338502e-06, + "loss": 0.2299, + "step": 6619 + }, + { + "epoch": 3.0776383077638307, + "grad_norm": 7.1639275550842285, + "learning_rate": 9.9933839993384e-06, + "loss": 0.6788, + "step": 6620 + }, + { + "epoch": 3.078103207810321, + "grad_norm": 6.967048645019531, + "learning_rate": 9.9933829993383e-06, + "loss": 0.3806, + "step": 6621 + }, + { + "epoch": 3.078568107856811, + "grad_norm": 4.224449157714844, + "learning_rate": 9.9933819993382e-06, + "loss": 0.4016, + "step": 6622 + }, + { + "epoch": 3.0790330079033006, + "grad_norm": 4.9757795333862305, + "learning_rate": 9.993380999338101e-06, + "loss": 0.2426, + "step": 6623 + }, + { + "epoch": 3.079497907949791, + "grad_norm": 6.442019939422607, + "learning_rate": 9.993379999338002e-06, + "loss": 0.1703, + "step": 6624 + }, + { + "epoch": 3.0799628079962806, + "grad_norm": 5.396347522735596, + "learning_rate": 9.9933789993379e-06, + "loss": 0.2888, + "step": 6625 + }, + { + "epoch": 3.080427708042771, + "grad_norm": 4.154814720153809, + "learning_rate": 9.993377999337801e-06, + "loss": 0.3644, + "step": 6626 + }, + { + "epoch": 3.0808926080892607, + "grad_norm": 3.590635299682617, + "learning_rate": 9.9933769993377e-06, + "loss": 0.2321, + "step": 6627 + }, + { + "epoch": 3.081357508135751, + "grad_norm": 5.560204029083252, + "learning_rate": 9.9933759993376e-06, + "loss": 0.5548, + "step": 6628 + }, + { + "epoch": 3.0818224081822407, + "grad_norm": 4.066268444061279, + "learning_rate": 9.993374999337501e-06, + "loss": 0.1776, + "step": 6629 + }, + { + "epoch": 3.082287308228731, + "grad_norm": 4.961853981018066, + "learning_rate": 9.9933739993374e-06, + "loss": 0.3241, + "step": 6630 + }, + { + "epoch": 3.082752208275221, + "grad_norm": 11.949103355407715, + "learning_rate": 9.993372999337301e-06, + "loss": 0.7831, + "step": 6631 + }, + { + "epoch": 3.083217108321711, + "grad_norm": 5.490677356719971, + "learning_rate": 9.9933719993372e-06, + "loss": 0.3613, + "step": 6632 + }, + { + "epoch": 3.083682008368201, + "grad_norm": 5.218023777008057, + "learning_rate": 9.9933709993371e-06, + "loss": 0.4242, + "step": 6633 + }, + { + "epoch": 3.0841469084146906, + "grad_norm": 6.027278423309326, + "learning_rate": 9.993369999337e-06, + "loss": 0.4318, + "step": 6634 + }, + { + "epoch": 3.084611808461181, + "grad_norm": 5.862231254577637, + "learning_rate": 9.9933689993369e-06, + "loss": 0.3954, + "step": 6635 + }, + { + "epoch": 3.0850767085076707, + "grad_norm": 7.290822505950928, + "learning_rate": 9.993367999336801e-06, + "loss": 0.3959, + "step": 6636 + }, + { + "epoch": 3.085541608554161, + "grad_norm": 4.937671184539795, + "learning_rate": 9.993366999336702e-06, + "loss": 0.4175, + "step": 6637 + }, + { + "epoch": 3.0860065086006507, + "grad_norm": 5.207125663757324, + "learning_rate": 9.9933659993366e-06, + "loss": 0.3157, + "step": 6638 + }, + { + "epoch": 3.086471408647141, + "grad_norm": 10.848355293273926, + "learning_rate": 9.993364999336501e-06, + "loss": 0.3642, + "step": 6639 + }, + { + "epoch": 3.086936308693631, + "grad_norm": 14.73580551147461, + "learning_rate": 9.9933639993364e-06, + "loss": 1.6689, + "step": 6640 + }, + { + "epoch": 3.087401208740121, + "grad_norm": 5.330877304077148, + "learning_rate": 9.993362999336301e-06, + "loss": 0.5104, + "step": 6641 + }, + { + "epoch": 3.087866108786611, + "grad_norm": 4.621266841888428, + "learning_rate": 9.993361999336202e-06, + "loss": 0.3476, + "step": 6642 + }, + { + "epoch": 3.088331008833101, + "grad_norm": 6.5357890129089355, + "learning_rate": 9.9933609993361e-06, + "loss": 0.5085, + "step": 6643 + }, + { + "epoch": 3.088795908879591, + "grad_norm": 8.865326881408691, + "learning_rate": 9.993359999336001e-06, + "loss": 0.1352, + "step": 6644 + }, + { + "epoch": 3.0892608089260807, + "grad_norm": 3.6077845096588135, + "learning_rate": 9.9933589993359e-06, + "loss": 0.3029, + "step": 6645 + }, + { + "epoch": 3.089725708972571, + "grad_norm": 4.545477390289307, + "learning_rate": 9.9933579993358e-06, + "loss": 0.2886, + "step": 6646 + }, + { + "epoch": 3.0901906090190607, + "grad_norm": 6.472936153411865, + "learning_rate": 9.9933569993357e-06, + "loss": 0.4, + "step": 6647 + }, + { + "epoch": 3.090655509065551, + "grad_norm": 4.86948299407959, + "learning_rate": 9.9933559993356e-06, + "loss": 0.2655, + "step": 6648 + }, + { + "epoch": 3.091120409112041, + "grad_norm": 6.905642986297607, + "learning_rate": 9.993354999335501e-06, + "loss": 0.261, + "step": 6649 + }, + { + "epoch": 3.091585309158531, + "grad_norm": 5.911109447479248, + "learning_rate": 9.993353999335402e-06, + "loss": 0.6145, + "step": 6650 + }, + { + "epoch": 3.092050209205021, + "grad_norm": 4.1263604164123535, + "learning_rate": 9.9933529993353e-06, + "loss": 0.2452, + "step": 6651 + }, + { + "epoch": 3.092515109251511, + "grad_norm": 4.702913761138916, + "learning_rate": 9.993351999335201e-06, + "loss": 0.2018, + "step": 6652 + }, + { + "epoch": 3.092980009298001, + "grad_norm": 5.553879737854004, + "learning_rate": 9.9933509993351e-06, + "loss": 0.297, + "step": 6653 + }, + { + "epoch": 3.093444909344491, + "grad_norm": 2.8480277061462402, + "learning_rate": 9.993349999335001e-06, + "loss": 0.1827, + "step": 6654 + }, + { + "epoch": 3.093909809390981, + "grad_norm": 3.8244502544403076, + "learning_rate": 9.993348999334902e-06, + "loss": 0.2449, + "step": 6655 + }, + { + "epoch": 3.0943747094374707, + "grad_norm": 8.145524978637695, + "learning_rate": 9.9933479993348e-06, + "loss": 0.3221, + "step": 6656 + }, + { + "epoch": 3.094839609483961, + "grad_norm": 3.6442337036132812, + "learning_rate": 9.993346999334701e-06, + "loss": 0.1527, + "step": 6657 + }, + { + "epoch": 3.0953045095304508, + "grad_norm": 3.561432361602783, + "learning_rate": 9.9933459993346e-06, + "loss": 0.1994, + "step": 6658 + }, + { + "epoch": 3.095769409576941, + "grad_norm": 6.127691745758057, + "learning_rate": 9.993344999334501e-06, + "loss": 0.3913, + "step": 6659 + }, + { + "epoch": 3.096234309623431, + "grad_norm": 6.34126615524292, + "learning_rate": 9.9933439993344e-06, + "loss": 0.5218, + "step": 6660 + }, + { + "epoch": 3.096699209669921, + "grad_norm": 4.239640235900879, + "learning_rate": 9.9933429993343e-06, + "loss": 0.1273, + "step": 6661 + }, + { + "epoch": 3.097164109716411, + "grad_norm": 4.876895427703857, + "learning_rate": 9.993341999334201e-06, + "loss": 0.2899, + "step": 6662 + }, + { + "epoch": 3.097629009762901, + "grad_norm": 3.7636654376983643, + "learning_rate": 9.993340999334102e-06, + "loss": 0.2835, + "step": 6663 + }, + { + "epoch": 3.098093909809391, + "grad_norm": 7.33107852935791, + "learning_rate": 9.993339999334e-06, + "loss": 0.5641, + "step": 6664 + }, + { + "epoch": 3.098558809855881, + "grad_norm": 4.240625381469727, + "learning_rate": 9.9933389993339e-06, + "loss": 0.2717, + "step": 6665 + }, + { + "epoch": 3.099023709902371, + "grad_norm": 6.169979572296143, + "learning_rate": 9.9933379993338e-06, + "loss": 0.4565, + "step": 6666 + }, + { + "epoch": 3.099488609948861, + "grad_norm": 7.624612808227539, + "learning_rate": 9.993336999333701e-06, + "loss": 0.7364, + "step": 6667 + }, + { + "epoch": 3.099953509995351, + "grad_norm": 5.230634689331055, + "learning_rate": 9.993335999333602e-06, + "loss": 0.3982, + "step": 6668 + }, + { + "epoch": 3.100418410041841, + "grad_norm": 4.010495185852051, + "learning_rate": 9.9933349993335e-06, + "loss": 0.3253, + "step": 6669 + }, + { + "epoch": 3.100883310088331, + "grad_norm": 3.5318620204925537, + "learning_rate": 9.993333999333401e-06, + "loss": 0.1304, + "step": 6670 + }, + { + "epoch": 3.101348210134821, + "grad_norm": 5.865382194519043, + "learning_rate": 9.9933329993333e-06, + "loss": 0.4363, + "step": 6671 + }, + { + "epoch": 3.101813110181311, + "grad_norm": 5.375895023345947, + "learning_rate": 9.993331999333201e-06, + "loss": 0.2773, + "step": 6672 + }, + { + "epoch": 3.102278010227801, + "grad_norm": 5.8542985916137695, + "learning_rate": 9.9933309993331e-06, + "loss": 0.3602, + "step": 6673 + }, + { + "epoch": 3.102742910274291, + "grad_norm": 4.881382942199707, + "learning_rate": 9.993329999333e-06, + "loss": 0.3292, + "step": 6674 + }, + { + "epoch": 3.103207810320781, + "grad_norm": 3.1607584953308105, + "learning_rate": 9.993328999332901e-06, + "loss": 0.1301, + "step": 6675 + }, + { + "epoch": 3.103672710367271, + "grad_norm": 2.835479497909546, + "learning_rate": 9.9933279993328e-06, + "loss": 0.1439, + "step": 6676 + }, + { + "epoch": 3.104137610413761, + "grad_norm": 4.939970016479492, + "learning_rate": 9.993326999332701e-06, + "loss": 0.2351, + "step": 6677 + }, + { + "epoch": 3.104602510460251, + "grad_norm": 7.4329094886779785, + "learning_rate": 9.9933259993326e-06, + "loss": 0.5607, + "step": 6678 + }, + { + "epoch": 3.105067410506741, + "grad_norm": 3.742504596710205, + "learning_rate": 9.9933249993325e-06, + "loss": 0.2543, + "step": 6679 + }, + { + "epoch": 3.105532310553231, + "grad_norm": 4.861566066741943, + "learning_rate": 9.993323999332401e-06, + "loss": 0.4295, + "step": 6680 + }, + { + "epoch": 3.105997210599721, + "grad_norm": 4.039530277252197, + "learning_rate": 9.993322999332302e-06, + "loss": 0.2467, + "step": 6681 + }, + { + "epoch": 3.106462110646211, + "grad_norm": 8.24579906463623, + "learning_rate": 9.9933219993322e-06, + "loss": 0.4909, + "step": 6682 + }, + { + "epoch": 3.106927010692701, + "grad_norm": 4.066402435302734, + "learning_rate": 9.993320999332101e-06, + "loss": 0.3118, + "step": 6683 + }, + { + "epoch": 3.107391910739191, + "grad_norm": 5.369564056396484, + "learning_rate": 9.993319999332e-06, + "loss": 0.3086, + "step": 6684 + }, + { + "epoch": 3.107856810785681, + "grad_norm": 7.720337390899658, + "learning_rate": 9.993318999331901e-06, + "loss": 0.5371, + "step": 6685 + }, + { + "epoch": 3.108321710832171, + "grad_norm": 4.509089946746826, + "learning_rate": 9.9933179993318e-06, + "loss": 0.2609, + "step": 6686 + }, + { + "epoch": 3.1087866108786613, + "grad_norm": 6.436683177947998, + "learning_rate": 9.9933169993317e-06, + "loss": 0.3294, + "step": 6687 + }, + { + "epoch": 3.109251510925151, + "grad_norm": 3.687229633331299, + "learning_rate": 9.993315999331601e-06, + "loss": 0.2931, + "step": 6688 + }, + { + "epoch": 3.1097164109716413, + "grad_norm": 6.274737358093262, + "learning_rate": 9.9933149993315e-06, + "loss": 0.4775, + "step": 6689 + }, + { + "epoch": 3.110181311018131, + "grad_norm": 3.840259313583374, + "learning_rate": 9.993313999331401e-06, + "loss": 0.2156, + "step": 6690 + }, + { + "epoch": 3.110646211064621, + "grad_norm": 5.491991996765137, + "learning_rate": 9.9933129993313e-06, + "loss": 0.4359, + "step": 6691 + }, + { + "epoch": 3.111111111111111, + "grad_norm": 5.080038547515869, + "learning_rate": 9.9933119993312e-06, + "loss": 0.2935, + "step": 6692 + }, + { + "epoch": 3.111576011157601, + "grad_norm": 3.5700533390045166, + "learning_rate": 9.993310999331101e-06, + "loss": 0.2282, + "step": 6693 + }, + { + "epoch": 3.112040911204091, + "grad_norm": 4.247894763946533, + "learning_rate": 9.993309999331002e-06, + "loss": 0.363, + "step": 6694 + }, + { + "epoch": 3.112505811250581, + "grad_norm": 11.936094284057617, + "learning_rate": 9.993308999330901e-06, + "loss": 0.3566, + "step": 6695 + }, + { + "epoch": 3.1129707112970713, + "grad_norm": 4.721043109893799, + "learning_rate": 9.9933079993308e-06, + "loss": 0.2992, + "step": 6696 + }, + { + "epoch": 3.113435611343561, + "grad_norm": 10.916254043579102, + "learning_rate": 9.9933069993307e-06, + "loss": 0.6151, + "step": 6697 + }, + { + "epoch": 3.1139005113900513, + "grad_norm": 4.367692947387695, + "learning_rate": 9.993305999330601e-06, + "loss": 0.2328, + "step": 6698 + }, + { + "epoch": 3.114365411436541, + "grad_norm": 6.346776962280273, + "learning_rate": 9.9933049993305e-06, + "loss": 0.4907, + "step": 6699 + }, + { + "epoch": 3.1148303114830314, + "grad_norm": 4.589867115020752, + "learning_rate": 9.9933039993304e-06, + "loss": 0.225, + "step": 6700 + }, + { + "epoch": 3.115295211529521, + "grad_norm": 8.203836441040039, + "learning_rate": 9.993302999330301e-06, + "loss": 0.6526, + "step": 6701 + }, + { + "epoch": 3.115760111576011, + "grad_norm": 6.359424114227295, + "learning_rate": 9.9933019993302e-06, + "loss": 0.4997, + "step": 6702 + }, + { + "epoch": 3.116225011622501, + "grad_norm": 4.97196626663208, + "learning_rate": 9.993300999330101e-06, + "loss": 0.2073, + "step": 6703 + }, + { + "epoch": 3.116689911668991, + "grad_norm": 4.365957736968994, + "learning_rate": 9.99329999933e-06, + "loss": 0.3991, + "step": 6704 + }, + { + "epoch": 3.1171548117154813, + "grad_norm": 4.135341644287109, + "learning_rate": 9.9932989993299e-06, + "loss": 0.2169, + "step": 6705 + }, + { + "epoch": 3.117619711761971, + "grad_norm": 4.481747627258301, + "learning_rate": 9.993297999329801e-06, + "loss": 0.2819, + "step": 6706 + }, + { + "epoch": 3.1180846118084613, + "grad_norm": 6.57716703414917, + "learning_rate": 9.9932969993297e-06, + "loss": 0.4696, + "step": 6707 + }, + { + "epoch": 3.118549511854951, + "grad_norm": 5.019410610198975, + "learning_rate": 9.993295999329601e-06, + "loss": 0.5069, + "step": 6708 + }, + { + "epoch": 3.1190144119014414, + "grad_norm": 5.758660316467285, + "learning_rate": 9.9932949993295e-06, + "loss": 0.4674, + "step": 6709 + }, + { + "epoch": 3.119479311947931, + "grad_norm": 4.02278995513916, + "learning_rate": 9.9932939993294e-06, + "loss": 0.243, + "step": 6710 + }, + { + "epoch": 3.1199442119944214, + "grad_norm": 4.838329792022705, + "learning_rate": 9.993292999329301e-06, + "loss": 0.3914, + "step": 6711 + }, + { + "epoch": 3.120409112040911, + "grad_norm": 4.370790958404541, + "learning_rate": 9.9932919993292e-06, + "loss": 0.2835, + "step": 6712 + }, + { + "epoch": 3.120874012087401, + "grad_norm": 9.933402061462402, + "learning_rate": 9.9932909993291e-06, + "loss": 0.507, + "step": 6713 + }, + { + "epoch": 3.1213389121338913, + "grad_norm": 5.9047088623046875, + "learning_rate": 9.993289999329002e-06, + "loss": 0.344, + "step": 6714 + }, + { + "epoch": 3.121803812180381, + "grad_norm": 6.1647748947143555, + "learning_rate": 9.9932889993289e-06, + "loss": 0.4631, + "step": 6715 + }, + { + "epoch": 3.1222687122268713, + "grad_norm": 5.158061981201172, + "learning_rate": 9.993287999328801e-06, + "loss": 0.4583, + "step": 6716 + }, + { + "epoch": 3.122733612273361, + "grad_norm": 6.2773871421813965, + "learning_rate": 9.9932869993287e-06, + "loss": 0.322, + "step": 6717 + }, + { + "epoch": 3.1231985123198513, + "grad_norm": 7.538154602050781, + "learning_rate": 9.9932859993286e-06, + "loss": 0.4011, + "step": 6718 + }, + { + "epoch": 3.123663412366341, + "grad_norm": 4.804591655731201, + "learning_rate": 9.993284999328501e-06, + "loss": 0.4657, + "step": 6719 + }, + { + "epoch": 3.1241283124128314, + "grad_norm": 4.528144359588623, + "learning_rate": 9.9932839993284e-06, + "loss": 0.1977, + "step": 6720 + }, + { + "epoch": 3.124593212459321, + "grad_norm": 4.273894786834717, + "learning_rate": 9.993282999328301e-06, + "loss": 0.2498, + "step": 6721 + }, + { + "epoch": 3.1250581125058114, + "grad_norm": 5.249650955200195, + "learning_rate": 9.9932819993282e-06, + "loss": 0.408, + "step": 6722 + }, + { + "epoch": 3.1255230125523012, + "grad_norm": 5.92936372756958, + "learning_rate": 9.9932809993281e-06, + "loss": 0.3033, + "step": 6723 + }, + { + "epoch": 3.125987912598791, + "grad_norm": 3.430634021759033, + "learning_rate": 9.993279999328001e-06, + "loss": 0.2061, + "step": 6724 + }, + { + "epoch": 3.1264528126452813, + "grad_norm": 5.001563549041748, + "learning_rate": 9.993278999327902e-06, + "loss": 0.2639, + "step": 6725 + }, + { + "epoch": 3.126917712691771, + "grad_norm": 9.804093360900879, + "learning_rate": 9.993277999327801e-06, + "loss": 0.7419, + "step": 6726 + }, + { + "epoch": 3.1273826127382613, + "grad_norm": 5.804643630981445, + "learning_rate": 9.993276999327702e-06, + "loss": 0.3076, + "step": 6727 + }, + { + "epoch": 3.127847512784751, + "grad_norm": 5.035150527954102, + "learning_rate": 9.9932759993276e-06, + "loss": 0.3291, + "step": 6728 + }, + { + "epoch": 3.1283124128312414, + "grad_norm": 5.5524115562438965, + "learning_rate": 9.9932749993275e-06, + "loss": 0.5545, + "step": 6729 + }, + { + "epoch": 3.128777312877731, + "grad_norm": 5.498424530029297, + "learning_rate": 9.9932739993274e-06, + "loss": 0.2734, + "step": 6730 + }, + { + "epoch": 3.1292422129242214, + "grad_norm": 4.715188980102539, + "learning_rate": 9.9932729993273e-06, + "loss": 0.2181, + "step": 6731 + }, + { + "epoch": 3.1297071129707112, + "grad_norm": 4.815274238586426, + "learning_rate": 9.993271999327201e-06, + "loss": 0.46, + "step": 6732 + }, + { + "epoch": 3.1301720130172015, + "grad_norm": 5.202629566192627, + "learning_rate": 9.9932709993271e-06, + "loss": 0.3711, + "step": 6733 + }, + { + "epoch": 3.1306369130636913, + "grad_norm": 6.924038410186768, + "learning_rate": 9.993269999327001e-06, + "loss": 0.6298, + "step": 6734 + }, + { + "epoch": 3.131101813110181, + "grad_norm": 3.7414779663085938, + "learning_rate": 9.9932689993269e-06, + "loss": 0.2155, + "step": 6735 + }, + { + "epoch": 3.1315667131566713, + "grad_norm": 4.886082649230957, + "learning_rate": 9.9932679993268e-06, + "loss": 0.1828, + "step": 6736 + }, + { + "epoch": 3.132031613203161, + "grad_norm": 6.001851558685303, + "learning_rate": 9.993266999326701e-06, + "loss": 0.2459, + "step": 6737 + }, + { + "epoch": 3.1324965132496514, + "grad_norm": 8.50053882598877, + "learning_rate": 9.993265999326602e-06, + "loss": 0.7365, + "step": 6738 + }, + { + "epoch": 3.132961413296141, + "grad_norm": 4.6217145919799805, + "learning_rate": 9.993264999326501e-06, + "loss": 0.3488, + "step": 6739 + }, + { + "epoch": 3.1334263133426314, + "grad_norm": 5.03187894821167, + "learning_rate": 9.9932639993264e-06, + "loss": 0.4106, + "step": 6740 + }, + { + "epoch": 3.1338912133891212, + "grad_norm": 6.417349815368652, + "learning_rate": 9.9932629993263e-06, + "loss": 0.4307, + "step": 6741 + }, + { + "epoch": 3.1343561134356115, + "grad_norm": 4.756894111633301, + "learning_rate": 9.9932619993262e-06, + "loss": 0.364, + "step": 6742 + }, + { + "epoch": 3.1348210134821013, + "grad_norm": 4.138437747955322, + "learning_rate": 9.9932609993261e-06, + "loss": 0.2445, + "step": 6743 + }, + { + "epoch": 3.1352859135285915, + "grad_norm": 5.349502086639404, + "learning_rate": 9.993259999326001e-06, + "loss": 0.3626, + "step": 6744 + }, + { + "epoch": 3.1357508135750813, + "grad_norm": 4.468254089355469, + "learning_rate": 9.993258999325902e-06, + "loss": 0.3494, + "step": 6745 + }, + { + "epoch": 3.1362157136215716, + "grad_norm": 3.9336326122283936, + "learning_rate": 9.9932579993258e-06, + "loss": 0.318, + "step": 6746 + }, + { + "epoch": 3.1366806136680614, + "grad_norm": 5.170281887054443, + "learning_rate": 9.993256999325701e-06, + "loss": 0.3898, + "step": 6747 + }, + { + "epoch": 3.137145513714551, + "grad_norm": 6.69973611831665, + "learning_rate": 9.9932559993256e-06, + "loss": 0.2763, + "step": 6748 + }, + { + "epoch": 3.1376104137610414, + "grad_norm": 8.962207794189453, + "learning_rate": 9.9932549993255e-06, + "loss": 0.7306, + "step": 6749 + }, + { + "epoch": 3.1380753138075312, + "grad_norm": 3.6556923389434814, + "learning_rate": 9.993253999325401e-06, + "loss": 0.2157, + "step": 6750 + }, + { + "epoch": 3.1385402138540215, + "grad_norm": 8.122370719909668, + "learning_rate": 9.9932529993253e-06, + "loss": 0.3838, + "step": 6751 + }, + { + "epoch": 3.1390051139005113, + "grad_norm": 6.447683811187744, + "learning_rate": 9.993251999325201e-06, + "loss": 0.5218, + "step": 6752 + }, + { + "epoch": 3.1394700139470015, + "grad_norm": 5.065829753875732, + "learning_rate": 9.9932509993251e-06, + "loss": 0.3662, + "step": 6753 + }, + { + "epoch": 3.1399349139934913, + "grad_norm": 6.836233615875244, + "learning_rate": 9.993249999325e-06, + "loss": 0.5444, + "step": 6754 + }, + { + "epoch": 3.1403998140399816, + "grad_norm": 7.065290927886963, + "learning_rate": 9.9932489993249e-06, + "loss": 0.2806, + "step": 6755 + }, + { + "epoch": 3.1408647140864714, + "grad_norm": 5.6937432289123535, + "learning_rate": 9.9932479993248e-06, + "loss": 0.4459, + "step": 6756 + }, + { + "epoch": 3.141329614132961, + "grad_norm": 4.683085918426514, + "learning_rate": 9.993246999324701e-06, + "loss": 0.2646, + "step": 6757 + }, + { + "epoch": 3.1417945141794514, + "grad_norm": 5.179460048675537, + "learning_rate": 9.993245999324602e-06, + "loss": 0.3342, + "step": 6758 + }, + { + "epoch": 3.1422594142259412, + "grad_norm": 5.815615177154541, + "learning_rate": 9.9932449993245e-06, + "loss": 0.3095, + "step": 6759 + }, + { + "epoch": 3.1427243142724315, + "grad_norm": 5.948932647705078, + "learning_rate": 9.993243999324401e-06, + "loss": 0.3206, + "step": 6760 + }, + { + "epoch": 3.1431892143189213, + "grad_norm": 4.19552755355835, + "learning_rate": 9.9932429993243e-06, + "loss": 0.2433, + "step": 6761 + }, + { + "epoch": 3.1436541143654115, + "grad_norm": 3.5680487155914307, + "learning_rate": 9.993241999324201e-06, + "loss": 0.2371, + "step": 6762 + }, + { + "epoch": 3.1441190144119013, + "grad_norm": 4.952763557434082, + "learning_rate": 9.993240999324102e-06, + "loss": 0.328, + "step": 6763 + }, + { + "epoch": 3.1445839144583916, + "grad_norm": 5.555406093597412, + "learning_rate": 9.993239999324e-06, + "loss": 0.4182, + "step": 6764 + }, + { + "epoch": 3.1450488145048814, + "grad_norm": 11.15718936920166, + "learning_rate": 9.993238999323901e-06, + "loss": 0.429, + "step": 6765 + }, + { + "epoch": 3.1455137145513716, + "grad_norm": 4.0041422843933105, + "learning_rate": 9.9932379993238e-06, + "loss": 0.3088, + "step": 6766 + }, + { + "epoch": 3.1459786145978614, + "grad_norm": 7.023111820220947, + "learning_rate": 9.9932369993237e-06, + "loss": 0.4551, + "step": 6767 + }, + { + "epoch": 3.1464435146443517, + "grad_norm": 4.039389610290527, + "learning_rate": 9.9932359993236e-06, + "loss": 0.2884, + "step": 6768 + }, + { + "epoch": 3.1469084146908415, + "grad_norm": 7.951324462890625, + "learning_rate": 9.9932349993235e-06, + "loss": 0.2732, + "step": 6769 + }, + { + "epoch": 3.1473733147373313, + "grad_norm": 5.611530780792236, + "learning_rate": 9.993233999323401e-06, + "loss": 0.433, + "step": 6770 + }, + { + "epoch": 3.1478382147838215, + "grad_norm": 3.525189161300659, + "learning_rate": 9.993232999323302e-06, + "loss": 0.1824, + "step": 6771 + }, + { + "epoch": 3.1483031148303113, + "grad_norm": 8.887323379516602, + "learning_rate": 9.9932319993232e-06, + "loss": 0.7269, + "step": 6772 + }, + { + "epoch": 3.1487680148768016, + "grad_norm": 5.663230895996094, + "learning_rate": 9.9932309993231e-06, + "loss": 0.2255, + "step": 6773 + }, + { + "epoch": 3.1492329149232914, + "grad_norm": 5.615072727203369, + "learning_rate": 9.993229999323e-06, + "loss": 0.6631, + "step": 6774 + }, + { + "epoch": 3.1496978149697816, + "grad_norm": 5.043954372406006, + "learning_rate": 9.993228999322901e-06, + "loss": 0.3037, + "step": 6775 + }, + { + "epoch": 3.1501627150162714, + "grad_norm": 3.239215612411499, + "learning_rate": 9.993227999322802e-06, + "loss": 0.2081, + "step": 6776 + }, + { + "epoch": 3.1506276150627617, + "grad_norm": 6.6497673988342285, + "learning_rate": 9.9932269993227e-06, + "loss": 0.5784, + "step": 6777 + }, + { + "epoch": 3.1510925151092515, + "grad_norm": 6.095803737640381, + "learning_rate": 9.993225999322601e-06, + "loss": 0.3225, + "step": 6778 + }, + { + "epoch": 3.1515574151557413, + "grad_norm": 4.636250019073486, + "learning_rate": 9.9932249993225e-06, + "loss": 0.2846, + "step": 6779 + }, + { + "epoch": 3.1520223152022315, + "grad_norm": 4.447196006774902, + "learning_rate": 9.993223999322401e-06, + "loss": 0.3132, + "step": 6780 + }, + { + "epoch": 3.1524872152487213, + "grad_norm": 5.582463264465332, + "learning_rate": 9.9932229993223e-06, + "loss": 0.4002, + "step": 6781 + }, + { + "epoch": 3.1529521152952116, + "grad_norm": 4.544326305389404, + "learning_rate": 9.9932219993222e-06, + "loss": 0.2391, + "step": 6782 + }, + { + "epoch": 3.1534170153417014, + "grad_norm": 4.3705058097839355, + "learning_rate": 9.993220999322101e-06, + "loss": 0.3387, + "step": 6783 + }, + { + "epoch": 3.1538819153881916, + "grad_norm": 3.6959118843078613, + "learning_rate": 9.993219999322e-06, + "loss": 0.1582, + "step": 6784 + }, + { + "epoch": 3.1543468154346814, + "grad_norm": 4.191586494445801, + "learning_rate": 9.9932189993219e-06, + "loss": 0.3063, + "step": 6785 + }, + { + "epoch": 3.1548117154811717, + "grad_norm": 5.841339588165283, + "learning_rate": 9.9932179993218e-06, + "loss": 0.4313, + "step": 6786 + }, + { + "epoch": 3.1552766155276615, + "grad_norm": 5.285788059234619, + "learning_rate": 9.9932169993217e-06, + "loss": 0.2523, + "step": 6787 + }, + { + "epoch": 3.1557415155741517, + "grad_norm": 8.864829063415527, + "learning_rate": 9.993215999321601e-06, + "loss": 0.6324, + "step": 6788 + }, + { + "epoch": 3.1562064156206415, + "grad_norm": 4.974559307098389, + "learning_rate": 9.993214999321502e-06, + "loss": 0.3892, + "step": 6789 + }, + { + "epoch": 3.1566713156671318, + "grad_norm": 5.631744384765625, + "learning_rate": 9.9932139993214e-06, + "loss": 0.4884, + "step": 6790 + }, + { + "epoch": 3.1571362157136216, + "grad_norm": 5.069777488708496, + "learning_rate": 9.993212999321301e-06, + "loss": 0.4207, + "step": 6791 + }, + { + "epoch": 3.1576011157601114, + "grad_norm": 5.65493631362915, + "learning_rate": 9.9932119993212e-06, + "loss": 0.426, + "step": 6792 + }, + { + "epoch": 3.1580660158066016, + "grad_norm": 5.737369060516357, + "learning_rate": 9.993210999321101e-06, + "loss": 0.1342, + "step": 6793 + }, + { + "epoch": 3.1585309158530914, + "grad_norm": 6.799437522888184, + "learning_rate": 9.993209999321002e-06, + "loss": 0.368, + "step": 6794 + }, + { + "epoch": 3.1589958158995817, + "grad_norm": 4.480295181274414, + "learning_rate": 9.9932089993209e-06, + "loss": 0.3284, + "step": 6795 + }, + { + "epoch": 3.1594607159460715, + "grad_norm": 4.30116605758667, + "learning_rate": 9.993207999320801e-06, + "loss": 0.3389, + "step": 6796 + }, + { + "epoch": 3.1599256159925617, + "grad_norm": 6.01076078414917, + "learning_rate": 9.9932069993207e-06, + "loss": 0.4462, + "step": 6797 + }, + { + "epoch": 3.1603905160390515, + "grad_norm": 5.9607014656066895, + "learning_rate": 9.993205999320601e-06, + "loss": 0.3403, + "step": 6798 + }, + { + "epoch": 3.1608554160855418, + "grad_norm": 12.708388328552246, + "learning_rate": 9.9932049993205e-06, + "loss": 0.7848, + "step": 6799 + }, + { + "epoch": 3.1613203161320316, + "grad_norm": 4.6961350440979, + "learning_rate": 9.9932039993204e-06, + "loss": 0.3482, + "step": 6800 + }, + { + "epoch": 3.161785216178522, + "grad_norm": 3.606445550918579, + "learning_rate": 9.993202999320301e-06, + "loss": 0.2016, + "step": 6801 + }, + { + "epoch": 3.1622501162250116, + "grad_norm": 7.878410816192627, + "learning_rate": 9.993201999320202e-06, + "loss": 0.4502, + "step": 6802 + }, + { + "epoch": 3.1627150162715014, + "grad_norm": 5.928516864776611, + "learning_rate": 9.9932009993201e-06, + "loss": 0.4085, + "step": 6803 + }, + { + "epoch": 3.1631799163179917, + "grad_norm": 4.832137107849121, + "learning_rate": 9.993199999320001e-06, + "loss": 0.3251, + "step": 6804 + }, + { + "epoch": 3.1636448163644815, + "grad_norm": 6.499179363250732, + "learning_rate": 9.9931989993199e-06, + "loss": 0.4947, + "step": 6805 + }, + { + "epoch": 3.1641097164109717, + "grad_norm": 5.956722259521484, + "learning_rate": 9.993197999319801e-06, + "loss": 0.2581, + "step": 6806 + }, + { + "epoch": 3.1645746164574615, + "grad_norm": 4.59526252746582, + "learning_rate": 9.993196999319702e-06, + "loss": 0.268, + "step": 6807 + }, + { + "epoch": 3.1650395165039518, + "grad_norm": 6.577779293060303, + "learning_rate": 9.9931959993196e-06, + "loss": 0.4107, + "step": 6808 + }, + { + "epoch": 3.1655044165504416, + "grad_norm": 4.656192302703857, + "learning_rate": 9.993194999319501e-06, + "loss": 0.3125, + "step": 6809 + }, + { + "epoch": 3.165969316596932, + "grad_norm": 4.153679847717285, + "learning_rate": 9.9931939993194e-06, + "loss": 0.1936, + "step": 6810 + }, + { + "epoch": 3.1664342166434216, + "grad_norm": 5.982085227966309, + "learning_rate": 9.993192999319301e-06, + "loss": 0.3388, + "step": 6811 + }, + { + "epoch": 3.166899116689912, + "grad_norm": 6.197073459625244, + "learning_rate": 9.9931919993192e-06, + "loss": 0.3422, + "step": 6812 + }, + { + "epoch": 3.1673640167364017, + "grad_norm": 8.55483627319336, + "learning_rate": 9.9931909993191e-06, + "loss": 0.4913, + "step": 6813 + }, + { + "epoch": 3.1678289167828915, + "grad_norm": 5.201732158660889, + "learning_rate": 9.993189999319001e-06, + "loss": 0.3099, + "step": 6814 + }, + { + "epoch": 3.1682938168293817, + "grad_norm": 6.58852481842041, + "learning_rate": 9.993188999318902e-06, + "loss": 0.7562, + "step": 6815 + }, + { + "epoch": 3.1687587168758715, + "grad_norm": 7.133097171783447, + "learning_rate": 9.9931879993188e-06, + "loss": 0.4156, + "step": 6816 + }, + { + "epoch": 3.1692236169223618, + "grad_norm": 3.0845084190368652, + "learning_rate": 9.9931869993187e-06, + "loss": 0.0759, + "step": 6817 + }, + { + "epoch": 3.1696885169688516, + "grad_norm": 5.318312168121338, + "learning_rate": 9.9931859993186e-06, + "loss": 0.3671, + "step": 6818 + }, + { + "epoch": 3.170153417015342, + "grad_norm": 5.307732105255127, + "learning_rate": 9.993184999318501e-06, + "loss": 0.4075, + "step": 6819 + }, + { + "epoch": 3.1706183170618316, + "grad_norm": 5.554150581359863, + "learning_rate": 9.993183999318402e-06, + "loss": 0.3352, + "step": 6820 + }, + { + "epoch": 3.171083217108322, + "grad_norm": 6.529595375061035, + "learning_rate": 9.9931829993183e-06, + "loss": 0.4902, + "step": 6821 + }, + { + "epoch": 3.1715481171548117, + "grad_norm": 5.1010355949401855, + "learning_rate": 9.993181999318201e-06, + "loss": 0.4009, + "step": 6822 + }, + { + "epoch": 3.172013017201302, + "grad_norm": 4.569424152374268, + "learning_rate": 9.9931809993181e-06, + "loss": 0.3013, + "step": 6823 + }, + { + "epoch": 3.1724779172477917, + "grad_norm": 4.452034950256348, + "learning_rate": 9.993179999318001e-06, + "loss": 0.3952, + "step": 6824 + }, + { + "epoch": 3.172942817294282, + "grad_norm": 6.811328411102295, + "learning_rate": 9.9931789993179e-06, + "loss": 0.3491, + "step": 6825 + }, + { + "epoch": 3.1734077173407718, + "grad_norm": 5.143378257751465, + "learning_rate": 9.9931779993178e-06, + "loss": 0.3154, + "step": 6826 + }, + { + "epoch": 3.1738726173872616, + "grad_norm": 4.065126895904541, + "learning_rate": 9.993176999317701e-06, + "loss": 0.2795, + "step": 6827 + }, + { + "epoch": 3.174337517433752, + "grad_norm": 5.569243907928467, + "learning_rate": 9.9931759993176e-06, + "loss": 0.2541, + "step": 6828 + }, + { + "epoch": 3.1748024174802416, + "grad_norm": 4.122230529785156, + "learning_rate": 9.993174999317501e-06, + "loss": 0.382, + "step": 6829 + }, + { + "epoch": 3.175267317526732, + "grad_norm": 5.281576156616211, + "learning_rate": 9.9931739993174e-06, + "loss": 0.4995, + "step": 6830 + }, + { + "epoch": 3.1757322175732217, + "grad_norm": 4.370349407196045, + "learning_rate": 9.9931729993173e-06, + "loss": 0.2297, + "step": 6831 + }, + { + "epoch": 3.176197117619712, + "grad_norm": 8.512421607971191, + "learning_rate": 9.993171999317201e-06, + "loss": 0.6389, + "step": 6832 + }, + { + "epoch": 3.1766620176662017, + "grad_norm": 4.6678056716918945, + "learning_rate": 9.993170999317102e-06, + "loss": 0.3121, + "step": 6833 + }, + { + "epoch": 3.177126917712692, + "grad_norm": 3.618093252182007, + "learning_rate": 9.993169999317e-06, + "loss": 0.2604, + "step": 6834 + }, + { + "epoch": 3.1775918177591818, + "grad_norm": 6.484758377075195, + "learning_rate": 9.993168999316901e-06, + "loss": 0.3658, + "step": 6835 + }, + { + "epoch": 3.1780567178056716, + "grad_norm": 3.790207862854004, + "learning_rate": 9.9931679993168e-06, + "loss": 0.169, + "step": 6836 + }, + { + "epoch": 3.178521617852162, + "grad_norm": 6.360755443572998, + "learning_rate": 9.9931669993167e-06, + "loss": 0.4589, + "step": 6837 + }, + { + "epoch": 3.1789865178986516, + "grad_norm": 4.891109466552734, + "learning_rate": 9.9931659993166e-06, + "loss": 0.31, + "step": 6838 + }, + { + "epoch": 3.179451417945142, + "grad_norm": 9.53514289855957, + "learning_rate": 9.9931649993165e-06, + "loss": 0.6155, + "step": 6839 + }, + { + "epoch": 3.1799163179916317, + "grad_norm": 4.419282913208008, + "learning_rate": 9.993163999316401e-06, + "loss": 0.2652, + "step": 6840 + }, + { + "epoch": 3.180381218038122, + "grad_norm": 5.4829421043396, + "learning_rate": 9.9931629993163e-06, + "loss": 0.3646, + "step": 6841 + }, + { + "epoch": 3.1808461180846117, + "grad_norm": 6.647782325744629, + "learning_rate": 9.993161999316201e-06, + "loss": 0.7335, + "step": 6842 + }, + { + "epoch": 3.181311018131102, + "grad_norm": 7.30598783493042, + "learning_rate": 9.9931609993161e-06, + "loss": 0.3445, + "step": 6843 + }, + { + "epoch": 3.1817759181775918, + "grad_norm": 6.134942054748535, + "learning_rate": 9.993159999316e-06, + "loss": 0.2431, + "step": 6844 + }, + { + "epoch": 3.182240818224082, + "grad_norm": 3.4310529232025146, + "learning_rate": 9.993158999315901e-06, + "loss": 0.232, + "step": 6845 + }, + { + "epoch": 3.182705718270572, + "grad_norm": 7.366221904754639, + "learning_rate": 9.993157999315802e-06, + "loss": 0.423, + "step": 6846 + }, + { + "epoch": 3.183170618317062, + "grad_norm": 7.887004375457764, + "learning_rate": 9.993156999315701e-06, + "loss": 0.6343, + "step": 6847 + }, + { + "epoch": 3.183635518363552, + "grad_norm": 6.2456841468811035, + "learning_rate": 9.9931559993156e-06, + "loss": 0.4062, + "step": 6848 + }, + { + "epoch": 3.1841004184100417, + "grad_norm": 4.950629234313965, + "learning_rate": 9.9931549993155e-06, + "loss": 0.3869, + "step": 6849 + }, + { + "epoch": 3.184565318456532, + "grad_norm": 8.829743385314941, + "learning_rate": 9.9931539993154e-06, + "loss": 0.5363, + "step": 6850 + }, + { + "epoch": 3.1850302185030217, + "grad_norm": 5.805711269378662, + "learning_rate": 9.9931529993153e-06, + "loss": 0.4992, + "step": 6851 + }, + { + "epoch": 3.185495118549512, + "grad_norm": 4.474420547485352, + "learning_rate": 9.9931519993152e-06, + "loss": 0.4607, + "step": 6852 + }, + { + "epoch": 3.1859600185960018, + "grad_norm": 6.4735426902771, + "learning_rate": 9.993150999315101e-06, + "loss": 0.5225, + "step": 6853 + }, + { + "epoch": 3.186424918642492, + "grad_norm": 4.275472164154053, + "learning_rate": 9.993149999315e-06, + "loss": 0.2714, + "step": 6854 + }, + { + "epoch": 3.186889818688982, + "grad_norm": 5.431915283203125, + "learning_rate": 9.993148999314901e-06, + "loss": 0.3291, + "step": 6855 + }, + { + "epoch": 3.187354718735472, + "grad_norm": 5.241943836212158, + "learning_rate": 9.9931479993148e-06, + "loss": 0.4717, + "step": 6856 + }, + { + "epoch": 3.187819618781962, + "grad_norm": 5.586805820465088, + "learning_rate": 9.9931469993147e-06, + "loss": 0.5157, + "step": 6857 + }, + { + "epoch": 3.1882845188284517, + "grad_norm": 4.58669376373291, + "learning_rate": 9.993145999314601e-06, + "loss": 0.2679, + "step": 6858 + }, + { + "epoch": 3.188749418874942, + "grad_norm": 6.433187961578369, + "learning_rate": 9.9931449993145e-06, + "loss": 0.4046, + "step": 6859 + }, + { + "epoch": 3.1892143189214317, + "grad_norm": 9.940901756286621, + "learning_rate": 9.993143999314401e-06, + "loss": 0.5798, + "step": 6860 + }, + { + "epoch": 3.189679218967922, + "grad_norm": 3.4177775382995605, + "learning_rate": 9.9931429993143e-06, + "loss": 0.1503, + "step": 6861 + }, + { + "epoch": 3.1901441190144117, + "grad_norm": 5.0486907958984375, + "learning_rate": 9.9931419993142e-06, + "loss": 0.5146, + "step": 6862 + }, + { + "epoch": 3.190609019060902, + "grad_norm": 4.5880818367004395, + "learning_rate": 9.993140999314101e-06, + "loss": 0.3046, + "step": 6863 + }, + { + "epoch": 3.191073919107392, + "grad_norm": 5.074894905090332, + "learning_rate": 9.993139999314e-06, + "loss": 0.3795, + "step": 6864 + }, + { + "epoch": 3.191538819153882, + "grad_norm": 8.115262031555176, + "learning_rate": 9.993138999313901e-06, + "loss": 0.2213, + "step": 6865 + }, + { + "epoch": 3.192003719200372, + "grad_norm": 6.894105911254883, + "learning_rate": 9.993137999313802e-06, + "loss": 0.4399, + "step": 6866 + }, + { + "epoch": 3.192468619246862, + "grad_norm": 12.240442276000977, + "learning_rate": 9.9931369993137e-06, + "loss": 0.3977, + "step": 6867 + }, + { + "epoch": 3.192933519293352, + "grad_norm": 6.680671215057373, + "learning_rate": 9.993135999313601e-06, + "loss": 0.4231, + "step": 6868 + }, + { + "epoch": 3.193398419339842, + "grad_norm": 9.278061866760254, + "learning_rate": 9.9931349993135e-06, + "loss": 0.7168, + "step": 6869 + }, + { + "epoch": 3.193863319386332, + "grad_norm": 4.04360294342041, + "learning_rate": 9.9931339993134e-06, + "loss": 0.2941, + "step": 6870 + }, + { + "epoch": 3.1943282194328217, + "grad_norm": 3.7392489910125732, + "learning_rate": 9.993132999313301e-06, + "loss": 0.2094, + "step": 6871 + }, + { + "epoch": 3.194793119479312, + "grad_norm": 4.248105525970459, + "learning_rate": 9.9931319993132e-06, + "loss": 0.2999, + "step": 6872 + }, + { + "epoch": 3.195258019525802, + "grad_norm": 3.8264784812927246, + "learning_rate": 9.993130999313101e-06, + "loss": 0.1871, + "step": 6873 + }, + { + "epoch": 3.195722919572292, + "grad_norm": 7.411715030670166, + "learning_rate": 9.993129999313e-06, + "loss": 0.4173, + "step": 6874 + }, + { + "epoch": 3.196187819618782, + "grad_norm": 5.187680244445801, + "learning_rate": 9.9931289993129e-06, + "loss": 0.2869, + "step": 6875 + }, + { + "epoch": 3.196652719665272, + "grad_norm": 3.9406628608703613, + "learning_rate": 9.993127999312801e-06, + "loss": 0.2812, + "step": 6876 + }, + { + "epoch": 3.197117619711762, + "grad_norm": 3.9507522583007812, + "learning_rate": 9.9931269993127e-06, + "loss": 0.1818, + "step": 6877 + }, + { + "epoch": 3.197582519758252, + "grad_norm": 5.069579124450684, + "learning_rate": 9.993125999312601e-06, + "loss": 0.3433, + "step": 6878 + }, + { + "epoch": 3.198047419804742, + "grad_norm": 6.651803016662598, + "learning_rate": 9.993124999312502e-06, + "loss": 0.7468, + "step": 6879 + }, + { + "epoch": 3.198512319851232, + "grad_norm": 5.779275894165039, + "learning_rate": 9.9931239993124e-06, + "loss": 0.5368, + "step": 6880 + }, + { + "epoch": 3.198977219897722, + "grad_norm": 5.464309215545654, + "learning_rate": 9.9931229993123e-06, + "loss": 0.4693, + "step": 6881 + }, + { + "epoch": 3.199442119944212, + "grad_norm": 7.378155708312988, + "learning_rate": 9.9931219993122e-06, + "loss": 0.7373, + "step": 6882 + }, + { + "epoch": 3.199907019990702, + "grad_norm": 6.544994831085205, + "learning_rate": 9.993120999312101e-06, + "loss": 0.536, + "step": 6883 + }, + { + "epoch": 3.200371920037192, + "grad_norm": 4.757709980010986, + "learning_rate": 9.993119999312002e-06, + "loss": 0.308, + "step": 6884 + }, + { + "epoch": 3.200836820083682, + "grad_norm": 4.180904388427734, + "learning_rate": 9.9931189993119e-06, + "loss": 0.4063, + "step": 6885 + }, + { + "epoch": 3.201301720130172, + "grad_norm": 4.365720272064209, + "learning_rate": 9.993117999311801e-06, + "loss": 0.3514, + "step": 6886 + }, + { + "epoch": 3.201766620176662, + "grad_norm": 4.634856700897217, + "learning_rate": 9.9931169993117e-06, + "loss": 0.3361, + "step": 6887 + }, + { + "epoch": 3.202231520223152, + "grad_norm": 5.499013423919678, + "learning_rate": 9.9931159993116e-06, + "loss": 0.4125, + "step": 6888 + }, + { + "epoch": 3.202696420269642, + "grad_norm": 7.19282865524292, + "learning_rate": 9.993114999311501e-06, + "loss": 0.6003, + "step": 6889 + }, + { + "epoch": 3.203161320316132, + "grad_norm": 5.790989398956299, + "learning_rate": 9.993113999311402e-06, + "loss": 0.3475, + "step": 6890 + }, + { + "epoch": 3.2036262203626222, + "grad_norm": 5.059851169586182, + "learning_rate": 9.993112999311301e-06, + "loss": 0.3675, + "step": 6891 + }, + { + "epoch": 3.204091120409112, + "grad_norm": 4.738916397094727, + "learning_rate": 9.9931119993112e-06, + "loss": 0.4486, + "step": 6892 + }, + { + "epoch": 3.204556020455602, + "grad_norm": 5.804294586181641, + "learning_rate": 9.9931109993111e-06, + "loss": 0.4237, + "step": 6893 + }, + { + "epoch": 3.205020920502092, + "grad_norm": 8.35043716430664, + "learning_rate": 9.993109999311e-06, + "loss": 0.3271, + "step": 6894 + }, + { + "epoch": 3.205485820548582, + "grad_norm": 4.390626430511475, + "learning_rate": 9.9931089993109e-06, + "loss": 0.2751, + "step": 6895 + }, + { + "epoch": 3.205950720595072, + "grad_norm": 8.889118194580078, + "learning_rate": 9.993107999310801e-06, + "loss": 0.8029, + "step": 6896 + }, + { + "epoch": 3.206415620641562, + "grad_norm": 3.8362913131713867, + "learning_rate": 9.993106999310702e-06, + "loss": 0.2096, + "step": 6897 + }, + { + "epoch": 3.206880520688052, + "grad_norm": 4.884056568145752, + "learning_rate": 9.9931059993106e-06, + "loss": 0.3467, + "step": 6898 + }, + { + "epoch": 3.207345420734542, + "grad_norm": 5.8976593017578125, + "learning_rate": 9.993104999310501e-06, + "loss": 0.532, + "step": 6899 + }, + { + "epoch": 3.2078103207810322, + "grad_norm": 6.452750205993652, + "learning_rate": 9.9931039993104e-06, + "loss": 0.3889, + "step": 6900 + }, + { + "epoch": 3.208275220827522, + "grad_norm": 3.4920709133148193, + "learning_rate": 9.9931029993103e-06, + "loss": 0.2423, + "step": 6901 + }, + { + "epoch": 3.2087401208740123, + "grad_norm": 5.981256008148193, + "learning_rate": 9.993101999310201e-06, + "loss": 0.467, + "step": 6902 + }, + { + "epoch": 3.209205020920502, + "grad_norm": 9.109092712402344, + "learning_rate": 9.9931009993101e-06, + "loss": 0.297, + "step": 6903 + }, + { + "epoch": 3.2096699209669923, + "grad_norm": 5.094177722930908, + "learning_rate": 9.993099999310001e-06, + "loss": 0.3569, + "step": 6904 + }, + { + "epoch": 3.210134821013482, + "grad_norm": 5.189803600311279, + "learning_rate": 9.9930989993099e-06, + "loss": 0.3083, + "step": 6905 + }, + { + "epoch": 3.210599721059972, + "grad_norm": 5.464649200439453, + "learning_rate": 9.9930979993098e-06, + "loss": 0.3012, + "step": 6906 + }, + { + "epoch": 3.211064621106462, + "grad_norm": 4.23825740814209, + "learning_rate": 9.9930969993097e-06, + "loss": 0.3767, + "step": 6907 + }, + { + "epoch": 3.211529521152952, + "grad_norm": 8.473114013671875, + "learning_rate": 9.9930959993096e-06, + "loss": 0.3731, + "step": 6908 + }, + { + "epoch": 3.2119944211994422, + "grad_norm": 4.325814247131348, + "learning_rate": 9.993094999309501e-06, + "loss": 0.284, + "step": 6909 + }, + { + "epoch": 3.212459321245932, + "grad_norm": 5.460343360900879, + "learning_rate": 9.993093999309402e-06, + "loss": 0.5377, + "step": 6910 + }, + { + "epoch": 3.2129242212924223, + "grad_norm": 5.167452812194824, + "learning_rate": 9.9930929993093e-06, + "loss": 0.4028, + "step": 6911 + }, + { + "epoch": 3.213389121338912, + "grad_norm": 4.970824718475342, + "learning_rate": 9.993091999309201e-06, + "loss": 0.3226, + "step": 6912 + }, + { + "epoch": 3.2138540213854023, + "grad_norm": 1.934969186782837, + "learning_rate": 9.9930909993091e-06, + "loss": 0.0812, + "step": 6913 + }, + { + "epoch": 3.214318921431892, + "grad_norm": 6.795027732849121, + "learning_rate": 9.993089999309001e-06, + "loss": 0.5647, + "step": 6914 + }, + { + "epoch": 3.214783821478382, + "grad_norm": 5.9535908699035645, + "learning_rate": 9.993088999308902e-06, + "loss": 0.6589, + "step": 6915 + }, + { + "epoch": 3.215248721524872, + "grad_norm": 9.36307144165039, + "learning_rate": 9.9930879993088e-06, + "loss": 0.6349, + "step": 6916 + }, + { + "epoch": 3.215713621571362, + "grad_norm": 4.263976573944092, + "learning_rate": 9.993086999308701e-06, + "loss": 0.2595, + "step": 6917 + }, + { + "epoch": 3.2161785216178522, + "grad_norm": 2.903012752532959, + "learning_rate": 9.9930859993086e-06, + "loss": 0.1529, + "step": 6918 + }, + { + "epoch": 3.216643421664342, + "grad_norm": 4.345538139343262, + "learning_rate": 9.9930849993085e-06, + "loss": 0.3461, + "step": 6919 + }, + { + "epoch": 3.2171083217108323, + "grad_norm": 5.157414436340332, + "learning_rate": 9.9930839993084e-06, + "loss": 0.3101, + "step": 6920 + }, + { + "epoch": 3.217573221757322, + "grad_norm": 5.145853042602539, + "learning_rate": 9.9930829993083e-06, + "loss": 0.3319, + "step": 6921 + }, + { + "epoch": 3.2180381218038123, + "grad_norm": 4.123612403869629, + "learning_rate": 9.993081999308201e-06, + "loss": 0.1825, + "step": 6922 + }, + { + "epoch": 3.218503021850302, + "grad_norm": 6.370307922363281, + "learning_rate": 9.993080999308102e-06, + "loss": 0.2398, + "step": 6923 + }, + { + "epoch": 3.2189679218967924, + "grad_norm": 5.6264753341674805, + "learning_rate": 9.993079999308e-06, + "loss": 0.2967, + "step": 6924 + }, + { + "epoch": 3.219432821943282, + "grad_norm": 4.952605724334717, + "learning_rate": 9.9930789993079e-06, + "loss": 0.231, + "step": 6925 + }, + { + "epoch": 3.2198977219897724, + "grad_norm": 4.521495342254639, + "learning_rate": 9.9930779993078e-06, + "loss": 0.3235, + "step": 6926 + }, + { + "epoch": 3.2203626220362622, + "grad_norm": 3.773470163345337, + "learning_rate": 9.993076999307701e-06, + "loss": 0.1673, + "step": 6927 + }, + { + "epoch": 3.220827522082752, + "grad_norm": 5.192686080932617, + "learning_rate": 9.993075999307602e-06, + "loss": 0.283, + "step": 6928 + }, + { + "epoch": 3.2212924221292423, + "grad_norm": 7.695489406585693, + "learning_rate": 9.9930749993075e-06, + "loss": 0.2548, + "step": 6929 + }, + { + "epoch": 3.221757322175732, + "grad_norm": 2.8902621269226074, + "learning_rate": 9.993073999307401e-06, + "loss": 0.1194, + "step": 6930 + }, + { + "epoch": 3.2222222222222223, + "grad_norm": 3.8391528129577637, + "learning_rate": 9.9930729993073e-06, + "loss": 0.2212, + "step": 6931 + }, + { + "epoch": 3.222687122268712, + "grad_norm": 5.953438758850098, + "learning_rate": 9.993071999307201e-06, + "loss": 0.382, + "step": 6932 + }, + { + "epoch": 3.2231520223152024, + "grad_norm": 6.928483963012695, + "learning_rate": 9.9930709993071e-06, + "loss": 0.5972, + "step": 6933 + }, + { + "epoch": 3.223616922361692, + "grad_norm": 4.361401081085205, + "learning_rate": 9.993069999307e-06, + "loss": 0.3013, + "step": 6934 + }, + { + "epoch": 3.2240818224081824, + "grad_norm": 5.122338771820068, + "learning_rate": 9.993068999306901e-06, + "loss": 0.2943, + "step": 6935 + }, + { + "epoch": 3.224546722454672, + "grad_norm": 6.669522285461426, + "learning_rate": 9.9930679993068e-06, + "loss": 0.5075, + "step": 6936 + }, + { + "epoch": 3.225011622501162, + "grad_norm": 6.5787858963012695, + "learning_rate": 9.9930669993067e-06, + "loss": 0.4488, + "step": 6937 + }, + { + "epoch": 3.2254765225476523, + "grad_norm": 3.6072895526885986, + "learning_rate": 9.9930659993066e-06, + "loss": 0.2071, + "step": 6938 + }, + { + "epoch": 3.225941422594142, + "grad_norm": 7.4755964279174805, + "learning_rate": 9.9930649993065e-06, + "loss": 0.5631, + "step": 6939 + }, + { + "epoch": 3.2264063226406323, + "grad_norm": 5.190778732299805, + "learning_rate": 9.993063999306401e-06, + "loss": 0.5085, + "step": 6940 + }, + { + "epoch": 3.226871222687122, + "grad_norm": 3.449674129486084, + "learning_rate": 9.993062999306302e-06, + "loss": 0.1576, + "step": 6941 + }, + { + "epoch": 3.2273361227336124, + "grad_norm": 6.808865070343018, + "learning_rate": 9.9930619993062e-06, + "loss": 0.4797, + "step": 6942 + }, + { + "epoch": 3.227801022780102, + "grad_norm": 4.6402506828308105, + "learning_rate": 9.993060999306101e-06, + "loss": 0.337, + "step": 6943 + }, + { + "epoch": 3.2282659228265924, + "grad_norm": 6.648484706878662, + "learning_rate": 9.993059999306e-06, + "loss": 0.4245, + "step": 6944 + }, + { + "epoch": 3.228730822873082, + "grad_norm": 3.1401543617248535, + "learning_rate": 9.993058999305901e-06, + "loss": 0.1323, + "step": 6945 + }, + { + "epoch": 3.2291957229195725, + "grad_norm": 5.389408111572266, + "learning_rate": 9.9930579993058e-06, + "loss": 0.4663, + "step": 6946 + }, + { + "epoch": 3.2296606229660623, + "grad_norm": 4.845921039581299, + "learning_rate": 9.9930569993057e-06, + "loss": 0.4491, + "step": 6947 + }, + { + "epoch": 3.2301255230125525, + "grad_norm": 7.609282970428467, + "learning_rate": 9.993055999305601e-06, + "loss": 0.4048, + "step": 6948 + }, + { + "epoch": 3.2305904230590423, + "grad_norm": 4.963563919067383, + "learning_rate": 9.9930549993055e-06, + "loss": 0.4054, + "step": 6949 + }, + { + "epoch": 3.231055323105532, + "grad_norm": 8.165877342224121, + "learning_rate": 9.993053999305401e-06, + "loss": 0.8201, + "step": 6950 + }, + { + "epoch": 3.2315202231520224, + "grad_norm": 3.4888994693756104, + "learning_rate": 9.9930529993053e-06, + "loss": 0.2338, + "step": 6951 + }, + { + "epoch": 3.231985123198512, + "grad_norm": 4.559362411499023, + "learning_rate": 9.9930519993052e-06, + "loss": 0.424, + "step": 6952 + }, + { + "epoch": 3.2324500232450024, + "grad_norm": 2.078925132751465, + "learning_rate": 9.993050999305101e-06, + "loss": 0.0952, + "step": 6953 + }, + { + "epoch": 3.232914923291492, + "grad_norm": 4.824130058288574, + "learning_rate": 9.993049999305002e-06, + "loss": 0.3237, + "step": 6954 + }, + { + "epoch": 3.2333798233379825, + "grad_norm": 5.023068428039551, + "learning_rate": 9.9930489993049e-06, + "loss": 0.458, + "step": 6955 + }, + { + "epoch": 3.2338447233844723, + "grad_norm": 5.5444746017456055, + "learning_rate": 9.993047999304801e-06, + "loss": 0.3544, + "step": 6956 + }, + { + "epoch": 3.2343096234309625, + "grad_norm": 4.122495174407959, + "learning_rate": 9.9930469993047e-06, + "loss": 0.2027, + "step": 6957 + }, + { + "epoch": 3.2347745234774523, + "grad_norm": 6.011197566986084, + "learning_rate": 9.993045999304601e-06, + "loss": 0.4022, + "step": 6958 + }, + { + "epoch": 3.2352394235239426, + "grad_norm": 4.305301666259766, + "learning_rate": 9.993044999304502e-06, + "loss": 0.1498, + "step": 6959 + }, + { + "epoch": 3.2357043235704324, + "grad_norm": 7.256709575653076, + "learning_rate": 9.9930439993044e-06, + "loss": 0.5093, + "step": 6960 + }, + { + "epoch": 3.236169223616922, + "grad_norm": 6.202186107635498, + "learning_rate": 9.993042999304301e-06, + "loss": 0.3288, + "step": 6961 + }, + { + "epoch": 3.2366341236634124, + "grad_norm": 9.68070125579834, + "learning_rate": 9.9930419993042e-06, + "loss": 0.4071, + "step": 6962 + }, + { + "epoch": 3.237099023709902, + "grad_norm": 4.52505350112915, + "learning_rate": 9.993040999304101e-06, + "loss": 0.3003, + "step": 6963 + }, + { + "epoch": 3.2375639237563925, + "grad_norm": 6.138084888458252, + "learning_rate": 9.993039999304e-06, + "loss": 0.3813, + "step": 6964 + }, + { + "epoch": 3.2380288238028823, + "grad_norm": 8.11023998260498, + "learning_rate": 9.9930389993039e-06, + "loss": 0.4004, + "step": 6965 + }, + { + "epoch": 3.2384937238493725, + "grad_norm": 6.743671417236328, + "learning_rate": 9.993037999303801e-06, + "loss": 0.6266, + "step": 6966 + }, + { + "epoch": 3.2389586238958623, + "grad_norm": 3.940708637237549, + "learning_rate": 9.993036999303702e-06, + "loss": 0.2727, + "step": 6967 + }, + { + "epoch": 3.2394235239423526, + "grad_norm": 4.581711292266846, + "learning_rate": 9.993035999303601e-06, + "loss": 0.3341, + "step": 6968 + }, + { + "epoch": 3.2398884239888424, + "grad_norm": 9.34502124786377, + "learning_rate": 9.9930349993035e-06, + "loss": 0.403, + "step": 6969 + }, + { + "epoch": 3.2403533240353326, + "grad_norm": 3.5180065631866455, + "learning_rate": 9.9930339993034e-06, + "loss": 0.2005, + "step": 6970 + }, + { + "epoch": 3.2408182240818224, + "grad_norm": 4.969961166381836, + "learning_rate": 9.993032999303301e-06, + "loss": 0.3547, + "step": 6971 + }, + { + "epoch": 3.241283124128312, + "grad_norm": 9.843235969543457, + "learning_rate": 9.993031999303202e-06, + "loss": 0.5333, + "step": 6972 + }, + { + "epoch": 3.2417480241748025, + "grad_norm": 5.6974406242370605, + "learning_rate": 9.9930309993031e-06, + "loss": 0.4374, + "step": 6973 + }, + { + "epoch": 3.2422129242212923, + "grad_norm": 5.926609516143799, + "learning_rate": 9.993029999303001e-06, + "loss": 0.3951, + "step": 6974 + }, + { + "epoch": 3.2426778242677825, + "grad_norm": 7.764004230499268, + "learning_rate": 9.9930289993029e-06, + "loss": 0.4355, + "step": 6975 + }, + { + "epoch": 3.2431427243142723, + "grad_norm": 4.0907511711120605, + "learning_rate": 9.993027999302801e-06, + "loss": 0.2228, + "step": 6976 + }, + { + "epoch": 3.2436076243607626, + "grad_norm": 3.8172504901885986, + "learning_rate": 9.9930269993027e-06, + "loss": 0.2435, + "step": 6977 + }, + { + "epoch": 3.2440725244072524, + "grad_norm": 5.676449775695801, + "learning_rate": 9.9930259993026e-06, + "loss": 0.4978, + "step": 6978 + }, + { + "epoch": 3.2445374244537426, + "grad_norm": 5.947616100311279, + "learning_rate": 9.993024999302501e-06, + "loss": 0.3539, + "step": 6979 + }, + { + "epoch": 3.2450023245002324, + "grad_norm": 4.995340347290039, + "learning_rate": 9.9930239993024e-06, + "loss": 0.357, + "step": 6980 + }, + { + "epoch": 3.2454672245467227, + "grad_norm": 4.144567489624023, + "learning_rate": 9.993022999302301e-06, + "loss": 0.2858, + "step": 6981 + }, + { + "epoch": 3.2459321245932125, + "grad_norm": 5.37592887878418, + "learning_rate": 9.9930219993022e-06, + "loss": 0.3808, + "step": 6982 + }, + { + "epoch": 3.2463970246397027, + "grad_norm": 6.6671953201293945, + "learning_rate": 9.9930209993021e-06, + "loss": 0.5996, + "step": 6983 + }, + { + "epoch": 3.2468619246861925, + "grad_norm": 7.265436172485352, + "learning_rate": 9.993019999302001e-06, + "loss": 0.3559, + "step": 6984 + }, + { + "epoch": 3.2473268247326823, + "grad_norm": 4.155337333679199, + "learning_rate": 9.993018999301902e-06, + "loss": 0.3335, + "step": 6985 + }, + { + "epoch": 3.2477917247791726, + "grad_norm": 4.344486713409424, + "learning_rate": 9.993017999301801e-06, + "loss": 0.2454, + "step": 6986 + }, + { + "epoch": 3.2482566248256624, + "grad_norm": 6.268126964569092, + "learning_rate": 9.993016999301702e-06, + "loss": 0.3836, + "step": 6987 + }, + { + "epoch": 3.2487215248721526, + "grad_norm": 6.198549747467041, + "learning_rate": 9.9930159993016e-06, + "loss": 0.334, + "step": 6988 + }, + { + "epoch": 3.2491864249186424, + "grad_norm": 7.105047225952148, + "learning_rate": 9.9930149993015e-06, + "loss": 0.507, + "step": 6989 + }, + { + "epoch": 3.2496513249651326, + "grad_norm": 3.4613425731658936, + "learning_rate": 9.9930139993014e-06, + "loss": 0.1916, + "step": 6990 + }, + { + "epoch": 3.2501162250116225, + "grad_norm": 5.602560997009277, + "learning_rate": 9.9930129993013e-06, + "loss": 0.277, + "step": 6991 + }, + { + "epoch": 3.2505811250581127, + "grad_norm": 8.649177551269531, + "learning_rate": 9.993011999301201e-06, + "loss": 0.4844, + "step": 6992 + }, + { + "epoch": 3.2510460251046025, + "grad_norm": 7.901052474975586, + "learning_rate": 9.9930109993011e-06, + "loss": 0.6648, + "step": 6993 + }, + { + "epoch": 3.2515109251510923, + "grad_norm": 4.134114742279053, + "learning_rate": 9.993009999301001e-06, + "loss": 0.3097, + "step": 6994 + }, + { + "epoch": 3.2519758251975825, + "grad_norm": 4.311406135559082, + "learning_rate": 9.9930089993009e-06, + "loss": 0.3713, + "step": 6995 + }, + { + "epoch": 3.2524407252440724, + "grad_norm": 6.801614284515381, + "learning_rate": 9.9930079993008e-06, + "loss": 0.5511, + "step": 6996 + }, + { + "epoch": 3.2529056252905626, + "grad_norm": 7.714344024658203, + "learning_rate": 9.993006999300701e-06, + "loss": 0.5081, + "step": 6997 + }, + { + "epoch": 3.2533705253370524, + "grad_norm": 5.966427803039551, + "learning_rate": 9.993005999300602e-06, + "loss": 0.3727, + "step": 6998 + }, + { + "epoch": 3.2538354253835426, + "grad_norm": 2.4910175800323486, + "learning_rate": 9.993004999300501e-06, + "loss": 0.082, + "step": 6999 + }, + { + "epoch": 3.2543003254300324, + "grad_norm": 7.61922550201416, + "learning_rate": 9.9930039993004e-06, + "loss": 0.6636, + "step": 7000 + }, + { + "epoch": 3.2547652254765227, + "grad_norm": 4.820358753204346, + "learning_rate": 9.9930029993003e-06, + "loss": 0.4298, + "step": 7001 + }, + { + "epoch": 3.2552301255230125, + "grad_norm": 3.953868865966797, + "learning_rate": 9.9930019993002e-06, + "loss": 0.1517, + "step": 7002 + }, + { + "epoch": 3.2556950255695027, + "grad_norm": 4.425828456878662, + "learning_rate": 9.9930009993001e-06, + "loss": 0.3554, + "step": 7003 + }, + { + "epoch": 3.2561599256159925, + "grad_norm": 5.382966995239258, + "learning_rate": 9.9929999993e-06, + "loss": 0.2964, + "step": 7004 + }, + { + "epoch": 3.256624825662483, + "grad_norm": 2.778439521789551, + "learning_rate": 9.992998999299901e-06, + "loss": 0.1266, + "step": 7005 + }, + { + "epoch": 3.2570897257089726, + "grad_norm": 5.036290645599365, + "learning_rate": 9.9929979992998e-06, + "loss": 0.282, + "step": 7006 + }, + { + "epoch": 3.2575546257554624, + "grad_norm": 8.429975509643555, + "learning_rate": 9.992996999299701e-06, + "loss": 0.2657, + "step": 7007 + }, + { + "epoch": 3.2580195258019526, + "grad_norm": 6.234624862670898, + "learning_rate": 9.9929959992996e-06, + "loss": 0.4381, + "step": 7008 + }, + { + "epoch": 3.2584844258484424, + "grad_norm": 5.573365688323975, + "learning_rate": 9.9929949992995e-06, + "loss": 0.4443, + "step": 7009 + }, + { + "epoch": 3.2589493258949327, + "grad_norm": 4.633677005767822, + "learning_rate": 9.992993999299401e-06, + "loss": 0.3262, + "step": 7010 + }, + { + "epoch": 3.2594142259414225, + "grad_norm": 6.045985221862793, + "learning_rate": 9.9929929992993e-06, + "loss": 0.4317, + "step": 7011 + }, + { + "epoch": 3.2598791259879127, + "grad_norm": 9.68837833404541, + "learning_rate": 9.992991999299201e-06, + "loss": 0.7301, + "step": 7012 + }, + { + "epoch": 3.2603440260344025, + "grad_norm": 7.632606506347656, + "learning_rate": 9.9929909992991e-06, + "loss": 0.545, + "step": 7013 + }, + { + "epoch": 3.260808926080893, + "grad_norm": 6.901191711425781, + "learning_rate": 9.992989999299e-06, + "loss": 0.3025, + "step": 7014 + }, + { + "epoch": 3.2612738261273826, + "grad_norm": 5.18479061126709, + "learning_rate": 9.9929889992989e-06, + "loss": 0.4363, + "step": 7015 + }, + { + "epoch": 3.2617387261738724, + "grad_norm": 4.71715784072876, + "learning_rate": 9.9929879992988e-06, + "loss": 0.2606, + "step": 7016 + }, + { + "epoch": 3.2622036262203626, + "grad_norm": 7.022663116455078, + "learning_rate": 9.992986999298701e-06, + "loss": 0.3016, + "step": 7017 + }, + { + "epoch": 3.2626685262668524, + "grad_norm": 8.71817398071289, + "learning_rate": 9.992985999298602e-06, + "loss": 0.8255, + "step": 7018 + }, + { + "epoch": 3.2631334263133427, + "grad_norm": 4.846087455749512, + "learning_rate": 9.9929849992985e-06, + "loss": 0.295, + "step": 7019 + }, + { + "epoch": 3.2635983263598325, + "grad_norm": 7.973678112030029, + "learning_rate": 9.992983999298401e-06, + "loss": 0.3077, + "step": 7020 + }, + { + "epoch": 3.2640632264063227, + "grad_norm": 3.770050525665283, + "learning_rate": 9.9929829992983e-06, + "loss": 0.1948, + "step": 7021 + }, + { + "epoch": 3.2645281264528125, + "grad_norm": 5.503271102905273, + "learning_rate": 9.9929819992982e-06, + "loss": 0.42, + "step": 7022 + }, + { + "epoch": 3.264993026499303, + "grad_norm": 5.57569694519043, + "learning_rate": 9.992980999298101e-06, + "loss": 0.4809, + "step": 7023 + }, + { + "epoch": 3.2654579265457926, + "grad_norm": 7.73209285736084, + "learning_rate": 9.992979999298e-06, + "loss": 0.4791, + "step": 7024 + }, + { + "epoch": 3.265922826592283, + "grad_norm": 6.191632270812988, + "learning_rate": 9.992978999297901e-06, + "loss": 0.5342, + "step": 7025 + }, + { + "epoch": 3.2663877266387726, + "grad_norm": 5.3311238288879395, + "learning_rate": 9.9929779992978e-06, + "loss": 0.332, + "step": 7026 + }, + { + "epoch": 3.266852626685263, + "grad_norm": 7.018301486968994, + "learning_rate": 9.9929769992977e-06, + "loss": 0.5311, + "step": 7027 + }, + { + "epoch": 3.2673175267317527, + "grad_norm": 3.5040788650512695, + "learning_rate": 9.992975999297601e-06, + "loss": 0.1421, + "step": 7028 + }, + { + "epoch": 3.2677824267782425, + "grad_norm": 6.696314811706543, + "learning_rate": 9.9929749992975e-06, + "loss": 0.3438, + "step": 7029 + }, + { + "epoch": 3.2682473268247327, + "grad_norm": 6.502747535705566, + "learning_rate": 9.992973999297401e-06, + "loss": 0.5685, + "step": 7030 + }, + { + "epoch": 3.2687122268712225, + "grad_norm": 4.715396881103516, + "learning_rate": 9.992972999297302e-06, + "loss": 0.3499, + "step": 7031 + }, + { + "epoch": 3.269177126917713, + "grad_norm": 5.480454444885254, + "learning_rate": 9.9929719992972e-06, + "loss": 0.2899, + "step": 7032 + }, + { + "epoch": 3.2696420269642026, + "grad_norm": 6.629595756530762, + "learning_rate": 9.9929709992971e-06, + "loss": 0.5902, + "step": 7033 + }, + { + "epoch": 3.270106927010693, + "grad_norm": 8.413932800292969, + "learning_rate": 9.992969999297e-06, + "loss": 0.558, + "step": 7034 + }, + { + "epoch": 3.2705718270571826, + "grad_norm": 6.275195598602295, + "learning_rate": 9.992968999296901e-06, + "loss": 0.5055, + "step": 7035 + }, + { + "epoch": 3.271036727103673, + "grad_norm": 4.448215484619141, + "learning_rate": 9.992967999296802e-06, + "loss": 0.3381, + "step": 7036 + }, + { + "epoch": 3.2715016271501627, + "grad_norm": 4.521892070770264, + "learning_rate": 9.9929669992967e-06, + "loss": 0.3232, + "step": 7037 + }, + { + "epoch": 3.2719665271966525, + "grad_norm": 4.856974124908447, + "learning_rate": 9.992965999296601e-06, + "loss": 0.2178, + "step": 7038 + }, + { + "epoch": 3.2724314272431427, + "grad_norm": 4.43786096572876, + "learning_rate": 9.9929649992965e-06, + "loss": 0.2677, + "step": 7039 + }, + { + "epoch": 3.272896327289633, + "grad_norm": 4.976840496063232, + "learning_rate": 9.9929639992964e-06, + "loss": 0.368, + "step": 7040 + }, + { + "epoch": 3.273361227336123, + "grad_norm": 5.820474147796631, + "learning_rate": 9.992962999296301e-06, + "loss": 0.3323, + "step": 7041 + }, + { + "epoch": 3.2738261273826126, + "grad_norm": 4.846644401550293, + "learning_rate": 9.9929619992962e-06, + "loss": 0.2981, + "step": 7042 + }, + { + "epoch": 3.274291027429103, + "grad_norm": 8.071024894714355, + "learning_rate": 9.992960999296101e-06, + "loss": 0.3266, + "step": 7043 + }, + { + "epoch": 3.2747559274755926, + "grad_norm": 3.652794122695923, + "learning_rate": 9.992959999296e-06, + "loss": 0.1633, + "step": 7044 + }, + { + "epoch": 3.275220827522083, + "grad_norm": 5.793461799621582, + "learning_rate": 9.9929589992959e-06, + "loss": 0.492, + "step": 7045 + }, + { + "epoch": 3.2756857275685727, + "grad_norm": 5.986874580383301, + "learning_rate": 9.9929579992958e-06, + "loss": 0.363, + "step": 7046 + }, + { + "epoch": 3.276150627615063, + "grad_norm": 4.464538097381592, + "learning_rate": 9.9929569992957e-06, + "loss": 0.2879, + "step": 7047 + }, + { + "epoch": 3.2766155276615527, + "grad_norm": 5.350561618804932, + "learning_rate": 9.992955999295601e-06, + "loss": 0.4152, + "step": 7048 + }, + { + "epoch": 3.277080427708043, + "grad_norm": 6.069313049316406, + "learning_rate": 9.992954999295502e-06, + "loss": 0.3526, + "step": 7049 + }, + { + "epoch": 3.2775453277545328, + "grad_norm": 8.807419776916504, + "learning_rate": 9.9929539992954e-06, + "loss": 0.6115, + "step": 7050 + }, + { + "epoch": 3.2780102278010226, + "grad_norm": 6.615990161895752, + "learning_rate": 9.992952999295301e-06, + "loss": 0.4091, + "step": 7051 + }, + { + "epoch": 3.278475127847513, + "grad_norm": 4.9589619636535645, + "learning_rate": 9.9929519992952e-06, + "loss": 0.2743, + "step": 7052 + }, + { + "epoch": 3.2789400278940026, + "grad_norm": 8.457770347595215, + "learning_rate": 9.992950999295101e-06, + "loss": 0.4484, + "step": 7053 + }, + { + "epoch": 3.279404927940493, + "grad_norm": 4.359564304351807, + "learning_rate": 9.992949999295002e-06, + "loss": 0.2717, + "step": 7054 + }, + { + "epoch": 3.2798698279869827, + "grad_norm": 5.892794609069824, + "learning_rate": 9.9929489992949e-06, + "loss": 0.6357, + "step": 7055 + }, + { + "epoch": 3.280334728033473, + "grad_norm": 7.855461597442627, + "learning_rate": 9.992947999294801e-06, + "loss": 0.2997, + "step": 7056 + }, + { + "epoch": 3.2807996280799627, + "grad_norm": 6.876903057098389, + "learning_rate": 9.9929469992947e-06, + "loss": 0.259, + "step": 7057 + }, + { + "epoch": 3.281264528126453, + "grad_norm": 2.9605610370635986, + "learning_rate": 9.9929459992946e-06, + "loss": 0.1016, + "step": 7058 + }, + { + "epoch": 3.2817294281729428, + "grad_norm": 4.2991437911987305, + "learning_rate": 9.9929449992945e-06, + "loss": 0.2906, + "step": 7059 + }, + { + "epoch": 3.2821943282194326, + "grad_norm": 5.183125972747803, + "learning_rate": 9.9929439992944e-06, + "loss": 0.4401, + "step": 7060 + }, + { + "epoch": 3.282659228265923, + "grad_norm": 4.247367858886719, + "learning_rate": 9.992942999294301e-06, + "loss": 0.2631, + "step": 7061 + }, + { + "epoch": 3.283124128312413, + "grad_norm": 7.154563903808594, + "learning_rate": 9.992941999294202e-06, + "loss": 0.4872, + "step": 7062 + }, + { + "epoch": 3.283589028358903, + "grad_norm": 5.335598945617676, + "learning_rate": 9.9929409992941e-06, + "loss": 0.5016, + "step": 7063 + }, + { + "epoch": 3.2840539284053927, + "grad_norm": 3.393186330795288, + "learning_rate": 9.992939999294001e-06, + "loss": 0.2624, + "step": 7064 + }, + { + "epoch": 3.284518828451883, + "grad_norm": 5.064007759094238, + "learning_rate": 9.9929389992939e-06, + "loss": 0.3214, + "step": 7065 + }, + { + "epoch": 3.2849837284983727, + "grad_norm": 8.812816619873047, + "learning_rate": 9.992937999293801e-06, + "loss": 0.3717, + "step": 7066 + }, + { + "epoch": 3.285448628544863, + "grad_norm": 5.000180244445801, + "learning_rate": 9.992936999293702e-06, + "loss": 0.3399, + "step": 7067 + }, + { + "epoch": 3.2859135285913528, + "grad_norm": 8.624418258666992, + "learning_rate": 9.9929359992936e-06, + "loss": 0.5136, + "step": 7068 + }, + { + "epoch": 3.286378428637843, + "grad_norm": 4.992156982421875, + "learning_rate": 9.992934999293501e-06, + "loss": 0.2585, + "step": 7069 + }, + { + "epoch": 3.286843328684333, + "grad_norm": 4.879831790924072, + "learning_rate": 9.9929339992934e-06, + "loss": 0.3556, + "step": 7070 + }, + { + "epoch": 3.287308228730823, + "grad_norm": 5.8806328773498535, + "learning_rate": 9.992932999293301e-06, + "loss": 0.3026, + "step": 7071 + }, + { + "epoch": 3.287773128777313, + "grad_norm": 6.108302116394043, + "learning_rate": 9.9929319992932e-06, + "loss": 0.3299, + "step": 7072 + }, + { + "epoch": 3.2882380288238027, + "grad_norm": 4.369602680206299, + "learning_rate": 9.9929309992931e-06, + "loss": 0.3497, + "step": 7073 + }, + { + "epoch": 3.288702928870293, + "grad_norm": 4.552048683166504, + "learning_rate": 9.992929999293001e-06, + "loss": 0.2657, + "step": 7074 + }, + { + "epoch": 3.2891678289167827, + "grad_norm": 7.007179260253906, + "learning_rate": 9.992928999292902e-06, + "loss": 0.4884, + "step": 7075 + }, + { + "epoch": 3.289632728963273, + "grad_norm": 6.648036479949951, + "learning_rate": 9.9929279992928e-06, + "loss": 0.8325, + "step": 7076 + }, + { + "epoch": 3.2900976290097628, + "grad_norm": 5.164874076843262, + "learning_rate": 9.9929269992927e-06, + "loss": 0.2191, + "step": 7077 + }, + { + "epoch": 3.290562529056253, + "grad_norm": 4.142219543457031, + "learning_rate": 9.9929259992926e-06, + "loss": 0.2999, + "step": 7078 + }, + { + "epoch": 3.291027429102743, + "grad_norm": 7.885228633880615, + "learning_rate": 9.992924999292501e-06, + "loss": 0.2938, + "step": 7079 + }, + { + "epoch": 3.291492329149233, + "grad_norm": 3.969054937362671, + "learning_rate": 9.992923999292402e-06, + "loss": 0.1248, + "step": 7080 + }, + { + "epoch": 3.291957229195723, + "grad_norm": 5.645511150360107, + "learning_rate": 9.9929229992923e-06, + "loss": 0.2757, + "step": 7081 + }, + { + "epoch": 3.292422129242213, + "grad_norm": 4.42717170715332, + "learning_rate": 9.992921999292201e-06, + "loss": 0.3227, + "step": 7082 + }, + { + "epoch": 3.292887029288703, + "grad_norm": 6.890261173248291, + "learning_rate": 9.9929209992921e-06, + "loss": 0.4175, + "step": 7083 + }, + { + "epoch": 3.293351929335193, + "grad_norm": 15.088212966918945, + "learning_rate": 9.992919999292001e-06, + "loss": 0.7984, + "step": 7084 + }, + { + "epoch": 3.293816829381683, + "grad_norm": 8.462029457092285, + "learning_rate": 9.9929189992919e-06, + "loss": 0.3388, + "step": 7085 + }, + { + "epoch": 3.2942817294281728, + "grad_norm": 5.793883800506592, + "learning_rate": 9.9929179992918e-06, + "loss": 0.2698, + "step": 7086 + }, + { + "epoch": 3.294746629474663, + "grad_norm": 5.529741287231445, + "learning_rate": 9.992916999291701e-06, + "loss": 0.4375, + "step": 7087 + }, + { + "epoch": 3.295211529521153, + "grad_norm": 5.274472713470459, + "learning_rate": 9.9929159992916e-06, + "loss": 0.4267, + "step": 7088 + }, + { + "epoch": 3.295676429567643, + "grad_norm": 3.4059345722198486, + "learning_rate": 9.9929149992915e-06, + "loss": 0.2055, + "step": 7089 + }, + { + "epoch": 3.296141329614133, + "grad_norm": 4.581259250640869, + "learning_rate": 9.9929139992914e-06, + "loss": 0.2969, + "step": 7090 + }, + { + "epoch": 3.296606229660623, + "grad_norm": 4.673484802246094, + "learning_rate": 9.9929129992913e-06, + "loss": 0.3069, + "step": 7091 + }, + { + "epoch": 3.297071129707113, + "grad_norm": 6.060430526733398, + "learning_rate": 9.992911999291201e-06, + "loss": 0.4168, + "step": 7092 + }, + { + "epoch": 3.297536029753603, + "grad_norm": 9.532958030700684, + "learning_rate": 9.992910999291102e-06, + "loss": 0.4241, + "step": 7093 + }, + { + "epoch": 3.298000929800093, + "grad_norm": 6.807565689086914, + "learning_rate": 9.992909999291e-06, + "loss": 0.3647, + "step": 7094 + }, + { + "epoch": 3.2984658298465828, + "grad_norm": 6.15658712387085, + "learning_rate": 9.992908999290901e-06, + "loss": 0.3455, + "step": 7095 + }, + { + "epoch": 3.298930729893073, + "grad_norm": 4.5608062744140625, + "learning_rate": 9.9929079992908e-06, + "loss": 0.3349, + "step": 7096 + }, + { + "epoch": 3.299395629939563, + "grad_norm": 5.84521484375, + "learning_rate": 9.992906999290701e-06, + "loss": 0.4849, + "step": 7097 + }, + { + "epoch": 3.299860529986053, + "grad_norm": 7.781924247741699, + "learning_rate": 9.9929059992906e-06, + "loss": 0.3911, + "step": 7098 + }, + { + "epoch": 3.300325430032543, + "grad_norm": 18.889545440673828, + "learning_rate": 9.9929049992905e-06, + "loss": 1.0333, + "step": 7099 + }, + { + "epoch": 3.300790330079033, + "grad_norm": 4.7282915115356445, + "learning_rate": 9.992903999290401e-06, + "loss": 0.3144, + "step": 7100 + }, + { + "epoch": 3.301255230125523, + "grad_norm": 6.872983455657959, + "learning_rate": 9.9929029992903e-06, + "loss": 0.9193, + "step": 7101 + }, + { + "epoch": 3.301720130172013, + "grad_norm": 5.8271684646606445, + "learning_rate": 9.992901999290201e-06, + "loss": 0.468, + "step": 7102 + }, + { + "epoch": 3.302185030218503, + "grad_norm": 3.556776762008667, + "learning_rate": 9.9929009992901e-06, + "loss": 0.213, + "step": 7103 + }, + { + "epoch": 3.302649930264993, + "grad_norm": 4.69654655456543, + "learning_rate": 9.99289999929e-06, + "loss": 0.4238, + "step": 7104 + }, + { + "epoch": 3.303114830311483, + "grad_norm": 5.7100510597229, + "learning_rate": 9.992898999289901e-06, + "loss": 0.3677, + "step": 7105 + }, + { + "epoch": 3.3035797303579733, + "grad_norm": 4.508598804473877, + "learning_rate": 9.992897999289802e-06, + "loss": 0.2359, + "step": 7106 + }, + { + "epoch": 3.304044630404463, + "grad_norm": 2.726353645324707, + "learning_rate": 9.9928969992897e-06, + "loss": 0.1485, + "step": 7107 + }, + { + "epoch": 3.304509530450953, + "grad_norm": 4.946287631988525, + "learning_rate": 9.992895999289601e-06, + "loss": 0.3042, + "step": 7108 + }, + { + "epoch": 3.304974430497443, + "grad_norm": 4.6941118240356445, + "learning_rate": 9.9928949992895e-06, + "loss": 0.2756, + "step": 7109 + }, + { + "epoch": 3.305439330543933, + "grad_norm": 5.053120136260986, + "learning_rate": 9.992893999289401e-06, + "loss": 0.3897, + "step": 7110 + }, + { + "epoch": 3.305904230590423, + "grad_norm": 5.22045373916626, + "learning_rate": 9.9928929992893e-06, + "loss": 0.3191, + "step": 7111 + }, + { + "epoch": 3.306369130636913, + "grad_norm": 5.393928527832031, + "learning_rate": 9.9928919992892e-06, + "loss": 0.3244, + "step": 7112 + }, + { + "epoch": 3.306834030683403, + "grad_norm": 3.37488055229187, + "learning_rate": 9.992890999289101e-06, + "loss": 0.2562, + "step": 7113 + }, + { + "epoch": 3.307298930729893, + "grad_norm": 7.062465667724609, + "learning_rate": 9.992889999289e-06, + "loss": 0.5556, + "step": 7114 + }, + { + "epoch": 3.3077638307763833, + "grad_norm": 5.3545050621032715, + "learning_rate": 9.992888999288901e-06, + "loss": 0.2759, + "step": 7115 + }, + { + "epoch": 3.308228730822873, + "grad_norm": 4.317400932312012, + "learning_rate": 9.9928879992888e-06, + "loss": 0.274, + "step": 7116 + }, + { + "epoch": 3.308693630869363, + "grad_norm": 3.7157440185546875, + "learning_rate": 9.9928869992887e-06, + "loss": 0.135, + "step": 7117 + }, + { + "epoch": 3.309158530915853, + "grad_norm": 5.832762718200684, + "learning_rate": 9.992885999288601e-06, + "loss": 0.3774, + "step": 7118 + }, + { + "epoch": 3.309623430962343, + "grad_norm": 5.74177885055542, + "learning_rate": 9.9928849992885e-06, + "loss": 0.4028, + "step": 7119 + }, + { + "epoch": 3.310088331008833, + "grad_norm": 7.056390285491943, + "learning_rate": 9.992883999288401e-06, + "loss": 0.3843, + "step": 7120 + }, + { + "epoch": 3.310553231055323, + "grad_norm": 5.820388317108154, + "learning_rate": 9.9928829992883e-06, + "loss": 0.4549, + "step": 7121 + }, + { + "epoch": 3.311018131101813, + "grad_norm": 6.930276393890381, + "learning_rate": 9.9928819992882e-06, + "loss": 0.5377, + "step": 7122 + }, + { + "epoch": 3.311483031148303, + "grad_norm": 3.7393758296966553, + "learning_rate": 9.992880999288101e-06, + "loss": 0.2503, + "step": 7123 + }, + { + "epoch": 3.3119479311947932, + "grad_norm": 6.503400802612305, + "learning_rate": 9.992879999288002e-06, + "loss": 0.424, + "step": 7124 + }, + { + "epoch": 3.312412831241283, + "grad_norm": 5.805760860443115, + "learning_rate": 9.9928789992879e-06, + "loss": 0.1355, + "step": 7125 + }, + { + "epoch": 3.3128777312877733, + "grad_norm": 7.944647312164307, + "learning_rate": 9.992877999287801e-06, + "loss": 0.5412, + "step": 7126 + }, + { + "epoch": 3.313342631334263, + "grad_norm": 7.050544738769531, + "learning_rate": 9.9928769992877e-06, + "loss": 0.6012, + "step": 7127 + }, + { + "epoch": 3.3138075313807533, + "grad_norm": 4.440800666809082, + "learning_rate": 9.992875999287601e-06, + "loss": 0.3171, + "step": 7128 + }, + { + "epoch": 3.314272431427243, + "grad_norm": 3.416196346282959, + "learning_rate": 9.9928749992875e-06, + "loss": 0.1503, + "step": 7129 + }, + { + "epoch": 3.314737331473733, + "grad_norm": 5.8075079917907715, + "learning_rate": 9.9928739992874e-06, + "loss": 0.3404, + "step": 7130 + }, + { + "epoch": 3.315202231520223, + "grad_norm": 2.8923656940460205, + "learning_rate": 9.992872999287301e-06, + "loss": 0.1267, + "step": 7131 + }, + { + "epoch": 3.315667131566713, + "grad_norm": 7.179591655731201, + "learning_rate": 9.9928719992872e-06, + "loss": 0.5651, + "step": 7132 + }, + { + "epoch": 3.3161320316132032, + "grad_norm": 6.7078351974487305, + "learning_rate": 9.992870999287101e-06, + "loss": 0.6126, + "step": 7133 + }, + { + "epoch": 3.316596931659693, + "grad_norm": 4.089321136474609, + "learning_rate": 9.992869999287e-06, + "loss": 0.2553, + "step": 7134 + }, + { + "epoch": 3.3170618317061833, + "grad_norm": 3.997345209121704, + "learning_rate": 9.9928689992869e-06, + "loss": 0.329, + "step": 7135 + }, + { + "epoch": 3.317526731752673, + "grad_norm": 7.429440498352051, + "learning_rate": 9.992867999286801e-06, + "loss": 0.4641, + "step": 7136 + }, + { + "epoch": 3.3179916317991633, + "grad_norm": 6.38182258605957, + "learning_rate": 9.992866999286702e-06, + "loss": 0.4974, + "step": 7137 + }, + { + "epoch": 3.318456531845653, + "grad_norm": 4.0163726806640625, + "learning_rate": 9.992865999286601e-06, + "loss": 0.203, + "step": 7138 + }, + { + "epoch": 3.318921431892143, + "grad_norm": 12.52319049835205, + "learning_rate": 9.992864999286502e-06, + "loss": 1.7862, + "step": 7139 + }, + { + "epoch": 3.319386331938633, + "grad_norm": 5.43958044052124, + "learning_rate": 9.9928639992864e-06, + "loss": 0.3782, + "step": 7140 + }, + { + "epoch": 3.3198512319851234, + "grad_norm": 4.443405628204346, + "learning_rate": 9.9928629992863e-06, + "loss": 0.2177, + "step": 7141 + }, + { + "epoch": 3.3203161320316132, + "grad_norm": 6.099630832672119, + "learning_rate": 9.9928619992862e-06, + "loss": 0.5137, + "step": 7142 + }, + { + "epoch": 3.320781032078103, + "grad_norm": 4.72268533706665, + "learning_rate": 9.9928609992861e-06, + "loss": 0.4075, + "step": 7143 + }, + { + "epoch": 3.3212459321245933, + "grad_norm": 4.827924728393555, + "learning_rate": 9.992859999286001e-06, + "loss": 0.3276, + "step": 7144 + }, + { + "epoch": 3.321710832171083, + "grad_norm": 4.31882905960083, + "learning_rate": 9.9928589992859e-06, + "loss": 0.3754, + "step": 7145 + }, + { + "epoch": 3.3221757322175733, + "grad_norm": 5.198692798614502, + "learning_rate": 9.992857999285801e-06, + "loss": 0.3067, + "step": 7146 + }, + { + "epoch": 3.322640632264063, + "grad_norm": 7.190765857696533, + "learning_rate": 9.9928569992857e-06, + "loss": 0.5969, + "step": 7147 + }, + { + "epoch": 3.3231055323105534, + "grad_norm": 4.867579936981201, + "learning_rate": 9.9928559992856e-06, + "loss": 0.2975, + "step": 7148 + }, + { + "epoch": 3.323570432357043, + "grad_norm": 3.955671787261963, + "learning_rate": 9.992854999285501e-06, + "loss": 0.153, + "step": 7149 + }, + { + "epoch": 3.3240353324035334, + "grad_norm": 3.546349287033081, + "learning_rate": 9.992853999285402e-06, + "loss": 0.1491, + "step": 7150 + }, + { + "epoch": 3.3245002324500232, + "grad_norm": 6.570027828216553, + "learning_rate": 9.992852999285301e-06, + "loss": 0.3127, + "step": 7151 + }, + { + "epoch": 3.324965132496513, + "grad_norm": 4.731911659240723, + "learning_rate": 9.9928519992852e-06, + "loss": 0.243, + "step": 7152 + }, + { + "epoch": 3.3254300325430033, + "grad_norm": 7.384954452514648, + "learning_rate": 9.9928509992851e-06, + "loss": 0.4533, + "step": 7153 + }, + { + "epoch": 3.325894932589493, + "grad_norm": 6.960841178894043, + "learning_rate": 9.992849999285e-06, + "loss": 0.3398, + "step": 7154 + }, + { + "epoch": 3.3263598326359833, + "grad_norm": 5.402586460113525, + "learning_rate": 9.9928489992849e-06, + "loss": 0.3642, + "step": 7155 + }, + { + "epoch": 3.326824732682473, + "grad_norm": 7.585971832275391, + "learning_rate": 9.992847999284801e-06, + "loss": 0.5571, + "step": 7156 + }, + { + "epoch": 3.3272896327289634, + "grad_norm": 6.100735187530518, + "learning_rate": 9.992846999284702e-06, + "loss": 0.5597, + "step": 7157 + }, + { + "epoch": 3.327754532775453, + "grad_norm": 4.701981544494629, + "learning_rate": 9.9928459992846e-06, + "loss": 0.2242, + "step": 7158 + }, + { + "epoch": 3.3282194328219434, + "grad_norm": 8.395452499389648, + "learning_rate": 9.992844999284501e-06, + "loss": 0.447, + "step": 7159 + }, + { + "epoch": 3.3286843328684332, + "grad_norm": 5.1271467208862305, + "learning_rate": 9.9928439992844e-06, + "loss": 0.3499, + "step": 7160 + }, + { + "epoch": 3.3291492329149235, + "grad_norm": 5.693853378295898, + "learning_rate": 9.9928429992843e-06, + "loss": 0.526, + "step": 7161 + }, + { + "epoch": 3.3296141329614133, + "grad_norm": 7.292142391204834, + "learning_rate": 9.992841999284201e-06, + "loss": 0.4113, + "step": 7162 + }, + { + "epoch": 3.3300790330079035, + "grad_norm": 5.163674831390381, + "learning_rate": 9.9928409992841e-06, + "loss": 0.2636, + "step": 7163 + }, + { + "epoch": 3.3305439330543933, + "grad_norm": 6.928340435028076, + "learning_rate": 9.992839999284001e-06, + "loss": 0.7799, + "step": 7164 + }, + { + "epoch": 3.331008833100883, + "grad_norm": 7.435180187225342, + "learning_rate": 9.9928389992839e-06, + "loss": 0.4328, + "step": 7165 + }, + { + "epoch": 3.3314737331473734, + "grad_norm": 2.9298946857452393, + "learning_rate": 9.9928379992838e-06, + "loss": 0.1828, + "step": 7166 + }, + { + "epoch": 3.331938633193863, + "grad_norm": 7.50698709487915, + "learning_rate": 9.9928369992837e-06, + "loss": 0.6674, + "step": 7167 + }, + { + "epoch": 3.3324035332403534, + "grad_norm": 5.525143623352051, + "learning_rate": 9.9928359992836e-06, + "loss": 0.3723, + "step": 7168 + }, + { + "epoch": 3.3328684332868432, + "grad_norm": 9.264496803283691, + "learning_rate": 9.992834999283501e-06, + "loss": 0.5561, + "step": 7169 + }, + { + "epoch": 3.3333333333333335, + "grad_norm": 4.73998498916626, + "learning_rate": 9.992833999283402e-06, + "loss": 0.2577, + "step": 7170 + }, + { + "epoch": 3.3337982333798233, + "grad_norm": 4.427800178527832, + "learning_rate": 9.9928329992833e-06, + "loss": 0.403, + "step": 7171 + }, + { + "epoch": 3.3342631334263135, + "grad_norm": 4.966070652008057, + "learning_rate": 9.992831999283201e-06, + "loss": 0.3364, + "step": 7172 + }, + { + "epoch": 3.3347280334728033, + "grad_norm": 8.248250007629395, + "learning_rate": 9.9928309992831e-06, + "loss": 0.5732, + "step": 7173 + }, + { + "epoch": 3.335192933519293, + "grad_norm": 4.556979656219482, + "learning_rate": 9.992829999283e-06, + "loss": 0.2875, + "step": 7174 + }, + { + "epoch": 3.3356578335657834, + "grad_norm": 4.775640964508057, + "learning_rate": 9.992828999282902e-06, + "loss": 0.2518, + "step": 7175 + }, + { + "epoch": 3.336122733612273, + "grad_norm": 4.525611877441406, + "learning_rate": 9.9928279992828e-06, + "loss": 0.4067, + "step": 7176 + }, + { + "epoch": 3.3365876336587634, + "grad_norm": 4.498414516448975, + "learning_rate": 9.992826999282701e-06, + "loss": 0.2899, + "step": 7177 + }, + { + "epoch": 3.3370525337052532, + "grad_norm": 6.267294883728027, + "learning_rate": 9.9928259992826e-06, + "loss": 0.3763, + "step": 7178 + }, + { + "epoch": 3.3375174337517435, + "grad_norm": 4.8090362548828125, + "learning_rate": 9.9928249992825e-06, + "loss": 0.3144, + "step": 7179 + }, + { + "epoch": 3.3379823337982333, + "grad_norm": 3.490628957748413, + "learning_rate": 9.9928239992824e-06, + "loss": 0.2338, + "step": 7180 + }, + { + "epoch": 3.3384472338447235, + "grad_norm": 4.300594806671143, + "learning_rate": 9.9928229992823e-06, + "loss": 0.2452, + "step": 7181 + }, + { + "epoch": 3.3389121338912133, + "grad_norm": 3.9250149726867676, + "learning_rate": 9.992821999282201e-06, + "loss": 0.2872, + "step": 7182 + }, + { + "epoch": 3.3393770339377036, + "grad_norm": 6.807643890380859, + "learning_rate": 9.992820999282102e-06, + "loss": 0.2403, + "step": 7183 + }, + { + "epoch": 3.3398419339841934, + "grad_norm": 7.976385593414307, + "learning_rate": 9.992819999282e-06, + "loss": 0.5929, + "step": 7184 + }, + { + "epoch": 3.3403068340306836, + "grad_norm": 3.87593412399292, + "learning_rate": 9.9928189992819e-06, + "loss": 0.1515, + "step": 7185 + }, + { + "epoch": 3.3407717340771734, + "grad_norm": 3.7020835876464844, + "learning_rate": 9.9928179992818e-06, + "loss": 0.2408, + "step": 7186 + }, + { + "epoch": 3.3412366341236632, + "grad_norm": 12.699322700500488, + "learning_rate": 9.992816999281701e-06, + "loss": 0.3699, + "step": 7187 + }, + { + "epoch": 3.3417015341701535, + "grad_norm": 6.450118064880371, + "learning_rate": 9.992815999281602e-06, + "loss": 0.6147, + "step": 7188 + }, + { + "epoch": 3.3421664342166433, + "grad_norm": 4.085748672485352, + "learning_rate": 9.9928149992815e-06, + "loss": 0.2558, + "step": 7189 + }, + { + "epoch": 3.3426313342631335, + "grad_norm": 4.718233585357666, + "learning_rate": 9.992813999281401e-06, + "loss": 0.313, + "step": 7190 + }, + { + "epoch": 3.3430962343096233, + "grad_norm": 3.3834614753723145, + "learning_rate": 9.9928129992813e-06, + "loss": 0.2224, + "step": 7191 + }, + { + "epoch": 3.3435611343561136, + "grad_norm": 6.37933874130249, + "learning_rate": 9.9928119992812e-06, + "loss": 0.3864, + "step": 7192 + }, + { + "epoch": 3.3440260344026034, + "grad_norm": 6.207315921783447, + "learning_rate": 9.992810999281101e-06, + "loss": 0.6424, + "step": 7193 + }, + { + "epoch": 3.3444909344490936, + "grad_norm": 8.934574127197266, + "learning_rate": 9.992809999281e-06, + "loss": 0.2342, + "step": 7194 + }, + { + "epoch": 3.3449558344955834, + "grad_norm": 2.9811108112335205, + "learning_rate": 9.992808999280901e-06, + "loss": 0.1503, + "step": 7195 + }, + { + "epoch": 3.3454207345420732, + "grad_norm": 5.188720226287842, + "learning_rate": 9.9928079992808e-06, + "loss": 0.5179, + "step": 7196 + }, + { + "epoch": 3.3458856345885635, + "grad_norm": 6.444155216217041, + "learning_rate": 9.9928069992807e-06, + "loss": 0.5846, + "step": 7197 + }, + { + "epoch": 3.3463505346350533, + "grad_norm": 5.557119369506836, + "learning_rate": 9.9928059992806e-06, + "loss": 0.4735, + "step": 7198 + }, + { + "epoch": 3.3468154346815435, + "grad_norm": 6.161092281341553, + "learning_rate": 9.9928049992805e-06, + "loss": 0.2991, + "step": 7199 + }, + { + "epoch": 3.3472803347280333, + "grad_norm": 5.401047229766846, + "learning_rate": 9.992803999280401e-06, + "loss": 0.2672, + "step": 7200 + }, + { + "epoch": 3.3477452347745236, + "grad_norm": 4.134251117706299, + "learning_rate": 9.992802999280302e-06, + "loss": 0.259, + "step": 7201 + }, + { + "epoch": 3.3482101348210134, + "grad_norm": 5.905023574829102, + "learning_rate": 9.9928019992802e-06, + "loss": 0.4499, + "step": 7202 + }, + { + "epoch": 3.3486750348675036, + "grad_norm": 3.2132632732391357, + "learning_rate": 9.992800999280101e-06, + "loss": 0.1644, + "step": 7203 + }, + { + "epoch": 3.3491399349139934, + "grad_norm": 4.692431449890137, + "learning_rate": 9.99279999928e-06, + "loss": 0.2849, + "step": 7204 + }, + { + "epoch": 3.3496048349604837, + "grad_norm": 5.384753227233887, + "learning_rate": 9.992798999279901e-06, + "loss": 0.3845, + "step": 7205 + }, + { + "epoch": 3.3500697350069735, + "grad_norm": 6.0444722175598145, + "learning_rate": 9.992797999279802e-06, + "loss": 0.4563, + "step": 7206 + }, + { + "epoch": 3.3505346350534637, + "grad_norm": 7.037875175476074, + "learning_rate": 9.9927969992797e-06, + "loss": 0.5405, + "step": 7207 + }, + { + "epoch": 3.3509995350999535, + "grad_norm": 4.733831882476807, + "learning_rate": 9.992795999279601e-06, + "loss": 0.2928, + "step": 7208 + }, + { + "epoch": 3.3514644351464433, + "grad_norm": 5.4814019203186035, + "learning_rate": 9.9927949992795e-06, + "loss": 0.3167, + "step": 7209 + }, + { + "epoch": 3.3519293351929336, + "grad_norm": 3.9713521003723145, + "learning_rate": 9.9927939992794e-06, + "loss": 0.2548, + "step": 7210 + }, + { + "epoch": 3.3523942352394234, + "grad_norm": 6.179991245269775, + "learning_rate": 9.9927929992793e-06, + "loss": 0.4184, + "step": 7211 + }, + { + "epoch": 3.3528591352859136, + "grad_norm": 6.872418403625488, + "learning_rate": 9.9927919992792e-06, + "loss": 0.2075, + "step": 7212 + }, + { + "epoch": 3.3533240353324034, + "grad_norm": 3.8280727863311768, + "learning_rate": 9.992790999279101e-06, + "loss": 0.1936, + "step": 7213 + }, + { + "epoch": 3.3537889353788937, + "grad_norm": 5.660084247589111, + "learning_rate": 9.992789999279002e-06, + "loss": 0.3416, + "step": 7214 + }, + { + "epoch": 3.3542538354253835, + "grad_norm": 7.681581020355225, + "learning_rate": 9.9927889992789e-06, + "loss": 0.3884, + "step": 7215 + }, + { + "epoch": 3.3547187354718737, + "grad_norm": 5.403874397277832, + "learning_rate": 9.992787999278801e-06, + "loss": 0.4482, + "step": 7216 + }, + { + "epoch": 3.3551836355183635, + "grad_norm": 3.7312517166137695, + "learning_rate": 9.9927869992787e-06, + "loss": 0.1623, + "step": 7217 + }, + { + "epoch": 3.3556485355648533, + "grad_norm": 4.794027805328369, + "learning_rate": 9.992785999278601e-06, + "loss": 0.2831, + "step": 7218 + }, + { + "epoch": 3.3561134356113436, + "grad_norm": 5.9126105308532715, + "learning_rate": 9.992784999278502e-06, + "loss": 0.5723, + "step": 7219 + }, + { + "epoch": 3.356578335657834, + "grad_norm": 10.870989799499512, + "learning_rate": 9.9927839992784e-06, + "loss": 0.4502, + "step": 7220 + }, + { + "epoch": 3.3570432357043236, + "grad_norm": 6.532385349273682, + "learning_rate": 9.992782999278301e-06, + "loss": 0.2986, + "step": 7221 + }, + { + "epoch": 3.3575081357508134, + "grad_norm": 5.4800567626953125, + "learning_rate": 9.9927819992782e-06, + "loss": 0.3299, + "step": 7222 + }, + { + "epoch": 3.3579730357973037, + "grad_norm": 4.538123607635498, + "learning_rate": 9.992780999278101e-06, + "loss": 0.3352, + "step": 7223 + }, + { + "epoch": 3.3584379358437935, + "grad_norm": 6.09683084487915, + "learning_rate": 9.992779999278e-06, + "loss": 0.4971, + "step": 7224 + }, + { + "epoch": 3.3589028358902837, + "grad_norm": 4.735533237457275, + "learning_rate": 9.9927789992779e-06, + "loss": 0.3992, + "step": 7225 + }, + { + "epoch": 3.3593677359367735, + "grad_norm": 5.060419082641602, + "learning_rate": 9.992777999277801e-06, + "loss": 0.306, + "step": 7226 + }, + { + "epoch": 3.3598326359832638, + "grad_norm": 5.692996025085449, + "learning_rate": 9.992776999277702e-06, + "loss": 0.3223, + "step": 7227 + }, + { + "epoch": 3.3602975360297536, + "grad_norm": 7.028481483459473, + "learning_rate": 9.9927759992776e-06, + "loss": 0.7808, + "step": 7228 + }, + { + "epoch": 3.360762436076244, + "grad_norm": 11.23828125, + "learning_rate": 9.9927749992775e-06, + "loss": 0.9717, + "step": 7229 + }, + { + "epoch": 3.3612273361227336, + "grad_norm": 7.074308395385742, + "learning_rate": 9.9927739992774e-06, + "loss": 0.5136, + "step": 7230 + }, + { + "epoch": 3.3616922361692234, + "grad_norm": 6.7666120529174805, + "learning_rate": 9.992772999277301e-06, + "loss": 0.6448, + "step": 7231 + }, + { + "epoch": 3.3621571362157137, + "grad_norm": 7.405423164367676, + "learning_rate": 9.992771999277202e-06, + "loss": 0.6049, + "step": 7232 + }, + { + "epoch": 3.3626220362622035, + "grad_norm": 4.290537357330322, + "learning_rate": 9.9927709992771e-06, + "loss": 0.2136, + "step": 7233 + }, + { + "epoch": 3.3630869363086937, + "grad_norm": 5.97194766998291, + "learning_rate": 9.992769999277001e-06, + "loss": 0.4184, + "step": 7234 + }, + { + "epoch": 3.3635518363551835, + "grad_norm": 5.423845291137695, + "learning_rate": 9.9927689992769e-06, + "loss": 0.4539, + "step": 7235 + }, + { + "epoch": 3.3640167364016738, + "grad_norm": 5.234395503997803, + "learning_rate": 9.992767999276801e-06, + "loss": 0.4603, + "step": 7236 + }, + { + "epoch": 3.3644816364481636, + "grad_norm": 5.233429908752441, + "learning_rate": 9.9927669992767e-06, + "loss": 0.2894, + "step": 7237 + }, + { + "epoch": 3.364946536494654, + "grad_norm": 8.405197143554688, + "learning_rate": 9.9927659992766e-06, + "loss": 0.4983, + "step": 7238 + }, + { + "epoch": 3.3654114365411436, + "grad_norm": 6.2007551193237305, + "learning_rate": 9.992764999276501e-06, + "loss": 0.3665, + "step": 7239 + }, + { + "epoch": 3.3658763365876334, + "grad_norm": 5.87105655670166, + "learning_rate": 9.9927639992764e-06, + "loss": 0.2424, + "step": 7240 + }, + { + "epoch": 3.3663412366341237, + "grad_norm": 5.837576866149902, + "learning_rate": 9.992762999276301e-06, + "loss": 0.518, + "step": 7241 + }, + { + "epoch": 3.366806136680614, + "grad_norm": 3.5358598232269287, + "learning_rate": 9.9927619992762e-06, + "loss": 0.2098, + "step": 7242 + }, + { + "epoch": 3.3672710367271037, + "grad_norm": 4.846395492553711, + "learning_rate": 9.9927609992761e-06, + "loss": 0.4092, + "step": 7243 + }, + { + "epoch": 3.3677359367735935, + "grad_norm": 4.580552577972412, + "learning_rate": 9.992759999276001e-06, + "loss": 0.2579, + "step": 7244 + }, + { + "epoch": 3.3682008368200838, + "grad_norm": 5.651418685913086, + "learning_rate": 9.992758999275902e-06, + "loss": 0.1762, + "step": 7245 + }, + { + "epoch": 3.3686657368665736, + "grad_norm": 5.333477020263672, + "learning_rate": 9.9927579992758e-06, + "loss": 0.249, + "step": 7246 + }, + { + "epoch": 3.369130636913064, + "grad_norm": 8.7036771774292, + "learning_rate": 9.992756999275701e-06, + "loss": 0.4592, + "step": 7247 + }, + { + "epoch": 3.3695955369595536, + "grad_norm": 3.8849422931671143, + "learning_rate": 9.9927559992756e-06, + "loss": 0.3463, + "step": 7248 + }, + { + "epoch": 3.370060437006044, + "grad_norm": 3.8918862342834473, + "learning_rate": 9.992754999275501e-06, + "loss": 0.2749, + "step": 7249 + }, + { + "epoch": 3.3705253370525337, + "grad_norm": 8.016485214233398, + "learning_rate": 9.9927539992754e-06, + "loss": 0.2805, + "step": 7250 + }, + { + "epoch": 3.370990237099024, + "grad_norm": 4.3206071853637695, + "learning_rate": 9.9927529992753e-06, + "loss": 0.3064, + "step": 7251 + }, + { + "epoch": 3.3714551371455137, + "grad_norm": 5.782379627227783, + "learning_rate": 9.992751999275201e-06, + "loss": 0.3771, + "step": 7252 + }, + { + "epoch": 3.3719200371920035, + "grad_norm": 8.222039222717285, + "learning_rate": 9.9927509992751e-06, + "loss": 0.3946, + "step": 7253 + }, + { + "epoch": 3.3723849372384938, + "grad_norm": 5.943896770477295, + "learning_rate": 9.992749999275001e-06, + "loss": 0.4466, + "step": 7254 + }, + { + "epoch": 3.3728498372849836, + "grad_norm": 5.345794677734375, + "learning_rate": 9.9927489992749e-06, + "loss": 0.3686, + "step": 7255 + }, + { + "epoch": 3.373314737331474, + "grad_norm": 7.114373683929443, + "learning_rate": 9.9927479992748e-06, + "loss": 0.4436, + "step": 7256 + }, + { + "epoch": 3.3737796373779636, + "grad_norm": 5.577574729919434, + "learning_rate": 9.992746999274701e-06, + "loss": 0.3681, + "step": 7257 + }, + { + "epoch": 3.374244537424454, + "grad_norm": 4.456395626068115, + "learning_rate": 9.992745999274602e-06, + "loss": 0.3382, + "step": 7258 + }, + { + "epoch": 3.3747094374709437, + "grad_norm": 3.87697696685791, + "learning_rate": 9.992744999274501e-06, + "loss": 0.142, + "step": 7259 + }, + { + "epoch": 3.375174337517434, + "grad_norm": 5.326664924621582, + "learning_rate": 9.9927439992744e-06, + "loss": 0.2934, + "step": 7260 + }, + { + "epoch": 3.3756392375639237, + "grad_norm": 8.706461906433105, + "learning_rate": 9.9927429992743e-06, + "loss": 0.7247, + "step": 7261 + }, + { + "epoch": 3.376104137610414, + "grad_norm": 4.6228814125061035, + "learning_rate": 9.992741999274201e-06, + "loss": 0.3501, + "step": 7262 + }, + { + "epoch": 3.3765690376569037, + "grad_norm": 5.667902946472168, + "learning_rate": 9.9927409992741e-06, + "loss": 0.2636, + "step": 7263 + }, + { + "epoch": 3.377033937703394, + "grad_norm": 3.437988042831421, + "learning_rate": 9.992739999274e-06, + "loss": 0.2706, + "step": 7264 + }, + { + "epoch": 3.377498837749884, + "grad_norm": 5.365514278411865, + "learning_rate": 9.992738999273901e-06, + "loss": 0.3779, + "step": 7265 + }, + { + "epoch": 3.3779637377963736, + "grad_norm": 8.348333358764648, + "learning_rate": 9.9927379992738e-06, + "loss": 0.4788, + "step": 7266 + }, + { + "epoch": 3.378428637842864, + "grad_norm": 7.542592525482178, + "learning_rate": 9.992736999273701e-06, + "loss": 0.3468, + "step": 7267 + }, + { + "epoch": 3.3788935378893536, + "grad_norm": 6.22538948059082, + "learning_rate": 9.9927359992736e-06, + "loss": 0.3646, + "step": 7268 + }, + { + "epoch": 3.379358437935844, + "grad_norm": 3.7032418251037598, + "learning_rate": 9.9927349992735e-06, + "loss": 0.2986, + "step": 7269 + }, + { + "epoch": 3.3798233379823337, + "grad_norm": 5.423750400543213, + "learning_rate": 9.992733999273401e-06, + "loss": 0.3913, + "step": 7270 + }, + { + "epoch": 3.380288238028824, + "grad_norm": 6.531966686248779, + "learning_rate": 9.9927329992733e-06, + "loss": 0.1782, + "step": 7271 + }, + { + "epoch": 3.3807531380753137, + "grad_norm": 4.41934871673584, + "learning_rate": 9.992731999273201e-06, + "loss": 0.2341, + "step": 7272 + }, + { + "epoch": 3.381218038121804, + "grad_norm": 10.203694343566895, + "learning_rate": 9.9927309992731e-06, + "loss": 0.5977, + "step": 7273 + }, + { + "epoch": 3.381682938168294, + "grad_norm": 4.717371940612793, + "learning_rate": 9.992729999273e-06, + "loss": 0.2369, + "step": 7274 + }, + { + "epoch": 3.3821478382147836, + "grad_norm": 4.531966686248779, + "learning_rate": 9.992728999272901e-06, + "loss": 0.3524, + "step": 7275 + }, + { + "epoch": 3.382612738261274, + "grad_norm": 5.830392360687256, + "learning_rate": 9.9927279992728e-06, + "loss": 0.4283, + "step": 7276 + }, + { + "epoch": 3.3830776383077636, + "grad_norm": 10.072427749633789, + "learning_rate": 9.9927269992727e-06, + "loss": 0.6407, + "step": 7277 + }, + { + "epoch": 3.383542538354254, + "grad_norm": 5.3291192054748535, + "learning_rate": 9.992725999272601e-06, + "loss": 0.406, + "step": 7278 + }, + { + "epoch": 3.3840074384007437, + "grad_norm": 4.226301670074463, + "learning_rate": 9.9927249992725e-06, + "loss": 0.3298, + "step": 7279 + }, + { + "epoch": 3.384472338447234, + "grad_norm": 6.171749114990234, + "learning_rate": 9.992723999272401e-06, + "loss": 0.4411, + "step": 7280 + }, + { + "epoch": 3.3849372384937237, + "grad_norm": 6.385472774505615, + "learning_rate": 9.9927229992723e-06, + "loss": 0.591, + "step": 7281 + }, + { + "epoch": 3.385402138540214, + "grad_norm": 5.15615177154541, + "learning_rate": 9.9927219992722e-06, + "loss": 0.3167, + "step": 7282 + }, + { + "epoch": 3.385867038586704, + "grad_norm": 8.263187408447266, + "learning_rate": 9.992720999272101e-06, + "loss": 0.6248, + "step": 7283 + }, + { + "epoch": 3.386331938633194, + "grad_norm": 8.134255409240723, + "learning_rate": 9.992719999272e-06, + "loss": 0.5456, + "step": 7284 + }, + { + "epoch": 3.386796838679684, + "grad_norm": 7.657655239105225, + "learning_rate": 9.992718999271901e-06, + "loss": 0.5447, + "step": 7285 + }, + { + "epoch": 3.387261738726174, + "grad_norm": 5.071996212005615, + "learning_rate": 9.9927179992718e-06, + "loss": 0.3683, + "step": 7286 + }, + { + "epoch": 3.387726638772664, + "grad_norm": 5.461882591247559, + "learning_rate": 9.9927169992717e-06, + "loss": 0.2548, + "step": 7287 + }, + { + "epoch": 3.3881915388191537, + "grad_norm": 5.79278039932251, + "learning_rate": 9.992715999271601e-06, + "loss": 0.3958, + "step": 7288 + }, + { + "epoch": 3.388656438865644, + "grad_norm": 3.8422694206237793, + "learning_rate": 9.992714999271502e-06, + "loss": 0.2409, + "step": 7289 + }, + { + "epoch": 3.3891213389121337, + "grad_norm": 2.9219160079956055, + "learning_rate": 9.992713999271401e-06, + "loss": 0.1632, + "step": 7290 + }, + { + "epoch": 3.389586238958624, + "grad_norm": 5.852606296539307, + "learning_rate": 9.992712999271302e-06, + "loss": 0.5516, + "step": 7291 + }, + { + "epoch": 3.390051139005114, + "grad_norm": 5.341282844543457, + "learning_rate": 9.9927119992712e-06, + "loss": 0.3857, + "step": 7292 + }, + { + "epoch": 3.390516039051604, + "grad_norm": 5.190321922302246, + "learning_rate": 9.9927109992711e-06, + "loss": 0.483, + "step": 7293 + }, + { + "epoch": 3.390980939098094, + "grad_norm": 4.446769714355469, + "learning_rate": 9.992709999271e-06, + "loss": 0.2377, + "step": 7294 + }, + { + "epoch": 3.391445839144584, + "grad_norm": 3.9845383167266846, + "learning_rate": 9.9927089992709e-06, + "loss": 0.3856, + "step": 7295 + }, + { + "epoch": 3.391910739191074, + "grad_norm": 8.653193473815918, + "learning_rate": 9.992707999270801e-06, + "loss": 0.7877, + "step": 7296 + }, + { + "epoch": 3.3923756392375637, + "grad_norm": 11.286497116088867, + "learning_rate": 9.9927069992707e-06, + "loss": 0.4782, + "step": 7297 + }, + { + "epoch": 3.392840539284054, + "grad_norm": 4.983236789703369, + "learning_rate": 9.992705999270601e-06, + "loss": 0.4879, + "step": 7298 + }, + { + "epoch": 3.393305439330544, + "grad_norm": 6.228190898895264, + "learning_rate": 9.9927049992705e-06, + "loss": 0.4533, + "step": 7299 + }, + { + "epoch": 3.393770339377034, + "grad_norm": 7.462355136871338, + "learning_rate": 9.9927039992704e-06, + "loss": 0.4387, + "step": 7300 + }, + { + "epoch": 3.394235239423524, + "grad_norm": 6.715991497039795, + "learning_rate": 9.992702999270301e-06, + "loss": 0.5856, + "step": 7301 + }, + { + "epoch": 3.394700139470014, + "grad_norm": 3.251962900161743, + "learning_rate": 9.992701999270202e-06, + "loss": 0.2757, + "step": 7302 + }, + { + "epoch": 3.395165039516504, + "grad_norm": 3.7031288146972656, + "learning_rate": 9.992700999270101e-06, + "loss": 0.2352, + "step": 7303 + }, + { + "epoch": 3.395629939562994, + "grad_norm": 5.326620101928711, + "learning_rate": 9.99269999927e-06, + "loss": 0.5216, + "step": 7304 + }, + { + "epoch": 3.396094839609484, + "grad_norm": 7.646754741668701, + "learning_rate": 9.9926989992699e-06, + "loss": 0.4322, + "step": 7305 + }, + { + "epoch": 3.396559739655974, + "grad_norm": 3.593364715576172, + "learning_rate": 9.9926979992698e-06, + "loss": 0.2411, + "step": 7306 + }, + { + "epoch": 3.397024639702464, + "grad_norm": 4.200722694396973, + "learning_rate": 9.9926969992697e-06, + "loss": 0.253, + "step": 7307 + }, + { + "epoch": 3.397489539748954, + "grad_norm": 7.571817398071289, + "learning_rate": 9.992695999269601e-06, + "loss": 0.7568, + "step": 7308 + }, + { + "epoch": 3.397954439795444, + "grad_norm": 3.059370994567871, + "learning_rate": 9.992694999269502e-06, + "loss": 0.0884, + "step": 7309 + }, + { + "epoch": 3.398419339841934, + "grad_norm": 5.471438884735107, + "learning_rate": 9.9926939992694e-06, + "loss": 0.4073, + "step": 7310 + }, + { + "epoch": 3.398884239888424, + "grad_norm": 5.542733192443848, + "learning_rate": 9.992692999269301e-06, + "loss": 0.4668, + "step": 7311 + }, + { + "epoch": 3.399349139934914, + "grad_norm": 7.819967269897461, + "learning_rate": 9.9926919992692e-06, + "loss": 0.2927, + "step": 7312 + }, + { + "epoch": 3.399814039981404, + "grad_norm": 4.457143783569336, + "learning_rate": 9.9926909992691e-06, + "loss": 0.2615, + "step": 7313 + }, + { + "epoch": 3.400278940027894, + "grad_norm": 5.62274169921875, + "learning_rate": 9.992689999269001e-06, + "loss": 0.3318, + "step": 7314 + }, + { + "epoch": 3.400743840074384, + "grad_norm": 5.971207141876221, + "learning_rate": 9.9926889992689e-06, + "loss": 0.4131, + "step": 7315 + }, + { + "epoch": 3.401208740120874, + "grad_norm": 5.634144306182861, + "learning_rate": 9.992687999268801e-06, + "loss": 0.3533, + "step": 7316 + }, + { + "epoch": 3.401673640167364, + "grad_norm": 3.773710250854492, + "learning_rate": 9.9926869992687e-06, + "loss": 0.2008, + "step": 7317 + }, + { + "epoch": 3.402138540213854, + "grad_norm": 5.165319442749023, + "learning_rate": 9.9926859992686e-06, + "loss": 0.4078, + "step": 7318 + }, + { + "epoch": 3.402603440260344, + "grad_norm": 5.933526992797852, + "learning_rate": 9.9926849992685e-06, + "loss": 0.4016, + "step": 7319 + }, + { + "epoch": 3.403068340306834, + "grad_norm": 7.230973720550537, + "learning_rate": 9.9926839992684e-06, + "loss": 0.2693, + "step": 7320 + }, + { + "epoch": 3.4035332403533243, + "grad_norm": 5.83908224105835, + "learning_rate": 9.992682999268301e-06, + "loss": 0.4337, + "step": 7321 + }, + { + "epoch": 3.403998140399814, + "grad_norm": 4.7375006675720215, + "learning_rate": 9.992681999268202e-06, + "loss": 0.3383, + "step": 7322 + }, + { + "epoch": 3.404463040446304, + "grad_norm": 4.208312034606934, + "learning_rate": 9.9926809992681e-06, + "loss": 0.2404, + "step": 7323 + }, + { + "epoch": 3.404927940492794, + "grad_norm": 4.614388942718506, + "learning_rate": 9.992679999268001e-06, + "loss": 0.2644, + "step": 7324 + }, + { + "epoch": 3.405392840539284, + "grad_norm": 5.116252899169922, + "learning_rate": 9.9926789992679e-06, + "loss": 0.5203, + "step": 7325 + }, + { + "epoch": 3.405857740585774, + "grad_norm": 2.3926103115081787, + "learning_rate": 9.992677999267801e-06, + "loss": 0.0774, + "step": 7326 + }, + { + "epoch": 3.406322640632264, + "grad_norm": 5.220162391662598, + "learning_rate": 9.992676999267702e-06, + "loss": 0.1321, + "step": 7327 + }, + { + "epoch": 3.4067875406787542, + "grad_norm": 4.693324089050293, + "learning_rate": 9.9926759992676e-06, + "loss": 0.2323, + "step": 7328 + }, + { + "epoch": 3.407252440725244, + "grad_norm": 5.715918064117432, + "learning_rate": 9.992674999267501e-06, + "loss": 0.2687, + "step": 7329 + }, + { + "epoch": 3.4077173407717343, + "grad_norm": 5.933376312255859, + "learning_rate": 9.9926739992674e-06, + "loss": 0.3882, + "step": 7330 + }, + { + "epoch": 3.408182240818224, + "grad_norm": 5.441254138946533, + "learning_rate": 9.9926729992673e-06, + "loss": 0.3454, + "step": 7331 + }, + { + "epoch": 3.408647140864714, + "grad_norm": 5.45803165435791, + "learning_rate": 9.9926719992672e-06, + "loss": 0.356, + "step": 7332 + }, + { + "epoch": 3.409112040911204, + "grad_norm": 5.161346912384033, + "learning_rate": 9.9926709992671e-06, + "loss": 0.4006, + "step": 7333 + }, + { + "epoch": 3.409576940957694, + "grad_norm": 6.624203681945801, + "learning_rate": 9.992669999267001e-06, + "loss": 0.3075, + "step": 7334 + }, + { + "epoch": 3.410041841004184, + "grad_norm": 4.214599609375, + "learning_rate": 9.992668999266902e-06, + "loss": 0.2665, + "step": 7335 + }, + { + "epoch": 3.410506741050674, + "grad_norm": 5.4114484786987305, + "learning_rate": 9.9926679992668e-06, + "loss": 0.3544, + "step": 7336 + }, + { + "epoch": 3.410971641097164, + "grad_norm": 5.368140697479248, + "learning_rate": 9.9926669992667e-06, + "loss": 0.5146, + "step": 7337 + }, + { + "epoch": 3.411436541143654, + "grad_norm": 7.018730640411377, + "learning_rate": 9.9926659992666e-06, + "loss": 0.447, + "step": 7338 + }, + { + "epoch": 3.4119014411901443, + "grad_norm": 6.572421073913574, + "learning_rate": 9.992664999266501e-06, + "loss": 0.2991, + "step": 7339 + }, + { + "epoch": 3.412366341236634, + "grad_norm": 4.404265403747559, + "learning_rate": 9.992663999266402e-06, + "loss": 0.2399, + "step": 7340 + }, + { + "epoch": 3.4128312412831243, + "grad_norm": 8.851370811462402, + "learning_rate": 9.9926629992663e-06, + "loss": 0.7392, + "step": 7341 + }, + { + "epoch": 3.413296141329614, + "grad_norm": 5.3929595947265625, + "learning_rate": 9.992661999266201e-06, + "loss": 0.4382, + "step": 7342 + }, + { + "epoch": 3.4137610413761044, + "grad_norm": 9.048551559448242, + "learning_rate": 9.9926609992661e-06, + "loss": 0.4786, + "step": 7343 + }, + { + "epoch": 3.414225941422594, + "grad_norm": 5.110420227050781, + "learning_rate": 9.992659999266001e-06, + "loss": 0.3302, + "step": 7344 + }, + { + "epoch": 3.414690841469084, + "grad_norm": 6.276269912719727, + "learning_rate": 9.9926589992659e-06, + "loss": 0.4459, + "step": 7345 + }, + { + "epoch": 3.415155741515574, + "grad_norm": 4.289761543273926, + "learning_rate": 9.9926579992658e-06, + "loss": 0.2779, + "step": 7346 + }, + { + "epoch": 3.415620641562064, + "grad_norm": 4.60464334487915, + "learning_rate": 9.992656999265701e-06, + "loss": 0.3157, + "step": 7347 + }, + { + "epoch": 3.4160855416085543, + "grad_norm": 12.446413040161133, + "learning_rate": 9.9926559992656e-06, + "loss": 0.6774, + "step": 7348 + }, + { + "epoch": 3.416550441655044, + "grad_norm": 6.253787517547607, + "learning_rate": 9.9926549992655e-06, + "loss": 0.4584, + "step": 7349 + }, + { + "epoch": 3.4170153417015343, + "grad_norm": 7.264977931976318, + "learning_rate": 9.9926539992654e-06, + "loss": 0.4554, + "step": 7350 + }, + { + "epoch": 3.417480241748024, + "grad_norm": 4.614774227142334, + "learning_rate": 9.9926529992653e-06, + "loss": 0.1358, + "step": 7351 + }, + { + "epoch": 3.4179451417945144, + "grad_norm": 5.305474281311035, + "learning_rate": 9.992651999265201e-06, + "loss": 0.302, + "step": 7352 + }, + { + "epoch": 3.418410041841004, + "grad_norm": 3.9909720420837402, + "learning_rate": 9.992650999265102e-06, + "loss": 0.2123, + "step": 7353 + }, + { + "epoch": 3.418874941887494, + "grad_norm": 3.75413179397583, + "learning_rate": 9.992649999265e-06, + "loss": 0.1848, + "step": 7354 + }, + { + "epoch": 3.419339841933984, + "grad_norm": 2.8738977909088135, + "learning_rate": 9.992648999264901e-06, + "loss": 0.0877, + "step": 7355 + }, + { + "epoch": 3.419804741980474, + "grad_norm": 5.895839691162109, + "learning_rate": 9.9926479992648e-06, + "loss": 0.3719, + "step": 7356 + }, + { + "epoch": 3.4202696420269643, + "grad_norm": 4.76785135269165, + "learning_rate": 9.992646999264701e-06, + "loss": 0.3095, + "step": 7357 + }, + { + "epoch": 3.420734542073454, + "grad_norm": 4.0182929039001465, + "learning_rate": 9.992645999264602e-06, + "loss": 0.2503, + "step": 7358 + }, + { + "epoch": 3.4211994421199443, + "grad_norm": 5.581352233886719, + "learning_rate": 9.9926449992645e-06, + "loss": 0.3701, + "step": 7359 + }, + { + "epoch": 3.421664342166434, + "grad_norm": 6.4922919273376465, + "learning_rate": 9.992643999264401e-06, + "loss": 0.5036, + "step": 7360 + }, + { + "epoch": 3.4221292422129244, + "grad_norm": 4.201266765594482, + "learning_rate": 9.9926429992643e-06, + "loss": 0.1863, + "step": 7361 + }, + { + "epoch": 3.422594142259414, + "grad_norm": 6.399768829345703, + "learning_rate": 9.9926419992642e-06, + "loss": 0.3551, + "step": 7362 + }, + { + "epoch": 3.4230590423059044, + "grad_norm": 5.328676700592041, + "learning_rate": 9.9926409992641e-06, + "loss": 0.2364, + "step": 7363 + }, + { + "epoch": 3.423523942352394, + "grad_norm": 5.003631114959717, + "learning_rate": 9.992639999264e-06, + "loss": 0.2425, + "step": 7364 + }, + { + "epoch": 3.4239888423988845, + "grad_norm": 8.361297607421875, + "learning_rate": 9.992638999263901e-06, + "loss": 0.6064, + "step": 7365 + }, + { + "epoch": 3.4244537424453743, + "grad_norm": 3.8811283111572266, + "learning_rate": 9.992637999263802e-06, + "loss": 0.2587, + "step": 7366 + }, + { + "epoch": 3.424918642491864, + "grad_norm": 6.456524848937988, + "learning_rate": 9.9926369992637e-06, + "loss": 0.3813, + "step": 7367 + }, + { + "epoch": 3.4253835425383543, + "grad_norm": 5.622960567474365, + "learning_rate": 9.992635999263601e-06, + "loss": 0.3986, + "step": 7368 + }, + { + "epoch": 3.425848442584844, + "grad_norm": 6.316389560699463, + "learning_rate": 9.9926349992635e-06, + "loss": 0.5093, + "step": 7369 + }, + { + "epoch": 3.4263133426313344, + "grad_norm": 7.78672981262207, + "learning_rate": 9.992633999263401e-06, + "loss": 0.3573, + "step": 7370 + }, + { + "epoch": 3.426778242677824, + "grad_norm": 4.927061557769775, + "learning_rate": 9.992632999263302e-06, + "loss": 0.5294, + "step": 7371 + }, + { + "epoch": 3.4272431427243144, + "grad_norm": 8.152297019958496, + "learning_rate": 9.9926319992632e-06, + "loss": 0.3098, + "step": 7372 + }, + { + "epoch": 3.427708042770804, + "grad_norm": 5.331478595733643, + "learning_rate": 9.992630999263101e-06, + "loss": 0.2972, + "step": 7373 + }, + { + "epoch": 3.4281729428172945, + "grad_norm": 4.282097816467285, + "learning_rate": 9.992629999263e-06, + "loss": 0.2353, + "step": 7374 + }, + { + "epoch": 3.4286378428637843, + "grad_norm": 6.452651023864746, + "learning_rate": 9.992628999262901e-06, + "loss": 0.4723, + "step": 7375 + }, + { + "epoch": 3.429102742910274, + "grad_norm": 4.785149097442627, + "learning_rate": 9.9926279992628e-06, + "loss": 0.3546, + "step": 7376 + }, + { + "epoch": 3.4295676429567643, + "grad_norm": 3.8610289096832275, + "learning_rate": 9.9926269992627e-06, + "loss": 0.2652, + "step": 7377 + }, + { + "epoch": 3.430032543003254, + "grad_norm": 5.376312255859375, + "learning_rate": 9.992625999262601e-06, + "loss": 0.4371, + "step": 7378 + }, + { + "epoch": 3.4304974430497444, + "grad_norm": 5.140150547027588, + "learning_rate": 9.992624999262502e-06, + "loss": 0.3516, + "step": 7379 + }, + { + "epoch": 3.430962343096234, + "grad_norm": 5.51079797744751, + "learning_rate": 9.9926239992624e-06, + "loss": 0.2426, + "step": 7380 + }, + { + "epoch": 3.4314272431427244, + "grad_norm": 5.3276143074035645, + "learning_rate": 9.9926229992623e-06, + "loss": 0.4063, + "step": 7381 + }, + { + "epoch": 3.431892143189214, + "grad_norm": 5.270547866821289, + "learning_rate": 9.9926219992622e-06, + "loss": 0.4102, + "step": 7382 + }, + { + "epoch": 3.4323570432357045, + "grad_norm": 6.233720302581787, + "learning_rate": 9.992620999262101e-06, + "loss": 0.2126, + "step": 7383 + }, + { + "epoch": 3.4328219432821943, + "grad_norm": 4.725608825683594, + "learning_rate": 9.992619999262002e-06, + "loss": 0.2319, + "step": 7384 + }, + { + "epoch": 3.4332868433286845, + "grad_norm": 2.648751974105835, + "learning_rate": 9.9926189992619e-06, + "loss": 0.114, + "step": 7385 + }, + { + "epoch": 3.4337517433751743, + "grad_norm": 4.0474982261657715, + "learning_rate": 9.992617999261801e-06, + "loss": 0.2872, + "step": 7386 + }, + { + "epoch": 3.4342166434216645, + "grad_norm": 4.755678653717041, + "learning_rate": 9.9926169992617e-06, + "loss": 0.2156, + "step": 7387 + }, + { + "epoch": 3.4346815434681544, + "grad_norm": 4.830945014953613, + "learning_rate": 9.992615999261601e-06, + "loss": 0.2847, + "step": 7388 + }, + { + "epoch": 3.435146443514644, + "grad_norm": 7.2333598136901855, + "learning_rate": 9.9926149992615e-06, + "loss": 0.6017, + "step": 7389 + }, + { + "epoch": 3.4356113435611344, + "grad_norm": 5.243837833404541, + "learning_rate": 9.9926139992614e-06, + "loss": 0.2041, + "step": 7390 + }, + { + "epoch": 3.436076243607624, + "grad_norm": 6.5054731369018555, + "learning_rate": 9.992612999261301e-06, + "loss": 0.3434, + "step": 7391 + }, + { + "epoch": 3.4365411436541144, + "grad_norm": 2.7672297954559326, + "learning_rate": 9.9926119992612e-06, + "loss": 0.0713, + "step": 7392 + }, + { + "epoch": 3.4370060437006043, + "grad_norm": 5.038865566253662, + "learning_rate": 9.992610999261101e-06, + "loss": 0.4376, + "step": 7393 + }, + { + "epoch": 3.4374709437470945, + "grad_norm": 6.14592981338501, + "learning_rate": 9.992609999261e-06, + "loss": 0.2209, + "step": 7394 + }, + { + "epoch": 3.4379358437935843, + "grad_norm": 6.8827104568481445, + "learning_rate": 9.9926089992609e-06, + "loss": 0.5509, + "step": 7395 + }, + { + "epoch": 3.4384007438400745, + "grad_norm": 7.4498443603515625, + "learning_rate": 9.992607999260801e-06, + "loss": 0.4797, + "step": 7396 + }, + { + "epoch": 3.4388656438865643, + "grad_norm": 5.262049198150635, + "learning_rate": 9.992606999260702e-06, + "loss": 0.3674, + "step": 7397 + }, + { + "epoch": 3.439330543933054, + "grad_norm": 6.564810276031494, + "learning_rate": 9.9926059992606e-06, + "loss": 0.56, + "step": 7398 + }, + { + "epoch": 3.4397954439795444, + "grad_norm": 5.824267864227295, + "learning_rate": 9.992604999260501e-06, + "loss": 0.3213, + "step": 7399 + }, + { + "epoch": 3.4402603440260346, + "grad_norm": 5.817203044891357, + "learning_rate": 9.9926039992604e-06, + "loss": 0.2702, + "step": 7400 + }, + { + "epoch": 3.4407252440725244, + "grad_norm": 4.4692606925964355, + "learning_rate": 9.9926029992603e-06, + "loss": 0.3396, + "step": 7401 + }, + { + "epoch": 3.4411901441190142, + "grad_norm": 5.3800129890441895, + "learning_rate": 9.9926019992602e-06, + "loss": 0.4325, + "step": 7402 + }, + { + "epoch": 3.4416550441655045, + "grad_norm": 5.75887393951416, + "learning_rate": 9.9926009992601e-06, + "loss": 0.4828, + "step": 7403 + }, + { + "epoch": 3.4421199442119943, + "grad_norm": 6.744731426239014, + "learning_rate": 9.992599999260001e-06, + "loss": 0.5116, + "step": 7404 + }, + { + "epoch": 3.4425848442584845, + "grad_norm": 6.0845255851745605, + "learning_rate": 9.9925989992599e-06, + "loss": 0.444, + "step": 7405 + }, + { + "epoch": 3.4430497443049743, + "grad_norm": 11.037041664123535, + "learning_rate": 9.992597999259801e-06, + "loss": 0.8184, + "step": 7406 + }, + { + "epoch": 3.4435146443514646, + "grad_norm": 4.065990447998047, + "learning_rate": 9.9925969992597e-06, + "loss": 0.2966, + "step": 7407 + }, + { + "epoch": 3.4439795443979544, + "grad_norm": 4.941860675811768, + "learning_rate": 9.9925959992596e-06, + "loss": 0.3895, + "step": 7408 + }, + { + "epoch": 3.4444444444444446, + "grad_norm": 6.298384666442871, + "learning_rate": 9.992594999259501e-06, + "loss": 0.4277, + "step": 7409 + }, + { + "epoch": 3.4449093444909344, + "grad_norm": 5.007673740386963, + "learning_rate": 9.992593999259402e-06, + "loss": 0.2949, + "step": 7410 + }, + { + "epoch": 3.4453742445374242, + "grad_norm": 4.67044734954834, + "learning_rate": 9.992592999259301e-06, + "loss": 0.2155, + "step": 7411 + }, + { + "epoch": 3.4458391445839145, + "grad_norm": 6.1636857986450195, + "learning_rate": 9.9925919992592e-06, + "loss": 0.325, + "step": 7412 + }, + { + "epoch": 3.4463040446304043, + "grad_norm": 6.776901721954346, + "learning_rate": 9.9925909992591e-06, + "loss": 0.6294, + "step": 7413 + }, + { + "epoch": 3.4467689446768945, + "grad_norm": 8.158512115478516, + "learning_rate": 9.992589999259e-06, + "loss": 0.3201, + "step": 7414 + }, + { + "epoch": 3.4472338447233843, + "grad_norm": 4.482407093048096, + "learning_rate": 9.9925889992589e-06, + "loss": 0.4552, + "step": 7415 + }, + { + "epoch": 3.4476987447698746, + "grad_norm": 6.139918327331543, + "learning_rate": 9.9925879992588e-06, + "loss": 0.7135, + "step": 7416 + }, + { + "epoch": 3.4481636448163644, + "grad_norm": 4.3017578125, + "learning_rate": 9.992586999258701e-06, + "loss": 0.2835, + "step": 7417 + }, + { + "epoch": 3.4486285448628546, + "grad_norm": 5.1276960372924805, + "learning_rate": 9.9925859992586e-06, + "loss": 0.3627, + "step": 7418 + }, + { + "epoch": 3.4490934449093444, + "grad_norm": 4.140954494476318, + "learning_rate": 9.992584999258501e-06, + "loss": 0.2523, + "step": 7419 + }, + { + "epoch": 3.4495583449558347, + "grad_norm": 5.316616058349609, + "learning_rate": 9.9925839992584e-06, + "loss": 0.4941, + "step": 7420 + }, + { + "epoch": 3.4500232450023245, + "grad_norm": 4.7725958824157715, + "learning_rate": 9.9925829992583e-06, + "loss": 0.3156, + "step": 7421 + }, + { + "epoch": 3.4504881450488147, + "grad_norm": 5.419440746307373, + "learning_rate": 9.992581999258201e-06, + "loss": 0.4905, + "step": 7422 + }, + { + "epoch": 3.4509530450953045, + "grad_norm": 4.302948474884033, + "learning_rate": 9.9925809992581e-06, + "loss": 0.3122, + "step": 7423 + }, + { + "epoch": 3.4514179451417943, + "grad_norm": 5.017322540283203, + "learning_rate": 9.992579999258001e-06, + "loss": 0.2484, + "step": 7424 + }, + { + "epoch": 3.4518828451882846, + "grad_norm": 6.13444709777832, + "learning_rate": 9.9925789992579e-06, + "loss": 0.4859, + "step": 7425 + }, + { + "epoch": 3.4523477452347744, + "grad_norm": 6.175135612487793, + "learning_rate": 9.9925779992578e-06, + "loss": 0.5105, + "step": 7426 + }, + { + "epoch": 3.4528126452812646, + "grad_norm": 4.823066711425781, + "learning_rate": 9.992576999257701e-06, + "loss": 0.3206, + "step": 7427 + }, + { + "epoch": 3.4532775453277544, + "grad_norm": 6.413930892944336, + "learning_rate": 9.9925759992576e-06, + "loss": 0.4191, + "step": 7428 + }, + { + "epoch": 3.4537424453742447, + "grad_norm": 4.841460704803467, + "learning_rate": 9.992574999257501e-06, + "loss": 0.4413, + "step": 7429 + }, + { + "epoch": 3.4542073454207345, + "grad_norm": 6.143120288848877, + "learning_rate": 9.992573999257402e-06, + "loss": 0.4082, + "step": 7430 + }, + { + "epoch": 3.4546722454672247, + "grad_norm": 3.9598259925842285, + "learning_rate": 9.9925729992573e-06, + "loss": 0.1558, + "step": 7431 + }, + { + "epoch": 3.4551371455137145, + "grad_norm": 5.418359279632568, + "learning_rate": 9.992571999257201e-06, + "loss": 0.447, + "step": 7432 + }, + { + "epoch": 3.4556020455602043, + "grad_norm": 4.898799896240234, + "learning_rate": 9.9925709992571e-06, + "loss": 0.434, + "step": 7433 + }, + { + "epoch": 3.4560669456066946, + "grad_norm": 5.986307621002197, + "learning_rate": 9.992569999257e-06, + "loss": 0.2999, + "step": 7434 + }, + { + "epoch": 3.4565318456531844, + "grad_norm": 4.469932556152344, + "learning_rate": 9.992568999256901e-06, + "loss": 0.2812, + "step": 7435 + }, + { + "epoch": 3.4569967456996746, + "grad_norm": 6.509045600891113, + "learning_rate": 9.9925679992568e-06, + "loss": 0.5399, + "step": 7436 + }, + { + "epoch": 3.4574616457461644, + "grad_norm": 5.924966812133789, + "learning_rate": 9.992566999256701e-06, + "loss": 0.4603, + "step": 7437 + }, + { + "epoch": 3.4579265457926547, + "grad_norm": 3.1588943004608154, + "learning_rate": 9.9925659992566e-06, + "loss": 0.1601, + "step": 7438 + }, + { + "epoch": 3.4583914458391445, + "grad_norm": 3.475330114364624, + "learning_rate": 9.9925649992565e-06, + "loss": 0.2647, + "step": 7439 + }, + { + "epoch": 3.4588563458856347, + "grad_norm": 3.8628923892974854, + "learning_rate": 9.992563999256401e-06, + "loss": 0.2716, + "step": 7440 + }, + { + "epoch": 3.4593212459321245, + "grad_norm": 5.030220031738281, + "learning_rate": 9.9925629992563e-06, + "loss": 0.2854, + "step": 7441 + }, + { + "epoch": 3.459786145978615, + "grad_norm": 5.644862174987793, + "learning_rate": 9.992561999256201e-06, + "loss": 0.281, + "step": 7442 + }, + { + "epoch": 3.4602510460251046, + "grad_norm": 4.3885369300842285, + "learning_rate": 9.992560999256102e-06, + "loss": 0.2164, + "step": 7443 + }, + { + "epoch": 3.460715946071595, + "grad_norm": 6.476162433624268, + "learning_rate": 9.992559999256e-06, + "loss": 0.3249, + "step": 7444 + }, + { + "epoch": 3.4611808461180846, + "grad_norm": 7.916736602783203, + "learning_rate": 9.9925589992559e-06, + "loss": 0.5739, + "step": 7445 + }, + { + "epoch": 3.4616457461645744, + "grad_norm": 4.808040142059326, + "learning_rate": 9.9925579992558e-06, + "loss": 0.2666, + "step": 7446 + }, + { + "epoch": 3.4621106462110647, + "grad_norm": 3.2435102462768555, + "learning_rate": 9.992556999255701e-06, + "loss": 0.1926, + "step": 7447 + }, + { + "epoch": 3.4625755462575545, + "grad_norm": 4.213681221008301, + "learning_rate": 9.992555999255602e-06, + "loss": 0.2779, + "step": 7448 + }, + { + "epoch": 3.4630404463040447, + "grad_norm": 4.866991996765137, + "learning_rate": 9.9925549992555e-06, + "loss": 0.3611, + "step": 7449 + }, + { + "epoch": 3.4635053463505345, + "grad_norm": 4.5816330909729, + "learning_rate": 9.992553999255401e-06, + "loss": 0.3118, + "step": 7450 + }, + { + "epoch": 3.4639702463970248, + "grad_norm": 7.067690849304199, + "learning_rate": 9.9925529992553e-06, + "loss": 0.5338, + "step": 7451 + }, + { + "epoch": 3.4644351464435146, + "grad_norm": 5.436336994171143, + "learning_rate": 9.9925519992552e-06, + "loss": 0.3402, + "step": 7452 + }, + { + "epoch": 3.464900046490005, + "grad_norm": 4.67192268371582, + "learning_rate": 9.992550999255101e-06, + "loss": 0.5476, + "step": 7453 + }, + { + "epoch": 3.4653649465364946, + "grad_norm": 5.544418811798096, + "learning_rate": 9.992549999255002e-06, + "loss": 0.4147, + "step": 7454 + }, + { + "epoch": 3.4658298465829844, + "grad_norm": 4.372371196746826, + "learning_rate": 9.992548999254901e-06, + "loss": 0.2603, + "step": 7455 + }, + { + "epoch": 3.4662947466294747, + "grad_norm": 7.083232402801514, + "learning_rate": 9.9925479992548e-06, + "loss": 0.4714, + "step": 7456 + }, + { + "epoch": 3.4667596466759645, + "grad_norm": 5.801002025604248, + "learning_rate": 9.9925469992547e-06, + "loss": 0.3227, + "step": 7457 + }, + { + "epoch": 3.4672245467224547, + "grad_norm": 7.34351110458374, + "learning_rate": 9.9925459992546e-06, + "loss": 0.6497, + "step": 7458 + }, + { + "epoch": 3.4676894467689445, + "grad_norm": 7.7269768714904785, + "learning_rate": 9.9925449992545e-06, + "loss": 0.6247, + "step": 7459 + }, + { + "epoch": 3.4681543468154348, + "grad_norm": 6.334197044372559, + "learning_rate": 9.992543999254401e-06, + "loss": 0.4184, + "step": 7460 + }, + { + "epoch": 3.4686192468619246, + "grad_norm": 3.7734971046447754, + "learning_rate": 9.992542999254302e-06, + "loss": 0.2257, + "step": 7461 + }, + { + "epoch": 3.469084146908415, + "grad_norm": 4.667993545532227, + "learning_rate": 9.9925419992542e-06, + "loss": 0.3179, + "step": 7462 + }, + { + "epoch": 3.4695490469549046, + "grad_norm": 4.694325923919678, + "learning_rate": 9.992540999254101e-06, + "loss": 0.2607, + "step": 7463 + }, + { + "epoch": 3.470013947001395, + "grad_norm": 9.551027297973633, + "learning_rate": 9.992539999254e-06, + "loss": 0.7468, + "step": 7464 + }, + { + "epoch": 3.4704788470478847, + "grad_norm": 3.32059907913208, + "learning_rate": 9.9925389992539e-06, + "loss": 0.2611, + "step": 7465 + }, + { + "epoch": 3.470943747094375, + "grad_norm": 7.640324592590332, + "learning_rate": 9.992537999253801e-06, + "loss": 0.388, + "step": 7466 + }, + { + "epoch": 3.4714086471408647, + "grad_norm": 5.672605514526367, + "learning_rate": 9.9925369992537e-06, + "loss": 0.6466, + "step": 7467 + }, + { + "epoch": 3.4718735471873545, + "grad_norm": 3.8170316219329834, + "learning_rate": 9.992535999253601e-06, + "loss": 0.197, + "step": 7468 + }, + { + "epoch": 3.4723384472338448, + "grad_norm": 7.109022617340088, + "learning_rate": 9.9925349992535e-06, + "loss": 0.3597, + "step": 7469 + }, + { + "epoch": 3.4728033472803346, + "grad_norm": 4.211771488189697, + "learning_rate": 9.9925339992534e-06, + "loss": 0.2652, + "step": 7470 + }, + { + "epoch": 3.473268247326825, + "grad_norm": 6.650413990020752, + "learning_rate": 9.9925329992533e-06, + "loss": 0.6821, + "step": 7471 + }, + { + "epoch": 3.4737331473733146, + "grad_norm": 4.260782718658447, + "learning_rate": 9.9925319992532e-06, + "loss": 0.1477, + "step": 7472 + }, + { + "epoch": 3.474198047419805, + "grad_norm": 5.0482988357543945, + "learning_rate": 9.992530999253101e-06, + "loss": 0.2499, + "step": 7473 + }, + { + "epoch": 3.4746629474662947, + "grad_norm": 3.340669870376587, + "learning_rate": 9.992529999253002e-06, + "loss": 0.1703, + "step": 7474 + }, + { + "epoch": 3.475127847512785, + "grad_norm": 4.299245357513428, + "learning_rate": 9.9925289992529e-06, + "loss": 0.3221, + "step": 7475 + }, + { + "epoch": 3.4755927475592747, + "grad_norm": 3.8189585208892822, + "learning_rate": 9.992527999252801e-06, + "loss": 0.183, + "step": 7476 + }, + { + "epoch": 3.4760576476057645, + "grad_norm": 5.101255893707275, + "learning_rate": 9.9925269992527e-06, + "loss": 0.2146, + "step": 7477 + }, + { + "epoch": 3.4765225476522548, + "grad_norm": 5.987974643707275, + "learning_rate": 9.992525999252601e-06, + "loss": 0.2661, + "step": 7478 + }, + { + "epoch": 3.476987447698745, + "grad_norm": 6.236059665679932, + "learning_rate": 9.992524999252502e-06, + "loss": 0.765, + "step": 7479 + }, + { + "epoch": 3.477452347745235, + "grad_norm": 4.779110431671143, + "learning_rate": 9.9925239992524e-06, + "loss": 0.2943, + "step": 7480 + }, + { + "epoch": 3.4779172477917246, + "grad_norm": 4.471465110778809, + "learning_rate": 9.992522999252301e-06, + "loss": 0.2198, + "step": 7481 + }, + { + "epoch": 3.478382147838215, + "grad_norm": 4.236536026000977, + "learning_rate": 9.9925219992522e-06, + "loss": 0.2691, + "step": 7482 + }, + { + "epoch": 3.4788470478847047, + "grad_norm": 5.242837429046631, + "learning_rate": 9.9925209992521e-06, + "loss": 0.5497, + "step": 7483 + }, + { + "epoch": 3.479311947931195, + "grad_norm": 4.901486396789551, + "learning_rate": 9.992519999252e-06, + "loss": 0.4249, + "step": 7484 + }, + { + "epoch": 3.4797768479776847, + "grad_norm": 5.072946548461914, + "learning_rate": 9.9925189992519e-06, + "loss": 0.4992, + "step": 7485 + }, + { + "epoch": 3.480241748024175, + "grad_norm": 5.42798376083374, + "learning_rate": 9.992517999251801e-06, + "loss": 0.4646, + "step": 7486 + }, + { + "epoch": 3.4807066480706648, + "grad_norm": 4.347535610198975, + "learning_rate": 9.992516999251702e-06, + "loss": 0.3416, + "step": 7487 + }, + { + "epoch": 3.481171548117155, + "grad_norm": 6.357630252838135, + "learning_rate": 9.9925159992516e-06, + "loss": 0.5089, + "step": 7488 + }, + { + "epoch": 3.481636448163645, + "grad_norm": 5.378523349761963, + "learning_rate": 9.9925149992515e-06, + "loss": 0.2199, + "step": 7489 + }, + { + "epoch": 3.4821013482101346, + "grad_norm": 3.6668787002563477, + "learning_rate": 9.9925139992514e-06, + "loss": 0.0549, + "step": 7490 + }, + { + "epoch": 3.482566248256625, + "grad_norm": 9.41571044921875, + "learning_rate": 9.992512999251301e-06, + "loss": 0.3159, + "step": 7491 + }, + { + "epoch": 3.4830311483031147, + "grad_norm": 8.37813663482666, + "learning_rate": 9.992511999251202e-06, + "loss": 0.5898, + "step": 7492 + }, + { + "epoch": 3.483496048349605, + "grad_norm": 6.434249401092529, + "learning_rate": 9.9925109992511e-06, + "loss": 0.2803, + "step": 7493 + }, + { + "epoch": 3.4839609483960947, + "grad_norm": 9.25659465789795, + "learning_rate": 9.992509999251001e-06, + "loss": 0.4351, + "step": 7494 + }, + { + "epoch": 3.484425848442585, + "grad_norm": 7.10609245300293, + "learning_rate": 9.9925089992509e-06, + "loss": 0.4859, + "step": 7495 + }, + { + "epoch": 3.4848907484890748, + "grad_norm": 9.918537139892578, + "learning_rate": 9.992507999250801e-06, + "loss": 0.4059, + "step": 7496 + }, + { + "epoch": 3.485355648535565, + "grad_norm": 3.3138158321380615, + "learning_rate": 9.9925069992507e-06, + "loss": 0.1193, + "step": 7497 + }, + { + "epoch": 3.485820548582055, + "grad_norm": 5.2466721534729, + "learning_rate": 9.9925059992506e-06, + "loss": 0.2707, + "step": 7498 + }, + { + "epoch": 3.4862854486285446, + "grad_norm": 5.622682094573975, + "learning_rate": 9.992504999250501e-06, + "loss": 0.4192, + "step": 7499 + }, + { + "epoch": 3.486750348675035, + "grad_norm": 4.5102219581604, + "learning_rate": 9.9925039992504e-06, + "loss": 0.2288, + "step": 7500 + }, + { + "epoch": 3.487215248721525, + "grad_norm": 4.473243713378906, + "learning_rate": 9.9925029992503e-06, + "loss": 0.3202, + "step": 7501 + }, + { + "epoch": 3.487680148768015, + "grad_norm": 3.5176641941070557, + "learning_rate": 9.9925019992502e-06, + "loss": 0.2144, + "step": 7502 + }, + { + "epoch": 3.4881450488145047, + "grad_norm": 4.110781192779541, + "learning_rate": 9.9925009992501e-06, + "loss": 0.2514, + "step": 7503 + }, + { + "epoch": 3.488609948860995, + "grad_norm": 4.817535877227783, + "learning_rate": 9.992499999250001e-06, + "loss": 0.3235, + "step": 7504 + }, + { + "epoch": 3.4890748489074848, + "grad_norm": 3.6134257316589355, + "learning_rate": 9.992498999249902e-06, + "loss": 0.1517, + "step": 7505 + }, + { + "epoch": 3.489539748953975, + "grad_norm": 9.954682350158691, + "learning_rate": 9.9924979992498e-06, + "loss": 0.7582, + "step": 7506 + }, + { + "epoch": 3.490004649000465, + "grad_norm": 3.6836140155792236, + "learning_rate": 9.992496999249701e-06, + "loss": 0.2013, + "step": 7507 + }, + { + "epoch": 3.490469549046955, + "grad_norm": 6.250258922576904, + "learning_rate": 9.9924959992496e-06, + "loss": 0.6699, + "step": 7508 + }, + { + "epoch": 3.490934449093445, + "grad_norm": 8.46849250793457, + "learning_rate": 9.992494999249501e-06, + "loss": 0.1916, + "step": 7509 + }, + { + "epoch": 3.491399349139935, + "grad_norm": 4.726973056793213, + "learning_rate": 9.9924939992494e-06, + "loss": 0.2461, + "step": 7510 + }, + { + "epoch": 3.491864249186425, + "grad_norm": 5.674875736236572, + "learning_rate": 9.9924929992493e-06, + "loss": 0.467, + "step": 7511 + }, + { + "epoch": 3.4923291492329147, + "grad_norm": 7.405993461608887, + "learning_rate": 9.992491999249201e-06, + "loss": 0.4506, + "step": 7512 + }, + { + "epoch": 3.492794049279405, + "grad_norm": 5.397034168243408, + "learning_rate": 9.9924909992491e-06, + "loss": 0.3844, + "step": 7513 + }, + { + "epoch": 3.4932589493258948, + "grad_norm": 4.352041721343994, + "learning_rate": 9.992489999249001e-06, + "loss": 0.1532, + "step": 7514 + }, + { + "epoch": 3.493723849372385, + "grad_norm": 5.969458103179932, + "learning_rate": 9.9924889992489e-06, + "loss": 0.2865, + "step": 7515 + }, + { + "epoch": 3.494188749418875, + "grad_norm": 5.330113410949707, + "learning_rate": 9.9924879992488e-06, + "loss": 0.2988, + "step": 7516 + }, + { + "epoch": 3.494653649465365, + "grad_norm": 5.584452152252197, + "learning_rate": 9.992486999248701e-06, + "loss": 0.3002, + "step": 7517 + }, + { + "epoch": 3.495118549511855, + "grad_norm": 7.344271183013916, + "learning_rate": 9.992485999248602e-06, + "loss": 0.4773, + "step": 7518 + }, + { + "epoch": 3.495583449558345, + "grad_norm": 6.850968360900879, + "learning_rate": 9.9924849992485e-06, + "loss": 0.312, + "step": 7519 + }, + { + "epoch": 3.496048349604835, + "grad_norm": 5.17326545715332, + "learning_rate": 9.992483999248401e-06, + "loss": 0.267, + "step": 7520 + }, + { + "epoch": 3.496513249651325, + "grad_norm": 5.942327499389648, + "learning_rate": 9.9924829992483e-06, + "loss": 0.5242, + "step": 7521 + }, + { + "epoch": 3.496978149697815, + "grad_norm": 6.222740650177002, + "learning_rate": 9.992481999248201e-06, + "loss": 0.5044, + "step": 7522 + }, + { + "epoch": 3.497443049744305, + "grad_norm": 6.255983352661133, + "learning_rate": 9.992480999248102e-06, + "loss": 0.3862, + "step": 7523 + }, + { + "epoch": 3.497907949790795, + "grad_norm": 5.12504243850708, + "learning_rate": 9.992479999248e-06, + "loss": 0.4604, + "step": 7524 + }, + { + "epoch": 3.498372849837285, + "grad_norm": 7.578052997589111, + "learning_rate": 9.992478999247901e-06, + "loss": 0.3283, + "step": 7525 + }, + { + "epoch": 3.498837749883775, + "grad_norm": 5.982784271240234, + "learning_rate": 9.9924779992478e-06, + "loss": 0.4589, + "step": 7526 + }, + { + "epoch": 3.499302649930265, + "grad_norm": 5.779885768890381, + "learning_rate": 9.992476999247701e-06, + "loss": 0.6434, + "step": 7527 + }, + { + "epoch": 3.499767549976755, + "grad_norm": 5.807971000671387, + "learning_rate": 9.9924759992476e-06, + "loss": 0.2759, + "step": 7528 + }, + { + "epoch": 3.500232450023245, + "grad_norm": 6.026835918426514, + "learning_rate": 9.9924749992475e-06, + "loss": 0.3421, + "step": 7529 + }, + { + "epoch": 3.500697350069735, + "grad_norm": 6.945743560791016, + "learning_rate": 9.992473999247401e-06, + "loss": 0.6942, + "step": 7530 + }, + { + "epoch": 3.501162250116225, + "grad_norm": 7.158663272857666, + "learning_rate": 9.9924729992473e-06, + "loss": 0.3828, + "step": 7531 + }, + { + "epoch": 3.501627150162715, + "grad_norm": 6.392284393310547, + "learning_rate": 9.992471999247201e-06, + "loss": 0.4699, + "step": 7532 + }, + { + "epoch": 3.502092050209205, + "grad_norm": 6.807513236999512, + "learning_rate": 9.9924709992471e-06, + "loss": 0.7348, + "step": 7533 + }, + { + "epoch": 3.502556950255695, + "grad_norm": 5.278719902038574, + "learning_rate": 9.992469999247e-06, + "loss": 0.4449, + "step": 7534 + }, + { + "epoch": 3.503021850302185, + "grad_norm": 6.689767837524414, + "learning_rate": 9.992468999246901e-06, + "loss": 0.2973, + "step": 7535 + }, + { + "epoch": 3.5034867503486753, + "grad_norm": 3.1043784618377686, + "learning_rate": 9.992467999246802e-06, + "loss": 0.2328, + "step": 7536 + }, + { + "epoch": 3.503951650395165, + "grad_norm": 2.4310522079467773, + "learning_rate": 9.9924669992467e-06, + "loss": 0.1331, + "step": 7537 + }, + { + "epoch": 3.504416550441655, + "grad_norm": 7.192633628845215, + "learning_rate": 9.992465999246601e-06, + "loss": 0.2649, + "step": 7538 + }, + { + "epoch": 3.504881450488145, + "grad_norm": 6.571774959564209, + "learning_rate": 9.9924649992465e-06, + "loss": 0.3428, + "step": 7539 + }, + { + "epoch": 3.505346350534635, + "grad_norm": 10.743792533874512, + "learning_rate": 9.992463999246401e-06, + "loss": 0.3121, + "step": 7540 + }, + { + "epoch": 3.505811250581125, + "grad_norm": 7.486828327178955, + "learning_rate": 9.9924629992463e-06, + "loss": 0.4902, + "step": 7541 + }, + { + "epoch": 3.506276150627615, + "grad_norm": 5.605395793914795, + "learning_rate": 9.9924619992462e-06, + "loss": 0.4233, + "step": 7542 + }, + { + "epoch": 3.506741050674105, + "grad_norm": 6.3868489265441895, + "learning_rate": 9.992460999246101e-06, + "loss": 0.4726, + "step": 7543 + }, + { + "epoch": 3.507205950720595, + "grad_norm": 5.204576015472412, + "learning_rate": 9.992459999246e-06, + "loss": 0.4826, + "step": 7544 + }, + { + "epoch": 3.5076708507670853, + "grad_norm": 5.729098796844482, + "learning_rate": 9.992458999245901e-06, + "loss": 0.3916, + "step": 7545 + }, + { + "epoch": 3.508135750813575, + "grad_norm": 5.020167827606201, + "learning_rate": 9.9924579992458e-06, + "loss": 0.2405, + "step": 7546 + }, + { + "epoch": 3.508600650860065, + "grad_norm": 4.609969615936279, + "learning_rate": 9.9924569992457e-06, + "loss": 0.2251, + "step": 7547 + }, + { + "epoch": 3.509065550906555, + "grad_norm": 5.3215484619140625, + "learning_rate": 9.992455999245601e-06, + "loss": 0.5857, + "step": 7548 + }, + { + "epoch": 3.509530450953045, + "grad_norm": 5.302478790283203, + "learning_rate": 9.992454999245502e-06, + "loss": 0.2344, + "step": 7549 + }, + { + "epoch": 3.509995350999535, + "grad_norm": 6.720122337341309, + "learning_rate": 9.9924539992454e-06, + "loss": 0.4202, + "step": 7550 + }, + { + "epoch": 3.510460251046025, + "grad_norm": 4.106143474578857, + "learning_rate": 9.992452999245301e-06, + "loss": 0.2463, + "step": 7551 + }, + { + "epoch": 3.5109251510925152, + "grad_norm": 7.146397113800049, + "learning_rate": 9.9924519992452e-06, + "loss": 0.452, + "step": 7552 + }, + { + "epoch": 3.511390051139005, + "grad_norm": 5.078897953033447, + "learning_rate": 9.9924509992451e-06, + "loss": 0.5677, + "step": 7553 + }, + { + "epoch": 3.5118549511854953, + "grad_norm": 6.280213832855225, + "learning_rate": 9.992449999245e-06, + "loss": 0.4406, + "step": 7554 + }, + { + "epoch": 3.512319851231985, + "grad_norm": 6.84803581237793, + "learning_rate": 9.9924489992449e-06, + "loss": 0.765, + "step": 7555 + }, + { + "epoch": 3.512784751278475, + "grad_norm": 6.2521772384643555, + "learning_rate": 9.992447999244801e-06, + "loss": 0.2831, + "step": 7556 + }, + { + "epoch": 3.513249651324965, + "grad_norm": 5.674371719360352, + "learning_rate": 9.9924469992447e-06, + "loss": 0.2918, + "step": 7557 + }, + { + "epoch": 3.5137145513714554, + "grad_norm": 11.38718318939209, + "learning_rate": 9.992445999244601e-06, + "loss": 0.7108, + "step": 7558 + }, + { + "epoch": 3.514179451417945, + "grad_norm": 5.612961292266846, + "learning_rate": 9.9924449992445e-06, + "loss": 0.3414, + "step": 7559 + }, + { + "epoch": 3.514644351464435, + "grad_norm": 5.065802574157715, + "learning_rate": 9.9924439992444e-06, + "loss": 0.2559, + "step": 7560 + }, + { + "epoch": 3.5151092515109252, + "grad_norm": 3.654656410217285, + "learning_rate": 9.992442999244301e-06, + "loss": 0.1291, + "step": 7561 + }, + { + "epoch": 3.515574151557415, + "grad_norm": 7.417534828186035, + "learning_rate": 9.992441999244202e-06, + "loss": 0.5973, + "step": 7562 + }, + { + "epoch": 3.5160390516039053, + "grad_norm": 4.965782642364502, + "learning_rate": 9.992440999244101e-06, + "loss": 0.2323, + "step": 7563 + }, + { + "epoch": 3.516503951650395, + "grad_norm": 5.83182954788208, + "learning_rate": 9.992439999244e-06, + "loss": 0.4873, + "step": 7564 + }, + { + "epoch": 3.5169688516968853, + "grad_norm": 4.007747650146484, + "learning_rate": 9.9924389992439e-06, + "loss": 0.2498, + "step": 7565 + }, + { + "epoch": 3.517433751743375, + "grad_norm": 5.411154747009277, + "learning_rate": 9.9924379992438e-06, + "loss": 0.2653, + "step": 7566 + }, + { + "epoch": 3.5178986517898654, + "grad_norm": 5.128386497497559, + "learning_rate": 9.9924369992437e-06, + "loss": 0.2721, + "step": 7567 + }, + { + "epoch": 3.518363551836355, + "grad_norm": 5.498985290527344, + "learning_rate": 9.9924359992436e-06, + "loss": 0.4196, + "step": 7568 + }, + { + "epoch": 3.518828451882845, + "grad_norm": 5.001932144165039, + "learning_rate": 9.992434999243501e-06, + "loss": 0.2308, + "step": 7569 + }, + { + "epoch": 3.5192933519293352, + "grad_norm": 4.930384635925293, + "learning_rate": 9.9924339992434e-06, + "loss": 0.4799, + "step": 7570 + }, + { + "epoch": 3.519758251975825, + "grad_norm": 5.255136966705322, + "learning_rate": 9.992432999243301e-06, + "loss": 0.4873, + "step": 7571 + }, + { + "epoch": 3.5202231520223153, + "grad_norm": 5.64469575881958, + "learning_rate": 9.9924319992432e-06, + "loss": 0.4398, + "step": 7572 + }, + { + "epoch": 3.520688052068805, + "grad_norm": 5.838398456573486, + "learning_rate": 9.9924309992431e-06, + "loss": 0.5086, + "step": 7573 + }, + { + "epoch": 3.5211529521152953, + "grad_norm": 6.709354400634766, + "learning_rate": 9.992429999243001e-06, + "loss": 0.4935, + "step": 7574 + }, + { + "epoch": 3.521617852161785, + "grad_norm": 8.19444751739502, + "learning_rate": 9.9924289992429e-06, + "loss": 0.6073, + "step": 7575 + }, + { + "epoch": 3.5220827522082754, + "grad_norm": 9.587715148925781, + "learning_rate": 9.992427999242801e-06, + "loss": 0.8848, + "step": 7576 + }, + { + "epoch": 3.522547652254765, + "grad_norm": 5.340432167053223, + "learning_rate": 9.9924269992427e-06, + "loss": 0.4726, + "step": 7577 + }, + { + "epoch": 3.523012552301255, + "grad_norm": 8.366827964782715, + "learning_rate": 9.9924259992426e-06, + "loss": 0.2291, + "step": 7578 + }, + { + "epoch": 3.5234774523477452, + "grad_norm": 6.83625602722168, + "learning_rate": 9.9924249992425e-06, + "loss": 0.2993, + "step": 7579 + }, + { + "epoch": 3.5239423523942355, + "grad_norm": 4.221080303192139, + "learning_rate": 9.9924239992424e-06, + "loss": 0.348, + "step": 7580 + }, + { + "epoch": 3.5244072524407253, + "grad_norm": 9.074591636657715, + "learning_rate": 9.992422999242301e-06, + "loss": 0.4351, + "step": 7581 + }, + { + "epoch": 3.524872152487215, + "grad_norm": 5.430827617645264, + "learning_rate": 9.992421999242202e-06, + "loss": 0.4526, + "step": 7582 + }, + { + "epoch": 3.5253370525337053, + "grad_norm": 5.8866987228393555, + "learning_rate": 9.9924209992421e-06, + "loss": 0.4975, + "step": 7583 + }, + { + "epoch": 3.525801952580195, + "grad_norm": 9.494888305664062, + "learning_rate": 9.992419999242001e-06, + "loss": 0.3602, + "step": 7584 + }, + { + "epoch": 3.5262668526266854, + "grad_norm": 5.148182392120361, + "learning_rate": 9.9924189992419e-06, + "loss": 0.2609, + "step": 7585 + }, + { + "epoch": 3.526731752673175, + "grad_norm": 8.037521362304688, + "learning_rate": 9.9924179992418e-06, + "loss": 0.465, + "step": 7586 + }, + { + "epoch": 3.5271966527196654, + "grad_norm": 8.594850540161133, + "learning_rate": 9.992416999241701e-06, + "loss": 0.3521, + "step": 7587 + }, + { + "epoch": 3.5276615527661552, + "grad_norm": 7.0236406326293945, + "learning_rate": 9.9924159992416e-06, + "loss": 0.4085, + "step": 7588 + }, + { + "epoch": 3.5281264528126455, + "grad_norm": 6.993546009063721, + "learning_rate": 9.992414999241501e-06, + "loss": 0.7707, + "step": 7589 + }, + { + "epoch": 3.5285913528591353, + "grad_norm": 6.596684455871582, + "learning_rate": 9.9924139992414e-06, + "loss": 0.5551, + "step": 7590 + }, + { + "epoch": 3.529056252905625, + "grad_norm": 4.843433856964111, + "learning_rate": 9.9924129992413e-06, + "loss": 0.2074, + "step": 7591 + }, + { + "epoch": 3.5295211529521153, + "grad_norm": 6.264275074005127, + "learning_rate": 9.992411999241201e-06, + "loss": 0.29, + "step": 7592 + }, + { + "epoch": 3.5299860529986056, + "grad_norm": 10.678971290588379, + "learning_rate": 9.9924109992411e-06, + "loss": 0.3036, + "step": 7593 + }, + { + "epoch": 3.5304509530450954, + "grad_norm": 7.167612075805664, + "learning_rate": 9.992409999241001e-06, + "loss": 0.3203, + "step": 7594 + }, + { + "epoch": 3.530915853091585, + "grad_norm": 5.56974983215332, + "learning_rate": 9.992408999240902e-06, + "loss": 0.5326, + "step": 7595 + }, + { + "epoch": 3.5313807531380754, + "grad_norm": 3.556096076965332, + "learning_rate": 9.9924079992408e-06, + "loss": 0.1328, + "step": 7596 + }, + { + "epoch": 3.5318456531845652, + "grad_norm": 8.992554664611816, + "learning_rate": 9.9924069992407e-06, + "loss": 0.6211, + "step": 7597 + }, + { + "epoch": 3.5323105532310555, + "grad_norm": 7.251515865325928, + "learning_rate": 9.9924059992406e-06, + "loss": 0.442, + "step": 7598 + }, + { + "epoch": 3.5327754532775453, + "grad_norm": 6.571633338928223, + "learning_rate": 9.992404999240501e-06, + "loss": 0.3467, + "step": 7599 + }, + { + "epoch": 3.533240353324035, + "grad_norm": 3.699164390563965, + "learning_rate": 9.992403999240402e-06, + "loss": 0.1444, + "step": 7600 + }, + { + "epoch": 3.5337052533705253, + "grad_norm": 4.803199768066406, + "learning_rate": 9.9924029992403e-06, + "loss": 0.4397, + "step": 7601 + }, + { + "epoch": 3.5341701534170156, + "grad_norm": 3.988075017929077, + "learning_rate": 9.992401999240201e-06, + "loss": 0.2231, + "step": 7602 + }, + { + "epoch": 3.5346350534635054, + "grad_norm": 4.934415817260742, + "learning_rate": 9.9924009992401e-06, + "loss": 0.5691, + "step": 7603 + }, + { + "epoch": 3.535099953509995, + "grad_norm": 4.917923450469971, + "learning_rate": 9.99239999924e-06, + "loss": 0.3545, + "step": 7604 + }, + { + "epoch": 3.5355648535564854, + "grad_norm": 5.91126823425293, + "learning_rate": 9.992398999239901e-06, + "loss": 0.4652, + "step": 7605 + }, + { + "epoch": 3.5360297536029752, + "grad_norm": 4.475883960723877, + "learning_rate": 9.9923979992398e-06, + "loss": 0.2209, + "step": 7606 + }, + { + "epoch": 3.5364946536494655, + "grad_norm": 2.822233200073242, + "learning_rate": 9.992396999239701e-06, + "loss": 0.1237, + "step": 7607 + }, + { + "epoch": 3.5369595536959553, + "grad_norm": 4.918337345123291, + "learning_rate": 9.9923959992396e-06, + "loss": 0.2799, + "step": 7608 + }, + { + "epoch": 3.5374244537424455, + "grad_norm": 5.111923694610596, + "learning_rate": 9.9923949992395e-06, + "loss": 0.517, + "step": 7609 + }, + { + "epoch": 3.5378893537889353, + "grad_norm": 4.548210144042969, + "learning_rate": 9.9923939992394e-06, + "loss": 0.2573, + "step": 7610 + }, + { + "epoch": 3.5383542538354256, + "grad_norm": 4.8437299728393555, + "learning_rate": 9.9923929992393e-06, + "loss": 0.307, + "step": 7611 + }, + { + "epoch": 3.5388191538819154, + "grad_norm": 4.919524192810059, + "learning_rate": 9.992391999239201e-06, + "loss": 0.3919, + "step": 7612 + }, + { + "epoch": 3.539284053928405, + "grad_norm": 3.706881046295166, + "learning_rate": 9.992390999239102e-06, + "loss": 0.2215, + "step": 7613 + }, + { + "epoch": 3.5397489539748954, + "grad_norm": 6.718112468719482, + "learning_rate": 9.992389999239e-06, + "loss": 0.5942, + "step": 7614 + }, + { + "epoch": 3.5402138540213857, + "grad_norm": 5.596782207489014, + "learning_rate": 9.992388999238901e-06, + "loss": 0.2547, + "step": 7615 + }, + { + "epoch": 3.5406787540678755, + "grad_norm": 7.511495113372803, + "learning_rate": 9.9923879992388e-06, + "loss": 0.2589, + "step": 7616 + }, + { + "epoch": 3.5411436541143653, + "grad_norm": 4.346659183502197, + "learning_rate": 9.992386999238701e-06, + "loss": 0.2924, + "step": 7617 + }, + { + "epoch": 3.5416085541608555, + "grad_norm": 6.312626838684082, + "learning_rate": 9.992385999238602e-06, + "loss": 0.2226, + "step": 7618 + }, + { + "epoch": 3.5420734542073453, + "grad_norm": 4.598736763000488, + "learning_rate": 9.9923849992385e-06, + "loss": 0.3745, + "step": 7619 + }, + { + "epoch": 3.5425383542538356, + "grad_norm": 4.80695104598999, + "learning_rate": 9.992383999238401e-06, + "loss": 0.2311, + "step": 7620 + }, + { + "epoch": 3.5430032543003254, + "grad_norm": 6.100685119628906, + "learning_rate": 9.9923829992383e-06, + "loss": 0.3863, + "step": 7621 + }, + { + "epoch": 3.543468154346815, + "grad_norm": 3.7354066371917725, + "learning_rate": 9.9923819992382e-06, + "loss": 0.2736, + "step": 7622 + }, + { + "epoch": 3.5439330543933054, + "grad_norm": 5.626888275146484, + "learning_rate": 9.9923809992381e-06, + "loss": 0.2564, + "step": 7623 + }, + { + "epoch": 3.5443979544397957, + "grad_norm": 4.886547565460205, + "learning_rate": 9.992379999238e-06, + "loss": 0.3378, + "step": 7624 + }, + { + "epoch": 3.5448628544862855, + "grad_norm": 4.786961555480957, + "learning_rate": 9.992378999237901e-06, + "loss": 0.3246, + "step": 7625 + }, + { + "epoch": 3.5453277545327753, + "grad_norm": 7.718458652496338, + "learning_rate": 9.992377999237802e-06, + "loss": 0.4131, + "step": 7626 + }, + { + "epoch": 3.5457926545792655, + "grad_norm": 2.384321451187134, + "learning_rate": 9.9923769992377e-06, + "loss": 0.0858, + "step": 7627 + }, + { + "epoch": 3.5462575546257553, + "grad_norm": 4.876363277435303, + "learning_rate": 9.992375999237601e-06, + "loss": 0.1873, + "step": 7628 + }, + { + "epoch": 3.5467224546722456, + "grad_norm": 5.2514424324035645, + "learning_rate": 9.9923749992375e-06, + "loss": 0.4548, + "step": 7629 + }, + { + "epoch": 3.5471873547187354, + "grad_norm": 3.7413580417633057, + "learning_rate": 9.992373999237401e-06, + "loss": 0.1953, + "step": 7630 + }, + { + "epoch": 3.5476522547652256, + "grad_norm": 9.237991333007812, + "learning_rate": 9.992372999237302e-06, + "loss": 0.5384, + "step": 7631 + }, + { + "epoch": 3.5481171548117154, + "grad_norm": 6.2664618492126465, + "learning_rate": 9.9923719992372e-06, + "loss": 0.677, + "step": 7632 + }, + { + "epoch": 3.5485820548582057, + "grad_norm": 7.027975082397461, + "learning_rate": 9.992370999237101e-06, + "loss": 0.5471, + "step": 7633 + }, + { + "epoch": 3.5490469549046955, + "grad_norm": 11.277908325195312, + "learning_rate": 9.992369999237e-06, + "loss": 0.3011, + "step": 7634 + }, + { + "epoch": 3.5495118549511853, + "grad_norm": 3.7706594467163086, + "learning_rate": 9.9923689992369e-06, + "loss": 0.184, + "step": 7635 + }, + { + "epoch": 3.5499767549976755, + "grad_norm": 5.447098255157471, + "learning_rate": 9.9923679992368e-06, + "loss": 0.4149, + "step": 7636 + }, + { + "epoch": 3.5504416550441658, + "grad_norm": 4.850311279296875, + "learning_rate": 9.9923669992367e-06, + "loss": 0.4141, + "step": 7637 + }, + { + "epoch": 3.5509065550906556, + "grad_norm": 3.4160006046295166, + "learning_rate": 9.992365999236601e-06, + "loss": 0.1956, + "step": 7638 + }, + { + "epoch": 3.5513714551371454, + "grad_norm": 4.216894149780273, + "learning_rate": 9.992364999236502e-06, + "loss": 0.3854, + "step": 7639 + }, + { + "epoch": 3.5518363551836356, + "grad_norm": 8.097980499267578, + "learning_rate": 9.9923639992364e-06, + "loss": 0.4914, + "step": 7640 + }, + { + "epoch": 3.5523012552301254, + "grad_norm": 3.937295913696289, + "learning_rate": 9.9923629992363e-06, + "loss": 0.2856, + "step": 7641 + }, + { + "epoch": 3.5527661552766157, + "grad_norm": 5.884511947631836, + "learning_rate": 9.9923619992362e-06, + "loss": 0.2193, + "step": 7642 + }, + { + "epoch": 3.5532310553231055, + "grad_norm": 5.6018900871276855, + "learning_rate": 9.992360999236101e-06, + "loss": 0.4083, + "step": 7643 + }, + { + "epoch": 3.5536959553695953, + "grad_norm": 4.757307529449463, + "learning_rate": 9.992359999236002e-06, + "loss": 0.2968, + "step": 7644 + }, + { + "epoch": 3.5541608554160855, + "grad_norm": 4.502208709716797, + "learning_rate": 9.9923589992359e-06, + "loss": 0.4121, + "step": 7645 + }, + { + "epoch": 3.5546257554625758, + "grad_norm": 7.2621941566467285, + "learning_rate": 9.992357999235801e-06, + "loss": 0.7061, + "step": 7646 + }, + { + "epoch": 3.5550906555090656, + "grad_norm": 4.529594421386719, + "learning_rate": 9.9923569992357e-06, + "loss": 0.3407, + "step": 7647 + }, + { + "epoch": 3.5555555555555554, + "grad_norm": 6.0386857986450195, + "learning_rate": 9.992355999235601e-06, + "loss": 0.5821, + "step": 7648 + }, + { + "epoch": 3.5560204556020456, + "grad_norm": 5.958247184753418, + "learning_rate": 9.9923549992355e-06, + "loss": 0.5951, + "step": 7649 + }, + { + "epoch": 3.5564853556485354, + "grad_norm": 4.541553497314453, + "learning_rate": 9.9923539992354e-06, + "loss": 0.2044, + "step": 7650 + }, + { + "epoch": 3.5569502556950257, + "grad_norm": 7.10795259475708, + "learning_rate": 9.992352999235301e-06, + "loss": 0.3204, + "step": 7651 + }, + { + "epoch": 3.5574151557415155, + "grad_norm": 4.937320232391357, + "learning_rate": 9.9923519992352e-06, + "loss": 0.2889, + "step": 7652 + }, + { + "epoch": 3.5578800557880057, + "grad_norm": 5.679959774017334, + "learning_rate": 9.9923509992351e-06, + "loss": 0.314, + "step": 7653 + }, + { + "epoch": 3.5583449558344955, + "grad_norm": 6.989568710327148, + "learning_rate": 9.992349999235e-06, + "loss": 0.5531, + "step": 7654 + }, + { + "epoch": 3.5588098558809858, + "grad_norm": 5.315104961395264, + "learning_rate": 9.9923489992349e-06, + "loss": 0.3019, + "step": 7655 + }, + { + "epoch": 3.5592747559274756, + "grad_norm": 4.595499038696289, + "learning_rate": 9.992347999234801e-06, + "loss": 0.2191, + "step": 7656 + }, + { + "epoch": 3.5597396559739654, + "grad_norm": 4.084751605987549, + "learning_rate": 9.992346999234702e-06, + "loss": 0.2557, + "step": 7657 + }, + { + "epoch": 3.5602045560204556, + "grad_norm": 4.678102493286133, + "learning_rate": 9.9923459992346e-06, + "loss": 0.3643, + "step": 7658 + }, + { + "epoch": 3.560669456066946, + "grad_norm": 8.708230018615723, + "learning_rate": 9.992344999234501e-06, + "loss": 0.4073, + "step": 7659 + }, + { + "epoch": 3.5611343561134357, + "grad_norm": 5.489797592163086, + "learning_rate": 9.9923439992344e-06, + "loss": 0.3649, + "step": 7660 + }, + { + "epoch": 3.5615992561599255, + "grad_norm": 3.5926766395568848, + "learning_rate": 9.992342999234301e-06, + "loss": 0.1311, + "step": 7661 + }, + { + "epoch": 3.5620641562064157, + "grad_norm": 5.295966625213623, + "learning_rate": 9.9923419992342e-06, + "loss": 0.3516, + "step": 7662 + }, + { + "epoch": 3.5625290562529055, + "grad_norm": 4.87642240524292, + "learning_rate": 9.9923409992341e-06, + "loss": 0.2834, + "step": 7663 + }, + { + "epoch": 3.5629939562993957, + "grad_norm": 4.357379913330078, + "learning_rate": 9.992339999234001e-06, + "loss": 0.2797, + "step": 7664 + }, + { + "epoch": 3.5634588563458856, + "grad_norm": 3.899825096130371, + "learning_rate": 9.9923389992339e-06, + "loss": 0.2203, + "step": 7665 + }, + { + "epoch": 3.563923756392376, + "grad_norm": 7.021658420562744, + "learning_rate": 9.992337999233801e-06, + "loss": 0.6521, + "step": 7666 + }, + { + "epoch": 3.5643886564388656, + "grad_norm": 6.699857234954834, + "learning_rate": 9.9923369992337e-06, + "loss": 0.3541, + "step": 7667 + }, + { + "epoch": 3.564853556485356, + "grad_norm": 5.998404502868652, + "learning_rate": 9.9923359992336e-06, + "loss": 0.3622, + "step": 7668 + }, + { + "epoch": 3.5653184565318456, + "grad_norm": 4.153057098388672, + "learning_rate": 9.992334999233501e-06, + "loss": 0.2721, + "step": 7669 + }, + { + "epoch": 3.5657833565783355, + "grad_norm": 4.376986026763916, + "learning_rate": 9.992333999233402e-06, + "loss": 0.2846, + "step": 7670 + }, + { + "epoch": 3.5662482566248257, + "grad_norm": 3.8964781761169434, + "learning_rate": 9.9923329992333e-06, + "loss": 0.2579, + "step": 7671 + }, + { + "epoch": 3.566713156671316, + "grad_norm": 4.389614582061768, + "learning_rate": 9.9923319992332e-06, + "loss": 0.3377, + "step": 7672 + }, + { + "epoch": 3.5671780567178057, + "grad_norm": 3.260104179382324, + "learning_rate": 9.9923309992331e-06, + "loss": 0.259, + "step": 7673 + }, + { + "epoch": 3.5676429567642955, + "grad_norm": 6.5744123458862305, + "learning_rate": 9.992329999233001e-06, + "loss": 0.3917, + "step": 7674 + }, + { + "epoch": 3.568107856810786, + "grad_norm": 6.699383735656738, + "learning_rate": 9.9923289992329e-06, + "loss": 0.6281, + "step": 7675 + }, + { + "epoch": 3.5685727568572756, + "grad_norm": 5.591184616088867, + "learning_rate": 9.9923279992328e-06, + "loss": 0.456, + "step": 7676 + }, + { + "epoch": 3.569037656903766, + "grad_norm": 8.110832214355469, + "learning_rate": 9.992326999232701e-06, + "loss": 0.632, + "step": 7677 + }, + { + "epoch": 3.5695025569502556, + "grad_norm": 4.557453155517578, + "learning_rate": 9.9923259992326e-06, + "loss": 0.185, + "step": 7678 + }, + { + "epoch": 3.5699674569967454, + "grad_norm": 5.21386194229126, + "learning_rate": 9.992324999232501e-06, + "loss": 0.3056, + "step": 7679 + }, + { + "epoch": 3.5704323570432357, + "grad_norm": 4.298923969268799, + "learning_rate": 9.9923239992324e-06, + "loss": 0.2526, + "step": 7680 + }, + { + "epoch": 3.570897257089726, + "grad_norm": 6.933964729309082, + "learning_rate": 9.9923229992323e-06, + "loss": 0.4068, + "step": 7681 + }, + { + "epoch": 3.5713621571362157, + "grad_norm": 4.225578308105469, + "learning_rate": 9.992321999232201e-06, + "loss": 0.2535, + "step": 7682 + }, + { + "epoch": 3.5718270571827055, + "grad_norm": 9.618477821350098, + "learning_rate": 9.9923209992321e-06, + "loss": 0.3866, + "step": 7683 + }, + { + "epoch": 3.572291957229196, + "grad_norm": 5.340106010437012, + "learning_rate": 9.992319999232001e-06, + "loss": 0.4406, + "step": 7684 + }, + { + "epoch": 3.5727568572756856, + "grad_norm": 8.615200996398926, + "learning_rate": 9.9923189992319e-06, + "loss": 0.5977, + "step": 7685 + }, + { + "epoch": 3.573221757322176, + "grad_norm": 6.510963439941406, + "learning_rate": 9.9923179992318e-06, + "loss": 0.3743, + "step": 7686 + }, + { + "epoch": 3.5736866573686656, + "grad_norm": 5.155163764953613, + "learning_rate": 9.992316999231701e-06, + "loss": 0.36, + "step": 7687 + }, + { + "epoch": 3.574151557415156, + "grad_norm": 3.3346681594848633, + "learning_rate": 9.992315999231602e-06, + "loss": 0.1474, + "step": 7688 + }, + { + "epoch": 3.5746164574616457, + "grad_norm": 4.467373847961426, + "learning_rate": 9.9923149992315e-06, + "loss": 0.2877, + "step": 7689 + }, + { + "epoch": 3.575081357508136, + "grad_norm": 5.136261463165283, + "learning_rate": 9.992313999231401e-06, + "loss": 0.4746, + "step": 7690 + }, + { + "epoch": 3.5755462575546257, + "grad_norm": 5.221466541290283, + "learning_rate": 9.9923129992313e-06, + "loss": 0.4181, + "step": 7691 + }, + { + "epoch": 3.5760111576011155, + "grad_norm": 4.943102836608887, + "learning_rate": 9.992311999231201e-06, + "loss": 0.1566, + "step": 7692 + }, + { + "epoch": 3.576476057647606, + "grad_norm": 5.5060133934021, + "learning_rate": 9.9923109992311e-06, + "loss": 0.3242, + "step": 7693 + }, + { + "epoch": 3.576940957694096, + "grad_norm": 7.300727367401123, + "learning_rate": 9.992309999231e-06, + "loss": 0.3442, + "step": 7694 + }, + { + "epoch": 3.577405857740586, + "grad_norm": 4.941981315612793, + "learning_rate": 9.992308999230901e-06, + "loss": 0.2886, + "step": 7695 + }, + { + "epoch": 3.5778707577870756, + "grad_norm": 5.935998439788818, + "learning_rate": 9.9923079992308e-06, + "loss": 0.5177, + "step": 7696 + }, + { + "epoch": 3.578335657833566, + "grad_norm": 5.194117546081543, + "learning_rate": 9.992306999230701e-06, + "loss": 0.4121, + "step": 7697 + }, + { + "epoch": 3.5788005578800557, + "grad_norm": 4.320762634277344, + "learning_rate": 9.9923059992306e-06, + "loss": 0.2271, + "step": 7698 + }, + { + "epoch": 3.579265457926546, + "grad_norm": 4.800943851470947, + "learning_rate": 9.9923049992305e-06, + "loss": 0.3551, + "step": 7699 + }, + { + "epoch": 3.5797303579730357, + "grad_norm": 5.179135322570801, + "learning_rate": 9.992303999230401e-06, + "loss": 0.3415, + "step": 7700 + }, + { + "epoch": 3.5801952580195255, + "grad_norm": 4.966032028198242, + "learning_rate": 9.992302999230302e-06, + "loss": 0.2914, + "step": 7701 + }, + { + "epoch": 3.580660158066016, + "grad_norm": 5.833197593688965, + "learning_rate": 9.992301999230201e-06, + "loss": 0.3158, + "step": 7702 + }, + { + "epoch": 3.581125058112506, + "grad_norm": 5.497900009155273, + "learning_rate": 9.992300999230102e-06, + "loss": 0.177, + "step": 7703 + }, + { + "epoch": 3.581589958158996, + "grad_norm": 4.739005088806152, + "learning_rate": 9.99229999923e-06, + "loss": 0.3739, + "step": 7704 + }, + { + "epoch": 3.5820548582054856, + "grad_norm": 5.817606449127197, + "learning_rate": 9.9922989992299e-06, + "loss": 0.3427, + "step": 7705 + }, + { + "epoch": 3.582519758251976, + "grad_norm": 4.200488567352295, + "learning_rate": 9.9922979992298e-06, + "loss": 0.2816, + "step": 7706 + }, + { + "epoch": 3.5829846582984657, + "grad_norm": 4.654976844787598, + "learning_rate": 9.9922969992297e-06, + "loss": 0.3991, + "step": 7707 + }, + { + "epoch": 3.583449558344956, + "grad_norm": 6.431690692901611, + "learning_rate": 9.992295999229601e-06, + "loss": 0.4039, + "step": 7708 + }, + { + "epoch": 3.5839144583914457, + "grad_norm": 9.210029602050781, + "learning_rate": 9.9922949992295e-06, + "loss": 0.9733, + "step": 7709 + }, + { + "epoch": 3.584379358437936, + "grad_norm": 3.884315013885498, + "learning_rate": 9.992293999229401e-06, + "loss": 0.1703, + "step": 7710 + }, + { + "epoch": 3.584844258484426, + "grad_norm": 5.696138858795166, + "learning_rate": 9.9922929992293e-06, + "loss": 0.3615, + "step": 7711 + }, + { + "epoch": 3.585309158530916, + "grad_norm": 5.485318183898926, + "learning_rate": 9.9922919992292e-06, + "loss": 0.3372, + "step": 7712 + }, + { + "epoch": 3.585774058577406, + "grad_norm": 6.581882476806641, + "learning_rate": 9.992290999229101e-06, + "loss": 0.4436, + "step": 7713 + }, + { + "epoch": 3.5862389586238956, + "grad_norm": 7.477715969085693, + "learning_rate": 9.992289999229002e-06, + "loss": 0.6387, + "step": 7714 + }, + { + "epoch": 3.586703858670386, + "grad_norm": 5.072680473327637, + "learning_rate": 9.992288999228901e-06, + "loss": 0.3645, + "step": 7715 + }, + { + "epoch": 3.587168758716876, + "grad_norm": 4.754345893859863, + "learning_rate": 9.9922879992288e-06, + "loss": 0.367, + "step": 7716 + }, + { + "epoch": 3.587633658763366, + "grad_norm": 8.206717491149902, + "learning_rate": 9.9922869992287e-06, + "loss": 0.3244, + "step": 7717 + }, + { + "epoch": 3.5880985588098557, + "grad_norm": 12.580855369567871, + "learning_rate": 9.9922859992286e-06, + "loss": 0.3755, + "step": 7718 + }, + { + "epoch": 3.588563458856346, + "grad_norm": 8.856315612792969, + "learning_rate": 9.9922849992285e-06, + "loss": 0.6988, + "step": 7719 + }, + { + "epoch": 3.589028358902836, + "grad_norm": 5.7649455070495605, + "learning_rate": 9.992283999228401e-06, + "loss": 0.7617, + "step": 7720 + }, + { + "epoch": 3.589493258949326, + "grad_norm": 5.480391502380371, + "learning_rate": 9.992282999228302e-06, + "loss": 0.3109, + "step": 7721 + }, + { + "epoch": 3.589958158995816, + "grad_norm": 5.874509334564209, + "learning_rate": 9.9922819992282e-06, + "loss": 0.4862, + "step": 7722 + }, + { + "epoch": 3.5904230590423056, + "grad_norm": 6.981359481811523, + "learning_rate": 9.992280999228101e-06, + "loss": 0.4713, + "step": 7723 + }, + { + "epoch": 3.590887959088796, + "grad_norm": 5.761654376983643, + "learning_rate": 9.992279999228e-06, + "loss": 0.5181, + "step": 7724 + }, + { + "epoch": 3.591352859135286, + "grad_norm": 5.179229259490967, + "learning_rate": 9.9922789992279e-06, + "loss": 0.1723, + "step": 7725 + }, + { + "epoch": 3.591817759181776, + "grad_norm": 4.975121021270752, + "learning_rate": 9.992277999227801e-06, + "loss": 0.28, + "step": 7726 + }, + { + "epoch": 3.5922826592282657, + "grad_norm": 5.438337802886963, + "learning_rate": 9.9922769992277e-06, + "loss": 0.4312, + "step": 7727 + }, + { + "epoch": 3.592747559274756, + "grad_norm": 6.175811767578125, + "learning_rate": 9.992275999227601e-06, + "loss": 0.6929, + "step": 7728 + }, + { + "epoch": 3.5932124593212458, + "grad_norm": 4.661355495452881, + "learning_rate": 9.9922749992275e-06, + "loss": 0.4292, + "step": 7729 + }, + { + "epoch": 3.593677359367736, + "grad_norm": 3.6983299255371094, + "learning_rate": 9.9922739992274e-06, + "loss": 0.345, + "step": 7730 + }, + { + "epoch": 3.594142259414226, + "grad_norm": 3.261418104171753, + "learning_rate": 9.9922729992273e-06, + "loss": 0.1725, + "step": 7731 + }, + { + "epoch": 3.594607159460716, + "grad_norm": 4.587123394012451, + "learning_rate": 9.9922719992272e-06, + "loss": 0.3356, + "step": 7732 + }, + { + "epoch": 3.595072059507206, + "grad_norm": 6.912278175354004, + "learning_rate": 9.992270999227101e-06, + "loss": 0.5309, + "step": 7733 + }, + { + "epoch": 3.595536959553696, + "grad_norm": 5.379083156585693, + "learning_rate": 9.992269999227002e-06, + "loss": 0.3195, + "step": 7734 + }, + { + "epoch": 3.596001859600186, + "grad_norm": 4.436120510101318, + "learning_rate": 9.9922689992269e-06, + "loss": 0.3975, + "step": 7735 + }, + { + "epoch": 3.5964667596466757, + "grad_norm": 5.984475612640381, + "learning_rate": 9.992267999226801e-06, + "loss": 0.4459, + "step": 7736 + }, + { + "epoch": 3.596931659693166, + "grad_norm": 4.712123870849609, + "learning_rate": 9.9922669992267e-06, + "loss": 0.2693, + "step": 7737 + }, + { + "epoch": 3.597396559739656, + "grad_norm": 4.934001445770264, + "learning_rate": 9.9922659992266e-06, + "loss": 0.3269, + "step": 7738 + }, + { + "epoch": 3.597861459786146, + "grad_norm": 4.467751979827881, + "learning_rate": 9.992264999226501e-06, + "loss": 0.2733, + "step": 7739 + }, + { + "epoch": 3.598326359832636, + "grad_norm": 4.503422260284424, + "learning_rate": 9.9922639992264e-06, + "loss": 0.3372, + "step": 7740 + }, + { + "epoch": 3.598791259879126, + "grad_norm": 3.9818949699401855, + "learning_rate": 9.992262999226301e-06, + "loss": 0.2589, + "step": 7741 + }, + { + "epoch": 3.599256159925616, + "grad_norm": 5.4304585456848145, + "learning_rate": 9.9922619992262e-06, + "loss": 0.3693, + "step": 7742 + }, + { + "epoch": 3.599721059972106, + "grad_norm": 6.803290843963623, + "learning_rate": 9.9922609992261e-06, + "loss": 0.4531, + "step": 7743 + }, + { + "epoch": 3.600185960018596, + "grad_norm": 6.348935604095459, + "learning_rate": 9.992259999226e-06, + "loss": 0.3031, + "step": 7744 + }, + { + "epoch": 3.600650860065086, + "grad_norm": 5.715703964233398, + "learning_rate": 9.9922589992259e-06, + "loss": 0.4585, + "step": 7745 + }, + { + "epoch": 3.601115760111576, + "grad_norm": 3.928260326385498, + "learning_rate": 9.992257999225801e-06, + "loss": 0.2119, + "step": 7746 + }, + { + "epoch": 3.601580660158066, + "grad_norm": 6.860134124755859, + "learning_rate": 9.992256999225702e-06, + "loss": 0.3349, + "step": 7747 + }, + { + "epoch": 3.602045560204556, + "grad_norm": 4.085231781005859, + "learning_rate": 9.9922559992256e-06, + "loss": 0.2701, + "step": 7748 + }, + { + "epoch": 3.602510460251046, + "grad_norm": 3.4699361324310303, + "learning_rate": 9.9922549992255e-06, + "loss": 0.1545, + "step": 7749 + }, + { + "epoch": 3.602975360297536, + "grad_norm": 7.107826232910156, + "learning_rate": 9.9922539992254e-06, + "loss": 0.4542, + "step": 7750 + }, + { + "epoch": 3.603440260344026, + "grad_norm": 5.522230625152588, + "learning_rate": 9.992252999225301e-06, + "loss": 0.1141, + "step": 7751 + }, + { + "epoch": 3.603905160390516, + "grad_norm": 4.582331657409668, + "learning_rate": 9.992251999225202e-06, + "loss": 0.3124, + "step": 7752 + }, + { + "epoch": 3.604370060437006, + "grad_norm": 10.794573783874512, + "learning_rate": 9.9922509992251e-06, + "loss": 0.7539, + "step": 7753 + }, + { + "epoch": 3.604834960483496, + "grad_norm": 4.111024856567383, + "learning_rate": 9.992249999225001e-06, + "loss": 0.3645, + "step": 7754 + }, + { + "epoch": 3.605299860529986, + "grad_norm": 6.089413166046143, + "learning_rate": 9.9922489992249e-06, + "loss": 0.1489, + "step": 7755 + }, + { + "epoch": 3.605764760576476, + "grad_norm": 7.947140693664551, + "learning_rate": 9.9922479992248e-06, + "loss": 0.7059, + "step": 7756 + }, + { + "epoch": 3.606229660622966, + "grad_norm": 4.065314769744873, + "learning_rate": 9.992246999224701e-06, + "loss": 0.205, + "step": 7757 + }, + { + "epoch": 3.606694560669456, + "grad_norm": 6.1632399559021, + "learning_rate": 9.9922459992246e-06, + "loss": 0.6584, + "step": 7758 + }, + { + "epoch": 3.607159460715946, + "grad_norm": 4.3146562576293945, + "learning_rate": 9.992244999224501e-06, + "loss": 0.2831, + "step": 7759 + }, + { + "epoch": 3.6076243607624363, + "grad_norm": 5.536690711975098, + "learning_rate": 9.9922439992244e-06, + "loss": 0.2801, + "step": 7760 + }, + { + "epoch": 3.608089260808926, + "grad_norm": 6.127683639526367, + "learning_rate": 9.9922429992243e-06, + "loss": 0.4021, + "step": 7761 + }, + { + "epoch": 3.608554160855416, + "grad_norm": 2.830590009689331, + "learning_rate": 9.9922419992242e-06, + "loss": 0.0771, + "step": 7762 + }, + { + "epoch": 3.609019060901906, + "grad_norm": 3.8270554542541504, + "learning_rate": 9.9922409992241e-06, + "loss": 0.2421, + "step": 7763 + }, + { + "epoch": 3.609483960948396, + "grad_norm": 5.191068172454834, + "learning_rate": 9.992239999224001e-06, + "loss": 0.263, + "step": 7764 + }, + { + "epoch": 3.609948860994886, + "grad_norm": 10.380234718322754, + "learning_rate": 9.992238999223902e-06, + "loss": 0.6194, + "step": 7765 + }, + { + "epoch": 3.610413761041376, + "grad_norm": 4.096468448638916, + "learning_rate": 9.9922379992238e-06, + "loss": 0.2303, + "step": 7766 + }, + { + "epoch": 3.6108786610878663, + "grad_norm": 4.3341145515441895, + "learning_rate": 9.992236999223701e-06, + "loss": 0.299, + "step": 7767 + }, + { + "epoch": 3.611343561134356, + "grad_norm": 7.0861711502075195, + "learning_rate": 9.9922359992236e-06, + "loss": 0.4366, + "step": 7768 + }, + { + "epoch": 3.6118084611808463, + "grad_norm": 5.588967800140381, + "learning_rate": 9.992234999223501e-06, + "loss": 0.333, + "step": 7769 + }, + { + "epoch": 3.612273361227336, + "grad_norm": 7.820956230163574, + "learning_rate": 9.992233999223402e-06, + "loss": 1.2061, + "step": 7770 + }, + { + "epoch": 3.612738261273826, + "grad_norm": 6.1866912841796875, + "learning_rate": 9.9922329992233e-06, + "loss": 0.7549, + "step": 7771 + }, + { + "epoch": 3.613203161320316, + "grad_norm": 6.81213903427124, + "learning_rate": 9.992231999223201e-06, + "loss": 0.5298, + "step": 7772 + }, + { + "epoch": 3.6136680613668064, + "grad_norm": 4.383517742156982, + "learning_rate": 9.9922309992231e-06, + "loss": 0.3756, + "step": 7773 + }, + { + "epoch": 3.614132961413296, + "grad_norm": 8.462226867675781, + "learning_rate": 9.992229999223e-06, + "loss": 0.528, + "step": 7774 + }, + { + "epoch": 3.614597861459786, + "grad_norm": 12.879956245422363, + "learning_rate": 9.9922289992229e-06, + "loss": 0.1208, + "step": 7775 + }, + { + "epoch": 3.6150627615062763, + "grad_norm": 4.106274604797363, + "learning_rate": 9.9922279992228e-06, + "loss": 0.2283, + "step": 7776 + }, + { + "epoch": 3.615527661552766, + "grad_norm": 5.536767482757568, + "learning_rate": 9.992226999222701e-06, + "loss": 0.2539, + "step": 7777 + }, + { + "epoch": 3.6159925615992563, + "grad_norm": 4.358861446380615, + "learning_rate": 9.992225999222602e-06, + "loss": 0.1817, + "step": 7778 + }, + { + "epoch": 3.616457461645746, + "grad_norm": 7.3513288497924805, + "learning_rate": 9.9922249992225e-06, + "loss": 0.4852, + "step": 7779 + }, + { + "epoch": 3.616922361692236, + "grad_norm": 5.381132125854492, + "learning_rate": 9.992223999222401e-06, + "loss": 0.3607, + "step": 7780 + }, + { + "epoch": 3.617387261738726, + "grad_norm": 4.626286029815674, + "learning_rate": 9.9922229992223e-06, + "loss": 0.4348, + "step": 7781 + }, + { + "epoch": 3.6178521617852164, + "grad_norm": 5.327314853668213, + "learning_rate": 9.992221999222201e-06, + "loss": 0.3832, + "step": 7782 + }, + { + "epoch": 3.618317061831706, + "grad_norm": 4.065019130706787, + "learning_rate": 9.992220999222102e-06, + "loss": 0.233, + "step": 7783 + }, + { + "epoch": 3.618781961878196, + "grad_norm": 5.424410343170166, + "learning_rate": 9.992219999222e-06, + "loss": 0.4302, + "step": 7784 + }, + { + "epoch": 3.6192468619246863, + "grad_norm": 5.755020618438721, + "learning_rate": 9.992218999221901e-06, + "loss": 0.3404, + "step": 7785 + }, + { + "epoch": 3.619711761971176, + "grad_norm": 6.148556709289551, + "learning_rate": 9.9922179992218e-06, + "loss": 0.2837, + "step": 7786 + }, + { + "epoch": 3.6201766620176663, + "grad_norm": 3.9078524112701416, + "learning_rate": 9.992216999221701e-06, + "loss": 0.1625, + "step": 7787 + }, + { + "epoch": 3.620641562064156, + "grad_norm": 4.906919956207275, + "learning_rate": 9.9922159992216e-06, + "loss": 0.3251, + "step": 7788 + }, + { + "epoch": 3.6211064621106464, + "grad_norm": 3.2012856006622314, + "learning_rate": 9.9922149992215e-06, + "loss": 0.1075, + "step": 7789 + }, + { + "epoch": 3.621571362157136, + "grad_norm": 5.7078447341918945, + "learning_rate": 9.992213999221401e-06, + "loss": 0.3721, + "step": 7790 + }, + { + "epoch": 3.6220362622036264, + "grad_norm": 3.709510564804077, + "learning_rate": 9.992212999221302e-06, + "loss": 0.2128, + "step": 7791 + }, + { + "epoch": 3.622501162250116, + "grad_norm": 4.014693737030029, + "learning_rate": 9.9922119992212e-06, + "loss": 0.2432, + "step": 7792 + }, + { + "epoch": 3.622966062296606, + "grad_norm": 3.577404737472534, + "learning_rate": 9.9922109992211e-06, + "loss": 0.2254, + "step": 7793 + }, + { + "epoch": 3.6234309623430963, + "grad_norm": 7.611312389373779, + "learning_rate": 9.992209999221e-06, + "loss": 0.5972, + "step": 7794 + }, + { + "epoch": 3.6238958623895865, + "grad_norm": 9.258544921875, + "learning_rate": 9.992208999220901e-06, + "loss": 0.6245, + "step": 7795 + }, + { + "epoch": 3.6243607624360763, + "grad_norm": 4.790771961212158, + "learning_rate": 9.992207999220802e-06, + "loss": 0.2868, + "step": 7796 + }, + { + "epoch": 3.624825662482566, + "grad_norm": 5.487369060516357, + "learning_rate": 9.9922069992207e-06, + "loss": 0.3392, + "step": 7797 + }, + { + "epoch": 3.6252905625290563, + "grad_norm": 6.224166393280029, + "learning_rate": 9.992205999220601e-06, + "loss": 0.2542, + "step": 7798 + }, + { + "epoch": 3.625755462575546, + "grad_norm": 5.30756950378418, + "learning_rate": 9.9922049992205e-06, + "loss": 0.3478, + "step": 7799 + }, + { + "epoch": 3.6262203626220364, + "grad_norm": 6.792389869689941, + "learning_rate": 9.992203999220401e-06, + "loss": 0.3537, + "step": 7800 + }, + { + "epoch": 3.626685262668526, + "grad_norm": 5.018935680389404, + "learning_rate": 9.9922029992203e-06, + "loss": 0.2456, + "step": 7801 + }, + { + "epoch": 3.627150162715016, + "grad_norm": 9.025073051452637, + "learning_rate": 9.9922019992202e-06, + "loss": 0.6366, + "step": 7802 + }, + { + "epoch": 3.6276150627615062, + "grad_norm": 7.2631659507751465, + "learning_rate": 9.992200999220101e-06, + "loss": 0.2872, + "step": 7803 + }, + { + "epoch": 3.6280799628079965, + "grad_norm": 3.471667766571045, + "learning_rate": 9.99219999922e-06, + "loss": 0.1637, + "step": 7804 + }, + { + "epoch": 3.6285448628544863, + "grad_norm": 2.9803712368011475, + "learning_rate": 9.992198999219901e-06, + "loss": 0.1464, + "step": 7805 + }, + { + "epoch": 3.629009762900976, + "grad_norm": 5.270829677581787, + "learning_rate": 9.9921979992198e-06, + "loss": 0.2833, + "step": 7806 + }, + { + "epoch": 3.6294746629474663, + "grad_norm": 4.921884059906006, + "learning_rate": 9.9921969992197e-06, + "loss": 0.328, + "step": 7807 + }, + { + "epoch": 3.629939562993956, + "grad_norm": 5.01820182800293, + "learning_rate": 9.992195999219601e-06, + "loss": 0.4047, + "step": 7808 + }, + { + "epoch": 3.6304044630404464, + "grad_norm": 4.442251682281494, + "learning_rate": 9.992194999219502e-06, + "loss": 0.1669, + "step": 7809 + }, + { + "epoch": 3.630869363086936, + "grad_norm": 2.997722864151001, + "learning_rate": 9.9921939992194e-06, + "loss": 0.1717, + "step": 7810 + }, + { + "epoch": 3.6313342631334264, + "grad_norm": 4.580142021179199, + "learning_rate": 9.992192999219301e-06, + "loss": 0.2218, + "step": 7811 + }, + { + "epoch": 3.6317991631799162, + "grad_norm": 6.099818706512451, + "learning_rate": 9.9921919992192e-06, + "loss": 0.4013, + "step": 7812 + }, + { + "epoch": 3.6322640632264065, + "grad_norm": 4.629861831665039, + "learning_rate": 9.9921909992191e-06, + "loss": 0.2417, + "step": 7813 + }, + { + "epoch": 3.6327289632728963, + "grad_norm": 7.314176082611084, + "learning_rate": 9.992189999219e-06, + "loss": 0.3865, + "step": 7814 + }, + { + "epoch": 3.633193863319386, + "grad_norm": 4.082392692565918, + "learning_rate": 9.9921889992189e-06, + "loss": 0.1925, + "step": 7815 + }, + { + "epoch": 3.6336587633658763, + "grad_norm": 6.543286323547363, + "learning_rate": 9.992187999218801e-06, + "loss": 0.4544, + "step": 7816 + }, + { + "epoch": 3.6341236634123666, + "grad_norm": 6.520508289337158, + "learning_rate": 9.9921869992187e-06, + "loss": 0.3854, + "step": 7817 + }, + { + "epoch": 3.6345885634588564, + "grad_norm": 11.645988464355469, + "learning_rate": 9.992185999218601e-06, + "loss": 0.5833, + "step": 7818 + }, + { + "epoch": 3.635053463505346, + "grad_norm": 4.940713405609131, + "learning_rate": 9.9921849992185e-06, + "loss": 0.275, + "step": 7819 + }, + { + "epoch": 3.6355183635518364, + "grad_norm": 4.332046985626221, + "learning_rate": 9.9921839992184e-06, + "loss": 0.3677, + "step": 7820 + }, + { + "epoch": 3.6359832635983262, + "grad_norm": 6.757460117340088, + "learning_rate": 9.992182999218301e-06, + "loss": 0.438, + "step": 7821 + }, + { + "epoch": 3.6364481636448165, + "grad_norm": 7.974299430847168, + "learning_rate": 9.992181999218202e-06, + "loss": 0.1704, + "step": 7822 + }, + { + "epoch": 3.6369130636913063, + "grad_norm": 5.697950839996338, + "learning_rate": 9.9921809992181e-06, + "loss": 0.3856, + "step": 7823 + }, + { + "epoch": 3.6373779637377965, + "grad_norm": 4.528530120849609, + "learning_rate": 9.992179999218e-06, + "loss": 0.2879, + "step": 7824 + }, + { + "epoch": 3.6378428637842863, + "grad_norm": 5.187484264373779, + "learning_rate": 9.9921789992179e-06, + "loss": 0.4089, + "step": 7825 + }, + { + "epoch": 3.6383077638307766, + "grad_norm": 5.711635589599609, + "learning_rate": 9.992177999217801e-06, + "loss": 0.4909, + "step": 7826 + }, + { + "epoch": 3.6387726638772664, + "grad_norm": 5.756196975708008, + "learning_rate": 9.9921769992177e-06, + "loss": 0.4141, + "step": 7827 + }, + { + "epoch": 3.639237563923756, + "grad_norm": 3.8830618858337402, + "learning_rate": 9.9921759992176e-06, + "loss": 0.3024, + "step": 7828 + }, + { + "epoch": 3.6397024639702464, + "grad_norm": 4.319950103759766, + "learning_rate": 9.992174999217501e-06, + "loss": 0.3005, + "step": 7829 + }, + { + "epoch": 3.6401673640167362, + "grad_norm": 4.036876678466797, + "learning_rate": 9.9921739992174e-06, + "loss": 0.185, + "step": 7830 + }, + { + "epoch": 3.6406322640632265, + "grad_norm": 4.8755927085876465, + "learning_rate": 9.992172999217301e-06, + "loss": 0.3285, + "step": 7831 + }, + { + "epoch": 3.6410971641097163, + "grad_norm": 4.478252410888672, + "learning_rate": 9.9921719992172e-06, + "loss": 0.3403, + "step": 7832 + }, + { + "epoch": 3.6415620641562065, + "grad_norm": 5.593343734741211, + "learning_rate": 9.9921709992171e-06, + "loss": 0.1877, + "step": 7833 + }, + { + "epoch": 3.6420269642026963, + "grad_norm": 6.74568510055542, + "learning_rate": 9.992169999217001e-06, + "loss": 0.6181, + "step": 7834 + }, + { + "epoch": 3.6424918642491866, + "grad_norm": 7.964699745178223, + "learning_rate": 9.9921689992169e-06, + "loss": 0.4689, + "step": 7835 + }, + { + "epoch": 3.6429567642956764, + "grad_norm": 4.929378032684326, + "learning_rate": 9.992167999216801e-06, + "loss": 0.4026, + "step": 7836 + }, + { + "epoch": 3.643421664342166, + "grad_norm": 8.78721809387207, + "learning_rate": 9.9921669992167e-06, + "loss": 0.4038, + "step": 7837 + }, + { + "epoch": 3.6438865643886564, + "grad_norm": 3.471409559249878, + "learning_rate": 9.9921659992166e-06, + "loss": 0.1382, + "step": 7838 + }, + { + "epoch": 3.6443514644351467, + "grad_norm": 4.201255798339844, + "learning_rate": 9.992164999216501e-06, + "loss": 0.3504, + "step": 7839 + }, + { + "epoch": 3.6448163644816365, + "grad_norm": 7.568720817565918, + "learning_rate": 9.9921639992164e-06, + "loss": 0.4267, + "step": 7840 + }, + { + "epoch": 3.6452812645281263, + "grad_norm": 5.75497579574585, + "learning_rate": 9.9921629992163e-06, + "loss": 0.5128, + "step": 7841 + }, + { + "epoch": 3.6457461645746165, + "grad_norm": 8.036436080932617, + "learning_rate": 9.992161999216201e-06, + "loss": 0.4789, + "step": 7842 + }, + { + "epoch": 3.6462110646211063, + "grad_norm": 8.494091033935547, + "learning_rate": 9.9921609992161e-06, + "loss": 0.7363, + "step": 7843 + }, + { + "epoch": 3.6466759646675966, + "grad_norm": 6.891714572906494, + "learning_rate": 9.992159999216001e-06, + "loss": 0.3921, + "step": 7844 + }, + { + "epoch": 3.6471408647140864, + "grad_norm": 5.83414888381958, + "learning_rate": 9.9921589992159e-06, + "loss": 0.3917, + "step": 7845 + }, + { + "epoch": 3.6476057647605766, + "grad_norm": 5.282825946807861, + "learning_rate": 9.9921579992158e-06, + "loss": 0.4018, + "step": 7846 + }, + { + "epoch": 3.6480706648070664, + "grad_norm": 5.395559310913086, + "learning_rate": 9.992156999215701e-06, + "loss": 0.384, + "step": 7847 + }, + { + "epoch": 3.6485355648535567, + "grad_norm": 6.815792083740234, + "learning_rate": 9.9921559992156e-06, + "loss": 0.4386, + "step": 7848 + }, + { + "epoch": 3.6490004649000465, + "grad_norm": 6.868009090423584, + "learning_rate": 9.992154999215501e-06, + "loss": 0.3241, + "step": 7849 + }, + { + "epoch": 3.6494653649465363, + "grad_norm": 4.987998962402344, + "learning_rate": 9.9921539992154e-06, + "loss": 0.4361, + "step": 7850 + }, + { + "epoch": 3.6499302649930265, + "grad_norm": 4.583704948425293, + "learning_rate": 9.9921529992153e-06, + "loss": 0.2954, + "step": 7851 + }, + { + "epoch": 3.6503951650395168, + "grad_norm": 4.562130451202393, + "learning_rate": 9.992151999215201e-06, + "loss": 0.194, + "step": 7852 + }, + { + "epoch": 3.6508600650860066, + "grad_norm": 7.124180316925049, + "learning_rate": 9.992150999215102e-06, + "loss": 0.7636, + "step": 7853 + }, + { + "epoch": 3.6513249651324964, + "grad_norm": 4.040521621704102, + "learning_rate": 9.992149999215001e-06, + "loss": 0.1829, + "step": 7854 + }, + { + "epoch": 3.6517898651789866, + "grad_norm": 4.314542770385742, + "learning_rate": 9.992148999214902e-06, + "loss": 0.2311, + "step": 7855 + }, + { + "epoch": 3.6522547652254764, + "grad_norm": 5.068157196044922, + "learning_rate": 9.9921479992148e-06, + "loss": 0.2567, + "step": 7856 + }, + { + "epoch": 3.6527196652719667, + "grad_norm": 3.7811875343322754, + "learning_rate": 9.9921469992147e-06, + "loss": 0.1985, + "step": 7857 + }, + { + "epoch": 3.6531845653184565, + "grad_norm": 8.170882225036621, + "learning_rate": 9.9921459992146e-06, + "loss": 0.5211, + "step": 7858 + }, + { + "epoch": 3.6536494653649463, + "grad_norm": 6.791769981384277, + "learning_rate": 9.9921449992145e-06, + "loss": 0.421, + "step": 7859 + }, + { + "epoch": 3.6541143654114365, + "grad_norm": 5.037191867828369, + "learning_rate": 9.992143999214401e-06, + "loss": 0.5529, + "step": 7860 + }, + { + "epoch": 3.6545792654579268, + "grad_norm": 6.049614906311035, + "learning_rate": 9.9921429992143e-06, + "loss": 0.4769, + "step": 7861 + }, + { + "epoch": 3.6550441655044166, + "grad_norm": 9.839277267456055, + "learning_rate": 9.992141999214201e-06, + "loss": 0.5678, + "step": 7862 + }, + { + "epoch": 3.6555090655509064, + "grad_norm": 3.230536699295044, + "learning_rate": 9.9921409992141e-06, + "loss": 0.2134, + "step": 7863 + }, + { + "epoch": 3.6559739655973966, + "grad_norm": 4.07127571105957, + "learning_rate": 9.992139999214e-06, + "loss": 0.2859, + "step": 7864 + }, + { + "epoch": 3.6564388656438864, + "grad_norm": 5.5447468757629395, + "learning_rate": 9.992138999213901e-06, + "loss": 0.4641, + "step": 7865 + }, + { + "epoch": 3.6569037656903767, + "grad_norm": 5.83288049697876, + "learning_rate": 9.992137999213802e-06, + "loss": 0.3507, + "step": 7866 + }, + { + "epoch": 3.6573686657368665, + "grad_norm": 7.7215495109558105, + "learning_rate": 9.992136999213701e-06, + "loss": 0.4493, + "step": 7867 + }, + { + "epoch": 3.6578335657833567, + "grad_norm": 4.695730686187744, + "learning_rate": 9.9921359992136e-06, + "loss": 0.5608, + "step": 7868 + }, + { + "epoch": 3.6582984658298465, + "grad_norm": 5.450352191925049, + "learning_rate": 9.9921349992135e-06, + "loss": 0.332, + "step": 7869 + }, + { + "epoch": 3.6587633658763368, + "grad_norm": 5.861698150634766, + "learning_rate": 9.9921339992134e-06, + "loss": 0.5543, + "step": 7870 + }, + { + "epoch": 3.6592282659228266, + "grad_norm": 5.233304500579834, + "learning_rate": 9.9921329992133e-06, + "loss": 0.2033, + "step": 7871 + }, + { + "epoch": 3.6596931659693164, + "grad_norm": 4.932979106903076, + "learning_rate": 9.992131999213201e-06, + "loss": 0.3552, + "step": 7872 + }, + { + "epoch": 3.6601580660158066, + "grad_norm": 3.8209996223449707, + "learning_rate": 9.992130999213102e-06, + "loss": 0.1969, + "step": 7873 + }, + { + "epoch": 3.660622966062297, + "grad_norm": 6.417337417602539, + "learning_rate": 9.992129999213e-06, + "loss": 0.5429, + "step": 7874 + }, + { + "epoch": 3.6610878661087867, + "grad_norm": 3.8892364501953125, + "learning_rate": 9.992128999212901e-06, + "loss": 0.2284, + "step": 7875 + }, + { + "epoch": 3.6615527661552765, + "grad_norm": 7.659767150878906, + "learning_rate": 9.9921279992128e-06, + "loss": 0.3263, + "step": 7876 + }, + { + "epoch": 3.6620176662017667, + "grad_norm": 4.399847984313965, + "learning_rate": 9.9921269992127e-06, + "loss": 0.2012, + "step": 7877 + }, + { + "epoch": 3.6624825662482565, + "grad_norm": 5.3279242515563965, + "learning_rate": 9.992125999212601e-06, + "loss": 0.4817, + "step": 7878 + }, + { + "epoch": 3.6629474662947468, + "grad_norm": 3.8096089363098145, + "learning_rate": 9.9921249992125e-06, + "loss": 0.2205, + "step": 7879 + }, + { + "epoch": 3.6634123663412366, + "grad_norm": 8.4736967086792, + "learning_rate": 9.992123999212401e-06, + "loss": 0.3025, + "step": 7880 + }, + { + "epoch": 3.6638772663877264, + "grad_norm": 3.4296915531158447, + "learning_rate": 9.9921229992123e-06, + "loss": 0.1568, + "step": 7881 + }, + { + "epoch": 3.6643421664342166, + "grad_norm": 5.1946234703063965, + "learning_rate": 9.9921219992122e-06, + "loss": 0.4438, + "step": 7882 + }, + { + "epoch": 3.664807066480707, + "grad_norm": 7.153382301330566, + "learning_rate": 9.9921209992121e-06, + "loss": 0.9987, + "step": 7883 + }, + { + "epoch": 3.6652719665271967, + "grad_norm": 4.758907794952393, + "learning_rate": 9.992119999212e-06, + "loss": 0.4233, + "step": 7884 + }, + { + "epoch": 3.6657368665736865, + "grad_norm": 7.073054313659668, + "learning_rate": 9.992118999211901e-06, + "loss": 0.4713, + "step": 7885 + }, + { + "epoch": 3.6662017666201767, + "grad_norm": 5.592694282531738, + "learning_rate": 9.992117999211802e-06, + "loss": 0.4288, + "step": 7886 + }, + { + "epoch": 3.6666666666666665, + "grad_norm": 9.169283866882324, + "learning_rate": 9.9921169992117e-06, + "loss": 0.5141, + "step": 7887 + }, + { + "epoch": 3.6671315667131568, + "grad_norm": 6.007300853729248, + "learning_rate": 9.992115999211601e-06, + "loss": 0.3259, + "step": 7888 + }, + { + "epoch": 3.6675964667596466, + "grad_norm": 4.49038028717041, + "learning_rate": 9.9921149992115e-06, + "loss": 0.2638, + "step": 7889 + }, + { + "epoch": 3.668061366806137, + "grad_norm": 6.08632755279541, + "learning_rate": 9.992113999211401e-06, + "loss": 0.3751, + "step": 7890 + }, + { + "epoch": 3.6685262668526266, + "grad_norm": 5.053904056549072, + "learning_rate": 9.992112999211302e-06, + "loss": 0.2023, + "step": 7891 + }, + { + "epoch": 3.668991166899117, + "grad_norm": 3.9729716777801514, + "learning_rate": 9.9921119992112e-06, + "loss": 0.2425, + "step": 7892 + }, + { + "epoch": 3.6694560669456067, + "grad_norm": 5.574167251586914, + "learning_rate": 9.992110999211101e-06, + "loss": 0.3941, + "step": 7893 + }, + { + "epoch": 3.6699209669920965, + "grad_norm": 5.910741806030273, + "learning_rate": 9.992109999211e-06, + "loss": 0.4911, + "step": 7894 + }, + { + "epoch": 3.6703858670385867, + "grad_norm": 8.12578010559082, + "learning_rate": 9.9921089992109e-06, + "loss": 0.3196, + "step": 7895 + }, + { + "epoch": 3.670850767085077, + "grad_norm": 6.086568832397461, + "learning_rate": 9.9921079992108e-06, + "loss": 0.3528, + "step": 7896 + }, + { + "epoch": 3.6713156671315668, + "grad_norm": 5.328800678253174, + "learning_rate": 9.9921069992107e-06, + "loss": 0.3848, + "step": 7897 + }, + { + "epoch": 3.6717805671780566, + "grad_norm": 7.500800609588623, + "learning_rate": 9.992105999210601e-06, + "loss": 0.5824, + "step": 7898 + }, + { + "epoch": 3.672245467224547, + "grad_norm": 4.0139312744140625, + "learning_rate": 9.992104999210502e-06, + "loss": 0.2298, + "step": 7899 + }, + { + "epoch": 3.6727103672710366, + "grad_norm": 5.405986785888672, + "learning_rate": 9.9921039992104e-06, + "loss": 0.3918, + "step": 7900 + }, + { + "epoch": 3.673175267317527, + "grad_norm": 9.800149917602539, + "learning_rate": 9.9921029992103e-06, + "loss": 0.5491, + "step": 7901 + }, + { + "epoch": 3.6736401673640167, + "grad_norm": 9.345376968383789, + "learning_rate": 9.9921019992102e-06, + "loss": 0.469, + "step": 7902 + }, + { + "epoch": 3.6741050674105065, + "grad_norm": 6.769668102264404, + "learning_rate": 9.992100999210101e-06, + "loss": 0.6265, + "step": 7903 + }, + { + "epoch": 3.6745699674569967, + "grad_norm": 6.146004676818848, + "learning_rate": 9.992099999210002e-06, + "loss": 0.2927, + "step": 7904 + }, + { + "epoch": 3.675034867503487, + "grad_norm": 2.342219352722168, + "learning_rate": 9.9920989992099e-06, + "loss": 0.1261, + "step": 7905 + }, + { + "epoch": 3.6754997675499768, + "grad_norm": 5.387047290802002, + "learning_rate": 9.992097999209801e-06, + "loss": 0.2122, + "step": 7906 + }, + { + "epoch": 3.6759646675964666, + "grad_norm": 4.8292951583862305, + "learning_rate": 9.9920969992097e-06, + "loss": 0.1554, + "step": 7907 + }, + { + "epoch": 3.676429567642957, + "grad_norm": 10.27605152130127, + "learning_rate": 9.992095999209601e-06, + "loss": 0.5241, + "step": 7908 + }, + { + "epoch": 3.6768944676894466, + "grad_norm": 6.507298469543457, + "learning_rate": 9.9920949992095e-06, + "loss": 0.5365, + "step": 7909 + }, + { + "epoch": 3.677359367735937, + "grad_norm": 6.372810363769531, + "learning_rate": 9.9920939992094e-06, + "loss": 0.4642, + "step": 7910 + }, + { + "epoch": 3.6778242677824267, + "grad_norm": 6.050292491912842, + "learning_rate": 9.992092999209301e-06, + "loss": 0.5224, + "step": 7911 + }, + { + "epoch": 3.678289167828917, + "grad_norm": 5.080482006072998, + "learning_rate": 9.9920919992092e-06, + "loss": 0.2471, + "step": 7912 + }, + { + "epoch": 3.6787540678754067, + "grad_norm": 6.267346382141113, + "learning_rate": 9.9920909992091e-06, + "loss": 0.4297, + "step": 7913 + }, + { + "epoch": 3.679218967921897, + "grad_norm": 7.330696105957031, + "learning_rate": 9.992089999209e-06, + "loss": 0.4267, + "step": 7914 + }, + { + "epoch": 3.6796838679683868, + "grad_norm": 6.521805286407471, + "learning_rate": 9.9920889992089e-06, + "loss": 0.5629, + "step": 7915 + }, + { + "epoch": 3.6801487680148766, + "grad_norm": 6.312394142150879, + "learning_rate": 9.992087999208801e-06, + "loss": 0.361, + "step": 7916 + }, + { + "epoch": 3.680613668061367, + "grad_norm": 13.453668594360352, + "learning_rate": 9.992086999208702e-06, + "loss": 0.473, + "step": 7917 + }, + { + "epoch": 3.681078568107857, + "grad_norm": 6.073001384735107, + "learning_rate": 9.9920859992086e-06, + "loss": 0.3915, + "step": 7918 + }, + { + "epoch": 3.681543468154347, + "grad_norm": 6.120200157165527, + "learning_rate": 9.992084999208501e-06, + "loss": 0.5171, + "step": 7919 + }, + { + "epoch": 3.6820083682008367, + "grad_norm": 3.782270908355713, + "learning_rate": 9.9920839992084e-06, + "loss": 0.2479, + "step": 7920 + }, + { + "epoch": 3.682473268247327, + "grad_norm": 5.658958435058594, + "learning_rate": 9.992082999208301e-06, + "loss": 0.4285, + "step": 7921 + }, + { + "epoch": 3.6829381682938167, + "grad_norm": 5.702939510345459, + "learning_rate": 9.992081999208202e-06, + "loss": 0.4481, + "step": 7922 + }, + { + "epoch": 3.683403068340307, + "grad_norm": 5.193851470947266, + "learning_rate": 9.9920809992081e-06, + "loss": 0.2341, + "step": 7923 + }, + { + "epoch": 3.6838679683867968, + "grad_norm": 6.206783771514893, + "learning_rate": 9.992079999208001e-06, + "loss": 0.3395, + "step": 7924 + }, + { + "epoch": 3.684332868433287, + "grad_norm": 6.655522346496582, + "learning_rate": 9.9920789992079e-06, + "loss": 0.2816, + "step": 7925 + }, + { + "epoch": 3.684797768479777, + "grad_norm": 7.313681602478027, + "learning_rate": 9.9920779992078e-06, + "loss": 0.3796, + "step": 7926 + }, + { + "epoch": 3.685262668526267, + "grad_norm": 6.625879287719727, + "learning_rate": 9.9920769992077e-06, + "loss": 0.5663, + "step": 7927 + }, + { + "epoch": 3.685727568572757, + "grad_norm": 4.741537570953369, + "learning_rate": 9.9920759992076e-06, + "loss": 0.3705, + "step": 7928 + }, + { + "epoch": 3.6861924686192467, + "grad_norm": 6.560869216918945, + "learning_rate": 9.992074999207501e-06, + "loss": 0.3098, + "step": 7929 + }, + { + "epoch": 3.686657368665737, + "grad_norm": 5.480988025665283, + "learning_rate": 9.992073999207402e-06, + "loss": 0.2279, + "step": 7930 + }, + { + "epoch": 3.687122268712227, + "grad_norm": 6.254319667816162, + "learning_rate": 9.9920729992073e-06, + "loss": 0.4647, + "step": 7931 + }, + { + "epoch": 3.687587168758717, + "grad_norm": 5.539151191711426, + "learning_rate": 9.992071999207201e-06, + "loss": 0.3912, + "step": 7932 + }, + { + "epoch": 3.6880520688052068, + "grad_norm": 3.621699094772339, + "learning_rate": 9.9920709992071e-06, + "loss": 0.1646, + "step": 7933 + }, + { + "epoch": 3.688516968851697, + "grad_norm": 5.801355361938477, + "learning_rate": 9.992069999207001e-06, + "loss": 0.2984, + "step": 7934 + }, + { + "epoch": 3.688981868898187, + "grad_norm": 6.7842488288879395, + "learning_rate": 9.992068999206902e-06, + "loss": 0.2926, + "step": 7935 + }, + { + "epoch": 3.689446768944677, + "grad_norm": 4.052796840667725, + "learning_rate": 9.9920679992068e-06, + "loss": 0.1981, + "step": 7936 + }, + { + "epoch": 3.689911668991167, + "grad_norm": 6.438063621520996, + "learning_rate": 9.992066999206701e-06, + "loss": 0.3421, + "step": 7937 + }, + { + "epoch": 3.6903765690376567, + "grad_norm": 11.977290153503418, + "learning_rate": 9.9920659992066e-06, + "loss": 0.2475, + "step": 7938 + }, + { + "epoch": 3.690841469084147, + "grad_norm": 3.37925124168396, + "learning_rate": 9.992064999206501e-06, + "loss": 0.1817, + "step": 7939 + }, + { + "epoch": 3.691306369130637, + "grad_norm": 3.9392380714416504, + "learning_rate": 9.9920639992064e-06, + "loss": 0.2766, + "step": 7940 + }, + { + "epoch": 3.691771269177127, + "grad_norm": 3.5388379096984863, + "learning_rate": 9.9920629992063e-06, + "loss": 0.1384, + "step": 7941 + }, + { + "epoch": 3.6922361692236167, + "grad_norm": 5.932239055633545, + "learning_rate": 9.992061999206201e-06, + "loss": 0.4268, + "step": 7942 + }, + { + "epoch": 3.692701069270107, + "grad_norm": 5.658578872680664, + "learning_rate": 9.992060999206102e-06, + "loss": 0.309, + "step": 7943 + }, + { + "epoch": 3.693165969316597, + "grad_norm": 7.851151466369629, + "learning_rate": 9.992059999206e-06, + "loss": 0.6857, + "step": 7944 + }, + { + "epoch": 3.693630869363087, + "grad_norm": 5.691410064697266, + "learning_rate": 9.9920589992059e-06, + "loss": 0.4365, + "step": 7945 + }, + { + "epoch": 3.694095769409577, + "grad_norm": 6.776597499847412, + "learning_rate": 9.9920579992058e-06, + "loss": 0.5577, + "step": 7946 + }, + { + "epoch": 3.694560669456067, + "grad_norm": 7.204743385314941, + "learning_rate": 9.992056999205701e-06, + "loss": 0.3256, + "step": 7947 + }, + { + "epoch": 3.695025569502557, + "grad_norm": 6.008859634399414, + "learning_rate": 9.992055999205602e-06, + "loss": 0.4881, + "step": 7948 + }, + { + "epoch": 3.695490469549047, + "grad_norm": 4.528019428253174, + "learning_rate": 9.9920549992055e-06, + "loss": 0.4168, + "step": 7949 + }, + { + "epoch": 3.695955369595537, + "grad_norm": 4.476529598236084, + "learning_rate": 9.992053999205401e-06, + "loss": 0.2039, + "step": 7950 + }, + { + "epoch": 3.6964202696420267, + "grad_norm": 5.452652931213379, + "learning_rate": 9.9920529992053e-06, + "loss": 0.2531, + "step": 7951 + }, + { + "epoch": 3.696885169688517, + "grad_norm": 5.218283176422119, + "learning_rate": 9.992051999205201e-06, + "loss": 0.3007, + "step": 7952 + }, + { + "epoch": 3.6973500697350072, + "grad_norm": 4.919886589050293, + "learning_rate": 9.9920509992051e-06, + "loss": 0.428, + "step": 7953 + }, + { + "epoch": 3.697814969781497, + "grad_norm": 5.479310512542725, + "learning_rate": 9.992049999205e-06, + "loss": 0.3241, + "step": 7954 + }, + { + "epoch": 3.698279869827987, + "grad_norm": 3.869878053665161, + "learning_rate": 9.992048999204901e-06, + "loss": 0.2883, + "step": 7955 + }, + { + "epoch": 3.698744769874477, + "grad_norm": 7.128423690795898, + "learning_rate": 9.9920479992048e-06, + "loss": 0.4241, + "step": 7956 + }, + { + "epoch": 3.699209669920967, + "grad_norm": 4.714370250701904, + "learning_rate": 9.992046999204701e-06, + "loss": 0.2893, + "step": 7957 + }, + { + "epoch": 3.699674569967457, + "grad_norm": 4.809313774108887, + "learning_rate": 9.9920459992046e-06, + "loss": 0.303, + "step": 7958 + }, + { + "epoch": 3.700139470013947, + "grad_norm": 5.62995719909668, + "learning_rate": 9.9920449992045e-06, + "loss": 0.4003, + "step": 7959 + }, + { + "epoch": 3.7006043700604367, + "grad_norm": 4.173465251922607, + "learning_rate": 9.992043999204401e-06, + "loss": 0.2845, + "step": 7960 + }, + { + "epoch": 3.701069270106927, + "grad_norm": 4.334497928619385, + "learning_rate": 9.992042999204302e-06, + "loss": 0.2695, + "step": 7961 + }, + { + "epoch": 3.7015341701534172, + "grad_norm": 13.239160537719727, + "learning_rate": 9.9920419992042e-06, + "loss": 1.1198, + "step": 7962 + }, + { + "epoch": 3.701999070199907, + "grad_norm": 6.954886436462402, + "learning_rate": 9.992040999204101e-06, + "loss": 0.4882, + "step": 7963 + }, + { + "epoch": 3.702463970246397, + "grad_norm": 6.9476494789123535, + "learning_rate": 9.992039999204e-06, + "loss": 0.6257, + "step": 7964 + }, + { + "epoch": 3.702928870292887, + "grad_norm": 5.29102897644043, + "learning_rate": 9.9920389992039e-06, + "loss": 0.3518, + "step": 7965 + }, + { + "epoch": 3.703393770339377, + "grad_norm": 4.529397487640381, + "learning_rate": 9.9920379992038e-06, + "loss": 0.2549, + "step": 7966 + }, + { + "epoch": 3.703858670385867, + "grad_norm": 4.649232387542725, + "learning_rate": 9.9920369992037e-06, + "loss": 0.3456, + "step": 7967 + }, + { + "epoch": 3.704323570432357, + "grad_norm": 4.518601417541504, + "learning_rate": 9.992035999203601e-06, + "loss": 0.4483, + "step": 7968 + }, + { + "epoch": 3.704788470478847, + "grad_norm": 5.3148193359375, + "learning_rate": 9.9920349992035e-06, + "loss": 0.2458, + "step": 7969 + }, + { + "epoch": 3.705253370525337, + "grad_norm": 4.948920249938965, + "learning_rate": 9.992033999203401e-06, + "loss": 0.3079, + "step": 7970 + }, + { + "epoch": 3.7057182705718272, + "grad_norm": 4.2183356285095215, + "learning_rate": 9.9920329992033e-06, + "loss": 0.2449, + "step": 7971 + }, + { + "epoch": 3.706183170618317, + "grad_norm": 4.861098766326904, + "learning_rate": 9.9920319992032e-06, + "loss": 0.2593, + "step": 7972 + }, + { + "epoch": 3.706648070664807, + "grad_norm": 4.5616841316223145, + "learning_rate": 9.992030999203101e-06, + "loss": 0.4046, + "step": 7973 + }, + { + "epoch": 3.707112970711297, + "grad_norm": 3.8976552486419678, + "learning_rate": 9.992029999203002e-06, + "loss": 0.1918, + "step": 7974 + }, + { + "epoch": 3.7075778707577873, + "grad_norm": 6.797491073608398, + "learning_rate": 9.992028999202901e-06, + "loss": 0.6562, + "step": 7975 + }, + { + "epoch": 3.708042770804277, + "grad_norm": 9.046483039855957, + "learning_rate": 9.9920279992028e-06, + "loss": 0.6572, + "step": 7976 + }, + { + "epoch": 3.708507670850767, + "grad_norm": 4.787805557250977, + "learning_rate": 9.9920269992027e-06, + "loss": 0.1668, + "step": 7977 + }, + { + "epoch": 3.708972570897257, + "grad_norm": 3.2057366371154785, + "learning_rate": 9.9920259992026e-06, + "loss": 0.1343, + "step": 7978 + }, + { + "epoch": 3.709437470943747, + "grad_norm": 4.630867958068848, + "learning_rate": 9.9920249992025e-06, + "loss": 0.2357, + "step": 7979 + }, + { + "epoch": 3.7099023709902372, + "grad_norm": 3.840024948120117, + "learning_rate": 9.9920239992024e-06, + "loss": 0.2559, + "step": 7980 + }, + { + "epoch": 3.710367271036727, + "grad_norm": 3.3976316452026367, + "learning_rate": 9.992022999202301e-06, + "loss": 0.1466, + "step": 7981 + }, + { + "epoch": 3.710832171083217, + "grad_norm": 4.363095760345459, + "learning_rate": 9.9920219992022e-06, + "loss": 0.3571, + "step": 7982 + }, + { + "epoch": 3.711297071129707, + "grad_norm": 4.788299560546875, + "learning_rate": 9.992020999202101e-06, + "loss": 0.3283, + "step": 7983 + }, + { + "epoch": 3.7117619711761973, + "grad_norm": 7.369814872741699, + "learning_rate": 9.992019999202e-06, + "loss": 0.2847, + "step": 7984 + }, + { + "epoch": 3.712226871222687, + "grad_norm": 5.9970383644104, + "learning_rate": 9.9920189992019e-06, + "loss": 0.6132, + "step": 7985 + }, + { + "epoch": 3.712691771269177, + "grad_norm": 7.197362899780273, + "learning_rate": 9.992017999201801e-06, + "loss": 0.334, + "step": 7986 + }, + { + "epoch": 3.713156671315667, + "grad_norm": 3.5254106521606445, + "learning_rate": 9.9920169992017e-06, + "loss": 0.2458, + "step": 7987 + }, + { + "epoch": 3.713621571362157, + "grad_norm": 4.152889728546143, + "learning_rate": 9.992015999201601e-06, + "loss": 0.2655, + "step": 7988 + }, + { + "epoch": 3.7140864714086472, + "grad_norm": 12.91289234161377, + "learning_rate": 9.9920149992015e-06, + "loss": 0.2963, + "step": 7989 + }, + { + "epoch": 3.714551371455137, + "grad_norm": 7.219899654388428, + "learning_rate": 9.9920139992014e-06, + "loss": 0.4368, + "step": 7990 + }, + { + "epoch": 3.7150162715016273, + "grad_norm": 4.22091817855835, + "learning_rate": 9.992012999201301e-06, + "loss": 0.2198, + "step": 7991 + }, + { + "epoch": 3.715481171548117, + "grad_norm": 5.380424976348877, + "learning_rate": 9.9920119992012e-06, + "loss": 0.249, + "step": 7992 + }, + { + "epoch": 3.7159460715946073, + "grad_norm": 11.121798515319824, + "learning_rate": 9.992010999201101e-06, + "loss": 0.5172, + "step": 7993 + }, + { + "epoch": 3.716410971641097, + "grad_norm": 3.830350160598755, + "learning_rate": 9.992009999201002e-06, + "loss": 0.2476, + "step": 7994 + }, + { + "epoch": 3.716875871687587, + "grad_norm": 6.4717583656311035, + "learning_rate": 9.9920089992009e-06, + "loss": 0.5192, + "step": 7995 + }, + { + "epoch": 3.717340771734077, + "grad_norm": 5.197819232940674, + "learning_rate": 9.992007999200801e-06, + "loss": 0.3514, + "step": 7996 + }, + { + "epoch": 3.7178056717805674, + "grad_norm": 8.841588020324707, + "learning_rate": 9.9920069992007e-06, + "loss": 0.6527, + "step": 7997 + }, + { + "epoch": 3.7182705718270572, + "grad_norm": 5.709453105926514, + "learning_rate": 9.9920059992006e-06, + "loss": 0.4612, + "step": 7998 + }, + { + "epoch": 3.718735471873547, + "grad_norm": 6.177268028259277, + "learning_rate": 9.992004999200501e-06, + "loss": 0.4754, + "step": 7999 + }, + { + "epoch": 3.7192003719200373, + "grad_norm": 6.176748275756836, + "learning_rate": 9.9920039992004e-06, + "loss": 0.2664, + "step": 8000 + }, + { + "epoch": 3.719665271966527, + "grad_norm": 6.845928192138672, + "learning_rate": 9.992002999200301e-06, + "loss": 0.4216, + "step": 8001 + }, + { + "epoch": 3.7201301720130173, + "grad_norm": 4.113222599029541, + "learning_rate": 9.9920019992002e-06, + "loss": 0.3244, + "step": 8002 + }, + { + "epoch": 3.720595072059507, + "grad_norm": 4.272891521453857, + "learning_rate": 9.9920009992001e-06, + "loss": 0.1797, + "step": 8003 + }, + { + "epoch": 3.7210599721059974, + "grad_norm": 4.367763042449951, + "learning_rate": 9.991999999200001e-06, + "loss": 0.2307, + "step": 8004 + }, + { + "epoch": 3.721524872152487, + "grad_norm": 4.812140464782715, + "learning_rate": 9.991998999199902e-06, + "loss": 0.3383, + "step": 8005 + }, + { + "epoch": 3.7219897721989774, + "grad_norm": 8.618048667907715, + "learning_rate": 9.991997999199801e-06, + "loss": 0.3567, + "step": 8006 + }, + { + "epoch": 3.7224546722454672, + "grad_norm": 5.297152996063232, + "learning_rate": 9.991996999199702e-06, + "loss": 0.3983, + "step": 8007 + }, + { + "epoch": 3.722919572291957, + "grad_norm": 7.455430507659912, + "learning_rate": 9.9919959991996e-06, + "loss": 0.2902, + "step": 8008 + }, + { + "epoch": 3.7233844723384473, + "grad_norm": 4.227490425109863, + "learning_rate": 9.9919949991995e-06, + "loss": 0.2124, + "step": 8009 + }, + { + "epoch": 3.723849372384937, + "grad_norm": 7.9852118492126465, + "learning_rate": 9.9919939991994e-06, + "loss": 0.5909, + "step": 8010 + }, + { + "epoch": 3.7243142724314273, + "grad_norm": 5.516506671905518, + "learning_rate": 9.9919929991993e-06, + "loss": 0.4246, + "step": 8011 + }, + { + "epoch": 3.724779172477917, + "grad_norm": 5.213747024536133, + "learning_rate": 9.991991999199201e-06, + "loss": 0.3301, + "step": 8012 + }, + { + "epoch": 3.7252440725244074, + "grad_norm": 5.093119144439697, + "learning_rate": 9.9919909991991e-06, + "loss": 0.2908, + "step": 8013 + }, + { + "epoch": 3.725708972570897, + "grad_norm": 6.366689682006836, + "learning_rate": 9.991989999199001e-06, + "loss": 0.4661, + "step": 8014 + }, + { + "epoch": 3.7261738726173874, + "grad_norm": 6.466591835021973, + "learning_rate": 9.9919889991989e-06, + "loss": 0.4623, + "step": 8015 + }, + { + "epoch": 3.726638772663877, + "grad_norm": 3.978868246078491, + "learning_rate": 9.9919879991988e-06, + "loss": 0.2174, + "step": 8016 + }, + { + "epoch": 3.727103672710367, + "grad_norm": 4.842835426330566, + "learning_rate": 9.991986999198701e-06, + "loss": 0.3161, + "step": 8017 + }, + { + "epoch": 3.7275685727568573, + "grad_norm": 7.421835422515869, + "learning_rate": 9.991985999198602e-06, + "loss": 0.4251, + "step": 8018 + }, + { + "epoch": 3.7280334728033475, + "grad_norm": 4.108036041259766, + "learning_rate": 9.991984999198501e-06, + "loss": 0.2506, + "step": 8019 + }, + { + "epoch": 3.7284983728498373, + "grad_norm": 4.541064739227295, + "learning_rate": 9.9919839991984e-06, + "loss": 0.2972, + "step": 8020 + }, + { + "epoch": 3.728963272896327, + "grad_norm": 3.029160737991333, + "learning_rate": 9.9919829991983e-06, + "loss": 0.1189, + "step": 8021 + }, + { + "epoch": 3.7294281729428174, + "grad_norm": 5.2684173583984375, + "learning_rate": 9.9919819991982e-06, + "loss": 0.2921, + "step": 8022 + }, + { + "epoch": 3.729893072989307, + "grad_norm": 6.385560035705566, + "learning_rate": 9.9919809991981e-06, + "loss": 0.5285, + "step": 8023 + }, + { + "epoch": 3.7303579730357974, + "grad_norm": 4.699007034301758, + "learning_rate": 9.991979999198001e-06, + "loss": 0.3628, + "step": 8024 + }, + { + "epoch": 3.730822873082287, + "grad_norm": 4.49936056137085, + "learning_rate": 9.991978999197902e-06, + "loss": 0.2935, + "step": 8025 + }, + { + "epoch": 3.7312877731287775, + "grad_norm": 8.727194786071777, + "learning_rate": 9.9919779991978e-06, + "loss": 0.5249, + "step": 8026 + }, + { + "epoch": 3.7317526731752673, + "grad_norm": 5.743704795837402, + "learning_rate": 9.991976999197701e-06, + "loss": 0.307, + "step": 8027 + }, + { + "epoch": 3.7322175732217575, + "grad_norm": 4.499845027923584, + "learning_rate": 9.9919759991976e-06, + "loss": 0.3428, + "step": 8028 + }, + { + "epoch": 3.7326824732682473, + "grad_norm": 3.9486193656921387, + "learning_rate": 9.9919749991975e-06, + "loss": 0.2025, + "step": 8029 + }, + { + "epoch": 3.733147373314737, + "grad_norm": 5.079177379608154, + "learning_rate": 9.991973999197401e-06, + "loss": 0.3074, + "step": 8030 + }, + { + "epoch": 3.7336122733612274, + "grad_norm": 3.687547445297241, + "learning_rate": 9.9919729991973e-06, + "loss": 0.1208, + "step": 8031 + }, + { + "epoch": 3.7340771734077176, + "grad_norm": 6.419219970703125, + "learning_rate": 9.991971999197201e-06, + "loss": 0.5859, + "step": 8032 + }, + { + "epoch": 3.7345420734542074, + "grad_norm": 5.058046817779541, + "learning_rate": 9.9919709991971e-06, + "loss": 0.3633, + "step": 8033 + }, + { + "epoch": 3.735006973500697, + "grad_norm": 4.5347747802734375, + "learning_rate": 9.991969999197e-06, + "loss": 0.401, + "step": 8034 + }, + { + "epoch": 3.7354718735471875, + "grad_norm": 5.677745819091797, + "learning_rate": 9.9919689991969e-06, + "loss": 0.3438, + "step": 8035 + }, + { + "epoch": 3.7359367735936773, + "grad_norm": 6.212587833404541, + "learning_rate": 9.9919679991968e-06, + "loss": 0.4309, + "step": 8036 + }, + { + "epoch": 3.7364016736401675, + "grad_norm": 8.272089004516602, + "learning_rate": 9.991966999196701e-06, + "loss": 0.3152, + "step": 8037 + }, + { + "epoch": 3.7368665736866573, + "grad_norm": 7.761667728424072, + "learning_rate": 9.991965999196602e-06, + "loss": 0.3104, + "step": 8038 + }, + { + "epoch": 3.737331473733147, + "grad_norm": 6.2225260734558105, + "learning_rate": 9.9919649991965e-06, + "loss": 0.3434, + "step": 8039 + }, + { + "epoch": 3.7377963737796374, + "grad_norm": 6.427236557006836, + "learning_rate": 9.991963999196401e-06, + "loss": 0.3782, + "step": 8040 + }, + { + "epoch": 3.7382612738261276, + "grad_norm": 5.13834810256958, + "learning_rate": 9.9919629991963e-06, + "loss": 0.4215, + "step": 8041 + }, + { + "epoch": 3.7387261738726174, + "grad_norm": 4.406197547912598, + "learning_rate": 9.991961999196201e-06, + "loss": 0.2905, + "step": 8042 + }, + { + "epoch": 3.739191073919107, + "grad_norm": 4.520762920379639, + "learning_rate": 9.991960999196102e-06, + "loss": 0.3348, + "step": 8043 + }, + { + "epoch": 3.7396559739655975, + "grad_norm": 6.007953643798828, + "learning_rate": 9.991959999196e-06, + "loss": 0.3523, + "step": 8044 + }, + { + "epoch": 3.7401208740120873, + "grad_norm": 4.514919281005859, + "learning_rate": 9.991958999195901e-06, + "loss": 0.3518, + "step": 8045 + }, + { + "epoch": 3.7405857740585775, + "grad_norm": 4.91103982925415, + "learning_rate": 9.9919579991958e-06, + "loss": 0.3602, + "step": 8046 + }, + { + "epoch": 3.7410506741050673, + "grad_norm": 2.9053821563720703, + "learning_rate": 9.9919569991957e-06, + "loss": 0.1469, + "step": 8047 + }, + { + "epoch": 3.7415155741515576, + "grad_norm": 5.262271404266357, + "learning_rate": 9.9919559991956e-06, + "loss": 0.5384, + "step": 8048 + }, + { + "epoch": 3.7419804741980474, + "grad_norm": 8.624787330627441, + "learning_rate": 9.9919549991955e-06, + "loss": 0.7743, + "step": 8049 + }, + { + "epoch": 3.7424453742445376, + "grad_norm": 6.963454723358154, + "learning_rate": 9.991953999195401e-06, + "loss": 0.4591, + "step": 8050 + }, + { + "epoch": 3.7429102742910274, + "grad_norm": 5.335914134979248, + "learning_rate": 9.991952999195302e-06, + "loss": 0.4524, + "step": 8051 + }, + { + "epoch": 3.743375174337517, + "grad_norm": 4.95686674118042, + "learning_rate": 9.9919519991952e-06, + "loss": 0.1934, + "step": 8052 + }, + { + "epoch": 3.7438400743840075, + "grad_norm": 7.204280853271484, + "learning_rate": 9.9919509991951e-06, + "loss": 0.268, + "step": 8053 + }, + { + "epoch": 3.7443049744304977, + "grad_norm": 4.488373756408691, + "learning_rate": 9.991949999195e-06, + "loss": 0.2817, + "step": 8054 + }, + { + "epoch": 3.7447698744769875, + "grad_norm": 12.885416030883789, + "learning_rate": 9.991948999194901e-06, + "loss": 0.3049, + "step": 8055 + }, + { + "epoch": 3.7452347745234773, + "grad_norm": 4.420548439025879, + "learning_rate": 9.991947999194802e-06, + "loss": 0.2324, + "step": 8056 + }, + { + "epoch": 3.7456996745699676, + "grad_norm": 7.134016990661621, + "learning_rate": 9.9919469991947e-06, + "loss": 0.163, + "step": 8057 + }, + { + "epoch": 3.7461645746164574, + "grad_norm": 6.951475620269775, + "learning_rate": 9.991945999194601e-06, + "loss": 0.45, + "step": 8058 + }, + { + "epoch": 3.7466294746629476, + "grad_norm": 6.3646931648254395, + "learning_rate": 9.9919449991945e-06, + "loss": 0.3973, + "step": 8059 + }, + { + "epoch": 3.7470943747094374, + "grad_norm": 7.211939334869385, + "learning_rate": 9.991943999194401e-06, + "loss": 0.4861, + "step": 8060 + }, + { + "epoch": 3.747559274755927, + "grad_norm": 7.991515159606934, + "learning_rate": 9.9919429991943e-06, + "loss": 0.6078, + "step": 8061 + }, + { + "epoch": 3.7480241748024175, + "grad_norm": 13.904059410095215, + "learning_rate": 9.9919419991942e-06, + "loss": 0.5042, + "step": 8062 + }, + { + "epoch": 3.7484890748489077, + "grad_norm": 3.5945303440093994, + "learning_rate": 9.991940999194101e-06, + "loss": 0.1561, + "step": 8063 + }, + { + "epoch": 3.7489539748953975, + "grad_norm": 6.002649784088135, + "learning_rate": 9.991939999194e-06, + "loss": 0.316, + "step": 8064 + }, + { + "epoch": 3.7494188749418873, + "grad_norm": 6.725522994995117, + "learning_rate": 9.9919389991939e-06, + "loss": 0.3546, + "step": 8065 + }, + { + "epoch": 3.7498837749883775, + "grad_norm": 3.1589107513427734, + "learning_rate": 9.9919379991938e-06, + "loss": 0.1762, + "step": 8066 + }, + { + "epoch": 3.7503486750348674, + "grad_norm": 7.343104362487793, + "learning_rate": 9.9919369991937e-06, + "loss": 0.51, + "step": 8067 + }, + { + "epoch": 3.7508135750813576, + "grad_norm": 5.071706771850586, + "learning_rate": 9.991935999193601e-06, + "loss": 0.46, + "step": 8068 + }, + { + "epoch": 3.7512784751278474, + "grad_norm": 2.9009130001068115, + "learning_rate": 9.991934999193502e-06, + "loss": 0.0952, + "step": 8069 + }, + { + "epoch": 3.7517433751743376, + "grad_norm": 5.036277770996094, + "learning_rate": 9.9919339991934e-06, + "loss": 0.3645, + "step": 8070 + }, + { + "epoch": 3.7522082752208274, + "grad_norm": 6.640346050262451, + "learning_rate": 9.991932999193301e-06, + "loss": 0.5146, + "step": 8071 + }, + { + "epoch": 3.7526731752673177, + "grad_norm": 5.385578632354736, + "learning_rate": 9.9919319991932e-06, + "loss": 0.5569, + "step": 8072 + }, + { + "epoch": 3.7531380753138075, + "grad_norm": 6.610107898712158, + "learning_rate": 9.991930999193101e-06, + "loss": 0.2989, + "step": 8073 + }, + { + "epoch": 3.7536029753602973, + "grad_norm": 5.157229423522949, + "learning_rate": 9.991929999193002e-06, + "loss": 0.3072, + "step": 8074 + }, + { + "epoch": 3.7540678754067875, + "grad_norm": 6.107309818267822, + "learning_rate": 9.9919289991929e-06, + "loss": 0.4473, + "step": 8075 + }, + { + "epoch": 3.754532775453278, + "grad_norm": 3.841618776321411, + "learning_rate": 9.991927999192801e-06, + "loss": 0.1781, + "step": 8076 + }, + { + "epoch": 3.7549976754997676, + "grad_norm": 5.903085231781006, + "learning_rate": 9.9919269991927e-06, + "loss": 0.3705, + "step": 8077 + }, + { + "epoch": 3.7554625755462574, + "grad_norm": 5.430007457733154, + "learning_rate": 9.991925999192601e-06, + "loss": 0.364, + "step": 8078 + }, + { + "epoch": 3.7559274755927476, + "grad_norm": 5.885687828063965, + "learning_rate": 9.9919249991925e-06, + "loss": 0.2631, + "step": 8079 + }, + { + "epoch": 3.7563923756392374, + "grad_norm": 6.355769634246826, + "learning_rate": 9.9919239991924e-06, + "loss": 0.4093, + "step": 8080 + }, + { + "epoch": 3.7568572756857277, + "grad_norm": 6.8736653327941895, + "learning_rate": 9.991922999192301e-06, + "loss": 0.6786, + "step": 8081 + }, + { + "epoch": 3.7573221757322175, + "grad_norm": 7.0991621017456055, + "learning_rate": 9.991921999192202e-06, + "loss": 0.5063, + "step": 8082 + }, + { + "epoch": 3.7577870757787077, + "grad_norm": 4.684720039367676, + "learning_rate": 9.9919209991921e-06, + "loss": 0.2437, + "step": 8083 + }, + { + "epoch": 3.7582519758251975, + "grad_norm": 5.663974761962891, + "learning_rate": 9.991919999192001e-06, + "loss": 0.3251, + "step": 8084 + }, + { + "epoch": 3.758716875871688, + "grad_norm": 4.35833215713501, + "learning_rate": 9.9919189991919e-06, + "loss": 0.1092, + "step": 8085 + }, + { + "epoch": 3.7591817759181776, + "grad_norm": 4.615569591522217, + "learning_rate": 9.991917999191801e-06, + "loss": 0.2416, + "step": 8086 + }, + { + "epoch": 3.7596466759646674, + "grad_norm": 5.93233585357666, + "learning_rate": 9.991916999191702e-06, + "loss": 0.2934, + "step": 8087 + }, + { + "epoch": 3.7601115760111576, + "grad_norm": 5.262429237365723, + "learning_rate": 9.9919159991916e-06, + "loss": 0.2176, + "step": 8088 + }, + { + "epoch": 3.7605764760576474, + "grad_norm": 4.122246265411377, + "learning_rate": 9.991914999191501e-06, + "loss": 0.2846, + "step": 8089 + }, + { + "epoch": 3.7610413761041377, + "grad_norm": 3.243411064147949, + "learning_rate": 9.9919139991914e-06, + "loss": 0.1686, + "step": 8090 + }, + { + "epoch": 3.7615062761506275, + "grad_norm": 4.239469051361084, + "learning_rate": 9.991912999191301e-06, + "loss": 0.3607, + "step": 8091 + }, + { + "epoch": 3.7619711761971177, + "grad_norm": 5.2663893699646, + "learning_rate": 9.9919119991912e-06, + "loss": 0.4196, + "step": 8092 + }, + { + "epoch": 3.7624360762436075, + "grad_norm": 6.568402290344238, + "learning_rate": 9.9919109991911e-06, + "loss": 0.4942, + "step": 8093 + }, + { + "epoch": 3.762900976290098, + "grad_norm": 7.341641426086426, + "learning_rate": 9.991909999191001e-06, + "loss": 0.501, + "step": 8094 + }, + { + "epoch": 3.7633658763365876, + "grad_norm": 5.216006755828857, + "learning_rate": 9.9919089991909e-06, + "loss": 0.334, + "step": 8095 + }, + { + "epoch": 3.7638307763830774, + "grad_norm": 5.454625606536865, + "learning_rate": 9.9919079991908e-06, + "loss": 0.464, + "step": 8096 + }, + { + "epoch": 3.7642956764295676, + "grad_norm": 7.69526481628418, + "learning_rate": 9.9919069991907e-06, + "loss": 0.5147, + "step": 8097 + }, + { + "epoch": 3.764760576476058, + "grad_norm": 4.423125267028809, + "learning_rate": 9.9919059991906e-06, + "loss": 0.3226, + "step": 8098 + }, + { + "epoch": 3.7652254765225477, + "grad_norm": 6.282346248626709, + "learning_rate": 9.991904999190501e-06, + "loss": 0.3379, + "step": 8099 + }, + { + "epoch": 3.7656903765690375, + "grad_norm": 4.563490390777588, + "learning_rate": 9.991903999190402e-06, + "loss": 0.2831, + "step": 8100 + }, + { + "epoch": 3.7661552766155277, + "grad_norm": 5.885722637176514, + "learning_rate": 9.9919029991903e-06, + "loss": 0.4099, + "step": 8101 + }, + { + "epoch": 3.7666201766620175, + "grad_norm": 4.867854118347168, + "learning_rate": 9.991901999190201e-06, + "loss": 0.3218, + "step": 8102 + }, + { + "epoch": 3.767085076708508, + "grad_norm": 6.929715156555176, + "learning_rate": 9.9919009991901e-06, + "loss": 0.3539, + "step": 8103 + }, + { + "epoch": 3.7675499767549976, + "grad_norm": 6.689059257507324, + "learning_rate": 9.991899999190001e-06, + "loss": 0.7044, + "step": 8104 + }, + { + "epoch": 3.768014876801488, + "grad_norm": 4.039724826812744, + "learning_rate": 9.9918989991899e-06, + "loss": 0.2901, + "step": 8105 + }, + { + "epoch": 3.7684797768479776, + "grad_norm": 6.5488457679748535, + "learning_rate": 9.9918979991898e-06, + "loss": 0.2289, + "step": 8106 + }, + { + "epoch": 3.768944676894468, + "grad_norm": 7.6928277015686035, + "learning_rate": 9.991896999189701e-06, + "loss": 0.4489, + "step": 8107 + }, + { + "epoch": 3.7694095769409577, + "grad_norm": 4.631412506103516, + "learning_rate": 9.9918959991896e-06, + "loss": 0.351, + "step": 8108 + }, + { + "epoch": 3.7698744769874475, + "grad_norm": 6.016086578369141, + "learning_rate": 9.991894999189501e-06, + "loss": 0.3642, + "step": 8109 + }, + { + "epoch": 3.7703393770339377, + "grad_norm": 5.571109771728516, + "learning_rate": 9.9918939991894e-06, + "loss": 0.4966, + "step": 8110 + }, + { + "epoch": 3.770804277080428, + "grad_norm": 4.746151447296143, + "learning_rate": 9.9918929991893e-06, + "loss": 0.3332, + "step": 8111 + }, + { + "epoch": 3.771269177126918, + "grad_norm": 7.483921527862549, + "learning_rate": 9.991891999189201e-06, + "loss": 0.4132, + "step": 8112 + }, + { + "epoch": 3.7717340771734076, + "grad_norm": 3.431380271911621, + "learning_rate": 9.991890999189102e-06, + "loss": 0.1714, + "step": 8113 + }, + { + "epoch": 3.772198977219898, + "grad_norm": 8.017375946044922, + "learning_rate": 9.991889999189e-06, + "loss": 0.6975, + "step": 8114 + }, + { + "epoch": 3.7726638772663876, + "grad_norm": 8.923957824707031, + "learning_rate": 9.991888999188901e-06, + "loss": 0.5231, + "step": 8115 + }, + { + "epoch": 3.773128777312878, + "grad_norm": 5.916432857513428, + "learning_rate": 9.9918879991888e-06, + "loss": 0.3225, + "step": 8116 + }, + { + "epoch": 3.7735936773593677, + "grad_norm": 7.148556709289551, + "learning_rate": 9.9918869991887e-06, + "loss": 0.4211, + "step": 8117 + }, + { + "epoch": 3.7740585774058575, + "grad_norm": 8.07442569732666, + "learning_rate": 9.9918859991886e-06, + "loss": 0.7491, + "step": 8118 + }, + { + "epoch": 3.7745234774523477, + "grad_norm": 3.342984199523926, + "learning_rate": 9.9918849991885e-06, + "loss": 0.1887, + "step": 8119 + }, + { + "epoch": 3.774988377498838, + "grad_norm": 8.077452659606934, + "learning_rate": 9.991883999188401e-06, + "loss": 0.5208, + "step": 8120 + }, + { + "epoch": 3.775453277545328, + "grad_norm": 6.835738182067871, + "learning_rate": 9.9918829991883e-06, + "loss": 0.5383, + "step": 8121 + }, + { + "epoch": 3.7759181775918176, + "grad_norm": 5.31669807434082, + "learning_rate": 9.991881999188201e-06, + "loss": 0.4108, + "step": 8122 + }, + { + "epoch": 3.776383077638308, + "grad_norm": 4.2677812576293945, + "learning_rate": 9.9918809991881e-06, + "loss": 0.4082, + "step": 8123 + }, + { + "epoch": 3.7768479776847976, + "grad_norm": 3.7172889709472656, + "learning_rate": 9.991879999188e-06, + "loss": 0.2246, + "step": 8124 + }, + { + "epoch": 3.777312877731288, + "grad_norm": 7.142573833465576, + "learning_rate": 9.991878999187901e-06, + "loss": 0.7742, + "step": 8125 + }, + { + "epoch": 3.7777777777777777, + "grad_norm": 4.696048736572266, + "learning_rate": 9.991877999187802e-06, + "loss": 0.2705, + "step": 8126 + }, + { + "epoch": 3.778242677824268, + "grad_norm": 5.815554141998291, + "learning_rate": 9.991876999187701e-06, + "loss": 0.4479, + "step": 8127 + }, + { + "epoch": 3.7787075778707577, + "grad_norm": 7.8807053565979, + "learning_rate": 9.9918759991876e-06, + "loss": 0.2491, + "step": 8128 + }, + { + "epoch": 3.779172477917248, + "grad_norm": 4.82457160949707, + "learning_rate": 9.9918749991875e-06, + "loss": 0.3532, + "step": 8129 + }, + { + "epoch": 3.7796373779637378, + "grad_norm": 6.770020484924316, + "learning_rate": 9.9918739991874e-06, + "loss": 0.6539, + "step": 8130 + }, + { + "epoch": 3.7801022780102276, + "grad_norm": 7.955662727355957, + "learning_rate": 9.9918729991873e-06, + "loss": 0.5014, + "step": 8131 + }, + { + "epoch": 3.780567178056718, + "grad_norm": 3.493274688720703, + "learning_rate": 9.9918719991872e-06, + "loss": 0.1089, + "step": 8132 + }, + { + "epoch": 3.781032078103208, + "grad_norm": 3.3759355545043945, + "learning_rate": 9.991870999187101e-06, + "loss": 0.1237, + "step": 8133 + }, + { + "epoch": 3.781496978149698, + "grad_norm": 8.706205368041992, + "learning_rate": 9.991869999187e-06, + "loss": 0.6735, + "step": 8134 + }, + { + "epoch": 3.7819618781961877, + "grad_norm": 6.206429958343506, + "learning_rate": 9.991868999186901e-06, + "loss": 0.4945, + "step": 8135 + }, + { + "epoch": 3.782426778242678, + "grad_norm": 6.217094421386719, + "learning_rate": 9.9918679991868e-06, + "loss": 0.4858, + "step": 8136 + }, + { + "epoch": 3.7828916782891677, + "grad_norm": 5.932711124420166, + "learning_rate": 9.9918669991867e-06, + "loss": 0.2773, + "step": 8137 + }, + { + "epoch": 3.783356578335658, + "grad_norm": 4.43695592880249, + "learning_rate": 9.991865999186601e-06, + "loss": 0.3211, + "step": 8138 + }, + { + "epoch": 3.7838214783821478, + "grad_norm": 7.308570861816406, + "learning_rate": 9.9918649991865e-06, + "loss": 0.4353, + "step": 8139 + }, + { + "epoch": 3.7842863784286376, + "grad_norm": 3.821450710296631, + "learning_rate": 9.991863999186401e-06, + "loss": 0.2892, + "step": 8140 + }, + { + "epoch": 3.784751278475128, + "grad_norm": 5.715663909912109, + "learning_rate": 9.9918629991863e-06, + "loss": 0.427, + "step": 8141 + }, + { + "epoch": 3.785216178521618, + "grad_norm": 5.688313961029053, + "learning_rate": 9.9918619991862e-06, + "loss": 0.2658, + "step": 8142 + }, + { + "epoch": 3.785681078568108, + "grad_norm": 9.063982963562012, + "learning_rate": 9.9918609991861e-06, + "loss": 0.2291, + "step": 8143 + }, + { + "epoch": 3.7861459786145977, + "grad_norm": 5.532311916351318, + "learning_rate": 9.991859999186e-06, + "loss": 0.254, + "step": 8144 + }, + { + "epoch": 3.786610878661088, + "grad_norm": 5.771093845367432, + "learning_rate": 9.991858999185901e-06, + "loss": 0.4363, + "step": 8145 + }, + { + "epoch": 3.7870757787075777, + "grad_norm": 5.412066459655762, + "learning_rate": 9.991857999185802e-06, + "loss": 0.3644, + "step": 8146 + }, + { + "epoch": 3.787540678754068, + "grad_norm": 5.971200466156006, + "learning_rate": 9.9918569991857e-06, + "loss": 0.3025, + "step": 8147 + }, + { + "epoch": 3.7880055788005578, + "grad_norm": 10.05216121673584, + "learning_rate": 9.991855999185601e-06, + "loss": 0.617, + "step": 8148 + }, + { + "epoch": 3.788470478847048, + "grad_norm": 6.410074234008789, + "learning_rate": 9.9918549991855e-06, + "loss": 0.5336, + "step": 8149 + }, + { + "epoch": 3.788935378893538, + "grad_norm": 5.161968231201172, + "learning_rate": 9.9918539991854e-06, + "loss": 0.3141, + "step": 8150 + }, + { + "epoch": 3.789400278940028, + "grad_norm": 7.111895561218262, + "learning_rate": 9.991852999185301e-06, + "loss": 0.5287, + "step": 8151 + }, + { + "epoch": 3.789865178986518, + "grad_norm": 4.827610015869141, + "learning_rate": 9.9918519991852e-06, + "loss": 0.2471, + "step": 8152 + }, + { + "epoch": 3.7903300790330077, + "grad_norm": 3.989518165588379, + "learning_rate": 9.991850999185101e-06, + "loss": 0.2111, + "step": 8153 + }, + { + "epoch": 3.790794979079498, + "grad_norm": 4.018933296203613, + "learning_rate": 9.991849999185e-06, + "loss": 0.3112, + "step": 8154 + }, + { + "epoch": 3.791259879125988, + "grad_norm": 4.694886684417725, + "learning_rate": 9.9918489991849e-06, + "loss": 0.2919, + "step": 8155 + }, + { + "epoch": 3.791724779172478, + "grad_norm": 7.153122901916504, + "learning_rate": 9.991847999184801e-06, + "loss": 0.5738, + "step": 8156 + }, + { + "epoch": 3.7921896792189678, + "grad_norm": 5.53386116027832, + "learning_rate": 9.9918469991847e-06, + "loss": 0.5017, + "step": 8157 + }, + { + "epoch": 3.792654579265458, + "grad_norm": 5.629660606384277, + "learning_rate": 9.991845999184601e-06, + "loss": 0.5436, + "step": 8158 + }, + { + "epoch": 3.793119479311948, + "grad_norm": 6.127092361450195, + "learning_rate": 9.991844999184502e-06, + "loss": 0.3859, + "step": 8159 + }, + { + "epoch": 3.793584379358438, + "grad_norm": 5.964624881744385, + "learning_rate": 9.9918439991844e-06, + "loss": 0.1519, + "step": 8160 + }, + { + "epoch": 3.794049279404928, + "grad_norm": 7.4236555099487305, + "learning_rate": 9.9918429991843e-06, + "loss": 0.45, + "step": 8161 + }, + { + "epoch": 3.7945141794514177, + "grad_norm": 4.987776279449463, + "learning_rate": 9.9918419991842e-06, + "loss": 0.3335, + "step": 8162 + }, + { + "epoch": 3.794979079497908, + "grad_norm": 6.362944602966309, + "learning_rate": 9.991840999184101e-06, + "loss": 0.4796, + "step": 8163 + }, + { + "epoch": 3.795443979544398, + "grad_norm": 4.971039772033691, + "learning_rate": 9.991839999184002e-06, + "loss": 0.3894, + "step": 8164 + }, + { + "epoch": 3.795908879590888, + "grad_norm": 13.021949768066406, + "learning_rate": 9.9918389991839e-06, + "loss": 0.4485, + "step": 8165 + }, + { + "epoch": 3.7963737796373778, + "grad_norm": 5.558749675750732, + "learning_rate": 9.991837999183801e-06, + "loss": 0.2693, + "step": 8166 + }, + { + "epoch": 3.796838679683868, + "grad_norm": 5.32520866394043, + "learning_rate": 9.9918369991837e-06, + "loss": 0.1603, + "step": 8167 + }, + { + "epoch": 3.797303579730358, + "grad_norm": 11.403828620910645, + "learning_rate": 9.9918359991836e-06, + "loss": 0.5831, + "step": 8168 + }, + { + "epoch": 3.797768479776848, + "grad_norm": 9.45832347869873, + "learning_rate": 9.991834999183501e-06, + "loss": 1.0407, + "step": 8169 + }, + { + "epoch": 3.798233379823338, + "grad_norm": 5.192744731903076, + "learning_rate": 9.991833999183402e-06, + "loss": 0.3749, + "step": 8170 + }, + { + "epoch": 3.798698279869828, + "grad_norm": 3.3078391551971436, + "learning_rate": 9.991832999183301e-06, + "loss": 0.1562, + "step": 8171 + }, + { + "epoch": 3.799163179916318, + "grad_norm": 6.665186405181885, + "learning_rate": 9.9918319991832e-06, + "loss": 0.68, + "step": 8172 + }, + { + "epoch": 3.799628079962808, + "grad_norm": 6.452627182006836, + "learning_rate": 9.9918309991831e-06, + "loss": 0.4762, + "step": 8173 + }, + { + "epoch": 3.800092980009298, + "grad_norm": 5.174325466156006, + "learning_rate": 9.991829999183e-06, + "loss": 0.431, + "step": 8174 + }, + { + "epoch": 3.8005578800557878, + "grad_norm": 5.993022918701172, + "learning_rate": 9.9918289991829e-06, + "loss": 0.4923, + "step": 8175 + }, + { + "epoch": 3.801022780102278, + "grad_norm": 6.624042510986328, + "learning_rate": 9.991827999182801e-06, + "loss": 0.6317, + "step": 8176 + }, + { + "epoch": 3.8014876801487683, + "grad_norm": 4.565954685211182, + "learning_rate": 9.991826999182702e-06, + "loss": 0.4062, + "step": 8177 + }, + { + "epoch": 3.801952580195258, + "grad_norm": 9.301985740661621, + "learning_rate": 9.9918259991826e-06, + "loss": 0.4239, + "step": 8178 + }, + { + "epoch": 3.802417480241748, + "grad_norm": 6.208896636962891, + "learning_rate": 9.991824999182501e-06, + "loss": 0.6251, + "step": 8179 + }, + { + "epoch": 3.802882380288238, + "grad_norm": 3.9886913299560547, + "learning_rate": 9.9918239991824e-06, + "loss": 0.1895, + "step": 8180 + }, + { + "epoch": 3.803347280334728, + "grad_norm": 6.544905185699463, + "learning_rate": 9.991822999182301e-06, + "loss": 0.477, + "step": 8181 + }, + { + "epoch": 3.803812180381218, + "grad_norm": 6.017119407653809, + "learning_rate": 9.991821999182202e-06, + "loss": 0.4904, + "step": 8182 + }, + { + "epoch": 3.804277080427708, + "grad_norm": 7.7362060546875, + "learning_rate": 9.9918209991821e-06, + "loss": 0.508, + "step": 8183 + }, + { + "epoch": 3.804741980474198, + "grad_norm": 8.651890754699707, + "learning_rate": 9.991819999182001e-06, + "loss": 0.9958, + "step": 8184 + }, + { + "epoch": 3.805206880520688, + "grad_norm": 9.325671195983887, + "learning_rate": 9.9918189991819e-06, + "loss": 0.3709, + "step": 8185 + }, + { + "epoch": 3.8056717805671783, + "grad_norm": 4.694086074829102, + "learning_rate": 9.9918179991818e-06, + "loss": 0.24, + "step": 8186 + }, + { + "epoch": 3.806136680613668, + "grad_norm": 6.447721481323242, + "learning_rate": 9.9918169991817e-06, + "loss": 0.3183, + "step": 8187 + }, + { + "epoch": 3.806601580660158, + "grad_norm": 4.532567977905273, + "learning_rate": 9.9918159991816e-06, + "loss": 0.3213, + "step": 8188 + }, + { + "epoch": 3.807066480706648, + "grad_norm": 5.785600662231445, + "learning_rate": 9.991814999181501e-06, + "loss": 0.4717, + "step": 8189 + }, + { + "epoch": 3.8075313807531384, + "grad_norm": 6.645782470703125, + "learning_rate": 9.991813999181402e-06, + "loss": 0.5007, + "step": 8190 + }, + { + "epoch": 3.807996280799628, + "grad_norm": 6.125485897064209, + "learning_rate": 9.9918129991813e-06, + "loss": 0.5418, + "step": 8191 + }, + { + "epoch": 3.808461180846118, + "grad_norm": 5.771772861480713, + "learning_rate": 9.991811999181201e-06, + "loss": 0.5822, + "step": 8192 + }, + { + "epoch": 3.808926080892608, + "grad_norm": 7.862780570983887, + "learning_rate": 9.9918109991811e-06, + "loss": 0.5861, + "step": 8193 + }, + { + "epoch": 3.809390980939098, + "grad_norm": 4.228445529937744, + "learning_rate": 9.991809999181001e-06, + "loss": 0.3261, + "step": 8194 + }, + { + "epoch": 3.8098558809855883, + "grad_norm": 4.1231255531311035, + "learning_rate": 9.991808999180902e-06, + "loss": 0.4097, + "step": 8195 + }, + { + "epoch": 3.810320781032078, + "grad_norm": 5.9804582595825195, + "learning_rate": 9.9918079991808e-06, + "loss": 0.483, + "step": 8196 + }, + { + "epoch": 3.810785681078568, + "grad_norm": 6.795562267303467, + "learning_rate": 9.991806999180701e-06, + "loss": 0.6678, + "step": 8197 + }, + { + "epoch": 3.811250581125058, + "grad_norm": 4.693277835845947, + "learning_rate": 9.9918059991806e-06, + "loss": 0.4059, + "step": 8198 + }, + { + "epoch": 3.8117154811715483, + "grad_norm": Infinity, + "learning_rate": 9.9918059991806e-06, + "loss": 0.4188, + "step": 8199 + }, + { + "epoch": 3.812180381218038, + "grad_norm": 4.39414644241333, + "learning_rate": 9.9918049991805e-06, + "loss": 0.2156, + "step": 8200 + }, + { + "epoch": 3.812645281264528, + "grad_norm": 3.601694107055664, + "learning_rate": 9.9918039991804e-06, + "loss": 0.1983, + "step": 8201 + }, + { + "epoch": 3.813110181311018, + "grad_norm": 6.244479179382324, + "learning_rate": 9.9918029991803e-06, + "loss": 0.4231, + "step": 8202 + }, + { + "epoch": 3.813575081357508, + "grad_norm": 6.152253150939941, + "learning_rate": 9.991801999180201e-06, + "loss": 0.4543, + "step": 8203 + }, + { + "epoch": 3.8140399814039982, + "grad_norm": 3.7803685665130615, + "learning_rate": 9.991800999180102e-06, + "loss": 0.1777, + "step": 8204 + }, + { + "epoch": 3.814504881450488, + "grad_norm": 5.624416351318359, + "learning_rate": 9.99179999918e-06, + "loss": 0.3957, + "step": 8205 + }, + { + "epoch": 3.8149697814969783, + "grad_norm": 6.067422389984131, + "learning_rate": 9.9917989991799e-06, + "loss": 0.5395, + "step": 8206 + }, + { + "epoch": 3.815434681543468, + "grad_norm": 5.237438201904297, + "learning_rate": 9.9917979991798e-06, + "loss": 0.3733, + "step": 8207 + }, + { + "epoch": 3.8158995815899583, + "grad_norm": 7.096942901611328, + "learning_rate": 9.991796999179701e-06, + "loss": 0.4588, + "step": 8208 + }, + { + "epoch": 3.816364481636448, + "grad_norm": 3.8403849601745605, + "learning_rate": 9.991795999179602e-06, + "loss": 0.2841, + "step": 8209 + }, + { + "epoch": 3.816829381682938, + "grad_norm": 3.710087299346924, + "learning_rate": 9.9917949991795e-06, + "loss": 0.1875, + "step": 8210 + }, + { + "epoch": 3.817294281729428, + "grad_norm": 15.27264404296875, + "learning_rate": 9.991793999179401e-06, + "loss": 0.2148, + "step": 8211 + }, + { + "epoch": 3.8177591817759184, + "grad_norm": 5.612453460693359, + "learning_rate": 9.9917929991793e-06, + "loss": 0.4659, + "step": 8212 + }, + { + "epoch": 3.8182240818224082, + "grad_norm": 6.678035259246826, + "learning_rate": 9.991791999179201e-06, + "loss": 0.564, + "step": 8213 + }, + { + "epoch": 3.818688981868898, + "grad_norm": 5.174407958984375, + "learning_rate": 9.9917909991791e-06, + "loss": 0.2187, + "step": 8214 + }, + { + "epoch": 3.8191538819153883, + "grad_norm": 4.076975345611572, + "learning_rate": 9.991789999179e-06, + "loss": 0.3355, + "step": 8215 + }, + { + "epoch": 3.819618781961878, + "grad_norm": 7.479708194732666, + "learning_rate": 9.991788999178901e-06, + "loss": 0.4558, + "step": 8216 + }, + { + "epoch": 3.8200836820083683, + "grad_norm": 3.7375895977020264, + "learning_rate": 9.9917879991788e-06, + "loss": 0.2228, + "step": 8217 + }, + { + "epoch": 3.820548582054858, + "grad_norm": 5.601219654083252, + "learning_rate": 9.9917869991787e-06, + "loss": 0.2636, + "step": 8218 + }, + { + "epoch": 3.821013482101348, + "grad_norm": 4.592341423034668, + "learning_rate": 9.9917859991786e-06, + "loss": 0.4382, + "step": 8219 + }, + { + "epoch": 3.821478382147838, + "grad_norm": 3.928349494934082, + "learning_rate": 9.9917849991785e-06, + "loss": 0.2075, + "step": 8220 + }, + { + "epoch": 3.8219432821943284, + "grad_norm": 5.733017444610596, + "learning_rate": 9.991783999178401e-06, + "loss": 0.5183, + "step": 8221 + }, + { + "epoch": 3.8224081822408182, + "grad_norm": 3.5023868083953857, + "learning_rate": 9.991782999178302e-06, + "loss": 0.135, + "step": 8222 + }, + { + "epoch": 3.822873082287308, + "grad_norm": 5.316474437713623, + "learning_rate": 9.9917819991782e-06, + "loss": 0.3317, + "step": 8223 + }, + { + "epoch": 3.8233379823337983, + "grad_norm": 5.395368576049805, + "learning_rate": 9.991780999178101e-06, + "loss": 0.4433, + "step": 8224 + }, + { + "epoch": 3.823802882380288, + "grad_norm": 4.4077558517456055, + "learning_rate": 9.991779999178e-06, + "loss": 0.2246, + "step": 8225 + }, + { + "epoch": 3.8242677824267783, + "grad_norm": 5.5794525146484375, + "learning_rate": 9.991778999177901e-06, + "loss": 0.2762, + "step": 8226 + }, + { + "epoch": 3.824732682473268, + "grad_norm": 6.07852029800415, + "learning_rate": 9.9917779991778e-06, + "loss": 0.5578, + "step": 8227 + }, + { + "epoch": 3.8251975825197584, + "grad_norm": 4.226799964904785, + "learning_rate": 9.9917769991777e-06, + "loss": 0.1613, + "step": 8228 + }, + { + "epoch": 3.825662482566248, + "grad_norm": 5.297264575958252, + "learning_rate": 9.991775999177601e-06, + "loss": 0.4428, + "step": 8229 + }, + { + "epoch": 3.8261273826127384, + "grad_norm": 5.152735233306885, + "learning_rate": 9.9917749991775e-06, + "loss": 0.3602, + "step": 8230 + }, + { + "epoch": 3.8265922826592282, + "grad_norm": 5.48353385925293, + "learning_rate": 9.991773999177401e-06, + "loss": 0.3473, + "step": 8231 + }, + { + "epoch": 3.827057182705718, + "grad_norm": 4.763925075531006, + "learning_rate": 9.9917729991773e-06, + "loss": 0.2724, + "step": 8232 + }, + { + "epoch": 3.8275220827522083, + "grad_norm": 4.947959899902344, + "learning_rate": 9.9917719991772e-06, + "loss": 0.3834, + "step": 8233 + }, + { + "epoch": 3.8279869827986985, + "grad_norm": 5.471530914306641, + "learning_rate": 9.991770999177101e-06, + "loss": 0.4296, + "step": 8234 + }, + { + "epoch": 3.8284518828451883, + "grad_norm": 6.315694332122803, + "learning_rate": 9.991769999177002e-06, + "loss": 0.4369, + "step": 8235 + }, + { + "epoch": 3.828916782891678, + "grad_norm": 6.408628940582275, + "learning_rate": 9.9917689991769e-06, + "loss": 0.4416, + "step": 8236 + }, + { + "epoch": 3.8293816829381684, + "grad_norm": 6.2156662940979, + "learning_rate": 9.9917679991768e-06, + "loss": 0.3811, + "step": 8237 + }, + { + "epoch": 3.829846582984658, + "grad_norm": 5.006252765655518, + "learning_rate": 9.9917669991767e-06, + "loss": 0.4499, + "step": 8238 + }, + { + "epoch": 3.8303114830311484, + "grad_norm": 5.907594203948975, + "learning_rate": 9.991765999176601e-06, + "loss": 0.649, + "step": 8239 + }, + { + "epoch": 3.8307763830776382, + "grad_norm": 4.2194294929504395, + "learning_rate": 9.991764999176502e-06, + "loss": 0.2588, + "step": 8240 + }, + { + "epoch": 3.831241283124128, + "grad_norm": 6.464730262756348, + "learning_rate": 9.9917639991764e-06, + "loss": 0.4516, + "step": 8241 + }, + { + "epoch": 3.8317061831706183, + "grad_norm": 4.625704288482666, + "learning_rate": 9.991762999176301e-06, + "loss": 0.4055, + "step": 8242 + }, + { + "epoch": 3.8321710832171085, + "grad_norm": 6.47548770904541, + "learning_rate": 9.9917619991762e-06, + "loss": 0.2351, + "step": 8243 + }, + { + "epoch": 3.8326359832635983, + "grad_norm": 6.962953567504883, + "learning_rate": 9.991760999176101e-06, + "loss": 0.6839, + "step": 8244 + }, + { + "epoch": 3.833100883310088, + "grad_norm": 3.3670566082000732, + "learning_rate": 9.991759999176e-06, + "loss": 0.1812, + "step": 8245 + }, + { + "epoch": 3.8335657833565784, + "grad_norm": 3.6820197105407715, + "learning_rate": 9.9917589991759e-06, + "loss": 0.2166, + "step": 8246 + }, + { + "epoch": 3.834030683403068, + "grad_norm": 4.790648937225342, + "learning_rate": 9.991757999175801e-06, + "loss": 0.2664, + "step": 8247 + }, + { + "epoch": 3.8344955834495584, + "grad_norm": 10.90504264831543, + "learning_rate": 9.9917569991757e-06, + "loss": 0.3784, + "step": 8248 + }, + { + "epoch": 3.8349604834960482, + "grad_norm": 4.676076412200928, + "learning_rate": 9.991755999175601e-06, + "loss": 0.2977, + "step": 8249 + }, + { + "epoch": 3.8354253835425385, + "grad_norm": 6.763779640197754, + "learning_rate": 9.9917549991755e-06, + "loss": 0.4594, + "step": 8250 + }, + { + "epoch": 3.8358902835890283, + "grad_norm": 4.117458820343018, + "learning_rate": 9.9917539991754e-06, + "loss": 0.3292, + "step": 8251 + }, + { + "epoch": 3.8363551836355185, + "grad_norm": 6.074709415435791, + "learning_rate": 9.991752999175301e-06, + "loss": 0.4799, + "step": 8252 + }, + { + "epoch": 3.8368200836820083, + "grad_norm": 4.186805725097656, + "learning_rate": 9.991751999175202e-06, + "loss": 0.2833, + "step": 8253 + }, + { + "epoch": 3.837284983728498, + "grad_norm": 7.7090253829956055, + "learning_rate": 9.9917509991751e-06, + "loss": 0.504, + "step": 8254 + }, + { + "epoch": 3.8377498837749884, + "grad_norm": 4.700527667999268, + "learning_rate": 9.991749999175001e-06, + "loss": 0.2247, + "step": 8255 + }, + { + "epoch": 3.8382147838214786, + "grad_norm": 10.584052085876465, + "learning_rate": 9.9917489991749e-06, + "loss": 0.6452, + "step": 8256 + }, + { + "epoch": 3.8386796838679684, + "grad_norm": 6.164209365844727, + "learning_rate": 9.991747999174801e-06, + "loss": 0.6151, + "step": 8257 + }, + { + "epoch": 3.8391445839144582, + "grad_norm": 5.848930358886719, + "learning_rate": 9.9917469991747e-06, + "loss": 0.3968, + "step": 8258 + }, + { + "epoch": 3.8396094839609485, + "grad_norm": 4.540421962738037, + "learning_rate": 9.9917459991746e-06, + "loss": 0.2945, + "step": 8259 + }, + { + "epoch": 3.8400743840074383, + "grad_norm": 4.639928817749023, + "learning_rate": 9.991744999174501e-06, + "loss": 0.2535, + "step": 8260 + }, + { + "epoch": 3.8405392840539285, + "grad_norm": 8.08181381225586, + "learning_rate": 9.9917439991744e-06, + "loss": 0.4031, + "step": 8261 + }, + { + "epoch": 3.8410041841004183, + "grad_norm": 7.393013000488281, + "learning_rate": 9.991742999174301e-06, + "loss": 0.238, + "step": 8262 + }, + { + "epoch": 3.8414690841469086, + "grad_norm": 6.270169258117676, + "learning_rate": 9.9917419991742e-06, + "loss": 0.5542, + "step": 8263 + }, + { + "epoch": 3.8419339841933984, + "grad_norm": 9.440079689025879, + "learning_rate": 9.9917409991741e-06, + "loss": 0.8507, + "step": 8264 + }, + { + "epoch": 3.8423988842398886, + "grad_norm": 5.649724006652832, + "learning_rate": 9.991739999174001e-06, + "loss": 0.2456, + "step": 8265 + }, + { + "epoch": 3.8428637842863784, + "grad_norm": 7.024329662322998, + "learning_rate": 9.991738999173902e-06, + "loss": 0.5136, + "step": 8266 + }, + { + "epoch": 3.8433286843328682, + "grad_norm": 5.502714157104492, + "learning_rate": 9.991737999173801e-06, + "loss": 0.2362, + "step": 8267 + }, + { + "epoch": 3.8437935843793585, + "grad_norm": 7.447957515716553, + "learning_rate": 9.991736999173702e-06, + "loss": 0.4357, + "step": 8268 + }, + { + "epoch": 3.8442584844258483, + "grad_norm": 7.1393609046936035, + "learning_rate": 9.9917359991736e-06, + "loss": 0.3374, + "step": 8269 + }, + { + "epoch": 3.8447233844723385, + "grad_norm": 4.133737087249756, + "learning_rate": 9.9917349991735e-06, + "loss": 0.2778, + "step": 8270 + }, + { + "epoch": 3.8451882845188283, + "grad_norm": 5.290931701660156, + "learning_rate": 9.9917339991734e-06, + "loss": 0.2129, + "step": 8271 + }, + { + "epoch": 3.8456531845653186, + "grad_norm": 5.605026721954346, + "learning_rate": 9.9917329991733e-06, + "loss": 0.3315, + "step": 8272 + }, + { + "epoch": 3.8461180846118084, + "grad_norm": 8.069546699523926, + "learning_rate": 9.991731999173201e-06, + "loss": 0.4912, + "step": 8273 + }, + { + "epoch": 3.8465829846582986, + "grad_norm": 7.085963726043701, + "learning_rate": 9.9917309991731e-06, + "loss": 0.4006, + "step": 8274 + }, + { + "epoch": 3.8470478847047884, + "grad_norm": 5.915323734283447, + "learning_rate": 9.991729999173001e-06, + "loss": 0.23, + "step": 8275 + }, + { + "epoch": 3.8475127847512782, + "grad_norm": 6.0223846435546875, + "learning_rate": 9.9917289991729e-06, + "loss": 0.5034, + "step": 8276 + }, + { + "epoch": 3.8479776847977685, + "grad_norm": 5.236659526824951, + "learning_rate": 9.9917279991728e-06, + "loss": 0.2543, + "step": 8277 + }, + { + "epoch": 3.8484425848442587, + "grad_norm": 4.3556623458862305, + "learning_rate": 9.991726999172701e-06, + "loss": 0.3542, + "step": 8278 + }, + { + "epoch": 3.8489074848907485, + "grad_norm": 9.363306999206543, + "learning_rate": 9.991725999172602e-06, + "loss": 0.4579, + "step": 8279 + }, + { + "epoch": 3.8493723849372383, + "grad_norm": 2.8011744022369385, + "learning_rate": 9.991724999172501e-06, + "loss": 0.1651, + "step": 8280 + }, + { + "epoch": 3.8498372849837286, + "grad_norm": 6.710410118103027, + "learning_rate": 9.9917239991724e-06, + "loss": 0.2686, + "step": 8281 + }, + { + "epoch": 3.8503021850302184, + "grad_norm": 6.206579685211182, + "learning_rate": 9.9917229991723e-06, + "loss": 0.2756, + "step": 8282 + }, + { + "epoch": 3.8507670850767086, + "grad_norm": 5.1598591804504395, + "learning_rate": 9.9917219991722e-06, + "loss": 0.3266, + "step": 8283 + }, + { + "epoch": 3.8512319851231984, + "grad_norm": 5.648209571838379, + "learning_rate": 9.9917209991721e-06, + "loss": 0.4469, + "step": 8284 + }, + { + "epoch": 3.8516968851696887, + "grad_norm": 5.464639186859131, + "learning_rate": 9.991719999172e-06, + "loss": 0.4692, + "step": 8285 + }, + { + "epoch": 3.8521617852161785, + "grad_norm": 5.778827667236328, + "learning_rate": 9.991718999171901e-06, + "loss": 0.3, + "step": 8286 + }, + { + "epoch": 3.8526266852626687, + "grad_norm": 8.213700294494629, + "learning_rate": 9.9917179991718e-06, + "loss": 0.468, + "step": 8287 + }, + { + "epoch": 3.8530915853091585, + "grad_norm": 5.999714374542236, + "learning_rate": 9.991716999171701e-06, + "loss": 0.341, + "step": 8288 + }, + { + "epoch": 3.8535564853556483, + "grad_norm": 6.062065124511719, + "learning_rate": 9.9917159991716e-06, + "loss": 0.4173, + "step": 8289 + }, + { + "epoch": 3.8540213854021386, + "grad_norm": 7.9165568351745605, + "learning_rate": 9.9917149991715e-06, + "loss": 0.2297, + "step": 8290 + }, + { + "epoch": 3.854486285448629, + "grad_norm": 5.892551422119141, + "learning_rate": 9.991713999171401e-06, + "loss": 0.3891, + "step": 8291 + }, + { + "epoch": 3.8549511854951186, + "grad_norm": 3.910585880279541, + "learning_rate": 9.9917129991713e-06, + "loss": 0.1966, + "step": 8292 + }, + { + "epoch": 3.8554160855416084, + "grad_norm": 5.469883918762207, + "learning_rate": 9.991711999171201e-06, + "loss": 0.43, + "step": 8293 + }, + { + "epoch": 3.8558809855880987, + "grad_norm": 6.449491024017334, + "learning_rate": 9.9917109991711e-06, + "loss": 0.5105, + "step": 8294 + }, + { + "epoch": 3.8563458856345885, + "grad_norm": 7.2353057861328125, + "learning_rate": 9.991709999171e-06, + "loss": 0.7395, + "step": 8295 + }, + { + "epoch": 3.8568107856810787, + "grad_norm": 5.013767719268799, + "learning_rate": 9.9917089991709e-06, + "loss": 0.3198, + "step": 8296 + }, + { + "epoch": 3.8572756857275685, + "grad_norm": 4.852630615234375, + "learning_rate": 9.9917079991708e-06, + "loss": 0.1771, + "step": 8297 + }, + { + "epoch": 3.8577405857740583, + "grad_norm": 5.234140872955322, + "learning_rate": 9.991706999170701e-06, + "loss": 0.4189, + "step": 8298 + }, + { + "epoch": 3.8582054858205486, + "grad_norm": 6.761181831359863, + "learning_rate": 9.991705999170602e-06, + "loss": 0.3044, + "step": 8299 + }, + { + "epoch": 3.858670385867039, + "grad_norm": 5.03357458114624, + "learning_rate": 9.9917049991705e-06, + "loss": 0.3763, + "step": 8300 + }, + { + "epoch": 3.8591352859135286, + "grad_norm": 6.6676459312438965, + "learning_rate": 9.991703999170401e-06, + "loss": 0.3913, + "step": 8301 + }, + { + "epoch": 3.8596001859600184, + "grad_norm": 5.299663543701172, + "learning_rate": 9.9917029991703e-06, + "loss": 0.1857, + "step": 8302 + }, + { + "epoch": 3.8600650860065087, + "grad_norm": 5.2261738777160645, + "learning_rate": 9.9917019991702e-06, + "loss": 0.2864, + "step": 8303 + }, + { + "epoch": 3.8605299860529985, + "grad_norm": 5.716352939605713, + "learning_rate": 9.991700999170101e-06, + "loss": 0.4118, + "step": 8304 + }, + { + "epoch": 3.8609948860994887, + "grad_norm": 5.325473308563232, + "learning_rate": 9.99169999917e-06, + "loss": 0.476, + "step": 8305 + }, + { + "epoch": 3.8614597861459785, + "grad_norm": 7.458709716796875, + "learning_rate": 9.991698999169901e-06, + "loss": 0.621, + "step": 8306 + }, + { + "epoch": 3.8619246861924688, + "grad_norm": 4.428379058837891, + "learning_rate": 9.9916979991698e-06, + "loss": 0.2847, + "step": 8307 + }, + { + "epoch": 3.8623895862389586, + "grad_norm": 4.549989223480225, + "learning_rate": 9.9916969991697e-06, + "loss": 0.275, + "step": 8308 + }, + { + "epoch": 3.862854486285449, + "grad_norm": 4.291851997375488, + "learning_rate": 9.991695999169601e-06, + "loss": 0.2728, + "step": 8309 + }, + { + "epoch": 3.8633193863319386, + "grad_norm": 4.309354305267334, + "learning_rate": 9.9916949991695e-06, + "loss": 0.1537, + "step": 8310 + }, + { + "epoch": 3.8637842863784284, + "grad_norm": 7.807085990905762, + "learning_rate": 9.991693999169401e-06, + "loss": 0.7694, + "step": 8311 + }, + { + "epoch": 3.8642491864249187, + "grad_norm": 7.068685054779053, + "learning_rate": 9.991692999169302e-06, + "loss": 0.3689, + "step": 8312 + }, + { + "epoch": 3.864714086471409, + "grad_norm": 5.7681450843811035, + "learning_rate": 9.9916919991692e-06, + "loss": 0.3131, + "step": 8313 + }, + { + "epoch": 3.8651789865178987, + "grad_norm": 4.2349090576171875, + "learning_rate": 9.9916909991691e-06, + "loss": 0.2832, + "step": 8314 + }, + { + "epoch": 3.8656438865643885, + "grad_norm": 2.8669545650482178, + "learning_rate": 9.991689999169e-06, + "loss": 0.1384, + "step": 8315 + }, + { + "epoch": 3.8661087866108788, + "grad_norm": 4.709055423736572, + "learning_rate": 9.991688999168901e-06, + "loss": 0.3209, + "step": 8316 + }, + { + "epoch": 3.8665736866573686, + "grad_norm": 5.297349452972412, + "learning_rate": 9.991687999168802e-06, + "loss": 0.4739, + "step": 8317 + }, + { + "epoch": 3.867038586703859, + "grad_norm": 6.1172099113464355, + "learning_rate": 9.9916869991687e-06, + "loss": 0.4459, + "step": 8318 + }, + { + "epoch": 3.8675034867503486, + "grad_norm": 8.413237571716309, + "learning_rate": 9.991685999168601e-06, + "loss": 0.5875, + "step": 8319 + }, + { + "epoch": 3.8679683867968384, + "grad_norm": 4.828882694244385, + "learning_rate": 9.9916849991685e-06, + "loss": 0.1145, + "step": 8320 + }, + { + "epoch": 3.8684332868433287, + "grad_norm": 5.105487823486328, + "learning_rate": 9.9916839991684e-06, + "loss": 0.3546, + "step": 8321 + }, + { + "epoch": 3.868898186889819, + "grad_norm": 10.293889045715332, + "learning_rate": 9.991682999168301e-06, + "loss": 0.5209, + "step": 8322 + }, + { + "epoch": 3.8693630869363087, + "grad_norm": 3.507774591445923, + "learning_rate": 9.9916819991682e-06, + "loss": 0.2181, + "step": 8323 + }, + { + "epoch": 3.8698279869827985, + "grad_norm": 4.772382736206055, + "learning_rate": 9.991680999168101e-06, + "loss": 0.467, + "step": 8324 + }, + { + "epoch": 3.8702928870292888, + "grad_norm": 6.016023635864258, + "learning_rate": 9.991679999168e-06, + "loss": 0.4323, + "step": 8325 + }, + { + "epoch": 3.8707577870757786, + "grad_norm": 6.963773250579834, + "learning_rate": 9.9916789991679e-06, + "loss": 0.4624, + "step": 8326 + }, + { + "epoch": 3.871222687122269, + "grad_norm": 5.605326175689697, + "learning_rate": 9.9916779991678e-06, + "loss": 0.3078, + "step": 8327 + }, + { + "epoch": 3.8716875871687586, + "grad_norm": 5.435598850250244, + "learning_rate": 9.9916769991677e-06, + "loss": 0.3223, + "step": 8328 + }, + { + "epoch": 3.872152487215249, + "grad_norm": 5.751889705657959, + "learning_rate": 9.991675999167601e-06, + "loss": 0.2125, + "step": 8329 + }, + { + "epoch": 3.8726173872617387, + "grad_norm": 4.698618412017822, + "learning_rate": 9.991674999167502e-06, + "loss": 0.2293, + "step": 8330 + }, + { + "epoch": 3.873082287308229, + "grad_norm": 5.052998065948486, + "learning_rate": 9.9916739991674e-06, + "loss": 0.2241, + "step": 8331 + }, + { + "epoch": 3.8735471873547187, + "grad_norm": 5.27571439743042, + "learning_rate": 9.991672999167301e-06, + "loss": 0.4098, + "step": 8332 + }, + { + "epoch": 3.8740120874012085, + "grad_norm": 2.6227359771728516, + "learning_rate": 9.9916719991672e-06, + "loss": 0.1509, + "step": 8333 + }, + { + "epoch": 3.8744769874476988, + "grad_norm": 5.928910255432129, + "learning_rate": 9.991670999167101e-06, + "loss": 0.5388, + "step": 8334 + }, + { + "epoch": 3.874941887494189, + "grad_norm": 5.398646831512451, + "learning_rate": 9.991669999167002e-06, + "loss": 0.4, + "step": 8335 + }, + { + "epoch": 3.875406787540679, + "grad_norm": 6.563229084014893, + "learning_rate": 9.9916689991669e-06, + "loss": 0.4692, + "step": 8336 + }, + { + "epoch": 3.8758716875871686, + "grad_norm": 4.566469192504883, + "learning_rate": 9.991667999166801e-06, + "loss": 0.3682, + "step": 8337 + }, + { + "epoch": 3.876336587633659, + "grad_norm": 4.308078289031982, + "learning_rate": 9.9916669991667e-06, + "loss": 0.3619, + "step": 8338 + }, + { + "epoch": 3.8768014876801487, + "grad_norm": 5.596317291259766, + "learning_rate": 9.9916659991666e-06, + "loss": 0.2805, + "step": 8339 + }, + { + "epoch": 3.877266387726639, + "grad_norm": 5.324923038482666, + "learning_rate": 9.9916649991665e-06, + "loss": 0.3983, + "step": 8340 + }, + { + "epoch": 3.8777312877731287, + "grad_norm": 8.854681015014648, + "learning_rate": 9.9916639991664e-06, + "loss": 0.5899, + "step": 8341 + }, + { + "epoch": 3.878196187819619, + "grad_norm": 7.422252655029297, + "learning_rate": 9.991662999166301e-06, + "loss": 0.4803, + "step": 8342 + }, + { + "epoch": 3.8786610878661087, + "grad_norm": 8.48043155670166, + "learning_rate": 9.991661999166202e-06, + "loss": 0.8115, + "step": 8343 + }, + { + "epoch": 3.879125987912599, + "grad_norm": 7.092176914215088, + "learning_rate": 9.9916609991661e-06, + "loss": 0.3498, + "step": 8344 + }, + { + "epoch": 3.879590887959089, + "grad_norm": 4.7996063232421875, + "learning_rate": 9.991659999166001e-06, + "loss": 0.3867, + "step": 8345 + }, + { + "epoch": 3.8800557880055786, + "grad_norm": 4.7708001136779785, + "learning_rate": 9.9916589991659e-06, + "loss": 0.18, + "step": 8346 + }, + { + "epoch": 3.880520688052069, + "grad_norm": 4.274286270141602, + "learning_rate": 9.991657999165801e-06, + "loss": 0.2243, + "step": 8347 + }, + { + "epoch": 3.8809855880985586, + "grad_norm": 7.842041015625, + "learning_rate": 9.991656999165702e-06, + "loss": 0.5847, + "step": 8348 + }, + { + "epoch": 3.881450488145049, + "grad_norm": 7.21961784362793, + "learning_rate": 9.9916559991656e-06, + "loss": 0.3532, + "step": 8349 + }, + { + "epoch": 3.8819153881915387, + "grad_norm": 4.49582052230835, + "learning_rate": 9.991654999165501e-06, + "loss": 0.3069, + "step": 8350 + }, + { + "epoch": 3.882380288238029, + "grad_norm": 6.700623989105225, + "learning_rate": 9.9916539991654e-06, + "loss": 0.35, + "step": 8351 + }, + { + "epoch": 3.8828451882845187, + "grad_norm": 3.6530277729034424, + "learning_rate": 9.991652999165301e-06, + "loss": 0.1892, + "step": 8352 + }, + { + "epoch": 3.883310088331009, + "grad_norm": 7.816925048828125, + "learning_rate": 9.9916519991652e-06, + "loss": 0.6435, + "step": 8353 + }, + { + "epoch": 3.883774988377499, + "grad_norm": 6.3144941329956055, + "learning_rate": 9.9916509991651e-06, + "loss": 0.2957, + "step": 8354 + }, + { + "epoch": 3.8842398884239886, + "grad_norm": 5.624675750732422, + "learning_rate": 9.991649999165001e-06, + "loss": 0.322, + "step": 8355 + }, + { + "epoch": 3.884704788470479, + "grad_norm": 7.897968769073486, + "learning_rate": 9.991648999164902e-06, + "loss": 0.7694, + "step": 8356 + }, + { + "epoch": 3.885169688516969, + "grad_norm": 5.637426376342773, + "learning_rate": 9.9916479991648e-06, + "loss": 0.2669, + "step": 8357 + }, + { + "epoch": 3.885634588563459, + "grad_norm": 5.424076080322266, + "learning_rate": 9.9916469991647e-06, + "loss": 0.5354, + "step": 8358 + }, + { + "epoch": 3.8860994886099487, + "grad_norm": 4.303739070892334, + "learning_rate": 9.9916459991646e-06, + "loss": 0.2093, + "step": 8359 + }, + { + "epoch": 3.886564388656439, + "grad_norm": 6.037886142730713, + "learning_rate": 9.991644999164501e-06, + "loss": 0.3516, + "step": 8360 + }, + { + "epoch": 3.8870292887029287, + "grad_norm": 7.747163772583008, + "learning_rate": 9.991643999164402e-06, + "loss": 0.5092, + "step": 8361 + }, + { + "epoch": 3.887494188749419, + "grad_norm": 4.304412841796875, + "learning_rate": 9.9916429991643e-06, + "loss": 0.226, + "step": 8362 + }, + { + "epoch": 3.887959088795909, + "grad_norm": 3.5156917572021484, + "learning_rate": 9.991641999164201e-06, + "loss": 0.1421, + "step": 8363 + }, + { + "epoch": 3.888423988842399, + "grad_norm": 4.399404048919678, + "learning_rate": 9.9916409991641e-06, + "loss": 0.2916, + "step": 8364 + }, + { + "epoch": 3.888888888888889, + "grad_norm": 9.019577980041504, + "learning_rate": 9.991639999164001e-06, + "loss": 0.9334, + "step": 8365 + }, + { + "epoch": 3.889353788935379, + "grad_norm": 5.55741024017334, + "learning_rate": 9.9916389991639e-06, + "loss": 0.3153, + "step": 8366 + }, + { + "epoch": 3.889818688981869, + "grad_norm": 4.875741958618164, + "learning_rate": 9.9916379991638e-06, + "loss": 0.3701, + "step": 8367 + }, + { + "epoch": 3.8902835890283587, + "grad_norm": 5.721108436584473, + "learning_rate": 9.991636999163701e-06, + "loss": 0.5189, + "step": 8368 + }, + { + "epoch": 3.890748489074849, + "grad_norm": 6.520559310913086, + "learning_rate": 9.9916359991636e-06, + "loss": 0.8428, + "step": 8369 + }, + { + "epoch": 3.891213389121339, + "grad_norm": 5.888576030731201, + "learning_rate": 9.991634999163501e-06, + "loss": 0.6334, + "step": 8370 + }, + { + "epoch": 3.891678289167829, + "grad_norm": 4.701350688934326, + "learning_rate": 9.9916339991634e-06, + "loss": 0.3475, + "step": 8371 + }, + { + "epoch": 3.892143189214319, + "grad_norm": 4.905146598815918, + "learning_rate": 9.9916329991633e-06, + "loss": 0.3498, + "step": 8372 + }, + { + "epoch": 3.892608089260809, + "grad_norm": 13.75506591796875, + "learning_rate": 9.991631999163201e-06, + "loss": 0.626, + "step": 8373 + }, + { + "epoch": 3.893072989307299, + "grad_norm": 6.471831321716309, + "learning_rate": 9.991630999163102e-06, + "loss": 0.4792, + "step": 8374 + }, + { + "epoch": 3.893537889353789, + "grad_norm": 5.770045280456543, + "learning_rate": 9.991629999163e-06, + "loss": 0.2341, + "step": 8375 + }, + { + "epoch": 3.894002789400279, + "grad_norm": 5.801808834075928, + "learning_rate": 9.991628999162901e-06, + "loss": 0.4804, + "step": 8376 + }, + { + "epoch": 3.8944676894467687, + "grad_norm": 5.181506633758545, + "learning_rate": 9.9916279991628e-06, + "loss": 0.5404, + "step": 8377 + }, + { + "epoch": 3.894932589493259, + "grad_norm": 7.849767208099365, + "learning_rate": 9.9916269991627e-06, + "loss": 0.595, + "step": 8378 + }, + { + "epoch": 3.895397489539749, + "grad_norm": 7.208399295806885, + "learning_rate": 9.9916259991626e-06, + "loss": 0.3596, + "step": 8379 + }, + { + "epoch": 3.895862389586239, + "grad_norm": 6.729241847991943, + "learning_rate": 9.9916249991625e-06, + "loss": 0.6276, + "step": 8380 + }, + { + "epoch": 3.896327289632729, + "grad_norm": 5.882885932922363, + "learning_rate": 9.991623999162401e-06, + "loss": 0.2958, + "step": 8381 + }, + { + "epoch": 3.896792189679219, + "grad_norm": 5.3682966232299805, + "learning_rate": 9.9916229991623e-06, + "loss": 0.2743, + "step": 8382 + }, + { + "epoch": 3.897257089725709, + "grad_norm": 7.023048400878906, + "learning_rate": 9.991621999162201e-06, + "loss": 0.4012, + "step": 8383 + }, + { + "epoch": 3.897721989772199, + "grad_norm": 4.363863468170166, + "learning_rate": 9.9916209991621e-06, + "loss": 0.3904, + "step": 8384 + }, + { + "epoch": 3.898186889818689, + "grad_norm": 7.766023635864258, + "learning_rate": 9.991619999162e-06, + "loss": 0.236, + "step": 8385 + }, + { + "epoch": 3.898651789865179, + "grad_norm": 6.47257137298584, + "learning_rate": 9.991618999161901e-06, + "loss": 0.4834, + "step": 8386 + }, + { + "epoch": 3.899116689911669, + "grad_norm": 4.821719169616699, + "learning_rate": 9.991617999161802e-06, + "loss": 0.2823, + "step": 8387 + }, + { + "epoch": 3.899581589958159, + "grad_norm": 5.721146583557129, + "learning_rate": 9.9916169991617e-06, + "loss": 0.3662, + "step": 8388 + }, + { + "epoch": 3.900046490004649, + "grad_norm": 5.532712936401367, + "learning_rate": 9.9916159991616e-06, + "loss": 0.4136, + "step": 8389 + }, + { + "epoch": 3.900511390051139, + "grad_norm": 5.69556188583374, + "learning_rate": 9.9916149991615e-06, + "loss": 0.3391, + "step": 8390 + }, + { + "epoch": 3.900976290097629, + "grad_norm": 5.388271331787109, + "learning_rate": 9.991613999161401e-06, + "loss": 0.4875, + "step": 8391 + }, + { + "epoch": 3.9014411901441193, + "grad_norm": 6.473782539367676, + "learning_rate": 9.9916129991613e-06, + "loss": 0.3977, + "step": 8392 + }, + { + "epoch": 3.901906090190609, + "grad_norm": 4.521950721740723, + "learning_rate": 9.9916119991612e-06, + "loss": 0.377, + "step": 8393 + }, + { + "epoch": 3.902370990237099, + "grad_norm": 5.477321147918701, + "learning_rate": 9.991610999161101e-06, + "loss": 0.2487, + "step": 8394 + }, + { + "epoch": 3.902835890283589, + "grad_norm": 6.069614887237549, + "learning_rate": 9.991609999161e-06, + "loss": 0.2833, + "step": 8395 + }, + { + "epoch": 3.903300790330079, + "grad_norm": 6.01432991027832, + "learning_rate": 9.991608999160901e-06, + "loss": 0.6557, + "step": 8396 + }, + { + "epoch": 3.903765690376569, + "grad_norm": 4.568993091583252, + "learning_rate": 9.9916079991608e-06, + "loss": 0.3058, + "step": 8397 + }, + { + "epoch": 3.904230590423059, + "grad_norm": 4.276514530181885, + "learning_rate": 9.9916069991607e-06, + "loss": 0.3107, + "step": 8398 + }, + { + "epoch": 3.904695490469549, + "grad_norm": 4.927193641662598, + "learning_rate": 9.991605999160601e-06, + "loss": 0.3177, + "step": 8399 + }, + { + "epoch": 3.905160390516039, + "grad_norm": 6.916048049926758, + "learning_rate": 9.9916049991605e-06, + "loss": 0.5227, + "step": 8400 + }, + { + "epoch": 3.9056252905625293, + "grad_norm": 6.781970500946045, + "learning_rate": 9.991603999160401e-06, + "loss": 0.3614, + "step": 8401 + }, + { + "epoch": 3.906090190609019, + "grad_norm": 4.233449459075928, + "learning_rate": 9.9916029991603e-06, + "loss": 0.1901, + "step": 8402 + }, + { + "epoch": 3.906555090655509, + "grad_norm": 7.395065784454346, + "learning_rate": 9.9916019991602e-06, + "loss": 0.4091, + "step": 8403 + }, + { + "epoch": 3.907019990701999, + "grad_norm": 5.6491923332214355, + "learning_rate": 9.991600999160101e-06, + "loss": 0.3613, + "step": 8404 + }, + { + "epoch": 3.907484890748489, + "grad_norm": 3.7696473598480225, + "learning_rate": 9.991599999160002e-06, + "loss": 0.2162, + "step": 8405 + }, + { + "epoch": 3.907949790794979, + "grad_norm": 5.922464370727539, + "learning_rate": 9.9915989991599e-06, + "loss": 0.5128, + "step": 8406 + }, + { + "epoch": 3.908414690841469, + "grad_norm": 7.227541446685791, + "learning_rate": 9.991597999159801e-06, + "loss": 0.4873, + "step": 8407 + }, + { + "epoch": 3.908879590887959, + "grad_norm": 5.606561183929443, + "learning_rate": 9.9915969991597e-06, + "loss": 0.5178, + "step": 8408 + }, + { + "epoch": 3.909344490934449, + "grad_norm": 5.9396562576293945, + "learning_rate": 9.991595999159601e-06, + "loss": 0.3964, + "step": 8409 + }, + { + "epoch": 3.9098093909809393, + "grad_norm": 5.959894180297852, + "learning_rate": 9.9915949991595e-06, + "loss": 0.4294, + "step": 8410 + }, + { + "epoch": 3.910274291027429, + "grad_norm": 3.8554341793060303, + "learning_rate": 9.9915939991594e-06, + "loss": 0.1922, + "step": 8411 + }, + { + "epoch": 3.910739191073919, + "grad_norm": 6.438398361206055, + "learning_rate": 9.991592999159301e-06, + "loss": 0.4714, + "step": 8412 + }, + { + "epoch": 3.911204091120409, + "grad_norm": 4.562563419342041, + "learning_rate": 9.9915919991592e-06, + "loss": 0.1964, + "step": 8413 + }, + { + "epoch": 3.9116689911668994, + "grad_norm": 6.0288519859313965, + "learning_rate": 9.991590999159101e-06, + "loss": 0.4469, + "step": 8414 + }, + { + "epoch": 3.912133891213389, + "grad_norm": 4.444561004638672, + "learning_rate": 9.991589999159e-06, + "loss": 0.2978, + "step": 8415 + }, + { + "epoch": 3.912598791259879, + "grad_norm": 6.620169162750244, + "learning_rate": 9.9915889991589e-06, + "loss": 0.4809, + "step": 8416 + }, + { + "epoch": 3.913063691306369, + "grad_norm": 34.10903549194336, + "learning_rate": 9.991587999158801e-06, + "loss": 0.3517, + "step": 8417 + }, + { + "epoch": 3.913528591352859, + "grad_norm": 3.6566193103790283, + "learning_rate": 9.991586999158702e-06, + "loss": 0.3381, + "step": 8418 + }, + { + "epoch": 3.9139934913993493, + "grad_norm": 7.539011478424072, + "learning_rate": 9.991585999158601e-06, + "loss": 0.4963, + "step": 8419 + }, + { + "epoch": 3.914458391445839, + "grad_norm": 5.035881042480469, + "learning_rate": 9.991584999158502e-06, + "loss": 0.2718, + "step": 8420 + }, + { + "epoch": 3.914923291492329, + "grad_norm": 4.337118148803711, + "learning_rate": 9.9915839991584e-06, + "loss": 0.3696, + "step": 8421 + }, + { + "epoch": 3.915388191538819, + "grad_norm": 6.511277198791504, + "learning_rate": 9.9915829991583e-06, + "loss": 0.5262, + "step": 8422 + }, + { + "epoch": 3.9158530915853094, + "grad_norm": 6.773167610168457, + "learning_rate": 9.9915819991582e-06, + "loss": 0.5835, + "step": 8423 + }, + { + "epoch": 3.916317991631799, + "grad_norm": 7.626578330993652, + "learning_rate": 9.9915809991581e-06, + "loss": 0.6261, + "step": 8424 + }, + { + "epoch": 3.916782891678289, + "grad_norm": 8.32001781463623, + "learning_rate": 9.991579999158001e-06, + "loss": 0.3111, + "step": 8425 + }, + { + "epoch": 3.917247791724779, + "grad_norm": 2.7876298427581787, + "learning_rate": 9.9915789991579e-06, + "loss": 0.0902, + "step": 8426 + }, + { + "epoch": 3.917712691771269, + "grad_norm": 6.623093128204346, + "learning_rate": 9.991577999157801e-06, + "loss": 0.291, + "step": 8427 + }, + { + "epoch": 3.9181775918177593, + "grad_norm": 6.207101821899414, + "learning_rate": 9.9915769991577e-06, + "loss": 0.5305, + "step": 8428 + }, + { + "epoch": 3.918642491864249, + "grad_norm": 2.8205559253692627, + "learning_rate": 9.9915759991576e-06, + "loss": 0.1353, + "step": 8429 + }, + { + "epoch": 3.9191073919107393, + "grad_norm": 9.05270004272461, + "learning_rate": 9.991574999157501e-06, + "loss": 0.4292, + "step": 8430 + }, + { + "epoch": 3.919572291957229, + "grad_norm": 7.122960567474365, + "learning_rate": 9.991573999157402e-06, + "loss": 0.2912, + "step": 8431 + }, + { + "epoch": 3.9200371920037194, + "grad_norm": 6.225837230682373, + "learning_rate": 9.991572999157301e-06, + "loss": 0.3732, + "step": 8432 + }, + { + "epoch": 3.920502092050209, + "grad_norm": 6.0032782554626465, + "learning_rate": 9.9915719991572e-06, + "loss": 0.5904, + "step": 8433 + }, + { + "epoch": 3.920966992096699, + "grad_norm": 3.7207021713256836, + "learning_rate": 9.9915709991571e-06, + "loss": 0.3183, + "step": 8434 + }, + { + "epoch": 3.921431892143189, + "grad_norm": 8.108747482299805, + "learning_rate": 9.991569999157e-06, + "loss": 0.806, + "step": 8435 + }, + { + "epoch": 3.9218967921896795, + "grad_norm": 4.770956993103027, + "learning_rate": 9.9915689991569e-06, + "loss": 0.3479, + "step": 8436 + }, + { + "epoch": 3.9223616922361693, + "grad_norm": 5.7126312255859375, + "learning_rate": 9.991567999156801e-06, + "loss": 0.272, + "step": 8437 + }, + { + "epoch": 3.922826592282659, + "grad_norm": 5.869287967681885, + "learning_rate": 9.991566999156702e-06, + "loss": 0.5639, + "step": 8438 + }, + { + "epoch": 3.9232914923291493, + "grad_norm": 5.020052433013916, + "learning_rate": 9.9915659991566e-06, + "loss": 0.3827, + "step": 8439 + }, + { + "epoch": 3.923756392375639, + "grad_norm": 7.482132911682129, + "learning_rate": 9.991564999156501e-06, + "loss": 0.7056, + "step": 8440 + }, + { + "epoch": 3.9242212924221294, + "grad_norm": 6.751254081726074, + "learning_rate": 9.9915639991564e-06, + "loss": 0.3525, + "step": 8441 + }, + { + "epoch": 3.924686192468619, + "grad_norm": 6.214183330535889, + "learning_rate": 9.9915629991563e-06, + "loss": 0.1792, + "step": 8442 + }, + { + "epoch": 3.9251510925151094, + "grad_norm": 4.198868751525879, + "learning_rate": 9.991561999156201e-06, + "loss": 0.2242, + "step": 8443 + }, + { + "epoch": 3.925615992561599, + "grad_norm": 4.374971389770508, + "learning_rate": 9.9915609991561e-06, + "loss": 0.2879, + "step": 8444 + }, + { + "epoch": 3.9260808926080895, + "grad_norm": 4.995963096618652, + "learning_rate": 9.991559999156001e-06, + "loss": 0.3072, + "step": 8445 + }, + { + "epoch": 3.9265457926545793, + "grad_norm": 5.24547004699707, + "learning_rate": 9.9915589991559e-06, + "loss": 0.3235, + "step": 8446 + }, + { + "epoch": 3.927010692701069, + "grad_norm": 4.5687408447265625, + "learning_rate": 9.9915579991558e-06, + "loss": 0.3617, + "step": 8447 + }, + { + "epoch": 3.9274755927475593, + "grad_norm": 6.397469997406006, + "learning_rate": 9.9915569991557e-06, + "loss": 0.27, + "step": 8448 + }, + { + "epoch": 3.9279404927940496, + "grad_norm": 7.627802848815918, + "learning_rate": 9.9915559991556e-06, + "loss": 0.5787, + "step": 8449 + }, + { + "epoch": 3.9284053928405394, + "grad_norm": 3.4110846519470215, + "learning_rate": 9.991554999155501e-06, + "loss": 0.1743, + "step": 8450 + }, + { + "epoch": 3.928870292887029, + "grad_norm": 3.7379815578460693, + "learning_rate": 9.991553999155402e-06, + "loss": 0.224, + "step": 8451 + }, + { + "epoch": 3.9293351929335194, + "grad_norm": 4.184046268463135, + "learning_rate": 9.9915529991553e-06, + "loss": 0.3808, + "step": 8452 + }, + { + "epoch": 3.929800092980009, + "grad_norm": 11.323447227478027, + "learning_rate": 9.991551999155201e-06, + "loss": 0.439, + "step": 8453 + }, + { + "epoch": 3.9302649930264995, + "grad_norm": 4.60951042175293, + "learning_rate": 9.9915509991551e-06, + "loss": 0.3097, + "step": 8454 + }, + { + "epoch": 3.9307298930729893, + "grad_norm": 5.065977573394775, + "learning_rate": 9.991549999155001e-06, + "loss": 0.3314, + "step": 8455 + }, + { + "epoch": 3.931194793119479, + "grad_norm": 10.82296371459961, + "learning_rate": 9.991548999154902e-06, + "loss": 0.648, + "step": 8456 + }, + { + "epoch": 3.9316596931659693, + "grad_norm": 4.356511116027832, + "learning_rate": 9.9915479991548e-06, + "loss": 0.3538, + "step": 8457 + }, + { + "epoch": 3.9321245932124596, + "grad_norm": 4.477720260620117, + "learning_rate": 9.991546999154701e-06, + "loss": 0.2973, + "step": 8458 + }, + { + "epoch": 3.9325894932589494, + "grad_norm": 5.266322135925293, + "learning_rate": 9.9915459991546e-06, + "loss": 0.382, + "step": 8459 + }, + { + "epoch": 3.933054393305439, + "grad_norm": 4.467319488525391, + "learning_rate": 9.9915449991545e-06, + "loss": 0.3132, + "step": 8460 + }, + { + "epoch": 3.9335192933519294, + "grad_norm": 2.888237953186035, + "learning_rate": 9.9915439991544e-06, + "loss": 0.0756, + "step": 8461 + }, + { + "epoch": 3.933984193398419, + "grad_norm": 5.80298376083374, + "learning_rate": 9.9915429991543e-06, + "loss": 0.3383, + "step": 8462 + }, + { + "epoch": 3.9344490934449095, + "grad_norm": 5.65889310836792, + "learning_rate": 9.991541999154201e-06, + "loss": 0.3993, + "step": 8463 + }, + { + "epoch": 3.9349139934913993, + "grad_norm": 7.05470609664917, + "learning_rate": 9.991540999154102e-06, + "loss": 0.5018, + "step": 8464 + }, + { + "epoch": 3.9353788935378895, + "grad_norm": 12.175347328186035, + "learning_rate": 9.991539999154e-06, + "loss": 0.3937, + "step": 8465 + }, + { + "epoch": 3.9358437935843793, + "grad_norm": 3.7388272285461426, + "learning_rate": 9.9915389991539e-06, + "loss": 0.1196, + "step": 8466 + }, + { + "epoch": 3.9363086936308695, + "grad_norm": 4.921021938323975, + "learning_rate": 9.9915379991538e-06, + "loss": 0.2692, + "step": 8467 + }, + { + "epoch": 3.9367735936773594, + "grad_norm": 5.782711505889893, + "learning_rate": 9.991536999153701e-06, + "loss": 0.3759, + "step": 8468 + }, + { + "epoch": 3.937238493723849, + "grad_norm": 4.273814678192139, + "learning_rate": 9.991535999153602e-06, + "loss": 0.3434, + "step": 8469 + }, + { + "epoch": 3.9377033937703394, + "grad_norm": 6.356419563293457, + "learning_rate": 9.9915349991535e-06, + "loss": 0.3172, + "step": 8470 + }, + { + "epoch": 3.9381682938168296, + "grad_norm": 3.618581771850586, + "learning_rate": 9.991533999153401e-06, + "loss": 0.2389, + "step": 8471 + }, + { + "epoch": 3.9386331938633194, + "grad_norm": 7.239921569824219, + "learning_rate": 9.9915329991533e-06, + "loss": 0.5527, + "step": 8472 + }, + { + "epoch": 3.9390980939098093, + "grad_norm": 7.268379211425781, + "learning_rate": 9.9915319991532e-06, + "loss": 0.5062, + "step": 8473 + }, + { + "epoch": 3.9395629939562995, + "grad_norm": 3.448476791381836, + "learning_rate": 9.991530999153101e-06, + "loss": 0.0988, + "step": 8474 + }, + { + "epoch": 3.9400278940027893, + "grad_norm": 5.209020614624023, + "learning_rate": 9.991529999153e-06, + "loss": 0.3963, + "step": 8475 + }, + { + "epoch": 3.9404927940492795, + "grad_norm": 8.775208473205566, + "learning_rate": 9.991528999152901e-06, + "loss": 0.3181, + "step": 8476 + }, + { + "epoch": 3.9409576940957693, + "grad_norm": 3.743950128555298, + "learning_rate": 9.9915279991528e-06, + "loss": 0.2099, + "step": 8477 + }, + { + "epoch": 3.941422594142259, + "grad_norm": 3.610405206680298, + "learning_rate": 9.9915269991527e-06, + "loss": 0.2134, + "step": 8478 + }, + { + "epoch": 3.9418874941887494, + "grad_norm": 3.172701120376587, + "learning_rate": 9.9915259991526e-06, + "loss": 0.1262, + "step": 8479 + }, + { + "epoch": 3.9423523942352396, + "grad_norm": 3.66875958442688, + "learning_rate": 9.9915249991525e-06, + "loss": 0.1714, + "step": 8480 + }, + { + "epoch": 3.9428172942817294, + "grad_norm": 7.76987361907959, + "learning_rate": 9.991523999152401e-06, + "loss": 0.4669, + "step": 8481 + }, + { + "epoch": 3.9432821943282192, + "grad_norm": 3.800935745239258, + "learning_rate": 9.991522999152302e-06, + "loss": 0.1979, + "step": 8482 + }, + { + "epoch": 3.9437470943747095, + "grad_norm": 4.9012861251831055, + "learning_rate": 9.9915219991522e-06, + "loss": 0.3822, + "step": 8483 + }, + { + "epoch": 3.9442119944211993, + "grad_norm": 3.0564982891082764, + "learning_rate": 9.991520999152101e-06, + "loss": 0.1434, + "step": 8484 + }, + { + "epoch": 3.9446768944676895, + "grad_norm": 5.408553123474121, + "learning_rate": 9.991519999152e-06, + "loss": 0.4928, + "step": 8485 + }, + { + "epoch": 3.9451417945141793, + "grad_norm": 15.092555046081543, + "learning_rate": 9.991518999151901e-06, + "loss": 0.5432, + "step": 8486 + }, + { + "epoch": 3.9456066945606696, + "grad_norm": 7.326084136962891, + "learning_rate": 9.991517999151802e-06, + "loss": 0.3943, + "step": 8487 + }, + { + "epoch": 3.9460715946071594, + "grad_norm": 5.9425482749938965, + "learning_rate": 9.9915169991517e-06, + "loss": 0.2594, + "step": 8488 + }, + { + "epoch": 3.9465364946536496, + "grad_norm": 4.223303318023682, + "learning_rate": 9.991515999151601e-06, + "loss": 0.3029, + "step": 8489 + }, + { + "epoch": 3.9470013947001394, + "grad_norm": 7.315976142883301, + "learning_rate": 9.9915149991515e-06, + "loss": 0.2481, + "step": 8490 + }, + { + "epoch": 3.9474662947466292, + "grad_norm": 5.94426155090332, + "learning_rate": 9.9915139991514e-06, + "loss": 0.4799, + "step": 8491 + }, + { + "epoch": 3.9479311947931195, + "grad_norm": 5.403470039367676, + "learning_rate": 9.9915129991513e-06, + "loss": 0.4805, + "step": 8492 + }, + { + "epoch": 3.9483960948396097, + "grad_norm": 6.6591644287109375, + "learning_rate": 9.9915119991512e-06, + "loss": 0.4593, + "step": 8493 + }, + { + "epoch": 3.9488609948860995, + "grad_norm": 4.174047470092773, + "learning_rate": 9.991510999151101e-06, + "loss": 0.3219, + "step": 8494 + }, + { + "epoch": 3.9493258949325893, + "grad_norm": 5.1999053955078125, + "learning_rate": 9.991509999151002e-06, + "loss": 0.2874, + "step": 8495 + }, + { + "epoch": 3.9497907949790796, + "grad_norm": 5.581734657287598, + "learning_rate": 9.9915089991509e-06, + "loss": 0.3597, + "step": 8496 + }, + { + "epoch": 3.9502556950255694, + "grad_norm": 5.048576354980469, + "learning_rate": 9.991507999150801e-06, + "loss": 0.4101, + "step": 8497 + }, + { + "epoch": 3.9507205950720596, + "grad_norm": 5.49753475189209, + "learning_rate": 9.9915069991507e-06, + "loss": 0.3445, + "step": 8498 + }, + { + "epoch": 3.9511854951185494, + "grad_norm": 4.062564373016357, + "learning_rate": 9.991505999150601e-06, + "loss": 0.1747, + "step": 8499 + }, + { + "epoch": 3.9516503951650392, + "grad_norm": 15.640562057495117, + "learning_rate": 9.991504999150502e-06, + "loss": 0.3631, + "step": 8500 + }, + { + "epoch": 3.9521152952115295, + "grad_norm": 6.772848606109619, + "learning_rate": 9.9915039991504e-06, + "loss": 0.3416, + "step": 8501 + }, + { + "epoch": 3.9525801952580197, + "grad_norm": 3.5016682147979736, + "learning_rate": 9.991502999150301e-06, + "loss": 0.1713, + "step": 8502 + }, + { + "epoch": 3.9530450953045095, + "grad_norm": 4.698338508605957, + "learning_rate": 9.9915019991502e-06, + "loss": 0.3538, + "step": 8503 + }, + { + "epoch": 3.9535099953509993, + "grad_norm": 5.521304607391357, + "learning_rate": 9.991500999150101e-06, + "loss": 0.4348, + "step": 8504 + }, + { + "epoch": 3.9539748953974896, + "grad_norm": 9.77878189086914, + "learning_rate": 9.99149999915e-06, + "loss": 0.4607, + "step": 8505 + }, + { + "epoch": 3.9544397954439794, + "grad_norm": 4.8643083572387695, + "learning_rate": 9.9914989991499e-06, + "loss": 0.3056, + "step": 8506 + }, + { + "epoch": 3.9549046954904696, + "grad_norm": 3.608048439025879, + "learning_rate": 9.991497999149801e-06, + "loss": 0.1417, + "step": 8507 + }, + { + "epoch": 3.9553695955369594, + "grad_norm": 4.495584011077881, + "learning_rate": 9.991496999149702e-06, + "loss": 0.3594, + "step": 8508 + }, + { + "epoch": 3.9558344955834497, + "grad_norm": 6.6447882652282715, + "learning_rate": 9.9914959991496e-06, + "loss": 0.4574, + "step": 8509 + }, + { + "epoch": 3.9562993956299395, + "grad_norm": 2.606096029281616, + "learning_rate": 9.9914949991495e-06, + "loss": 0.1001, + "step": 8510 + }, + { + "epoch": 3.9567642956764297, + "grad_norm": 4.736668109893799, + "learning_rate": 9.9914939991494e-06, + "loss": 0.2512, + "step": 8511 + }, + { + "epoch": 3.9572291957229195, + "grad_norm": 7.077473163604736, + "learning_rate": 9.991492999149301e-06, + "loss": 0.4679, + "step": 8512 + }, + { + "epoch": 3.9576940957694093, + "grad_norm": 4.2238616943359375, + "learning_rate": 9.991491999149202e-06, + "loss": 0.2612, + "step": 8513 + }, + { + "epoch": 3.9581589958158996, + "grad_norm": 7.071758270263672, + "learning_rate": 9.9914909991491e-06, + "loss": 0.5491, + "step": 8514 + }, + { + "epoch": 3.95862389586239, + "grad_norm": 9.740255355834961, + "learning_rate": 9.991489999149001e-06, + "loss": 0.6357, + "step": 8515 + }, + { + "epoch": 3.9590887959088796, + "grad_norm": 4.13477087020874, + "learning_rate": 9.9914889991489e-06, + "loss": 0.2053, + "step": 8516 + }, + { + "epoch": 3.9595536959553694, + "grad_norm": 5.220872402191162, + "learning_rate": 9.991487999148801e-06, + "loss": 0.2952, + "step": 8517 + }, + { + "epoch": 3.9600185960018597, + "grad_norm": 4.801597595214844, + "learning_rate": 9.9914869991487e-06, + "loss": 0.2851, + "step": 8518 + }, + { + "epoch": 3.9604834960483495, + "grad_norm": 5.786384582519531, + "learning_rate": 9.9914859991486e-06, + "loss": 0.473, + "step": 8519 + }, + { + "epoch": 3.9609483960948397, + "grad_norm": 5.794051647186279, + "learning_rate": 9.991484999148501e-06, + "loss": 0.3281, + "step": 8520 + }, + { + "epoch": 3.9614132961413295, + "grad_norm": 9.087976455688477, + "learning_rate": 9.9914839991484e-06, + "loss": 0.6646, + "step": 8521 + }, + { + "epoch": 3.96187819618782, + "grad_norm": 5.129822254180908, + "learning_rate": 9.991482999148301e-06, + "loss": 0.2472, + "step": 8522 + }, + { + "epoch": 3.9623430962343096, + "grad_norm": 4.492726802825928, + "learning_rate": 9.9914819991482e-06, + "loss": 0.2814, + "step": 8523 + }, + { + "epoch": 3.9628079962808, + "grad_norm": 4.724972248077393, + "learning_rate": 9.9914809991481e-06, + "loss": 0.2664, + "step": 8524 + }, + { + "epoch": 3.9632728963272896, + "grad_norm": 5.558962345123291, + "learning_rate": 9.991479999148001e-06, + "loss": 0.3646, + "step": 8525 + }, + { + "epoch": 3.9637377963737794, + "grad_norm": 6.554358959197998, + "learning_rate": 9.991478999147902e-06, + "loss": 0.5103, + "step": 8526 + }, + { + "epoch": 3.9642026964202697, + "grad_norm": 4.201560020446777, + "learning_rate": 9.9914779991478e-06, + "loss": 0.1929, + "step": 8527 + }, + { + "epoch": 3.9646675964667595, + "grad_norm": 4.385227203369141, + "learning_rate": 9.991476999147701e-06, + "loss": 0.2839, + "step": 8528 + }, + { + "epoch": 3.9651324965132497, + "grad_norm": 5.761413097381592, + "learning_rate": 9.9914759991476e-06, + "loss": 0.2678, + "step": 8529 + }, + { + "epoch": 3.9655973965597395, + "grad_norm": 4.234961032867432, + "learning_rate": 9.9914749991475e-06, + "loss": 0.132, + "step": 8530 + }, + { + "epoch": 3.9660622966062298, + "grad_norm": 4.444221019744873, + "learning_rate": 9.9914739991474e-06, + "loss": 0.3408, + "step": 8531 + }, + { + "epoch": 3.9665271966527196, + "grad_norm": 3.0091261863708496, + "learning_rate": 9.9914729991473e-06, + "loss": 0.1485, + "step": 8532 + }, + { + "epoch": 3.96699209669921, + "grad_norm": 4.555489540100098, + "learning_rate": 9.991471999147201e-06, + "loss": 0.2135, + "step": 8533 + }, + { + "epoch": 3.9674569967456996, + "grad_norm": 5.563653945922852, + "learning_rate": 9.9914709991471e-06, + "loss": 0.2717, + "step": 8534 + }, + { + "epoch": 3.9679218967921894, + "grad_norm": 6.385219097137451, + "learning_rate": 9.991469999147001e-06, + "loss": 0.2311, + "step": 8535 + }, + { + "epoch": 3.9683867968386797, + "grad_norm": 5.151920318603516, + "learning_rate": 9.9914689991469e-06, + "loss": 0.2242, + "step": 8536 + }, + { + "epoch": 3.96885169688517, + "grad_norm": 5.665719985961914, + "learning_rate": 9.9914679991468e-06, + "loss": 0.2872, + "step": 8537 + }, + { + "epoch": 3.9693165969316597, + "grad_norm": 4.2066192626953125, + "learning_rate": 9.991466999146701e-06, + "loss": 0.2079, + "step": 8538 + }, + { + "epoch": 3.9697814969781495, + "grad_norm": 6.196699142456055, + "learning_rate": 9.991465999146602e-06, + "loss": 0.2307, + "step": 8539 + }, + { + "epoch": 3.9702463970246398, + "grad_norm": 4.188365459442139, + "learning_rate": 9.991464999146501e-06, + "loss": 0.2383, + "step": 8540 + }, + { + "epoch": 3.9707112970711296, + "grad_norm": 7.484074592590332, + "learning_rate": 9.9914639991464e-06, + "loss": 0.6707, + "step": 8541 + }, + { + "epoch": 3.97117619711762, + "grad_norm": 4.108683109283447, + "learning_rate": 9.9914629991463e-06, + "loss": 0.2209, + "step": 8542 + }, + { + "epoch": 3.9716410971641096, + "grad_norm": 6.02879524230957, + "learning_rate": 9.991461999146201e-06, + "loss": 0.3274, + "step": 8543 + }, + { + "epoch": 3.9721059972106, + "grad_norm": 7.439096927642822, + "learning_rate": 9.9914609991461e-06, + "loss": 0.578, + "step": 8544 + }, + { + "epoch": 3.9725708972570897, + "grad_norm": 7.496358871459961, + "learning_rate": 9.991459999146e-06, + "loss": 0.5286, + "step": 8545 + }, + { + "epoch": 3.97303579730358, + "grad_norm": 11.283209800720215, + "learning_rate": 9.991458999145901e-06, + "loss": 0.379, + "step": 8546 + }, + { + "epoch": 3.9735006973500697, + "grad_norm": 5.975988864898682, + "learning_rate": 9.9914579991458e-06, + "loss": 0.3674, + "step": 8547 + }, + { + "epoch": 3.9739655973965595, + "grad_norm": 6.980700969696045, + "learning_rate": 9.991456999145701e-06, + "loss": 0.6345, + "step": 8548 + }, + { + "epoch": 3.9744304974430498, + "grad_norm": 5.774702548980713, + "learning_rate": 9.9914559991456e-06, + "loss": 0.3846, + "step": 8549 + }, + { + "epoch": 3.97489539748954, + "grad_norm": 6.727471351623535, + "learning_rate": 9.9914549991455e-06, + "loss": 0.3295, + "step": 8550 + }, + { + "epoch": 3.97536029753603, + "grad_norm": 4.625946044921875, + "learning_rate": 9.991453999145401e-06, + "loss": 0.3147, + "step": 8551 + }, + { + "epoch": 3.9758251975825196, + "grad_norm": 6.85690975189209, + "learning_rate": 9.9914529991453e-06, + "loss": 0.5449, + "step": 8552 + }, + { + "epoch": 3.97629009762901, + "grad_norm": 5.8419294357299805, + "learning_rate": 9.991451999145201e-06, + "loss": 0.4977, + "step": 8553 + }, + { + "epoch": 3.9767549976754997, + "grad_norm": 4.3220696449279785, + "learning_rate": 9.9914509991451e-06, + "loss": 0.1916, + "step": 8554 + }, + { + "epoch": 3.97721989772199, + "grad_norm": 4.831447601318359, + "learning_rate": 9.991449999145e-06, + "loss": 0.4331, + "step": 8555 + }, + { + "epoch": 3.9776847977684797, + "grad_norm": 4.957176685333252, + "learning_rate": 9.991448999144901e-06, + "loss": 0.3871, + "step": 8556 + }, + { + "epoch": 3.9781496978149695, + "grad_norm": 5.0559401512146, + "learning_rate": 9.9914479991448e-06, + "loss": 0.4284, + "step": 8557 + }, + { + "epoch": 3.9786145978614598, + "grad_norm": 6.2175612449646, + "learning_rate": 9.9914469991447e-06, + "loss": 0.4452, + "step": 8558 + }, + { + "epoch": 3.97907949790795, + "grad_norm": 4.618626117706299, + "learning_rate": 9.991445999144602e-06, + "loss": 0.2901, + "step": 8559 + }, + { + "epoch": 3.97954439795444, + "grad_norm": 3.7352354526519775, + "learning_rate": 9.9914449991445e-06, + "loss": 0.2031, + "step": 8560 + }, + { + "epoch": 3.9800092980009296, + "grad_norm": 7.217040061950684, + "learning_rate": 9.991443999144401e-06, + "loss": 0.5287, + "step": 8561 + }, + { + "epoch": 3.98047419804742, + "grad_norm": 4.81499719619751, + "learning_rate": 9.9914429991443e-06, + "loss": 0.2482, + "step": 8562 + }, + { + "epoch": 3.9809390980939097, + "grad_norm": 4.427807331085205, + "learning_rate": 9.9914419991442e-06, + "loss": 0.2597, + "step": 8563 + }, + { + "epoch": 3.9814039981404, + "grad_norm": 5.9440016746521, + "learning_rate": 9.991440999144101e-06, + "loss": 0.3904, + "step": 8564 + }, + { + "epoch": 3.9818688981868897, + "grad_norm": 6.778812885284424, + "learning_rate": 9.991439999144e-06, + "loss": 0.4739, + "step": 8565 + }, + { + "epoch": 3.98233379823338, + "grad_norm": 5.225628852844238, + "learning_rate": 9.991438999143901e-06, + "loss": 0.3583, + "step": 8566 + }, + { + "epoch": 3.9827986982798698, + "grad_norm": 4.6378326416015625, + "learning_rate": 9.9914379991438e-06, + "loss": 0.3731, + "step": 8567 + }, + { + "epoch": 3.98326359832636, + "grad_norm": 6.473048686981201, + "learning_rate": 9.9914369991437e-06, + "loss": 0.4882, + "step": 8568 + }, + { + "epoch": 3.98372849837285, + "grad_norm": 4.402724266052246, + "learning_rate": 9.991435999143601e-06, + "loss": 0.3867, + "step": 8569 + }, + { + "epoch": 3.9841933984193396, + "grad_norm": 3.6428611278533936, + "learning_rate": 9.991434999143502e-06, + "loss": 0.1778, + "step": 8570 + }, + { + "epoch": 3.98465829846583, + "grad_norm": 8.05639934539795, + "learning_rate": 9.991433999143401e-06, + "loss": 0.4675, + "step": 8571 + }, + { + "epoch": 3.98512319851232, + "grad_norm": 32.359901428222656, + "learning_rate": 9.991432999143302e-06, + "loss": 0.1651, + "step": 8572 + }, + { + "epoch": 3.98558809855881, + "grad_norm": 8.305622100830078, + "learning_rate": 9.9914319991432e-06, + "loss": 0.7999, + "step": 8573 + }, + { + "epoch": 3.9860529986052997, + "grad_norm": 6.070267677307129, + "learning_rate": 9.9914309991431e-06, + "loss": 0.4206, + "step": 8574 + }, + { + "epoch": 3.98651789865179, + "grad_norm": 5.795929431915283, + "learning_rate": 9.991429999143e-06, + "loss": 0.35, + "step": 8575 + }, + { + "epoch": 3.9869827986982798, + "grad_norm": 4.733736515045166, + "learning_rate": 9.9914289991429e-06, + "loss": 0.2493, + "step": 8576 + }, + { + "epoch": 3.98744769874477, + "grad_norm": 7.228445053100586, + "learning_rate": 9.991427999142801e-06, + "loss": 0.5082, + "step": 8577 + }, + { + "epoch": 3.98791259879126, + "grad_norm": 5.993122577667236, + "learning_rate": 9.9914269991427e-06, + "loss": 0.399, + "step": 8578 + }, + { + "epoch": 3.9883774988377496, + "grad_norm": 5.109029769897461, + "learning_rate": 9.991425999142601e-06, + "loss": 0.3835, + "step": 8579 + }, + { + "epoch": 3.98884239888424, + "grad_norm": 5.767646312713623, + "learning_rate": 9.9914249991425e-06, + "loss": 0.338, + "step": 8580 + }, + { + "epoch": 3.98930729893073, + "grad_norm": 5.323148250579834, + "learning_rate": 9.9914239991424e-06, + "loss": 0.2603, + "step": 8581 + }, + { + "epoch": 3.98977219897722, + "grad_norm": 6.677481651306152, + "learning_rate": 9.991422999142301e-06, + "loss": 0.2305, + "step": 8582 + }, + { + "epoch": 3.9902370990237097, + "grad_norm": 5.5061140060424805, + "learning_rate": 9.991421999142202e-06, + "loss": 0.3332, + "step": 8583 + }, + { + "epoch": 3.9907019990702, + "grad_norm": 4.673863887786865, + "learning_rate": 9.991420999142101e-06, + "loss": 0.3059, + "step": 8584 + }, + { + "epoch": 3.9911668991166898, + "grad_norm": 4.621357440948486, + "learning_rate": 9.991419999142e-06, + "loss": 0.2149, + "step": 8585 + }, + { + "epoch": 3.99163179916318, + "grad_norm": 5.325860500335693, + "learning_rate": 9.9914189991419e-06, + "loss": 0.1646, + "step": 8586 + }, + { + "epoch": 3.99209669920967, + "grad_norm": 5.257841110229492, + "learning_rate": 9.9914179991418e-06, + "loss": 0.4528, + "step": 8587 + }, + { + "epoch": 3.99256159925616, + "grad_norm": 5.73211145401001, + "learning_rate": 9.9914169991417e-06, + "loss": 0.247, + "step": 8588 + }, + { + "epoch": 3.99302649930265, + "grad_norm": 5.323985576629639, + "learning_rate": 9.991415999141601e-06, + "loss": 0.2698, + "step": 8589 + }, + { + "epoch": 3.99349139934914, + "grad_norm": 6.080668926239014, + "learning_rate": 9.991414999141502e-06, + "loss": 0.2783, + "step": 8590 + }, + { + "epoch": 3.99395629939563, + "grad_norm": 6.122230529785156, + "learning_rate": 9.9914139991414e-06, + "loss": 0.4544, + "step": 8591 + }, + { + "epoch": 3.9944211994421197, + "grad_norm": 3.395712375640869, + "learning_rate": 9.991412999141301e-06, + "loss": 0.1756, + "step": 8592 + }, + { + "epoch": 3.99488609948861, + "grad_norm": 4.871823787689209, + "learning_rate": 9.9914119991412e-06, + "loss": 0.2849, + "step": 8593 + }, + { + "epoch": 3.9953509995351, + "grad_norm": 3.7678544521331787, + "learning_rate": 9.9914109991411e-06, + "loss": 0.1794, + "step": 8594 + }, + { + "epoch": 3.99581589958159, + "grad_norm": 10.889883995056152, + "learning_rate": 9.991409999141001e-06, + "loss": 0.4015, + "step": 8595 + }, + { + "epoch": 3.99628079962808, + "grad_norm": 6.349218368530273, + "learning_rate": 9.9914089991409e-06, + "loss": 0.4993, + "step": 8596 + }, + { + "epoch": 3.99674569967457, + "grad_norm": 6.27474308013916, + "learning_rate": 9.991407999140801e-06, + "loss": 0.3769, + "step": 8597 + }, + { + "epoch": 3.99721059972106, + "grad_norm": 8.47109603881836, + "learning_rate": 9.9914069991407e-06, + "loss": 0.5774, + "step": 8598 + }, + { + "epoch": 3.99767549976755, + "grad_norm": 5.892036437988281, + "learning_rate": 9.9914059991406e-06, + "loss": 0.3303, + "step": 8599 + }, + { + "epoch": 3.99814039981404, + "grad_norm": 6.108529090881348, + "learning_rate": 9.9914049991405e-06, + "loss": 0.3429, + "step": 8600 + }, + { + "epoch": 3.99860529986053, + "grad_norm": 6.202589511871338, + "learning_rate": 9.9914039991404e-06, + "loss": 0.4616, + "step": 8601 + }, + { + "epoch": 3.99907019990702, + "grad_norm": 4.393272876739502, + "learning_rate": 9.991402999140301e-06, + "loss": 0.1887, + "step": 8602 + }, + { + "epoch": 3.99953509995351, + "grad_norm": 5.726790904998779, + "learning_rate": 9.991401999140202e-06, + "loss": 0.3528, + "step": 8603 + }, + { + "epoch": 4.0, + "grad_norm": 5.867901802062988, + "learning_rate": 9.9914009991401e-06, + "loss": 0.4372, + "step": 8604 + }, + { + "epoch": 4.00046490004649, + "grad_norm": 5.187891006469727, + "learning_rate": 9.991399999140001e-06, + "loss": 0.5663, + "step": 8605 + }, + { + "epoch": 4.00092980009298, + "grad_norm": 5.950860500335693, + "learning_rate": 9.9913989991399e-06, + "loss": 0.4672, + "step": 8606 + }, + { + "epoch": 4.00139470013947, + "grad_norm": 4.8320536613464355, + "learning_rate": 9.991397999139801e-06, + "loss": 0.3684, + "step": 8607 + }, + { + "epoch": 4.00185960018596, + "grad_norm": 4.109183311462402, + "learning_rate": 9.991396999139702e-06, + "loss": 0.1127, + "step": 8608 + }, + { + "epoch": 4.00232450023245, + "grad_norm": 4.985267639160156, + "learning_rate": 9.9913959991396e-06, + "loss": 0.2863, + "step": 8609 + }, + { + "epoch": 4.00278940027894, + "grad_norm": 6.051086902618408, + "learning_rate": 9.991394999139501e-06, + "loss": 0.3192, + "step": 8610 + }, + { + "epoch": 4.00325430032543, + "grad_norm": 5.303004264831543, + "learning_rate": 9.9913939991394e-06, + "loss": 0.2532, + "step": 8611 + }, + { + "epoch": 4.00371920037192, + "grad_norm": 4.901470184326172, + "learning_rate": 9.9913929991393e-06, + "loss": 0.4679, + "step": 8612 + }, + { + "epoch": 4.00418410041841, + "grad_norm": 4.076053142547607, + "learning_rate": 9.9913919991392e-06, + "loss": 0.2727, + "step": 8613 + }, + { + "epoch": 4.0046490004649, + "grad_norm": 4.103943824768066, + "learning_rate": 9.9913909991391e-06, + "loss": 0.2781, + "step": 8614 + }, + { + "epoch": 4.0051139005113905, + "grad_norm": 5.813246250152588, + "learning_rate": 9.991389999139001e-06, + "loss": 0.3669, + "step": 8615 + }, + { + "epoch": 4.00557880055788, + "grad_norm": 6.515575885772705, + "learning_rate": 9.991388999138902e-06, + "loss": 0.3182, + "step": 8616 + }, + { + "epoch": 4.00604370060437, + "grad_norm": 3.325073003768921, + "learning_rate": 9.9913879991388e-06, + "loss": 0.2029, + "step": 8617 + }, + { + "epoch": 4.00650860065086, + "grad_norm": 5.203667640686035, + "learning_rate": 9.9913869991387e-06, + "loss": 0.3095, + "step": 8618 + }, + { + "epoch": 4.00697350069735, + "grad_norm": 3.1638309955596924, + "learning_rate": 9.9913859991386e-06, + "loss": 0.185, + "step": 8619 + }, + { + "epoch": 4.00743840074384, + "grad_norm": 5.068531513214111, + "learning_rate": 9.991384999138501e-06, + "loss": 0.5541, + "step": 8620 + }, + { + "epoch": 4.00790330079033, + "grad_norm": 4.684556007385254, + "learning_rate": 9.991383999138402e-06, + "loss": 0.2452, + "step": 8621 + }, + { + "epoch": 4.00836820083682, + "grad_norm": 6.760135650634766, + "learning_rate": 9.9913829991383e-06, + "loss": 0.4358, + "step": 8622 + }, + { + "epoch": 4.00883310088331, + "grad_norm": 3.253958225250244, + "learning_rate": 9.991381999138201e-06, + "loss": 0.0973, + "step": 8623 + }, + { + "epoch": 4.0092980009298005, + "grad_norm": 4.3201093673706055, + "learning_rate": 9.9913809991381e-06, + "loss": 0.2082, + "step": 8624 + }, + { + "epoch": 4.00976290097629, + "grad_norm": 4.246491432189941, + "learning_rate": 9.991379999138001e-06, + "loss": 0.2203, + "step": 8625 + }, + { + "epoch": 4.01022780102278, + "grad_norm": 8.05374813079834, + "learning_rate": 9.9913789991379e-06, + "loss": 0.4358, + "step": 8626 + }, + { + "epoch": 4.01069270106927, + "grad_norm": 5.256668567657471, + "learning_rate": 9.9913779991378e-06, + "loss": 0.4198, + "step": 8627 + }, + { + "epoch": 4.01115760111576, + "grad_norm": 5.227355480194092, + "learning_rate": 9.991376999137701e-06, + "loss": 0.4656, + "step": 8628 + }, + { + "epoch": 4.01162250116225, + "grad_norm": 3.9588303565979004, + "learning_rate": 9.9913759991376e-06, + "loss": 0.2249, + "step": 8629 + }, + { + "epoch": 4.01208740120874, + "grad_norm": 4.0043253898620605, + "learning_rate": 9.9913749991375e-06, + "loss": 0.1837, + "step": 8630 + }, + { + "epoch": 4.01255230125523, + "grad_norm": 8.527580261230469, + "learning_rate": 9.9913739991374e-06, + "loss": 0.6355, + "step": 8631 + }, + { + "epoch": 4.01301720130172, + "grad_norm": 10.546321868896484, + "learning_rate": 9.9913729991373e-06, + "loss": 0.3739, + "step": 8632 + }, + { + "epoch": 4.0134821013482105, + "grad_norm": 5.7071356773376465, + "learning_rate": 9.991371999137201e-06, + "loss": 0.4417, + "step": 8633 + }, + { + "epoch": 4.0139470013947, + "grad_norm": 11.17810344696045, + "learning_rate": 9.991370999137102e-06, + "loss": 0.6927, + "step": 8634 + }, + { + "epoch": 4.01441190144119, + "grad_norm": 4.833978652954102, + "learning_rate": 9.991369999137e-06, + "loss": 0.3242, + "step": 8635 + }, + { + "epoch": 4.01487680148768, + "grad_norm": 6.374232292175293, + "learning_rate": 9.991368999136901e-06, + "loss": 0.4722, + "step": 8636 + }, + { + "epoch": 4.015341701534171, + "grad_norm": 7.027050018310547, + "learning_rate": 9.9913679991368e-06, + "loss": 0.3924, + "step": 8637 + }, + { + "epoch": 4.01580660158066, + "grad_norm": 3.5113179683685303, + "learning_rate": 9.991366999136701e-06, + "loss": 0.171, + "step": 8638 + }, + { + "epoch": 4.01627150162715, + "grad_norm": 6.112441539764404, + "learning_rate": 9.991365999136602e-06, + "loss": 0.2908, + "step": 8639 + }, + { + "epoch": 4.01673640167364, + "grad_norm": 5.1256537437438965, + "learning_rate": 9.9913649991365e-06, + "loss": 0.1808, + "step": 8640 + }, + { + "epoch": 4.01720130172013, + "grad_norm": 4.693594455718994, + "learning_rate": 9.991363999136401e-06, + "loss": 0.176, + "step": 8641 + }, + { + "epoch": 4.0176662017666205, + "grad_norm": 6.986037254333496, + "learning_rate": 9.9913629991363e-06, + "loss": 0.4379, + "step": 8642 + }, + { + "epoch": 4.01813110181311, + "grad_norm": 7.422395706176758, + "learning_rate": 9.991361999136201e-06, + "loss": 0.588, + "step": 8643 + }, + { + "epoch": 4.0185960018596, + "grad_norm": 4.2769670486450195, + "learning_rate": 9.9913609991361e-06, + "loss": 0.1929, + "step": 8644 + }, + { + "epoch": 4.01906090190609, + "grad_norm": 4.313206672668457, + "learning_rate": 9.991359999136e-06, + "loss": 0.1973, + "step": 8645 + }, + { + "epoch": 4.019525801952581, + "grad_norm": 6.814403057098389, + "learning_rate": 9.991358999135901e-06, + "loss": 0.2321, + "step": 8646 + }, + { + "epoch": 4.01999070199907, + "grad_norm": 1.9174803495407104, + "learning_rate": 9.991357999135802e-06, + "loss": 0.0929, + "step": 8647 + }, + { + "epoch": 4.02045560204556, + "grad_norm": 5.194610595703125, + "learning_rate": 9.9913569991357e-06, + "loss": 0.2258, + "step": 8648 + }, + { + "epoch": 4.02092050209205, + "grad_norm": 4.322388172149658, + "learning_rate": 9.991355999135601e-06, + "loss": 0.2373, + "step": 8649 + }, + { + "epoch": 4.02138540213854, + "grad_norm": 5.233599662780762, + "learning_rate": 9.9913549991355e-06, + "loss": 0.272, + "step": 8650 + }, + { + "epoch": 4.0218503021850305, + "grad_norm": 5.142061233520508, + "learning_rate": 9.991353999135401e-06, + "loss": 0.2832, + "step": 8651 + }, + { + "epoch": 4.02231520223152, + "grad_norm": 5.7198100090026855, + "learning_rate": 9.991352999135302e-06, + "loss": 0.3599, + "step": 8652 + }, + { + "epoch": 4.02278010227801, + "grad_norm": 5.3642754554748535, + "learning_rate": 9.9913519991352e-06, + "loss": 0.2524, + "step": 8653 + }, + { + "epoch": 4.0232450023245, + "grad_norm": 3.077054262161255, + "learning_rate": 9.991350999135101e-06, + "loss": 0.1724, + "step": 8654 + }, + { + "epoch": 4.023709902370991, + "grad_norm": 4.753485202789307, + "learning_rate": 9.991349999135e-06, + "loss": 0.228, + "step": 8655 + }, + { + "epoch": 4.02417480241748, + "grad_norm": 5.652727127075195, + "learning_rate": 9.991348999134901e-06, + "loss": 0.5214, + "step": 8656 + }, + { + "epoch": 4.02463970246397, + "grad_norm": 4.327095031738281, + "learning_rate": 9.9913479991348e-06, + "loss": 0.1735, + "step": 8657 + }, + { + "epoch": 4.02510460251046, + "grad_norm": 4.165304660797119, + "learning_rate": 9.9913469991347e-06, + "loss": 0.1451, + "step": 8658 + }, + { + "epoch": 4.025569502556951, + "grad_norm": 6.0113348960876465, + "learning_rate": 9.991345999134601e-06, + "loss": 0.393, + "step": 8659 + }, + { + "epoch": 4.0260344026034405, + "grad_norm": 6.444305896759033, + "learning_rate": 9.9913449991345e-06, + "loss": 0.3344, + "step": 8660 + }, + { + "epoch": 4.02649930264993, + "grad_norm": 3.2772088050842285, + "learning_rate": 9.9913439991344e-06, + "loss": 0.1003, + "step": 8661 + }, + { + "epoch": 4.02696420269642, + "grad_norm": 2.957395553588867, + "learning_rate": 9.9913429991343e-06, + "loss": 0.1221, + "step": 8662 + }, + { + "epoch": 4.02742910274291, + "grad_norm": 6.021089553833008, + "learning_rate": 9.9913419991342e-06, + "loss": 0.3278, + "step": 8663 + }, + { + "epoch": 4.027894002789401, + "grad_norm": 5.935059547424316, + "learning_rate": 9.991340999134101e-06, + "loss": 0.4063, + "step": 8664 + }, + { + "epoch": 4.02835890283589, + "grad_norm": 8.302364349365234, + "learning_rate": 9.991339999134002e-06, + "loss": 0.499, + "step": 8665 + }, + { + "epoch": 4.02882380288238, + "grad_norm": 8.222038269042969, + "learning_rate": 9.9913389991339e-06, + "loss": 0.5486, + "step": 8666 + }, + { + "epoch": 4.02928870292887, + "grad_norm": 4.9365949630737305, + "learning_rate": 9.991337999133801e-06, + "loss": 0.3697, + "step": 8667 + }, + { + "epoch": 4.029753602975361, + "grad_norm": 7.696190357208252, + "learning_rate": 9.9913369991337e-06, + "loss": 0.3918, + "step": 8668 + }, + { + "epoch": 4.0302185030218505, + "grad_norm": 3.8297672271728516, + "learning_rate": 9.991335999133601e-06, + "loss": 0.1695, + "step": 8669 + }, + { + "epoch": 4.03068340306834, + "grad_norm": 5.684112548828125, + "learning_rate": 9.9913349991335e-06, + "loss": 0.409, + "step": 8670 + }, + { + "epoch": 4.03114830311483, + "grad_norm": 3.5197582244873047, + "learning_rate": 9.9913339991334e-06, + "loss": 0.0898, + "step": 8671 + }, + { + "epoch": 4.03161320316132, + "grad_norm": 4.188271999359131, + "learning_rate": 9.991332999133301e-06, + "loss": 0.2626, + "step": 8672 + }, + { + "epoch": 4.032078103207811, + "grad_norm": 5.148788928985596, + "learning_rate": 9.9913319991332e-06, + "loss": 0.3208, + "step": 8673 + }, + { + "epoch": 4.0325430032543, + "grad_norm": 6.853492736816406, + "learning_rate": 9.991330999133101e-06, + "loss": 0.3443, + "step": 8674 + }, + { + "epoch": 4.03300790330079, + "grad_norm": 7.490018367767334, + "learning_rate": 9.991329999133e-06, + "loss": 0.465, + "step": 8675 + }, + { + "epoch": 4.03347280334728, + "grad_norm": 5.811092853546143, + "learning_rate": 9.9913289991329e-06, + "loss": 0.2587, + "step": 8676 + }, + { + "epoch": 4.033937703393771, + "grad_norm": 5.507937908172607, + "learning_rate": 9.991327999132801e-06, + "loss": 0.2781, + "step": 8677 + }, + { + "epoch": 4.0344026034402605, + "grad_norm": 10.831843376159668, + "learning_rate": 9.991326999132702e-06, + "loss": 0.4715, + "step": 8678 + }, + { + "epoch": 4.03486750348675, + "grad_norm": 3.8370964527130127, + "learning_rate": 9.9913259991326e-06, + "loss": 0.0882, + "step": 8679 + }, + { + "epoch": 4.03533240353324, + "grad_norm": 5.249598503112793, + "learning_rate": 9.991324999132501e-06, + "loss": 0.4494, + "step": 8680 + }, + { + "epoch": 4.035797303579731, + "grad_norm": 5.6947221755981445, + "learning_rate": 9.9913239991324e-06, + "loss": 0.2601, + "step": 8681 + }, + { + "epoch": 4.036262203626221, + "grad_norm": 5.569007396697998, + "learning_rate": 9.9913229991323e-06, + "loss": 0.1145, + "step": 8682 + }, + { + "epoch": 4.03672710367271, + "grad_norm": 4.697781085968018, + "learning_rate": 9.9913219991322e-06, + "loss": 0.2116, + "step": 8683 + }, + { + "epoch": 4.0371920037192, + "grad_norm": 4.213014602661133, + "learning_rate": 9.9913209991321e-06, + "loss": 0.3495, + "step": 8684 + }, + { + "epoch": 4.03765690376569, + "grad_norm": 8.369109153747559, + "learning_rate": 9.991319999132001e-06, + "loss": 0.2844, + "step": 8685 + }, + { + "epoch": 4.038121803812181, + "grad_norm": 7.441469192504883, + "learning_rate": 9.9913189991319e-06, + "loss": 0.4657, + "step": 8686 + }, + { + "epoch": 4.0385867038586705, + "grad_norm": 3.812077760696411, + "learning_rate": 9.991317999131801e-06, + "loss": 0.1998, + "step": 8687 + }, + { + "epoch": 4.03905160390516, + "grad_norm": 3.4826109409332275, + "learning_rate": 9.9913169991317e-06, + "loss": 0.1507, + "step": 8688 + }, + { + "epoch": 4.03951650395165, + "grad_norm": 5.82440185546875, + "learning_rate": 9.9913159991316e-06, + "loss": 0.4444, + "step": 8689 + }, + { + "epoch": 4.039981403998141, + "grad_norm": 4.371851921081543, + "learning_rate": 9.991314999131501e-06, + "loss": 0.2686, + "step": 8690 + }, + { + "epoch": 4.040446304044631, + "grad_norm": 4.21297550201416, + "learning_rate": 9.991313999131402e-06, + "loss": 0.1851, + "step": 8691 + }, + { + "epoch": 4.04091120409112, + "grad_norm": 4.9405012130737305, + "learning_rate": 9.991312999131301e-06, + "loss": 0.3732, + "step": 8692 + }, + { + "epoch": 4.04137610413761, + "grad_norm": 8.477139472961426, + "learning_rate": 9.9913119991312e-06, + "loss": 0.442, + "step": 8693 + }, + { + "epoch": 4.0418410041841, + "grad_norm": 3.714805841445923, + "learning_rate": 9.9913109991311e-06, + "loss": 0.2921, + "step": 8694 + }, + { + "epoch": 4.042305904230591, + "grad_norm": 3.7897043228149414, + "learning_rate": 9.991309999131e-06, + "loss": 0.0688, + "step": 8695 + }, + { + "epoch": 4.0427708042770805, + "grad_norm": 4.406866550445557, + "learning_rate": 9.9913089991309e-06, + "loss": 0.3563, + "step": 8696 + }, + { + "epoch": 4.04323570432357, + "grad_norm": 4.63918924331665, + "learning_rate": 9.9913079991308e-06, + "loss": 0.2137, + "step": 8697 + }, + { + "epoch": 4.04370060437006, + "grad_norm": 6.932974815368652, + "learning_rate": 9.991306999130701e-06, + "loss": 0.3984, + "step": 8698 + }, + { + "epoch": 4.044165504416551, + "grad_norm": 5.462420463562012, + "learning_rate": 9.9913059991306e-06, + "loss": 0.4014, + "step": 8699 + }, + { + "epoch": 4.044630404463041, + "grad_norm": 5.66067361831665, + "learning_rate": 9.991304999130501e-06, + "loss": 0.2267, + "step": 8700 + }, + { + "epoch": 4.04509530450953, + "grad_norm": 2.781317710876465, + "learning_rate": 9.9913039991304e-06, + "loss": 0.1005, + "step": 8701 + }, + { + "epoch": 4.04556020455602, + "grad_norm": 6.7585530281066895, + "learning_rate": 9.9913029991303e-06, + "loss": 0.584, + "step": 8702 + }, + { + "epoch": 4.046025104602511, + "grad_norm": 5.346123695373535, + "learning_rate": 9.991301999130201e-06, + "loss": 0.2868, + "step": 8703 + }, + { + "epoch": 4.046490004649001, + "grad_norm": 3.5586841106414795, + "learning_rate": 9.9913009991301e-06, + "loss": 0.3196, + "step": 8704 + }, + { + "epoch": 4.0469549046954905, + "grad_norm": 6.6799397468566895, + "learning_rate": 9.991299999130001e-06, + "loss": 0.3662, + "step": 8705 + }, + { + "epoch": 4.04741980474198, + "grad_norm": 3.313336133956909, + "learning_rate": 9.9912989991299e-06, + "loss": 0.2238, + "step": 8706 + }, + { + "epoch": 4.04788470478847, + "grad_norm": 3.212474822998047, + "learning_rate": 9.9912979991298e-06, + "loss": 0.1481, + "step": 8707 + }, + { + "epoch": 4.048349604834961, + "grad_norm": 5.439833641052246, + "learning_rate": 9.991296999129701e-06, + "loss": 0.2959, + "step": 8708 + }, + { + "epoch": 4.048814504881451, + "grad_norm": 6.791680335998535, + "learning_rate": 9.9912959991296e-06, + "loss": 0.2863, + "step": 8709 + }, + { + "epoch": 4.04927940492794, + "grad_norm": 7.155752182006836, + "learning_rate": 9.991294999129501e-06, + "loss": 0.256, + "step": 8710 + }, + { + "epoch": 4.04974430497443, + "grad_norm": 4.139428615570068, + "learning_rate": 9.991293999129402e-06, + "loss": 0.2227, + "step": 8711 + }, + { + "epoch": 4.050209205020921, + "grad_norm": 2.822733163833618, + "learning_rate": 9.9912929991293e-06, + "loss": 0.0943, + "step": 8712 + }, + { + "epoch": 4.050674105067411, + "grad_norm": 6.384578227996826, + "learning_rate": 9.991291999129201e-06, + "loss": 0.4025, + "step": 8713 + }, + { + "epoch": 4.0511390051139005, + "grad_norm": 4.935233116149902, + "learning_rate": 9.9912909991291e-06, + "loss": 0.305, + "step": 8714 + }, + { + "epoch": 4.05160390516039, + "grad_norm": 5.896101474761963, + "learning_rate": 9.991289999129e-06, + "loss": 0.3169, + "step": 8715 + }, + { + "epoch": 4.052068805206881, + "grad_norm": 8.032756805419922, + "learning_rate": 9.991288999128901e-06, + "loss": 0.3821, + "step": 8716 + }, + { + "epoch": 4.052533705253371, + "grad_norm": 4.666884899139404, + "learning_rate": 9.9912879991288e-06, + "loss": 0.2742, + "step": 8717 + }, + { + "epoch": 4.052998605299861, + "grad_norm": 4.245198726654053, + "learning_rate": 9.991286999128701e-06, + "loss": 0.2766, + "step": 8718 + }, + { + "epoch": 4.05346350534635, + "grad_norm": 5.824294090270996, + "learning_rate": 9.9912859991286e-06, + "loss": 0.2851, + "step": 8719 + }, + { + "epoch": 4.05392840539284, + "grad_norm": 9.294486999511719, + "learning_rate": 9.9912849991285e-06, + "loss": 0.5356, + "step": 8720 + }, + { + "epoch": 4.054393305439331, + "grad_norm": 6.033421516418457, + "learning_rate": 9.991283999128401e-06, + "loss": 0.3486, + "step": 8721 + }, + { + "epoch": 4.054858205485821, + "grad_norm": 5.244217395782471, + "learning_rate": 9.9912829991283e-06, + "loss": 0.257, + "step": 8722 + }, + { + "epoch": 4.0553231055323105, + "grad_norm": 6.787354469299316, + "learning_rate": 9.991281999128201e-06, + "loss": 0.5185, + "step": 8723 + }, + { + "epoch": 4.0557880055788, + "grad_norm": 5.0287580490112305, + "learning_rate": 9.991280999128102e-06, + "loss": 0.1475, + "step": 8724 + }, + { + "epoch": 4.056252905625291, + "grad_norm": 8.2864351272583, + "learning_rate": 9.991279999128e-06, + "loss": 0.3436, + "step": 8725 + }, + { + "epoch": 4.056717805671781, + "grad_norm": 3.084397554397583, + "learning_rate": 9.9912789991279e-06, + "loss": 0.1786, + "step": 8726 + }, + { + "epoch": 4.0571827057182706, + "grad_norm": 5.651367664337158, + "learning_rate": 9.9912779991278e-06, + "loss": 0.3459, + "step": 8727 + }, + { + "epoch": 4.05764760576476, + "grad_norm": 3.583228826522827, + "learning_rate": 9.991276999127701e-06, + "loss": 0.1937, + "step": 8728 + }, + { + "epoch": 4.05811250581125, + "grad_norm": 4.2244157791137695, + "learning_rate": 9.991275999127602e-06, + "loss": 0.203, + "step": 8729 + }, + { + "epoch": 4.058577405857741, + "grad_norm": 4.705873966217041, + "learning_rate": 9.9912749991275e-06, + "loss": 0.1973, + "step": 8730 + }, + { + "epoch": 4.059042305904231, + "grad_norm": 6.1338419914245605, + "learning_rate": 9.991273999127401e-06, + "loss": 0.2295, + "step": 8731 + }, + { + "epoch": 4.0595072059507205, + "grad_norm": 8.145644187927246, + "learning_rate": 9.9912729991273e-06, + "loss": 0.2946, + "step": 8732 + }, + { + "epoch": 4.05997210599721, + "grad_norm": 5.177975654602051, + "learning_rate": 9.9912719991272e-06, + "loss": 0.3274, + "step": 8733 + }, + { + "epoch": 4.060437006043701, + "grad_norm": 4.647773742675781, + "learning_rate": 9.991270999127101e-06, + "loss": 0.2948, + "step": 8734 + }, + { + "epoch": 4.060901906090191, + "grad_norm": 4.529958248138428, + "learning_rate": 9.991269999127002e-06, + "loss": 0.2754, + "step": 8735 + }, + { + "epoch": 4.0613668061366806, + "grad_norm": 6.9650068283081055, + "learning_rate": 9.991268999126901e-06, + "loss": 0.4427, + "step": 8736 + }, + { + "epoch": 4.06183170618317, + "grad_norm": 4.637859344482422, + "learning_rate": 9.9912679991268e-06, + "loss": 0.2432, + "step": 8737 + }, + { + "epoch": 4.062296606229661, + "grad_norm": 4.5785346031188965, + "learning_rate": 9.9912669991267e-06, + "loss": 0.188, + "step": 8738 + }, + { + "epoch": 4.062761506276151, + "grad_norm": 7.371599197387695, + "learning_rate": 9.9912659991266e-06, + "loss": 0.2717, + "step": 8739 + }, + { + "epoch": 4.063226406322641, + "grad_norm": 5.677791595458984, + "learning_rate": 9.9912649991265e-06, + "loss": 0.398, + "step": 8740 + }, + { + "epoch": 4.0636913063691305, + "grad_norm": 6.475762367248535, + "learning_rate": 9.991263999126401e-06, + "loss": 0.4622, + "step": 8741 + }, + { + "epoch": 4.06415620641562, + "grad_norm": 3.690894842147827, + "learning_rate": 9.991262999126302e-06, + "loss": 0.1641, + "step": 8742 + }, + { + "epoch": 4.064621106462111, + "grad_norm": 3.374966859817505, + "learning_rate": 9.9912619991262e-06, + "loss": 0.1496, + "step": 8743 + }, + { + "epoch": 4.065086006508601, + "grad_norm": 4.797285556793213, + "learning_rate": 9.991260999126101e-06, + "loss": 0.2577, + "step": 8744 + }, + { + "epoch": 4.0655509065550905, + "grad_norm": 3.5871763229370117, + "learning_rate": 9.991259999126e-06, + "loss": 0.1355, + "step": 8745 + }, + { + "epoch": 4.06601580660158, + "grad_norm": 4.721572399139404, + "learning_rate": 9.9912589991259e-06, + "loss": 0.1684, + "step": 8746 + }, + { + "epoch": 4.066480706648071, + "grad_norm": 3.457909107208252, + "learning_rate": 9.991257999125801e-06, + "loss": 0.1609, + "step": 8747 + }, + { + "epoch": 4.066945606694561, + "grad_norm": 5.959712982177734, + "learning_rate": 9.9912569991257e-06, + "loss": 0.4256, + "step": 8748 + }, + { + "epoch": 4.067410506741051, + "grad_norm": 4.462769031524658, + "learning_rate": 9.991255999125601e-06, + "loss": 0.2527, + "step": 8749 + }, + { + "epoch": 4.0678754067875404, + "grad_norm": 5.463027000427246, + "learning_rate": 9.9912549991255e-06, + "loss": 0.3409, + "step": 8750 + }, + { + "epoch": 4.06834030683403, + "grad_norm": 4.407458782196045, + "learning_rate": 9.9912539991254e-06, + "loss": 0.321, + "step": 8751 + }, + { + "epoch": 4.068805206880521, + "grad_norm": 6.564177513122559, + "learning_rate": 9.9912529991253e-06, + "loss": 0.2736, + "step": 8752 + }, + { + "epoch": 4.069270106927011, + "grad_norm": 2.780529737472534, + "learning_rate": 9.9912519991252e-06, + "loss": 0.1563, + "step": 8753 + }, + { + "epoch": 4.0697350069735005, + "grad_norm": 4.1208672523498535, + "learning_rate": 9.991250999125101e-06, + "loss": 0.2612, + "step": 8754 + }, + { + "epoch": 4.07019990701999, + "grad_norm": 3.569709539413452, + "learning_rate": 9.991249999125002e-06, + "loss": 0.1196, + "step": 8755 + }, + { + "epoch": 4.070664807066481, + "grad_norm": 5.429008960723877, + "learning_rate": 9.9912489991249e-06, + "loss": 0.1795, + "step": 8756 + }, + { + "epoch": 4.071129707112971, + "grad_norm": 4.251283168792725, + "learning_rate": 9.991247999124801e-06, + "loss": 0.1751, + "step": 8757 + }, + { + "epoch": 4.071594607159461, + "grad_norm": 2.9475550651550293, + "learning_rate": 9.9912469991247e-06, + "loss": 0.1078, + "step": 8758 + }, + { + "epoch": 4.0720595072059504, + "grad_norm": 3.892535924911499, + "learning_rate": 9.991245999124601e-06, + "loss": 0.1343, + "step": 8759 + }, + { + "epoch": 4.072524407252441, + "grad_norm": 8.41137981414795, + "learning_rate": 9.991244999124502e-06, + "loss": 0.5577, + "step": 8760 + }, + { + "epoch": 4.072989307298931, + "grad_norm": 6.672977924346924, + "learning_rate": 9.9912439991244e-06, + "loss": 0.601, + "step": 8761 + }, + { + "epoch": 4.073454207345421, + "grad_norm": 4.674140453338623, + "learning_rate": 9.991242999124301e-06, + "loss": 0.211, + "step": 8762 + }, + { + "epoch": 4.0739191073919105, + "grad_norm": 4.7865471839904785, + "learning_rate": 9.9912419991242e-06, + "loss": 0.2324, + "step": 8763 + }, + { + "epoch": 4.0743840074384, + "grad_norm": 3.07725191116333, + "learning_rate": 9.9912409991241e-06, + "loss": 0.1541, + "step": 8764 + }, + { + "epoch": 4.074848907484891, + "grad_norm": 6.369438171386719, + "learning_rate": 9.991239999124e-06, + "loss": 0.4517, + "step": 8765 + }, + { + "epoch": 4.075313807531381, + "grad_norm": 4.789504528045654, + "learning_rate": 9.9912389991239e-06, + "loss": 0.1976, + "step": 8766 + }, + { + "epoch": 4.075778707577871, + "grad_norm": 4.585958957672119, + "learning_rate": 9.991237999123801e-06, + "loss": 0.2189, + "step": 8767 + }, + { + "epoch": 4.07624360762436, + "grad_norm": 4.5338826179504395, + "learning_rate": 9.991236999123702e-06, + "loss": 0.3315, + "step": 8768 + }, + { + "epoch": 4.076708507670851, + "grad_norm": 4.848607540130615, + "learning_rate": 9.9912359991236e-06, + "loss": 0.2725, + "step": 8769 + }, + { + "epoch": 4.077173407717341, + "grad_norm": 5.090126991271973, + "learning_rate": 9.9912349991235e-06, + "loss": 0.2903, + "step": 8770 + }, + { + "epoch": 4.077638307763831, + "grad_norm": 4.61611270904541, + "learning_rate": 9.9912339991234e-06, + "loss": 0.2773, + "step": 8771 + }, + { + "epoch": 4.0781032078103205, + "grad_norm": 4.503048419952393, + "learning_rate": 9.991232999123301e-06, + "loss": 0.2181, + "step": 8772 + }, + { + "epoch": 4.07856810785681, + "grad_norm": 5.908793926239014, + "learning_rate": 9.991231999123202e-06, + "loss": 0.3063, + "step": 8773 + }, + { + "epoch": 4.079033007903301, + "grad_norm": 5.420773983001709, + "learning_rate": 9.9912309991231e-06, + "loss": 0.3528, + "step": 8774 + }, + { + "epoch": 4.079497907949791, + "grad_norm": 5.181894779205322, + "learning_rate": 9.991229999123001e-06, + "loss": 0.3112, + "step": 8775 + }, + { + "epoch": 4.079962807996281, + "grad_norm": 5.403919696807861, + "learning_rate": 9.9912289991229e-06, + "loss": 0.3731, + "step": 8776 + }, + { + "epoch": 4.08042770804277, + "grad_norm": 4.653299808502197, + "learning_rate": 9.991227999122801e-06, + "loss": 0.3045, + "step": 8777 + }, + { + "epoch": 4.080892608089261, + "grad_norm": 6.771958827972412, + "learning_rate": 9.9912269991227e-06, + "loss": 0.4309, + "step": 8778 + }, + { + "epoch": 4.081357508135751, + "grad_norm": 4.336580753326416, + "learning_rate": 9.9912259991226e-06, + "loss": 0.2187, + "step": 8779 + }, + { + "epoch": 4.081822408182241, + "grad_norm": 5.43284273147583, + "learning_rate": 9.991224999122501e-06, + "loss": 0.3879, + "step": 8780 + }, + { + "epoch": 4.0822873082287305, + "grad_norm": 4.96766471862793, + "learning_rate": 9.9912239991224e-06, + "loss": 0.4006, + "step": 8781 + }, + { + "epoch": 4.082752208275221, + "grad_norm": 3.9851417541503906, + "learning_rate": 9.9912229991223e-06, + "loss": 0.2006, + "step": 8782 + }, + { + "epoch": 4.083217108321711, + "grad_norm": 5.284991264343262, + "learning_rate": 9.9912219991222e-06, + "loss": 0.3897, + "step": 8783 + }, + { + "epoch": 4.083682008368201, + "grad_norm": 3.3888626098632812, + "learning_rate": 9.9912209991221e-06, + "loss": 0.1887, + "step": 8784 + }, + { + "epoch": 4.084146908414691, + "grad_norm": 7.645252227783203, + "learning_rate": 9.991219999122001e-06, + "loss": 0.3752, + "step": 8785 + }, + { + "epoch": 4.08461180846118, + "grad_norm": 2.601284980773926, + "learning_rate": 9.991218999121902e-06, + "loss": 0.124, + "step": 8786 + }, + { + "epoch": 4.085076708507671, + "grad_norm": 4.9229512214660645, + "learning_rate": 9.9912179991218e-06, + "loss": 0.3119, + "step": 8787 + }, + { + "epoch": 4.085541608554161, + "grad_norm": 5.074150562286377, + "learning_rate": 9.991216999121701e-06, + "loss": 0.2765, + "step": 8788 + }, + { + "epoch": 4.086006508600651, + "grad_norm": 4.149901390075684, + "learning_rate": 9.9912159991216e-06, + "loss": 0.2837, + "step": 8789 + }, + { + "epoch": 4.0864714086471405, + "grad_norm": 6.546433925628662, + "learning_rate": 9.991214999121501e-06, + "loss": 0.2924, + "step": 8790 + }, + { + "epoch": 4.086936308693631, + "grad_norm": 7.0956645011901855, + "learning_rate": 9.9912139991214e-06, + "loss": 0.2673, + "step": 8791 + }, + { + "epoch": 4.087401208740121, + "grad_norm": 4.049623489379883, + "learning_rate": 9.9912129991213e-06, + "loss": 0.2844, + "step": 8792 + }, + { + "epoch": 4.087866108786611, + "grad_norm": 9.457386016845703, + "learning_rate": 9.991211999121201e-06, + "loss": 0.7284, + "step": 8793 + }, + { + "epoch": 4.088331008833101, + "grad_norm": 4.123287200927734, + "learning_rate": 9.9912109991211e-06, + "loss": 0.1135, + "step": 8794 + }, + { + "epoch": 4.088795908879591, + "grad_norm": 6.534922122955322, + "learning_rate": 9.991209999121001e-06, + "loss": 0.2821, + "step": 8795 + }, + { + "epoch": 4.089260808926081, + "grad_norm": 3.3568129539489746, + "learning_rate": 9.9912089991209e-06, + "loss": 0.114, + "step": 8796 + }, + { + "epoch": 4.089725708972571, + "grad_norm": 4.911664009094238, + "learning_rate": 9.9912079991208e-06, + "loss": 0.2591, + "step": 8797 + }, + { + "epoch": 4.090190609019061, + "grad_norm": 7.097163200378418, + "learning_rate": 9.991206999120701e-06, + "loss": 0.5933, + "step": 8798 + }, + { + "epoch": 4.0906555090655505, + "grad_norm": 3.4891180992126465, + "learning_rate": 9.991205999120602e-06, + "loss": 0.1642, + "step": 8799 + }, + { + "epoch": 4.091120409112041, + "grad_norm": 5.048712253570557, + "learning_rate": 9.9912049991205e-06, + "loss": 0.2907, + "step": 8800 + }, + { + "epoch": 4.091585309158531, + "grad_norm": 4.193267345428467, + "learning_rate": 9.9912039991204e-06, + "loss": 0.2857, + "step": 8801 + }, + { + "epoch": 4.092050209205021, + "grad_norm": 4.559300899505615, + "learning_rate": 9.9912029991203e-06, + "loss": 0.2149, + "step": 8802 + }, + { + "epoch": 4.092515109251511, + "grad_norm": 5.6075053215026855, + "learning_rate": 9.991201999120201e-06, + "loss": 0.2337, + "step": 8803 + }, + { + "epoch": 4.092980009298001, + "grad_norm": 3.158883571624756, + "learning_rate": 9.991200999120102e-06, + "loss": 0.1546, + "step": 8804 + }, + { + "epoch": 4.093444909344491, + "grad_norm": 3.0328941345214844, + "learning_rate": 9.99119999912e-06, + "loss": 0.1224, + "step": 8805 + }, + { + "epoch": 4.093909809390981, + "grad_norm": 4.668888568878174, + "learning_rate": 9.991198999119901e-06, + "loss": 0.2426, + "step": 8806 + }, + { + "epoch": 4.094374709437471, + "grad_norm": 5.474318504333496, + "learning_rate": 9.9911979991198e-06, + "loss": 0.2763, + "step": 8807 + }, + { + "epoch": 4.0948396094839605, + "grad_norm": 5.869375705718994, + "learning_rate": 9.991196999119701e-06, + "loss": 0.2167, + "step": 8808 + }, + { + "epoch": 4.095304509530451, + "grad_norm": 6.323030948638916, + "learning_rate": 9.9911959991196e-06, + "loss": 0.5097, + "step": 8809 + }, + { + "epoch": 4.095769409576941, + "grad_norm": 6.353596210479736, + "learning_rate": 9.9911949991195e-06, + "loss": 0.2732, + "step": 8810 + }, + { + "epoch": 4.096234309623431, + "grad_norm": 3.66831111907959, + "learning_rate": 9.991193999119401e-06, + "loss": 0.1361, + "step": 8811 + }, + { + "epoch": 4.096699209669921, + "grad_norm": 6.0280537605285645, + "learning_rate": 9.9911929991193e-06, + "loss": 0.147, + "step": 8812 + }, + { + "epoch": 4.097164109716411, + "grad_norm": 3.2730424404144287, + "learning_rate": 9.991191999119201e-06, + "loss": 0.1605, + "step": 8813 + }, + { + "epoch": 4.097629009762901, + "grad_norm": 9.030681610107422, + "learning_rate": 9.9911909991191e-06, + "loss": 0.3556, + "step": 8814 + }, + { + "epoch": 4.098093909809391, + "grad_norm": 4.86299991607666, + "learning_rate": 9.991189999119e-06, + "loss": 0.2806, + "step": 8815 + }, + { + "epoch": 4.098558809855881, + "grad_norm": 5.028418064117432, + "learning_rate": 9.991188999118901e-06, + "loss": 0.2059, + "step": 8816 + }, + { + "epoch": 4.099023709902371, + "grad_norm": 6.16750955581665, + "learning_rate": 9.991187999118802e-06, + "loss": 0.604, + "step": 8817 + }, + { + "epoch": 4.099488609948861, + "grad_norm": 5.964094638824463, + "learning_rate": 9.9911869991187e-06, + "loss": 0.2717, + "step": 8818 + }, + { + "epoch": 4.099953509995351, + "grad_norm": 8.65963077545166, + "learning_rate": 9.991185999118601e-06, + "loss": 0.1506, + "step": 8819 + }, + { + "epoch": 4.100418410041841, + "grad_norm": 6.257133960723877, + "learning_rate": 9.9911849991185e-06, + "loss": 0.4845, + "step": 8820 + }, + { + "epoch": 4.100883310088331, + "grad_norm": 3.7918412685394287, + "learning_rate": 9.991183999118401e-06, + "loss": 0.1939, + "step": 8821 + }, + { + "epoch": 4.101348210134821, + "grad_norm": 7.833853244781494, + "learning_rate": 9.9911829991183e-06, + "loss": 0.6141, + "step": 8822 + }, + { + "epoch": 4.101813110181311, + "grad_norm": 6.6111626625061035, + "learning_rate": 9.9911819991182e-06, + "loss": 0.4323, + "step": 8823 + }, + { + "epoch": 4.102278010227801, + "grad_norm": 4.242384433746338, + "learning_rate": 9.991180999118101e-06, + "loss": 0.3077, + "step": 8824 + }, + { + "epoch": 4.102742910274291, + "grad_norm": 7.845676422119141, + "learning_rate": 9.991179999118e-06, + "loss": 0.8525, + "step": 8825 + }, + { + "epoch": 4.103207810320781, + "grad_norm": 4.602939605712891, + "learning_rate": 9.991178999117901e-06, + "loss": 0.2121, + "step": 8826 + }, + { + "epoch": 4.103672710367271, + "grad_norm": 5.852104663848877, + "learning_rate": 9.9911779991178e-06, + "loss": 0.2795, + "step": 8827 + }, + { + "epoch": 4.104137610413761, + "grad_norm": 3.3364455699920654, + "learning_rate": 9.9911769991177e-06, + "loss": 0.1358, + "step": 8828 + }, + { + "epoch": 4.104602510460251, + "grad_norm": 5.337404727935791, + "learning_rate": 9.991175999117601e-06, + "loss": 0.3531, + "step": 8829 + }, + { + "epoch": 4.105067410506741, + "grad_norm": 6.053722858428955, + "learning_rate": 9.991174999117502e-06, + "loss": 0.2774, + "step": 8830 + }, + { + "epoch": 4.105532310553231, + "grad_norm": 4.6524505615234375, + "learning_rate": 9.991173999117401e-06, + "loss": 0.3456, + "step": 8831 + }, + { + "epoch": 4.105997210599721, + "grad_norm": 4.670603275299072, + "learning_rate": 9.991172999117302e-06, + "loss": 0.2107, + "step": 8832 + }, + { + "epoch": 4.106462110646211, + "grad_norm": 4.72681188583374, + "learning_rate": 9.9911719991172e-06, + "loss": 0.3066, + "step": 8833 + }, + { + "epoch": 4.106927010692701, + "grad_norm": 5.9829511642456055, + "learning_rate": 9.9911709991171e-06, + "loss": 0.42, + "step": 8834 + }, + { + "epoch": 4.107391910739191, + "grad_norm": 5.394476890563965, + "learning_rate": 9.991169999117e-06, + "loss": 0.2105, + "step": 8835 + }, + { + "epoch": 4.107856810785681, + "grad_norm": 3.662372350692749, + "learning_rate": 9.9911689991169e-06, + "loss": 0.098, + "step": 8836 + }, + { + "epoch": 4.108321710832171, + "grad_norm": 5.0897088050842285, + "learning_rate": 9.991167999116801e-06, + "loss": 0.2862, + "step": 8837 + }, + { + "epoch": 4.108786610878661, + "grad_norm": 2.8145358562469482, + "learning_rate": 9.9911669991167e-06, + "loss": 0.1355, + "step": 8838 + }, + { + "epoch": 4.1092515109251515, + "grad_norm": 9.01266860961914, + "learning_rate": 9.991165999116601e-06, + "loss": 0.7505, + "step": 8839 + }, + { + "epoch": 4.109716410971641, + "grad_norm": 6.921653747558594, + "learning_rate": 9.9911649991165e-06, + "loss": 0.3947, + "step": 8840 + }, + { + "epoch": 4.110181311018131, + "grad_norm": 3.6825785636901855, + "learning_rate": 9.9911639991164e-06, + "loss": 0.098, + "step": 8841 + }, + { + "epoch": 4.110646211064621, + "grad_norm": 5.968442916870117, + "learning_rate": 9.991162999116301e-06, + "loss": 0.3925, + "step": 8842 + }, + { + "epoch": 4.111111111111111, + "grad_norm": 4.099778652191162, + "learning_rate": 9.991161999116202e-06, + "loss": 0.1736, + "step": 8843 + }, + { + "epoch": 4.111576011157601, + "grad_norm": 7.656275749206543, + "learning_rate": 9.991160999116101e-06, + "loss": 0.458, + "step": 8844 + }, + { + "epoch": 4.112040911204091, + "grad_norm": 6.037096977233887, + "learning_rate": 9.991159999116e-06, + "loss": 0.3868, + "step": 8845 + }, + { + "epoch": 4.112505811250581, + "grad_norm": 3.8739380836486816, + "learning_rate": 9.9911589991159e-06, + "loss": 0.1117, + "step": 8846 + }, + { + "epoch": 4.112970711297071, + "grad_norm": 4.4978346824646, + "learning_rate": 9.9911579991158e-06, + "loss": 0.2602, + "step": 8847 + }, + { + "epoch": 4.1134356113435615, + "grad_norm": 6.656520843505859, + "learning_rate": 9.9911569991157e-06, + "loss": 0.5867, + "step": 8848 + }, + { + "epoch": 4.113900511390051, + "grad_norm": 4.955191612243652, + "learning_rate": 9.9911559991156e-06, + "loss": 0.1793, + "step": 8849 + }, + { + "epoch": 4.114365411436541, + "grad_norm": 3.303095817565918, + "learning_rate": 9.991154999115501e-06, + "loss": 0.1508, + "step": 8850 + }, + { + "epoch": 4.114830311483031, + "grad_norm": 4.588221073150635, + "learning_rate": 9.9911539991154e-06, + "loss": 0.1911, + "step": 8851 + }, + { + "epoch": 4.115295211529521, + "grad_norm": 5.479280471801758, + "learning_rate": 9.991152999115301e-06, + "loss": 0.1326, + "step": 8852 + }, + { + "epoch": 4.115760111576011, + "grad_norm": 7.363409519195557, + "learning_rate": 9.9911519991152e-06, + "loss": 0.3801, + "step": 8853 + }, + { + "epoch": 4.116225011622501, + "grad_norm": 3.321371555328369, + "learning_rate": 9.9911509991151e-06, + "loss": 0.1667, + "step": 8854 + }, + { + "epoch": 4.116689911668991, + "grad_norm": 5.274712085723877, + "learning_rate": 9.991149999115001e-06, + "loss": 0.2472, + "step": 8855 + }, + { + "epoch": 4.117154811715481, + "grad_norm": 4.423489570617676, + "learning_rate": 9.9911489991149e-06, + "loss": 0.2407, + "step": 8856 + }, + { + "epoch": 4.1176197117619715, + "grad_norm": 4.815450668334961, + "learning_rate": 9.991147999114801e-06, + "loss": 0.218, + "step": 8857 + }, + { + "epoch": 4.118084611808461, + "grad_norm": 5.9585065841674805, + "learning_rate": 9.9911469991147e-06, + "loss": 0.1011, + "step": 8858 + }, + { + "epoch": 4.118549511854951, + "grad_norm": 3.7663233280181885, + "learning_rate": 9.9911459991146e-06, + "loss": 0.1477, + "step": 8859 + }, + { + "epoch": 4.119014411901441, + "grad_norm": 5.066274642944336, + "learning_rate": 9.9911449991145e-06, + "loss": 0.3568, + "step": 8860 + }, + { + "epoch": 4.119479311947932, + "grad_norm": 2.6969895362854004, + "learning_rate": 9.9911439991144e-06, + "loss": 0.0939, + "step": 8861 + }, + { + "epoch": 4.119944211994421, + "grad_norm": 6.569159984588623, + "learning_rate": 9.991142999114301e-06, + "loss": 0.291, + "step": 8862 + }, + { + "epoch": 4.120409112040911, + "grad_norm": 4.89948844909668, + "learning_rate": 9.991141999114202e-06, + "loss": 0.1928, + "step": 8863 + }, + { + "epoch": 4.120874012087401, + "grad_norm": 4.6510419845581055, + "learning_rate": 9.9911409991141e-06, + "loss": 0.3267, + "step": 8864 + }, + { + "epoch": 4.121338912133891, + "grad_norm": 6.333298683166504, + "learning_rate": 9.991139999114001e-06, + "loss": 0.5523, + "step": 8865 + }, + { + "epoch": 4.1218038121803815, + "grad_norm": 9.17754077911377, + "learning_rate": 9.9911389991139e-06, + "loss": 0.2829, + "step": 8866 + }, + { + "epoch": 4.122268712226871, + "grad_norm": 3.132869243621826, + "learning_rate": 9.9911379991138e-06, + "loss": 0.1434, + "step": 8867 + }, + { + "epoch": 4.122733612273361, + "grad_norm": 5.4712042808532715, + "learning_rate": 9.991136999113701e-06, + "loss": 0.2493, + "step": 8868 + }, + { + "epoch": 4.123198512319851, + "grad_norm": 1.8518601655960083, + "learning_rate": 9.9911359991136e-06, + "loss": 0.0421, + "step": 8869 + }, + { + "epoch": 4.123663412366342, + "grad_norm": 3.6210598945617676, + "learning_rate": 9.991134999113501e-06, + "loss": 0.0961, + "step": 8870 + }, + { + "epoch": 4.124128312412831, + "grad_norm": 4.720934867858887, + "learning_rate": 9.9911339991134e-06, + "loss": 0.3238, + "step": 8871 + }, + { + "epoch": 4.124593212459321, + "grad_norm": 7.584634304046631, + "learning_rate": 9.9911329991133e-06, + "loss": 0.4622, + "step": 8872 + }, + { + "epoch": 4.125058112505811, + "grad_norm": 5.586373329162598, + "learning_rate": 9.991131999113201e-06, + "loss": 0.289, + "step": 8873 + }, + { + "epoch": 4.125523012552302, + "grad_norm": 4.238238334655762, + "learning_rate": 9.9911309991131e-06, + "loss": 0.155, + "step": 8874 + }, + { + "epoch": 4.1259879125987915, + "grad_norm": 3.541146755218506, + "learning_rate": 9.991129999113001e-06, + "loss": 0.1694, + "step": 8875 + }, + { + "epoch": 4.126452812645281, + "grad_norm": 4.982387065887451, + "learning_rate": 9.991128999112902e-06, + "loss": 0.3243, + "step": 8876 + }, + { + "epoch": 4.126917712691771, + "grad_norm": 5.932204246520996, + "learning_rate": 9.9911279991128e-06, + "loss": 0.3091, + "step": 8877 + }, + { + "epoch": 4.127382612738261, + "grad_norm": 3.936114549636841, + "learning_rate": 9.9911269991127e-06, + "loss": 0.1139, + "step": 8878 + }, + { + "epoch": 4.127847512784752, + "grad_norm": 6.001041412353516, + "learning_rate": 9.9911259991126e-06, + "loss": 0.3516, + "step": 8879 + }, + { + "epoch": 4.128312412831241, + "grad_norm": 6.538389205932617, + "learning_rate": 9.991124999112501e-06, + "loss": 0.4568, + "step": 8880 + }, + { + "epoch": 4.128777312877731, + "grad_norm": 9.517478942871094, + "learning_rate": 9.991123999112402e-06, + "loss": 0.5645, + "step": 8881 + }, + { + "epoch": 4.129242212924221, + "grad_norm": 5.263302326202393, + "learning_rate": 9.9911229991123e-06, + "loss": 0.5028, + "step": 8882 + }, + { + "epoch": 4.129707112970712, + "grad_norm": 6.137864589691162, + "learning_rate": 9.991121999112201e-06, + "loss": 0.3047, + "step": 8883 + }, + { + "epoch": 4.1301720130172015, + "grad_norm": 5.871241092681885, + "learning_rate": 9.9911209991121e-06, + "loss": 0.3473, + "step": 8884 + }, + { + "epoch": 4.130636913063691, + "grad_norm": 6.858383655548096, + "learning_rate": 9.991119999112e-06, + "loss": 0.1856, + "step": 8885 + }, + { + "epoch": 4.131101813110181, + "grad_norm": 12.492576599121094, + "learning_rate": 9.991118999111901e-06, + "loss": 0.4893, + "step": 8886 + }, + { + "epoch": 4.131566713156671, + "grad_norm": 7.994753837585449, + "learning_rate": 9.9911179991118e-06, + "loss": 0.2812, + "step": 8887 + }, + { + "epoch": 4.132031613203162, + "grad_norm": 4.069161891937256, + "learning_rate": 9.991116999111701e-06, + "loss": 0.1989, + "step": 8888 + }, + { + "epoch": 4.132496513249651, + "grad_norm": 5.798177719116211, + "learning_rate": 9.9911159991116e-06, + "loss": 0.4213, + "step": 8889 + }, + { + "epoch": 4.132961413296141, + "grad_norm": 5.374795436859131, + "learning_rate": 9.9911149991115e-06, + "loss": 0.3144, + "step": 8890 + }, + { + "epoch": 4.133426313342631, + "grad_norm": 3.911576747894287, + "learning_rate": 9.9911139991114e-06, + "loss": 0.1391, + "step": 8891 + }, + { + "epoch": 4.133891213389122, + "grad_norm": 5.401784896850586, + "learning_rate": 9.9911129991113e-06, + "loss": 0.2956, + "step": 8892 + }, + { + "epoch": 4.1343561134356115, + "grad_norm": 6.497187614440918, + "learning_rate": 9.991111999111201e-06, + "loss": 0.4214, + "step": 8893 + }, + { + "epoch": 4.134821013482101, + "grad_norm": 7.035581111907959, + "learning_rate": 9.991110999111102e-06, + "loss": 0.5309, + "step": 8894 + }, + { + "epoch": 4.135285913528591, + "grad_norm": 4.720493316650391, + "learning_rate": 9.991109999111e-06, + "loss": 0.2884, + "step": 8895 + }, + { + "epoch": 4.135750813575082, + "grad_norm": 5.1056413650512695, + "learning_rate": 9.991108999110901e-06, + "loss": 0.3335, + "step": 8896 + }, + { + "epoch": 4.136215713621572, + "grad_norm": 3.7777562141418457, + "learning_rate": 9.9911079991108e-06, + "loss": 0.112, + "step": 8897 + }, + { + "epoch": 4.136680613668061, + "grad_norm": 4.788946628570557, + "learning_rate": 9.991106999110701e-06, + "loss": 0.2693, + "step": 8898 + }, + { + "epoch": 4.137145513714551, + "grad_norm": 3.116621494293213, + "learning_rate": 9.991105999110602e-06, + "loss": 0.176, + "step": 8899 + }, + { + "epoch": 4.137610413761041, + "grad_norm": 4.7862467765808105, + "learning_rate": 9.9911049991105e-06, + "loss": 0.4028, + "step": 8900 + }, + { + "epoch": 4.138075313807532, + "grad_norm": 6.449438095092773, + "learning_rate": 9.991103999110401e-06, + "loss": 0.2521, + "step": 8901 + }, + { + "epoch": 4.1385402138540215, + "grad_norm": 6.887516021728516, + "learning_rate": 9.9911029991103e-06, + "loss": 0.2968, + "step": 8902 + }, + { + "epoch": 4.139005113900511, + "grad_norm": 3.6100895404815674, + "learning_rate": 9.9911019991102e-06, + "loss": 0.1563, + "step": 8903 + }, + { + "epoch": 4.139470013947001, + "grad_norm": 6.632273197174072, + "learning_rate": 9.9911009991101e-06, + "loss": 0.2892, + "step": 8904 + }, + { + "epoch": 4.139934913993492, + "grad_norm": 4.8746514320373535, + "learning_rate": 9.99109999911e-06, + "loss": 0.3362, + "step": 8905 + }, + { + "epoch": 4.140399814039982, + "grad_norm": 4.088988304138184, + "learning_rate": 9.991098999109901e-06, + "loss": 0.1985, + "step": 8906 + }, + { + "epoch": 4.140864714086471, + "grad_norm": 5.267740249633789, + "learning_rate": 9.991097999109802e-06, + "loss": 0.2943, + "step": 8907 + }, + { + "epoch": 4.141329614132961, + "grad_norm": 8.782347679138184, + "learning_rate": 9.9910969991097e-06, + "loss": 0.2469, + "step": 8908 + }, + { + "epoch": 4.141794514179451, + "grad_norm": 5.319448471069336, + "learning_rate": 9.991095999109601e-06, + "loss": 0.1165, + "step": 8909 + }, + { + "epoch": 4.142259414225942, + "grad_norm": 5.415376663208008, + "learning_rate": 9.9910949991095e-06, + "loss": 0.4549, + "step": 8910 + }, + { + "epoch": 4.1427243142724315, + "grad_norm": 5.96784782409668, + "learning_rate": 9.991093999109401e-06, + "loss": 0.3861, + "step": 8911 + }, + { + "epoch": 4.143189214318921, + "grad_norm": 5.742704391479492, + "learning_rate": 9.991092999109302e-06, + "loss": 0.2476, + "step": 8912 + }, + { + "epoch": 4.143654114365411, + "grad_norm": 5.654045104980469, + "learning_rate": 9.9910919991092e-06, + "loss": 0.2497, + "step": 8913 + }, + { + "epoch": 4.144119014411902, + "grad_norm": 6.224580764770508, + "learning_rate": 9.991090999109101e-06, + "loss": 0.4735, + "step": 8914 + }, + { + "epoch": 4.144583914458392, + "grad_norm": 5.517715930938721, + "learning_rate": 9.991089999109e-06, + "loss": 0.2971, + "step": 8915 + }, + { + "epoch": 4.145048814504881, + "grad_norm": 3.3827223777770996, + "learning_rate": 9.991088999108901e-06, + "loss": 0.1604, + "step": 8916 + }, + { + "epoch": 4.145513714551371, + "grad_norm": 6.000076770782471, + "learning_rate": 9.9910879991088e-06, + "loss": 0.2737, + "step": 8917 + }, + { + "epoch": 4.145978614597862, + "grad_norm": 4.994681358337402, + "learning_rate": 9.9910869991087e-06, + "loss": 0.1512, + "step": 8918 + }, + { + "epoch": 4.146443514644352, + "grad_norm": 7.949263095855713, + "learning_rate": 9.991085999108601e-06, + "loss": 0.3927, + "step": 8919 + }, + { + "epoch": 4.1469084146908415, + "grad_norm": 5.698633670806885, + "learning_rate": 9.991084999108502e-06, + "loss": 0.3563, + "step": 8920 + }, + { + "epoch": 4.147373314737331, + "grad_norm": 6.220924377441406, + "learning_rate": 9.9910839991084e-06, + "loss": 0.3166, + "step": 8921 + }, + { + "epoch": 4.147838214783821, + "grad_norm": 4.510117530822754, + "learning_rate": 9.9910829991083e-06, + "loss": 0.2023, + "step": 8922 + }, + { + "epoch": 4.148303114830312, + "grad_norm": 8.472702026367188, + "learning_rate": 9.9910819991082e-06, + "loss": 0.2865, + "step": 8923 + }, + { + "epoch": 4.148768014876802, + "grad_norm": 3.8774220943450928, + "learning_rate": 9.991080999108101e-06, + "loss": 0.2166, + "step": 8924 + }, + { + "epoch": 4.149232914923291, + "grad_norm": 6.401660442352295, + "learning_rate": 9.991079999108002e-06, + "loss": 0.1949, + "step": 8925 + }, + { + "epoch": 4.149697814969781, + "grad_norm": 6.241265296936035, + "learning_rate": 9.9910789991079e-06, + "loss": 0.3067, + "step": 8926 + }, + { + "epoch": 4.150162715016272, + "grad_norm": 4.2010908126831055, + "learning_rate": 9.991077999107801e-06, + "loss": 0.2833, + "step": 8927 + }, + { + "epoch": 4.150627615062762, + "grad_norm": 4.502042770385742, + "learning_rate": 9.9910769991077e-06, + "loss": 0.1977, + "step": 8928 + }, + { + "epoch": 4.1510925151092515, + "grad_norm": 4.660582065582275, + "learning_rate": 9.991075999107601e-06, + "loss": 0.2128, + "step": 8929 + }, + { + "epoch": 4.151557415155741, + "grad_norm": 5.246844291687012, + "learning_rate": 9.9910749991075e-06, + "loss": 0.3581, + "step": 8930 + }, + { + "epoch": 4.152022315202231, + "grad_norm": 5.824151515960693, + "learning_rate": 9.9910739991074e-06, + "loss": 0.1834, + "step": 8931 + }, + { + "epoch": 4.152487215248722, + "grad_norm": 3.859910011291504, + "learning_rate": 9.991072999107301e-06, + "loss": 0.1699, + "step": 8932 + }, + { + "epoch": 4.152952115295212, + "grad_norm": 6.29049825668335, + "learning_rate": 9.9910719991072e-06, + "loss": 0.3951, + "step": 8933 + }, + { + "epoch": 4.153417015341701, + "grad_norm": 5.513393402099609, + "learning_rate": 9.9910709991071e-06, + "loss": 0.3708, + "step": 8934 + }, + { + "epoch": 4.153881915388191, + "grad_norm": 4.990880489349365, + "learning_rate": 9.991069999107e-06, + "loss": 0.2508, + "step": 8935 + }, + { + "epoch": 4.154346815434682, + "grad_norm": 4.5450520515441895, + "learning_rate": 9.9910689991069e-06, + "loss": 0.1858, + "step": 8936 + }, + { + "epoch": 4.154811715481172, + "grad_norm": 6.152183532714844, + "learning_rate": 9.991067999106801e-06, + "loss": 0.2929, + "step": 8937 + }, + { + "epoch": 4.1552766155276615, + "grad_norm": 4.445064067840576, + "learning_rate": 9.991066999106702e-06, + "loss": 0.2787, + "step": 8938 + }, + { + "epoch": 4.155741515574151, + "grad_norm": 5.19802713394165, + "learning_rate": 9.9910659991066e-06, + "loss": 0.3232, + "step": 8939 + }, + { + "epoch": 4.156206415620642, + "grad_norm": 6.597072601318359, + "learning_rate": 9.991064999106501e-06, + "loss": 0.4684, + "step": 8940 + }, + { + "epoch": 4.156671315667132, + "grad_norm": 4.788276672363281, + "learning_rate": 9.9910639991064e-06, + "loss": 0.2185, + "step": 8941 + }, + { + "epoch": 4.157136215713622, + "grad_norm": 3.4085400104522705, + "learning_rate": 9.991062999106301e-06, + "loss": 0.1115, + "step": 8942 + }, + { + "epoch": 4.157601115760111, + "grad_norm": 3.9234964847564697, + "learning_rate": 9.9910619991062e-06, + "loss": 0.2294, + "step": 8943 + }, + { + "epoch": 4.158066015806601, + "grad_norm": 14.798604965209961, + "learning_rate": 9.9910609991061e-06, + "loss": 0.2962, + "step": 8944 + }, + { + "epoch": 4.158530915853092, + "grad_norm": 5.9229254722595215, + "learning_rate": 9.991059999106001e-06, + "loss": 0.2325, + "step": 8945 + }, + { + "epoch": 4.158995815899582, + "grad_norm": 4.300853729248047, + "learning_rate": 9.9910589991059e-06, + "loss": 0.1926, + "step": 8946 + }, + { + "epoch": 4.1594607159460715, + "grad_norm": 9.15042781829834, + "learning_rate": 9.991057999105801e-06, + "loss": 0.6327, + "step": 8947 + }, + { + "epoch": 4.159925615992561, + "grad_norm": 9.906756401062012, + "learning_rate": 9.9910569991057e-06, + "loss": 0.4706, + "step": 8948 + }, + { + "epoch": 4.160390516039052, + "grad_norm": 4.741153240203857, + "learning_rate": 9.9910559991056e-06, + "loss": 0.083, + "step": 8949 + }, + { + "epoch": 4.160855416085542, + "grad_norm": 4.206657886505127, + "learning_rate": 9.991054999105501e-06, + "loss": 0.167, + "step": 8950 + }, + { + "epoch": 4.161320316132032, + "grad_norm": 8.398709297180176, + "learning_rate": 9.991053999105402e-06, + "loss": 0.8059, + "step": 8951 + }, + { + "epoch": 4.161785216178521, + "grad_norm": 3.6398420333862305, + "learning_rate": 9.9910529991053e-06, + "loss": 0.2175, + "step": 8952 + }, + { + "epoch": 4.162250116225012, + "grad_norm": 8.980164527893066, + "learning_rate": 9.9910519991052e-06, + "loss": 0.2972, + "step": 8953 + }, + { + "epoch": 4.162715016271502, + "grad_norm": 4.660053253173828, + "learning_rate": 9.9910509991051e-06, + "loss": 0.2102, + "step": 8954 + }, + { + "epoch": 4.163179916317992, + "grad_norm": 4.3213300704956055, + "learning_rate": 9.991049999105001e-06, + "loss": 0.1296, + "step": 8955 + }, + { + "epoch": 4.1636448163644815, + "grad_norm": 5.103208541870117, + "learning_rate": 9.9910489991049e-06, + "loss": 0.4425, + "step": 8956 + }, + { + "epoch": 4.164109716410971, + "grad_norm": 5.454392433166504, + "learning_rate": 9.9910479991048e-06, + "loss": 0.174, + "step": 8957 + }, + { + "epoch": 4.164574616457462, + "grad_norm": 5.552009582519531, + "learning_rate": 9.991046999104701e-06, + "loss": 0.3146, + "step": 8958 + }, + { + "epoch": 4.165039516503952, + "grad_norm": 4.358377933502197, + "learning_rate": 9.9910459991046e-06, + "loss": 0.2324, + "step": 8959 + }, + { + "epoch": 4.165504416550442, + "grad_norm": 4.985982894897461, + "learning_rate": 9.991044999104501e-06, + "loss": 0.3435, + "step": 8960 + }, + { + "epoch": 4.165969316596931, + "grad_norm": 7.018010139465332, + "learning_rate": 9.9910439991044e-06, + "loss": 0.3994, + "step": 8961 + }, + { + "epoch": 4.166434216643422, + "grad_norm": 5.7416253089904785, + "learning_rate": 9.9910429991043e-06, + "loss": 0.546, + "step": 8962 + }, + { + "epoch": 4.166899116689912, + "grad_norm": 4.415569305419922, + "learning_rate": 9.991041999104201e-06, + "loss": 0.2611, + "step": 8963 + }, + { + "epoch": 4.167364016736402, + "grad_norm": 7.8724751472473145, + "learning_rate": 9.9910409991041e-06, + "loss": 0.3635, + "step": 8964 + }, + { + "epoch": 4.1678289167828915, + "grad_norm": 4.970385551452637, + "learning_rate": 9.991039999104001e-06, + "loss": 0.1401, + "step": 8965 + }, + { + "epoch": 4.168293816829381, + "grad_norm": 4.441365718841553, + "learning_rate": 9.9910389991039e-06, + "loss": 0.2348, + "step": 8966 + }, + { + "epoch": 4.168758716875872, + "grad_norm": 5.775564193725586, + "learning_rate": 9.9910379991038e-06, + "loss": 0.3203, + "step": 8967 + }, + { + "epoch": 4.169223616922362, + "grad_norm": 4.876204013824463, + "learning_rate": 9.991036999103701e-06, + "loss": 0.1849, + "step": 8968 + }, + { + "epoch": 4.169688516968852, + "grad_norm": 6.13627815246582, + "learning_rate": 9.991035999103602e-06, + "loss": 0.3572, + "step": 8969 + }, + { + "epoch": 4.170153417015341, + "grad_norm": 2.409457206726074, + "learning_rate": 9.9910349991035e-06, + "loss": 0.022, + "step": 8970 + }, + { + "epoch": 4.170618317061832, + "grad_norm": 5.174119472503662, + "learning_rate": 9.991033999103401e-06, + "loss": 0.3126, + "step": 8971 + }, + { + "epoch": 4.171083217108322, + "grad_norm": 3.625722885131836, + "learning_rate": 9.9910329991033e-06, + "loss": 0.1455, + "step": 8972 + }, + { + "epoch": 4.171548117154812, + "grad_norm": 4.205105304718018, + "learning_rate": 9.991031999103201e-06, + "loss": 0.2402, + "step": 8973 + }, + { + "epoch": 4.1720130172013015, + "grad_norm": 6.3815789222717285, + "learning_rate": 9.9910309991031e-06, + "loss": 0.2763, + "step": 8974 + }, + { + "epoch": 4.172477917247791, + "grad_norm": 3.3168530464172363, + "learning_rate": 9.991029999103e-06, + "loss": 0.1299, + "step": 8975 + }, + { + "epoch": 4.172942817294282, + "grad_norm": 5.67842960357666, + "learning_rate": 9.991028999102901e-06, + "loss": 0.4145, + "step": 8976 + }, + { + "epoch": 4.173407717340772, + "grad_norm": 7.400556564331055, + "learning_rate": 9.9910279991028e-06, + "loss": 0.3254, + "step": 8977 + }, + { + "epoch": 4.173872617387262, + "grad_norm": 4.291192531585693, + "learning_rate": 9.991026999102701e-06, + "loss": 0.2224, + "step": 8978 + }, + { + "epoch": 4.174337517433751, + "grad_norm": 3.816438913345337, + "learning_rate": 9.9910259991026e-06, + "loss": 0.2408, + "step": 8979 + }, + { + "epoch": 4.174802417480242, + "grad_norm": 4.6216020584106445, + "learning_rate": 9.9910249991025e-06, + "loss": 0.2478, + "step": 8980 + }, + { + "epoch": 4.175267317526732, + "grad_norm": 6.4333038330078125, + "learning_rate": 9.991023999102401e-06, + "loss": 0.3052, + "step": 8981 + }, + { + "epoch": 4.175732217573222, + "grad_norm": 5.129361629486084, + "learning_rate": 9.991022999102302e-06, + "loss": 0.3979, + "step": 8982 + }, + { + "epoch": 4.1761971176197115, + "grad_norm": 4.206441402435303, + "learning_rate": 9.991021999102201e-06, + "loss": 0.2261, + "step": 8983 + }, + { + "epoch": 4.176662017666202, + "grad_norm": 4.3597869873046875, + "learning_rate": 9.991020999102102e-06, + "loss": 0.2572, + "step": 8984 + }, + { + "epoch": 4.177126917712692, + "grad_norm": 4.755560874938965, + "learning_rate": 9.991019999102e-06, + "loss": 0.2106, + "step": 8985 + }, + { + "epoch": 4.177591817759182, + "grad_norm": 4.5581183433532715, + "learning_rate": 9.9910189991019e-06, + "loss": 0.2814, + "step": 8986 + }, + { + "epoch": 4.178056717805672, + "grad_norm": 4.982365131378174, + "learning_rate": 9.9910179991018e-06, + "loss": 0.2037, + "step": 8987 + }, + { + "epoch": 4.178521617852161, + "grad_norm": 5.159357070922852, + "learning_rate": 9.9910169991017e-06, + "loss": 0.3452, + "step": 8988 + }, + { + "epoch": 4.178986517898652, + "grad_norm": 3.3193702697753906, + "learning_rate": 9.991015999101601e-06, + "loss": 0.1702, + "step": 8989 + }, + { + "epoch": 4.179451417945142, + "grad_norm": 4.312126159667969, + "learning_rate": 9.9910149991015e-06, + "loss": 0.2092, + "step": 8990 + }, + { + "epoch": 4.179916317991632, + "grad_norm": 3.923757314682007, + "learning_rate": 9.991013999101401e-06, + "loss": 0.1933, + "step": 8991 + }, + { + "epoch": 4.1803812180381215, + "grad_norm": 3.8278613090515137, + "learning_rate": 9.9910129991013e-06, + "loss": 0.1731, + "step": 8992 + }, + { + "epoch": 4.180846118084612, + "grad_norm": 5.032049179077148, + "learning_rate": 9.9910119991012e-06, + "loss": 0.197, + "step": 8993 + }, + { + "epoch": 4.181311018131102, + "grad_norm": 5.232093811035156, + "learning_rate": 9.991010999101101e-06, + "loss": 0.1677, + "step": 8994 + }, + { + "epoch": 4.181775918177592, + "grad_norm": 20.766948699951172, + "learning_rate": 9.991009999101002e-06, + "loss": 0.3967, + "step": 8995 + }, + { + "epoch": 4.182240818224082, + "grad_norm": 9.027753829956055, + "learning_rate": 9.991008999100901e-06, + "loss": 0.1914, + "step": 8996 + }, + { + "epoch": 4.182705718270572, + "grad_norm": 2.866612672805786, + "learning_rate": 9.9910079991008e-06, + "loss": 0.0851, + "step": 8997 + }, + { + "epoch": 4.183170618317062, + "grad_norm": 5.4446187019348145, + "learning_rate": 9.9910069991007e-06, + "loss": 0.3322, + "step": 8998 + }, + { + "epoch": 4.183635518363552, + "grad_norm": 6.723505020141602, + "learning_rate": 9.9910059991006e-06, + "loss": 0.2658, + "step": 8999 + }, + { + "epoch": 4.184100418410042, + "grad_norm": 13.74229621887207, + "learning_rate": 9.9910049991005e-06, + "loss": 0.2765, + "step": 9000 + }, + { + "epoch": 4.1845653184565315, + "grad_norm": 6.2287917137146, + "learning_rate": 9.991003999100401e-06, + "loss": 0.364, + "step": 9001 + }, + { + "epoch": 4.185030218503022, + "grad_norm": 6.4185686111450195, + "learning_rate": 9.991002999100302e-06, + "loss": 0.2624, + "step": 9002 + }, + { + "epoch": 4.185495118549512, + "grad_norm": 7.251617431640625, + "learning_rate": 9.9910019991002e-06, + "loss": 0.2574, + "step": 9003 + }, + { + "epoch": 4.185960018596002, + "grad_norm": 6.6324968338012695, + "learning_rate": 9.991000999100101e-06, + "loss": 0.1613, + "step": 9004 + }, + { + "epoch": 4.1864249186424916, + "grad_norm": 6.436995506286621, + "learning_rate": 9.9909999991e-06, + "loss": 0.2683, + "step": 9005 + }, + { + "epoch": 4.186889818688982, + "grad_norm": 6.929717063903809, + "learning_rate": 9.9909989990999e-06, + "loss": 0.5643, + "step": 9006 + }, + { + "epoch": 4.187354718735472, + "grad_norm": 5.3065571784973145, + "learning_rate": 9.990997999099801e-06, + "loss": 0.2216, + "step": 9007 + }, + { + "epoch": 4.187819618781962, + "grad_norm": 6.022818565368652, + "learning_rate": 9.9909969990997e-06, + "loss": 0.4914, + "step": 9008 + }, + { + "epoch": 4.188284518828452, + "grad_norm": 5.944896697998047, + "learning_rate": 9.990995999099601e-06, + "loss": 0.4091, + "step": 9009 + }, + { + "epoch": 4.1887494188749415, + "grad_norm": 7.577328205108643, + "learning_rate": 9.9909949990995e-06, + "loss": 0.2491, + "step": 9010 + }, + { + "epoch": 4.189214318921432, + "grad_norm": 7.082871913909912, + "learning_rate": 9.9909939990994e-06, + "loss": 0.2262, + "step": 9011 + }, + { + "epoch": 4.189679218967922, + "grad_norm": 4.323507785797119, + "learning_rate": 9.9909929990993e-06, + "loss": 0.246, + "step": 9012 + }, + { + "epoch": 4.190144119014412, + "grad_norm": 4.677602291107178, + "learning_rate": 9.9909919990992e-06, + "loss": 0.2482, + "step": 9013 + }, + { + "epoch": 4.1906090190609016, + "grad_norm": 6.2854323387146, + "learning_rate": 9.990990999099101e-06, + "loss": 0.5409, + "step": 9014 + }, + { + "epoch": 4.191073919107392, + "grad_norm": 9.16693115234375, + "learning_rate": 9.990989999099002e-06, + "loss": 0.4877, + "step": 9015 + }, + { + "epoch": 4.191538819153882, + "grad_norm": 4.119822978973389, + "learning_rate": 9.9909889990989e-06, + "loss": 0.1923, + "step": 9016 + }, + { + "epoch": 4.192003719200372, + "grad_norm": 6.332491874694824, + "learning_rate": 9.990987999098801e-06, + "loss": 0.3867, + "step": 9017 + }, + { + "epoch": 4.192468619246862, + "grad_norm": 4.066377639770508, + "learning_rate": 9.9909869990987e-06, + "loss": 0.2135, + "step": 9018 + }, + { + "epoch": 4.192933519293352, + "grad_norm": 7.526035308837891, + "learning_rate": 9.9909859990986e-06, + "loss": 0.4218, + "step": 9019 + }, + { + "epoch": 4.193398419339842, + "grad_norm": 6.471800327301025, + "learning_rate": 9.990984999098502e-06, + "loss": 0.3376, + "step": 9020 + }, + { + "epoch": 4.193863319386332, + "grad_norm": 6.051138401031494, + "learning_rate": 9.9909839990984e-06, + "loss": 0.2641, + "step": 9021 + }, + { + "epoch": 4.194328219432822, + "grad_norm": 4.884809970855713, + "learning_rate": 9.990982999098301e-06, + "loss": 0.3895, + "step": 9022 + }, + { + "epoch": 4.1947931194793115, + "grad_norm": 6.79915714263916, + "learning_rate": 9.9909819990982e-06, + "loss": 0.6021, + "step": 9023 + }, + { + "epoch": 4.195258019525802, + "grad_norm": 4.265320777893066, + "learning_rate": 9.9909809990981e-06, + "loss": 0.3162, + "step": 9024 + }, + { + "epoch": 4.195722919572292, + "grad_norm": 4.055388927459717, + "learning_rate": 9.990979999098e-06, + "loss": 0.1995, + "step": 9025 + }, + { + "epoch": 4.196187819618782, + "grad_norm": 5.270259380340576, + "learning_rate": 9.9909789990979e-06, + "loss": 0.1971, + "step": 9026 + }, + { + "epoch": 4.196652719665272, + "grad_norm": 4.777965545654297, + "learning_rate": 9.990977999097801e-06, + "loss": 0.257, + "step": 9027 + }, + { + "epoch": 4.197117619711762, + "grad_norm": 5.851052761077881, + "learning_rate": 9.990976999097702e-06, + "loss": 0.3386, + "step": 9028 + }, + { + "epoch": 4.197582519758252, + "grad_norm": 5.540870189666748, + "learning_rate": 9.9909759990976e-06, + "loss": 0.3112, + "step": 9029 + }, + { + "epoch": 4.198047419804742, + "grad_norm": 11.217273712158203, + "learning_rate": 9.9909749990975e-06, + "loss": 0.4501, + "step": 9030 + }, + { + "epoch": 4.198512319851232, + "grad_norm": 3.557976245880127, + "learning_rate": 9.9909739990974e-06, + "loss": 0.1807, + "step": 9031 + }, + { + "epoch": 4.198977219897722, + "grad_norm": 4.819978713989258, + "learning_rate": 9.990972999097301e-06, + "loss": 0.2161, + "step": 9032 + }, + { + "epoch": 4.199442119944212, + "grad_norm": 9.464043617248535, + "learning_rate": 9.990971999097202e-06, + "loss": 0.4111, + "step": 9033 + }, + { + "epoch": 4.199907019990702, + "grad_norm": 4.79041051864624, + "learning_rate": 9.9909709990971e-06, + "loss": 0.2652, + "step": 9034 + }, + { + "epoch": 4.200371920037192, + "grad_norm": 3.6280055046081543, + "learning_rate": 9.990969999097001e-06, + "loss": 0.1214, + "step": 9035 + }, + { + "epoch": 4.200836820083682, + "grad_norm": 5.3487548828125, + "learning_rate": 9.9909689990969e-06, + "loss": 0.3843, + "step": 9036 + }, + { + "epoch": 4.201301720130172, + "grad_norm": 4.494670867919922, + "learning_rate": 9.9909679990968e-06, + "loss": 0.2261, + "step": 9037 + }, + { + "epoch": 4.201766620176662, + "grad_norm": 5.299595355987549, + "learning_rate": 9.990966999096701e-06, + "loss": 0.2325, + "step": 9038 + }, + { + "epoch": 4.202231520223152, + "grad_norm": 5.454095840454102, + "learning_rate": 9.9909659990966e-06, + "loss": 0.2504, + "step": 9039 + }, + { + "epoch": 4.202696420269642, + "grad_norm": 4.057345390319824, + "learning_rate": 9.990964999096501e-06, + "loss": 0.1684, + "step": 9040 + }, + { + "epoch": 4.203161320316132, + "grad_norm": 6.164270877838135, + "learning_rate": 9.9909639990964e-06, + "loss": 0.3177, + "step": 9041 + }, + { + "epoch": 4.203626220362622, + "grad_norm": 5.558006286621094, + "learning_rate": 9.9909629990963e-06, + "loss": 0.2607, + "step": 9042 + }, + { + "epoch": 4.204091120409112, + "grad_norm": 5.724677562713623, + "learning_rate": 9.9909619990962e-06, + "loss": 0.3358, + "step": 9043 + }, + { + "epoch": 4.204556020455602, + "grad_norm": 5.6917314529418945, + "learning_rate": 9.9909609990961e-06, + "loss": 0.2368, + "step": 9044 + }, + { + "epoch": 4.205020920502092, + "grad_norm": 10.058784484863281, + "learning_rate": 9.990959999096001e-06, + "loss": 0.6958, + "step": 9045 + }, + { + "epoch": 4.205485820548582, + "grad_norm": 7.632631301879883, + "learning_rate": 9.990958999095902e-06, + "loss": 0.3809, + "step": 9046 + }, + { + "epoch": 4.205950720595072, + "grad_norm": 3.972992420196533, + "learning_rate": 9.9909579990958e-06, + "loss": 0.1388, + "step": 9047 + }, + { + "epoch": 4.206415620641562, + "grad_norm": 8.376639366149902, + "learning_rate": 9.990956999095701e-06, + "loss": 0.4181, + "step": 9048 + }, + { + "epoch": 4.206880520688052, + "grad_norm": 6.041959762573242, + "learning_rate": 9.9909559990956e-06, + "loss": 0.3713, + "step": 9049 + }, + { + "epoch": 4.207345420734542, + "grad_norm": 9.015778541564941, + "learning_rate": 9.990954999095501e-06, + "loss": 0.5036, + "step": 9050 + }, + { + "epoch": 4.207810320781032, + "grad_norm": 3.043288469314575, + "learning_rate": 9.990953999095402e-06, + "loss": 0.1608, + "step": 9051 + }, + { + "epoch": 4.208275220827522, + "grad_norm": 5.752023696899414, + "learning_rate": 9.9909529990953e-06, + "loss": 0.4575, + "step": 9052 + }, + { + "epoch": 4.208740120874012, + "grad_norm": 4.462798118591309, + "learning_rate": 9.990951999095201e-06, + "loss": 0.2484, + "step": 9053 + }, + { + "epoch": 4.209205020920502, + "grad_norm": 4.700591087341309, + "learning_rate": 9.9909509990951e-06, + "loss": 0.163, + "step": 9054 + }, + { + "epoch": 4.209669920966992, + "grad_norm": 7.532468318939209, + "learning_rate": 9.990949999095e-06, + "loss": 0.4169, + "step": 9055 + }, + { + "epoch": 4.210134821013482, + "grad_norm": 5.053748607635498, + "learning_rate": 9.9909489990949e-06, + "loss": 0.3182, + "step": 9056 + }, + { + "epoch": 4.210599721059972, + "grad_norm": 5.757434844970703, + "learning_rate": 9.9909479990948e-06, + "loss": 0.4231, + "step": 9057 + }, + { + "epoch": 4.211064621106462, + "grad_norm": 3.1254231929779053, + "learning_rate": 9.990946999094701e-06, + "loss": 0.0842, + "step": 9058 + }, + { + "epoch": 4.211529521152952, + "grad_norm": 8.687654495239258, + "learning_rate": 9.990945999094602e-06, + "loss": 0.4025, + "step": 9059 + }, + { + "epoch": 4.211994421199442, + "grad_norm": 5.709409236907959, + "learning_rate": 9.9909449990945e-06, + "loss": 0.2768, + "step": 9060 + }, + { + "epoch": 4.212459321245932, + "grad_norm": 7.9854888916015625, + "learning_rate": 9.990943999094401e-06, + "loss": 0.1778, + "step": 9061 + }, + { + "epoch": 4.212924221292422, + "grad_norm": 5.062191963195801, + "learning_rate": 9.9909429990943e-06, + "loss": 0.2387, + "step": 9062 + }, + { + "epoch": 4.2133891213389125, + "grad_norm": 7.304651260375977, + "learning_rate": 9.990941999094201e-06, + "loss": 0.3628, + "step": 9063 + }, + { + "epoch": 4.213854021385402, + "grad_norm": 3.488395929336548, + "learning_rate": 9.990940999094102e-06, + "loss": 0.1218, + "step": 9064 + }, + { + "epoch": 4.214318921431892, + "grad_norm": 5.6360859870910645, + "learning_rate": 9.990939999094e-06, + "loss": 0.3011, + "step": 9065 + }, + { + "epoch": 4.214783821478382, + "grad_norm": 6.087290287017822, + "learning_rate": 9.990938999093901e-06, + "loss": 0.2245, + "step": 9066 + }, + { + "epoch": 4.215248721524872, + "grad_norm": 4.708183288574219, + "learning_rate": 9.9909379990938e-06, + "loss": 0.1737, + "step": 9067 + }, + { + "epoch": 4.215713621571362, + "grad_norm": 3.6711204051971436, + "learning_rate": 9.990936999093701e-06, + "loss": 0.1525, + "step": 9068 + }, + { + "epoch": 4.216178521617852, + "grad_norm": 6.454403877258301, + "learning_rate": 9.9909359990936e-06, + "loss": 0.4374, + "step": 9069 + }, + { + "epoch": 4.216643421664342, + "grad_norm": 8.401972770690918, + "learning_rate": 9.9909349990935e-06, + "loss": 0.7305, + "step": 9070 + }, + { + "epoch": 4.217108321710832, + "grad_norm": 4.053170680999756, + "learning_rate": 9.990933999093401e-06, + "loss": 0.1076, + "step": 9071 + }, + { + "epoch": 4.2175732217573225, + "grad_norm": 5.499892711639404, + "learning_rate": 9.990932999093302e-06, + "loss": 0.2614, + "step": 9072 + }, + { + "epoch": 4.218038121803812, + "grad_norm": 5.171576499938965, + "learning_rate": 9.9909319990932e-06, + "loss": 0.4286, + "step": 9073 + }, + { + "epoch": 4.218503021850302, + "grad_norm": 6.178049564361572, + "learning_rate": 9.9909309990931e-06, + "loss": 0.3308, + "step": 9074 + }, + { + "epoch": 4.218967921896792, + "grad_norm": 5.45896053314209, + "learning_rate": 9.990929999093e-06, + "loss": 0.2504, + "step": 9075 + }, + { + "epoch": 4.219432821943283, + "grad_norm": 6.674102783203125, + "learning_rate": 9.990928999092901e-06, + "loss": 0.5378, + "step": 9076 + }, + { + "epoch": 4.219897721989772, + "grad_norm": 5.770936489105225, + "learning_rate": 9.990927999092802e-06, + "loss": 0.2545, + "step": 9077 + }, + { + "epoch": 4.220362622036262, + "grad_norm": 5.010940074920654, + "learning_rate": 9.9909269990927e-06, + "loss": 0.173, + "step": 9078 + }, + { + "epoch": 4.220827522082752, + "grad_norm": 5.313000202178955, + "learning_rate": 9.990925999092601e-06, + "loss": 0.2389, + "step": 9079 + }, + { + "epoch": 4.221292422129242, + "grad_norm": 6.7529754638671875, + "learning_rate": 9.9909249990925e-06, + "loss": 0.5174, + "step": 9080 + }, + { + "epoch": 4.2217573221757325, + "grad_norm": 5.224708557128906, + "learning_rate": 9.990923999092401e-06, + "loss": 0.2104, + "step": 9081 + }, + { + "epoch": 4.222222222222222, + "grad_norm": 4.027459144592285, + "learning_rate": 9.9909229990923e-06, + "loss": 0.2323, + "step": 9082 + }, + { + "epoch": 4.222687122268712, + "grad_norm": 3.637232780456543, + "learning_rate": 9.9909219990922e-06, + "loss": 0.1091, + "step": 9083 + }, + { + "epoch": 4.223152022315202, + "grad_norm": 6.8331475257873535, + "learning_rate": 9.990920999092101e-06, + "loss": 0.4761, + "step": 9084 + }, + { + "epoch": 4.223616922361693, + "grad_norm": 3.409963607788086, + "learning_rate": 9.990919999092e-06, + "loss": 0.1278, + "step": 9085 + }, + { + "epoch": 4.224081822408182, + "grad_norm": 6.222006797790527, + "learning_rate": 9.990918999091901e-06, + "loss": 0.3433, + "step": 9086 + }, + { + "epoch": 4.224546722454672, + "grad_norm": 5.702883720397949, + "learning_rate": 9.9909179990918e-06, + "loss": 0.2072, + "step": 9087 + }, + { + "epoch": 4.225011622501162, + "grad_norm": 5.711672306060791, + "learning_rate": 9.9909169990917e-06, + "loss": 0.2149, + "step": 9088 + }, + { + "epoch": 4.225476522547652, + "grad_norm": 6.332920551300049, + "learning_rate": 9.990915999091601e-06, + "loss": 0.2333, + "step": 9089 + }, + { + "epoch": 4.2259414225941425, + "grad_norm": 5.916647911071777, + "learning_rate": 9.990914999091502e-06, + "loss": 0.421, + "step": 9090 + }, + { + "epoch": 4.226406322640632, + "grad_norm": 5.179713249206543, + "learning_rate": 9.9909139990914e-06, + "loss": 0.3795, + "step": 9091 + }, + { + "epoch": 4.226871222687122, + "grad_norm": 5.229369640350342, + "learning_rate": 9.990912999091301e-06, + "loss": 0.3039, + "step": 9092 + }, + { + "epoch": 4.227336122733612, + "grad_norm": 3.0160903930664062, + "learning_rate": 9.9909119990912e-06, + "loss": 0.0803, + "step": 9093 + }, + { + "epoch": 4.227801022780103, + "grad_norm": 2.608663320541382, + "learning_rate": 9.9909109990911e-06, + "loss": 0.0533, + "step": 9094 + }, + { + "epoch": 4.228265922826592, + "grad_norm": 5.4456963539123535, + "learning_rate": 9.990909999091e-06, + "loss": 0.4829, + "step": 9095 + }, + { + "epoch": 4.228730822873082, + "grad_norm": 5.243575572967529, + "learning_rate": 9.9909089990909e-06, + "loss": 0.227, + "step": 9096 + }, + { + "epoch": 4.229195722919572, + "grad_norm": 8.115047454833984, + "learning_rate": 9.990907999090801e-06, + "loss": 0.3605, + "step": 9097 + }, + { + "epoch": 4.229660622966063, + "grad_norm": 4.7021989822387695, + "learning_rate": 9.9909069990907e-06, + "loss": 0.2205, + "step": 9098 + }, + { + "epoch": 4.2301255230125525, + "grad_norm": 4.700892448425293, + "learning_rate": 9.990905999090601e-06, + "loss": 0.3202, + "step": 9099 + }, + { + "epoch": 4.230590423059042, + "grad_norm": 13.137456893920898, + "learning_rate": 9.9909049990905e-06, + "loss": 0.3175, + "step": 9100 + }, + { + "epoch": 4.231055323105532, + "grad_norm": 11.466451644897461, + "learning_rate": 9.9909039990904e-06, + "loss": 0.6851, + "step": 9101 + }, + { + "epoch": 4.231520223152022, + "grad_norm": 8.836050987243652, + "learning_rate": 9.990902999090301e-06, + "loss": 0.1831, + "step": 9102 + }, + { + "epoch": 4.231985123198513, + "grad_norm": 5.615365505218506, + "learning_rate": 9.990901999090202e-06, + "loss": 0.363, + "step": 9103 + }, + { + "epoch": 4.232450023245002, + "grad_norm": 3.1955959796905518, + "learning_rate": 9.990900999090101e-06, + "loss": 0.1079, + "step": 9104 + }, + { + "epoch": 4.232914923291492, + "grad_norm": 4.916450500488281, + "learning_rate": 9.99089999909e-06, + "loss": 0.3637, + "step": 9105 + }, + { + "epoch": 4.233379823337982, + "grad_norm": 5.601739406585693, + "learning_rate": 9.9908989990899e-06, + "loss": 0.4215, + "step": 9106 + }, + { + "epoch": 4.233844723384473, + "grad_norm": 10.762868881225586, + "learning_rate": 9.990897999089801e-06, + "loss": 0.1583, + "step": 9107 + }, + { + "epoch": 4.2343096234309625, + "grad_norm": 4.324827671051025, + "learning_rate": 9.9908969990897e-06, + "loss": 0.2245, + "step": 9108 + }, + { + "epoch": 4.234774523477452, + "grad_norm": 6.57759428024292, + "learning_rate": 9.9908959990896e-06, + "loss": 0.3615, + "step": 9109 + }, + { + "epoch": 4.235239423523942, + "grad_norm": 7.889616966247559, + "learning_rate": 9.990894999089501e-06, + "loss": 0.4495, + "step": 9110 + }, + { + "epoch": 4.235704323570433, + "grad_norm": 6.613075256347656, + "learning_rate": 9.9908939990894e-06, + "loss": 0.4474, + "step": 9111 + }, + { + "epoch": 4.236169223616923, + "grad_norm": 6.190244197845459, + "learning_rate": 9.990892999089301e-06, + "loss": 0.2484, + "step": 9112 + }, + { + "epoch": 4.236634123663412, + "grad_norm": 5.175865173339844, + "learning_rate": 9.9908919990892e-06, + "loss": 0.3388, + "step": 9113 + }, + { + "epoch": 4.237099023709902, + "grad_norm": 6.747528076171875, + "learning_rate": 9.9908909990891e-06, + "loss": 0.4123, + "step": 9114 + }, + { + "epoch": 4.237563923756392, + "grad_norm": 3.902466297149658, + "learning_rate": 9.990889999089001e-06, + "loss": 0.2191, + "step": 9115 + }, + { + "epoch": 4.238028823802883, + "grad_norm": 6.839108943939209, + "learning_rate": 9.9908889990889e-06, + "loss": 0.5668, + "step": 9116 + }, + { + "epoch": 4.2384937238493725, + "grad_norm": 2.7377824783325195, + "learning_rate": 9.990887999088801e-06, + "loss": 0.1127, + "step": 9117 + }, + { + "epoch": 4.238958623895862, + "grad_norm": 5.954606533050537, + "learning_rate": 9.9908869990887e-06, + "loss": 0.3204, + "step": 9118 + }, + { + "epoch": 4.239423523942352, + "grad_norm": 5.404450416564941, + "learning_rate": 9.9908859990886e-06, + "loss": 0.1958, + "step": 9119 + }, + { + "epoch": 4.239888423988843, + "grad_norm": 3.2624690532684326, + "learning_rate": 9.990884999088501e-06, + "loss": 0.1134, + "step": 9120 + }, + { + "epoch": 4.240353324035333, + "grad_norm": 8.519404411315918, + "learning_rate": 9.9908839990884e-06, + "loss": 0.7162, + "step": 9121 + }, + { + "epoch": 4.240818224081822, + "grad_norm": 5.670749187469482, + "learning_rate": 9.9908829990883e-06, + "loss": 0.4059, + "step": 9122 + }, + { + "epoch": 4.241283124128312, + "grad_norm": 6.416330814361572, + "learning_rate": 9.990881999088201e-06, + "loss": 0.3211, + "step": 9123 + }, + { + "epoch": 4.241748024174802, + "grad_norm": 10.222604751586914, + "learning_rate": 9.9908809990881e-06, + "loss": 0.1495, + "step": 9124 + }, + { + "epoch": 4.242212924221293, + "grad_norm": 6.513195991516113, + "learning_rate": 9.990879999088001e-06, + "loss": 0.5442, + "step": 9125 + }, + { + "epoch": 4.2426778242677825, + "grad_norm": 5.854039669036865, + "learning_rate": 9.9908789990879e-06, + "loss": 0.2649, + "step": 9126 + }, + { + "epoch": 4.243142724314272, + "grad_norm": 4.13105583190918, + "learning_rate": 9.9908779990878e-06, + "loss": 0.2359, + "step": 9127 + }, + { + "epoch": 4.243607624360762, + "grad_norm": 3.6795012950897217, + "learning_rate": 9.990876999087701e-06, + "loss": 0.1363, + "step": 9128 + }, + { + "epoch": 4.244072524407253, + "grad_norm": 3.971576690673828, + "learning_rate": 9.9908759990876e-06, + "loss": 0.1842, + "step": 9129 + }, + { + "epoch": 4.244537424453743, + "grad_norm": 4.4947333335876465, + "learning_rate": 9.990874999087501e-06, + "loss": 0.2688, + "step": 9130 + }, + { + "epoch": 4.245002324500232, + "grad_norm": 1.9853668212890625, + "learning_rate": 9.9908739990874e-06, + "loss": 0.0469, + "step": 9131 + }, + { + "epoch": 4.245467224546722, + "grad_norm": 5.472909927368164, + "learning_rate": 9.9908729990873e-06, + "loss": 0.3123, + "step": 9132 + }, + { + "epoch": 4.245932124593212, + "grad_norm": 4.742130756378174, + "learning_rate": 9.990871999087201e-06, + "loss": 0.2736, + "step": 9133 + }, + { + "epoch": 4.246397024639703, + "grad_norm": 3.3045477867126465, + "learning_rate": 9.990870999087102e-06, + "loss": 0.0908, + "step": 9134 + }, + { + "epoch": 4.2468619246861925, + "grad_norm": 5.793573379516602, + "learning_rate": 9.990869999087001e-06, + "loss": 0.5934, + "step": 9135 + }, + { + "epoch": 4.247326824732682, + "grad_norm": 5.949992656707764, + "learning_rate": 9.990868999086902e-06, + "loss": 0.2843, + "step": 9136 + }, + { + "epoch": 4.247791724779172, + "grad_norm": 3.8891470432281494, + "learning_rate": 9.9908679990868e-06, + "loss": 0.1672, + "step": 9137 + }, + { + "epoch": 4.248256624825663, + "grad_norm": 10.029850006103516, + "learning_rate": 9.9908669990867e-06, + "loss": 0.3161, + "step": 9138 + }, + { + "epoch": 4.248721524872153, + "grad_norm": 6.0796732902526855, + "learning_rate": 9.9908659990866e-06, + "loss": 0.5325, + "step": 9139 + }, + { + "epoch": 4.249186424918642, + "grad_norm": 3.3972036838531494, + "learning_rate": 9.9908649990865e-06, + "loss": 0.26, + "step": 9140 + }, + { + "epoch": 4.249651324965132, + "grad_norm": 3.7338786125183105, + "learning_rate": 9.990863999086401e-06, + "loss": 0.1688, + "step": 9141 + }, + { + "epoch": 4.250116225011623, + "grad_norm": 6.261459827423096, + "learning_rate": 9.9908629990863e-06, + "loss": 0.3343, + "step": 9142 + }, + { + "epoch": 4.250581125058113, + "grad_norm": 4.94160270690918, + "learning_rate": 9.990861999086201e-06, + "loss": 0.277, + "step": 9143 + }, + { + "epoch": 4.2510460251046025, + "grad_norm": 7.942879676818848, + "learning_rate": 9.9908609990861e-06, + "loss": 0.1838, + "step": 9144 + }, + { + "epoch": 4.251510925151092, + "grad_norm": 5.208940029144287, + "learning_rate": 9.990859999086e-06, + "loss": 0.3079, + "step": 9145 + }, + { + "epoch": 4.251975825197582, + "grad_norm": 3.214811325073242, + "learning_rate": 9.990858999085901e-06, + "loss": 0.1054, + "step": 9146 + }, + { + "epoch": 4.252440725244073, + "grad_norm": 6.568121910095215, + "learning_rate": 9.990857999085802e-06, + "loss": 0.4093, + "step": 9147 + }, + { + "epoch": 4.252905625290563, + "grad_norm": 4.941591739654541, + "learning_rate": 9.990856999085701e-06, + "loss": 0.1402, + "step": 9148 + }, + { + "epoch": 4.253370525337052, + "grad_norm": 7.040532112121582, + "learning_rate": 9.9908559990856e-06, + "loss": 0.2694, + "step": 9149 + }, + { + "epoch": 4.253835425383542, + "grad_norm": 6.789215564727783, + "learning_rate": 9.9908549990855e-06, + "loss": 0.3757, + "step": 9150 + }, + { + "epoch": 4.254300325430033, + "grad_norm": 7.148548126220703, + "learning_rate": 9.9908539990854e-06, + "loss": 0.4159, + "step": 9151 + }, + { + "epoch": 4.254765225476523, + "grad_norm": 5.775414943695068, + "learning_rate": 9.9908529990853e-06, + "loss": 0.3975, + "step": 9152 + }, + { + "epoch": 4.2552301255230125, + "grad_norm": 6.203925609588623, + "learning_rate": 9.990851999085201e-06, + "loss": 0.3152, + "step": 9153 + }, + { + "epoch": 4.255695025569502, + "grad_norm": 8.234856605529785, + "learning_rate": 9.990850999085102e-06, + "loss": 0.4232, + "step": 9154 + }, + { + "epoch": 4.256159925615993, + "grad_norm": 5.295629501342773, + "learning_rate": 9.990849999085e-06, + "loss": 0.2422, + "step": 9155 + }, + { + "epoch": 4.256624825662483, + "grad_norm": 12.025513648986816, + "learning_rate": 9.990848999084901e-06, + "loss": 0.4529, + "step": 9156 + }, + { + "epoch": 4.257089725708973, + "grad_norm": 4.422934055328369, + "learning_rate": 9.9908479990848e-06, + "loss": 0.1701, + "step": 9157 + }, + { + "epoch": 4.257554625755462, + "grad_norm": 8.42046070098877, + "learning_rate": 9.9908469990847e-06, + "loss": 0.432, + "step": 9158 + }, + { + "epoch": 4.258019525801952, + "grad_norm": 8.815457344055176, + "learning_rate": 9.990845999084601e-06, + "loss": 0.504, + "step": 9159 + }, + { + "epoch": 4.258484425848443, + "grad_norm": 6.278488636016846, + "learning_rate": 9.9908449990845e-06, + "loss": 0.2188, + "step": 9160 + }, + { + "epoch": 4.258949325894933, + "grad_norm": 7.327540397644043, + "learning_rate": 9.990843999084401e-06, + "loss": 0.3974, + "step": 9161 + }, + { + "epoch": 4.2594142259414225, + "grad_norm": 12.628582000732422, + "learning_rate": 9.9908429990843e-06, + "loss": 0.8734, + "step": 9162 + }, + { + "epoch": 4.259879125987912, + "grad_norm": 4.841217517852783, + "learning_rate": 9.9908419990842e-06, + "loss": 0.2397, + "step": 9163 + }, + { + "epoch": 4.260344026034403, + "grad_norm": 4.854616165161133, + "learning_rate": 9.9908409990841e-06, + "loss": 0.1764, + "step": 9164 + }, + { + "epoch": 4.260808926080893, + "grad_norm": 4.033719062805176, + "learning_rate": 9.990839999084e-06, + "loss": 0.2543, + "step": 9165 + }, + { + "epoch": 4.261273826127383, + "grad_norm": 6.415824890136719, + "learning_rate": 9.990838999083901e-06, + "loss": 0.3441, + "step": 9166 + }, + { + "epoch": 4.261738726173872, + "grad_norm": 6.019629001617432, + "learning_rate": 9.990837999083802e-06, + "loss": 0.46, + "step": 9167 + }, + { + "epoch": 4.262203626220362, + "grad_norm": 5.705272197723389, + "learning_rate": 9.9908369990837e-06, + "loss": 0.2442, + "step": 9168 + }, + { + "epoch": 4.262668526266853, + "grad_norm": 4.762701988220215, + "learning_rate": 9.990835999083601e-06, + "loss": 0.2628, + "step": 9169 + }, + { + "epoch": 4.263133426313343, + "grad_norm": 5.65317440032959, + "learning_rate": 9.9908349990835e-06, + "loss": 0.4394, + "step": 9170 + }, + { + "epoch": 4.2635983263598325, + "grad_norm": 6.996469020843506, + "learning_rate": 9.990833999083401e-06, + "loss": 0.1208, + "step": 9171 + }, + { + "epoch": 4.264063226406322, + "grad_norm": 6.2982177734375, + "learning_rate": 9.990832999083302e-06, + "loss": 0.4624, + "step": 9172 + }, + { + "epoch": 4.264528126452813, + "grad_norm": 7.141604900360107, + "learning_rate": 9.9908319990832e-06, + "loss": 0.2615, + "step": 9173 + }, + { + "epoch": 4.264993026499303, + "grad_norm": 6.5479607582092285, + "learning_rate": 9.990830999083101e-06, + "loss": 0.3128, + "step": 9174 + }, + { + "epoch": 4.265457926545793, + "grad_norm": 5.132667064666748, + "learning_rate": 9.990829999083e-06, + "loss": 0.4133, + "step": 9175 + }, + { + "epoch": 4.265922826592282, + "grad_norm": 7.832809925079346, + "learning_rate": 9.9908289990829e-06, + "loss": 0.3358, + "step": 9176 + }, + { + "epoch": 4.266387726638772, + "grad_norm": 7.317866325378418, + "learning_rate": 9.9908279990828e-06, + "loss": 0.2572, + "step": 9177 + }, + { + "epoch": 4.266852626685263, + "grad_norm": 7.692831516265869, + "learning_rate": 9.9908269990827e-06, + "loss": 0.3721, + "step": 9178 + }, + { + "epoch": 4.267317526731753, + "grad_norm": 5.049525260925293, + "learning_rate": 9.990825999082601e-06, + "loss": 0.3656, + "step": 9179 + }, + { + "epoch": 4.2677824267782425, + "grad_norm": 4.234042167663574, + "learning_rate": 9.990824999082502e-06, + "loss": 0.2014, + "step": 9180 + }, + { + "epoch": 4.268247326824732, + "grad_norm": 6.876502990722656, + "learning_rate": 9.9908239990824e-06, + "loss": 0.3655, + "step": 9181 + }, + { + "epoch": 4.268712226871223, + "grad_norm": 4.257472991943359, + "learning_rate": 9.9908229990823e-06, + "loss": 0.1369, + "step": 9182 + }, + { + "epoch": 4.269177126917713, + "grad_norm": 5.146773338317871, + "learning_rate": 9.9908219990822e-06, + "loss": 0.1848, + "step": 9183 + }, + { + "epoch": 4.269642026964203, + "grad_norm": 5.730294227600098, + "learning_rate": 9.990820999082101e-06, + "loss": 0.2221, + "step": 9184 + }, + { + "epoch": 4.270106927010692, + "grad_norm": 3.8838090896606445, + "learning_rate": 9.990819999082002e-06, + "loss": 0.2349, + "step": 9185 + }, + { + "epoch": 4.270571827057183, + "grad_norm": 4.594846725463867, + "learning_rate": 9.9908189990819e-06, + "loss": 0.327, + "step": 9186 + }, + { + "epoch": 4.271036727103673, + "grad_norm": 5.181523323059082, + "learning_rate": 9.990817999081801e-06, + "loss": 0.4024, + "step": 9187 + }, + { + "epoch": 4.271501627150163, + "grad_norm": 5.886035919189453, + "learning_rate": 9.9908169990817e-06, + "loss": 0.4147, + "step": 9188 + }, + { + "epoch": 4.2719665271966525, + "grad_norm": 3.766950845718384, + "learning_rate": 9.990815999081601e-06, + "loss": 0.1836, + "step": 9189 + }, + { + "epoch": 4.272431427243143, + "grad_norm": 7.905692100524902, + "learning_rate": 9.9908149990815e-06, + "loss": 0.3068, + "step": 9190 + }, + { + "epoch": 4.272896327289633, + "grad_norm": 6.935624599456787, + "learning_rate": 9.9908139990814e-06, + "loss": 0.5238, + "step": 9191 + }, + { + "epoch": 4.273361227336123, + "grad_norm": 3.9440531730651855, + "learning_rate": 9.990812999081301e-06, + "loss": 0.2426, + "step": 9192 + }, + { + "epoch": 4.273826127382613, + "grad_norm": 6.430100440979004, + "learning_rate": 9.9908119990812e-06, + "loss": 0.4228, + "step": 9193 + }, + { + "epoch": 4.274291027429102, + "grad_norm": 4.647778511047363, + "learning_rate": 9.9908109990811e-06, + "loss": 0.1801, + "step": 9194 + }, + { + "epoch": 4.274755927475593, + "grad_norm": 4.007472038269043, + "learning_rate": 9.990809999081e-06, + "loss": 0.0974, + "step": 9195 + }, + { + "epoch": 4.275220827522083, + "grad_norm": 5.81678581237793, + "learning_rate": 9.9908089990809e-06, + "loss": 0.231, + "step": 9196 + }, + { + "epoch": 4.275685727568573, + "grad_norm": 5.132024765014648, + "learning_rate": 9.990807999080801e-06, + "loss": 0.2362, + "step": 9197 + }, + { + "epoch": 4.2761506276150625, + "grad_norm": 2.735595941543579, + "learning_rate": 9.990806999080702e-06, + "loss": 0.0606, + "step": 9198 + }, + { + "epoch": 4.276615527661553, + "grad_norm": 4.334656238555908, + "learning_rate": 9.9908059990806e-06, + "loss": 0.1879, + "step": 9199 + }, + { + "epoch": 4.277080427708043, + "grad_norm": 6.6732659339904785, + "learning_rate": 9.990804999080501e-06, + "loss": 0.4468, + "step": 9200 + }, + { + "epoch": 4.277545327754533, + "grad_norm": 5.442502498626709, + "learning_rate": 9.9908039990804e-06, + "loss": 0.1037, + "step": 9201 + }, + { + "epoch": 4.278010227801023, + "grad_norm": 7.484472751617432, + "learning_rate": 9.990802999080301e-06, + "loss": 0.3804, + "step": 9202 + }, + { + "epoch": 4.278475127847512, + "grad_norm": 6.926887035369873, + "learning_rate": 9.990801999080202e-06, + "loss": 0.2939, + "step": 9203 + }, + { + "epoch": 4.278940027894003, + "grad_norm": 7.013240337371826, + "learning_rate": 9.9908009990801e-06, + "loss": 0.458, + "step": 9204 + }, + { + "epoch": 4.279404927940493, + "grad_norm": 6.321433067321777, + "learning_rate": 9.990799999080001e-06, + "loss": 0.2175, + "step": 9205 + }, + { + "epoch": 4.279869827986983, + "grad_norm": 6.339756965637207, + "learning_rate": 9.9907989990799e-06, + "loss": 0.4204, + "step": 9206 + }, + { + "epoch": 4.2803347280334725, + "grad_norm": 3.902871608734131, + "learning_rate": 9.9907979990798e-06, + "loss": 0.1095, + "step": 9207 + }, + { + "epoch": 4.280799628079963, + "grad_norm": 5.477300643920898, + "learning_rate": 9.9907969990797e-06, + "loss": 0.469, + "step": 9208 + }, + { + "epoch": 4.281264528126453, + "grad_norm": 3.6807100772857666, + "learning_rate": 9.9907959990796e-06, + "loss": 0.1875, + "step": 9209 + }, + { + "epoch": 4.281729428172943, + "grad_norm": 5.120092868804932, + "learning_rate": 9.990794999079501e-06, + "loss": 0.3704, + "step": 9210 + }, + { + "epoch": 4.282194328219433, + "grad_norm": 4.305742263793945, + "learning_rate": 9.990793999079402e-06, + "loss": 0.2115, + "step": 9211 + }, + { + "epoch": 4.282659228265922, + "grad_norm": 9.160598754882812, + "learning_rate": 9.9907929990793e-06, + "loss": 0.4676, + "step": 9212 + }, + { + "epoch": 4.283124128312413, + "grad_norm": 6.427425384521484, + "learning_rate": 9.990791999079201e-06, + "loss": 0.5227, + "step": 9213 + }, + { + "epoch": 4.283589028358903, + "grad_norm": 4.833632946014404, + "learning_rate": 9.9907909990791e-06, + "loss": 0.2115, + "step": 9214 + }, + { + "epoch": 4.284053928405393, + "grad_norm": 6.816425323486328, + "learning_rate": 9.990789999079001e-06, + "loss": 0.2488, + "step": 9215 + }, + { + "epoch": 4.2845188284518825, + "grad_norm": 4.860684871673584, + "learning_rate": 9.990788999078902e-06, + "loss": 0.1877, + "step": 9216 + }, + { + "epoch": 4.284983728498373, + "grad_norm": 8.138521194458008, + "learning_rate": 9.9907879990788e-06, + "loss": 0.329, + "step": 9217 + }, + { + "epoch": 4.285448628544863, + "grad_norm": 5.367650508880615, + "learning_rate": 9.990786999078701e-06, + "loss": 0.2377, + "step": 9218 + }, + { + "epoch": 4.285913528591353, + "grad_norm": Infinity, + "learning_rate": 9.990786999078701e-06, + "loss": 0.2425, + "step": 9219 + }, + { + "epoch": 4.286378428637843, + "grad_norm": 3.9401447772979736, + "learning_rate": 9.9907859990786e-06, + "loss": 0.2192, + "step": 9220 + }, + { + "epoch": 4.286843328684333, + "grad_norm": 10.086082458496094, + "learning_rate": 9.990784999078501e-06, + "loss": 0.5298, + "step": 9221 + }, + { + "epoch": 4.287308228730823, + "grad_norm": 4.567083835601807, + "learning_rate": 9.9907839990784e-06, + "loss": 0.1719, + "step": 9222 + }, + { + "epoch": 4.287773128777313, + "grad_norm": 7.926868438720703, + "learning_rate": 9.9907829990783e-06, + "loss": 0.6719, + "step": 9223 + }, + { + "epoch": 4.288238028823803, + "grad_norm": 6.230688095092773, + "learning_rate": 9.990781999078201e-06, + "loss": 0.3325, + "step": 9224 + }, + { + "epoch": 4.2887029288702925, + "grad_norm": 4.003109455108643, + "learning_rate": 9.9907809990781e-06, + "loss": 0.2371, + "step": 9225 + }, + { + "epoch": 4.289167828916783, + "grad_norm": 10.535419464111328, + "learning_rate": 9.990779999078e-06, + "loss": 0.4783, + "step": 9226 + }, + { + "epoch": 4.289632728963273, + "grad_norm": 5.941984176635742, + "learning_rate": 9.9907789990779e-06, + "loss": 0.3764, + "step": 9227 + }, + { + "epoch": 4.290097629009763, + "grad_norm": 2.9820406436920166, + "learning_rate": 9.9907779990778e-06, + "loss": 0.1382, + "step": 9228 + }, + { + "epoch": 4.290562529056253, + "grad_norm": 4.4714508056640625, + "learning_rate": 9.990776999077701e-06, + "loss": 0.4002, + "step": 9229 + }, + { + "epoch": 4.291027429102743, + "grad_norm": 5.267688274383545, + "learning_rate": 9.990775999077602e-06, + "loss": 0.2831, + "step": 9230 + }, + { + "epoch": 4.291492329149233, + "grad_norm": 6.244569778442383, + "learning_rate": 9.9907749990775e-06, + "loss": 0.4544, + "step": 9231 + }, + { + "epoch": 4.291957229195723, + "grad_norm": 4.063471794128418, + "learning_rate": 9.990773999077401e-06, + "loss": 0.1753, + "step": 9232 + }, + { + "epoch": 4.292422129242213, + "grad_norm": 5.914006233215332, + "learning_rate": 9.9907729990773e-06, + "loss": 0.2558, + "step": 9233 + }, + { + "epoch": 4.292887029288703, + "grad_norm": 7.226432800292969, + "learning_rate": 9.990771999077201e-06, + "loss": 0.3566, + "step": 9234 + }, + { + "epoch": 4.293351929335193, + "grad_norm": 4.469247817993164, + "learning_rate": 9.9907709990771e-06, + "loss": 0.3479, + "step": 9235 + }, + { + "epoch": 4.293816829381683, + "grad_norm": 4.610983371734619, + "learning_rate": 9.990769999077e-06, + "loss": 0.1624, + "step": 9236 + }, + { + "epoch": 4.294281729428173, + "grad_norm": 6.292228698730469, + "learning_rate": 9.990768999076901e-06, + "loss": 0.2157, + "step": 9237 + }, + { + "epoch": 4.294746629474663, + "grad_norm": 7.7406182289123535, + "learning_rate": 9.9907679990768e-06, + "loss": 0.197, + "step": 9238 + }, + { + "epoch": 4.295211529521153, + "grad_norm": 6.733648300170898, + "learning_rate": 9.990766999076701e-06, + "loss": 0.4014, + "step": 9239 + }, + { + "epoch": 4.295676429567643, + "grad_norm": 6.295163154602051, + "learning_rate": 9.9907659990766e-06, + "loss": 0.5005, + "step": 9240 + }, + { + "epoch": 4.296141329614133, + "grad_norm": 4.0203423500061035, + "learning_rate": 9.9907649990765e-06, + "loss": 0.2115, + "step": 9241 + }, + { + "epoch": 4.296606229660623, + "grad_norm": 4.25365686416626, + "learning_rate": 9.990763999076401e-06, + "loss": 0.1801, + "step": 9242 + }, + { + "epoch": 4.297071129707113, + "grad_norm": 3.1802031993865967, + "learning_rate": 9.990762999076302e-06, + "loss": 0.1097, + "step": 9243 + }, + { + "epoch": 4.297536029753603, + "grad_norm": 7.316305637359619, + "learning_rate": 9.9907619990762e-06, + "loss": 0.37, + "step": 9244 + }, + { + "epoch": 4.298000929800093, + "grad_norm": 7.655298709869385, + "learning_rate": 9.990760999076101e-06, + "loss": 0.4713, + "step": 9245 + }, + { + "epoch": 4.298465829846583, + "grad_norm": 3.372042655944824, + "learning_rate": 9.990759999076e-06, + "loss": 0.1299, + "step": 9246 + }, + { + "epoch": 4.298930729893073, + "grad_norm": 4.930408477783203, + "learning_rate": 9.9907589990759e-06, + "loss": 0.1594, + "step": 9247 + }, + { + "epoch": 4.299395629939563, + "grad_norm": 5.591670513153076, + "learning_rate": 9.9907579990758e-06, + "loss": 0.3299, + "step": 9248 + }, + { + "epoch": 4.299860529986053, + "grad_norm": 6.844636917114258, + "learning_rate": 9.9907569990757e-06, + "loss": 0.3992, + "step": 9249 + }, + { + "epoch": 4.300325430032543, + "grad_norm": 7.708812713623047, + "learning_rate": 9.990755999075601e-06, + "loss": 0.1854, + "step": 9250 + }, + { + "epoch": 4.300790330079033, + "grad_norm": 10.874711990356445, + "learning_rate": 9.9907549990755e-06, + "loss": 0.8454, + "step": 9251 + }, + { + "epoch": 4.301255230125523, + "grad_norm": 6.003425121307373, + "learning_rate": 9.990753999075401e-06, + "loss": 0.239, + "step": 9252 + }, + { + "epoch": 4.301720130172013, + "grad_norm": 6.019950866699219, + "learning_rate": 9.9907529990753e-06, + "loss": 0.4294, + "step": 9253 + }, + { + "epoch": 4.302185030218503, + "grad_norm": 2.8541126251220703, + "learning_rate": 9.9907519990752e-06, + "loss": 0.0947, + "step": 9254 + }, + { + "epoch": 4.302649930264993, + "grad_norm": 4.084362983703613, + "learning_rate": 9.990750999075101e-06, + "loss": 0.1823, + "step": 9255 + }, + { + "epoch": 4.303114830311483, + "grad_norm": 4.350910186767578, + "learning_rate": 9.990749999075002e-06, + "loss": 0.1807, + "step": 9256 + }, + { + "epoch": 4.303579730357973, + "grad_norm": 6.786457538604736, + "learning_rate": 9.990748999074901e-06, + "loss": 0.5981, + "step": 9257 + }, + { + "epoch": 4.304044630404463, + "grad_norm": 8.137569427490234, + "learning_rate": 9.9907479990748e-06, + "loss": 0.4812, + "step": 9258 + }, + { + "epoch": 4.304509530450953, + "grad_norm": 6.815385818481445, + "learning_rate": 9.9907469990747e-06, + "loss": 0.2004, + "step": 9259 + }, + { + "epoch": 4.304974430497443, + "grad_norm": 4.711209297180176, + "learning_rate": 9.9907459990746e-06, + "loss": 0.147, + "step": 9260 + }, + { + "epoch": 4.305439330543933, + "grad_norm": 3.806340217590332, + "learning_rate": 9.9907449990745e-06, + "loss": 0.1529, + "step": 9261 + }, + { + "epoch": 4.305904230590423, + "grad_norm": 5.0257720947265625, + "learning_rate": 9.9907439990744e-06, + "loss": 0.3751, + "step": 9262 + }, + { + "epoch": 4.306369130636913, + "grad_norm": 3.640495538711548, + "learning_rate": 9.990742999074301e-06, + "loss": 0.1662, + "step": 9263 + }, + { + "epoch": 4.306834030683403, + "grad_norm": 5.046971321105957, + "learning_rate": 9.9907419990742e-06, + "loss": 0.201, + "step": 9264 + }, + { + "epoch": 4.3072989307298934, + "grad_norm": 5.387972831726074, + "learning_rate": 9.990740999074101e-06, + "loss": 0.2899, + "step": 9265 + }, + { + "epoch": 4.307763830776383, + "grad_norm": 8.825787544250488, + "learning_rate": 9.990739999074e-06, + "loss": 0.4392, + "step": 9266 + }, + { + "epoch": 4.308228730822873, + "grad_norm": 7.778316020965576, + "learning_rate": 9.9907389990739e-06, + "loss": 0.3044, + "step": 9267 + }, + { + "epoch": 4.308693630869363, + "grad_norm": 6.831406593322754, + "learning_rate": 9.990737999073801e-06, + "loss": 0.4429, + "step": 9268 + }, + { + "epoch": 4.3091585309158535, + "grad_norm": 3.579096555709839, + "learning_rate": 9.9907369990737e-06, + "loss": 0.1623, + "step": 9269 + }, + { + "epoch": 4.309623430962343, + "grad_norm": 13.366432189941406, + "learning_rate": 9.990735999073601e-06, + "loss": 0.3199, + "step": 9270 + }, + { + "epoch": 4.310088331008833, + "grad_norm": 4.74901008605957, + "learning_rate": 9.9907349990735e-06, + "loss": 0.3276, + "step": 9271 + }, + { + "epoch": 4.310553231055323, + "grad_norm": 14.433035850524902, + "learning_rate": 9.9907339990734e-06, + "loss": 0.2178, + "step": 9272 + }, + { + "epoch": 4.311018131101813, + "grad_norm": 5.693557262420654, + "learning_rate": 9.990732999073301e-06, + "loss": 0.3485, + "step": 9273 + }, + { + "epoch": 4.3114830311483034, + "grad_norm": 6.674008846282959, + "learning_rate": 9.9907319990732e-06, + "loss": 0.4076, + "step": 9274 + }, + { + "epoch": 4.311947931194793, + "grad_norm": 5.843663215637207, + "learning_rate": 9.990730999073101e-06, + "loss": 0.3231, + "step": 9275 + }, + { + "epoch": 4.312412831241283, + "grad_norm": 4.286233425140381, + "learning_rate": 9.990729999073002e-06, + "loss": 0.1839, + "step": 9276 + }, + { + "epoch": 4.312877731287773, + "grad_norm": 7.0464301109313965, + "learning_rate": 9.9907289990729e-06, + "loss": 0.2812, + "step": 9277 + }, + { + "epoch": 4.3133426313342635, + "grad_norm": 8.79800033569336, + "learning_rate": 9.990727999072801e-06, + "loss": 0.464, + "step": 9278 + }, + { + "epoch": 4.313807531380753, + "grad_norm": 5.187692642211914, + "learning_rate": 9.9907269990727e-06, + "loss": 0.2649, + "step": 9279 + }, + { + "epoch": 4.314272431427243, + "grad_norm": 3.9387366771698, + "learning_rate": 9.9907259990726e-06, + "loss": 0.0882, + "step": 9280 + }, + { + "epoch": 4.314737331473733, + "grad_norm": 7.011622428894043, + "learning_rate": 9.990724999072501e-06, + "loss": 0.5035, + "step": 9281 + }, + { + "epoch": 4.315202231520223, + "grad_norm": 5.031150817871094, + "learning_rate": 9.9907239990724e-06, + "loss": 0.3071, + "step": 9282 + }, + { + "epoch": 4.315667131566713, + "grad_norm": 6.591774940490723, + "learning_rate": 9.990722999072301e-06, + "loss": 0.5008, + "step": 9283 + }, + { + "epoch": 4.316132031613203, + "grad_norm": 5.779001712799072, + "learning_rate": 9.9907219990722e-06, + "loss": 0.2398, + "step": 9284 + }, + { + "epoch": 4.316596931659693, + "grad_norm": 6.116891860961914, + "learning_rate": 9.9907209990721e-06, + "loss": 0.4633, + "step": 9285 + }, + { + "epoch": 4.317061831706183, + "grad_norm": 4.105109691619873, + "learning_rate": 9.990719999072001e-06, + "loss": 0.1929, + "step": 9286 + }, + { + "epoch": 4.3175267317526735, + "grad_norm": 5.0127716064453125, + "learning_rate": 9.9907189990719e-06, + "loss": 0.2838, + "step": 9287 + }, + { + "epoch": 4.317991631799163, + "grad_norm": 6.492511749267578, + "learning_rate": 9.990717999071801e-06, + "loss": 0.3909, + "step": 9288 + }, + { + "epoch": 4.318456531845653, + "grad_norm": 7.960748195648193, + "learning_rate": 9.990716999071702e-06, + "loss": 0.4504, + "step": 9289 + }, + { + "epoch": 4.318921431892143, + "grad_norm": 5.409559726715088, + "learning_rate": 9.9907159990716e-06, + "loss": 0.3166, + "step": 9290 + }, + { + "epoch": 4.319386331938633, + "grad_norm": 5.096149444580078, + "learning_rate": 9.9907149990715e-06, + "loss": 0.2353, + "step": 9291 + }, + { + "epoch": 4.319851231985123, + "grad_norm": 4.803279399871826, + "learning_rate": 9.9907139990714e-06, + "loss": 0.3154, + "step": 9292 + }, + { + "epoch": 4.320316132031613, + "grad_norm": 4.734849452972412, + "learning_rate": 9.990712999071301e-06, + "loss": 0.2355, + "step": 9293 + }, + { + "epoch": 4.320781032078103, + "grad_norm": 3.2359795570373535, + "learning_rate": 9.990711999071202e-06, + "loss": 0.1105, + "step": 9294 + }, + { + "epoch": 4.321245932124593, + "grad_norm": 6.759286880493164, + "learning_rate": 9.9907109990711e-06, + "loss": 0.4749, + "step": 9295 + }, + { + "epoch": 4.3217108321710835, + "grad_norm": 5.346119403839111, + "learning_rate": 9.990709999071001e-06, + "loss": 0.2366, + "step": 9296 + }, + { + "epoch": 4.322175732217573, + "grad_norm": 4.140191555023193, + "learning_rate": 9.9907089990709e-06, + "loss": 0.2608, + "step": 9297 + }, + { + "epoch": 4.322640632264063, + "grad_norm": 4.27674674987793, + "learning_rate": 9.9907079990708e-06, + "loss": 0.2647, + "step": 9298 + }, + { + "epoch": 4.323105532310553, + "grad_norm": 5.969515323638916, + "learning_rate": 9.990706999070701e-06, + "loss": 0.4309, + "step": 9299 + }, + { + "epoch": 4.323570432357044, + "grad_norm": 6.636129856109619, + "learning_rate": 9.990705999070602e-06, + "loss": 0.1966, + "step": 9300 + }, + { + "epoch": 4.324035332403533, + "grad_norm": 4.941951274871826, + "learning_rate": 9.990704999070501e-06, + "loss": 0.2875, + "step": 9301 + }, + { + "epoch": 4.324500232450023, + "grad_norm": 4.190491676330566, + "learning_rate": 9.9907039990704e-06, + "loss": 0.1572, + "step": 9302 + }, + { + "epoch": 4.324965132496513, + "grad_norm": 5.345705032348633, + "learning_rate": 9.9907029990703e-06, + "loss": 0.386, + "step": 9303 + }, + { + "epoch": 4.325430032543003, + "grad_norm": 3.4374887943267822, + "learning_rate": 9.9907019990702e-06, + "loss": 0.1646, + "step": 9304 + }, + { + "epoch": 4.3258949325894935, + "grad_norm": 6.216460704803467, + "learning_rate": 9.9907009990701e-06, + "loss": 0.3187, + "step": 9305 + }, + { + "epoch": 4.326359832635983, + "grad_norm": 5.124750137329102, + "learning_rate": 9.990699999070001e-06, + "loss": 0.2714, + "step": 9306 + }, + { + "epoch": 4.326824732682473, + "grad_norm": 6.7482452392578125, + "learning_rate": 9.990698999069902e-06, + "loss": 0.2304, + "step": 9307 + }, + { + "epoch": 4.327289632728963, + "grad_norm": 3.7776269912719727, + "learning_rate": 9.9906979990698e-06, + "loss": 0.1562, + "step": 9308 + }, + { + "epoch": 4.327754532775454, + "grad_norm": 2.4090983867645264, + "learning_rate": 9.990696999069701e-06, + "loss": 0.0735, + "step": 9309 + }, + { + "epoch": 4.328219432821943, + "grad_norm": 5.096031665802002, + "learning_rate": 9.9906959990696e-06, + "loss": 0.439, + "step": 9310 + }, + { + "epoch": 4.328684332868433, + "grad_norm": 6.883484363555908, + "learning_rate": 9.9906949990695e-06, + "loss": 0.4532, + "step": 9311 + }, + { + "epoch": 4.329149232914923, + "grad_norm": 5.467026233673096, + "learning_rate": 9.990693999069401e-06, + "loss": 0.3524, + "step": 9312 + }, + { + "epoch": 4.329614132961414, + "grad_norm": 3.08858323097229, + "learning_rate": 9.9906929990693e-06, + "loss": 0.1469, + "step": 9313 + }, + { + "epoch": 4.3300790330079035, + "grad_norm": 5.062413692474365, + "learning_rate": 9.990691999069201e-06, + "loss": 0.2931, + "step": 9314 + }, + { + "epoch": 4.330543933054393, + "grad_norm": 4.075096130371094, + "learning_rate": 9.9906909990691e-06, + "loss": 0.241, + "step": 9315 + }, + { + "epoch": 4.331008833100883, + "grad_norm": 4.453060150146484, + "learning_rate": 9.990689999069e-06, + "loss": 0.2608, + "step": 9316 + }, + { + "epoch": 4.331473733147373, + "grad_norm": 3.8194892406463623, + "learning_rate": 9.9906889990689e-06, + "loss": 0.1711, + "step": 9317 + }, + { + "epoch": 4.331938633193864, + "grad_norm": 4.258971214294434, + "learning_rate": 9.9906879990688e-06, + "loss": 0.1253, + "step": 9318 + }, + { + "epoch": 4.332403533240353, + "grad_norm": 4.296113967895508, + "learning_rate": 9.990686999068701e-06, + "loss": 0.1096, + "step": 9319 + }, + { + "epoch": 4.332868433286843, + "grad_norm": 4.202164173126221, + "learning_rate": 9.990685999068602e-06, + "loss": 0.1421, + "step": 9320 + }, + { + "epoch": 4.333333333333333, + "grad_norm": 4.842818737030029, + "learning_rate": 9.9906849990685e-06, + "loss": 0.2334, + "step": 9321 + }, + { + "epoch": 4.333798233379824, + "grad_norm": 9.502901077270508, + "learning_rate": 9.990683999068401e-06, + "loss": 0.3145, + "step": 9322 + }, + { + "epoch": 4.3342631334263135, + "grad_norm": 6.061665058135986, + "learning_rate": 9.9906829990683e-06, + "loss": 0.21, + "step": 9323 + }, + { + "epoch": 4.334728033472803, + "grad_norm": 3.991321325302124, + "learning_rate": 9.990681999068201e-06, + "loss": 0.1687, + "step": 9324 + }, + { + "epoch": 4.335192933519293, + "grad_norm": 5.023401260375977, + "learning_rate": 9.990680999068102e-06, + "loss": 0.258, + "step": 9325 + }, + { + "epoch": 4.335657833565783, + "grad_norm": 3.2131474018096924, + "learning_rate": 9.990679999068e-06, + "loss": 0.1459, + "step": 9326 + }, + { + "epoch": 4.336122733612274, + "grad_norm": 5.015263557434082, + "learning_rate": 9.990678999067901e-06, + "loss": 0.359, + "step": 9327 + }, + { + "epoch": 4.336587633658763, + "grad_norm": 8.016777038574219, + "learning_rate": 9.9906779990678e-06, + "loss": 0.4199, + "step": 9328 + }, + { + "epoch": 4.337052533705253, + "grad_norm": 6.223348140716553, + "learning_rate": 9.9906769990677e-06, + "loss": 0.1372, + "step": 9329 + }, + { + "epoch": 4.337517433751743, + "grad_norm": 6.401369571685791, + "learning_rate": 9.9906759990676e-06, + "loss": 0.3741, + "step": 9330 + }, + { + "epoch": 4.337982333798234, + "grad_norm": 6.882705211639404, + "learning_rate": 9.9906749990675e-06, + "loss": 0.429, + "step": 9331 + }, + { + "epoch": 4.3384472338447235, + "grad_norm": 6.791640281677246, + "learning_rate": 9.990673999067401e-06, + "loss": 0.5275, + "step": 9332 + }, + { + "epoch": 4.338912133891213, + "grad_norm": 4.3797736167907715, + "learning_rate": 9.990672999067302e-06, + "loss": 0.1998, + "step": 9333 + }, + { + "epoch": 4.339377033937703, + "grad_norm": 5.151256084442139, + "learning_rate": 9.9906719990672e-06, + "loss": 0.2026, + "step": 9334 + }, + { + "epoch": 4.339841933984193, + "grad_norm": 6.747164726257324, + "learning_rate": 9.9906709990671e-06, + "loss": 0.4399, + "step": 9335 + }, + { + "epoch": 4.340306834030684, + "grad_norm": 3.081212282180786, + "learning_rate": 9.990669999067e-06, + "loss": 0.1073, + "step": 9336 + }, + { + "epoch": 4.340771734077173, + "grad_norm": 5.7197113037109375, + "learning_rate": 9.990668999066901e-06, + "loss": 0.1951, + "step": 9337 + }, + { + "epoch": 4.341236634123663, + "grad_norm": 5.01422643661499, + "learning_rate": 9.990667999066802e-06, + "loss": 0.1744, + "step": 9338 + }, + { + "epoch": 4.341701534170153, + "grad_norm": 6.941490650177002, + "learning_rate": 9.9906669990667e-06, + "loss": 0.2698, + "step": 9339 + }, + { + "epoch": 4.342166434216644, + "grad_norm": 7.5025482177734375, + "learning_rate": 9.990665999066601e-06, + "loss": 0.246, + "step": 9340 + }, + { + "epoch": 4.3426313342631335, + "grad_norm": 2.4544289112091064, + "learning_rate": 9.9906649990665e-06, + "loss": 0.0967, + "step": 9341 + }, + { + "epoch": 4.343096234309623, + "grad_norm": 6.325511455535889, + "learning_rate": 9.990663999066401e-06, + "loss": 0.2721, + "step": 9342 + }, + { + "epoch": 4.343561134356113, + "grad_norm": 5.437417507171631, + "learning_rate": 9.9906629990663e-06, + "loss": 0.2336, + "step": 9343 + }, + { + "epoch": 4.344026034402604, + "grad_norm": 5.999084949493408, + "learning_rate": 9.9906619990662e-06, + "loss": 0.4355, + "step": 9344 + }, + { + "epoch": 4.344490934449094, + "grad_norm": 7.445609092712402, + "learning_rate": 9.990660999066101e-06, + "loss": 0.5212, + "step": 9345 + }, + { + "epoch": 4.344955834495583, + "grad_norm": 7.636040687561035, + "learning_rate": 9.990659999066e-06, + "loss": 0.4319, + "step": 9346 + }, + { + "epoch": 4.345420734542073, + "grad_norm": 3.682497024536133, + "learning_rate": 9.9906589990659e-06, + "loss": 0.1073, + "step": 9347 + }, + { + "epoch": 4.345885634588564, + "grad_norm": 6.048182487487793, + "learning_rate": 9.9906579990658e-06, + "loss": 0.3306, + "step": 9348 + }, + { + "epoch": 4.346350534635054, + "grad_norm": 5.411661624908447, + "learning_rate": 9.9906569990657e-06, + "loss": 0.3261, + "step": 9349 + }, + { + "epoch": 4.3468154346815435, + "grad_norm": 7.715559482574463, + "learning_rate": 9.990655999065601e-06, + "loss": 0.3449, + "step": 9350 + }, + { + "epoch": 4.347280334728033, + "grad_norm": 7.110766410827637, + "learning_rate": 9.990654999065502e-06, + "loss": 0.5005, + "step": 9351 + }, + { + "epoch": 4.347745234774523, + "grad_norm": 5.196442604064941, + "learning_rate": 9.9906539990654e-06, + "loss": 0.3596, + "step": 9352 + }, + { + "epoch": 4.348210134821014, + "grad_norm": 5.061161041259766, + "learning_rate": 9.990652999065301e-06, + "loss": 0.2912, + "step": 9353 + }, + { + "epoch": 4.348675034867504, + "grad_norm": 5.426323413848877, + "learning_rate": 9.9906519990652e-06, + "loss": 0.1831, + "step": 9354 + }, + { + "epoch": 4.349139934913993, + "grad_norm": 5.573718547821045, + "learning_rate": 9.990650999065101e-06, + "loss": 0.2946, + "step": 9355 + }, + { + "epoch": 4.349604834960483, + "grad_norm": 3.681009531021118, + "learning_rate": 9.990649999065e-06, + "loss": 0.1234, + "step": 9356 + }, + { + "epoch": 4.350069735006974, + "grad_norm": 5.912764549255371, + "learning_rate": 9.9906489990649e-06, + "loss": 0.6027, + "step": 9357 + }, + { + "epoch": 4.350534635053464, + "grad_norm": 4.522889137268066, + "learning_rate": 9.990647999064801e-06, + "loss": 0.1538, + "step": 9358 + }, + { + "epoch": 4.3509995350999535, + "grad_norm": 5.12624979019165, + "learning_rate": 9.9906469990647e-06, + "loss": 0.2646, + "step": 9359 + }, + { + "epoch": 4.351464435146443, + "grad_norm": 4.0559210777282715, + "learning_rate": 9.990645999064601e-06, + "loss": 0.1972, + "step": 9360 + }, + { + "epoch": 4.351929335192933, + "grad_norm": 3.193959951400757, + "learning_rate": 9.9906449990645e-06, + "loss": 0.1026, + "step": 9361 + }, + { + "epoch": 4.352394235239424, + "grad_norm": 3.731513500213623, + "learning_rate": 9.9906439990644e-06, + "loss": 0.2121, + "step": 9362 + }, + { + "epoch": 4.352859135285914, + "grad_norm": 7.739988327026367, + "learning_rate": 9.990642999064301e-06, + "loss": 0.6465, + "step": 9363 + }, + { + "epoch": 4.353324035332403, + "grad_norm": 5.447793483734131, + "learning_rate": 9.990641999064202e-06, + "loss": 0.1563, + "step": 9364 + }, + { + "epoch": 4.353788935378893, + "grad_norm": 5.112318992614746, + "learning_rate": 9.9906409990641e-06, + "loss": 0.3676, + "step": 9365 + }, + { + "epoch": 4.354253835425384, + "grad_norm": 8.899246215820312, + "learning_rate": 9.990639999064e-06, + "loss": 0.5243, + "step": 9366 + }, + { + "epoch": 4.354718735471874, + "grad_norm": 6.975677490234375, + "learning_rate": 9.9906389990639e-06, + "loss": 0.3936, + "step": 9367 + }, + { + "epoch": 4.3551836355183635, + "grad_norm": 4.047145366668701, + "learning_rate": 9.990637999063801e-06, + "loss": 0.2437, + "step": 9368 + }, + { + "epoch": 4.355648535564853, + "grad_norm": 6.360062599182129, + "learning_rate": 9.990636999063702e-06, + "loss": 0.2865, + "step": 9369 + }, + { + "epoch": 4.356113435611343, + "grad_norm": 4.206220626831055, + "learning_rate": 9.9906359990636e-06, + "loss": 0.1982, + "step": 9370 + }, + { + "epoch": 4.356578335657834, + "grad_norm": 5.417168140411377, + "learning_rate": 9.990634999063501e-06, + "loss": 0.3569, + "step": 9371 + }, + { + "epoch": 4.357043235704324, + "grad_norm": 6.508734703063965, + "learning_rate": 9.9906339990634e-06, + "loss": 0.3989, + "step": 9372 + }, + { + "epoch": 4.357508135750813, + "grad_norm": 6.738999843597412, + "learning_rate": 9.990632999063301e-06, + "loss": 0.3308, + "step": 9373 + }, + { + "epoch": 4.357973035797303, + "grad_norm": 7.314848899841309, + "learning_rate": 9.9906319990632e-06, + "loss": 0.5191, + "step": 9374 + }, + { + "epoch": 4.358437935843794, + "grad_norm": 7.2942681312561035, + "learning_rate": 9.9906309990631e-06, + "loss": 0.3248, + "step": 9375 + }, + { + "epoch": 4.358902835890284, + "grad_norm": 4.8715386390686035, + "learning_rate": 9.990629999063001e-06, + "loss": 0.3605, + "step": 9376 + }, + { + "epoch": 4.3593677359367735, + "grad_norm": 6.162871837615967, + "learning_rate": 9.9906289990629e-06, + "loss": 0.373, + "step": 9377 + }, + { + "epoch": 4.359832635983263, + "grad_norm": 3.610119342803955, + "learning_rate": 9.990627999062801e-06, + "loss": 0.2046, + "step": 9378 + }, + { + "epoch": 4.360297536029754, + "grad_norm": 4.25026798248291, + "learning_rate": 9.9906269990627e-06, + "loss": 0.2016, + "step": 9379 + }, + { + "epoch": 4.360762436076244, + "grad_norm": 8.606371879577637, + "learning_rate": 9.9906259990626e-06, + "loss": 0.4641, + "step": 9380 + }, + { + "epoch": 4.361227336122734, + "grad_norm": 4.705204963684082, + "learning_rate": 9.990624999062501e-06, + "loss": 0.2529, + "step": 9381 + }, + { + "epoch": 4.361692236169223, + "grad_norm": 10.19314956665039, + "learning_rate": 9.990623999062402e-06, + "loss": 0.49, + "step": 9382 + }, + { + "epoch": 4.362157136215713, + "grad_norm": 5.132131099700928, + "learning_rate": 9.9906229990623e-06, + "loss": 0.2632, + "step": 9383 + }, + { + "epoch": 4.362622036262204, + "grad_norm": 5.047766208648682, + "learning_rate": 9.990621999062201e-06, + "loss": 0.3195, + "step": 9384 + }, + { + "epoch": 4.363086936308694, + "grad_norm": 6.451709747314453, + "learning_rate": 9.9906209990621e-06, + "loss": 0.339, + "step": 9385 + }, + { + "epoch": 4.3635518363551835, + "grad_norm": 8.89260482788086, + "learning_rate": 9.990619999062001e-06, + "loss": 0.4746, + "step": 9386 + }, + { + "epoch": 4.364016736401673, + "grad_norm": 4.524385929107666, + "learning_rate": 9.9906189990619e-06, + "loss": 0.2208, + "step": 9387 + }, + { + "epoch": 4.364481636448164, + "grad_norm": 5.924479961395264, + "learning_rate": 9.9906179990618e-06, + "loss": 0.3381, + "step": 9388 + }, + { + "epoch": 4.364946536494654, + "grad_norm": 7.361903667449951, + "learning_rate": 9.990616999061701e-06, + "loss": 0.342, + "step": 9389 + }, + { + "epoch": 4.365411436541144, + "grad_norm": 4.126504421234131, + "learning_rate": 9.9906159990616e-06, + "loss": 0.203, + "step": 9390 + }, + { + "epoch": 4.365876336587633, + "grad_norm": 4.928638458251953, + "learning_rate": 9.990614999061501e-06, + "loss": 0.3872, + "step": 9391 + }, + { + "epoch": 4.366341236634124, + "grad_norm": 3.2656033039093018, + "learning_rate": 9.9906139990614e-06, + "loss": 0.097, + "step": 9392 + }, + { + "epoch": 4.366806136680614, + "grad_norm": 3.5415115356445312, + "learning_rate": 9.9906129990613e-06, + "loss": 0.2195, + "step": 9393 + }, + { + "epoch": 4.367271036727104, + "grad_norm": 8.442273139953613, + "learning_rate": 9.990611999061201e-06, + "loss": 0.3395, + "step": 9394 + }, + { + "epoch": 4.3677359367735935, + "grad_norm": 5.286134719848633, + "learning_rate": 9.990610999061102e-06, + "loss": 0.3012, + "step": 9395 + }, + { + "epoch": 4.368200836820083, + "grad_norm": 5.363455295562744, + "learning_rate": 9.990609999061e-06, + "loss": 0.27, + "step": 9396 + }, + { + "epoch": 4.368665736866574, + "grad_norm": 4.240932941436768, + "learning_rate": 9.990608999060901e-06, + "loss": 0.2273, + "step": 9397 + }, + { + "epoch": 4.369130636913064, + "grad_norm": 2.7542974948883057, + "learning_rate": 9.9906079990608e-06, + "loss": 0.0661, + "step": 9398 + }, + { + "epoch": 4.369595536959554, + "grad_norm": 4.479103088378906, + "learning_rate": 9.9906069990607e-06, + "loss": 0.2236, + "step": 9399 + }, + { + "epoch": 4.370060437006043, + "grad_norm": 5.758334636688232, + "learning_rate": 9.9906059990606e-06, + "loss": 0.2707, + "step": 9400 + }, + { + "epoch": 4.370525337052534, + "grad_norm": 8.513476371765137, + "learning_rate": 9.9906049990605e-06, + "loss": 0.5767, + "step": 9401 + }, + { + "epoch": 4.370990237099024, + "grad_norm": 3.433082342147827, + "learning_rate": 9.990603999060401e-06, + "loss": 0.131, + "step": 9402 + }, + { + "epoch": 4.371455137145514, + "grad_norm": 4.371907711029053, + "learning_rate": 9.9906029990603e-06, + "loss": 0.2635, + "step": 9403 + }, + { + "epoch": 4.3719200371920035, + "grad_norm": 4.420289993286133, + "learning_rate": 9.990601999060201e-06, + "loss": 0.167, + "step": 9404 + }, + { + "epoch": 4.372384937238493, + "grad_norm": 5.996499061584473, + "learning_rate": 9.9906009990601e-06, + "loss": 0.3925, + "step": 9405 + }, + { + "epoch": 4.372849837284984, + "grad_norm": 4.734436988830566, + "learning_rate": 9.99059999906e-06, + "loss": 0.1607, + "step": 9406 + }, + { + "epoch": 4.373314737331474, + "grad_norm": 4.437458515167236, + "learning_rate": 9.990598999059901e-06, + "loss": 0.245, + "step": 9407 + }, + { + "epoch": 4.373779637377964, + "grad_norm": 4.833254337310791, + "learning_rate": 9.990597999059802e-06, + "loss": 0.1756, + "step": 9408 + }, + { + "epoch": 4.374244537424453, + "grad_norm": 4.666719436645508, + "learning_rate": 9.990596999059701e-06, + "loss": 0.4279, + "step": 9409 + }, + { + "epoch": 4.374709437470944, + "grad_norm": 4.09827184677124, + "learning_rate": 9.9905959990596e-06, + "loss": 0.1305, + "step": 9410 + }, + { + "epoch": 4.375174337517434, + "grad_norm": 3.019618511199951, + "learning_rate": 9.9905949990595e-06, + "loss": 0.0739, + "step": 9411 + }, + { + "epoch": 4.375639237563924, + "grad_norm": 3.7535433769226074, + "learning_rate": 9.9905939990594e-06, + "loss": 0.1727, + "step": 9412 + }, + { + "epoch": 4.3761041376104135, + "grad_norm": 4.007974624633789, + "learning_rate": 9.9905929990593e-06, + "loss": 0.2004, + "step": 9413 + }, + { + "epoch": 4.376569037656903, + "grad_norm": 4.2427215576171875, + "learning_rate": 9.9905919990592e-06, + "loss": 0.2727, + "step": 9414 + }, + { + "epoch": 4.377033937703394, + "grad_norm": 6.285948276519775, + "learning_rate": 9.990590999059101e-06, + "loss": 0.3239, + "step": 9415 + }, + { + "epoch": 4.377498837749884, + "grad_norm": 3.9789092540740967, + "learning_rate": 9.990589999059e-06, + "loss": 0.2488, + "step": 9416 + }, + { + "epoch": 4.377963737796374, + "grad_norm": 4.26491117477417, + "learning_rate": 9.990588999058901e-06, + "loss": 0.2565, + "step": 9417 + }, + { + "epoch": 4.378428637842863, + "grad_norm": 7.31703519821167, + "learning_rate": 9.9905879990588e-06, + "loss": 0.4454, + "step": 9418 + }, + { + "epoch": 4.378893537889354, + "grad_norm": 11.300634384155273, + "learning_rate": 9.9905869990587e-06, + "loss": 0.2233, + "step": 9419 + }, + { + "epoch": 4.379358437935844, + "grad_norm": 6.285191535949707, + "learning_rate": 9.990585999058601e-06, + "loss": 0.5016, + "step": 9420 + }, + { + "epoch": 4.379823337982334, + "grad_norm": 2.3894805908203125, + "learning_rate": 9.9905849990585e-06, + "loss": 0.0812, + "step": 9421 + }, + { + "epoch": 4.3802882380288235, + "grad_norm": 4.662092208862305, + "learning_rate": 9.990583999058401e-06, + "loss": 0.1827, + "step": 9422 + }, + { + "epoch": 4.380753138075314, + "grad_norm": 4.452382564544678, + "learning_rate": 9.9905829990583e-06, + "loss": 0.1649, + "step": 9423 + }, + { + "epoch": 4.381218038121804, + "grad_norm": 4.727802753448486, + "learning_rate": 9.9905819990582e-06, + "loss": 0.1585, + "step": 9424 + }, + { + "epoch": 4.381682938168294, + "grad_norm": 4.297463417053223, + "learning_rate": 9.9905809990581e-06, + "loss": 0.2322, + "step": 9425 + }, + { + "epoch": 4.382147838214784, + "grad_norm": 4.979161262512207, + "learning_rate": 9.990579999058e-06, + "loss": 0.2003, + "step": 9426 + }, + { + "epoch": 4.382612738261274, + "grad_norm": 4.801449775695801, + "learning_rate": 9.990578999057901e-06, + "loss": 0.2607, + "step": 9427 + }, + { + "epoch": 4.383077638307764, + "grad_norm": 3.7941131591796875, + "learning_rate": 9.990577999057802e-06, + "loss": 0.1651, + "step": 9428 + }, + { + "epoch": 4.383542538354254, + "grad_norm": 4.639232158660889, + "learning_rate": 9.9905769990577e-06, + "loss": 0.241, + "step": 9429 + }, + { + "epoch": 4.384007438400744, + "grad_norm": 9.987279891967773, + "learning_rate": 9.990575999057601e-06, + "loss": 0.2227, + "step": 9430 + }, + { + "epoch": 4.3844723384472335, + "grad_norm": 8.727543830871582, + "learning_rate": 9.9905749990575e-06, + "loss": 0.4192, + "step": 9431 + }, + { + "epoch": 4.384937238493724, + "grad_norm": 4.297917366027832, + "learning_rate": 9.9905739990574e-06, + "loss": 0.1117, + "step": 9432 + }, + { + "epoch": 4.385402138540214, + "grad_norm": 4.389893531799316, + "learning_rate": 9.990572999057301e-06, + "loss": 0.273, + "step": 9433 + }, + { + "epoch": 4.385867038586704, + "grad_norm": 4.805945873260498, + "learning_rate": 9.9905719990572e-06, + "loss": 0.315, + "step": 9434 + }, + { + "epoch": 4.386331938633194, + "grad_norm": 6.49200963973999, + "learning_rate": 9.990570999057101e-06, + "loss": 0.5029, + "step": 9435 + }, + { + "epoch": 4.386796838679684, + "grad_norm": 7.841855049133301, + "learning_rate": 9.990569999057e-06, + "loss": 0.5306, + "step": 9436 + }, + { + "epoch": 4.387261738726174, + "grad_norm": 5.69293737411499, + "learning_rate": 9.9905689990569e-06, + "loss": 0.4033, + "step": 9437 + }, + { + "epoch": 4.387726638772664, + "grad_norm": 3.7381086349487305, + "learning_rate": 9.990567999056801e-06, + "loss": 0.182, + "step": 9438 + }, + { + "epoch": 4.388191538819154, + "grad_norm": 4.657268524169922, + "learning_rate": 9.9905669990567e-06, + "loss": 0.1517, + "step": 9439 + }, + { + "epoch": 4.3886564388656435, + "grad_norm": 6.010471343994141, + "learning_rate": 9.990565999056601e-06, + "loss": 0.2951, + "step": 9440 + }, + { + "epoch": 4.389121338912134, + "grad_norm": 5.338073253631592, + "learning_rate": 9.990564999056502e-06, + "loss": 0.3254, + "step": 9441 + }, + { + "epoch": 4.389586238958624, + "grad_norm": 5.618577003479004, + "learning_rate": 9.9905639990564e-06, + "loss": 0.3696, + "step": 9442 + }, + { + "epoch": 4.390051139005114, + "grad_norm": 6.162380695343018, + "learning_rate": 9.9905629990563e-06, + "loss": 0.2391, + "step": 9443 + }, + { + "epoch": 4.390516039051604, + "grad_norm": 6.071844100952148, + "learning_rate": 9.9905619990562e-06, + "loss": 0.3198, + "step": 9444 + }, + { + "epoch": 4.390980939098094, + "grad_norm": 5.471048355102539, + "learning_rate": 9.990560999056101e-06, + "loss": 0.5063, + "step": 9445 + }, + { + "epoch": 4.391445839144584, + "grad_norm": 5.88077974319458, + "learning_rate": 9.990559999056002e-06, + "loss": 0.3003, + "step": 9446 + }, + { + "epoch": 4.391910739191074, + "grad_norm": 7.853032112121582, + "learning_rate": 9.9905589990559e-06, + "loss": 0.266, + "step": 9447 + }, + { + "epoch": 4.392375639237564, + "grad_norm": 4.462108135223389, + "learning_rate": 9.990557999055801e-06, + "loss": 0.1949, + "step": 9448 + }, + { + "epoch": 4.3928405392840535, + "grad_norm": 5.120703220367432, + "learning_rate": 9.9905569990557e-06, + "loss": 0.2582, + "step": 9449 + }, + { + "epoch": 4.393305439330544, + "grad_norm": 6.66316556930542, + "learning_rate": 9.9905559990556e-06, + "loss": 0.3626, + "step": 9450 + }, + { + "epoch": 4.393770339377034, + "grad_norm": 5.563345909118652, + "learning_rate": 9.990554999055501e-06, + "loss": 0.3854, + "step": 9451 + }, + { + "epoch": 4.394235239423524, + "grad_norm": 5.98907470703125, + "learning_rate": 9.9905539990554e-06, + "loss": 0.2779, + "step": 9452 + }, + { + "epoch": 4.394700139470014, + "grad_norm": 7.043375492095947, + "learning_rate": 9.990552999055301e-06, + "loss": 0.3958, + "step": 9453 + }, + { + "epoch": 4.395165039516504, + "grad_norm": 8.110495567321777, + "learning_rate": 9.9905519990552e-06, + "loss": 0.3556, + "step": 9454 + }, + { + "epoch": 4.395629939562994, + "grad_norm": 5.840389728546143, + "learning_rate": 9.9905509990551e-06, + "loss": 0.3132, + "step": 9455 + }, + { + "epoch": 4.396094839609484, + "grad_norm": 4.020063400268555, + "learning_rate": 9.990549999055e-06, + "loss": 0.1796, + "step": 9456 + }, + { + "epoch": 4.396559739655974, + "grad_norm": 4.972419738769531, + "learning_rate": 9.9905489990549e-06, + "loss": 0.1822, + "step": 9457 + }, + { + "epoch": 4.397024639702464, + "grad_norm": 6.910674095153809, + "learning_rate": 9.990547999054801e-06, + "loss": 0.2634, + "step": 9458 + }, + { + "epoch": 4.397489539748954, + "grad_norm": 3.9454989433288574, + "learning_rate": 9.990546999054702e-06, + "loss": 0.1862, + "step": 9459 + }, + { + "epoch": 4.397954439795444, + "grad_norm": 5.614216327667236, + "learning_rate": 9.9905459990546e-06, + "loss": 0.3678, + "step": 9460 + }, + { + "epoch": 4.398419339841934, + "grad_norm": 5.6874775886535645, + "learning_rate": 9.990544999054501e-06, + "loss": 0.425, + "step": 9461 + }, + { + "epoch": 4.398884239888424, + "grad_norm": 6.73253870010376, + "learning_rate": 9.9905439990544e-06, + "loss": 0.3094, + "step": 9462 + }, + { + "epoch": 4.399349139934914, + "grad_norm": 5.554123878479004, + "learning_rate": 9.990542999054301e-06, + "loss": 0.1264, + "step": 9463 + }, + { + "epoch": 4.399814039981404, + "grad_norm": 6.379730701446533, + "learning_rate": 9.990541999054202e-06, + "loss": 0.2514, + "step": 9464 + }, + { + "epoch": 4.400278940027894, + "grad_norm": 7.299145698547363, + "learning_rate": 9.9905409990541e-06, + "loss": 0.2578, + "step": 9465 + }, + { + "epoch": 4.400743840074384, + "grad_norm": 8.711135864257812, + "learning_rate": 9.990539999054001e-06, + "loss": 0.2926, + "step": 9466 + }, + { + "epoch": 4.401208740120874, + "grad_norm": 6.469375133514404, + "learning_rate": 9.9905389990539e-06, + "loss": 0.3203, + "step": 9467 + }, + { + "epoch": 4.401673640167364, + "grad_norm": 3.598311185836792, + "learning_rate": 9.9905379990538e-06, + "loss": 0.1674, + "step": 9468 + }, + { + "epoch": 4.402138540213854, + "grad_norm": 5.037461280822754, + "learning_rate": 9.9905369990537e-06, + "loss": 0.2674, + "step": 9469 + }, + { + "epoch": 4.402603440260344, + "grad_norm": 4.523031711578369, + "learning_rate": 9.9905359990536e-06, + "loss": 0.2834, + "step": 9470 + }, + { + "epoch": 4.4030683403068345, + "grad_norm": 4.388763427734375, + "learning_rate": 9.990534999053501e-06, + "loss": 0.2649, + "step": 9471 + }, + { + "epoch": 4.403533240353324, + "grad_norm": 5.234048843383789, + "learning_rate": 9.990533999053402e-06, + "loss": 0.3406, + "step": 9472 + }, + { + "epoch": 4.403998140399814, + "grad_norm": 3.518911600112915, + "learning_rate": 9.9905329990533e-06, + "loss": 0.1604, + "step": 9473 + }, + { + "epoch": 4.404463040446304, + "grad_norm": 6.618965148925781, + "learning_rate": 9.990531999053201e-06, + "loss": 0.319, + "step": 9474 + }, + { + "epoch": 4.404927940492794, + "grad_norm": 8.537836074829102, + "learning_rate": 9.9905309990531e-06, + "loss": 0.4015, + "step": 9475 + }, + { + "epoch": 4.405392840539284, + "grad_norm": 5.966643333435059, + "learning_rate": 9.990529999053001e-06, + "loss": 0.4278, + "step": 9476 + }, + { + "epoch": 4.405857740585774, + "grad_norm": 11.175469398498535, + "learning_rate": 9.990528999052902e-06, + "loss": 0.4154, + "step": 9477 + }, + { + "epoch": 4.406322640632264, + "grad_norm": 4.804361820220947, + "learning_rate": 9.9905279990528e-06, + "loss": 0.2729, + "step": 9478 + }, + { + "epoch": 4.406787540678754, + "grad_norm": 7.961976051330566, + "learning_rate": 9.990526999052701e-06, + "loss": 0.4369, + "step": 9479 + }, + { + "epoch": 4.4072524407252445, + "grad_norm": 9.583279609680176, + "learning_rate": 9.9905259990526e-06, + "loss": 0.3269, + "step": 9480 + }, + { + "epoch": 4.407717340771734, + "grad_norm": 9.539289474487305, + "learning_rate": 9.9905249990525e-06, + "loss": 0.5418, + "step": 9481 + }, + { + "epoch": 4.408182240818224, + "grad_norm": 8.643261909484863, + "learning_rate": 9.9905239990524e-06, + "loss": 0.2497, + "step": 9482 + }, + { + "epoch": 4.408647140864714, + "grad_norm": 6.665125370025635, + "learning_rate": 9.9905229990523e-06, + "loss": 0.5001, + "step": 9483 + }, + { + "epoch": 4.409112040911204, + "grad_norm": 4.462886810302734, + "learning_rate": 9.990521999052201e-06, + "loss": 0.2139, + "step": 9484 + }, + { + "epoch": 4.409576940957694, + "grad_norm": 8.421347618103027, + "learning_rate": 9.990520999052102e-06, + "loss": 0.3383, + "step": 9485 + }, + { + "epoch": 4.410041841004184, + "grad_norm": 2.0724403858184814, + "learning_rate": 9.990519999052e-06, + "loss": 0.0578, + "step": 9486 + }, + { + "epoch": 4.410506741050674, + "grad_norm": 6.092759609222412, + "learning_rate": 9.9905189990519e-06, + "loss": 0.4387, + "step": 9487 + }, + { + "epoch": 4.410971641097164, + "grad_norm": 3.1315386295318604, + "learning_rate": 9.9905179990518e-06, + "loss": 0.1054, + "step": 9488 + }, + { + "epoch": 4.4114365411436545, + "grad_norm": 6.856549263000488, + "learning_rate": 9.990516999051701e-06, + "loss": 0.657, + "step": 9489 + }, + { + "epoch": 4.411901441190144, + "grad_norm": 6.579289436340332, + "learning_rate": 9.990515999051602e-06, + "loss": 0.4226, + "step": 9490 + }, + { + "epoch": 4.412366341236634, + "grad_norm": 6.476588726043701, + "learning_rate": 9.9905149990515e-06, + "loss": 0.4084, + "step": 9491 + }, + { + "epoch": 4.412831241283124, + "grad_norm": 6.765704154968262, + "learning_rate": 9.990513999051401e-06, + "loss": 0.3111, + "step": 9492 + }, + { + "epoch": 4.413296141329614, + "grad_norm": 5.718696594238281, + "learning_rate": 9.9905129990513e-06, + "loss": 0.1211, + "step": 9493 + }, + { + "epoch": 4.413761041376104, + "grad_norm": 4.952746391296387, + "learning_rate": 9.990511999051201e-06, + "loss": 0.1374, + "step": 9494 + }, + { + "epoch": 4.414225941422594, + "grad_norm": 9.237079620361328, + "learning_rate": 9.9905109990511e-06, + "loss": 0.4792, + "step": 9495 + }, + { + "epoch": 4.414690841469084, + "grad_norm": 5.9244232177734375, + "learning_rate": 9.990509999051e-06, + "loss": 0.2768, + "step": 9496 + }, + { + "epoch": 4.415155741515574, + "grad_norm": 5.778731346130371, + "learning_rate": 9.990508999050901e-06, + "loss": 0.2255, + "step": 9497 + }, + { + "epoch": 4.4156206415620645, + "grad_norm": 4.911518573760986, + "learning_rate": 9.9905079990508e-06, + "loss": 0.1835, + "step": 9498 + }, + { + "epoch": 4.416085541608554, + "grad_norm": 6.2037458419799805, + "learning_rate": 9.9905069990507e-06, + "loss": 0.5264, + "step": 9499 + }, + { + "epoch": 4.416550441655044, + "grad_norm": 5.8498921394348145, + "learning_rate": 9.9905059990506e-06, + "loss": 0.5912, + "step": 9500 + }, + { + "epoch": 4.417015341701534, + "grad_norm": 4.443543910980225, + "learning_rate": 9.9905049990505e-06, + "loss": 0.355, + "step": 9501 + }, + { + "epoch": 4.417480241748025, + "grad_norm": 4.56294584274292, + "learning_rate": 9.990503999050401e-06, + "loss": 0.2509, + "step": 9502 + }, + { + "epoch": 4.417945141794514, + "grad_norm": 4.14104700088501, + "learning_rate": 9.990502999050302e-06, + "loss": 0.1587, + "step": 9503 + }, + { + "epoch": 4.418410041841004, + "grad_norm": 5.2696614265441895, + "learning_rate": 9.9905019990502e-06, + "loss": 0.2084, + "step": 9504 + }, + { + "epoch": 4.418874941887494, + "grad_norm": 5.3362555503845215, + "learning_rate": 9.990500999050101e-06, + "loss": 0.2747, + "step": 9505 + }, + { + "epoch": 4.419339841933985, + "grad_norm": 4.280484199523926, + "learning_rate": 9.99049999905e-06, + "loss": 0.2241, + "step": 9506 + }, + { + "epoch": 4.4198047419804745, + "grad_norm": 9.916972160339355, + "learning_rate": 9.990498999049901e-06, + "loss": 0.5118, + "step": 9507 + }, + { + "epoch": 4.420269642026964, + "grad_norm": 6.613603591918945, + "learning_rate": 9.9904979990498e-06, + "loss": 0.2407, + "step": 9508 + }, + { + "epoch": 4.420734542073454, + "grad_norm": 5.786176681518555, + "learning_rate": 9.9904969990497e-06, + "loss": 0.3462, + "step": 9509 + }, + { + "epoch": 4.421199442119944, + "grad_norm": 5.378159999847412, + "learning_rate": 9.990495999049601e-06, + "loss": 0.3295, + "step": 9510 + }, + { + "epoch": 4.421664342166435, + "grad_norm": 9.562945365905762, + "learning_rate": 9.9904949990495e-06, + "loss": 0.7532, + "step": 9511 + }, + { + "epoch": 4.422129242212924, + "grad_norm": 11.025872230529785, + "learning_rate": 9.990493999049401e-06, + "loss": 0.373, + "step": 9512 + }, + { + "epoch": 4.422594142259414, + "grad_norm": 3.8801424503326416, + "learning_rate": 9.9904929990493e-06, + "loss": 0.1145, + "step": 9513 + }, + { + "epoch": 4.423059042305904, + "grad_norm": 3.1304333209991455, + "learning_rate": 9.9904919990492e-06, + "loss": 0.1692, + "step": 9514 + }, + { + "epoch": 4.423523942352395, + "grad_norm": 5.663626194000244, + "learning_rate": 9.990490999049101e-06, + "loss": 0.3086, + "step": 9515 + }, + { + "epoch": 4.4239888423988845, + "grad_norm": 6.723112106323242, + "learning_rate": 9.990489999049002e-06, + "loss": 0.4277, + "step": 9516 + }, + { + "epoch": 4.424453742445374, + "grad_norm": 4.236299514770508, + "learning_rate": 9.9904889990489e-06, + "loss": 0.1424, + "step": 9517 + }, + { + "epoch": 4.424918642491864, + "grad_norm": 4.9631571769714355, + "learning_rate": 9.9904879990488e-06, + "loss": 0.4112, + "step": 9518 + }, + { + "epoch": 4.425383542538354, + "grad_norm": 6.380121231079102, + "learning_rate": 9.9904869990487e-06, + "loss": 0.3123, + "step": 9519 + }, + { + "epoch": 4.4258484425848446, + "grad_norm": 7.2036566734313965, + "learning_rate": 9.990485999048601e-06, + "loss": 0.2941, + "step": 9520 + }, + { + "epoch": 4.426313342631334, + "grad_norm": 3.822943687438965, + "learning_rate": 9.9904849990485e-06, + "loss": 0.2731, + "step": 9521 + }, + { + "epoch": 4.426778242677824, + "grad_norm": 4.453098297119141, + "learning_rate": 9.9904839990484e-06, + "loss": 0.2005, + "step": 9522 + }, + { + "epoch": 4.427243142724314, + "grad_norm": 4.251066207885742, + "learning_rate": 9.990482999048301e-06, + "loss": 0.2665, + "step": 9523 + }, + { + "epoch": 4.427708042770805, + "grad_norm": 5.449945449829102, + "learning_rate": 9.9904819990482e-06, + "loss": 0.283, + "step": 9524 + }, + { + "epoch": 4.4281729428172945, + "grad_norm": 6.2776875495910645, + "learning_rate": 9.990480999048101e-06, + "loss": 0.3766, + "step": 9525 + }, + { + "epoch": 4.428637842863784, + "grad_norm": 7.0520830154418945, + "learning_rate": 9.990479999048e-06, + "loss": 0.2729, + "step": 9526 + }, + { + "epoch": 4.429102742910274, + "grad_norm": 4.289215087890625, + "learning_rate": 9.9904789990479e-06, + "loss": 0.1754, + "step": 9527 + }, + { + "epoch": 4.429567642956764, + "grad_norm": 7.705321788787842, + "learning_rate": 9.990477999047801e-06, + "loss": 0.3127, + "step": 9528 + }, + { + "epoch": 4.4300325430032546, + "grad_norm": 3.4477007389068604, + "learning_rate": 9.9904769990477e-06, + "loss": 0.0826, + "step": 9529 + }, + { + "epoch": 4.430497443049744, + "grad_norm": 4.812204360961914, + "learning_rate": 9.990475999047601e-06, + "loss": 0.2319, + "step": 9530 + }, + { + "epoch": 4.430962343096234, + "grad_norm": 8.141561508178711, + "learning_rate": 9.9904749990475e-06, + "loss": 0.3174, + "step": 9531 + }, + { + "epoch": 4.431427243142724, + "grad_norm": 11.05268669128418, + "learning_rate": 9.9904739990474e-06, + "loss": 0.388, + "step": 9532 + }, + { + "epoch": 4.431892143189215, + "grad_norm": 6.499638080596924, + "learning_rate": 9.990472999047301e-06, + "loss": 0.3626, + "step": 9533 + }, + { + "epoch": 4.4323570432357045, + "grad_norm": 5.827358722686768, + "learning_rate": 9.990471999047202e-06, + "loss": 0.3984, + "step": 9534 + }, + { + "epoch": 4.432821943282194, + "grad_norm": 4.755474090576172, + "learning_rate": 9.9904709990471e-06, + "loss": 0.1683, + "step": 9535 + }, + { + "epoch": 4.433286843328684, + "grad_norm": 5.132297515869141, + "learning_rate": 9.990469999047001e-06, + "loss": 0.3195, + "step": 9536 + }, + { + "epoch": 4.433751743375175, + "grad_norm": 3.286807060241699, + "learning_rate": 9.9904689990469e-06, + "loss": 0.085, + "step": 9537 + }, + { + "epoch": 4.4342166434216645, + "grad_norm": 5.878546237945557, + "learning_rate": 9.990467999046801e-06, + "loss": 0.2321, + "step": 9538 + }, + { + "epoch": 4.434681543468154, + "grad_norm": 7.278792858123779, + "learning_rate": 9.9904669990467e-06, + "loss": 0.3436, + "step": 9539 + }, + { + "epoch": 4.435146443514644, + "grad_norm": 11.31108283996582, + "learning_rate": 9.9904659990466e-06, + "loss": 0.336, + "step": 9540 + }, + { + "epoch": 4.435611343561134, + "grad_norm": 3.776059865951538, + "learning_rate": 9.990464999046501e-06, + "loss": 0.1227, + "step": 9541 + }, + { + "epoch": 4.436076243607625, + "grad_norm": 5.3172078132629395, + "learning_rate": 9.9904639990464e-06, + "loss": 0.3249, + "step": 9542 + }, + { + "epoch": 4.4365411436541144, + "grad_norm": 5.287046432495117, + "learning_rate": 9.990462999046301e-06, + "loss": 0.3738, + "step": 9543 + }, + { + "epoch": 4.437006043700604, + "grad_norm": 3.956681966781616, + "learning_rate": 9.9904619990462e-06, + "loss": 0.1846, + "step": 9544 + }, + { + "epoch": 4.437470943747094, + "grad_norm": 5.795361042022705, + "learning_rate": 9.9904609990461e-06, + "loss": 0.3161, + "step": 9545 + }, + { + "epoch": 4.437935843793585, + "grad_norm": 5.852274417877197, + "learning_rate": 9.990459999046001e-06, + "loss": 0.2198, + "step": 9546 + }, + { + "epoch": 4.4384007438400745, + "grad_norm": 5.983644962310791, + "learning_rate": 9.990458999045902e-06, + "loss": 0.2668, + "step": 9547 + }, + { + "epoch": 4.438865643886564, + "grad_norm": 4.042856216430664, + "learning_rate": 9.990457999045801e-06, + "loss": 0.0883, + "step": 9548 + }, + { + "epoch": 4.439330543933054, + "grad_norm": 5.200505256652832, + "learning_rate": 9.990456999045702e-06, + "loss": 0.24, + "step": 9549 + }, + { + "epoch": 4.439795443979545, + "grad_norm": 5.055543422698975, + "learning_rate": 9.9904559990456e-06, + "loss": 0.3074, + "step": 9550 + }, + { + "epoch": 4.440260344026035, + "grad_norm": 7.484160423278809, + "learning_rate": 9.9904549990455e-06, + "loss": 0.4254, + "step": 9551 + }, + { + "epoch": 4.4407252440725244, + "grad_norm": 8.574478149414062, + "learning_rate": 9.9904539990454e-06, + "loss": 0.9775, + "step": 9552 + }, + { + "epoch": 4.441190144119014, + "grad_norm": 6.1918864250183105, + "learning_rate": 9.9904529990453e-06, + "loss": 0.3323, + "step": 9553 + }, + { + "epoch": 4.441655044165504, + "grad_norm": 4.672555923461914, + "learning_rate": 9.990451999045201e-06, + "loss": 0.3358, + "step": 9554 + }, + { + "epoch": 4.442119944211995, + "grad_norm": 4.280801296234131, + "learning_rate": 9.9904509990451e-06, + "loss": 0.3202, + "step": 9555 + }, + { + "epoch": 4.4425848442584845, + "grad_norm": 4.0252461433410645, + "learning_rate": 9.990449999045001e-06, + "loss": 0.2002, + "step": 9556 + }, + { + "epoch": 4.443049744304974, + "grad_norm": 4.265021800994873, + "learning_rate": 9.9904489990449e-06, + "loss": 0.2684, + "step": 9557 + }, + { + "epoch": 4.443514644351464, + "grad_norm": 4.88045597076416, + "learning_rate": 9.9904479990448e-06, + "loss": 0.2038, + "step": 9558 + }, + { + "epoch": 4.443979544397955, + "grad_norm": 4.873003959655762, + "learning_rate": 9.990446999044701e-06, + "loss": 0.1933, + "step": 9559 + }, + { + "epoch": 4.444444444444445, + "grad_norm": 3.363645553588867, + "learning_rate": 9.990445999044602e-06, + "loss": 0.1226, + "step": 9560 + }, + { + "epoch": 4.4449093444909344, + "grad_norm": 5.755237579345703, + "learning_rate": 9.990444999044501e-06, + "loss": 0.4317, + "step": 9561 + }, + { + "epoch": 4.445374244537424, + "grad_norm": 6.437377452850342, + "learning_rate": 9.9904439990444e-06, + "loss": 0.4205, + "step": 9562 + }, + { + "epoch": 4.445839144583914, + "grad_norm": 5.237550258636475, + "learning_rate": 9.9904429990443e-06, + "loss": 0.3365, + "step": 9563 + }, + { + "epoch": 4.446304044630405, + "grad_norm": 4.679872512817383, + "learning_rate": 9.9904419990442e-06, + "loss": 0.221, + "step": 9564 + }, + { + "epoch": 4.4467689446768945, + "grad_norm": 4.2989277839660645, + "learning_rate": 9.9904409990441e-06, + "loss": 0.311, + "step": 9565 + }, + { + "epoch": 4.447233844723384, + "grad_norm": 10.141091346740723, + "learning_rate": 9.990439999044001e-06, + "loss": 0.3009, + "step": 9566 + }, + { + "epoch": 4.447698744769874, + "grad_norm": 6.7296271324157715, + "learning_rate": 9.990438999043902e-06, + "loss": 0.239, + "step": 9567 + }, + { + "epoch": 4.448163644816365, + "grad_norm": 5.502957820892334, + "learning_rate": 9.9904379990438e-06, + "loss": 0.3077, + "step": 9568 + }, + { + "epoch": 4.448628544862855, + "grad_norm": 4.042819976806641, + "learning_rate": 9.990436999043701e-06, + "loss": 0.2057, + "step": 9569 + }, + { + "epoch": 4.449093444909344, + "grad_norm": 4.889708518981934, + "learning_rate": 9.9904359990436e-06, + "loss": 0.2748, + "step": 9570 + }, + { + "epoch": 4.449558344955834, + "grad_norm": 5.754184246063232, + "learning_rate": 9.9904349990435e-06, + "loss": 0.4627, + "step": 9571 + }, + { + "epoch": 4.450023245002324, + "grad_norm": 8.105818748474121, + "learning_rate": 9.990433999043401e-06, + "loss": 0.6609, + "step": 9572 + }, + { + "epoch": 4.450488145048815, + "grad_norm": 4.194422245025635, + "learning_rate": 9.9904329990433e-06, + "loss": 0.1992, + "step": 9573 + }, + { + "epoch": 4.4509530450953045, + "grad_norm": 3.155341625213623, + "learning_rate": 9.990431999043201e-06, + "loss": 0.1172, + "step": 9574 + }, + { + "epoch": 4.451417945141794, + "grad_norm": 7.623612880706787, + "learning_rate": 9.9904309990431e-06, + "loss": 0.384, + "step": 9575 + }, + { + "epoch": 4.451882845188284, + "grad_norm": 6.461071014404297, + "learning_rate": 9.990429999043e-06, + "loss": 0.3281, + "step": 9576 + }, + { + "epoch": 4.452347745234775, + "grad_norm": 6.1012959480285645, + "learning_rate": 9.9904289990429e-06, + "loss": 0.4692, + "step": 9577 + }, + { + "epoch": 4.452812645281265, + "grad_norm": 2.438063383102417, + "learning_rate": 9.9904279990428e-06, + "loss": 0.0719, + "step": 9578 + }, + { + "epoch": 4.453277545327754, + "grad_norm": 4.738842487335205, + "learning_rate": 9.990426999042701e-06, + "loss": 0.2442, + "step": 9579 + }, + { + "epoch": 4.453742445374244, + "grad_norm": 7.835455894470215, + "learning_rate": 9.990425999042602e-06, + "loss": 0.4421, + "step": 9580 + }, + { + "epoch": 4.454207345420735, + "grad_norm": 4.474701404571533, + "learning_rate": 9.9904249990425e-06, + "loss": 0.2463, + "step": 9581 + }, + { + "epoch": 4.454672245467225, + "grad_norm": 4.785681247711182, + "learning_rate": 9.990423999042401e-06, + "loss": 0.3238, + "step": 9582 + }, + { + "epoch": 4.4551371455137145, + "grad_norm": 5.447755336761475, + "learning_rate": 9.9904229990423e-06, + "loss": 0.3445, + "step": 9583 + }, + { + "epoch": 4.455602045560204, + "grad_norm": 2.990201234817505, + "learning_rate": 9.9904219990422e-06, + "loss": 0.1236, + "step": 9584 + }, + { + "epoch": 4.456066945606695, + "grad_norm": 5.440340042114258, + "learning_rate": 9.990420999042101e-06, + "loss": 0.3745, + "step": 9585 + }, + { + "epoch": 4.456531845653185, + "grad_norm": 3.8630869388580322, + "learning_rate": 9.990419999042e-06, + "loss": 0.2396, + "step": 9586 + }, + { + "epoch": 4.456996745699675, + "grad_norm": 5.252440929412842, + "learning_rate": 9.990418999041901e-06, + "loss": 0.3169, + "step": 9587 + }, + { + "epoch": 4.457461645746164, + "grad_norm": 3.897005319595337, + "learning_rate": 9.9904179990418e-06, + "loss": 0.1843, + "step": 9588 + }, + { + "epoch": 4.457926545792654, + "grad_norm": 7.618522644042969, + "learning_rate": 9.9904169990417e-06, + "loss": 0.2897, + "step": 9589 + }, + { + "epoch": 4.458391445839145, + "grad_norm": 5.85942268371582, + "learning_rate": 9.9904159990416e-06, + "loss": 0.2849, + "step": 9590 + }, + { + "epoch": 4.458856345885635, + "grad_norm": 5.3805131912231445, + "learning_rate": 9.9904149990415e-06, + "loss": 0.266, + "step": 9591 + }, + { + "epoch": 4.4593212459321245, + "grad_norm": 8.814221382141113, + "learning_rate": 9.990413999041401e-06, + "loss": 0.5152, + "step": 9592 + }, + { + "epoch": 4.459786145978614, + "grad_norm": 4.937765598297119, + "learning_rate": 9.990412999041302e-06, + "loss": 0.3127, + "step": 9593 + }, + { + "epoch": 4.460251046025105, + "grad_norm": 6.001560688018799, + "learning_rate": 9.9904119990412e-06, + "loss": 0.454, + "step": 9594 + }, + { + "epoch": 4.460715946071595, + "grad_norm": 5.558451175689697, + "learning_rate": 9.9904109990411e-06, + "loss": 0.4135, + "step": 9595 + }, + { + "epoch": 4.461180846118085, + "grad_norm": 2.6303176879882812, + "learning_rate": 9.990409999041e-06, + "loss": 0.1273, + "step": 9596 + }, + { + "epoch": 4.461645746164574, + "grad_norm": 4.5178375244140625, + "learning_rate": 9.990408999040901e-06, + "loss": 0.2057, + "step": 9597 + }, + { + "epoch": 4.462110646211064, + "grad_norm": 6.212421417236328, + "learning_rate": 9.990407999040802e-06, + "loss": 0.2157, + "step": 9598 + }, + { + "epoch": 4.462575546257555, + "grad_norm": 7.124369144439697, + "learning_rate": 9.9904069990407e-06, + "loss": 0.2513, + "step": 9599 + }, + { + "epoch": 4.463040446304045, + "grad_norm": 3.942070484161377, + "learning_rate": 9.990405999040601e-06, + "loss": 0.1744, + "step": 9600 + }, + { + "epoch": 4.4635053463505345, + "grad_norm": 5.608713150024414, + "learning_rate": 9.9904049990405e-06, + "loss": 0.361, + "step": 9601 + }, + { + "epoch": 4.463970246397024, + "grad_norm": 4.4525651931762695, + "learning_rate": 9.9904039990404e-06, + "loss": 0.221, + "step": 9602 + }, + { + "epoch": 4.464435146443515, + "grad_norm": 5.6285600662231445, + "learning_rate": 9.990402999040301e-06, + "loss": 0.3066, + "step": 9603 + }, + { + "epoch": 4.464900046490005, + "grad_norm": 6.516317844390869, + "learning_rate": 9.9904019990402e-06, + "loss": 0.3185, + "step": 9604 + }, + { + "epoch": 4.465364946536495, + "grad_norm": 5.950333118438721, + "learning_rate": 9.990400999040101e-06, + "loss": 0.3347, + "step": 9605 + }, + { + "epoch": 4.465829846582984, + "grad_norm": 4.985559463500977, + "learning_rate": 9.99039999904e-06, + "loss": 0.2302, + "step": 9606 + }, + { + "epoch": 4.466294746629474, + "grad_norm": 4.718813419342041, + "learning_rate": 9.9903989990399e-06, + "loss": 0.145, + "step": 9607 + }, + { + "epoch": 4.466759646675965, + "grad_norm": 5.455022811889648, + "learning_rate": 9.9903979990398e-06, + "loss": 0.2428, + "step": 9608 + }, + { + "epoch": 4.467224546722455, + "grad_norm": 5.833909034729004, + "learning_rate": 9.9903969990397e-06, + "loss": 0.4762, + "step": 9609 + }, + { + "epoch": 4.4676894467689445, + "grad_norm": 3.8171560764312744, + "learning_rate": 9.990395999039601e-06, + "loss": 0.1521, + "step": 9610 + }, + { + "epoch": 4.468154346815434, + "grad_norm": 6.71713399887085, + "learning_rate": 9.990394999039502e-06, + "loss": 0.3296, + "step": 9611 + }, + { + "epoch": 4.468619246861925, + "grad_norm": 18.759204864501953, + "learning_rate": 9.9903939990394e-06, + "loss": 0.4249, + "step": 9612 + }, + { + "epoch": 4.469084146908415, + "grad_norm": 5.862446308135986, + "learning_rate": 9.990392999039301e-06, + "loss": 0.311, + "step": 9613 + }, + { + "epoch": 4.469549046954905, + "grad_norm": 4.9544219970703125, + "learning_rate": 9.9903919990392e-06, + "loss": 0.1211, + "step": 9614 + }, + { + "epoch": 4.470013947001394, + "grad_norm": 7.47583532333374, + "learning_rate": 9.990390999039101e-06, + "loss": 0.5565, + "step": 9615 + }, + { + "epoch": 4.470478847047885, + "grad_norm": 6.380617141723633, + "learning_rate": 9.990389999039002e-06, + "loss": 0.3647, + "step": 9616 + }, + { + "epoch": 4.470943747094375, + "grad_norm": 3.7088112831115723, + "learning_rate": 9.9903889990389e-06, + "loss": 0.1764, + "step": 9617 + }, + { + "epoch": 4.471408647140865, + "grad_norm": 12.42190170288086, + "learning_rate": 9.990387999038801e-06, + "loss": 0.3931, + "step": 9618 + }, + { + "epoch": 4.4718735471873545, + "grad_norm": 5.999908924102783, + "learning_rate": 9.9903869990387e-06, + "loss": 0.4029, + "step": 9619 + }, + { + "epoch": 4.472338447233844, + "grad_norm": 4.127218246459961, + "learning_rate": 9.9903859990386e-06, + "loss": 0.2258, + "step": 9620 + }, + { + "epoch": 4.472803347280335, + "grad_norm": 4.760138511657715, + "learning_rate": 9.9903849990385e-06, + "loss": 0.2853, + "step": 9621 + }, + { + "epoch": 4.473268247326825, + "grad_norm": 5.585207939147949, + "learning_rate": 9.9903839990384e-06, + "loss": 0.176, + "step": 9622 + }, + { + "epoch": 4.473733147373315, + "grad_norm": 5.235278129577637, + "learning_rate": 9.990382999038301e-06, + "loss": 0.3815, + "step": 9623 + }, + { + "epoch": 4.474198047419804, + "grad_norm": 7.135595321655273, + "learning_rate": 9.990381999038202e-06, + "loss": 0.3134, + "step": 9624 + }, + { + "epoch": 4.474662947466295, + "grad_norm": 5.322930335998535, + "learning_rate": 9.9903809990381e-06, + "loss": 0.1349, + "step": 9625 + }, + { + "epoch": 4.475127847512785, + "grad_norm": 4.763232231140137, + "learning_rate": 9.990379999038001e-06, + "loss": 0.2557, + "step": 9626 + }, + { + "epoch": 4.475592747559275, + "grad_norm": 5.388132572174072, + "learning_rate": 9.9903789990379e-06, + "loss": 0.2186, + "step": 9627 + }, + { + "epoch": 4.4760576476057645, + "grad_norm": 4.179927349090576, + "learning_rate": 9.990377999037801e-06, + "loss": 0.1705, + "step": 9628 + }, + { + "epoch": 4.476522547652255, + "grad_norm": 5.746971130371094, + "learning_rate": 9.990376999037702e-06, + "loss": 0.418, + "step": 9629 + }, + { + "epoch": 4.476987447698745, + "grad_norm": 5.134687423706055, + "learning_rate": 9.9903759990376e-06, + "loss": 0.3618, + "step": 9630 + }, + { + "epoch": 4.477452347745235, + "grad_norm": 4.79599142074585, + "learning_rate": 9.990374999037501e-06, + "loss": 0.3156, + "step": 9631 + }, + { + "epoch": 4.477917247791725, + "grad_norm": 6.211194038391113, + "learning_rate": 9.9903739990374e-06, + "loss": 0.2823, + "step": 9632 + }, + { + "epoch": 4.478382147838214, + "grad_norm": 4.374379634857178, + "learning_rate": 9.990372999037301e-06, + "loss": 0.2746, + "step": 9633 + }, + { + "epoch": 4.478847047884705, + "grad_norm": 3.723395347595215, + "learning_rate": 9.9903719990372e-06, + "loss": 0.1705, + "step": 9634 + }, + { + "epoch": 4.479311947931195, + "grad_norm": 5.431520462036133, + "learning_rate": 9.9903709990371e-06, + "loss": 0.2923, + "step": 9635 + }, + { + "epoch": 4.479776847977685, + "grad_norm": 5.248690128326416, + "learning_rate": 9.990369999037001e-06, + "loss": 0.3451, + "step": 9636 + }, + { + "epoch": 4.4802417480241745, + "grad_norm": 4.876051902770996, + "learning_rate": 9.990368999036902e-06, + "loss": 0.3336, + "step": 9637 + }, + { + "epoch": 4.480706648070665, + "grad_norm": 4.602251052856445, + "learning_rate": 9.9903679990368e-06, + "loss": 0.1765, + "step": 9638 + }, + { + "epoch": 4.481171548117155, + "grad_norm": 10.103796005249023, + "learning_rate": 9.9903669990367e-06, + "loss": 0.6582, + "step": 9639 + }, + { + "epoch": 4.481636448163645, + "grad_norm": 9.17137622833252, + "learning_rate": 9.9903659990366e-06, + "loss": 0.4734, + "step": 9640 + }, + { + "epoch": 4.482101348210135, + "grad_norm": 3.227964401245117, + "learning_rate": 9.990364999036501e-06, + "loss": 0.1577, + "step": 9641 + }, + { + "epoch": 4.482566248256624, + "grad_norm": 5.502265930175781, + "learning_rate": 9.990363999036402e-06, + "loss": 0.1544, + "step": 9642 + }, + { + "epoch": 4.483031148303115, + "grad_norm": 6.8828606605529785, + "learning_rate": 9.9903629990363e-06, + "loss": 0.3788, + "step": 9643 + }, + { + "epoch": 4.483496048349605, + "grad_norm": 10.599496841430664, + "learning_rate": 9.990361999036201e-06, + "loss": 0.5422, + "step": 9644 + }, + { + "epoch": 4.483960948396095, + "grad_norm": 6.7085652351379395, + "learning_rate": 9.9903609990361e-06, + "loss": 0.2955, + "step": 9645 + }, + { + "epoch": 4.4844258484425845, + "grad_norm": 6.9726362228393555, + "learning_rate": 9.990359999036001e-06, + "loss": 0.3316, + "step": 9646 + }, + { + "epoch": 4.484890748489075, + "grad_norm": 4.827299118041992, + "learning_rate": 9.9903589990359e-06, + "loss": 0.2612, + "step": 9647 + }, + { + "epoch": 4.485355648535565, + "grad_norm": 11.006776809692383, + "learning_rate": 9.9903579990358e-06, + "loss": 0.3574, + "step": 9648 + }, + { + "epoch": 4.485820548582055, + "grad_norm": 7.392117500305176, + "learning_rate": 9.990356999035701e-06, + "loss": 0.5517, + "step": 9649 + }, + { + "epoch": 4.486285448628545, + "grad_norm": 7.588341236114502, + "learning_rate": 9.9903559990356e-06, + "loss": 0.1624, + "step": 9650 + }, + { + "epoch": 4.486750348675034, + "grad_norm": 6.194585800170898, + "learning_rate": 9.990354999035501e-06, + "loss": 0.2575, + "step": 9651 + }, + { + "epoch": 4.487215248721525, + "grad_norm": 5.7238078117370605, + "learning_rate": 9.9903539990354e-06, + "loss": 0.2169, + "step": 9652 + }, + { + "epoch": 4.487680148768015, + "grad_norm": 5.854093551635742, + "learning_rate": 9.9903529990353e-06, + "loss": 0.2277, + "step": 9653 + }, + { + "epoch": 4.488145048814505, + "grad_norm": 5.407196521759033, + "learning_rate": 9.990351999035201e-06, + "loss": 0.2478, + "step": 9654 + }, + { + "epoch": 4.4886099488609945, + "grad_norm": 7.569406509399414, + "learning_rate": 9.990350999035102e-06, + "loss": 0.3328, + "step": 9655 + }, + { + "epoch": 4.489074848907485, + "grad_norm": 3.690749406814575, + "learning_rate": 9.990349999035e-06, + "loss": 0.1588, + "step": 9656 + }, + { + "epoch": 4.489539748953975, + "grad_norm": 5.555450916290283, + "learning_rate": 9.990348999034901e-06, + "loss": 0.3107, + "step": 9657 + }, + { + "epoch": 4.490004649000465, + "grad_norm": 5.188071250915527, + "learning_rate": 9.9903479990348e-06, + "loss": 0.2877, + "step": 9658 + }, + { + "epoch": 4.490469549046955, + "grad_norm": 5.54628849029541, + "learning_rate": 9.9903469990347e-06, + "loss": 0.4423, + "step": 9659 + }, + { + "epoch": 4.490934449093445, + "grad_norm": 4.231020927429199, + "learning_rate": 9.9903459990346e-06, + "loss": 0.2166, + "step": 9660 + }, + { + "epoch": 4.491399349139935, + "grad_norm": 5.207355976104736, + "learning_rate": 9.9903449990345e-06, + "loss": 0.3063, + "step": 9661 + }, + { + "epoch": 4.491864249186425, + "grad_norm": 5.822172164916992, + "learning_rate": 9.990343999034401e-06, + "loss": 0.5006, + "step": 9662 + }, + { + "epoch": 4.492329149232915, + "grad_norm": 7.007233619689941, + "learning_rate": 9.9903429990343e-06, + "loss": 0.119, + "step": 9663 + }, + { + "epoch": 4.492794049279405, + "grad_norm": 4.704606056213379, + "learning_rate": 9.990341999034201e-06, + "loss": 0.1596, + "step": 9664 + }, + { + "epoch": 4.493258949325895, + "grad_norm": 5.500551223754883, + "learning_rate": 9.9903409990341e-06, + "loss": 0.359, + "step": 9665 + }, + { + "epoch": 4.493723849372385, + "grad_norm": 3.620495080947876, + "learning_rate": 9.990339999034e-06, + "loss": 0.1379, + "step": 9666 + }, + { + "epoch": 4.494188749418875, + "grad_norm": 3.9646501541137695, + "learning_rate": 9.990338999033901e-06, + "loss": 0.1884, + "step": 9667 + }, + { + "epoch": 4.494653649465365, + "grad_norm": 3.0159714221954346, + "learning_rate": 9.990337999033802e-06, + "loss": 0.1002, + "step": 9668 + }, + { + "epoch": 4.495118549511855, + "grad_norm": 4.8505072593688965, + "learning_rate": 9.9903369990337e-06, + "loss": 0.2959, + "step": 9669 + }, + { + "epoch": 4.495583449558345, + "grad_norm": 3.808603525161743, + "learning_rate": 9.9903359990336e-06, + "loss": 0.1965, + "step": 9670 + }, + { + "epoch": 4.496048349604835, + "grad_norm": 7.265925884246826, + "learning_rate": 9.9903349990335e-06, + "loss": 0.2403, + "step": 9671 + }, + { + "epoch": 4.496513249651325, + "grad_norm": 4.822021961212158, + "learning_rate": 9.990333999033401e-06, + "loss": 0.2082, + "step": 9672 + }, + { + "epoch": 4.496978149697815, + "grad_norm": 4.57530403137207, + "learning_rate": 9.9903329990333e-06, + "loss": 0.149, + "step": 9673 + }, + { + "epoch": 4.497443049744305, + "grad_norm": 7.784276962280273, + "learning_rate": 9.9903319990332e-06, + "loss": 0.3899, + "step": 9674 + }, + { + "epoch": 4.497907949790795, + "grad_norm": 6.273117542266846, + "learning_rate": 9.990330999033101e-06, + "loss": 0.2574, + "step": 9675 + }, + { + "epoch": 4.498372849837285, + "grad_norm": 5.179259777069092, + "learning_rate": 9.990329999033e-06, + "loss": 0.2787, + "step": 9676 + }, + { + "epoch": 4.498837749883775, + "grad_norm": 7.371156215667725, + "learning_rate": 9.990328999032901e-06, + "loss": 0.3921, + "step": 9677 + }, + { + "epoch": 4.499302649930265, + "grad_norm": 4.558537006378174, + "learning_rate": 9.9903279990328e-06, + "loss": 0.08, + "step": 9678 + }, + { + "epoch": 4.499767549976755, + "grad_norm": 5.556728363037109, + "learning_rate": 9.9903269990327e-06, + "loss": 0.296, + "step": 9679 + }, + { + "epoch": 4.500232450023245, + "grad_norm": 5.022894859313965, + "learning_rate": 9.990325999032601e-06, + "loss": 0.1751, + "step": 9680 + }, + { + "epoch": 4.500697350069735, + "grad_norm": 8.056923866271973, + "learning_rate": 9.9903249990325e-06, + "loss": 0.5653, + "step": 9681 + }, + { + "epoch": 4.501162250116225, + "grad_norm": 6.936756610870361, + "learning_rate": 9.990323999032401e-06, + "loss": 0.4229, + "step": 9682 + }, + { + "epoch": 4.501627150162715, + "grad_norm": 6.227275371551514, + "learning_rate": 9.9903229990323e-06, + "loss": 0.2914, + "step": 9683 + }, + { + "epoch": 4.502092050209205, + "grad_norm": 5.1601338386535645, + "learning_rate": 9.9903219990322e-06, + "loss": 0.3511, + "step": 9684 + }, + { + "epoch": 4.502556950255695, + "grad_norm": 3.730722665786743, + "learning_rate": 9.990320999032101e-06, + "loss": 0.0739, + "step": 9685 + }, + { + "epoch": 4.503021850302185, + "grad_norm": 6.096288681030273, + "learning_rate": 9.990319999032e-06, + "loss": 0.2735, + "step": 9686 + }, + { + "epoch": 4.503486750348675, + "grad_norm": 4.998672962188721, + "learning_rate": 9.9903189990319e-06, + "loss": 0.1621, + "step": 9687 + }, + { + "epoch": 4.503951650395165, + "grad_norm": 9.602455139160156, + "learning_rate": 9.990317999031801e-06, + "loss": 0.7968, + "step": 9688 + }, + { + "epoch": 4.504416550441655, + "grad_norm": 5.884352684020996, + "learning_rate": 9.9903169990317e-06, + "loss": 0.2604, + "step": 9689 + }, + { + "epoch": 4.504881450488145, + "grad_norm": 9.114278793334961, + "learning_rate": 9.990315999031601e-06, + "loss": 0.2791, + "step": 9690 + }, + { + "epoch": 4.505346350534635, + "grad_norm": 2.786947727203369, + "learning_rate": 9.9903149990315e-06, + "loss": 0.1257, + "step": 9691 + }, + { + "epoch": 4.505811250581125, + "grad_norm": 5.688967704772949, + "learning_rate": 9.9903139990314e-06, + "loss": 0.2897, + "step": 9692 + }, + { + "epoch": 4.506276150627615, + "grad_norm": 4.8223676681518555, + "learning_rate": 9.990312999031301e-06, + "loss": 0.2469, + "step": 9693 + }, + { + "epoch": 4.506741050674105, + "grad_norm": 5.46897554397583, + "learning_rate": 9.9903119990312e-06, + "loss": 0.3504, + "step": 9694 + }, + { + "epoch": 4.507205950720595, + "grad_norm": 5.1948041915893555, + "learning_rate": 9.990310999031101e-06, + "loss": 0.3042, + "step": 9695 + }, + { + "epoch": 4.507670850767085, + "grad_norm": 6.260500431060791, + "learning_rate": 9.990309999031e-06, + "loss": 0.3992, + "step": 9696 + }, + { + "epoch": 4.508135750813575, + "grad_norm": 9.760674476623535, + "learning_rate": 9.9903089990309e-06, + "loss": 0.5959, + "step": 9697 + }, + { + "epoch": 4.508600650860065, + "grad_norm": 6.937314033508301, + "learning_rate": 9.990307999030801e-06, + "loss": 0.6142, + "step": 9698 + }, + { + "epoch": 4.509065550906556, + "grad_norm": 7.0272603034973145, + "learning_rate": 9.990306999030702e-06, + "loss": 0.6187, + "step": 9699 + }, + { + "epoch": 4.509530450953045, + "grad_norm": 4.943532466888428, + "learning_rate": 9.990305999030601e-06, + "loss": 0.2506, + "step": 9700 + }, + { + "epoch": 4.509995350999535, + "grad_norm": 6.5492472648620605, + "learning_rate": 9.990304999030502e-06, + "loss": 0.1942, + "step": 9701 + }, + { + "epoch": 4.510460251046025, + "grad_norm": 5.492170333862305, + "learning_rate": 9.9903039990304e-06, + "loss": 0.4313, + "step": 9702 + }, + { + "epoch": 4.510925151092515, + "grad_norm": 6.858870983123779, + "learning_rate": 9.9903029990303e-06, + "loss": 0.474, + "step": 9703 + }, + { + "epoch": 4.5113900511390055, + "grad_norm": 4.437234401702881, + "learning_rate": 9.9903019990302e-06, + "loss": 0.1674, + "step": 9704 + }, + { + "epoch": 4.511854951185495, + "grad_norm": 6.2445783615112305, + "learning_rate": 9.9903009990301e-06, + "loss": 0.3037, + "step": 9705 + }, + { + "epoch": 4.512319851231985, + "grad_norm": 5.635870933532715, + "learning_rate": 9.990299999030001e-06, + "loss": 0.2884, + "step": 9706 + }, + { + "epoch": 4.512784751278475, + "grad_norm": 4.119051933288574, + "learning_rate": 9.9902989990299e-06, + "loss": 0.2907, + "step": 9707 + }, + { + "epoch": 4.513249651324966, + "grad_norm": 3.730886936187744, + "learning_rate": 9.990297999029801e-06, + "loss": 0.209, + "step": 9708 + }, + { + "epoch": 4.513714551371455, + "grad_norm": 5.4583420753479, + "learning_rate": 9.9902969990297e-06, + "loss": 0.3314, + "step": 9709 + }, + { + "epoch": 4.514179451417945, + "grad_norm": 7.272864818572998, + "learning_rate": 9.9902959990296e-06, + "loss": 0.6037, + "step": 9710 + }, + { + "epoch": 4.514644351464435, + "grad_norm": 4.852536678314209, + "learning_rate": 9.990294999029501e-06, + "loss": 0.2744, + "step": 9711 + }, + { + "epoch": 4.515109251510925, + "grad_norm": 7.770708084106445, + "learning_rate": 9.990293999029402e-06, + "loss": 0.1522, + "step": 9712 + }, + { + "epoch": 4.5155741515574155, + "grad_norm": 4.282055377960205, + "learning_rate": 9.990292999029301e-06, + "loss": 0.1803, + "step": 9713 + }, + { + "epoch": 4.516039051603905, + "grad_norm": 5.607873916625977, + "learning_rate": 9.9902919990292e-06, + "loss": 0.3323, + "step": 9714 + }, + { + "epoch": 4.516503951650395, + "grad_norm": 5.0373430252075195, + "learning_rate": 9.9902909990291e-06, + "loss": 0.1784, + "step": 9715 + }, + { + "epoch": 4.516968851696885, + "grad_norm": 9.26966381072998, + "learning_rate": 9.990289999029e-06, + "loss": 0.3544, + "step": 9716 + }, + { + "epoch": 4.517433751743376, + "grad_norm": 5.0968475341796875, + "learning_rate": 9.9902889990289e-06, + "loss": 0.2115, + "step": 9717 + }, + { + "epoch": 4.517898651789865, + "grad_norm": 5.6568217277526855, + "learning_rate": 9.990287999028801e-06, + "loss": 0.1465, + "step": 9718 + }, + { + "epoch": 4.518363551836355, + "grad_norm": 5.267033100128174, + "learning_rate": 9.990286999028702e-06, + "loss": 0.3245, + "step": 9719 + }, + { + "epoch": 4.518828451882845, + "grad_norm": 4.288541316986084, + "learning_rate": 9.9902859990286e-06, + "loss": 0.2088, + "step": 9720 + }, + { + "epoch": 4.519293351929335, + "grad_norm": 3.349764585494995, + "learning_rate": 9.990284999028501e-06, + "loss": 0.1092, + "step": 9721 + }, + { + "epoch": 4.5197582519758255, + "grad_norm": 5.419585704803467, + "learning_rate": 9.9902839990284e-06, + "loss": 0.3585, + "step": 9722 + }, + { + "epoch": 4.520223152022315, + "grad_norm": 5.420953750610352, + "learning_rate": 9.9902829990283e-06, + "loss": 0.34, + "step": 9723 + }, + { + "epoch": 4.520688052068805, + "grad_norm": 4.368344783782959, + "learning_rate": 9.990281999028201e-06, + "loss": 0.1262, + "step": 9724 + }, + { + "epoch": 4.521152952115295, + "grad_norm": 6.4953107833862305, + "learning_rate": 9.9902809990281e-06, + "loss": 0.3504, + "step": 9725 + }, + { + "epoch": 4.521617852161786, + "grad_norm": 5.624021530151367, + "learning_rate": 9.990279999028001e-06, + "loss": 0.3111, + "step": 9726 + }, + { + "epoch": 4.522082752208275, + "grad_norm": 6.860439777374268, + "learning_rate": 9.9902789990279e-06, + "loss": 0.2092, + "step": 9727 + }, + { + "epoch": 4.522547652254765, + "grad_norm": 10.417460441589355, + "learning_rate": 9.9902779990278e-06, + "loss": 0.4389, + "step": 9728 + }, + { + "epoch": 4.523012552301255, + "grad_norm": 6.429449558258057, + "learning_rate": 9.9902769990277e-06, + "loss": 0.2518, + "step": 9729 + }, + { + "epoch": 4.523477452347745, + "grad_norm": 6.166568756103516, + "learning_rate": 9.9902759990276e-06, + "loss": 0.2682, + "step": 9730 + }, + { + "epoch": 4.5239423523942355, + "grad_norm": 6.132393836975098, + "learning_rate": 9.990274999027501e-06, + "loss": 0.3982, + "step": 9731 + }, + { + "epoch": 4.524407252440725, + "grad_norm": 9.18830394744873, + "learning_rate": 9.990273999027402e-06, + "loss": 0.2891, + "step": 9732 + }, + { + "epoch": 4.524872152487215, + "grad_norm": 5.998446464538574, + "learning_rate": 9.9902729990273e-06, + "loss": 0.4707, + "step": 9733 + }, + { + "epoch": 4.525337052533705, + "grad_norm": 7.887923240661621, + "learning_rate": 9.990271999027201e-06, + "loss": 0.4429, + "step": 9734 + }, + { + "epoch": 4.525801952580196, + "grad_norm": 7.620444297790527, + "learning_rate": 9.9902709990271e-06, + "loss": 0.3367, + "step": 9735 + }, + { + "epoch": 4.526266852626685, + "grad_norm": 5.9371209144592285, + "learning_rate": 9.990269999027001e-06, + "loss": 0.3778, + "step": 9736 + }, + { + "epoch": 4.526731752673175, + "grad_norm": 5.7201433181762695, + "learning_rate": 9.990268999026902e-06, + "loss": 0.2869, + "step": 9737 + }, + { + "epoch": 4.527196652719665, + "grad_norm": 5.752798080444336, + "learning_rate": 9.9902679990268e-06, + "loss": 0.3904, + "step": 9738 + }, + { + "epoch": 4.527661552766155, + "grad_norm": 3.837947368621826, + "learning_rate": 9.990266999026701e-06, + "loss": 0.1618, + "step": 9739 + }, + { + "epoch": 4.5281264528126455, + "grad_norm": 6.164106369018555, + "learning_rate": 9.9902659990266e-06, + "loss": 0.4948, + "step": 9740 + }, + { + "epoch": 4.528591352859135, + "grad_norm": 4.0481672286987305, + "learning_rate": 9.9902649990265e-06, + "loss": 0.1705, + "step": 9741 + }, + { + "epoch": 4.529056252905625, + "grad_norm": 5.901339054107666, + "learning_rate": 9.9902639990264e-06, + "loss": 0.3683, + "step": 9742 + }, + { + "epoch": 4.529521152952116, + "grad_norm": 4.869748115539551, + "learning_rate": 9.9902629990263e-06, + "loss": 0.3052, + "step": 9743 + }, + { + "epoch": 4.529986052998606, + "grad_norm": 4.7747979164123535, + "learning_rate": 9.990261999026201e-06, + "loss": 0.1875, + "step": 9744 + }, + { + "epoch": 4.530450953045095, + "grad_norm": 3.88774037361145, + "learning_rate": 9.990260999026102e-06, + "loss": 0.1635, + "step": 9745 + }, + { + "epoch": 4.530915853091585, + "grad_norm": 6.826028823852539, + "learning_rate": 9.990259999026e-06, + "loss": 0.4002, + "step": 9746 + }, + { + "epoch": 4.531380753138075, + "grad_norm": 9.607813835144043, + "learning_rate": 9.9902589990259e-06, + "loss": 0.2772, + "step": 9747 + }, + { + "epoch": 4.531845653184566, + "grad_norm": 3.48193359375, + "learning_rate": 9.9902579990258e-06, + "loss": 0.1221, + "step": 9748 + }, + { + "epoch": 4.5323105532310555, + "grad_norm": 7.509178638458252, + "learning_rate": 9.990256999025701e-06, + "loss": 0.5678, + "step": 9749 + }, + { + "epoch": 4.532775453277545, + "grad_norm": 5.52692174911499, + "learning_rate": 9.990255999025602e-06, + "loss": 0.2694, + "step": 9750 + }, + { + "epoch": 4.533240353324035, + "grad_norm": 4.771557331085205, + "learning_rate": 9.9902549990255e-06, + "loss": 0.2154, + "step": 9751 + }, + { + "epoch": 4.533705253370526, + "grad_norm": 4.323147296905518, + "learning_rate": 9.990253999025401e-06, + "loss": 0.3099, + "step": 9752 + }, + { + "epoch": 4.534170153417016, + "grad_norm": 4.852615833282471, + "learning_rate": 9.9902529990253e-06, + "loss": 0.192, + "step": 9753 + }, + { + "epoch": 4.534635053463505, + "grad_norm": 4.751824855804443, + "learning_rate": 9.990251999025201e-06, + "loss": 0.3522, + "step": 9754 + }, + { + "epoch": 4.535099953509995, + "grad_norm": 4.813645839691162, + "learning_rate": 9.9902509990251e-06, + "loss": 0.2323, + "step": 9755 + }, + { + "epoch": 4.535564853556485, + "grad_norm": 4.886682033538818, + "learning_rate": 9.990249999025e-06, + "loss": 0.1945, + "step": 9756 + }, + { + "epoch": 4.536029753602976, + "grad_norm": 4.975283145904541, + "learning_rate": 9.990248999024901e-06, + "loss": 0.0927, + "step": 9757 + }, + { + "epoch": 4.5364946536494655, + "grad_norm": 8.35230541229248, + "learning_rate": 9.9902479990248e-06, + "loss": 0.2267, + "step": 9758 + }, + { + "epoch": 4.536959553695955, + "grad_norm": 3.986302137374878, + "learning_rate": 9.9902469990247e-06, + "loss": 0.1783, + "step": 9759 + }, + { + "epoch": 4.537424453742445, + "grad_norm": 3.554260492324829, + "learning_rate": 9.9902459990246e-06, + "loss": 0.2008, + "step": 9760 + }, + { + "epoch": 4.537889353788936, + "grad_norm": 5.357598781585693, + "learning_rate": 9.9902449990245e-06, + "loss": 0.377, + "step": 9761 + }, + { + "epoch": 4.538354253835426, + "grad_norm": 4.842860221862793, + "learning_rate": 9.990243999024401e-06, + "loss": 0.2188, + "step": 9762 + }, + { + "epoch": 4.538819153881915, + "grad_norm": 7.982436180114746, + "learning_rate": 9.990242999024302e-06, + "loss": 0.2112, + "step": 9763 + }, + { + "epoch": 4.539284053928405, + "grad_norm": 4.585967063903809, + "learning_rate": 9.9902419990242e-06, + "loss": 0.2185, + "step": 9764 + }, + { + "epoch": 4.539748953974895, + "grad_norm": 3.6830360889434814, + "learning_rate": 9.990240999024101e-06, + "loss": 0.134, + "step": 9765 + }, + { + "epoch": 4.540213854021386, + "grad_norm": 9.236247062683105, + "learning_rate": 9.990239999024e-06, + "loss": 0.5505, + "step": 9766 + }, + { + "epoch": 4.5406787540678755, + "grad_norm": 7.141378402709961, + "learning_rate": 9.990238999023901e-06, + "loss": 0.3391, + "step": 9767 + }, + { + "epoch": 4.541143654114365, + "grad_norm": 4.540724277496338, + "learning_rate": 9.990237999023802e-06, + "loss": 0.1865, + "step": 9768 + }, + { + "epoch": 4.541608554160855, + "grad_norm": 9.932229042053223, + "learning_rate": 9.9902369990237e-06, + "loss": 0.5408, + "step": 9769 + }, + { + "epoch": 4.542073454207346, + "grad_norm": 7.37249755859375, + "learning_rate": 9.990235999023601e-06, + "loss": 0.3523, + "step": 9770 + }, + { + "epoch": 4.542538354253836, + "grad_norm": 5.50372838973999, + "learning_rate": 9.9902349990235e-06, + "loss": 0.201, + "step": 9771 + }, + { + "epoch": 4.543003254300325, + "grad_norm": 6.519275188446045, + "learning_rate": 9.9902339990234e-06, + "loss": 0.2558, + "step": 9772 + }, + { + "epoch": 4.543468154346815, + "grad_norm": 6.195790767669678, + "learning_rate": 9.9902329990233e-06, + "loss": 0.3368, + "step": 9773 + }, + { + "epoch": 4.543933054393305, + "grad_norm": 5.35660982131958, + "learning_rate": 9.9902319990232e-06, + "loss": 0.2449, + "step": 9774 + }, + { + "epoch": 4.544397954439796, + "grad_norm": 4.220370769500732, + "learning_rate": 9.990230999023101e-06, + "loss": 0.137, + "step": 9775 + }, + { + "epoch": 4.5448628544862855, + "grad_norm": 4.248498439788818, + "learning_rate": 9.990229999023002e-06, + "loss": 0.2307, + "step": 9776 + }, + { + "epoch": 4.545327754532775, + "grad_norm": 4.542881488800049, + "learning_rate": 9.9902289990229e-06, + "loss": 0.2355, + "step": 9777 + }, + { + "epoch": 4.545792654579266, + "grad_norm": 4.441304683685303, + "learning_rate": 9.990227999022801e-06, + "loss": 0.2643, + "step": 9778 + }, + { + "epoch": 4.546257554625756, + "grad_norm": 5.124523162841797, + "learning_rate": 9.9902269990227e-06, + "loss": 0.1813, + "step": 9779 + }, + { + "epoch": 4.546722454672246, + "grad_norm": 8.305695533752441, + "learning_rate": 9.990225999022601e-06, + "loss": 0.5333, + "step": 9780 + }, + { + "epoch": 4.547187354718735, + "grad_norm": 4.345355987548828, + "learning_rate": 9.990224999022502e-06, + "loss": 0.267, + "step": 9781 + }, + { + "epoch": 4.547652254765225, + "grad_norm": 6.707678318023682, + "learning_rate": 9.9902239990224e-06, + "loss": 0.4822, + "step": 9782 + }, + { + "epoch": 4.548117154811716, + "grad_norm": 5.779963970184326, + "learning_rate": 9.990222999022301e-06, + "loss": 0.407, + "step": 9783 + }, + { + "epoch": 4.548582054858206, + "grad_norm": 6.577779293060303, + "learning_rate": 9.9902219990222e-06, + "loss": 0.3185, + "step": 9784 + }, + { + "epoch": 4.5490469549046955, + "grad_norm": 7.3964433670043945, + "learning_rate": 9.990220999022101e-06, + "loss": 0.4879, + "step": 9785 + }, + { + "epoch": 4.549511854951185, + "grad_norm": 7.115274429321289, + "learning_rate": 9.990219999022e-06, + "loss": 0.4551, + "step": 9786 + }, + { + "epoch": 4.549976754997676, + "grad_norm": 6.92748498916626, + "learning_rate": 9.9902189990219e-06, + "loss": 0.2945, + "step": 9787 + }, + { + "epoch": 4.550441655044166, + "grad_norm": 7.909730434417725, + "learning_rate": 9.990217999021801e-06, + "loss": 0.1732, + "step": 9788 + }, + { + "epoch": 4.550906555090656, + "grad_norm": 10.342153549194336, + "learning_rate": 9.9902169990217e-06, + "loss": 1.2797, + "step": 9789 + }, + { + "epoch": 4.551371455137145, + "grad_norm": 3.847731590270996, + "learning_rate": 9.9902159990216e-06, + "loss": 0.1456, + "step": 9790 + }, + { + "epoch": 4.551836355183635, + "grad_norm": 9.006233215332031, + "learning_rate": 9.9902149990215e-06, + "loss": 0.3593, + "step": 9791 + }, + { + "epoch": 4.552301255230126, + "grad_norm": 13.119215965270996, + "learning_rate": 9.9902139990214e-06, + "loss": 0.5007, + "step": 9792 + }, + { + "epoch": 4.552766155276616, + "grad_norm": 5.852987766265869, + "learning_rate": 9.990212999021301e-06, + "loss": 0.3555, + "step": 9793 + }, + { + "epoch": 4.5532310553231055, + "grad_norm": 5.929834842681885, + "learning_rate": 9.990211999021202e-06, + "loss": 0.2176, + "step": 9794 + }, + { + "epoch": 4.553695955369595, + "grad_norm": 3.933621883392334, + "learning_rate": 9.9902109990211e-06, + "loss": 0.179, + "step": 9795 + }, + { + "epoch": 4.554160855416086, + "grad_norm": 4.8619489669799805, + "learning_rate": 9.990209999021001e-06, + "loss": 0.1899, + "step": 9796 + }, + { + "epoch": 4.554625755462576, + "grad_norm": 5.051548004150391, + "learning_rate": 9.9902089990209e-06, + "loss": 0.2185, + "step": 9797 + }, + { + "epoch": 4.5550906555090656, + "grad_norm": 5.934837818145752, + "learning_rate": 9.990207999020801e-06, + "loss": 0.3039, + "step": 9798 + }, + { + "epoch": 4.555555555555555, + "grad_norm": 3.612056016921997, + "learning_rate": 9.9902069990207e-06, + "loss": 0.1623, + "step": 9799 + }, + { + "epoch": 4.556020455602045, + "grad_norm": 5.716867923736572, + "learning_rate": 9.9902059990206e-06, + "loss": 0.4537, + "step": 9800 + }, + { + "epoch": 4.556485355648536, + "grad_norm": 6.189187049865723, + "learning_rate": 9.990204999020501e-06, + "loss": 0.2882, + "step": 9801 + }, + { + "epoch": 4.556950255695026, + "grad_norm": 8.066007614135742, + "learning_rate": 9.9902039990204e-06, + "loss": 0.2207, + "step": 9802 + }, + { + "epoch": 4.5574151557415155, + "grad_norm": 5.732692718505859, + "learning_rate": 9.990202999020301e-06, + "loss": 0.1159, + "step": 9803 + }, + { + "epoch": 4.557880055788005, + "grad_norm": 6.428135871887207, + "learning_rate": 9.9902019990202e-06, + "loss": 0.1757, + "step": 9804 + }, + { + "epoch": 4.558344955834496, + "grad_norm": 4.377623081207275, + "learning_rate": 9.9902009990201e-06, + "loss": 0.1905, + "step": 9805 + }, + { + "epoch": 4.558809855880986, + "grad_norm": 6.57150936126709, + "learning_rate": 9.990199999020001e-06, + "loss": 0.5019, + "step": 9806 + }, + { + "epoch": 4.5592747559274756, + "grad_norm": 5.208494186401367, + "learning_rate": 9.990198999019902e-06, + "loss": 0.2765, + "step": 9807 + }, + { + "epoch": 4.559739655973965, + "grad_norm": 6.668722629547119, + "learning_rate": 9.9901979990198e-06, + "loss": 0.5128, + "step": 9808 + }, + { + "epoch": 4.560204556020455, + "grad_norm": 6.818478584289551, + "learning_rate": 9.990196999019701e-06, + "loss": 0.4026, + "step": 9809 + }, + { + "epoch": 4.560669456066946, + "grad_norm": 5.945124626159668, + "learning_rate": 9.9901959990196e-06, + "loss": 0.3624, + "step": 9810 + }, + { + "epoch": 4.561134356113436, + "grad_norm": 9.425885200500488, + "learning_rate": 9.9901949990195e-06, + "loss": 0.1608, + "step": 9811 + }, + { + "epoch": 4.5615992561599255, + "grad_norm": 4.257078170776367, + "learning_rate": 9.9901939990194e-06, + "loss": 0.1971, + "step": 9812 + }, + { + "epoch": 4.562064156206415, + "grad_norm": 3.7925865650177, + "learning_rate": 9.9901929990193e-06, + "loss": 0.1233, + "step": 9813 + }, + { + "epoch": 4.562529056252906, + "grad_norm": 8.644584655761719, + "learning_rate": 9.990191999019201e-06, + "loss": 0.23, + "step": 9814 + }, + { + "epoch": 4.562993956299396, + "grad_norm": 5.444163799285889, + "learning_rate": 9.9901909990191e-06, + "loss": 0.2334, + "step": 9815 + }, + { + "epoch": 4.5634588563458856, + "grad_norm": 4.471383571624756, + "learning_rate": 9.990189999019001e-06, + "loss": 0.1414, + "step": 9816 + }, + { + "epoch": 4.563923756392375, + "grad_norm": 5.6957902908325195, + "learning_rate": 9.9901889990189e-06, + "loss": 0.3713, + "step": 9817 + }, + { + "epoch": 4.564388656438865, + "grad_norm": 3.851926565170288, + "learning_rate": 9.9901879990188e-06, + "loss": 0.2381, + "step": 9818 + }, + { + "epoch": 4.564853556485356, + "grad_norm": 5.525265693664551, + "learning_rate": 9.990186999018701e-06, + "loss": 0.3745, + "step": 9819 + }, + { + "epoch": 4.565318456531846, + "grad_norm": 5.817408084869385, + "learning_rate": 9.990185999018602e-06, + "loss": 0.5055, + "step": 9820 + }, + { + "epoch": 4.5657833565783355, + "grad_norm": 4.327104091644287, + "learning_rate": 9.990184999018501e-06, + "loss": 0.1249, + "step": 9821 + }, + { + "epoch": 4.566248256624826, + "grad_norm": 5.7990641593933105, + "learning_rate": 9.9901839990184e-06, + "loss": 0.2122, + "step": 9822 + }, + { + "epoch": 4.566713156671316, + "grad_norm": 5.852689743041992, + "learning_rate": 9.9901829990183e-06, + "loss": 0.191, + "step": 9823 + }, + { + "epoch": 4.567178056717806, + "grad_norm": 5.138922214508057, + "learning_rate": 9.9901819990182e-06, + "loss": 0.2835, + "step": 9824 + }, + { + "epoch": 4.5676429567642955, + "grad_norm": 7.762726306915283, + "learning_rate": 9.9901809990181e-06, + "loss": 0.4031, + "step": 9825 + }, + { + "epoch": 4.568107856810785, + "grad_norm": 5.896322727203369, + "learning_rate": 9.990179999018e-06, + "loss": 0.4459, + "step": 9826 + }, + { + "epoch": 4.568572756857276, + "grad_norm": 8.361231803894043, + "learning_rate": 9.990178999017901e-06, + "loss": 0.5047, + "step": 9827 + }, + { + "epoch": 4.569037656903766, + "grad_norm": 4.74375581741333, + "learning_rate": 9.9901779990178e-06, + "loss": 0.204, + "step": 9828 + }, + { + "epoch": 4.569502556950256, + "grad_norm": 3.201080560684204, + "learning_rate": 9.990176999017701e-06, + "loss": 0.114, + "step": 9829 + }, + { + "epoch": 4.5699674569967454, + "grad_norm": 7.147031784057617, + "learning_rate": 9.9901759990176e-06, + "loss": 0.5634, + "step": 9830 + }, + { + "epoch": 4.570432357043236, + "grad_norm": 4.230187892913818, + "learning_rate": 9.9901749990175e-06, + "loss": 0.2479, + "step": 9831 + }, + { + "epoch": 4.570897257089726, + "grad_norm": 4.739376544952393, + "learning_rate": 9.990173999017401e-06, + "loss": 0.2465, + "step": 9832 + }, + { + "epoch": 4.571362157136216, + "grad_norm": 4.361557483673096, + "learning_rate": 9.9901729990173e-06, + "loss": 0.2498, + "step": 9833 + }, + { + "epoch": 4.5718270571827055, + "grad_norm": 3.711965322494507, + "learning_rate": 9.990171999017201e-06, + "loss": 0.2203, + "step": 9834 + }, + { + "epoch": 4.572291957229195, + "grad_norm": 8.510491371154785, + "learning_rate": 9.9901709990171e-06, + "loss": 0.5078, + "step": 9835 + }, + { + "epoch": 4.572756857275686, + "grad_norm": 6.793416976928711, + "learning_rate": 9.990169999017e-06, + "loss": 0.3345, + "step": 9836 + }, + { + "epoch": 4.573221757322176, + "grad_norm": 3.7524335384368896, + "learning_rate": 9.990168999016901e-06, + "loss": 0.1634, + "step": 9837 + }, + { + "epoch": 4.573686657368666, + "grad_norm": 4.536583423614502, + "learning_rate": 9.9901679990168e-06, + "loss": 0.2152, + "step": 9838 + }, + { + "epoch": 4.5741515574151554, + "grad_norm": 6.696046829223633, + "learning_rate": 9.990166999016701e-06, + "loss": 0.2627, + "step": 9839 + }, + { + "epoch": 4.574616457461646, + "grad_norm": 6.201166152954102, + "learning_rate": 9.990165999016602e-06, + "loss": 0.3235, + "step": 9840 + }, + { + "epoch": 4.575081357508136, + "grad_norm": 4.359444618225098, + "learning_rate": 9.9901649990165e-06, + "loss": 0.1812, + "step": 9841 + }, + { + "epoch": 4.575546257554626, + "grad_norm": 5.356270790100098, + "learning_rate": 9.990163999016401e-06, + "loss": 0.2909, + "step": 9842 + }, + { + "epoch": 4.5760111576011155, + "grad_norm": 9.451204299926758, + "learning_rate": 9.9901629990163e-06, + "loss": 0.2996, + "step": 9843 + }, + { + "epoch": 4.576476057647605, + "grad_norm": 5.4078192710876465, + "learning_rate": 9.9901619990162e-06, + "loss": 0.2604, + "step": 9844 + }, + { + "epoch": 4.576940957694096, + "grad_norm": 4.083951473236084, + "learning_rate": 9.990160999016101e-06, + "loss": 0.1963, + "step": 9845 + }, + { + "epoch": 4.577405857740586, + "grad_norm": 5.62297248840332, + "learning_rate": 9.990159999016e-06, + "loss": 0.5511, + "step": 9846 + }, + { + "epoch": 4.577870757787076, + "grad_norm": 6.285838603973389, + "learning_rate": 9.990158999015901e-06, + "loss": 0.3914, + "step": 9847 + }, + { + "epoch": 4.578335657833565, + "grad_norm": 6.000421047210693, + "learning_rate": 9.9901579990158e-06, + "loss": 0.333, + "step": 9848 + }, + { + "epoch": 4.578800557880056, + "grad_norm": 6.456029891967773, + "learning_rate": 9.9901569990157e-06, + "loss": 0.1523, + "step": 9849 + }, + { + "epoch": 4.579265457926546, + "grad_norm": 6.772982120513916, + "learning_rate": 9.990155999015601e-06, + "loss": 0.2886, + "step": 9850 + }, + { + "epoch": 4.579730357973036, + "grad_norm": 4.549906253814697, + "learning_rate": 9.9901549990155e-06, + "loss": 0.1739, + "step": 9851 + }, + { + "epoch": 4.5801952580195255, + "grad_norm": 4.9693074226379395, + "learning_rate": 9.990153999015401e-06, + "loss": 0.3088, + "step": 9852 + }, + { + "epoch": 4.580660158066015, + "grad_norm": 4.182807445526123, + "learning_rate": 9.990152999015302e-06, + "loss": 0.1316, + "step": 9853 + }, + { + "epoch": 4.581125058112506, + "grad_norm": 7.225632190704346, + "learning_rate": 9.9901519990152e-06, + "loss": 0.5021, + "step": 9854 + }, + { + "epoch": 4.581589958158996, + "grad_norm": 6.704801559448242, + "learning_rate": 9.9901509990151e-06, + "loss": 0.2105, + "step": 9855 + }, + { + "epoch": 4.582054858205486, + "grad_norm": 5.9185004234313965, + "learning_rate": 9.990149999015e-06, + "loss": 0.2073, + "step": 9856 + }, + { + "epoch": 4.582519758251975, + "grad_norm": 4.988205909729004, + "learning_rate": 9.9901489990149e-06, + "loss": 0.2649, + "step": 9857 + }, + { + "epoch": 4.582984658298466, + "grad_norm": 6.5569610595703125, + "learning_rate": 9.990147999014801e-06, + "loss": 0.4296, + "step": 9858 + }, + { + "epoch": 4.583449558344956, + "grad_norm": 6.726095676422119, + "learning_rate": 9.9901469990147e-06, + "loss": 0.4272, + "step": 9859 + }, + { + "epoch": 4.583914458391446, + "grad_norm": 6.036686420440674, + "learning_rate": 9.990145999014601e-06, + "loss": 0.2053, + "step": 9860 + }, + { + "epoch": 4.5843793584379355, + "grad_norm": 10.155668258666992, + "learning_rate": 9.9901449990145e-06, + "loss": 0.272, + "step": 9861 + }, + { + "epoch": 4.584844258484426, + "grad_norm": 7.315460205078125, + "learning_rate": 9.9901439990144e-06, + "loss": 0.4945, + "step": 9862 + }, + { + "epoch": 4.585309158530916, + "grad_norm": 4.732814311981201, + "learning_rate": 9.990142999014301e-06, + "loss": 0.2556, + "step": 9863 + }, + { + "epoch": 4.585774058577406, + "grad_norm": 3.1694977283477783, + "learning_rate": 9.990141999014202e-06, + "loss": 0.1171, + "step": 9864 + }, + { + "epoch": 4.586238958623896, + "grad_norm": 6.63001012802124, + "learning_rate": 9.990140999014101e-06, + "loss": 0.4426, + "step": 9865 + }, + { + "epoch": 4.586703858670386, + "grad_norm": 5.6373395919799805, + "learning_rate": 9.990139999014e-06, + "loss": 0.1936, + "step": 9866 + }, + { + "epoch": 4.587168758716876, + "grad_norm": 7.075153350830078, + "learning_rate": 9.9901389990139e-06, + "loss": 0.4868, + "step": 9867 + }, + { + "epoch": 4.587633658763366, + "grad_norm": 6.817625522613525, + "learning_rate": 9.9901379990138e-06, + "loss": 0.5413, + "step": 9868 + }, + { + "epoch": 4.588098558809856, + "grad_norm": 4.861026763916016, + "learning_rate": 9.9901369990137e-06, + "loss": 0.3423, + "step": 9869 + }, + { + "epoch": 4.5885634588563455, + "grad_norm": 7.53151798248291, + "learning_rate": 9.990135999013601e-06, + "loss": 0.6093, + "step": 9870 + }, + { + "epoch": 4.589028358902836, + "grad_norm": 5.53064489364624, + "learning_rate": 9.990134999013502e-06, + "loss": 0.305, + "step": 9871 + }, + { + "epoch": 4.589493258949326, + "grad_norm": 5.205107688903809, + "learning_rate": 9.9901339990134e-06, + "loss": 0.2323, + "step": 9872 + }, + { + "epoch": 4.589958158995816, + "grad_norm": 5.915637969970703, + "learning_rate": 9.990132999013301e-06, + "loss": 0.3696, + "step": 9873 + }, + { + "epoch": 4.590423059042306, + "grad_norm": 8.188227653503418, + "learning_rate": 9.9901319990132e-06, + "loss": 0.3117, + "step": 9874 + }, + { + "epoch": 4.590887959088796, + "grad_norm": 7.2137227058410645, + "learning_rate": 9.9901309990131e-06, + "loss": 0.5192, + "step": 9875 + }, + { + "epoch": 4.591352859135286, + "grad_norm": 6.484154224395752, + "learning_rate": 9.990129999013001e-06, + "loss": 0.3862, + "step": 9876 + }, + { + "epoch": 4.591817759181776, + "grad_norm": 10.0237455368042, + "learning_rate": 9.9901289990129e-06, + "loss": 0.5619, + "step": 9877 + }, + { + "epoch": 4.592282659228266, + "grad_norm": 9.93196964263916, + "learning_rate": 9.990127999012801e-06, + "loss": 0.2668, + "step": 9878 + }, + { + "epoch": 4.5927475592747555, + "grad_norm": 6.1892900466918945, + "learning_rate": 9.9901269990127e-06, + "loss": 0.3073, + "step": 9879 + }, + { + "epoch": 4.593212459321246, + "grad_norm": 3.8001508712768555, + "learning_rate": 9.9901259990126e-06, + "loss": 0.202, + "step": 9880 + }, + { + "epoch": 4.593677359367736, + "grad_norm": 6.122893333435059, + "learning_rate": 9.9901249990125e-06, + "loss": 0.3795, + "step": 9881 + }, + { + "epoch": 4.594142259414226, + "grad_norm": 4.140259742736816, + "learning_rate": 9.9901239990124e-06, + "loss": 0.224, + "step": 9882 + }, + { + "epoch": 4.594607159460716, + "grad_norm": 2.7351417541503906, + "learning_rate": 9.990122999012301e-06, + "loss": 0.0968, + "step": 9883 + }, + { + "epoch": 4.595072059507206, + "grad_norm": 2.9134652614593506, + "learning_rate": 9.990121999012202e-06, + "loss": 0.1141, + "step": 9884 + }, + { + "epoch": 4.595536959553696, + "grad_norm": 6.950939655303955, + "learning_rate": 9.9901209990121e-06, + "loss": 0.2101, + "step": 9885 + }, + { + "epoch": 4.596001859600186, + "grad_norm": 12.501660346984863, + "learning_rate": 9.990119999012001e-06, + "loss": 0.6091, + "step": 9886 + }, + { + "epoch": 4.596466759646676, + "grad_norm": 5.938127040863037, + "learning_rate": 9.9901189990119e-06, + "loss": 0.3225, + "step": 9887 + }, + { + "epoch": 4.5969316596931655, + "grad_norm": 4.265416622161865, + "learning_rate": 9.990117999011801e-06, + "loss": 0.1652, + "step": 9888 + }, + { + "epoch": 4.597396559739656, + "grad_norm": 6.143935680389404, + "learning_rate": 9.990116999011702e-06, + "loss": 0.2427, + "step": 9889 + }, + { + "epoch": 4.597861459786146, + "grad_norm": 4.776899814605713, + "learning_rate": 9.9901159990116e-06, + "loss": 0.2114, + "step": 9890 + }, + { + "epoch": 4.598326359832636, + "grad_norm": 7.035226821899414, + "learning_rate": 9.990114999011501e-06, + "loss": 0.2026, + "step": 9891 + }, + { + "epoch": 4.598791259879126, + "grad_norm": 4.719570159912109, + "learning_rate": 9.9901139990114e-06, + "loss": 0.1997, + "step": 9892 + }, + { + "epoch": 4.599256159925616, + "grad_norm": 4.987359046936035, + "learning_rate": 9.9901129990113e-06, + "loss": 0.3148, + "step": 9893 + }, + { + "epoch": 4.599721059972106, + "grad_norm": 6.826108932495117, + "learning_rate": 9.9901119990112e-06, + "loss": 0.4696, + "step": 9894 + }, + { + "epoch": 4.600185960018596, + "grad_norm": 8.679643630981445, + "learning_rate": 9.9901109990111e-06, + "loss": 0.2325, + "step": 9895 + }, + { + "epoch": 4.600650860065086, + "grad_norm": 3.9828405380249023, + "learning_rate": 9.990109999011001e-06, + "loss": 0.1545, + "step": 9896 + }, + { + "epoch": 4.6011157601115755, + "grad_norm": 5.383347988128662, + "learning_rate": 9.990108999010902e-06, + "loss": 0.2418, + "step": 9897 + }, + { + "epoch": 4.601580660158066, + "grad_norm": 6.117810249328613, + "learning_rate": 9.9901079990108e-06, + "loss": 0.1849, + "step": 9898 + }, + { + "epoch": 4.602045560204556, + "grad_norm": 7.060995578765869, + "learning_rate": 9.9901069990107e-06, + "loss": 0.1904, + "step": 9899 + }, + { + "epoch": 4.602510460251046, + "grad_norm": 5.35364294052124, + "learning_rate": 9.9901059990106e-06, + "loss": 0.3223, + "step": 9900 + }, + { + "epoch": 4.6029753602975365, + "grad_norm": 4.4571943283081055, + "learning_rate": 9.990104999010501e-06, + "loss": 0.177, + "step": 9901 + }, + { + "epoch": 4.603440260344026, + "grad_norm": 7.893532752990723, + "learning_rate": 9.990103999010402e-06, + "loss": 0.3672, + "step": 9902 + }, + { + "epoch": 4.603905160390516, + "grad_norm": 13.113173484802246, + "learning_rate": 9.9901029990103e-06, + "loss": 0.5315, + "step": 9903 + }, + { + "epoch": 4.604370060437006, + "grad_norm": 6.3094282150268555, + "learning_rate": 9.990101999010201e-06, + "loss": 0.3227, + "step": 9904 + }, + { + "epoch": 4.604834960483496, + "grad_norm": 10.11447525024414, + "learning_rate": 9.9901009990101e-06, + "loss": 0.2309, + "step": 9905 + }, + { + "epoch": 4.605299860529986, + "grad_norm": 4.064146995544434, + "learning_rate": 9.990099999010001e-06, + "loss": 0.1784, + "step": 9906 + }, + { + "epoch": 4.605764760576476, + "grad_norm": 6.9054059982299805, + "learning_rate": 9.9900989990099e-06, + "loss": 0.5288, + "step": 9907 + }, + { + "epoch": 4.606229660622966, + "grad_norm": 3.7099502086639404, + "learning_rate": 9.9900979990098e-06, + "loss": 0.1093, + "step": 9908 + }, + { + "epoch": 4.606694560669456, + "grad_norm": 7.481320381164551, + "learning_rate": 9.990096999009701e-06, + "loss": 0.2376, + "step": 9909 + }, + { + "epoch": 4.6071594607159465, + "grad_norm": 7.725322246551514, + "learning_rate": 9.9900959990096e-06, + "loss": 0.4317, + "step": 9910 + }, + { + "epoch": 4.607624360762436, + "grad_norm": 4.495022773742676, + "learning_rate": 9.9900949990095e-06, + "loss": 0.2268, + "step": 9911 + }, + { + "epoch": 4.608089260808926, + "grad_norm": 5.0628156661987305, + "learning_rate": 9.9900939990094e-06, + "loss": 0.199, + "step": 9912 + }, + { + "epoch": 4.608554160855416, + "grad_norm": 4.920501232147217, + "learning_rate": 9.9900929990093e-06, + "loss": 0.1241, + "step": 9913 + }, + { + "epoch": 4.609019060901906, + "grad_norm": 7.14857816696167, + "learning_rate": 9.990091999009201e-06, + "loss": 0.4791, + "step": 9914 + }, + { + "epoch": 4.609483960948396, + "grad_norm": 4.603165149688721, + "learning_rate": 9.990090999009102e-06, + "loss": 0.2415, + "step": 9915 + }, + { + "epoch": 4.609948860994886, + "grad_norm": 5.159396648406982, + "learning_rate": 9.990089999009e-06, + "loss": 0.3772, + "step": 9916 + }, + { + "epoch": 4.610413761041376, + "grad_norm": 2.7627453804016113, + "learning_rate": 9.990088999008901e-06, + "loss": 0.0732, + "step": 9917 + }, + { + "epoch": 4.610878661087866, + "grad_norm": 4.2941060066223145, + "learning_rate": 9.9900879990088e-06, + "loss": 0.1758, + "step": 9918 + }, + { + "epoch": 4.6113435611343565, + "grad_norm": 4.9388861656188965, + "learning_rate": 9.990086999008701e-06, + "loss": 0.1471, + "step": 9919 + }, + { + "epoch": 4.611808461180846, + "grad_norm": 6.929171562194824, + "learning_rate": 9.9900859990086e-06, + "loss": 0.3845, + "step": 9920 + }, + { + "epoch": 4.612273361227336, + "grad_norm": 5.064842700958252, + "learning_rate": 9.9900849990085e-06, + "loss": 0.2577, + "step": 9921 + }, + { + "epoch": 4.612738261273826, + "grad_norm": 5.7552103996276855, + "learning_rate": 9.990083999008401e-06, + "loss": 0.19, + "step": 9922 + }, + { + "epoch": 4.613203161320316, + "grad_norm": 7.172431945800781, + "learning_rate": 9.9900829990083e-06, + "loss": 0.4528, + "step": 9923 + }, + { + "epoch": 4.613668061366806, + "grad_norm": 7.3805766105651855, + "learning_rate": 9.990081999008201e-06, + "loss": 0.3399, + "step": 9924 + }, + { + "epoch": 4.614132961413296, + "grad_norm": 7.596665382385254, + "learning_rate": 9.9900809990081e-06, + "loss": 0.4702, + "step": 9925 + }, + { + "epoch": 4.614597861459786, + "grad_norm": 8.405858993530273, + "learning_rate": 9.990079999008e-06, + "loss": 0.3615, + "step": 9926 + }, + { + "epoch": 4.615062761506276, + "grad_norm": 5.563106536865234, + "learning_rate": 9.990078999007901e-06, + "loss": 0.2974, + "step": 9927 + }, + { + "epoch": 4.6155276615527665, + "grad_norm": 6.441150665283203, + "learning_rate": 9.990077999007802e-06, + "loss": 0.3726, + "step": 9928 + }, + { + "epoch": 4.615992561599256, + "grad_norm": 3.8875327110290527, + "learning_rate": 9.9900769990077e-06, + "loss": 0.161, + "step": 9929 + }, + { + "epoch": 4.616457461645746, + "grad_norm": 5.215025424957275, + "learning_rate": 9.9900759990076e-06, + "loss": 0.2696, + "step": 9930 + }, + { + "epoch": 4.616922361692236, + "grad_norm": 5.941627025604248, + "learning_rate": 9.9900749990075e-06, + "loss": 0.2736, + "step": 9931 + }, + { + "epoch": 4.617387261738726, + "grad_norm": 4.92324161529541, + "learning_rate": 9.990073999007401e-06, + "loss": 0.1081, + "step": 9932 + }, + { + "epoch": 4.617852161785216, + "grad_norm": 5.396844387054443, + "learning_rate": 9.990072999007302e-06, + "loss": 0.3129, + "step": 9933 + }, + { + "epoch": 4.618317061831706, + "grad_norm": 6.771300315856934, + "learning_rate": 9.9900719990072e-06, + "loss": 0.2264, + "step": 9934 + }, + { + "epoch": 4.618781961878196, + "grad_norm": 5.304585933685303, + "learning_rate": 9.990070999007101e-06, + "loss": 0.2947, + "step": 9935 + }, + { + "epoch": 4.619246861924686, + "grad_norm": 3.9442265033721924, + "learning_rate": 9.990069999007e-06, + "loss": 0.2283, + "step": 9936 + }, + { + "epoch": 4.6197117619711765, + "grad_norm": 4.271894931793213, + "learning_rate": 9.990068999006901e-06, + "loss": 0.1576, + "step": 9937 + }, + { + "epoch": 4.620176662017666, + "grad_norm": 3.9841041564941406, + "learning_rate": 9.9900679990068e-06, + "loss": 0.1913, + "step": 9938 + }, + { + "epoch": 4.620641562064156, + "grad_norm": 7.034554958343506, + "learning_rate": 9.9900669990067e-06, + "loss": 0.2302, + "step": 9939 + }, + { + "epoch": 4.621106462110646, + "grad_norm": 4.628777980804443, + "learning_rate": 9.990065999006601e-06, + "loss": 0.2769, + "step": 9940 + }, + { + "epoch": 4.621571362157137, + "grad_norm": 5.517336368560791, + "learning_rate": 9.9900649990065e-06, + "loss": 0.385, + "step": 9941 + }, + { + "epoch": 4.622036262203626, + "grad_norm": 5.074097633361816, + "learning_rate": 9.9900639990064e-06, + "loss": 0.2688, + "step": 9942 + }, + { + "epoch": 4.622501162250116, + "grad_norm": 5.004227161407471, + "learning_rate": 9.9900629990063e-06, + "loss": 0.2205, + "step": 9943 + }, + { + "epoch": 4.622966062296606, + "grad_norm": 7.325369358062744, + "learning_rate": 9.9900619990062e-06, + "loss": 0.6636, + "step": 9944 + }, + { + "epoch": 4.623430962343097, + "grad_norm": 5.915350914001465, + "learning_rate": 9.990060999006101e-06, + "loss": 0.4287, + "step": 9945 + }, + { + "epoch": 4.6238958623895865, + "grad_norm": 7.22466516494751, + "learning_rate": 9.990059999006002e-06, + "loss": 0.5609, + "step": 9946 + }, + { + "epoch": 4.624360762436076, + "grad_norm": 5.788186550140381, + "learning_rate": 9.9900589990059e-06, + "loss": 0.3766, + "step": 9947 + }, + { + "epoch": 4.624825662482566, + "grad_norm": 3.5520193576812744, + "learning_rate": 9.990057999005801e-06, + "loss": 0.134, + "step": 9948 + }, + { + "epoch": 4.625290562529056, + "grad_norm": 6.438917636871338, + "learning_rate": 9.9900569990057e-06, + "loss": 0.384, + "step": 9949 + }, + { + "epoch": 4.625755462575547, + "grad_norm": 3.9809484481811523, + "learning_rate": 9.990055999005601e-06, + "loss": 0.2251, + "step": 9950 + }, + { + "epoch": 4.626220362622036, + "grad_norm": 5.19096040725708, + "learning_rate": 9.9900549990055e-06, + "loss": 0.3359, + "step": 9951 + }, + { + "epoch": 4.626685262668526, + "grad_norm": 5.781331539154053, + "learning_rate": 9.9900539990054e-06, + "loss": 0.3029, + "step": 9952 + }, + { + "epoch": 4.627150162715016, + "grad_norm": 11.600139617919922, + "learning_rate": 9.990052999005301e-06, + "loss": 0.4256, + "step": 9953 + }, + { + "epoch": 4.627615062761507, + "grad_norm": 5.056423664093018, + "learning_rate": 9.9900519990052e-06, + "loss": 0.2356, + "step": 9954 + }, + { + "epoch": 4.6280799628079965, + "grad_norm": 4.093839168548584, + "learning_rate": 9.990050999005101e-06, + "loss": 0.1738, + "step": 9955 + }, + { + "epoch": 4.628544862854486, + "grad_norm": 5.9289937019348145, + "learning_rate": 9.990049999005e-06, + "loss": 0.3399, + "step": 9956 + }, + { + "epoch": 4.629009762900976, + "grad_norm": 4.994576454162598, + "learning_rate": 9.9900489990049e-06, + "loss": 0.2698, + "step": 9957 + }, + { + "epoch": 4.629474662947466, + "grad_norm": 6.250811576843262, + "learning_rate": 9.990047999004801e-06, + "loss": 0.3389, + "step": 9958 + }, + { + "epoch": 4.629939562993957, + "grad_norm": 7.065188407897949, + "learning_rate": 9.990046999004702e-06, + "loss": 0.4879, + "step": 9959 + }, + { + "epoch": 4.630404463040446, + "grad_norm": 7.68949031829834, + "learning_rate": 9.9900459990046e-06, + "loss": 0.2536, + "step": 9960 + }, + { + "epoch": 4.630869363086936, + "grad_norm": 6.364133358001709, + "learning_rate": 9.990044999004501e-06, + "loss": 0.3877, + "step": 9961 + }, + { + "epoch": 4.631334263133426, + "grad_norm": 9.774145126342773, + "learning_rate": 9.9900439990044e-06, + "loss": 0.4234, + "step": 9962 + }, + { + "epoch": 4.631799163179917, + "grad_norm": 7.602027416229248, + "learning_rate": 9.9900429990043e-06, + "loss": 0.6025, + "step": 9963 + }, + { + "epoch": 4.6322640632264065, + "grad_norm": 4.552061080932617, + "learning_rate": 9.9900419990042e-06, + "loss": 0.1893, + "step": 9964 + }, + { + "epoch": 4.632728963272896, + "grad_norm": 5.195285797119141, + "learning_rate": 9.9900409990041e-06, + "loss": 0.3264, + "step": 9965 + }, + { + "epoch": 4.633193863319386, + "grad_norm": 6.087814807891846, + "learning_rate": 9.990039999004001e-06, + "loss": 0.2693, + "step": 9966 + }, + { + "epoch": 4.633658763365876, + "grad_norm": 6.767018795013428, + "learning_rate": 9.9900389990039e-06, + "loss": 0.371, + "step": 9967 + }, + { + "epoch": 4.634123663412367, + "grad_norm": 6.278478622436523, + "learning_rate": 9.990037999003801e-06, + "loss": 0.104, + "step": 9968 + }, + { + "epoch": 4.634588563458856, + "grad_norm": 5.508037090301514, + "learning_rate": 9.9900369990037e-06, + "loss": 0.164, + "step": 9969 + }, + { + "epoch": 4.635053463505346, + "grad_norm": 4.744035720825195, + "learning_rate": 9.9900359990036e-06, + "loss": 0.2286, + "step": 9970 + }, + { + "epoch": 4.635518363551836, + "grad_norm": 6.1795973777771, + "learning_rate": 9.990034999003501e-06, + "loss": 0.3907, + "step": 9971 + }, + { + "epoch": 4.635983263598327, + "grad_norm": 4.883414268493652, + "learning_rate": 9.990033999003402e-06, + "loss": 0.2018, + "step": 9972 + }, + { + "epoch": 4.6364481636448165, + "grad_norm": 3.8181445598602295, + "learning_rate": 9.990032999003301e-06, + "loss": 0.2017, + "step": 9973 + }, + { + "epoch": 4.636913063691306, + "grad_norm": 8.137370109558105, + "learning_rate": 9.9900319990032e-06, + "loss": 0.3922, + "step": 9974 + }, + { + "epoch": 4.637377963737796, + "grad_norm": 9.298128128051758, + "learning_rate": 9.9900309990031e-06, + "loss": 0.1942, + "step": 9975 + }, + { + "epoch": 4.637842863784286, + "grad_norm": 5.756299018859863, + "learning_rate": 9.990029999003e-06, + "loss": 0.2459, + "step": 9976 + }, + { + "epoch": 4.638307763830777, + "grad_norm": 5.353652000427246, + "learning_rate": 9.9900289990029e-06, + "loss": 0.2674, + "step": 9977 + }, + { + "epoch": 4.638772663877266, + "grad_norm": 7.421026706695557, + "learning_rate": 9.9900279990028e-06, + "loss": 0.4591, + "step": 9978 + }, + { + "epoch": 4.639237563923756, + "grad_norm": 4.9482316970825195, + "learning_rate": 9.990026999002701e-06, + "loss": 0.2143, + "step": 9979 + }, + { + "epoch": 4.639702463970247, + "grad_norm": 5.192266464233398, + "learning_rate": 9.9900259990026e-06, + "loss": 0.3308, + "step": 9980 + }, + { + "epoch": 4.640167364016737, + "grad_norm": 5.253842830657959, + "learning_rate": 9.990024999002501e-06, + "loss": 0.3715, + "step": 9981 + }, + { + "epoch": 4.6406322640632265, + "grad_norm": 10.867000579833984, + "learning_rate": 9.9900239990024e-06, + "loss": 0.6818, + "step": 9982 + }, + { + "epoch": 4.641097164109716, + "grad_norm": 5.154342174530029, + "learning_rate": 9.9900229990023e-06, + "loss": 0.2982, + "step": 9983 + }, + { + "epoch": 4.641562064156206, + "grad_norm": 7.224852085113525, + "learning_rate": 9.990021999002201e-06, + "loss": 0.2418, + "step": 9984 + }, + { + "epoch": 4.642026964202697, + "grad_norm": 9.98491096496582, + "learning_rate": 9.9900209990021e-06, + "loss": 0.4113, + "step": 9985 + }, + { + "epoch": 4.642491864249187, + "grad_norm": 5.604720592498779, + "learning_rate": 9.990019999002001e-06, + "loss": 0.3092, + "step": 9986 + }, + { + "epoch": 4.642956764295676, + "grad_norm": 14.885810852050781, + "learning_rate": 9.9900189990019e-06, + "loss": 0.1476, + "step": 9987 + }, + { + "epoch": 4.643421664342166, + "grad_norm": 1.9427903890609741, + "learning_rate": 9.9900179990018e-06, + "loss": 0.0332, + "step": 9988 + }, + { + "epoch": 4.643886564388657, + "grad_norm": 4.760478973388672, + "learning_rate": 9.9900169990017e-06, + "loss": 0.2736, + "step": 9989 + }, + { + "epoch": 4.644351464435147, + "grad_norm": 10.003270149230957, + "learning_rate": 9.9900159990016e-06, + "loss": 0.3096, + "step": 9990 + }, + { + "epoch": 4.6448163644816365, + "grad_norm": 5.008742332458496, + "learning_rate": 9.990014999001501e-06, + "loss": 0.1944, + "step": 9991 + }, + { + "epoch": 4.645281264528126, + "grad_norm": 7.06533145904541, + "learning_rate": 9.990013999001402e-06, + "loss": 0.2071, + "step": 9992 + }, + { + "epoch": 4.645746164574616, + "grad_norm": 10.087481498718262, + "learning_rate": 9.9900129990013e-06, + "loss": 0.503, + "step": 9993 + }, + { + "epoch": 4.646211064621107, + "grad_norm": 3.8420324325561523, + "learning_rate": 9.990011999001201e-06, + "loss": 0.2828, + "step": 9994 + }, + { + "epoch": 4.646675964667597, + "grad_norm": 6.961453914642334, + "learning_rate": 9.9900109990011e-06, + "loss": 0.4602, + "step": 9995 + }, + { + "epoch": 4.647140864714086, + "grad_norm": 6.430562973022461, + "learning_rate": 9.990009999001e-06, + "loss": 0.5125, + "step": 9996 + }, + { + "epoch": 4.647605764760576, + "grad_norm": 6.465627670288086, + "learning_rate": 9.990008999000901e-06, + "loss": 0.3076, + "step": 9997 + }, + { + "epoch": 4.648070664807067, + "grad_norm": 5.87154483795166, + "learning_rate": 9.9900079990008e-06, + "loss": 0.1934, + "step": 9998 + }, + { + "epoch": 4.648535564853557, + "grad_norm": 4.928010940551758, + "learning_rate": 9.990006999000701e-06, + "loss": 0.2911, + "step": 9999 + }, + { + "epoch": 4.6490004649000465, + "grad_norm": 5.482219219207764, + "learning_rate": 9.9900059990006e-06, + "loss": 0.2413, + "step": 10000 + }, + { + "epoch": 4.649465364946536, + "grad_norm": 6.5229668617248535, + "learning_rate": 9.9900049990005e-06, + "loss": 0.6839, + "step": 10001 + }, + { + "epoch": 4.649930264993026, + "grad_norm": 3.674919843673706, + "learning_rate": 9.990003999000401e-06, + "loss": 0.0927, + "step": 10002 + }, + { + "epoch": 4.650395165039517, + "grad_norm": 4.0809326171875, + "learning_rate": 9.9900029990003e-06, + "loss": 0.2161, + "step": 10003 + }, + { + "epoch": 4.650860065086007, + "grad_norm": 4.066190719604492, + "learning_rate": 9.990001999000201e-06, + "loss": 0.194, + "step": 10004 + }, + { + "epoch": 4.651324965132496, + "grad_norm": 6.210089206695557, + "learning_rate": 9.990000999000102e-06, + "loss": 0.0744, + "step": 10005 + }, + { + "epoch": 4.651789865178986, + "grad_norm": 4.921350479125977, + "learning_rate": 9.989999999e-06, + "loss": 0.2431, + "step": 10006 + }, + { + "epoch": 4.652254765225477, + "grad_norm": 4.811419486999512, + "learning_rate": 9.9899989989999e-06, + "loss": 0.3042, + "step": 10007 + }, + { + "epoch": 4.652719665271967, + "grad_norm": 4.524742603302002, + "learning_rate": 9.9899979989998e-06, + "loss": 0.25, + "step": 10008 + }, + { + "epoch": 4.6531845653184565, + "grad_norm": 7.781065940856934, + "learning_rate": 9.989996998999701e-06, + "loss": 0.2078, + "step": 10009 + }, + { + "epoch": 4.653649465364946, + "grad_norm": 10.205381393432617, + "learning_rate": 9.989995998999602e-06, + "loss": 0.786, + "step": 10010 + }, + { + "epoch": 4.654114365411436, + "grad_norm": 3.8357696533203125, + "learning_rate": 9.9899949989995e-06, + "loss": 0.1892, + "step": 10011 + }, + { + "epoch": 4.654579265457927, + "grad_norm": 4.298732280731201, + "learning_rate": 9.989993998999401e-06, + "loss": 0.2082, + "step": 10012 + }, + { + "epoch": 4.655044165504417, + "grad_norm": 5.309329032897949, + "learning_rate": 9.9899929989993e-06, + "loss": 0.4328, + "step": 10013 + }, + { + "epoch": 4.655509065550906, + "grad_norm": 7.149444580078125, + "learning_rate": 9.9899919989992e-06, + "loss": 0.4565, + "step": 10014 + }, + { + "epoch": 4.655973965597396, + "grad_norm": 8.694665908813477, + "learning_rate": 9.989990998999101e-06, + "loss": 0.5988, + "step": 10015 + }, + { + "epoch": 4.656438865643887, + "grad_norm": 5.744207859039307, + "learning_rate": 9.989989998999e-06, + "loss": 0.2744, + "step": 10016 + }, + { + "epoch": 4.656903765690377, + "grad_norm": 5.148853302001953, + "learning_rate": 9.989988998998901e-06, + "loss": 0.215, + "step": 10017 + }, + { + "epoch": 4.6573686657368665, + "grad_norm": 5.931779384613037, + "learning_rate": 9.9899879989988e-06, + "loss": 0.291, + "step": 10018 + }, + { + "epoch": 4.657833565783356, + "grad_norm": 6.384768486022949, + "learning_rate": 9.9899869989987e-06, + "loss": 0.4923, + "step": 10019 + }, + { + "epoch": 4.658298465829847, + "grad_norm": 4.437007427215576, + "learning_rate": 9.9899859989986e-06, + "loss": 0.2742, + "step": 10020 + }, + { + "epoch": 4.658763365876337, + "grad_norm": 5.914467811584473, + "learning_rate": 9.9899849989985e-06, + "loss": 0.4981, + "step": 10021 + }, + { + "epoch": 4.659228265922827, + "grad_norm": 12.50883960723877, + "learning_rate": 9.989983998998401e-06, + "loss": 0.4561, + "step": 10022 + }, + { + "epoch": 4.659693165969316, + "grad_norm": 4.040650844573975, + "learning_rate": 9.989982998998302e-06, + "loss": 0.1433, + "step": 10023 + }, + { + "epoch": 4.660158066015807, + "grad_norm": 9.842466354370117, + "learning_rate": 9.9899819989982e-06, + "loss": 0.3869, + "step": 10024 + }, + { + "epoch": 4.660622966062297, + "grad_norm": 6.331165313720703, + "learning_rate": 9.989980998998101e-06, + "loss": 0.4645, + "step": 10025 + }, + { + "epoch": 4.661087866108787, + "grad_norm": 5.1406402587890625, + "learning_rate": 9.989979998998e-06, + "loss": 0.3178, + "step": 10026 + }, + { + "epoch": 4.6615527661552765, + "grad_norm": 11.097774505615234, + "learning_rate": 9.989978998997901e-06, + "loss": 0.4539, + "step": 10027 + }, + { + "epoch": 4.662017666201766, + "grad_norm": 4.482564449310303, + "learning_rate": 9.989977998997802e-06, + "loss": 0.204, + "step": 10028 + }, + { + "epoch": 4.662482566248257, + "grad_norm": 6.38792085647583, + "learning_rate": 9.9899769989977e-06, + "loss": 0.4179, + "step": 10029 + }, + { + "epoch": 4.662947466294747, + "grad_norm": 4.8290839195251465, + "learning_rate": 9.989975998997601e-06, + "loss": 0.293, + "step": 10030 + }, + { + "epoch": 4.663412366341237, + "grad_norm": 5.380973815917969, + "learning_rate": 9.9899749989975e-06, + "loss": 0.3013, + "step": 10031 + }, + { + "epoch": 4.663877266387726, + "grad_norm": 4.544403553009033, + "learning_rate": 9.9899739989974e-06, + "loss": 0.1854, + "step": 10032 + }, + { + "epoch": 4.664342166434217, + "grad_norm": 5.517350196838379, + "learning_rate": 9.9899729989973e-06, + "loss": 0.1818, + "step": 10033 + }, + { + "epoch": 4.664807066480707, + "grad_norm": 2.7143502235412598, + "learning_rate": 9.9899719989972e-06, + "loss": 0.0785, + "step": 10034 + }, + { + "epoch": 4.665271966527197, + "grad_norm": 5.037646293640137, + "learning_rate": 9.989970998997101e-06, + "loss": 0.2086, + "step": 10035 + }, + { + "epoch": 4.6657368665736865, + "grad_norm": 4.595767021179199, + "learning_rate": 9.989969998997002e-06, + "loss": 0.2976, + "step": 10036 + }, + { + "epoch": 4.666201766620176, + "grad_norm": 6.702356338500977, + "learning_rate": 9.9899689989969e-06, + "loss": 0.4468, + "step": 10037 + }, + { + "epoch": 4.666666666666667, + "grad_norm": 4.238291263580322, + "learning_rate": 9.989967998996801e-06, + "loss": 0.2434, + "step": 10038 + }, + { + "epoch": 4.667131566713157, + "grad_norm": 5.662033557891846, + "learning_rate": 9.9899669989967e-06, + "loss": 0.4297, + "step": 10039 + }, + { + "epoch": 4.667596466759647, + "grad_norm": 5.205219268798828, + "learning_rate": 9.989965998996601e-06, + "loss": 0.2692, + "step": 10040 + }, + { + "epoch": 4.668061366806136, + "grad_norm": 3.7984676361083984, + "learning_rate": 9.989964998996502e-06, + "loss": 0.1588, + "step": 10041 + }, + { + "epoch": 4.668526266852627, + "grad_norm": 7.364230632781982, + "learning_rate": 9.9899639989964e-06, + "loss": 0.6, + "step": 10042 + }, + { + "epoch": 4.668991166899117, + "grad_norm": 3.1974689960479736, + "learning_rate": 9.989962998996301e-06, + "loss": 0.1283, + "step": 10043 + }, + { + "epoch": 4.669456066945607, + "grad_norm": 5.23352575302124, + "learning_rate": 9.9899619989962e-06, + "loss": 0.2857, + "step": 10044 + }, + { + "epoch": 4.6699209669920965, + "grad_norm": 4.932262420654297, + "learning_rate": 9.9899609989961e-06, + "loss": 0.2212, + "step": 10045 + }, + { + "epoch": 4.670385867038586, + "grad_norm": 5.8839545249938965, + "learning_rate": 9.989959998996e-06, + "loss": 0.2178, + "step": 10046 + }, + { + "epoch": 4.670850767085077, + "grad_norm": 8.183121681213379, + "learning_rate": 9.9899589989959e-06, + "loss": 0.5075, + "step": 10047 + }, + { + "epoch": 4.671315667131567, + "grad_norm": 5.382597923278809, + "learning_rate": 9.989957998995801e-06, + "loss": 0.4078, + "step": 10048 + }, + { + "epoch": 4.671780567178057, + "grad_norm": 1.9119994640350342, + "learning_rate": 9.989956998995702e-06, + "loss": 0.0681, + "step": 10049 + }, + { + "epoch": 4.672245467224546, + "grad_norm": 3.5565991401672363, + "learning_rate": 9.9899559989956e-06, + "loss": 0.1352, + "step": 10050 + }, + { + "epoch": 4.672710367271037, + "grad_norm": 2.296748638153076, + "learning_rate": 9.9899549989955e-06, + "loss": 0.0822, + "step": 10051 + }, + { + "epoch": 4.673175267317527, + "grad_norm": 5.7103471755981445, + "learning_rate": 9.9899539989954e-06, + "loss": 0.3278, + "step": 10052 + }, + { + "epoch": 4.673640167364017, + "grad_norm": 4.961726188659668, + "learning_rate": 9.989952998995301e-06, + "loss": 0.264, + "step": 10053 + }, + { + "epoch": 4.6741050674105065, + "grad_norm": 6.442715167999268, + "learning_rate": 9.989951998995202e-06, + "loss": 0.6341, + "step": 10054 + }, + { + "epoch": 4.674569967456996, + "grad_norm": 4.071779251098633, + "learning_rate": 9.9899509989951e-06, + "loss": 0.1532, + "step": 10055 + }, + { + "epoch": 4.675034867503487, + "grad_norm": 12.871149063110352, + "learning_rate": 9.989949998995001e-06, + "loss": 0.5229, + "step": 10056 + }, + { + "epoch": 4.675499767549977, + "grad_norm": 8.403976440429688, + "learning_rate": 9.9899489989949e-06, + "loss": 0.3191, + "step": 10057 + }, + { + "epoch": 4.675964667596467, + "grad_norm": 5.963927745819092, + "learning_rate": 9.989947998994801e-06, + "loss": 0.4589, + "step": 10058 + }, + { + "epoch": 4.676429567642957, + "grad_norm": 5.702792644500732, + "learning_rate": 9.9899469989947e-06, + "loss": 0.2056, + "step": 10059 + }, + { + "epoch": 4.676894467689447, + "grad_norm": 7.6041259765625, + "learning_rate": 9.9899459989946e-06, + "loss": 0.7646, + "step": 10060 + }, + { + "epoch": 4.677359367735937, + "grad_norm": 5.844473361968994, + "learning_rate": 9.989944998994501e-06, + "loss": 0.2709, + "step": 10061 + }, + { + "epoch": 4.677824267782427, + "grad_norm": 4.896490573883057, + "learning_rate": 9.9899439989944e-06, + "loss": 0.2078, + "step": 10062 + }, + { + "epoch": 4.6782891678289165, + "grad_norm": 6.7272562980651855, + "learning_rate": 9.9899429989943e-06, + "loss": 0.3049, + "step": 10063 + }, + { + "epoch": 4.678754067875407, + "grad_norm": 5.414809703826904, + "learning_rate": 9.9899419989942e-06, + "loss": 0.3726, + "step": 10064 + }, + { + "epoch": 4.679218967921897, + "grad_norm": 6.0896077156066895, + "learning_rate": 9.9899409989941e-06, + "loss": 0.3892, + "step": 10065 + }, + { + "epoch": 4.679683867968387, + "grad_norm": 9.05351448059082, + "learning_rate": 9.989939998994001e-06, + "loss": 0.5001, + "step": 10066 + }, + { + "epoch": 4.680148768014877, + "grad_norm": 5.524998664855957, + "learning_rate": 9.989938998993902e-06, + "loss": 0.2011, + "step": 10067 + }, + { + "epoch": 4.680613668061367, + "grad_norm": 4.153557777404785, + "learning_rate": 9.9899379989938e-06, + "loss": 0.1411, + "step": 10068 + }, + { + "epoch": 4.681078568107857, + "grad_norm": 5.92114782333374, + "learning_rate": 9.989936998993701e-06, + "loss": 0.3895, + "step": 10069 + }, + { + "epoch": 4.681543468154347, + "grad_norm": 6.466803550720215, + "learning_rate": 9.9899359989936e-06, + "loss": 0.5742, + "step": 10070 + }, + { + "epoch": 4.682008368200837, + "grad_norm": 4.404356002807617, + "learning_rate": 9.989934998993501e-06, + "loss": 0.1602, + "step": 10071 + }, + { + "epoch": 4.6824732682473265, + "grad_norm": 4.399075984954834, + "learning_rate": 9.9899339989934e-06, + "loss": 0.2654, + "step": 10072 + }, + { + "epoch": 4.682938168293817, + "grad_norm": 7.1238813400268555, + "learning_rate": 9.9899329989933e-06, + "loss": 0.3393, + "step": 10073 + }, + { + "epoch": 4.683403068340307, + "grad_norm": 5.918613910675049, + "learning_rate": 9.989931998993201e-06, + "loss": 0.3586, + "step": 10074 + }, + { + "epoch": 4.683867968386797, + "grad_norm": 5.709161281585693, + "learning_rate": 9.9899309989931e-06, + "loss": 0.1304, + "step": 10075 + }, + { + "epoch": 4.684332868433287, + "grad_norm": 5.586008071899414, + "learning_rate": 9.989929998993001e-06, + "loss": 0.2725, + "step": 10076 + }, + { + "epoch": 4.684797768479777, + "grad_norm": 4.618127822875977, + "learning_rate": 9.9899289989929e-06, + "loss": 0.2451, + "step": 10077 + }, + { + "epoch": 4.685262668526267, + "grad_norm": 6.24320650100708, + "learning_rate": 9.9899279989928e-06, + "loss": 0.393, + "step": 10078 + }, + { + "epoch": 4.685727568572757, + "grad_norm": 2.4287471771240234, + "learning_rate": 9.989926998992701e-06, + "loss": 0.0949, + "step": 10079 + }, + { + "epoch": 4.686192468619247, + "grad_norm": 5.417907238006592, + "learning_rate": 9.989925998992602e-06, + "loss": 0.4429, + "step": 10080 + }, + { + "epoch": 4.6866573686657365, + "grad_norm": 5.612710475921631, + "learning_rate": 9.9899249989925e-06, + "loss": 0.2696, + "step": 10081 + }, + { + "epoch": 4.687122268712227, + "grad_norm": 4.866976737976074, + "learning_rate": 9.9899239989924e-06, + "loss": 0.2476, + "step": 10082 + }, + { + "epoch": 4.687587168758717, + "grad_norm": 3.8902783393859863, + "learning_rate": 9.9899229989923e-06, + "loss": 0.1839, + "step": 10083 + }, + { + "epoch": 4.688052068805207, + "grad_norm": 2.696998119354248, + "learning_rate": 9.989921998992201e-06, + "loss": 0.1099, + "step": 10084 + }, + { + "epoch": 4.6885169688516966, + "grad_norm": 5.521135330200195, + "learning_rate": 9.9899209989921e-06, + "loss": 0.2656, + "step": 10085 + }, + { + "epoch": 4.688981868898187, + "grad_norm": 3.7167112827301025, + "learning_rate": 9.989919998992e-06, + "loss": 0.1377, + "step": 10086 + }, + { + "epoch": 4.689446768944677, + "grad_norm": 8.825045585632324, + "learning_rate": 9.989918998991901e-06, + "loss": 0.3723, + "step": 10087 + }, + { + "epoch": 4.689911668991167, + "grad_norm": 7.508024215698242, + "learning_rate": 9.9899179989918e-06, + "loss": 0.57, + "step": 10088 + }, + { + "epoch": 4.690376569037657, + "grad_norm": 4.631096363067627, + "learning_rate": 9.989916998991701e-06, + "loss": 0.292, + "step": 10089 + }, + { + "epoch": 4.6908414690841465, + "grad_norm": 8.559922218322754, + "learning_rate": 9.9899159989916e-06, + "loss": 0.451, + "step": 10090 + }, + { + "epoch": 4.691306369130637, + "grad_norm": 5.556093215942383, + "learning_rate": 9.9899149989915e-06, + "loss": 0.237, + "step": 10091 + }, + { + "epoch": 4.691771269177127, + "grad_norm": 6.0215744972229, + "learning_rate": 9.989913998991401e-06, + "loss": 0.3451, + "step": 10092 + }, + { + "epoch": 4.692236169223617, + "grad_norm": 4.1961236000061035, + "learning_rate": 9.9899129989913e-06, + "loss": 0.2316, + "step": 10093 + }, + { + "epoch": 4.6927010692701066, + "grad_norm": 4.473472595214844, + "learning_rate": 9.989911998991201e-06, + "loss": 0.1965, + "step": 10094 + }, + { + "epoch": 4.693165969316597, + "grad_norm": 6.115008354187012, + "learning_rate": 9.9899109989911e-06, + "loss": 0.3387, + "step": 10095 + }, + { + "epoch": 4.693630869363087, + "grad_norm": 4.3575615882873535, + "learning_rate": 9.989909998991e-06, + "loss": 0.1979, + "step": 10096 + }, + { + "epoch": 4.694095769409577, + "grad_norm": 4.794384002685547, + "learning_rate": 9.989908998990901e-06, + "loss": 0.2621, + "step": 10097 + }, + { + "epoch": 4.694560669456067, + "grad_norm": 6.353578567504883, + "learning_rate": 9.989907998990802e-06, + "loss": 0.3626, + "step": 10098 + }, + { + "epoch": 4.695025569502557, + "grad_norm": 4.63088846206665, + "learning_rate": 9.9899069989907e-06, + "loss": 0.2058, + "step": 10099 + }, + { + "epoch": 4.695490469549047, + "grad_norm": 6.1675333976745605, + "learning_rate": 9.989905998990601e-06, + "loss": 0.2225, + "step": 10100 + }, + { + "epoch": 4.695955369595537, + "grad_norm": 4.948912620544434, + "learning_rate": 9.9899049989905e-06, + "loss": 0.2758, + "step": 10101 + }, + { + "epoch": 4.696420269642027, + "grad_norm": 5.384538650512695, + "learning_rate": 9.989903998990401e-06, + "loss": 0.3402, + "step": 10102 + }, + { + "epoch": 4.696885169688517, + "grad_norm": 5.6933369636535645, + "learning_rate": 9.9899029989903e-06, + "loss": 0.3099, + "step": 10103 + }, + { + "epoch": 4.697350069735007, + "grad_norm": 2.9708709716796875, + "learning_rate": 9.9899019989902e-06, + "loss": 0.1196, + "step": 10104 + }, + { + "epoch": 4.697814969781497, + "grad_norm": 3.36348032951355, + "learning_rate": 9.989900998990101e-06, + "loss": 0.07, + "step": 10105 + }, + { + "epoch": 4.698279869827987, + "grad_norm": 5.895326614379883, + "learning_rate": 9.98989999899e-06, + "loss": 0.33, + "step": 10106 + }, + { + "epoch": 4.698744769874477, + "grad_norm": 5.903149604797363, + "learning_rate": 9.989898998989901e-06, + "loss": 0.2693, + "step": 10107 + }, + { + "epoch": 4.699209669920967, + "grad_norm": 3.211176633834839, + "learning_rate": 9.9898979989898e-06, + "loss": 0.1354, + "step": 10108 + }, + { + "epoch": 4.699674569967457, + "grad_norm": 5.941539764404297, + "learning_rate": 9.9898969989897e-06, + "loss": 0.3471, + "step": 10109 + }, + { + "epoch": 4.700139470013947, + "grad_norm": 3.8898894786834717, + "learning_rate": 9.989895998989601e-06, + "loss": 0.0963, + "step": 10110 + }, + { + "epoch": 4.700604370060437, + "grad_norm": 4.890740394592285, + "learning_rate": 9.989894998989502e-06, + "loss": 0.2963, + "step": 10111 + }, + { + "epoch": 4.701069270106927, + "grad_norm": 6.7644147872924805, + "learning_rate": 9.989893998989401e-06, + "loss": 0.4119, + "step": 10112 + }, + { + "epoch": 4.701534170153417, + "grad_norm": 2.6077821254730225, + "learning_rate": 9.989892998989302e-06, + "loss": 0.0863, + "step": 10113 + }, + { + "epoch": 4.701999070199907, + "grad_norm": 4.900104522705078, + "learning_rate": 9.9898919989892e-06, + "loss": 0.1921, + "step": 10114 + }, + { + "epoch": 4.702463970246397, + "grad_norm": 6.255355358123779, + "learning_rate": 9.9898909989891e-06, + "loss": 0.3095, + "step": 10115 + }, + { + "epoch": 4.702928870292887, + "grad_norm": 4.0261640548706055, + "learning_rate": 9.989889998989e-06, + "loss": 0.2235, + "step": 10116 + }, + { + "epoch": 4.703393770339377, + "grad_norm": 5.719488143920898, + "learning_rate": 9.9898889989889e-06, + "loss": 0.228, + "step": 10117 + }, + { + "epoch": 4.703858670385867, + "grad_norm": 10.645210266113281, + "learning_rate": 9.989887998988801e-06, + "loss": 0.3745, + "step": 10118 + }, + { + "epoch": 4.704323570432357, + "grad_norm": 5.400542259216309, + "learning_rate": 9.9898869989887e-06, + "loss": 0.352, + "step": 10119 + }, + { + "epoch": 4.704788470478847, + "grad_norm": 4.575652599334717, + "learning_rate": 9.989885998988601e-06, + "loss": 0.188, + "step": 10120 + }, + { + "epoch": 4.705253370525337, + "grad_norm": 4.5066657066345215, + "learning_rate": 9.9898849989885e-06, + "loss": 0.3521, + "step": 10121 + }, + { + "epoch": 4.705718270571827, + "grad_norm": 4.368104457855225, + "learning_rate": 9.9898839989884e-06, + "loss": 0.2061, + "step": 10122 + }, + { + "epoch": 4.706183170618317, + "grad_norm": 4.3087897300720215, + "learning_rate": 9.989882998988301e-06, + "loss": 0.1937, + "step": 10123 + }, + { + "epoch": 4.706648070664807, + "grad_norm": 5.850965976715088, + "learning_rate": 9.989881998988202e-06, + "loss": 0.3989, + "step": 10124 + }, + { + "epoch": 4.707112970711297, + "grad_norm": 5.4770989418029785, + "learning_rate": 9.989880998988101e-06, + "loss": 0.3456, + "step": 10125 + }, + { + "epoch": 4.707577870757787, + "grad_norm": 5.073973655700684, + "learning_rate": 9.989879998988e-06, + "loss": 0.2521, + "step": 10126 + }, + { + "epoch": 4.708042770804277, + "grad_norm": 16.630908966064453, + "learning_rate": 9.9898789989879e-06, + "loss": 0.4364, + "step": 10127 + }, + { + "epoch": 4.708507670850767, + "grad_norm": 4.663699150085449, + "learning_rate": 9.9898779989878e-06, + "loss": 0.1814, + "step": 10128 + }, + { + "epoch": 4.708972570897257, + "grad_norm": 5.04444694519043, + "learning_rate": 9.9898769989877e-06, + "loss": 0.3432, + "step": 10129 + }, + { + "epoch": 4.709437470943747, + "grad_norm": 4.627487659454346, + "learning_rate": 9.9898759989876e-06, + "loss": 0.1411, + "step": 10130 + }, + { + "epoch": 4.709902370990237, + "grad_norm": 4.423231601715088, + "learning_rate": 9.989874998987502e-06, + "loss": 0.1418, + "step": 10131 + }, + { + "epoch": 4.710367271036727, + "grad_norm": 8.245758056640625, + "learning_rate": 9.9898739989874e-06, + "loss": 0.4831, + "step": 10132 + }, + { + "epoch": 4.710832171083217, + "grad_norm": 3.1921234130859375, + "learning_rate": 9.989872998987301e-06, + "loss": 0.1895, + "step": 10133 + }, + { + "epoch": 4.711297071129707, + "grad_norm": 8.158321380615234, + "learning_rate": 9.9898719989872e-06, + "loss": 0.3088, + "step": 10134 + }, + { + "epoch": 4.711761971176197, + "grad_norm": 7.242746829986572, + "learning_rate": 9.9898709989871e-06, + "loss": 0.3845, + "step": 10135 + }, + { + "epoch": 4.712226871222687, + "grad_norm": 5.094803333282471, + "learning_rate": 9.989869998987001e-06, + "loss": 0.3675, + "step": 10136 + }, + { + "epoch": 4.712691771269177, + "grad_norm": 4.991396427154541, + "learning_rate": 9.9898689989869e-06, + "loss": 0.2085, + "step": 10137 + }, + { + "epoch": 4.713156671315668, + "grad_norm": 5.32935905456543, + "learning_rate": 9.989867998986801e-06, + "loss": 0.2344, + "step": 10138 + }, + { + "epoch": 4.713621571362157, + "grad_norm": 9.661227226257324, + "learning_rate": 9.9898669989867e-06, + "loss": 0.3986, + "step": 10139 + }, + { + "epoch": 4.714086471408647, + "grad_norm": 5.522648334503174, + "learning_rate": 9.9898659989866e-06, + "loss": 0.3249, + "step": 10140 + }, + { + "epoch": 4.714551371455137, + "grad_norm": 7.800415992736816, + "learning_rate": 9.9898649989865e-06, + "loss": 0.6231, + "step": 10141 + }, + { + "epoch": 4.715016271501627, + "grad_norm": 4.259583950042725, + "learning_rate": 9.9898639989864e-06, + "loss": 0.2249, + "step": 10142 + }, + { + "epoch": 4.7154811715481175, + "grad_norm": 6.153963565826416, + "learning_rate": 9.989862998986301e-06, + "loss": 0.3067, + "step": 10143 + }, + { + "epoch": 4.715946071594607, + "grad_norm": 4.465085506439209, + "learning_rate": 9.989861998986202e-06, + "loss": 0.3302, + "step": 10144 + }, + { + "epoch": 4.716410971641097, + "grad_norm": 10.300276756286621, + "learning_rate": 9.9898609989861e-06, + "loss": 0.1963, + "step": 10145 + }, + { + "epoch": 4.716875871687587, + "grad_norm": 4.444118022918701, + "learning_rate": 9.989859998986001e-06, + "loss": 0.2502, + "step": 10146 + }, + { + "epoch": 4.717340771734078, + "grad_norm": 3.950871229171753, + "learning_rate": 9.9898589989859e-06, + "loss": 0.1558, + "step": 10147 + }, + { + "epoch": 4.717805671780567, + "grad_norm": 4.537194728851318, + "learning_rate": 9.9898579989858e-06, + "loss": 0.3515, + "step": 10148 + }, + { + "epoch": 4.718270571827057, + "grad_norm": 3.31709885597229, + "learning_rate": 9.989856998985701e-06, + "loss": 0.1695, + "step": 10149 + }, + { + "epoch": 4.718735471873547, + "grad_norm": 6.060608863830566, + "learning_rate": 9.9898559989856e-06, + "loss": 0.4309, + "step": 10150 + }, + { + "epoch": 4.719200371920037, + "grad_norm": 3.9916558265686035, + "learning_rate": 9.989854998985501e-06, + "loss": 0.0988, + "step": 10151 + }, + { + "epoch": 4.7196652719665275, + "grad_norm": 4.954975128173828, + "learning_rate": 9.9898539989854e-06, + "loss": 0.2214, + "step": 10152 + }, + { + "epoch": 4.720130172013017, + "grad_norm": 5.04109001159668, + "learning_rate": 9.9898529989853e-06, + "loss": 0.3559, + "step": 10153 + }, + { + "epoch": 4.720595072059507, + "grad_norm": 8.53763484954834, + "learning_rate": 9.9898519989852e-06, + "loss": 0.9715, + "step": 10154 + }, + { + "epoch": 4.721059972105997, + "grad_norm": 4.456722259521484, + "learning_rate": 9.9898509989851e-06, + "loss": 0.1916, + "step": 10155 + }, + { + "epoch": 4.721524872152488, + "grad_norm": 4.329704761505127, + "learning_rate": 9.989849998985001e-06, + "loss": 0.2766, + "step": 10156 + }, + { + "epoch": 4.721989772198977, + "grad_norm": 3.0235910415649414, + "learning_rate": 9.989848998984902e-06, + "loss": 0.1713, + "step": 10157 + }, + { + "epoch": 4.722454672245467, + "grad_norm": 6.465124130249023, + "learning_rate": 9.9898479989848e-06, + "loss": 0.2319, + "step": 10158 + }, + { + "epoch": 4.722919572291957, + "grad_norm": 6.498291492462158, + "learning_rate": 9.9898469989847e-06, + "loss": 0.3425, + "step": 10159 + }, + { + "epoch": 4.723384472338447, + "grad_norm": 5.362537384033203, + "learning_rate": 9.9898459989846e-06, + "loss": 0.334, + "step": 10160 + }, + { + "epoch": 4.7238493723849375, + "grad_norm": 4.107555389404297, + "learning_rate": 9.989844998984501e-06, + "loss": 0.1979, + "step": 10161 + }, + { + "epoch": 4.724314272431427, + "grad_norm": 5.0656280517578125, + "learning_rate": 9.989843998984402e-06, + "loss": 0.2211, + "step": 10162 + }, + { + "epoch": 4.724779172477917, + "grad_norm": 4.809146404266357, + "learning_rate": 9.9898429989843e-06, + "loss": 0.2314, + "step": 10163 + }, + { + "epoch": 4.725244072524407, + "grad_norm": 7.69077205657959, + "learning_rate": 9.989841998984201e-06, + "loss": 0.7087, + "step": 10164 + }, + { + "epoch": 4.725708972570898, + "grad_norm": 5.61195707321167, + "learning_rate": 9.9898409989841e-06, + "loss": 0.2292, + "step": 10165 + }, + { + "epoch": 4.726173872617387, + "grad_norm": 6.571619510650635, + "learning_rate": 9.989839998984e-06, + "loss": 0.4162, + "step": 10166 + }, + { + "epoch": 4.726638772663877, + "grad_norm": 7.998994827270508, + "learning_rate": 9.989838998983901e-06, + "loss": 0.2428, + "step": 10167 + }, + { + "epoch": 4.727103672710367, + "grad_norm": 7.189952850341797, + "learning_rate": 9.9898379989838e-06, + "loss": 0.3477, + "step": 10168 + }, + { + "epoch": 4.727568572756857, + "grad_norm": 5.074331760406494, + "learning_rate": 9.989836998983701e-06, + "loss": 0.2128, + "step": 10169 + }, + { + "epoch": 4.7280334728033475, + "grad_norm": 3.9942848682403564, + "learning_rate": 9.9898359989836e-06, + "loss": 0.2139, + "step": 10170 + }, + { + "epoch": 4.728498372849837, + "grad_norm": 6.418593406677246, + "learning_rate": 9.9898349989835e-06, + "loss": 0.5933, + "step": 10171 + }, + { + "epoch": 4.728963272896327, + "grad_norm": 3.9356236457824707, + "learning_rate": 9.9898339989834e-06, + "loss": 0.1574, + "step": 10172 + }, + { + "epoch": 4.729428172942817, + "grad_norm": 4.732483863830566, + "learning_rate": 9.9898329989833e-06, + "loss": 0.2809, + "step": 10173 + }, + { + "epoch": 4.729893072989308, + "grad_norm": 3.1233928203582764, + "learning_rate": 9.989831998983201e-06, + "loss": 0.085, + "step": 10174 + }, + { + "epoch": 4.730357973035797, + "grad_norm": 7.240025997161865, + "learning_rate": 9.989830998983102e-06, + "loss": 0.3223, + "step": 10175 + }, + { + "epoch": 4.730822873082287, + "grad_norm": 4.338541507720947, + "learning_rate": 9.989829998983e-06, + "loss": 0.2031, + "step": 10176 + }, + { + "epoch": 4.731287773128777, + "grad_norm": 5.236254692077637, + "learning_rate": 9.989828998982901e-06, + "loss": 0.2239, + "step": 10177 + }, + { + "epoch": 4.731752673175267, + "grad_norm": 4.543619155883789, + "learning_rate": 9.9898279989828e-06, + "loss": 0.1821, + "step": 10178 + }, + { + "epoch": 4.7322175732217575, + "grad_norm": 5.187551021575928, + "learning_rate": 9.989826998982701e-06, + "loss": 0.1584, + "step": 10179 + }, + { + "epoch": 4.732682473268247, + "grad_norm": 5.944474697113037, + "learning_rate": 9.989825998982602e-06, + "loss": 0.6205, + "step": 10180 + }, + { + "epoch": 4.733147373314737, + "grad_norm": 3.4649691581726074, + "learning_rate": 9.9898249989825e-06, + "loss": 0.146, + "step": 10181 + }, + { + "epoch": 4.733612273361228, + "grad_norm": 5.8810296058654785, + "learning_rate": 9.989823998982401e-06, + "loss": 0.1544, + "step": 10182 + }, + { + "epoch": 4.734077173407718, + "grad_norm": 5.450099945068359, + "learning_rate": 9.9898229989823e-06, + "loss": 0.4406, + "step": 10183 + }, + { + "epoch": 4.734542073454207, + "grad_norm": 5.919989109039307, + "learning_rate": 9.9898219989822e-06, + "loss": 0.2765, + "step": 10184 + }, + { + "epoch": 4.735006973500697, + "grad_norm": 4.40645694732666, + "learning_rate": 9.9898209989821e-06, + "loss": 0.2516, + "step": 10185 + }, + { + "epoch": 4.735471873547187, + "grad_norm": 9.803914070129395, + "learning_rate": 9.989819998982e-06, + "loss": 0.3602, + "step": 10186 + }, + { + "epoch": 4.735936773593678, + "grad_norm": 8.645854949951172, + "learning_rate": 9.989818998981901e-06, + "loss": 0.4618, + "step": 10187 + }, + { + "epoch": 4.7364016736401675, + "grad_norm": 7.577649116516113, + "learning_rate": 9.989817998981802e-06, + "loss": 0.5216, + "step": 10188 + }, + { + "epoch": 4.736866573686657, + "grad_norm": 3.7191479206085205, + "learning_rate": 9.9898169989817e-06, + "loss": 0.2261, + "step": 10189 + }, + { + "epoch": 4.737331473733147, + "grad_norm": 5.904477119445801, + "learning_rate": 9.989815998981601e-06, + "loss": 0.3004, + "step": 10190 + }, + { + "epoch": 4.737796373779638, + "grad_norm": 5.375982284545898, + "learning_rate": 9.9898149989815e-06, + "loss": 0.2984, + "step": 10191 + }, + { + "epoch": 4.738261273826128, + "grad_norm": 3.6880555152893066, + "learning_rate": 9.989813998981401e-06, + "loss": 0.1877, + "step": 10192 + }, + { + "epoch": 4.738726173872617, + "grad_norm": 7.873799800872803, + "learning_rate": 9.989812998981302e-06, + "loss": 0.4555, + "step": 10193 + }, + { + "epoch": 4.739191073919107, + "grad_norm": 5.476081371307373, + "learning_rate": 9.9898119989812e-06, + "loss": 0.3888, + "step": 10194 + }, + { + "epoch": 4.739655973965597, + "grad_norm": 4.765103816986084, + "learning_rate": 9.989810998981101e-06, + "loss": 0.217, + "step": 10195 + }, + { + "epoch": 4.740120874012088, + "grad_norm": 5.006093502044678, + "learning_rate": 9.989809998981e-06, + "loss": 0.2175, + "step": 10196 + }, + { + "epoch": 4.7405857740585775, + "grad_norm": 4.624588489532471, + "learning_rate": 9.989808998980901e-06, + "loss": 0.175, + "step": 10197 + }, + { + "epoch": 4.741050674105067, + "grad_norm": 5.682192802429199, + "learning_rate": 9.9898079989808e-06, + "loss": 0.3788, + "step": 10198 + }, + { + "epoch": 4.741515574151557, + "grad_norm": 7.556668758392334, + "learning_rate": 9.9898069989807e-06, + "loss": 0.2713, + "step": 10199 + }, + { + "epoch": 4.741980474198048, + "grad_norm": 6.980066299438477, + "learning_rate": 9.989805998980601e-06, + "loss": 0.2978, + "step": 10200 + }, + { + "epoch": 4.742445374244538, + "grad_norm": 4.991690158843994, + "learning_rate": 9.989804998980502e-06, + "loss": 0.2278, + "step": 10201 + }, + { + "epoch": 4.742910274291027, + "grad_norm": 5.453930854797363, + "learning_rate": 9.9898039989804e-06, + "loss": 0.1731, + "step": 10202 + }, + { + "epoch": 4.743375174337517, + "grad_norm": 4.508658409118652, + "learning_rate": 9.9898029989803e-06, + "loss": 0.217, + "step": 10203 + }, + { + "epoch": 4.743840074384007, + "grad_norm": 3.978783130645752, + "learning_rate": 9.9898019989802e-06, + "loss": 0.1691, + "step": 10204 + }, + { + "epoch": 4.744304974430498, + "grad_norm": 6.293287754058838, + "learning_rate": 9.989800998980101e-06, + "loss": 0.1821, + "step": 10205 + }, + { + "epoch": 4.7447698744769875, + "grad_norm": 7.757944583892822, + "learning_rate": 9.989799998980002e-06, + "loss": 0.2869, + "step": 10206 + }, + { + "epoch": 4.745234774523477, + "grad_norm": 5.471498012542725, + "learning_rate": 9.9897989989799e-06, + "loss": 0.2531, + "step": 10207 + }, + { + "epoch": 4.745699674569967, + "grad_norm": 5.596552848815918, + "learning_rate": 9.989797998979801e-06, + "loss": 0.2836, + "step": 10208 + }, + { + "epoch": 4.746164574616458, + "grad_norm": 4.205079555511475, + "learning_rate": 9.9897969989797e-06, + "loss": 0.2016, + "step": 10209 + }, + { + "epoch": 4.746629474662948, + "grad_norm": 5.242405891418457, + "learning_rate": 9.989795998979601e-06, + "loss": 0.3805, + "step": 10210 + }, + { + "epoch": 4.747094374709437, + "grad_norm": 9.118929862976074, + "learning_rate": 9.9897949989795e-06, + "loss": 0.3183, + "step": 10211 + }, + { + "epoch": 4.747559274755927, + "grad_norm": 9.67721939086914, + "learning_rate": 9.9897939989794e-06, + "loss": 0.3759, + "step": 10212 + }, + { + "epoch": 4.748024174802417, + "grad_norm": 5.13372802734375, + "learning_rate": 9.989792998979301e-06, + "loss": 0.3644, + "step": 10213 + }, + { + "epoch": 4.748489074848908, + "grad_norm": 5.348263263702393, + "learning_rate": 9.9897919989792e-06, + "loss": 0.2902, + "step": 10214 + }, + { + "epoch": 4.7489539748953975, + "grad_norm": 4.146857261657715, + "learning_rate": 9.989790998979101e-06, + "loss": 0.2514, + "step": 10215 + }, + { + "epoch": 4.749418874941887, + "grad_norm": 4.002388954162598, + "learning_rate": 9.989789998979e-06, + "loss": 0.1364, + "step": 10216 + }, + { + "epoch": 4.749883774988378, + "grad_norm": 11.365304946899414, + "learning_rate": 9.9897889989789e-06, + "loss": 0.2661, + "step": 10217 + }, + { + "epoch": 4.750348675034868, + "grad_norm": 6.07337760925293, + "learning_rate": 9.989787998978801e-06, + "loss": 0.387, + "step": 10218 + }, + { + "epoch": 4.750813575081358, + "grad_norm": 5.772143840789795, + "learning_rate": 9.989786998978702e-06, + "loss": 0.3877, + "step": 10219 + }, + { + "epoch": 4.751278475127847, + "grad_norm": 7.664778232574463, + "learning_rate": 9.9897859989786e-06, + "loss": 0.5464, + "step": 10220 + }, + { + "epoch": 4.751743375174337, + "grad_norm": 8.150472640991211, + "learning_rate": 9.989784998978501e-06, + "loss": 0.5107, + "step": 10221 + }, + { + "epoch": 4.752208275220828, + "grad_norm": 4.255146026611328, + "learning_rate": 9.9897839989784e-06, + "loss": 0.2032, + "step": 10222 + }, + { + "epoch": 4.752673175267318, + "grad_norm": 5.053378582000732, + "learning_rate": 9.9897829989783e-06, + "loss": 0.2759, + "step": 10223 + }, + { + "epoch": 4.7531380753138075, + "grad_norm": 4.76231575012207, + "learning_rate": 9.9897819989782e-06, + "loss": 0.2492, + "step": 10224 + }, + { + "epoch": 4.753602975360297, + "grad_norm": 5.496225833892822, + "learning_rate": 9.9897809989781e-06, + "loss": 0.3519, + "step": 10225 + }, + { + "epoch": 4.754067875406788, + "grad_norm": 6.430568695068359, + "learning_rate": 9.989779998978001e-06, + "loss": 0.3129, + "step": 10226 + }, + { + "epoch": 4.754532775453278, + "grad_norm": 5.833431720733643, + "learning_rate": 9.9897789989779e-06, + "loss": 0.3169, + "step": 10227 + }, + { + "epoch": 4.754997675499768, + "grad_norm": 4.429482460021973, + "learning_rate": 9.989777998977801e-06, + "loss": 0.254, + "step": 10228 + }, + { + "epoch": 4.755462575546257, + "grad_norm": 6.142752170562744, + "learning_rate": 9.9897769989777e-06, + "loss": 0.3442, + "step": 10229 + }, + { + "epoch": 4.755927475592747, + "grad_norm": 4.759104251861572, + "learning_rate": 9.9897759989776e-06, + "loss": 0.1962, + "step": 10230 + }, + { + "epoch": 4.756392375639238, + "grad_norm": 6.277011394500732, + "learning_rate": 9.989774998977501e-06, + "loss": 0.3638, + "step": 10231 + }, + { + "epoch": 4.756857275685728, + "grad_norm": 8.701483726501465, + "learning_rate": 9.989773998977402e-06, + "loss": 0.3481, + "step": 10232 + }, + { + "epoch": 4.7573221757322175, + "grad_norm": 7.493145942687988, + "learning_rate": 9.9897729989773e-06, + "loss": 0.5616, + "step": 10233 + }, + { + "epoch": 4.757787075778707, + "grad_norm": 7.621062278747559, + "learning_rate": 9.9897719989772e-06, + "loss": 0.3667, + "step": 10234 + }, + { + "epoch": 4.758251975825198, + "grad_norm": 7.028818130493164, + "learning_rate": 9.9897709989771e-06, + "loss": 0.5584, + "step": 10235 + }, + { + "epoch": 4.758716875871688, + "grad_norm": 6.865070819854736, + "learning_rate": 9.989769998977001e-06, + "loss": 0.2823, + "step": 10236 + }, + { + "epoch": 4.759181775918178, + "grad_norm": 4.778927326202393, + "learning_rate": 9.9897689989769e-06, + "loss": 0.248, + "step": 10237 + }, + { + "epoch": 4.759646675964667, + "grad_norm": 4.916318416595459, + "learning_rate": 9.9897679989768e-06, + "loss": 0.3009, + "step": 10238 + }, + { + "epoch": 4.760111576011157, + "grad_norm": 6.273286819458008, + "learning_rate": 9.989766998976701e-06, + "loss": 0.4905, + "step": 10239 + }, + { + "epoch": 4.760576476057648, + "grad_norm": 2.6960220336914062, + "learning_rate": 9.9897659989766e-06, + "loss": 0.0737, + "step": 10240 + }, + { + "epoch": 4.761041376104138, + "grad_norm": 7.060067176818848, + "learning_rate": 9.989764998976501e-06, + "loss": 0.3495, + "step": 10241 + }, + { + "epoch": 4.7615062761506275, + "grad_norm": 5.734601974487305, + "learning_rate": 9.9897639989764e-06, + "loss": 0.6211, + "step": 10242 + }, + { + "epoch": 4.761971176197117, + "grad_norm": 3.6581573486328125, + "learning_rate": 9.9897629989763e-06, + "loss": 0.1962, + "step": 10243 + }, + { + "epoch": 4.762436076243608, + "grad_norm": 4.976374626159668, + "learning_rate": 9.989761998976201e-06, + "loss": 0.1927, + "step": 10244 + }, + { + "epoch": 4.762900976290098, + "grad_norm": 5.234469413757324, + "learning_rate": 9.9897609989761e-06, + "loss": 0.2598, + "step": 10245 + }, + { + "epoch": 4.763365876336588, + "grad_norm": 15.412802696228027, + "learning_rate": 9.989759998976001e-06, + "loss": 1.6497, + "step": 10246 + }, + { + "epoch": 4.763830776383077, + "grad_norm": 7.260754585266113, + "learning_rate": 9.9897589989759e-06, + "loss": 0.4888, + "step": 10247 + }, + { + "epoch": 4.764295676429567, + "grad_norm": 5.560030460357666, + "learning_rate": 9.9897579989758e-06, + "loss": 0.2733, + "step": 10248 + }, + { + "epoch": 4.764760576476058, + "grad_norm": 7.703429698944092, + "learning_rate": 9.989756998975701e-06, + "loss": 0.4217, + "step": 10249 + }, + { + "epoch": 4.765225476522548, + "grad_norm": 3.863698720932007, + "learning_rate": 9.9897559989756e-06, + "loss": 0.1907, + "step": 10250 + }, + { + "epoch": 4.7656903765690375, + "grad_norm": 4.568164348602295, + "learning_rate": 9.9897549989755e-06, + "loss": 0.2558, + "step": 10251 + }, + { + "epoch": 4.766155276615527, + "grad_norm": 7.286221504211426, + "learning_rate": 9.989753998975401e-06, + "loss": 0.5139, + "step": 10252 + }, + { + "epoch": 4.766620176662018, + "grad_norm": 2.8576178550720215, + "learning_rate": 9.9897529989753e-06, + "loss": 0.0757, + "step": 10253 + }, + { + "epoch": 4.767085076708508, + "grad_norm": 4.252397060394287, + "learning_rate": 9.989751998975201e-06, + "loss": 0.1204, + "step": 10254 + }, + { + "epoch": 4.767549976754998, + "grad_norm": 6.672576427459717, + "learning_rate": 9.9897509989751e-06, + "loss": 0.3353, + "step": 10255 + }, + { + "epoch": 4.768014876801487, + "grad_norm": 5.50213623046875, + "learning_rate": 9.989749998975e-06, + "loss": 0.4124, + "step": 10256 + }, + { + "epoch": 4.768479776847977, + "grad_norm": 5.388737201690674, + "learning_rate": 9.989748998974901e-06, + "loss": 0.1211, + "step": 10257 + }, + { + "epoch": 4.768944676894468, + "grad_norm": 4.462337017059326, + "learning_rate": 9.9897479989748e-06, + "loss": 0.3193, + "step": 10258 + }, + { + "epoch": 4.769409576940958, + "grad_norm": 7.364530086517334, + "learning_rate": 9.989746998974701e-06, + "loss": 0.3373, + "step": 10259 + }, + { + "epoch": 4.7698744769874475, + "grad_norm": 3.857630491256714, + "learning_rate": 9.9897459989746e-06, + "loss": 0.2245, + "step": 10260 + }, + { + "epoch": 4.770339377033938, + "grad_norm": 4.176729202270508, + "learning_rate": 9.9897449989745e-06, + "loss": 0.2277, + "step": 10261 + }, + { + "epoch": 4.770804277080428, + "grad_norm": 3.9040634632110596, + "learning_rate": 9.989743998974401e-06, + "loss": 0.2263, + "step": 10262 + }, + { + "epoch": 4.771269177126918, + "grad_norm": 5.11561918258667, + "learning_rate": 9.989742998974302e-06, + "loss": 0.1583, + "step": 10263 + }, + { + "epoch": 4.771734077173408, + "grad_norm": 4.485804557800293, + "learning_rate": 9.989741998974201e-06, + "loss": 0.1977, + "step": 10264 + }, + { + "epoch": 4.772198977219897, + "grad_norm": 5.4461164474487305, + "learning_rate": 9.989740998974102e-06, + "loss": 0.2785, + "step": 10265 + }, + { + "epoch": 4.772663877266388, + "grad_norm": 4.975518226623535, + "learning_rate": 9.989739998974e-06, + "loss": 0.3145, + "step": 10266 + }, + { + "epoch": 4.773128777312878, + "grad_norm": 5.959756851196289, + "learning_rate": 9.9897389989739e-06, + "loss": 0.3346, + "step": 10267 + }, + { + "epoch": 4.773593677359368, + "grad_norm": 4.205581188201904, + "learning_rate": 9.9897379989738e-06, + "loss": 0.2507, + "step": 10268 + }, + { + "epoch": 4.7740585774058575, + "grad_norm": 3.0003511905670166, + "learning_rate": 9.9897369989737e-06, + "loss": 0.1045, + "step": 10269 + }, + { + "epoch": 4.774523477452348, + "grad_norm": 5.892990589141846, + "learning_rate": 9.989735998973601e-06, + "loss": 0.3475, + "step": 10270 + }, + { + "epoch": 4.774988377498838, + "grad_norm": 4.639866828918457, + "learning_rate": 9.9897349989735e-06, + "loss": 0.2062, + "step": 10271 + }, + { + "epoch": 4.775453277545328, + "grad_norm": 4.558016300201416, + "learning_rate": 9.989733998973401e-06, + "loss": 0.1964, + "step": 10272 + }, + { + "epoch": 4.775918177591818, + "grad_norm": 5.274717807769775, + "learning_rate": 9.9897329989733e-06, + "loss": 0.3097, + "step": 10273 + }, + { + "epoch": 4.776383077638307, + "grad_norm": 3.4059178829193115, + "learning_rate": 9.9897319989732e-06, + "loss": 0.1772, + "step": 10274 + }, + { + "epoch": 4.776847977684798, + "grad_norm": 4.261287689208984, + "learning_rate": 9.989730998973101e-06, + "loss": 0.2668, + "step": 10275 + }, + { + "epoch": 4.777312877731288, + "grad_norm": 7.283013343811035, + "learning_rate": 9.989729998973002e-06, + "loss": 0.4666, + "step": 10276 + }, + { + "epoch": 4.777777777777778, + "grad_norm": 5.198851108551025, + "learning_rate": 9.989728998972901e-06, + "loss": 0.2465, + "step": 10277 + }, + { + "epoch": 4.7782426778242675, + "grad_norm": 4.630976676940918, + "learning_rate": 9.9897279989728e-06, + "loss": 0.2259, + "step": 10278 + }, + { + "epoch": 4.778707577870758, + "grad_norm": 5.366891860961914, + "learning_rate": 9.9897269989727e-06, + "loss": 0.2502, + "step": 10279 + }, + { + "epoch": 4.779172477917248, + "grad_norm": 4.3364644050598145, + "learning_rate": 9.9897259989726e-06, + "loss": 0.244, + "step": 10280 + }, + { + "epoch": 4.779637377963738, + "grad_norm": 4.646618843078613, + "learning_rate": 9.9897249989725e-06, + "loss": 0.1955, + "step": 10281 + }, + { + "epoch": 4.780102278010228, + "grad_norm": 5.4322404861450195, + "learning_rate": 9.989723998972401e-06, + "loss": 0.1216, + "step": 10282 + }, + { + "epoch": 4.780567178056717, + "grad_norm": 4.717141628265381, + "learning_rate": 9.989722998972302e-06, + "loss": 0.2493, + "step": 10283 + }, + { + "epoch": 4.781032078103208, + "grad_norm": 5.848628044128418, + "learning_rate": 9.9897219989722e-06, + "loss": 0.3736, + "step": 10284 + }, + { + "epoch": 4.781496978149698, + "grad_norm": 8.765608787536621, + "learning_rate": 9.989720998972101e-06, + "loss": 0.2161, + "step": 10285 + }, + { + "epoch": 4.781961878196188, + "grad_norm": 6.706182479858398, + "learning_rate": 9.989719998972e-06, + "loss": 0.4024, + "step": 10286 + }, + { + "epoch": 4.7824267782426775, + "grad_norm": 5.905442714691162, + "learning_rate": 9.9897189989719e-06, + "loss": 0.2885, + "step": 10287 + }, + { + "epoch": 4.782891678289168, + "grad_norm": 7.066157817840576, + "learning_rate": 9.989717998971801e-06, + "loss": 0.5417, + "step": 10288 + }, + { + "epoch": 4.783356578335658, + "grad_norm": 3.519846200942993, + "learning_rate": 9.9897169989717e-06, + "loss": 0.1581, + "step": 10289 + }, + { + "epoch": 4.783821478382148, + "grad_norm": 6.622713565826416, + "learning_rate": 9.989715998971601e-06, + "loss": 0.5107, + "step": 10290 + }, + { + "epoch": 4.784286378428638, + "grad_norm": 12.627467155456543, + "learning_rate": 9.9897149989715e-06, + "loss": 0.2851, + "step": 10291 + }, + { + "epoch": 4.784751278475127, + "grad_norm": 3.584202527999878, + "learning_rate": 9.9897139989714e-06, + "loss": 0.0994, + "step": 10292 + }, + { + "epoch": 4.785216178521618, + "grad_norm": 4.709413528442383, + "learning_rate": 9.9897129989713e-06, + "loss": 0.2497, + "step": 10293 + }, + { + "epoch": 4.785681078568108, + "grad_norm": 7.007564067840576, + "learning_rate": 9.9897119989712e-06, + "loss": 0.1222, + "step": 10294 + }, + { + "epoch": 4.786145978614598, + "grad_norm": 6.508622646331787, + "learning_rate": 9.989710998971101e-06, + "loss": 0.5258, + "step": 10295 + }, + { + "epoch": 4.786610878661088, + "grad_norm": 4.181931018829346, + "learning_rate": 9.989709998971002e-06, + "loss": 0.2146, + "step": 10296 + }, + { + "epoch": 4.787075778707578, + "grad_norm": 4.210280895233154, + "learning_rate": 9.9897089989709e-06, + "loss": 0.1828, + "step": 10297 + }, + { + "epoch": 4.787540678754068, + "grad_norm": 4.102425575256348, + "learning_rate": 9.989707998970801e-06, + "loss": 0.1319, + "step": 10298 + }, + { + "epoch": 4.788005578800558, + "grad_norm": 5.090095043182373, + "learning_rate": 9.9897069989707e-06, + "loss": 0.1617, + "step": 10299 + }, + { + "epoch": 4.788470478847048, + "grad_norm": 6.479487419128418, + "learning_rate": 9.989705998970601e-06, + "loss": 0.1191, + "step": 10300 + }, + { + "epoch": 4.788935378893538, + "grad_norm": 5.136975288391113, + "learning_rate": 9.989704998970502e-06, + "loss": 0.2348, + "step": 10301 + }, + { + "epoch": 4.789400278940028, + "grad_norm": 6.150022029876709, + "learning_rate": 9.9897039989704e-06, + "loss": 0.3916, + "step": 10302 + }, + { + "epoch": 4.789865178986518, + "grad_norm": 8.694212913513184, + "learning_rate": 9.989702998970301e-06, + "loss": 0.2666, + "step": 10303 + }, + { + "epoch": 4.790330079033008, + "grad_norm": 16.472753524780273, + "learning_rate": 9.9897019989702e-06, + "loss": 0.3136, + "step": 10304 + }, + { + "epoch": 4.790794979079498, + "grad_norm": 2.9992527961730957, + "learning_rate": 9.9897009989701e-06, + "loss": 0.0715, + "step": 10305 + }, + { + "epoch": 4.791259879125988, + "grad_norm": 5.0886993408203125, + "learning_rate": 9.98969999897e-06, + "loss": 0.2484, + "step": 10306 + }, + { + "epoch": 4.791724779172478, + "grad_norm": 5.163442134857178, + "learning_rate": 9.9896989989699e-06, + "loss": 0.2673, + "step": 10307 + }, + { + "epoch": 4.792189679218968, + "grad_norm": 7.8088154792785645, + "learning_rate": 9.989697998969801e-06, + "loss": 0.3513, + "step": 10308 + }, + { + "epoch": 4.792654579265458, + "grad_norm": 3.1403305530548096, + "learning_rate": 9.989696998969702e-06, + "loss": 0.1, + "step": 10309 + }, + { + "epoch": 4.793119479311948, + "grad_norm": 5.88418436050415, + "learning_rate": 9.9896959989696e-06, + "loss": 0.4549, + "step": 10310 + }, + { + "epoch": 4.793584379358438, + "grad_norm": 13.390607833862305, + "learning_rate": 9.9896949989695e-06, + "loss": 0.3831, + "step": 10311 + }, + { + "epoch": 4.794049279404928, + "grad_norm": 5.436734676361084, + "learning_rate": 9.9896939989694e-06, + "loss": 0.2278, + "step": 10312 + }, + { + "epoch": 4.794514179451418, + "grad_norm": 2.7320332527160645, + "learning_rate": 9.989692998969301e-06, + "loss": 0.0885, + "step": 10313 + }, + { + "epoch": 4.794979079497908, + "grad_norm": 4.687387943267822, + "learning_rate": 9.989691998969202e-06, + "loss": 0.2641, + "step": 10314 + }, + { + "epoch": 4.795443979544398, + "grad_norm": 5.924389362335205, + "learning_rate": 9.9896909989691e-06, + "loss": 0.2618, + "step": 10315 + }, + { + "epoch": 4.795908879590888, + "grad_norm": 4.582963943481445, + "learning_rate": 9.989689998969001e-06, + "loss": 0.1791, + "step": 10316 + }, + { + "epoch": 4.796373779637378, + "grad_norm": 5.415489673614502, + "learning_rate": 9.9896889989689e-06, + "loss": 0.2504, + "step": 10317 + }, + { + "epoch": 4.796838679683868, + "grad_norm": 7.285367488861084, + "learning_rate": 9.9896879989688e-06, + "loss": 0.1214, + "step": 10318 + }, + { + "epoch": 4.797303579730358, + "grad_norm": 6.1690568923950195, + "learning_rate": 9.9896869989687e-06, + "loss": 0.2471, + "step": 10319 + }, + { + "epoch": 4.797768479776848, + "grad_norm": 6.996684551239014, + "learning_rate": 9.9896859989686e-06, + "loss": 0.4278, + "step": 10320 + }, + { + "epoch": 4.798233379823338, + "grad_norm": 4.54746150970459, + "learning_rate": 9.989684998968501e-06, + "loss": 0.3336, + "step": 10321 + }, + { + "epoch": 4.798698279869828, + "grad_norm": 7.378709316253662, + "learning_rate": 9.9896839989684e-06, + "loss": 0.3127, + "step": 10322 + }, + { + "epoch": 4.799163179916318, + "grad_norm": 5.875476837158203, + "learning_rate": 9.9896829989683e-06, + "loss": 0.2714, + "step": 10323 + }, + { + "epoch": 4.799628079962808, + "grad_norm": 5.235526084899902, + "learning_rate": 9.9896819989682e-06, + "loss": 0.3548, + "step": 10324 + }, + { + "epoch": 4.800092980009298, + "grad_norm": 3.8220102787017822, + "learning_rate": 9.9896809989681e-06, + "loss": 0.1563, + "step": 10325 + }, + { + "epoch": 4.800557880055788, + "grad_norm": 6.722143173217773, + "learning_rate": 9.989679998968001e-06, + "loss": 0.4341, + "step": 10326 + }, + { + "epoch": 4.801022780102278, + "grad_norm": 4.673069477081299, + "learning_rate": 9.989678998967902e-06, + "loss": 0.169, + "step": 10327 + }, + { + "epoch": 4.801487680148768, + "grad_norm": 4.552925109863281, + "learning_rate": 9.9896779989678e-06, + "loss": 0.2101, + "step": 10328 + }, + { + "epoch": 4.801952580195258, + "grad_norm": 6.654307842254639, + "learning_rate": 9.989676998967701e-06, + "loss": 0.6036, + "step": 10329 + }, + { + "epoch": 4.802417480241748, + "grad_norm": 4.486143589019775, + "learning_rate": 9.9896759989676e-06, + "loss": 0.158, + "step": 10330 + }, + { + "epoch": 4.802882380288238, + "grad_norm": 3.761824131011963, + "learning_rate": 9.989674998967501e-06, + "loss": 0.2075, + "step": 10331 + }, + { + "epoch": 4.803347280334728, + "grad_norm": 4.90961217880249, + "learning_rate": 9.989673998967402e-06, + "loss": 0.3002, + "step": 10332 + }, + { + "epoch": 4.803812180381218, + "grad_norm": 4.317301273345947, + "learning_rate": 9.9896729989673e-06, + "loss": 0.1254, + "step": 10333 + }, + { + "epoch": 4.804277080427708, + "grad_norm": 6.123403072357178, + "learning_rate": 9.989671998967201e-06, + "loss": 0.2535, + "step": 10334 + }, + { + "epoch": 4.804741980474198, + "grad_norm": 7.122442722320557, + "learning_rate": 9.9896709989671e-06, + "loss": 0.2068, + "step": 10335 + }, + { + "epoch": 4.805206880520688, + "grad_norm": 6.296876907348633, + "learning_rate": 9.989669998967e-06, + "loss": 0.2407, + "step": 10336 + }, + { + "epoch": 4.805671780567178, + "grad_norm": 5.082381248474121, + "learning_rate": 9.9896689989669e-06, + "loss": 0.1952, + "step": 10337 + }, + { + "epoch": 4.806136680613668, + "grad_norm": 9.021323204040527, + "learning_rate": 9.9896679989668e-06, + "loss": 0.413, + "step": 10338 + }, + { + "epoch": 4.806601580660158, + "grad_norm": 8.785185813903809, + "learning_rate": 9.989666998966701e-06, + "loss": 0.7141, + "step": 10339 + }, + { + "epoch": 4.8070664807066485, + "grad_norm": 7.088465213775635, + "learning_rate": 9.989665998966602e-06, + "loss": 0.3804, + "step": 10340 + }, + { + "epoch": 4.807531380753138, + "grad_norm": 5.660850524902344, + "learning_rate": 9.9896649989665e-06, + "loss": 0.4429, + "step": 10341 + }, + { + "epoch": 4.807996280799628, + "grad_norm": 5.650301456451416, + "learning_rate": 9.989663998966401e-06, + "loss": 0.1883, + "step": 10342 + }, + { + "epoch": 4.808461180846118, + "grad_norm": 6.5172882080078125, + "learning_rate": 9.9896629989663e-06, + "loss": 0.2649, + "step": 10343 + }, + { + "epoch": 4.808926080892608, + "grad_norm": 4.557281970977783, + "learning_rate": 9.989661998966201e-06, + "loss": 0.2875, + "step": 10344 + }, + { + "epoch": 4.8093909809390984, + "grad_norm": 6.8862996101379395, + "learning_rate": 9.989660998966102e-06, + "loss": 0.4255, + "step": 10345 + }, + { + "epoch": 4.809855880985588, + "grad_norm": 6.785994529724121, + "learning_rate": 9.989659998966e-06, + "loss": 0.4141, + "step": 10346 + }, + { + "epoch": 4.810320781032078, + "grad_norm": 5.02195930480957, + "learning_rate": 9.989658998965901e-06, + "loss": 0.1436, + "step": 10347 + }, + { + "epoch": 4.810785681078568, + "grad_norm": 4.84826135635376, + "learning_rate": 9.9896579989658e-06, + "loss": 0.1999, + "step": 10348 + }, + { + "epoch": 4.8112505811250585, + "grad_norm": 3.8807756900787354, + "learning_rate": 9.989656998965701e-06, + "loss": 0.1348, + "step": 10349 + }, + { + "epoch": 4.811715481171548, + "grad_norm": 4.333731651306152, + "learning_rate": 9.9896559989656e-06, + "loss": 0.2316, + "step": 10350 + }, + { + "epoch": 4.812180381218038, + "grad_norm": 4.149636268615723, + "learning_rate": 9.9896549989655e-06, + "loss": 0.1563, + "step": 10351 + }, + { + "epoch": 4.812645281264528, + "grad_norm": 7.400701999664307, + "learning_rate": 9.989653998965401e-06, + "loss": 0.4179, + "step": 10352 + }, + { + "epoch": 4.813110181311018, + "grad_norm": 3.3049263954162598, + "learning_rate": 9.9896529989653e-06, + "loss": 0.1129, + "step": 10353 + }, + { + "epoch": 4.8135750813575084, + "grad_norm": 4.629458427429199, + "learning_rate": 9.9896519989652e-06, + "loss": 0.1839, + "step": 10354 + }, + { + "epoch": 4.814039981403998, + "grad_norm": 7.513705730438232, + "learning_rate": 9.9896509989651e-06, + "loss": 0.4414, + "step": 10355 + }, + { + "epoch": 4.814504881450488, + "grad_norm": 7.3541579246521, + "learning_rate": 9.989649998965e-06, + "loss": 0.4039, + "step": 10356 + }, + { + "epoch": 4.814969781496978, + "grad_norm": 5.876742362976074, + "learning_rate": 9.989648998964901e-06, + "loss": 0.4677, + "step": 10357 + }, + { + "epoch": 4.8154346815434685, + "grad_norm": 5.0651936531066895, + "learning_rate": 9.989647998964802e-06, + "loss": 0.2343, + "step": 10358 + }, + { + "epoch": 4.815899581589958, + "grad_norm": 11.274005889892578, + "learning_rate": 9.9896469989647e-06, + "loss": 0.6476, + "step": 10359 + }, + { + "epoch": 4.816364481636448, + "grad_norm": 6.015191078186035, + "learning_rate": 9.989645998964601e-06, + "loss": 0.3423, + "step": 10360 + }, + { + "epoch": 4.816829381682938, + "grad_norm": 5.644052028656006, + "learning_rate": 9.9896449989645e-06, + "loss": 0.3013, + "step": 10361 + }, + { + "epoch": 4.817294281729428, + "grad_norm": 4.674873352050781, + "learning_rate": 9.989643998964401e-06, + "loss": 0.2079, + "step": 10362 + }, + { + "epoch": 4.817759181775918, + "grad_norm": 4.883729457855225, + "learning_rate": 9.9896429989643e-06, + "loss": 0.2487, + "step": 10363 + }, + { + "epoch": 4.818224081822408, + "grad_norm": 8.84382152557373, + "learning_rate": 9.9896419989642e-06, + "loss": 0.5618, + "step": 10364 + }, + { + "epoch": 4.818688981868898, + "grad_norm": 6.2472991943359375, + "learning_rate": 9.989640998964101e-06, + "loss": 0.4818, + "step": 10365 + }, + { + "epoch": 4.819153881915388, + "grad_norm": 9.98538589477539, + "learning_rate": 9.989639998964e-06, + "loss": 0.3849, + "step": 10366 + }, + { + "epoch": 4.8196187819618785, + "grad_norm": 4.217313766479492, + "learning_rate": 9.989638998963901e-06, + "loss": 0.2189, + "step": 10367 + }, + { + "epoch": 4.820083682008368, + "grad_norm": 8.117847442626953, + "learning_rate": 9.9896379989638e-06, + "loss": 0.3298, + "step": 10368 + }, + { + "epoch": 4.820548582054858, + "grad_norm": 5.974038124084473, + "learning_rate": 9.9896369989637e-06, + "loss": 0.1195, + "step": 10369 + }, + { + "epoch": 4.821013482101348, + "grad_norm": 8.68582820892334, + "learning_rate": 9.989635998963601e-06, + "loss": 0.5546, + "step": 10370 + }, + { + "epoch": 4.821478382147838, + "grad_norm": 6.432646751403809, + "learning_rate": 9.989634998963502e-06, + "loss": 0.456, + "step": 10371 + }, + { + "epoch": 4.821943282194328, + "grad_norm": 7.155545711517334, + "learning_rate": 9.9896339989634e-06, + "loss": 0.4373, + "step": 10372 + }, + { + "epoch": 4.822408182240818, + "grad_norm": 5.637589454650879, + "learning_rate": 9.989632998963301e-06, + "loss": 0.3062, + "step": 10373 + }, + { + "epoch": 4.822873082287308, + "grad_norm": 6.00918436050415, + "learning_rate": 9.9896319989632e-06, + "loss": 0.4951, + "step": 10374 + }, + { + "epoch": 4.823337982333798, + "grad_norm": 5.071533203125, + "learning_rate": 9.9896309989631e-06, + "loss": 0.2822, + "step": 10375 + }, + { + "epoch": 4.8238028823802885, + "grad_norm": 4.952261924743652, + "learning_rate": 9.989629998963e-06, + "loss": 0.3434, + "step": 10376 + }, + { + "epoch": 4.824267782426778, + "grad_norm": 7.157773017883301, + "learning_rate": 9.9896289989629e-06, + "loss": 0.4416, + "step": 10377 + }, + { + "epoch": 4.824732682473268, + "grad_norm": 13.420231819152832, + "learning_rate": 9.989627998962801e-06, + "loss": 0.2617, + "step": 10378 + }, + { + "epoch": 4.825197582519758, + "grad_norm": 5.640274524688721, + "learning_rate": 9.9896269989627e-06, + "loss": 0.4495, + "step": 10379 + }, + { + "epoch": 4.825662482566249, + "grad_norm": 5.598100662231445, + "learning_rate": 9.989625998962601e-06, + "loss": 0.3232, + "step": 10380 + }, + { + "epoch": 4.826127382612738, + "grad_norm": 2.6856799125671387, + "learning_rate": 9.9896249989625e-06, + "loss": 0.0852, + "step": 10381 + }, + { + "epoch": 4.826592282659228, + "grad_norm": 5.115905284881592, + "learning_rate": 9.9896239989624e-06, + "loss": 0.3651, + "step": 10382 + }, + { + "epoch": 4.827057182705718, + "grad_norm": 4.247076511383057, + "learning_rate": 9.989622998962301e-06, + "loss": 0.2508, + "step": 10383 + }, + { + "epoch": 4.827522082752209, + "grad_norm": 4.1382341384887695, + "learning_rate": 9.989621998962202e-06, + "loss": 0.2019, + "step": 10384 + }, + { + "epoch": 4.8279869827986985, + "grad_norm": 4.585038661956787, + "learning_rate": 9.989620998962101e-06, + "loss": 0.1874, + "step": 10385 + }, + { + "epoch": 4.828451882845188, + "grad_norm": 6.073827266693115, + "learning_rate": 9.989619998962e-06, + "loss": 0.2873, + "step": 10386 + }, + { + "epoch": 4.828916782891678, + "grad_norm": 10.057088851928711, + "learning_rate": 9.9896189989619e-06, + "loss": 0.3648, + "step": 10387 + }, + { + "epoch": 4.829381682938168, + "grad_norm": 5.116932392120361, + "learning_rate": 9.9896179989618e-06, + "loss": 0.1649, + "step": 10388 + }, + { + "epoch": 4.829846582984659, + "grad_norm": 6.067676544189453, + "learning_rate": 9.9896169989617e-06, + "loss": 0.4372, + "step": 10389 + }, + { + "epoch": 4.830311483031148, + "grad_norm": 4.6457200050354, + "learning_rate": 9.9896159989616e-06, + "loss": 0.2218, + "step": 10390 + }, + { + "epoch": 4.830776383077638, + "grad_norm": 5.042115211486816, + "learning_rate": 9.989614998961501e-06, + "loss": 0.3692, + "step": 10391 + }, + { + "epoch": 4.831241283124128, + "grad_norm": 6.73656702041626, + "learning_rate": 9.9896139989614e-06, + "loss": 0.3674, + "step": 10392 + }, + { + "epoch": 4.831706183170619, + "grad_norm": 5.849018573760986, + "learning_rate": 9.989612998961301e-06, + "loss": 0.3865, + "step": 10393 + }, + { + "epoch": 4.8321710832171085, + "grad_norm": 4.100821495056152, + "learning_rate": 9.9896119989612e-06, + "loss": 0.137, + "step": 10394 + }, + { + "epoch": 4.832635983263598, + "grad_norm": 4.18739652633667, + "learning_rate": 9.9896109989611e-06, + "loss": 0.2708, + "step": 10395 + }, + { + "epoch": 4.833100883310088, + "grad_norm": 4.120202541351318, + "learning_rate": 9.989609998961001e-06, + "loss": 0.175, + "step": 10396 + }, + { + "epoch": 4.833565783356578, + "grad_norm": 5.073047637939453, + "learning_rate": 9.9896089989609e-06, + "loss": 0.2131, + "step": 10397 + }, + { + "epoch": 4.834030683403069, + "grad_norm": 4.004355430603027, + "learning_rate": 9.989607998960801e-06, + "loss": 0.1386, + "step": 10398 + }, + { + "epoch": 4.834495583449558, + "grad_norm": 6.268239498138428, + "learning_rate": 9.9896069989607e-06, + "loss": 0.397, + "step": 10399 + }, + { + "epoch": 4.834960483496048, + "grad_norm": 5.871522426605225, + "learning_rate": 9.9896059989606e-06, + "loss": 0.2086, + "step": 10400 + }, + { + "epoch": 4.835425383542538, + "grad_norm": 6.964033126831055, + "learning_rate": 9.989604998960501e-06, + "loss": 0.3621, + "step": 10401 + }, + { + "epoch": 4.835890283589029, + "grad_norm": 5.822314262390137, + "learning_rate": 9.9896039989604e-06, + "loss": 0.3135, + "step": 10402 + }, + { + "epoch": 4.8363551836355185, + "grad_norm": 6.144704341888428, + "learning_rate": 9.9896029989603e-06, + "loss": 0.346, + "step": 10403 + }, + { + "epoch": 4.836820083682008, + "grad_norm": 4.050589561462402, + "learning_rate": 9.989601998960202e-06, + "loss": 0.1773, + "step": 10404 + }, + { + "epoch": 4.837284983728498, + "grad_norm": 2.918236255645752, + "learning_rate": 9.9896009989601e-06, + "loss": 0.1026, + "step": 10405 + }, + { + "epoch": 4.837749883774988, + "grad_norm": 10.859235763549805, + "learning_rate": 9.989599998960001e-06, + "loss": 0.7172, + "step": 10406 + }, + { + "epoch": 4.838214783821479, + "grad_norm": 5.077722549438477, + "learning_rate": 9.9895989989599e-06, + "loss": 0.3426, + "step": 10407 + }, + { + "epoch": 4.838679683867968, + "grad_norm": 5.4993462562561035, + "learning_rate": 9.9895979989598e-06, + "loss": 0.3737, + "step": 10408 + }, + { + "epoch": 4.839144583914458, + "grad_norm": 7.522674083709717, + "learning_rate": 9.989596998959701e-06, + "loss": 0.5472, + "step": 10409 + }, + { + "epoch": 4.839609483960948, + "grad_norm": 5.721911430358887, + "learning_rate": 9.9895959989596e-06, + "loss": 0.334, + "step": 10410 + }, + { + "epoch": 4.840074384007439, + "grad_norm": 5.135854244232178, + "learning_rate": 9.989594998959501e-06, + "loss": 0.2174, + "step": 10411 + }, + { + "epoch": 4.8405392840539285, + "grad_norm": 5.13712215423584, + "learning_rate": 9.9895939989594e-06, + "loss": 0.2708, + "step": 10412 + }, + { + "epoch": 4.841004184100418, + "grad_norm": 5.638555526733398, + "learning_rate": 9.9895929989593e-06, + "loss": 0.414, + "step": 10413 + }, + { + "epoch": 4.841469084146908, + "grad_norm": 7.197741508483887, + "learning_rate": 9.989591998959201e-06, + "loss": 0.5902, + "step": 10414 + }, + { + "epoch": 4.841933984193398, + "grad_norm": 3.9451560974121094, + "learning_rate": 9.9895909989591e-06, + "loss": 0.2102, + "step": 10415 + }, + { + "epoch": 4.842398884239889, + "grad_norm": 8.112188339233398, + "learning_rate": 9.989589998959001e-06, + "loss": 0.5591, + "step": 10416 + }, + { + "epoch": 4.842863784286378, + "grad_norm": 8.389025688171387, + "learning_rate": 9.989588998958902e-06, + "loss": 0.5929, + "step": 10417 + }, + { + "epoch": 4.843328684332868, + "grad_norm": 5.389368057250977, + "learning_rate": 9.9895879989588e-06, + "loss": 0.3288, + "step": 10418 + }, + { + "epoch": 4.843793584379359, + "grad_norm": 10.47960090637207, + "learning_rate": 9.9895869989587e-06, + "loss": 0.6848, + "step": 10419 + }, + { + "epoch": 4.844258484425849, + "grad_norm": 4.86142110824585, + "learning_rate": 9.9895859989586e-06, + "loss": 0.2761, + "step": 10420 + }, + { + "epoch": 4.8447233844723385, + "grad_norm": 6.414243221282959, + "learning_rate": 9.9895849989585e-06, + "loss": 0.2818, + "step": 10421 + }, + { + "epoch": 4.845188284518828, + "grad_norm": 8.567687034606934, + "learning_rate": 9.989583998958401e-06, + "loss": 0.8804, + "step": 10422 + }, + { + "epoch": 4.845653184565318, + "grad_norm": 4.861550807952881, + "learning_rate": 9.9895829989583e-06, + "loss": 0.3415, + "step": 10423 + }, + { + "epoch": 4.846118084611809, + "grad_norm": 4.091798782348633, + "learning_rate": 9.989581998958201e-06, + "loss": 0.3084, + "step": 10424 + }, + { + "epoch": 4.846582984658299, + "grad_norm": 9.728804588317871, + "learning_rate": 9.9895809989581e-06, + "loss": 0.3779, + "step": 10425 + }, + { + "epoch": 4.847047884704788, + "grad_norm": 3.2440335750579834, + "learning_rate": 9.989579998958e-06, + "loss": 0.0986, + "step": 10426 + }, + { + "epoch": 4.847512784751278, + "grad_norm": 5.1234588623046875, + "learning_rate": 9.989578998957901e-06, + "loss": 0.2963, + "step": 10427 + }, + { + "epoch": 4.847977684797769, + "grad_norm": 5.52407693862915, + "learning_rate": 9.989577998957802e-06, + "loss": 0.2616, + "step": 10428 + }, + { + "epoch": 4.848442584844259, + "grad_norm": 7.011279106140137, + "learning_rate": 9.989576998957701e-06, + "loss": 0.5661, + "step": 10429 + }, + { + "epoch": 4.8489074848907485, + "grad_norm": 5.122862339019775, + "learning_rate": 9.9895759989576e-06, + "loss": 0.1247, + "step": 10430 + }, + { + "epoch": 4.849372384937238, + "grad_norm": 5.528585433959961, + "learning_rate": 9.9895749989575e-06, + "loss": 0.2973, + "step": 10431 + }, + { + "epoch": 4.849837284983728, + "grad_norm": 5.749905109405518, + "learning_rate": 9.9895739989574e-06, + "loss": 0.2526, + "step": 10432 + }, + { + "epoch": 4.850302185030219, + "grad_norm": 8.057342529296875, + "learning_rate": 9.9895729989573e-06, + "loss": 0.7167, + "step": 10433 + }, + { + "epoch": 4.850767085076709, + "grad_norm": 4.676977634429932, + "learning_rate": 9.989571998957201e-06, + "loss": 0.2327, + "step": 10434 + }, + { + "epoch": 4.851231985123198, + "grad_norm": 8.10913372039795, + "learning_rate": 9.989570998957102e-06, + "loss": 0.1812, + "step": 10435 + }, + { + "epoch": 4.851696885169688, + "grad_norm": 5.7972941398620605, + "learning_rate": 9.989569998957e-06, + "loss": 0.3306, + "step": 10436 + }, + { + "epoch": 4.852161785216179, + "grad_norm": 5.604498863220215, + "learning_rate": 9.989568998956901e-06, + "loss": 0.3734, + "step": 10437 + }, + { + "epoch": 4.852626685262669, + "grad_norm": 4.599228858947754, + "learning_rate": 9.9895679989568e-06, + "loss": 0.2473, + "step": 10438 + }, + { + "epoch": 4.8530915853091585, + "grad_norm": 5.1325578689575195, + "learning_rate": 9.9895669989567e-06, + "loss": 0.2246, + "step": 10439 + }, + { + "epoch": 4.853556485355648, + "grad_norm": 5.795267581939697, + "learning_rate": 9.989565998956601e-06, + "loss": 0.3302, + "step": 10440 + }, + { + "epoch": 4.854021385402138, + "grad_norm": 4.784295558929443, + "learning_rate": 9.9895649989565e-06, + "loss": 0.1821, + "step": 10441 + }, + { + "epoch": 4.854486285448629, + "grad_norm": 7.045658588409424, + "learning_rate": 9.989563998956401e-06, + "loss": 0.2404, + "step": 10442 + }, + { + "epoch": 4.854951185495119, + "grad_norm": 5.334747791290283, + "learning_rate": 9.9895629989563e-06, + "loss": 0.3571, + "step": 10443 + }, + { + "epoch": 4.855416085541608, + "grad_norm": 4.814773082733154, + "learning_rate": 9.9895619989562e-06, + "loss": 0.2373, + "step": 10444 + }, + { + "epoch": 4.855880985588098, + "grad_norm": 3.9083659648895264, + "learning_rate": 9.9895609989561e-06, + "loss": 0.2171, + "step": 10445 + }, + { + "epoch": 4.856345885634589, + "grad_norm": 6.048440456390381, + "learning_rate": 9.989559998956e-06, + "loss": 0.2403, + "step": 10446 + }, + { + "epoch": 4.856810785681079, + "grad_norm": 4.305479526519775, + "learning_rate": 9.989558998955901e-06, + "loss": 0.2797, + "step": 10447 + }, + { + "epoch": 4.8572756857275685, + "grad_norm": 6.051976680755615, + "learning_rate": 9.989557998955802e-06, + "loss": 0.2345, + "step": 10448 + }, + { + "epoch": 4.857740585774058, + "grad_norm": 4.404318332672119, + "learning_rate": 9.9895569989557e-06, + "loss": 0.1702, + "step": 10449 + }, + { + "epoch": 4.858205485820548, + "grad_norm": 6.133153915405273, + "learning_rate": 9.989555998955601e-06, + "loss": 0.4534, + "step": 10450 + }, + { + "epoch": 4.858670385867039, + "grad_norm": 4.28267765045166, + "learning_rate": 9.9895549989555e-06, + "loss": 0.1893, + "step": 10451 + }, + { + "epoch": 4.859135285913529, + "grad_norm": 8.280362129211426, + "learning_rate": 9.989553998955401e-06, + "loss": 0.3725, + "step": 10452 + }, + { + "epoch": 4.859600185960018, + "grad_norm": 9.341215133666992, + "learning_rate": 9.989552998955302e-06, + "loss": 0.5347, + "step": 10453 + }, + { + "epoch": 4.860065086006508, + "grad_norm": 7.108531951904297, + "learning_rate": 9.9895519989552e-06, + "loss": 0.6339, + "step": 10454 + }, + { + "epoch": 4.860529986052999, + "grad_norm": 5.933206558227539, + "learning_rate": 9.989550998955101e-06, + "loss": 0.2723, + "step": 10455 + }, + { + "epoch": 4.860994886099489, + "grad_norm": 5.56339168548584, + "learning_rate": 9.989549998955e-06, + "loss": 0.3806, + "step": 10456 + }, + { + "epoch": 4.8614597861459785, + "grad_norm": 4.5538506507873535, + "learning_rate": 9.9895489989549e-06, + "loss": 0.2251, + "step": 10457 + }, + { + "epoch": 4.861924686192468, + "grad_norm": 4.902257919311523, + "learning_rate": 9.9895479989548e-06, + "loss": 0.2984, + "step": 10458 + }, + { + "epoch": 4.862389586238959, + "grad_norm": 3.4506659507751465, + "learning_rate": 9.9895469989547e-06, + "loss": 0.106, + "step": 10459 + }, + { + "epoch": 4.862854486285449, + "grad_norm": 4.564420700073242, + "learning_rate": 9.989545998954601e-06, + "loss": 0.1796, + "step": 10460 + }, + { + "epoch": 4.863319386331939, + "grad_norm": 7.344308376312256, + "learning_rate": 9.989544998954502e-06, + "loss": 0.3535, + "step": 10461 + }, + { + "epoch": 4.863784286378428, + "grad_norm": 3.7129082679748535, + "learning_rate": 9.9895439989544e-06, + "loss": 0.1306, + "step": 10462 + }, + { + "epoch": 4.864249186424919, + "grad_norm": 6.495152950286865, + "learning_rate": 9.9895429989543e-06, + "loss": 0.5127, + "step": 10463 + }, + { + "epoch": 4.864714086471409, + "grad_norm": 6.85738468170166, + "learning_rate": 9.9895419989542e-06, + "loss": 0.4561, + "step": 10464 + }, + { + "epoch": 4.865178986517899, + "grad_norm": 6.099296569824219, + "learning_rate": 9.989540998954101e-06, + "loss": 0.1049, + "step": 10465 + }, + { + "epoch": 4.8656438865643885, + "grad_norm": 6.358236789703369, + "learning_rate": 9.989539998954002e-06, + "loss": 0.3002, + "step": 10466 + }, + { + "epoch": 4.866108786610878, + "grad_norm": 7.32103157043457, + "learning_rate": 9.9895389989539e-06, + "loss": 0.3976, + "step": 10467 + }, + { + "epoch": 4.866573686657369, + "grad_norm": 6.535960674285889, + "learning_rate": 9.989537998953801e-06, + "loss": 0.345, + "step": 10468 + }, + { + "epoch": 4.867038586703859, + "grad_norm": 6.607537746429443, + "learning_rate": 9.9895369989537e-06, + "loss": 0.3194, + "step": 10469 + }, + { + "epoch": 4.867503486750349, + "grad_norm": 4.253188610076904, + "learning_rate": 9.989535998953601e-06, + "loss": 0.2622, + "step": 10470 + }, + { + "epoch": 4.867968386796838, + "grad_norm": 5.267176151275635, + "learning_rate": 9.9895349989535e-06, + "loss": 0.2983, + "step": 10471 + }, + { + "epoch": 4.868433286843329, + "grad_norm": 7.186878681182861, + "learning_rate": 9.9895339989534e-06, + "loss": 0.3953, + "step": 10472 + }, + { + "epoch": 4.868898186889819, + "grad_norm": 3.3339431285858154, + "learning_rate": 9.989532998953301e-06, + "loss": 0.1146, + "step": 10473 + }, + { + "epoch": 4.869363086936309, + "grad_norm": 5.065343379974365, + "learning_rate": 9.9895319989532e-06, + "loss": 0.3493, + "step": 10474 + }, + { + "epoch": 4.8698279869827985, + "grad_norm": 4.721461772918701, + "learning_rate": 9.9895309989531e-06, + "loss": 0.1941, + "step": 10475 + }, + { + "epoch": 4.870292887029288, + "grad_norm": 7.274620056152344, + "learning_rate": 9.989529998953e-06, + "loss": 0.5274, + "step": 10476 + }, + { + "epoch": 4.870757787075779, + "grad_norm": 4.746023654937744, + "learning_rate": 9.9895289989529e-06, + "loss": 0.1883, + "step": 10477 + }, + { + "epoch": 4.871222687122269, + "grad_norm": 7.432840824127197, + "learning_rate": 9.989527998952801e-06, + "loss": 0.3432, + "step": 10478 + }, + { + "epoch": 4.871687587168759, + "grad_norm": 5.670336723327637, + "learning_rate": 9.989526998952702e-06, + "loss": 0.427, + "step": 10479 + }, + { + "epoch": 4.872152487215248, + "grad_norm": 5.519477367401123, + "learning_rate": 9.9895259989526e-06, + "loss": 0.3094, + "step": 10480 + }, + { + "epoch": 4.872617387261739, + "grad_norm": 12.957469940185547, + "learning_rate": 9.989524998952501e-06, + "loss": 1.2266, + "step": 10481 + }, + { + "epoch": 4.873082287308229, + "grad_norm": 4.010883331298828, + "learning_rate": 9.9895239989524e-06, + "loss": 0.1736, + "step": 10482 + }, + { + "epoch": 4.873547187354719, + "grad_norm": 4.552490234375, + "learning_rate": 9.989522998952301e-06, + "loss": 0.27, + "step": 10483 + }, + { + "epoch": 4.8740120874012085, + "grad_norm": 4.834558963775635, + "learning_rate": 9.9895219989522e-06, + "loss": 0.2027, + "step": 10484 + }, + { + "epoch": 4.874476987447698, + "grad_norm": 7.39992094039917, + "learning_rate": 9.9895209989521e-06, + "loss": 0.5024, + "step": 10485 + }, + { + "epoch": 4.874941887494189, + "grad_norm": 5.3884663581848145, + "learning_rate": 9.989519998952001e-06, + "loss": 0.3233, + "step": 10486 + }, + { + "epoch": 4.875406787540679, + "grad_norm": 6.157980442047119, + "learning_rate": 9.9895189989519e-06, + "loss": 0.3066, + "step": 10487 + }, + { + "epoch": 4.875871687587169, + "grad_norm": 5.420973777770996, + "learning_rate": 9.989517998951801e-06, + "loss": 0.3078, + "step": 10488 + }, + { + "epoch": 4.876336587633658, + "grad_norm": 8.14017105102539, + "learning_rate": 9.9895169989517e-06, + "loss": 0.219, + "step": 10489 + }, + { + "epoch": 4.876801487680149, + "grad_norm": 5.353580474853516, + "learning_rate": 9.9895159989516e-06, + "loss": 0.2803, + "step": 10490 + }, + { + "epoch": 4.877266387726639, + "grad_norm": 3.6561200618743896, + "learning_rate": 9.989514998951501e-06, + "loss": 0.1587, + "step": 10491 + }, + { + "epoch": 4.877731287773129, + "grad_norm": 3.5330257415771484, + "learning_rate": 9.989513998951402e-06, + "loss": 0.0504, + "step": 10492 + }, + { + "epoch": 4.8781961878196185, + "grad_norm": 4.540982246398926, + "learning_rate": 9.9895129989513e-06, + "loss": 0.2356, + "step": 10493 + }, + { + "epoch": 4.878661087866108, + "grad_norm": 9.90265941619873, + "learning_rate": 9.9895119989512e-06, + "loss": 0.6032, + "step": 10494 + }, + { + "epoch": 4.879125987912599, + "grad_norm": 5.057967185974121, + "learning_rate": 9.9895109989511e-06, + "loss": 0.3071, + "step": 10495 + }, + { + "epoch": 4.879590887959089, + "grad_norm": 3.3327581882476807, + "learning_rate": 9.989509998951001e-06, + "loss": 0.0955, + "step": 10496 + }, + { + "epoch": 4.880055788005579, + "grad_norm": 5.611229419708252, + "learning_rate": 9.989508998950902e-06, + "loss": 0.3155, + "step": 10497 + }, + { + "epoch": 4.880520688052069, + "grad_norm": 4.952291488647461, + "learning_rate": 9.9895079989508e-06, + "loss": 0.3285, + "step": 10498 + }, + { + "epoch": 4.880985588098559, + "grad_norm": 5.5484938621521, + "learning_rate": 9.989506998950701e-06, + "loss": 0.3603, + "step": 10499 + }, + { + "epoch": 4.881450488145049, + "grad_norm": 6.042341709136963, + "learning_rate": 9.9895059989506e-06, + "loss": 0.4568, + "step": 10500 + }, + { + "epoch": 4.881915388191539, + "grad_norm": 6.050080299377441, + "learning_rate": 9.989504998950501e-06, + "loss": 0.3986, + "step": 10501 + }, + { + "epoch": 4.8823802882380285, + "grad_norm": 5.026501655578613, + "learning_rate": 9.9895039989504e-06, + "loss": 0.3373, + "step": 10502 + }, + { + "epoch": 4.882845188284519, + "grad_norm": 4.85551643371582, + "learning_rate": 9.9895029989503e-06, + "loss": 0.3484, + "step": 10503 + }, + { + "epoch": 4.883310088331009, + "grad_norm": 5.398210048675537, + "learning_rate": 9.989501998950201e-06, + "loss": 0.2926, + "step": 10504 + }, + { + "epoch": 4.883774988377499, + "grad_norm": 4.266400337219238, + "learning_rate": 9.9895009989501e-06, + "loss": 0.0915, + "step": 10505 + }, + { + "epoch": 4.884239888423989, + "grad_norm": 4.588543891906738, + "learning_rate": 9.98949999895e-06, + "loss": 0.1814, + "step": 10506 + }, + { + "epoch": 4.884704788470479, + "grad_norm": 5.518252372741699, + "learning_rate": 9.9894989989499e-06, + "loss": 0.4063, + "step": 10507 + }, + { + "epoch": 4.885169688516969, + "grad_norm": 7.445760726928711, + "learning_rate": 9.9894979989498e-06, + "loss": 0.4119, + "step": 10508 + }, + { + "epoch": 4.885634588563459, + "grad_norm": 6.482110977172852, + "learning_rate": 9.989496998949701e-06, + "loss": 0.3614, + "step": 10509 + }, + { + "epoch": 4.886099488609949, + "grad_norm": 4.764923572540283, + "learning_rate": 9.989495998949602e-06, + "loss": 0.2063, + "step": 10510 + }, + { + "epoch": 4.8865643886564385, + "grad_norm": 6.132972240447998, + "learning_rate": 9.9894949989495e-06, + "loss": 0.4662, + "step": 10511 + }, + { + "epoch": 4.887029288702929, + "grad_norm": 5.867253303527832, + "learning_rate": 9.989493998949401e-06, + "loss": 0.3303, + "step": 10512 + }, + { + "epoch": 4.887494188749419, + "grad_norm": 6.810372829437256, + "learning_rate": 9.9894929989493e-06, + "loss": 0.2569, + "step": 10513 + }, + { + "epoch": 4.887959088795909, + "grad_norm": 9.638239860534668, + "learning_rate": 9.989491998949201e-06, + "loss": 0.3349, + "step": 10514 + }, + { + "epoch": 4.888423988842399, + "grad_norm": 9.678293228149414, + "learning_rate": 9.9894909989491e-06, + "loss": 0.4873, + "step": 10515 + }, + { + "epoch": 4.888888888888889, + "grad_norm": 5.787105560302734, + "learning_rate": 9.989489998949e-06, + "loss": 0.4064, + "step": 10516 + }, + { + "epoch": 4.889353788935379, + "grad_norm": 5.487607002258301, + "learning_rate": 9.989488998948901e-06, + "loss": 0.3443, + "step": 10517 + }, + { + "epoch": 4.889818688981869, + "grad_norm": 4.199826240539551, + "learning_rate": 9.9894879989488e-06, + "loss": 0.0962, + "step": 10518 + }, + { + "epoch": 4.890283589028359, + "grad_norm": 7.831302642822266, + "learning_rate": 9.989486998948701e-06, + "loss": 0.4288, + "step": 10519 + }, + { + "epoch": 4.8907484890748485, + "grad_norm": 6.5926432609558105, + "learning_rate": 9.9894859989486e-06, + "loss": 0.2624, + "step": 10520 + }, + { + "epoch": 4.891213389121339, + "grad_norm": 4.537871360778809, + "learning_rate": 9.9894849989485e-06, + "loss": 0.2827, + "step": 10521 + }, + { + "epoch": 4.891678289167829, + "grad_norm": 9.249168395996094, + "learning_rate": 9.989483998948401e-06, + "loss": 0.2043, + "step": 10522 + }, + { + "epoch": 4.892143189214319, + "grad_norm": 4.330883026123047, + "learning_rate": 9.989482998948302e-06, + "loss": 0.21, + "step": 10523 + }, + { + "epoch": 4.892608089260809, + "grad_norm": 4.660626411437988, + "learning_rate": 9.9894819989482e-06, + "loss": 0.2785, + "step": 10524 + }, + { + "epoch": 4.893072989307299, + "grad_norm": 6.561645030975342, + "learning_rate": 9.989480998948101e-06, + "loss": 0.416, + "step": 10525 + }, + { + "epoch": 4.893537889353789, + "grad_norm": 4.958625316619873, + "learning_rate": 9.989479998948e-06, + "loss": 0.2631, + "step": 10526 + }, + { + "epoch": 4.894002789400279, + "grad_norm": 2.9375908374786377, + "learning_rate": 9.9894789989479e-06, + "loss": 0.0958, + "step": 10527 + }, + { + "epoch": 4.894467689446769, + "grad_norm": 5.4310736656188965, + "learning_rate": 9.9894779989478e-06, + "loss": 0.3474, + "step": 10528 + }, + { + "epoch": 4.8949325894932585, + "grad_norm": 6.8583784103393555, + "learning_rate": 9.9894769989477e-06, + "loss": 0.3776, + "step": 10529 + }, + { + "epoch": 4.895397489539749, + "grad_norm": 7.892832279205322, + "learning_rate": 9.989475998947601e-06, + "loss": 0.4946, + "step": 10530 + }, + { + "epoch": 4.895862389586239, + "grad_norm": 8.737588882446289, + "learning_rate": 9.9894749989475e-06, + "loss": 0.3866, + "step": 10531 + }, + { + "epoch": 4.896327289632729, + "grad_norm": 5.724647045135498, + "learning_rate": 9.989473998947401e-06, + "loss": 0.255, + "step": 10532 + }, + { + "epoch": 4.896792189679219, + "grad_norm": 6.296819686889648, + "learning_rate": 9.9894729989473e-06, + "loss": 0.4524, + "step": 10533 + }, + { + "epoch": 4.897257089725709, + "grad_norm": 6.243406772613525, + "learning_rate": 9.9894719989472e-06, + "loss": 0.4415, + "step": 10534 + }, + { + "epoch": 4.897721989772199, + "grad_norm": 5.334801197052002, + "learning_rate": 9.989470998947101e-06, + "loss": 0.2314, + "step": 10535 + }, + { + "epoch": 4.898186889818689, + "grad_norm": 6.422966003417969, + "learning_rate": 9.989469998947002e-06, + "loss": 0.4576, + "step": 10536 + }, + { + "epoch": 4.898651789865179, + "grad_norm": 9.060644149780273, + "learning_rate": 9.989468998946901e-06, + "loss": 0.2993, + "step": 10537 + }, + { + "epoch": 4.899116689911669, + "grad_norm": 5.209707260131836, + "learning_rate": 9.9894679989468e-06, + "loss": 0.3432, + "step": 10538 + }, + { + "epoch": 4.899581589958159, + "grad_norm": 5.125683784484863, + "learning_rate": 9.9894669989467e-06, + "loss": 0.1224, + "step": 10539 + }, + { + "epoch": 4.900046490004649, + "grad_norm": 5.708930492401123, + "learning_rate": 9.9894659989466e-06, + "loss": 0.3027, + "step": 10540 + }, + { + "epoch": 4.900511390051139, + "grad_norm": 4.718201637268066, + "learning_rate": 9.9894649989465e-06, + "loss": 0.3076, + "step": 10541 + }, + { + "epoch": 4.9009762900976295, + "grad_norm": 5.6802544593811035, + "learning_rate": 9.9894639989464e-06, + "loss": 0.2749, + "step": 10542 + }, + { + "epoch": 4.901441190144119, + "grad_norm": 7.435101509094238, + "learning_rate": 9.989462998946301e-06, + "loss": 0.3278, + "step": 10543 + }, + { + "epoch": 4.901906090190609, + "grad_norm": 5.884664058685303, + "learning_rate": 9.9894619989462e-06, + "loss": 0.3519, + "step": 10544 + }, + { + "epoch": 4.902370990237099, + "grad_norm": 5.551841735839844, + "learning_rate": 9.989460998946101e-06, + "loss": 0.1472, + "step": 10545 + }, + { + "epoch": 4.902835890283589, + "grad_norm": 6.4578118324279785, + "learning_rate": 9.989459998946e-06, + "loss": 0.3394, + "step": 10546 + }, + { + "epoch": 4.903300790330079, + "grad_norm": 4.912051200866699, + "learning_rate": 9.9894589989459e-06, + "loss": 0.3055, + "step": 10547 + }, + { + "epoch": 4.903765690376569, + "grad_norm": 4.005723476409912, + "learning_rate": 9.989457998945801e-06, + "loss": 0.1798, + "step": 10548 + }, + { + "epoch": 4.904230590423059, + "grad_norm": 5.3203959465026855, + "learning_rate": 9.9894569989457e-06, + "loss": 0.3288, + "step": 10549 + }, + { + "epoch": 4.904695490469549, + "grad_norm": 4.41676664352417, + "learning_rate": 9.989455998945601e-06, + "loss": 0.2027, + "step": 10550 + }, + { + "epoch": 4.9051603905160395, + "grad_norm": 8.230388641357422, + "learning_rate": 9.9894549989455e-06, + "loss": 0.4097, + "step": 10551 + }, + { + "epoch": 4.905625290562529, + "grad_norm": 3.3890862464904785, + "learning_rate": 9.9894539989454e-06, + "loss": 0.1312, + "step": 10552 + }, + { + "epoch": 4.906090190609019, + "grad_norm": 4.61184024810791, + "learning_rate": 9.9894529989453e-06, + "loss": 0.3231, + "step": 10553 + }, + { + "epoch": 4.906555090655509, + "grad_norm": 4.564674377441406, + "learning_rate": 9.9894519989452e-06, + "loss": 0.1169, + "step": 10554 + }, + { + "epoch": 4.907019990701999, + "grad_norm": 6.257568836212158, + "learning_rate": 9.989450998945101e-06, + "loss": 0.371, + "step": 10555 + }, + { + "epoch": 4.907484890748489, + "grad_norm": 3.6141955852508545, + "learning_rate": 9.989449998945002e-06, + "loss": 0.11, + "step": 10556 + }, + { + "epoch": 4.907949790794979, + "grad_norm": 4.629188537597656, + "learning_rate": 9.9894489989449e-06, + "loss": 0.1997, + "step": 10557 + }, + { + "epoch": 4.908414690841469, + "grad_norm": 8.69906997680664, + "learning_rate": 9.989447998944801e-06, + "loss": 0.5964, + "step": 10558 + }, + { + "epoch": 4.908879590887959, + "grad_norm": 10.639261245727539, + "learning_rate": 9.9894469989447e-06, + "loss": 0.431, + "step": 10559 + }, + { + "epoch": 4.9093444909344495, + "grad_norm": 6.052586078643799, + "learning_rate": 9.9894459989446e-06, + "loss": 0.2948, + "step": 10560 + }, + { + "epoch": 4.909809390980939, + "grad_norm": 5.099608421325684, + "learning_rate": 9.989444998944501e-06, + "loss": 0.3946, + "step": 10561 + }, + { + "epoch": 4.910274291027429, + "grad_norm": 6.175998687744141, + "learning_rate": 9.9894439989444e-06, + "loss": 0.3476, + "step": 10562 + }, + { + "epoch": 4.910739191073919, + "grad_norm": 5.189296722412109, + "learning_rate": 9.989442998944301e-06, + "loss": 0.2565, + "step": 10563 + }, + { + "epoch": 4.911204091120409, + "grad_norm": 7.555948734283447, + "learning_rate": 9.9894419989442e-06, + "loss": 0.4141, + "step": 10564 + }, + { + "epoch": 4.911668991166899, + "grad_norm": 4.998579978942871, + "learning_rate": 9.9894409989441e-06, + "loss": 0.2234, + "step": 10565 + }, + { + "epoch": 4.912133891213389, + "grad_norm": 9.95439624786377, + "learning_rate": 9.989439998944001e-06, + "loss": 0.5388, + "step": 10566 + }, + { + "epoch": 4.912598791259879, + "grad_norm": 4.566199779510498, + "learning_rate": 9.9894389989439e-06, + "loss": 0.2775, + "step": 10567 + }, + { + "epoch": 4.913063691306369, + "grad_norm": 5.884324550628662, + "learning_rate": 9.989437998943801e-06, + "loss": 0.2339, + "step": 10568 + }, + { + "epoch": 4.9135285913528595, + "grad_norm": 5.9989118576049805, + "learning_rate": 9.989436998943702e-06, + "loss": 0.3151, + "step": 10569 + }, + { + "epoch": 4.913993491399349, + "grad_norm": 5.891808032989502, + "learning_rate": 9.9894359989436e-06, + "loss": 0.4068, + "step": 10570 + }, + { + "epoch": 4.914458391445839, + "grad_norm": 4.897436141967773, + "learning_rate": 9.9894349989435e-06, + "loss": 0.1947, + "step": 10571 + }, + { + "epoch": 4.914923291492329, + "grad_norm": 6.194758415222168, + "learning_rate": 9.9894339989434e-06, + "loss": 0.2842, + "step": 10572 + }, + { + "epoch": 4.915388191538819, + "grad_norm": 5.090296745300293, + "learning_rate": 9.989432998943301e-06, + "loss": 0.2258, + "step": 10573 + }, + { + "epoch": 4.915853091585309, + "grad_norm": 5.345414638519287, + "learning_rate": 9.989431998943202e-06, + "loss": 0.2699, + "step": 10574 + }, + { + "epoch": 4.916317991631799, + "grad_norm": 6.013545513153076, + "learning_rate": 9.9894309989431e-06, + "loss": 0.1895, + "step": 10575 + }, + { + "epoch": 4.916782891678289, + "grad_norm": 7.442067623138428, + "learning_rate": 9.989429998943001e-06, + "loss": 0.191, + "step": 10576 + }, + { + "epoch": 4.91724779172478, + "grad_norm": 6.5143632888793945, + "learning_rate": 9.9894289989429e-06, + "loss": 0.3207, + "step": 10577 + }, + { + "epoch": 4.9177126917712695, + "grad_norm": 5.564217567443848, + "learning_rate": 9.9894279989428e-06, + "loss": 0.282, + "step": 10578 + }, + { + "epoch": 4.918177591817759, + "grad_norm": 7.707867622375488, + "learning_rate": 9.989426998942701e-06, + "loss": 0.5223, + "step": 10579 + }, + { + "epoch": 4.918642491864249, + "grad_norm": 5.1799635887146, + "learning_rate": 9.989425998942602e-06, + "loss": 0.2783, + "step": 10580 + }, + { + "epoch": 4.919107391910739, + "grad_norm": 4.548023223876953, + "learning_rate": 9.989424998942501e-06, + "loss": 0.2425, + "step": 10581 + }, + { + "epoch": 4.91957229195723, + "grad_norm": 6.330061912536621, + "learning_rate": 9.9894239989424e-06, + "loss": 0.3455, + "step": 10582 + }, + { + "epoch": 4.920037192003719, + "grad_norm": 6.407558917999268, + "learning_rate": 9.9894229989423e-06, + "loss": 0.3578, + "step": 10583 + }, + { + "epoch": 4.920502092050209, + "grad_norm": 6.738530158996582, + "learning_rate": 9.9894219989422e-06, + "loss": 0.4287, + "step": 10584 + }, + { + "epoch": 4.920966992096699, + "grad_norm": 8.287714958190918, + "learning_rate": 9.9894209989421e-06, + "loss": 0.394, + "step": 10585 + }, + { + "epoch": 4.92143189214319, + "grad_norm": 6.706441402435303, + "learning_rate": 9.989419998942001e-06, + "loss": 0.4772, + "step": 10586 + }, + { + "epoch": 4.9218967921896795, + "grad_norm": 2.8545048236846924, + "learning_rate": 9.989418998941902e-06, + "loss": 0.0482, + "step": 10587 + }, + { + "epoch": 4.922361692236169, + "grad_norm": 3.0848798751831055, + "learning_rate": 9.9894179989418e-06, + "loss": 0.0949, + "step": 10588 + }, + { + "epoch": 4.922826592282659, + "grad_norm": 6.055418491363525, + "learning_rate": 9.989416998941701e-06, + "loss": 0.5135, + "step": 10589 + }, + { + "epoch": 4.923291492329149, + "grad_norm": 4.398275375366211, + "learning_rate": 9.9894159989416e-06, + "loss": 0.2088, + "step": 10590 + }, + { + "epoch": 4.92375639237564, + "grad_norm": 3.2465147972106934, + "learning_rate": 9.9894149989415e-06, + "loss": 0.1268, + "step": 10591 + }, + { + "epoch": 4.924221292422129, + "grad_norm": 5.617159366607666, + "learning_rate": 9.989413998941402e-06, + "loss": 0.3973, + "step": 10592 + }, + { + "epoch": 4.924686192468619, + "grad_norm": 9.342296600341797, + "learning_rate": 9.9894129989413e-06, + "loss": 0.451, + "step": 10593 + }, + { + "epoch": 4.925151092515109, + "grad_norm": 5.223686218261719, + "learning_rate": 9.989411998941201e-06, + "loss": 0.1808, + "step": 10594 + }, + { + "epoch": 4.9256159925616, + "grad_norm": 6.996793270111084, + "learning_rate": 9.9894109989411e-06, + "loss": 0.2123, + "step": 10595 + }, + { + "epoch": 4.9260808926080895, + "grad_norm": 5.562061309814453, + "learning_rate": 9.989409998941e-06, + "loss": 0.2949, + "step": 10596 + }, + { + "epoch": 4.926545792654579, + "grad_norm": 4.007220268249512, + "learning_rate": 9.9894089989409e-06, + "loss": 0.1651, + "step": 10597 + }, + { + "epoch": 4.927010692701069, + "grad_norm": 3.034804344177246, + "learning_rate": 9.9894079989408e-06, + "loss": 0.0791, + "step": 10598 + }, + { + "epoch": 4.927475592747559, + "grad_norm": 7.132925033569336, + "learning_rate": 9.989406998940701e-06, + "loss": 0.3835, + "step": 10599 + }, + { + "epoch": 4.9279404927940496, + "grad_norm": 5.805181503295898, + "learning_rate": 9.989405998940602e-06, + "loss": 0.3729, + "step": 10600 + }, + { + "epoch": 4.928405392840539, + "grad_norm": 6.6857686042785645, + "learning_rate": 9.9894049989405e-06, + "loss": 0.5186, + "step": 10601 + }, + { + "epoch": 4.928870292887029, + "grad_norm": 5.868505477905273, + "learning_rate": 9.989403998940401e-06, + "loss": 0.1615, + "step": 10602 + }, + { + "epoch": 4.929335192933519, + "grad_norm": 4.727272987365723, + "learning_rate": 9.9894029989403e-06, + "loss": 0.106, + "step": 10603 + }, + { + "epoch": 4.92980009298001, + "grad_norm": 6.5863847732543945, + "learning_rate": 9.989401998940201e-06, + "loss": 0.1432, + "step": 10604 + }, + { + "epoch": 4.9302649930264995, + "grad_norm": 5.2050299644470215, + "learning_rate": 9.989400998940102e-06, + "loss": 0.1345, + "step": 10605 + }, + { + "epoch": 4.930729893072989, + "grad_norm": 4.057097434997559, + "learning_rate": 9.98939999894e-06, + "loss": 0.2438, + "step": 10606 + }, + { + "epoch": 4.931194793119479, + "grad_norm": 9.318946838378906, + "learning_rate": 9.989398998939901e-06, + "loss": 0.3546, + "step": 10607 + }, + { + "epoch": 4.931659693165969, + "grad_norm": 5.652122974395752, + "learning_rate": 9.9893979989398e-06, + "loss": 0.2902, + "step": 10608 + }, + { + "epoch": 4.9321245932124596, + "grad_norm": 4.994584560394287, + "learning_rate": 9.9893969989397e-06, + "loss": 0.289, + "step": 10609 + }, + { + "epoch": 4.932589493258949, + "grad_norm": 4.060024738311768, + "learning_rate": 9.9893959989396e-06, + "loss": 0.1068, + "step": 10610 + }, + { + "epoch": 4.933054393305439, + "grad_norm": 4.987418174743652, + "learning_rate": 9.9893949989395e-06, + "loss": 0.2635, + "step": 10611 + }, + { + "epoch": 4.933519293351929, + "grad_norm": 5.64089298248291, + "learning_rate": 9.989393998939401e-06, + "loss": 0.3195, + "step": 10612 + }, + { + "epoch": 4.93398419339842, + "grad_norm": 6.394387245178223, + "learning_rate": 9.989392998939302e-06, + "loss": 0.3042, + "step": 10613 + }, + { + "epoch": 4.9344490934449095, + "grad_norm": 10.501381874084473, + "learning_rate": 9.9893919989392e-06, + "loss": 0.7512, + "step": 10614 + }, + { + "epoch": 4.934913993491399, + "grad_norm": 5.728550434112549, + "learning_rate": 9.9893909989391e-06, + "loss": 0.4003, + "step": 10615 + }, + { + "epoch": 4.935378893537889, + "grad_norm": 5.515946388244629, + "learning_rate": 9.989389998939e-06, + "loss": 0.2383, + "step": 10616 + }, + { + "epoch": 4.93584379358438, + "grad_norm": 4.785565376281738, + "learning_rate": 9.989388998938901e-06, + "loss": 0.215, + "step": 10617 + }, + { + "epoch": 4.9363086936308695, + "grad_norm": 7.552608013153076, + "learning_rate": 9.989387998938802e-06, + "loss": 0.4091, + "step": 10618 + }, + { + "epoch": 4.936773593677359, + "grad_norm": 4.807866096496582, + "learning_rate": 9.9893869989387e-06, + "loss": 0.2372, + "step": 10619 + }, + { + "epoch": 4.937238493723849, + "grad_norm": 5.967960834503174, + "learning_rate": 9.989385998938601e-06, + "loss": 0.211, + "step": 10620 + }, + { + "epoch": 4.93770339377034, + "grad_norm": 6.02923059463501, + "learning_rate": 9.9893849989385e-06, + "loss": 0.4907, + "step": 10621 + }, + { + "epoch": 4.93816829381683, + "grad_norm": 5.534809112548828, + "learning_rate": 9.989383998938401e-06, + "loss": 0.2068, + "step": 10622 + }, + { + "epoch": 4.9386331938633194, + "grad_norm": 7.995306968688965, + "learning_rate": 9.9893829989383e-06, + "loss": 0.7649, + "step": 10623 + }, + { + "epoch": 4.939098093909809, + "grad_norm": 7.1569342613220215, + "learning_rate": 9.9893819989382e-06, + "loss": 0.4351, + "step": 10624 + }, + { + "epoch": 4.939562993956299, + "grad_norm": 5.895422458648682, + "learning_rate": 9.989380998938101e-06, + "loss": 0.4062, + "step": 10625 + }, + { + "epoch": 4.94002789400279, + "grad_norm": 5.170822620391846, + "learning_rate": 9.989379998938e-06, + "loss": 0.3449, + "step": 10626 + }, + { + "epoch": 4.9404927940492795, + "grad_norm": 7.087022304534912, + "learning_rate": 9.9893789989379e-06, + "loss": 0.2364, + "step": 10627 + }, + { + "epoch": 4.940957694095769, + "grad_norm": 6.806859970092773, + "learning_rate": 9.9893779989378e-06, + "loss": 0.5404, + "step": 10628 + }, + { + "epoch": 4.941422594142259, + "grad_norm": 6.9973931312561035, + "learning_rate": 9.9893769989377e-06, + "loss": 0.3969, + "step": 10629 + }, + { + "epoch": 4.94188749418875, + "grad_norm": 5.394634246826172, + "learning_rate": 9.989375998937601e-06, + "loss": 0.3056, + "step": 10630 + }, + { + "epoch": 4.94235239423524, + "grad_norm": 5.733177661895752, + "learning_rate": 9.989374998937502e-06, + "loss": 0.4799, + "step": 10631 + }, + { + "epoch": 4.9428172942817294, + "grad_norm": 4.735509395599365, + "learning_rate": 9.9893739989374e-06, + "loss": 0.2756, + "step": 10632 + }, + { + "epoch": 4.943282194328219, + "grad_norm": 4.932458400726318, + "learning_rate": 9.989372998937301e-06, + "loss": 0.2673, + "step": 10633 + }, + { + "epoch": 4.943747094374709, + "grad_norm": 8.441084861755371, + "learning_rate": 9.9893719989372e-06, + "loss": 0.4066, + "step": 10634 + }, + { + "epoch": 4.9442119944212, + "grad_norm": 4.6635422706604, + "learning_rate": 9.989370998937101e-06, + "loss": 0.3052, + "step": 10635 + }, + { + "epoch": 4.9446768944676895, + "grad_norm": 5.227269649505615, + "learning_rate": 9.989369998937e-06, + "loss": 0.2958, + "step": 10636 + }, + { + "epoch": 4.945141794514179, + "grad_norm": 6.229074954986572, + "learning_rate": 9.9893689989369e-06, + "loss": 0.2646, + "step": 10637 + }, + { + "epoch": 4.945606694560669, + "grad_norm": 7.7406535148620605, + "learning_rate": 9.989367998936801e-06, + "loss": 0.2834, + "step": 10638 + }, + { + "epoch": 4.94607159460716, + "grad_norm": 6.820900917053223, + "learning_rate": 9.9893669989367e-06, + "loss": 0.3617, + "step": 10639 + }, + { + "epoch": 4.94653649465365, + "grad_norm": 3.9623329639434814, + "learning_rate": 9.989365998936601e-06, + "loss": 0.1299, + "step": 10640 + }, + { + "epoch": 4.947001394700139, + "grad_norm": 6.350567817687988, + "learning_rate": 9.9893649989365e-06, + "loss": 0.2333, + "step": 10641 + }, + { + "epoch": 4.947466294746629, + "grad_norm": 6.164994716644287, + "learning_rate": 9.9893639989364e-06, + "loss": 0.3349, + "step": 10642 + }, + { + "epoch": 4.947931194793119, + "grad_norm": 4.264883995056152, + "learning_rate": 9.989362998936301e-06, + "loss": 0.29, + "step": 10643 + }, + { + "epoch": 4.94839609483961, + "grad_norm": 4.04359769821167, + "learning_rate": 9.989361998936202e-06, + "loss": 0.1655, + "step": 10644 + }, + { + "epoch": 4.9488609948860995, + "grad_norm": 5.274411678314209, + "learning_rate": 9.9893609989361e-06, + "loss": 0.3611, + "step": 10645 + }, + { + "epoch": 4.949325894932589, + "grad_norm": 3.0151216983795166, + "learning_rate": 9.989359998936e-06, + "loss": 0.0802, + "step": 10646 + }, + { + "epoch": 4.949790794979079, + "grad_norm": 6.992415428161621, + "learning_rate": 9.9893589989359e-06, + "loss": 0.3022, + "step": 10647 + }, + { + "epoch": 4.95025569502557, + "grad_norm": 5.178832530975342, + "learning_rate": 9.989357998935801e-06, + "loss": 0.1465, + "step": 10648 + }, + { + "epoch": 4.95072059507206, + "grad_norm": 5.834742069244385, + "learning_rate": 9.9893569989357e-06, + "loss": 0.374, + "step": 10649 + }, + { + "epoch": 4.951185495118549, + "grad_norm": 5.619356155395508, + "learning_rate": 9.9893559989356e-06, + "loss": 0.2606, + "step": 10650 + }, + { + "epoch": 4.951650395165039, + "grad_norm": 7.204737186431885, + "learning_rate": 9.989354998935501e-06, + "loss": 0.3851, + "step": 10651 + }, + { + "epoch": 4.952115295211529, + "grad_norm": 5.272088050842285, + "learning_rate": 9.9893539989354e-06, + "loss": 0.2902, + "step": 10652 + }, + { + "epoch": 4.95258019525802, + "grad_norm": 5.031311988830566, + "learning_rate": 9.989352998935301e-06, + "loss": 0.2527, + "step": 10653 + }, + { + "epoch": 4.9530450953045095, + "grad_norm": 9.052371978759766, + "learning_rate": 9.9893519989352e-06, + "loss": 0.2873, + "step": 10654 + }, + { + "epoch": 4.953509995350999, + "grad_norm": 5.5227742195129395, + "learning_rate": 9.9893509989351e-06, + "loss": 0.3322, + "step": 10655 + }, + { + "epoch": 4.95397489539749, + "grad_norm": 6.847362041473389, + "learning_rate": 9.989349998935001e-06, + "loss": 0.3938, + "step": 10656 + }, + { + "epoch": 4.95443979544398, + "grad_norm": 7.1030120849609375, + "learning_rate": 9.9893489989349e-06, + "loss": 0.332, + "step": 10657 + }, + { + "epoch": 4.95490469549047, + "grad_norm": 8.039301872253418, + "learning_rate": 9.989347998934801e-06, + "loss": 0.3079, + "step": 10658 + }, + { + "epoch": 4.955369595536959, + "grad_norm": 10.109467506408691, + "learning_rate": 9.9893469989347e-06, + "loss": 0.5062, + "step": 10659 + }, + { + "epoch": 4.955834495583449, + "grad_norm": 6.728675365447998, + "learning_rate": 9.9893459989346e-06, + "loss": 0.4035, + "step": 10660 + }, + { + "epoch": 4.95629939562994, + "grad_norm": 15.359380722045898, + "learning_rate": 9.989344998934501e-06, + "loss": 0.408, + "step": 10661 + }, + { + "epoch": 4.95676429567643, + "grad_norm": 4.256460666656494, + "learning_rate": 9.989343998934402e-06, + "loss": 0.1171, + "step": 10662 + }, + { + "epoch": 4.9572291957229195, + "grad_norm": 6.270087242126465, + "learning_rate": 9.9893429989343e-06, + "loss": 0.3975, + "step": 10663 + }, + { + "epoch": 4.957694095769409, + "grad_norm": 5.812521934509277, + "learning_rate": 9.989341998934201e-06, + "loss": 0.3569, + "step": 10664 + }, + { + "epoch": 4.9581589958159, + "grad_norm": 4.983554363250732, + "learning_rate": 9.9893409989341e-06, + "loss": 0.2295, + "step": 10665 + }, + { + "epoch": 4.95862389586239, + "grad_norm": 5.368256568908691, + "learning_rate": 9.989339998934001e-06, + "loss": 0.3533, + "step": 10666 + }, + { + "epoch": 4.95908879590888, + "grad_norm": 6.082484722137451, + "learning_rate": 9.9893389989339e-06, + "loss": 0.42, + "step": 10667 + }, + { + "epoch": 4.959553695955369, + "grad_norm": 6.238668441772461, + "learning_rate": 9.9893379989338e-06, + "loss": 0.2247, + "step": 10668 + }, + { + "epoch": 4.960018596001859, + "grad_norm": 6.221591472625732, + "learning_rate": 9.989336998933701e-06, + "loss": 0.2962, + "step": 10669 + }, + { + "epoch": 4.96048349604835, + "grad_norm": 5.139761447906494, + "learning_rate": 9.9893359989336e-06, + "loss": 0.2986, + "step": 10670 + }, + { + "epoch": 4.96094839609484, + "grad_norm": 4.747359752655029, + "learning_rate": 9.989334998933501e-06, + "loss": 0.2853, + "step": 10671 + }, + { + "epoch": 4.9614132961413295, + "grad_norm": 6.250209331512451, + "learning_rate": 9.9893339989334e-06, + "loss": 0.3666, + "step": 10672 + }, + { + "epoch": 4.961878196187819, + "grad_norm": 8.00385570526123, + "learning_rate": 9.9893329989333e-06, + "loss": 0.3949, + "step": 10673 + }, + { + "epoch": 4.96234309623431, + "grad_norm": 3.754854679107666, + "learning_rate": 9.989331998933201e-06, + "loss": 0.2093, + "step": 10674 + }, + { + "epoch": 4.9628079962808, + "grad_norm": 5.553319454193115, + "learning_rate": 9.989330998933102e-06, + "loss": 0.1256, + "step": 10675 + }, + { + "epoch": 4.96327289632729, + "grad_norm": 6.881571292877197, + "learning_rate": 9.989329998933001e-06, + "loss": 0.2813, + "step": 10676 + }, + { + "epoch": 4.963737796373779, + "grad_norm": 4.349122524261475, + "learning_rate": 9.989328998932902e-06, + "loss": 0.2853, + "step": 10677 + }, + { + "epoch": 4.964202696420269, + "grad_norm": 5.5583815574646, + "learning_rate": 9.9893279989328e-06, + "loss": 0.2551, + "step": 10678 + }, + { + "epoch": 4.96466759646676, + "grad_norm": 3.657226085662842, + "learning_rate": 9.9893269989327e-06, + "loss": 0.2198, + "step": 10679 + }, + { + "epoch": 4.96513249651325, + "grad_norm": 6.376910209655762, + "learning_rate": 9.9893259989326e-06, + "loss": 0.3233, + "step": 10680 + }, + { + "epoch": 4.9655973965597395, + "grad_norm": 2.979121446609497, + "learning_rate": 9.9893249989325e-06, + "loss": 0.165, + "step": 10681 + }, + { + "epoch": 4.966062296606229, + "grad_norm": 8.382972717285156, + "learning_rate": 9.989323998932401e-06, + "loss": 0.3058, + "step": 10682 + }, + { + "epoch": 4.96652719665272, + "grad_norm": 6.374612331390381, + "learning_rate": 9.9893229989323e-06, + "loss": 0.3165, + "step": 10683 + }, + { + "epoch": 4.96699209669921, + "grad_norm": 9.222305297851562, + "learning_rate": 9.989321998932201e-06, + "loss": 0.1436, + "step": 10684 + }, + { + "epoch": 4.9674569967457, + "grad_norm": 6.465320110321045, + "learning_rate": 9.9893209989321e-06, + "loss": 0.4558, + "step": 10685 + }, + { + "epoch": 4.967921896792189, + "grad_norm": 5.325767517089844, + "learning_rate": 9.989319998932e-06, + "loss": 0.207, + "step": 10686 + }, + { + "epoch": 4.968386796838679, + "grad_norm": 5.133421897888184, + "learning_rate": 9.989318998931901e-06, + "loss": 0.2894, + "step": 10687 + }, + { + "epoch": 4.96885169688517, + "grad_norm": 8.011754989624023, + "learning_rate": 9.989317998931802e-06, + "loss": 0.4994, + "step": 10688 + }, + { + "epoch": 4.96931659693166, + "grad_norm": 4.237745761871338, + "learning_rate": 9.989316998931701e-06, + "loss": 0.1875, + "step": 10689 + }, + { + "epoch": 4.9697814969781495, + "grad_norm": 8.093353271484375, + "learning_rate": 9.9893159989316e-06, + "loss": 0.5852, + "step": 10690 + }, + { + "epoch": 4.970246397024639, + "grad_norm": 5.847987174987793, + "learning_rate": 9.9893149989315e-06, + "loss": 0.245, + "step": 10691 + }, + { + "epoch": 4.97071129707113, + "grad_norm": 6.503767490386963, + "learning_rate": 9.9893139989314e-06, + "loss": 0.2792, + "step": 10692 + }, + { + "epoch": 4.97117619711762, + "grad_norm": 9.72176742553711, + "learning_rate": 9.9893129989313e-06, + "loss": 0.2821, + "step": 10693 + }, + { + "epoch": 4.97164109716411, + "grad_norm": 3.2901926040649414, + "learning_rate": 9.9893119989312e-06, + "loss": 0.1699, + "step": 10694 + }, + { + "epoch": 4.972105997210599, + "grad_norm": 4.909605979919434, + "learning_rate": 9.989310998931101e-06, + "loss": 0.1813, + "step": 10695 + }, + { + "epoch": 4.972570897257089, + "grad_norm": 3.614509105682373, + "learning_rate": 9.989309998931e-06, + "loss": 0.1432, + "step": 10696 + }, + { + "epoch": 4.97303579730358, + "grad_norm": 11.494155883789062, + "learning_rate": 9.989308998930901e-06, + "loss": 0.3413, + "step": 10697 + }, + { + "epoch": 4.97350069735007, + "grad_norm": 4.234290599822998, + "learning_rate": 9.9893079989308e-06, + "loss": 0.1083, + "step": 10698 + }, + { + "epoch": 4.9739655973965595, + "grad_norm": 5.378146648406982, + "learning_rate": 9.9893069989307e-06, + "loss": 0.31, + "step": 10699 + }, + { + "epoch": 4.97443049744305, + "grad_norm": 4.880207061767578, + "learning_rate": 9.989305998930601e-06, + "loss": 0.2571, + "step": 10700 + }, + { + "epoch": 4.97489539748954, + "grad_norm": 5.262909412384033, + "learning_rate": 9.9893049989305e-06, + "loss": 0.2903, + "step": 10701 + }, + { + "epoch": 4.97536029753603, + "grad_norm": 6.390863418579102, + "learning_rate": 9.989303998930401e-06, + "loss": 0.4723, + "step": 10702 + }, + { + "epoch": 4.97582519758252, + "grad_norm": 5.12352180480957, + "learning_rate": 9.9893029989303e-06, + "loss": 0.3571, + "step": 10703 + }, + { + "epoch": 4.976290097629009, + "grad_norm": 5.300191879272461, + "learning_rate": 9.9893019989302e-06, + "loss": 0.3682, + "step": 10704 + }, + { + "epoch": 4.9767549976755, + "grad_norm": 9.685931205749512, + "learning_rate": 9.9893009989301e-06, + "loss": 0.6054, + "step": 10705 + }, + { + "epoch": 4.97721989772199, + "grad_norm": 2.7388012409210205, + "learning_rate": 9.98929999893e-06, + "loss": 0.089, + "step": 10706 + }, + { + "epoch": 4.97768479776848, + "grad_norm": 4.301443576812744, + "learning_rate": 9.989298998929901e-06, + "loss": 0.1343, + "step": 10707 + }, + { + "epoch": 4.9781496978149695, + "grad_norm": 5.676952838897705, + "learning_rate": 9.989297998929802e-06, + "loss": 0.1536, + "step": 10708 + }, + { + "epoch": 4.97861459786146, + "grad_norm": 6.037975311279297, + "learning_rate": 9.9892969989297e-06, + "loss": 0.239, + "step": 10709 + }, + { + "epoch": 4.97907949790795, + "grad_norm": 4.153989791870117, + "learning_rate": 9.989295998929601e-06, + "loss": 0.1999, + "step": 10710 + }, + { + "epoch": 4.97954439795444, + "grad_norm": 8.262985229492188, + "learning_rate": 9.9892949989295e-06, + "loss": 0.6167, + "step": 10711 + }, + { + "epoch": 4.98000929800093, + "grad_norm": 6.708693981170654, + "learning_rate": 9.9892939989294e-06, + "loss": 0.4407, + "step": 10712 + }, + { + "epoch": 4.980474198047419, + "grad_norm": 5.573073387145996, + "learning_rate": 9.989292998929301e-06, + "loss": 0.2619, + "step": 10713 + }, + { + "epoch": 4.98093909809391, + "grad_norm": 6.916827201843262, + "learning_rate": 9.9892919989292e-06, + "loss": 0.3782, + "step": 10714 + }, + { + "epoch": 4.9814039981404, + "grad_norm": 9.44460391998291, + "learning_rate": 9.989290998929101e-06, + "loss": 0.2818, + "step": 10715 + }, + { + "epoch": 4.98186889818689, + "grad_norm": 6.758260250091553, + "learning_rate": 9.989289998929e-06, + "loss": 0.3696, + "step": 10716 + }, + { + "epoch": 4.9823337982333795, + "grad_norm": 5.506319046020508, + "learning_rate": 9.9892889989289e-06, + "loss": 0.2434, + "step": 10717 + }, + { + "epoch": 4.98279869827987, + "grad_norm": 6.735909938812256, + "learning_rate": 9.9892879989288e-06, + "loss": 0.3158, + "step": 10718 + }, + { + "epoch": 4.98326359832636, + "grad_norm": 5.464966773986816, + "learning_rate": 9.9892869989287e-06, + "loss": 0.3329, + "step": 10719 + }, + { + "epoch": 4.98372849837285, + "grad_norm": 6.641495704650879, + "learning_rate": 9.989285998928601e-06, + "loss": 0.2891, + "step": 10720 + }, + { + "epoch": 4.98419339841934, + "grad_norm": 4.149350166320801, + "learning_rate": 9.989284998928502e-06, + "loss": 0.1497, + "step": 10721 + }, + { + "epoch": 4.984658298465829, + "grad_norm": 5.930226802825928, + "learning_rate": 9.9892839989284e-06, + "loss": 0.3964, + "step": 10722 + }, + { + "epoch": 4.98512319851232, + "grad_norm": 4.408370018005371, + "learning_rate": 9.9892829989283e-06, + "loss": 0.2289, + "step": 10723 + }, + { + "epoch": 4.98558809855881, + "grad_norm": 5.287887096405029, + "learning_rate": 9.9892819989282e-06, + "loss": 0.2649, + "step": 10724 + }, + { + "epoch": 4.9860529986053, + "grad_norm": 6.638917922973633, + "learning_rate": 9.989280998928101e-06, + "loss": 0.3616, + "step": 10725 + }, + { + "epoch": 4.9865178986517895, + "grad_norm": 4.984868049621582, + "learning_rate": 9.989279998928002e-06, + "loss": 0.3466, + "step": 10726 + }, + { + "epoch": 4.98698279869828, + "grad_norm": 6.3288421630859375, + "learning_rate": 9.9892789989279e-06, + "loss": 0.4193, + "step": 10727 + }, + { + "epoch": 4.98744769874477, + "grad_norm": 5.31939172744751, + "learning_rate": 9.989277998927801e-06, + "loss": 0.2844, + "step": 10728 + }, + { + "epoch": 4.98791259879126, + "grad_norm": 6.296680927276611, + "learning_rate": 9.9892769989277e-06, + "loss": 0.3113, + "step": 10729 + }, + { + "epoch": 4.98837749883775, + "grad_norm": 8.697906494140625, + "learning_rate": 9.9892759989276e-06, + "loss": 0.3344, + "step": 10730 + }, + { + "epoch": 4.988842398884239, + "grad_norm": 4.601659774780273, + "learning_rate": 9.989274998927501e-06, + "loss": 0.2584, + "step": 10731 + }, + { + "epoch": 4.98930729893073, + "grad_norm": 9.228649139404297, + "learning_rate": 9.9892739989274e-06, + "loss": 0.3859, + "step": 10732 + }, + { + "epoch": 4.98977219897722, + "grad_norm": 6.51139497756958, + "learning_rate": 9.989272998927301e-06, + "loss": 0.3792, + "step": 10733 + }, + { + "epoch": 4.99023709902371, + "grad_norm": 15.657153129577637, + "learning_rate": 9.9892719989272e-06, + "loss": 0.6705, + "step": 10734 + }, + { + "epoch": 4.9907019990702, + "grad_norm": 8.29842758178711, + "learning_rate": 9.9892709989271e-06, + "loss": 0.6919, + "step": 10735 + }, + { + "epoch": 4.99116689911669, + "grad_norm": 6.695074558258057, + "learning_rate": 9.989269998927e-06, + "loss": 0.5526, + "step": 10736 + }, + { + "epoch": 4.99163179916318, + "grad_norm": 5.410065174102783, + "learning_rate": 9.9892689989269e-06, + "loss": 0.3106, + "step": 10737 + }, + { + "epoch": 4.99209669920967, + "grad_norm": 5.386198043823242, + "learning_rate": 9.989267998926801e-06, + "loss": 0.415, + "step": 10738 + }, + { + "epoch": 4.99256159925616, + "grad_norm": 4.29514217376709, + "learning_rate": 9.989266998926702e-06, + "loss": 0.194, + "step": 10739 + }, + { + "epoch": 4.99302649930265, + "grad_norm": 5.498699188232422, + "learning_rate": 9.9892659989266e-06, + "loss": 0.2784, + "step": 10740 + }, + { + "epoch": 4.99349139934914, + "grad_norm": 3.460218667984009, + "learning_rate": 9.989264998926501e-06, + "loss": 0.1795, + "step": 10741 + }, + { + "epoch": 4.99395629939563, + "grad_norm": 3.7269787788391113, + "learning_rate": 9.9892639989264e-06, + "loss": 0.1934, + "step": 10742 + }, + { + "epoch": 4.99442119944212, + "grad_norm": 5.97410249710083, + "learning_rate": 9.989262998926301e-06, + "loss": 0.3792, + "step": 10743 + }, + { + "epoch": 4.99488609948861, + "grad_norm": 10.607665061950684, + "learning_rate": 9.989261998926202e-06, + "loss": 0.3785, + "step": 10744 + }, + { + "epoch": 4.9953509995351, + "grad_norm": 5.124697208404541, + "learning_rate": 9.9892609989261e-06, + "loss": 0.182, + "step": 10745 + }, + { + "epoch": 4.99581589958159, + "grad_norm": 6.272425174713135, + "learning_rate": 9.989259998926001e-06, + "loss": 0.4142, + "step": 10746 + }, + { + "epoch": 4.99628079962808, + "grad_norm": 7.037789344787598, + "learning_rate": 9.9892589989259e-06, + "loss": 0.3537, + "step": 10747 + }, + { + "epoch": 4.99674569967457, + "grad_norm": 5.568120002746582, + "learning_rate": 9.9892579989258e-06, + "loss": 0.4383, + "step": 10748 + }, + { + "epoch": 4.99721059972106, + "grad_norm": 5.049176216125488, + "learning_rate": 9.9892569989257e-06, + "loss": 0.3142, + "step": 10749 + }, + { + "epoch": 4.99767549976755, + "grad_norm": 13.366802215576172, + "learning_rate": 9.9892559989256e-06, + "loss": 0.287, + "step": 10750 + }, + { + "epoch": 4.99814039981404, + "grad_norm": 9.79520320892334, + "learning_rate": 9.989254998925501e-06, + "loss": 0.1934, + "step": 10751 + }, + { + "epoch": 4.99860529986053, + "grad_norm": 5.138782978057861, + "learning_rate": 9.989253998925402e-06, + "loss": 0.2804, + "step": 10752 + }, + { + "epoch": 4.99907019990702, + "grad_norm": 4.838087558746338, + "learning_rate": 9.9892529989253e-06, + "loss": 0.2887, + "step": 10753 + }, + { + "epoch": 4.99953509995351, + "grad_norm": 8.788103103637695, + "learning_rate": 9.989251998925201e-06, + "loss": 0.6174, + "step": 10754 + }, + { + "epoch": 5.0, + "grad_norm": 4.841432094573975, + "learning_rate": 9.9892509989251e-06, + "loss": 0.2044, + "step": 10755 + }, + { + "epoch": 5.00046490004649, + "grad_norm": 6.191995620727539, + "learning_rate": 9.989249998925001e-06, + "loss": 0.365, + "step": 10756 + }, + { + "epoch": 5.00092980009298, + "grad_norm": 3.719809055328369, + "learning_rate": 9.989248998924902e-06, + "loss": 0.1931, + "step": 10757 + }, + { + "epoch": 5.00139470013947, + "grad_norm": 6.056956768035889, + "learning_rate": 9.9892479989248e-06, + "loss": 0.5285, + "step": 10758 + }, + { + "epoch": 5.00185960018596, + "grad_norm": 4.468196392059326, + "learning_rate": 9.989246998924701e-06, + "loss": 0.2712, + "step": 10759 + }, + { + "epoch": 5.00232450023245, + "grad_norm": 3.1195485591888428, + "learning_rate": 9.9892459989246e-06, + "loss": 0.1412, + "step": 10760 + }, + { + "epoch": 5.00278940027894, + "grad_norm": 7.420111179351807, + "learning_rate": 9.989244998924501e-06, + "loss": 0.1511, + "step": 10761 + }, + { + "epoch": 5.00325430032543, + "grad_norm": 5.29220724105835, + "learning_rate": 9.9892439989244e-06, + "loss": 0.2375, + "step": 10762 + }, + { + "epoch": 5.00371920037192, + "grad_norm": 3.5416905879974365, + "learning_rate": 9.9892429989243e-06, + "loss": 0.2001, + "step": 10763 + }, + { + "epoch": 5.00418410041841, + "grad_norm": 5.1379265785217285, + "learning_rate": 9.989241998924201e-06, + "loss": 0.2257, + "step": 10764 + }, + { + "epoch": 5.0046490004649, + "grad_norm": 4.846053600311279, + "learning_rate": 9.989240998924102e-06, + "loss": 0.2295, + "step": 10765 + }, + { + "epoch": 5.0051139005113905, + "grad_norm": 5.3436079025268555, + "learning_rate": 9.989239998924e-06, + "loss": 0.164, + "step": 10766 + }, + { + "epoch": 5.00557880055788, + "grad_norm": 9.09115219116211, + "learning_rate": 9.9892389989239e-06, + "loss": 0.4309, + "step": 10767 + }, + { + "epoch": 5.00604370060437, + "grad_norm": 6.017485618591309, + "learning_rate": 9.9892379989238e-06, + "loss": 0.1795, + "step": 10768 + }, + { + "epoch": 5.00650860065086, + "grad_norm": 4.162603378295898, + "learning_rate": 9.989236998923701e-06, + "loss": 0.1772, + "step": 10769 + }, + { + "epoch": 5.00697350069735, + "grad_norm": 1.5138994455337524, + "learning_rate": 9.989235998923602e-06, + "loss": 0.0348, + "step": 10770 + }, + { + "epoch": 5.00743840074384, + "grad_norm": 5.371227264404297, + "learning_rate": 9.9892349989235e-06, + "loss": 0.198, + "step": 10771 + }, + { + "epoch": 5.00790330079033, + "grad_norm": 6.011104106903076, + "learning_rate": 9.989233998923401e-06, + "loss": 0.189, + "step": 10772 + }, + { + "epoch": 5.00836820083682, + "grad_norm": 6.20469856262207, + "learning_rate": 9.9892329989233e-06, + "loss": 0.4383, + "step": 10773 + }, + { + "epoch": 5.00883310088331, + "grad_norm": 4.417482852935791, + "learning_rate": 9.989231998923201e-06, + "loss": 0.2078, + "step": 10774 + }, + { + "epoch": 5.0092980009298005, + "grad_norm": 3.130995512008667, + "learning_rate": 9.9892309989231e-06, + "loss": 0.1513, + "step": 10775 + }, + { + "epoch": 5.00976290097629, + "grad_norm": 3.7239911556243896, + "learning_rate": 9.989229998923e-06, + "loss": 0.1883, + "step": 10776 + }, + { + "epoch": 5.01022780102278, + "grad_norm": 3.6356921195983887, + "learning_rate": 9.989228998922901e-06, + "loss": 0.1544, + "step": 10777 + }, + { + "epoch": 5.01069270106927, + "grad_norm": 3.5350968837738037, + "learning_rate": 9.9892279989228e-06, + "loss": 0.1178, + "step": 10778 + }, + { + "epoch": 5.01115760111576, + "grad_norm": 4.422608852386475, + "learning_rate": 9.9892269989227e-06, + "loss": 0.1681, + "step": 10779 + }, + { + "epoch": 5.01162250116225, + "grad_norm": 3.505815029144287, + "learning_rate": 9.9892259989226e-06, + "loss": 0.106, + "step": 10780 + }, + { + "epoch": 5.01208740120874, + "grad_norm": 4.702446460723877, + "learning_rate": 9.9892249989225e-06, + "loss": 0.1998, + "step": 10781 + }, + { + "epoch": 5.01255230125523, + "grad_norm": 6.0295538902282715, + "learning_rate": 9.989223998922401e-06, + "loss": 0.2025, + "step": 10782 + }, + { + "epoch": 5.01301720130172, + "grad_norm": 5.394504547119141, + "learning_rate": 9.989222998922302e-06, + "loss": 0.4127, + "step": 10783 + }, + { + "epoch": 5.0134821013482105, + "grad_norm": 6.307869911193848, + "learning_rate": 9.9892219989222e-06, + "loss": 0.322, + "step": 10784 + }, + { + "epoch": 5.0139470013947, + "grad_norm": 5.426379680633545, + "learning_rate": 9.989220998922101e-06, + "loss": 0.3135, + "step": 10785 + }, + { + "epoch": 5.01441190144119, + "grad_norm": 5.508844375610352, + "learning_rate": 9.989219998922e-06, + "loss": 0.1798, + "step": 10786 + }, + { + "epoch": 5.01487680148768, + "grad_norm": 6.824454307556152, + "learning_rate": 9.9892189989219e-06, + "loss": 0.3654, + "step": 10787 + }, + { + "epoch": 5.015341701534171, + "grad_norm": 2.6391854286193848, + "learning_rate": 9.9892179989218e-06, + "loss": 0.07, + "step": 10788 + }, + { + "epoch": 5.01580660158066, + "grad_norm": 9.82484245300293, + "learning_rate": 9.9892169989217e-06, + "loss": 0.2669, + "step": 10789 + }, + { + "epoch": 5.01627150162715, + "grad_norm": 3.79213547706604, + "learning_rate": 9.989215998921601e-06, + "loss": 0.1013, + "step": 10790 + }, + { + "epoch": 5.01673640167364, + "grad_norm": 4.499590873718262, + "learning_rate": 9.9892149989215e-06, + "loss": 0.2374, + "step": 10791 + }, + { + "epoch": 5.01720130172013, + "grad_norm": 4.819867134094238, + "learning_rate": 9.989213998921401e-06, + "loss": 0.1943, + "step": 10792 + }, + { + "epoch": 5.0176662017666205, + "grad_norm": 5.1394500732421875, + "learning_rate": 9.9892129989213e-06, + "loss": 0.1555, + "step": 10793 + }, + { + "epoch": 5.01813110181311, + "grad_norm": 4.894166469573975, + "learning_rate": 9.9892119989212e-06, + "loss": 0.2197, + "step": 10794 + }, + { + "epoch": 5.0185960018596, + "grad_norm": 4.111968517303467, + "learning_rate": 9.989210998921101e-06, + "loss": 0.1508, + "step": 10795 + }, + { + "epoch": 5.01906090190609, + "grad_norm": 8.019947052001953, + "learning_rate": 9.989209998921002e-06, + "loss": 0.3204, + "step": 10796 + }, + { + "epoch": 5.019525801952581, + "grad_norm": 4.295135974884033, + "learning_rate": 9.9892089989209e-06, + "loss": 0.1901, + "step": 10797 + }, + { + "epoch": 5.01999070199907, + "grad_norm": 7.0832366943359375, + "learning_rate": 9.9892079989208e-06, + "loss": 0.2225, + "step": 10798 + }, + { + "epoch": 5.02045560204556, + "grad_norm": 5.400976657867432, + "learning_rate": 9.9892069989207e-06, + "loss": 0.4034, + "step": 10799 + }, + { + "epoch": 5.02092050209205, + "grad_norm": 3.5336716175079346, + "learning_rate": 9.989205998920601e-06, + "loss": 0.17, + "step": 10800 + }, + { + "epoch": 5.02138540213854, + "grad_norm": 7.374716758728027, + "learning_rate": 9.9892049989205e-06, + "loss": 0.3561, + "step": 10801 + }, + { + "epoch": 5.0218503021850305, + "grad_norm": 4.83047342300415, + "learning_rate": 9.9892039989204e-06, + "loss": 0.2876, + "step": 10802 + }, + { + "epoch": 5.02231520223152, + "grad_norm": 4.439749717712402, + "learning_rate": 9.989202998920301e-06, + "loss": 0.2819, + "step": 10803 + }, + { + "epoch": 5.02278010227801, + "grad_norm": 3.3789167404174805, + "learning_rate": 9.9892019989202e-06, + "loss": 0.179, + "step": 10804 + }, + { + "epoch": 5.0232450023245, + "grad_norm": 2.6123642921447754, + "learning_rate": 9.989200998920101e-06, + "loss": 0.0932, + "step": 10805 + }, + { + "epoch": 5.023709902370991, + "grad_norm": 5.610192775726318, + "learning_rate": 9.98919999892e-06, + "loss": 0.2706, + "step": 10806 + }, + { + "epoch": 5.02417480241748, + "grad_norm": 5.452523708343506, + "learning_rate": 9.9891989989199e-06, + "loss": 0.2257, + "step": 10807 + }, + { + "epoch": 5.02463970246397, + "grad_norm": 4.079238414764404, + "learning_rate": 9.989197998919801e-06, + "loss": 0.1783, + "step": 10808 + }, + { + "epoch": 5.02510460251046, + "grad_norm": 3.8127338886260986, + "learning_rate": 9.9891969989197e-06, + "loss": 0.1511, + "step": 10809 + }, + { + "epoch": 5.025569502556951, + "grad_norm": 4.766107559204102, + "learning_rate": 9.989195998919601e-06, + "loss": 0.1837, + "step": 10810 + }, + { + "epoch": 5.0260344026034405, + "grad_norm": 3.6954829692840576, + "learning_rate": 9.9891949989195e-06, + "loss": 0.0725, + "step": 10811 + }, + { + "epoch": 5.02649930264993, + "grad_norm": 6.456759929656982, + "learning_rate": 9.9891939989194e-06, + "loss": 0.2429, + "step": 10812 + }, + { + "epoch": 5.02696420269642, + "grad_norm": 3.261026620864868, + "learning_rate": 9.989192998919301e-06, + "loss": 0.1369, + "step": 10813 + }, + { + "epoch": 5.02742910274291, + "grad_norm": 6.791769981384277, + "learning_rate": 9.989191998919202e-06, + "loss": 0.1753, + "step": 10814 + }, + { + "epoch": 5.027894002789401, + "grad_norm": 6.487844944000244, + "learning_rate": 9.9891909989191e-06, + "loss": 0.4208, + "step": 10815 + }, + { + "epoch": 5.02835890283589, + "grad_norm": 4.28952169418335, + "learning_rate": 9.989189998919001e-06, + "loss": 0.1751, + "step": 10816 + }, + { + "epoch": 5.02882380288238, + "grad_norm": 5.520526885986328, + "learning_rate": 9.9891889989189e-06, + "loss": 0.2431, + "step": 10817 + }, + { + "epoch": 5.02928870292887, + "grad_norm": 3.1144895553588867, + "learning_rate": 9.989187998918801e-06, + "loss": 0.1287, + "step": 10818 + }, + { + "epoch": 5.029753602975361, + "grad_norm": 7.180879592895508, + "learning_rate": 9.9891869989187e-06, + "loss": 0.3298, + "step": 10819 + }, + { + "epoch": 5.0302185030218505, + "grad_norm": 4.610947132110596, + "learning_rate": 9.9891859989186e-06, + "loss": 0.1291, + "step": 10820 + }, + { + "epoch": 5.03068340306834, + "grad_norm": 8.148880004882812, + "learning_rate": 9.989184998918501e-06, + "loss": 0.3543, + "step": 10821 + }, + { + "epoch": 5.03114830311483, + "grad_norm": 5.425479412078857, + "learning_rate": 9.9891839989184e-06, + "loss": 0.2294, + "step": 10822 + }, + { + "epoch": 5.03161320316132, + "grad_norm": 5.168003082275391, + "learning_rate": 9.989182998918301e-06, + "loss": 0.1994, + "step": 10823 + }, + { + "epoch": 5.032078103207811, + "grad_norm": 4.458009719848633, + "learning_rate": 9.9891819989182e-06, + "loss": 0.1053, + "step": 10824 + }, + { + "epoch": 5.0325430032543, + "grad_norm": 4.328068256378174, + "learning_rate": 9.9891809989181e-06, + "loss": 0.221, + "step": 10825 + }, + { + "epoch": 5.03300790330079, + "grad_norm": 4.506632328033447, + "learning_rate": 9.989179998918001e-06, + "loss": 0.1641, + "step": 10826 + }, + { + "epoch": 5.03347280334728, + "grad_norm": 6.582061290740967, + "learning_rate": 9.989178998917902e-06, + "loss": 0.2119, + "step": 10827 + }, + { + "epoch": 5.033937703393771, + "grad_norm": 3.688702344894409, + "learning_rate": 9.989177998917801e-06, + "loss": 0.1443, + "step": 10828 + }, + { + "epoch": 5.0344026034402605, + "grad_norm": 3.386411190032959, + "learning_rate": 9.989176998917702e-06, + "loss": 0.1069, + "step": 10829 + }, + { + "epoch": 5.03486750348675, + "grad_norm": 3.3831751346588135, + "learning_rate": 9.9891759989176e-06, + "loss": 0.153, + "step": 10830 + }, + { + "epoch": 5.03533240353324, + "grad_norm": 5.545093059539795, + "learning_rate": 9.9891749989175e-06, + "loss": 0.1992, + "step": 10831 + }, + { + "epoch": 5.035797303579731, + "grad_norm": 5.330069541931152, + "learning_rate": 9.9891739989174e-06, + "loss": 0.1231, + "step": 10832 + }, + { + "epoch": 5.036262203626221, + "grad_norm": 7.277401447296143, + "learning_rate": 9.9891729989173e-06, + "loss": 0.2374, + "step": 10833 + }, + { + "epoch": 5.03672710367271, + "grad_norm": 4.9978437423706055, + "learning_rate": 9.989171998917201e-06, + "loss": 0.1456, + "step": 10834 + }, + { + "epoch": 5.0371920037192, + "grad_norm": 4.716688632965088, + "learning_rate": 9.9891709989171e-06, + "loss": 0.1485, + "step": 10835 + }, + { + "epoch": 5.03765690376569, + "grad_norm": 5.6527485847473145, + "learning_rate": 9.989169998917001e-06, + "loss": 0.1921, + "step": 10836 + }, + { + "epoch": 5.038121803812181, + "grad_norm": 5.381747722625732, + "learning_rate": 9.9891689989169e-06, + "loss": 0.2036, + "step": 10837 + }, + { + "epoch": 5.0385867038586705, + "grad_norm": 5.6524505615234375, + "learning_rate": 9.9891679989168e-06, + "loss": 0.221, + "step": 10838 + }, + { + "epoch": 5.03905160390516, + "grad_norm": 6.127008438110352, + "learning_rate": 9.989166998916701e-06, + "loss": 0.1796, + "step": 10839 + }, + { + "epoch": 5.03951650395165, + "grad_norm": 6.558064937591553, + "learning_rate": 9.989165998916602e-06, + "loss": 0.3242, + "step": 10840 + }, + { + "epoch": 5.039981403998141, + "grad_norm": 4.7265706062316895, + "learning_rate": 9.989164998916501e-06, + "loss": 0.2027, + "step": 10841 + }, + { + "epoch": 5.040446304044631, + "grad_norm": 5.1701836585998535, + "learning_rate": 9.9891639989164e-06, + "loss": 0.2656, + "step": 10842 + }, + { + "epoch": 5.04091120409112, + "grad_norm": 6.164311408996582, + "learning_rate": 9.9891629989163e-06, + "loss": 0.2839, + "step": 10843 + }, + { + "epoch": 5.04137610413761, + "grad_norm": 4.362523078918457, + "learning_rate": 9.9891619989162e-06, + "loss": 0.1244, + "step": 10844 + }, + { + "epoch": 5.0418410041841, + "grad_norm": 6.105111598968506, + "learning_rate": 9.9891609989161e-06, + "loss": 0.2866, + "step": 10845 + }, + { + "epoch": 5.042305904230591, + "grad_norm": 7.379046440124512, + "learning_rate": 9.989159998916001e-06, + "loss": 0.3516, + "step": 10846 + }, + { + "epoch": 5.0427708042770805, + "grad_norm": 2.5258991718292236, + "learning_rate": 9.989158998915902e-06, + "loss": 0.0565, + "step": 10847 + }, + { + "epoch": 5.04323570432357, + "grad_norm": 4.6019415855407715, + "learning_rate": 9.9891579989158e-06, + "loss": 0.3104, + "step": 10848 + }, + { + "epoch": 5.04370060437006, + "grad_norm": 4.560312271118164, + "learning_rate": 9.989156998915701e-06, + "loss": 0.1494, + "step": 10849 + }, + { + "epoch": 5.044165504416551, + "grad_norm": 7.6189093589782715, + "learning_rate": 9.9891559989156e-06, + "loss": 0.1385, + "step": 10850 + }, + { + "epoch": 5.044630404463041, + "grad_norm": 3.636852502822876, + "learning_rate": 9.9891549989155e-06, + "loss": 0.1284, + "step": 10851 + }, + { + "epoch": 5.04509530450953, + "grad_norm": 4.393270492553711, + "learning_rate": 9.989153998915401e-06, + "loss": 0.2172, + "step": 10852 + }, + { + "epoch": 5.04556020455602, + "grad_norm": 4.973272800445557, + "learning_rate": 9.9891529989153e-06, + "loss": 0.2744, + "step": 10853 + }, + { + "epoch": 5.046025104602511, + "grad_norm": 6.579732418060303, + "learning_rate": 9.989151998915201e-06, + "loss": 0.4849, + "step": 10854 + }, + { + "epoch": 5.046490004649001, + "grad_norm": 6.006589889526367, + "learning_rate": 9.9891509989151e-06, + "loss": 0.2498, + "step": 10855 + }, + { + "epoch": 5.0469549046954905, + "grad_norm": 5.55825662612915, + "learning_rate": 9.989149998915e-06, + "loss": 0.3311, + "step": 10856 + }, + { + "epoch": 5.04741980474198, + "grad_norm": 3.209650993347168, + "learning_rate": 9.9891489989149e-06, + "loss": 0.1453, + "step": 10857 + }, + { + "epoch": 5.04788470478847, + "grad_norm": 4.656797885894775, + "learning_rate": 9.9891479989148e-06, + "loss": 0.1383, + "step": 10858 + }, + { + "epoch": 5.048349604834961, + "grad_norm": 6.109928607940674, + "learning_rate": 9.989146998914701e-06, + "loss": 0.3506, + "step": 10859 + }, + { + "epoch": 5.048814504881451, + "grad_norm": 3.1676981449127197, + "learning_rate": 9.989145998914602e-06, + "loss": 0.0938, + "step": 10860 + }, + { + "epoch": 5.04927940492794, + "grad_norm": 4.897518157958984, + "learning_rate": 9.9891449989145e-06, + "loss": 0.2404, + "step": 10861 + }, + { + "epoch": 5.04974430497443, + "grad_norm": 2.091301202774048, + "learning_rate": 9.989143998914401e-06, + "loss": 0.0503, + "step": 10862 + }, + { + "epoch": 5.050209205020921, + "grad_norm": 5.7208781242370605, + "learning_rate": 9.9891429989143e-06, + "loss": 0.2772, + "step": 10863 + }, + { + "epoch": 5.050674105067411, + "grad_norm": 11.967724800109863, + "learning_rate": 9.9891419989142e-06, + "loss": 0.345, + "step": 10864 + }, + { + "epoch": 5.0511390051139005, + "grad_norm": 5.881536960601807, + "learning_rate": 9.989140998914102e-06, + "loss": 0.2431, + "step": 10865 + }, + { + "epoch": 5.05160390516039, + "grad_norm": 3.8786187171936035, + "learning_rate": 9.989139998914e-06, + "loss": 0.0918, + "step": 10866 + }, + { + "epoch": 5.052068805206881, + "grad_norm": 6.050811290740967, + "learning_rate": 9.989138998913901e-06, + "loss": 0.3884, + "step": 10867 + }, + { + "epoch": 5.052533705253371, + "grad_norm": 4.303178787231445, + "learning_rate": 9.9891379989138e-06, + "loss": 0.1724, + "step": 10868 + }, + { + "epoch": 5.052998605299861, + "grad_norm": 10.955704689025879, + "learning_rate": 9.9891369989137e-06, + "loss": 0.2454, + "step": 10869 + }, + { + "epoch": 5.05346350534635, + "grad_norm": 6.635936260223389, + "learning_rate": 9.9891359989136e-06, + "loss": 0.6696, + "step": 10870 + }, + { + "epoch": 5.05392840539284, + "grad_norm": 3.3349061012268066, + "learning_rate": 9.9891349989135e-06, + "loss": 0.0683, + "step": 10871 + }, + { + "epoch": 5.054393305439331, + "grad_norm": 5.4715189933776855, + "learning_rate": 9.989133998913401e-06, + "loss": 0.2618, + "step": 10872 + }, + { + "epoch": 5.054858205485821, + "grad_norm": 4.6213483810424805, + "learning_rate": 9.989132998913302e-06, + "loss": 0.1702, + "step": 10873 + }, + { + "epoch": 5.0553231055323105, + "grad_norm": 4.720368385314941, + "learning_rate": 9.9891319989132e-06, + "loss": 0.1401, + "step": 10874 + }, + { + "epoch": 5.0557880055788, + "grad_norm": 3.1633574962615967, + "learning_rate": 9.9891309989131e-06, + "loss": 0.1111, + "step": 10875 + }, + { + "epoch": 5.056252905625291, + "grad_norm": 3.9524219036102295, + "learning_rate": 9.989129998913e-06, + "loss": 0.2462, + "step": 10876 + }, + { + "epoch": 5.056717805671781, + "grad_norm": 4.736701011657715, + "learning_rate": 9.989128998912901e-06, + "loss": 0.252, + "step": 10877 + }, + { + "epoch": 5.0571827057182706, + "grad_norm": 6.495545387268066, + "learning_rate": 9.989127998912802e-06, + "loss": 0.2698, + "step": 10878 + }, + { + "epoch": 5.05764760576476, + "grad_norm": 3.1815412044525146, + "learning_rate": 9.9891269989127e-06, + "loss": 0.058, + "step": 10879 + }, + { + "epoch": 5.05811250581125, + "grad_norm": 9.31912612915039, + "learning_rate": 9.989125998912601e-06, + "loss": 0.1839, + "step": 10880 + }, + { + "epoch": 5.058577405857741, + "grad_norm": 5.120662212371826, + "learning_rate": 9.9891249989125e-06, + "loss": 0.1996, + "step": 10881 + }, + { + "epoch": 5.059042305904231, + "grad_norm": 5.287334442138672, + "learning_rate": 9.9891239989124e-06, + "loss": 0.2126, + "step": 10882 + }, + { + "epoch": 5.0595072059507205, + "grad_norm": 5.2754034996032715, + "learning_rate": 9.989122998912301e-06, + "loss": 0.3476, + "step": 10883 + }, + { + "epoch": 5.05997210599721, + "grad_norm": 6.115639686584473, + "learning_rate": 9.9891219989122e-06, + "loss": 0.2648, + "step": 10884 + }, + { + "epoch": 5.060437006043701, + "grad_norm": 2.6118292808532715, + "learning_rate": 9.989120998912101e-06, + "loss": 0.0964, + "step": 10885 + }, + { + "epoch": 5.060901906090191, + "grad_norm": 5.11236047744751, + "learning_rate": 9.989119998912e-06, + "loss": 0.4147, + "step": 10886 + }, + { + "epoch": 5.0613668061366806, + "grad_norm": 5.554403305053711, + "learning_rate": 9.9891189989119e-06, + "loss": 0.3575, + "step": 10887 + }, + { + "epoch": 5.06183170618317, + "grad_norm": 5.524445056915283, + "learning_rate": 9.9891179989118e-06, + "loss": 0.203, + "step": 10888 + }, + { + "epoch": 5.062296606229661, + "grad_norm": 2.806211233139038, + "learning_rate": 9.9891169989117e-06, + "loss": 0.078, + "step": 10889 + }, + { + "epoch": 5.062761506276151, + "grad_norm": 5.5189127922058105, + "learning_rate": 9.989115998911601e-06, + "loss": 0.3288, + "step": 10890 + }, + { + "epoch": 5.063226406322641, + "grad_norm": 5.635857105255127, + "learning_rate": 9.989114998911502e-06, + "loss": 0.2076, + "step": 10891 + }, + { + "epoch": 5.0636913063691305, + "grad_norm": 8.123834609985352, + "learning_rate": 9.9891139989114e-06, + "loss": 0.2639, + "step": 10892 + }, + { + "epoch": 5.06415620641562, + "grad_norm": 5.85235595703125, + "learning_rate": 9.989112998911301e-06, + "loss": 0.3312, + "step": 10893 + }, + { + "epoch": 5.064621106462111, + "grad_norm": 5.563394069671631, + "learning_rate": 9.9891119989112e-06, + "loss": 0.1603, + "step": 10894 + }, + { + "epoch": 5.065086006508601, + "grad_norm": 4.148224353790283, + "learning_rate": 9.989110998911101e-06, + "loss": 0.1721, + "step": 10895 + }, + { + "epoch": 5.0655509065550905, + "grad_norm": 4.421968460083008, + "learning_rate": 9.989109998911002e-06, + "loss": 0.1961, + "step": 10896 + }, + { + "epoch": 5.06601580660158, + "grad_norm": 7.981452941894531, + "learning_rate": 9.9891089989109e-06, + "loss": 0.3098, + "step": 10897 + }, + { + "epoch": 5.066480706648071, + "grad_norm": 4.132032871246338, + "learning_rate": 9.989107998910801e-06, + "loss": 0.192, + "step": 10898 + }, + { + "epoch": 5.066945606694561, + "grad_norm": 2.3669726848602295, + "learning_rate": 9.9891069989107e-06, + "loss": 0.0703, + "step": 10899 + }, + { + "epoch": 5.067410506741051, + "grad_norm": 6.738861083984375, + "learning_rate": 9.9891059989106e-06, + "loss": 0.2412, + "step": 10900 + }, + { + "epoch": 5.0678754067875404, + "grad_norm": 2.6452889442443848, + "learning_rate": 9.9891049989105e-06, + "loss": 0.0554, + "step": 10901 + }, + { + "epoch": 5.06834030683403, + "grad_norm": 4.5971550941467285, + "learning_rate": 9.9891039989104e-06, + "loss": 0.1723, + "step": 10902 + }, + { + "epoch": 5.068805206880521, + "grad_norm": 4.206604480743408, + "learning_rate": 9.989102998910301e-06, + "loss": 0.1591, + "step": 10903 + }, + { + "epoch": 5.069270106927011, + "grad_norm": 5.619023323059082, + "learning_rate": 9.989101998910202e-06, + "loss": 0.278, + "step": 10904 + }, + { + "epoch": 5.0697350069735005, + "grad_norm": 6.027493000030518, + "learning_rate": 9.9891009989101e-06, + "loss": 0.2744, + "step": 10905 + }, + { + "epoch": 5.07019990701999, + "grad_norm": 5.690732479095459, + "learning_rate": 9.989099998910001e-06, + "loss": 0.2659, + "step": 10906 + }, + { + "epoch": 5.070664807066481, + "grad_norm": 4.685452938079834, + "learning_rate": 9.9890989989099e-06, + "loss": 0.1827, + "step": 10907 + }, + { + "epoch": 5.071129707112971, + "grad_norm": 14.76069450378418, + "learning_rate": 9.989097998909801e-06, + "loss": 0.2252, + "step": 10908 + }, + { + "epoch": 5.071594607159461, + "grad_norm": 5.826590538024902, + "learning_rate": 9.989096998909702e-06, + "loss": 0.1686, + "step": 10909 + }, + { + "epoch": 5.0720595072059504, + "grad_norm": 4.397263526916504, + "learning_rate": 9.9890959989096e-06, + "loss": 0.091, + "step": 10910 + }, + { + "epoch": 5.072524407252441, + "grad_norm": 5.414210796356201, + "learning_rate": 9.989094998909501e-06, + "loss": 0.2401, + "step": 10911 + }, + { + "epoch": 5.072989307298931, + "grad_norm": 4.107457160949707, + "learning_rate": 9.9890939989094e-06, + "loss": 0.1209, + "step": 10912 + }, + { + "epoch": 5.073454207345421, + "grad_norm": 6.735073089599609, + "learning_rate": 9.989092998909301e-06, + "loss": 0.198, + "step": 10913 + }, + { + "epoch": 5.0739191073919105, + "grad_norm": 4.656951427459717, + "learning_rate": 9.9890919989092e-06, + "loss": 0.1871, + "step": 10914 + }, + { + "epoch": 5.0743840074384, + "grad_norm": 7.073060989379883, + "learning_rate": 9.9890909989091e-06, + "loss": 0.3785, + "step": 10915 + }, + { + "epoch": 5.074848907484891, + "grad_norm": 6.220086097717285, + "learning_rate": 9.989089998909001e-06, + "loss": 0.2321, + "step": 10916 + }, + { + "epoch": 5.075313807531381, + "grad_norm": 5.690938949584961, + "learning_rate": 9.9890889989089e-06, + "loss": 0.2532, + "step": 10917 + }, + { + "epoch": 5.075778707577871, + "grad_norm": 4.542809009552002, + "learning_rate": 9.9890879989088e-06, + "loss": 0.1913, + "step": 10918 + }, + { + "epoch": 5.07624360762436, + "grad_norm": 5.485464572906494, + "learning_rate": 9.9890869989087e-06, + "loss": 0.3155, + "step": 10919 + }, + { + "epoch": 5.076708507670851, + "grad_norm": 4.788780689239502, + "learning_rate": 9.9890859989086e-06, + "loss": 0.2155, + "step": 10920 + }, + { + "epoch": 5.077173407717341, + "grad_norm": 3.506711006164551, + "learning_rate": 9.989084998908501e-06, + "loss": 0.1376, + "step": 10921 + }, + { + "epoch": 5.077638307763831, + "grad_norm": 5.69832181930542, + "learning_rate": 9.989083998908402e-06, + "loss": 0.3782, + "step": 10922 + }, + { + "epoch": 5.0781032078103205, + "grad_norm": 4.565099239349365, + "learning_rate": 9.9890829989083e-06, + "loss": 0.2144, + "step": 10923 + }, + { + "epoch": 5.07856810785681, + "grad_norm": 6.250098705291748, + "learning_rate": 9.989081998908201e-06, + "loss": 0.3069, + "step": 10924 + }, + { + "epoch": 5.079033007903301, + "grad_norm": 6.198171138763428, + "learning_rate": 9.9890809989081e-06, + "loss": 0.1212, + "step": 10925 + }, + { + "epoch": 5.079497907949791, + "grad_norm": 9.311027526855469, + "learning_rate": 9.989079998908001e-06, + "loss": 0.5597, + "step": 10926 + }, + { + "epoch": 5.079962807996281, + "grad_norm": 6.082521915435791, + "learning_rate": 9.9890789989079e-06, + "loss": 0.2817, + "step": 10927 + }, + { + "epoch": 5.08042770804277, + "grad_norm": 3.6609768867492676, + "learning_rate": 9.9890779989078e-06, + "loss": 0.1722, + "step": 10928 + }, + { + "epoch": 5.080892608089261, + "grad_norm": 3.742036819458008, + "learning_rate": 9.989076998907701e-06, + "loss": 0.168, + "step": 10929 + }, + { + "epoch": 5.081357508135751, + "grad_norm": 3.893270254135132, + "learning_rate": 9.9890759989076e-06, + "loss": 0.1091, + "step": 10930 + }, + { + "epoch": 5.081822408182241, + "grad_norm": 4.7300124168396, + "learning_rate": 9.989074998907501e-06, + "loss": 0.3463, + "step": 10931 + }, + { + "epoch": 5.0822873082287305, + "grad_norm": 5.069123268127441, + "learning_rate": 9.9890739989074e-06, + "loss": 0.2491, + "step": 10932 + }, + { + "epoch": 5.082752208275221, + "grad_norm": 8.214990615844727, + "learning_rate": 9.9890729989073e-06, + "loss": 0.3135, + "step": 10933 + }, + { + "epoch": 5.083217108321711, + "grad_norm": 3.597642421722412, + "learning_rate": 9.989071998907201e-06, + "loss": 0.1687, + "step": 10934 + }, + { + "epoch": 5.083682008368201, + "grad_norm": 4.455666542053223, + "learning_rate": 9.989070998907102e-06, + "loss": 0.2079, + "step": 10935 + }, + { + "epoch": 5.084146908414691, + "grad_norm": 2.3220014572143555, + "learning_rate": 9.989069998907e-06, + "loss": 0.0618, + "step": 10936 + }, + { + "epoch": 5.08461180846118, + "grad_norm": 4.500897407531738, + "learning_rate": 9.989068998906901e-06, + "loss": 0.2228, + "step": 10937 + }, + { + "epoch": 5.085076708507671, + "grad_norm": 7.462706565856934, + "learning_rate": 9.9890679989068e-06, + "loss": 0.3419, + "step": 10938 + }, + { + "epoch": 5.085541608554161, + "grad_norm": 4.9777984619140625, + "learning_rate": 9.9890669989067e-06, + "loss": 0.2239, + "step": 10939 + }, + { + "epoch": 5.086006508600651, + "grad_norm": 4.486581325531006, + "learning_rate": 9.9890659989066e-06, + "loss": 0.248, + "step": 10940 + }, + { + "epoch": 5.0864714086471405, + "grad_norm": 3.0501787662506104, + "learning_rate": 9.9890649989065e-06, + "loss": 0.1528, + "step": 10941 + }, + { + "epoch": 5.086936308693631, + "grad_norm": 9.551643371582031, + "learning_rate": 9.989063998906401e-06, + "loss": 0.2131, + "step": 10942 + }, + { + "epoch": 5.087401208740121, + "grad_norm": 4.419058799743652, + "learning_rate": 9.9890629989063e-06, + "loss": 0.1556, + "step": 10943 + }, + { + "epoch": 5.087866108786611, + "grad_norm": 6.20059871673584, + "learning_rate": 9.989061998906201e-06, + "loss": 0.4201, + "step": 10944 + }, + { + "epoch": 5.088331008833101, + "grad_norm": 6.093903541564941, + "learning_rate": 9.9890609989061e-06, + "loss": 0.2251, + "step": 10945 + }, + { + "epoch": 5.088795908879591, + "grad_norm": 6.80393648147583, + "learning_rate": 9.989059998906e-06, + "loss": 0.2847, + "step": 10946 + }, + { + "epoch": 5.089260808926081, + "grad_norm": 4.359868049621582, + "learning_rate": 9.989058998905901e-06, + "loss": 0.1717, + "step": 10947 + }, + { + "epoch": 5.089725708972571, + "grad_norm": 3.74621844291687, + "learning_rate": 9.989057998905802e-06, + "loss": 0.0843, + "step": 10948 + }, + { + "epoch": 5.090190609019061, + "grad_norm": 5.931319236755371, + "learning_rate": 9.989056998905701e-06, + "loss": 0.478, + "step": 10949 + }, + { + "epoch": 5.0906555090655505, + "grad_norm": 5.279189586639404, + "learning_rate": 9.9890559989056e-06, + "loss": 0.2519, + "step": 10950 + }, + { + "epoch": 5.091120409112041, + "grad_norm": 5.301680564880371, + "learning_rate": 9.9890549989055e-06, + "loss": 0.2817, + "step": 10951 + }, + { + "epoch": 5.091585309158531, + "grad_norm": 5.018486022949219, + "learning_rate": 9.9890539989054e-06, + "loss": 0.2113, + "step": 10952 + }, + { + "epoch": 5.092050209205021, + "grad_norm": 5.352835178375244, + "learning_rate": 9.9890529989053e-06, + "loss": 0.2663, + "step": 10953 + }, + { + "epoch": 5.092515109251511, + "grad_norm": 6.343761920928955, + "learning_rate": 9.9890519989052e-06, + "loss": 0.1672, + "step": 10954 + }, + { + "epoch": 5.092980009298001, + "grad_norm": 5.411853313446045, + "learning_rate": 9.989050998905101e-06, + "loss": 0.2766, + "step": 10955 + }, + { + "epoch": 5.093444909344491, + "grad_norm": 6.411820411682129, + "learning_rate": 9.989049998905e-06, + "loss": 0.2644, + "step": 10956 + }, + { + "epoch": 5.093909809390981, + "grad_norm": 6.1422810554504395, + "learning_rate": 9.989048998904901e-06, + "loss": 0.17, + "step": 10957 + }, + { + "epoch": 5.094374709437471, + "grad_norm": 6.158684730529785, + "learning_rate": 9.9890479989048e-06, + "loss": 0.3061, + "step": 10958 + }, + { + "epoch": 5.0948396094839605, + "grad_norm": 9.30392074584961, + "learning_rate": 9.9890469989047e-06, + "loss": 0.2074, + "step": 10959 + }, + { + "epoch": 5.095304509530451, + "grad_norm": 4.135946273803711, + "learning_rate": 9.989045998904601e-06, + "loss": 0.1095, + "step": 10960 + }, + { + "epoch": 5.095769409576941, + "grad_norm": 5.662919044494629, + "learning_rate": 9.9890449989045e-06, + "loss": 0.277, + "step": 10961 + }, + { + "epoch": 5.096234309623431, + "grad_norm": 2.6449642181396484, + "learning_rate": 9.989043998904401e-06, + "loss": 0.0535, + "step": 10962 + }, + { + "epoch": 5.096699209669921, + "grad_norm": 5.776620864868164, + "learning_rate": 9.9890429989043e-06, + "loss": 0.2404, + "step": 10963 + }, + { + "epoch": 5.097164109716411, + "grad_norm": 5.340521335601807, + "learning_rate": 9.9890419989042e-06, + "loss": 0.1521, + "step": 10964 + }, + { + "epoch": 5.097629009762901, + "grad_norm": 4.12189245223999, + "learning_rate": 9.989040998904101e-06, + "loss": 0.2268, + "step": 10965 + }, + { + "epoch": 5.098093909809391, + "grad_norm": 4.649481296539307, + "learning_rate": 9.989039998904e-06, + "loss": 0.2091, + "step": 10966 + }, + { + "epoch": 5.098558809855881, + "grad_norm": 6.480130195617676, + "learning_rate": 9.9890389989039e-06, + "loss": 0.4889, + "step": 10967 + }, + { + "epoch": 5.099023709902371, + "grad_norm": 4.798299789428711, + "learning_rate": 9.989037998903801e-06, + "loss": 0.2531, + "step": 10968 + }, + { + "epoch": 5.099488609948861, + "grad_norm": 5.682989597320557, + "learning_rate": 9.9890369989037e-06, + "loss": 0.3453, + "step": 10969 + }, + { + "epoch": 5.099953509995351, + "grad_norm": 3.2616679668426514, + "learning_rate": 9.989035998903601e-06, + "loss": 0.0715, + "step": 10970 + }, + { + "epoch": 5.100418410041841, + "grad_norm": 4.276516437530518, + "learning_rate": 9.9890349989035e-06, + "loss": 0.2039, + "step": 10971 + }, + { + "epoch": 5.100883310088331, + "grad_norm": 4.691619873046875, + "learning_rate": 9.9890339989034e-06, + "loss": 0.1708, + "step": 10972 + }, + { + "epoch": 5.101348210134821, + "grad_norm": 4.6381683349609375, + "learning_rate": 9.989032998903301e-06, + "loss": 0.1798, + "step": 10973 + }, + { + "epoch": 5.101813110181311, + "grad_norm": 5.212643146514893, + "learning_rate": 9.9890319989032e-06, + "loss": 0.2507, + "step": 10974 + }, + { + "epoch": 5.102278010227801, + "grad_norm": 5.975442886352539, + "learning_rate": 9.989030998903101e-06, + "loss": 0.3574, + "step": 10975 + }, + { + "epoch": 5.102742910274291, + "grad_norm": 5.433172702789307, + "learning_rate": 9.989029998903e-06, + "loss": 0.2904, + "step": 10976 + }, + { + "epoch": 5.103207810320781, + "grad_norm": 6.165915012359619, + "learning_rate": 9.9890289989029e-06, + "loss": 0.3044, + "step": 10977 + }, + { + "epoch": 5.103672710367271, + "grad_norm": 5.738190174102783, + "learning_rate": 9.989027998902801e-06, + "loss": 0.3259, + "step": 10978 + }, + { + "epoch": 5.104137610413761, + "grad_norm": 11.194450378417969, + "learning_rate": 9.989026998902702e-06, + "loss": 0.4363, + "step": 10979 + }, + { + "epoch": 5.104602510460251, + "grad_norm": 5.9392194747924805, + "learning_rate": 9.989025998902601e-06, + "loss": 0.3714, + "step": 10980 + }, + { + "epoch": 5.105067410506741, + "grad_norm": 3.5345139503479004, + "learning_rate": 9.989024998902502e-06, + "loss": 0.1357, + "step": 10981 + }, + { + "epoch": 5.105532310553231, + "grad_norm": 5.003402233123779, + "learning_rate": 9.9890239989024e-06, + "loss": 0.1686, + "step": 10982 + }, + { + "epoch": 5.105997210599721, + "grad_norm": 3.4271233081817627, + "learning_rate": 9.9890229989023e-06, + "loss": 0.1525, + "step": 10983 + }, + { + "epoch": 5.106462110646211, + "grad_norm": 3.829040050506592, + "learning_rate": 9.9890219989022e-06, + "loss": 0.1582, + "step": 10984 + }, + { + "epoch": 5.106927010692701, + "grad_norm": 10.760604858398438, + "learning_rate": 9.9890209989021e-06, + "loss": 0.3155, + "step": 10985 + }, + { + "epoch": 5.107391910739191, + "grad_norm": 4.281643867492676, + "learning_rate": 9.989019998902001e-06, + "loss": 0.1431, + "step": 10986 + }, + { + "epoch": 5.107856810785681, + "grad_norm": 8.062453269958496, + "learning_rate": 9.9890189989019e-06, + "loss": 0.3364, + "step": 10987 + }, + { + "epoch": 5.108321710832171, + "grad_norm": 6.668592929840088, + "learning_rate": 9.989017998901801e-06, + "loss": 0.2164, + "step": 10988 + }, + { + "epoch": 5.108786610878661, + "grad_norm": 4.037251949310303, + "learning_rate": 9.9890169989017e-06, + "loss": 0.1225, + "step": 10989 + }, + { + "epoch": 5.1092515109251515, + "grad_norm": 4.8719892501831055, + "learning_rate": 9.9890159989016e-06, + "loss": 0.2654, + "step": 10990 + }, + { + "epoch": 5.109716410971641, + "grad_norm": 13.147176742553711, + "learning_rate": 9.989014998901501e-06, + "loss": 0.4705, + "step": 10991 + }, + { + "epoch": 5.110181311018131, + "grad_norm": 5.730197429656982, + "learning_rate": 9.989013998901402e-06, + "loss": 0.1073, + "step": 10992 + }, + { + "epoch": 5.110646211064621, + "grad_norm": 6.201664447784424, + "learning_rate": 9.989012998901301e-06, + "loss": 0.3461, + "step": 10993 + }, + { + "epoch": 5.111111111111111, + "grad_norm": 5.196551322937012, + "learning_rate": 9.9890119989012e-06, + "loss": 0.2331, + "step": 10994 + }, + { + "epoch": 5.111576011157601, + "grad_norm": 4.537326335906982, + "learning_rate": 9.9890109989011e-06, + "loss": 0.156, + "step": 10995 + }, + { + "epoch": 5.112040911204091, + "grad_norm": 4.235353946685791, + "learning_rate": 9.989009998901e-06, + "loss": 0.2234, + "step": 10996 + }, + { + "epoch": 5.112505811250581, + "grad_norm": 5.293135643005371, + "learning_rate": 9.9890089989009e-06, + "loss": 0.2913, + "step": 10997 + }, + { + "epoch": 5.112970711297071, + "grad_norm": 4.878074645996094, + "learning_rate": 9.989007998900801e-06, + "loss": 0.1236, + "step": 10998 + }, + { + "epoch": 5.1134356113435615, + "grad_norm": 4.401830673217773, + "learning_rate": 9.989006998900702e-06, + "loss": 0.2114, + "step": 10999 + }, + { + "epoch": 5.113900511390051, + "grad_norm": 4.473524570465088, + "learning_rate": 9.9890059989006e-06, + "loss": 0.2054, + "step": 11000 + }, + { + "epoch": 5.114365411436541, + "grad_norm": 3.560638427734375, + "learning_rate": 9.989004998900501e-06, + "loss": 0.0919, + "step": 11001 + }, + { + "epoch": 5.114830311483031, + "grad_norm": 2.4894065856933594, + "learning_rate": 9.9890039989004e-06, + "loss": 0.0528, + "step": 11002 + }, + { + "epoch": 5.115295211529521, + "grad_norm": 4.213519096374512, + "learning_rate": 9.9890029989003e-06, + "loss": 0.1387, + "step": 11003 + }, + { + "epoch": 5.115760111576011, + "grad_norm": 5.333871841430664, + "learning_rate": 9.989001998900201e-06, + "loss": 0.1448, + "step": 11004 + }, + { + "epoch": 5.116225011622501, + "grad_norm": 4.6752119064331055, + "learning_rate": 9.9890009989001e-06, + "loss": 0.1442, + "step": 11005 + }, + { + "epoch": 5.116689911668991, + "grad_norm": 4.252805709838867, + "learning_rate": 9.988999998900001e-06, + "loss": 0.1022, + "step": 11006 + }, + { + "epoch": 5.117154811715481, + "grad_norm": 6.107954502105713, + "learning_rate": 9.9889989988999e-06, + "loss": 0.427, + "step": 11007 + }, + { + "epoch": 5.1176197117619715, + "grad_norm": 9.729720115661621, + "learning_rate": 9.9889979988998e-06, + "loss": 0.3035, + "step": 11008 + }, + { + "epoch": 5.118084611808461, + "grad_norm": 5.383641242980957, + "learning_rate": 9.9889969988997e-06, + "loss": 0.242, + "step": 11009 + }, + { + "epoch": 5.118549511854951, + "grad_norm": 1.0586435794830322, + "learning_rate": 9.9889959988996e-06, + "loss": 0.0148, + "step": 11010 + }, + { + "epoch": 5.119014411901441, + "grad_norm": 5.568345546722412, + "learning_rate": 9.988994998899501e-06, + "loss": 0.083, + "step": 11011 + }, + { + "epoch": 5.119479311947932, + "grad_norm": 5.598052024841309, + "learning_rate": 9.988993998899402e-06, + "loss": 0.1994, + "step": 11012 + }, + { + "epoch": 5.119944211994421, + "grad_norm": 4.776963233947754, + "learning_rate": 9.9889929988993e-06, + "loss": 0.1739, + "step": 11013 + }, + { + "epoch": 5.120409112040911, + "grad_norm": 4.708008766174316, + "learning_rate": 9.988991998899201e-06, + "loss": 0.2193, + "step": 11014 + }, + { + "epoch": 5.120874012087401, + "grad_norm": 6.032577037811279, + "learning_rate": 9.9889909988991e-06, + "loss": 0.2871, + "step": 11015 + }, + { + "epoch": 5.121338912133891, + "grad_norm": 10.949178695678711, + "learning_rate": 9.988989998899001e-06, + "loss": 0.566, + "step": 11016 + }, + { + "epoch": 5.1218038121803815, + "grad_norm": 5.272647380828857, + "learning_rate": 9.988988998898902e-06, + "loss": 0.3021, + "step": 11017 + }, + { + "epoch": 5.122268712226871, + "grad_norm": 3.6669867038726807, + "learning_rate": 9.9889879988988e-06, + "loss": 0.1373, + "step": 11018 + }, + { + "epoch": 5.122733612273361, + "grad_norm": 6.896965503692627, + "learning_rate": 9.988986998898701e-06, + "loss": 0.1554, + "step": 11019 + }, + { + "epoch": 5.123198512319851, + "grad_norm": 4.378571510314941, + "learning_rate": 9.9889859988986e-06, + "loss": 0.1241, + "step": 11020 + }, + { + "epoch": 5.123663412366342, + "grad_norm": 5.48827600479126, + "learning_rate": 9.9889849988985e-06, + "loss": 0.1715, + "step": 11021 + }, + { + "epoch": 5.124128312412831, + "grad_norm": 8.108386993408203, + "learning_rate": 9.9889839988984e-06, + "loss": 0.1819, + "step": 11022 + }, + { + "epoch": 5.124593212459321, + "grad_norm": 6.329641342163086, + "learning_rate": 9.9889829988983e-06, + "loss": 0.2733, + "step": 11023 + }, + { + "epoch": 5.125058112505811, + "grad_norm": 9.044734001159668, + "learning_rate": 9.988981998898201e-06, + "loss": 0.3485, + "step": 11024 + }, + { + "epoch": 5.125523012552302, + "grad_norm": 6.25254487991333, + "learning_rate": 9.988980998898102e-06, + "loss": 0.319, + "step": 11025 + }, + { + "epoch": 5.1259879125987915, + "grad_norm": 15.924102783203125, + "learning_rate": 9.988979998898e-06, + "loss": 0.285, + "step": 11026 + }, + { + "epoch": 5.126452812645281, + "grad_norm": 5.369789123535156, + "learning_rate": 9.9889789988979e-06, + "loss": 0.2685, + "step": 11027 + }, + { + "epoch": 5.126917712691771, + "grad_norm": 4.515631198883057, + "learning_rate": 9.9889779988978e-06, + "loss": 0.2285, + "step": 11028 + }, + { + "epoch": 5.127382612738261, + "grad_norm": 1.986396074295044, + "learning_rate": 9.988976998897701e-06, + "loss": 0.0596, + "step": 11029 + }, + { + "epoch": 5.127847512784752, + "grad_norm": 4.629096984863281, + "learning_rate": 9.988975998897602e-06, + "loss": 0.2572, + "step": 11030 + }, + { + "epoch": 5.128312412831241, + "grad_norm": 8.407731056213379, + "learning_rate": 9.9889749988975e-06, + "loss": 0.2442, + "step": 11031 + }, + { + "epoch": 5.128777312877731, + "grad_norm": 4.826585292816162, + "learning_rate": 9.988973998897401e-06, + "loss": 0.1521, + "step": 11032 + }, + { + "epoch": 5.129242212924221, + "grad_norm": 5.2128095626831055, + "learning_rate": 9.9889729988973e-06, + "loss": 0.2831, + "step": 11033 + }, + { + "epoch": 5.129707112970712, + "grad_norm": 6.873355388641357, + "learning_rate": 9.988971998897201e-06, + "loss": 0.3273, + "step": 11034 + }, + { + "epoch": 5.1301720130172015, + "grad_norm": 8.282797813415527, + "learning_rate": 9.9889709988971e-06, + "loss": 0.4078, + "step": 11035 + }, + { + "epoch": 5.130636913063691, + "grad_norm": 4.779570579528809, + "learning_rate": 9.988969998897e-06, + "loss": 0.1372, + "step": 11036 + }, + { + "epoch": 5.131101813110181, + "grad_norm": 8.496000289916992, + "learning_rate": 9.988968998896901e-06, + "loss": 0.3156, + "step": 11037 + }, + { + "epoch": 5.131566713156671, + "grad_norm": 5.348788738250732, + "learning_rate": 9.9889679988968e-06, + "loss": 0.2986, + "step": 11038 + }, + { + "epoch": 5.132031613203162, + "grad_norm": 6.060985088348389, + "learning_rate": 9.9889669988967e-06, + "loss": 0.1479, + "step": 11039 + }, + { + "epoch": 5.132496513249651, + "grad_norm": 5.619966983795166, + "learning_rate": 9.9889659988966e-06, + "loss": 0.2043, + "step": 11040 + }, + { + "epoch": 5.132961413296141, + "grad_norm": 9.392255783081055, + "learning_rate": 9.9889649988965e-06, + "loss": 0.2597, + "step": 11041 + }, + { + "epoch": 5.133426313342631, + "grad_norm": 4.074686050415039, + "learning_rate": 9.988963998896401e-06, + "loss": 0.1781, + "step": 11042 + }, + { + "epoch": 5.133891213389122, + "grad_norm": 3.9252400398254395, + "learning_rate": 9.988962998896302e-06, + "loss": 0.1394, + "step": 11043 + }, + { + "epoch": 5.1343561134356115, + "grad_norm": 2.017900228500366, + "learning_rate": 9.9889619988962e-06, + "loss": 0.0586, + "step": 11044 + }, + { + "epoch": 5.134821013482101, + "grad_norm": 8.82087516784668, + "learning_rate": 9.988960998896101e-06, + "loss": 0.2807, + "step": 11045 + }, + { + "epoch": 5.135285913528591, + "grad_norm": 4.697834491729736, + "learning_rate": 9.988959998896e-06, + "loss": 0.2059, + "step": 11046 + }, + { + "epoch": 5.135750813575082, + "grad_norm": 4.378930568695068, + "learning_rate": 9.988958998895901e-06, + "loss": 0.2203, + "step": 11047 + }, + { + "epoch": 5.136215713621572, + "grad_norm": 6.685934543609619, + "learning_rate": 9.988957998895802e-06, + "loss": 0.31, + "step": 11048 + }, + { + "epoch": 5.136680613668061, + "grad_norm": 14.056069374084473, + "learning_rate": 9.9889569988957e-06, + "loss": 0.3654, + "step": 11049 + }, + { + "epoch": 5.137145513714551, + "grad_norm": 4.367588996887207, + "learning_rate": 9.988955998895601e-06, + "loss": 0.166, + "step": 11050 + }, + { + "epoch": 5.137610413761041, + "grad_norm": 5.4987993240356445, + "learning_rate": 9.9889549988955e-06, + "loss": 0.1421, + "step": 11051 + }, + { + "epoch": 5.138075313807532, + "grad_norm": 5.094279766082764, + "learning_rate": 9.9889539988954e-06, + "loss": 0.2539, + "step": 11052 + }, + { + "epoch": 5.1385402138540215, + "grad_norm": 7.470536708831787, + "learning_rate": 9.9889529988953e-06, + "loss": 0.2803, + "step": 11053 + }, + { + "epoch": 5.139005113900511, + "grad_norm": 5.614582061767578, + "learning_rate": 9.9889519988952e-06, + "loss": 0.1252, + "step": 11054 + }, + { + "epoch": 5.139470013947001, + "grad_norm": 8.498311996459961, + "learning_rate": 9.988950998895101e-06, + "loss": 0.5399, + "step": 11055 + }, + { + "epoch": 5.139934913993492, + "grad_norm": 5.4970479011535645, + "learning_rate": 9.988949998895002e-06, + "loss": 0.2672, + "step": 11056 + }, + { + "epoch": 5.140399814039982, + "grad_norm": 6.037197113037109, + "learning_rate": 9.9889489988949e-06, + "loss": 0.2687, + "step": 11057 + }, + { + "epoch": 5.140864714086471, + "grad_norm": 5.324801921844482, + "learning_rate": 9.9889479988948e-06, + "loss": 0.2945, + "step": 11058 + }, + { + "epoch": 5.141329614132961, + "grad_norm": 7.823080539703369, + "learning_rate": 9.9889469988947e-06, + "loss": 0.2967, + "step": 11059 + }, + { + "epoch": 5.141794514179451, + "grad_norm": 6.997341632843018, + "learning_rate": 9.988945998894601e-06, + "loss": 0.4836, + "step": 11060 + }, + { + "epoch": 5.142259414225942, + "grad_norm": 8.044424057006836, + "learning_rate": 9.988944998894502e-06, + "loss": 0.4236, + "step": 11061 + }, + { + "epoch": 5.1427243142724315, + "grad_norm": 4.557175636291504, + "learning_rate": 9.9889439988944e-06, + "loss": 0.1566, + "step": 11062 + }, + { + "epoch": 5.143189214318921, + "grad_norm": 6.40741491317749, + "learning_rate": 9.988942998894301e-06, + "loss": 0.3878, + "step": 11063 + }, + { + "epoch": 5.143654114365411, + "grad_norm": 3.645609140396118, + "learning_rate": 9.9889419988942e-06, + "loss": 0.1515, + "step": 11064 + }, + { + "epoch": 5.144119014411902, + "grad_norm": 6.536356449127197, + "learning_rate": 9.988940998894101e-06, + "loss": 0.4271, + "step": 11065 + }, + { + "epoch": 5.144583914458392, + "grad_norm": 4.730187892913818, + "learning_rate": 9.988939998894e-06, + "loss": 0.1884, + "step": 11066 + }, + { + "epoch": 5.145048814504881, + "grad_norm": 6.036203384399414, + "learning_rate": 9.9889389988939e-06, + "loss": 0.3098, + "step": 11067 + }, + { + "epoch": 5.145513714551371, + "grad_norm": 3.307776689529419, + "learning_rate": 9.988937998893801e-06, + "loss": 0.1249, + "step": 11068 + }, + { + "epoch": 5.145978614597862, + "grad_norm": 3.9013876914978027, + "learning_rate": 9.9889369988937e-06, + "loss": 0.1723, + "step": 11069 + }, + { + "epoch": 5.146443514644352, + "grad_norm": 5.665116310119629, + "learning_rate": 9.9889359988936e-06, + "loss": 0.283, + "step": 11070 + }, + { + "epoch": 5.1469084146908415, + "grad_norm": 7.263914585113525, + "learning_rate": 9.9889349988935e-06, + "loss": 0.3549, + "step": 11071 + }, + { + "epoch": 5.147373314737331, + "grad_norm": 10.341962814331055, + "learning_rate": 9.9889339988934e-06, + "loss": 0.3185, + "step": 11072 + }, + { + "epoch": 5.147838214783821, + "grad_norm": 5.022977828979492, + "learning_rate": 9.988932998893301e-06, + "loss": 0.1904, + "step": 11073 + }, + { + "epoch": 5.148303114830312, + "grad_norm": 4.93693733215332, + "learning_rate": 9.988931998893202e-06, + "loss": 0.2446, + "step": 11074 + }, + { + "epoch": 5.148768014876802, + "grad_norm": 5.2442731857299805, + "learning_rate": 9.9889309988931e-06, + "loss": 0.2569, + "step": 11075 + }, + { + "epoch": 5.149232914923291, + "grad_norm": 4.281988143920898, + "learning_rate": 9.988929998893001e-06, + "loss": 0.2685, + "step": 11076 + }, + { + "epoch": 5.149697814969781, + "grad_norm": 10.21973991394043, + "learning_rate": 9.9889289988929e-06, + "loss": 0.25, + "step": 11077 + }, + { + "epoch": 5.150162715016272, + "grad_norm": 4.778295993804932, + "learning_rate": 9.988927998892801e-06, + "loss": 0.215, + "step": 11078 + }, + { + "epoch": 5.150627615062762, + "grad_norm": 4.962299823760986, + "learning_rate": 9.9889269988927e-06, + "loss": 0.1944, + "step": 11079 + }, + { + "epoch": 5.1510925151092515, + "grad_norm": 4.421217441558838, + "learning_rate": 9.9889259988926e-06, + "loss": 0.2113, + "step": 11080 + }, + { + "epoch": 5.151557415155741, + "grad_norm": 6.102321624755859, + "learning_rate": 9.988924998892501e-06, + "loss": 0.1747, + "step": 11081 + }, + { + "epoch": 5.152022315202231, + "grad_norm": 4.897612571716309, + "learning_rate": 9.9889239988924e-06, + "loss": 0.1872, + "step": 11082 + }, + { + "epoch": 5.152487215248722, + "grad_norm": 3.281313896179199, + "learning_rate": 9.988922998892301e-06, + "loss": 0.1078, + "step": 11083 + }, + { + "epoch": 5.152952115295212, + "grad_norm": 5.120100975036621, + "learning_rate": 9.9889219988922e-06, + "loss": 0.229, + "step": 11084 + }, + { + "epoch": 5.153417015341701, + "grad_norm": 7.300202369689941, + "learning_rate": 9.9889209988921e-06, + "loss": 0.4749, + "step": 11085 + }, + { + "epoch": 5.153881915388191, + "grad_norm": 3.593825340270996, + "learning_rate": 9.988919998892001e-06, + "loss": 0.1265, + "step": 11086 + }, + { + "epoch": 5.154346815434682, + "grad_norm": 3.155224084854126, + "learning_rate": 9.988918998891902e-06, + "loss": 0.1229, + "step": 11087 + }, + { + "epoch": 5.154811715481172, + "grad_norm": 5.029443740844727, + "learning_rate": 9.9889179988918e-06, + "loss": 0.193, + "step": 11088 + }, + { + "epoch": 5.1552766155276615, + "grad_norm": 2.969827651977539, + "learning_rate": 9.988916998891701e-06, + "loss": 0.0554, + "step": 11089 + }, + { + "epoch": 5.155741515574151, + "grad_norm": 5.693967342376709, + "learning_rate": 9.9889159988916e-06, + "loss": 0.1791, + "step": 11090 + }, + { + "epoch": 5.156206415620642, + "grad_norm": 3.1898093223571777, + "learning_rate": 9.9889149988915e-06, + "loss": 0.1023, + "step": 11091 + }, + { + "epoch": 5.156671315667132, + "grad_norm": 12.8561372756958, + "learning_rate": 9.9889139988914e-06, + "loss": 0.3385, + "step": 11092 + }, + { + "epoch": 5.157136215713622, + "grad_norm": 5.796658039093018, + "learning_rate": 9.9889129988913e-06, + "loss": 0.2939, + "step": 11093 + }, + { + "epoch": 5.157601115760111, + "grad_norm": 5.0882887840271, + "learning_rate": 9.988911998891201e-06, + "loss": 0.2437, + "step": 11094 + }, + { + "epoch": 5.158066015806601, + "grad_norm": 4.720968246459961, + "learning_rate": 9.9889109988911e-06, + "loss": 0.1928, + "step": 11095 + }, + { + "epoch": 5.158530915853092, + "grad_norm": 5.928635597229004, + "learning_rate": 9.988909998891001e-06, + "loss": 0.31, + "step": 11096 + }, + { + "epoch": 5.158995815899582, + "grad_norm": 3.1893343925476074, + "learning_rate": 9.9889089988909e-06, + "loss": 0.0922, + "step": 11097 + }, + { + "epoch": 5.1594607159460715, + "grad_norm": 5.936867713928223, + "learning_rate": 9.9889079988908e-06, + "loss": 0.1291, + "step": 11098 + }, + { + "epoch": 5.159925615992561, + "grad_norm": 3.8787825107574463, + "learning_rate": 9.988906998890701e-06, + "loss": 0.1385, + "step": 11099 + }, + { + "epoch": 5.160390516039052, + "grad_norm": 5.587488174438477, + "learning_rate": 9.988905998890602e-06, + "loss": 0.2901, + "step": 11100 + }, + { + "epoch": 5.160855416085542, + "grad_norm": 6.400516986846924, + "learning_rate": 9.988904998890501e-06, + "loss": 0.3402, + "step": 11101 + }, + { + "epoch": 5.161320316132032, + "grad_norm": 5.208285331726074, + "learning_rate": 9.9889039988904e-06, + "loss": 0.2355, + "step": 11102 + }, + { + "epoch": 5.161785216178521, + "grad_norm": 3.180446147918701, + "learning_rate": 9.9889029988903e-06, + "loss": 0.0931, + "step": 11103 + }, + { + "epoch": 5.162250116225012, + "grad_norm": 7.323156356811523, + "learning_rate": 9.9889019988902e-06, + "loss": 0.2501, + "step": 11104 + }, + { + "epoch": 5.162715016271502, + "grad_norm": 5.463550090789795, + "learning_rate": 9.9889009988901e-06, + "loss": 0.4268, + "step": 11105 + }, + { + "epoch": 5.163179916317992, + "grad_norm": 7.131437301635742, + "learning_rate": 9.98889999889e-06, + "loss": 0.2835, + "step": 11106 + }, + { + "epoch": 5.1636448163644815, + "grad_norm": 3.579371452331543, + "learning_rate": 9.988898998889901e-06, + "loss": 0.148, + "step": 11107 + }, + { + "epoch": 5.164109716410971, + "grad_norm": 12.738482475280762, + "learning_rate": 9.9888979988898e-06, + "loss": 0.1775, + "step": 11108 + }, + { + "epoch": 5.164574616457462, + "grad_norm": 6.6138691902160645, + "learning_rate": 9.988896998889701e-06, + "loss": 0.3227, + "step": 11109 + }, + { + "epoch": 5.165039516503952, + "grad_norm": 5.79180908203125, + "learning_rate": 9.9888959988896e-06, + "loss": 0.3505, + "step": 11110 + }, + { + "epoch": 5.165504416550442, + "grad_norm": 5.8532938957214355, + "learning_rate": 9.9888949988895e-06, + "loss": 0.2706, + "step": 11111 + }, + { + "epoch": 5.165969316596931, + "grad_norm": 4.48431396484375, + "learning_rate": 9.988893998889401e-06, + "loss": 0.2301, + "step": 11112 + }, + { + "epoch": 5.166434216643422, + "grad_norm": 3.8308522701263428, + "learning_rate": 9.9888929988893e-06, + "loss": 0.0985, + "step": 11113 + }, + { + "epoch": 5.166899116689912, + "grad_norm": 4.18000602722168, + "learning_rate": 9.988891998889201e-06, + "loss": 0.1479, + "step": 11114 + }, + { + "epoch": 5.167364016736402, + "grad_norm": 2.7904462814331055, + "learning_rate": 9.9888909988891e-06, + "loss": 0.0829, + "step": 11115 + }, + { + "epoch": 5.1678289167828915, + "grad_norm": 5.081626892089844, + "learning_rate": 9.988889998889e-06, + "loss": 0.2383, + "step": 11116 + }, + { + "epoch": 5.168293816829381, + "grad_norm": 5.723305702209473, + "learning_rate": 9.988888998888901e-06, + "loss": 0.3431, + "step": 11117 + }, + { + "epoch": 5.168758716875872, + "grad_norm": 5.451626777648926, + "learning_rate": 9.9888879988888e-06, + "loss": 0.271, + "step": 11118 + }, + { + "epoch": 5.169223616922362, + "grad_norm": 4.260531902313232, + "learning_rate": 9.988886998888701e-06, + "loss": 0.22, + "step": 11119 + }, + { + "epoch": 5.169688516968852, + "grad_norm": 10.967809677124023, + "learning_rate": 9.988885998888602e-06, + "loss": 0.1483, + "step": 11120 + }, + { + "epoch": 5.170153417015341, + "grad_norm": 5.307269096374512, + "learning_rate": 9.9888849988885e-06, + "loss": 0.2083, + "step": 11121 + }, + { + "epoch": 5.170618317061832, + "grad_norm": 3.9050607681274414, + "learning_rate": 9.988883998888401e-06, + "loss": 0.0729, + "step": 11122 + }, + { + "epoch": 5.171083217108322, + "grad_norm": 9.15485954284668, + "learning_rate": 9.9888829988883e-06, + "loss": 0.2859, + "step": 11123 + }, + { + "epoch": 5.171548117154812, + "grad_norm": 4.3648457527160645, + "learning_rate": 9.9888819988882e-06, + "loss": 0.1443, + "step": 11124 + }, + { + "epoch": 5.1720130172013015, + "grad_norm": 7.236875057220459, + "learning_rate": 9.988880998888101e-06, + "loss": 0.4437, + "step": 11125 + }, + { + "epoch": 5.172477917247791, + "grad_norm": 4.2478203773498535, + "learning_rate": 9.988879998888e-06, + "loss": 0.1719, + "step": 11126 + }, + { + "epoch": 5.172942817294282, + "grad_norm": 4.832932949066162, + "learning_rate": 9.988878998887901e-06, + "loss": 0.1582, + "step": 11127 + }, + { + "epoch": 5.173407717340772, + "grad_norm": 6.499814033508301, + "learning_rate": 9.9888779988878e-06, + "loss": 0.3823, + "step": 11128 + }, + { + "epoch": 5.173872617387262, + "grad_norm": 9.746864318847656, + "learning_rate": 9.9888769988877e-06, + "loss": 0.3717, + "step": 11129 + }, + { + "epoch": 5.174337517433751, + "grad_norm": 4.087283134460449, + "learning_rate": 9.988875998887601e-06, + "loss": 0.1264, + "step": 11130 + }, + { + "epoch": 5.174802417480242, + "grad_norm": 5.759976387023926, + "learning_rate": 9.9888749988875e-06, + "loss": 0.197, + "step": 11131 + }, + { + "epoch": 5.175267317526732, + "grad_norm": 7.53461217880249, + "learning_rate": 9.988873998887401e-06, + "loss": 0.4223, + "step": 11132 + }, + { + "epoch": 5.175732217573222, + "grad_norm": 5.644089698791504, + "learning_rate": 9.988872998887302e-06, + "loss": 0.3052, + "step": 11133 + }, + { + "epoch": 5.1761971176197115, + "grad_norm": 7.563586235046387, + "learning_rate": 9.9888719988872e-06, + "loss": 0.1459, + "step": 11134 + }, + { + "epoch": 5.176662017666202, + "grad_norm": 4.008749961853027, + "learning_rate": 9.9888709988871e-06, + "loss": 0.1525, + "step": 11135 + }, + { + "epoch": 5.177126917712692, + "grad_norm": 7.258610725402832, + "learning_rate": 9.988869998887e-06, + "loss": 0.463, + "step": 11136 + }, + { + "epoch": 5.177591817759182, + "grad_norm": 3.554080009460449, + "learning_rate": 9.988868998886901e-06, + "loss": 0.1188, + "step": 11137 + }, + { + "epoch": 5.178056717805672, + "grad_norm": 2.794637680053711, + "learning_rate": 9.988867998886802e-06, + "loss": 0.0865, + "step": 11138 + }, + { + "epoch": 5.178521617852161, + "grad_norm": 6.892092227935791, + "learning_rate": 9.9888669988867e-06, + "loss": 0.3804, + "step": 11139 + }, + { + "epoch": 5.178986517898652, + "grad_norm": 6.258517742156982, + "learning_rate": 9.988865998886601e-06, + "loss": 0.3445, + "step": 11140 + }, + { + "epoch": 5.179451417945142, + "grad_norm": 4.612583637237549, + "learning_rate": 9.9888649988865e-06, + "loss": 0.2336, + "step": 11141 + }, + { + "epoch": 5.179916317991632, + "grad_norm": 6.942617416381836, + "learning_rate": 9.9888639988864e-06, + "loss": 0.4623, + "step": 11142 + }, + { + "epoch": 5.1803812180381215, + "grad_norm": 2.6085782051086426, + "learning_rate": 9.988862998886301e-06, + "loss": 0.0896, + "step": 11143 + }, + { + "epoch": 5.180846118084612, + "grad_norm": 4.447371006011963, + "learning_rate": 9.988861998886202e-06, + "loss": 0.1324, + "step": 11144 + }, + { + "epoch": 5.181311018131102, + "grad_norm": 6.150737762451172, + "learning_rate": 9.988860998886101e-06, + "loss": 0.3601, + "step": 11145 + }, + { + "epoch": 5.181775918177592, + "grad_norm": 2.902573823928833, + "learning_rate": 9.988859998886e-06, + "loss": 0.0892, + "step": 11146 + }, + { + "epoch": 5.182240818224082, + "grad_norm": 8.790007591247559, + "learning_rate": 9.9888589988859e-06, + "loss": 0.3633, + "step": 11147 + }, + { + "epoch": 5.182705718270572, + "grad_norm": 4.4394755363464355, + "learning_rate": 9.9888579988858e-06, + "loss": 0.2352, + "step": 11148 + }, + { + "epoch": 5.183170618317062, + "grad_norm": 13.191089630126953, + "learning_rate": 9.9888569988857e-06, + "loss": 0.244, + "step": 11149 + }, + { + "epoch": 5.183635518363552, + "grad_norm": 4.595800876617432, + "learning_rate": 9.988855998885601e-06, + "loss": 0.0953, + "step": 11150 + }, + { + "epoch": 5.184100418410042, + "grad_norm": 6.0978522300720215, + "learning_rate": 9.988854998885502e-06, + "loss": 0.2583, + "step": 11151 + }, + { + "epoch": 5.1845653184565315, + "grad_norm": 4.615416049957275, + "learning_rate": 9.9888539988854e-06, + "loss": 0.1798, + "step": 11152 + }, + { + "epoch": 5.185030218503022, + "grad_norm": 4.2585344314575195, + "learning_rate": 9.988852998885301e-06, + "loss": 0.1554, + "step": 11153 + }, + { + "epoch": 5.185495118549512, + "grad_norm": 4.255918979644775, + "learning_rate": 9.9888519988852e-06, + "loss": 0.3027, + "step": 11154 + }, + { + "epoch": 5.185960018596002, + "grad_norm": 3.458585023880005, + "learning_rate": 9.9888509988851e-06, + "loss": 0.1085, + "step": 11155 + }, + { + "epoch": 5.1864249186424916, + "grad_norm": 8.720917701721191, + "learning_rate": 9.988849998885001e-06, + "loss": 0.3626, + "step": 11156 + }, + { + "epoch": 5.186889818688982, + "grad_norm": 4.2652740478515625, + "learning_rate": 9.9888489988849e-06, + "loss": 0.115, + "step": 11157 + }, + { + "epoch": 5.187354718735472, + "grad_norm": 4.965098857879639, + "learning_rate": 9.988847998884801e-06, + "loss": 0.2033, + "step": 11158 + }, + { + "epoch": 5.187819618781962, + "grad_norm": 2.5085060596466064, + "learning_rate": 9.9888469988847e-06, + "loss": 0.0669, + "step": 11159 + }, + { + "epoch": 5.188284518828452, + "grad_norm": 4.271401405334473, + "learning_rate": 9.9888459988846e-06, + "loss": 0.1508, + "step": 11160 + }, + { + "epoch": 5.1887494188749415, + "grad_norm": 3.0039892196655273, + "learning_rate": 9.9888449988845e-06, + "loss": 0.0764, + "step": 11161 + }, + { + "epoch": 5.189214318921432, + "grad_norm": 5.98223352432251, + "learning_rate": 9.9888439988844e-06, + "loss": 0.2756, + "step": 11162 + }, + { + "epoch": 5.189679218967922, + "grad_norm": 7.659090995788574, + "learning_rate": 9.988842998884301e-06, + "loss": 0.356, + "step": 11163 + }, + { + "epoch": 5.190144119014412, + "grad_norm": 10.049285888671875, + "learning_rate": 9.988841998884202e-06, + "loss": 0.3891, + "step": 11164 + }, + { + "epoch": 5.1906090190609016, + "grad_norm": 7.195994853973389, + "learning_rate": 9.9888409988841e-06, + "loss": 0.3971, + "step": 11165 + }, + { + "epoch": 5.191073919107392, + "grad_norm": 7.650528430938721, + "learning_rate": 9.988839998884001e-06, + "loss": 0.4431, + "step": 11166 + }, + { + "epoch": 5.191538819153882, + "grad_norm": 2.70320463180542, + "learning_rate": 9.9888389988839e-06, + "loss": 0.0507, + "step": 11167 + }, + { + "epoch": 5.192003719200372, + "grad_norm": 5.744325160980225, + "learning_rate": 9.988837998883801e-06, + "loss": 0.1621, + "step": 11168 + }, + { + "epoch": 5.192468619246862, + "grad_norm": 4.654499053955078, + "learning_rate": 9.988836998883702e-06, + "loss": 0.2221, + "step": 11169 + }, + { + "epoch": 5.192933519293352, + "grad_norm": 6.311840534210205, + "learning_rate": 9.9888359988836e-06, + "loss": 0.4173, + "step": 11170 + }, + { + "epoch": 5.193398419339842, + "grad_norm": 7.05149507522583, + "learning_rate": 9.988834998883501e-06, + "loss": 0.1878, + "step": 11171 + }, + { + "epoch": 5.193863319386332, + "grad_norm": 6.80274772644043, + "learning_rate": 9.9888339988834e-06, + "loss": 0.4147, + "step": 11172 + }, + { + "epoch": 5.194328219432822, + "grad_norm": 3.227658271789551, + "learning_rate": 9.9888329988833e-06, + "loss": 0.0748, + "step": 11173 + }, + { + "epoch": 5.1947931194793115, + "grad_norm": 4.105044364929199, + "learning_rate": 9.9888319988832e-06, + "loss": 0.1005, + "step": 11174 + }, + { + "epoch": 5.195258019525802, + "grad_norm": 5.9500508308410645, + "learning_rate": 9.9888309988831e-06, + "loss": 0.198, + "step": 11175 + }, + { + "epoch": 5.195722919572292, + "grad_norm": 3.6847922801971436, + "learning_rate": 9.988829998883001e-06, + "loss": 0.1974, + "step": 11176 + }, + { + "epoch": 5.196187819618782, + "grad_norm": 7.310481548309326, + "learning_rate": 9.988828998882902e-06, + "loss": 0.3622, + "step": 11177 + }, + { + "epoch": 5.196652719665272, + "grad_norm": 3.8456921577453613, + "learning_rate": 9.9888279988828e-06, + "loss": 0.1086, + "step": 11178 + }, + { + "epoch": 5.197117619711762, + "grad_norm": 5.098194122314453, + "learning_rate": 9.9888269988827e-06, + "loss": 0.2208, + "step": 11179 + }, + { + "epoch": 5.197582519758252, + "grad_norm": 6.882020950317383, + "learning_rate": 9.9888259988826e-06, + "loss": 0.4444, + "step": 11180 + }, + { + "epoch": 5.198047419804742, + "grad_norm": 4.05751371383667, + "learning_rate": 9.988824998882501e-06, + "loss": 0.1616, + "step": 11181 + }, + { + "epoch": 5.198512319851232, + "grad_norm": 5.211989879608154, + "learning_rate": 9.988823998882402e-06, + "loss": 0.3446, + "step": 11182 + }, + { + "epoch": 5.198977219897722, + "grad_norm": 12.07154369354248, + "learning_rate": 9.9888229988823e-06, + "loss": 0.1289, + "step": 11183 + }, + { + "epoch": 5.199442119944212, + "grad_norm": 5.286795616149902, + "learning_rate": 9.988821998882201e-06, + "loss": 0.1868, + "step": 11184 + }, + { + "epoch": 5.199907019990702, + "grad_norm": 3.2029967308044434, + "learning_rate": 9.9888209988821e-06, + "loss": 0.08, + "step": 11185 + }, + { + "epoch": 5.200371920037192, + "grad_norm": 3.259214162826538, + "learning_rate": 9.988819998882001e-06, + "loss": 0.1055, + "step": 11186 + }, + { + "epoch": 5.200836820083682, + "grad_norm": 3.509380578994751, + "learning_rate": 9.9888189988819e-06, + "loss": 0.0739, + "step": 11187 + }, + { + "epoch": 5.201301720130172, + "grad_norm": 5.141380786895752, + "learning_rate": 9.9888179988818e-06, + "loss": 0.1489, + "step": 11188 + }, + { + "epoch": 5.201766620176662, + "grad_norm": 14.507543563842773, + "learning_rate": 9.988816998881701e-06, + "loss": 0.269, + "step": 11189 + }, + { + "epoch": 5.202231520223152, + "grad_norm": 6.959283351898193, + "learning_rate": 9.9888159988816e-06, + "loss": 0.2359, + "step": 11190 + }, + { + "epoch": 5.202696420269642, + "grad_norm": 6.133047580718994, + "learning_rate": 9.9888149988815e-06, + "loss": 0.3461, + "step": 11191 + }, + { + "epoch": 5.203161320316132, + "grad_norm": 4.934201240539551, + "learning_rate": 9.9888139988814e-06, + "loss": 0.2056, + "step": 11192 + }, + { + "epoch": 5.203626220362622, + "grad_norm": 5.998302936553955, + "learning_rate": 9.9888129988813e-06, + "loss": 0.218, + "step": 11193 + }, + { + "epoch": 5.204091120409112, + "grad_norm": 5.131719589233398, + "learning_rate": 9.988811998881201e-06, + "loss": 0.2513, + "step": 11194 + }, + { + "epoch": 5.204556020455602, + "grad_norm": 3.9869184494018555, + "learning_rate": 9.988810998881102e-06, + "loss": 0.1164, + "step": 11195 + }, + { + "epoch": 5.205020920502092, + "grad_norm": 4.805441856384277, + "learning_rate": 9.988809998881e-06, + "loss": 0.1968, + "step": 11196 + }, + { + "epoch": 5.205485820548582, + "grad_norm": 5.277163028717041, + "learning_rate": 9.988808998880901e-06, + "loss": 0.2406, + "step": 11197 + }, + { + "epoch": 5.205950720595072, + "grad_norm": 7.0483479499816895, + "learning_rate": 9.9888079988808e-06, + "loss": 0.2915, + "step": 11198 + }, + { + "epoch": 5.206415620641562, + "grad_norm": 17.221988677978516, + "learning_rate": 9.988806998880701e-06, + "loss": 0.2052, + "step": 11199 + }, + { + "epoch": 5.206880520688052, + "grad_norm": 9.446733474731445, + "learning_rate": 9.9888059988806e-06, + "loss": 0.534, + "step": 11200 + }, + { + "epoch": 5.207345420734542, + "grad_norm": 6.250046253204346, + "learning_rate": 9.9888049988805e-06, + "loss": 0.0916, + "step": 11201 + }, + { + "epoch": 5.207810320781032, + "grad_norm": 2.977079153060913, + "learning_rate": 9.988803998880401e-06, + "loss": 0.0661, + "step": 11202 + }, + { + "epoch": 5.208275220827522, + "grad_norm": 3.9858195781707764, + "learning_rate": 9.9888029988803e-06, + "loss": 0.183, + "step": 11203 + }, + { + "epoch": 5.208740120874012, + "grad_norm": 7.604218006134033, + "learning_rate": 9.988801998880201e-06, + "loss": 0.3046, + "step": 11204 + }, + { + "epoch": 5.209205020920502, + "grad_norm": 4.375697135925293, + "learning_rate": 9.9888009988801e-06, + "loss": 0.2919, + "step": 11205 + }, + { + "epoch": 5.209669920966992, + "grad_norm": 5.0768256187438965, + "learning_rate": 9.98879999888e-06, + "loss": 0.1637, + "step": 11206 + }, + { + "epoch": 5.210134821013482, + "grad_norm": 5.971208572387695, + "learning_rate": 9.988798998879901e-06, + "loss": 0.3139, + "step": 11207 + }, + { + "epoch": 5.210599721059972, + "grad_norm": 5.576565742492676, + "learning_rate": 9.988797998879802e-06, + "loss": 0.1944, + "step": 11208 + }, + { + "epoch": 5.211064621106462, + "grad_norm": 5.880520820617676, + "learning_rate": 9.9887969988797e-06, + "loss": 0.4231, + "step": 11209 + }, + { + "epoch": 5.211529521152952, + "grad_norm": 7.038687229156494, + "learning_rate": 9.9887959988796e-06, + "loss": 0.1842, + "step": 11210 + }, + { + "epoch": 5.211994421199442, + "grad_norm": 4.057091236114502, + "learning_rate": 9.9887949988795e-06, + "loss": 0.1942, + "step": 11211 + }, + { + "epoch": 5.212459321245932, + "grad_norm": 4.242066860198975, + "learning_rate": 9.988793998879401e-06, + "loss": 0.1463, + "step": 11212 + }, + { + "epoch": 5.212924221292422, + "grad_norm": 2.9622905254364014, + "learning_rate": 9.988792998879302e-06, + "loss": 0.1105, + "step": 11213 + }, + { + "epoch": 5.2133891213389125, + "grad_norm": 5.708853244781494, + "learning_rate": 9.9887919988792e-06, + "loss": 0.2834, + "step": 11214 + }, + { + "epoch": 5.213854021385402, + "grad_norm": 4.4569878578186035, + "learning_rate": 9.988790998879101e-06, + "loss": 0.1816, + "step": 11215 + }, + { + "epoch": 5.214318921431892, + "grad_norm": 4.346212387084961, + "learning_rate": 9.988789998879e-06, + "loss": 0.1717, + "step": 11216 + }, + { + "epoch": 5.214783821478382, + "grad_norm": 5.284202575683594, + "learning_rate": 9.988788998878901e-06, + "loss": 0.1892, + "step": 11217 + }, + { + "epoch": 5.215248721524872, + "grad_norm": 4.3254499435424805, + "learning_rate": 9.9887879988788e-06, + "loss": 0.2142, + "step": 11218 + }, + { + "epoch": 5.215713621571362, + "grad_norm": 3.6190037727355957, + "learning_rate": 9.9887869988787e-06, + "loss": 0.1811, + "step": 11219 + }, + { + "epoch": 5.216178521617852, + "grad_norm": 6.141335487365723, + "learning_rate": 9.988785998878601e-06, + "loss": 0.3328, + "step": 11220 + }, + { + "epoch": 5.216643421664342, + "grad_norm": 5.1036696434021, + "learning_rate": 9.9887849988785e-06, + "loss": 0.1701, + "step": 11221 + }, + { + "epoch": 5.217108321710832, + "grad_norm": 3.9306442737579346, + "learning_rate": 9.988783998878401e-06, + "loss": 0.1726, + "step": 11222 + }, + { + "epoch": 5.2175732217573225, + "grad_norm": 4.831483840942383, + "learning_rate": 9.9887829988783e-06, + "loss": 0.1982, + "step": 11223 + }, + { + "epoch": 5.218038121803812, + "grad_norm": 4.941807270050049, + "learning_rate": 9.9887819988782e-06, + "loss": 0.1864, + "step": 11224 + }, + { + "epoch": 5.218503021850302, + "grad_norm": 7.100167751312256, + "learning_rate": 9.988780998878101e-06, + "loss": 0.396, + "step": 11225 + }, + { + "epoch": 5.218967921896792, + "grad_norm": 5.78481388092041, + "learning_rate": 9.988779998878002e-06, + "loss": 0.2824, + "step": 11226 + }, + { + "epoch": 5.219432821943283, + "grad_norm": 7.009482383728027, + "learning_rate": 9.9887789988779e-06, + "loss": 0.4024, + "step": 11227 + }, + { + "epoch": 5.219897721989772, + "grad_norm": 4.262627601623535, + "learning_rate": 9.988777998877801e-06, + "loss": 0.2055, + "step": 11228 + }, + { + "epoch": 5.220362622036262, + "grad_norm": 3.706676959991455, + "learning_rate": 9.9887769988777e-06, + "loss": 0.1618, + "step": 11229 + }, + { + "epoch": 5.220827522082752, + "grad_norm": 5.881284713745117, + "learning_rate": 9.988775998877601e-06, + "loss": 0.3357, + "step": 11230 + }, + { + "epoch": 5.221292422129242, + "grad_norm": 5.94756555557251, + "learning_rate": 9.9887749988775e-06, + "loss": 0.2669, + "step": 11231 + }, + { + "epoch": 5.2217573221757325, + "grad_norm": 3.225766658782959, + "learning_rate": 9.9887739988774e-06, + "loss": 0.0894, + "step": 11232 + }, + { + "epoch": 5.222222222222222, + "grad_norm": 6.933812618255615, + "learning_rate": 9.988772998877301e-06, + "loss": 0.4053, + "step": 11233 + }, + { + "epoch": 5.222687122268712, + "grad_norm": 4.061377048492432, + "learning_rate": 9.9887719988772e-06, + "loss": 0.1628, + "step": 11234 + }, + { + "epoch": 5.223152022315202, + "grad_norm": 5.862953186035156, + "learning_rate": 9.988770998877101e-06, + "loss": 0.147, + "step": 11235 + }, + { + "epoch": 5.223616922361693, + "grad_norm": 9.013343811035156, + "learning_rate": 9.988769998877e-06, + "loss": 0.338, + "step": 11236 + }, + { + "epoch": 5.224081822408182, + "grad_norm": 5.1486310958862305, + "learning_rate": 9.9887689988769e-06, + "loss": 0.2114, + "step": 11237 + }, + { + "epoch": 5.224546722454672, + "grad_norm": 4.148470401763916, + "learning_rate": 9.988767998876801e-06, + "loss": 0.116, + "step": 11238 + }, + { + "epoch": 5.225011622501162, + "grad_norm": 5.206708908081055, + "learning_rate": 9.988766998876702e-06, + "loss": 0.1867, + "step": 11239 + }, + { + "epoch": 5.225476522547652, + "grad_norm": 5.62797212600708, + "learning_rate": 9.9887659988766e-06, + "loss": 0.1616, + "step": 11240 + }, + { + "epoch": 5.2259414225941425, + "grad_norm": 7.6237335205078125, + "learning_rate": 9.988764998876501e-06, + "loss": 0.1078, + "step": 11241 + }, + { + "epoch": 5.226406322640632, + "grad_norm": 7.3529791831970215, + "learning_rate": 9.9887639988764e-06, + "loss": 0.4039, + "step": 11242 + }, + { + "epoch": 5.226871222687122, + "grad_norm": 6.70905876159668, + "learning_rate": 9.9887629988763e-06, + "loss": 0.4628, + "step": 11243 + }, + { + "epoch": 5.227336122733612, + "grad_norm": 5.741565227508545, + "learning_rate": 9.9887619988762e-06, + "loss": 0.2486, + "step": 11244 + }, + { + "epoch": 5.227801022780103, + "grad_norm": 6.280969619750977, + "learning_rate": 9.9887609988761e-06, + "loss": 0.1876, + "step": 11245 + }, + { + "epoch": 5.228265922826592, + "grad_norm": 3.5481033325195312, + "learning_rate": 9.988759998876001e-06, + "loss": 0.1374, + "step": 11246 + }, + { + "epoch": 5.228730822873082, + "grad_norm": 5.601710796356201, + "learning_rate": 9.9887589988759e-06, + "loss": 0.2617, + "step": 11247 + }, + { + "epoch": 5.229195722919572, + "grad_norm": 4.651913166046143, + "learning_rate": 9.988757998875801e-06, + "loss": 0.1733, + "step": 11248 + }, + { + "epoch": 5.229660622966063, + "grad_norm": 3.7965681552886963, + "learning_rate": 9.9887569988757e-06, + "loss": 0.1505, + "step": 11249 + }, + { + "epoch": 5.2301255230125525, + "grad_norm": 4.922436237335205, + "learning_rate": 9.9887559988756e-06, + "loss": 0.1742, + "step": 11250 + }, + { + "epoch": 5.230590423059042, + "grad_norm": 8.500234603881836, + "learning_rate": 9.988754998875501e-06, + "loss": 0.5883, + "step": 11251 + }, + { + "epoch": 5.231055323105532, + "grad_norm": 6.172390937805176, + "learning_rate": 9.988753998875402e-06, + "loss": 0.2745, + "step": 11252 + }, + { + "epoch": 5.231520223152022, + "grad_norm": 7.641575813293457, + "learning_rate": 9.988752998875301e-06, + "loss": 0.3302, + "step": 11253 + }, + { + "epoch": 5.231985123198513, + "grad_norm": 4.8586106300354, + "learning_rate": 9.9887519988752e-06, + "loss": 0.1747, + "step": 11254 + }, + { + "epoch": 5.232450023245002, + "grad_norm": 4.392371654510498, + "learning_rate": 9.9887509988751e-06, + "loss": 0.1493, + "step": 11255 + }, + { + "epoch": 5.232914923291492, + "grad_norm": 5.735185623168945, + "learning_rate": 9.988749998875e-06, + "loss": 0.4229, + "step": 11256 + }, + { + "epoch": 5.233379823337982, + "grad_norm": 6.719430923461914, + "learning_rate": 9.9887489988749e-06, + "loss": 0.2063, + "step": 11257 + }, + { + "epoch": 5.233844723384473, + "grad_norm": 6.995728969573975, + "learning_rate": 9.9887479988748e-06, + "loss": 0.3135, + "step": 11258 + }, + { + "epoch": 5.2343096234309625, + "grad_norm": 5.916481018066406, + "learning_rate": 9.988746998874701e-06, + "loss": 0.3237, + "step": 11259 + }, + { + "epoch": 5.234774523477452, + "grad_norm": 9.325631141662598, + "learning_rate": 9.9887459988746e-06, + "loss": 0.5369, + "step": 11260 + }, + { + "epoch": 5.235239423523942, + "grad_norm": 4.738605499267578, + "learning_rate": 9.988744998874501e-06, + "loss": 0.1162, + "step": 11261 + }, + { + "epoch": 5.235704323570433, + "grad_norm": 5.920001983642578, + "learning_rate": 9.9887439988744e-06, + "loss": 0.3604, + "step": 11262 + }, + { + "epoch": 5.236169223616923, + "grad_norm": 3.525059938430786, + "learning_rate": 9.9887429988743e-06, + "loss": 0.1024, + "step": 11263 + }, + { + "epoch": 5.236634123663412, + "grad_norm": 3.9877820014953613, + "learning_rate": 9.988741998874201e-06, + "loss": 0.1504, + "step": 11264 + }, + { + "epoch": 5.237099023709902, + "grad_norm": 3.4977357387542725, + "learning_rate": 9.9887409988741e-06, + "loss": 0.1253, + "step": 11265 + }, + { + "epoch": 5.237563923756392, + "grad_norm": 4.156935214996338, + "learning_rate": 9.988739998874001e-06, + "loss": 0.1549, + "step": 11266 + }, + { + "epoch": 5.238028823802883, + "grad_norm": 8.608430862426758, + "learning_rate": 9.9887389988739e-06, + "loss": 0.2834, + "step": 11267 + }, + { + "epoch": 5.2384937238493725, + "grad_norm": 4.823823928833008, + "learning_rate": 9.9887379988738e-06, + "loss": 0.1442, + "step": 11268 + }, + { + "epoch": 5.238958623895862, + "grad_norm": 7.750701427459717, + "learning_rate": 9.9887369988737e-06, + "loss": 0.3264, + "step": 11269 + }, + { + "epoch": 5.239423523942352, + "grad_norm": 3.6072885990142822, + "learning_rate": 9.9887359988736e-06, + "loss": 0.0926, + "step": 11270 + }, + { + "epoch": 5.239888423988843, + "grad_norm": 5.455996990203857, + "learning_rate": 9.988734998873501e-06, + "loss": 0.1257, + "step": 11271 + }, + { + "epoch": 5.240353324035333, + "grad_norm": 5.618294715881348, + "learning_rate": 9.988733998873402e-06, + "loss": 0.2273, + "step": 11272 + }, + { + "epoch": 5.240818224081822, + "grad_norm": 7.025243759155273, + "learning_rate": 9.9887329988733e-06, + "loss": 0.4037, + "step": 11273 + }, + { + "epoch": 5.241283124128312, + "grad_norm": 5.270616054534912, + "learning_rate": 9.988731998873201e-06, + "loss": 0.2346, + "step": 11274 + }, + { + "epoch": 5.241748024174802, + "grad_norm": 5.047996997833252, + "learning_rate": 9.9887309988731e-06, + "loss": 0.2726, + "step": 11275 + }, + { + "epoch": 5.242212924221293, + "grad_norm": 10.023916244506836, + "learning_rate": 9.988729998873e-06, + "loss": 0.6421, + "step": 11276 + }, + { + "epoch": 5.2426778242677825, + "grad_norm": 6.351077556610107, + "learning_rate": 9.988728998872901e-06, + "loss": 0.2607, + "step": 11277 + }, + { + "epoch": 5.243142724314272, + "grad_norm": 6.762391567230225, + "learning_rate": 9.9887279988728e-06, + "loss": 0.3626, + "step": 11278 + }, + { + "epoch": 5.243607624360762, + "grad_norm": 4.583886623382568, + "learning_rate": 9.988726998872701e-06, + "loss": 0.1761, + "step": 11279 + }, + { + "epoch": 5.244072524407253, + "grad_norm": 5.1529035568237305, + "learning_rate": 9.9887259988726e-06, + "loss": 0.2655, + "step": 11280 + }, + { + "epoch": 5.244537424453743, + "grad_norm": 18.37993621826172, + "learning_rate": 9.9887249988725e-06, + "loss": 0.075, + "step": 11281 + }, + { + "epoch": 5.245002324500232, + "grad_norm": 5.506463527679443, + "learning_rate": 9.988723998872401e-06, + "loss": 0.252, + "step": 11282 + }, + { + "epoch": 5.245467224546722, + "grad_norm": 5.7914581298828125, + "learning_rate": 9.9887229988723e-06, + "loss": 0.4507, + "step": 11283 + }, + { + "epoch": 5.245932124593212, + "grad_norm": 5.318521976470947, + "learning_rate": 9.988721998872201e-06, + "loss": 0.2006, + "step": 11284 + }, + { + "epoch": 5.246397024639703, + "grad_norm": 7.511234283447266, + "learning_rate": 9.988720998872102e-06, + "loss": 0.4445, + "step": 11285 + }, + { + "epoch": 5.2468619246861925, + "grad_norm": 4.831170082092285, + "learning_rate": 9.988719998872e-06, + "loss": 0.2242, + "step": 11286 + }, + { + "epoch": 5.247326824732682, + "grad_norm": 5.0571770668029785, + "learning_rate": 9.9887189988719e-06, + "loss": 0.1758, + "step": 11287 + }, + { + "epoch": 5.247791724779172, + "grad_norm": 4.789941310882568, + "learning_rate": 9.9887179988718e-06, + "loss": 0.1084, + "step": 11288 + }, + { + "epoch": 5.248256624825663, + "grad_norm": 5.290782928466797, + "learning_rate": 9.988716998871701e-06, + "loss": 0.1735, + "step": 11289 + }, + { + "epoch": 5.248721524872153, + "grad_norm": 4.357005596160889, + "learning_rate": 9.988715998871602e-06, + "loss": 0.163, + "step": 11290 + }, + { + "epoch": 5.249186424918642, + "grad_norm": 6.35297155380249, + "learning_rate": 9.9887149988715e-06, + "loss": 0.2464, + "step": 11291 + }, + { + "epoch": 5.249651324965132, + "grad_norm": 7.363187789916992, + "learning_rate": 9.988713998871401e-06, + "loss": 0.2001, + "step": 11292 + }, + { + "epoch": 5.250116225011623, + "grad_norm": 4.694131851196289, + "learning_rate": 9.9887129988713e-06, + "loss": 0.3088, + "step": 11293 + }, + { + "epoch": 5.250581125058113, + "grad_norm": 4.266948699951172, + "learning_rate": 9.9887119988712e-06, + "loss": 0.1761, + "step": 11294 + }, + { + "epoch": 5.2510460251046025, + "grad_norm": 2.3711085319519043, + "learning_rate": 9.988710998871101e-06, + "loss": 0.0614, + "step": 11295 + }, + { + "epoch": 5.251510925151092, + "grad_norm": 8.07818603515625, + "learning_rate": 9.988709998871e-06, + "loss": 0.3547, + "step": 11296 + }, + { + "epoch": 5.251975825197582, + "grad_norm": 6.063901901245117, + "learning_rate": 9.988708998870901e-06, + "loss": 0.3046, + "step": 11297 + }, + { + "epoch": 5.252440725244073, + "grad_norm": 4.078821659088135, + "learning_rate": 9.9887079988708e-06, + "loss": 0.1638, + "step": 11298 + }, + { + "epoch": 5.252905625290563, + "grad_norm": 4.919605731964111, + "learning_rate": 9.9887069988707e-06, + "loss": 0.1529, + "step": 11299 + }, + { + "epoch": 5.253370525337052, + "grad_norm": 5.7946672439575195, + "learning_rate": 9.9887059988706e-06, + "loss": 0.2622, + "step": 11300 + }, + { + "epoch": 5.253835425383542, + "grad_norm": 4.620631694793701, + "learning_rate": 9.9887049988705e-06, + "loss": 0.2021, + "step": 11301 + }, + { + "epoch": 5.254300325430033, + "grad_norm": 7.6417460441589355, + "learning_rate": 9.988703998870401e-06, + "loss": 0.2598, + "step": 11302 + }, + { + "epoch": 5.254765225476523, + "grad_norm": 3.9041831493377686, + "learning_rate": 9.988702998870302e-06, + "loss": 0.0928, + "step": 11303 + }, + { + "epoch": 5.2552301255230125, + "grad_norm": 3.423391819000244, + "learning_rate": 9.9887019988702e-06, + "loss": 0.0551, + "step": 11304 + }, + { + "epoch": 5.255695025569502, + "grad_norm": 5.830896854400635, + "learning_rate": 9.988700998870101e-06, + "loss": 0.3955, + "step": 11305 + }, + { + "epoch": 5.256159925615993, + "grad_norm": 4.270524978637695, + "learning_rate": 9.98869999887e-06, + "loss": 0.1147, + "step": 11306 + }, + { + "epoch": 5.256624825662483, + "grad_norm": 6.809309959411621, + "learning_rate": 9.988698998869901e-06, + "loss": 0.4227, + "step": 11307 + }, + { + "epoch": 5.257089725708973, + "grad_norm": 9.400073051452637, + "learning_rate": 9.988697998869802e-06, + "loss": 0.5065, + "step": 11308 + }, + { + "epoch": 5.257554625755462, + "grad_norm": 5.0789666175842285, + "learning_rate": 9.9886969988697e-06, + "loss": 0.2389, + "step": 11309 + }, + { + "epoch": 5.258019525801952, + "grad_norm": 2.593022584915161, + "learning_rate": 9.988695998869601e-06, + "loss": 0.0695, + "step": 11310 + }, + { + "epoch": 5.258484425848443, + "grad_norm": 5.2460761070251465, + "learning_rate": 9.9886949988695e-06, + "loss": 0.254, + "step": 11311 + }, + { + "epoch": 5.258949325894933, + "grad_norm": 3.118698835372925, + "learning_rate": 9.9886939988694e-06, + "loss": 0.0757, + "step": 11312 + }, + { + "epoch": 5.2594142259414225, + "grad_norm": 2.5115272998809814, + "learning_rate": 9.9886929988693e-06, + "loss": 0.0585, + "step": 11313 + }, + { + "epoch": 5.259879125987912, + "grad_norm": 6.62284517288208, + "learning_rate": 9.9886919988692e-06, + "loss": 0.4154, + "step": 11314 + }, + { + "epoch": 5.260344026034403, + "grad_norm": 4.407756328582764, + "learning_rate": 9.988690998869101e-06, + "loss": 0.1244, + "step": 11315 + }, + { + "epoch": 5.260808926080893, + "grad_norm": 6.421946048736572, + "learning_rate": 9.988689998869002e-06, + "loss": 0.3972, + "step": 11316 + }, + { + "epoch": 5.261273826127383, + "grad_norm": 5.008045196533203, + "learning_rate": 9.9886889988689e-06, + "loss": 0.1853, + "step": 11317 + }, + { + "epoch": 5.261738726173872, + "grad_norm": 2.9609477519989014, + "learning_rate": 9.988687998868801e-06, + "loss": 0.1026, + "step": 11318 + }, + { + "epoch": 5.262203626220362, + "grad_norm": 4.095687389373779, + "learning_rate": 9.9886869988687e-06, + "loss": 0.1866, + "step": 11319 + }, + { + "epoch": 5.262668526266853, + "grad_norm": 3.8819987773895264, + "learning_rate": 9.988685998868601e-06, + "loss": 0.2005, + "step": 11320 + }, + { + "epoch": 5.263133426313343, + "grad_norm": 6.031750202178955, + "learning_rate": 9.988684998868502e-06, + "loss": 0.2609, + "step": 11321 + }, + { + "epoch": 5.2635983263598325, + "grad_norm": 4.1515655517578125, + "learning_rate": 9.9886839988684e-06, + "loss": 0.2112, + "step": 11322 + }, + { + "epoch": 5.264063226406322, + "grad_norm": 5.306697845458984, + "learning_rate": 9.988682998868301e-06, + "loss": 0.2474, + "step": 11323 + }, + { + "epoch": 5.264528126452813, + "grad_norm": 4.812448024749756, + "learning_rate": 9.9886819988682e-06, + "loss": 0.1312, + "step": 11324 + }, + { + "epoch": 5.264993026499303, + "grad_norm": 7.489837646484375, + "learning_rate": 9.9886809988681e-06, + "loss": 0.4088, + "step": 11325 + }, + { + "epoch": 5.265457926545793, + "grad_norm": 4.731039047241211, + "learning_rate": 9.988679998868e-06, + "loss": 0.1758, + "step": 11326 + }, + { + "epoch": 5.265922826592282, + "grad_norm": 5.092525959014893, + "learning_rate": 9.9886789988679e-06, + "loss": 0.1713, + "step": 11327 + }, + { + "epoch": 5.266387726638772, + "grad_norm": 4.248818397521973, + "learning_rate": 9.988677998867801e-06, + "loss": 0.2409, + "step": 11328 + }, + { + "epoch": 5.266852626685263, + "grad_norm": 6.0682477951049805, + "learning_rate": 9.988676998867702e-06, + "loss": 0.1985, + "step": 11329 + }, + { + "epoch": 5.267317526731753, + "grad_norm": 5.200296878814697, + "learning_rate": 9.9886759988676e-06, + "loss": 0.247, + "step": 11330 + }, + { + "epoch": 5.2677824267782425, + "grad_norm": 4.649752140045166, + "learning_rate": 9.9886749988675e-06, + "loss": 0.1329, + "step": 11331 + }, + { + "epoch": 5.268247326824732, + "grad_norm": 5.783690452575684, + "learning_rate": 9.9886739988674e-06, + "loss": 0.2677, + "step": 11332 + }, + { + "epoch": 5.268712226871223, + "grad_norm": 5.360842704772949, + "learning_rate": 9.988672998867301e-06, + "loss": 0.2382, + "step": 11333 + }, + { + "epoch": 5.269177126917713, + "grad_norm": 4.936532020568848, + "learning_rate": 9.988671998867202e-06, + "loss": 0.2391, + "step": 11334 + }, + { + "epoch": 5.269642026964203, + "grad_norm": 5.175439357757568, + "learning_rate": 9.9886709988671e-06, + "loss": 0.2101, + "step": 11335 + }, + { + "epoch": 5.270106927010692, + "grad_norm": 3.0586376190185547, + "learning_rate": 9.988669998867001e-06, + "loss": 0.1015, + "step": 11336 + }, + { + "epoch": 5.270571827057183, + "grad_norm": 5.459699630737305, + "learning_rate": 9.9886689988669e-06, + "loss": 0.1994, + "step": 11337 + }, + { + "epoch": 5.271036727103673, + "grad_norm": 5.494594573974609, + "learning_rate": 9.988667998866801e-06, + "loss": 0.2836, + "step": 11338 + }, + { + "epoch": 5.271501627150163, + "grad_norm": 5.834506511688232, + "learning_rate": 9.9886669988667e-06, + "loss": 0.2575, + "step": 11339 + }, + { + "epoch": 5.2719665271966525, + "grad_norm": 4.308287620544434, + "learning_rate": 9.9886659988666e-06, + "loss": 0.1927, + "step": 11340 + }, + { + "epoch": 5.272431427243143, + "grad_norm": 4.737879276275635, + "learning_rate": 9.988664998866501e-06, + "loss": 0.3048, + "step": 11341 + }, + { + "epoch": 5.272896327289633, + "grad_norm": 5.044849395751953, + "learning_rate": 9.9886639988664e-06, + "loss": 0.1687, + "step": 11342 + }, + { + "epoch": 5.273361227336123, + "grad_norm": 4.815458297729492, + "learning_rate": 9.9886629988663e-06, + "loss": 0.1823, + "step": 11343 + }, + { + "epoch": 5.273826127382613, + "grad_norm": 3.8577358722686768, + "learning_rate": 9.9886619988662e-06, + "loss": 0.1273, + "step": 11344 + }, + { + "epoch": 5.274291027429102, + "grad_norm": 3.5323469638824463, + "learning_rate": 9.9886609988661e-06, + "loss": 0.131, + "step": 11345 + }, + { + "epoch": 5.274755927475593, + "grad_norm": 4.51802396774292, + "learning_rate": 9.988659998866001e-06, + "loss": 0.1964, + "step": 11346 + }, + { + "epoch": 5.275220827522083, + "grad_norm": 6.024505615234375, + "learning_rate": 9.988658998865902e-06, + "loss": 0.3405, + "step": 11347 + }, + { + "epoch": 5.275685727568573, + "grad_norm": 6.382872104644775, + "learning_rate": 9.9886579988658e-06, + "loss": 0.2717, + "step": 11348 + }, + { + "epoch": 5.2761506276150625, + "grad_norm": 4.897765159606934, + "learning_rate": 9.988656998865701e-06, + "loss": 0.1818, + "step": 11349 + }, + { + "epoch": 5.276615527661553, + "grad_norm": 5.253246784210205, + "learning_rate": 9.9886559988656e-06, + "loss": 0.2199, + "step": 11350 + }, + { + "epoch": 5.277080427708043, + "grad_norm": 3.9910943508148193, + "learning_rate": 9.988654998865501e-06, + "loss": 0.094, + "step": 11351 + }, + { + "epoch": 5.277545327754533, + "grad_norm": 8.389840126037598, + "learning_rate": 9.9886539988654e-06, + "loss": 0.1112, + "step": 11352 + }, + { + "epoch": 5.278010227801023, + "grad_norm": 8.216795921325684, + "learning_rate": 9.9886529988653e-06, + "loss": 0.3552, + "step": 11353 + }, + { + "epoch": 5.278475127847512, + "grad_norm": 5.645856857299805, + "learning_rate": 9.988651998865201e-06, + "loss": 0.2244, + "step": 11354 + }, + { + "epoch": 5.278940027894003, + "grad_norm": 5.31505012512207, + "learning_rate": 9.9886509988651e-06, + "loss": 0.2577, + "step": 11355 + }, + { + "epoch": 5.279404927940493, + "grad_norm": 6.825138092041016, + "learning_rate": 9.988649998865001e-06, + "loss": 0.1662, + "step": 11356 + }, + { + "epoch": 5.279869827986983, + "grad_norm": 4.737704753875732, + "learning_rate": 9.9886489988649e-06, + "loss": 0.1061, + "step": 11357 + }, + { + "epoch": 5.2803347280334725, + "grad_norm": 10.4564208984375, + "learning_rate": 9.9886479988648e-06, + "loss": 0.4916, + "step": 11358 + }, + { + "epoch": 5.280799628079963, + "grad_norm": 3.316824436187744, + "learning_rate": 9.988646998864701e-06, + "loss": 0.0927, + "step": 11359 + }, + { + "epoch": 5.281264528126453, + "grad_norm": 7.309389591217041, + "learning_rate": 9.988645998864602e-06, + "loss": 0.2432, + "step": 11360 + }, + { + "epoch": 5.281729428172943, + "grad_norm": 7.0967559814453125, + "learning_rate": 9.9886449988645e-06, + "loss": 0.2784, + "step": 11361 + }, + { + "epoch": 5.282194328219433, + "grad_norm": 5.003580570220947, + "learning_rate": 9.9886439988644e-06, + "loss": 0.1667, + "step": 11362 + }, + { + "epoch": 5.282659228265922, + "grad_norm": 6.307856559753418, + "learning_rate": 9.9886429988643e-06, + "loss": 0.2157, + "step": 11363 + }, + { + "epoch": 5.283124128312413, + "grad_norm": 3.1351494789123535, + "learning_rate": 9.988641998864201e-06, + "loss": 0.112, + "step": 11364 + }, + { + "epoch": 5.283589028358903, + "grad_norm": 3.5027248859405518, + "learning_rate": 9.9886409988641e-06, + "loss": 0.0986, + "step": 11365 + }, + { + "epoch": 5.284053928405393, + "grad_norm": 8.319189071655273, + "learning_rate": 9.988639998864e-06, + "loss": 0.3639, + "step": 11366 + }, + { + "epoch": 5.2845188284518825, + "grad_norm": NaN, + "learning_rate": 9.988639998864e-06, + "loss": 0.3052, + "step": 11367 + }, + { + "epoch": 5.284983728498373, + "grad_norm": 6.291802883148193, + "learning_rate": 9.988638998863901e-06, + "loss": 0.3063, + "step": 11368 + }, + { + "epoch": 5.285448628544863, + "grad_norm": 4.8774542808532715, + "learning_rate": 9.9886379988638e-06, + "loss": 0.2085, + "step": 11369 + }, + { + "epoch": 5.285913528591353, + "grad_norm": 4.875803470611572, + "learning_rate": 9.988636998863701e-06, + "loss": 0.225, + "step": 11370 + }, + { + "epoch": 5.286378428637843, + "grad_norm": 2.1277496814727783, + "learning_rate": 9.9886359988636e-06, + "loss": 0.0286, + "step": 11371 + }, + { + "epoch": 5.286843328684333, + "grad_norm": 4.869632244110107, + "learning_rate": 9.9886349988635e-06, + "loss": 0.1803, + "step": 11372 + }, + { + "epoch": 5.287308228730823, + "grad_norm": 6.588778972625732, + "learning_rate": 9.988633998863401e-06, + "loss": 0.2591, + "step": 11373 + }, + { + "epoch": 5.287773128777313, + "grad_norm": 4.665684223175049, + "learning_rate": 9.9886329988633e-06, + "loss": 0.1691, + "step": 11374 + }, + { + "epoch": 5.288238028823803, + "grad_norm": 4.152441024780273, + "learning_rate": 9.988631998863201e-06, + "loss": 0.1183, + "step": 11375 + }, + { + "epoch": 5.2887029288702925, + "grad_norm": 3.5535945892333984, + "learning_rate": 9.9886309988631e-06, + "loss": 0.148, + "step": 11376 + }, + { + "epoch": 5.289167828916783, + "grad_norm": 3.9374170303344727, + "learning_rate": 9.988629998863e-06, + "loss": 0.1373, + "step": 11377 + }, + { + "epoch": 5.289632728963273, + "grad_norm": 6.800112724304199, + "learning_rate": 9.988628998862901e-06, + "loss": 0.4457, + "step": 11378 + }, + { + "epoch": 5.290097629009763, + "grad_norm": 5.370677947998047, + "learning_rate": 9.988627998862802e-06, + "loss": 0.3039, + "step": 11379 + }, + { + "epoch": 5.290562529056253, + "grad_norm": 5.848599910736084, + "learning_rate": 9.9886269988627e-06, + "loss": 0.3494, + "step": 11380 + }, + { + "epoch": 5.291027429102743, + "grad_norm": 4.796615123748779, + "learning_rate": 9.988625998862601e-06, + "loss": 0.1928, + "step": 11381 + }, + { + "epoch": 5.291492329149233, + "grad_norm": 4.303055763244629, + "learning_rate": 9.9886249988625e-06, + "loss": 0.1994, + "step": 11382 + }, + { + "epoch": 5.291957229195723, + "grad_norm": 5.784064292907715, + "learning_rate": 9.988623998862401e-06, + "loss": 0.1654, + "step": 11383 + }, + { + "epoch": 5.292422129242213, + "grad_norm": 5.1817522048950195, + "learning_rate": 9.9886229988623e-06, + "loss": 0.2936, + "step": 11384 + }, + { + "epoch": 5.292887029288703, + "grad_norm": 4.2649149894714355, + "learning_rate": 9.9886219988622e-06, + "loss": 0.1967, + "step": 11385 + }, + { + "epoch": 5.293351929335193, + "grad_norm": 7.174921989440918, + "learning_rate": 9.988620998862101e-06, + "loss": 0.4959, + "step": 11386 + }, + { + "epoch": 5.293816829381683, + "grad_norm": 5.532524108886719, + "learning_rate": 9.988619998862e-06, + "loss": 0.2412, + "step": 11387 + }, + { + "epoch": 5.294281729428173, + "grad_norm": 5.143608570098877, + "learning_rate": 9.988618998861901e-06, + "loss": 0.1686, + "step": 11388 + }, + { + "epoch": 5.294746629474663, + "grad_norm": 5.43477725982666, + "learning_rate": 9.9886179988618e-06, + "loss": 0.1996, + "step": 11389 + }, + { + "epoch": 5.295211529521153, + "grad_norm": 8.350754737854004, + "learning_rate": 9.9886169988617e-06, + "loss": 0.2467, + "step": 11390 + }, + { + "epoch": 5.295676429567643, + "grad_norm": 3.074315309524536, + "learning_rate": 9.988615998861601e-06, + "loss": 0.0836, + "step": 11391 + }, + { + "epoch": 5.296141329614133, + "grad_norm": 5.7832207679748535, + "learning_rate": 9.988614998861502e-06, + "loss": 0.1761, + "step": 11392 + }, + { + "epoch": 5.296606229660623, + "grad_norm": 6.159127712249756, + "learning_rate": 9.988613998861401e-06, + "loss": 0.2514, + "step": 11393 + }, + { + "epoch": 5.297071129707113, + "grad_norm": 3.2296018600463867, + "learning_rate": 9.988612998861302e-06, + "loss": 0.1582, + "step": 11394 + }, + { + "epoch": 5.297536029753603, + "grad_norm": 5.896690845489502, + "learning_rate": 9.9886119988612e-06, + "loss": 0.1716, + "step": 11395 + }, + { + "epoch": 5.298000929800093, + "grad_norm": 6.739692211151123, + "learning_rate": 9.9886109988611e-06, + "loss": 0.3257, + "step": 11396 + }, + { + "epoch": 5.298465829846583, + "grad_norm": 7.128353595733643, + "learning_rate": 9.988609998861e-06, + "loss": 0.3888, + "step": 11397 + }, + { + "epoch": 5.298930729893073, + "grad_norm": 5.12960147857666, + "learning_rate": 9.9886089988609e-06, + "loss": 0.1731, + "step": 11398 + }, + { + "epoch": 5.299395629939563, + "grad_norm": 8.75456714630127, + "learning_rate": 9.988607998860801e-06, + "loss": 1.0301, + "step": 11399 + }, + { + "epoch": 5.299860529986053, + "grad_norm": 6.423065185546875, + "learning_rate": 9.9886069988607e-06, + "loss": 0.2291, + "step": 11400 + }, + { + "epoch": 5.300325430032543, + "grad_norm": 3.693483352661133, + "learning_rate": 9.988605998860601e-06, + "loss": 0.159, + "step": 11401 + }, + { + "epoch": 5.300790330079033, + "grad_norm": 6.277737140655518, + "learning_rate": 9.9886049988605e-06, + "loss": 0.0933, + "step": 11402 + }, + { + "epoch": 5.301255230125523, + "grad_norm": 5.813694477081299, + "learning_rate": 9.9886039988604e-06, + "loss": 0.352, + "step": 11403 + }, + { + "epoch": 5.301720130172013, + "grad_norm": 5.311845779418945, + "learning_rate": 9.988602998860301e-06, + "loss": 0.2127, + "step": 11404 + }, + { + "epoch": 5.302185030218503, + "grad_norm": 6.1414923667907715, + "learning_rate": 9.988601998860202e-06, + "loss": 0.4066, + "step": 11405 + }, + { + "epoch": 5.302649930264993, + "grad_norm": 6.6968817710876465, + "learning_rate": 9.988600998860101e-06, + "loss": 0.4553, + "step": 11406 + }, + { + "epoch": 5.303114830311483, + "grad_norm": 5.422366142272949, + "learning_rate": 9.98859999886e-06, + "loss": 0.1506, + "step": 11407 + }, + { + "epoch": 5.303579730357973, + "grad_norm": 5.239831924438477, + "learning_rate": 9.9885989988599e-06, + "loss": 0.2981, + "step": 11408 + }, + { + "epoch": 5.304044630404463, + "grad_norm": 4.88203763961792, + "learning_rate": 9.9885979988598e-06, + "loss": 0.1508, + "step": 11409 + }, + { + "epoch": 5.304509530450953, + "grad_norm": 4.713435173034668, + "learning_rate": 9.9885969988597e-06, + "loss": 0.1579, + "step": 11410 + }, + { + "epoch": 5.304974430497443, + "grad_norm": 5.862536907196045, + "learning_rate": 9.988595998859601e-06, + "loss": 0.3228, + "step": 11411 + }, + { + "epoch": 5.305439330543933, + "grad_norm": 3.339515447616577, + "learning_rate": 9.988594998859502e-06, + "loss": 0.2053, + "step": 11412 + }, + { + "epoch": 5.305904230590423, + "grad_norm": 4.76357889175415, + "learning_rate": 9.9885939988594e-06, + "loss": 0.1547, + "step": 11413 + }, + { + "epoch": 5.306369130636913, + "grad_norm": 7.781790256500244, + "learning_rate": 9.988592998859301e-06, + "loss": 0.4158, + "step": 11414 + }, + { + "epoch": 5.306834030683403, + "grad_norm": 3.687427043914795, + "learning_rate": 9.9885919988592e-06, + "loss": 0.0966, + "step": 11415 + }, + { + "epoch": 5.3072989307298934, + "grad_norm": 5.8535685539245605, + "learning_rate": 9.9885909988591e-06, + "loss": 0.2808, + "step": 11416 + }, + { + "epoch": 5.307763830776383, + "grad_norm": 5.606578826904297, + "learning_rate": 9.988589998859001e-06, + "loss": 0.2605, + "step": 11417 + }, + { + "epoch": 5.308228730822873, + "grad_norm": 4.981384754180908, + "learning_rate": 9.9885889988589e-06, + "loss": 0.1025, + "step": 11418 + }, + { + "epoch": 5.308693630869363, + "grad_norm": 4.992426872253418, + "learning_rate": 9.988587998858801e-06, + "loss": 0.2223, + "step": 11419 + }, + { + "epoch": 5.3091585309158535, + "grad_norm": 2.9984986782073975, + "learning_rate": 9.9885869988587e-06, + "loss": 0.0966, + "step": 11420 + }, + { + "epoch": 5.309623430962343, + "grad_norm": 5.1089019775390625, + "learning_rate": 9.9885859988586e-06, + "loss": 0.1526, + "step": 11421 + }, + { + "epoch": 5.310088331008833, + "grad_norm": 7.29833459854126, + "learning_rate": 9.9885849988585e-06, + "loss": 0.1707, + "step": 11422 + }, + { + "epoch": 5.310553231055323, + "grad_norm": 9.416308403015137, + "learning_rate": 9.9885839988584e-06, + "loss": 0.3697, + "step": 11423 + }, + { + "epoch": 5.311018131101813, + "grad_norm": 5.300662994384766, + "learning_rate": 9.988582998858301e-06, + "loss": 0.2926, + "step": 11424 + }, + { + "epoch": 5.3114830311483034, + "grad_norm": 10.526965141296387, + "learning_rate": 9.988581998858202e-06, + "loss": 0.1493, + "step": 11425 + }, + { + "epoch": 5.311947931194793, + "grad_norm": 7.320295333862305, + "learning_rate": 9.9885809988581e-06, + "loss": 0.1023, + "step": 11426 + }, + { + "epoch": 5.312412831241283, + "grad_norm": 11.074324607849121, + "learning_rate": 9.988579998858001e-06, + "loss": 0.1122, + "step": 11427 + }, + { + "epoch": 5.312877731287773, + "grad_norm": 4.438880443572998, + "learning_rate": 9.9885789988579e-06, + "loss": 0.2415, + "step": 11428 + }, + { + "epoch": 5.3133426313342635, + "grad_norm": 4.751554489135742, + "learning_rate": 9.9885779988578e-06, + "loss": 0.1906, + "step": 11429 + }, + { + "epoch": 5.313807531380753, + "grad_norm": 4.314353942871094, + "learning_rate": 9.988576998857701e-06, + "loss": 0.1166, + "step": 11430 + }, + { + "epoch": 5.314272431427243, + "grad_norm": 3.6077253818511963, + "learning_rate": 9.9885759988576e-06, + "loss": 0.1313, + "step": 11431 + }, + { + "epoch": 5.314737331473733, + "grad_norm": 4.147696018218994, + "learning_rate": 9.988574998857501e-06, + "loss": 0.1855, + "step": 11432 + }, + { + "epoch": 5.315202231520223, + "grad_norm": 11.354342460632324, + "learning_rate": 9.9885739988574e-06, + "loss": 0.4141, + "step": 11433 + }, + { + "epoch": 5.315667131566713, + "grad_norm": 5.5447096824646, + "learning_rate": 9.9885729988573e-06, + "loss": 0.3851, + "step": 11434 + }, + { + "epoch": 5.316132031613203, + "grad_norm": 4.81862735748291, + "learning_rate": 9.9885719988572e-06, + "loss": 0.2085, + "step": 11435 + }, + { + "epoch": 5.316596931659693, + "grad_norm": 6.0084404945373535, + "learning_rate": 9.9885709988571e-06, + "loss": 0.3488, + "step": 11436 + }, + { + "epoch": 5.317061831706183, + "grad_norm": 5.7704758644104, + "learning_rate": 9.988569998857001e-06, + "loss": 0.1165, + "step": 11437 + }, + { + "epoch": 5.3175267317526735, + "grad_norm": 5.067534923553467, + "learning_rate": 9.988568998856902e-06, + "loss": 0.2498, + "step": 11438 + }, + { + "epoch": 5.317991631799163, + "grad_norm": 3.8502535820007324, + "learning_rate": 9.9885679988568e-06, + "loss": 0.0975, + "step": 11439 + }, + { + "epoch": 5.318456531845653, + "grad_norm": 6.7389607429504395, + "learning_rate": 9.9885669988567e-06, + "loss": 0.259, + "step": 11440 + }, + { + "epoch": 5.318921431892143, + "grad_norm": 5.445252418518066, + "learning_rate": 9.9885659988566e-06, + "loss": 0.2185, + "step": 11441 + }, + { + "epoch": 5.319386331938633, + "grad_norm": 5.538417339324951, + "learning_rate": 9.988564998856501e-06, + "loss": 0.3264, + "step": 11442 + }, + { + "epoch": 5.319851231985123, + "grad_norm": 4.44834566116333, + "learning_rate": 9.988563998856402e-06, + "loss": 0.237, + "step": 11443 + }, + { + "epoch": 5.320316132031613, + "grad_norm": 3.8278706073760986, + "learning_rate": 9.9885629988563e-06, + "loss": 0.1102, + "step": 11444 + }, + { + "epoch": 5.320781032078103, + "grad_norm": 4.6261773109436035, + "learning_rate": 9.988561998856201e-06, + "loss": 0.2485, + "step": 11445 + }, + { + "epoch": 5.321245932124593, + "grad_norm": 4.103870391845703, + "learning_rate": 9.9885609988561e-06, + "loss": 0.1781, + "step": 11446 + }, + { + "epoch": 5.3217108321710835, + "grad_norm": 3.667344808578491, + "learning_rate": 9.988559998856e-06, + "loss": 0.1325, + "step": 11447 + }, + { + "epoch": 5.322175732217573, + "grad_norm": 6.545688629150391, + "learning_rate": 9.988558998855901e-06, + "loss": 0.3265, + "step": 11448 + }, + { + "epoch": 5.322640632264063, + "grad_norm": 17.04507064819336, + "learning_rate": 9.9885579988558e-06, + "loss": 0.5395, + "step": 11449 + }, + { + "epoch": 5.323105532310553, + "grad_norm": 9.22514820098877, + "learning_rate": 9.988556998855701e-06, + "loss": 0.2, + "step": 11450 + }, + { + "epoch": 5.323570432357044, + "grad_norm": 5.033051490783691, + "learning_rate": 9.9885559988556e-06, + "loss": 0.1557, + "step": 11451 + }, + { + "epoch": 5.324035332403533, + "grad_norm": 4.370913028717041, + "learning_rate": 9.9885549988555e-06, + "loss": 0.1674, + "step": 11452 + }, + { + "epoch": 5.324500232450023, + "grad_norm": 4.335698127746582, + "learning_rate": 9.9885539988554e-06, + "loss": 0.1478, + "step": 11453 + }, + { + "epoch": 5.324965132496513, + "grad_norm": 4.181303977966309, + "learning_rate": 9.9885529988553e-06, + "loss": 0.0897, + "step": 11454 + }, + { + "epoch": 5.325430032543003, + "grad_norm": 6.529539585113525, + "learning_rate": 9.988551998855201e-06, + "loss": 0.2553, + "step": 11455 + }, + { + "epoch": 5.3258949325894935, + "grad_norm": 9.346054077148438, + "learning_rate": 9.988550998855102e-06, + "loss": 0.2752, + "step": 11456 + }, + { + "epoch": 5.326359832635983, + "grad_norm": 4.396026611328125, + "learning_rate": 9.988549998855e-06, + "loss": 0.1483, + "step": 11457 + }, + { + "epoch": 5.326824732682473, + "grad_norm": 4.639784336090088, + "learning_rate": 9.988548998854901e-06, + "loss": 0.1838, + "step": 11458 + }, + { + "epoch": 5.327289632728963, + "grad_norm": 7.957401752471924, + "learning_rate": 9.9885479988548e-06, + "loss": 0.3733, + "step": 11459 + }, + { + "epoch": 5.327754532775454, + "grad_norm": 6.137650489807129, + "learning_rate": 9.988546998854701e-06, + "loss": 0.2714, + "step": 11460 + }, + { + "epoch": 5.328219432821943, + "grad_norm": 4.14930534362793, + "learning_rate": 9.988545998854602e-06, + "loss": 0.121, + "step": 11461 + }, + { + "epoch": 5.328684332868433, + "grad_norm": 4.686304092407227, + "learning_rate": 9.9885449988545e-06, + "loss": 0.2193, + "step": 11462 + }, + { + "epoch": 5.329149232914923, + "grad_norm": 3.976921558380127, + "learning_rate": 9.988543998854401e-06, + "loss": 0.1566, + "step": 11463 + }, + { + "epoch": 5.329614132961414, + "grad_norm": 4.694478988647461, + "learning_rate": 9.9885429988543e-06, + "loss": 0.2218, + "step": 11464 + }, + { + "epoch": 5.3300790330079035, + "grad_norm": 6.74687385559082, + "learning_rate": 9.9885419988542e-06, + "loss": 0.3053, + "step": 11465 + }, + { + "epoch": 5.330543933054393, + "grad_norm": 5.058104038238525, + "learning_rate": 9.9885409988541e-06, + "loss": 0.1382, + "step": 11466 + }, + { + "epoch": 5.331008833100883, + "grad_norm": 6.16771936416626, + "learning_rate": 9.988539998854e-06, + "loss": 0.3378, + "step": 11467 + }, + { + "epoch": 5.331473733147373, + "grad_norm": 3.6866812705993652, + "learning_rate": 9.988538998853901e-06, + "loss": 0.1354, + "step": 11468 + }, + { + "epoch": 5.331938633193864, + "grad_norm": 4.217073917388916, + "learning_rate": 9.988537998853802e-06, + "loss": 0.2144, + "step": 11469 + }, + { + "epoch": 5.332403533240353, + "grad_norm": 11.105052947998047, + "learning_rate": 9.9885369988537e-06, + "loss": 0.5832, + "step": 11470 + }, + { + "epoch": 5.332868433286843, + "grad_norm": 4.895779132843018, + "learning_rate": 9.988535998853601e-06, + "loss": 0.2121, + "step": 11471 + }, + { + "epoch": 5.333333333333333, + "grad_norm": 6.044084072113037, + "learning_rate": 9.9885349988535e-06, + "loss": 0.3423, + "step": 11472 + }, + { + "epoch": 5.333798233379824, + "grad_norm": 6.801644802093506, + "learning_rate": 9.988533998853401e-06, + "loss": 0.2504, + "step": 11473 + }, + { + "epoch": 5.3342631334263135, + "grad_norm": 8.422395706176758, + "learning_rate": 9.988532998853302e-06, + "loss": 0.2596, + "step": 11474 + }, + { + "epoch": 5.334728033472803, + "grad_norm": 7.930394649505615, + "learning_rate": 9.9885319988532e-06, + "loss": 0.2828, + "step": 11475 + }, + { + "epoch": 5.335192933519293, + "grad_norm": 6.94292688369751, + "learning_rate": 9.988530998853101e-06, + "loss": 0.291, + "step": 11476 + }, + { + "epoch": 5.335657833565783, + "grad_norm": 7.571208953857422, + "learning_rate": 9.988529998853e-06, + "loss": 0.5843, + "step": 11477 + }, + { + "epoch": 5.336122733612274, + "grad_norm": 5.276540279388428, + "learning_rate": 9.988528998852901e-06, + "loss": 0.2471, + "step": 11478 + }, + { + "epoch": 5.336587633658763, + "grad_norm": 3.566516637802124, + "learning_rate": 9.9885279988528e-06, + "loss": 0.0769, + "step": 11479 + }, + { + "epoch": 5.337052533705253, + "grad_norm": 6.06998872756958, + "learning_rate": 9.9885269988527e-06, + "loss": 0.3523, + "step": 11480 + }, + { + "epoch": 5.337517433751743, + "grad_norm": 7.198936462402344, + "learning_rate": 9.988525998852601e-06, + "loss": 0.2658, + "step": 11481 + }, + { + "epoch": 5.337982333798234, + "grad_norm": 5.002989292144775, + "learning_rate": 9.9885249988525e-06, + "loss": 0.1903, + "step": 11482 + }, + { + "epoch": 5.3384472338447235, + "grad_norm": 6.445676803588867, + "learning_rate": 9.9885239988524e-06, + "loss": 0.2832, + "step": 11483 + }, + { + "epoch": 5.338912133891213, + "grad_norm": 7.800590991973877, + "learning_rate": 9.9885229988523e-06, + "loss": 0.368, + "step": 11484 + }, + { + "epoch": 5.339377033937703, + "grad_norm": 3.8075599670410156, + "learning_rate": 9.9885219988522e-06, + "loss": 0.1744, + "step": 11485 + }, + { + "epoch": 5.339841933984193, + "grad_norm": 7.64102029800415, + "learning_rate": 9.988520998852101e-06, + "loss": 0.3205, + "step": 11486 + }, + { + "epoch": 5.340306834030684, + "grad_norm": 10.291045188903809, + "learning_rate": 9.988519998852002e-06, + "loss": 0.2217, + "step": 11487 + }, + { + "epoch": 5.340771734077173, + "grad_norm": 5.172906875610352, + "learning_rate": 9.9885189988519e-06, + "loss": 0.2099, + "step": 11488 + }, + { + "epoch": 5.341236634123663, + "grad_norm": 9.908321380615234, + "learning_rate": 9.988517998851801e-06, + "loss": 0.6274, + "step": 11489 + }, + { + "epoch": 5.341701534170153, + "grad_norm": 6.044211387634277, + "learning_rate": 9.9885169988517e-06, + "loss": 0.238, + "step": 11490 + }, + { + "epoch": 5.342166434216644, + "grad_norm": 6.896169662475586, + "learning_rate": 9.988515998851601e-06, + "loss": 0.3258, + "step": 11491 + }, + { + "epoch": 5.3426313342631335, + "grad_norm": 4.215007305145264, + "learning_rate": 9.9885149988515e-06, + "loss": 0.126, + "step": 11492 + }, + { + "epoch": 5.343096234309623, + "grad_norm": 5.653061389923096, + "learning_rate": 9.9885139988514e-06, + "loss": 0.2504, + "step": 11493 + }, + { + "epoch": 5.343561134356113, + "grad_norm": 5.266270637512207, + "learning_rate": 9.988512998851301e-06, + "loss": 0.1605, + "step": 11494 + }, + { + "epoch": 5.344026034402604, + "grad_norm": 5.572588920593262, + "learning_rate": 9.9885119988512e-06, + "loss": 0.243, + "step": 11495 + }, + { + "epoch": 5.344490934449094, + "grad_norm": 5.763850688934326, + "learning_rate": 9.988510998851101e-06, + "loss": 0.2506, + "step": 11496 + }, + { + "epoch": 5.344955834495583, + "grad_norm": 5.222442150115967, + "learning_rate": 9.988509998851e-06, + "loss": 0.143, + "step": 11497 + }, + { + "epoch": 5.345420734542073, + "grad_norm": 4.217872142791748, + "learning_rate": 9.9885089988509e-06, + "loss": 0.1222, + "step": 11498 + }, + { + "epoch": 5.345885634588564, + "grad_norm": 3.4354724884033203, + "learning_rate": 9.988507998850801e-06, + "loss": 0.1077, + "step": 11499 + }, + { + "epoch": 5.346350534635054, + "grad_norm": 5.037121772766113, + "learning_rate": 9.988506998850702e-06, + "loss": 0.2648, + "step": 11500 + }, + { + "epoch": 5.3468154346815435, + "grad_norm": 5.6433258056640625, + "learning_rate": 9.9885059988506e-06, + "loss": 0.2011, + "step": 11501 + }, + { + "epoch": 5.347280334728033, + "grad_norm": 5.084986686706543, + "learning_rate": 9.988504998850501e-06, + "loss": 0.1424, + "step": 11502 + }, + { + "epoch": 5.347745234774523, + "grad_norm": 3.682866096496582, + "learning_rate": 9.9885039988504e-06, + "loss": 0.0975, + "step": 11503 + }, + { + "epoch": 5.348210134821014, + "grad_norm": 12.318397521972656, + "learning_rate": 9.9885029988503e-06, + "loss": 0.4609, + "step": 11504 + }, + { + "epoch": 5.348675034867504, + "grad_norm": 4.7003254890441895, + "learning_rate": 9.9885019988502e-06, + "loss": 0.1932, + "step": 11505 + }, + { + "epoch": 5.349139934913993, + "grad_norm": 5.988404750823975, + "learning_rate": 9.9885009988501e-06, + "loss": 0.1528, + "step": 11506 + }, + { + "epoch": 5.349604834960483, + "grad_norm": 7.154895782470703, + "learning_rate": 9.988499998850001e-06, + "loss": 0.2033, + "step": 11507 + }, + { + "epoch": 5.350069735006974, + "grad_norm": 5.483614921569824, + "learning_rate": 9.9884989988499e-06, + "loss": 0.1903, + "step": 11508 + }, + { + "epoch": 5.350534635053464, + "grad_norm": 5.86358642578125, + "learning_rate": 9.988497998849801e-06, + "loss": 0.3935, + "step": 11509 + }, + { + "epoch": 5.3509995350999535, + "grad_norm": 4.501586437225342, + "learning_rate": 9.9884969988497e-06, + "loss": 0.1692, + "step": 11510 + }, + { + "epoch": 5.351464435146443, + "grad_norm": 4.8650007247924805, + "learning_rate": 9.9884959988496e-06, + "loss": 0.1845, + "step": 11511 + }, + { + "epoch": 5.351929335192933, + "grad_norm": 5.9380879402160645, + "learning_rate": 9.988494998849501e-06, + "loss": 0.2999, + "step": 11512 + }, + { + "epoch": 5.352394235239424, + "grad_norm": 8.699713706970215, + "learning_rate": 9.988493998849402e-06, + "loss": 0.4004, + "step": 11513 + }, + { + "epoch": 5.352859135285914, + "grad_norm": 5.8971381187438965, + "learning_rate": 9.9884929988493e-06, + "loss": 0.244, + "step": 11514 + }, + { + "epoch": 5.353324035332403, + "grad_norm": 4.325484275817871, + "learning_rate": 9.9884919988492e-06, + "loss": 0.1285, + "step": 11515 + }, + { + "epoch": 5.353788935378893, + "grad_norm": 3.774961471557617, + "learning_rate": 9.9884909988491e-06, + "loss": 0.0962, + "step": 11516 + }, + { + "epoch": 5.354253835425384, + "grad_norm": 7.186316013336182, + "learning_rate": 9.988489998849001e-06, + "loss": 0.3778, + "step": 11517 + }, + { + "epoch": 5.354718735471874, + "grad_norm": 9.230391502380371, + "learning_rate": 9.9884889988489e-06, + "loss": 0.4092, + "step": 11518 + }, + { + "epoch": 5.3551836355183635, + "grad_norm": 2.9089345932006836, + "learning_rate": 9.9884879988488e-06, + "loss": 0.0672, + "step": 11519 + }, + { + "epoch": 5.355648535564853, + "grad_norm": 5.427067756652832, + "learning_rate": 9.988486998848701e-06, + "loss": 0.2106, + "step": 11520 + }, + { + "epoch": 5.356113435611343, + "grad_norm": 6.120594024658203, + "learning_rate": 9.9884859988486e-06, + "loss": 0.3035, + "step": 11521 + }, + { + "epoch": 5.356578335657834, + "grad_norm": 6.050205230712891, + "learning_rate": 9.988484998848501e-06, + "loss": 0.1206, + "step": 11522 + }, + { + "epoch": 5.357043235704324, + "grad_norm": 7.417928218841553, + "learning_rate": 9.9884839988484e-06, + "loss": 0.3028, + "step": 11523 + }, + { + "epoch": 5.357508135750813, + "grad_norm": 7.7776970863342285, + "learning_rate": 9.9884829988483e-06, + "loss": 0.1655, + "step": 11524 + }, + { + "epoch": 5.357973035797303, + "grad_norm": 5.7949748039245605, + "learning_rate": 9.988481998848201e-06, + "loss": 0.0996, + "step": 11525 + }, + { + "epoch": 5.358437935843794, + "grad_norm": 7.2245402336120605, + "learning_rate": 9.9884809988481e-06, + "loss": 0.2991, + "step": 11526 + }, + { + "epoch": 5.358902835890284, + "grad_norm": 3.407254934310913, + "learning_rate": 9.988479998848001e-06, + "loss": 0.0857, + "step": 11527 + }, + { + "epoch": 5.3593677359367735, + "grad_norm": 5.6627702713012695, + "learning_rate": 9.9884789988479e-06, + "loss": 0.2287, + "step": 11528 + }, + { + "epoch": 5.359832635983263, + "grad_norm": 4.941817283630371, + "learning_rate": 9.9884779988478e-06, + "loss": 0.172, + "step": 11529 + }, + { + "epoch": 5.360297536029754, + "grad_norm": 4.932900905609131, + "learning_rate": 9.988476998847701e-06, + "loss": 0.1816, + "step": 11530 + }, + { + "epoch": 5.360762436076244, + "grad_norm": 6.494765281677246, + "learning_rate": 9.9884759988476e-06, + "loss": 0.2609, + "step": 11531 + }, + { + "epoch": 5.361227336122734, + "grad_norm": 6.170744895935059, + "learning_rate": 9.9884749988475e-06, + "loss": 0.2766, + "step": 11532 + }, + { + "epoch": 5.361692236169223, + "grad_norm": 2.6762030124664307, + "learning_rate": 9.988473998847401e-06, + "loss": 0.0426, + "step": 11533 + }, + { + "epoch": 5.362157136215713, + "grad_norm": 8.593914985656738, + "learning_rate": 9.9884729988473e-06, + "loss": 0.1524, + "step": 11534 + }, + { + "epoch": 5.362622036262204, + "grad_norm": 2.2357048988342285, + "learning_rate": 9.988471998847201e-06, + "loss": 0.0423, + "step": 11535 + }, + { + "epoch": 5.363086936308694, + "grad_norm": 2.896690845489502, + "learning_rate": 9.9884709988471e-06, + "loss": 0.0784, + "step": 11536 + }, + { + "epoch": 5.3635518363551835, + "grad_norm": 4.818882465362549, + "learning_rate": 9.988469998847e-06, + "loss": 0.1649, + "step": 11537 + }, + { + "epoch": 5.364016736401673, + "grad_norm": 3.850334644317627, + "learning_rate": 9.988468998846901e-06, + "loss": 0.1405, + "step": 11538 + }, + { + "epoch": 5.364481636448164, + "grad_norm": 5.075009346008301, + "learning_rate": 9.9884679988468e-06, + "loss": 0.1858, + "step": 11539 + }, + { + "epoch": 5.364946536494654, + "grad_norm": 5.754823207855225, + "learning_rate": 9.988466998846701e-06, + "loss": 0.1763, + "step": 11540 + }, + { + "epoch": 5.365411436541144, + "grad_norm": 7.14142370223999, + "learning_rate": 9.9884659988466e-06, + "loss": 0.4806, + "step": 11541 + }, + { + "epoch": 5.365876336587633, + "grad_norm": 8.672120094299316, + "learning_rate": 9.9884649988465e-06, + "loss": 0.4465, + "step": 11542 + }, + { + "epoch": 5.366341236634124, + "grad_norm": 4.908422946929932, + "learning_rate": 9.988463998846401e-06, + "loss": 0.2184, + "step": 11543 + }, + { + "epoch": 5.366806136680614, + "grad_norm": 4.5941338539123535, + "learning_rate": 9.988462998846302e-06, + "loss": 0.1605, + "step": 11544 + }, + { + "epoch": 5.367271036727104, + "grad_norm": 2.8960556983947754, + "learning_rate": 9.988461998846201e-06, + "loss": 0.0607, + "step": 11545 + }, + { + "epoch": 5.3677359367735935, + "grad_norm": 6.19329309463501, + "learning_rate": 9.988460998846102e-06, + "loss": 0.3375, + "step": 11546 + }, + { + "epoch": 5.368200836820083, + "grad_norm": 6.445839881896973, + "learning_rate": 9.988459998846e-06, + "loss": 0.2632, + "step": 11547 + }, + { + "epoch": 5.368665736866574, + "grad_norm": 7.800458908081055, + "learning_rate": 9.9884589988459e-06, + "loss": 0.4102, + "step": 11548 + }, + { + "epoch": 5.369130636913064, + "grad_norm": 5.492624282836914, + "learning_rate": 9.9884579988458e-06, + "loss": 0.2377, + "step": 11549 + }, + { + "epoch": 5.369595536959554, + "grad_norm": 4.827147006988525, + "learning_rate": 9.9884569988457e-06, + "loss": 0.1398, + "step": 11550 + }, + { + "epoch": 5.370060437006043, + "grad_norm": 2.1672160625457764, + "learning_rate": 9.988455998845601e-06, + "loss": 0.0368, + "step": 11551 + }, + { + "epoch": 5.370525337052534, + "grad_norm": 6.195812225341797, + "learning_rate": 9.9884549988455e-06, + "loss": 0.2056, + "step": 11552 + }, + { + "epoch": 5.370990237099024, + "grad_norm": 5.668805122375488, + "learning_rate": 9.988453998845401e-06, + "loss": 0.2587, + "step": 11553 + }, + { + "epoch": 5.371455137145514, + "grad_norm": 3.599525213241577, + "learning_rate": 9.9884529988453e-06, + "loss": 0.0942, + "step": 11554 + }, + { + "epoch": 5.3719200371920035, + "grad_norm": 4.667754650115967, + "learning_rate": 9.9884519988452e-06, + "loss": 0.1362, + "step": 11555 + }, + { + "epoch": 5.372384937238493, + "grad_norm": 3.6325440406799316, + "learning_rate": 9.988450998845101e-06, + "loss": 0.1038, + "step": 11556 + }, + { + "epoch": 5.372849837284984, + "grad_norm": 5.800827503204346, + "learning_rate": 9.988449998845002e-06, + "loss": 0.2338, + "step": 11557 + }, + { + "epoch": 5.373314737331474, + "grad_norm": 7.200035572052002, + "learning_rate": 9.988448998844901e-06, + "loss": 0.2601, + "step": 11558 + }, + { + "epoch": 5.373779637377964, + "grad_norm": 5.070812225341797, + "learning_rate": 9.9884479988448e-06, + "loss": 0.1273, + "step": 11559 + }, + { + "epoch": 5.374244537424453, + "grad_norm": 4.238870143890381, + "learning_rate": 9.9884469988447e-06, + "loss": 0.1286, + "step": 11560 + }, + { + "epoch": 5.374709437470944, + "grad_norm": 4.630924224853516, + "learning_rate": 9.9884459988446e-06, + "loss": 0.2396, + "step": 11561 + }, + { + "epoch": 5.375174337517434, + "grad_norm": 4.082691192626953, + "learning_rate": 9.9884449988445e-06, + "loss": 0.1766, + "step": 11562 + }, + { + "epoch": 5.375639237563924, + "grad_norm": 7.70106315612793, + "learning_rate": 9.988443998844401e-06, + "loss": 0.0958, + "step": 11563 + }, + { + "epoch": 5.3761041376104135, + "grad_norm": 2.517233371734619, + "learning_rate": 9.988442998844302e-06, + "loss": 0.0467, + "step": 11564 + }, + { + "epoch": 5.376569037656903, + "grad_norm": 3.0288326740264893, + "learning_rate": 9.9884419988442e-06, + "loss": 0.0467, + "step": 11565 + }, + { + "epoch": 5.377033937703394, + "grad_norm": 5.03764533996582, + "learning_rate": 9.988440998844101e-06, + "loss": 0.1978, + "step": 11566 + }, + { + "epoch": 5.377498837749884, + "grad_norm": 4.95982551574707, + "learning_rate": 9.988439998844e-06, + "loss": 0.1856, + "step": 11567 + }, + { + "epoch": 5.377963737796374, + "grad_norm": 6.78348970413208, + "learning_rate": 9.9884389988439e-06, + "loss": 0.3547, + "step": 11568 + }, + { + "epoch": 5.378428637842863, + "grad_norm": 5.146393775939941, + "learning_rate": 9.988437998843801e-06, + "loss": 0.2366, + "step": 11569 + }, + { + "epoch": 5.378893537889354, + "grad_norm": 3.6061511039733887, + "learning_rate": 9.9884369988437e-06, + "loss": 0.1141, + "step": 11570 + }, + { + "epoch": 5.379358437935844, + "grad_norm": 7.100583553314209, + "learning_rate": 9.988435998843601e-06, + "loss": 0.3731, + "step": 11571 + }, + { + "epoch": 5.379823337982334, + "grad_norm": 5.350821018218994, + "learning_rate": 9.9884349988435e-06, + "loss": 0.2561, + "step": 11572 + }, + { + "epoch": 5.3802882380288235, + "grad_norm": 5.800881862640381, + "learning_rate": 9.9884339988434e-06, + "loss": 0.3032, + "step": 11573 + }, + { + "epoch": 5.380753138075314, + "grad_norm": 4.5363664627075195, + "learning_rate": 9.9884329988433e-06, + "loss": 0.1995, + "step": 11574 + }, + { + "epoch": 5.381218038121804, + "grad_norm": 3.757936477661133, + "learning_rate": 9.9884319988432e-06, + "loss": 0.0918, + "step": 11575 + }, + { + "epoch": 5.381682938168294, + "grad_norm": 6.519900321960449, + "learning_rate": 9.988430998843101e-06, + "loss": 0.2007, + "step": 11576 + }, + { + "epoch": 5.382147838214784, + "grad_norm": 5.198527812957764, + "learning_rate": 9.988429998843002e-06, + "loss": 0.0461, + "step": 11577 + }, + { + "epoch": 5.382612738261274, + "grad_norm": 5.768898963928223, + "learning_rate": 9.9884289988429e-06, + "loss": 0.2365, + "step": 11578 + }, + { + "epoch": 5.383077638307764, + "grad_norm": 6.267490386962891, + "learning_rate": 9.988427998842801e-06, + "loss": 0.2946, + "step": 11579 + }, + { + "epoch": 5.383542538354254, + "grad_norm": 6.099261283874512, + "learning_rate": 9.9884269988427e-06, + "loss": 0.2969, + "step": 11580 + }, + { + "epoch": 5.384007438400744, + "grad_norm": 4.708968639373779, + "learning_rate": 9.988425998842601e-06, + "loss": 0.1339, + "step": 11581 + }, + { + "epoch": 5.3844723384472335, + "grad_norm": 4.411438465118408, + "learning_rate": 9.988424998842502e-06, + "loss": 0.1712, + "step": 11582 + }, + { + "epoch": 5.384937238493724, + "grad_norm": 4.1093316078186035, + "learning_rate": 9.9884239988424e-06, + "loss": 0.1429, + "step": 11583 + }, + { + "epoch": 5.385402138540214, + "grad_norm": 5.464073657989502, + "learning_rate": 9.988422998842301e-06, + "loss": 0.2305, + "step": 11584 + }, + { + "epoch": 5.385867038586704, + "grad_norm": 6.313300609588623, + "learning_rate": 9.9884219988422e-06, + "loss": 0.1499, + "step": 11585 + }, + { + "epoch": 5.386331938633194, + "grad_norm": 6.017398357391357, + "learning_rate": 9.9884209988421e-06, + "loss": 0.2565, + "step": 11586 + }, + { + "epoch": 5.386796838679684, + "grad_norm": 9.53261661529541, + "learning_rate": 9.988419998842e-06, + "loss": 0.3777, + "step": 11587 + }, + { + "epoch": 5.387261738726174, + "grad_norm": 4.953032493591309, + "learning_rate": 9.9884189988419e-06, + "loss": 0.1538, + "step": 11588 + }, + { + "epoch": 5.387726638772664, + "grad_norm": 8.115865707397461, + "learning_rate": 9.988417998841801e-06, + "loss": 0.0924, + "step": 11589 + }, + { + "epoch": 5.388191538819154, + "grad_norm": 4.138685703277588, + "learning_rate": 9.988416998841702e-06, + "loss": 0.1036, + "step": 11590 + }, + { + "epoch": 5.3886564388656435, + "grad_norm": 6.284145832061768, + "learning_rate": 9.9884159988416e-06, + "loss": 0.3303, + "step": 11591 + }, + { + "epoch": 5.389121338912134, + "grad_norm": 6.975838661193848, + "learning_rate": 9.9884149988415e-06, + "loss": 0.4102, + "step": 11592 + }, + { + "epoch": 5.389586238958624, + "grad_norm": 6.620131492614746, + "learning_rate": 9.9884139988414e-06, + "loss": 0.2134, + "step": 11593 + }, + { + "epoch": 5.390051139005114, + "grad_norm": 6.67706298828125, + "learning_rate": 9.988412998841301e-06, + "loss": 0.3135, + "step": 11594 + }, + { + "epoch": 5.390516039051604, + "grad_norm": 5.043463230133057, + "learning_rate": 9.988411998841202e-06, + "loss": 0.252, + "step": 11595 + }, + { + "epoch": 5.390980939098094, + "grad_norm": 4.105141639709473, + "learning_rate": 9.9884109988411e-06, + "loss": 0.1057, + "step": 11596 + }, + { + "epoch": 5.391445839144584, + "grad_norm": 4.589963436126709, + "learning_rate": 9.988409998841001e-06, + "loss": 0.1544, + "step": 11597 + }, + { + "epoch": 5.391910739191074, + "grad_norm": 5.6422271728515625, + "learning_rate": 9.9884089988409e-06, + "loss": 0.305, + "step": 11598 + }, + { + "epoch": 5.392375639237564, + "grad_norm": 6.373017311096191, + "learning_rate": 9.988407998840801e-06, + "loss": 0.3308, + "step": 11599 + }, + { + "epoch": 5.3928405392840535, + "grad_norm": 7.353903293609619, + "learning_rate": 9.9884069988407e-06, + "loss": 0.1919, + "step": 11600 + }, + { + "epoch": 5.393305439330544, + "grad_norm": 5.448550224304199, + "learning_rate": 9.9884059988406e-06, + "loss": 0.1986, + "step": 11601 + }, + { + "epoch": 5.393770339377034, + "grad_norm": 6.200387001037598, + "learning_rate": 9.988404998840501e-06, + "loss": 0.3247, + "step": 11602 + }, + { + "epoch": 5.394235239423524, + "grad_norm": 6.424281120300293, + "learning_rate": 9.9884039988404e-06, + "loss": 0.3031, + "step": 11603 + }, + { + "epoch": 5.394700139470014, + "grad_norm": 4.768149375915527, + "learning_rate": 9.9884029988403e-06, + "loss": 0.1832, + "step": 11604 + }, + { + "epoch": 5.395165039516504, + "grad_norm": 6.022848606109619, + "learning_rate": 9.9884019988402e-06, + "loss": 0.1202, + "step": 11605 + }, + { + "epoch": 5.395629939562994, + "grad_norm": 8.86095905303955, + "learning_rate": 9.9884009988401e-06, + "loss": 0.4532, + "step": 11606 + }, + { + "epoch": 5.396094839609484, + "grad_norm": 5.042858123779297, + "learning_rate": 9.988399998840001e-06, + "loss": 0.1429, + "step": 11607 + }, + { + "epoch": 5.396559739655974, + "grad_norm": 8.33321762084961, + "learning_rate": 9.988398998839902e-06, + "loss": 0.5175, + "step": 11608 + }, + { + "epoch": 5.397024639702464, + "grad_norm": 3.9160828590393066, + "learning_rate": 9.9883979988398e-06, + "loss": 0.1264, + "step": 11609 + }, + { + "epoch": 5.397489539748954, + "grad_norm": 4.288855075836182, + "learning_rate": 9.988396998839701e-06, + "loss": 0.1015, + "step": 11610 + }, + { + "epoch": 5.397954439795444, + "grad_norm": 5.489004611968994, + "learning_rate": 9.9883959988396e-06, + "loss": 0.1724, + "step": 11611 + }, + { + "epoch": 5.398419339841934, + "grad_norm": 6.84521484375, + "learning_rate": 9.988394998839501e-06, + "loss": 0.381, + "step": 11612 + }, + { + "epoch": 5.398884239888424, + "grad_norm": 3.983449697494507, + "learning_rate": 9.988393998839402e-06, + "loss": 0.1349, + "step": 11613 + }, + { + "epoch": 5.399349139934914, + "grad_norm": 5.0917582511901855, + "learning_rate": 9.9883929988393e-06, + "loss": 0.2596, + "step": 11614 + }, + { + "epoch": 5.399814039981404, + "grad_norm": 4.452930927276611, + "learning_rate": 9.988391998839201e-06, + "loss": 0.2032, + "step": 11615 + }, + { + "epoch": 5.400278940027894, + "grad_norm": 5.121654510498047, + "learning_rate": 9.9883909988391e-06, + "loss": 0.2632, + "step": 11616 + }, + { + "epoch": 5.400743840074384, + "grad_norm": 6.039309978485107, + "learning_rate": 9.988389998839e-06, + "loss": 0.4045, + "step": 11617 + }, + { + "epoch": 5.401208740120874, + "grad_norm": 3.2968008518218994, + "learning_rate": 9.9883889988389e-06, + "loss": 0.1494, + "step": 11618 + }, + { + "epoch": 5.401673640167364, + "grad_norm": 10.329607963562012, + "learning_rate": 9.9883879988388e-06, + "loss": 0.3013, + "step": 11619 + }, + { + "epoch": 5.402138540213854, + "grad_norm": 7.854704856872559, + "learning_rate": 9.988386998838701e-06, + "loss": 0.3393, + "step": 11620 + }, + { + "epoch": 5.402603440260344, + "grad_norm": 2.703993082046509, + "learning_rate": 9.988385998838602e-06, + "loss": 0.0766, + "step": 11621 + }, + { + "epoch": 5.4030683403068345, + "grad_norm": 4.904130458831787, + "learning_rate": 9.9883849988385e-06, + "loss": 0.2651, + "step": 11622 + }, + { + "epoch": 5.403533240353324, + "grad_norm": 5.197391510009766, + "learning_rate": 9.9883839988384e-06, + "loss": 0.3531, + "step": 11623 + }, + { + "epoch": 5.403998140399814, + "grad_norm": 5.1171040534973145, + "learning_rate": 9.9883829988383e-06, + "loss": 0.2859, + "step": 11624 + }, + { + "epoch": 5.404463040446304, + "grad_norm": 5.712289810180664, + "learning_rate": 9.988381998838201e-06, + "loss": 0.1959, + "step": 11625 + }, + { + "epoch": 5.404927940492794, + "grad_norm": 5.478761196136475, + "learning_rate": 9.988380998838102e-06, + "loss": 0.3746, + "step": 11626 + }, + { + "epoch": 5.405392840539284, + "grad_norm": 6.291998386383057, + "learning_rate": 9.988379998838e-06, + "loss": 0.3604, + "step": 11627 + }, + { + "epoch": 5.405857740585774, + "grad_norm": 8.164362907409668, + "learning_rate": 9.988378998837901e-06, + "loss": 0.6021, + "step": 11628 + }, + { + "epoch": 5.406322640632264, + "grad_norm": 6.320788860321045, + "learning_rate": 9.9883779988378e-06, + "loss": 0.2814, + "step": 11629 + }, + { + "epoch": 5.406787540678754, + "grad_norm": 6.952826499938965, + "learning_rate": 9.988376998837701e-06, + "loss": 0.2326, + "step": 11630 + }, + { + "epoch": 5.4072524407252445, + "grad_norm": 6.232759952545166, + "learning_rate": 9.9883759988376e-06, + "loss": 0.2146, + "step": 11631 + }, + { + "epoch": 5.407717340771734, + "grad_norm": 2.175123453140259, + "learning_rate": 9.9883749988375e-06, + "loss": 0.0389, + "step": 11632 + }, + { + "epoch": 5.408182240818224, + "grad_norm": 5.730978965759277, + "learning_rate": 9.988373998837401e-06, + "loss": 0.271, + "step": 11633 + }, + { + "epoch": 5.408647140864714, + "grad_norm": 5.823145389556885, + "learning_rate": 9.9883729988373e-06, + "loss": 0.3309, + "step": 11634 + }, + { + "epoch": 5.409112040911204, + "grad_norm": 8.123746871948242, + "learning_rate": 9.9883719988372e-06, + "loss": 0.3817, + "step": 11635 + }, + { + "epoch": 5.409576940957694, + "grad_norm": 5.432088375091553, + "learning_rate": 9.9883709988371e-06, + "loss": 0.2579, + "step": 11636 + }, + { + "epoch": 5.410041841004184, + "grad_norm": 4.25083589553833, + "learning_rate": 9.988369998837e-06, + "loss": 0.1612, + "step": 11637 + }, + { + "epoch": 5.410506741050674, + "grad_norm": 6.264277935028076, + "learning_rate": 9.988368998836901e-06, + "loss": 0.3049, + "step": 11638 + }, + { + "epoch": 5.410971641097164, + "grad_norm": 7.023777961730957, + "learning_rate": 9.988367998836802e-06, + "loss": 0.225, + "step": 11639 + }, + { + "epoch": 5.4114365411436545, + "grad_norm": 4.479142189025879, + "learning_rate": 9.9883669988367e-06, + "loss": 0.1235, + "step": 11640 + }, + { + "epoch": 5.411901441190144, + "grad_norm": 3.8510472774505615, + "learning_rate": 9.988365998836601e-06, + "loss": 0.0889, + "step": 11641 + }, + { + "epoch": 5.412366341236634, + "grad_norm": 3.696093797683716, + "learning_rate": 9.9883649988365e-06, + "loss": 0.0695, + "step": 11642 + }, + { + "epoch": 5.412831241283124, + "grad_norm": 5.125757217407227, + "learning_rate": 9.988363998836401e-06, + "loss": 0.2085, + "step": 11643 + }, + { + "epoch": 5.413296141329614, + "grad_norm": 8.317493438720703, + "learning_rate": 9.9883629988363e-06, + "loss": 0.5507, + "step": 11644 + }, + { + "epoch": 5.413761041376104, + "grad_norm": 5.192529678344727, + "learning_rate": 9.9883619988362e-06, + "loss": 0.1388, + "step": 11645 + }, + { + "epoch": 5.414225941422594, + "grad_norm": 4.157790660858154, + "learning_rate": 9.988360998836101e-06, + "loss": 0.1342, + "step": 11646 + }, + { + "epoch": 5.414690841469084, + "grad_norm": 5.719890594482422, + "learning_rate": 9.988359998836e-06, + "loss": 0.1703, + "step": 11647 + }, + { + "epoch": 5.415155741515574, + "grad_norm": 5.126612186431885, + "learning_rate": 9.988358998835901e-06, + "loss": 0.1507, + "step": 11648 + }, + { + "epoch": 5.4156206415620645, + "grad_norm": 4.996531009674072, + "learning_rate": 9.9883579988358e-06, + "loss": 0.1439, + "step": 11649 + }, + { + "epoch": 5.416085541608554, + "grad_norm": 4.3580217361450195, + "learning_rate": 9.9883569988357e-06, + "loss": 0.1093, + "step": 11650 + }, + { + "epoch": 5.416550441655044, + "grad_norm": 7.118924617767334, + "learning_rate": 9.988355998835601e-06, + "loss": 0.239, + "step": 11651 + }, + { + "epoch": 5.417015341701534, + "grad_norm": 6.569984436035156, + "learning_rate": 9.988354998835502e-06, + "loss": 0.4205, + "step": 11652 + }, + { + "epoch": 5.417480241748025, + "grad_norm": 5.183562278747559, + "learning_rate": 9.9883539988354e-06, + "loss": 0.1903, + "step": 11653 + }, + { + "epoch": 5.417945141794514, + "grad_norm": 3.668884515762329, + "learning_rate": 9.988352998835301e-06, + "loss": 0.1483, + "step": 11654 + }, + { + "epoch": 5.418410041841004, + "grad_norm": 11.526994705200195, + "learning_rate": 9.9883519988352e-06, + "loss": 0.2925, + "step": 11655 + }, + { + "epoch": 5.418874941887494, + "grad_norm": 4.87384557723999, + "learning_rate": 9.9883509988351e-06, + "loss": 0.2157, + "step": 11656 + }, + { + "epoch": 5.419339841933985, + "grad_norm": 5.328398704528809, + "learning_rate": 9.988349998835e-06, + "loss": 0.2281, + "step": 11657 + }, + { + "epoch": 5.4198047419804745, + "grad_norm": 7.763597011566162, + "learning_rate": 9.9883489988349e-06, + "loss": 0.2995, + "step": 11658 + }, + { + "epoch": 5.420269642026964, + "grad_norm": 6.961370468139648, + "learning_rate": 9.988347998834801e-06, + "loss": 0.2078, + "step": 11659 + }, + { + "epoch": 5.420734542073454, + "grad_norm": 4.066214561462402, + "learning_rate": 9.9883469988347e-06, + "loss": 0.091, + "step": 11660 + }, + { + "epoch": 5.421199442119944, + "grad_norm": 4.750272274017334, + "learning_rate": 9.988345998834601e-06, + "loss": 0.1779, + "step": 11661 + }, + { + "epoch": 5.421664342166435, + "grad_norm": 5.388063907623291, + "learning_rate": 9.9883449988345e-06, + "loss": 0.2043, + "step": 11662 + }, + { + "epoch": 5.422129242212924, + "grad_norm": 5.253352642059326, + "learning_rate": 9.9883439988344e-06, + "loss": 0.3401, + "step": 11663 + }, + { + "epoch": 5.422594142259414, + "grad_norm": 6.8428850173950195, + "learning_rate": 9.988342998834301e-06, + "loss": 0.2782, + "step": 11664 + }, + { + "epoch": 5.423059042305904, + "grad_norm": 4.784128665924072, + "learning_rate": 9.988341998834202e-06, + "loss": 0.1447, + "step": 11665 + }, + { + "epoch": 5.423523942352395, + "grad_norm": 4.061262130737305, + "learning_rate": 9.988340998834101e-06, + "loss": 0.1326, + "step": 11666 + }, + { + "epoch": 5.4239888423988845, + "grad_norm": 5.231024265289307, + "learning_rate": 9.988339998834e-06, + "loss": 0.2281, + "step": 11667 + }, + { + "epoch": 5.424453742445374, + "grad_norm": 4.479249000549316, + "learning_rate": 9.9883389988339e-06, + "loss": 0.1927, + "step": 11668 + }, + { + "epoch": 5.424918642491864, + "grad_norm": 5.3464837074279785, + "learning_rate": 9.9883379988338e-06, + "loss": 0.081, + "step": 11669 + }, + { + "epoch": 5.425383542538354, + "grad_norm": 6.224452972412109, + "learning_rate": 9.9883369988337e-06, + "loss": 0.265, + "step": 11670 + }, + { + "epoch": 5.4258484425848446, + "grad_norm": 6.594588279724121, + "learning_rate": 9.9883359988336e-06, + "loss": 0.245, + "step": 11671 + }, + { + "epoch": 5.426313342631334, + "grad_norm": 6.229089260101318, + "learning_rate": 9.988334998833501e-06, + "loss": 0.1891, + "step": 11672 + }, + { + "epoch": 5.426778242677824, + "grad_norm": 4.2358880043029785, + "learning_rate": 9.9883339988334e-06, + "loss": 0.1899, + "step": 11673 + }, + { + "epoch": 5.427243142724314, + "grad_norm": 4.0818915367126465, + "learning_rate": 9.988332998833301e-06, + "loss": 0.1411, + "step": 11674 + }, + { + "epoch": 5.427708042770805, + "grad_norm": 6.446625709533691, + "learning_rate": 9.9883319988332e-06, + "loss": 0.3471, + "step": 11675 + }, + { + "epoch": 5.4281729428172945, + "grad_norm": 6.033839225769043, + "learning_rate": 9.9883309988331e-06, + "loss": 0.2167, + "step": 11676 + }, + { + "epoch": 5.428637842863784, + "grad_norm": 4.887164115905762, + "learning_rate": 9.988329998833001e-06, + "loss": 0.1644, + "step": 11677 + }, + { + "epoch": 5.429102742910274, + "grad_norm": 2.762540340423584, + "learning_rate": 9.9883289988329e-06, + "loss": 0.1007, + "step": 11678 + }, + { + "epoch": 5.429567642956764, + "grad_norm": 7.321948051452637, + "learning_rate": 9.988327998832801e-06, + "loss": 0.3762, + "step": 11679 + }, + { + "epoch": 5.4300325430032546, + "grad_norm": 6.261663436889648, + "learning_rate": 9.9883269988327e-06, + "loss": 0.1403, + "step": 11680 + }, + { + "epoch": 5.430497443049744, + "grad_norm": 3.2547008991241455, + "learning_rate": 9.9883259988326e-06, + "loss": 0.1211, + "step": 11681 + }, + { + "epoch": 5.430962343096234, + "grad_norm": 5.138772964477539, + "learning_rate": 9.988324998832501e-06, + "loss": 0.1971, + "step": 11682 + }, + { + "epoch": 5.431427243142724, + "grad_norm": 21.532745361328125, + "learning_rate": 9.9883239988324e-06, + "loss": 0.4338, + "step": 11683 + }, + { + "epoch": 5.431892143189215, + "grad_norm": 6.697549343109131, + "learning_rate": 9.988322998832301e-06, + "loss": 0.3514, + "step": 11684 + }, + { + "epoch": 5.4323570432357045, + "grad_norm": 6.949622631072998, + "learning_rate": 9.988321998832202e-06, + "loss": 0.1765, + "step": 11685 + }, + { + "epoch": 5.432821943282194, + "grad_norm": 4.07832145690918, + "learning_rate": 9.9883209988321e-06, + "loss": 0.1003, + "step": 11686 + }, + { + "epoch": 5.433286843328684, + "grad_norm": 4.894749164581299, + "learning_rate": 9.988319998832001e-06, + "loss": 0.1229, + "step": 11687 + }, + { + "epoch": 5.433751743375175, + "grad_norm": 6.4177656173706055, + "learning_rate": 9.9883189988319e-06, + "loss": 0.2092, + "step": 11688 + }, + { + "epoch": 5.4342166434216645, + "grad_norm": 4.959311485290527, + "learning_rate": 9.9883179988318e-06, + "loss": 0.2291, + "step": 11689 + }, + { + "epoch": 5.434681543468154, + "grad_norm": 5.784992218017578, + "learning_rate": 9.988316998831701e-06, + "loss": 0.108, + "step": 11690 + }, + { + "epoch": 5.435146443514644, + "grad_norm": 4.162249565124512, + "learning_rate": 9.9883159988316e-06, + "loss": 0.156, + "step": 11691 + }, + { + "epoch": 5.435611343561134, + "grad_norm": 9.474480628967285, + "learning_rate": 9.988314998831501e-06, + "loss": 0.3254, + "step": 11692 + }, + { + "epoch": 5.436076243607625, + "grad_norm": 6.249318599700928, + "learning_rate": 9.9883139988314e-06, + "loss": 0.2582, + "step": 11693 + }, + { + "epoch": 5.4365411436541144, + "grad_norm": 4.441227436065674, + "learning_rate": 9.9883129988313e-06, + "loss": 0.1478, + "step": 11694 + }, + { + "epoch": 5.437006043700604, + "grad_norm": 6.556535720825195, + "learning_rate": 9.988311998831201e-06, + "loss": 0.1229, + "step": 11695 + }, + { + "epoch": 5.437470943747094, + "grad_norm": 6.5516438484191895, + "learning_rate": 9.9883109988311e-06, + "loss": 0.2932, + "step": 11696 + }, + { + "epoch": 5.437935843793585, + "grad_norm": 5.008204460144043, + "learning_rate": 9.988309998831001e-06, + "loss": 0.198, + "step": 11697 + }, + { + "epoch": 5.4384007438400745, + "grad_norm": 5.805911540985107, + "learning_rate": 9.988308998830902e-06, + "loss": 0.3358, + "step": 11698 + }, + { + "epoch": 5.438865643886564, + "grad_norm": 5.688365936279297, + "learning_rate": 9.9883079988308e-06, + "loss": 0.3107, + "step": 11699 + }, + { + "epoch": 5.439330543933054, + "grad_norm": 5.49184513092041, + "learning_rate": 9.9883069988307e-06, + "loss": 0.271, + "step": 11700 + }, + { + "epoch": 5.439795443979545, + "grad_norm": 6.185006141662598, + "learning_rate": 9.9883059988306e-06, + "loss": 0.4044, + "step": 11701 + }, + { + "epoch": 5.440260344026035, + "grad_norm": 4.553122043609619, + "learning_rate": 9.9883049988305e-06, + "loss": 0.0983, + "step": 11702 + }, + { + "epoch": 5.4407252440725244, + "grad_norm": 9.65876579284668, + "learning_rate": 9.988303998830401e-06, + "loss": 0.4126, + "step": 11703 + }, + { + "epoch": 5.441190144119014, + "grad_norm": 2.416816234588623, + "learning_rate": 9.9883029988303e-06, + "loss": 0.0368, + "step": 11704 + }, + { + "epoch": 5.441655044165504, + "grad_norm": 7.184269428253174, + "learning_rate": 9.988301998830201e-06, + "loss": 0.1142, + "step": 11705 + }, + { + "epoch": 5.442119944211995, + "grad_norm": 3.485236406326294, + "learning_rate": 9.9883009988301e-06, + "loss": 0.0969, + "step": 11706 + }, + { + "epoch": 5.4425848442584845, + "grad_norm": 4.141102313995361, + "learning_rate": 9.98829999883e-06, + "loss": 0.1207, + "step": 11707 + }, + { + "epoch": 5.443049744304974, + "grad_norm": 4.27523136138916, + "learning_rate": 9.988298998829901e-06, + "loss": 0.1852, + "step": 11708 + }, + { + "epoch": 5.443514644351464, + "grad_norm": 4.029647350311279, + "learning_rate": 9.988297998829802e-06, + "loss": 0.1287, + "step": 11709 + }, + { + "epoch": 5.443979544397955, + "grad_norm": 5.985950946807861, + "learning_rate": 9.988296998829701e-06, + "loss": 0.2122, + "step": 11710 + }, + { + "epoch": 5.444444444444445, + "grad_norm": 3.823913812637329, + "learning_rate": 9.9882959988296e-06, + "loss": 0.1262, + "step": 11711 + }, + { + "epoch": 5.4449093444909344, + "grad_norm": 4.5288848876953125, + "learning_rate": 9.9882949988295e-06, + "loss": 0.1888, + "step": 11712 + }, + { + "epoch": 5.445374244537424, + "grad_norm": 3.780447244644165, + "learning_rate": 9.9882939988294e-06, + "loss": 0.1206, + "step": 11713 + }, + { + "epoch": 5.445839144583914, + "grad_norm": 3.5801632404327393, + "learning_rate": 9.9882929988293e-06, + "loss": 0.1611, + "step": 11714 + }, + { + "epoch": 5.446304044630405, + "grad_norm": 5.144158840179443, + "learning_rate": 9.988291998829201e-06, + "loss": 0.2186, + "step": 11715 + }, + { + "epoch": 5.4467689446768945, + "grad_norm": 4.338473796844482, + "learning_rate": 9.988290998829102e-06, + "loss": 0.1045, + "step": 11716 + }, + { + "epoch": 5.447233844723384, + "grad_norm": 7.129174709320068, + "learning_rate": 9.988289998829e-06, + "loss": 0.2287, + "step": 11717 + }, + { + "epoch": 5.447698744769874, + "grad_norm": 4.676663398742676, + "learning_rate": 9.988288998828901e-06, + "loss": 0.1708, + "step": 11718 + }, + { + "epoch": 5.448163644816365, + "grad_norm": 6.784872055053711, + "learning_rate": 9.9882879988288e-06, + "loss": 0.2744, + "step": 11719 + }, + { + "epoch": 5.448628544862855, + "grad_norm": 4.553268909454346, + "learning_rate": 9.9882869988287e-06, + "loss": 0.1856, + "step": 11720 + }, + { + "epoch": 5.449093444909344, + "grad_norm": 8.117475509643555, + "learning_rate": 9.988285998828601e-06, + "loss": 0.4382, + "step": 11721 + }, + { + "epoch": 5.449558344955834, + "grad_norm": 3.9686436653137207, + "learning_rate": 9.9882849988285e-06, + "loss": 0.1734, + "step": 11722 + }, + { + "epoch": 5.450023245002324, + "grad_norm": 6.2698140144348145, + "learning_rate": 9.988283998828401e-06, + "loss": 0.4015, + "step": 11723 + }, + { + "epoch": 5.450488145048815, + "grad_norm": 4.762072563171387, + "learning_rate": 9.9882829988283e-06, + "loss": 0.1633, + "step": 11724 + }, + { + "epoch": 5.4509530450953045, + "grad_norm": 4.866327285766602, + "learning_rate": 9.9882819988282e-06, + "loss": 0.1489, + "step": 11725 + }, + { + "epoch": 5.451417945141794, + "grad_norm": 4.769167423248291, + "learning_rate": 9.9882809988281e-06, + "loss": 0.2309, + "step": 11726 + }, + { + "epoch": 5.451882845188284, + "grad_norm": 6.35584020614624, + "learning_rate": 9.988279998828e-06, + "loss": 0.2104, + "step": 11727 + }, + { + "epoch": 5.452347745234775, + "grad_norm": 4.590717315673828, + "learning_rate": 9.988278998827901e-06, + "loss": 0.1443, + "step": 11728 + }, + { + "epoch": 5.452812645281265, + "grad_norm": 4.641330718994141, + "learning_rate": 9.988277998827802e-06, + "loss": 0.0729, + "step": 11729 + }, + { + "epoch": 5.453277545327754, + "grad_norm": 7.443795204162598, + "learning_rate": 9.9882769988277e-06, + "loss": 0.2583, + "step": 11730 + }, + { + "epoch": 5.453742445374244, + "grad_norm": 11.001696586608887, + "learning_rate": 9.988275998827601e-06, + "loss": 0.3054, + "step": 11731 + }, + { + "epoch": 5.454207345420735, + "grad_norm": 3.841464042663574, + "learning_rate": 9.9882749988275e-06, + "loss": 0.1377, + "step": 11732 + }, + { + "epoch": 5.454672245467225, + "grad_norm": 6.389795303344727, + "learning_rate": 9.988273998827401e-06, + "loss": 0.2845, + "step": 11733 + }, + { + "epoch": 5.4551371455137145, + "grad_norm": 5.171884536743164, + "learning_rate": 9.988272998827302e-06, + "loss": 0.288, + "step": 11734 + }, + { + "epoch": 5.455602045560204, + "grad_norm": 6.120428085327148, + "learning_rate": 9.9882719988272e-06, + "loss": 0.2034, + "step": 11735 + }, + { + "epoch": 5.456066945606695, + "grad_norm": 5.468534469604492, + "learning_rate": 9.988270998827101e-06, + "loss": 0.1836, + "step": 11736 + }, + { + "epoch": 5.456531845653185, + "grad_norm": 8.48369026184082, + "learning_rate": 9.988269998827e-06, + "loss": 0.3006, + "step": 11737 + }, + { + "epoch": 5.456996745699675, + "grad_norm": 4.248183250427246, + "learning_rate": 9.9882689988269e-06, + "loss": 0.1978, + "step": 11738 + }, + { + "epoch": 5.457461645746164, + "grad_norm": 5.860454559326172, + "learning_rate": 9.9882679988268e-06, + "loss": 0.3973, + "step": 11739 + }, + { + "epoch": 5.457926545792654, + "grad_norm": 5.120257377624512, + "learning_rate": 9.9882669988267e-06, + "loss": 0.2826, + "step": 11740 + }, + { + "epoch": 5.458391445839145, + "grad_norm": 6.115530490875244, + "learning_rate": 9.988265998826601e-06, + "loss": 0.5503, + "step": 11741 + }, + { + "epoch": 5.458856345885635, + "grad_norm": 6.719173431396484, + "learning_rate": 9.988264998826502e-06, + "loss": 0.1132, + "step": 11742 + }, + { + "epoch": 5.4593212459321245, + "grad_norm": 4.703110218048096, + "learning_rate": 9.9882639988264e-06, + "loss": 0.1782, + "step": 11743 + }, + { + "epoch": 5.459786145978614, + "grad_norm": 3.553277015686035, + "learning_rate": 9.9882629988263e-06, + "loss": 0.1027, + "step": 11744 + }, + { + "epoch": 5.460251046025105, + "grad_norm": 5.47097110748291, + "learning_rate": 9.9882619988262e-06, + "loss": 0.1903, + "step": 11745 + }, + { + "epoch": 5.460715946071595, + "grad_norm": 4.904457092285156, + "learning_rate": 9.988260998826101e-06, + "loss": 0.2342, + "step": 11746 + }, + { + "epoch": 5.461180846118085, + "grad_norm": 7.827671051025391, + "learning_rate": 9.988259998826002e-06, + "loss": 0.3447, + "step": 11747 + }, + { + "epoch": 5.461645746164574, + "grad_norm": 3.8737218379974365, + "learning_rate": 9.9882589988259e-06, + "loss": 0.1748, + "step": 11748 + }, + { + "epoch": 5.462110646211064, + "grad_norm": 5.207533359527588, + "learning_rate": 9.988257998825801e-06, + "loss": 0.2378, + "step": 11749 + }, + { + "epoch": 5.462575546257555, + "grad_norm": 10.409561157226562, + "learning_rate": 9.9882569988257e-06, + "loss": 0.1438, + "step": 11750 + }, + { + "epoch": 5.463040446304045, + "grad_norm": 6.315216541290283, + "learning_rate": 9.988255998825601e-06, + "loss": 0.2443, + "step": 11751 + }, + { + "epoch": 5.4635053463505345, + "grad_norm": 4.046448707580566, + "learning_rate": 9.9882549988255e-06, + "loss": 0.1307, + "step": 11752 + }, + { + "epoch": 5.463970246397024, + "grad_norm": 4.128967761993408, + "learning_rate": 9.9882539988254e-06, + "loss": 0.1239, + "step": 11753 + }, + { + "epoch": 5.464435146443515, + "grad_norm": 7.017398357391357, + "learning_rate": 9.988252998825301e-06, + "loss": 0.4106, + "step": 11754 + }, + { + "epoch": 5.464900046490005, + "grad_norm": 5.0106520652771, + "learning_rate": 9.9882519988252e-06, + "loss": 0.2336, + "step": 11755 + }, + { + "epoch": 5.465364946536495, + "grad_norm": 3.476212978363037, + "learning_rate": 9.9882509988251e-06, + "loss": 0.1054, + "step": 11756 + }, + { + "epoch": 5.465829846582984, + "grad_norm": 6.939982891082764, + "learning_rate": 9.988249998825e-06, + "loss": 0.3949, + "step": 11757 + }, + { + "epoch": 5.466294746629474, + "grad_norm": 3.9505650997161865, + "learning_rate": 9.9882489988249e-06, + "loss": 0.1379, + "step": 11758 + }, + { + "epoch": 5.466759646675965, + "grad_norm": 5.498077392578125, + "learning_rate": 9.988247998824801e-06, + "loss": 0.2041, + "step": 11759 + }, + { + "epoch": 5.467224546722455, + "grad_norm": 5.94424295425415, + "learning_rate": 9.988246998824702e-06, + "loss": 0.3058, + "step": 11760 + }, + { + "epoch": 5.4676894467689445, + "grad_norm": 4.9201273918151855, + "learning_rate": 9.9882459988246e-06, + "loss": 0.1957, + "step": 11761 + }, + { + "epoch": 5.468154346815434, + "grad_norm": 3.2039108276367188, + "learning_rate": 9.988244998824501e-06, + "loss": 0.0456, + "step": 11762 + }, + { + "epoch": 5.468619246861925, + "grad_norm": 6.1804656982421875, + "learning_rate": 9.9882439988244e-06, + "loss": 0.2418, + "step": 11763 + }, + { + "epoch": 5.469084146908415, + "grad_norm": 2.3915812969207764, + "learning_rate": 9.988242998824301e-06, + "loss": 0.073, + "step": 11764 + }, + { + "epoch": 5.469549046954905, + "grad_norm": 4.962741374969482, + "learning_rate": 9.9882419988242e-06, + "loss": 0.1604, + "step": 11765 + }, + { + "epoch": 5.470013947001394, + "grad_norm": 11.043110847473145, + "learning_rate": 9.9882409988241e-06, + "loss": 0.538, + "step": 11766 + }, + { + "epoch": 5.470478847047885, + "grad_norm": 3.0069386959075928, + "learning_rate": 9.988239998824001e-06, + "loss": 0.133, + "step": 11767 + }, + { + "epoch": 5.470943747094375, + "grad_norm": 8.10533332824707, + "learning_rate": 9.9882389988239e-06, + "loss": 0.1764, + "step": 11768 + }, + { + "epoch": 5.471408647140865, + "grad_norm": 5.788373947143555, + "learning_rate": 9.988237998823801e-06, + "loss": 0.2655, + "step": 11769 + }, + { + "epoch": 5.4718735471873545, + "grad_norm": 7.765659809112549, + "learning_rate": 9.9882369988237e-06, + "loss": 0.4575, + "step": 11770 + }, + { + "epoch": 5.472338447233844, + "grad_norm": 5.7245283126831055, + "learning_rate": 9.9882359988236e-06, + "loss": 0.2655, + "step": 11771 + }, + { + "epoch": 5.472803347280335, + "grad_norm": 5.857509613037109, + "learning_rate": 9.988234998823501e-06, + "loss": 0.3181, + "step": 11772 + }, + { + "epoch": 5.473268247326825, + "grad_norm": 3.738898754119873, + "learning_rate": 9.988233998823402e-06, + "loss": 0.097, + "step": 11773 + }, + { + "epoch": 5.473733147373315, + "grad_norm": 3.8388586044311523, + "learning_rate": 9.9882329988233e-06, + "loss": 0.0662, + "step": 11774 + }, + { + "epoch": 5.474198047419804, + "grad_norm": 6.565121173858643, + "learning_rate": 9.9882319988232e-06, + "loss": 0.1889, + "step": 11775 + }, + { + "epoch": 5.474662947466295, + "grad_norm": 6.095057010650635, + "learning_rate": 9.9882309988231e-06, + "loss": 0.065, + "step": 11776 + }, + { + "epoch": 5.475127847512785, + "grad_norm": 4.42427396774292, + "learning_rate": 9.988229998823001e-06, + "loss": 0.1866, + "step": 11777 + }, + { + "epoch": 5.475592747559275, + "grad_norm": 3.817558765411377, + "learning_rate": 9.988228998822902e-06, + "loss": 0.1193, + "step": 11778 + }, + { + "epoch": 5.4760576476057645, + "grad_norm": 7.23942756652832, + "learning_rate": 9.9882279988228e-06, + "loss": 0.457, + "step": 11779 + }, + { + "epoch": 5.476522547652255, + "grad_norm": 3.459167718887329, + "learning_rate": 9.988226998822701e-06, + "loss": 0.1439, + "step": 11780 + }, + { + "epoch": 5.476987447698745, + "grad_norm": 4.6238484382629395, + "learning_rate": 9.9882259988226e-06, + "loss": 0.2442, + "step": 11781 + }, + { + "epoch": 5.477452347745235, + "grad_norm": 4.639708995819092, + "learning_rate": 9.988224998822501e-06, + "loss": 0.1649, + "step": 11782 + }, + { + "epoch": 5.477917247791725, + "grad_norm": 7.796631336212158, + "learning_rate": 9.9882239988224e-06, + "loss": 0.2115, + "step": 11783 + }, + { + "epoch": 5.478382147838214, + "grad_norm": 4.9806036949157715, + "learning_rate": 9.9882229988223e-06, + "loss": 0.2109, + "step": 11784 + }, + { + "epoch": 5.478847047884705, + "grad_norm": 3.990816116333008, + "learning_rate": 9.988221998822201e-06, + "loss": 0.1128, + "step": 11785 + }, + { + "epoch": 5.479311947931195, + "grad_norm": 8.111685752868652, + "learning_rate": 9.9882209988221e-06, + "loss": 0.3024, + "step": 11786 + }, + { + "epoch": 5.479776847977685, + "grad_norm": 3.0105104446411133, + "learning_rate": 9.988219998822001e-06, + "loss": 0.0775, + "step": 11787 + }, + { + "epoch": 5.4802417480241745, + "grad_norm": 4.849283218383789, + "learning_rate": 9.9882189988219e-06, + "loss": 0.1521, + "step": 11788 + }, + { + "epoch": 5.480706648070665, + "grad_norm": 4.538661479949951, + "learning_rate": 9.9882179988218e-06, + "loss": 0.1723, + "step": 11789 + }, + { + "epoch": 5.481171548117155, + "grad_norm": 5.594282627105713, + "learning_rate": 9.988216998821701e-06, + "loss": 0.0893, + "step": 11790 + }, + { + "epoch": 5.481636448163645, + "grad_norm": 3.692772626876831, + "learning_rate": 9.988215998821602e-06, + "loss": 0.1185, + "step": 11791 + }, + { + "epoch": 5.482101348210135, + "grad_norm": 3.652064800262451, + "learning_rate": 9.9882149988215e-06, + "loss": 0.0563, + "step": 11792 + }, + { + "epoch": 5.482566248256624, + "grad_norm": 6.843670845031738, + "learning_rate": 9.988213998821401e-06, + "loss": 0.3106, + "step": 11793 + }, + { + "epoch": 5.483031148303115, + "grad_norm": 5.205461502075195, + "learning_rate": 9.9882129988213e-06, + "loss": 0.2296, + "step": 11794 + }, + { + "epoch": 5.483496048349605, + "grad_norm": 3.569822311401367, + "learning_rate": 9.988211998821201e-06, + "loss": 0.0757, + "step": 11795 + }, + { + "epoch": 5.483960948396095, + "grad_norm": 5.7106475830078125, + "learning_rate": 9.9882109988211e-06, + "loss": 0.1605, + "step": 11796 + }, + { + "epoch": 5.4844258484425845, + "grad_norm": 4.050415992736816, + "learning_rate": 9.988209998821e-06, + "loss": 0.103, + "step": 11797 + }, + { + "epoch": 5.484890748489075, + "grad_norm": 8.021778106689453, + "learning_rate": 9.988208998820901e-06, + "loss": 0.1454, + "step": 11798 + }, + { + "epoch": 5.485355648535565, + "grad_norm": 5.518946647644043, + "learning_rate": 9.9882079988208e-06, + "loss": 0.2518, + "step": 11799 + }, + { + "epoch": 5.485820548582055, + "grad_norm": 4.853835105895996, + "learning_rate": 9.988206998820701e-06, + "loss": 0.1631, + "step": 11800 + }, + { + "epoch": 5.486285448628545, + "grad_norm": 5.438633918762207, + "learning_rate": 9.9882059988206e-06, + "loss": 0.178, + "step": 11801 + }, + { + "epoch": 5.486750348675034, + "grad_norm": 7.082118988037109, + "learning_rate": 9.9882049988205e-06, + "loss": 0.4128, + "step": 11802 + }, + { + "epoch": 5.487215248721525, + "grad_norm": 4.222187519073486, + "learning_rate": 9.988203998820401e-06, + "loss": 0.1126, + "step": 11803 + }, + { + "epoch": 5.487680148768015, + "grad_norm": 6.140316009521484, + "learning_rate": 9.988202998820302e-06, + "loss": 0.2213, + "step": 11804 + }, + { + "epoch": 5.488145048814505, + "grad_norm": 6.782186508178711, + "learning_rate": 9.9882019988202e-06, + "loss": 0.4073, + "step": 11805 + }, + { + "epoch": 5.4886099488609945, + "grad_norm": 6.400522232055664, + "learning_rate": 9.988200998820101e-06, + "loss": 0.2938, + "step": 11806 + }, + { + "epoch": 5.489074848907485, + "grad_norm": 5.592568874359131, + "learning_rate": 9.98819999882e-06, + "loss": 0.3251, + "step": 11807 + }, + { + "epoch": 5.489539748953975, + "grad_norm": 8.03565788269043, + "learning_rate": 9.9881989988199e-06, + "loss": 0.577, + "step": 11808 + }, + { + "epoch": 5.490004649000465, + "grad_norm": 6.66098165512085, + "learning_rate": 9.9881979988198e-06, + "loss": 0.2126, + "step": 11809 + }, + { + "epoch": 5.490469549046955, + "grad_norm": 3.691121816635132, + "learning_rate": 9.9881969988197e-06, + "loss": 0.1088, + "step": 11810 + }, + { + "epoch": 5.490934449093445, + "grad_norm": 5.926982879638672, + "learning_rate": 9.988195998819601e-06, + "loss": 0.278, + "step": 11811 + }, + { + "epoch": 5.491399349139935, + "grad_norm": 5.155257225036621, + "learning_rate": 9.9881949988195e-06, + "loss": 0.2056, + "step": 11812 + }, + { + "epoch": 5.491864249186425, + "grad_norm": 4.27501106262207, + "learning_rate": 9.988193998819401e-06, + "loss": 0.1507, + "step": 11813 + }, + { + "epoch": 5.492329149232915, + "grad_norm": 5.650412559509277, + "learning_rate": 9.9881929988193e-06, + "loss": 0.296, + "step": 11814 + }, + { + "epoch": 5.492794049279405, + "grad_norm": 6.310373306274414, + "learning_rate": 9.9881919988192e-06, + "loss": 0.2542, + "step": 11815 + }, + { + "epoch": 5.493258949325895, + "grad_norm": 7.011479377746582, + "learning_rate": 9.988190998819101e-06, + "loss": 0.2155, + "step": 11816 + }, + { + "epoch": 5.493723849372385, + "grad_norm": 4.234103202819824, + "learning_rate": 9.988189998819002e-06, + "loss": 0.1495, + "step": 11817 + }, + { + "epoch": 5.494188749418875, + "grad_norm": 6.587925910949707, + "learning_rate": 9.988188998818901e-06, + "loss": 0.4353, + "step": 11818 + }, + { + "epoch": 5.494653649465365, + "grad_norm": 4.469338417053223, + "learning_rate": 9.9881879988188e-06, + "loss": 0.1994, + "step": 11819 + }, + { + "epoch": 5.495118549511855, + "grad_norm": 6.708706378936768, + "learning_rate": 9.9881869988187e-06, + "loss": 0.271, + "step": 11820 + }, + { + "epoch": 5.495583449558345, + "grad_norm": 4.016130447387695, + "learning_rate": 9.9881859988186e-06, + "loss": 0.1693, + "step": 11821 + }, + { + "epoch": 5.496048349604835, + "grad_norm": 4.185458183288574, + "learning_rate": 9.9881849988185e-06, + "loss": 0.0565, + "step": 11822 + }, + { + "epoch": 5.496513249651325, + "grad_norm": 9.321775436401367, + "learning_rate": 9.9881839988184e-06, + "loss": 0.2075, + "step": 11823 + }, + { + "epoch": 5.496978149697815, + "grad_norm": 5.191105842590332, + "learning_rate": 9.988182998818301e-06, + "loss": 0.1097, + "step": 11824 + }, + { + "epoch": 5.497443049744305, + "grad_norm": 4.606657981872559, + "learning_rate": 9.9881819988182e-06, + "loss": 0.1356, + "step": 11825 + }, + { + "epoch": 5.497907949790795, + "grad_norm": 6.995884418487549, + "learning_rate": 9.988180998818101e-06, + "loss": 0.3957, + "step": 11826 + }, + { + "epoch": 5.498372849837285, + "grad_norm": 4.878742218017578, + "learning_rate": 9.988179998818e-06, + "loss": 0.2081, + "step": 11827 + }, + { + "epoch": 5.498837749883775, + "grad_norm": 5.461201190948486, + "learning_rate": 9.9881789988179e-06, + "loss": 0.3298, + "step": 11828 + }, + { + "epoch": 5.499302649930265, + "grad_norm": 5.635296821594238, + "learning_rate": 9.988177998817801e-06, + "loss": 0.2164, + "step": 11829 + }, + { + "epoch": 5.499767549976755, + "grad_norm": 4.079861640930176, + "learning_rate": 9.9881769988177e-06, + "loss": 0.1059, + "step": 11830 + }, + { + "epoch": 5.500232450023245, + "grad_norm": 5.396640777587891, + "learning_rate": 9.988175998817601e-06, + "loss": 0.3156, + "step": 11831 + }, + { + "epoch": 5.500697350069735, + "grad_norm": 7.124979496002197, + "learning_rate": 9.9881749988175e-06, + "loss": 0.1353, + "step": 11832 + }, + { + "epoch": 5.501162250116225, + "grad_norm": 3.081632614135742, + "learning_rate": 9.9881739988174e-06, + "loss": 0.1117, + "step": 11833 + }, + { + "epoch": 5.501627150162715, + "grad_norm": 4.656660556793213, + "learning_rate": 9.9881729988173e-06, + "loss": 0.2017, + "step": 11834 + }, + { + "epoch": 5.502092050209205, + "grad_norm": 4.697283744812012, + "learning_rate": 9.9881719988172e-06, + "loss": 0.2431, + "step": 11835 + }, + { + "epoch": 5.502556950255695, + "grad_norm": 5.10834264755249, + "learning_rate": 9.988170998817101e-06, + "loss": 0.2153, + "step": 11836 + }, + { + "epoch": 5.503021850302185, + "grad_norm": 4.985957145690918, + "learning_rate": 9.988169998817002e-06, + "loss": 0.2325, + "step": 11837 + }, + { + "epoch": 5.503486750348675, + "grad_norm": 6.461579322814941, + "learning_rate": 9.9881689988169e-06, + "loss": 0.2633, + "step": 11838 + }, + { + "epoch": 5.503951650395165, + "grad_norm": 5.618546962738037, + "learning_rate": 9.988167998816801e-06, + "loss": 0.2357, + "step": 11839 + }, + { + "epoch": 5.504416550441655, + "grad_norm": 6.858516693115234, + "learning_rate": 9.9881669988167e-06, + "loss": 0.3445, + "step": 11840 + }, + { + "epoch": 5.504881450488145, + "grad_norm": 4.182948589324951, + "learning_rate": 9.9881659988166e-06, + "loss": 0.0949, + "step": 11841 + }, + { + "epoch": 5.505346350534635, + "grad_norm": 2.9011878967285156, + "learning_rate": 9.988164998816501e-06, + "loss": 0.0616, + "step": 11842 + }, + { + "epoch": 5.505811250581125, + "grad_norm": 3.0179457664489746, + "learning_rate": 9.9881639988164e-06, + "loss": 0.0939, + "step": 11843 + }, + { + "epoch": 5.506276150627615, + "grad_norm": 5.250738620758057, + "learning_rate": 9.988162998816301e-06, + "loss": 0.2127, + "step": 11844 + }, + { + "epoch": 5.506741050674105, + "grad_norm": 6.4255690574646, + "learning_rate": 9.9881619988162e-06, + "loss": 0.2699, + "step": 11845 + }, + { + "epoch": 5.507205950720595, + "grad_norm": 4.9578399658203125, + "learning_rate": 9.9881609988161e-06, + "loss": 0.2153, + "step": 11846 + }, + { + "epoch": 5.507670850767085, + "grad_norm": 6.205576419830322, + "learning_rate": 9.988159998816001e-06, + "loss": 0.3232, + "step": 11847 + }, + { + "epoch": 5.508135750813575, + "grad_norm": 4.117396354675293, + "learning_rate": 9.9881589988159e-06, + "loss": 0.1836, + "step": 11848 + }, + { + "epoch": 5.508600650860065, + "grad_norm": 4.691512584686279, + "learning_rate": 9.988157998815801e-06, + "loss": 0.1581, + "step": 11849 + }, + { + "epoch": 5.509065550906556, + "grad_norm": 4.90568208694458, + "learning_rate": 9.988156998815702e-06, + "loss": 0.1923, + "step": 11850 + }, + { + "epoch": 5.509530450953045, + "grad_norm": 3.795103073120117, + "learning_rate": 9.9881559988156e-06, + "loss": 0.1176, + "step": 11851 + }, + { + "epoch": 5.509995350999535, + "grad_norm": 9.500988960266113, + "learning_rate": 9.9881549988155e-06, + "loss": 0.7678, + "step": 11852 + }, + { + "epoch": 5.510460251046025, + "grad_norm": 7.213641166687012, + "learning_rate": 9.9881539988154e-06, + "loss": 0.1281, + "step": 11853 + }, + { + "epoch": 5.510925151092515, + "grad_norm": 8.590194702148438, + "learning_rate": 9.988152998815301e-06, + "loss": 0.3213, + "step": 11854 + }, + { + "epoch": 5.5113900511390055, + "grad_norm": 4.973840713500977, + "learning_rate": 9.988151998815202e-06, + "loss": 0.2132, + "step": 11855 + }, + { + "epoch": 5.511854951185495, + "grad_norm": 6.186428070068359, + "learning_rate": 9.9881509988151e-06, + "loss": 0.4591, + "step": 11856 + }, + { + "epoch": 5.512319851231985, + "grad_norm": 7.787641525268555, + "learning_rate": 9.988149998815001e-06, + "loss": 0.3631, + "step": 11857 + }, + { + "epoch": 5.512784751278475, + "grad_norm": 4.858275413513184, + "learning_rate": 9.9881489988149e-06, + "loss": 0.2367, + "step": 11858 + }, + { + "epoch": 5.513249651324966, + "grad_norm": 5.323333263397217, + "learning_rate": 9.9881479988148e-06, + "loss": 0.2803, + "step": 11859 + }, + { + "epoch": 5.513714551371455, + "grad_norm": 3.536757230758667, + "learning_rate": 9.988146998814701e-06, + "loss": 0.1035, + "step": 11860 + }, + { + "epoch": 5.514179451417945, + "grad_norm": 7.384530067443848, + "learning_rate": 9.9881459988146e-06, + "loss": 0.1451, + "step": 11861 + }, + { + "epoch": 5.514644351464435, + "grad_norm": 10.221990585327148, + "learning_rate": 9.988144998814501e-06, + "loss": 0.2582, + "step": 11862 + }, + { + "epoch": 5.515109251510925, + "grad_norm": 6.540007591247559, + "learning_rate": 9.9881439988144e-06, + "loss": 0.2968, + "step": 11863 + }, + { + "epoch": 5.5155741515574155, + "grad_norm": 4.191531658172607, + "learning_rate": 9.9881429988143e-06, + "loss": 0.1826, + "step": 11864 + }, + { + "epoch": 5.516039051603905, + "grad_norm": 7.191685199737549, + "learning_rate": 9.9881419988142e-06, + "loss": 0.3109, + "step": 11865 + }, + { + "epoch": 5.516503951650395, + "grad_norm": 7.243817329406738, + "learning_rate": 9.9881409988141e-06, + "loss": 0.2709, + "step": 11866 + }, + { + "epoch": 5.516968851696885, + "grad_norm": 5.22511100769043, + "learning_rate": 9.988139998814001e-06, + "loss": 0.2002, + "step": 11867 + }, + { + "epoch": 5.517433751743376, + "grad_norm": 7.335038661956787, + "learning_rate": 9.988138998813902e-06, + "loss": 0.3264, + "step": 11868 + }, + { + "epoch": 5.517898651789865, + "grad_norm": 3.1621041297912598, + "learning_rate": 9.9881379988138e-06, + "loss": 0.0653, + "step": 11869 + }, + { + "epoch": 5.518363551836355, + "grad_norm": 4.706296920776367, + "learning_rate": 9.988136998813701e-06, + "loss": 0.2893, + "step": 11870 + }, + { + "epoch": 5.518828451882845, + "grad_norm": 3.92549204826355, + "learning_rate": 9.9881359988136e-06, + "loss": 0.121, + "step": 11871 + }, + { + "epoch": 5.519293351929335, + "grad_norm": 3.8971526622772217, + "learning_rate": 9.988134998813501e-06, + "loss": 0.1477, + "step": 11872 + }, + { + "epoch": 5.5197582519758255, + "grad_norm": 3.6961052417755127, + "learning_rate": 9.988133998813402e-06, + "loss": 0.1185, + "step": 11873 + }, + { + "epoch": 5.520223152022315, + "grad_norm": 6.976910591125488, + "learning_rate": 9.9881329988133e-06, + "loss": 0.2608, + "step": 11874 + }, + { + "epoch": 5.520688052068805, + "grad_norm": 5.267163276672363, + "learning_rate": 9.988131998813201e-06, + "loss": 0.3029, + "step": 11875 + }, + { + "epoch": 5.521152952115295, + "grad_norm": 5.683195114135742, + "learning_rate": 9.9881309988131e-06, + "loss": 0.1988, + "step": 11876 + }, + { + "epoch": 5.521617852161786, + "grad_norm": 6.222949504852295, + "learning_rate": 9.988129998813e-06, + "loss": 0.2682, + "step": 11877 + }, + { + "epoch": 5.522082752208275, + "grad_norm": 4.060539245605469, + "learning_rate": 9.9881289988129e-06, + "loss": 0.1734, + "step": 11878 + }, + { + "epoch": 5.522547652254765, + "grad_norm": 3.891127347946167, + "learning_rate": 9.9881279988128e-06, + "loss": 0.0768, + "step": 11879 + }, + { + "epoch": 5.523012552301255, + "grad_norm": 4.3992743492126465, + "learning_rate": 9.988126998812701e-06, + "loss": 0.2283, + "step": 11880 + }, + { + "epoch": 5.523477452347745, + "grad_norm": 5.169523239135742, + "learning_rate": 9.988125998812602e-06, + "loss": 0.1615, + "step": 11881 + }, + { + "epoch": 5.5239423523942355, + "grad_norm": 8.273761749267578, + "learning_rate": 9.9881249988125e-06, + "loss": 0.5121, + "step": 11882 + }, + { + "epoch": 5.524407252440725, + "grad_norm": 4.372474193572998, + "learning_rate": 9.988123998812401e-06, + "loss": 0.1653, + "step": 11883 + }, + { + "epoch": 5.524872152487215, + "grad_norm": 5.379973411560059, + "learning_rate": 9.9881229988123e-06, + "loss": 0.2195, + "step": 11884 + }, + { + "epoch": 5.525337052533705, + "grad_norm": 5.251799583435059, + "learning_rate": 9.988121998812201e-06, + "loss": 0.2436, + "step": 11885 + }, + { + "epoch": 5.525801952580196, + "grad_norm": 7.376361846923828, + "learning_rate": 9.988120998812102e-06, + "loss": 0.3032, + "step": 11886 + }, + { + "epoch": 5.526266852626685, + "grad_norm": 7.0229878425598145, + "learning_rate": 9.988119998812e-06, + "loss": 0.268, + "step": 11887 + }, + { + "epoch": 5.526731752673175, + "grad_norm": 5.505502223968506, + "learning_rate": 9.988118998811901e-06, + "loss": 0.2361, + "step": 11888 + }, + { + "epoch": 5.527196652719665, + "grad_norm": 5.1093645095825195, + "learning_rate": 9.9881179988118e-06, + "loss": 0.2238, + "step": 11889 + }, + { + "epoch": 5.527661552766155, + "grad_norm": 5.2623724937438965, + "learning_rate": 9.9881169988117e-06, + "loss": 0.1646, + "step": 11890 + }, + { + "epoch": 5.5281264528126455, + "grad_norm": 4.476867198944092, + "learning_rate": 9.9881159988116e-06, + "loss": 0.2154, + "step": 11891 + }, + { + "epoch": 5.528591352859135, + "grad_norm": 8.415599822998047, + "learning_rate": 9.9881149988115e-06, + "loss": 0.3784, + "step": 11892 + }, + { + "epoch": 5.529056252905625, + "grad_norm": 5.897237777709961, + "learning_rate": 9.988113998811401e-06, + "loss": 0.2911, + "step": 11893 + }, + { + "epoch": 5.529521152952116, + "grad_norm": 4.824792385101318, + "learning_rate": 9.988112998811302e-06, + "loss": 0.2269, + "step": 11894 + }, + { + "epoch": 5.529986052998606, + "grad_norm": 6.974609851837158, + "learning_rate": 9.9881119988112e-06, + "loss": 0.2488, + "step": 11895 + }, + { + "epoch": 5.530450953045095, + "grad_norm": 5.139242649078369, + "learning_rate": 9.9881109988111e-06, + "loss": 0.1539, + "step": 11896 + }, + { + "epoch": 5.530915853091585, + "grad_norm": 5.204833030700684, + "learning_rate": 9.988109998811e-06, + "loss": 0.1862, + "step": 11897 + }, + { + "epoch": 5.531380753138075, + "grad_norm": 5.261721611022949, + "learning_rate": 9.988108998810901e-06, + "loss": 0.2463, + "step": 11898 + }, + { + "epoch": 5.531845653184566, + "grad_norm": 6.926687240600586, + "learning_rate": 9.988107998810802e-06, + "loss": 0.3576, + "step": 11899 + }, + { + "epoch": 5.5323105532310555, + "grad_norm": 7.111121654510498, + "learning_rate": 9.9881069988107e-06, + "loss": 0.21, + "step": 11900 + }, + { + "epoch": 5.532775453277545, + "grad_norm": 5.247806072235107, + "learning_rate": 9.988105998810601e-06, + "loss": 0.2491, + "step": 11901 + }, + { + "epoch": 5.533240353324035, + "grad_norm": 4.127379417419434, + "learning_rate": 9.9881049988105e-06, + "loss": 0.1281, + "step": 11902 + }, + { + "epoch": 5.533705253370526, + "grad_norm": 6.000222206115723, + "learning_rate": 9.988103998810401e-06, + "loss": 0.4018, + "step": 11903 + }, + { + "epoch": 5.534170153417016, + "grad_norm": 5.28493070602417, + "learning_rate": 9.9881029988103e-06, + "loss": 0.2214, + "step": 11904 + }, + { + "epoch": 5.534635053463505, + "grad_norm": 5.788666248321533, + "learning_rate": 9.9881019988102e-06, + "loss": 0.2997, + "step": 11905 + }, + { + "epoch": 5.535099953509995, + "grad_norm": 6.052770137786865, + "learning_rate": 9.988100998810101e-06, + "loss": 0.257, + "step": 11906 + }, + { + "epoch": 5.535564853556485, + "grad_norm": 3.6280055046081543, + "learning_rate": 9.98809999881e-06, + "loss": 0.1185, + "step": 11907 + }, + { + "epoch": 5.536029753602976, + "grad_norm": 6.036645412445068, + "learning_rate": 9.9880989988099e-06, + "loss": 0.2909, + "step": 11908 + }, + { + "epoch": 5.5364946536494655, + "grad_norm": 1.773085355758667, + "learning_rate": 9.9880979988098e-06, + "loss": 0.0262, + "step": 11909 + }, + { + "epoch": 5.536959553695955, + "grad_norm": 7.551063060760498, + "learning_rate": 9.9880969988097e-06, + "loss": 0.3751, + "step": 11910 + }, + { + "epoch": 5.537424453742445, + "grad_norm": 3.8692495822906494, + "learning_rate": 9.988095998809601e-06, + "loss": 0.1397, + "step": 11911 + }, + { + "epoch": 5.537889353788936, + "grad_norm": 6.455897331237793, + "learning_rate": 9.988094998809502e-06, + "loss": 0.292, + "step": 11912 + }, + { + "epoch": 5.538354253835426, + "grad_norm": 8.9745454788208, + "learning_rate": 9.9880939988094e-06, + "loss": 0.3041, + "step": 11913 + }, + { + "epoch": 5.538819153881915, + "grad_norm": 4.355361461639404, + "learning_rate": 9.988092998809301e-06, + "loss": 0.0872, + "step": 11914 + }, + { + "epoch": 5.539284053928405, + "grad_norm": 3.4168314933776855, + "learning_rate": 9.9880919988092e-06, + "loss": 0.0627, + "step": 11915 + }, + { + "epoch": 5.539748953974895, + "grad_norm": 4.269414901733398, + "learning_rate": 9.988090998809101e-06, + "loss": 0.1274, + "step": 11916 + }, + { + "epoch": 5.540213854021386, + "grad_norm": 6.159637928009033, + "learning_rate": 9.988089998809e-06, + "loss": 0.3049, + "step": 11917 + }, + { + "epoch": 5.5406787540678755, + "grad_norm": 5.626351833343506, + "learning_rate": 9.9880889988089e-06, + "loss": 0.2121, + "step": 11918 + }, + { + "epoch": 5.541143654114365, + "grad_norm": 5.216669082641602, + "learning_rate": 9.988087998808801e-06, + "loss": 0.2103, + "step": 11919 + }, + { + "epoch": 5.541608554160855, + "grad_norm": 5.6927490234375, + "learning_rate": 9.9880869988087e-06, + "loss": 0.1462, + "step": 11920 + }, + { + "epoch": 5.542073454207346, + "grad_norm": 4.363626003265381, + "learning_rate": 9.988085998808601e-06, + "loss": 0.1308, + "step": 11921 + }, + { + "epoch": 5.542538354253836, + "grad_norm": 5.538067817687988, + "learning_rate": 9.9880849988085e-06, + "loss": 0.2659, + "step": 11922 + }, + { + "epoch": 5.543003254300325, + "grad_norm": 5.269493103027344, + "learning_rate": 9.9880839988084e-06, + "loss": 0.2547, + "step": 11923 + }, + { + "epoch": 5.543468154346815, + "grad_norm": 4.736871719360352, + "learning_rate": 9.988082998808301e-06, + "loss": 0.1981, + "step": 11924 + }, + { + "epoch": 5.543933054393305, + "grad_norm": 5.928048610687256, + "learning_rate": 9.988081998808202e-06, + "loss": 0.2378, + "step": 11925 + }, + { + "epoch": 5.544397954439796, + "grad_norm": 6.308376789093018, + "learning_rate": 9.9880809988081e-06, + "loss": 0.2936, + "step": 11926 + }, + { + "epoch": 5.5448628544862855, + "grad_norm": 4.531064510345459, + "learning_rate": 9.988079998808e-06, + "loss": 0.1084, + "step": 11927 + }, + { + "epoch": 5.545327754532775, + "grad_norm": 6.277875900268555, + "learning_rate": 9.9880789988079e-06, + "loss": 0.2088, + "step": 11928 + }, + { + "epoch": 5.545792654579266, + "grad_norm": 4.898411750793457, + "learning_rate": 9.988077998807801e-06, + "loss": 0.2343, + "step": 11929 + }, + { + "epoch": 5.546257554625756, + "grad_norm": 4.626470565795898, + "learning_rate": 9.9880769988077e-06, + "loss": 0.1645, + "step": 11930 + }, + { + "epoch": 5.546722454672246, + "grad_norm": 6.30303430557251, + "learning_rate": 9.9880759988076e-06, + "loss": 0.372, + "step": 11931 + }, + { + "epoch": 5.547187354718735, + "grad_norm": 4.8220133781433105, + "learning_rate": 9.988074998807501e-06, + "loss": 0.1731, + "step": 11932 + }, + { + "epoch": 5.547652254765225, + "grad_norm": 4.677704811096191, + "learning_rate": 9.9880739988074e-06, + "loss": 0.1866, + "step": 11933 + }, + { + "epoch": 5.548117154811716, + "grad_norm": 3.8809196949005127, + "learning_rate": 9.988072998807301e-06, + "loss": 0.1769, + "step": 11934 + }, + { + "epoch": 5.548582054858206, + "grad_norm": 2.980208396911621, + "learning_rate": 9.9880719988072e-06, + "loss": 0.0718, + "step": 11935 + }, + { + "epoch": 5.5490469549046955, + "grad_norm": 5.099897384643555, + "learning_rate": 9.9880709988071e-06, + "loss": 0.1815, + "step": 11936 + }, + { + "epoch": 5.549511854951185, + "grad_norm": 8.035919189453125, + "learning_rate": 9.988069998807001e-06, + "loss": 0.3116, + "step": 11937 + }, + { + "epoch": 5.549976754997676, + "grad_norm": 5.187145233154297, + "learning_rate": 9.9880689988069e-06, + "loss": 0.0903, + "step": 11938 + }, + { + "epoch": 5.550441655044166, + "grad_norm": 7.045681953430176, + "learning_rate": 9.988067998806801e-06, + "loss": 0.3244, + "step": 11939 + }, + { + "epoch": 5.550906555090656, + "grad_norm": 5.2543487548828125, + "learning_rate": 9.9880669988067e-06, + "loss": 0.3014, + "step": 11940 + }, + { + "epoch": 5.551371455137145, + "grad_norm": 8.45932674407959, + "learning_rate": 9.9880659988066e-06, + "loss": 0.2214, + "step": 11941 + }, + { + "epoch": 5.551836355183635, + "grad_norm": 2.5649471282958984, + "learning_rate": 9.988064998806501e-06, + "loss": 0.0417, + "step": 11942 + }, + { + "epoch": 5.552301255230126, + "grad_norm": 3.0446481704711914, + "learning_rate": 9.988063998806402e-06, + "loss": 0.1122, + "step": 11943 + }, + { + "epoch": 5.552766155276616, + "grad_norm": 3.9455370903015137, + "learning_rate": 9.9880629988063e-06, + "loss": 0.1535, + "step": 11944 + }, + { + "epoch": 5.5532310553231055, + "grad_norm": 4.811606407165527, + "learning_rate": 9.988061998806201e-06, + "loss": 0.0999, + "step": 11945 + }, + { + "epoch": 5.553695955369595, + "grad_norm": 7.993954658508301, + "learning_rate": 9.9880609988061e-06, + "loss": 0.1822, + "step": 11946 + }, + { + "epoch": 5.554160855416086, + "grad_norm": 7.21120023727417, + "learning_rate": 9.988059998806001e-06, + "loss": 0.378, + "step": 11947 + }, + { + "epoch": 5.554625755462576, + "grad_norm": 5.650484561920166, + "learning_rate": 9.9880589988059e-06, + "loss": 0.2344, + "step": 11948 + }, + { + "epoch": 5.5550906555090656, + "grad_norm": 6.251757621765137, + "learning_rate": 9.9880579988058e-06, + "loss": 0.2908, + "step": 11949 + }, + { + "epoch": 5.555555555555555, + "grad_norm": 4.847829818725586, + "learning_rate": 9.988056998805701e-06, + "loss": 0.1665, + "step": 11950 + }, + { + "epoch": 5.556020455602045, + "grad_norm": 6.858452796936035, + "learning_rate": 9.9880559988056e-06, + "loss": 0.2072, + "step": 11951 + }, + { + "epoch": 5.556485355648536, + "grad_norm": 6.203792095184326, + "learning_rate": 9.988054998805501e-06, + "loss": 0.231, + "step": 11952 + }, + { + "epoch": 5.556950255695026, + "grad_norm": 1.8342820405960083, + "learning_rate": 9.9880539988054e-06, + "loss": 0.0542, + "step": 11953 + }, + { + "epoch": 5.5574151557415155, + "grad_norm": 5.448204040527344, + "learning_rate": 9.9880529988053e-06, + "loss": 0.3003, + "step": 11954 + }, + { + "epoch": 5.557880055788005, + "grad_norm": 6.673495292663574, + "learning_rate": 9.988051998805201e-06, + "loss": 0.4916, + "step": 11955 + }, + { + "epoch": 5.558344955834496, + "grad_norm": 6.343385219573975, + "learning_rate": 9.988050998805102e-06, + "loss": 0.1367, + "step": 11956 + }, + { + "epoch": 5.558809855880986, + "grad_norm": 6.160974979400635, + "learning_rate": 9.988049998805001e-06, + "loss": 0.2448, + "step": 11957 + }, + { + "epoch": 5.5592747559274756, + "grad_norm": 6.0344414710998535, + "learning_rate": 9.988048998804902e-06, + "loss": 0.175, + "step": 11958 + }, + { + "epoch": 5.559739655973965, + "grad_norm": 5.9064764976501465, + "learning_rate": 9.9880479988048e-06, + "loss": 0.2973, + "step": 11959 + }, + { + "epoch": 5.560204556020455, + "grad_norm": 5.440276145935059, + "learning_rate": 9.9880469988047e-06, + "loss": 0.0849, + "step": 11960 + }, + { + "epoch": 5.560669456066946, + "grad_norm": 4.295125484466553, + "learning_rate": 9.9880459988046e-06, + "loss": 0.1681, + "step": 11961 + }, + { + "epoch": 5.561134356113436, + "grad_norm": 5.259362697601318, + "learning_rate": 9.9880449988045e-06, + "loss": 0.122, + "step": 11962 + }, + { + "epoch": 5.5615992561599255, + "grad_norm": 6.638518333435059, + "learning_rate": 9.988043998804401e-06, + "loss": 0.4045, + "step": 11963 + }, + { + "epoch": 5.562064156206415, + "grad_norm": 16.462135314941406, + "learning_rate": 9.9880429988043e-06, + "loss": 0.4857, + "step": 11964 + }, + { + "epoch": 5.562529056252906, + "grad_norm": 6.3951826095581055, + "learning_rate": 9.988041998804201e-06, + "loss": 0.3364, + "step": 11965 + }, + { + "epoch": 5.562993956299396, + "grad_norm": 6.560571670532227, + "learning_rate": 9.9880409988041e-06, + "loss": 0.2726, + "step": 11966 + }, + { + "epoch": 5.5634588563458856, + "grad_norm": 7.113731861114502, + "learning_rate": 9.988039998804e-06, + "loss": 0.2186, + "step": 11967 + }, + { + "epoch": 5.563923756392375, + "grad_norm": 6.85111141204834, + "learning_rate": 9.988038998803901e-06, + "loss": 0.3803, + "step": 11968 + }, + { + "epoch": 5.564388656438865, + "grad_norm": 3.871673345565796, + "learning_rate": 9.988037998803802e-06, + "loss": 0.2303, + "step": 11969 + }, + { + "epoch": 5.564853556485356, + "grad_norm": 4.933372497558594, + "learning_rate": 9.988036998803701e-06, + "loss": 0.1334, + "step": 11970 + }, + { + "epoch": 5.565318456531846, + "grad_norm": 3.8796122074127197, + "learning_rate": 9.9880359988036e-06, + "loss": 0.1231, + "step": 11971 + }, + { + "epoch": 5.5657833565783355, + "grad_norm": 4.1396098136901855, + "learning_rate": 9.9880349988035e-06, + "loss": 0.2088, + "step": 11972 + }, + { + "epoch": 5.566248256624826, + "grad_norm": 5.625781536102295, + "learning_rate": 9.9880339988034e-06, + "loss": 0.2502, + "step": 11973 + }, + { + "epoch": 5.566713156671316, + "grad_norm": 7.201853275299072, + "learning_rate": 9.9880329988033e-06, + "loss": 0.2476, + "step": 11974 + }, + { + "epoch": 5.567178056717806, + "grad_norm": 4.139517784118652, + "learning_rate": 9.9880319988032e-06, + "loss": 0.1787, + "step": 11975 + }, + { + "epoch": 5.5676429567642955, + "grad_norm": 4.134814262390137, + "learning_rate": 9.988030998803102e-06, + "loss": 0.2007, + "step": 11976 + }, + { + "epoch": 5.568107856810785, + "grad_norm": 4.816324710845947, + "learning_rate": 9.988029998803e-06, + "loss": 0.1463, + "step": 11977 + }, + { + "epoch": 5.568572756857276, + "grad_norm": 7.482572078704834, + "learning_rate": 9.988028998802901e-06, + "loss": 0.401, + "step": 11978 + }, + { + "epoch": 5.569037656903766, + "grad_norm": 7.139183044433594, + "learning_rate": 9.9880279988028e-06, + "loss": 0.4422, + "step": 11979 + }, + { + "epoch": 5.569502556950256, + "grad_norm": 5.719149589538574, + "learning_rate": 9.9880269988027e-06, + "loss": 0.1902, + "step": 11980 + }, + { + "epoch": 5.5699674569967454, + "grad_norm": 5.741128921508789, + "learning_rate": 9.988025998802601e-06, + "loss": 0.2868, + "step": 11981 + }, + { + "epoch": 5.570432357043236, + "grad_norm": 6.3551483154296875, + "learning_rate": 9.9880249988025e-06, + "loss": 0.0982, + "step": 11982 + }, + { + "epoch": 5.570897257089726, + "grad_norm": 3.1773722171783447, + "learning_rate": 9.988023998802401e-06, + "loss": 0.1109, + "step": 11983 + }, + { + "epoch": 5.571362157136216, + "grad_norm": 6.149084568023682, + "learning_rate": 9.9880229988023e-06, + "loss": 0.2407, + "step": 11984 + }, + { + "epoch": 5.5718270571827055, + "grad_norm": 3.799926519393921, + "learning_rate": 9.9880219988022e-06, + "loss": 0.1926, + "step": 11985 + }, + { + "epoch": 5.572291957229195, + "grad_norm": 5.30614709854126, + "learning_rate": 9.9880209988021e-06, + "loss": 0.2305, + "step": 11986 + }, + { + "epoch": 5.572756857275686, + "grad_norm": 3.924403429031372, + "learning_rate": 9.988019998802e-06, + "loss": 0.1809, + "step": 11987 + }, + { + "epoch": 5.573221757322176, + "grad_norm": 5.402607440948486, + "learning_rate": 9.988018998801901e-06, + "loss": 0.2037, + "step": 11988 + }, + { + "epoch": 5.573686657368666, + "grad_norm": 6.410706996917725, + "learning_rate": 9.988017998801802e-06, + "loss": 0.2989, + "step": 11989 + }, + { + "epoch": 5.5741515574151554, + "grad_norm": 7.541596412658691, + "learning_rate": 9.9880169988017e-06, + "loss": 0.3066, + "step": 11990 + }, + { + "epoch": 5.574616457461646, + "grad_norm": 5.701383113861084, + "learning_rate": 9.988015998801601e-06, + "loss": 0.3742, + "step": 11991 + }, + { + "epoch": 5.575081357508136, + "grad_norm": 6.351964950561523, + "learning_rate": 9.9880149988015e-06, + "loss": 0.158, + "step": 11992 + }, + { + "epoch": 5.575546257554626, + "grad_norm": 4.90049934387207, + "learning_rate": 9.9880139988014e-06, + "loss": 0.0909, + "step": 11993 + }, + { + "epoch": 5.5760111576011155, + "grad_norm": 3.4115400314331055, + "learning_rate": 9.988012998801301e-06, + "loss": 0.1254, + "step": 11994 + }, + { + "epoch": 5.576476057647605, + "grad_norm": 3.5040364265441895, + "learning_rate": 9.9880119988012e-06, + "loss": 0.1108, + "step": 11995 + }, + { + "epoch": 5.576940957694096, + "grad_norm": 4.115536689758301, + "learning_rate": 9.988010998801101e-06, + "loss": 0.1221, + "step": 11996 + }, + { + "epoch": 5.577405857740586, + "grad_norm": 5.245213031768799, + "learning_rate": 9.988009998801e-06, + "loss": 0.2318, + "step": 11997 + }, + { + "epoch": 5.577870757787076, + "grad_norm": 5.289319038391113, + "learning_rate": 9.9880089988009e-06, + "loss": 0.2141, + "step": 11998 + }, + { + "epoch": 5.578335657833565, + "grad_norm": 6.522698879241943, + "learning_rate": 9.9880079988008e-06, + "loss": 0.2728, + "step": 11999 + }, + { + "epoch": 5.578800557880056, + "grad_norm": 4.338969707489014, + "learning_rate": 9.9880069988007e-06, + "loss": 0.0986, + "step": 12000 + }, + { + "epoch": 5.579265457926546, + "grad_norm": 5.908843994140625, + "learning_rate": 9.988005998800601e-06, + "loss": 0.2932, + "step": 12001 + }, + { + "epoch": 5.579730357973036, + "grad_norm": 6.4242658615112305, + "learning_rate": 9.988004998800502e-06, + "loss": 0.3603, + "step": 12002 + }, + { + "epoch": 5.5801952580195255, + "grad_norm": 3.594499349594116, + "learning_rate": 9.9880039988004e-06, + "loss": 0.0876, + "step": 12003 + }, + { + "epoch": 5.580660158066015, + "grad_norm": 6.688094139099121, + "learning_rate": 9.9880029988003e-06, + "loss": 0.4038, + "step": 12004 + }, + { + "epoch": 5.581125058112506, + "grad_norm": 3.976475954055786, + "learning_rate": 9.9880019988002e-06, + "loss": 0.087, + "step": 12005 + }, + { + "epoch": 5.581589958158996, + "grad_norm": 7.769976615905762, + "learning_rate": 9.988000998800101e-06, + "loss": 0.4691, + "step": 12006 + }, + { + "epoch": 5.582054858205486, + "grad_norm": 7.552331447601318, + "learning_rate": 9.987999998800002e-06, + "loss": 0.2521, + "step": 12007 + }, + { + "epoch": 5.582519758251975, + "grad_norm": 6.582062721252441, + "learning_rate": 9.9879989987999e-06, + "loss": 0.2505, + "step": 12008 + }, + { + "epoch": 5.582984658298466, + "grad_norm": 2.571413278579712, + "learning_rate": 9.987997998799801e-06, + "loss": 0.0635, + "step": 12009 + }, + { + "epoch": 5.583449558344956, + "grad_norm": 7.887553691864014, + "learning_rate": 9.9879969987997e-06, + "loss": 0.1486, + "step": 12010 + }, + { + "epoch": 5.583914458391446, + "grad_norm": 5.926156044006348, + "learning_rate": 9.9879959987996e-06, + "loss": 0.2351, + "step": 12011 + }, + { + "epoch": 5.5843793584379355, + "grad_norm": 2.3953189849853516, + "learning_rate": 9.987994998799501e-06, + "loss": 0.0665, + "step": 12012 + }, + { + "epoch": 5.584844258484426, + "grad_norm": 7.414058685302734, + "learning_rate": 9.9879939987994e-06, + "loss": 0.1637, + "step": 12013 + }, + { + "epoch": 5.585309158530916, + "grad_norm": 6.253831386566162, + "learning_rate": 9.987992998799301e-06, + "loss": 0.3472, + "step": 12014 + }, + { + "epoch": 5.585774058577406, + "grad_norm": 3.8922159671783447, + "learning_rate": 9.9879919987992e-06, + "loss": 0.1473, + "step": 12015 + }, + { + "epoch": 5.586238958623896, + "grad_norm": 5.595746994018555, + "learning_rate": 9.9879909987991e-06, + "loss": 0.2124, + "step": 12016 + }, + { + "epoch": 5.586703858670386, + "grad_norm": 4.584840297698975, + "learning_rate": 9.987989998799e-06, + "loss": 0.1616, + "step": 12017 + }, + { + "epoch": 5.587168758716876, + "grad_norm": 4.873195648193359, + "learning_rate": 9.9879889987989e-06, + "loss": 0.1639, + "step": 12018 + }, + { + "epoch": 5.587633658763366, + "grad_norm": 8.109397888183594, + "learning_rate": 9.987987998798801e-06, + "loss": 0.3381, + "step": 12019 + }, + { + "epoch": 5.588098558809856, + "grad_norm": 8.836730003356934, + "learning_rate": 9.987986998798702e-06, + "loss": 0.2453, + "step": 12020 + }, + { + "epoch": 5.5885634588563455, + "grad_norm": 7.81411600112915, + "learning_rate": 9.9879859987986e-06, + "loss": 0.2018, + "step": 12021 + }, + { + "epoch": 5.589028358902836, + "grad_norm": 4.396930694580078, + "learning_rate": 9.987984998798501e-06, + "loss": 0.2209, + "step": 12022 + }, + { + "epoch": 5.589493258949326, + "grad_norm": 6.004605293273926, + "learning_rate": 9.9879839987984e-06, + "loss": 0.4254, + "step": 12023 + }, + { + "epoch": 5.589958158995816, + "grad_norm": 6.1044793128967285, + "learning_rate": 9.987982998798301e-06, + "loss": 0.259, + "step": 12024 + }, + { + "epoch": 5.590423059042306, + "grad_norm": 4.235157012939453, + "learning_rate": 9.987981998798202e-06, + "loss": 0.0801, + "step": 12025 + }, + { + "epoch": 5.590887959088796, + "grad_norm": 6.391194820404053, + "learning_rate": 9.9879809987981e-06, + "loss": 0.1718, + "step": 12026 + }, + { + "epoch": 5.591352859135286, + "grad_norm": 3.2365808486938477, + "learning_rate": 9.987979998798001e-06, + "loss": 0.0802, + "step": 12027 + }, + { + "epoch": 5.591817759181776, + "grad_norm": 6.244013786315918, + "learning_rate": 9.9879789987979e-06, + "loss": 0.3295, + "step": 12028 + }, + { + "epoch": 5.592282659228266, + "grad_norm": 11.217549324035645, + "learning_rate": 9.9879779987978e-06, + "loss": 0.2983, + "step": 12029 + }, + { + "epoch": 5.5927475592747555, + "grad_norm": 5.5549397468566895, + "learning_rate": 9.9879769987977e-06, + "loss": 0.2498, + "step": 12030 + }, + { + "epoch": 5.593212459321246, + "grad_norm": 6.322498798370361, + "learning_rate": 9.9879759987976e-06, + "loss": 0.226, + "step": 12031 + }, + { + "epoch": 5.593677359367736, + "grad_norm": 5.346972465515137, + "learning_rate": 9.987974998797501e-06, + "loss": 0.2188, + "step": 12032 + }, + { + "epoch": 5.594142259414226, + "grad_norm": 11.103867530822754, + "learning_rate": 9.987973998797402e-06, + "loss": 0.2872, + "step": 12033 + }, + { + "epoch": 5.594607159460716, + "grad_norm": 3.752974510192871, + "learning_rate": 9.9879729987973e-06, + "loss": 0.1411, + "step": 12034 + }, + { + "epoch": 5.595072059507206, + "grad_norm": 5.205402374267578, + "learning_rate": 9.987971998797201e-06, + "loss": 0.2837, + "step": 12035 + }, + { + "epoch": 5.595536959553696, + "grad_norm": 5.344376087188721, + "learning_rate": 9.9879709987971e-06, + "loss": 0.2579, + "step": 12036 + }, + { + "epoch": 5.596001859600186, + "grad_norm": 6.252267837524414, + "learning_rate": 9.987969998797001e-06, + "loss": 0.3755, + "step": 12037 + }, + { + "epoch": 5.596466759646676, + "grad_norm": 3.6271584033966064, + "learning_rate": 9.987968998796902e-06, + "loss": 0.0965, + "step": 12038 + }, + { + "epoch": 5.5969316596931655, + "grad_norm": 3.8377785682678223, + "learning_rate": 9.9879679987968e-06, + "loss": 0.1008, + "step": 12039 + }, + { + "epoch": 5.597396559739656, + "grad_norm": 5.498049259185791, + "learning_rate": 9.987966998796701e-06, + "loss": 0.2885, + "step": 12040 + }, + { + "epoch": 5.597861459786146, + "grad_norm": 9.598413467407227, + "learning_rate": 9.9879659987966e-06, + "loss": 0.2289, + "step": 12041 + }, + { + "epoch": 5.598326359832636, + "grad_norm": 4.1004486083984375, + "learning_rate": 9.987964998796501e-06, + "loss": 0.1754, + "step": 12042 + }, + { + "epoch": 5.598791259879126, + "grad_norm": 4.469529628753662, + "learning_rate": 9.9879639987964e-06, + "loss": 0.1831, + "step": 12043 + }, + { + "epoch": 5.599256159925616, + "grad_norm": 3.3027796745300293, + "learning_rate": 9.9879629987963e-06, + "loss": 0.091, + "step": 12044 + }, + { + "epoch": 5.599721059972106, + "grad_norm": 5.76466703414917, + "learning_rate": 9.987961998796201e-06, + "loss": 0.3577, + "step": 12045 + }, + { + "epoch": 5.600185960018596, + "grad_norm": 6.774229526519775, + "learning_rate": 9.9879609987961e-06, + "loss": 0.334, + "step": 12046 + }, + { + "epoch": 5.600650860065086, + "grad_norm": 9.472699165344238, + "learning_rate": 9.987959998796e-06, + "loss": 0.2805, + "step": 12047 + }, + { + "epoch": 5.6011157601115755, + "grad_norm": 7.66081428527832, + "learning_rate": 9.9879589987959e-06, + "loss": 0.2952, + "step": 12048 + }, + { + "epoch": 5.601580660158066, + "grad_norm": 6.901862144470215, + "learning_rate": 9.9879579987958e-06, + "loss": 0.2992, + "step": 12049 + }, + { + "epoch": 5.602045560204556, + "grad_norm": 6.091846466064453, + "learning_rate": 9.987956998795701e-06, + "loss": 0.2561, + "step": 12050 + }, + { + "epoch": 5.602510460251046, + "grad_norm": 5.207871437072754, + "learning_rate": 9.987955998795602e-06, + "loss": 0.1431, + "step": 12051 + }, + { + "epoch": 5.6029753602975365, + "grad_norm": 6.428591728210449, + "learning_rate": 9.9879549987955e-06, + "loss": 0.2245, + "step": 12052 + }, + { + "epoch": 5.603440260344026, + "grad_norm": 10.51169490814209, + "learning_rate": 9.987953998795401e-06, + "loss": 0.3345, + "step": 12053 + }, + { + "epoch": 5.603905160390516, + "grad_norm": 9.408761978149414, + "learning_rate": 9.9879529987953e-06, + "loss": 0.4702, + "step": 12054 + }, + { + "epoch": 5.604370060437006, + "grad_norm": 6.641358375549316, + "learning_rate": 9.987951998795201e-06, + "loss": 0.3478, + "step": 12055 + }, + { + "epoch": 5.604834960483496, + "grad_norm": 5.670695781707764, + "learning_rate": 9.9879509987951e-06, + "loss": 0.3014, + "step": 12056 + }, + { + "epoch": 5.605299860529986, + "grad_norm": 9.009382247924805, + "learning_rate": 9.987949998795e-06, + "loss": 0.6662, + "step": 12057 + }, + { + "epoch": 5.605764760576476, + "grad_norm": 6.360233783721924, + "learning_rate": 9.987948998794901e-06, + "loss": 0.3926, + "step": 12058 + }, + { + "epoch": 5.606229660622966, + "grad_norm": 6.552149772644043, + "learning_rate": 9.9879479987948e-06, + "loss": 0.2211, + "step": 12059 + }, + { + "epoch": 5.606694560669456, + "grad_norm": 8.844947814941406, + "learning_rate": 9.987946998794701e-06, + "loss": 0.1873, + "step": 12060 + }, + { + "epoch": 5.6071594607159465, + "grad_norm": 9.221221923828125, + "learning_rate": 9.9879459987946e-06, + "loss": 0.5239, + "step": 12061 + }, + { + "epoch": 5.607624360762436, + "grad_norm": 4.712094306945801, + "learning_rate": 9.9879449987945e-06, + "loss": 0.1853, + "step": 12062 + }, + { + "epoch": 5.608089260808926, + "grad_norm": 7.691887855529785, + "learning_rate": 9.987943998794401e-06, + "loss": 0.3615, + "step": 12063 + }, + { + "epoch": 5.608554160855416, + "grad_norm": 3.4483587741851807, + "learning_rate": 9.987942998794302e-06, + "loss": 0.1026, + "step": 12064 + }, + { + "epoch": 5.609019060901906, + "grad_norm": 8.358811378479004, + "learning_rate": 9.9879419987942e-06, + "loss": 0.4653, + "step": 12065 + }, + { + "epoch": 5.609483960948396, + "grad_norm": 6.21994161605835, + "learning_rate": 9.987940998794101e-06, + "loss": 0.2062, + "step": 12066 + }, + { + "epoch": 5.609948860994886, + "grad_norm": 3.9071388244628906, + "learning_rate": 9.987939998794e-06, + "loss": 0.1288, + "step": 12067 + }, + { + "epoch": 5.610413761041376, + "grad_norm": 5.966552734375, + "learning_rate": 9.9879389987939e-06, + "loss": 0.1863, + "step": 12068 + }, + { + "epoch": 5.610878661087866, + "grad_norm": 4.562688827514648, + "learning_rate": 9.9879379987938e-06, + "loss": 0.1558, + "step": 12069 + }, + { + "epoch": 5.6113435611343565, + "grad_norm": 4.746169090270996, + "learning_rate": 9.9879369987937e-06, + "loss": 0.2564, + "step": 12070 + }, + { + "epoch": 5.611808461180846, + "grad_norm": 5.645890712738037, + "learning_rate": 9.987935998793601e-06, + "loss": 0.1944, + "step": 12071 + }, + { + "epoch": 5.612273361227336, + "grad_norm": 7.209900379180908, + "learning_rate": 9.9879349987935e-06, + "loss": 0.3508, + "step": 12072 + }, + { + "epoch": 5.612738261273826, + "grad_norm": 7.092768669128418, + "learning_rate": 9.987933998793401e-06, + "loss": 0.4314, + "step": 12073 + }, + { + "epoch": 5.613203161320316, + "grad_norm": 5.911160945892334, + "learning_rate": 9.9879329987933e-06, + "loss": 0.4204, + "step": 12074 + }, + { + "epoch": 5.613668061366806, + "grad_norm": 5.233985900878906, + "learning_rate": 9.9879319987932e-06, + "loss": 0.2673, + "step": 12075 + }, + { + "epoch": 5.614132961413296, + "grad_norm": 4.448670864105225, + "learning_rate": 9.987930998793101e-06, + "loss": 0.2429, + "step": 12076 + }, + { + "epoch": 5.614597861459786, + "grad_norm": 6.653746604919434, + "learning_rate": 9.987929998793002e-06, + "loss": 0.4985, + "step": 12077 + }, + { + "epoch": 5.615062761506276, + "grad_norm": 6.003925800323486, + "learning_rate": 9.9879289987929e-06, + "loss": 0.4704, + "step": 12078 + }, + { + "epoch": 5.6155276615527665, + "grad_norm": 5.276743412017822, + "learning_rate": 9.9879279987928e-06, + "loss": 0.216, + "step": 12079 + }, + { + "epoch": 5.615992561599256, + "grad_norm": 5.48940372467041, + "learning_rate": 9.9879269987927e-06, + "loss": 0.3554, + "step": 12080 + }, + { + "epoch": 5.616457461645746, + "grad_norm": 6.488648414611816, + "learning_rate": 9.987925998792601e-06, + "loss": 0.356, + "step": 12081 + }, + { + "epoch": 5.616922361692236, + "grad_norm": 7.183485507965088, + "learning_rate": 9.9879249987925e-06, + "loss": 0.2824, + "step": 12082 + }, + { + "epoch": 5.617387261738726, + "grad_norm": 5.251055717468262, + "learning_rate": 9.9879239987924e-06, + "loss": 0.2299, + "step": 12083 + }, + { + "epoch": 5.617852161785216, + "grad_norm": 4.613152980804443, + "learning_rate": 9.987922998792301e-06, + "loss": 0.2354, + "step": 12084 + }, + { + "epoch": 5.618317061831706, + "grad_norm": 6.062521934509277, + "learning_rate": 9.9879219987922e-06, + "loss": 0.2868, + "step": 12085 + }, + { + "epoch": 5.618781961878196, + "grad_norm": 8.554082870483398, + "learning_rate": 9.987920998792101e-06, + "loss": 0.4052, + "step": 12086 + }, + { + "epoch": 5.619246861924686, + "grad_norm": 7.863063812255859, + "learning_rate": 9.987919998792e-06, + "loss": 0.26, + "step": 12087 + }, + { + "epoch": 5.6197117619711765, + "grad_norm": 6.7960286140441895, + "learning_rate": 9.9879189987919e-06, + "loss": 0.2655, + "step": 12088 + }, + { + "epoch": 5.620176662017666, + "grad_norm": 7.207371234893799, + "learning_rate": 9.987917998791801e-06, + "loss": 0.1254, + "step": 12089 + }, + { + "epoch": 5.620641562064156, + "grad_norm": 6.313011646270752, + "learning_rate": 9.9879169987917e-06, + "loss": 0.2839, + "step": 12090 + }, + { + "epoch": 5.621106462110646, + "grad_norm": 7.033502101898193, + "learning_rate": 9.987915998791601e-06, + "loss": 0.538, + "step": 12091 + }, + { + "epoch": 5.621571362157137, + "grad_norm": 6.70368766784668, + "learning_rate": 9.9879149987915e-06, + "loss": 0.2771, + "step": 12092 + }, + { + "epoch": 5.622036262203626, + "grad_norm": 6.65900182723999, + "learning_rate": 9.9879139987914e-06, + "loss": 0.1641, + "step": 12093 + }, + { + "epoch": 5.622501162250116, + "grad_norm": 4.379800319671631, + "learning_rate": 9.987912998791301e-06, + "loss": 0.1831, + "step": 12094 + }, + { + "epoch": 5.622966062296606, + "grad_norm": 5.674220561981201, + "learning_rate": 9.9879119987912e-06, + "loss": 0.1658, + "step": 12095 + }, + { + "epoch": 5.623430962343097, + "grad_norm": 2.3677468299865723, + "learning_rate": 9.9879109987911e-06, + "loss": 0.0444, + "step": 12096 + }, + { + "epoch": 5.6238958623895865, + "grad_norm": 3.4500176906585693, + "learning_rate": 9.987909998791001e-06, + "loss": 0.0645, + "step": 12097 + }, + { + "epoch": 5.624360762436076, + "grad_norm": 9.529106140136719, + "learning_rate": 9.9879089987909e-06, + "loss": 0.1332, + "step": 12098 + }, + { + "epoch": 5.624825662482566, + "grad_norm": 5.552257537841797, + "learning_rate": 9.987907998790801e-06, + "loss": 0.2943, + "step": 12099 + }, + { + "epoch": 5.625290562529056, + "grad_norm": 6.016002178192139, + "learning_rate": 9.9879069987907e-06, + "loss": 0.237, + "step": 12100 + }, + { + "epoch": 5.625755462575547, + "grad_norm": 5.0285325050354, + "learning_rate": 9.9879059987906e-06, + "loss": 0.1651, + "step": 12101 + }, + { + "epoch": 5.626220362622036, + "grad_norm": 5.009092330932617, + "learning_rate": 9.987904998790501e-06, + "loss": 0.2979, + "step": 12102 + }, + { + "epoch": 5.626685262668526, + "grad_norm": 9.453598976135254, + "learning_rate": 9.9879039987904e-06, + "loss": 0.5485, + "step": 12103 + }, + { + "epoch": 5.627150162715016, + "grad_norm": 5.54573392868042, + "learning_rate": 9.987902998790301e-06, + "loss": 0.0772, + "step": 12104 + }, + { + "epoch": 5.627615062761507, + "grad_norm": 4.191933631896973, + "learning_rate": 9.9879019987902e-06, + "loss": 0.1332, + "step": 12105 + }, + { + "epoch": 5.6280799628079965, + "grad_norm": 7.037622928619385, + "learning_rate": 9.9879009987901e-06, + "loss": 0.3116, + "step": 12106 + }, + { + "epoch": 5.628544862854486, + "grad_norm": 7.898891448974609, + "learning_rate": 9.987899998790001e-06, + "loss": 0.6076, + "step": 12107 + }, + { + "epoch": 5.629009762900976, + "grad_norm": 6.855409622192383, + "learning_rate": 9.987898998789902e-06, + "loss": 0.3477, + "step": 12108 + }, + { + "epoch": 5.629474662947466, + "grad_norm": 4.450735092163086, + "learning_rate": 9.987897998789801e-06, + "loss": 0.1709, + "step": 12109 + }, + { + "epoch": 5.629939562993957, + "grad_norm": 4.493161678314209, + "learning_rate": 9.987896998789702e-06, + "loss": 0.142, + "step": 12110 + }, + { + "epoch": 5.630404463040446, + "grad_norm": 7.731836318969727, + "learning_rate": 9.9878959987896e-06, + "loss": 0.3, + "step": 12111 + }, + { + "epoch": 5.630869363086936, + "grad_norm": 7.770173072814941, + "learning_rate": 9.9878949987895e-06, + "loss": 0.1826, + "step": 12112 + }, + { + "epoch": 5.631334263133426, + "grad_norm": 8.868731498718262, + "learning_rate": 9.9878939987894e-06, + "loss": 0.3356, + "step": 12113 + }, + { + "epoch": 5.631799163179917, + "grad_norm": 5.594869136810303, + "learning_rate": 9.9878929987893e-06, + "loss": 0.2704, + "step": 12114 + }, + { + "epoch": 5.6322640632264065, + "grad_norm": 6.441303730010986, + "learning_rate": 9.987891998789201e-06, + "loss": 0.2953, + "step": 12115 + }, + { + "epoch": 5.632728963272896, + "grad_norm": 4.902121543884277, + "learning_rate": 9.9878909987891e-06, + "loss": 0.1602, + "step": 12116 + }, + { + "epoch": 5.633193863319386, + "grad_norm": 4.733346939086914, + "learning_rate": 9.987889998789001e-06, + "loss": 0.1386, + "step": 12117 + }, + { + "epoch": 5.633658763365876, + "grad_norm": 5.142831802368164, + "learning_rate": 9.9878889987889e-06, + "loss": 0.346, + "step": 12118 + }, + { + "epoch": 5.634123663412367, + "grad_norm": 3.768841505050659, + "learning_rate": 9.9878879987888e-06, + "loss": 0.1013, + "step": 12119 + }, + { + "epoch": 5.634588563458856, + "grad_norm": 5.335822582244873, + "learning_rate": 9.987886998788701e-06, + "loss": 0.2893, + "step": 12120 + }, + { + "epoch": 5.635053463505346, + "grad_norm": 4.371828079223633, + "learning_rate": 9.987885998788602e-06, + "loss": 0.1274, + "step": 12121 + }, + { + "epoch": 5.635518363551836, + "grad_norm": 7.929278373718262, + "learning_rate": 9.987884998788501e-06, + "loss": 0.258, + "step": 12122 + }, + { + "epoch": 5.635983263598327, + "grad_norm": 7.34299373626709, + "learning_rate": 9.9878839987884e-06, + "loss": 0.4235, + "step": 12123 + }, + { + "epoch": 5.6364481636448165, + "grad_norm": 5.3779826164245605, + "learning_rate": 9.9878829987883e-06, + "loss": 0.2435, + "step": 12124 + }, + { + "epoch": 5.636913063691306, + "grad_norm": 6.52425479888916, + "learning_rate": 9.9878819987882e-06, + "loss": 0.1438, + "step": 12125 + }, + { + "epoch": 5.637377963737796, + "grad_norm": 6.547791957855225, + "learning_rate": 9.9878809987881e-06, + "loss": 0.2116, + "step": 12126 + }, + { + "epoch": 5.637842863784286, + "grad_norm": 4.581368446350098, + "learning_rate": 9.987879998788001e-06, + "loss": 0.1121, + "step": 12127 + }, + { + "epoch": 5.638307763830777, + "grad_norm": 5.284690856933594, + "learning_rate": 9.987878998787902e-06, + "loss": 0.3881, + "step": 12128 + }, + { + "epoch": 5.638772663877266, + "grad_norm": 4.926149368286133, + "learning_rate": 9.9878779987878e-06, + "loss": 0.18, + "step": 12129 + }, + { + "epoch": 5.639237563923756, + "grad_norm": 10.881048202514648, + "learning_rate": 9.987876998787701e-06, + "loss": 0.506, + "step": 12130 + }, + { + "epoch": 5.639702463970247, + "grad_norm": 4.3643927574157715, + "learning_rate": 9.9878759987876e-06, + "loss": 0.1964, + "step": 12131 + }, + { + "epoch": 5.640167364016737, + "grad_norm": 7.647458076477051, + "learning_rate": 9.9878749987875e-06, + "loss": 0.1027, + "step": 12132 + }, + { + "epoch": 5.6406322640632265, + "grad_norm": 3.878375291824341, + "learning_rate": 9.987873998787401e-06, + "loss": 0.0952, + "step": 12133 + }, + { + "epoch": 5.641097164109716, + "grad_norm": 6.7078962326049805, + "learning_rate": 9.9878729987873e-06, + "loss": 0.4115, + "step": 12134 + }, + { + "epoch": 5.641562064156206, + "grad_norm": 7.175532817840576, + "learning_rate": 9.987871998787201e-06, + "loss": 0.3327, + "step": 12135 + }, + { + "epoch": 5.642026964202697, + "grad_norm": 5.468133449554443, + "learning_rate": 9.9878709987871e-06, + "loss": 0.1562, + "step": 12136 + }, + { + "epoch": 5.642491864249187, + "grad_norm": 4.766102313995361, + "learning_rate": 9.987869998787e-06, + "loss": 0.2405, + "step": 12137 + }, + { + "epoch": 5.642956764295676, + "grad_norm": 3.825403928756714, + "learning_rate": 9.9878689987869e-06, + "loss": 0.1317, + "step": 12138 + }, + { + "epoch": 5.643421664342166, + "grad_norm": 6.9739789962768555, + "learning_rate": 9.9878679987868e-06, + "loss": 0.1552, + "step": 12139 + }, + { + "epoch": 5.643886564388657, + "grad_norm": 4.932211399078369, + "learning_rate": 9.987866998786701e-06, + "loss": 0.1538, + "step": 12140 + }, + { + "epoch": 5.644351464435147, + "grad_norm": 8.369454383850098, + "learning_rate": 9.987865998786602e-06, + "loss": 0.4616, + "step": 12141 + }, + { + "epoch": 5.6448163644816365, + "grad_norm": 5.163234233856201, + "learning_rate": 9.9878649987865e-06, + "loss": 0.1855, + "step": 12142 + }, + { + "epoch": 5.645281264528126, + "grad_norm": 5.256067752838135, + "learning_rate": 9.987863998786401e-06, + "loss": 0.2061, + "step": 12143 + }, + { + "epoch": 5.645746164574616, + "grad_norm": 6.421163558959961, + "learning_rate": 9.9878629987863e-06, + "loss": 0.3733, + "step": 12144 + }, + { + "epoch": 5.646211064621107, + "grad_norm": 5.472813606262207, + "learning_rate": 9.987861998786201e-06, + "loss": 0.2738, + "step": 12145 + }, + { + "epoch": 5.646675964667597, + "grad_norm": 5.610746383666992, + "learning_rate": 9.987860998786102e-06, + "loss": 0.2767, + "step": 12146 + }, + { + "epoch": 5.647140864714086, + "grad_norm": 12.513467788696289, + "learning_rate": 9.987859998786e-06, + "loss": 0.4063, + "step": 12147 + }, + { + "epoch": 5.647605764760576, + "grad_norm": 6.167660713195801, + "learning_rate": 9.987858998785901e-06, + "loss": 0.3228, + "step": 12148 + }, + { + "epoch": 5.648070664807067, + "grad_norm": 4.360177993774414, + "learning_rate": 9.9878579987858e-06, + "loss": 0.2093, + "step": 12149 + }, + { + "epoch": 5.648535564853557, + "grad_norm": 4.024086952209473, + "learning_rate": 9.9878569987857e-06, + "loss": 0.1438, + "step": 12150 + }, + { + "epoch": 5.6490004649000465, + "grad_norm": 6.762012958526611, + "learning_rate": 9.9878559987856e-06, + "loss": 0.2632, + "step": 12151 + }, + { + "epoch": 5.649465364946536, + "grad_norm": 5.468650817871094, + "learning_rate": 9.9878549987855e-06, + "loss": 0.1991, + "step": 12152 + }, + { + "epoch": 5.649930264993026, + "grad_norm": 4.144965171813965, + "learning_rate": 9.987853998785401e-06, + "loss": 0.2107, + "step": 12153 + }, + { + "epoch": 5.650395165039517, + "grad_norm": 2.2107906341552734, + "learning_rate": 9.987852998785302e-06, + "loss": 0.0396, + "step": 12154 + }, + { + "epoch": 5.650860065086007, + "grad_norm": 5.084787845611572, + "learning_rate": 9.9878519987852e-06, + "loss": 0.2376, + "step": 12155 + }, + { + "epoch": 5.651324965132496, + "grad_norm": 6.855254650115967, + "learning_rate": 9.9878509987851e-06, + "loss": 0.3615, + "step": 12156 + }, + { + "epoch": 5.651789865178986, + "grad_norm": 5.206773281097412, + "learning_rate": 9.987849998785e-06, + "loss": 0.2092, + "step": 12157 + }, + { + "epoch": 5.652254765225477, + "grad_norm": 4.500396728515625, + "learning_rate": 9.987848998784901e-06, + "loss": 0.1452, + "step": 12158 + }, + { + "epoch": 5.652719665271967, + "grad_norm": 5.23947811126709, + "learning_rate": 9.987847998784802e-06, + "loss": 0.2666, + "step": 12159 + }, + { + "epoch": 5.6531845653184565, + "grad_norm": 4.468201160430908, + "learning_rate": 9.9878469987847e-06, + "loss": 0.2346, + "step": 12160 + }, + { + "epoch": 5.653649465364946, + "grad_norm": 7.320878982543945, + "learning_rate": 9.987845998784601e-06, + "loss": 0.2564, + "step": 12161 + }, + { + "epoch": 5.654114365411436, + "grad_norm": 6.066131591796875, + "learning_rate": 9.9878449987845e-06, + "loss": 0.2478, + "step": 12162 + }, + { + "epoch": 5.654579265457927, + "grad_norm": 3.665699005126953, + "learning_rate": 9.9878439987844e-06, + "loss": 0.1668, + "step": 12163 + }, + { + "epoch": 5.655044165504417, + "grad_norm": 7.104134559631348, + "learning_rate": 9.9878429987843e-06, + "loss": 0.4407, + "step": 12164 + }, + { + "epoch": 5.655509065550906, + "grad_norm": 6.577054500579834, + "learning_rate": 9.9878419987842e-06, + "loss": 0.2451, + "step": 12165 + }, + { + "epoch": 5.655973965597396, + "grad_norm": 3.6896729469299316, + "learning_rate": 9.987840998784101e-06, + "loss": 0.0625, + "step": 12166 + }, + { + "epoch": 5.656438865643887, + "grad_norm": 3.8052618503570557, + "learning_rate": 9.987839998784e-06, + "loss": 0.0996, + "step": 12167 + }, + { + "epoch": 5.656903765690377, + "grad_norm": 3.5800459384918213, + "learning_rate": 9.9878389987839e-06, + "loss": 0.1213, + "step": 12168 + }, + { + "epoch": 5.6573686657368665, + "grad_norm": 6.950285911560059, + "learning_rate": 9.9878379987838e-06, + "loss": 0.3336, + "step": 12169 + }, + { + "epoch": 5.657833565783356, + "grad_norm": 6.409906387329102, + "learning_rate": 9.9878369987837e-06, + "loss": 0.1362, + "step": 12170 + }, + { + "epoch": 5.658298465829847, + "grad_norm": 4.061538219451904, + "learning_rate": 9.987835998783601e-06, + "loss": 0.1038, + "step": 12171 + }, + { + "epoch": 5.658763365876337, + "grad_norm": 4.729556083679199, + "learning_rate": 9.987834998783502e-06, + "loss": 0.1848, + "step": 12172 + }, + { + "epoch": 5.659228265922827, + "grad_norm": 4.854077339172363, + "learning_rate": 9.9878339987834e-06, + "loss": 0.1519, + "step": 12173 + }, + { + "epoch": 5.659693165969316, + "grad_norm": 8.809391021728516, + "learning_rate": 9.987832998783301e-06, + "loss": 0.1409, + "step": 12174 + }, + { + "epoch": 5.660158066015807, + "grad_norm": 5.441283226013184, + "learning_rate": 9.9878319987832e-06, + "loss": 0.2094, + "step": 12175 + }, + { + "epoch": 5.660622966062297, + "grad_norm": 7.386356830596924, + "learning_rate": 9.987830998783101e-06, + "loss": 0.1779, + "step": 12176 + }, + { + "epoch": 5.661087866108787, + "grad_norm": 5.34227991104126, + "learning_rate": 9.987829998783002e-06, + "loss": 0.1591, + "step": 12177 + }, + { + "epoch": 5.6615527661552765, + "grad_norm": 5.566285610198975, + "learning_rate": 9.9878289987829e-06, + "loss": 0.2558, + "step": 12178 + }, + { + "epoch": 5.662017666201766, + "grad_norm": 9.179396629333496, + "learning_rate": 9.987827998782801e-06, + "loss": 0.1607, + "step": 12179 + }, + { + "epoch": 5.662482566248257, + "grad_norm": 7.373863697052002, + "learning_rate": 9.9878269987827e-06, + "loss": 0.2435, + "step": 12180 + }, + { + "epoch": 5.662947466294747, + "grad_norm": 5.15784215927124, + "learning_rate": 9.9878259987826e-06, + "loss": 0.2897, + "step": 12181 + }, + { + "epoch": 5.663412366341237, + "grad_norm": 4.810065746307373, + "learning_rate": 9.9878249987825e-06, + "loss": 0.1567, + "step": 12182 + }, + { + "epoch": 5.663877266387726, + "grad_norm": 8.360816955566406, + "learning_rate": 9.9878239987824e-06, + "loss": 0.3623, + "step": 12183 + }, + { + "epoch": 5.664342166434217, + "grad_norm": 7.9466753005981445, + "learning_rate": 9.987822998782301e-06, + "loss": 0.3348, + "step": 12184 + }, + { + "epoch": 5.664807066480707, + "grad_norm": 4.351609230041504, + "learning_rate": 9.987821998782202e-06, + "loss": 0.1714, + "step": 12185 + }, + { + "epoch": 5.665271966527197, + "grad_norm": 7.0288166999816895, + "learning_rate": 9.9878209987821e-06, + "loss": 0.1936, + "step": 12186 + }, + { + "epoch": 5.6657368665736865, + "grad_norm": 4.74945068359375, + "learning_rate": 9.987819998782e-06, + "loss": 0.1828, + "step": 12187 + }, + { + "epoch": 5.666201766620176, + "grad_norm": 4.641119003295898, + "learning_rate": 9.9878189987819e-06, + "loss": 0.1973, + "step": 12188 + }, + { + "epoch": 5.666666666666667, + "grad_norm": 4.041248321533203, + "learning_rate": 9.987817998781801e-06, + "loss": 0.1711, + "step": 12189 + }, + { + "epoch": 5.667131566713157, + "grad_norm": 5.0018630027771, + "learning_rate": 9.987816998781702e-06, + "loss": 0.1774, + "step": 12190 + }, + { + "epoch": 5.667596466759647, + "grad_norm": 9.788407325744629, + "learning_rate": 9.9878159987816e-06, + "loss": 0.0632, + "step": 12191 + }, + { + "epoch": 5.668061366806136, + "grad_norm": 13.650562286376953, + "learning_rate": 9.987814998781501e-06, + "loss": 0.3518, + "step": 12192 + }, + { + "epoch": 5.668526266852627, + "grad_norm": 6.547540187835693, + "learning_rate": 9.9878139987814e-06, + "loss": 0.3901, + "step": 12193 + }, + { + "epoch": 5.668991166899117, + "grad_norm": 6.651985168457031, + "learning_rate": 9.987812998781301e-06, + "loss": 0.439, + "step": 12194 + }, + { + "epoch": 5.669456066945607, + "grad_norm": 7.829564094543457, + "learning_rate": 9.9878119987812e-06, + "loss": 0.3543, + "step": 12195 + }, + { + "epoch": 5.6699209669920965, + "grad_norm": 4.992544174194336, + "learning_rate": 9.9878109987811e-06, + "loss": 0.166, + "step": 12196 + }, + { + "epoch": 5.670385867038586, + "grad_norm": 4.9284281730651855, + "learning_rate": 9.987809998781001e-06, + "loss": 0.208, + "step": 12197 + }, + { + "epoch": 5.670850767085077, + "grad_norm": 4.2297043800354, + "learning_rate": 9.9878089987809e-06, + "loss": 0.1389, + "step": 12198 + }, + { + "epoch": 5.671315667131567, + "grad_norm": 4.402933597564697, + "learning_rate": 9.9878079987808e-06, + "loss": 0.1798, + "step": 12199 + }, + { + "epoch": 5.671780567178057, + "grad_norm": 3.298433542251587, + "learning_rate": 9.9878069987807e-06, + "loss": 0.14, + "step": 12200 + }, + { + "epoch": 5.672245467224546, + "grad_norm": 4.409950256347656, + "learning_rate": 9.9878059987806e-06, + "loss": 0.168, + "step": 12201 + }, + { + "epoch": 5.672710367271037, + "grad_norm": 7.642734050750732, + "learning_rate": 9.987804998780501e-06, + "loss": 0.7267, + "step": 12202 + }, + { + "epoch": 5.673175267317527, + "grad_norm": 4.240837097167969, + "learning_rate": 9.987803998780402e-06, + "loss": 0.1302, + "step": 12203 + }, + { + "epoch": 5.673640167364017, + "grad_norm": 6.3225016593933105, + "learning_rate": 9.9878029987803e-06, + "loss": 0.088, + "step": 12204 + }, + { + "epoch": 5.6741050674105065, + "grad_norm": 8.40183162689209, + "learning_rate": 9.987801998780201e-06, + "loss": 0.3154, + "step": 12205 + }, + { + "epoch": 5.674569967456996, + "grad_norm": 4.348914623260498, + "learning_rate": 9.9878009987801e-06, + "loss": 0.1851, + "step": 12206 + }, + { + "epoch": 5.675034867503487, + "grad_norm": 6.199214935302734, + "learning_rate": 9.987799998780001e-06, + "loss": 0.2674, + "step": 12207 + }, + { + "epoch": 5.675499767549977, + "grad_norm": 3.853344440460205, + "learning_rate": 9.9877989987799e-06, + "loss": 0.1928, + "step": 12208 + }, + { + "epoch": 5.675964667596467, + "grad_norm": 4.8550872802734375, + "learning_rate": 9.9877979987798e-06, + "loss": 0.1543, + "step": 12209 + }, + { + "epoch": 5.676429567642957, + "grad_norm": 3.9370415210723877, + "learning_rate": 9.987796998779701e-06, + "loss": 0.1639, + "step": 12210 + }, + { + "epoch": 5.676894467689447, + "grad_norm": 4.215131759643555, + "learning_rate": 9.9877959987796e-06, + "loss": 0.1244, + "step": 12211 + }, + { + "epoch": 5.677359367735937, + "grad_norm": 5.81602144241333, + "learning_rate": 9.987794998779501e-06, + "loss": 0.2984, + "step": 12212 + }, + { + "epoch": 5.677824267782427, + "grad_norm": 3.541670799255371, + "learning_rate": 9.9877939987794e-06, + "loss": 0.0745, + "step": 12213 + }, + { + "epoch": 5.6782891678289165, + "grad_norm": 3.6452224254608154, + "learning_rate": 9.9877929987793e-06, + "loss": 0.123, + "step": 12214 + }, + { + "epoch": 5.678754067875407, + "grad_norm": 8.43023681640625, + "learning_rate": 9.987791998779201e-06, + "loss": 0.263, + "step": 12215 + }, + { + "epoch": 5.679218967921897, + "grad_norm": 4.295901298522949, + "learning_rate": 9.987790998779102e-06, + "loss": 0.1772, + "step": 12216 + }, + { + "epoch": 5.679683867968387, + "grad_norm": 6.585404396057129, + "learning_rate": 9.987789998779e-06, + "loss": 0.3675, + "step": 12217 + }, + { + "epoch": 5.680148768014877, + "grad_norm": 7.553507328033447, + "learning_rate": 9.987788998778901e-06, + "loss": 0.2595, + "step": 12218 + }, + { + "epoch": 5.680613668061367, + "grad_norm": 7.375125885009766, + "learning_rate": 9.9877879987788e-06, + "loss": 0.4939, + "step": 12219 + }, + { + "epoch": 5.681078568107857, + "grad_norm": 3.0202372074127197, + "learning_rate": 9.9877869987787e-06, + "loss": 0.0465, + "step": 12220 + }, + { + "epoch": 5.681543468154347, + "grad_norm": 5.113184928894043, + "learning_rate": 9.9877859987786e-06, + "loss": 0.2114, + "step": 12221 + }, + { + "epoch": 5.682008368200837, + "grad_norm": 4.239744186401367, + "learning_rate": 9.9877849987785e-06, + "loss": 0.1683, + "step": 12222 + }, + { + "epoch": 5.6824732682473265, + "grad_norm": 8.872769355773926, + "learning_rate": 9.987783998778401e-06, + "loss": 0.399, + "step": 12223 + }, + { + "epoch": 5.682938168293817, + "grad_norm": 5.305861473083496, + "learning_rate": 9.9877829987783e-06, + "loss": 0.3383, + "step": 12224 + }, + { + "epoch": 5.683403068340307, + "grad_norm": 5.2532501220703125, + "learning_rate": 9.987781998778201e-06, + "loss": 0.2892, + "step": 12225 + }, + { + "epoch": 5.683867968386797, + "grad_norm": 5.020745277404785, + "learning_rate": 9.9877809987781e-06, + "loss": 0.1442, + "step": 12226 + }, + { + "epoch": 5.684332868433287, + "grad_norm": 3.258969783782959, + "learning_rate": 9.987779998778e-06, + "loss": 0.1303, + "step": 12227 + }, + { + "epoch": 5.684797768479777, + "grad_norm": 7.004039764404297, + "learning_rate": 9.987778998777901e-06, + "loss": 0.223, + "step": 12228 + }, + { + "epoch": 5.685262668526267, + "grad_norm": 5.015078544616699, + "learning_rate": 9.987777998777802e-06, + "loss": 0.147, + "step": 12229 + }, + { + "epoch": 5.685727568572757, + "grad_norm": 4.721847057342529, + "learning_rate": 9.987776998777701e-06, + "loss": 0.2118, + "step": 12230 + }, + { + "epoch": 5.686192468619247, + "grad_norm": 8.045034408569336, + "learning_rate": 9.9877759987776e-06, + "loss": 0.2892, + "step": 12231 + }, + { + "epoch": 5.6866573686657365, + "grad_norm": 3.610717535018921, + "learning_rate": 9.9877749987775e-06, + "loss": 0.103, + "step": 12232 + }, + { + "epoch": 5.687122268712227, + "grad_norm": 9.69802188873291, + "learning_rate": 9.9877739987774e-06, + "loss": 0.3539, + "step": 12233 + }, + { + "epoch": 5.687587168758717, + "grad_norm": 4.52254581451416, + "learning_rate": 9.9877729987773e-06, + "loss": 0.1562, + "step": 12234 + }, + { + "epoch": 5.688052068805207, + "grad_norm": 9.672887802124023, + "learning_rate": 9.9877719987772e-06, + "loss": 0.2778, + "step": 12235 + }, + { + "epoch": 5.6885169688516966, + "grad_norm": 3.77785062789917, + "learning_rate": 9.987770998777101e-06, + "loss": 0.1067, + "step": 12236 + }, + { + "epoch": 5.688981868898187, + "grad_norm": 5.987009048461914, + "learning_rate": 9.987769998777e-06, + "loss": 0.1461, + "step": 12237 + }, + { + "epoch": 5.689446768944677, + "grad_norm": 8.667975425720215, + "learning_rate": 9.987768998776901e-06, + "loss": 0.4844, + "step": 12238 + }, + { + "epoch": 5.689911668991167, + "grad_norm": 4.318647384643555, + "learning_rate": 9.9877679987768e-06, + "loss": 0.1968, + "step": 12239 + }, + { + "epoch": 5.690376569037657, + "grad_norm": 5.238731384277344, + "learning_rate": 9.9877669987767e-06, + "loss": 0.2135, + "step": 12240 + }, + { + "epoch": 5.6908414690841465, + "grad_norm": 6.705209732055664, + "learning_rate": 9.987765998776601e-06, + "loss": 0.2776, + "step": 12241 + }, + { + "epoch": 5.691306369130637, + "grad_norm": 5.220598220825195, + "learning_rate": 9.9877649987765e-06, + "loss": 0.2173, + "step": 12242 + }, + { + "epoch": 5.691771269177127, + "grad_norm": 6.753205299377441, + "learning_rate": 9.987763998776401e-06, + "loss": 0.3659, + "step": 12243 + }, + { + "epoch": 5.692236169223617, + "grad_norm": 5.8297247886657715, + "learning_rate": 9.9877629987763e-06, + "loss": 0.0649, + "step": 12244 + }, + { + "epoch": 5.6927010692701066, + "grad_norm": 5.836606025695801, + "learning_rate": 9.9877619987762e-06, + "loss": 0.1572, + "step": 12245 + }, + { + "epoch": 5.693165969316597, + "grad_norm": 3.919064998626709, + "learning_rate": 9.987760998776101e-06, + "loss": 0.0799, + "step": 12246 + }, + { + "epoch": 5.693630869363087, + "grad_norm": 4.535991668701172, + "learning_rate": 9.987759998776e-06, + "loss": 0.1885, + "step": 12247 + }, + { + "epoch": 5.694095769409577, + "grad_norm": 3.626476526260376, + "learning_rate": 9.987758998775901e-06, + "loss": 0.0937, + "step": 12248 + }, + { + "epoch": 5.694560669456067, + "grad_norm": 6.428524017333984, + "learning_rate": 9.987757998775802e-06, + "loss": 0.247, + "step": 12249 + }, + { + "epoch": 5.695025569502557, + "grad_norm": 5.019030570983887, + "learning_rate": 9.9877569987757e-06, + "loss": 0.1411, + "step": 12250 + }, + { + "epoch": 5.695490469549047, + "grad_norm": 10.022470474243164, + "learning_rate": 9.987755998775601e-06, + "loss": 0.271, + "step": 12251 + }, + { + "epoch": 5.695955369595537, + "grad_norm": 4.873880863189697, + "learning_rate": 9.9877549987755e-06, + "loss": 0.2716, + "step": 12252 + }, + { + "epoch": 5.696420269642027, + "grad_norm": 3.68658709526062, + "learning_rate": 9.9877539987754e-06, + "loss": 0.0867, + "step": 12253 + }, + { + "epoch": 5.696885169688517, + "grad_norm": 6.226521015167236, + "learning_rate": 9.987752998775301e-06, + "loss": 0.2598, + "step": 12254 + }, + { + "epoch": 5.697350069735007, + "grad_norm": 4.340602397918701, + "learning_rate": 9.9877519987752e-06, + "loss": 0.1623, + "step": 12255 + }, + { + "epoch": 5.697814969781497, + "grad_norm": 6.126380443572998, + "learning_rate": 9.987750998775101e-06, + "loss": 0.3725, + "step": 12256 + }, + { + "epoch": 5.698279869827987, + "grad_norm": 4.85648250579834, + "learning_rate": 9.987749998775e-06, + "loss": 0.1878, + "step": 12257 + }, + { + "epoch": 5.698744769874477, + "grad_norm": 7.415361404418945, + "learning_rate": 9.9877489987749e-06, + "loss": 0.2426, + "step": 12258 + }, + { + "epoch": 5.699209669920967, + "grad_norm": 4.332111358642578, + "learning_rate": 9.987747998774801e-06, + "loss": 0.1994, + "step": 12259 + }, + { + "epoch": 5.699674569967457, + "grad_norm": 5.079491138458252, + "learning_rate": 9.9877469987747e-06, + "loss": 0.192, + "step": 12260 + }, + { + "epoch": 5.700139470013947, + "grad_norm": 4.465485572814941, + "learning_rate": 9.987745998774601e-06, + "loss": 0.1507, + "step": 12261 + }, + { + "epoch": 5.700604370060437, + "grad_norm": 5.735915184020996, + "learning_rate": 9.987744998774502e-06, + "loss": 0.329, + "step": 12262 + }, + { + "epoch": 5.701069270106927, + "grad_norm": 5.690445899963379, + "learning_rate": 9.9877439987744e-06, + "loss": 0.1907, + "step": 12263 + }, + { + "epoch": 5.701534170153417, + "grad_norm": 4.357534885406494, + "learning_rate": 9.9877429987743e-06, + "loss": 0.1896, + "step": 12264 + }, + { + "epoch": 5.701999070199907, + "grad_norm": 6.174241065979004, + "learning_rate": 9.9877419987742e-06, + "loss": 0.3723, + "step": 12265 + }, + { + "epoch": 5.702463970246397, + "grad_norm": 5.4910502433776855, + "learning_rate": 9.9877409987741e-06, + "loss": 0.3044, + "step": 12266 + }, + { + "epoch": 5.702928870292887, + "grad_norm": 6.820669651031494, + "learning_rate": 9.987739998774001e-06, + "loss": 0.2521, + "step": 12267 + }, + { + "epoch": 5.703393770339377, + "grad_norm": 5.16483211517334, + "learning_rate": 9.9877389987739e-06, + "loss": 0.2476, + "step": 12268 + }, + { + "epoch": 5.703858670385867, + "grad_norm": 5.981675624847412, + "learning_rate": 9.987737998773801e-06, + "loss": 0.0763, + "step": 12269 + }, + { + "epoch": 5.704323570432357, + "grad_norm": 3.861016273498535, + "learning_rate": 9.9877369987737e-06, + "loss": 0.1376, + "step": 12270 + }, + { + "epoch": 5.704788470478847, + "grad_norm": 4.683557033538818, + "learning_rate": 9.9877359987736e-06, + "loss": 0.2843, + "step": 12271 + }, + { + "epoch": 5.705253370525337, + "grad_norm": 8.07517147064209, + "learning_rate": 9.987734998773501e-06, + "loss": 0.3657, + "step": 12272 + }, + { + "epoch": 5.705718270571827, + "grad_norm": 6.103357315063477, + "learning_rate": 9.987733998773402e-06, + "loss": 0.2751, + "step": 12273 + }, + { + "epoch": 5.706183170618317, + "grad_norm": 3.5387303829193115, + "learning_rate": 9.987732998773301e-06, + "loss": 0.0669, + "step": 12274 + }, + { + "epoch": 5.706648070664807, + "grad_norm": 8.836590766906738, + "learning_rate": 9.9877319987732e-06, + "loss": 0.1636, + "step": 12275 + }, + { + "epoch": 5.707112970711297, + "grad_norm": 5.722933292388916, + "learning_rate": 9.9877309987731e-06, + "loss": 0.1256, + "step": 12276 + }, + { + "epoch": 5.707577870757787, + "grad_norm": 6.602475643157959, + "learning_rate": 9.987729998773e-06, + "loss": 0.2354, + "step": 12277 + }, + { + "epoch": 5.708042770804277, + "grad_norm": 3.704749584197998, + "learning_rate": 9.9877289987729e-06, + "loss": 0.1207, + "step": 12278 + }, + { + "epoch": 5.708507670850767, + "grad_norm": 4.861956596374512, + "learning_rate": 9.987727998772801e-06, + "loss": 0.1796, + "step": 12279 + }, + { + "epoch": 5.708972570897257, + "grad_norm": 4.697651386260986, + "learning_rate": 9.987726998772702e-06, + "loss": 0.1121, + "step": 12280 + }, + { + "epoch": 5.709437470943747, + "grad_norm": 10.117223739624023, + "learning_rate": 9.9877259987726e-06, + "loss": 0.4975, + "step": 12281 + }, + { + "epoch": 5.709902370990237, + "grad_norm": 6.532720565795898, + "learning_rate": 9.987724998772501e-06, + "loss": 0.342, + "step": 12282 + }, + { + "epoch": 5.710367271036727, + "grad_norm": 8.550371170043945, + "learning_rate": 9.9877239987724e-06, + "loss": 0.2761, + "step": 12283 + }, + { + "epoch": 5.710832171083217, + "grad_norm": 4.225372791290283, + "learning_rate": 9.9877229987723e-06, + "loss": 0.098, + "step": 12284 + }, + { + "epoch": 5.711297071129707, + "grad_norm": 6.26847505569458, + "learning_rate": 9.987721998772201e-06, + "loss": 0.2038, + "step": 12285 + }, + { + "epoch": 5.711761971176197, + "grad_norm": 8.373675346374512, + "learning_rate": 9.9877209987721e-06, + "loss": 0.4453, + "step": 12286 + }, + { + "epoch": 5.712226871222687, + "grad_norm": 3.5122499465942383, + "learning_rate": 9.987719998772001e-06, + "loss": 0.1127, + "step": 12287 + }, + { + "epoch": 5.712691771269177, + "grad_norm": 4.890333652496338, + "learning_rate": 9.9877189987719e-06, + "loss": 0.1342, + "step": 12288 + }, + { + "epoch": 5.713156671315668, + "grad_norm": 4.182945728302002, + "learning_rate": 9.9877179987718e-06, + "loss": 0.0864, + "step": 12289 + }, + { + "epoch": 5.713621571362157, + "grad_norm": 7.677793979644775, + "learning_rate": 9.9877169987717e-06, + "loss": 0.4593, + "step": 12290 + }, + { + "epoch": 5.714086471408647, + "grad_norm": 5.73891019821167, + "learning_rate": 9.9877159987716e-06, + "loss": 0.2248, + "step": 12291 + }, + { + "epoch": 5.714551371455137, + "grad_norm": 5.947330951690674, + "learning_rate": 9.987714998771501e-06, + "loss": 0.238, + "step": 12292 + }, + { + "epoch": 5.715016271501627, + "grad_norm": 4.50628662109375, + "learning_rate": 9.987713998771402e-06, + "loss": 0.1975, + "step": 12293 + }, + { + "epoch": 5.7154811715481175, + "grad_norm": 6.686439037322998, + "learning_rate": 9.9877129987713e-06, + "loss": 0.34, + "step": 12294 + }, + { + "epoch": 5.715946071594607, + "grad_norm": 4.653482437133789, + "learning_rate": 9.987711998771201e-06, + "loss": 0.2072, + "step": 12295 + }, + { + "epoch": 5.716410971641097, + "grad_norm": 6.063654899597168, + "learning_rate": 9.9877109987711e-06, + "loss": 0.3679, + "step": 12296 + }, + { + "epoch": 5.716875871687587, + "grad_norm": 4.910088062286377, + "learning_rate": 9.987709998771001e-06, + "loss": 0.1415, + "step": 12297 + }, + { + "epoch": 5.717340771734078, + "grad_norm": 5.989445209503174, + "learning_rate": 9.987708998770902e-06, + "loss": 0.3332, + "step": 12298 + }, + { + "epoch": 5.717805671780567, + "grad_norm": 6.476961612701416, + "learning_rate": 9.9877079987708e-06, + "loss": 0.3517, + "step": 12299 + }, + { + "epoch": 5.718270571827057, + "grad_norm": 4.397124290466309, + "learning_rate": 9.987706998770701e-06, + "loss": 0.108, + "step": 12300 + }, + { + "epoch": 5.718735471873547, + "grad_norm": 4.821004867553711, + "learning_rate": 9.9877059987706e-06, + "loss": 0.1181, + "step": 12301 + }, + { + "epoch": 5.719200371920037, + "grad_norm": 5.703532695770264, + "learning_rate": 9.9877049987705e-06, + "loss": 0.265, + "step": 12302 + }, + { + "epoch": 5.7196652719665275, + "grad_norm": 6.2228007316589355, + "learning_rate": 9.9877039987704e-06, + "loss": 0.1569, + "step": 12303 + }, + { + "epoch": 5.720130172013017, + "grad_norm": 7.355706214904785, + "learning_rate": 9.9877029987703e-06, + "loss": 0.1791, + "step": 12304 + }, + { + "epoch": 5.720595072059507, + "grad_norm": 4.688656806945801, + "learning_rate": 9.987701998770201e-06, + "loss": 0.1271, + "step": 12305 + }, + { + "epoch": 5.721059972105997, + "grad_norm": 3.6016035079956055, + "learning_rate": 9.987700998770102e-06, + "loss": 0.0992, + "step": 12306 + }, + { + "epoch": 5.721524872152488, + "grad_norm": 9.021588325500488, + "learning_rate": 9.98769999877e-06, + "loss": 0.1572, + "step": 12307 + }, + { + "epoch": 5.721989772198977, + "grad_norm": 6.627445220947266, + "learning_rate": 9.9876989987699e-06, + "loss": 0.2005, + "step": 12308 + }, + { + "epoch": 5.722454672245467, + "grad_norm": 8.461775779724121, + "learning_rate": 9.9876979987698e-06, + "loss": 0.2585, + "step": 12309 + }, + { + "epoch": 5.722919572291957, + "grad_norm": 3.856313705444336, + "learning_rate": 9.987696998769701e-06, + "loss": 0.0856, + "step": 12310 + }, + { + "epoch": 5.723384472338447, + "grad_norm": 5.99852991104126, + "learning_rate": 9.987695998769602e-06, + "loss": 0.4572, + "step": 12311 + }, + { + "epoch": 5.7238493723849375, + "grad_norm": 6.165585041046143, + "learning_rate": 9.9876949987695e-06, + "loss": 0.2354, + "step": 12312 + }, + { + "epoch": 5.724314272431427, + "grad_norm": 5.301091194152832, + "learning_rate": 9.987693998769401e-06, + "loss": 0.1347, + "step": 12313 + }, + { + "epoch": 5.724779172477917, + "grad_norm": 4.260205268859863, + "learning_rate": 9.9876929987693e-06, + "loss": 0.1289, + "step": 12314 + }, + { + "epoch": 5.725244072524407, + "grad_norm": 5.940729141235352, + "learning_rate": 9.987691998769201e-06, + "loss": 0.2513, + "step": 12315 + }, + { + "epoch": 5.725708972570898, + "grad_norm": 5.960992813110352, + "learning_rate": 9.9876909987691e-06, + "loss": 0.1761, + "step": 12316 + }, + { + "epoch": 5.726173872617387, + "grad_norm": 7.186306476593018, + "learning_rate": 9.987689998769e-06, + "loss": 0.3497, + "step": 12317 + }, + { + "epoch": 5.726638772663877, + "grad_norm": 7.063317775726318, + "learning_rate": 9.987688998768901e-06, + "loss": 0.3009, + "step": 12318 + }, + { + "epoch": 5.727103672710367, + "grad_norm": 4.123445510864258, + "learning_rate": 9.9876879987688e-06, + "loss": 0.1723, + "step": 12319 + }, + { + "epoch": 5.727568572756857, + "grad_norm": 4.893242359161377, + "learning_rate": 9.9876869987687e-06, + "loss": 0.1755, + "step": 12320 + }, + { + "epoch": 5.7280334728033475, + "grad_norm": 4.390624046325684, + "learning_rate": 9.9876859987686e-06, + "loss": 0.1558, + "step": 12321 + }, + { + "epoch": 5.728498372849837, + "grad_norm": 8.018492698669434, + "learning_rate": 9.9876849987685e-06, + "loss": 0.4473, + "step": 12322 + }, + { + "epoch": 5.728963272896327, + "grad_norm": 4.644896984100342, + "learning_rate": 9.987683998768401e-06, + "loss": 0.1687, + "step": 12323 + }, + { + "epoch": 5.729428172942817, + "grad_norm": 4.115877628326416, + "learning_rate": 9.987682998768302e-06, + "loss": 0.185, + "step": 12324 + }, + { + "epoch": 5.729893072989308, + "grad_norm": 7.268655300140381, + "learning_rate": 9.9876819987682e-06, + "loss": 0.1852, + "step": 12325 + }, + { + "epoch": 5.730357973035797, + "grad_norm": 7.234052658081055, + "learning_rate": 9.987680998768101e-06, + "loss": 0.331, + "step": 12326 + }, + { + "epoch": 5.730822873082287, + "grad_norm": 5.211357593536377, + "learning_rate": 9.987679998768e-06, + "loss": 0.2613, + "step": 12327 + }, + { + "epoch": 5.731287773128777, + "grad_norm": 11.777497291564941, + "learning_rate": 9.987678998767901e-06, + "loss": 0.3207, + "step": 12328 + }, + { + "epoch": 5.731752673175267, + "grad_norm": 6.726089954376221, + "learning_rate": 9.9876779987678e-06, + "loss": 0.3325, + "step": 12329 + }, + { + "epoch": 5.7322175732217575, + "grad_norm": 3.6874747276306152, + "learning_rate": 9.9876769987677e-06, + "loss": 0.0812, + "step": 12330 + }, + { + "epoch": 5.732682473268247, + "grad_norm": 3.7308430671691895, + "learning_rate": 9.987675998767601e-06, + "loss": 0.0655, + "step": 12331 + }, + { + "epoch": 5.733147373314737, + "grad_norm": 4.524864196777344, + "learning_rate": 9.9876749987675e-06, + "loss": 0.1532, + "step": 12332 + }, + { + "epoch": 5.733612273361228, + "grad_norm": 5.418142795562744, + "learning_rate": 9.987673998767401e-06, + "loss": 0.2064, + "step": 12333 + }, + { + "epoch": 5.734077173407718, + "grad_norm": 3.8504607677459717, + "learning_rate": 9.9876729987673e-06, + "loss": 0.1265, + "step": 12334 + }, + { + "epoch": 5.734542073454207, + "grad_norm": 6.455188274383545, + "learning_rate": 9.9876719987672e-06, + "loss": 0.1851, + "step": 12335 + }, + { + "epoch": 5.735006973500697, + "grad_norm": 7.217745304107666, + "learning_rate": 9.987670998767101e-06, + "loss": 0.1757, + "step": 12336 + }, + { + "epoch": 5.735471873547187, + "grad_norm": 8.103734970092773, + "learning_rate": 9.987669998767002e-06, + "loss": 0.3657, + "step": 12337 + }, + { + "epoch": 5.735936773593678, + "grad_norm": 9.280885696411133, + "learning_rate": 9.9876689987669e-06, + "loss": 0.2756, + "step": 12338 + }, + { + "epoch": 5.7364016736401675, + "grad_norm": 8.648914337158203, + "learning_rate": 9.9876679987668e-06, + "loss": 0.2688, + "step": 12339 + }, + { + "epoch": 5.736866573686657, + "grad_norm": 6.978066444396973, + "learning_rate": 9.9876669987667e-06, + "loss": 0.215, + "step": 12340 + }, + { + "epoch": 5.737331473733147, + "grad_norm": 9.973934173583984, + "learning_rate": 9.987665998766601e-06, + "loss": 0.784, + "step": 12341 + }, + { + "epoch": 5.737796373779638, + "grad_norm": 5.557358264923096, + "learning_rate": 9.987664998766502e-06, + "loss": 0.3333, + "step": 12342 + }, + { + "epoch": 5.738261273826128, + "grad_norm": 5.765900135040283, + "learning_rate": 9.9876639987664e-06, + "loss": 0.246, + "step": 12343 + }, + { + "epoch": 5.738726173872617, + "grad_norm": 9.409724235534668, + "learning_rate": 9.987662998766301e-06, + "loss": 0.6331, + "step": 12344 + }, + { + "epoch": 5.739191073919107, + "grad_norm": 6.468050479888916, + "learning_rate": 9.9876619987662e-06, + "loss": 0.3016, + "step": 12345 + }, + { + "epoch": 5.739655973965597, + "grad_norm": 4.973047733306885, + "learning_rate": 9.987660998766101e-06, + "loss": 0.1748, + "step": 12346 + }, + { + "epoch": 5.740120874012088, + "grad_norm": 4.211171627044678, + "learning_rate": 9.987659998766e-06, + "loss": 0.2075, + "step": 12347 + }, + { + "epoch": 5.7405857740585775, + "grad_norm": 5.724331378936768, + "learning_rate": 9.9876589987659e-06, + "loss": 0.1476, + "step": 12348 + }, + { + "epoch": 5.741050674105067, + "grad_norm": 6.871070861816406, + "learning_rate": 9.987657998765801e-06, + "loss": 0.2711, + "step": 12349 + }, + { + "epoch": 5.741515574151557, + "grad_norm": 3.883481740951538, + "learning_rate": 9.9876569987657e-06, + "loss": 0.0528, + "step": 12350 + }, + { + "epoch": 5.741980474198048, + "grad_norm": 6.314239978790283, + "learning_rate": 9.9876559987656e-06, + "loss": 0.2367, + "step": 12351 + }, + { + "epoch": 5.742445374244538, + "grad_norm": 3.42179536819458, + "learning_rate": 9.9876549987655e-06, + "loss": 0.1055, + "step": 12352 + }, + { + "epoch": 5.742910274291027, + "grad_norm": 5.16832160949707, + "learning_rate": 9.9876539987654e-06, + "loss": 0.1797, + "step": 12353 + }, + { + "epoch": 5.743375174337517, + "grad_norm": 5.5555100440979, + "learning_rate": 9.987652998765301e-06, + "loss": 0.2139, + "step": 12354 + }, + { + "epoch": 5.743840074384007, + "grad_norm": 7.3680548667907715, + "learning_rate": 9.987651998765202e-06, + "loss": 0.4215, + "step": 12355 + }, + { + "epoch": 5.744304974430498, + "grad_norm": 4.929687976837158, + "learning_rate": 9.9876509987651e-06, + "loss": 0.2323, + "step": 12356 + }, + { + "epoch": 5.7447698744769875, + "grad_norm": 4.43997859954834, + "learning_rate": 9.987649998765001e-06, + "loss": 0.1719, + "step": 12357 + }, + { + "epoch": 5.745234774523477, + "grad_norm": 4.159209728240967, + "learning_rate": 9.9876489987649e-06, + "loss": 0.0859, + "step": 12358 + }, + { + "epoch": 5.745699674569967, + "grad_norm": 5.941527366638184, + "learning_rate": 9.987647998764801e-06, + "loss": 0.3021, + "step": 12359 + }, + { + "epoch": 5.746164574616458, + "grad_norm": 7.225183010101318, + "learning_rate": 9.9876469987647e-06, + "loss": 0.3174, + "step": 12360 + }, + { + "epoch": 5.746629474662948, + "grad_norm": 10.401921272277832, + "learning_rate": 9.9876459987646e-06, + "loss": 0.5178, + "step": 12361 + }, + { + "epoch": 5.747094374709437, + "grad_norm": 5.367983341217041, + "learning_rate": 9.987644998764501e-06, + "loss": 0.1215, + "step": 12362 + }, + { + "epoch": 5.747559274755927, + "grad_norm": 3.5973448753356934, + "learning_rate": 9.9876439987644e-06, + "loss": 0.0616, + "step": 12363 + }, + { + "epoch": 5.748024174802417, + "grad_norm": 5.352502822875977, + "learning_rate": 9.987642998764301e-06, + "loss": 0.2278, + "step": 12364 + }, + { + "epoch": 5.748489074848908, + "grad_norm": 5.025609016418457, + "learning_rate": 9.9876419987642e-06, + "loss": 0.2092, + "step": 12365 + }, + { + "epoch": 5.7489539748953975, + "grad_norm": 5.916660308837891, + "learning_rate": 9.9876409987641e-06, + "loss": 0.1752, + "step": 12366 + }, + { + "epoch": 5.749418874941887, + "grad_norm": 6.446516990661621, + "learning_rate": 9.987639998764001e-06, + "loss": 0.3006, + "step": 12367 + }, + { + "epoch": 5.749883774988378, + "grad_norm": 3.8067405223846436, + "learning_rate": 9.987638998763902e-06, + "loss": 0.1606, + "step": 12368 + }, + { + "epoch": 5.750348675034868, + "grad_norm": 4.43110990524292, + "learning_rate": 9.9876379987638e-06, + "loss": 0.1073, + "step": 12369 + }, + { + "epoch": 5.750813575081358, + "grad_norm": 7.460370063781738, + "learning_rate": 9.987636998763701e-06, + "loss": 0.4119, + "step": 12370 + }, + { + "epoch": 5.751278475127847, + "grad_norm": 5.12186861038208, + "learning_rate": 9.9876359987636e-06, + "loss": 0.1387, + "step": 12371 + }, + { + "epoch": 5.751743375174337, + "grad_norm": 3.1420176029205322, + "learning_rate": 9.9876349987635e-06, + "loss": 0.091, + "step": 12372 + }, + { + "epoch": 5.752208275220828, + "grad_norm": 8.141134262084961, + "learning_rate": 9.9876339987634e-06, + "loss": 0.3683, + "step": 12373 + }, + { + "epoch": 5.752673175267318, + "grad_norm": 5.7108845710754395, + "learning_rate": 9.9876329987633e-06, + "loss": 0.1935, + "step": 12374 + }, + { + "epoch": 5.7531380753138075, + "grad_norm": 5.221104145050049, + "learning_rate": 9.987631998763201e-06, + "loss": 0.2182, + "step": 12375 + }, + { + "epoch": 5.753602975360297, + "grad_norm": 2.951561450958252, + "learning_rate": 9.9876309987631e-06, + "loss": 0.0879, + "step": 12376 + }, + { + "epoch": 5.754067875406788, + "grad_norm": 12.013527870178223, + "learning_rate": 9.987629998763001e-06, + "loss": 0.2115, + "step": 12377 + }, + { + "epoch": 5.754532775453278, + "grad_norm": 7.004714488983154, + "learning_rate": 9.9876289987629e-06, + "loss": 0.2586, + "step": 12378 + }, + { + "epoch": 5.754997675499768, + "grad_norm": 4.180604934692383, + "learning_rate": 9.9876279987628e-06, + "loss": 0.1731, + "step": 12379 + }, + { + "epoch": 5.755462575546257, + "grad_norm": 4.886331558227539, + "learning_rate": 9.987626998762701e-06, + "loss": 0.2056, + "step": 12380 + }, + { + "epoch": 5.755927475592747, + "grad_norm": 8.31771183013916, + "learning_rate": 9.987625998762602e-06, + "loss": 0.5479, + "step": 12381 + }, + { + "epoch": 5.756392375639238, + "grad_norm": 6.760024070739746, + "learning_rate": 9.987624998762501e-06, + "loss": 0.3738, + "step": 12382 + }, + { + "epoch": 5.756857275685728, + "grad_norm": 6.014089584350586, + "learning_rate": 9.9876239987624e-06, + "loss": 0.2959, + "step": 12383 + }, + { + "epoch": 5.7573221757322175, + "grad_norm": 5.45219612121582, + "learning_rate": 9.9876229987623e-06, + "loss": 0.3043, + "step": 12384 + }, + { + "epoch": 5.757787075778707, + "grad_norm": 6.87967586517334, + "learning_rate": 9.9876219987622e-06, + "loss": 0.2794, + "step": 12385 + }, + { + "epoch": 5.758251975825198, + "grad_norm": 5.888274192810059, + "learning_rate": 9.9876209987621e-06, + "loss": 0.3092, + "step": 12386 + }, + { + "epoch": 5.758716875871688, + "grad_norm": 5.625553607940674, + "learning_rate": 9.987619998762e-06, + "loss": 0.1454, + "step": 12387 + }, + { + "epoch": 5.759181775918178, + "grad_norm": 2.748180389404297, + "learning_rate": 9.987618998761901e-06, + "loss": 0.1006, + "step": 12388 + }, + { + "epoch": 5.759646675964667, + "grad_norm": 4.664681434631348, + "learning_rate": 9.9876179987618e-06, + "loss": 0.1784, + "step": 12389 + }, + { + "epoch": 5.760111576011157, + "grad_norm": 4.951648235321045, + "learning_rate": 9.987616998761701e-06, + "loss": 0.2759, + "step": 12390 + }, + { + "epoch": 5.760576476057648, + "grad_norm": 29.57087516784668, + "learning_rate": 9.9876159987616e-06, + "loss": 0.2248, + "step": 12391 + }, + { + "epoch": 5.761041376104138, + "grad_norm": 5.126036643981934, + "learning_rate": 9.9876149987615e-06, + "loss": 0.1424, + "step": 12392 + }, + { + "epoch": 5.7615062761506275, + "grad_norm": 10.739431381225586, + "learning_rate": 9.987613998761401e-06, + "loss": 0.6768, + "step": 12393 + }, + { + "epoch": 5.761971176197117, + "grad_norm": 2.8288955688476562, + "learning_rate": 9.9876129987613e-06, + "loss": 0.1219, + "step": 12394 + }, + { + "epoch": 5.762436076243608, + "grad_norm": 4.620542049407959, + "learning_rate": 9.987611998761201e-06, + "loss": 0.1522, + "step": 12395 + }, + { + "epoch": 5.762900976290098, + "grad_norm": 6.925504684448242, + "learning_rate": 9.9876109987611e-06, + "loss": 0.2946, + "step": 12396 + }, + { + "epoch": 5.763365876336588, + "grad_norm": 5.793567180633545, + "learning_rate": 9.987609998761e-06, + "loss": 0.2709, + "step": 12397 + }, + { + "epoch": 5.763830776383077, + "grad_norm": 6.030614376068115, + "learning_rate": 9.9876089987609e-06, + "loss": 0.275, + "step": 12398 + }, + { + "epoch": 5.764295676429567, + "grad_norm": 5.003636837005615, + "learning_rate": 9.9876079987608e-06, + "loss": 0.2643, + "step": 12399 + }, + { + "epoch": 5.764760576476058, + "grad_norm": 2.8558409214019775, + "learning_rate": 9.987606998760701e-06, + "loss": 0.0671, + "step": 12400 + }, + { + "epoch": 5.765225476522548, + "grad_norm": 4.64305305480957, + "learning_rate": 9.987605998760602e-06, + "loss": 0.1398, + "step": 12401 + }, + { + "epoch": 5.7656903765690375, + "grad_norm": 5.3029093742370605, + "learning_rate": 9.9876049987605e-06, + "loss": 0.2826, + "step": 12402 + }, + { + "epoch": 5.766155276615527, + "grad_norm": 6.1825737953186035, + "learning_rate": 9.987603998760401e-06, + "loss": 0.41, + "step": 12403 + }, + { + "epoch": 5.766620176662018, + "grad_norm": 4.844883441925049, + "learning_rate": 9.9876029987603e-06, + "loss": 0.2773, + "step": 12404 + }, + { + "epoch": 5.767085076708508, + "grad_norm": 8.289796829223633, + "learning_rate": 9.9876019987602e-06, + "loss": 0.4234, + "step": 12405 + }, + { + "epoch": 5.767549976754998, + "grad_norm": 3.811534881591797, + "learning_rate": 9.987600998760101e-06, + "loss": 0.1019, + "step": 12406 + }, + { + "epoch": 5.768014876801487, + "grad_norm": 4.86077356338501, + "learning_rate": 9.98759999876e-06, + "loss": 0.1838, + "step": 12407 + }, + { + "epoch": 5.768479776847977, + "grad_norm": 6.002965450286865, + "learning_rate": 9.987598998759901e-06, + "loss": 0.2937, + "step": 12408 + }, + { + "epoch": 5.768944676894468, + "grad_norm": 6.118922233581543, + "learning_rate": 9.9875979987598e-06, + "loss": 0.4007, + "step": 12409 + }, + { + "epoch": 5.769409576940958, + "grad_norm": 6.462087631225586, + "learning_rate": 9.9875969987597e-06, + "loss": 0.3231, + "step": 12410 + }, + { + "epoch": 5.7698744769874475, + "grad_norm": 7.890545845031738, + "learning_rate": 9.987595998759601e-06, + "loss": 0.2506, + "step": 12411 + }, + { + "epoch": 5.770339377033938, + "grad_norm": 8.089654922485352, + "learning_rate": 9.9875949987595e-06, + "loss": 0.6599, + "step": 12412 + }, + { + "epoch": 5.770804277080428, + "grad_norm": 6.064271926879883, + "learning_rate": 9.987593998759401e-06, + "loss": 0.3393, + "step": 12413 + }, + { + "epoch": 5.771269177126918, + "grad_norm": 5.445112228393555, + "learning_rate": 9.987592998759302e-06, + "loss": 0.3287, + "step": 12414 + }, + { + "epoch": 5.771734077173408, + "grad_norm": 4.813650608062744, + "learning_rate": 9.9875919987592e-06, + "loss": 0.1772, + "step": 12415 + }, + { + "epoch": 5.772198977219897, + "grad_norm": 5.328141212463379, + "learning_rate": 9.9875909987591e-06, + "loss": 0.1928, + "step": 12416 + }, + { + "epoch": 5.772663877266388, + "grad_norm": 9.331888198852539, + "learning_rate": 9.987589998759e-06, + "loss": 0.3501, + "step": 12417 + }, + { + "epoch": 5.773128777312878, + "grad_norm": 8.09553050994873, + "learning_rate": 9.987588998758901e-06, + "loss": 0.6303, + "step": 12418 + }, + { + "epoch": 5.773593677359368, + "grad_norm": 6.701943397521973, + "learning_rate": 9.987587998758802e-06, + "loss": 0.3921, + "step": 12419 + }, + { + "epoch": 5.7740585774058575, + "grad_norm": 12.861113548278809, + "learning_rate": 9.9875869987587e-06, + "loss": 0.3202, + "step": 12420 + }, + { + "epoch": 5.774523477452348, + "grad_norm": 7.541524410247803, + "learning_rate": 9.987585998758601e-06, + "loss": 0.2064, + "step": 12421 + }, + { + "epoch": 5.774988377498838, + "grad_norm": 5.090665817260742, + "learning_rate": 9.9875849987585e-06, + "loss": 0.1528, + "step": 12422 + }, + { + "epoch": 5.775453277545328, + "grad_norm": 7.168289661407471, + "learning_rate": 9.9875839987584e-06, + "loss": 0.4335, + "step": 12423 + }, + { + "epoch": 5.775918177591818, + "grad_norm": 5.171245098114014, + "learning_rate": 9.987582998758301e-06, + "loss": 0.2772, + "step": 12424 + }, + { + "epoch": 5.776383077638307, + "grad_norm": 1.9049663543701172, + "learning_rate": 9.9875819987582e-06, + "loss": 0.0439, + "step": 12425 + }, + { + "epoch": 5.776847977684798, + "grad_norm": 4.611715316772461, + "learning_rate": 9.987580998758101e-06, + "loss": 0.2168, + "step": 12426 + }, + { + "epoch": 5.777312877731288, + "grad_norm": 8.141369819641113, + "learning_rate": 9.987579998758e-06, + "loss": 0.1745, + "step": 12427 + }, + { + "epoch": 5.777777777777778, + "grad_norm": 4.347712993621826, + "learning_rate": 9.9875789987579e-06, + "loss": 0.1152, + "step": 12428 + }, + { + "epoch": 5.7782426778242675, + "grad_norm": 8.2291898727417, + "learning_rate": 9.9875779987578e-06, + "loss": 0.5409, + "step": 12429 + }, + { + "epoch": 5.778707577870758, + "grad_norm": 6.73563289642334, + "learning_rate": 9.9875769987577e-06, + "loss": 0.2835, + "step": 12430 + }, + { + "epoch": 5.779172477917248, + "grad_norm": 5.568181037902832, + "learning_rate": 9.987575998757601e-06, + "loss": 0.3647, + "step": 12431 + }, + { + "epoch": 5.779637377963738, + "grad_norm": 6.9285502433776855, + "learning_rate": 9.987574998757502e-06, + "loss": 0.2872, + "step": 12432 + }, + { + "epoch": 5.780102278010228, + "grad_norm": 6.399144649505615, + "learning_rate": 9.9875739987574e-06, + "loss": 0.199, + "step": 12433 + }, + { + "epoch": 5.780567178056717, + "grad_norm": 7.335097789764404, + "learning_rate": 9.987572998757301e-06, + "loss": 0.3714, + "step": 12434 + }, + { + "epoch": 5.781032078103208, + "grad_norm": 2.797732353210449, + "learning_rate": 9.9875719987572e-06, + "loss": 0.07, + "step": 12435 + }, + { + "epoch": 5.781496978149698, + "grad_norm": 8.380186080932617, + "learning_rate": 9.9875709987571e-06, + "loss": 0.3338, + "step": 12436 + }, + { + "epoch": 5.781961878196188, + "grad_norm": 4.663298606872559, + "learning_rate": 9.987569998757002e-06, + "loss": 0.2686, + "step": 12437 + }, + { + "epoch": 5.7824267782426775, + "grad_norm": 4.222248554229736, + "learning_rate": 9.9875689987569e-06, + "loss": 0.1693, + "step": 12438 + }, + { + "epoch": 5.782891678289168, + "grad_norm": 3.1187996864318848, + "learning_rate": 9.987567998756801e-06, + "loss": 0.1022, + "step": 12439 + }, + { + "epoch": 5.783356578335658, + "grad_norm": 6.929786682128906, + "learning_rate": 9.9875669987567e-06, + "loss": 0.2935, + "step": 12440 + }, + { + "epoch": 5.783821478382148, + "grad_norm": 7.347548484802246, + "learning_rate": 9.9875659987566e-06, + "loss": 0.38, + "step": 12441 + }, + { + "epoch": 5.784286378428638, + "grad_norm": 4.711312294006348, + "learning_rate": 9.9875649987565e-06, + "loss": 0.2482, + "step": 12442 + }, + { + "epoch": 5.784751278475127, + "grad_norm": 7.303674697875977, + "learning_rate": 9.9875639987564e-06, + "loss": 0.2815, + "step": 12443 + }, + { + "epoch": 5.785216178521618, + "grad_norm": 6.6603102684021, + "learning_rate": 9.987562998756301e-06, + "loss": 0.559, + "step": 12444 + }, + { + "epoch": 5.785681078568108, + "grad_norm": 5.250842571258545, + "learning_rate": 9.987561998756202e-06, + "loss": 0.4492, + "step": 12445 + }, + { + "epoch": 5.786145978614598, + "grad_norm": 3.1234192848205566, + "learning_rate": 9.9875609987561e-06, + "loss": 0.0686, + "step": 12446 + }, + { + "epoch": 5.786610878661088, + "grad_norm": 9.004293441772461, + "learning_rate": 9.987559998756001e-06, + "loss": 0.357, + "step": 12447 + }, + { + "epoch": 5.787075778707578, + "grad_norm": 4.1695709228515625, + "learning_rate": 9.9875589987559e-06, + "loss": 0.1125, + "step": 12448 + }, + { + "epoch": 5.787540678754068, + "grad_norm": 4.0777716636657715, + "learning_rate": 9.987557998755801e-06, + "loss": 0.0858, + "step": 12449 + }, + { + "epoch": 5.788005578800558, + "grad_norm": 7.69725227355957, + "learning_rate": 9.987556998755702e-06, + "loss": 0.4679, + "step": 12450 + }, + { + "epoch": 5.788470478847048, + "grad_norm": 3.5070393085479736, + "learning_rate": 9.9875559987556e-06, + "loss": 0.1055, + "step": 12451 + }, + { + "epoch": 5.788935378893538, + "grad_norm": 8.086538314819336, + "learning_rate": 9.987554998755501e-06, + "loss": 0.3769, + "step": 12452 + }, + { + "epoch": 5.789400278940028, + "grad_norm": 3.337506055831909, + "learning_rate": 9.9875539987554e-06, + "loss": 0.1198, + "step": 12453 + }, + { + "epoch": 5.789865178986518, + "grad_norm": 4.275012016296387, + "learning_rate": 9.9875529987553e-06, + "loss": 0.1683, + "step": 12454 + }, + { + "epoch": 5.790330079033008, + "grad_norm": 3.3894619941711426, + "learning_rate": 9.9875519987552e-06, + "loss": 0.0855, + "step": 12455 + }, + { + "epoch": 5.790794979079498, + "grad_norm": 8.863667488098145, + "learning_rate": 9.9875509987551e-06, + "loss": 0.5107, + "step": 12456 + }, + { + "epoch": 5.791259879125988, + "grad_norm": 3.450399875640869, + "learning_rate": 9.987549998755001e-06, + "loss": 0.0876, + "step": 12457 + }, + { + "epoch": 5.791724779172478, + "grad_norm": 7.711775302886963, + "learning_rate": 9.9875489987549e-06, + "loss": 0.3866, + "step": 12458 + }, + { + "epoch": 5.792189679218968, + "grad_norm": 4.928114414215088, + "learning_rate": 9.9875479987548e-06, + "loss": 0.201, + "step": 12459 + }, + { + "epoch": 5.792654579265458, + "grad_norm": 7.268225193023682, + "learning_rate": 9.9875469987547e-06, + "loss": 0.4807, + "step": 12460 + }, + { + "epoch": 5.793119479311948, + "grad_norm": 3.7829031944274902, + "learning_rate": 9.9875459987546e-06, + "loss": 0.1167, + "step": 12461 + }, + { + "epoch": 5.793584379358438, + "grad_norm": 7.63043212890625, + "learning_rate": 9.987544998754501e-06, + "loss": 0.3515, + "step": 12462 + }, + { + "epoch": 5.794049279404928, + "grad_norm": 6.469924449920654, + "learning_rate": 9.987543998754402e-06, + "loss": 0.3424, + "step": 12463 + }, + { + "epoch": 5.794514179451418, + "grad_norm": 5.469061374664307, + "learning_rate": 9.9875429987543e-06, + "loss": 0.1896, + "step": 12464 + }, + { + "epoch": 5.794979079497908, + "grad_norm": 6.197627067565918, + "learning_rate": 9.987541998754201e-06, + "loss": 0.1831, + "step": 12465 + }, + { + "epoch": 5.795443979544398, + "grad_norm": 3.3606069087982178, + "learning_rate": 9.9875409987541e-06, + "loss": 0.1082, + "step": 12466 + }, + { + "epoch": 5.795908879590888, + "grad_norm": 6.406025409698486, + "learning_rate": 9.987539998754001e-06, + "loss": 0.1863, + "step": 12467 + }, + { + "epoch": 5.796373779637378, + "grad_norm": 4.21198844909668, + "learning_rate": 9.9875389987539e-06, + "loss": 0.0894, + "step": 12468 + }, + { + "epoch": 5.796838679683868, + "grad_norm": 5.479652404785156, + "learning_rate": 9.9875379987538e-06, + "loss": 0.3261, + "step": 12469 + }, + { + "epoch": 5.797303579730358, + "grad_norm": 5.274005889892578, + "learning_rate": 9.987536998753701e-06, + "loss": 0.224, + "step": 12470 + }, + { + "epoch": 5.797768479776848, + "grad_norm": 5.336792945861816, + "learning_rate": 9.9875359987536e-06, + "loss": 0.2035, + "step": 12471 + }, + { + "epoch": 5.798233379823338, + "grad_norm": 4.435632228851318, + "learning_rate": 9.9875349987535e-06, + "loss": 0.0799, + "step": 12472 + }, + { + "epoch": 5.798698279869828, + "grad_norm": 6.448853015899658, + "learning_rate": 9.9875339987534e-06, + "loss": 0.289, + "step": 12473 + }, + { + "epoch": 5.799163179916318, + "grad_norm": 3.5549681186676025, + "learning_rate": 9.9875329987533e-06, + "loss": 0.1091, + "step": 12474 + }, + { + "epoch": 5.799628079962808, + "grad_norm": 13.117223739624023, + "learning_rate": 9.987531998753201e-06, + "loss": 0.3879, + "step": 12475 + }, + { + "epoch": 5.800092980009298, + "grad_norm": 5.703701019287109, + "learning_rate": 9.987530998753102e-06, + "loss": 0.3014, + "step": 12476 + }, + { + "epoch": 5.800557880055788, + "grad_norm": 7.0445780754089355, + "learning_rate": 9.987529998753e-06, + "loss": 0.1706, + "step": 12477 + }, + { + "epoch": 5.801022780102278, + "grad_norm": 7.509753227233887, + "learning_rate": 9.987528998752901e-06, + "loss": 0.205, + "step": 12478 + }, + { + "epoch": 5.801487680148768, + "grad_norm": 7.507884502410889, + "learning_rate": 9.9875279987528e-06, + "loss": 0.4352, + "step": 12479 + }, + { + "epoch": 5.801952580195258, + "grad_norm": 4.9829254150390625, + "learning_rate": 9.987526998752701e-06, + "loss": 0.1551, + "step": 12480 + }, + { + "epoch": 5.802417480241748, + "grad_norm": 4.129003524780273, + "learning_rate": 9.9875259987526e-06, + "loss": 0.1706, + "step": 12481 + }, + { + "epoch": 5.802882380288238, + "grad_norm": 6.030624866485596, + "learning_rate": 9.9875249987525e-06, + "loss": 0.2561, + "step": 12482 + }, + { + "epoch": 5.803347280334728, + "grad_norm": 4.498439311981201, + "learning_rate": 9.987523998752401e-06, + "loss": 0.2268, + "step": 12483 + }, + { + "epoch": 5.803812180381218, + "grad_norm": 6.9184956550598145, + "learning_rate": 9.9875229987523e-06, + "loss": 0.4523, + "step": 12484 + }, + { + "epoch": 5.804277080427708, + "grad_norm": 6.239058017730713, + "learning_rate": 9.987521998752201e-06, + "loss": 0.3443, + "step": 12485 + }, + { + "epoch": 5.804741980474198, + "grad_norm": 6.762894630432129, + "learning_rate": 9.9875209987521e-06, + "loss": 0.2761, + "step": 12486 + }, + { + "epoch": 5.805206880520688, + "grad_norm": 6.542507648468018, + "learning_rate": 9.987519998752e-06, + "loss": 0.2806, + "step": 12487 + }, + { + "epoch": 5.805671780567178, + "grad_norm": 7.453958511352539, + "learning_rate": 9.987518998751901e-06, + "loss": 0.3363, + "step": 12488 + }, + { + "epoch": 5.806136680613668, + "grad_norm": 4.329196929931641, + "learning_rate": 9.987517998751802e-06, + "loss": 0.2362, + "step": 12489 + }, + { + "epoch": 5.806601580660158, + "grad_norm": 10.896238327026367, + "learning_rate": 9.9875169987517e-06, + "loss": 0.3338, + "step": 12490 + }, + { + "epoch": 5.8070664807066485, + "grad_norm": 11.164793968200684, + "learning_rate": 9.9875159987516e-06, + "loss": 0.2286, + "step": 12491 + }, + { + "epoch": 5.807531380753138, + "grad_norm": 4.81406831741333, + "learning_rate": 9.9875149987515e-06, + "loss": 0.1818, + "step": 12492 + }, + { + "epoch": 5.807996280799628, + "grad_norm": 6.529824256896973, + "learning_rate": 9.987513998751401e-06, + "loss": 0.2582, + "step": 12493 + }, + { + "epoch": 5.808461180846118, + "grad_norm": 4.561014652252197, + "learning_rate": 9.9875129987513e-06, + "loss": 0.262, + "step": 12494 + }, + { + "epoch": 5.808926080892608, + "grad_norm": 4.151762008666992, + "learning_rate": 9.9875119987512e-06, + "loss": 0.1916, + "step": 12495 + }, + { + "epoch": 5.8093909809390984, + "grad_norm": 5.5507965087890625, + "learning_rate": 9.987510998751101e-06, + "loss": 0.1639, + "step": 12496 + }, + { + "epoch": 5.809855880985588, + "grad_norm": 6.636428356170654, + "learning_rate": 9.987509998751e-06, + "loss": 0.1729, + "step": 12497 + }, + { + "epoch": 5.810320781032078, + "grad_norm": 6.5795817375183105, + "learning_rate": 9.987508998750901e-06, + "loss": 0.3952, + "step": 12498 + }, + { + "epoch": 5.810785681078568, + "grad_norm": 4.817804336547852, + "learning_rate": 9.9875079987508e-06, + "loss": 0.2677, + "step": 12499 + }, + { + "epoch": 5.8112505811250585, + "grad_norm": 8.617138862609863, + "learning_rate": 9.9875069987507e-06, + "loss": 0.4101, + "step": 12500 + }, + { + "epoch": 5.811715481171548, + "grad_norm": 4.877542972564697, + "learning_rate": 9.987505998750601e-06, + "loss": 0.2088, + "step": 12501 + }, + { + "epoch": 5.812180381218038, + "grad_norm": 7.311886787414551, + "learning_rate": 9.9875049987505e-06, + "loss": 0.1309, + "step": 12502 + }, + { + "epoch": 5.812645281264528, + "grad_norm": 4.078858852386475, + "learning_rate": 9.987503998750401e-06, + "loss": 0.1419, + "step": 12503 + }, + { + "epoch": 5.813110181311018, + "grad_norm": 4.7640557289123535, + "learning_rate": 9.9875029987503e-06, + "loss": 0.1032, + "step": 12504 + }, + { + "epoch": 5.8135750813575084, + "grad_norm": 4.083815574645996, + "learning_rate": 9.9875019987502e-06, + "loss": 0.1886, + "step": 12505 + }, + { + "epoch": 5.814039981403998, + "grad_norm": 5.425678253173828, + "learning_rate": 9.987500998750101e-06, + "loss": 0.2284, + "step": 12506 + }, + { + "epoch": 5.814504881450488, + "grad_norm": 7.384053707122803, + "learning_rate": 9.987499998750002e-06, + "loss": 0.3905, + "step": 12507 + }, + { + "epoch": 5.814969781496978, + "grad_norm": 6.361844062805176, + "learning_rate": 9.9874989987499e-06, + "loss": 0.2013, + "step": 12508 + }, + { + "epoch": 5.8154346815434685, + "grad_norm": 4.047618865966797, + "learning_rate": 9.987497998749801e-06, + "loss": 0.1511, + "step": 12509 + }, + { + "epoch": 5.815899581589958, + "grad_norm": 4.293447494506836, + "learning_rate": 9.9874969987497e-06, + "loss": 0.1678, + "step": 12510 + }, + { + "epoch": 5.816364481636448, + "grad_norm": 6.532107830047607, + "learning_rate": 9.987495998749601e-06, + "loss": 0.1559, + "step": 12511 + }, + { + "epoch": 5.816829381682938, + "grad_norm": 8.059553146362305, + "learning_rate": 9.9874949987495e-06, + "loss": 0.4509, + "step": 12512 + }, + { + "epoch": 5.817294281729428, + "grad_norm": 4.646440029144287, + "learning_rate": 9.9874939987494e-06, + "loss": 0.1397, + "step": 12513 + }, + { + "epoch": 5.817759181775918, + "grad_norm": 8.125102996826172, + "learning_rate": 9.987492998749301e-06, + "loss": 0.4565, + "step": 12514 + }, + { + "epoch": 5.818224081822408, + "grad_norm": 6.269319534301758, + "learning_rate": 9.9874919987492e-06, + "loss": 0.2043, + "step": 12515 + }, + { + "epoch": 5.818688981868898, + "grad_norm": 5.548915386199951, + "learning_rate": 9.987490998749101e-06, + "loss": 0.1619, + "step": 12516 + }, + { + "epoch": 5.819153881915388, + "grad_norm": 3.8942179679870605, + "learning_rate": 9.987489998749e-06, + "loss": 0.0809, + "step": 12517 + }, + { + "epoch": 5.8196187819618785, + "grad_norm": 6.000967502593994, + "learning_rate": 9.9874889987489e-06, + "loss": 0.313, + "step": 12518 + }, + { + "epoch": 5.820083682008368, + "grad_norm": 2.344953775405884, + "learning_rate": 9.987487998748801e-06, + "loss": 0.0414, + "step": 12519 + }, + { + "epoch": 5.820548582054858, + "grad_norm": 5.151068210601807, + "learning_rate": 9.987486998748702e-06, + "loss": 0.4119, + "step": 12520 + }, + { + "epoch": 5.821013482101348, + "grad_norm": 4.117042541503906, + "learning_rate": 9.987485998748601e-06, + "loss": 0.1073, + "step": 12521 + }, + { + "epoch": 5.821478382147838, + "grad_norm": 6.20100212097168, + "learning_rate": 9.987484998748502e-06, + "loss": 0.2687, + "step": 12522 + }, + { + "epoch": 5.821943282194328, + "grad_norm": 7.049336910247803, + "learning_rate": 9.9874839987484e-06, + "loss": 0.4093, + "step": 12523 + }, + { + "epoch": 5.822408182240818, + "grad_norm": 6.19357967376709, + "learning_rate": 9.9874829987483e-06, + "loss": 0.2304, + "step": 12524 + }, + { + "epoch": 5.822873082287308, + "grad_norm": 2.866101026535034, + "learning_rate": 9.9874819987482e-06, + "loss": 0.0705, + "step": 12525 + }, + { + "epoch": 5.823337982333798, + "grad_norm": 11.28722858428955, + "learning_rate": 9.9874809987481e-06, + "loss": 0.3212, + "step": 12526 + }, + { + "epoch": 5.8238028823802885, + "grad_norm": 3.418658494949341, + "learning_rate": 9.987479998748001e-06, + "loss": 0.1267, + "step": 12527 + }, + { + "epoch": 5.824267782426778, + "grad_norm": 5.8687286376953125, + "learning_rate": 9.9874789987479e-06, + "loss": 0.3097, + "step": 12528 + }, + { + "epoch": 5.824732682473268, + "grad_norm": 5.784587860107422, + "learning_rate": 9.987477998747801e-06, + "loss": 0.219, + "step": 12529 + }, + { + "epoch": 5.825197582519758, + "grad_norm": 3.7181222438812256, + "learning_rate": 9.9874769987477e-06, + "loss": 0.0912, + "step": 12530 + }, + { + "epoch": 5.825662482566249, + "grad_norm": 7.563018321990967, + "learning_rate": 9.9874759987476e-06, + "loss": 0.3192, + "step": 12531 + }, + { + "epoch": 5.826127382612738, + "grad_norm": 6.024291038513184, + "learning_rate": 9.987474998747501e-06, + "loss": 0.3098, + "step": 12532 + }, + { + "epoch": 5.826592282659228, + "grad_norm": 9.752391815185547, + "learning_rate": 9.987473998747402e-06, + "loss": 0.2854, + "step": 12533 + }, + { + "epoch": 5.827057182705718, + "grad_norm": 5.970663547515869, + "learning_rate": 9.987472998747301e-06, + "loss": 0.32, + "step": 12534 + }, + { + "epoch": 5.827522082752209, + "grad_norm": 5.691788196563721, + "learning_rate": 9.9874719987472e-06, + "loss": 0.3198, + "step": 12535 + }, + { + "epoch": 5.8279869827986985, + "grad_norm": 6.344272613525391, + "learning_rate": 9.9874709987471e-06, + "loss": 0.2202, + "step": 12536 + }, + { + "epoch": 5.828451882845188, + "grad_norm": 8.094742774963379, + "learning_rate": 9.987469998747e-06, + "loss": 0.419, + "step": 12537 + }, + { + "epoch": 5.828916782891678, + "grad_norm": 5.414390563964844, + "learning_rate": 9.9874689987469e-06, + "loss": 0.2572, + "step": 12538 + }, + { + "epoch": 5.829381682938168, + "grad_norm": 4.404802322387695, + "learning_rate": 9.9874679987468e-06, + "loss": 0.1389, + "step": 12539 + }, + { + "epoch": 5.829846582984659, + "grad_norm": 9.355874061584473, + "learning_rate": 9.987466998746701e-06, + "loss": 0.3993, + "step": 12540 + }, + { + "epoch": 5.830311483031148, + "grad_norm": 3.07627534866333, + "learning_rate": 9.9874659987466e-06, + "loss": 0.126, + "step": 12541 + }, + { + "epoch": 5.830776383077638, + "grad_norm": 5.4625935554504395, + "learning_rate": 9.987464998746501e-06, + "loss": 0.1783, + "step": 12542 + }, + { + "epoch": 5.831241283124128, + "grad_norm": 4.9798078536987305, + "learning_rate": 9.9874639987464e-06, + "loss": 0.2051, + "step": 12543 + }, + { + "epoch": 5.831706183170619, + "grad_norm": 7.294951915740967, + "learning_rate": 9.9874629987463e-06, + "loss": 0.2448, + "step": 12544 + }, + { + "epoch": 5.8321710832171085, + "grad_norm": 4.862918376922607, + "learning_rate": 9.987461998746201e-06, + "loss": 0.2197, + "step": 12545 + }, + { + "epoch": 5.832635983263598, + "grad_norm": 5.606544494628906, + "learning_rate": 9.9874609987461e-06, + "loss": 0.1898, + "step": 12546 + }, + { + "epoch": 5.833100883310088, + "grad_norm": 8.373903274536133, + "learning_rate": 9.987459998746001e-06, + "loss": 0.3656, + "step": 12547 + }, + { + "epoch": 5.833565783356578, + "grad_norm": 4.718838691711426, + "learning_rate": 9.9874589987459e-06, + "loss": 0.2063, + "step": 12548 + }, + { + "epoch": 5.834030683403069, + "grad_norm": 4.781779766082764, + "learning_rate": 9.9874579987458e-06, + "loss": 0.2011, + "step": 12549 + }, + { + "epoch": 5.834495583449558, + "grad_norm": 5.613640785217285, + "learning_rate": 9.9874569987457e-06, + "loss": 0.251, + "step": 12550 + }, + { + "epoch": 5.834960483496048, + "grad_norm": 6.9897966384887695, + "learning_rate": 9.9874559987456e-06, + "loss": 0.3434, + "step": 12551 + }, + { + "epoch": 5.835425383542538, + "grad_norm": 4.4634318351745605, + "learning_rate": 9.987454998745501e-06, + "loss": 0.1069, + "step": 12552 + }, + { + "epoch": 5.835890283589029, + "grad_norm": 4.9961395263671875, + "learning_rate": 9.987453998745402e-06, + "loss": 0.2185, + "step": 12553 + }, + { + "epoch": 5.8363551836355185, + "grad_norm": 5.771387100219727, + "learning_rate": 9.9874529987453e-06, + "loss": 0.0925, + "step": 12554 + }, + { + "epoch": 5.836820083682008, + "grad_norm": 6.899661064147949, + "learning_rate": 9.987451998745201e-06, + "loss": 0.387, + "step": 12555 + }, + { + "epoch": 5.837284983728498, + "grad_norm": 5.838687896728516, + "learning_rate": 9.9874509987451e-06, + "loss": 0.2609, + "step": 12556 + }, + { + "epoch": 5.837749883774988, + "grad_norm": 6.468730926513672, + "learning_rate": 9.987449998745e-06, + "loss": 0.3211, + "step": 12557 + }, + { + "epoch": 5.838214783821479, + "grad_norm": 8.202398300170898, + "learning_rate": 9.987448998744901e-06, + "loss": 0.5019, + "step": 12558 + }, + { + "epoch": 5.838679683867968, + "grad_norm": 6.685708522796631, + "learning_rate": 9.9874479987448e-06, + "loss": 0.1339, + "step": 12559 + }, + { + "epoch": 5.839144583914458, + "grad_norm": 9.71675968170166, + "learning_rate": 9.987446998744701e-06, + "loss": 0.4166, + "step": 12560 + }, + { + "epoch": 5.839609483960948, + "grad_norm": 4.481266498565674, + "learning_rate": 9.9874459987446e-06, + "loss": 0.3133, + "step": 12561 + }, + { + "epoch": 5.840074384007439, + "grad_norm": 5.0895562171936035, + "learning_rate": 9.9874449987445e-06, + "loss": 0.1833, + "step": 12562 + }, + { + "epoch": 5.8405392840539285, + "grad_norm": 6.3899664878845215, + "learning_rate": 9.9874439987444e-06, + "loss": 0.4046, + "step": 12563 + }, + { + "epoch": 5.841004184100418, + "grad_norm": 4.254045486450195, + "learning_rate": 9.9874429987443e-06, + "loss": 0.1674, + "step": 12564 + }, + { + "epoch": 5.841469084146908, + "grad_norm": 3.5619921684265137, + "learning_rate": 9.987441998744201e-06, + "loss": 0.1017, + "step": 12565 + }, + { + "epoch": 5.841933984193398, + "grad_norm": 5.07499885559082, + "learning_rate": 9.987440998744102e-06, + "loss": 0.2129, + "step": 12566 + }, + { + "epoch": 5.842398884239889, + "grad_norm": 5.416986465454102, + "learning_rate": 9.987439998744e-06, + "loss": 0.1945, + "step": 12567 + }, + { + "epoch": 5.842863784286378, + "grad_norm": 4.233792304992676, + "learning_rate": 9.9874389987439e-06, + "loss": 0.1881, + "step": 12568 + }, + { + "epoch": 5.843328684332868, + "grad_norm": 8.592947959899902, + "learning_rate": 9.9874379987438e-06, + "loss": 0.3712, + "step": 12569 + }, + { + "epoch": 5.843793584379359, + "grad_norm": 5.623281955718994, + "learning_rate": 9.987436998743701e-06, + "loss": 0.2592, + "step": 12570 + }, + { + "epoch": 5.844258484425849, + "grad_norm": 5.442298412322998, + "learning_rate": 9.987435998743602e-06, + "loss": 0.2801, + "step": 12571 + }, + { + "epoch": 5.8447233844723385, + "grad_norm": 4.691169738769531, + "learning_rate": 9.9874349987435e-06, + "loss": 0.1868, + "step": 12572 + }, + { + "epoch": 5.845188284518828, + "grad_norm": 5.756775856018066, + "learning_rate": 9.987433998743401e-06, + "loss": 0.2307, + "step": 12573 + }, + { + "epoch": 5.845653184565318, + "grad_norm": 4.8849029541015625, + "learning_rate": 9.9874329987433e-06, + "loss": 0.2608, + "step": 12574 + }, + { + "epoch": 5.846118084611809, + "grad_norm": 2.4429240226745605, + "learning_rate": 9.9874319987432e-06, + "loss": 0.0512, + "step": 12575 + }, + { + "epoch": 5.846582984658299, + "grad_norm": 8.50363540649414, + "learning_rate": 9.987430998743101e-06, + "loss": 0.3949, + "step": 12576 + }, + { + "epoch": 5.847047884704788, + "grad_norm": 3.153726100921631, + "learning_rate": 9.987429998743e-06, + "loss": 0.0849, + "step": 12577 + }, + { + "epoch": 5.847512784751278, + "grad_norm": 3.367784023284912, + "learning_rate": 9.987428998742901e-06, + "loss": 0.0745, + "step": 12578 + }, + { + "epoch": 5.847977684797769, + "grad_norm": 9.744597434997559, + "learning_rate": 9.9874279987428e-06, + "loss": 0.281, + "step": 12579 + }, + { + "epoch": 5.848442584844259, + "grad_norm": 6.5878400802612305, + "learning_rate": 9.9874269987427e-06, + "loss": 0.2684, + "step": 12580 + }, + { + "epoch": 5.8489074848907485, + "grad_norm": 5.194019317626953, + "learning_rate": 9.9874259987426e-06, + "loss": 0.2993, + "step": 12581 + }, + { + "epoch": 5.849372384937238, + "grad_norm": 3.3391952514648438, + "learning_rate": 9.9874249987425e-06, + "loss": 0.1019, + "step": 12582 + }, + { + "epoch": 5.849837284983728, + "grad_norm": 5.232353210449219, + "learning_rate": 9.987423998742401e-06, + "loss": 0.1699, + "step": 12583 + }, + { + "epoch": 5.850302185030219, + "grad_norm": 6.269271373748779, + "learning_rate": 9.987422998742302e-06, + "loss": 0.2544, + "step": 12584 + }, + { + "epoch": 5.850767085076709, + "grad_norm": 7.8062744140625, + "learning_rate": 9.9874219987422e-06, + "loss": 0.4266, + "step": 12585 + }, + { + "epoch": 5.851231985123198, + "grad_norm": 8.801427841186523, + "learning_rate": 9.987420998742101e-06, + "loss": 0.398, + "step": 12586 + }, + { + "epoch": 5.851696885169688, + "grad_norm": 7.457001686096191, + "learning_rate": 9.987419998742e-06, + "loss": 0.3475, + "step": 12587 + }, + { + "epoch": 5.852161785216179, + "grad_norm": 5.528017520904541, + "learning_rate": 9.987418998741901e-06, + "loss": 0.2474, + "step": 12588 + }, + { + "epoch": 5.852626685262669, + "grad_norm": 5.877744197845459, + "learning_rate": 9.987417998741802e-06, + "loss": 0.2535, + "step": 12589 + }, + { + "epoch": 5.8530915853091585, + "grad_norm": 6.617264747619629, + "learning_rate": 9.9874169987417e-06, + "loss": 0.3029, + "step": 12590 + }, + { + "epoch": 5.853556485355648, + "grad_norm": 4.850255489349365, + "learning_rate": 9.987415998741601e-06, + "loss": 0.1427, + "step": 12591 + }, + { + "epoch": 5.854021385402138, + "grad_norm": 5.005951404571533, + "learning_rate": 9.9874149987415e-06, + "loss": 0.2376, + "step": 12592 + }, + { + "epoch": 5.854486285448629, + "grad_norm": 1.6603912115097046, + "learning_rate": 9.9874139987414e-06, + "loss": 0.0293, + "step": 12593 + }, + { + "epoch": 5.854951185495119, + "grad_norm": 4.254489421844482, + "learning_rate": 9.9874129987413e-06, + "loss": 0.1662, + "step": 12594 + }, + { + "epoch": 5.855416085541608, + "grad_norm": 4.006986618041992, + "learning_rate": 9.9874119987412e-06, + "loss": 0.1819, + "step": 12595 + }, + { + "epoch": 5.855880985588098, + "grad_norm": 7.318634510040283, + "learning_rate": 9.987410998741101e-06, + "loss": 0.3847, + "step": 12596 + }, + { + "epoch": 5.856345885634589, + "grad_norm": 7.455138683319092, + "learning_rate": 9.987409998741002e-06, + "loss": 0.5169, + "step": 12597 + }, + { + "epoch": 5.856810785681079, + "grad_norm": 4.838932514190674, + "learning_rate": 9.9874089987409e-06, + "loss": 0.2123, + "step": 12598 + }, + { + "epoch": 5.8572756857275685, + "grad_norm": 5.818620681762695, + "learning_rate": 9.9874079987408e-06, + "loss": 0.1334, + "step": 12599 + }, + { + "epoch": 5.857740585774058, + "grad_norm": 7.348050117492676, + "learning_rate": 9.9874069987407e-06, + "loss": 0.2683, + "step": 12600 + }, + { + "epoch": 5.858205485820548, + "grad_norm": 4.133739948272705, + "learning_rate": 9.987405998740601e-06, + "loss": 0.1393, + "step": 12601 + }, + { + "epoch": 5.858670385867039, + "grad_norm": 8.078782081604004, + "learning_rate": 9.987404998740502e-06, + "loss": 0.4264, + "step": 12602 + }, + { + "epoch": 5.859135285913529, + "grad_norm": 6.583529949188232, + "learning_rate": 9.9874039987404e-06, + "loss": 0.28, + "step": 12603 + }, + { + "epoch": 5.859600185960018, + "grad_norm": 4.228565216064453, + "learning_rate": 9.987402998740301e-06, + "loss": 0.1406, + "step": 12604 + }, + { + "epoch": 5.860065086006508, + "grad_norm": 4.486139297485352, + "learning_rate": 9.9874019987402e-06, + "loss": 0.2075, + "step": 12605 + }, + { + "epoch": 5.860529986052999, + "grad_norm": 4.522080898284912, + "learning_rate": 9.987400998740101e-06, + "loss": 0.1721, + "step": 12606 + }, + { + "epoch": 5.860994886099489, + "grad_norm": 4.928760528564453, + "learning_rate": 9.98739999874e-06, + "loss": 0.2305, + "step": 12607 + }, + { + "epoch": 5.8614597861459785, + "grad_norm": 10.683208465576172, + "learning_rate": 9.9873989987399e-06, + "loss": 0.3916, + "step": 12608 + }, + { + "epoch": 5.861924686192468, + "grad_norm": 5.35800313949585, + "learning_rate": 9.987397998739801e-06, + "loss": 0.2473, + "step": 12609 + }, + { + "epoch": 5.862389586238959, + "grad_norm": 7.710094928741455, + "learning_rate": 9.9873969987397e-06, + "loss": 0.3354, + "step": 12610 + }, + { + "epoch": 5.862854486285449, + "grad_norm": 11.94559383392334, + "learning_rate": 9.9873959987396e-06, + "loss": 1.0341, + "step": 12611 + }, + { + "epoch": 5.863319386331939, + "grad_norm": 4.231446266174316, + "learning_rate": 9.9873949987395e-06, + "loss": 0.269, + "step": 12612 + }, + { + "epoch": 5.863784286378428, + "grad_norm": 5.60918664932251, + "learning_rate": 9.9873939987394e-06, + "loss": 0.3067, + "step": 12613 + }, + { + "epoch": 5.864249186424919, + "grad_norm": 11.014208793640137, + "learning_rate": 9.987392998739301e-06, + "loss": 0.2628, + "step": 12614 + }, + { + "epoch": 5.864714086471409, + "grad_norm": 3.8924834728240967, + "learning_rate": 9.987391998739202e-06, + "loss": 0.096, + "step": 12615 + }, + { + "epoch": 5.865178986517899, + "grad_norm": 7.214598178863525, + "learning_rate": 9.9873909987391e-06, + "loss": 0.6328, + "step": 12616 + }, + { + "epoch": 5.8656438865643885, + "grad_norm": 5.826150417327881, + "learning_rate": 9.987389998739001e-06, + "loss": 0.2941, + "step": 12617 + }, + { + "epoch": 5.866108786610878, + "grad_norm": 5.330164909362793, + "learning_rate": 9.9873889987389e-06, + "loss": 0.21, + "step": 12618 + }, + { + "epoch": 5.866573686657369, + "grad_norm": 6.696218013763428, + "learning_rate": 9.987387998738801e-06, + "loss": 0.1887, + "step": 12619 + }, + { + "epoch": 5.867038586703859, + "grad_norm": 6.40448522567749, + "learning_rate": 9.9873869987387e-06, + "loss": 0.2813, + "step": 12620 + }, + { + "epoch": 5.867503486750349, + "grad_norm": 5.200531482696533, + "learning_rate": 9.9873859987386e-06, + "loss": 0.2918, + "step": 12621 + }, + { + "epoch": 5.867968386796838, + "grad_norm": 5.176632404327393, + "learning_rate": 9.987384998738501e-06, + "loss": 0.1741, + "step": 12622 + }, + { + "epoch": 5.868433286843329, + "grad_norm": 3.5072457790374756, + "learning_rate": 9.9873839987384e-06, + "loss": 0.1605, + "step": 12623 + }, + { + "epoch": 5.868898186889819, + "grad_norm": 8.648165702819824, + "learning_rate": 9.9873829987383e-06, + "loss": 0.4621, + "step": 12624 + }, + { + "epoch": 5.869363086936309, + "grad_norm": 7.302768707275391, + "learning_rate": 9.9873819987382e-06, + "loss": 0.2134, + "step": 12625 + }, + { + "epoch": 5.8698279869827985, + "grad_norm": 3.1708192825317383, + "learning_rate": 9.9873809987381e-06, + "loss": 0.1193, + "step": 12626 + }, + { + "epoch": 5.870292887029288, + "grad_norm": 4.11184549331665, + "learning_rate": 9.987379998738001e-06, + "loss": 0.1516, + "step": 12627 + }, + { + "epoch": 5.870757787075779, + "grad_norm": 7.104803562164307, + "learning_rate": 9.987378998737902e-06, + "loss": 0.3894, + "step": 12628 + }, + { + "epoch": 5.871222687122269, + "grad_norm": 4.667377948760986, + "learning_rate": 9.9873779987378e-06, + "loss": 0.1809, + "step": 12629 + }, + { + "epoch": 5.871687587168759, + "grad_norm": 6.704770088195801, + "learning_rate": 9.987376998737701e-06, + "loss": 0.257, + "step": 12630 + }, + { + "epoch": 5.872152487215248, + "grad_norm": 5.237026214599609, + "learning_rate": 9.9873759987376e-06, + "loss": 0.1935, + "step": 12631 + }, + { + "epoch": 5.872617387261739, + "grad_norm": 9.277749061584473, + "learning_rate": 9.9873749987375e-06, + "loss": 0.2979, + "step": 12632 + }, + { + "epoch": 5.873082287308229, + "grad_norm": 6.438263893127441, + "learning_rate": 9.9873739987374e-06, + "loss": 0.3849, + "step": 12633 + }, + { + "epoch": 5.873547187354719, + "grad_norm": 6.547405242919922, + "learning_rate": 9.9873729987373e-06, + "loss": 0.3111, + "step": 12634 + }, + { + "epoch": 5.8740120874012085, + "grad_norm": 7.88825798034668, + "learning_rate": 9.987371998737201e-06, + "loss": 0.3306, + "step": 12635 + }, + { + "epoch": 5.874476987447698, + "grad_norm": 3.7877321243286133, + "learning_rate": 9.9873709987371e-06, + "loss": 0.1317, + "step": 12636 + }, + { + "epoch": 5.874941887494189, + "grad_norm": 5.771093845367432, + "learning_rate": 9.987369998737001e-06, + "loss": 0.2633, + "step": 12637 + }, + { + "epoch": 5.875406787540679, + "grad_norm": 4.044275283813477, + "learning_rate": 9.9873689987369e-06, + "loss": 0.1302, + "step": 12638 + }, + { + "epoch": 5.875871687587169, + "grad_norm": 7.833385944366455, + "learning_rate": 9.9873679987368e-06, + "loss": 0.3466, + "step": 12639 + }, + { + "epoch": 5.876336587633658, + "grad_norm": 5.942773818969727, + "learning_rate": 9.987366998736701e-06, + "loss": 0.1273, + "step": 12640 + }, + { + "epoch": 5.876801487680149, + "grad_norm": 3.046100378036499, + "learning_rate": 9.987365998736602e-06, + "loss": 0.0857, + "step": 12641 + }, + { + "epoch": 5.877266387726639, + "grad_norm": 5.609141826629639, + "learning_rate": 9.9873649987365e-06, + "loss": 0.318, + "step": 12642 + }, + { + "epoch": 5.877731287773129, + "grad_norm": 7.4311041831970215, + "learning_rate": 9.9873639987364e-06, + "loss": 0.2598, + "step": 12643 + }, + { + "epoch": 5.8781961878196185, + "grad_norm": 12.312759399414062, + "learning_rate": 9.9873629987363e-06, + "loss": 0.3507, + "step": 12644 + }, + { + "epoch": 5.878661087866108, + "grad_norm": 2.583751916885376, + "learning_rate": 9.987361998736201e-06, + "loss": 0.0688, + "step": 12645 + }, + { + "epoch": 5.879125987912599, + "grad_norm": 6.780377388000488, + "learning_rate": 9.9873609987361e-06, + "loss": 0.1284, + "step": 12646 + }, + { + "epoch": 5.879590887959089, + "grad_norm": 5.913645267486572, + "learning_rate": 9.987359998736e-06, + "loss": 0.396, + "step": 12647 + }, + { + "epoch": 5.880055788005579, + "grad_norm": 4.230659008026123, + "learning_rate": 9.987358998735901e-06, + "loss": 0.1367, + "step": 12648 + }, + { + "epoch": 5.880520688052069, + "grad_norm": 5.352174282073975, + "learning_rate": 9.9873579987358e-06, + "loss": 0.1251, + "step": 12649 + }, + { + "epoch": 5.880985588098559, + "grad_norm": 9.005951881408691, + "learning_rate": 9.987356998735701e-06, + "loss": 0.4538, + "step": 12650 + }, + { + "epoch": 5.881450488145049, + "grad_norm": 4.7683587074279785, + "learning_rate": 9.9873559987356e-06, + "loss": 0.2379, + "step": 12651 + }, + { + "epoch": 5.881915388191539, + "grad_norm": 4.128711700439453, + "learning_rate": 9.9873549987355e-06, + "loss": 0.0481, + "step": 12652 + }, + { + "epoch": 5.8823802882380285, + "grad_norm": 5.001165390014648, + "learning_rate": 9.987353998735401e-06, + "loss": 0.1569, + "step": 12653 + }, + { + "epoch": 5.882845188284519, + "grad_norm": 7.247454643249512, + "learning_rate": 9.9873529987353e-06, + "loss": 0.2038, + "step": 12654 + }, + { + "epoch": 5.883310088331009, + "grad_norm": 5.9474873542785645, + "learning_rate": 9.987351998735201e-06, + "loss": 0.216, + "step": 12655 + }, + { + "epoch": 5.883774988377499, + "grad_norm": 5.139963626861572, + "learning_rate": 9.9873509987351e-06, + "loss": 0.1702, + "step": 12656 + }, + { + "epoch": 5.884239888423989, + "grad_norm": 2.2400600910186768, + "learning_rate": 9.987349998735e-06, + "loss": 0.054, + "step": 12657 + }, + { + "epoch": 5.884704788470479, + "grad_norm": 4.490512371063232, + "learning_rate": 9.987348998734901e-06, + "loss": 0.2306, + "step": 12658 + }, + { + "epoch": 5.885169688516969, + "grad_norm": 4.95164155960083, + "learning_rate": 9.9873479987348e-06, + "loss": 0.2002, + "step": 12659 + }, + { + "epoch": 5.885634588563459, + "grad_norm": 5.457334518432617, + "learning_rate": 9.9873469987347e-06, + "loss": 0.2833, + "step": 12660 + }, + { + "epoch": 5.886099488609949, + "grad_norm": 6.1700310707092285, + "learning_rate": 9.987345998734601e-06, + "loss": 0.2807, + "step": 12661 + }, + { + "epoch": 5.8865643886564385, + "grad_norm": 10.742548942565918, + "learning_rate": 9.9873449987345e-06, + "loss": 0.3608, + "step": 12662 + }, + { + "epoch": 5.887029288702929, + "grad_norm": 7.934668064117432, + "learning_rate": 9.987343998734401e-06, + "loss": 0.2948, + "step": 12663 + }, + { + "epoch": 5.887494188749419, + "grad_norm": 5.154878616333008, + "learning_rate": 9.9873429987343e-06, + "loss": 0.2496, + "step": 12664 + }, + { + "epoch": 5.887959088795909, + "grad_norm": 6.828437328338623, + "learning_rate": 9.9873419987342e-06, + "loss": 0.3034, + "step": 12665 + }, + { + "epoch": 5.888423988842399, + "grad_norm": 6.412062168121338, + "learning_rate": 9.987340998734101e-06, + "loss": 0.1974, + "step": 12666 + }, + { + "epoch": 5.888888888888889, + "grad_norm": 4.704667568206787, + "learning_rate": 9.987339998734e-06, + "loss": 0.1746, + "step": 12667 + }, + { + "epoch": 5.889353788935379, + "grad_norm": 6.026299953460693, + "learning_rate": 9.987338998733901e-06, + "loss": 0.2451, + "step": 12668 + }, + { + "epoch": 5.889818688981869, + "grad_norm": 7.814128875732422, + "learning_rate": 9.9873379987338e-06, + "loss": 0.4177, + "step": 12669 + }, + { + "epoch": 5.890283589028359, + "grad_norm": 3.180972099304199, + "learning_rate": 9.9873369987337e-06, + "loss": 0.0845, + "step": 12670 + }, + { + "epoch": 5.8907484890748485, + "grad_norm": 3.0690486431121826, + "learning_rate": 9.987335998733601e-06, + "loss": 0.06, + "step": 12671 + }, + { + "epoch": 5.891213389121339, + "grad_norm": 4.842381954193115, + "learning_rate": 9.987334998733502e-06, + "loss": 0.1853, + "step": 12672 + }, + { + "epoch": 5.891678289167829, + "grad_norm": 7.683355808258057, + "learning_rate": 9.987333998733401e-06, + "loss": 0.2908, + "step": 12673 + }, + { + "epoch": 5.892143189214319, + "grad_norm": 7.589159965515137, + "learning_rate": 9.987332998733302e-06, + "loss": 0.2577, + "step": 12674 + }, + { + "epoch": 5.892608089260809, + "grad_norm": 10.032113075256348, + "learning_rate": 9.9873319987332e-06, + "loss": 0.3377, + "step": 12675 + }, + { + "epoch": 5.893072989307299, + "grad_norm": 5.135961055755615, + "learning_rate": 9.9873309987331e-06, + "loss": 0.1964, + "step": 12676 + }, + { + "epoch": 5.893537889353789, + "grad_norm": 6.292044639587402, + "learning_rate": 9.987329998733e-06, + "loss": 0.2103, + "step": 12677 + }, + { + "epoch": 5.894002789400279, + "grad_norm": 6.33248233795166, + "learning_rate": 9.9873289987329e-06, + "loss": 0.258, + "step": 12678 + }, + { + "epoch": 5.894467689446769, + "grad_norm": 9.200496673583984, + "learning_rate": 9.987327998732801e-06, + "loss": 0.7409, + "step": 12679 + }, + { + "epoch": 5.8949325894932585, + "grad_norm": 7.4486541748046875, + "learning_rate": 9.9873269987327e-06, + "loss": 0.2515, + "step": 12680 + }, + { + "epoch": 5.895397489539749, + "grad_norm": 4.34213924407959, + "learning_rate": 9.987325998732601e-06, + "loss": 0.2195, + "step": 12681 + }, + { + "epoch": 5.895862389586239, + "grad_norm": 2.907067060470581, + "learning_rate": 9.9873249987325e-06, + "loss": 0.0706, + "step": 12682 + }, + { + "epoch": 5.896327289632729, + "grad_norm": 5.547394752502441, + "learning_rate": 9.9873239987324e-06, + "loss": 0.3007, + "step": 12683 + }, + { + "epoch": 5.896792189679219, + "grad_norm": 10.145522117614746, + "learning_rate": 9.987322998732301e-06, + "loss": 0.6912, + "step": 12684 + }, + { + "epoch": 5.897257089725709, + "grad_norm": 2.412108898162842, + "learning_rate": 9.987321998732202e-06, + "loss": 0.0608, + "step": 12685 + }, + { + "epoch": 5.897721989772199, + "grad_norm": 5.037460803985596, + "learning_rate": 9.987320998732101e-06, + "loss": 0.2216, + "step": 12686 + }, + { + "epoch": 5.898186889818689, + "grad_norm": 6.747386455535889, + "learning_rate": 9.987319998732e-06, + "loss": 0.2222, + "step": 12687 + }, + { + "epoch": 5.898651789865179, + "grad_norm": 7.532002925872803, + "learning_rate": 9.9873189987319e-06, + "loss": 0.1454, + "step": 12688 + }, + { + "epoch": 5.899116689911669, + "grad_norm": 6.438199996948242, + "learning_rate": 9.9873179987318e-06, + "loss": 0.3869, + "step": 12689 + }, + { + "epoch": 5.899581589958159, + "grad_norm": 5.080563545227051, + "learning_rate": 9.9873169987317e-06, + "loss": 0.2969, + "step": 12690 + }, + { + "epoch": 5.900046490004649, + "grad_norm": 7.819587230682373, + "learning_rate": 9.987315998731601e-06, + "loss": 0.4691, + "step": 12691 + }, + { + "epoch": 5.900511390051139, + "grad_norm": 6.146832466125488, + "learning_rate": 9.987314998731502e-06, + "loss": 0.277, + "step": 12692 + }, + { + "epoch": 5.9009762900976295, + "grad_norm": 3.505028486251831, + "learning_rate": 9.9873139987314e-06, + "loss": 0.1288, + "step": 12693 + }, + { + "epoch": 5.901441190144119, + "grad_norm": 4.1165547370910645, + "learning_rate": 9.987312998731301e-06, + "loss": 0.1227, + "step": 12694 + }, + { + "epoch": 5.901906090190609, + "grad_norm": 5.157564640045166, + "learning_rate": 9.9873119987312e-06, + "loss": 0.3743, + "step": 12695 + }, + { + "epoch": 5.902370990237099, + "grad_norm": 7.153842449188232, + "learning_rate": 9.9873109987311e-06, + "loss": 0.155, + "step": 12696 + }, + { + "epoch": 5.902835890283589, + "grad_norm": 13.821111679077148, + "learning_rate": 9.987309998731001e-06, + "loss": 0.1893, + "step": 12697 + }, + { + "epoch": 5.903300790330079, + "grad_norm": 6.6426191329956055, + "learning_rate": 9.9873089987309e-06, + "loss": 0.289, + "step": 12698 + }, + { + "epoch": 5.903765690376569, + "grad_norm": 5.034901142120361, + "learning_rate": 9.987307998730801e-06, + "loss": 0.2419, + "step": 12699 + }, + { + "epoch": 5.904230590423059, + "grad_norm": 7.912660598754883, + "learning_rate": 9.9873069987307e-06, + "loss": 0.3754, + "step": 12700 + }, + { + "epoch": 5.904695490469549, + "grad_norm": 16.24079704284668, + "learning_rate": 9.9873059987306e-06, + "loss": 1.374, + "step": 12701 + }, + { + "epoch": 5.9051603905160395, + "grad_norm": 5.195338249206543, + "learning_rate": 9.9873049987305e-06, + "loss": 0.1869, + "step": 12702 + }, + { + "epoch": 5.905625290562529, + "grad_norm": 3.8254690170288086, + "learning_rate": 9.9873039987304e-06, + "loss": 0.152, + "step": 12703 + }, + { + "epoch": 5.906090190609019, + "grad_norm": 11.86341667175293, + "learning_rate": 9.987302998730301e-06, + "loss": 0.4005, + "step": 12704 + }, + { + "epoch": 5.906555090655509, + "grad_norm": 5.758669376373291, + "learning_rate": 9.987301998730202e-06, + "loss": 0.2893, + "step": 12705 + }, + { + "epoch": 5.907019990701999, + "grad_norm": 4.579347133636475, + "learning_rate": 9.9873009987301e-06, + "loss": 0.1729, + "step": 12706 + }, + { + "epoch": 5.907484890748489, + "grad_norm": 4.913050174713135, + "learning_rate": 9.987299998730001e-06, + "loss": 0.1853, + "step": 12707 + }, + { + "epoch": 5.907949790794979, + "grad_norm": 6.1275529861450195, + "learning_rate": 9.9872989987299e-06, + "loss": 0.2032, + "step": 12708 + }, + { + "epoch": 5.908414690841469, + "grad_norm": 7.428422451019287, + "learning_rate": 9.987297998729801e-06, + "loss": 0.4829, + "step": 12709 + }, + { + "epoch": 5.908879590887959, + "grad_norm": 5.243499755859375, + "learning_rate": 9.987296998729702e-06, + "loss": 0.2912, + "step": 12710 + }, + { + "epoch": 5.9093444909344495, + "grad_norm": 3.2454261779785156, + "learning_rate": 9.9872959987296e-06, + "loss": 0.085, + "step": 12711 + }, + { + "epoch": 5.909809390980939, + "grad_norm": 8.548979759216309, + "learning_rate": 9.987294998729501e-06, + "loss": 0.3335, + "step": 12712 + }, + { + "epoch": 5.910274291027429, + "grad_norm": 6.252660274505615, + "learning_rate": 9.9872939987294e-06, + "loss": 0.2597, + "step": 12713 + }, + { + "epoch": 5.910739191073919, + "grad_norm": 4.725567817687988, + "learning_rate": 9.9872929987293e-06, + "loss": 0.207, + "step": 12714 + }, + { + "epoch": 5.911204091120409, + "grad_norm": 11.003080368041992, + "learning_rate": 9.9872919987292e-06, + "loss": 0.6996, + "step": 12715 + }, + { + "epoch": 5.911668991166899, + "grad_norm": 6.989077091217041, + "learning_rate": 9.9872909987291e-06, + "loss": 0.2658, + "step": 12716 + }, + { + "epoch": 5.912133891213389, + "grad_norm": 4.758357524871826, + "learning_rate": 9.987289998729001e-06, + "loss": 0.2266, + "step": 12717 + }, + { + "epoch": 5.912598791259879, + "grad_norm": 4.845225811004639, + "learning_rate": 9.987288998728902e-06, + "loss": 0.1773, + "step": 12718 + }, + { + "epoch": 5.913063691306369, + "grad_norm": 2.6350836753845215, + "learning_rate": 9.9872879987288e-06, + "loss": 0.0536, + "step": 12719 + }, + { + "epoch": 5.9135285913528595, + "grad_norm": 3.9353983402252197, + "learning_rate": 9.9872869987287e-06, + "loss": 0.1103, + "step": 12720 + }, + { + "epoch": 5.913993491399349, + "grad_norm": 9.108762741088867, + "learning_rate": 9.9872859987286e-06, + "loss": 0.5288, + "step": 12721 + }, + { + "epoch": 5.914458391445839, + "grad_norm": 5.212458610534668, + "learning_rate": 9.987284998728501e-06, + "loss": 0.2036, + "step": 12722 + }, + { + "epoch": 5.914923291492329, + "grad_norm": 4.911321640014648, + "learning_rate": 9.987283998728402e-06, + "loss": 0.2342, + "step": 12723 + }, + { + "epoch": 5.915388191538819, + "grad_norm": 4.83859920501709, + "learning_rate": 9.9872829987283e-06, + "loss": 0.2576, + "step": 12724 + }, + { + "epoch": 5.915853091585309, + "grad_norm": 6.436941623687744, + "learning_rate": 9.987281998728201e-06, + "loss": 0.3546, + "step": 12725 + }, + { + "epoch": 5.916317991631799, + "grad_norm": 4.605725288391113, + "learning_rate": 9.9872809987281e-06, + "loss": 0.1233, + "step": 12726 + }, + { + "epoch": 5.916782891678289, + "grad_norm": 5.32492208480835, + "learning_rate": 9.987279998728e-06, + "loss": 0.2333, + "step": 12727 + }, + { + "epoch": 5.91724779172478, + "grad_norm": 7.641358852386475, + "learning_rate": 9.9872789987279e-06, + "loss": 0.3779, + "step": 12728 + }, + { + "epoch": 5.9177126917712695, + "grad_norm": 5.4877519607543945, + "learning_rate": 9.9872779987278e-06, + "loss": 0.2467, + "step": 12729 + }, + { + "epoch": 5.918177591817759, + "grad_norm": 3.9627368450164795, + "learning_rate": 9.987276998727701e-06, + "loss": 0.0906, + "step": 12730 + }, + { + "epoch": 5.918642491864249, + "grad_norm": 6.842897891998291, + "learning_rate": 9.9872759987276e-06, + "loss": 0.377, + "step": 12731 + }, + { + "epoch": 5.919107391910739, + "grad_norm": 6.0924835205078125, + "learning_rate": 9.9872749987275e-06, + "loss": 0.3007, + "step": 12732 + }, + { + "epoch": 5.91957229195723, + "grad_norm": 5.99074649810791, + "learning_rate": 9.9872739987274e-06, + "loss": 0.2829, + "step": 12733 + }, + { + "epoch": 5.920037192003719, + "grad_norm": 11.582792282104492, + "learning_rate": 9.9872729987273e-06, + "loss": 0.5009, + "step": 12734 + }, + { + "epoch": 5.920502092050209, + "grad_norm": 7.622169017791748, + "learning_rate": 9.987271998727201e-06, + "loss": 0.3242, + "step": 12735 + }, + { + "epoch": 5.920966992096699, + "grad_norm": 3.5286808013916016, + "learning_rate": 9.987270998727102e-06, + "loss": 0.0935, + "step": 12736 + }, + { + "epoch": 5.92143189214319, + "grad_norm": 7.087674617767334, + "learning_rate": 9.987269998727e-06, + "loss": 0.3264, + "step": 12737 + }, + { + "epoch": 5.9218967921896795, + "grad_norm": 3.5139272212982178, + "learning_rate": 9.987268998726901e-06, + "loss": 0.0991, + "step": 12738 + }, + { + "epoch": 5.922361692236169, + "grad_norm": 5.642136096954346, + "learning_rate": 9.9872679987268e-06, + "loss": 0.3583, + "step": 12739 + }, + { + "epoch": 5.922826592282659, + "grad_norm": 13.110286712646484, + "learning_rate": 9.987266998726701e-06, + "loss": 0.264, + "step": 12740 + }, + { + "epoch": 5.923291492329149, + "grad_norm": 5.50581693649292, + "learning_rate": 9.987265998726602e-06, + "loss": 0.2403, + "step": 12741 + }, + { + "epoch": 5.92375639237564, + "grad_norm": 4.947712421417236, + "learning_rate": 9.9872649987265e-06, + "loss": 0.3002, + "step": 12742 + }, + { + "epoch": 5.924221292422129, + "grad_norm": 6.684628486633301, + "learning_rate": 9.987263998726401e-06, + "loss": 0.1996, + "step": 12743 + }, + { + "epoch": 5.924686192468619, + "grad_norm": 3.8016116619110107, + "learning_rate": 9.9872629987263e-06, + "loss": 0.0909, + "step": 12744 + }, + { + "epoch": 5.925151092515109, + "grad_norm": 4.188547134399414, + "learning_rate": 9.9872619987262e-06, + "loss": 0.1718, + "step": 12745 + }, + { + "epoch": 5.9256159925616, + "grad_norm": 5.448657512664795, + "learning_rate": 9.9872609987261e-06, + "loss": 0.1438, + "step": 12746 + }, + { + "epoch": 5.9260808926080895, + "grad_norm": 6.33152437210083, + "learning_rate": 9.987259998726e-06, + "loss": 0.2862, + "step": 12747 + }, + { + "epoch": 5.926545792654579, + "grad_norm": 5.433250427246094, + "learning_rate": 9.987258998725901e-06, + "loss": 0.226, + "step": 12748 + }, + { + "epoch": 5.927010692701069, + "grad_norm": 4.892200946807861, + "learning_rate": 9.987257998725802e-06, + "loss": 0.2969, + "step": 12749 + }, + { + "epoch": 5.927475592747559, + "grad_norm": 4.591060638427734, + "learning_rate": 9.9872569987257e-06, + "loss": 0.1636, + "step": 12750 + }, + { + "epoch": 5.9279404927940496, + "grad_norm": 6.068577766418457, + "learning_rate": 9.9872559987256e-06, + "loss": 0.2916, + "step": 12751 + }, + { + "epoch": 5.928405392840539, + "grad_norm": 3.7788074016571045, + "learning_rate": 9.9872549987255e-06, + "loss": 0.1583, + "step": 12752 + }, + { + "epoch": 5.928870292887029, + "grad_norm": 6.9061689376831055, + "learning_rate": 9.987253998725401e-06, + "loss": 0.2922, + "step": 12753 + }, + { + "epoch": 5.929335192933519, + "grad_norm": 5.583242893218994, + "learning_rate": 9.987252998725302e-06, + "loss": 0.1699, + "step": 12754 + }, + { + "epoch": 5.92980009298001, + "grad_norm": 3.52844500541687, + "learning_rate": 9.9872519987252e-06, + "loss": 0.1009, + "step": 12755 + }, + { + "epoch": 5.9302649930264995, + "grad_norm": 8.949033737182617, + "learning_rate": 9.987250998725101e-06, + "loss": 0.7229, + "step": 12756 + }, + { + "epoch": 5.930729893072989, + "grad_norm": 5.041260719299316, + "learning_rate": 9.987249998725e-06, + "loss": 0.2032, + "step": 12757 + }, + { + "epoch": 5.931194793119479, + "grad_norm": 3.826937675476074, + "learning_rate": 9.987248998724901e-06, + "loss": 0.2043, + "step": 12758 + }, + { + "epoch": 5.931659693165969, + "grad_norm": 2.836691379547119, + "learning_rate": 9.9872479987248e-06, + "loss": 0.0716, + "step": 12759 + }, + { + "epoch": 5.9321245932124596, + "grad_norm": 4.411704063415527, + "learning_rate": 9.9872469987247e-06, + "loss": 0.2714, + "step": 12760 + }, + { + "epoch": 5.932589493258949, + "grad_norm": 6.082199573516846, + "learning_rate": 9.987245998724601e-06, + "loss": 0.2225, + "step": 12761 + }, + { + "epoch": 5.933054393305439, + "grad_norm": 5.675535202026367, + "learning_rate": 9.9872449987245e-06, + "loss": 0.2834, + "step": 12762 + }, + { + "epoch": 5.933519293351929, + "grad_norm": 3.581169605255127, + "learning_rate": 9.9872439987244e-06, + "loss": 0.1374, + "step": 12763 + }, + { + "epoch": 5.93398419339842, + "grad_norm": 8.111862182617188, + "learning_rate": 9.9872429987243e-06, + "loss": 0.3425, + "step": 12764 + }, + { + "epoch": 5.9344490934449095, + "grad_norm": 6.470550537109375, + "learning_rate": 9.9872419987242e-06, + "loss": 0.1777, + "step": 12765 + }, + { + "epoch": 5.934913993491399, + "grad_norm": 4.889643669128418, + "learning_rate": 9.987240998724101e-06, + "loss": 0.1454, + "step": 12766 + }, + { + "epoch": 5.935378893537889, + "grad_norm": 4.630599498748779, + "learning_rate": 9.987239998724002e-06, + "loss": 0.2864, + "step": 12767 + }, + { + "epoch": 5.93584379358438, + "grad_norm": 3.3460817337036133, + "learning_rate": 9.9872389987239e-06, + "loss": 0.0562, + "step": 12768 + }, + { + "epoch": 5.9363086936308695, + "grad_norm": 5.313773155212402, + "learning_rate": 9.987237998723801e-06, + "loss": 0.2885, + "step": 12769 + }, + { + "epoch": 5.936773593677359, + "grad_norm": 5.242966651916504, + "learning_rate": 9.9872369987237e-06, + "loss": 0.1993, + "step": 12770 + }, + { + "epoch": 5.937238493723849, + "grad_norm": 4.876601219177246, + "learning_rate": 9.987235998723601e-06, + "loss": 0.2037, + "step": 12771 + }, + { + "epoch": 5.93770339377034, + "grad_norm": 6.895362854003906, + "learning_rate": 9.9872349987235e-06, + "loss": 0.3198, + "step": 12772 + }, + { + "epoch": 5.93816829381683, + "grad_norm": 5.259251594543457, + "learning_rate": 9.9872339987234e-06, + "loss": 0.2264, + "step": 12773 + }, + { + "epoch": 5.9386331938633194, + "grad_norm": 6.608183860778809, + "learning_rate": 9.987232998723301e-06, + "loss": 0.2559, + "step": 12774 + }, + { + "epoch": 5.939098093909809, + "grad_norm": 6.150241851806641, + "learning_rate": 9.9872319987232e-06, + "loss": 0.4305, + "step": 12775 + }, + { + "epoch": 5.939562993956299, + "grad_norm": 4.673802852630615, + "learning_rate": 9.987230998723101e-06, + "loss": 0.1443, + "step": 12776 + }, + { + "epoch": 5.94002789400279, + "grad_norm": 6.867313385009766, + "learning_rate": 9.987229998723e-06, + "loss": 0.2342, + "step": 12777 + }, + { + "epoch": 5.9404927940492795, + "grad_norm": 6.415299892425537, + "learning_rate": 9.9872289987229e-06, + "loss": 0.1134, + "step": 12778 + }, + { + "epoch": 5.940957694095769, + "grad_norm": 4.778309345245361, + "learning_rate": 9.987227998722801e-06, + "loss": 0.2017, + "step": 12779 + }, + { + "epoch": 5.941422594142259, + "grad_norm": 5.24874210357666, + "learning_rate": 9.987226998722702e-06, + "loss": 0.2459, + "step": 12780 + }, + { + "epoch": 5.94188749418875, + "grad_norm": 6.3668036460876465, + "learning_rate": 9.9872259987226e-06, + "loss": 0.4435, + "step": 12781 + }, + { + "epoch": 5.94235239423524, + "grad_norm": 5.350764274597168, + "learning_rate": 9.987224998722501e-06, + "loss": 0.2328, + "step": 12782 + }, + { + "epoch": 5.9428172942817294, + "grad_norm": 8.79177474975586, + "learning_rate": 9.9872239987224e-06, + "loss": 0.5497, + "step": 12783 + }, + { + "epoch": 5.943282194328219, + "grad_norm": 5.25726842880249, + "learning_rate": 9.9872229987223e-06, + "loss": 0.2595, + "step": 12784 + }, + { + "epoch": 5.943747094374709, + "grad_norm": 11.975218772888184, + "learning_rate": 9.9872219987222e-06, + "loss": 0.523, + "step": 12785 + }, + { + "epoch": 5.9442119944212, + "grad_norm": 4.904669284820557, + "learning_rate": 9.9872209987221e-06, + "loss": 0.1449, + "step": 12786 + }, + { + "epoch": 5.9446768944676895, + "grad_norm": 9.070215225219727, + "learning_rate": 9.987219998722001e-06, + "loss": 0.4041, + "step": 12787 + }, + { + "epoch": 5.945141794514179, + "grad_norm": 5.983394622802734, + "learning_rate": 9.9872189987219e-06, + "loss": 0.1921, + "step": 12788 + }, + { + "epoch": 5.945606694560669, + "grad_norm": 5.319808006286621, + "learning_rate": 9.987217998721801e-06, + "loss": 0.2775, + "step": 12789 + }, + { + "epoch": 5.94607159460716, + "grad_norm": 6.1101908683776855, + "learning_rate": 9.9872169987217e-06, + "loss": 0.2233, + "step": 12790 + }, + { + "epoch": 5.94653649465365, + "grad_norm": 4.888025283813477, + "learning_rate": 9.9872159987216e-06, + "loss": 0.159, + "step": 12791 + }, + { + "epoch": 5.947001394700139, + "grad_norm": 4.974210739135742, + "learning_rate": 9.987214998721501e-06, + "loss": 0.1706, + "step": 12792 + }, + { + "epoch": 5.947466294746629, + "grad_norm": 6.326246738433838, + "learning_rate": 9.987213998721402e-06, + "loss": 0.2803, + "step": 12793 + }, + { + "epoch": 5.947931194793119, + "grad_norm": 6.785887241363525, + "learning_rate": 9.987212998721301e-06, + "loss": 0.6383, + "step": 12794 + }, + { + "epoch": 5.94839609483961, + "grad_norm": 6.503730773925781, + "learning_rate": 9.9872119987212e-06, + "loss": 0.2771, + "step": 12795 + }, + { + "epoch": 5.9488609948860995, + "grad_norm": 6.642497539520264, + "learning_rate": 9.9872109987211e-06, + "loss": 0.3559, + "step": 12796 + }, + { + "epoch": 5.949325894932589, + "grad_norm": 3.9172470569610596, + "learning_rate": 9.987209998721e-06, + "loss": 0.2331, + "step": 12797 + }, + { + "epoch": 5.949790794979079, + "grad_norm": 10.16185474395752, + "learning_rate": 9.9872089987209e-06, + "loss": 0.1438, + "step": 12798 + }, + { + "epoch": 5.95025569502557, + "grad_norm": 5.951796054840088, + "learning_rate": 9.9872079987208e-06, + "loss": 0.2541, + "step": 12799 + }, + { + "epoch": 5.95072059507206, + "grad_norm": 4.339954853057861, + "learning_rate": 9.987206998720701e-06, + "loss": 0.2301, + "step": 12800 + }, + { + "epoch": 5.951185495118549, + "grad_norm": 4.982366561889648, + "learning_rate": 9.9872059987206e-06, + "loss": 0.3145, + "step": 12801 + }, + { + "epoch": 5.951650395165039, + "grad_norm": 13.058465957641602, + "learning_rate": 9.987204998720501e-06, + "loss": 0.1804, + "step": 12802 + }, + { + "epoch": 5.952115295211529, + "grad_norm": 5.107154369354248, + "learning_rate": 9.9872039987204e-06, + "loss": 0.3292, + "step": 12803 + }, + { + "epoch": 5.95258019525802, + "grad_norm": 4.228498935699463, + "learning_rate": 9.9872029987203e-06, + "loss": 0.1717, + "step": 12804 + }, + { + "epoch": 5.9530450953045095, + "grad_norm": 8.985190391540527, + "learning_rate": 9.987201998720201e-06, + "loss": 0.4642, + "step": 12805 + }, + { + "epoch": 5.953509995350999, + "grad_norm": 4.039201736450195, + "learning_rate": 9.9872009987201e-06, + "loss": 0.0989, + "step": 12806 + }, + { + "epoch": 5.95397489539749, + "grad_norm": 5.4272379875183105, + "learning_rate": 9.987199998720001e-06, + "loss": 0.3381, + "step": 12807 + }, + { + "epoch": 5.95443979544398, + "grad_norm": 3.3291752338409424, + "learning_rate": 9.9871989987199e-06, + "loss": 0.1286, + "step": 12808 + }, + { + "epoch": 5.95490469549047, + "grad_norm": 2.7388153076171875, + "learning_rate": 9.9871979987198e-06, + "loss": 0.087, + "step": 12809 + }, + { + "epoch": 5.955369595536959, + "grad_norm": 6.267618656158447, + "learning_rate": 9.987196998719701e-06, + "loss": 0.232, + "step": 12810 + }, + { + "epoch": 5.955834495583449, + "grad_norm": 7.113882064819336, + "learning_rate": 9.9871959987196e-06, + "loss": 0.3524, + "step": 12811 + }, + { + "epoch": 5.95629939562994, + "grad_norm": 4.745032787322998, + "learning_rate": 9.9871949987195e-06, + "loss": 0.1759, + "step": 12812 + }, + { + "epoch": 5.95676429567643, + "grad_norm": 5.255289077758789, + "learning_rate": 9.987193998719401e-06, + "loss": 0.1325, + "step": 12813 + }, + { + "epoch": 5.9572291957229195, + "grad_norm": 5.494358539581299, + "learning_rate": 9.9871929987193e-06, + "loss": 0.1467, + "step": 12814 + }, + { + "epoch": 5.957694095769409, + "grad_norm": 4.412227153778076, + "learning_rate": 9.987191998719201e-06, + "loss": 0.2037, + "step": 12815 + }, + { + "epoch": 5.9581589958159, + "grad_norm": 4.549781322479248, + "learning_rate": 9.9871909987191e-06, + "loss": 0.1672, + "step": 12816 + }, + { + "epoch": 5.95862389586239, + "grad_norm": 8.85688304901123, + "learning_rate": 9.987189998719e-06, + "loss": 0.5148, + "step": 12817 + }, + { + "epoch": 5.95908879590888, + "grad_norm": 6.992033958435059, + "learning_rate": 9.987188998718901e-06, + "loss": 0.548, + "step": 12818 + }, + { + "epoch": 5.959553695955369, + "grad_norm": 3.588487148284912, + "learning_rate": 9.9871879987188e-06, + "loss": 0.1606, + "step": 12819 + }, + { + "epoch": 5.960018596001859, + "grad_norm": 3.5528337955474854, + "learning_rate": 9.987186998718701e-06, + "loss": 0.0841, + "step": 12820 + }, + { + "epoch": 5.96048349604835, + "grad_norm": 5.812241554260254, + "learning_rate": 9.9871859987186e-06, + "loss": 0.2273, + "step": 12821 + }, + { + "epoch": 5.96094839609484, + "grad_norm": 7.5692901611328125, + "learning_rate": 9.9871849987185e-06, + "loss": 0.4339, + "step": 12822 + }, + { + "epoch": 5.9614132961413295, + "grad_norm": 5.511832237243652, + "learning_rate": 9.987183998718401e-06, + "loss": 0.161, + "step": 12823 + }, + { + "epoch": 5.961878196187819, + "grad_norm": 2.540522575378418, + "learning_rate": 9.9871829987183e-06, + "loss": 0.0667, + "step": 12824 + }, + { + "epoch": 5.96234309623431, + "grad_norm": 3.483137607574463, + "learning_rate": 9.987181998718201e-06, + "loss": 0.1342, + "step": 12825 + }, + { + "epoch": 5.9628079962808, + "grad_norm": 5.173486232757568, + "learning_rate": 9.987180998718102e-06, + "loss": 0.2146, + "step": 12826 + }, + { + "epoch": 5.96327289632729, + "grad_norm": 8.173789978027344, + "learning_rate": 9.987179998718e-06, + "loss": 0.3375, + "step": 12827 + }, + { + "epoch": 5.963737796373779, + "grad_norm": 4.468027114868164, + "learning_rate": 9.9871789987179e-06, + "loss": 0.1911, + "step": 12828 + }, + { + "epoch": 5.964202696420269, + "grad_norm": 3.425257444381714, + "learning_rate": 9.9871779987178e-06, + "loss": 0.1789, + "step": 12829 + }, + { + "epoch": 5.96466759646676, + "grad_norm": 6.417053699493408, + "learning_rate": 9.9871769987177e-06, + "loss": 0.2108, + "step": 12830 + }, + { + "epoch": 5.96513249651325, + "grad_norm": 4.080670356750488, + "learning_rate": 9.987175998717601e-06, + "loss": 0.168, + "step": 12831 + }, + { + "epoch": 5.9655973965597395, + "grad_norm": 4.8072333335876465, + "learning_rate": 9.9871749987175e-06, + "loss": 0.3293, + "step": 12832 + }, + { + "epoch": 5.966062296606229, + "grad_norm": 3.3861727714538574, + "learning_rate": 9.987173998717401e-06, + "loss": 0.1037, + "step": 12833 + }, + { + "epoch": 5.96652719665272, + "grad_norm": 6.346316337585449, + "learning_rate": 9.9871729987173e-06, + "loss": 0.344, + "step": 12834 + }, + { + "epoch": 5.96699209669921, + "grad_norm": 7.278039455413818, + "learning_rate": 9.9871719987172e-06, + "loss": 0.3134, + "step": 12835 + }, + { + "epoch": 5.9674569967457, + "grad_norm": 6.249755382537842, + "learning_rate": 9.987170998717101e-06, + "loss": 0.4414, + "step": 12836 + }, + { + "epoch": 5.967921896792189, + "grad_norm": 5.578820705413818, + "learning_rate": 9.987169998717002e-06, + "loss": 0.2226, + "step": 12837 + }, + { + "epoch": 5.968386796838679, + "grad_norm": 5.440159320831299, + "learning_rate": 9.987168998716901e-06, + "loss": 0.2675, + "step": 12838 + }, + { + "epoch": 5.96885169688517, + "grad_norm": 10.134894371032715, + "learning_rate": 9.9871679987168e-06, + "loss": 0.4018, + "step": 12839 + }, + { + "epoch": 5.96931659693166, + "grad_norm": 4.0183868408203125, + "learning_rate": 9.9871669987167e-06, + "loss": 0.1332, + "step": 12840 + }, + { + "epoch": 5.9697814969781495, + "grad_norm": 3.7357053756713867, + "learning_rate": 9.9871659987166e-06, + "loss": 0.1077, + "step": 12841 + }, + { + "epoch": 5.970246397024639, + "grad_norm": 5.913174152374268, + "learning_rate": 9.9871649987165e-06, + "loss": 0.1803, + "step": 12842 + }, + { + "epoch": 5.97071129707113, + "grad_norm": 4.274318695068359, + "learning_rate": 9.987163998716401e-06, + "loss": 0.1597, + "step": 12843 + }, + { + "epoch": 5.97117619711762, + "grad_norm": 7.548904895782471, + "learning_rate": 9.987162998716302e-06, + "loss": 0.4864, + "step": 12844 + }, + { + "epoch": 5.97164109716411, + "grad_norm": 4.479726314544678, + "learning_rate": 9.9871619987162e-06, + "loss": 0.2493, + "step": 12845 + }, + { + "epoch": 5.972105997210599, + "grad_norm": 7.3963236808776855, + "learning_rate": 9.987160998716101e-06, + "loss": 0.1929, + "step": 12846 + }, + { + "epoch": 5.972570897257089, + "grad_norm": 5.662945747375488, + "learning_rate": 9.987159998716e-06, + "loss": 0.1597, + "step": 12847 + }, + { + "epoch": 5.97303579730358, + "grad_norm": 4.727105617523193, + "learning_rate": 9.9871589987159e-06, + "loss": 0.1485, + "step": 12848 + }, + { + "epoch": 5.97350069735007, + "grad_norm": 3.2275474071502686, + "learning_rate": 9.987157998715801e-06, + "loss": 0.0808, + "step": 12849 + }, + { + "epoch": 5.9739655973965595, + "grad_norm": 4.105709075927734, + "learning_rate": 9.9871569987157e-06, + "loss": 0.1065, + "step": 12850 + }, + { + "epoch": 5.97443049744305, + "grad_norm": 4.220144271850586, + "learning_rate": 9.987155998715601e-06, + "loss": 0.1377, + "step": 12851 + }, + { + "epoch": 5.97489539748954, + "grad_norm": 4.755367755889893, + "learning_rate": 9.9871549987155e-06, + "loss": 0.185, + "step": 12852 + }, + { + "epoch": 5.97536029753603, + "grad_norm": 2.9607319831848145, + "learning_rate": 9.9871539987154e-06, + "loss": 0.0806, + "step": 12853 + }, + { + "epoch": 5.97582519758252, + "grad_norm": 5.5143208503723145, + "learning_rate": 9.9871529987153e-06, + "loss": 0.2586, + "step": 12854 + }, + { + "epoch": 5.976290097629009, + "grad_norm": 3.5573065280914307, + "learning_rate": 9.9871519987152e-06, + "loss": 0.1082, + "step": 12855 + }, + { + "epoch": 5.9767549976755, + "grad_norm": 6.653648853302002, + "learning_rate": 9.987150998715101e-06, + "loss": 0.2372, + "step": 12856 + }, + { + "epoch": 5.97721989772199, + "grad_norm": 2.8944571018218994, + "learning_rate": 9.987149998715002e-06, + "loss": 0.0698, + "step": 12857 + }, + { + "epoch": 5.97768479776848, + "grad_norm": 13.191024780273438, + "learning_rate": 9.9871489987149e-06, + "loss": 0.4314, + "step": 12858 + }, + { + "epoch": 5.9781496978149695, + "grad_norm": 4.242642879486084, + "learning_rate": 9.987147998714801e-06, + "loss": 0.1102, + "step": 12859 + }, + { + "epoch": 5.97861459786146, + "grad_norm": 5.786744594573975, + "learning_rate": 9.9871469987147e-06, + "loss": 0.2591, + "step": 12860 + }, + { + "epoch": 5.97907949790795, + "grad_norm": 9.730655670166016, + "learning_rate": 9.987145998714601e-06, + "loss": 0.3805, + "step": 12861 + }, + { + "epoch": 5.97954439795444, + "grad_norm": 7.153598785400391, + "learning_rate": 9.987144998714502e-06, + "loss": 0.3273, + "step": 12862 + }, + { + "epoch": 5.98000929800093, + "grad_norm": 6.6354522705078125, + "learning_rate": 9.9871439987144e-06, + "loss": 0.4131, + "step": 12863 + }, + { + "epoch": 5.980474198047419, + "grad_norm": 7.921511173248291, + "learning_rate": 9.987142998714301e-06, + "loss": 0.4063, + "step": 12864 + }, + { + "epoch": 5.98093909809391, + "grad_norm": 5.790859699249268, + "learning_rate": 9.9871419987142e-06, + "loss": 0.3345, + "step": 12865 + }, + { + "epoch": 5.9814039981404, + "grad_norm": 4.224299430847168, + "learning_rate": 9.9871409987141e-06, + "loss": 0.187, + "step": 12866 + }, + { + "epoch": 5.98186889818689, + "grad_norm": 3.5347049236297607, + "learning_rate": 9.987139998714e-06, + "loss": 0.1227, + "step": 12867 + }, + { + "epoch": 5.9823337982333795, + "grad_norm": 8.330761909484863, + "learning_rate": 9.9871389987139e-06, + "loss": 0.92, + "step": 12868 + }, + { + "epoch": 5.98279869827987, + "grad_norm": 11.20364761352539, + "learning_rate": 9.987137998713801e-06, + "loss": 0.5036, + "step": 12869 + }, + { + "epoch": 5.98326359832636, + "grad_norm": 6.993566036224365, + "learning_rate": 9.987136998713702e-06, + "loss": 0.3174, + "step": 12870 + }, + { + "epoch": 5.98372849837285, + "grad_norm": 3.635791063308716, + "learning_rate": 9.9871359987136e-06, + "loss": 0.0658, + "step": 12871 + }, + { + "epoch": 5.98419339841934, + "grad_norm": 9.118582725524902, + "learning_rate": 9.9871349987135e-06, + "loss": 0.2737, + "step": 12872 + }, + { + "epoch": 5.984658298465829, + "grad_norm": 3.9937260150909424, + "learning_rate": 9.9871339987134e-06, + "loss": 0.0955, + "step": 12873 + }, + { + "epoch": 5.98512319851232, + "grad_norm": 3.102027416229248, + "learning_rate": 9.987132998713301e-06, + "loss": 0.0818, + "step": 12874 + }, + { + "epoch": 5.98558809855881, + "grad_norm": 4.71635627746582, + "learning_rate": 9.987131998713202e-06, + "loss": 0.1477, + "step": 12875 + }, + { + "epoch": 5.9860529986053, + "grad_norm": 5.062005043029785, + "learning_rate": 9.9871309987131e-06, + "loss": 0.3564, + "step": 12876 + }, + { + "epoch": 5.9865178986517895, + "grad_norm": 9.045232772827148, + "learning_rate": 9.987129998713001e-06, + "loss": 0.4678, + "step": 12877 + }, + { + "epoch": 5.98698279869828, + "grad_norm": 4.179198265075684, + "learning_rate": 9.9871289987129e-06, + "loss": 0.1606, + "step": 12878 + }, + { + "epoch": 5.98744769874477, + "grad_norm": 5.782346725463867, + "learning_rate": 9.987127998712801e-06, + "loss": 0.2995, + "step": 12879 + }, + { + "epoch": 5.98791259879126, + "grad_norm": 5.717284202575684, + "learning_rate": 9.9871269987127e-06, + "loss": 0.2957, + "step": 12880 + }, + { + "epoch": 5.98837749883775, + "grad_norm": 4.299577236175537, + "learning_rate": 9.9871259987126e-06, + "loss": 0.1459, + "step": 12881 + }, + { + "epoch": 5.988842398884239, + "grad_norm": 4.682750701904297, + "learning_rate": 9.987124998712501e-06, + "loss": 0.1871, + "step": 12882 + }, + { + "epoch": 5.98930729893073, + "grad_norm": 3.4867448806762695, + "learning_rate": 9.9871239987124e-06, + "loss": 0.2141, + "step": 12883 + }, + { + "epoch": 5.98977219897722, + "grad_norm": 3.983729839324951, + "learning_rate": 9.9871229987123e-06, + "loss": 0.1932, + "step": 12884 + }, + { + "epoch": 5.99023709902371, + "grad_norm": 4.43710470199585, + "learning_rate": 9.9871219987122e-06, + "loss": 0.2144, + "step": 12885 + }, + { + "epoch": 5.9907019990702, + "grad_norm": 7.976717472076416, + "learning_rate": 9.9871209987121e-06, + "loss": 0.3497, + "step": 12886 + }, + { + "epoch": 5.99116689911669, + "grad_norm": 4.801514148712158, + "learning_rate": 9.987119998712001e-06, + "loss": 0.2325, + "step": 12887 + }, + { + "epoch": 5.99163179916318, + "grad_norm": 5.840821266174316, + "learning_rate": 9.987118998711902e-06, + "loss": 0.1042, + "step": 12888 + }, + { + "epoch": 5.99209669920967, + "grad_norm": 2.815063238143921, + "learning_rate": 9.9871179987118e-06, + "loss": 0.0769, + "step": 12889 + }, + { + "epoch": 5.99256159925616, + "grad_norm": 3.8412954807281494, + "learning_rate": 9.987116998711701e-06, + "loss": 0.1136, + "step": 12890 + }, + { + "epoch": 5.99302649930265, + "grad_norm": 6.379650592803955, + "learning_rate": 9.9871159987116e-06, + "loss": 0.3865, + "step": 12891 + }, + { + "epoch": 5.99349139934914, + "grad_norm": 4.482799530029297, + "learning_rate": 9.987114998711501e-06, + "loss": 0.0994, + "step": 12892 + }, + { + "epoch": 5.99395629939563, + "grad_norm": 4.308323860168457, + "learning_rate": 9.9871139987114e-06, + "loss": 0.1552, + "step": 12893 + }, + { + "epoch": 5.99442119944212, + "grad_norm": 5.539119720458984, + "learning_rate": 9.9871129987113e-06, + "loss": 0.083, + "step": 12894 + }, + { + "epoch": 5.99488609948861, + "grad_norm": 5.106306552886963, + "learning_rate": 9.987111998711201e-06, + "loss": 0.3104, + "step": 12895 + }, + { + "epoch": 5.9953509995351, + "grad_norm": 4.5551958084106445, + "learning_rate": 9.9871109987111e-06, + "loss": 0.2104, + "step": 12896 + }, + { + "epoch": 5.99581589958159, + "grad_norm": 3.9699926376342773, + "learning_rate": 9.987109998711e-06, + "loss": 0.147, + "step": 12897 + }, + { + "epoch": 5.99628079962808, + "grad_norm": 4.878609657287598, + "learning_rate": 9.9871089987109e-06, + "loss": 0.1893, + "step": 12898 + }, + { + "epoch": 5.99674569967457, + "grad_norm": 3.1628026962280273, + "learning_rate": 9.9871079987108e-06, + "loss": 0.0746, + "step": 12899 + }, + { + "epoch": 5.99721059972106, + "grad_norm": 4.731689929962158, + "learning_rate": 9.987106998710701e-06, + "loss": 0.1714, + "step": 12900 + }, + { + "epoch": 5.99767549976755, + "grad_norm": 3.894443988800049, + "learning_rate": 9.987105998710602e-06, + "loss": 0.1273, + "step": 12901 + }, + { + "epoch": 5.99814039981404, + "grad_norm": 5.64202356338501, + "learning_rate": 9.9871049987105e-06, + "loss": 0.2854, + "step": 12902 + }, + { + "epoch": 5.99860529986053, + "grad_norm": 6.301222324371338, + "learning_rate": 9.9871039987104e-06, + "loss": 0.182, + "step": 12903 + }, + { + "epoch": 5.99907019990702, + "grad_norm": 7.222960948944092, + "learning_rate": 9.9871029987103e-06, + "loss": 0.4143, + "step": 12904 + }, + { + "epoch": 5.99953509995351, + "grad_norm": 4.457801342010498, + "learning_rate": 9.987101998710201e-06, + "loss": 0.2649, + "step": 12905 + }, + { + "epoch": 6.0, + "grad_norm": 7.3542585372924805, + "learning_rate": 9.987100998710102e-06, + "loss": 0.1294, + "step": 12906 + }, + { + "epoch": 6.00046490004649, + "grad_norm": 5.543666362762451, + "learning_rate": 9.98709999871e-06, + "loss": 0.1546, + "step": 12907 + }, + { + "epoch": 6.00092980009298, + "grad_norm": 4.2462239265441895, + "learning_rate": 9.987098998709901e-06, + "loss": 0.2136, + "step": 12908 + }, + { + "epoch": 6.00139470013947, + "grad_norm": 6.742617130279541, + "learning_rate": 9.9870979987098e-06, + "loss": 0.319, + "step": 12909 + }, + { + "epoch": 6.00185960018596, + "grad_norm": 4.132018089294434, + "learning_rate": 9.987096998709701e-06, + "loss": 0.1245, + "step": 12910 + }, + { + "epoch": 6.00232450023245, + "grad_norm": 5.015681743621826, + "learning_rate": 9.9870959987096e-06, + "loss": 0.1286, + "step": 12911 + }, + { + "epoch": 6.00278940027894, + "grad_norm": 3.8832240104675293, + "learning_rate": 9.9870949987095e-06, + "loss": 0.1471, + "step": 12912 + }, + { + "epoch": 6.00325430032543, + "grad_norm": 5.405463218688965, + "learning_rate": 9.987093998709401e-06, + "loss": 0.2027, + "step": 12913 + }, + { + "epoch": 6.00371920037192, + "grad_norm": 5.107647895812988, + "learning_rate": 9.9870929987093e-06, + "loss": 0.1635, + "step": 12914 + }, + { + "epoch": 6.00418410041841, + "grad_norm": 4.3904314041137695, + "learning_rate": 9.9870919987092e-06, + "loss": 0.0616, + "step": 12915 + }, + { + "epoch": 6.0046490004649, + "grad_norm": 4.7297515869140625, + "learning_rate": 9.9870909987091e-06, + "loss": 0.0954, + "step": 12916 + }, + { + "epoch": 6.0051139005113905, + "grad_norm": 4.0568013191223145, + "learning_rate": 9.987089998709e-06, + "loss": 0.1069, + "step": 12917 + }, + { + "epoch": 6.00557880055788, + "grad_norm": 5.043839931488037, + "learning_rate": 9.987088998708901e-06, + "loss": 0.1971, + "step": 12918 + }, + { + "epoch": 6.00604370060437, + "grad_norm": 2.8936808109283447, + "learning_rate": 9.987087998708802e-06, + "loss": 0.0875, + "step": 12919 + }, + { + "epoch": 6.00650860065086, + "grad_norm": 2.916790008544922, + "learning_rate": 9.9870869987087e-06, + "loss": 0.0908, + "step": 12920 + }, + { + "epoch": 6.00697350069735, + "grad_norm": 4.822860240936279, + "learning_rate": 9.987085998708601e-06, + "loss": 0.077, + "step": 12921 + }, + { + "epoch": 6.00743840074384, + "grad_norm": 3.1821165084838867, + "learning_rate": 9.9870849987085e-06, + "loss": 0.1011, + "step": 12922 + }, + { + "epoch": 6.00790330079033, + "grad_norm": 5.202454566955566, + "learning_rate": 9.987083998708401e-06, + "loss": 0.1446, + "step": 12923 + }, + { + "epoch": 6.00836820083682, + "grad_norm": 3.1822285652160645, + "learning_rate": 9.9870829987083e-06, + "loss": 0.0846, + "step": 12924 + }, + { + "epoch": 6.00883310088331, + "grad_norm": 2.879415988922119, + "learning_rate": 9.9870819987082e-06, + "loss": 0.0611, + "step": 12925 + }, + { + "epoch": 6.0092980009298005, + "grad_norm": 4.186221122741699, + "learning_rate": 9.987080998708101e-06, + "loss": 0.1372, + "step": 12926 + }, + { + "epoch": 6.00976290097629, + "grad_norm": 4.449685573577881, + "learning_rate": 9.987079998708e-06, + "loss": 0.2053, + "step": 12927 + }, + { + "epoch": 6.01022780102278, + "grad_norm": 5.8079729080200195, + "learning_rate": 9.987078998707901e-06, + "loss": 0.0881, + "step": 12928 + }, + { + "epoch": 6.01069270106927, + "grad_norm": 4.9358229637146, + "learning_rate": 9.9870779987078e-06, + "loss": 0.2103, + "step": 12929 + }, + { + "epoch": 6.01115760111576, + "grad_norm": 6.814001560211182, + "learning_rate": 9.9870769987077e-06, + "loss": 0.242, + "step": 12930 + }, + { + "epoch": 6.01162250116225, + "grad_norm": 6.652743816375732, + "learning_rate": 9.987075998707601e-06, + "loss": 0.2035, + "step": 12931 + }, + { + "epoch": 6.01208740120874, + "grad_norm": 6.073646545410156, + "learning_rate": 9.987074998707502e-06, + "loss": 0.2311, + "step": 12932 + }, + { + "epoch": 6.01255230125523, + "grad_norm": 3.4660232067108154, + "learning_rate": 9.9870739987074e-06, + "loss": 0.0683, + "step": 12933 + }, + { + "epoch": 6.01301720130172, + "grad_norm": 4.8900227546691895, + "learning_rate": 9.987072998707301e-06, + "loss": 0.237, + "step": 12934 + }, + { + "epoch": 6.0134821013482105, + "grad_norm": 5.3841400146484375, + "learning_rate": 9.9870719987072e-06, + "loss": 0.1869, + "step": 12935 + }, + { + "epoch": 6.0139470013947, + "grad_norm": 7.802471160888672, + "learning_rate": 9.9870709987071e-06, + "loss": 0.3595, + "step": 12936 + }, + { + "epoch": 6.01441190144119, + "grad_norm": 5.221591949462891, + "learning_rate": 9.987069998707e-06, + "loss": 0.1545, + "step": 12937 + }, + { + "epoch": 6.01487680148768, + "grad_norm": 6.935675144195557, + "learning_rate": 9.9870689987069e-06, + "loss": 0.4244, + "step": 12938 + }, + { + "epoch": 6.015341701534171, + "grad_norm": 8.054855346679688, + "learning_rate": 9.987067998706801e-06, + "loss": 0.4642, + "step": 12939 + }, + { + "epoch": 6.01580660158066, + "grad_norm": 15.918935775756836, + "learning_rate": 9.9870669987067e-06, + "loss": 0.3468, + "step": 12940 + }, + { + "epoch": 6.01627150162715, + "grad_norm": 4.8025288581848145, + "learning_rate": 9.987065998706601e-06, + "loss": 0.2076, + "step": 12941 + }, + { + "epoch": 6.01673640167364, + "grad_norm": 6.262852191925049, + "learning_rate": 9.9870649987065e-06, + "loss": 0.2218, + "step": 12942 + }, + { + "epoch": 6.01720130172013, + "grad_norm": 3.095489740371704, + "learning_rate": 9.9870639987064e-06, + "loss": 0.0857, + "step": 12943 + }, + { + "epoch": 6.0176662017666205, + "grad_norm": 6.34553861618042, + "learning_rate": 9.987062998706301e-06, + "loss": 0.2351, + "step": 12944 + }, + { + "epoch": 6.01813110181311, + "grad_norm": 4.792500972747803, + "learning_rate": 9.987061998706202e-06, + "loss": 0.26, + "step": 12945 + }, + { + "epoch": 6.0185960018596, + "grad_norm": 3.822425127029419, + "learning_rate": 9.987060998706101e-06, + "loss": 0.1824, + "step": 12946 + }, + { + "epoch": 6.01906090190609, + "grad_norm": 4.670709609985352, + "learning_rate": 9.987059998706e-06, + "loss": 0.2295, + "step": 12947 + }, + { + "epoch": 6.019525801952581, + "grad_norm": 6.434863090515137, + "learning_rate": 9.9870589987059e-06, + "loss": 0.3224, + "step": 12948 + }, + { + "epoch": 6.01999070199907, + "grad_norm": 5.84604024887085, + "learning_rate": 9.9870579987058e-06, + "loss": 0.1231, + "step": 12949 + }, + { + "epoch": 6.02045560204556, + "grad_norm": 6.200417995452881, + "learning_rate": 9.9870569987057e-06, + "loss": 0.174, + "step": 12950 + }, + { + "epoch": 6.02092050209205, + "grad_norm": 7.354926586151123, + "learning_rate": 9.9870559987056e-06, + "loss": 0.0462, + "step": 12951 + }, + { + "epoch": 6.02138540213854, + "grad_norm": 4.988061904907227, + "learning_rate": 9.987054998705501e-06, + "loss": 0.2084, + "step": 12952 + }, + { + "epoch": 6.0218503021850305, + "grad_norm": 6.2401227951049805, + "learning_rate": 9.9870539987054e-06, + "loss": 0.2036, + "step": 12953 + }, + { + "epoch": 6.02231520223152, + "grad_norm": 5.2424139976501465, + "learning_rate": 9.987052998705301e-06, + "loss": 0.3192, + "step": 12954 + }, + { + "epoch": 6.02278010227801, + "grad_norm": 4.564380168914795, + "learning_rate": 9.9870519987052e-06, + "loss": 0.0618, + "step": 12955 + }, + { + "epoch": 6.0232450023245, + "grad_norm": 3.9677250385284424, + "learning_rate": 9.9870509987051e-06, + "loss": 0.0899, + "step": 12956 + }, + { + "epoch": 6.023709902370991, + "grad_norm": 3.8098037242889404, + "learning_rate": 9.987049998705001e-06, + "loss": 0.1419, + "step": 12957 + }, + { + "epoch": 6.02417480241748, + "grad_norm": 5.1115803718566895, + "learning_rate": 9.9870489987049e-06, + "loss": 0.33, + "step": 12958 + }, + { + "epoch": 6.02463970246397, + "grad_norm": 3.6489145755767822, + "learning_rate": 9.987047998704801e-06, + "loss": 0.1442, + "step": 12959 + }, + { + "epoch": 6.02510460251046, + "grad_norm": 5.432004928588867, + "learning_rate": 9.9870469987047e-06, + "loss": 0.1322, + "step": 12960 + }, + { + "epoch": 6.025569502556951, + "grad_norm": 5.347540378570557, + "learning_rate": 9.9870459987046e-06, + "loss": 0.2661, + "step": 12961 + }, + { + "epoch": 6.0260344026034405, + "grad_norm": 4.359226703643799, + "learning_rate": 9.9870449987045e-06, + "loss": 0.1256, + "step": 12962 + }, + { + "epoch": 6.02649930264993, + "grad_norm": 6.661477565765381, + "learning_rate": 9.9870439987044e-06, + "loss": 0.1731, + "step": 12963 + }, + { + "epoch": 6.02696420269642, + "grad_norm": 6.247859001159668, + "learning_rate": 9.987042998704301e-06, + "loss": 0.3026, + "step": 12964 + }, + { + "epoch": 6.02742910274291, + "grad_norm": 6.79817008972168, + "learning_rate": 9.987041998704202e-06, + "loss": 0.2587, + "step": 12965 + }, + { + "epoch": 6.027894002789401, + "grad_norm": 7.116866111755371, + "learning_rate": 9.9870409987041e-06, + "loss": 0.26, + "step": 12966 + }, + { + "epoch": 6.02835890283589, + "grad_norm": 6.913412570953369, + "learning_rate": 9.987039998704001e-06, + "loss": 0.145, + "step": 12967 + }, + { + "epoch": 6.02882380288238, + "grad_norm": 4.463513374328613, + "learning_rate": 9.9870389987039e-06, + "loss": 0.1818, + "step": 12968 + }, + { + "epoch": 6.02928870292887, + "grad_norm": 2.7265872955322266, + "learning_rate": 9.9870379987038e-06, + "loss": 0.0317, + "step": 12969 + }, + { + "epoch": 6.029753602975361, + "grad_norm": 4.646970272064209, + "learning_rate": 9.987036998703701e-06, + "loss": 0.1521, + "step": 12970 + }, + { + "epoch": 6.0302185030218505, + "grad_norm": 5.011460781097412, + "learning_rate": 9.9870359987036e-06, + "loss": 0.1976, + "step": 12971 + }, + { + "epoch": 6.03068340306834, + "grad_norm": 6.573915004730225, + "learning_rate": 9.987034998703501e-06, + "loss": 0.3301, + "step": 12972 + }, + { + "epoch": 6.03114830311483, + "grad_norm": 8.251087188720703, + "learning_rate": 9.9870339987034e-06, + "loss": 0.1546, + "step": 12973 + }, + { + "epoch": 6.03161320316132, + "grad_norm": 4.581360816955566, + "learning_rate": 9.9870329987033e-06, + "loss": 0.1713, + "step": 12974 + }, + { + "epoch": 6.032078103207811, + "grad_norm": 3.922471523284912, + "learning_rate": 9.987031998703201e-06, + "loss": 0.1492, + "step": 12975 + }, + { + "epoch": 6.0325430032543, + "grad_norm": 4.928203582763672, + "learning_rate": 9.9870309987031e-06, + "loss": 0.2115, + "step": 12976 + }, + { + "epoch": 6.03300790330079, + "grad_norm": 7.196179389953613, + "learning_rate": 9.987029998703001e-06, + "loss": 0.2676, + "step": 12977 + }, + { + "epoch": 6.03347280334728, + "grad_norm": 5.0425519943237305, + "learning_rate": 9.987028998702902e-06, + "loss": 0.0765, + "step": 12978 + }, + { + "epoch": 6.033937703393771, + "grad_norm": 6.963411808013916, + "learning_rate": 9.9870279987028e-06, + "loss": 0.4216, + "step": 12979 + }, + { + "epoch": 6.0344026034402605, + "grad_norm": 3.9600472450256348, + "learning_rate": 9.9870269987027e-06, + "loss": 0.1614, + "step": 12980 + }, + { + "epoch": 6.03486750348675, + "grad_norm": 7.596357822418213, + "learning_rate": 9.9870259987026e-06, + "loss": 0.2079, + "step": 12981 + }, + { + "epoch": 6.03533240353324, + "grad_norm": 5.017897605895996, + "learning_rate": 9.987024998702501e-06, + "loss": 0.193, + "step": 12982 + }, + { + "epoch": 6.035797303579731, + "grad_norm": 5.975855350494385, + "learning_rate": 9.987023998702402e-06, + "loss": 0.1099, + "step": 12983 + }, + { + "epoch": 6.036262203626221, + "grad_norm": 3.962777614593506, + "learning_rate": 9.9870229987023e-06, + "loss": 0.1471, + "step": 12984 + }, + { + "epoch": 6.03672710367271, + "grad_norm": 5.148624897003174, + "learning_rate": 9.987021998702201e-06, + "loss": 0.1741, + "step": 12985 + }, + { + "epoch": 6.0371920037192, + "grad_norm": 5.181689739227295, + "learning_rate": 9.9870209987021e-06, + "loss": 0.2532, + "step": 12986 + }, + { + "epoch": 6.03765690376569, + "grad_norm": 5.15285587310791, + "learning_rate": 9.987019998702e-06, + "loss": 0.1379, + "step": 12987 + }, + { + "epoch": 6.038121803812181, + "grad_norm": 5.932135581970215, + "learning_rate": 9.987018998701901e-06, + "loss": 0.2838, + "step": 12988 + }, + { + "epoch": 6.0385867038586705, + "grad_norm": 4.377217769622803, + "learning_rate": 9.9870179987018e-06, + "loss": 0.1714, + "step": 12989 + }, + { + "epoch": 6.03905160390516, + "grad_norm": 4.977087020874023, + "learning_rate": 9.987016998701701e-06, + "loss": 0.1384, + "step": 12990 + }, + { + "epoch": 6.03951650395165, + "grad_norm": 6.694918155670166, + "learning_rate": 9.9870159987016e-06, + "loss": 0.4094, + "step": 12991 + }, + { + "epoch": 6.039981403998141, + "grad_norm": 5.567802429199219, + "learning_rate": 9.9870149987015e-06, + "loss": 0.3547, + "step": 12992 + }, + { + "epoch": 6.040446304044631, + "grad_norm": 4.821101188659668, + "learning_rate": 9.9870139987014e-06, + "loss": 0.1827, + "step": 12993 + }, + { + "epoch": 6.04091120409112, + "grad_norm": 4.310473442077637, + "learning_rate": 9.9870129987013e-06, + "loss": 0.2031, + "step": 12994 + }, + { + "epoch": 6.04137610413761, + "grad_norm": 2.015943765640259, + "learning_rate": 9.987011998701201e-06, + "loss": 0.0323, + "step": 12995 + }, + { + "epoch": 6.0418410041841, + "grad_norm": 3.8898346424102783, + "learning_rate": 9.987010998701102e-06, + "loss": 0.0893, + "step": 12996 + }, + { + "epoch": 6.042305904230591, + "grad_norm": 6.516469478607178, + "learning_rate": 9.987009998701e-06, + "loss": 0.262, + "step": 12997 + }, + { + "epoch": 6.0427708042770805, + "grad_norm": 2.70215106010437, + "learning_rate": 9.987008998700901e-06, + "loss": 0.0878, + "step": 12998 + }, + { + "epoch": 6.04323570432357, + "grad_norm": 2.6528592109680176, + "learning_rate": 9.9870079987008e-06, + "loss": 0.0448, + "step": 12999 + }, + { + "epoch": 6.04370060437006, + "grad_norm": 7.274248123168945, + "learning_rate": 9.9870069987007e-06, + "loss": 0.2147, + "step": 13000 + }, + { + "epoch": 6.044165504416551, + "grad_norm": 4.854148864746094, + "learning_rate": 9.987005998700601e-06, + "loss": 0.1859, + "step": 13001 + }, + { + "epoch": 6.044630404463041, + "grad_norm": 4.1957855224609375, + "learning_rate": 9.9870049987005e-06, + "loss": 0.1468, + "step": 13002 + }, + { + "epoch": 6.04509530450953, + "grad_norm": 3.4478507041931152, + "learning_rate": 9.987003998700401e-06, + "loss": 0.065, + "step": 13003 + }, + { + "epoch": 6.04556020455602, + "grad_norm": 2.6331002712249756, + "learning_rate": 9.9870029987003e-06, + "loss": 0.0687, + "step": 13004 + }, + { + "epoch": 6.046025104602511, + "grad_norm": 3.84688401222229, + "learning_rate": 9.9870019987002e-06, + "loss": 0.1336, + "step": 13005 + }, + { + "epoch": 6.046490004649001, + "grad_norm": 6.827157497406006, + "learning_rate": 9.9870009987001e-06, + "loss": 0.4163, + "step": 13006 + }, + { + "epoch": 6.0469549046954905, + "grad_norm": 4.9909539222717285, + "learning_rate": 9.9869999987e-06, + "loss": 0.1573, + "step": 13007 + }, + { + "epoch": 6.04741980474198, + "grad_norm": 5.0493059158325195, + "learning_rate": 9.986998998699901e-06, + "loss": 0.1652, + "step": 13008 + }, + { + "epoch": 6.04788470478847, + "grad_norm": 4.569192886352539, + "learning_rate": 9.986997998699802e-06, + "loss": 0.1502, + "step": 13009 + }, + { + "epoch": 6.048349604834961, + "grad_norm": 6.278692722320557, + "learning_rate": 9.9869969986997e-06, + "loss": 0.1486, + "step": 13010 + }, + { + "epoch": 6.048814504881451, + "grad_norm": 5.668505668640137, + "learning_rate": 9.986995998699601e-06, + "loss": 0.1962, + "step": 13011 + }, + { + "epoch": 6.04927940492794, + "grad_norm": 5.8399128913879395, + "learning_rate": 9.9869949986995e-06, + "loss": 0.0978, + "step": 13012 + }, + { + "epoch": 6.04974430497443, + "grad_norm": 6.793979644775391, + "learning_rate": 9.986993998699401e-06, + "loss": 0.1329, + "step": 13013 + }, + { + "epoch": 6.050209205020921, + "grad_norm": 5.937116622924805, + "learning_rate": 9.986992998699302e-06, + "loss": 0.2968, + "step": 13014 + }, + { + "epoch": 6.050674105067411, + "grad_norm": 4.4742960929870605, + "learning_rate": 9.9869919986992e-06, + "loss": 0.1561, + "step": 13015 + }, + { + "epoch": 6.0511390051139005, + "grad_norm": 2.4968202114105225, + "learning_rate": 9.986990998699101e-06, + "loss": 0.0591, + "step": 13016 + }, + { + "epoch": 6.05160390516039, + "grad_norm": 5.047667026519775, + "learning_rate": 9.986989998699e-06, + "loss": 0.1988, + "step": 13017 + }, + { + "epoch": 6.052068805206881, + "grad_norm": 6.275325775146484, + "learning_rate": 9.9869889986989e-06, + "loss": 0.2381, + "step": 13018 + }, + { + "epoch": 6.052533705253371, + "grad_norm": 5.142940998077393, + "learning_rate": 9.9869879986988e-06, + "loss": 0.2133, + "step": 13019 + }, + { + "epoch": 6.052998605299861, + "grad_norm": 8.536040306091309, + "learning_rate": 9.9869869986987e-06, + "loss": 0.2006, + "step": 13020 + }, + { + "epoch": 6.05346350534635, + "grad_norm": 5.572248458862305, + "learning_rate": 9.986985998698601e-06, + "loss": 0.2092, + "step": 13021 + }, + { + "epoch": 6.05392840539284, + "grad_norm": 3.702615261077881, + "learning_rate": 9.9869849986985e-06, + "loss": 0.1294, + "step": 13022 + }, + { + "epoch": 6.054393305439331, + "grad_norm": 6.774037837982178, + "learning_rate": 9.9869839986984e-06, + "loss": 0.272, + "step": 13023 + }, + { + "epoch": 6.054858205485821, + "grad_norm": 6.6933488845825195, + "learning_rate": 9.9869829986983e-06, + "loss": 0.3372, + "step": 13024 + }, + { + "epoch": 6.0553231055323105, + "grad_norm": 4.049201965332031, + "learning_rate": 9.9869819986982e-06, + "loss": 0.1079, + "step": 13025 + }, + { + "epoch": 6.0557880055788, + "grad_norm": 4.6713032722473145, + "learning_rate": 9.986980998698101e-06, + "loss": 0.147, + "step": 13026 + }, + { + "epoch": 6.056252905625291, + "grad_norm": 2.936105251312256, + "learning_rate": 9.986979998698002e-06, + "loss": 0.0646, + "step": 13027 + }, + { + "epoch": 6.056717805671781, + "grad_norm": 8.695577621459961, + "learning_rate": 9.9869789986979e-06, + "loss": 0.325, + "step": 13028 + }, + { + "epoch": 6.0571827057182706, + "grad_norm": 4.000566482543945, + "learning_rate": 9.986977998697801e-06, + "loss": 0.1271, + "step": 13029 + }, + { + "epoch": 6.05764760576476, + "grad_norm": 5.342721939086914, + "learning_rate": 9.9869769986977e-06, + "loss": 0.1426, + "step": 13030 + }, + { + "epoch": 6.05811250581125, + "grad_norm": 5.29679536819458, + "learning_rate": 9.986975998697601e-06, + "loss": 0.1183, + "step": 13031 + }, + { + "epoch": 6.058577405857741, + "grad_norm": 4.139978885650635, + "learning_rate": 9.9869749986975e-06, + "loss": 0.1005, + "step": 13032 + }, + { + "epoch": 6.059042305904231, + "grad_norm": 3.1832118034362793, + "learning_rate": 9.9869739986974e-06, + "loss": 0.0882, + "step": 13033 + }, + { + "epoch": 6.0595072059507205, + "grad_norm": 4.119778633117676, + "learning_rate": 9.986972998697301e-06, + "loss": 0.1519, + "step": 13034 + }, + { + "epoch": 6.05997210599721, + "grad_norm": 2.0376596450805664, + "learning_rate": 9.9869719986972e-06, + "loss": 0.037, + "step": 13035 + }, + { + "epoch": 6.060437006043701, + "grad_norm": 4.921615123748779, + "learning_rate": 9.9869709986971e-06, + "loss": 0.183, + "step": 13036 + }, + { + "epoch": 6.060901906090191, + "grad_norm": 6.218265533447266, + "learning_rate": 9.986969998697e-06, + "loss": 0.1312, + "step": 13037 + }, + { + "epoch": 6.0613668061366806, + "grad_norm": 5.481038570404053, + "learning_rate": 9.9869689986969e-06, + "loss": 0.4512, + "step": 13038 + }, + { + "epoch": 6.06183170618317, + "grad_norm": 4.821275234222412, + "learning_rate": 9.986967998696801e-06, + "loss": 0.1484, + "step": 13039 + }, + { + "epoch": 6.062296606229661, + "grad_norm": 6.023797035217285, + "learning_rate": 9.986966998696702e-06, + "loss": 0.3042, + "step": 13040 + }, + { + "epoch": 6.062761506276151, + "grad_norm": 6.103550434112549, + "learning_rate": 9.9869659986966e-06, + "loss": 0.1178, + "step": 13041 + }, + { + "epoch": 6.063226406322641, + "grad_norm": 3.5748627185821533, + "learning_rate": 9.986964998696501e-06, + "loss": 0.0522, + "step": 13042 + }, + { + "epoch": 6.0636913063691305, + "grad_norm": 3.1520845890045166, + "learning_rate": 9.9869639986964e-06, + "loss": 0.0609, + "step": 13043 + }, + { + "epoch": 6.06415620641562, + "grad_norm": 2.994213104248047, + "learning_rate": 9.986962998696301e-06, + "loss": 0.028, + "step": 13044 + }, + { + "epoch": 6.064621106462111, + "grad_norm": 4.8634934425354, + "learning_rate": 9.9869619986962e-06, + "loss": 0.1156, + "step": 13045 + }, + { + "epoch": 6.065086006508601, + "grad_norm": 3.0091981887817383, + "learning_rate": 9.9869609986961e-06, + "loss": 0.0791, + "step": 13046 + }, + { + "epoch": 6.0655509065550905, + "grad_norm": 7.678324222564697, + "learning_rate": 9.986959998696001e-06, + "loss": 0.1382, + "step": 13047 + }, + { + "epoch": 6.06601580660158, + "grad_norm": 4.340754508972168, + "learning_rate": 9.9869589986959e-06, + "loss": 0.1444, + "step": 13048 + }, + { + "epoch": 6.066480706648071, + "grad_norm": 8.056041717529297, + "learning_rate": 9.986957998695801e-06, + "loss": 0.2497, + "step": 13049 + }, + { + "epoch": 6.066945606694561, + "grad_norm": 5.6112284660339355, + "learning_rate": 9.9869569986957e-06, + "loss": 0.1999, + "step": 13050 + }, + { + "epoch": 6.067410506741051, + "grad_norm": 1.740722417831421, + "learning_rate": 9.9869559986956e-06, + "loss": 0.042, + "step": 13051 + }, + { + "epoch": 6.0678754067875404, + "grad_norm": 2.920375347137451, + "learning_rate": 9.986954998695501e-06, + "loss": 0.0729, + "step": 13052 + }, + { + "epoch": 6.06834030683403, + "grad_norm": 8.071351051330566, + "learning_rate": 9.986953998695402e-06, + "loss": 0.3455, + "step": 13053 + }, + { + "epoch": 6.068805206880521, + "grad_norm": 4.873058795928955, + "learning_rate": 9.9869529986953e-06, + "loss": 0.2366, + "step": 13054 + }, + { + "epoch": 6.069270106927011, + "grad_norm": 5.433971881866455, + "learning_rate": 9.9869519986952e-06, + "loss": 0.2932, + "step": 13055 + }, + { + "epoch": 6.0697350069735005, + "grad_norm": 5.251003742218018, + "learning_rate": 9.9869509986951e-06, + "loss": 0.1752, + "step": 13056 + }, + { + "epoch": 6.07019990701999, + "grad_norm": 3.2699997425079346, + "learning_rate": 9.986949998695001e-06, + "loss": 0.0806, + "step": 13057 + }, + { + "epoch": 6.070664807066481, + "grad_norm": 6.647690773010254, + "learning_rate": 9.9869489986949e-06, + "loss": 0.2774, + "step": 13058 + }, + { + "epoch": 6.071129707112971, + "grad_norm": 1.3468209505081177, + "learning_rate": 9.9869479986948e-06, + "loss": 0.0188, + "step": 13059 + }, + { + "epoch": 6.071594607159461, + "grad_norm": 5.442947864532471, + "learning_rate": 9.986946998694701e-06, + "loss": 0.1464, + "step": 13060 + }, + { + "epoch": 6.0720595072059504, + "grad_norm": 4.723100662231445, + "learning_rate": 9.9869459986946e-06, + "loss": 0.1082, + "step": 13061 + }, + { + "epoch": 6.072524407252441, + "grad_norm": 4.750783920288086, + "learning_rate": 9.986944998694501e-06, + "loss": 0.0983, + "step": 13062 + }, + { + "epoch": 6.072989307298931, + "grad_norm": 5.991708278656006, + "learning_rate": 9.9869439986944e-06, + "loss": 0.2646, + "step": 13063 + }, + { + "epoch": 6.073454207345421, + "grad_norm": 4.590573787689209, + "learning_rate": 9.9869429986943e-06, + "loss": 0.2349, + "step": 13064 + }, + { + "epoch": 6.0739191073919105, + "grad_norm": 5.132450103759766, + "learning_rate": 9.986941998694201e-06, + "loss": 0.1835, + "step": 13065 + }, + { + "epoch": 6.0743840074384, + "grad_norm": 4.426754951477051, + "learning_rate": 9.9869409986941e-06, + "loss": 0.1455, + "step": 13066 + }, + { + "epoch": 6.074848907484891, + "grad_norm": 3.7779810428619385, + "learning_rate": 9.986939998694001e-06, + "loss": 0.0967, + "step": 13067 + }, + { + "epoch": 6.075313807531381, + "grad_norm": 6.824985504150391, + "learning_rate": 9.9869389986939e-06, + "loss": 0.1965, + "step": 13068 + }, + { + "epoch": 6.075778707577871, + "grad_norm": 6.7673821449279785, + "learning_rate": 9.9869379986938e-06, + "loss": 0.2737, + "step": 13069 + }, + { + "epoch": 6.07624360762436, + "grad_norm": 6.1961469650268555, + "learning_rate": 9.986936998693701e-06, + "loss": 0.1506, + "step": 13070 + }, + { + "epoch": 6.076708507670851, + "grad_norm": 3.2862350940704346, + "learning_rate": 9.986935998693602e-06, + "loss": 0.0652, + "step": 13071 + }, + { + "epoch": 6.077173407717341, + "grad_norm": 5.952411651611328, + "learning_rate": 9.9869349986935e-06, + "loss": 0.2113, + "step": 13072 + }, + { + "epoch": 6.077638307763831, + "grad_norm": 2.728151321411133, + "learning_rate": 9.986933998693401e-06, + "loss": 0.091, + "step": 13073 + }, + { + "epoch": 6.0781032078103205, + "grad_norm": 4.21885871887207, + "learning_rate": 9.9869329986933e-06, + "loss": 0.0809, + "step": 13074 + }, + { + "epoch": 6.07856810785681, + "grad_norm": 5.18699836730957, + "learning_rate": 9.986931998693201e-06, + "loss": 0.1144, + "step": 13075 + }, + { + "epoch": 6.079033007903301, + "grad_norm": 5.280211448669434, + "learning_rate": 9.9869309986931e-06, + "loss": 0.1774, + "step": 13076 + }, + { + "epoch": 6.079497907949791, + "grad_norm": 4.334612846374512, + "learning_rate": 9.986929998693e-06, + "loss": 0.1264, + "step": 13077 + }, + { + "epoch": 6.079962807996281, + "grad_norm": 3.5482051372528076, + "learning_rate": 9.986928998692901e-06, + "loss": 0.1146, + "step": 13078 + }, + { + "epoch": 6.08042770804277, + "grad_norm": 8.098623275756836, + "learning_rate": 9.9869279986928e-06, + "loss": 0.1507, + "step": 13079 + }, + { + "epoch": 6.080892608089261, + "grad_norm": 1.9113317728042603, + "learning_rate": 9.986926998692701e-06, + "loss": 0.038, + "step": 13080 + }, + { + "epoch": 6.081357508135751, + "grad_norm": 4.413683891296387, + "learning_rate": 9.9869259986926e-06, + "loss": 0.1635, + "step": 13081 + }, + { + "epoch": 6.081822408182241, + "grad_norm": 6.429877281188965, + "learning_rate": 9.9869249986925e-06, + "loss": 0.2868, + "step": 13082 + }, + { + "epoch": 6.0822873082287305, + "grad_norm": 5.311830520629883, + "learning_rate": 9.986923998692401e-06, + "loss": 0.2351, + "step": 13083 + }, + { + "epoch": 6.082752208275221, + "grad_norm": 3.410118579864502, + "learning_rate": 9.986922998692302e-06, + "loss": 0.0731, + "step": 13084 + }, + { + "epoch": 6.083217108321711, + "grad_norm": 4.020870208740234, + "learning_rate": 9.9869219986922e-06, + "loss": 0.1106, + "step": 13085 + }, + { + "epoch": 6.083682008368201, + "grad_norm": 4.759524822235107, + "learning_rate": 9.986920998692101e-06, + "loss": 0.245, + "step": 13086 + }, + { + "epoch": 6.084146908414691, + "grad_norm": 13.397867202758789, + "learning_rate": 9.986919998692e-06, + "loss": 0.2974, + "step": 13087 + }, + { + "epoch": 6.08461180846118, + "grad_norm": 4.69045877456665, + "learning_rate": 9.9869189986919e-06, + "loss": 0.1419, + "step": 13088 + }, + { + "epoch": 6.085076708507671, + "grad_norm": 4.705326557159424, + "learning_rate": 9.9869179986918e-06, + "loss": 0.1597, + "step": 13089 + }, + { + "epoch": 6.085541608554161, + "grad_norm": 1.8484731912612915, + "learning_rate": 9.9869169986917e-06, + "loss": 0.0324, + "step": 13090 + }, + { + "epoch": 6.086006508600651, + "grad_norm": 6.949005603790283, + "learning_rate": 9.986915998691601e-06, + "loss": 0.2195, + "step": 13091 + }, + { + "epoch": 6.0864714086471405, + "grad_norm": 7.247747898101807, + "learning_rate": 9.9869149986915e-06, + "loss": 0.3176, + "step": 13092 + }, + { + "epoch": 6.086936308693631, + "grad_norm": 4.605177402496338, + "learning_rate": 9.986913998691401e-06, + "loss": 0.1635, + "step": 13093 + }, + { + "epoch": 6.087401208740121, + "grad_norm": 4.852396011352539, + "learning_rate": 9.9869129986913e-06, + "loss": 0.1383, + "step": 13094 + }, + { + "epoch": 6.087866108786611, + "grad_norm": 4.706364631652832, + "learning_rate": 9.9869119986912e-06, + "loss": 0.1362, + "step": 13095 + }, + { + "epoch": 6.088331008833101, + "grad_norm": 2.840597152709961, + "learning_rate": 9.986910998691101e-06, + "loss": 0.0733, + "step": 13096 + }, + { + "epoch": 6.088795908879591, + "grad_norm": 5.05210018157959, + "learning_rate": 9.986909998691002e-06, + "loss": 0.1523, + "step": 13097 + }, + { + "epoch": 6.089260808926081, + "grad_norm": 4.306992530822754, + "learning_rate": 9.986908998690901e-06, + "loss": 0.1391, + "step": 13098 + }, + { + "epoch": 6.089725708972571, + "grad_norm": 6.12999153137207, + "learning_rate": 9.9869079986908e-06, + "loss": 0.0809, + "step": 13099 + }, + { + "epoch": 6.090190609019061, + "grad_norm": 4.649692535400391, + "learning_rate": 9.9869069986907e-06, + "loss": 0.0737, + "step": 13100 + }, + { + "epoch": 6.0906555090655505, + "grad_norm": 6.959776878356934, + "learning_rate": 9.9869059986906e-06, + "loss": 0.386, + "step": 13101 + }, + { + "epoch": 6.091120409112041, + "grad_norm": 3.0853114128112793, + "learning_rate": 9.9869049986905e-06, + "loss": 0.0912, + "step": 13102 + }, + { + "epoch": 6.091585309158531, + "grad_norm": 6.423983573913574, + "learning_rate": 9.9869039986904e-06, + "loss": 0.2365, + "step": 13103 + }, + { + "epoch": 6.092050209205021, + "grad_norm": 5.354743480682373, + "learning_rate": 9.986902998690301e-06, + "loss": 0.2004, + "step": 13104 + }, + { + "epoch": 6.092515109251511, + "grad_norm": 6.483980178833008, + "learning_rate": 9.9869019986902e-06, + "loss": 0.1948, + "step": 13105 + }, + { + "epoch": 6.092980009298001, + "grad_norm": 6.143059730529785, + "learning_rate": 9.986900998690101e-06, + "loss": 0.1917, + "step": 13106 + }, + { + "epoch": 6.093444909344491, + "grad_norm": 6.5618133544921875, + "learning_rate": 9.98689999869e-06, + "loss": 0.1951, + "step": 13107 + }, + { + "epoch": 6.093909809390981, + "grad_norm": 4.096187591552734, + "learning_rate": 9.9868989986899e-06, + "loss": 0.1137, + "step": 13108 + }, + { + "epoch": 6.094374709437471, + "grad_norm": 3.950471878051758, + "learning_rate": 9.986897998689801e-06, + "loss": 0.1022, + "step": 13109 + }, + { + "epoch": 6.0948396094839605, + "grad_norm": 5.6044020652771, + "learning_rate": 9.9868969986897e-06, + "loss": 0.2037, + "step": 13110 + }, + { + "epoch": 6.095304509530451, + "grad_norm": 6.34195613861084, + "learning_rate": 9.986895998689601e-06, + "loss": 0.2445, + "step": 13111 + }, + { + "epoch": 6.095769409576941, + "grad_norm": 8.765692710876465, + "learning_rate": 9.9868949986895e-06, + "loss": 0.3033, + "step": 13112 + }, + { + "epoch": 6.096234309623431, + "grad_norm": 6.040877819061279, + "learning_rate": 9.9868939986894e-06, + "loss": 0.1381, + "step": 13113 + }, + { + "epoch": 6.096699209669921, + "grad_norm": 5.146555423736572, + "learning_rate": 9.9868929986893e-06, + "loss": 0.1652, + "step": 13114 + }, + { + "epoch": 6.097164109716411, + "grad_norm": 4.535782337188721, + "learning_rate": 9.9868919986892e-06, + "loss": 0.1663, + "step": 13115 + }, + { + "epoch": 6.097629009762901, + "grad_norm": 6.644027233123779, + "learning_rate": 9.986890998689101e-06, + "loss": 0.2865, + "step": 13116 + }, + { + "epoch": 6.098093909809391, + "grad_norm": 6.0799479484558105, + "learning_rate": 9.986889998689002e-06, + "loss": 0.1837, + "step": 13117 + }, + { + "epoch": 6.098558809855881, + "grad_norm": 2.275451898574829, + "learning_rate": 9.9868889986889e-06, + "loss": 0.0531, + "step": 13118 + }, + { + "epoch": 6.099023709902371, + "grad_norm": 6.028979778289795, + "learning_rate": 9.986887998688801e-06, + "loss": 0.1305, + "step": 13119 + }, + { + "epoch": 6.099488609948861, + "grad_norm": 4.076606273651123, + "learning_rate": 9.9868869986887e-06, + "loss": 0.1145, + "step": 13120 + }, + { + "epoch": 6.099953509995351, + "grad_norm": 4.837757587432861, + "learning_rate": 9.9868859986886e-06, + "loss": 0.1813, + "step": 13121 + }, + { + "epoch": 6.100418410041841, + "grad_norm": 6.0423970222473145, + "learning_rate": 9.986884998688501e-06, + "loss": 0.2275, + "step": 13122 + }, + { + "epoch": 6.100883310088331, + "grad_norm": 5.936700344085693, + "learning_rate": 9.9868839986884e-06, + "loss": 0.2382, + "step": 13123 + }, + { + "epoch": 6.101348210134821, + "grad_norm": 3.936556577682495, + "learning_rate": 9.986882998688301e-06, + "loss": 0.1574, + "step": 13124 + }, + { + "epoch": 6.101813110181311, + "grad_norm": 6.923903465270996, + "learning_rate": 9.9868819986882e-06, + "loss": 0.3295, + "step": 13125 + }, + { + "epoch": 6.102278010227801, + "grad_norm": 6.630427837371826, + "learning_rate": 9.9868809986881e-06, + "loss": 0.2474, + "step": 13126 + }, + { + "epoch": 6.102742910274291, + "grad_norm": 6.709728240966797, + "learning_rate": 9.986879998688e-06, + "loss": 0.3202, + "step": 13127 + }, + { + "epoch": 6.103207810320781, + "grad_norm": 5.117430686950684, + "learning_rate": 9.9868789986879e-06, + "loss": 0.1475, + "step": 13128 + }, + { + "epoch": 6.103672710367271, + "grad_norm": 5.094918251037598, + "learning_rate": 9.986877998687801e-06, + "loss": 0.1719, + "step": 13129 + }, + { + "epoch": 6.104137610413761, + "grad_norm": 8.372584342956543, + "learning_rate": 9.986876998687702e-06, + "loss": 0.3188, + "step": 13130 + }, + { + "epoch": 6.104602510460251, + "grad_norm": 3.1189143657684326, + "learning_rate": 9.9868759986876e-06, + "loss": 0.0737, + "step": 13131 + }, + { + "epoch": 6.105067410506741, + "grad_norm": 4.838466167449951, + "learning_rate": 9.9868749986875e-06, + "loss": 0.1105, + "step": 13132 + }, + { + "epoch": 6.105532310553231, + "grad_norm": 3.135047197341919, + "learning_rate": 9.9868739986874e-06, + "loss": 0.0636, + "step": 13133 + }, + { + "epoch": 6.105997210599721, + "grad_norm": 4.852935791015625, + "learning_rate": 9.986872998687301e-06, + "loss": 0.2315, + "step": 13134 + }, + { + "epoch": 6.106462110646211, + "grad_norm": 5.232976913452148, + "learning_rate": 9.986871998687202e-06, + "loss": 0.2041, + "step": 13135 + }, + { + "epoch": 6.106927010692701, + "grad_norm": 2.0171728134155273, + "learning_rate": 9.9868709986871e-06, + "loss": 0.0375, + "step": 13136 + }, + { + "epoch": 6.107391910739191, + "grad_norm": 7.6682305335998535, + "learning_rate": 9.986869998687001e-06, + "loss": 0.4904, + "step": 13137 + }, + { + "epoch": 6.107856810785681, + "grad_norm": 2.9261815547943115, + "learning_rate": 9.9868689986869e-06, + "loss": 0.081, + "step": 13138 + }, + { + "epoch": 6.108321710832171, + "grad_norm": 4.41893196105957, + "learning_rate": 9.9868679986868e-06, + "loss": 0.1279, + "step": 13139 + }, + { + "epoch": 6.108786610878661, + "grad_norm": 5.340587139129639, + "learning_rate": 9.986866998686701e-06, + "loss": 0.2017, + "step": 13140 + }, + { + "epoch": 6.1092515109251515, + "grad_norm": 5.724814414978027, + "learning_rate": 9.9868659986866e-06, + "loss": 0.1668, + "step": 13141 + }, + { + "epoch": 6.109716410971641, + "grad_norm": 2.3839986324310303, + "learning_rate": 9.986864998686501e-06, + "loss": 0.0335, + "step": 13142 + }, + { + "epoch": 6.110181311018131, + "grad_norm": 6.237898349761963, + "learning_rate": 9.9868639986864e-06, + "loss": 0.206, + "step": 13143 + }, + { + "epoch": 6.110646211064621, + "grad_norm": 6.469804763793945, + "learning_rate": 9.9868629986863e-06, + "loss": 0.1345, + "step": 13144 + }, + { + "epoch": 6.111111111111111, + "grad_norm": 3.388871669769287, + "learning_rate": 9.9868619986862e-06, + "loss": 0.0503, + "step": 13145 + }, + { + "epoch": 6.111576011157601, + "grad_norm": 9.779781341552734, + "learning_rate": 9.9868609986861e-06, + "loss": 0.2112, + "step": 13146 + }, + { + "epoch": 6.112040911204091, + "grad_norm": 3.273343086242676, + "learning_rate": 9.986859998686001e-06, + "loss": 0.0753, + "step": 13147 + }, + { + "epoch": 6.112505811250581, + "grad_norm": 4.36407470703125, + "learning_rate": 9.986858998685902e-06, + "loss": 0.1314, + "step": 13148 + }, + { + "epoch": 6.112970711297071, + "grad_norm": 4.238972187042236, + "learning_rate": 9.9868579986858e-06, + "loss": 0.0801, + "step": 13149 + }, + { + "epoch": 6.1134356113435615, + "grad_norm": 3.5391430854797363, + "learning_rate": 9.986856998685701e-06, + "loss": 0.0807, + "step": 13150 + }, + { + "epoch": 6.113900511390051, + "grad_norm": 3.3957467079162598, + "learning_rate": 9.9868559986856e-06, + "loss": 0.0675, + "step": 13151 + }, + { + "epoch": 6.114365411436541, + "grad_norm": 5.117953777313232, + "learning_rate": 9.986854998685501e-06, + "loss": 0.1807, + "step": 13152 + }, + { + "epoch": 6.114830311483031, + "grad_norm": 7.168174743652344, + "learning_rate": 9.986853998685402e-06, + "loss": 0.339, + "step": 13153 + }, + { + "epoch": 6.115295211529521, + "grad_norm": 2.3203811645507812, + "learning_rate": 9.9868529986853e-06, + "loss": 0.0479, + "step": 13154 + }, + { + "epoch": 6.115760111576011, + "grad_norm": 4.908470630645752, + "learning_rate": 9.986851998685201e-06, + "loss": 0.2163, + "step": 13155 + }, + { + "epoch": 6.116225011622501, + "grad_norm": 5.110370635986328, + "learning_rate": 9.9868509986851e-06, + "loss": 0.1465, + "step": 13156 + }, + { + "epoch": 6.116689911668991, + "grad_norm": 9.383443832397461, + "learning_rate": 9.986849998685e-06, + "loss": 0.3497, + "step": 13157 + }, + { + "epoch": 6.117154811715481, + "grad_norm": 5.751209735870361, + "learning_rate": 9.9868489986849e-06, + "loss": 0.0908, + "step": 13158 + }, + { + "epoch": 6.1176197117619715, + "grad_norm": 7.574322700500488, + "learning_rate": 9.9868479986848e-06, + "loss": 0.2743, + "step": 13159 + }, + { + "epoch": 6.118084611808461, + "grad_norm": 6.208547592163086, + "learning_rate": 9.986846998684701e-06, + "loss": 0.1134, + "step": 13160 + }, + { + "epoch": 6.118549511854951, + "grad_norm": 5.73339319229126, + "learning_rate": 9.986845998684602e-06, + "loss": 0.1797, + "step": 13161 + }, + { + "epoch": 6.119014411901441, + "grad_norm": 4.5596394538879395, + "learning_rate": 9.9868449986845e-06, + "loss": 0.1384, + "step": 13162 + }, + { + "epoch": 6.119479311947932, + "grad_norm": 6.449049472808838, + "learning_rate": 9.9868439986844e-06, + "loss": 0.3192, + "step": 13163 + }, + { + "epoch": 6.119944211994421, + "grad_norm": 2.953328847885132, + "learning_rate": 9.9868429986843e-06, + "loss": 0.0566, + "step": 13164 + }, + { + "epoch": 6.120409112040911, + "grad_norm": 6.544865608215332, + "learning_rate": 9.986841998684201e-06, + "loss": 0.2892, + "step": 13165 + }, + { + "epoch": 6.120874012087401, + "grad_norm": 8.8678560256958, + "learning_rate": 9.986840998684102e-06, + "loss": 0.3304, + "step": 13166 + }, + { + "epoch": 6.121338912133891, + "grad_norm": 6.726463794708252, + "learning_rate": 9.986839998684e-06, + "loss": 0.1212, + "step": 13167 + }, + { + "epoch": 6.1218038121803815, + "grad_norm": 3.4886474609375, + "learning_rate": 9.986838998683901e-06, + "loss": 0.1105, + "step": 13168 + }, + { + "epoch": 6.122268712226871, + "grad_norm": 5.206578254699707, + "learning_rate": 9.9868379986838e-06, + "loss": 0.1223, + "step": 13169 + }, + { + "epoch": 6.122733612273361, + "grad_norm": 3.0496084690093994, + "learning_rate": 9.986836998683701e-06, + "loss": 0.0517, + "step": 13170 + }, + { + "epoch": 6.123198512319851, + "grad_norm": 3.1166532039642334, + "learning_rate": 9.9868359986836e-06, + "loss": 0.0711, + "step": 13171 + }, + { + "epoch": 6.123663412366342, + "grad_norm": 4.862766265869141, + "learning_rate": 9.9868349986835e-06, + "loss": 0.1832, + "step": 13172 + }, + { + "epoch": 6.124128312412831, + "grad_norm": 6.55179500579834, + "learning_rate": 9.986833998683401e-06, + "loss": 0.1965, + "step": 13173 + }, + { + "epoch": 6.124593212459321, + "grad_norm": 4.415182113647461, + "learning_rate": 9.9868329986833e-06, + "loss": 0.1313, + "step": 13174 + }, + { + "epoch": 6.125058112505811, + "grad_norm": 3.6423208713531494, + "learning_rate": 9.9868319986832e-06, + "loss": 0.0727, + "step": 13175 + }, + { + "epoch": 6.125523012552302, + "grad_norm": 5.801853656768799, + "learning_rate": 9.9868309986831e-06, + "loss": 0.1706, + "step": 13176 + }, + { + "epoch": 6.1259879125987915, + "grad_norm": 4.46264123916626, + "learning_rate": 9.986829998683e-06, + "loss": 0.1405, + "step": 13177 + }, + { + "epoch": 6.126452812645281, + "grad_norm": 4.315385818481445, + "learning_rate": 9.986828998682901e-06, + "loss": 0.1705, + "step": 13178 + }, + { + "epoch": 6.126917712691771, + "grad_norm": 5.902073383331299, + "learning_rate": 9.986827998682802e-06, + "loss": 0.1595, + "step": 13179 + }, + { + "epoch": 6.127382612738261, + "grad_norm": 5.152707099914551, + "learning_rate": 9.9868269986827e-06, + "loss": 0.1294, + "step": 13180 + }, + { + "epoch": 6.127847512784752, + "grad_norm": 8.904936790466309, + "learning_rate": 9.986825998682601e-06, + "loss": 0.1968, + "step": 13181 + }, + { + "epoch": 6.128312412831241, + "grad_norm": 10.73348331451416, + "learning_rate": 9.9868249986825e-06, + "loss": 0.2706, + "step": 13182 + }, + { + "epoch": 6.128777312877731, + "grad_norm": 6.0683698654174805, + "learning_rate": 9.986823998682401e-06, + "loss": 0.0984, + "step": 13183 + }, + { + "epoch": 6.129242212924221, + "grad_norm": 6.135708332061768, + "learning_rate": 9.9868229986823e-06, + "loss": 0.2254, + "step": 13184 + }, + { + "epoch": 6.129707112970712, + "grad_norm": 2.827915668487549, + "learning_rate": 9.9868219986822e-06, + "loss": 0.0693, + "step": 13185 + }, + { + "epoch": 6.1301720130172015, + "grad_norm": 5.02428674697876, + "learning_rate": 9.986820998682101e-06, + "loss": 0.2457, + "step": 13186 + }, + { + "epoch": 6.130636913063691, + "grad_norm": 4.582225322723389, + "learning_rate": 9.986819998682e-06, + "loss": 0.205, + "step": 13187 + }, + { + "epoch": 6.131101813110181, + "grad_norm": 4.879201889038086, + "learning_rate": 9.9868189986819e-06, + "loss": 0.1162, + "step": 13188 + }, + { + "epoch": 6.131566713156671, + "grad_norm": 5.080564498901367, + "learning_rate": 9.9868179986818e-06, + "loss": 0.1318, + "step": 13189 + }, + { + "epoch": 6.132031613203162, + "grad_norm": 6.888038635253906, + "learning_rate": 9.9868169986817e-06, + "loss": 0.1561, + "step": 13190 + }, + { + "epoch": 6.132496513249651, + "grad_norm": 5.181428909301758, + "learning_rate": 9.986815998681601e-06, + "loss": 0.2517, + "step": 13191 + }, + { + "epoch": 6.132961413296141, + "grad_norm": 6.216144561767578, + "learning_rate": 9.986814998681502e-06, + "loss": 0.2043, + "step": 13192 + }, + { + "epoch": 6.133426313342631, + "grad_norm": 2.6325345039367676, + "learning_rate": 9.9868139986814e-06, + "loss": 0.0493, + "step": 13193 + }, + { + "epoch": 6.133891213389122, + "grad_norm": 8.44499397277832, + "learning_rate": 9.986812998681301e-06, + "loss": 0.3242, + "step": 13194 + }, + { + "epoch": 6.1343561134356115, + "grad_norm": 9.09831428527832, + "learning_rate": 9.9868119986812e-06, + "loss": 0.2529, + "step": 13195 + }, + { + "epoch": 6.134821013482101, + "grad_norm": 3.685887575149536, + "learning_rate": 9.9868109986811e-06, + "loss": 0.1342, + "step": 13196 + }, + { + "epoch": 6.135285913528591, + "grad_norm": 3.7835521697998047, + "learning_rate": 9.986809998681e-06, + "loss": 0.107, + "step": 13197 + }, + { + "epoch": 6.135750813575082, + "grad_norm": 4.775981903076172, + "learning_rate": 9.9868089986809e-06, + "loss": 0.0254, + "step": 13198 + }, + { + "epoch": 6.136215713621572, + "grad_norm": 6.69658899307251, + "learning_rate": 9.986807998680801e-06, + "loss": 0.2616, + "step": 13199 + }, + { + "epoch": 6.136680613668061, + "grad_norm": 3.280240774154663, + "learning_rate": 9.9868069986807e-06, + "loss": 0.0628, + "step": 13200 + }, + { + "epoch": 6.137145513714551, + "grad_norm": 4.005375862121582, + "learning_rate": 9.986805998680601e-06, + "loss": 0.1283, + "step": 13201 + }, + { + "epoch": 6.137610413761041, + "grad_norm": 7.321502208709717, + "learning_rate": 9.9868049986805e-06, + "loss": 0.2221, + "step": 13202 + }, + { + "epoch": 6.138075313807532, + "grad_norm": 3.2569901943206787, + "learning_rate": 9.9868039986804e-06, + "loss": 0.1019, + "step": 13203 + }, + { + "epoch": 6.1385402138540215, + "grad_norm": 3.589160203933716, + "learning_rate": 9.986802998680301e-06, + "loss": 0.079, + "step": 13204 + }, + { + "epoch": 6.139005113900511, + "grad_norm": 3.16213059425354, + "learning_rate": 9.986801998680202e-06, + "loss": 0.0594, + "step": 13205 + }, + { + "epoch": 6.139470013947001, + "grad_norm": 6.447509765625, + "learning_rate": 9.9868009986801e-06, + "loss": 0.2221, + "step": 13206 + }, + { + "epoch": 6.139934913993492, + "grad_norm": 5.908288478851318, + "learning_rate": 9.98679999868e-06, + "loss": 0.1283, + "step": 13207 + }, + { + "epoch": 6.140399814039982, + "grad_norm": 5.362545967102051, + "learning_rate": 9.9867989986799e-06, + "loss": 0.2241, + "step": 13208 + }, + { + "epoch": 6.140864714086471, + "grad_norm": 8.743903160095215, + "learning_rate": 9.986797998679801e-06, + "loss": 0.0983, + "step": 13209 + }, + { + "epoch": 6.141329614132961, + "grad_norm": 5.334982872009277, + "learning_rate": 9.9867969986797e-06, + "loss": 0.1355, + "step": 13210 + }, + { + "epoch": 6.141794514179451, + "grad_norm": 3.44474458694458, + "learning_rate": 9.9867959986796e-06, + "loss": 0.1299, + "step": 13211 + }, + { + "epoch": 6.142259414225942, + "grad_norm": 4.825870990753174, + "learning_rate": 9.986794998679501e-06, + "loss": 0.1367, + "step": 13212 + }, + { + "epoch": 6.1427243142724315, + "grad_norm": 12.44098949432373, + "learning_rate": 9.9867939986794e-06, + "loss": 0.145, + "step": 13213 + }, + { + "epoch": 6.143189214318921, + "grad_norm": 4.882612705230713, + "learning_rate": 9.986792998679301e-06, + "loss": 0.1672, + "step": 13214 + }, + { + "epoch": 6.143654114365411, + "grad_norm": 4.5404744148254395, + "learning_rate": 9.9867919986792e-06, + "loss": 0.113, + "step": 13215 + }, + { + "epoch": 6.144119014411902, + "grad_norm": 7.899662017822266, + "learning_rate": 9.9867909986791e-06, + "loss": 0.4933, + "step": 13216 + }, + { + "epoch": 6.144583914458392, + "grad_norm": 6.132013320922852, + "learning_rate": 9.986789998679001e-06, + "loss": 0.2799, + "step": 13217 + }, + { + "epoch": 6.145048814504881, + "grad_norm": 6.742186546325684, + "learning_rate": 9.9867889986789e-06, + "loss": 0.1803, + "step": 13218 + }, + { + "epoch": 6.145513714551371, + "grad_norm": 5.477051258087158, + "learning_rate": 9.986787998678801e-06, + "loss": 0.182, + "step": 13219 + }, + { + "epoch": 6.145978614597862, + "grad_norm": 4.50266695022583, + "learning_rate": 9.9867869986787e-06, + "loss": 0.0912, + "step": 13220 + }, + { + "epoch": 6.146443514644352, + "grad_norm": 1.8227062225341797, + "learning_rate": 9.9867859986786e-06, + "loss": 0.028, + "step": 13221 + }, + { + "epoch": 6.1469084146908415, + "grad_norm": 8.349313735961914, + "learning_rate": 9.986784998678501e-06, + "loss": 0.3891, + "step": 13222 + }, + { + "epoch": 6.147373314737331, + "grad_norm": 4.122276782989502, + "learning_rate": 9.9867839986784e-06, + "loss": 0.0882, + "step": 13223 + }, + { + "epoch": 6.147838214783821, + "grad_norm": 7.696528434753418, + "learning_rate": 9.9867829986783e-06, + "loss": 0.4724, + "step": 13224 + }, + { + "epoch": 6.148303114830312, + "grad_norm": 5.2307562828063965, + "learning_rate": 9.986781998678201e-06, + "loss": 0.1311, + "step": 13225 + }, + { + "epoch": 6.148768014876802, + "grad_norm": 6.819816589355469, + "learning_rate": 9.9867809986781e-06, + "loss": 0.1075, + "step": 13226 + }, + { + "epoch": 6.149232914923291, + "grad_norm": 3.691213369369507, + "learning_rate": 9.986779998678001e-06, + "loss": 0.1671, + "step": 13227 + }, + { + "epoch": 6.149697814969781, + "grad_norm": 3.406118154525757, + "learning_rate": 9.9867789986779e-06, + "loss": 0.0804, + "step": 13228 + }, + { + "epoch": 6.150162715016272, + "grad_norm": 5.077033996582031, + "learning_rate": 9.9867779986778e-06, + "loss": 0.101, + "step": 13229 + }, + { + "epoch": 6.150627615062762, + "grad_norm": 5.638761520385742, + "learning_rate": 9.986776998677701e-06, + "loss": 0.0692, + "step": 13230 + }, + { + "epoch": 6.1510925151092515, + "grad_norm": 4.089870452880859, + "learning_rate": 9.9867759986776e-06, + "loss": 0.0681, + "step": 13231 + }, + { + "epoch": 6.151557415155741, + "grad_norm": 4.259609222412109, + "learning_rate": 9.986774998677501e-06, + "loss": 0.1148, + "step": 13232 + }, + { + "epoch": 6.152022315202231, + "grad_norm": 2.7673561573028564, + "learning_rate": 9.9867739986774e-06, + "loss": 0.0709, + "step": 13233 + }, + { + "epoch": 6.152487215248722, + "grad_norm": 7.581044673919678, + "learning_rate": 9.9867729986773e-06, + "loss": 0.2059, + "step": 13234 + }, + { + "epoch": 6.152952115295212, + "grad_norm": 4.935514450073242, + "learning_rate": 9.986771998677201e-06, + "loss": 0.2259, + "step": 13235 + }, + { + "epoch": 6.153417015341701, + "grad_norm": 4.358657360076904, + "learning_rate": 9.986770998677102e-06, + "loss": 0.166, + "step": 13236 + }, + { + "epoch": 6.153881915388191, + "grad_norm": 5.034963607788086, + "learning_rate": 9.986769998677001e-06, + "loss": 0.1765, + "step": 13237 + }, + { + "epoch": 6.154346815434682, + "grad_norm": 7.248738765716553, + "learning_rate": 9.986768998676902e-06, + "loss": 0.2762, + "step": 13238 + }, + { + "epoch": 6.154811715481172, + "grad_norm": 5.666942596435547, + "learning_rate": 9.9867679986768e-06, + "loss": 0.1329, + "step": 13239 + }, + { + "epoch": 6.1552766155276615, + "grad_norm": 5.1463303565979, + "learning_rate": 9.9867669986767e-06, + "loss": 0.2633, + "step": 13240 + }, + { + "epoch": 6.155741515574151, + "grad_norm": 4.667153835296631, + "learning_rate": 9.9867659986766e-06, + "loss": 0.1239, + "step": 13241 + }, + { + "epoch": 6.156206415620642, + "grad_norm": 4.555191516876221, + "learning_rate": 9.9867649986765e-06, + "loss": 0.1552, + "step": 13242 + }, + { + "epoch": 6.156671315667132, + "grad_norm": 5.178398609161377, + "learning_rate": 9.986763998676401e-06, + "loss": 0.1847, + "step": 13243 + }, + { + "epoch": 6.157136215713622, + "grad_norm": 6.337597846984863, + "learning_rate": 9.9867629986763e-06, + "loss": 0.1832, + "step": 13244 + }, + { + "epoch": 6.157601115760111, + "grad_norm": 3.4654524326324463, + "learning_rate": 9.986761998676201e-06, + "loss": 0.053, + "step": 13245 + }, + { + "epoch": 6.158066015806601, + "grad_norm": 5.52045202255249, + "learning_rate": 9.9867609986761e-06, + "loss": 0.1648, + "step": 13246 + }, + { + "epoch": 6.158530915853092, + "grad_norm": 4.556800365447998, + "learning_rate": 9.986759998676e-06, + "loss": 0.0919, + "step": 13247 + }, + { + "epoch": 6.158995815899582, + "grad_norm": 5.234555721282959, + "learning_rate": 9.986758998675901e-06, + "loss": 0.1504, + "step": 13248 + }, + { + "epoch": 6.1594607159460715, + "grad_norm": 6.6499199867248535, + "learning_rate": 9.986757998675802e-06, + "loss": 0.1809, + "step": 13249 + }, + { + "epoch": 6.159925615992561, + "grad_norm": 6.4717583656311035, + "learning_rate": 9.986756998675701e-06, + "loss": 0.2908, + "step": 13250 + }, + { + "epoch": 6.160390516039052, + "grad_norm": 7.0276384353637695, + "learning_rate": 9.9867559986756e-06, + "loss": 0.1584, + "step": 13251 + }, + { + "epoch": 6.160855416085542, + "grad_norm": 10.26512622833252, + "learning_rate": 9.9867549986755e-06, + "loss": 0.0643, + "step": 13252 + }, + { + "epoch": 6.161320316132032, + "grad_norm": 6.2089409828186035, + "learning_rate": 9.9867539986754e-06, + "loss": 0.1651, + "step": 13253 + }, + { + "epoch": 6.161785216178521, + "grad_norm": 5.001736640930176, + "learning_rate": 9.9867529986753e-06, + "loss": 0.143, + "step": 13254 + }, + { + "epoch": 6.162250116225012, + "grad_norm": 4.751472473144531, + "learning_rate": 9.986751998675201e-06, + "loss": 0.2339, + "step": 13255 + }, + { + "epoch": 6.162715016271502, + "grad_norm": 5.994742393493652, + "learning_rate": 9.986750998675102e-06, + "loss": 0.1948, + "step": 13256 + }, + { + "epoch": 6.163179916317992, + "grad_norm": 4.898187160491943, + "learning_rate": 9.986749998675e-06, + "loss": 0.2258, + "step": 13257 + }, + { + "epoch": 6.1636448163644815, + "grad_norm": 5.725803375244141, + "learning_rate": 9.986748998674901e-06, + "loss": 0.2566, + "step": 13258 + }, + { + "epoch": 6.164109716410971, + "grad_norm": 5.999340057373047, + "learning_rate": 9.9867479986748e-06, + "loss": 0.2329, + "step": 13259 + }, + { + "epoch": 6.164574616457462, + "grad_norm": 5.169058322906494, + "learning_rate": 9.9867469986747e-06, + "loss": 0.2268, + "step": 13260 + }, + { + "epoch": 6.165039516503952, + "grad_norm": 3.3551864624023438, + "learning_rate": 9.986745998674601e-06, + "loss": 0.0878, + "step": 13261 + }, + { + "epoch": 6.165504416550442, + "grad_norm": 5.428296089172363, + "learning_rate": 9.9867449986745e-06, + "loss": 0.1464, + "step": 13262 + }, + { + "epoch": 6.165969316596931, + "grad_norm": 5.550655841827393, + "learning_rate": 9.986743998674401e-06, + "loss": 0.0864, + "step": 13263 + }, + { + "epoch": 6.166434216643422, + "grad_norm": 5.110546588897705, + "learning_rate": 9.9867429986743e-06, + "loss": 0.0749, + "step": 13264 + }, + { + "epoch": 6.166899116689912, + "grad_norm": 4.738039493560791, + "learning_rate": 9.9867419986742e-06, + "loss": 0.1338, + "step": 13265 + }, + { + "epoch": 6.167364016736402, + "grad_norm": 5.1554412841796875, + "learning_rate": 9.9867409986741e-06, + "loss": 0.1307, + "step": 13266 + }, + { + "epoch": 6.1678289167828915, + "grad_norm": 3.452781915664673, + "learning_rate": 9.986739998674e-06, + "loss": 0.0554, + "step": 13267 + }, + { + "epoch": 6.168293816829381, + "grad_norm": 5.934895992279053, + "learning_rate": 9.986738998673901e-06, + "loss": 0.1379, + "step": 13268 + }, + { + "epoch": 6.168758716875872, + "grad_norm": 5.211171627044678, + "learning_rate": 9.986737998673802e-06, + "loss": 0.2158, + "step": 13269 + }, + { + "epoch": 6.169223616922362, + "grad_norm": 2.408538818359375, + "learning_rate": 9.9867369986737e-06, + "loss": 0.0455, + "step": 13270 + }, + { + "epoch": 6.169688516968852, + "grad_norm": 3.800182819366455, + "learning_rate": 9.986735998673601e-06, + "loss": 0.0873, + "step": 13271 + }, + { + "epoch": 6.170153417015341, + "grad_norm": 3.481947422027588, + "learning_rate": 9.9867349986735e-06, + "loss": 0.0741, + "step": 13272 + }, + { + "epoch": 6.170618317061832, + "grad_norm": 5.276212692260742, + "learning_rate": 9.9867339986734e-06, + "loss": 0.1772, + "step": 13273 + }, + { + "epoch": 6.171083217108322, + "grad_norm": 3.6179943084716797, + "learning_rate": 9.986732998673301e-06, + "loss": 0.1152, + "step": 13274 + }, + { + "epoch": 6.171548117154812, + "grad_norm": 4.155045509338379, + "learning_rate": 9.9867319986732e-06, + "loss": 0.1407, + "step": 13275 + }, + { + "epoch": 6.1720130172013015, + "grad_norm": 2.804837226867676, + "learning_rate": 9.986730998673101e-06, + "loss": 0.0536, + "step": 13276 + }, + { + "epoch": 6.172477917247791, + "grad_norm": 4.243673324584961, + "learning_rate": 9.986729998673e-06, + "loss": 0.1291, + "step": 13277 + }, + { + "epoch": 6.172942817294282, + "grad_norm": 6.454854965209961, + "learning_rate": 9.9867289986729e-06, + "loss": 0.3438, + "step": 13278 + }, + { + "epoch": 6.173407717340772, + "grad_norm": 6.852438449859619, + "learning_rate": 9.9867279986728e-06, + "loss": 0.3071, + "step": 13279 + }, + { + "epoch": 6.173872617387262, + "grad_norm": 6.039108753204346, + "learning_rate": 9.9867269986727e-06, + "loss": 0.2023, + "step": 13280 + }, + { + "epoch": 6.174337517433751, + "grad_norm": 2.0099270343780518, + "learning_rate": 9.986725998672601e-06, + "loss": 0.0363, + "step": 13281 + }, + { + "epoch": 6.174802417480242, + "grad_norm": 4.366040229797363, + "learning_rate": 9.986724998672502e-06, + "loss": 0.1042, + "step": 13282 + }, + { + "epoch": 6.175267317526732, + "grad_norm": 3.8125336170196533, + "learning_rate": 9.9867239986724e-06, + "loss": 0.0738, + "step": 13283 + }, + { + "epoch": 6.175732217573222, + "grad_norm": 5.546273708343506, + "learning_rate": 9.9867229986723e-06, + "loss": 0.069, + "step": 13284 + }, + { + "epoch": 6.1761971176197115, + "grad_norm": 2.1802093982696533, + "learning_rate": 9.9867219986722e-06, + "loss": 0.0353, + "step": 13285 + }, + { + "epoch": 6.176662017666202, + "grad_norm": 7.168664455413818, + "learning_rate": 9.986720998672101e-06, + "loss": 0.209, + "step": 13286 + }, + { + "epoch": 6.177126917712692, + "grad_norm": 5.364066123962402, + "learning_rate": 9.986719998672002e-06, + "loss": 0.1879, + "step": 13287 + }, + { + "epoch": 6.177591817759182, + "grad_norm": 5.784873962402344, + "learning_rate": 9.9867189986719e-06, + "loss": 0.1197, + "step": 13288 + }, + { + "epoch": 6.178056717805672, + "grad_norm": 2.7583768367767334, + "learning_rate": 9.986717998671801e-06, + "loss": 0.0447, + "step": 13289 + }, + { + "epoch": 6.178521617852161, + "grad_norm": 3.143463373184204, + "learning_rate": 9.9867169986717e-06, + "loss": 0.069, + "step": 13290 + }, + { + "epoch": 6.178986517898652, + "grad_norm": 23.147626876831055, + "learning_rate": 9.9867159986716e-06, + "loss": 0.1153, + "step": 13291 + }, + { + "epoch": 6.179451417945142, + "grad_norm": 3.393604278564453, + "learning_rate": 9.9867149986715e-06, + "loss": 0.0886, + "step": 13292 + }, + { + "epoch": 6.179916317991632, + "grad_norm": 8.777888298034668, + "learning_rate": 9.9867139986714e-06, + "loss": 0.2579, + "step": 13293 + }, + { + "epoch": 6.1803812180381215, + "grad_norm": 4.628684997558594, + "learning_rate": 9.986712998671301e-06, + "loss": 0.1375, + "step": 13294 + }, + { + "epoch": 6.180846118084612, + "grad_norm": 5.51901388168335, + "learning_rate": 9.9867119986712e-06, + "loss": 0.1429, + "step": 13295 + }, + { + "epoch": 6.181311018131102, + "grad_norm": 8.494633674621582, + "learning_rate": 9.9867109986711e-06, + "loss": 0.2827, + "step": 13296 + }, + { + "epoch": 6.181775918177592, + "grad_norm": 4.57603645324707, + "learning_rate": 9.986709998671e-06, + "loss": 0.1507, + "step": 13297 + }, + { + "epoch": 6.182240818224082, + "grad_norm": 4.977685451507568, + "learning_rate": 9.9867089986709e-06, + "loss": 0.1491, + "step": 13298 + }, + { + "epoch": 6.182705718270572, + "grad_norm": 4.7476582527160645, + "learning_rate": 9.986707998670801e-06, + "loss": 0.1099, + "step": 13299 + }, + { + "epoch": 6.183170618317062, + "grad_norm": 4.779635906219482, + "learning_rate": 9.986706998670702e-06, + "loss": 0.1143, + "step": 13300 + }, + { + "epoch": 6.183635518363552, + "grad_norm": 5.898509502410889, + "learning_rate": 9.9867059986706e-06, + "loss": 0.2197, + "step": 13301 + }, + { + "epoch": 6.184100418410042, + "grad_norm": 6.16341495513916, + "learning_rate": 9.986704998670501e-06, + "loss": 0.305, + "step": 13302 + }, + { + "epoch": 6.1845653184565315, + "grad_norm": 5.5401997566223145, + "learning_rate": 9.9867039986704e-06, + "loss": 0.0849, + "step": 13303 + }, + { + "epoch": 6.185030218503022, + "grad_norm": 10.38148021697998, + "learning_rate": 9.986702998670301e-06, + "loss": 0.2792, + "step": 13304 + }, + { + "epoch": 6.185495118549512, + "grad_norm": 5.90608549118042, + "learning_rate": 9.986701998670202e-06, + "loss": 0.1918, + "step": 13305 + }, + { + "epoch": 6.185960018596002, + "grad_norm": 6.024235725402832, + "learning_rate": 9.9867009986701e-06, + "loss": 0.1829, + "step": 13306 + }, + { + "epoch": 6.1864249186424916, + "grad_norm": 6.822093486785889, + "learning_rate": 9.986699998670001e-06, + "loss": 0.1834, + "step": 13307 + }, + { + "epoch": 6.186889818688982, + "grad_norm": 3.470916986465454, + "learning_rate": 9.9866989986699e-06, + "loss": 0.1244, + "step": 13308 + }, + { + "epoch": 6.187354718735472, + "grad_norm": 5.5666046142578125, + "learning_rate": 9.9866979986698e-06, + "loss": 0.103, + "step": 13309 + }, + { + "epoch": 6.187819618781962, + "grad_norm": 6.647640705108643, + "learning_rate": 9.9866969986697e-06, + "loss": 0.2411, + "step": 13310 + }, + { + "epoch": 6.188284518828452, + "grad_norm": 8.059183120727539, + "learning_rate": 9.9866959986696e-06, + "loss": 0.3759, + "step": 13311 + }, + { + "epoch": 6.1887494188749415, + "grad_norm": 4.33790922164917, + "learning_rate": 9.986694998669501e-06, + "loss": 0.1236, + "step": 13312 + }, + { + "epoch": 6.189214318921432, + "grad_norm": 11.693316459655762, + "learning_rate": 9.986693998669402e-06, + "loss": 0.2681, + "step": 13313 + }, + { + "epoch": 6.189679218967922, + "grad_norm": 11.332210540771484, + "learning_rate": 9.9866929986693e-06, + "loss": 0.237, + "step": 13314 + }, + { + "epoch": 6.190144119014412, + "grad_norm": 2.48305082321167, + "learning_rate": 9.9866919986692e-06, + "loss": 0.0392, + "step": 13315 + }, + { + "epoch": 6.1906090190609016, + "grad_norm": 3.8107552528381348, + "learning_rate": 9.9866909986691e-06, + "loss": 0.0793, + "step": 13316 + }, + { + "epoch": 6.191073919107392, + "grad_norm": 6.614158630371094, + "learning_rate": 9.986689998669001e-06, + "loss": 0.2499, + "step": 13317 + }, + { + "epoch": 6.191538819153882, + "grad_norm": 5.218815803527832, + "learning_rate": 9.986688998668902e-06, + "loss": 0.1727, + "step": 13318 + }, + { + "epoch": 6.192003719200372, + "grad_norm": 4.082332611083984, + "learning_rate": 9.9866879986688e-06, + "loss": 0.0883, + "step": 13319 + }, + { + "epoch": 6.192468619246862, + "grad_norm": 5.773778915405273, + "learning_rate": 9.986686998668701e-06, + "loss": 0.0952, + "step": 13320 + }, + { + "epoch": 6.192933519293352, + "grad_norm": 4.2636919021606445, + "learning_rate": 9.9866859986686e-06, + "loss": 0.1611, + "step": 13321 + }, + { + "epoch": 6.193398419339842, + "grad_norm": 5.245027542114258, + "learning_rate": 9.986684998668501e-06, + "loss": 0.1013, + "step": 13322 + }, + { + "epoch": 6.193863319386332, + "grad_norm": 3.678514242172241, + "learning_rate": 9.9866839986684e-06, + "loss": 0.0594, + "step": 13323 + }, + { + "epoch": 6.194328219432822, + "grad_norm": 5.443979263305664, + "learning_rate": 9.9866829986683e-06, + "loss": 0.1743, + "step": 13324 + }, + { + "epoch": 6.1947931194793115, + "grad_norm": 3.3876442909240723, + "learning_rate": 9.986681998668201e-06, + "loss": 0.045, + "step": 13325 + }, + { + "epoch": 6.195258019525802, + "grad_norm": 5.33097505569458, + "learning_rate": 9.9866809986681e-06, + "loss": 0.2338, + "step": 13326 + }, + { + "epoch": 6.195722919572292, + "grad_norm": 3.1805999279022217, + "learning_rate": 9.986679998668e-06, + "loss": 0.0549, + "step": 13327 + }, + { + "epoch": 6.196187819618782, + "grad_norm": 4.624533653259277, + "learning_rate": 9.9866789986679e-06, + "loss": 0.1215, + "step": 13328 + }, + { + "epoch": 6.196652719665272, + "grad_norm": 2.6580264568328857, + "learning_rate": 9.9866779986678e-06, + "loss": 0.0462, + "step": 13329 + }, + { + "epoch": 6.197117619711762, + "grad_norm": 2.9760899543762207, + "learning_rate": 9.986676998667701e-06, + "loss": 0.0516, + "step": 13330 + }, + { + "epoch": 6.197582519758252, + "grad_norm": 5.511728763580322, + "learning_rate": 9.986675998667602e-06, + "loss": 0.2545, + "step": 13331 + }, + { + "epoch": 6.198047419804742, + "grad_norm": 5.782730579376221, + "learning_rate": 9.9866749986675e-06, + "loss": 0.2418, + "step": 13332 + }, + { + "epoch": 6.198512319851232, + "grad_norm": 3.7498562335968018, + "learning_rate": 9.986673998667401e-06, + "loss": 0.123, + "step": 13333 + }, + { + "epoch": 6.198977219897722, + "grad_norm": 4.6311163902282715, + "learning_rate": 9.9866729986673e-06, + "loss": 0.1293, + "step": 13334 + }, + { + "epoch": 6.199442119944212, + "grad_norm": 7.428393840789795, + "learning_rate": 9.986671998667201e-06, + "loss": 0.4284, + "step": 13335 + }, + { + "epoch": 6.199907019990702, + "grad_norm": 8.7603178024292, + "learning_rate": 9.9866709986671e-06, + "loss": 0.4864, + "step": 13336 + }, + { + "epoch": 6.200371920037192, + "grad_norm": 3.4652976989746094, + "learning_rate": 9.986669998667e-06, + "loss": 0.0786, + "step": 13337 + }, + { + "epoch": 6.200836820083682, + "grad_norm": 5.055106163024902, + "learning_rate": 9.986668998666901e-06, + "loss": 0.0758, + "step": 13338 + }, + { + "epoch": 6.201301720130172, + "grad_norm": 4.149855613708496, + "learning_rate": 9.9866679986668e-06, + "loss": 0.0978, + "step": 13339 + }, + { + "epoch": 6.201766620176662, + "grad_norm": 6.377396106719971, + "learning_rate": 9.986666998666701e-06, + "loss": 0.3141, + "step": 13340 + }, + { + "epoch": 6.202231520223152, + "grad_norm": 6.249302864074707, + "learning_rate": 9.9866659986666e-06, + "loss": 0.3173, + "step": 13341 + }, + { + "epoch": 6.202696420269642, + "grad_norm": 4.0805559158325195, + "learning_rate": 9.9866649986665e-06, + "loss": 0.0982, + "step": 13342 + }, + { + "epoch": 6.203161320316132, + "grad_norm": 10.239228248596191, + "learning_rate": 9.986663998666401e-06, + "loss": 0.2193, + "step": 13343 + }, + { + "epoch": 6.203626220362622, + "grad_norm": 4.503380298614502, + "learning_rate": 9.986662998666302e-06, + "loss": 0.1487, + "step": 13344 + }, + { + "epoch": 6.204091120409112, + "grad_norm": 2.2262444496154785, + "learning_rate": 9.9866619986662e-06, + "loss": 0.0319, + "step": 13345 + }, + { + "epoch": 6.204556020455602, + "grad_norm": 4.445829391479492, + "learning_rate": 9.986660998666101e-06, + "loss": 0.0659, + "step": 13346 + }, + { + "epoch": 6.205020920502092, + "grad_norm": 4.1921916007995605, + "learning_rate": 9.986659998666e-06, + "loss": 0.1115, + "step": 13347 + }, + { + "epoch": 6.205485820548582, + "grad_norm": 7.249804973602295, + "learning_rate": 9.9866589986659e-06, + "loss": 0.1712, + "step": 13348 + }, + { + "epoch": 6.205950720595072, + "grad_norm": 4.624195575714111, + "learning_rate": 9.9866579986658e-06, + "loss": 0.126, + "step": 13349 + }, + { + "epoch": 6.206415620641562, + "grad_norm": 6.642866134643555, + "learning_rate": 9.9866569986657e-06, + "loss": 0.248, + "step": 13350 + }, + { + "epoch": 6.206880520688052, + "grad_norm": 8.089234352111816, + "learning_rate": 9.986655998665601e-06, + "loss": 0.1959, + "step": 13351 + }, + { + "epoch": 6.207345420734542, + "grad_norm": 7.837535858154297, + "learning_rate": 9.9866549986655e-06, + "loss": 0.421, + "step": 13352 + }, + { + "epoch": 6.207810320781032, + "grad_norm": 8.32485294342041, + "learning_rate": 9.986653998665401e-06, + "loss": 0.3073, + "step": 13353 + }, + { + "epoch": 6.208275220827522, + "grad_norm": 5.308872699737549, + "learning_rate": 9.9866529986653e-06, + "loss": 0.1225, + "step": 13354 + }, + { + "epoch": 6.208740120874012, + "grad_norm": 5.8398051261901855, + "learning_rate": 9.9866519986652e-06, + "loss": 0.281, + "step": 13355 + }, + { + "epoch": 6.209205020920502, + "grad_norm": 4.216181755065918, + "learning_rate": 9.986650998665101e-06, + "loss": 0.1369, + "step": 13356 + }, + { + "epoch": 6.209669920966992, + "grad_norm": 7.3136138916015625, + "learning_rate": 9.986649998665002e-06, + "loss": 0.3797, + "step": 13357 + }, + { + "epoch": 6.210134821013482, + "grad_norm": 5.583580493927002, + "learning_rate": 9.9866489986649e-06, + "loss": 0.3132, + "step": 13358 + }, + { + "epoch": 6.210599721059972, + "grad_norm": 5.1017374992370605, + "learning_rate": 9.9866479986648e-06, + "loss": 0.1834, + "step": 13359 + }, + { + "epoch": 6.211064621106462, + "grad_norm": 8.170920372009277, + "learning_rate": 9.9866469986647e-06, + "loss": 0.218, + "step": 13360 + }, + { + "epoch": 6.211529521152952, + "grad_norm": 6.806972503662109, + "learning_rate": 9.9866459986646e-06, + "loss": 0.3971, + "step": 13361 + }, + { + "epoch": 6.211994421199442, + "grad_norm": 3.1074631214141846, + "learning_rate": 9.9866449986645e-06, + "loss": 0.0893, + "step": 13362 + }, + { + "epoch": 6.212459321245932, + "grad_norm": 4.320138931274414, + "learning_rate": 9.9866439986644e-06, + "loss": 0.2208, + "step": 13363 + }, + { + "epoch": 6.212924221292422, + "grad_norm": 1.9753365516662598, + "learning_rate": 9.986642998664301e-06, + "loss": 0.041, + "step": 13364 + }, + { + "epoch": 6.2133891213389125, + "grad_norm": 9.632226943969727, + "learning_rate": 9.9866419986642e-06, + "loss": 0.2033, + "step": 13365 + }, + { + "epoch": 6.213854021385402, + "grad_norm": 6.307525157928467, + "learning_rate": 9.986640998664101e-06, + "loss": 0.1368, + "step": 13366 + }, + { + "epoch": 6.214318921431892, + "grad_norm": 3.9106814861297607, + "learning_rate": 9.986639998664e-06, + "loss": 0.1109, + "step": 13367 + }, + { + "epoch": 6.214783821478382, + "grad_norm": 3.9809911251068115, + "learning_rate": 9.9866389986639e-06, + "loss": 0.2039, + "step": 13368 + }, + { + "epoch": 6.215248721524872, + "grad_norm": 6.5070295333862305, + "learning_rate": 9.986637998663801e-06, + "loss": 0.1792, + "step": 13369 + }, + { + "epoch": 6.215713621571362, + "grad_norm": 5.792603492736816, + "learning_rate": 9.9866369986637e-06, + "loss": 0.2431, + "step": 13370 + }, + { + "epoch": 6.216178521617852, + "grad_norm": 5.075913906097412, + "learning_rate": 9.986635998663601e-06, + "loss": 0.1087, + "step": 13371 + }, + { + "epoch": 6.216643421664342, + "grad_norm": 6.401622295379639, + "learning_rate": 9.9866349986635e-06, + "loss": 0.1703, + "step": 13372 + }, + { + "epoch": 6.217108321710832, + "grad_norm": 4.413451671600342, + "learning_rate": 9.9866339986634e-06, + "loss": 0.1031, + "step": 13373 + }, + { + "epoch": 6.2175732217573225, + "grad_norm": 4.482436656951904, + "learning_rate": 9.986632998663301e-06, + "loss": 0.1435, + "step": 13374 + }, + { + "epoch": 6.218038121803812, + "grad_norm": 3.3624837398529053, + "learning_rate": 9.9866319986632e-06, + "loss": 0.1049, + "step": 13375 + }, + { + "epoch": 6.218503021850302, + "grad_norm": 3.3504691123962402, + "learning_rate": 9.9866309986631e-06, + "loss": 0.0842, + "step": 13376 + }, + { + "epoch": 6.218967921896792, + "grad_norm": 6.358952522277832, + "learning_rate": 9.986629998663001e-06, + "loss": 0.2042, + "step": 13377 + }, + { + "epoch": 6.219432821943283, + "grad_norm": 4.722170829772949, + "learning_rate": 9.9866289986629e-06, + "loss": 0.171, + "step": 13378 + }, + { + "epoch": 6.219897721989772, + "grad_norm": 5.334135055541992, + "learning_rate": 9.986627998662801e-06, + "loss": 0.1847, + "step": 13379 + }, + { + "epoch": 6.220362622036262, + "grad_norm": 5.33033561706543, + "learning_rate": 9.9866269986627e-06, + "loss": 0.1517, + "step": 13380 + }, + { + "epoch": 6.220827522082752, + "grad_norm": 2.2433202266693115, + "learning_rate": 9.9866259986626e-06, + "loss": 0.0378, + "step": 13381 + }, + { + "epoch": 6.221292422129242, + "grad_norm": 2.468580722808838, + "learning_rate": 9.986624998662501e-06, + "loss": 0.0732, + "step": 13382 + }, + { + "epoch": 6.2217573221757325, + "grad_norm": 5.241724491119385, + "learning_rate": 9.9866239986624e-06, + "loss": 0.1875, + "step": 13383 + }, + { + "epoch": 6.222222222222222, + "grad_norm": 5.740681171417236, + "learning_rate": 9.986622998662301e-06, + "loss": 0.1913, + "step": 13384 + }, + { + "epoch": 6.222687122268712, + "grad_norm": 4.188978672027588, + "learning_rate": 9.9866219986622e-06, + "loss": 0.1196, + "step": 13385 + }, + { + "epoch": 6.223152022315202, + "grad_norm": 6.779123306274414, + "learning_rate": 9.9866209986621e-06, + "loss": 0.0719, + "step": 13386 + }, + { + "epoch": 6.223616922361693, + "grad_norm": 6.499244213104248, + "learning_rate": 9.986619998662001e-06, + "loss": 0.22, + "step": 13387 + }, + { + "epoch": 6.224081822408182, + "grad_norm": 3.791806221008301, + "learning_rate": 9.986618998661902e-06, + "loss": 0.0877, + "step": 13388 + }, + { + "epoch": 6.224546722454672, + "grad_norm": 3.8381733894348145, + "learning_rate": 9.986617998661801e-06, + "loss": 0.0919, + "step": 13389 + }, + { + "epoch": 6.225011622501162, + "grad_norm": 8.093714714050293, + "learning_rate": 9.986616998661702e-06, + "loss": 0.2163, + "step": 13390 + }, + { + "epoch": 6.225476522547652, + "grad_norm": 8.167662620544434, + "learning_rate": 9.9866159986616e-06, + "loss": 0.4441, + "step": 13391 + }, + { + "epoch": 6.2259414225941425, + "grad_norm": 7.165520668029785, + "learning_rate": 9.9866149986615e-06, + "loss": 0.1622, + "step": 13392 + }, + { + "epoch": 6.226406322640632, + "grad_norm": 8.237751007080078, + "learning_rate": 9.9866139986614e-06, + "loss": 0.3991, + "step": 13393 + }, + { + "epoch": 6.226871222687122, + "grad_norm": 1.0884544849395752, + "learning_rate": 9.9866129986613e-06, + "loss": 0.0151, + "step": 13394 + }, + { + "epoch": 6.227336122733612, + "grad_norm": 5.30825138092041, + "learning_rate": 9.986611998661201e-06, + "loss": 0.1764, + "step": 13395 + }, + { + "epoch": 6.227801022780103, + "grad_norm": 2.9794869422912598, + "learning_rate": 9.9866109986611e-06, + "loss": 0.0484, + "step": 13396 + }, + { + "epoch": 6.228265922826592, + "grad_norm": 2.680201292037964, + "learning_rate": 9.986609998661001e-06, + "loss": 0.0896, + "step": 13397 + }, + { + "epoch": 6.228730822873082, + "grad_norm": 5.946937084197998, + "learning_rate": 9.9866089986609e-06, + "loss": 0.309, + "step": 13398 + }, + { + "epoch": 6.229195722919572, + "grad_norm": 3.826225996017456, + "learning_rate": 9.9866079986608e-06, + "loss": 0.1048, + "step": 13399 + }, + { + "epoch": 6.229660622966063, + "grad_norm": 6.461924076080322, + "learning_rate": 9.986606998660701e-06, + "loss": 0.2533, + "step": 13400 + }, + { + "epoch": 6.2301255230125525, + "grad_norm": 5.7256879806518555, + "learning_rate": 9.986605998660602e-06, + "loss": 0.2124, + "step": 13401 + }, + { + "epoch": 6.230590423059042, + "grad_norm": 4.547878742218018, + "learning_rate": 9.986604998660501e-06, + "loss": 0.1127, + "step": 13402 + }, + { + "epoch": 6.231055323105532, + "grad_norm": 3.486158609390259, + "learning_rate": 9.9866039986604e-06, + "loss": 0.0479, + "step": 13403 + }, + { + "epoch": 6.231520223152022, + "grad_norm": 3.0145835876464844, + "learning_rate": 9.9866029986603e-06, + "loss": 0.0631, + "step": 13404 + }, + { + "epoch": 6.231985123198513, + "grad_norm": 8.540023803710938, + "learning_rate": 9.9866019986602e-06, + "loss": 0.3549, + "step": 13405 + }, + { + "epoch": 6.232450023245002, + "grad_norm": 4.436973571777344, + "learning_rate": 9.9866009986601e-06, + "loss": 0.1877, + "step": 13406 + }, + { + "epoch": 6.232914923291492, + "grad_norm": 5.456325054168701, + "learning_rate": 9.986599998660001e-06, + "loss": 0.1016, + "step": 13407 + }, + { + "epoch": 6.233379823337982, + "grad_norm": 5.544440746307373, + "learning_rate": 9.986598998659902e-06, + "loss": 0.1725, + "step": 13408 + }, + { + "epoch": 6.233844723384473, + "grad_norm": 2.970773220062256, + "learning_rate": 9.9865979986598e-06, + "loss": 0.1082, + "step": 13409 + }, + { + "epoch": 6.2343096234309625, + "grad_norm": 3.16457462310791, + "learning_rate": 9.986596998659701e-06, + "loss": 0.0585, + "step": 13410 + }, + { + "epoch": 6.234774523477452, + "grad_norm": 8.287104606628418, + "learning_rate": 9.9865959986596e-06, + "loss": 0.2432, + "step": 13411 + }, + { + "epoch": 6.235239423523942, + "grad_norm": 3.5862178802490234, + "learning_rate": 9.9865949986595e-06, + "loss": 0.0788, + "step": 13412 + }, + { + "epoch": 6.235704323570433, + "grad_norm": 5.516677379608154, + "learning_rate": 9.986593998659401e-06, + "loss": 0.2109, + "step": 13413 + }, + { + "epoch": 6.236169223616923, + "grad_norm": 6.364782333374023, + "learning_rate": 9.9865929986593e-06, + "loss": 0.319, + "step": 13414 + }, + { + "epoch": 6.236634123663412, + "grad_norm": 16.735218048095703, + "learning_rate": 9.986591998659201e-06, + "loss": 0.7949, + "step": 13415 + }, + { + "epoch": 6.237099023709902, + "grad_norm": 4.805080890655518, + "learning_rate": 9.9865909986591e-06, + "loss": 0.0995, + "step": 13416 + }, + { + "epoch": 6.237563923756392, + "grad_norm": 5.779761791229248, + "learning_rate": 9.986589998659e-06, + "loss": 0.2714, + "step": 13417 + }, + { + "epoch": 6.238028823802883, + "grad_norm": 5.581011772155762, + "learning_rate": 9.9865889986589e-06, + "loss": 0.2066, + "step": 13418 + }, + { + "epoch": 6.2384937238493725, + "grad_norm": 10.638533592224121, + "learning_rate": 9.9865879986588e-06, + "loss": 0.4166, + "step": 13419 + }, + { + "epoch": 6.238958623895862, + "grad_norm": 2.182208776473999, + "learning_rate": 9.986586998658701e-06, + "loss": 0.0283, + "step": 13420 + }, + { + "epoch": 6.239423523942352, + "grad_norm": 3.327502727508545, + "learning_rate": 9.986585998658602e-06, + "loss": 0.0598, + "step": 13421 + }, + { + "epoch": 6.239888423988843, + "grad_norm": 8.015003204345703, + "learning_rate": 9.9865849986585e-06, + "loss": 0.4516, + "step": 13422 + }, + { + "epoch": 6.240353324035333, + "grad_norm": 2.2872231006622314, + "learning_rate": 9.986583998658401e-06, + "loss": 0.0377, + "step": 13423 + }, + { + "epoch": 6.240818224081822, + "grad_norm": 6.427635669708252, + "learning_rate": 9.9865829986583e-06, + "loss": 0.2028, + "step": 13424 + }, + { + "epoch": 6.241283124128312, + "grad_norm": 5.128576755523682, + "learning_rate": 9.986581998658201e-06, + "loss": 0.1613, + "step": 13425 + }, + { + "epoch": 6.241748024174802, + "grad_norm": 6.638921737670898, + "learning_rate": 9.986580998658102e-06, + "loss": 0.2783, + "step": 13426 + }, + { + "epoch": 6.242212924221293, + "grad_norm": 5.8597941398620605, + "learning_rate": 9.986579998658e-06, + "loss": 0.114, + "step": 13427 + }, + { + "epoch": 6.2426778242677825, + "grad_norm": 7.592766284942627, + "learning_rate": 9.986578998657901e-06, + "loss": 0.598, + "step": 13428 + }, + { + "epoch": 6.243142724314272, + "grad_norm": 4.865126609802246, + "learning_rate": 9.9865779986578e-06, + "loss": 0.1702, + "step": 13429 + }, + { + "epoch": 6.243607624360762, + "grad_norm": 7.5388383865356445, + "learning_rate": 9.9865769986577e-06, + "loss": 0.186, + "step": 13430 + }, + { + "epoch": 6.244072524407253, + "grad_norm": 10.063531875610352, + "learning_rate": 9.9865759986576e-06, + "loss": 0.1539, + "step": 13431 + }, + { + "epoch": 6.244537424453743, + "grad_norm": 6.460005760192871, + "learning_rate": 9.9865749986575e-06, + "loss": 0.1928, + "step": 13432 + }, + { + "epoch": 6.245002324500232, + "grad_norm": 4.5533928871154785, + "learning_rate": 9.986573998657401e-06, + "loss": 0.1063, + "step": 13433 + }, + { + "epoch": 6.245467224546722, + "grad_norm": 2.2123842239379883, + "learning_rate": 9.986572998657302e-06, + "loss": 0.0476, + "step": 13434 + }, + { + "epoch": 6.245932124593212, + "grad_norm": 3.5539441108703613, + "learning_rate": 9.9865719986572e-06, + "loss": 0.0843, + "step": 13435 + }, + { + "epoch": 6.246397024639703, + "grad_norm": 3.354921340942383, + "learning_rate": 9.9865709986571e-06, + "loss": 0.1177, + "step": 13436 + }, + { + "epoch": 6.2468619246861925, + "grad_norm": 3.8930230140686035, + "learning_rate": 9.986569998657e-06, + "loss": 0.1527, + "step": 13437 + }, + { + "epoch": 6.247326824732682, + "grad_norm": 7.376802921295166, + "learning_rate": 9.986568998656901e-06, + "loss": 0.4976, + "step": 13438 + }, + { + "epoch": 6.247791724779172, + "grad_norm": 4.031066417694092, + "learning_rate": 9.986567998656802e-06, + "loss": 0.0968, + "step": 13439 + }, + { + "epoch": 6.248256624825663, + "grad_norm": 6.822312831878662, + "learning_rate": 9.9865669986567e-06, + "loss": 0.1639, + "step": 13440 + }, + { + "epoch": 6.248721524872153, + "grad_norm": 4.408595085144043, + "learning_rate": 9.986565998656601e-06, + "loss": 0.191, + "step": 13441 + }, + { + "epoch": 6.249186424918642, + "grad_norm": 4.375720024108887, + "learning_rate": 9.9865649986565e-06, + "loss": 0.0739, + "step": 13442 + }, + { + "epoch": 6.249651324965132, + "grad_norm": 3.4069976806640625, + "learning_rate": 9.986563998656401e-06, + "loss": 0.0964, + "step": 13443 + }, + { + "epoch": 6.250116225011623, + "grad_norm": 2.9462451934814453, + "learning_rate": 9.9865629986563e-06, + "loss": 0.068, + "step": 13444 + }, + { + "epoch": 6.250581125058113, + "grad_norm": 3.348870277404785, + "learning_rate": 9.9865619986562e-06, + "loss": 0.0851, + "step": 13445 + }, + { + "epoch": 6.2510460251046025, + "grad_norm": 4.682251453399658, + "learning_rate": 9.986560998656101e-06, + "loss": 0.1516, + "step": 13446 + }, + { + "epoch": 6.251510925151092, + "grad_norm": 3.892423629760742, + "learning_rate": 9.986559998656e-06, + "loss": 0.1524, + "step": 13447 + }, + { + "epoch": 6.251975825197582, + "grad_norm": 6.846324443817139, + "learning_rate": 9.9865589986559e-06, + "loss": 0.2469, + "step": 13448 + }, + { + "epoch": 6.252440725244073, + "grad_norm": 7.134050369262695, + "learning_rate": 9.9865579986558e-06, + "loss": 0.2555, + "step": 13449 + }, + { + "epoch": 6.252905625290563, + "grad_norm": 6.116449356079102, + "learning_rate": 9.9865569986557e-06, + "loss": 0.2272, + "step": 13450 + }, + { + "epoch": 6.253370525337052, + "grad_norm": 7.704119682312012, + "learning_rate": 9.986555998655601e-06, + "loss": 0.2772, + "step": 13451 + }, + { + "epoch": 6.253835425383542, + "grad_norm": 4.365331172943115, + "learning_rate": 9.986554998655502e-06, + "loss": 0.1656, + "step": 13452 + }, + { + "epoch": 6.254300325430033, + "grad_norm": 4.981770038604736, + "learning_rate": 9.9865539986554e-06, + "loss": 0.1466, + "step": 13453 + }, + { + "epoch": 6.254765225476523, + "grad_norm": 4.106132984161377, + "learning_rate": 9.986552998655301e-06, + "loss": 0.1209, + "step": 13454 + }, + { + "epoch": 6.2552301255230125, + "grad_norm": 5.223456382751465, + "learning_rate": 9.9865519986552e-06, + "loss": 0.2207, + "step": 13455 + }, + { + "epoch": 6.255695025569502, + "grad_norm": 3.9357526302337646, + "learning_rate": 9.986550998655101e-06, + "loss": 0.0991, + "step": 13456 + }, + { + "epoch": 6.256159925615993, + "grad_norm": 6.464926719665527, + "learning_rate": 9.986549998655002e-06, + "loss": 0.188, + "step": 13457 + }, + { + "epoch": 6.256624825662483, + "grad_norm": 5.429508686065674, + "learning_rate": 9.9865489986549e-06, + "loss": 0.1884, + "step": 13458 + }, + { + "epoch": 6.257089725708973, + "grad_norm": 3.957267999649048, + "learning_rate": 9.986547998654801e-06, + "loss": 0.0971, + "step": 13459 + }, + { + "epoch": 6.257554625755462, + "grad_norm": 7.451467037200928, + "learning_rate": 9.9865469986547e-06, + "loss": 0.2938, + "step": 13460 + }, + { + "epoch": 6.258019525801952, + "grad_norm": 5.8767595291137695, + "learning_rate": 9.9865459986546e-06, + "loss": 0.0678, + "step": 13461 + }, + { + "epoch": 6.258484425848443, + "grad_norm": 5.8254170417785645, + "learning_rate": 9.9865449986545e-06, + "loss": 0.2464, + "step": 13462 + }, + { + "epoch": 6.258949325894933, + "grad_norm": 7.7581071853637695, + "learning_rate": 9.9865439986544e-06, + "loss": 0.2624, + "step": 13463 + }, + { + "epoch": 6.2594142259414225, + "grad_norm": 2.9721503257751465, + "learning_rate": 9.986542998654301e-06, + "loss": 0.0924, + "step": 13464 + }, + { + "epoch": 6.259879125987912, + "grad_norm": 4.444208145141602, + "learning_rate": 9.986541998654202e-06, + "loss": 0.1513, + "step": 13465 + }, + { + "epoch": 6.260344026034403, + "grad_norm": 5.937469482421875, + "learning_rate": 9.9865409986541e-06, + "loss": 0.166, + "step": 13466 + }, + { + "epoch": 6.260808926080893, + "grad_norm": 4.625209331512451, + "learning_rate": 9.986539998654e-06, + "loss": 0.1977, + "step": 13467 + }, + { + "epoch": 6.261273826127383, + "grad_norm": 3.539081573486328, + "learning_rate": 9.9865389986539e-06, + "loss": 0.1036, + "step": 13468 + }, + { + "epoch": 6.261738726173872, + "grad_norm": 5.68717622756958, + "learning_rate": 9.986537998653801e-06, + "loss": 0.0876, + "step": 13469 + }, + { + "epoch": 6.262203626220362, + "grad_norm": 5.522770404815674, + "learning_rate": 9.986536998653702e-06, + "loss": 0.2335, + "step": 13470 + }, + { + "epoch": 6.262668526266853, + "grad_norm": 5.1403021812438965, + "learning_rate": 9.9865359986536e-06, + "loss": 0.1365, + "step": 13471 + }, + { + "epoch": 6.263133426313343, + "grad_norm": 3.9500858783721924, + "learning_rate": 9.986534998653501e-06, + "loss": 0.0717, + "step": 13472 + }, + { + "epoch": 6.2635983263598325, + "grad_norm": 7.869617938995361, + "learning_rate": 9.9865339986534e-06, + "loss": 0.3798, + "step": 13473 + }, + { + "epoch": 6.264063226406322, + "grad_norm": 5.755587100982666, + "learning_rate": 9.986532998653301e-06, + "loss": 0.2211, + "step": 13474 + }, + { + "epoch": 6.264528126452813, + "grad_norm": 6.232015132904053, + "learning_rate": 9.9865319986532e-06, + "loss": 0.1269, + "step": 13475 + }, + { + "epoch": 6.264993026499303, + "grad_norm": 8.819134712219238, + "learning_rate": 9.9865309986531e-06, + "loss": 0.2765, + "step": 13476 + }, + { + "epoch": 6.265457926545793, + "grad_norm": 3.061758518218994, + "learning_rate": 9.986529998653001e-06, + "loss": 0.1063, + "step": 13477 + }, + { + "epoch": 6.265922826592282, + "grad_norm": 6.379294395446777, + "learning_rate": 9.9865289986529e-06, + "loss": 0.1853, + "step": 13478 + }, + { + "epoch": 6.266387726638772, + "grad_norm": 3.2461726665496826, + "learning_rate": 9.9865279986528e-06, + "loss": 0.0399, + "step": 13479 + }, + { + "epoch": 6.266852626685263, + "grad_norm": 4.3581671714782715, + "learning_rate": 9.9865269986527e-06, + "loss": 0.0852, + "step": 13480 + }, + { + "epoch": 6.267317526731753, + "grad_norm": 8.187448501586914, + "learning_rate": 9.9865259986526e-06, + "loss": 0.1371, + "step": 13481 + }, + { + "epoch": 6.2677824267782425, + "grad_norm": 9.892626762390137, + "learning_rate": 9.986524998652501e-06, + "loss": 0.2257, + "step": 13482 + }, + { + "epoch": 6.268247326824732, + "grad_norm": 4.097274303436279, + "learning_rate": 9.986523998652402e-06, + "loss": 0.0866, + "step": 13483 + }, + { + "epoch": 6.268712226871223, + "grad_norm": 5.382351875305176, + "learning_rate": 9.9865229986523e-06, + "loss": 0.1538, + "step": 13484 + }, + { + "epoch": 6.269177126917713, + "grad_norm": 2.7701237201690674, + "learning_rate": 9.986521998652201e-06, + "loss": 0.0562, + "step": 13485 + }, + { + "epoch": 6.269642026964203, + "grad_norm": 4.4751458168029785, + "learning_rate": 9.9865209986521e-06, + "loss": 0.1303, + "step": 13486 + }, + { + "epoch": 6.270106927010692, + "grad_norm": 4.480958938598633, + "learning_rate": 9.986519998652001e-06, + "loss": 0.0754, + "step": 13487 + }, + { + "epoch": 6.270571827057183, + "grad_norm": 4.7218146324157715, + "learning_rate": 9.9865189986519e-06, + "loss": 0.1371, + "step": 13488 + }, + { + "epoch": 6.271036727103673, + "grad_norm": 4.438849925994873, + "learning_rate": 9.9865179986518e-06, + "loss": 0.1065, + "step": 13489 + }, + { + "epoch": 6.271501627150163, + "grad_norm": 5.659519195556641, + "learning_rate": 9.986516998651701e-06, + "loss": 0.2329, + "step": 13490 + }, + { + "epoch": 6.2719665271966525, + "grad_norm": 3.0722644329071045, + "learning_rate": 9.9865159986516e-06, + "loss": 0.0485, + "step": 13491 + }, + { + "epoch": 6.272431427243143, + "grad_norm": 5.942834854125977, + "learning_rate": 9.986514998651501e-06, + "loss": 0.1304, + "step": 13492 + }, + { + "epoch": 6.272896327289633, + "grad_norm": 15.109264373779297, + "learning_rate": 9.9865139986514e-06, + "loss": 0.0952, + "step": 13493 + }, + { + "epoch": 6.273361227336123, + "grad_norm": 6.720978260040283, + "learning_rate": 9.9865129986513e-06, + "loss": 0.224, + "step": 13494 + }, + { + "epoch": 6.273826127382613, + "grad_norm": 3.5952844619750977, + "learning_rate": 9.986511998651201e-06, + "loss": 0.0623, + "step": 13495 + }, + { + "epoch": 6.274291027429102, + "grad_norm": 7.213622093200684, + "learning_rate": 9.986510998651102e-06, + "loss": 0.1635, + "step": 13496 + }, + { + "epoch": 6.274755927475593, + "grad_norm": 4.60811710357666, + "learning_rate": 9.986509998651e-06, + "loss": 0.1562, + "step": 13497 + }, + { + "epoch": 6.275220827522083, + "grad_norm": 5.242891311645508, + "learning_rate": 9.986508998650901e-06, + "loss": 0.201, + "step": 13498 + }, + { + "epoch": 6.275685727568573, + "grad_norm": 7.175049304962158, + "learning_rate": 9.9865079986508e-06, + "loss": 0.3646, + "step": 13499 + }, + { + "epoch": 6.2761506276150625, + "grad_norm": 5.0823140144348145, + "learning_rate": 9.9865069986507e-06, + "loss": 0.1057, + "step": 13500 + }, + { + "epoch": 6.276615527661553, + "grad_norm": 5.573178768157959, + "learning_rate": 9.9865059986506e-06, + "loss": 0.1586, + "step": 13501 + }, + { + "epoch": 6.277080427708043, + "grad_norm": 4.761231422424316, + "learning_rate": 9.9865049986505e-06, + "loss": 0.074, + "step": 13502 + }, + { + "epoch": 6.277545327754533, + "grad_norm": 8.763131141662598, + "learning_rate": 9.986503998650401e-06, + "loss": 0.1749, + "step": 13503 + }, + { + "epoch": 6.278010227801023, + "grad_norm": 4.337854385375977, + "learning_rate": 9.9865029986503e-06, + "loss": 0.1752, + "step": 13504 + }, + { + "epoch": 6.278475127847512, + "grad_norm": 5.839339256286621, + "learning_rate": 9.986501998650201e-06, + "loss": 0.1573, + "step": 13505 + }, + { + "epoch": 6.278940027894003, + "grad_norm": 5.989346504211426, + "learning_rate": 9.9865009986501e-06, + "loss": 0.2782, + "step": 13506 + }, + { + "epoch": 6.279404927940493, + "grad_norm": 4.208798408508301, + "learning_rate": 9.98649999865e-06, + "loss": 0.1025, + "step": 13507 + }, + { + "epoch": 6.279869827986983, + "grad_norm": 7.869119167327881, + "learning_rate": 9.986498998649901e-06, + "loss": 0.2266, + "step": 13508 + }, + { + "epoch": 6.2803347280334725, + "grad_norm": 5.871124744415283, + "learning_rate": 9.986497998649802e-06, + "loss": 0.1925, + "step": 13509 + }, + { + "epoch": 6.280799628079963, + "grad_norm": 5.513876438140869, + "learning_rate": 9.986496998649701e-06, + "loss": 0.2636, + "step": 13510 + }, + { + "epoch": 6.281264528126453, + "grad_norm": 2.1865930557250977, + "learning_rate": 9.9864959986496e-06, + "loss": 0.0634, + "step": 13511 + }, + { + "epoch": 6.281729428172943, + "grad_norm": 4.678464889526367, + "learning_rate": 9.9864949986495e-06, + "loss": 0.1718, + "step": 13512 + }, + { + "epoch": 6.282194328219433, + "grad_norm": 5.234842300415039, + "learning_rate": 9.9864939986494e-06, + "loss": 0.1293, + "step": 13513 + }, + { + "epoch": 6.282659228265922, + "grad_norm": 4.254791259765625, + "learning_rate": 9.9864929986493e-06, + "loss": 0.1717, + "step": 13514 + }, + { + "epoch": 6.283124128312413, + "grad_norm": 4.019006729125977, + "learning_rate": 9.9864919986492e-06, + "loss": 0.094, + "step": 13515 + }, + { + "epoch": 6.283589028358903, + "grad_norm": 9.295367240905762, + "learning_rate": 9.986490998649101e-06, + "loss": 0.2804, + "step": 13516 + }, + { + "epoch": 6.284053928405393, + "grad_norm": 6.914929389953613, + "learning_rate": 9.986489998649e-06, + "loss": 0.3505, + "step": 13517 + }, + { + "epoch": 6.2845188284518825, + "grad_norm": 4.854269504547119, + "learning_rate": 9.986488998648901e-06, + "loss": 0.2132, + "step": 13518 + }, + { + "epoch": 6.284983728498373, + "grad_norm": 3.64623761177063, + "learning_rate": 9.9864879986488e-06, + "loss": 0.086, + "step": 13519 + }, + { + "epoch": 6.285448628544863, + "grad_norm": 6.408176422119141, + "learning_rate": 9.9864869986487e-06, + "loss": 0.2953, + "step": 13520 + }, + { + "epoch": 6.285913528591353, + "grad_norm": 5.1396589279174805, + "learning_rate": 9.986485998648601e-06, + "loss": 0.1395, + "step": 13521 + }, + { + "epoch": 6.286378428637843, + "grad_norm": 5.61399507522583, + "learning_rate": 9.9864849986485e-06, + "loss": 0.2538, + "step": 13522 + }, + { + "epoch": 6.286843328684333, + "grad_norm": 5.674767017364502, + "learning_rate": 9.986483998648401e-06, + "loss": 0.2142, + "step": 13523 + }, + { + "epoch": 6.287308228730823, + "grad_norm": 6.053715229034424, + "learning_rate": 9.9864829986483e-06, + "loss": 0.0992, + "step": 13524 + }, + { + "epoch": 6.287773128777313, + "grad_norm": 5.251662254333496, + "learning_rate": 9.9864819986482e-06, + "loss": 0.1952, + "step": 13525 + }, + { + "epoch": 6.288238028823803, + "grad_norm": 5.805206775665283, + "learning_rate": 9.9864809986481e-06, + "loss": 0.3012, + "step": 13526 + }, + { + "epoch": 6.2887029288702925, + "grad_norm": 7.740405559539795, + "learning_rate": 9.986479998648e-06, + "loss": 0.118, + "step": 13527 + }, + { + "epoch": 6.289167828916783, + "grad_norm": 2.1529994010925293, + "learning_rate": 9.986478998647901e-06, + "loss": 0.0376, + "step": 13528 + }, + { + "epoch": 6.289632728963273, + "grad_norm": 6.216892242431641, + "learning_rate": 9.986477998647802e-06, + "loss": 0.281, + "step": 13529 + }, + { + "epoch": 6.290097629009763, + "grad_norm": 3.9519290924072266, + "learning_rate": 9.9864769986477e-06, + "loss": 0.1861, + "step": 13530 + }, + { + "epoch": 6.290562529056253, + "grad_norm": 5.657161235809326, + "learning_rate": 9.986475998647601e-06, + "loss": 0.2081, + "step": 13531 + }, + { + "epoch": 6.291027429102743, + "grad_norm": 7.705247402191162, + "learning_rate": 9.9864749986475e-06, + "loss": 0.1963, + "step": 13532 + }, + { + "epoch": 6.291492329149233, + "grad_norm": 5.366873264312744, + "learning_rate": 9.9864739986474e-06, + "loss": 0.1607, + "step": 13533 + }, + { + "epoch": 6.291957229195723, + "grad_norm": 6.507511138916016, + "learning_rate": 9.986472998647301e-06, + "loss": 0.3794, + "step": 13534 + }, + { + "epoch": 6.292422129242213, + "grad_norm": 6.949885368347168, + "learning_rate": 9.9864719986472e-06, + "loss": 0.2427, + "step": 13535 + }, + { + "epoch": 6.292887029288703, + "grad_norm": 5.210314750671387, + "learning_rate": 9.986470998647101e-06, + "loss": 0.1543, + "step": 13536 + }, + { + "epoch": 6.293351929335193, + "grad_norm": 6.768825531005859, + "learning_rate": 9.986469998647e-06, + "loss": 0.1876, + "step": 13537 + }, + { + "epoch": 6.293816829381683, + "grad_norm": 5.006449222564697, + "learning_rate": 9.9864689986469e-06, + "loss": 0.146, + "step": 13538 + }, + { + "epoch": 6.294281729428173, + "grad_norm": 5.197455883026123, + "learning_rate": 9.986467998646801e-06, + "loss": 0.1345, + "step": 13539 + }, + { + "epoch": 6.294746629474663, + "grad_norm": 1.7350943088531494, + "learning_rate": 9.9864669986467e-06, + "loss": 0.0206, + "step": 13540 + }, + { + "epoch": 6.295211529521153, + "grad_norm": 6.89002799987793, + "learning_rate": 9.986465998646601e-06, + "loss": 0.2585, + "step": 13541 + }, + { + "epoch": 6.295676429567643, + "grad_norm": 6.317755222320557, + "learning_rate": 9.986464998646502e-06, + "loss": 0.2763, + "step": 13542 + }, + { + "epoch": 6.296141329614133, + "grad_norm": 3.724987268447876, + "learning_rate": 9.9864639986464e-06, + "loss": 0.0922, + "step": 13543 + }, + { + "epoch": 6.296606229660623, + "grad_norm": 4.7444915771484375, + "learning_rate": 9.9864629986463e-06, + "loss": 0.1355, + "step": 13544 + }, + { + "epoch": 6.297071129707113, + "grad_norm": 5.62815523147583, + "learning_rate": 9.9864619986462e-06, + "loss": 0.2089, + "step": 13545 + }, + { + "epoch": 6.297536029753603, + "grad_norm": 11.71494197845459, + "learning_rate": 9.9864609986461e-06, + "loss": 0.3255, + "step": 13546 + }, + { + "epoch": 6.298000929800093, + "grad_norm": 5.555407524108887, + "learning_rate": 9.986459998646001e-06, + "loss": 0.1314, + "step": 13547 + }, + { + "epoch": 6.298465829846583, + "grad_norm": 6.292297840118408, + "learning_rate": 9.9864589986459e-06, + "loss": 0.1303, + "step": 13548 + }, + { + "epoch": 6.298930729893073, + "grad_norm": 4.349380016326904, + "learning_rate": 9.986457998645801e-06, + "loss": 0.1112, + "step": 13549 + }, + { + "epoch": 6.299395629939563, + "grad_norm": 3.262012481689453, + "learning_rate": 9.9864569986457e-06, + "loss": 0.0418, + "step": 13550 + }, + { + "epoch": 6.299860529986053, + "grad_norm": 5.407992839813232, + "learning_rate": 9.9864559986456e-06, + "loss": 0.201, + "step": 13551 + }, + { + "epoch": 6.300325430032543, + "grad_norm": 4.128430366516113, + "learning_rate": 9.986454998645501e-06, + "loss": 0.1239, + "step": 13552 + }, + { + "epoch": 6.300790330079033, + "grad_norm": 5.188741683959961, + "learning_rate": 9.986453998645402e-06, + "loss": 0.139, + "step": 13553 + }, + { + "epoch": 6.301255230125523, + "grad_norm": 5.5715250968933105, + "learning_rate": 9.986452998645301e-06, + "loss": 0.1164, + "step": 13554 + }, + { + "epoch": 6.301720130172013, + "grad_norm": 5.344173908233643, + "learning_rate": 9.9864519986452e-06, + "loss": 0.1114, + "step": 13555 + }, + { + "epoch": 6.302185030218503, + "grad_norm": 4.9710893630981445, + "learning_rate": 9.9864509986451e-06, + "loss": 0.2114, + "step": 13556 + }, + { + "epoch": 6.302649930264993, + "grad_norm": 2.913478136062622, + "learning_rate": 9.986449998645e-06, + "loss": 0.0914, + "step": 13557 + }, + { + "epoch": 6.303114830311483, + "grad_norm": 7.67628288269043, + "learning_rate": 9.9864489986449e-06, + "loss": 0.3159, + "step": 13558 + }, + { + "epoch": 6.303579730357973, + "grad_norm": 6.276856899261475, + "learning_rate": 9.986447998644801e-06, + "loss": 0.2477, + "step": 13559 + }, + { + "epoch": 6.304044630404463, + "grad_norm": 5.060609817504883, + "learning_rate": 9.986446998644702e-06, + "loss": 0.1533, + "step": 13560 + }, + { + "epoch": 6.304509530450953, + "grad_norm": 3.0355613231658936, + "learning_rate": 9.9864459986446e-06, + "loss": 0.0693, + "step": 13561 + }, + { + "epoch": 6.304974430497443, + "grad_norm": 4.156972408294678, + "learning_rate": 9.986444998644501e-06, + "loss": 0.1264, + "step": 13562 + }, + { + "epoch": 6.305439330543933, + "grad_norm": 5.025824546813965, + "learning_rate": 9.9864439986444e-06, + "loss": 0.1672, + "step": 13563 + }, + { + "epoch": 6.305904230590423, + "grad_norm": 3.142141342163086, + "learning_rate": 9.9864429986443e-06, + "loss": 0.0474, + "step": 13564 + }, + { + "epoch": 6.306369130636913, + "grad_norm": 4.238023281097412, + "learning_rate": 9.986441998644201e-06, + "loss": 0.1476, + "step": 13565 + }, + { + "epoch": 6.306834030683403, + "grad_norm": 5.698825359344482, + "learning_rate": 9.9864409986441e-06, + "loss": 0.2497, + "step": 13566 + }, + { + "epoch": 6.3072989307298934, + "grad_norm": 8.606704711914062, + "learning_rate": 9.986439998644001e-06, + "loss": 0.6241, + "step": 13567 + }, + { + "epoch": 6.307763830776383, + "grad_norm": 1.9699867963790894, + "learning_rate": 9.9864389986439e-06, + "loss": 0.031, + "step": 13568 + }, + { + "epoch": 6.308228730822873, + "grad_norm": 6.888940811157227, + "learning_rate": 9.9864379986438e-06, + "loss": 0.3845, + "step": 13569 + }, + { + "epoch": 6.308693630869363, + "grad_norm": 6.599706172943115, + "learning_rate": 9.9864369986437e-06, + "loss": 0.192, + "step": 13570 + }, + { + "epoch": 6.3091585309158535, + "grad_norm": 8.007706642150879, + "learning_rate": 9.9864359986436e-06, + "loss": 0.3382, + "step": 13571 + }, + { + "epoch": 6.309623430962343, + "grad_norm": 4.774594306945801, + "learning_rate": 9.986434998643501e-06, + "loss": 0.14, + "step": 13572 + }, + { + "epoch": 6.310088331008833, + "grad_norm": 4.855463027954102, + "learning_rate": 9.986433998643402e-06, + "loss": 0.1801, + "step": 13573 + }, + { + "epoch": 6.310553231055323, + "grad_norm": 5.4663987159729, + "learning_rate": 9.9864329986433e-06, + "loss": 0.2044, + "step": 13574 + }, + { + "epoch": 6.311018131101813, + "grad_norm": 3.847198724746704, + "learning_rate": 9.986431998643201e-06, + "loss": 0.1582, + "step": 13575 + }, + { + "epoch": 6.3114830311483034, + "grad_norm": 3.7825937271118164, + "learning_rate": 9.9864309986431e-06, + "loss": 0.0761, + "step": 13576 + }, + { + "epoch": 6.311947931194793, + "grad_norm": 6.192516326904297, + "learning_rate": 9.986429998643001e-06, + "loss": 0.1881, + "step": 13577 + }, + { + "epoch": 6.312412831241283, + "grad_norm": 6.654992580413818, + "learning_rate": 9.986428998642902e-06, + "loss": 0.2897, + "step": 13578 + }, + { + "epoch": 6.312877731287773, + "grad_norm": 7.575283527374268, + "learning_rate": 9.9864279986428e-06, + "loss": 0.2583, + "step": 13579 + }, + { + "epoch": 6.3133426313342635, + "grad_norm": 7.408782005310059, + "learning_rate": 9.986426998642701e-06, + "loss": 0.328, + "step": 13580 + }, + { + "epoch": 6.313807531380753, + "grad_norm": 2.718851327896118, + "learning_rate": 9.9864259986426e-06, + "loss": 0.057, + "step": 13581 + }, + { + "epoch": 6.314272431427243, + "grad_norm": 6.981540203094482, + "learning_rate": 9.9864249986425e-06, + "loss": 0.2262, + "step": 13582 + }, + { + "epoch": 6.314737331473733, + "grad_norm": 6.858368873596191, + "learning_rate": 9.9864239986424e-06, + "loss": 0.2284, + "step": 13583 + }, + { + "epoch": 6.315202231520223, + "grad_norm": 6.893659591674805, + "learning_rate": 9.9864229986423e-06, + "loss": 0.1812, + "step": 13584 + }, + { + "epoch": 6.315667131566713, + "grad_norm": 3.7212862968444824, + "learning_rate": 9.986421998642201e-06, + "loss": 0.0808, + "step": 13585 + }, + { + "epoch": 6.316132031613203, + "grad_norm": 6.016401767730713, + "learning_rate": 9.9864209986421e-06, + "loss": 0.1764, + "step": 13586 + }, + { + "epoch": 6.316596931659693, + "grad_norm": 5.712327003479004, + "learning_rate": 9.986419998642e-06, + "loss": 0.1705, + "step": 13587 + }, + { + "epoch": 6.317061831706183, + "grad_norm": 4.657220363616943, + "learning_rate": 9.9864189986419e-06, + "loss": 0.1162, + "step": 13588 + }, + { + "epoch": 6.3175267317526735, + "grad_norm": 4.667501926422119, + "learning_rate": 9.9864179986418e-06, + "loss": 0.0906, + "step": 13589 + }, + { + "epoch": 6.317991631799163, + "grad_norm": 13.366498947143555, + "learning_rate": 9.986416998641701e-06, + "loss": 0.1895, + "step": 13590 + }, + { + "epoch": 6.318456531845653, + "grad_norm": 5.10909366607666, + "learning_rate": 9.986415998641602e-06, + "loss": 0.0827, + "step": 13591 + }, + { + "epoch": 6.318921431892143, + "grad_norm": 7.298783302307129, + "learning_rate": 9.9864149986415e-06, + "loss": 0.3925, + "step": 13592 + }, + { + "epoch": 6.319386331938633, + "grad_norm": 3.80238676071167, + "learning_rate": 9.986413998641401e-06, + "loss": 0.1242, + "step": 13593 + }, + { + "epoch": 6.319851231985123, + "grad_norm": 7.605471134185791, + "learning_rate": 9.9864129986413e-06, + "loss": 0.3306, + "step": 13594 + }, + { + "epoch": 6.320316132031613, + "grad_norm": 6.058560371398926, + "learning_rate": 9.986411998641201e-06, + "loss": 0.3105, + "step": 13595 + }, + { + "epoch": 6.320781032078103, + "grad_norm": 5.592609882354736, + "learning_rate": 9.9864109986411e-06, + "loss": 0.2136, + "step": 13596 + }, + { + "epoch": 6.321245932124593, + "grad_norm": 11.164895057678223, + "learning_rate": 9.986409998641e-06, + "loss": 0.0998, + "step": 13597 + }, + { + "epoch": 6.3217108321710835, + "grad_norm": 7.169548034667969, + "learning_rate": 9.986408998640901e-06, + "loss": 0.2047, + "step": 13598 + }, + { + "epoch": 6.322175732217573, + "grad_norm": 2.3647541999816895, + "learning_rate": 9.9864079986408e-06, + "loss": 0.0444, + "step": 13599 + }, + { + "epoch": 6.322640632264063, + "grad_norm": 4.414066791534424, + "learning_rate": 9.9864069986407e-06, + "loss": 0.0659, + "step": 13600 + }, + { + "epoch": 6.323105532310553, + "grad_norm": 6.91382360458374, + "learning_rate": 9.9864059986406e-06, + "loss": 0.1772, + "step": 13601 + }, + { + "epoch": 6.323570432357044, + "grad_norm": 8.00886058807373, + "learning_rate": 9.9864049986405e-06, + "loss": 0.2153, + "step": 13602 + }, + { + "epoch": 6.324035332403533, + "grad_norm": 4.272266864776611, + "learning_rate": 9.986403998640401e-06, + "loss": 0.0737, + "step": 13603 + }, + { + "epoch": 6.324500232450023, + "grad_norm": 5.792821407318115, + "learning_rate": 9.986402998640302e-06, + "loss": 0.0896, + "step": 13604 + }, + { + "epoch": 6.324965132496513, + "grad_norm": 5.178670883178711, + "learning_rate": 9.9864019986402e-06, + "loss": 0.1411, + "step": 13605 + }, + { + "epoch": 6.325430032543003, + "grad_norm": 4.8874192237854, + "learning_rate": 9.986400998640101e-06, + "loss": 0.1777, + "step": 13606 + }, + { + "epoch": 6.3258949325894935, + "grad_norm": 4.342739105224609, + "learning_rate": 9.98639999864e-06, + "loss": 0.0931, + "step": 13607 + }, + { + "epoch": 6.326359832635983, + "grad_norm": 2.516030788421631, + "learning_rate": 9.986398998639901e-06, + "loss": 0.0332, + "step": 13608 + }, + { + "epoch": 6.326824732682473, + "grad_norm": 6.626985549926758, + "learning_rate": 9.9863979986398e-06, + "loss": 0.2869, + "step": 13609 + }, + { + "epoch": 6.327289632728963, + "grad_norm": 4.542298793792725, + "learning_rate": 9.9863969986397e-06, + "loss": 0.1131, + "step": 13610 + }, + { + "epoch": 6.327754532775454, + "grad_norm": 5.908045768737793, + "learning_rate": 9.986395998639601e-06, + "loss": 0.1086, + "step": 13611 + }, + { + "epoch": 6.328219432821943, + "grad_norm": 6.412275791168213, + "learning_rate": 9.9863949986395e-06, + "loss": 0.2898, + "step": 13612 + }, + { + "epoch": 6.328684332868433, + "grad_norm": 5.211395263671875, + "learning_rate": 9.986393998639401e-06, + "loss": 0.1754, + "step": 13613 + }, + { + "epoch": 6.329149232914923, + "grad_norm": 4.026699066162109, + "learning_rate": 9.9863929986393e-06, + "loss": 0.0966, + "step": 13614 + }, + { + "epoch": 6.329614132961414, + "grad_norm": 8.190552711486816, + "learning_rate": 9.9863919986392e-06, + "loss": 0.1183, + "step": 13615 + }, + { + "epoch": 6.3300790330079035, + "grad_norm": 5.25712776184082, + "learning_rate": 9.986390998639101e-06, + "loss": 0.1879, + "step": 13616 + }, + { + "epoch": 6.330543933054393, + "grad_norm": 6.5656232833862305, + "learning_rate": 9.986389998639002e-06, + "loss": 0.3532, + "step": 13617 + }, + { + "epoch": 6.331008833100883, + "grad_norm": 4.524208068847656, + "learning_rate": 9.9863889986389e-06, + "loss": 0.1259, + "step": 13618 + }, + { + "epoch": 6.331473733147373, + "grad_norm": 3.4463064670562744, + "learning_rate": 9.9863879986388e-06, + "loss": 0.0896, + "step": 13619 + }, + { + "epoch": 6.331938633193864, + "grad_norm": 4.270047664642334, + "learning_rate": 9.9863869986387e-06, + "loss": 0.1121, + "step": 13620 + }, + { + "epoch": 6.332403533240353, + "grad_norm": 6.480785369873047, + "learning_rate": 9.986385998638601e-06, + "loss": 0.1845, + "step": 13621 + }, + { + "epoch": 6.332868433286843, + "grad_norm": 5.779994964599609, + "learning_rate": 9.986384998638502e-06, + "loss": 0.0873, + "step": 13622 + }, + { + "epoch": 6.333333333333333, + "grad_norm": 5.440316200256348, + "learning_rate": 9.9863839986384e-06, + "loss": 0.1933, + "step": 13623 + }, + { + "epoch": 6.333798233379824, + "grad_norm": 2.9773783683776855, + "learning_rate": 9.986382998638301e-06, + "loss": 0.0549, + "step": 13624 + }, + { + "epoch": 6.3342631334263135, + "grad_norm": 4.802187442779541, + "learning_rate": 9.9863819986382e-06, + "loss": 0.1391, + "step": 13625 + }, + { + "epoch": 6.334728033472803, + "grad_norm": 5.212932109832764, + "learning_rate": 9.986380998638101e-06, + "loss": 0.1802, + "step": 13626 + }, + { + "epoch": 6.335192933519293, + "grad_norm": 4.717373847961426, + "learning_rate": 9.986379998638e-06, + "loss": 0.0889, + "step": 13627 + }, + { + "epoch": 6.335657833565783, + "grad_norm": 4.420691967010498, + "learning_rate": 9.9863789986379e-06, + "loss": 0.1582, + "step": 13628 + }, + { + "epoch": 6.336122733612274, + "grad_norm": 4.765825271606445, + "learning_rate": 9.986377998637801e-06, + "loss": 0.1564, + "step": 13629 + }, + { + "epoch": 6.336587633658763, + "grad_norm": 5.852580547332764, + "learning_rate": 9.9863769986377e-06, + "loss": 0.2224, + "step": 13630 + }, + { + "epoch": 6.337052533705253, + "grad_norm": 2.740885019302368, + "learning_rate": 9.986375998637601e-06, + "loss": 0.0472, + "step": 13631 + }, + { + "epoch": 6.337517433751743, + "grad_norm": 4.599647521972656, + "learning_rate": 9.9863749986375e-06, + "loss": 0.0819, + "step": 13632 + }, + { + "epoch": 6.337982333798234, + "grad_norm": 4.025115013122559, + "learning_rate": 9.9863739986374e-06, + "loss": 0.0543, + "step": 13633 + }, + { + "epoch": 6.3384472338447235, + "grad_norm": 4.857696533203125, + "learning_rate": 9.986372998637301e-06, + "loss": 0.1204, + "step": 13634 + }, + { + "epoch": 6.338912133891213, + "grad_norm": 6.4496169090271, + "learning_rate": 9.986371998637202e-06, + "loss": 0.1739, + "step": 13635 + }, + { + "epoch": 6.339377033937703, + "grad_norm": 2.43817138671875, + "learning_rate": 9.9863709986371e-06, + "loss": 0.0427, + "step": 13636 + }, + { + "epoch": 6.339841933984193, + "grad_norm": 3.9851644039154053, + "learning_rate": 9.986369998637001e-06, + "loss": 0.1009, + "step": 13637 + }, + { + "epoch": 6.340306834030684, + "grad_norm": 8.029928207397461, + "learning_rate": 9.9863689986369e-06, + "loss": 0.1517, + "step": 13638 + }, + { + "epoch": 6.340771734077173, + "grad_norm": 8.378901481628418, + "learning_rate": 9.986367998636801e-06, + "loss": 0.2194, + "step": 13639 + }, + { + "epoch": 6.341236634123663, + "grad_norm": 6.837408542633057, + "learning_rate": 9.9863669986367e-06, + "loss": 0.1826, + "step": 13640 + }, + { + "epoch": 6.341701534170153, + "grad_norm": 4.591335296630859, + "learning_rate": 9.9863659986366e-06, + "loss": 0.1437, + "step": 13641 + }, + { + "epoch": 6.342166434216644, + "grad_norm": 6.373098373413086, + "learning_rate": 9.986364998636501e-06, + "loss": 0.1964, + "step": 13642 + }, + { + "epoch": 6.3426313342631335, + "grad_norm": 4.794610500335693, + "learning_rate": 9.9863639986364e-06, + "loss": 0.2199, + "step": 13643 + }, + { + "epoch": 6.343096234309623, + "grad_norm": 6.310877799987793, + "learning_rate": 9.986362998636301e-06, + "loss": 0.3021, + "step": 13644 + }, + { + "epoch": 6.343561134356113, + "grad_norm": 5.84527587890625, + "learning_rate": 9.9863619986362e-06, + "loss": 0.2589, + "step": 13645 + }, + { + "epoch": 6.344026034402604, + "grad_norm": 6.187847137451172, + "learning_rate": 9.9863609986361e-06, + "loss": 0.2479, + "step": 13646 + }, + { + "epoch": 6.344490934449094, + "grad_norm": 6.446936130523682, + "learning_rate": 9.986359998636001e-06, + "loss": 0.3097, + "step": 13647 + }, + { + "epoch": 6.344955834495583, + "grad_norm": 8.457841873168945, + "learning_rate": 9.986358998635902e-06, + "loss": 0.644, + "step": 13648 + }, + { + "epoch": 6.345420734542073, + "grad_norm": 4.182990550994873, + "learning_rate": 9.9863579986358e-06, + "loss": 0.1554, + "step": 13649 + }, + { + "epoch": 6.345885634588564, + "grad_norm": 4.957285404205322, + "learning_rate": 9.986356998635701e-06, + "loss": 0.2154, + "step": 13650 + }, + { + "epoch": 6.346350534635054, + "grad_norm": 4.556777477264404, + "learning_rate": 9.9863559986356e-06, + "loss": 0.139, + "step": 13651 + }, + { + "epoch": 6.3468154346815435, + "grad_norm": 3.708836793899536, + "learning_rate": 9.9863549986355e-06, + "loss": 0.1135, + "step": 13652 + }, + { + "epoch": 6.347280334728033, + "grad_norm": 6.381546497344971, + "learning_rate": 9.9863539986354e-06, + "loss": 0.1584, + "step": 13653 + }, + { + "epoch": 6.347745234774523, + "grad_norm": 7.160576820373535, + "learning_rate": 9.9863529986353e-06, + "loss": 0.2436, + "step": 13654 + }, + { + "epoch": 6.348210134821014, + "grad_norm": 4.420085906982422, + "learning_rate": 9.986351998635201e-06, + "loss": 0.1395, + "step": 13655 + }, + { + "epoch": 6.348675034867504, + "grad_norm": 8.353377342224121, + "learning_rate": 9.9863509986351e-06, + "loss": 0.3062, + "step": 13656 + }, + { + "epoch": 6.349139934913993, + "grad_norm": 5.853734493255615, + "learning_rate": 9.986349998635001e-06, + "loss": 0.3501, + "step": 13657 + }, + { + "epoch": 6.349604834960483, + "grad_norm": 7.791342258453369, + "learning_rate": 9.9863489986349e-06, + "loss": 0.3824, + "step": 13658 + }, + { + "epoch": 6.350069735006974, + "grad_norm": 4.6690263748168945, + "learning_rate": 9.9863479986348e-06, + "loss": 0.1683, + "step": 13659 + }, + { + "epoch": 6.350534635053464, + "grad_norm": 5.715548992156982, + "learning_rate": 9.986346998634701e-06, + "loss": 0.1581, + "step": 13660 + }, + { + "epoch": 6.3509995350999535, + "grad_norm": 5.347758769989014, + "learning_rate": 9.986345998634602e-06, + "loss": 0.1971, + "step": 13661 + }, + { + "epoch": 6.351464435146443, + "grad_norm": 4.351200103759766, + "learning_rate": 9.986344998634501e-06, + "loss": 0.1177, + "step": 13662 + }, + { + "epoch": 6.351929335192933, + "grad_norm": 4.608420372009277, + "learning_rate": 9.9863439986344e-06, + "loss": 0.2004, + "step": 13663 + }, + { + "epoch": 6.352394235239424, + "grad_norm": 5.134815692901611, + "learning_rate": 9.9863429986343e-06, + "loss": 0.2375, + "step": 13664 + }, + { + "epoch": 6.352859135285914, + "grad_norm": 7.815572261810303, + "learning_rate": 9.9863419986342e-06, + "loss": 0.2617, + "step": 13665 + }, + { + "epoch": 6.353324035332403, + "grad_norm": 4.677488327026367, + "learning_rate": 9.9863409986341e-06, + "loss": 0.034, + "step": 13666 + }, + { + "epoch": 6.353788935378893, + "grad_norm": 6.9315290451049805, + "learning_rate": 9.986339998634e-06, + "loss": 0.123, + "step": 13667 + }, + { + "epoch": 6.354253835425384, + "grad_norm": 7.907310485839844, + "learning_rate": 9.986338998633901e-06, + "loss": 0.1888, + "step": 13668 + }, + { + "epoch": 6.354718735471874, + "grad_norm": 6.038448333740234, + "learning_rate": 9.9863379986338e-06, + "loss": 0.1846, + "step": 13669 + }, + { + "epoch": 6.3551836355183635, + "grad_norm": 3.0231375694274902, + "learning_rate": 9.986336998633701e-06, + "loss": 0.0548, + "step": 13670 + }, + { + "epoch": 6.355648535564853, + "grad_norm": 5.165258884429932, + "learning_rate": 9.9863359986336e-06, + "loss": 0.2356, + "step": 13671 + }, + { + "epoch": 6.356113435611343, + "grad_norm": 3.713829517364502, + "learning_rate": 9.9863349986335e-06, + "loss": 0.067, + "step": 13672 + }, + { + "epoch": 6.356578335657834, + "grad_norm": 12.70678424835205, + "learning_rate": 9.986333998633401e-06, + "loss": 0.1284, + "step": 13673 + }, + { + "epoch": 6.357043235704324, + "grad_norm": 5.605410099029541, + "learning_rate": 9.9863329986333e-06, + "loss": 0.2114, + "step": 13674 + }, + { + "epoch": 6.357508135750813, + "grad_norm": 10.069236755371094, + "learning_rate": 9.986331998633201e-06, + "loss": 0.2942, + "step": 13675 + }, + { + "epoch": 6.357973035797303, + "grad_norm": 3.9991002082824707, + "learning_rate": 9.9863309986331e-06, + "loss": 0.1145, + "step": 13676 + }, + { + "epoch": 6.358437935843794, + "grad_norm": 4.265126705169678, + "learning_rate": 9.986329998633e-06, + "loss": 0.1432, + "step": 13677 + }, + { + "epoch": 6.358902835890284, + "grad_norm": 4.7881855964660645, + "learning_rate": 9.9863289986329e-06, + "loss": 0.1815, + "step": 13678 + }, + { + "epoch": 6.3593677359367735, + "grad_norm": 4.6601948738098145, + "learning_rate": 9.9863279986328e-06, + "loss": 0.1321, + "step": 13679 + }, + { + "epoch": 6.359832635983263, + "grad_norm": 5.271836757659912, + "learning_rate": 9.986326998632701e-06, + "loss": 0.069, + "step": 13680 + }, + { + "epoch": 6.360297536029754, + "grad_norm": 7.0667572021484375, + "learning_rate": 9.986325998632602e-06, + "loss": 0.1777, + "step": 13681 + }, + { + "epoch": 6.360762436076244, + "grad_norm": 5.1777801513671875, + "learning_rate": 9.9863249986325e-06, + "loss": 0.1396, + "step": 13682 + }, + { + "epoch": 6.361227336122734, + "grad_norm": 4.050053596496582, + "learning_rate": 9.986323998632401e-06, + "loss": 0.125, + "step": 13683 + }, + { + "epoch": 6.361692236169223, + "grad_norm": 2.7054760456085205, + "learning_rate": 9.9863229986323e-06, + "loss": 0.0463, + "step": 13684 + }, + { + "epoch": 6.362157136215713, + "grad_norm": 4.304922103881836, + "learning_rate": 9.9863219986322e-06, + "loss": 0.1648, + "step": 13685 + }, + { + "epoch": 6.362622036262204, + "grad_norm": 2.092296838760376, + "learning_rate": 9.986320998632101e-06, + "loss": 0.0226, + "step": 13686 + }, + { + "epoch": 6.363086936308694, + "grad_norm": 5.5715742111206055, + "learning_rate": 9.986319998632e-06, + "loss": 0.193, + "step": 13687 + }, + { + "epoch": 6.3635518363551835, + "grad_norm": 3.871690511703491, + "learning_rate": 9.986318998631901e-06, + "loss": 0.1276, + "step": 13688 + }, + { + "epoch": 6.364016736401673, + "grad_norm": 3.0069522857666016, + "learning_rate": 9.9863179986318e-06, + "loss": 0.0654, + "step": 13689 + }, + { + "epoch": 6.364481636448164, + "grad_norm": 8.247714042663574, + "learning_rate": 9.9863169986317e-06, + "loss": 0.2452, + "step": 13690 + }, + { + "epoch": 6.364946536494654, + "grad_norm": 5.196975231170654, + "learning_rate": 9.986315998631601e-06, + "loss": 0.2086, + "step": 13691 + }, + { + "epoch": 6.365411436541144, + "grad_norm": 5.633697509765625, + "learning_rate": 9.9863149986315e-06, + "loss": 0.0834, + "step": 13692 + }, + { + "epoch": 6.365876336587633, + "grad_norm": 6.018944263458252, + "learning_rate": 9.986313998631401e-06, + "loss": 0.2185, + "step": 13693 + }, + { + "epoch": 6.366341236634124, + "grad_norm": 4.0812482833862305, + "learning_rate": 9.986312998631302e-06, + "loss": 0.0961, + "step": 13694 + }, + { + "epoch": 6.366806136680614, + "grad_norm": 3.77695369720459, + "learning_rate": 9.9863119986312e-06, + "loss": 0.0676, + "step": 13695 + }, + { + "epoch": 6.367271036727104, + "grad_norm": 2.7142109870910645, + "learning_rate": 9.9863109986311e-06, + "loss": 0.0271, + "step": 13696 + }, + { + "epoch": 6.3677359367735935, + "grad_norm": 4.080056667327881, + "learning_rate": 9.986309998631e-06, + "loss": 0.089, + "step": 13697 + }, + { + "epoch": 6.368200836820083, + "grad_norm": 4.137689590454102, + "learning_rate": 9.986308998630901e-06, + "loss": 0.1475, + "step": 13698 + }, + { + "epoch": 6.368665736866574, + "grad_norm": 2.351112127304077, + "learning_rate": 9.986307998630802e-06, + "loss": 0.0517, + "step": 13699 + }, + { + "epoch": 6.369130636913064, + "grad_norm": 4.762025833129883, + "learning_rate": 9.9863069986307e-06, + "loss": 0.2734, + "step": 13700 + }, + { + "epoch": 6.369595536959554, + "grad_norm": 5.161097526550293, + "learning_rate": 9.986305998630601e-06, + "loss": 0.172, + "step": 13701 + }, + { + "epoch": 6.370060437006043, + "grad_norm": 4.755797863006592, + "learning_rate": 9.9863049986305e-06, + "loss": 0.0577, + "step": 13702 + }, + { + "epoch": 6.370525337052534, + "grad_norm": 7.917960166931152, + "learning_rate": 9.9863039986304e-06, + "loss": 0.3958, + "step": 13703 + }, + { + "epoch": 6.370990237099024, + "grad_norm": 4.5492987632751465, + "learning_rate": 9.986302998630301e-06, + "loss": 0.1622, + "step": 13704 + }, + { + "epoch": 6.371455137145514, + "grad_norm": 6.176703929901123, + "learning_rate": 9.9863019986302e-06, + "loss": 0.1181, + "step": 13705 + }, + { + "epoch": 6.3719200371920035, + "grad_norm": 8.78862476348877, + "learning_rate": 9.986300998630101e-06, + "loss": 0.2781, + "step": 13706 + }, + { + "epoch": 6.372384937238493, + "grad_norm": 4.570718765258789, + "learning_rate": 9.98629999863e-06, + "loss": 0.1035, + "step": 13707 + }, + { + "epoch": 6.372849837284984, + "grad_norm": 3.6180226802825928, + "learning_rate": 9.9862989986299e-06, + "loss": 0.0749, + "step": 13708 + }, + { + "epoch": 6.373314737331474, + "grad_norm": 5.282434940338135, + "learning_rate": 9.9862979986298e-06, + "loss": 0.1574, + "step": 13709 + }, + { + "epoch": 6.373779637377964, + "grad_norm": 5.8578033447265625, + "learning_rate": 9.9862969986297e-06, + "loss": 0.2327, + "step": 13710 + }, + { + "epoch": 6.374244537424453, + "grad_norm": 5.098834037780762, + "learning_rate": 9.986295998629601e-06, + "loss": 0.2313, + "step": 13711 + }, + { + "epoch": 6.374709437470944, + "grad_norm": 6.800632476806641, + "learning_rate": 9.986294998629502e-06, + "loss": 0.2415, + "step": 13712 + }, + { + "epoch": 6.375174337517434, + "grad_norm": 3.9309792518615723, + "learning_rate": 9.9862939986294e-06, + "loss": 0.1133, + "step": 13713 + }, + { + "epoch": 6.375639237563924, + "grad_norm": 5.019199848175049, + "learning_rate": 9.986292998629301e-06, + "loss": 0.1597, + "step": 13714 + }, + { + "epoch": 6.3761041376104135, + "grad_norm": 8.860222816467285, + "learning_rate": 9.9862919986292e-06, + "loss": 0.3743, + "step": 13715 + }, + { + "epoch": 6.376569037656903, + "grad_norm": 4.7536211013793945, + "learning_rate": 9.986290998629101e-06, + "loss": 0.1688, + "step": 13716 + }, + { + "epoch": 6.377033937703394, + "grad_norm": 3.3162174224853516, + "learning_rate": 9.986289998629002e-06, + "loss": 0.0832, + "step": 13717 + }, + { + "epoch": 6.377498837749884, + "grad_norm": 7.117657661437988, + "learning_rate": 9.9862889986289e-06, + "loss": 0.4021, + "step": 13718 + }, + { + "epoch": 6.377963737796374, + "grad_norm": 4.16968297958374, + "learning_rate": 9.986287998628801e-06, + "loss": 0.1569, + "step": 13719 + }, + { + "epoch": 6.378428637842863, + "grad_norm": 5.492490768432617, + "learning_rate": 9.9862869986287e-06, + "loss": 0.1108, + "step": 13720 + }, + { + "epoch": 6.378893537889354, + "grad_norm": 4.837029933929443, + "learning_rate": 9.9862859986286e-06, + "loss": 0.1618, + "step": 13721 + }, + { + "epoch": 6.379358437935844, + "grad_norm": 3.248720169067383, + "learning_rate": 9.9862849986285e-06, + "loss": 0.0753, + "step": 13722 + }, + { + "epoch": 6.379823337982334, + "grad_norm": 7.74465274810791, + "learning_rate": 9.9862839986284e-06, + "loss": 0.4477, + "step": 13723 + }, + { + "epoch": 6.3802882380288235, + "grad_norm": 5.0269904136657715, + "learning_rate": 9.986282998628301e-06, + "loss": 0.1073, + "step": 13724 + }, + { + "epoch": 6.380753138075314, + "grad_norm": 3.739439010620117, + "learning_rate": 9.986281998628202e-06, + "loss": 0.1293, + "step": 13725 + }, + { + "epoch": 6.381218038121804, + "grad_norm": 6.871564865112305, + "learning_rate": 9.9862809986281e-06, + "loss": 0.229, + "step": 13726 + }, + { + "epoch": 6.381682938168294, + "grad_norm": 5.94610595703125, + "learning_rate": 9.986279998628e-06, + "loss": 0.3711, + "step": 13727 + }, + { + "epoch": 6.382147838214784, + "grad_norm": 4.032332897186279, + "learning_rate": 9.9862789986279e-06, + "loss": 0.1214, + "step": 13728 + }, + { + "epoch": 6.382612738261274, + "grad_norm": 7.485066890716553, + "learning_rate": 9.986277998627801e-06, + "loss": 0.2216, + "step": 13729 + }, + { + "epoch": 6.383077638307764, + "grad_norm": 5.208416938781738, + "learning_rate": 9.986276998627702e-06, + "loss": 0.1532, + "step": 13730 + }, + { + "epoch": 6.383542538354254, + "grad_norm": 4.613203525543213, + "learning_rate": 9.9862759986276e-06, + "loss": 0.1607, + "step": 13731 + }, + { + "epoch": 6.384007438400744, + "grad_norm": 4.497204303741455, + "learning_rate": 9.986274998627501e-06, + "loss": 0.1471, + "step": 13732 + }, + { + "epoch": 6.3844723384472335, + "grad_norm": 4.778932094573975, + "learning_rate": 9.9862739986274e-06, + "loss": 0.1401, + "step": 13733 + }, + { + "epoch": 6.384937238493724, + "grad_norm": 3.0431909561157227, + "learning_rate": 9.9862729986273e-06, + "loss": 0.0431, + "step": 13734 + }, + { + "epoch": 6.385402138540214, + "grad_norm": 4.88758659362793, + "learning_rate": 9.9862719986272e-06, + "loss": 0.155, + "step": 13735 + }, + { + "epoch": 6.385867038586704, + "grad_norm": 3.622624397277832, + "learning_rate": 9.9862709986271e-06, + "loss": 0.0717, + "step": 13736 + }, + { + "epoch": 6.386331938633194, + "grad_norm": 6.252020359039307, + "learning_rate": 9.986269998627001e-06, + "loss": 0.2382, + "step": 13737 + }, + { + "epoch": 6.386796838679684, + "grad_norm": 9.971556663513184, + "learning_rate": 9.9862689986269e-06, + "loss": 0.2363, + "step": 13738 + }, + { + "epoch": 6.387261738726174, + "grad_norm": 7.8283538818359375, + "learning_rate": 9.9862679986268e-06, + "loss": 0.3345, + "step": 13739 + }, + { + "epoch": 6.387726638772664, + "grad_norm": 3.3094911575317383, + "learning_rate": 9.9862669986267e-06, + "loss": 0.0907, + "step": 13740 + }, + { + "epoch": 6.388191538819154, + "grad_norm": 3.4280667304992676, + "learning_rate": 9.9862659986266e-06, + "loss": 0.0991, + "step": 13741 + }, + { + "epoch": 6.3886564388656435, + "grad_norm": 5.647937774658203, + "learning_rate": 9.986264998626501e-06, + "loss": 0.3194, + "step": 13742 + }, + { + "epoch": 6.389121338912134, + "grad_norm": 5.355607509613037, + "learning_rate": 9.986263998626402e-06, + "loss": 0.2077, + "step": 13743 + }, + { + "epoch": 6.389586238958624, + "grad_norm": 7.612819671630859, + "learning_rate": 9.9862629986263e-06, + "loss": 0.2198, + "step": 13744 + }, + { + "epoch": 6.390051139005114, + "grad_norm": 3.8954288959503174, + "learning_rate": 9.986261998626201e-06, + "loss": 0.0968, + "step": 13745 + }, + { + "epoch": 6.390516039051604, + "grad_norm": 7.209325790405273, + "learning_rate": 9.9862609986261e-06, + "loss": 0.3209, + "step": 13746 + }, + { + "epoch": 6.390980939098094, + "grad_norm": 5.897054195404053, + "learning_rate": 9.986259998626001e-06, + "loss": 0.1624, + "step": 13747 + }, + { + "epoch": 6.391445839144584, + "grad_norm": 2.950829267501831, + "learning_rate": 9.9862589986259e-06, + "loss": 0.0932, + "step": 13748 + }, + { + "epoch": 6.391910739191074, + "grad_norm": 4.403727054595947, + "learning_rate": 9.9862579986258e-06, + "loss": 0.081, + "step": 13749 + }, + { + "epoch": 6.392375639237564, + "grad_norm": 6.48130989074707, + "learning_rate": 9.986256998625701e-06, + "loss": 0.213, + "step": 13750 + }, + { + "epoch": 6.3928405392840535, + "grad_norm": 4.581518650054932, + "learning_rate": 9.9862559986256e-06, + "loss": 0.0921, + "step": 13751 + }, + { + "epoch": 6.393305439330544, + "grad_norm": 7.019681453704834, + "learning_rate": 9.9862549986255e-06, + "loss": 0.2636, + "step": 13752 + }, + { + "epoch": 6.393770339377034, + "grad_norm": 5.606003761291504, + "learning_rate": 9.9862539986254e-06, + "loss": 0.212, + "step": 13753 + }, + { + "epoch": 6.394235239423524, + "grad_norm": 4.853588104248047, + "learning_rate": 9.9862529986253e-06, + "loss": 0.1455, + "step": 13754 + }, + { + "epoch": 6.394700139470014, + "grad_norm": 4.3804612159729, + "learning_rate": 9.986251998625201e-06, + "loss": 0.1095, + "step": 13755 + }, + { + "epoch": 6.395165039516504, + "grad_norm": 5.685649394989014, + "learning_rate": 9.986250998625102e-06, + "loss": 0.2057, + "step": 13756 + }, + { + "epoch": 6.395629939562994, + "grad_norm": 6.451321125030518, + "learning_rate": 9.986249998625e-06, + "loss": 0.2385, + "step": 13757 + }, + { + "epoch": 6.396094839609484, + "grad_norm": 11.378137588500977, + "learning_rate": 9.986248998624901e-06, + "loss": 0.3395, + "step": 13758 + }, + { + "epoch": 6.396559739655974, + "grad_norm": 3.747086763381958, + "learning_rate": 9.9862479986248e-06, + "loss": 0.1401, + "step": 13759 + }, + { + "epoch": 6.397024639702464, + "grad_norm": 3.8762919902801514, + "learning_rate": 9.9862469986247e-06, + "loss": 0.1385, + "step": 13760 + }, + { + "epoch": 6.397489539748954, + "grad_norm": 4.939510822296143, + "learning_rate": 9.9862459986246e-06, + "loss": 0.2018, + "step": 13761 + }, + { + "epoch": 6.397954439795444, + "grad_norm": 6.882252216339111, + "learning_rate": 9.9862449986245e-06, + "loss": 0.2781, + "step": 13762 + }, + { + "epoch": 6.398419339841934, + "grad_norm": 4.468417167663574, + "learning_rate": 9.986243998624401e-06, + "loss": 0.1402, + "step": 13763 + }, + { + "epoch": 6.398884239888424, + "grad_norm": 5.220935344696045, + "learning_rate": 9.9862429986243e-06, + "loss": 0.2095, + "step": 13764 + }, + { + "epoch": 6.399349139934914, + "grad_norm": 4.037087917327881, + "learning_rate": 9.986241998624201e-06, + "loss": 0.1355, + "step": 13765 + }, + { + "epoch": 6.399814039981404, + "grad_norm": 8.825431823730469, + "learning_rate": 9.9862409986241e-06, + "loss": 0.2185, + "step": 13766 + }, + { + "epoch": 6.400278940027894, + "grad_norm": 4.958506107330322, + "learning_rate": 9.986239998624e-06, + "loss": 0.0704, + "step": 13767 + }, + { + "epoch": 6.400743840074384, + "grad_norm": 4.847679615020752, + "learning_rate": 9.986238998623901e-06, + "loss": 0.1424, + "step": 13768 + }, + { + "epoch": 6.401208740120874, + "grad_norm": 7.070382595062256, + "learning_rate": 9.986237998623802e-06, + "loss": 0.1179, + "step": 13769 + }, + { + "epoch": 6.401673640167364, + "grad_norm": 3.7282602787017822, + "learning_rate": 9.9862369986237e-06, + "loss": 0.065, + "step": 13770 + }, + { + "epoch": 6.402138540213854, + "grad_norm": 6.286600112915039, + "learning_rate": 9.9862359986236e-06, + "loss": 0.1508, + "step": 13771 + }, + { + "epoch": 6.402603440260344, + "grad_norm": 8.386133193969727, + "learning_rate": 9.9862349986235e-06, + "loss": 0.2353, + "step": 13772 + }, + { + "epoch": 6.4030683403068345, + "grad_norm": 7.44069766998291, + "learning_rate": 9.986233998623401e-06, + "loss": 0.2513, + "step": 13773 + }, + { + "epoch": 6.403533240353324, + "grad_norm": 11.777178764343262, + "learning_rate": 9.9862329986233e-06, + "loss": 0.4727, + "step": 13774 + }, + { + "epoch": 6.403998140399814, + "grad_norm": 5.133321762084961, + "learning_rate": 9.9862319986232e-06, + "loss": 0.1338, + "step": 13775 + }, + { + "epoch": 6.404463040446304, + "grad_norm": 5.83318567276001, + "learning_rate": 9.986230998623101e-06, + "loss": 0.1699, + "step": 13776 + }, + { + "epoch": 6.404927940492794, + "grad_norm": 4.323616981506348, + "learning_rate": 9.986229998623e-06, + "loss": 0.0649, + "step": 13777 + }, + { + "epoch": 6.405392840539284, + "grad_norm": 5.117777347564697, + "learning_rate": 9.986228998622901e-06, + "loss": 0.0599, + "step": 13778 + }, + { + "epoch": 6.405857740585774, + "grad_norm": 3.898286819458008, + "learning_rate": 9.9862279986228e-06, + "loss": 0.0951, + "step": 13779 + }, + { + "epoch": 6.406322640632264, + "grad_norm": 3.227283239364624, + "learning_rate": 9.9862269986227e-06, + "loss": 0.1073, + "step": 13780 + }, + { + "epoch": 6.406787540678754, + "grad_norm": 7.667980194091797, + "learning_rate": 9.986225998622601e-06, + "loss": 0.3402, + "step": 13781 + }, + { + "epoch": 6.4072524407252445, + "grad_norm": 3.820089817047119, + "learning_rate": 9.9862249986225e-06, + "loss": 0.0355, + "step": 13782 + }, + { + "epoch": 6.407717340771734, + "grad_norm": 4.588863372802734, + "learning_rate": 9.986223998622401e-06, + "loss": 0.2076, + "step": 13783 + }, + { + "epoch": 6.408182240818224, + "grad_norm": 5.909541606903076, + "learning_rate": 9.9862229986223e-06, + "loss": 0.1839, + "step": 13784 + }, + { + "epoch": 6.408647140864714, + "grad_norm": 2.4915273189544678, + "learning_rate": 9.9862219986222e-06, + "loss": 0.0446, + "step": 13785 + }, + { + "epoch": 6.409112040911204, + "grad_norm": 6.374502182006836, + "learning_rate": 9.986220998622101e-06, + "loss": 0.2069, + "step": 13786 + }, + { + "epoch": 6.409576940957694, + "grad_norm": 7.113396644592285, + "learning_rate": 9.986219998622002e-06, + "loss": 0.2299, + "step": 13787 + }, + { + "epoch": 6.410041841004184, + "grad_norm": 2.162137746810913, + "learning_rate": 9.9862189986219e-06, + "loss": 0.0428, + "step": 13788 + }, + { + "epoch": 6.410506741050674, + "grad_norm": 4.733645915985107, + "learning_rate": 9.986217998621801e-06, + "loss": 0.0827, + "step": 13789 + }, + { + "epoch": 6.410971641097164, + "grad_norm": 6.274186611175537, + "learning_rate": 9.9862169986217e-06, + "loss": 0.2213, + "step": 13790 + }, + { + "epoch": 6.4114365411436545, + "grad_norm": 5.850729465484619, + "learning_rate": 9.986215998621601e-06, + "loss": 0.1305, + "step": 13791 + }, + { + "epoch": 6.411901441190144, + "grad_norm": 7.094454765319824, + "learning_rate": 9.9862149986215e-06, + "loss": 0.2687, + "step": 13792 + }, + { + "epoch": 6.412366341236634, + "grad_norm": 6.082940101623535, + "learning_rate": 9.9862139986214e-06, + "loss": 0.2935, + "step": 13793 + }, + { + "epoch": 6.412831241283124, + "grad_norm": 4.520515441894531, + "learning_rate": 9.986212998621301e-06, + "loss": 0.1152, + "step": 13794 + }, + { + "epoch": 6.413296141329614, + "grad_norm": 3.6270651817321777, + "learning_rate": 9.9862119986212e-06, + "loss": 0.0954, + "step": 13795 + }, + { + "epoch": 6.413761041376104, + "grad_norm": 5.278481483459473, + "learning_rate": 9.986210998621101e-06, + "loss": 0.2099, + "step": 13796 + }, + { + "epoch": 6.414225941422594, + "grad_norm": 5.417101860046387, + "learning_rate": 9.986209998621e-06, + "loss": 0.1427, + "step": 13797 + }, + { + "epoch": 6.414690841469084, + "grad_norm": 3.912902593612671, + "learning_rate": 9.9862089986209e-06, + "loss": 0.0864, + "step": 13798 + }, + { + "epoch": 6.415155741515574, + "grad_norm": 5.7741498947143555, + "learning_rate": 9.986207998620801e-06, + "loss": 0.2518, + "step": 13799 + }, + { + "epoch": 6.4156206415620645, + "grad_norm": 4.156474590301514, + "learning_rate": 9.986206998620702e-06, + "loss": 0.1335, + "step": 13800 + }, + { + "epoch": 6.416085541608554, + "grad_norm": 3.7367522716522217, + "learning_rate": 9.986205998620601e-06, + "loss": 0.0842, + "step": 13801 + }, + { + "epoch": 6.416550441655044, + "grad_norm": 3.943842887878418, + "learning_rate": 9.986204998620502e-06, + "loss": 0.0622, + "step": 13802 + }, + { + "epoch": 6.417015341701534, + "grad_norm": 3.487840175628662, + "learning_rate": 9.9862039986204e-06, + "loss": 0.0731, + "step": 13803 + }, + { + "epoch": 6.417480241748025, + "grad_norm": 9.858806610107422, + "learning_rate": 9.9862029986203e-06, + "loss": 0.3755, + "step": 13804 + }, + { + "epoch": 6.417945141794514, + "grad_norm": 7.022698402404785, + "learning_rate": 9.9862019986202e-06, + "loss": 0.1575, + "step": 13805 + }, + { + "epoch": 6.418410041841004, + "grad_norm": 4.095978260040283, + "learning_rate": 9.9862009986201e-06, + "loss": 0.1181, + "step": 13806 + }, + { + "epoch": 6.418874941887494, + "grad_norm": 3.055605173110962, + "learning_rate": 9.986199998620001e-06, + "loss": 0.0858, + "step": 13807 + }, + { + "epoch": 6.419339841933985, + "grad_norm": 7.034362316131592, + "learning_rate": 9.9861989986199e-06, + "loss": 0.417, + "step": 13808 + }, + { + "epoch": 6.4198047419804745, + "grad_norm": 5.907016754150391, + "learning_rate": 9.986197998619801e-06, + "loss": 0.2197, + "step": 13809 + }, + { + "epoch": 6.420269642026964, + "grad_norm": 5.487029552459717, + "learning_rate": 9.9861969986197e-06, + "loss": 0.1022, + "step": 13810 + }, + { + "epoch": 6.420734542073454, + "grad_norm": 7.718933582305908, + "learning_rate": 9.9861959986196e-06, + "loss": 0.3087, + "step": 13811 + }, + { + "epoch": 6.421199442119944, + "grad_norm": 6.623459339141846, + "learning_rate": 9.986194998619501e-06, + "loss": 0.189, + "step": 13812 + }, + { + "epoch": 6.421664342166435, + "grad_norm": 10.76820182800293, + "learning_rate": 9.986193998619402e-06, + "loss": 0.112, + "step": 13813 + }, + { + "epoch": 6.422129242212924, + "grad_norm": 4.732797145843506, + "learning_rate": 9.986192998619301e-06, + "loss": 0.1744, + "step": 13814 + }, + { + "epoch": 6.422594142259414, + "grad_norm": 5.452641487121582, + "learning_rate": 9.9861919986192e-06, + "loss": 0.18, + "step": 13815 + }, + { + "epoch": 6.423059042305904, + "grad_norm": 4.1940178871154785, + "learning_rate": 9.9861909986191e-06, + "loss": 0.0907, + "step": 13816 + }, + { + "epoch": 6.423523942352395, + "grad_norm": 7.54440975189209, + "learning_rate": 9.986189998619e-06, + "loss": 0.286, + "step": 13817 + }, + { + "epoch": 6.4239888423988845, + "grad_norm": 3.22231125831604, + "learning_rate": 9.9861889986189e-06, + "loss": 0.1207, + "step": 13818 + }, + { + "epoch": 6.424453742445374, + "grad_norm": 4.593298435211182, + "learning_rate": 9.9861879986188e-06, + "loss": 0.1558, + "step": 13819 + }, + { + "epoch": 6.424918642491864, + "grad_norm": 6.4371018409729, + "learning_rate": 9.986186998618702e-06, + "loss": 0.1837, + "step": 13820 + }, + { + "epoch": 6.425383542538354, + "grad_norm": 5.178947925567627, + "learning_rate": 9.9861859986186e-06, + "loss": 0.2055, + "step": 13821 + }, + { + "epoch": 6.4258484425848446, + "grad_norm": 6.770873546600342, + "learning_rate": 9.986184998618501e-06, + "loss": 0.1768, + "step": 13822 + }, + { + "epoch": 6.426313342631334, + "grad_norm": 5.155829906463623, + "learning_rate": 9.9861839986184e-06, + "loss": 0.2123, + "step": 13823 + }, + { + "epoch": 6.426778242677824, + "grad_norm": 6.5538554191589355, + "learning_rate": 9.9861829986183e-06, + "loss": 0.2123, + "step": 13824 + }, + { + "epoch": 6.427243142724314, + "grad_norm": 4.642916202545166, + "learning_rate": 9.986181998618201e-06, + "loss": 0.153, + "step": 13825 + }, + { + "epoch": 6.427708042770805, + "grad_norm": 4.31644868850708, + "learning_rate": 9.9861809986181e-06, + "loss": 0.1123, + "step": 13826 + }, + { + "epoch": 6.4281729428172945, + "grad_norm": 3.0502567291259766, + "learning_rate": 9.986179998618001e-06, + "loss": 0.1017, + "step": 13827 + }, + { + "epoch": 6.428637842863784, + "grad_norm": 5.367300510406494, + "learning_rate": 9.9861789986179e-06, + "loss": 0.1425, + "step": 13828 + }, + { + "epoch": 6.429102742910274, + "grad_norm": 6.302215576171875, + "learning_rate": 9.9861779986178e-06, + "loss": 0.1744, + "step": 13829 + }, + { + "epoch": 6.429567642956764, + "grad_norm": 4.835635185241699, + "learning_rate": 9.9861769986177e-06, + "loss": 0.144, + "step": 13830 + }, + { + "epoch": 6.4300325430032546, + "grad_norm": 6.370391845703125, + "learning_rate": 9.9861759986176e-06, + "loss": 0.2107, + "step": 13831 + }, + { + "epoch": 6.430497443049744, + "grad_norm": 5.670284271240234, + "learning_rate": 9.986174998617501e-06, + "loss": 0.1774, + "step": 13832 + }, + { + "epoch": 6.430962343096234, + "grad_norm": 7.31503438949585, + "learning_rate": 9.986173998617402e-06, + "loss": 0.1853, + "step": 13833 + }, + { + "epoch": 6.431427243142724, + "grad_norm": 7.099430561065674, + "learning_rate": 9.9861729986173e-06, + "loss": 0.2701, + "step": 13834 + }, + { + "epoch": 6.431892143189215, + "grad_norm": 9.131795883178711, + "learning_rate": 9.986171998617201e-06, + "loss": 0.3097, + "step": 13835 + }, + { + "epoch": 6.4323570432357045, + "grad_norm": 4.851503849029541, + "learning_rate": 9.9861709986171e-06, + "loss": 0.174, + "step": 13836 + }, + { + "epoch": 6.432821943282194, + "grad_norm": 7.664724349975586, + "learning_rate": 9.986169998617e-06, + "loss": 0.1433, + "step": 13837 + }, + { + "epoch": 6.433286843328684, + "grad_norm": 6.005046367645264, + "learning_rate": 9.986168998616901e-06, + "loss": 0.3151, + "step": 13838 + }, + { + "epoch": 6.433751743375175, + "grad_norm": 5.493814945220947, + "learning_rate": 9.9861679986168e-06, + "loss": 0.1609, + "step": 13839 + }, + { + "epoch": 6.4342166434216645, + "grad_norm": 4.190836429595947, + "learning_rate": 9.986166998616701e-06, + "loss": 0.0994, + "step": 13840 + }, + { + "epoch": 6.434681543468154, + "grad_norm": 5.0900774002075195, + "learning_rate": 9.9861659986166e-06, + "loss": 0.1044, + "step": 13841 + }, + { + "epoch": 6.435146443514644, + "grad_norm": 4.489829063415527, + "learning_rate": 9.9861649986165e-06, + "loss": 0.111, + "step": 13842 + }, + { + "epoch": 6.435611343561134, + "grad_norm": 5.424627780914307, + "learning_rate": 9.9861639986164e-06, + "loss": 0.1535, + "step": 13843 + }, + { + "epoch": 6.436076243607625, + "grad_norm": 8.564680099487305, + "learning_rate": 9.9861629986163e-06, + "loss": 0.337, + "step": 13844 + }, + { + "epoch": 6.4365411436541144, + "grad_norm": 4.576708793640137, + "learning_rate": 9.986161998616201e-06, + "loss": 0.1165, + "step": 13845 + }, + { + "epoch": 6.437006043700604, + "grad_norm": 11.59167194366455, + "learning_rate": 9.986160998616102e-06, + "loss": 0.1022, + "step": 13846 + }, + { + "epoch": 6.437470943747094, + "grad_norm": 5.416695594787598, + "learning_rate": 9.986159998616e-06, + "loss": 0.292, + "step": 13847 + }, + { + "epoch": 6.437935843793585, + "grad_norm": 3.234802722930908, + "learning_rate": 9.9861589986159e-06, + "loss": 0.0947, + "step": 13848 + }, + { + "epoch": 6.4384007438400745, + "grad_norm": 4.673633575439453, + "learning_rate": 9.9861579986158e-06, + "loss": 0.1182, + "step": 13849 + }, + { + "epoch": 6.438865643886564, + "grad_norm": 10.27039623260498, + "learning_rate": 9.986156998615701e-06, + "loss": 0.4206, + "step": 13850 + }, + { + "epoch": 6.439330543933054, + "grad_norm": 4.272756576538086, + "learning_rate": 9.986155998615602e-06, + "loss": 0.1105, + "step": 13851 + }, + { + "epoch": 6.439795443979545, + "grad_norm": 5.042566299438477, + "learning_rate": 9.9861549986155e-06, + "loss": 0.1294, + "step": 13852 + }, + { + "epoch": 6.440260344026035, + "grad_norm": 5.42917013168335, + "learning_rate": 9.986153998615401e-06, + "loss": 0.1742, + "step": 13853 + }, + { + "epoch": 6.4407252440725244, + "grad_norm": 5.117829322814941, + "learning_rate": 9.9861529986153e-06, + "loss": 0.1867, + "step": 13854 + }, + { + "epoch": 6.441190144119014, + "grad_norm": 5.885809421539307, + "learning_rate": 9.9861519986152e-06, + "loss": 0.1889, + "step": 13855 + }, + { + "epoch": 6.441655044165504, + "grad_norm": 6.743411064147949, + "learning_rate": 9.986150998615101e-06, + "loss": 0.1961, + "step": 13856 + }, + { + "epoch": 6.442119944211995, + "grad_norm": 9.417527198791504, + "learning_rate": 9.986149998615e-06, + "loss": 0.1233, + "step": 13857 + }, + { + "epoch": 6.4425848442584845, + "grad_norm": 8.464372634887695, + "learning_rate": 9.986148998614901e-06, + "loss": 0.3196, + "step": 13858 + }, + { + "epoch": 6.443049744304974, + "grad_norm": 3.989945888519287, + "learning_rate": 9.9861479986148e-06, + "loss": 0.0832, + "step": 13859 + }, + { + "epoch": 6.443514644351464, + "grad_norm": 7.2496514320373535, + "learning_rate": 9.9861469986147e-06, + "loss": 0.1828, + "step": 13860 + }, + { + "epoch": 6.443979544397955, + "grad_norm": 3.3810317516326904, + "learning_rate": 9.9861459986146e-06, + "loss": 0.092, + "step": 13861 + }, + { + "epoch": 6.444444444444445, + "grad_norm": 3.533003807067871, + "learning_rate": 9.9861449986145e-06, + "loss": 0.0801, + "step": 13862 + }, + { + "epoch": 6.4449093444909344, + "grad_norm": 6.433197975158691, + "learning_rate": 9.986143998614401e-06, + "loss": 0.2749, + "step": 13863 + }, + { + "epoch": 6.445374244537424, + "grad_norm": 6.546753883361816, + "learning_rate": 9.986142998614302e-06, + "loss": 0.3525, + "step": 13864 + }, + { + "epoch": 6.445839144583914, + "grad_norm": 5.418191432952881, + "learning_rate": 9.9861419986142e-06, + "loss": 0.1875, + "step": 13865 + }, + { + "epoch": 6.446304044630405, + "grad_norm": 5.209221839904785, + "learning_rate": 9.986140998614101e-06, + "loss": 0.2124, + "step": 13866 + }, + { + "epoch": 6.4467689446768945, + "grad_norm": 4.934913635253906, + "learning_rate": 9.986139998614e-06, + "loss": 0.1317, + "step": 13867 + }, + { + "epoch": 6.447233844723384, + "grad_norm": 4.356419086456299, + "learning_rate": 9.986138998613901e-06, + "loss": 0.1201, + "step": 13868 + }, + { + "epoch": 6.447698744769874, + "grad_norm": 7.004451751708984, + "learning_rate": 9.986137998613802e-06, + "loss": 0.447, + "step": 13869 + }, + { + "epoch": 6.448163644816365, + "grad_norm": 4.589147567749023, + "learning_rate": 9.9861369986137e-06, + "loss": 0.1513, + "step": 13870 + }, + { + "epoch": 6.448628544862855, + "grad_norm": 3.3883557319641113, + "learning_rate": 9.986135998613601e-06, + "loss": 0.0871, + "step": 13871 + }, + { + "epoch": 6.449093444909344, + "grad_norm": 5.60577917098999, + "learning_rate": 9.9861349986135e-06, + "loss": 0.1579, + "step": 13872 + }, + { + "epoch": 6.449558344955834, + "grad_norm": 10.440072059631348, + "learning_rate": 9.9861339986134e-06, + "loss": 0.3007, + "step": 13873 + }, + { + "epoch": 6.450023245002324, + "grad_norm": 2.27520751953125, + "learning_rate": 9.9861329986133e-06, + "loss": 0.0285, + "step": 13874 + }, + { + "epoch": 6.450488145048815, + "grad_norm": 6.041921615600586, + "learning_rate": 9.9861319986132e-06, + "loss": 0.1726, + "step": 13875 + }, + { + "epoch": 6.4509530450953045, + "grad_norm": 6.354124546051025, + "learning_rate": 9.986130998613101e-06, + "loss": 0.2771, + "step": 13876 + }, + { + "epoch": 6.451417945141794, + "grad_norm": 5.529216289520264, + "learning_rate": 9.986129998613002e-06, + "loss": 0.1815, + "step": 13877 + }, + { + "epoch": 6.451882845188284, + "grad_norm": 4.902965545654297, + "learning_rate": 9.9861289986129e-06, + "loss": 0.1521, + "step": 13878 + }, + { + "epoch": 6.452347745234775, + "grad_norm": 6.654168605804443, + "learning_rate": 9.9861279986128e-06, + "loss": 0.3059, + "step": 13879 + }, + { + "epoch": 6.452812645281265, + "grad_norm": 7.053252696990967, + "learning_rate": 9.9861269986127e-06, + "loss": 0.1928, + "step": 13880 + }, + { + "epoch": 6.453277545327754, + "grad_norm": 3.7977004051208496, + "learning_rate": 9.986125998612601e-06, + "loss": 0.0571, + "step": 13881 + }, + { + "epoch": 6.453742445374244, + "grad_norm": 4.725167751312256, + "learning_rate": 9.986124998612502e-06, + "loss": 0.1256, + "step": 13882 + }, + { + "epoch": 6.454207345420735, + "grad_norm": 6.666332244873047, + "learning_rate": 9.9861239986124e-06, + "loss": 0.2185, + "step": 13883 + }, + { + "epoch": 6.454672245467225, + "grad_norm": 5.363935947418213, + "learning_rate": 9.986122998612301e-06, + "loss": 0.1471, + "step": 13884 + }, + { + "epoch": 6.4551371455137145, + "grad_norm": 2.2815914154052734, + "learning_rate": 9.9861219986122e-06, + "loss": 0.0349, + "step": 13885 + }, + { + "epoch": 6.455602045560204, + "grad_norm": 4.896013259887695, + "learning_rate": 9.986120998612101e-06, + "loss": 0.1228, + "step": 13886 + }, + { + "epoch": 6.456066945606695, + "grad_norm": 7.202281951904297, + "learning_rate": 9.986119998612e-06, + "loss": 0.1958, + "step": 13887 + }, + { + "epoch": 6.456531845653185, + "grad_norm": 3.8330600261688232, + "learning_rate": 9.9861189986119e-06, + "loss": 0.0826, + "step": 13888 + }, + { + "epoch": 6.456996745699675, + "grad_norm": 5.321171760559082, + "learning_rate": 9.986117998611801e-06, + "loss": 0.1413, + "step": 13889 + }, + { + "epoch": 6.457461645746164, + "grad_norm": 6.488718509674072, + "learning_rate": 9.9861169986117e-06, + "loss": 0.2956, + "step": 13890 + }, + { + "epoch": 6.457926545792654, + "grad_norm": 7.162005424499512, + "learning_rate": 9.9861159986116e-06, + "loss": 0.2377, + "step": 13891 + }, + { + "epoch": 6.458391445839145, + "grad_norm": 6.1173787117004395, + "learning_rate": 9.9861149986115e-06, + "loss": 0.1593, + "step": 13892 + }, + { + "epoch": 6.458856345885635, + "grad_norm": 3.81620454788208, + "learning_rate": 9.9861139986114e-06, + "loss": 0.1303, + "step": 13893 + }, + { + "epoch": 6.4593212459321245, + "grad_norm": 12.709479331970215, + "learning_rate": 9.986112998611301e-06, + "loss": 0.5302, + "step": 13894 + }, + { + "epoch": 6.459786145978614, + "grad_norm": 13.012849807739258, + "learning_rate": 9.986111998611202e-06, + "loss": 0.452, + "step": 13895 + }, + { + "epoch": 6.460251046025105, + "grad_norm": 3.705258846282959, + "learning_rate": 9.9861109986111e-06, + "loss": 0.0436, + "step": 13896 + }, + { + "epoch": 6.460715946071595, + "grad_norm": 6.431516170501709, + "learning_rate": 9.986109998611001e-06, + "loss": 0.2422, + "step": 13897 + }, + { + "epoch": 6.461180846118085, + "grad_norm": 6.044456481933594, + "learning_rate": 9.9861089986109e-06, + "loss": 0.097, + "step": 13898 + }, + { + "epoch": 6.461645746164574, + "grad_norm": 11.017590522766113, + "learning_rate": 9.986107998610801e-06, + "loss": 0.1998, + "step": 13899 + }, + { + "epoch": 6.462110646211064, + "grad_norm": 5.4606242179870605, + "learning_rate": 9.9861069986107e-06, + "loss": 0.2081, + "step": 13900 + }, + { + "epoch": 6.462575546257555, + "grad_norm": 6.158010005950928, + "learning_rate": 9.9861059986106e-06, + "loss": 0.2213, + "step": 13901 + }, + { + "epoch": 6.463040446304045, + "grad_norm": 3.0929133892059326, + "learning_rate": 9.986104998610501e-06, + "loss": 0.0798, + "step": 13902 + }, + { + "epoch": 6.4635053463505345, + "grad_norm": 4.239352703094482, + "learning_rate": 9.9861039986104e-06, + "loss": 0.1112, + "step": 13903 + }, + { + "epoch": 6.463970246397024, + "grad_norm": 5.712252140045166, + "learning_rate": 9.986102998610301e-06, + "loss": 0.1818, + "step": 13904 + }, + { + "epoch": 6.464435146443515, + "grad_norm": 7.7506489753723145, + "learning_rate": 9.9861019986102e-06, + "loss": 0.391, + "step": 13905 + }, + { + "epoch": 6.464900046490005, + "grad_norm": 4.013493537902832, + "learning_rate": 9.9861009986101e-06, + "loss": 0.0692, + "step": 13906 + }, + { + "epoch": 6.465364946536495, + "grad_norm": 6.499163627624512, + "learning_rate": 9.986099998610001e-06, + "loss": 0.1129, + "step": 13907 + }, + { + "epoch": 6.465829846582984, + "grad_norm": 5.834417819976807, + "learning_rate": 9.986098998609902e-06, + "loss": 0.1388, + "step": 13908 + }, + { + "epoch": 6.466294746629474, + "grad_norm": 5.740723133087158, + "learning_rate": 9.9860979986098e-06, + "loss": 0.2161, + "step": 13909 + }, + { + "epoch": 6.466759646675965, + "grad_norm": 5.271947383880615, + "learning_rate": 9.986096998609701e-06, + "loss": 0.1321, + "step": 13910 + }, + { + "epoch": 6.467224546722455, + "grad_norm": 6.681092739105225, + "learning_rate": 9.9860959986096e-06, + "loss": 0.171, + "step": 13911 + }, + { + "epoch": 6.4676894467689445, + "grad_norm": 5.052267551422119, + "learning_rate": 9.9860949986095e-06, + "loss": 0.0642, + "step": 13912 + }, + { + "epoch": 6.468154346815434, + "grad_norm": 6.933216094970703, + "learning_rate": 9.9860939986094e-06, + "loss": 0.1585, + "step": 13913 + }, + { + "epoch": 6.468619246861925, + "grad_norm": 4.4842448234558105, + "learning_rate": 9.9860929986093e-06, + "loss": 0.1324, + "step": 13914 + }, + { + "epoch": 6.469084146908415, + "grad_norm": 5.738238334655762, + "learning_rate": 9.986091998609201e-06, + "loss": 0.1751, + "step": 13915 + }, + { + "epoch": 6.469549046954905, + "grad_norm": 5.969188213348389, + "learning_rate": 9.9860909986091e-06, + "loss": 0.1447, + "step": 13916 + }, + { + "epoch": 6.470013947001394, + "grad_norm": 6.861186981201172, + "learning_rate": 9.986089998609001e-06, + "loss": 0.2972, + "step": 13917 + }, + { + "epoch": 6.470478847047885, + "grad_norm": 8.215736389160156, + "learning_rate": 9.9860889986089e-06, + "loss": 0.2189, + "step": 13918 + }, + { + "epoch": 6.470943747094375, + "grad_norm": 2.433022975921631, + "learning_rate": 9.9860879986088e-06, + "loss": 0.0453, + "step": 13919 + }, + { + "epoch": 6.471408647140865, + "grad_norm": 5.420531272888184, + "learning_rate": 9.986086998608701e-06, + "loss": 0.0966, + "step": 13920 + }, + { + "epoch": 6.4718735471873545, + "grad_norm": 13.516361236572266, + "learning_rate": 9.986085998608602e-06, + "loss": 0.5851, + "step": 13921 + }, + { + "epoch": 6.472338447233844, + "grad_norm": 6.583310127258301, + "learning_rate": 9.9860849986085e-06, + "loss": 0.2343, + "step": 13922 + }, + { + "epoch": 6.472803347280335, + "grad_norm": 7.245143413543701, + "learning_rate": 9.9860839986084e-06, + "loss": 0.3924, + "step": 13923 + }, + { + "epoch": 6.473268247326825, + "grad_norm": 3.5617971420288086, + "learning_rate": 9.9860829986083e-06, + "loss": 0.08, + "step": 13924 + }, + { + "epoch": 6.473733147373315, + "grad_norm": 5.621847152709961, + "learning_rate": 9.986081998608201e-06, + "loss": 0.154, + "step": 13925 + }, + { + "epoch": 6.474198047419804, + "grad_norm": 6.53911018371582, + "learning_rate": 9.9860809986081e-06, + "loss": 0.1832, + "step": 13926 + }, + { + "epoch": 6.474662947466295, + "grad_norm": 3.643695116043091, + "learning_rate": 9.986079998608e-06, + "loss": 0.0545, + "step": 13927 + }, + { + "epoch": 6.475127847512785, + "grad_norm": 5.30234432220459, + "learning_rate": 9.986078998607901e-06, + "loss": 0.1422, + "step": 13928 + }, + { + "epoch": 6.475592747559275, + "grad_norm": 4.437175750732422, + "learning_rate": 9.9860779986078e-06, + "loss": 0.1742, + "step": 13929 + }, + { + "epoch": 6.4760576476057645, + "grad_norm": 6.600363731384277, + "learning_rate": 9.986076998607701e-06, + "loss": 0.2207, + "step": 13930 + }, + { + "epoch": 6.476522547652255, + "grad_norm": 4.521572113037109, + "learning_rate": 9.9860759986076e-06, + "loss": 0.18, + "step": 13931 + }, + { + "epoch": 6.476987447698745, + "grad_norm": 5.423354148864746, + "learning_rate": 9.9860749986075e-06, + "loss": 0.0944, + "step": 13932 + }, + { + "epoch": 6.477452347745235, + "grad_norm": 4.153341770172119, + "learning_rate": 9.986073998607401e-06, + "loss": 0.0822, + "step": 13933 + }, + { + "epoch": 6.477917247791725, + "grad_norm": 7.784372329711914, + "learning_rate": 9.9860729986073e-06, + "loss": 0.2369, + "step": 13934 + }, + { + "epoch": 6.478382147838214, + "grad_norm": 3.8107409477233887, + "learning_rate": 9.986071998607201e-06, + "loss": 0.1164, + "step": 13935 + }, + { + "epoch": 6.478847047884705, + "grad_norm": 6.6765456199646, + "learning_rate": 9.9860709986071e-06, + "loss": 0.3339, + "step": 13936 + }, + { + "epoch": 6.479311947931195, + "grad_norm": 3.484020471572876, + "learning_rate": 9.986069998607e-06, + "loss": 0.0918, + "step": 13937 + }, + { + "epoch": 6.479776847977685, + "grad_norm": 3.317411184310913, + "learning_rate": 9.986068998606901e-06, + "loss": 0.1045, + "step": 13938 + }, + { + "epoch": 6.4802417480241745, + "grad_norm": 3.578789234161377, + "learning_rate": 9.9860679986068e-06, + "loss": 0.0863, + "step": 13939 + }, + { + "epoch": 6.480706648070665, + "grad_norm": 4.021416187286377, + "learning_rate": 9.9860669986067e-06, + "loss": 0.0837, + "step": 13940 + }, + { + "epoch": 6.481171548117155, + "grad_norm": 3.978848457336426, + "learning_rate": 9.986065998606601e-06, + "loss": 0.1084, + "step": 13941 + }, + { + "epoch": 6.481636448163645, + "grad_norm": 8.569194793701172, + "learning_rate": 9.9860649986065e-06, + "loss": 0.3146, + "step": 13942 + }, + { + "epoch": 6.482101348210135, + "grad_norm": 6.833261013031006, + "learning_rate": 9.986063998606401e-06, + "loss": 0.1965, + "step": 13943 + }, + { + "epoch": 6.482566248256624, + "grad_norm": 7.884079933166504, + "learning_rate": 9.9860629986063e-06, + "loss": 0.3124, + "step": 13944 + }, + { + "epoch": 6.483031148303115, + "grad_norm": 4.936991214752197, + "learning_rate": 9.9860619986062e-06, + "loss": 0.0888, + "step": 13945 + }, + { + "epoch": 6.483496048349605, + "grad_norm": 4.937537670135498, + "learning_rate": 9.986060998606101e-06, + "loss": 0.1681, + "step": 13946 + }, + { + "epoch": 6.483960948396095, + "grad_norm": 4.786613941192627, + "learning_rate": 9.986059998606e-06, + "loss": 0.1837, + "step": 13947 + }, + { + "epoch": 6.4844258484425845, + "grad_norm": 5.399908065795898, + "learning_rate": 9.986058998605901e-06, + "loss": 0.2133, + "step": 13948 + }, + { + "epoch": 6.484890748489075, + "grad_norm": 2.3229715824127197, + "learning_rate": 9.9860579986058e-06, + "loss": 0.0216, + "step": 13949 + }, + { + "epoch": 6.485355648535565, + "grad_norm": 4.822716236114502, + "learning_rate": 9.9860569986057e-06, + "loss": 0.1142, + "step": 13950 + }, + { + "epoch": 6.485820548582055, + "grad_norm": 5.477962017059326, + "learning_rate": 9.986055998605601e-06, + "loss": 0.1671, + "step": 13951 + }, + { + "epoch": 6.486285448628545, + "grad_norm": 4.800859451293945, + "learning_rate": 9.986054998605502e-06, + "loss": 0.1467, + "step": 13952 + }, + { + "epoch": 6.486750348675034, + "grad_norm": 4.05147123336792, + "learning_rate": 9.986053998605401e-06, + "loss": 0.1043, + "step": 13953 + }, + { + "epoch": 6.487215248721525, + "grad_norm": 6.722812175750732, + "learning_rate": 9.986052998605302e-06, + "loss": 0.2025, + "step": 13954 + }, + { + "epoch": 6.487680148768015, + "grad_norm": 5.880919456481934, + "learning_rate": 9.9860519986052e-06, + "loss": 0.1686, + "step": 13955 + }, + { + "epoch": 6.488145048814505, + "grad_norm": 6.088743686676025, + "learning_rate": 9.9860509986051e-06, + "loss": 0.1679, + "step": 13956 + }, + { + "epoch": 6.4886099488609945, + "grad_norm": 6.083298206329346, + "learning_rate": 9.986049998605e-06, + "loss": 0.2626, + "step": 13957 + }, + { + "epoch": 6.489074848907485, + "grad_norm": 4.018389701843262, + "learning_rate": 9.9860489986049e-06, + "loss": 0.0692, + "step": 13958 + }, + { + "epoch": 6.489539748953975, + "grad_norm": 8.028854370117188, + "learning_rate": 9.986047998604801e-06, + "loss": 0.4573, + "step": 13959 + }, + { + "epoch": 6.490004649000465, + "grad_norm": 6.252965927124023, + "learning_rate": 9.9860469986047e-06, + "loss": 0.0723, + "step": 13960 + }, + { + "epoch": 6.490469549046955, + "grad_norm": 4.8975911140441895, + "learning_rate": 9.986045998604601e-06, + "loss": 0.1041, + "step": 13961 + }, + { + "epoch": 6.490934449093445, + "grad_norm": 4.4440226554870605, + "learning_rate": 9.9860449986045e-06, + "loss": 0.0852, + "step": 13962 + }, + { + "epoch": 6.491399349139935, + "grad_norm": 3.7478861808776855, + "learning_rate": 9.9860439986044e-06, + "loss": 0.0969, + "step": 13963 + }, + { + "epoch": 6.491864249186425, + "grad_norm": 5.28074312210083, + "learning_rate": 9.986042998604301e-06, + "loss": 0.1535, + "step": 13964 + }, + { + "epoch": 6.492329149232915, + "grad_norm": 4.628222465515137, + "learning_rate": 9.986041998604202e-06, + "loss": 0.1123, + "step": 13965 + }, + { + "epoch": 6.492794049279405, + "grad_norm": 4.679562568664551, + "learning_rate": 9.986040998604101e-06, + "loss": 0.1995, + "step": 13966 + }, + { + "epoch": 6.493258949325895, + "grad_norm": 9.539258003234863, + "learning_rate": 9.986039998604e-06, + "loss": 0.1087, + "step": 13967 + }, + { + "epoch": 6.493723849372385, + "grad_norm": 15.301848411560059, + "learning_rate": 9.9860389986039e-06, + "loss": 0.3311, + "step": 13968 + }, + { + "epoch": 6.494188749418875, + "grad_norm": 4.93301248550415, + "learning_rate": 9.9860379986038e-06, + "loss": 0.1028, + "step": 13969 + }, + { + "epoch": 6.494653649465365, + "grad_norm": 6.462350845336914, + "learning_rate": 9.9860369986037e-06, + "loss": 0.2365, + "step": 13970 + }, + { + "epoch": 6.495118549511855, + "grad_norm": 6.2060465812683105, + "learning_rate": 9.986035998603601e-06, + "loss": 0.2254, + "step": 13971 + }, + { + "epoch": 6.495583449558345, + "grad_norm": 5.004571914672852, + "learning_rate": 9.986034998603502e-06, + "loss": 0.1504, + "step": 13972 + }, + { + "epoch": 6.496048349604835, + "grad_norm": 6.783838272094727, + "learning_rate": 9.9860339986034e-06, + "loss": 0.2264, + "step": 13973 + }, + { + "epoch": 6.496513249651325, + "grad_norm": 4.922215461730957, + "learning_rate": 9.986032998603301e-06, + "loss": 0.2663, + "step": 13974 + }, + { + "epoch": 6.496978149697815, + "grad_norm": 5.72924280166626, + "learning_rate": 9.9860319986032e-06, + "loss": 0.2192, + "step": 13975 + }, + { + "epoch": 6.497443049744305, + "grad_norm": 7.80324125289917, + "learning_rate": 9.9860309986031e-06, + "loss": 0.2275, + "step": 13976 + }, + { + "epoch": 6.497907949790795, + "grad_norm": 4.640398979187012, + "learning_rate": 9.986029998603001e-06, + "loss": 0.1614, + "step": 13977 + }, + { + "epoch": 6.498372849837285, + "grad_norm": 6.542014122009277, + "learning_rate": 9.9860289986029e-06, + "loss": 0.2803, + "step": 13978 + }, + { + "epoch": 6.498837749883775, + "grad_norm": 4.432976245880127, + "learning_rate": 9.986027998602801e-06, + "loss": 0.1326, + "step": 13979 + }, + { + "epoch": 6.499302649930265, + "grad_norm": 4.550360202789307, + "learning_rate": 9.9860269986027e-06, + "loss": 0.0935, + "step": 13980 + }, + { + "epoch": 6.499767549976755, + "grad_norm": 3.3999111652374268, + "learning_rate": 9.9860259986026e-06, + "loss": 0.078, + "step": 13981 + }, + { + "epoch": 6.500232450023245, + "grad_norm": 4.624937534332275, + "learning_rate": 9.9860249986025e-06, + "loss": 0.2059, + "step": 13982 + }, + { + "epoch": 6.500697350069735, + "grad_norm": 7.149195671081543, + "learning_rate": 9.9860239986024e-06, + "loss": 0.1961, + "step": 13983 + }, + { + "epoch": 6.501162250116225, + "grad_norm": 5.068437099456787, + "learning_rate": 9.986022998602301e-06, + "loss": 0.2225, + "step": 13984 + }, + { + "epoch": 6.501627150162715, + "grad_norm": 3.5932531356811523, + "learning_rate": 9.986021998602202e-06, + "loss": 0.0711, + "step": 13985 + }, + { + "epoch": 6.502092050209205, + "grad_norm": 3.0710339546203613, + "learning_rate": 9.9860209986021e-06, + "loss": 0.0628, + "step": 13986 + }, + { + "epoch": 6.502556950255695, + "grad_norm": 4.679804801940918, + "learning_rate": 9.986019998602001e-06, + "loss": 0.0863, + "step": 13987 + }, + { + "epoch": 6.503021850302185, + "grad_norm": 6.899979114532471, + "learning_rate": 9.9860189986019e-06, + "loss": 0.2638, + "step": 13988 + }, + { + "epoch": 6.503486750348675, + "grad_norm": 5.328759670257568, + "learning_rate": 9.986017998601801e-06, + "loss": 0.1143, + "step": 13989 + }, + { + "epoch": 6.503951650395165, + "grad_norm": 6.140453815460205, + "learning_rate": 9.986016998601702e-06, + "loss": 0.2809, + "step": 13990 + }, + { + "epoch": 6.504416550441655, + "grad_norm": 5.580905914306641, + "learning_rate": 9.9860159986016e-06, + "loss": 0.1923, + "step": 13991 + }, + { + "epoch": 6.504881450488145, + "grad_norm": 7.688439846038818, + "learning_rate": 9.986014998601501e-06, + "loss": 0.2326, + "step": 13992 + }, + { + "epoch": 6.505346350534635, + "grad_norm": 7.23765230178833, + "learning_rate": 9.9860139986014e-06, + "loss": 0.1728, + "step": 13993 + }, + { + "epoch": 6.505811250581125, + "grad_norm": 4.903200149536133, + "learning_rate": 9.9860129986013e-06, + "loss": 0.2302, + "step": 13994 + }, + { + "epoch": 6.506276150627615, + "grad_norm": 5.424860000610352, + "learning_rate": 9.9860119986012e-06, + "loss": 0.243, + "step": 13995 + }, + { + "epoch": 6.506741050674105, + "grad_norm": 8.153656005859375, + "learning_rate": 9.9860109986011e-06, + "loss": 0.2505, + "step": 13996 + }, + { + "epoch": 6.507205950720595, + "grad_norm": 5.068619251251221, + "learning_rate": 9.986009998601001e-06, + "loss": 0.1816, + "step": 13997 + }, + { + "epoch": 6.507670850767085, + "grad_norm": 5.629489421844482, + "learning_rate": 9.986008998600902e-06, + "loss": 0.1668, + "step": 13998 + }, + { + "epoch": 6.508135750813575, + "grad_norm": 5.510096073150635, + "learning_rate": 9.9860079986008e-06, + "loss": 0.2684, + "step": 13999 + }, + { + "epoch": 6.508600650860065, + "grad_norm": 2.803464889526367, + "learning_rate": 9.9860069986007e-06, + "loss": 0.0408, + "step": 14000 + }, + { + "epoch": 6.509065550906556, + "grad_norm": 6.214261054992676, + "learning_rate": 9.9860059986006e-06, + "loss": 0.1544, + "step": 14001 + }, + { + "epoch": 6.509530450953045, + "grad_norm": 4.586818218231201, + "learning_rate": 9.986004998600501e-06, + "loss": 0.1033, + "step": 14002 + }, + { + "epoch": 6.509995350999535, + "grad_norm": 5.558227062225342, + "learning_rate": 9.986003998600402e-06, + "loss": 0.1674, + "step": 14003 + }, + { + "epoch": 6.510460251046025, + "grad_norm": 4.1055779457092285, + "learning_rate": 9.9860029986003e-06, + "loss": 0.0952, + "step": 14004 + }, + { + "epoch": 6.510925151092515, + "grad_norm": 3.2225756645202637, + "learning_rate": 9.986001998600201e-06, + "loss": 0.0555, + "step": 14005 + }, + { + "epoch": 6.5113900511390055, + "grad_norm": 5.664281368255615, + "learning_rate": 9.9860009986001e-06, + "loss": 0.1001, + "step": 14006 + }, + { + "epoch": 6.511854951185495, + "grad_norm": 4.291551113128662, + "learning_rate": 9.9859999986e-06, + "loss": 0.1898, + "step": 14007 + }, + { + "epoch": 6.512319851231985, + "grad_norm": 5.013302803039551, + "learning_rate": 9.9859989985999e-06, + "loss": 0.0906, + "step": 14008 + }, + { + "epoch": 6.512784751278475, + "grad_norm": 8.832080841064453, + "learning_rate": 9.9859979985998e-06, + "loss": 0.3115, + "step": 14009 + }, + { + "epoch": 6.513249651324966, + "grad_norm": 5.953103065490723, + "learning_rate": 9.985996998599701e-06, + "loss": 0.2127, + "step": 14010 + }, + { + "epoch": 6.513714551371455, + "grad_norm": 5.713141918182373, + "learning_rate": 9.9859959985996e-06, + "loss": 0.239, + "step": 14011 + }, + { + "epoch": 6.514179451417945, + "grad_norm": 4.012436389923096, + "learning_rate": 9.9859949985995e-06, + "loss": 0.1099, + "step": 14012 + }, + { + "epoch": 6.514644351464435, + "grad_norm": 5.942927837371826, + "learning_rate": 9.9859939985994e-06, + "loss": 0.1555, + "step": 14013 + }, + { + "epoch": 6.515109251510925, + "grad_norm": 5.586328983306885, + "learning_rate": 9.9859929985993e-06, + "loss": 0.2065, + "step": 14014 + }, + { + "epoch": 6.5155741515574155, + "grad_norm": 5.298583030700684, + "learning_rate": 9.985991998599201e-06, + "loss": 0.1193, + "step": 14015 + }, + { + "epoch": 6.516039051603905, + "grad_norm": 5.0072102546691895, + "learning_rate": 9.985990998599102e-06, + "loss": 0.1543, + "step": 14016 + }, + { + "epoch": 6.516503951650395, + "grad_norm": 6.218194484710693, + "learning_rate": 9.985989998599e-06, + "loss": 0.1906, + "step": 14017 + }, + { + "epoch": 6.516968851696885, + "grad_norm": 4.99752950668335, + "learning_rate": 9.985988998598901e-06, + "loss": 0.1899, + "step": 14018 + }, + { + "epoch": 6.517433751743376, + "grad_norm": 3.858098030090332, + "learning_rate": 9.9859879985988e-06, + "loss": 0.0816, + "step": 14019 + }, + { + "epoch": 6.517898651789865, + "grad_norm": 4.64828634262085, + "learning_rate": 9.985986998598701e-06, + "loss": 0.2157, + "step": 14020 + }, + { + "epoch": 6.518363551836355, + "grad_norm": 6.570058345794678, + "learning_rate": 9.985985998598602e-06, + "loss": 0.1662, + "step": 14021 + }, + { + "epoch": 6.518828451882845, + "grad_norm": 6.553529262542725, + "learning_rate": 9.9859849985985e-06, + "loss": 0.2509, + "step": 14022 + }, + { + "epoch": 6.519293351929335, + "grad_norm": 4.200013637542725, + "learning_rate": 9.985983998598401e-06, + "loss": 0.0937, + "step": 14023 + }, + { + "epoch": 6.5197582519758255, + "grad_norm": 5.23903226852417, + "learning_rate": 9.9859829985983e-06, + "loss": 0.1397, + "step": 14024 + }, + { + "epoch": 6.520223152022315, + "grad_norm": 4.351957321166992, + "learning_rate": 9.9859819985982e-06, + "loss": 0.1544, + "step": 14025 + }, + { + "epoch": 6.520688052068805, + "grad_norm": 3.026801347732544, + "learning_rate": 9.9859809985981e-06, + "loss": 0.036, + "step": 14026 + }, + { + "epoch": 6.521152952115295, + "grad_norm": 4.230859279632568, + "learning_rate": 9.985979998598e-06, + "loss": 0.1617, + "step": 14027 + }, + { + "epoch": 6.521617852161786, + "grad_norm": 5.461236476898193, + "learning_rate": 9.985978998597901e-06, + "loss": 0.2313, + "step": 14028 + }, + { + "epoch": 6.522082752208275, + "grad_norm": 7.8145904541015625, + "learning_rate": 9.985977998597802e-06, + "loss": 0.3255, + "step": 14029 + }, + { + "epoch": 6.522547652254765, + "grad_norm": 2.8280441761016846, + "learning_rate": 9.9859769985977e-06, + "loss": 0.0553, + "step": 14030 + }, + { + "epoch": 6.523012552301255, + "grad_norm": 4.277159214019775, + "learning_rate": 9.9859759985976e-06, + "loss": 0.092, + "step": 14031 + }, + { + "epoch": 6.523477452347745, + "grad_norm": 5.481252670288086, + "learning_rate": 9.9859749985975e-06, + "loss": 0.2517, + "step": 14032 + }, + { + "epoch": 6.5239423523942355, + "grad_norm": 5.972659111022949, + "learning_rate": 9.985973998597401e-06, + "loss": 0.3378, + "step": 14033 + }, + { + "epoch": 6.524407252440725, + "grad_norm": 7.749546527862549, + "learning_rate": 9.985972998597302e-06, + "loss": 0.1745, + "step": 14034 + }, + { + "epoch": 6.524872152487215, + "grad_norm": 5.2994384765625, + "learning_rate": 9.9859719985972e-06, + "loss": 0.1387, + "step": 14035 + }, + { + "epoch": 6.525337052533705, + "grad_norm": 7.967743873596191, + "learning_rate": 9.985970998597101e-06, + "loss": 0.2891, + "step": 14036 + }, + { + "epoch": 6.525801952580196, + "grad_norm": 5.97635555267334, + "learning_rate": 9.985969998597e-06, + "loss": 0.2688, + "step": 14037 + }, + { + "epoch": 6.526266852626685, + "grad_norm": 3.9523978233337402, + "learning_rate": 9.985968998596901e-06, + "loss": 0.0811, + "step": 14038 + }, + { + "epoch": 6.526731752673175, + "grad_norm": 5.52859354019165, + "learning_rate": 9.9859679985968e-06, + "loss": 0.2618, + "step": 14039 + }, + { + "epoch": 6.527196652719665, + "grad_norm": 4.667165756225586, + "learning_rate": 9.9859669985967e-06, + "loss": 0.1887, + "step": 14040 + }, + { + "epoch": 6.527661552766155, + "grad_norm": 2.479710340499878, + "learning_rate": 9.985965998596601e-06, + "loss": 0.061, + "step": 14041 + }, + { + "epoch": 6.5281264528126455, + "grad_norm": 6.99791145324707, + "learning_rate": 9.9859649985965e-06, + "loss": 0.223, + "step": 14042 + }, + { + "epoch": 6.528591352859135, + "grad_norm": 4.581587791442871, + "learning_rate": 9.9859639985964e-06, + "loss": 0.2395, + "step": 14043 + }, + { + "epoch": 6.529056252905625, + "grad_norm": 6.769725322723389, + "learning_rate": 9.9859629985963e-06, + "loss": 0.2772, + "step": 14044 + }, + { + "epoch": 6.529521152952116, + "grad_norm": 8.153871536254883, + "learning_rate": 9.9859619985962e-06, + "loss": 0.1496, + "step": 14045 + }, + { + "epoch": 6.529986052998606, + "grad_norm": 4.535538673400879, + "learning_rate": 9.985960998596101e-06, + "loss": 0.1068, + "step": 14046 + }, + { + "epoch": 6.530450953045095, + "grad_norm": 3.579989194869995, + "learning_rate": 9.985959998596002e-06, + "loss": 0.112, + "step": 14047 + }, + { + "epoch": 6.530915853091585, + "grad_norm": 4.589475154876709, + "learning_rate": 9.9859589985959e-06, + "loss": 0.2143, + "step": 14048 + }, + { + "epoch": 6.531380753138075, + "grad_norm": 8.486577033996582, + "learning_rate": 9.985957998595801e-06, + "loss": 0.3033, + "step": 14049 + }, + { + "epoch": 6.531845653184566, + "grad_norm": 6.348498344421387, + "learning_rate": 9.9859569985957e-06, + "loss": 0.17, + "step": 14050 + }, + { + "epoch": 6.5323105532310555, + "grad_norm": 4.05194091796875, + "learning_rate": 9.985955998595601e-06, + "loss": 0.1099, + "step": 14051 + }, + { + "epoch": 6.532775453277545, + "grad_norm": 4.1816606521606445, + "learning_rate": 9.9859549985955e-06, + "loss": 0.1407, + "step": 14052 + }, + { + "epoch": 6.533240353324035, + "grad_norm": 4.552303314208984, + "learning_rate": 9.9859539985954e-06, + "loss": 0.0652, + "step": 14053 + }, + { + "epoch": 6.533705253370526, + "grad_norm": 4.625299453735352, + "learning_rate": 9.985952998595301e-06, + "loss": 0.1486, + "step": 14054 + }, + { + "epoch": 6.534170153417016, + "grad_norm": 8.75671672821045, + "learning_rate": 9.9859519985952e-06, + "loss": 0.3155, + "step": 14055 + }, + { + "epoch": 6.534635053463505, + "grad_norm": 4.853709697723389, + "learning_rate": 9.985950998595101e-06, + "loss": 0.2131, + "step": 14056 + }, + { + "epoch": 6.535099953509995, + "grad_norm": 10.926682472229004, + "learning_rate": 9.985949998595e-06, + "loss": 0.3426, + "step": 14057 + }, + { + "epoch": 6.535564853556485, + "grad_norm": 4.269689083099365, + "learning_rate": 9.9859489985949e-06, + "loss": 0.1342, + "step": 14058 + }, + { + "epoch": 6.536029753602976, + "grad_norm": 6.307546138763428, + "learning_rate": 9.985947998594801e-06, + "loss": 0.2352, + "step": 14059 + }, + { + "epoch": 6.5364946536494655, + "grad_norm": 8.760581970214844, + "learning_rate": 9.985946998594702e-06, + "loss": 0.2453, + "step": 14060 + }, + { + "epoch": 6.536959553695955, + "grad_norm": 5.717742443084717, + "learning_rate": 9.9859459985946e-06, + "loss": 0.2184, + "step": 14061 + }, + { + "epoch": 6.537424453742445, + "grad_norm": 6.3740057945251465, + "learning_rate": 9.985944998594501e-06, + "loss": 0.2719, + "step": 14062 + }, + { + "epoch": 6.537889353788936, + "grad_norm": 5.658763885498047, + "learning_rate": 9.9859439985944e-06, + "loss": 0.1297, + "step": 14063 + }, + { + "epoch": 6.538354253835426, + "grad_norm": 7.0126447677612305, + "learning_rate": 9.9859429985943e-06, + "loss": 0.3604, + "step": 14064 + }, + { + "epoch": 6.538819153881915, + "grad_norm": 4.17099666595459, + "learning_rate": 9.9859419985942e-06, + "loss": 0.098, + "step": 14065 + }, + { + "epoch": 6.539284053928405, + "grad_norm": 4.935799598693848, + "learning_rate": 9.9859409985941e-06, + "loss": 0.2318, + "step": 14066 + }, + { + "epoch": 6.539748953974895, + "grad_norm": 4.551596641540527, + "learning_rate": 9.985939998594001e-06, + "loss": 0.0761, + "step": 14067 + }, + { + "epoch": 6.540213854021386, + "grad_norm": 4.501790523529053, + "learning_rate": 9.9859389985939e-06, + "loss": 0.0781, + "step": 14068 + }, + { + "epoch": 6.5406787540678755, + "grad_norm": 4.471104145050049, + "learning_rate": 9.985937998593801e-06, + "loss": 0.133, + "step": 14069 + }, + { + "epoch": 6.541143654114365, + "grad_norm": 5.5495500564575195, + "learning_rate": 9.9859369985937e-06, + "loss": 0.1927, + "step": 14070 + }, + { + "epoch": 6.541608554160855, + "grad_norm": 5.954348564147949, + "learning_rate": 9.9859359985936e-06, + "loss": 0.1092, + "step": 14071 + }, + { + "epoch": 6.542073454207346, + "grad_norm": 4.757513999938965, + "learning_rate": 9.985934998593501e-06, + "loss": 0.1283, + "step": 14072 + }, + { + "epoch": 6.542538354253836, + "grad_norm": 2.701437473297119, + "learning_rate": 9.985933998593402e-06, + "loss": 0.0339, + "step": 14073 + }, + { + "epoch": 6.543003254300325, + "grad_norm": 6.471463680267334, + "learning_rate": 9.985932998593301e-06, + "loss": 0.2462, + "step": 14074 + }, + { + "epoch": 6.543468154346815, + "grad_norm": 3.4702227115631104, + "learning_rate": 9.9859319985932e-06, + "loss": 0.0852, + "step": 14075 + }, + { + "epoch": 6.543933054393305, + "grad_norm": 6.7038655281066895, + "learning_rate": 9.9859309985931e-06, + "loss": 0.0566, + "step": 14076 + }, + { + "epoch": 6.544397954439796, + "grad_norm": 6.779550552368164, + "learning_rate": 9.985929998593e-06, + "loss": 0.0879, + "step": 14077 + }, + { + "epoch": 6.5448628544862855, + "grad_norm": 5.014877796173096, + "learning_rate": 9.9859289985929e-06, + "loss": 0.165, + "step": 14078 + }, + { + "epoch": 6.545327754532775, + "grad_norm": 7.76627779006958, + "learning_rate": 9.9859279985928e-06, + "loss": 0.1868, + "step": 14079 + }, + { + "epoch": 6.545792654579266, + "grad_norm": 4.1883111000061035, + "learning_rate": 9.985926998592701e-06, + "loss": 0.1241, + "step": 14080 + }, + { + "epoch": 6.546257554625756, + "grad_norm": 12.983302116394043, + "learning_rate": 9.9859259985926e-06, + "loss": 0.4377, + "step": 14081 + }, + { + "epoch": 6.546722454672246, + "grad_norm": 8.426118850708008, + "learning_rate": 9.985924998592501e-06, + "loss": 0.3378, + "step": 14082 + }, + { + "epoch": 6.547187354718735, + "grad_norm": 6.341713905334473, + "learning_rate": 9.9859239985924e-06, + "loss": 0.255, + "step": 14083 + }, + { + "epoch": 6.547652254765225, + "grad_norm": 4.122277736663818, + "learning_rate": 9.9859229985923e-06, + "loss": 0.1227, + "step": 14084 + }, + { + "epoch": 6.548117154811716, + "grad_norm": 8.532907485961914, + "learning_rate": 9.985921998592201e-06, + "loss": 0.2217, + "step": 14085 + }, + { + "epoch": 6.548582054858206, + "grad_norm": 4.745917320251465, + "learning_rate": 9.9859209985921e-06, + "loss": 0.1676, + "step": 14086 + }, + { + "epoch": 6.5490469549046955, + "grad_norm": 3.0533740520477295, + "learning_rate": 9.985919998592001e-06, + "loss": 0.0653, + "step": 14087 + }, + { + "epoch": 6.549511854951185, + "grad_norm": 5.741941452026367, + "learning_rate": 9.9859189985919e-06, + "loss": 0.1001, + "step": 14088 + }, + { + "epoch": 6.549976754997676, + "grad_norm": 5.769320964813232, + "learning_rate": 9.9859179985918e-06, + "loss": 0.2796, + "step": 14089 + }, + { + "epoch": 6.550441655044166, + "grad_norm": 3.4218826293945312, + "learning_rate": 9.985916998591701e-06, + "loss": 0.0943, + "step": 14090 + }, + { + "epoch": 6.550906555090656, + "grad_norm": 7.008244037628174, + "learning_rate": 9.9859159985916e-06, + "loss": 0.2525, + "step": 14091 + }, + { + "epoch": 6.551371455137145, + "grad_norm": 4.385026454925537, + "learning_rate": 9.985914998591501e-06, + "loss": 0.1734, + "step": 14092 + }, + { + "epoch": 6.551836355183635, + "grad_norm": 8.564901351928711, + "learning_rate": 9.985913998591402e-06, + "loss": 0.2986, + "step": 14093 + }, + { + "epoch": 6.552301255230126, + "grad_norm": 12.491021156311035, + "learning_rate": 9.9859129985913e-06, + "loss": 0.1728, + "step": 14094 + }, + { + "epoch": 6.552766155276616, + "grad_norm": 3.105670928955078, + "learning_rate": 9.985911998591201e-06, + "loss": 0.054, + "step": 14095 + }, + { + "epoch": 6.5532310553231055, + "grad_norm": 5.940182209014893, + "learning_rate": 9.9859109985911e-06, + "loss": 0.2446, + "step": 14096 + }, + { + "epoch": 6.553695955369595, + "grad_norm": 4.671454906463623, + "learning_rate": 9.985909998591e-06, + "loss": 0.1568, + "step": 14097 + }, + { + "epoch": 6.554160855416086, + "grad_norm": 2.3144216537475586, + "learning_rate": 9.985908998590901e-06, + "loss": 0.0362, + "step": 14098 + }, + { + "epoch": 6.554625755462576, + "grad_norm": 3.683743953704834, + "learning_rate": 9.9859079985908e-06, + "loss": 0.1183, + "step": 14099 + }, + { + "epoch": 6.5550906555090656, + "grad_norm": 4.059123516082764, + "learning_rate": 9.985906998590701e-06, + "loss": 0.1235, + "step": 14100 + }, + { + "epoch": 6.555555555555555, + "grad_norm": 3.9274353981018066, + "learning_rate": 9.9859059985906e-06, + "loss": 0.0741, + "step": 14101 + }, + { + "epoch": 6.556020455602045, + "grad_norm": 5.027818202972412, + "learning_rate": 9.9859049985905e-06, + "loss": 0.1023, + "step": 14102 + }, + { + "epoch": 6.556485355648536, + "grad_norm": 6.72341775894165, + "learning_rate": 9.985903998590401e-06, + "loss": 0.1937, + "step": 14103 + }, + { + "epoch": 6.556950255695026, + "grad_norm": 4.907838821411133, + "learning_rate": 9.9859029985903e-06, + "loss": 0.1652, + "step": 14104 + }, + { + "epoch": 6.5574151557415155, + "grad_norm": 4.721332550048828, + "learning_rate": 9.985901998590201e-06, + "loss": 0.1432, + "step": 14105 + }, + { + "epoch": 6.557880055788005, + "grad_norm": 3.7466306686401367, + "learning_rate": 9.985900998590102e-06, + "loss": 0.0957, + "step": 14106 + }, + { + "epoch": 6.558344955834496, + "grad_norm": 3.419408082962036, + "learning_rate": 9.98589999859e-06, + "loss": 0.1246, + "step": 14107 + }, + { + "epoch": 6.558809855880986, + "grad_norm": 8.156244277954102, + "learning_rate": 9.9858989985899e-06, + "loss": 0.1296, + "step": 14108 + }, + { + "epoch": 6.5592747559274756, + "grad_norm": 4.253251552581787, + "learning_rate": 9.9858979985898e-06, + "loss": 0.1177, + "step": 14109 + }, + { + "epoch": 6.559739655973965, + "grad_norm": 6.867100238800049, + "learning_rate": 9.9858969985897e-06, + "loss": 0.2026, + "step": 14110 + }, + { + "epoch": 6.560204556020455, + "grad_norm": 8.266400337219238, + "learning_rate": 9.985895998589601e-06, + "loss": 0.2914, + "step": 14111 + }, + { + "epoch": 6.560669456066946, + "grad_norm": 5.996044635772705, + "learning_rate": 9.9858949985895e-06, + "loss": 0.2105, + "step": 14112 + }, + { + "epoch": 6.561134356113436, + "grad_norm": 5.59277868270874, + "learning_rate": 9.985893998589401e-06, + "loss": 0.1337, + "step": 14113 + }, + { + "epoch": 6.5615992561599255, + "grad_norm": 4.605599403381348, + "learning_rate": 9.9858929985893e-06, + "loss": 0.1541, + "step": 14114 + }, + { + "epoch": 6.562064156206415, + "grad_norm": 6.599727153778076, + "learning_rate": 9.9858919985892e-06, + "loss": 0.1762, + "step": 14115 + }, + { + "epoch": 6.562529056252906, + "grad_norm": 6.215143203735352, + "learning_rate": 9.985890998589101e-06, + "loss": 0.2325, + "step": 14116 + }, + { + "epoch": 6.562993956299396, + "grad_norm": 5.9201202392578125, + "learning_rate": 9.985889998589002e-06, + "loss": 0.1932, + "step": 14117 + }, + { + "epoch": 6.5634588563458856, + "grad_norm": 5.748934268951416, + "learning_rate": 9.985888998588901e-06, + "loss": 0.2086, + "step": 14118 + }, + { + "epoch": 6.563923756392375, + "grad_norm": 7.195235729217529, + "learning_rate": 9.9858879985888e-06, + "loss": 0.2124, + "step": 14119 + }, + { + "epoch": 6.564388656438865, + "grad_norm": 5.834964275360107, + "learning_rate": 9.9858869985887e-06, + "loss": 0.1861, + "step": 14120 + }, + { + "epoch": 6.564853556485356, + "grad_norm": 3.5310609340667725, + "learning_rate": 9.9858859985886e-06, + "loss": 0.0897, + "step": 14121 + }, + { + "epoch": 6.565318456531846, + "grad_norm": 6.207219123840332, + "learning_rate": 9.9858849985885e-06, + "loss": 0.1873, + "step": 14122 + }, + { + "epoch": 6.5657833565783355, + "grad_norm": 8.72341537475586, + "learning_rate": 9.985883998588401e-06, + "loss": 0.235, + "step": 14123 + }, + { + "epoch": 6.566248256624826, + "grad_norm": 3.8916122913360596, + "learning_rate": 9.985882998588302e-06, + "loss": 0.0869, + "step": 14124 + }, + { + "epoch": 6.566713156671316, + "grad_norm": 3.4438722133636475, + "learning_rate": 9.9858819985882e-06, + "loss": 0.0647, + "step": 14125 + }, + { + "epoch": 6.567178056717806, + "grad_norm": 5.427014350891113, + "learning_rate": 9.985880998588101e-06, + "loss": 0.2027, + "step": 14126 + }, + { + "epoch": 6.5676429567642955, + "grad_norm": 5.942643642425537, + "learning_rate": 9.985879998588e-06, + "loss": 0.2851, + "step": 14127 + }, + { + "epoch": 6.568107856810785, + "grad_norm": 13.85257339477539, + "learning_rate": 9.9858789985879e-06, + "loss": 0.1647, + "step": 14128 + }, + { + "epoch": 6.568572756857276, + "grad_norm": 3.9647934436798096, + "learning_rate": 9.985877998587801e-06, + "loss": 0.1223, + "step": 14129 + }, + { + "epoch": 6.569037656903766, + "grad_norm": 8.449615478515625, + "learning_rate": 9.9858769985877e-06, + "loss": 0.2597, + "step": 14130 + }, + { + "epoch": 6.569502556950256, + "grad_norm": 6.326356887817383, + "learning_rate": 9.985875998587601e-06, + "loss": 0.201, + "step": 14131 + }, + { + "epoch": 6.5699674569967454, + "grad_norm": 5.645620822906494, + "learning_rate": 9.9858749985875e-06, + "loss": 0.1243, + "step": 14132 + }, + { + "epoch": 6.570432357043236, + "grad_norm": 2.3388335704803467, + "learning_rate": 9.9858739985874e-06, + "loss": 0.0404, + "step": 14133 + }, + { + "epoch": 6.570897257089726, + "grad_norm": 3.7626678943634033, + "learning_rate": 9.9858729985873e-06, + "loss": 0.0845, + "step": 14134 + }, + { + "epoch": 6.571362157136216, + "grad_norm": 6.430394649505615, + "learning_rate": 9.9858719985872e-06, + "loss": 0.2765, + "step": 14135 + }, + { + "epoch": 6.5718270571827055, + "grad_norm": 6.61348295211792, + "learning_rate": 9.985870998587101e-06, + "loss": 0.1751, + "step": 14136 + }, + { + "epoch": 6.572291957229195, + "grad_norm": 4.841862678527832, + "learning_rate": 9.985869998587002e-06, + "loss": 0.1508, + "step": 14137 + }, + { + "epoch": 6.572756857275686, + "grad_norm": 5.72727108001709, + "learning_rate": 9.9858689985869e-06, + "loss": 0.1437, + "step": 14138 + }, + { + "epoch": 6.573221757322176, + "grad_norm": 5.029782295227051, + "learning_rate": 9.985867998586801e-06, + "loss": 0.1655, + "step": 14139 + }, + { + "epoch": 6.573686657368666, + "grad_norm": 5.883672714233398, + "learning_rate": 9.9858669985867e-06, + "loss": 0.1973, + "step": 14140 + }, + { + "epoch": 6.5741515574151554, + "grad_norm": 3.0313522815704346, + "learning_rate": 9.985865998586601e-06, + "loss": 0.1294, + "step": 14141 + }, + { + "epoch": 6.574616457461646, + "grad_norm": 4.895979881286621, + "learning_rate": 9.985864998586502e-06, + "loss": 0.084, + "step": 14142 + }, + { + "epoch": 6.575081357508136, + "grad_norm": 5.383480072021484, + "learning_rate": 9.9858639985864e-06, + "loss": 0.1401, + "step": 14143 + }, + { + "epoch": 6.575546257554626, + "grad_norm": 5.049272060394287, + "learning_rate": 9.985862998586301e-06, + "loss": 0.1027, + "step": 14144 + }, + { + "epoch": 6.5760111576011155, + "grad_norm": 8.554121017456055, + "learning_rate": 9.9858619985862e-06, + "loss": 0.3873, + "step": 14145 + }, + { + "epoch": 6.576476057647605, + "grad_norm": 4.93639612197876, + "learning_rate": 9.9858609985861e-06, + "loss": 0.0669, + "step": 14146 + }, + { + "epoch": 6.576940957694096, + "grad_norm": 3.5920372009277344, + "learning_rate": 9.985859998586e-06, + "loss": 0.1274, + "step": 14147 + }, + { + "epoch": 6.577405857740586, + "grad_norm": 6.279752254486084, + "learning_rate": 9.9858589985859e-06, + "loss": 0.2764, + "step": 14148 + }, + { + "epoch": 6.577870757787076, + "grad_norm": 8.050798416137695, + "learning_rate": 9.985857998585801e-06, + "loss": 0.2404, + "step": 14149 + }, + { + "epoch": 6.578335657833565, + "grad_norm": 4.765344619750977, + "learning_rate": 9.9858569985857e-06, + "loss": 0.1535, + "step": 14150 + }, + { + "epoch": 6.578800557880056, + "grad_norm": 6.669418811798096, + "learning_rate": 9.9858559985856e-06, + "loss": 0.3639, + "step": 14151 + }, + { + "epoch": 6.579265457926546, + "grad_norm": 8.77941608428955, + "learning_rate": 9.9858549985855e-06, + "loss": 0.348, + "step": 14152 + }, + { + "epoch": 6.579730357973036, + "grad_norm": 5.1632304191589355, + "learning_rate": 9.9858539985854e-06, + "loss": 0.103, + "step": 14153 + }, + { + "epoch": 6.5801952580195255, + "grad_norm": 3.4326350688934326, + "learning_rate": 9.985852998585301e-06, + "loss": 0.0896, + "step": 14154 + }, + { + "epoch": 6.580660158066015, + "grad_norm": 3.8922178745269775, + "learning_rate": 9.985851998585202e-06, + "loss": 0.1499, + "step": 14155 + }, + { + "epoch": 6.581125058112506, + "grad_norm": 7.530762672424316, + "learning_rate": 9.9858509985851e-06, + "loss": 0.4252, + "step": 14156 + }, + { + "epoch": 6.581589958158996, + "grad_norm": 3.6872031688690186, + "learning_rate": 9.985849998585001e-06, + "loss": 0.1212, + "step": 14157 + }, + { + "epoch": 6.582054858205486, + "grad_norm": 4.947414398193359, + "learning_rate": 9.9858489985849e-06, + "loss": 0.1475, + "step": 14158 + }, + { + "epoch": 6.582519758251975, + "grad_norm": 7.1828694343566895, + "learning_rate": 9.985847998584801e-06, + "loss": 0.2421, + "step": 14159 + }, + { + "epoch": 6.582984658298466, + "grad_norm": 4.327578544616699, + "learning_rate": 9.9858469985847e-06, + "loss": 0.0887, + "step": 14160 + }, + { + "epoch": 6.583449558344956, + "grad_norm": 7.081507682800293, + "learning_rate": 9.9858459985846e-06, + "loss": 0.1992, + "step": 14161 + }, + { + "epoch": 6.583914458391446, + "grad_norm": 6.5038580894470215, + "learning_rate": 9.985844998584501e-06, + "loss": 0.1683, + "step": 14162 + }, + { + "epoch": 6.5843793584379355, + "grad_norm": 4.26938009262085, + "learning_rate": 9.9858439985844e-06, + "loss": 0.1298, + "step": 14163 + }, + { + "epoch": 6.584844258484426, + "grad_norm": 4.462749481201172, + "learning_rate": 9.9858429985843e-06, + "loss": 0.0992, + "step": 14164 + }, + { + "epoch": 6.585309158530916, + "grad_norm": 4.064214706420898, + "learning_rate": 9.9858419985842e-06, + "loss": 0.1007, + "step": 14165 + }, + { + "epoch": 6.585774058577406, + "grad_norm": 5.205019950866699, + "learning_rate": 9.9858409985841e-06, + "loss": 0.1261, + "step": 14166 + }, + { + "epoch": 6.586238958623896, + "grad_norm": 2.9730594158172607, + "learning_rate": 9.985839998584001e-06, + "loss": 0.0528, + "step": 14167 + }, + { + "epoch": 6.586703858670386, + "grad_norm": 6.759396076202393, + "learning_rate": 9.985838998583902e-06, + "loss": 0.1664, + "step": 14168 + }, + { + "epoch": 6.587168758716876, + "grad_norm": 5.4650726318359375, + "learning_rate": 9.9858379985838e-06, + "loss": 0.2003, + "step": 14169 + }, + { + "epoch": 6.587633658763366, + "grad_norm": 6.787993907928467, + "learning_rate": 9.985836998583701e-06, + "loss": 0.1659, + "step": 14170 + }, + { + "epoch": 6.588098558809856, + "grad_norm": 5.604122638702393, + "learning_rate": 9.9858359985836e-06, + "loss": 0.1653, + "step": 14171 + }, + { + "epoch": 6.5885634588563455, + "grad_norm": 9.898481369018555, + "learning_rate": 9.985834998583501e-06, + "loss": 0.2348, + "step": 14172 + }, + { + "epoch": 6.589028358902836, + "grad_norm": 9.60004711151123, + "learning_rate": 9.9858339985834e-06, + "loss": 0.2418, + "step": 14173 + }, + { + "epoch": 6.589493258949326, + "grad_norm": 6.572298526763916, + "learning_rate": 9.9858329985833e-06, + "loss": 0.2709, + "step": 14174 + }, + { + "epoch": 6.589958158995816, + "grad_norm": 5.389463901519775, + "learning_rate": 9.985831998583201e-06, + "loss": 0.1591, + "step": 14175 + }, + { + "epoch": 6.590423059042306, + "grad_norm": 5.149554252624512, + "learning_rate": 9.9858309985831e-06, + "loss": 0.1658, + "step": 14176 + }, + { + "epoch": 6.590887959088796, + "grad_norm": 7.258159160614014, + "learning_rate": 9.985829998583001e-06, + "loss": 0.2001, + "step": 14177 + }, + { + "epoch": 6.591352859135286, + "grad_norm": 5.8093438148498535, + "learning_rate": 9.9858289985829e-06, + "loss": 0.0566, + "step": 14178 + }, + { + "epoch": 6.591817759181776, + "grad_norm": 5.209879398345947, + "learning_rate": 9.9858279985828e-06, + "loss": 0.1319, + "step": 14179 + }, + { + "epoch": 6.592282659228266, + "grad_norm": 4.743191242218018, + "learning_rate": 9.985826998582701e-06, + "loss": 0.1886, + "step": 14180 + }, + { + "epoch": 6.5927475592747555, + "grad_norm": 4.5994873046875, + "learning_rate": 9.985825998582602e-06, + "loss": 0.1363, + "step": 14181 + }, + { + "epoch": 6.593212459321246, + "grad_norm": 4.637462139129639, + "learning_rate": 9.9858249985825e-06, + "loss": 0.0709, + "step": 14182 + }, + { + "epoch": 6.593677359367736, + "grad_norm": 3.9401679039001465, + "learning_rate": 9.9858239985824e-06, + "loss": 0.152, + "step": 14183 + }, + { + "epoch": 6.594142259414226, + "grad_norm": 3.3218579292297363, + "learning_rate": 9.9858229985823e-06, + "loss": 0.0587, + "step": 14184 + }, + { + "epoch": 6.594607159460716, + "grad_norm": 3.9424071311950684, + "learning_rate": 9.985821998582201e-06, + "loss": 0.0767, + "step": 14185 + }, + { + "epoch": 6.595072059507206, + "grad_norm": 5.236071586608887, + "learning_rate": 9.985820998582102e-06, + "loss": 0.178, + "step": 14186 + }, + { + "epoch": 6.595536959553696, + "grad_norm": 4.428808689117432, + "learning_rate": 9.985819998582e-06, + "loss": 0.101, + "step": 14187 + }, + { + "epoch": 6.596001859600186, + "grad_norm": 6.722596168518066, + "learning_rate": 9.985818998581901e-06, + "loss": 0.3074, + "step": 14188 + }, + { + "epoch": 6.596466759646676, + "grad_norm": 8.382952690124512, + "learning_rate": 9.9858179985818e-06, + "loss": 0.1646, + "step": 14189 + }, + { + "epoch": 6.5969316596931655, + "grad_norm": 3.3122799396514893, + "learning_rate": 9.985816998581701e-06, + "loss": 0.097, + "step": 14190 + }, + { + "epoch": 6.597396559739656, + "grad_norm": 5.256296634674072, + "learning_rate": 9.9858159985816e-06, + "loss": 0.1698, + "step": 14191 + }, + { + "epoch": 6.597861459786146, + "grad_norm": 3.3128833770751953, + "learning_rate": 9.9858149985815e-06, + "loss": 0.0499, + "step": 14192 + }, + { + "epoch": 6.598326359832636, + "grad_norm": 3.494614362716675, + "learning_rate": 9.985813998581401e-06, + "loss": 0.065, + "step": 14193 + }, + { + "epoch": 6.598791259879126, + "grad_norm": 6.149482250213623, + "learning_rate": 9.9858129985813e-06, + "loss": 0.3105, + "step": 14194 + }, + { + "epoch": 6.599256159925616, + "grad_norm": 5.839262962341309, + "learning_rate": 9.9858119985812e-06, + "loss": 0.291, + "step": 14195 + }, + { + "epoch": 6.599721059972106, + "grad_norm": 5.177228927612305, + "learning_rate": 9.9858109985811e-06, + "loss": 0.2176, + "step": 14196 + }, + { + "epoch": 6.600185960018596, + "grad_norm": 5.139754772186279, + "learning_rate": 9.985809998581e-06, + "loss": 0.226, + "step": 14197 + }, + { + "epoch": 6.600650860065086, + "grad_norm": 7.8893351554870605, + "learning_rate": 9.985808998580901e-06, + "loss": 0.16, + "step": 14198 + }, + { + "epoch": 6.6011157601115755, + "grad_norm": 6.663599014282227, + "learning_rate": 9.985807998580802e-06, + "loss": 0.1918, + "step": 14199 + }, + { + "epoch": 6.601580660158066, + "grad_norm": 7.70486307144165, + "learning_rate": 9.9858069985807e-06, + "loss": 0.2465, + "step": 14200 + }, + { + "epoch": 6.602045560204556, + "grad_norm": 4.93821907043457, + "learning_rate": 9.985805998580601e-06, + "loss": 0.1409, + "step": 14201 + }, + { + "epoch": 6.602510460251046, + "grad_norm": 9.40152359008789, + "learning_rate": 9.9858049985805e-06, + "loss": 0.3607, + "step": 14202 + }, + { + "epoch": 6.6029753602975365, + "grad_norm": 5.429227352142334, + "learning_rate": 9.985803998580401e-06, + "loss": 0.0794, + "step": 14203 + }, + { + "epoch": 6.603440260344026, + "grad_norm": 8.168869018554688, + "learning_rate": 9.9858029985803e-06, + "loss": 0.3903, + "step": 14204 + }, + { + "epoch": 6.603905160390516, + "grad_norm": 4.5090107917785645, + "learning_rate": 9.9858019985802e-06, + "loss": 0.0907, + "step": 14205 + }, + { + "epoch": 6.604370060437006, + "grad_norm": 3.9886739253997803, + "learning_rate": 9.985800998580101e-06, + "loss": 0.0999, + "step": 14206 + }, + { + "epoch": 6.604834960483496, + "grad_norm": 4.54143762588501, + "learning_rate": 9.98579999858e-06, + "loss": 0.1407, + "step": 14207 + }, + { + "epoch": 6.605299860529986, + "grad_norm": 5.272350311279297, + "learning_rate": 9.985798998579901e-06, + "loss": 0.2295, + "step": 14208 + }, + { + "epoch": 6.605764760576476, + "grad_norm": 5.539490222930908, + "learning_rate": 9.9857979985798e-06, + "loss": 0.1334, + "step": 14209 + }, + { + "epoch": 6.606229660622966, + "grad_norm": 32.955928802490234, + "learning_rate": 9.9857969985797e-06, + "loss": 0.4676, + "step": 14210 + }, + { + "epoch": 6.606694560669456, + "grad_norm": 6.805792808532715, + "learning_rate": 9.985795998579601e-06, + "loss": 0.3341, + "step": 14211 + }, + { + "epoch": 6.6071594607159465, + "grad_norm": 5.9815802574157715, + "learning_rate": 9.985794998579502e-06, + "loss": 0.2456, + "step": 14212 + }, + { + "epoch": 6.607624360762436, + "grad_norm": 8.417433738708496, + "learning_rate": 9.9857939985794e-06, + "loss": 0.0876, + "step": 14213 + }, + { + "epoch": 6.608089260808926, + "grad_norm": 2.9972143173217773, + "learning_rate": 9.985792998579301e-06, + "loss": 0.0473, + "step": 14214 + }, + { + "epoch": 6.608554160855416, + "grad_norm": 5.137458324432373, + "learning_rate": 9.9857919985792e-06, + "loss": 0.1695, + "step": 14215 + }, + { + "epoch": 6.609019060901906, + "grad_norm": 4.709127902984619, + "learning_rate": 9.9857909985791e-06, + "loss": 0.102, + "step": 14216 + }, + { + "epoch": 6.609483960948396, + "grad_norm": 7.304725646972656, + "learning_rate": 9.985789998579e-06, + "loss": 0.455, + "step": 14217 + }, + { + "epoch": 6.609948860994886, + "grad_norm": 6.961213111877441, + "learning_rate": 9.9857889985789e-06, + "loss": 0.2509, + "step": 14218 + }, + { + "epoch": 6.610413761041376, + "grad_norm": 3.85322904586792, + "learning_rate": 9.985787998578801e-06, + "loss": 0.0734, + "step": 14219 + }, + { + "epoch": 6.610878661087866, + "grad_norm": 7.088387966156006, + "learning_rate": 9.9857869985787e-06, + "loss": 0.2619, + "step": 14220 + }, + { + "epoch": 6.6113435611343565, + "grad_norm": 4.838310718536377, + "learning_rate": 9.985785998578601e-06, + "loss": 0.1985, + "step": 14221 + }, + { + "epoch": 6.611808461180846, + "grad_norm": 4.027819633483887, + "learning_rate": 9.9857849985785e-06, + "loss": 0.181, + "step": 14222 + }, + { + "epoch": 6.612273361227336, + "grad_norm": 6.477940559387207, + "learning_rate": 9.9857839985784e-06, + "loss": 0.2583, + "step": 14223 + }, + { + "epoch": 6.612738261273826, + "grad_norm": 8.809449195861816, + "learning_rate": 9.985782998578301e-06, + "loss": 0.2505, + "step": 14224 + }, + { + "epoch": 6.613203161320316, + "grad_norm": 8.067248344421387, + "learning_rate": 9.985781998578202e-06, + "loss": 0.284, + "step": 14225 + }, + { + "epoch": 6.613668061366806, + "grad_norm": 9.488947868347168, + "learning_rate": 9.985780998578101e-06, + "loss": 0.5005, + "step": 14226 + }, + { + "epoch": 6.614132961413296, + "grad_norm": 4.881049156188965, + "learning_rate": 9.985779998578e-06, + "loss": 0.1465, + "step": 14227 + }, + { + "epoch": 6.614597861459786, + "grad_norm": 6.650556564331055, + "learning_rate": 9.9857789985779e-06, + "loss": 0.2258, + "step": 14228 + }, + { + "epoch": 6.615062761506276, + "grad_norm": 5.1784844398498535, + "learning_rate": 9.9857779985778e-06, + "loss": 0.2016, + "step": 14229 + }, + { + "epoch": 6.6155276615527665, + "grad_norm": 6.614979267120361, + "learning_rate": 9.9857769985777e-06, + "loss": 0.2697, + "step": 14230 + }, + { + "epoch": 6.615992561599256, + "grad_norm": 7.382466793060303, + "learning_rate": 9.9857759985776e-06, + "loss": 0.3709, + "step": 14231 + }, + { + "epoch": 6.616457461645746, + "grad_norm": 5.755544662475586, + "learning_rate": 9.985774998577501e-06, + "loss": 0.1753, + "step": 14232 + }, + { + "epoch": 6.616922361692236, + "grad_norm": 6.867260456085205, + "learning_rate": 9.9857739985774e-06, + "loss": 0.2904, + "step": 14233 + }, + { + "epoch": 6.617387261738726, + "grad_norm": 7.4532318115234375, + "learning_rate": 9.985772998577301e-06, + "loss": 0.1488, + "step": 14234 + }, + { + "epoch": 6.617852161785216, + "grad_norm": 3.0806663036346436, + "learning_rate": 9.9857719985772e-06, + "loss": 0.0732, + "step": 14235 + }, + { + "epoch": 6.618317061831706, + "grad_norm": 6.596668720245361, + "learning_rate": 9.9857709985771e-06, + "loss": 0.2467, + "step": 14236 + }, + { + "epoch": 6.618781961878196, + "grad_norm": 2.039546489715576, + "learning_rate": 9.985769998577001e-06, + "loss": 0.0189, + "step": 14237 + }, + { + "epoch": 6.619246861924686, + "grad_norm": 6.118122100830078, + "learning_rate": 9.9857689985769e-06, + "loss": 0.1961, + "step": 14238 + }, + { + "epoch": 6.6197117619711765, + "grad_norm": 4.699887752532959, + "learning_rate": 9.985767998576801e-06, + "loss": 0.2227, + "step": 14239 + }, + { + "epoch": 6.620176662017666, + "grad_norm": 3.8652565479278564, + "learning_rate": 9.9857669985767e-06, + "loss": 0.0864, + "step": 14240 + }, + { + "epoch": 6.620641562064156, + "grad_norm": 5.750995635986328, + "learning_rate": 9.9857659985766e-06, + "loss": 0.2052, + "step": 14241 + }, + { + "epoch": 6.621106462110646, + "grad_norm": 6.14533805847168, + "learning_rate": 9.9857649985765e-06, + "loss": 0.1432, + "step": 14242 + }, + { + "epoch": 6.621571362157137, + "grad_norm": 6.63939094543457, + "learning_rate": 9.9857639985764e-06, + "loss": 0.2929, + "step": 14243 + }, + { + "epoch": 6.622036262203626, + "grad_norm": 8.600132942199707, + "learning_rate": 9.985762998576301e-06, + "loss": 0.3874, + "step": 14244 + }, + { + "epoch": 6.622501162250116, + "grad_norm": 6.978946685791016, + "learning_rate": 9.985761998576202e-06, + "loss": 0.204, + "step": 14245 + }, + { + "epoch": 6.622966062296606, + "grad_norm": 9.720340728759766, + "learning_rate": 9.9857609985761e-06, + "loss": 0.2971, + "step": 14246 + }, + { + "epoch": 6.623430962343097, + "grad_norm": 3.3148269653320312, + "learning_rate": 9.985759998576001e-06, + "loss": 0.0833, + "step": 14247 + }, + { + "epoch": 6.6238958623895865, + "grad_norm": 6.138078212738037, + "learning_rate": 9.9857589985759e-06, + "loss": 0.265, + "step": 14248 + }, + { + "epoch": 6.624360762436076, + "grad_norm": 5.76439905166626, + "learning_rate": 9.9857579985758e-06, + "loss": 0.2508, + "step": 14249 + }, + { + "epoch": 6.624825662482566, + "grad_norm": 7.9002838134765625, + "learning_rate": 9.985756998575701e-06, + "loss": 0.2091, + "step": 14250 + }, + { + "epoch": 6.625290562529056, + "grad_norm": 3.2670390605926514, + "learning_rate": 9.9857559985756e-06, + "loss": 0.0796, + "step": 14251 + }, + { + "epoch": 6.625755462575547, + "grad_norm": 6.3838677406311035, + "learning_rate": 9.985754998575501e-06, + "loss": 0.1434, + "step": 14252 + }, + { + "epoch": 6.626220362622036, + "grad_norm": 6.293051719665527, + "learning_rate": 9.9857539985754e-06, + "loss": 0.2384, + "step": 14253 + }, + { + "epoch": 6.626685262668526, + "grad_norm": 5.439201354980469, + "learning_rate": 9.9857529985753e-06, + "loss": 0.1364, + "step": 14254 + }, + { + "epoch": 6.627150162715016, + "grad_norm": 15.153665542602539, + "learning_rate": 9.985751998575201e-06, + "loss": 0.2768, + "step": 14255 + }, + { + "epoch": 6.627615062761507, + "grad_norm": 5.290113925933838, + "learning_rate": 9.9857509985751e-06, + "loss": 0.1487, + "step": 14256 + }, + { + "epoch": 6.6280799628079965, + "grad_norm": 6.840270042419434, + "learning_rate": 9.985749998575001e-06, + "loss": 0.2261, + "step": 14257 + }, + { + "epoch": 6.628544862854486, + "grad_norm": 4.603517055511475, + "learning_rate": 9.985748998574902e-06, + "loss": 0.1031, + "step": 14258 + }, + { + "epoch": 6.629009762900976, + "grad_norm": 5.798144817352295, + "learning_rate": 9.9857479985748e-06, + "loss": 0.1414, + "step": 14259 + }, + { + "epoch": 6.629474662947466, + "grad_norm": 4.530493259429932, + "learning_rate": 9.9857469985747e-06, + "loss": 0.1598, + "step": 14260 + }, + { + "epoch": 6.629939562993957, + "grad_norm": 5.681118488311768, + "learning_rate": 9.9857459985746e-06, + "loss": 0.1659, + "step": 14261 + }, + { + "epoch": 6.630404463040446, + "grad_norm": 3.6677329540252686, + "learning_rate": 9.985744998574501e-06, + "loss": 0.098, + "step": 14262 + }, + { + "epoch": 6.630869363086936, + "grad_norm": 4.879074573516846, + "learning_rate": 9.985743998574402e-06, + "loss": 0.1223, + "step": 14263 + }, + { + "epoch": 6.631334263133426, + "grad_norm": 5.0671515464782715, + "learning_rate": 9.9857429985743e-06, + "loss": 0.1776, + "step": 14264 + }, + { + "epoch": 6.631799163179917, + "grad_norm": 4.654116630554199, + "learning_rate": 9.985741998574201e-06, + "loss": 0.2327, + "step": 14265 + }, + { + "epoch": 6.6322640632264065, + "grad_norm": 6.1154093742370605, + "learning_rate": 9.9857409985741e-06, + "loss": 0.24, + "step": 14266 + }, + { + "epoch": 6.632728963272896, + "grad_norm": 5.3906402587890625, + "learning_rate": 9.985739998574e-06, + "loss": 0.0663, + "step": 14267 + }, + { + "epoch": 6.633193863319386, + "grad_norm": 4.524909496307373, + "learning_rate": 9.985738998573901e-06, + "loss": 0.1119, + "step": 14268 + }, + { + "epoch": 6.633658763365876, + "grad_norm": 4.894222259521484, + "learning_rate": 9.9857379985738e-06, + "loss": 0.1243, + "step": 14269 + }, + { + "epoch": 6.634123663412367, + "grad_norm": 7.271808624267578, + "learning_rate": 9.985736998573701e-06, + "loss": 0.3089, + "step": 14270 + }, + { + "epoch": 6.634588563458856, + "grad_norm": 4.89233922958374, + "learning_rate": 9.9857359985736e-06, + "loss": 0.1039, + "step": 14271 + }, + { + "epoch": 6.635053463505346, + "grad_norm": 6.701788425445557, + "learning_rate": 9.9857349985735e-06, + "loss": 0.3197, + "step": 14272 + }, + { + "epoch": 6.635518363551836, + "grad_norm": 5.422712802886963, + "learning_rate": 9.9857339985734e-06, + "loss": 0.2143, + "step": 14273 + }, + { + "epoch": 6.635983263598327, + "grad_norm": 11.130729675292969, + "learning_rate": 9.9857329985733e-06, + "loss": 0.1875, + "step": 14274 + }, + { + "epoch": 6.6364481636448165, + "grad_norm": 5.191876411437988, + "learning_rate": 9.985731998573201e-06, + "loss": 0.3209, + "step": 14275 + }, + { + "epoch": 6.636913063691306, + "grad_norm": 3.476752758026123, + "learning_rate": 9.985730998573102e-06, + "loss": 0.0731, + "step": 14276 + }, + { + "epoch": 6.637377963737796, + "grad_norm": 7.258742809295654, + "learning_rate": 9.985729998573e-06, + "loss": 0.3406, + "step": 14277 + }, + { + "epoch": 6.637842863784286, + "grad_norm": 7.326430320739746, + "learning_rate": 9.985728998572901e-06, + "loss": 0.1483, + "step": 14278 + }, + { + "epoch": 6.638307763830777, + "grad_norm": 7.87614631652832, + "learning_rate": 9.9857279985728e-06, + "loss": 0.3661, + "step": 14279 + }, + { + "epoch": 6.638772663877266, + "grad_norm": 3.3273816108703613, + "learning_rate": 9.9857269985727e-06, + "loss": 0.0913, + "step": 14280 + }, + { + "epoch": 6.639237563923756, + "grad_norm": 6.687269687652588, + "learning_rate": 9.985725998572602e-06, + "loss": 0.1898, + "step": 14281 + }, + { + "epoch": 6.639702463970247, + "grad_norm": 7.794702529907227, + "learning_rate": 9.9857249985725e-06, + "loss": 0.3721, + "step": 14282 + }, + { + "epoch": 6.640167364016737, + "grad_norm": 4.061814785003662, + "learning_rate": 9.985723998572401e-06, + "loss": 0.1237, + "step": 14283 + }, + { + "epoch": 6.6406322640632265, + "grad_norm": 7.472151756286621, + "learning_rate": 9.9857229985723e-06, + "loss": 0.268, + "step": 14284 + }, + { + "epoch": 6.641097164109716, + "grad_norm": 5.255150318145752, + "learning_rate": 9.9857219985722e-06, + "loss": 0.1865, + "step": 14285 + }, + { + "epoch": 6.641562064156206, + "grad_norm": 4.649481296539307, + "learning_rate": 9.9857209985721e-06, + "loss": 0.1248, + "step": 14286 + }, + { + "epoch": 6.642026964202697, + "grad_norm": 6.542163848876953, + "learning_rate": 9.985719998572e-06, + "loss": 0.307, + "step": 14287 + }, + { + "epoch": 6.642491864249187, + "grad_norm": 5.524788856506348, + "learning_rate": 9.985718998571901e-06, + "loss": 0.1996, + "step": 14288 + }, + { + "epoch": 6.642956764295676, + "grad_norm": 5.607463836669922, + "learning_rate": 9.985717998571802e-06, + "loss": 0.1305, + "step": 14289 + }, + { + "epoch": 6.643421664342166, + "grad_norm": 4.160898685455322, + "learning_rate": 9.9857169985717e-06, + "loss": 0.0981, + "step": 14290 + }, + { + "epoch": 6.643886564388657, + "grad_norm": 3.2975354194641113, + "learning_rate": 9.9857159985716e-06, + "loss": 0.0705, + "step": 14291 + }, + { + "epoch": 6.644351464435147, + "grad_norm": 5.144852638244629, + "learning_rate": 9.9857149985715e-06, + "loss": 0.1637, + "step": 14292 + }, + { + "epoch": 6.6448163644816365, + "grad_norm": 5.105834484100342, + "learning_rate": 9.985713998571401e-06, + "loss": 0.1674, + "step": 14293 + }, + { + "epoch": 6.645281264528126, + "grad_norm": 5.740676403045654, + "learning_rate": 9.985712998571302e-06, + "loss": 0.1666, + "step": 14294 + }, + { + "epoch": 6.645746164574616, + "grad_norm": 11.188611030578613, + "learning_rate": 9.9857119985712e-06, + "loss": 0.4372, + "step": 14295 + }, + { + "epoch": 6.646211064621107, + "grad_norm": 1.9829277992248535, + "learning_rate": 9.985710998571101e-06, + "loss": 0.0399, + "step": 14296 + }, + { + "epoch": 6.646675964667597, + "grad_norm": 9.538044929504395, + "learning_rate": 9.985709998571e-06, + "loss": 0.364, + "step": 14297 + }, + { + "epoch": 6.647140864714086, + "grad_norm": 7.034855842590332, + "learning_rate": 9.9857089985709e-06, + "loss": 0.2375, + "step": 14298 + }, + { + "epoch": 6.647605764760576, + "grad_norm": 7.1064276695251465, + "learning_rate": 9.9857079985708e-06, + "loss": 0.2221, + "step": 14299 + }, + { + "epoch": 6.648070664807067, + "grad_norm": 7.0441107749938965, + "learning_rate": 9.9857069985707e-06, + "loss": 0.3764, + "step": 14300 + }, + { + "epoch": 6.648535564853557, + "grad_norm": 5.242641925811768, + "learning_rate": 9.985705998570601e-06, + "loss": 0.1437, + "step": 14301 + }, + { + "epoch": 6.6490004649000465, + "grad_norm": 3.774461030960083, + "learning_rate": 9.9857049985705e-06, + "loss": 0.0679, + "step": 14302 + }, + { + "epoch": 6.649465364946536, + "grad_norm": 8.38928508758545, + "learning_rate": 9.9857039985704e-06, + "loss": 0.3863, + "step": 14303 + }, + { + "epoch": 6.649930264993026, + "grad_norm": 9.130131721496582, + "learning_rate": 9.9857029985703e-06, + "loss": 0.3504, + "step": 14304 + }, + { + "epoch": 6.650395165039517, + "grad_norm": 5.607924461364746, + "learning_rate": 9.9857019985702e-06, + "loss": 0.2685, + "step": 14305 + }, + { + "epoch": 6.650860065086007, + "grad_norm": 4.922170639038086, + "learning_rate": 9.985700998570101e-06, + "loss": 0.1973, + "step": 14306 + }, + { + "epoch": 6.651324965132496, + "grad_norm": 6.768627166748047, + "learning_rate": 9.985699998570002e-06, + "loss": 0.0884, + "step": 14307 + }, + { + "epoch": 6.651789865178986, + "grad_norm": 3.491962194442749, + "learning_rate": 9.9856989985699e-06, + "loss": 0.0948, + "step": 14308 + }, + { + "epoch": 6.652254765225477, + "grad_norm": 8.642671585083008, + "learning_rate": 9.985697998569801e-06, + "loss": 0.3412, + "step": 14309 + }, + { + "epoch": 6.652719665271967, + "grad_norm": 4.3749213218688965, + "learning_rate": 9.9856969985697e-06, + "loss": 0.1302, + "step": 14310 + }, + { + "epoch": 6.6531845653184565, + "grad_norm": 4.230331897735596, + "learning_rate": 9.985695998569601e-06, + "loss": 0.1044, + "step": 14311 + }, + { + "epoch": 6.653649465364946, + "grad_norm": 4.802365779876709, + "learning_rate": 9.9856949985695e-06, + "loss": 0.1049, + "step": 14312 + }, + { + "epoch": 6.654114365411436, + "grad_norm": 4.351428031921387, + "learning_rate": 9.9856939985694e-06, + "loss": 0.116, + "step": 14313 + }, + { + "epoch": 6.654579265457927, + "grad_norm": 7.47017240524292, + "learning_rate": 9.985692998569301e-06, + "loss": 0.2263, + "step": 14314 + }, + { + "epoch": 6.655044165504417, + "grad_norm": 6.281847953796387, + "learning_rate": 9.9856919985692e-06, + "loss": 0.2304, + "step": 14315 + }, + { + "epoch": 6.655509065550906, + "grad_norm": 6.937926769256592, + "learning_rate": 9.9856909985691e-06, + "loss": 0.2855, + "step": 14316 + }, + { + "epoch": 6.655973965597396, + "grad_norm": 8.480049133300781, + "learning_rate": 9.985689998569e-06, + "loss": 0.4564, + "step": 14317 + }, + { + "epoch": 6.656438865643887, + "grad_norm": 7.871381759643555, + "learning_rate": 9.9856889985689e-06, + "loss": 0.2966, + "step": 14318 + }, + { + "epoch": 6.656903765690377, + "grad_norm": 5.683877468109131, + "learning_rate": 9.985687998568801e-06, + "loss": 0.1091, + "step": 14319 + }, + { + "epoch": 6.6573686657368665, + "grad_norm": 5.827840805053711, + "learning_rate": 9.985686998568702e-06, + "loss": 0.1501, + "step": 14320 + }, + { + "epoch": 6.657833565783356, + "grad_norm": 5.749847888946533, + "learning_rate": 9.9856859985686e-06, + "loss": 0.1762, + "step": 14321 + }, + { + "epoch": 6.658298465829847, + "grad_norm": 5.1470136642456055, + "learning_rate": 9.985684998568501e-06, + "loss": 0.2352, + "step": 14322 + }, + { + "epoch": 6.658763365876337, + "grad_norm": 7.742006778717041, + "learning_rate": 9.9856839985684e-06, + "loss": 0.1399, + "step": 14323 + }, + { + "epoch": 6.659228265922827, + "grad_norm": 5.58524751663208, + "learning_rate": 9.985682998568301e-06, + "loss": 0.1526, + "step": 14324 + }, + { + "epoch": 6.659693165969316, + "grad_norm": 4.800800800323486, + "learning_rate": 9.9856819985682e-06, + "loss": 0.201, + "step": 14325 + }, + { + "epoch": 6.660158066015807, + "grad_norm": 3.812044143676758, + "learning_rate": 9.9856809985681e-06, + "loss": 0.0964, + "step": 14326 + }, + { + "epoch": 6.660622966062297, + "grad_norm": 6.9203572273254395, + "learning_rate": 9.985679998568001e-06, + "loss": 0.2247, + "step": 14327 + }, + { + "epoch": 6.661087866108787, + "grad_norm": 6.914507865905762, + "learning_rate": 9.9856789985679e-06, + "loss": 0.1671, + "step": 14328 + }, + { + "epoch": 6.6615527661552765, + "grad_norm": 5.491410255432129, + "learning_rate": 9.985677998567801e-06, + "loss": 0.2273, + "step": 14329 + }, + { + "epoch": 6.662017666201766, + "grad_norm": 2.9040591716766357, + "learning_rate": 9.9856769985677e-06, + "loss": 0.0693, + "step": 14330 + }, + { + "epoch": 6.662482566248257, + "grad_norm": 2.7864465713500977, + "learning_rate": 9.9856759985676e-06, + "loss": 0.048, + "step": 14331 + }, + { + "epoch": 6.662947466294747, + "grad_norm": 5.1083478927612305, + "learning_rate": 9.985674998567501e-06, + "loss": 0.196, + "step": 14332 + }, + { + "epoch": 6.663412366341237, + "grad_norm": 9.617266654968262, + "learning_rate": 9.985673998567402e-06, + "loss": 0.2713, + "step": 14333 + }, + { + "epoch": 6.663877266387726, + "grad_norm": 3.6975998878479004, + "learning_rate": 9.9856729985673e-06, + "loss": 0.0939, + "step": 14334 + }, + { + "epoch": 6.664342166434217, + "grad_norm": 5.542099475860596, + "learning_rate": 9.9856719985672e-06, + "loss": 0.1332, + "step": 14335 + }, + { + "epoch": 6.664807066480707, + "grad_norm": 6.244939804077148, + "learning_rate": 9.9856709985671e-06, + "loss": 0.3657, + "step": 14336 + }, + { + "epoch": 6.665271966527197, + "grad_norm": 8.568846702575684, + "learning_rate": 9.985669998567001e-06, + "loss": 0.3486, + "step": 14337 + }, + { + "epoch": 6.6657368665736865, + "grad_norm": 5.574740886688232, + "learning_rate": 9.9856689985669e-06, + "loss": 0.2187, + "step": 14338 + }, + { + "epoch": 6.666201766620176, + "grad_norm": 3.740440845489502, + "learning_rate": 9.9856679985668e-06, + "loss": 0.0539, + "step": 14339 + }, + { + "epoch": 6.666666666666667, + "grad_norm": 4.5341339111328125, + "learning_rate": 9.985666998566701e-06, + "loss": 0.1647, + "step": 14340 + }, + { + "epoch": 6.667131566713157, + "grad_norm": 4.611595153808594, + "learning_rate": 9.9856659985666e-06, + "loss": 0.1249, + "step": 14341 + }, + { + "epoch": 6.667596466759647, + "grad_norm": 8.378387451171875, + "learning_rate": 9.985664998566501e-06, + "loss": 0.2736, + "step": 14342 + }, + { + "epoch": 6.668061366806136, + "grad_norm": 2.7614431381225586, + "learning_rate": 9.9856639985664e-06, + "loss": 0.0623, + "step": 14343 + }, + { + "epoch": 6.668526266852627, + "grad_norm": 8.821901321411133, + "learning_rate": 9.9856629985663e-06, + "loss": 0.3457, + "step": 14344 + }, + { + "epoch": 6.668991166899117, + "grad_norm": 3.1455421447753906, + "learning_rate": 9.985661998566201e-06, + "loss": 0.0611, + "step": 14345 + }, + { + "epoch": 6.669456066945607, + "grad_norm": 4.751759052276611, + "learning_rate": 9.9856609985661e-06, + "loss": 0.0991, + "step": 14346 + }, + { + "epoch": 6.6699209669920965, + "grad_norm": 3.398252248764038, + "learning_rate": 9.985659998566001e-06, + "loss": 0.0349, + "step": 14347 + }, + { + "epoch": 6.670385867038586, + "grad_norm": 7.13319730758667, + "learning_rate": 9.9856589985659e-06, + "loss": 0.1787, + "step": 14348 + }, + { + "epoch": 6.670850767085077, + "grad_norm": 6.822784900665283, + "learning_rate": 9.9856579985658e-06, + "loss": 0.2085, + "step": 14349 + }, + { + "epoch": 6.671315667131567, + "grad_norm": 8.513250350952148, + "learning_rate": 9.985656998565701e-06, + "loss": 0.3025, + "step": 14350 + }, + { + "epoch": 6.671780567178057, + "grad_norm": 2.7307660579681396, + "learning_rate": 9.985655998565602e-06, + "loss": 0.041, + "step": 14351 + }, + { + "epoch": 6.672245467224546, + "grad_norm": 7.189734935760498, + "learning_rate": 9.9856549985655e-06, + "loss": 0.4497, + "step": 14352 + }, + { + "epoch": 6.672710367271037, + "grad_norm": 6.667140960693359, + "learning_rate": 9.985653998565401e-06, + "loss": 0.2225, + "step": 14353 + }, + { + "epoch": 6.673175267317527, + "grad_norm": 8.029425621032715, + "learning_rate": 9.9856529985653e-06, + "loss": 0.3175, + "step": 14354 + }, + { + "epoch": 6.673640167364017, + "grad_norm": 6.883372783660889, + "learning_rate": 9.985651998565201e-06, + "loss": 0.3335, + "step": 14355 + }, + { + "epoch": 6.6741050674105065, + "grad_norm": 7.316263675689697, + "learning_rate": 9.9856509985651e-06, + "loss": 0.1746, + "step": 14356 + }, + { + "epoch": 6.674569967456996, + "grad_norm": 4.68596887588501, + "learning_rate": 9.985649998565e-06, + "loss": 0.07, + "step": 14357 + }, + { + "epoch": 6.675034867503487, + "grad_norm": 9.458901405334473, + "learning_rate": 9.985648998564901e-06, + "loss": 0.2572, + "step": 14358 + }, + { + "epoch": 6.675499767549977, + "grad_norm": 4.766493320465088, + "learning_rate": 9.9856479985648e-06, + "loss": 0.133, + "step": 14359 + }, + { + "epoch": 6.675964667596467, + "grad_norm": 7.037400245666504, + "learning_rate": 9.985646998564701e-06, + "loss": 0.2792, + "step": 14360 + }, + { + "epoch": 6.676429567642957, + "grad_norm": 3.7772202491760254, + "learning_rate": 9.9856459985646e-06, + "loss": 0.1093, + "step": 14361 + }, + { + "epoch": 6.676894467689447, + "grad_norm": 6.569474697113037, + "learning_rate": 9.9856449985645e-06, + "loss": 0.3006, + "step": 14362 + }, + { + "epoch": 6.677359367735937, + "grad_norm": 3.5180466175079346, + "learning_rate": 9.985643998564401e-06, + "loss": 0.0645, + "step": 14363 + }, + { + "epoch": 6.677824267782427, + "grad_norm": 7.975086212158203, + "learning_rate": 9.985642998564302e-06, + "loss": 0.1092, + "step": 14364 + }, + { + "epoch": 6.6782891678289165, + "grad_norm": 7.850602626800537, + "learning_rate": 9.985641998564201e-06, + "loss": 0.2374, + "step": 14365 + }, + { + "epoch": 6.678754067875407, + "grad_norm": 2.679093599319458, + "learning_rate": 9.985640998564102e-06, + "loss": 0.0528, + "step": 14366 + }, + { + "epoch": 6.679218967921897, + "grad_norm": 6.027029037475586, + "learning_rate": 9.985639998564e-06, + "loss": 0.1896, + "step": 14367 + }, + { + "epoch": 6.679683867968387, + "grad_norm": 3.133366823196411, + "learning_rate": 9.9856389985639e-06, + "loss": 0.0773, + "step": 14368 + }, + { + "epoch": 6.680148768014877, + "grad_norm": 2.582664728164673, + "learning_rate": 9.9856379985638e-06, + "loss": 0.0525, + "step": 14369 + }, + { + "epoch": 6.680613668061367, + "grad_norm": 3.823504686355591, + "learning_rate": 9.9856369985637e-06, + "loss": 0.1401, + "step": 14370 + }, + { + "epoch": 6.681078568107857, + "grad_norm": 8.3486328125, + "learning_rate": 9.985635998563601e-06, + "loss": 0.3534, + "step": 14371 + }, + { + "epoch": 6.681543468154347, + "grad_norm": 8.052895545959473, + "learning_rate": 9.9856349985635e-06, + "loss": 0.1937, + "step": 14372 + }, + { + "epoch": 6.682008368200837, + "grad_norm": 6.327796459197998, + "learning_rate": 9.985633998563401e-06, + "loss": 0.0957, + "step": 14373 + }, + { + "epoch": 6.6824732682473265, + "grad_norm": 4.323028087615967, + "learning_rate": 9.9856329985633e-06, + "loss": 0.125, + "step": 14374 + }, + { + "epoch": 6.682938168293817, + "grad_norm": 5.0326738357543945, + "learning_rate": 9.9856319985632e-06, + "loss": 0.1676, + "step": 14375 + }, + { + "epoch": 6.683403068340307, + "grad_norm": 5.2501726150512695, + "learning_rate": 9.985630998563101e-06, + "loss": 0.2146, + "step": 14376 + }, + { + "epoch": 6.683867968386797, + "grad_norm": 4.881181240081787, + "learning_rate": 9.985629998563002e-06, + "loss": 0.1045, + "step": 14377 + }, + { + "epoch": 6.684332868433287, + "grad_norm": 5.623345375061035, + "learning_rate": 9.985628998562901e-06, + "loss": 0.2496, + "step": 14378 + }, + { + "epoch": 6.684797768479777, + "grad_norm": 2.9169576168060303, + "learning_rate": 9.9856279985628e-06, + "loss": 0.0583, + "step": 14379 + }, + { + "epoch": 6.685262668526267, + "grad_norm": 4.8905253410339355, + "learning_rate": 9.9856269985627e-06, + "loss": 0.1601, + "step": 14380 + }, + { + "epoch": 6.685727568572757, + "grad_norm": 6.526998043060303, + "learning_rate": 9.9856259985626e-06, + "loss": 0.2441, + "step": 14381 + }, + { + "epoch": 6.686192468619247, + "grad_norm": 5.004676342010498, + "learning_rate": 9.9856249985625e-06, + "loss": 0.1428, + "step": 14382 + }, + { + "epoch": 6.6866573686657365, + "grad_norm": 5.613152503967285, + "learning_rate": 9.9856239985624e-06, + "loss": 0.2516, + "step": 14383 + }, + { + "epoch": 6.687122268712227, + "grad_norm": 6.395786762237549, + "learning_rate": 9.985622998562301e-06, + "loss": 0.2757, + "step": 14384 + }, + { + "epoch": 6.687587168758717, + "grad_norm": 9.563292503356934, + "learning_rate": 9.9856219985622e-06, + "loss": 0.2443, + "step": 14385 + }, + { + "epoch": 6.688052068805207, + "grad_norm": 3.8715922832489014, + "learning_rate": 9.985620998562101e-06, + "loss": 0.1113, + "step": 14386 + }, + { + "epoch": 6.6885169688516966, + "grad_norm": 4.171772480010986, + "learning_rate": 9.985619998562e-06, + "loss": 0.1093, + "step": 14387 + }, + { + "epoch": 6.688981868898187, + "grad_norm": 4.815847396850586, + "learning_rate": 9.9856189985619e-06, + "loss": 0.1631, + "step": 14388 + }, + { + "epoch": 6.689446768944677, + "grad_norm": 4.6942877769470215, + "learning_rate": 9.985617998561801e-06, + "loss": 0.126, + "step": 14389 + }, + { + "epoch": 6.689911668991167, + "grad_norm": 5.5921454429626465, + "learning_rate": 9.9856169985617e-06, + "loss": 0.2241, + "step": 14390 + }, + { + "epoch": 6.690376569037657, + "grad_norm": 5.200544834136963, + "learning_rate": 9.985615998561601e-06, + "loss": 0.1017, + "step": 14391 + }, + { + "epoch": 6.6908414690841465, + "grad_norm": 9.237876892089844, + "learning_rate": 9.9856149985615e-06, + "loss": 0.1653, + "step": 14392 + }, + { + "epoch": 6.691306369130637, + "grad_norm": 5.314124584197998, + "learning_rate": 9.9856139985614e-06, + "loss": 0.1965, + "step": 14393 + }, + { + "epoch": 6.691771269177127, + "grad_norm": 6.416880130767822, + "learning_rate": 9.9856129985613e-06, + "loss": 0.1418, + "step": 14394 + }, + { + "epoch": 6.692236169223617, + "grad_norm": 3.4239494800567627, + "learning_rate": 9.9856119985612e-06, + "loss": 0.1009, + "step": 14395 + }, + { + "epoch": 6.6927010692701066, + "grad_norm": 4.200303554534912, + "learning_rate": 9.985610998561101e-06, + "loss": 0.0672, + "step": 14396 + }, + { + "epoch": 6.693165969316597, + "grad_norm": 9.006250381469727, + "learning_rate": 9.985609998561002e-06, + "loss": 0.3893, + "step": 14397 + }, + { + "epoch": 6.693630869363087, + "grad_norm": 5.053436279296875, + "learning_rate": 9.9856089985609e-06, + "loss": 0.1628, + "step": 14398 + }, + { + "epoch": 6.694095769409577, + "grad_norm": 7.242534160614014, + "learning_rate": 9.985607998560801e-06, + "loss": 0.2867, + "step": 14399 + }, + { + "epoch": 6.694560669456067, + "grad_norm": 8.1412992477417, + "learning_rate": 9.9856069985607e-06, + "loss": 0.1807, + "step": 14400 + }, + { + "epoch": 6.695025569502557, + "grad_norm": 5.417262554168701, + "learning_rate": 9.9856059985606e-06, + "loss": 0.1104, + "step": 14401 + }, + { + "epoch": 6.695490469549047, + "grad_norm": 5.854702472686768, + "learning_rate": 9.985604998560501e-06, + "loss": 0.2066, + "step": 14402 + }, + { + "epoch": 6.695955369595537, + "grad_norm": 8.04275131225586, + "learning_rate": 9.9856039985604e-06, + "loss": 0.407, + "step": 14403 + }, + { + "epoch": 6.696420269642027, + "grad_norm": 8.57281494140625, + "learning_rate": 9.985602998560301e-06, + "loss": 0.2021, + "step": 14404 + }, + { + "epoch": 6.696885169688517, + "grad_norm": 4.597344875335693, + "learning_rate": 9.9856019985602e-06, + "loss": 0.1165, + "step": 14405 + }, + { + "epoch": 6.697350069735007, + "grad_norm": 6.081541538238525, + "learning_rate": 9.9856009985601e-06, + "loss": 0.248, + "step": 14406 + }, + { + "epoch": 6.697814969781497, + "grad_norm": 6.728316783905029, + "learning_rate": 9.98559999856e-06, + "loss": 0.3412, + "step": 14407 + }, + { + "epoch": 6.698279869827987, + "grad_norm": 6.183267593383789, + "learning_rate": 9.9855989985599e-06, + "loss": 0.2438, + "step": 14408 + }, + { + "epoch": 6.698744769874477, + "grad_norm": 6.571442604064941, + "learning_rate": 9.985597998559801e-06, + "loss": 0.2964, + "step": 14409 + }, + { + "epoch": 6.699209669920967, + "grad_norm": 7.9455647468566895, + "learning_rate": 9.985596998559702e-06, + "loss": 0.3054, + "step": 14410 + }, + { + "epoch": 6.699674569967457, + "grad_norm": 3.6632933616638184, + "learning_rate": 9.9855959985596e-06, + "loss": 0.1036, + "step": 14411 + }, + { + "epoch": 6.700139470013947, + "grad_norm": 4.222466468811035, + "learning_rate": 9.9855949985595e-06, + "loss": 0.1196, + "step": 14412 + }, + { + "epoch": 6.700604370060437, + "grad_norm": 3.3321101665496826, + "learning_rate": 9.9855939985594e-06, + "loss": 0.0448, + "step": 14413 + }, + { + "epoch": 6.701069270106927, + "grad_norm": 4.206692218780518, + "learning_rate": 9.985592998559301e-06, + "loss": 0.1077, + "step": 14414 + }, + { + "epoch": 6.701534170153417, + "grad_norm": 4.220547676086426, + "learning_rate": 9.985591998559202e-06, + "loss": 0.1146, + "step": 14415 + }, + { + "epoch": 6.701999070199907, + "grad_norm": 6.0354156494140625, + "learning_rate": 9.9855909985591e-06, + "loss": 0.2171, + "step": 14416 + }, + { + "epoch": 6.702463970246397, + "grad_norm": 2.916334390640259, + "learning_rate": 9.985589998559001e-06, + "loss": 0.0612, + "step": 14417 + }, + { + "epoch": 6.702928870292887, + "grad_norm": 6.039182662963867, + "learning_rate": 9.9855889985589e-06, + "loss": 0.3136, + "step": 14418 + }, + { + "epoch": 6.703393770339377, + "grad_norm": 5.832716464996338, + "learning_rate": 9.9855879985588e-06, + "loss": 0.125, + "step": 14419 + }, + { + "epoch": 6.703858670385867, + "grad_norm": 4.660895347595215, + "learning_rate": 9.985586998558701e-06, + "loss": 0.1739, + "step": 14420 + }, + { + "epoch": 6.704323570432357, + "grad_norm": 5.695956230163574, + "learning_rate": 9.9855859985586e-06, + "loss": 0.2094, + "step": 14421 + }, + { + "epoch": 6.704788470478847, + "grad_norm": 10.541629791259766, + "learning_rate": 9.985584998558501e-06, + "loss": 0.644, + "step": 14422 + }, + { + "epoch": 6.705253370525337, + "grad_norm": 4.1029582023620605, + "learning_rate": 9.9855839985584e-06, + "loss": 0.1468, + "step": 14423 + }, + { + "epoch": 6.705718270571827, + "grad_norm": 9.500137329101562, + "learning_rate": 9.9855829985583e-06, + "loss": 0.2355, + "step": 14424 + }, + { + "epoch": 6.706183170618317, + "grad_norm": 7.433060646057129, + "learning_rate": 9.9855819985582e-06, + "loss": 0.1524, + "step": 14425 + }, + { + "epoch": 6.706648070664807, + "grad_norm": 4.015411853790283, + "learning_rate": 9.9855809985581e-06, + "loss": 0.1368, + "step": 14426 + }, + { + "epoch": 6.707112970711297, + "grad_norm": 3.473832845687866, + "learning_rate": 9.985579998558001e-06, + "loss": 0.0686, + "step": 14427 + }, + { + "epoch": 6.707577870757787, + "grad_norm": 7.056102275848389, + "learning_rate": 9.985578998557902e-06, + "loss": 0.3647, + "step": 14428 + }, + { + "epoch": 6.708042770804277, + "grad_norm": 4.209188938140869, + "learning_rate": 9.9855779985578e-06, + "loss": 0.076, + "step": 14429 + }, + { + "epoch": 6.708507670850767, + "grad_norm": 3.209874153137207, + "learning_rate": 9.985576998557701e-06, + "loss": 0.0764, + "step": 14430 + }, + { + "epoch": 6.708972570897257, + "grad_norm": 5.656899452209473, + "learning_rate": 9.9855759985576e-06, + "loss": 0.2515, + "step": 14431 + }, + { + "epoch": 6.709437470943747, + "grad_norm": 5.760035991668701, + "learning_rate": 9.985574998557501e-06, + "loss": 0.1324, + "step": 14432 + }, + { + "epoch": 6.709902370990237, + "grad_norm": 3.8732802867889404, + "learning_rate": 9.985573998557402e-06, + "loss": 0.1003, + "step": 14433 + }, + { + "epoch": 6.710367271036727, + "grad_norm": 5.802309989929199, + "learning_rate": 9.9855729985573e-06, + "loss": 0.1229, + "step": 14434 + }, + { + "epoch": 6.710832171083217, + "grad_norm": 4.162820816040039, + "learning_rate": 9.985571998557201e-06, + "loss": 0.1376, + "step": 14435 + }, + { + "epoch": 6.711297071129707, + "grad_norm": 1.3749891519546509, + "learning_rate": 9.9855709985571e-06, + "loss": 0.0195, + "step": 14436 + }, + { + "epoch": 6.711761971176197, + "grad_norm": 3.8679211139678955, + "learning_rate": 9.985569998557e-06, + "loss": 0.0958, + "step": 14437 + }, + { + "epoch": 6.712226871222687, + "grad_norm": 5.357542514801025, + "learning_rate": 9.9855689985569e-06, + "loss": 0.1824, + "step": 14438 + }, + { + "epoch": 6.712691771269177, + "grad_norm": 5.355941295623779, + "learning_rate": 9.9855679985568e-06, + "loss": 0.2664, + "step": 14439 + }, + { + "epoch": 6.713156671315668, + "grad_norm": 2.932762622833252, + "learning_rate": 9.985566998556701e-06, + "loss": 0.0618, + "step": 14440 + }, + { + "epoch": 6.713621571362157, + "grad_norm": 5.555026054382324, + "learning_rate": 9.985565998556602e-06, + "loss": 0.3059, + "step": 14441 + }, + { + "epoch": 6.714086471408647, + "grad_norm": 7.181087493896484, + "learning_rate": 9.9855649985565e-06, + "loss": 0.183, + "step": 14442 + }, + { + "epoch": 6.714551371455137, + "grad_norm": 4.707934379577637, + "learning_rate": 9.9855639985564e-06, + "loss": 0.0656, + "step": 14443 + }, + { + "epoch": 6.715016271501627, + "grad_norm": 4.308955192565918, + "learning_rate": 9.9855629985563e-06, + "loss": 0.167, + "step": 14444 + }, + { + "epoch": 6.7154811715481175, + "grad_norm": 3.670091390609741, + "learning_rate": 9.985561998556201e-06, + "loss": 0.0624, + "step": 14445 + }, + { + "epoch": 6.715946071594607, + "grad_norm": 3.298886299133301, + "learning_rate": 9.985560998556102e-06, + "loss": 0.0538, + "step": 14446 + }, + { + "epoch": 6.716410971641097, + "grad_norm": 5.476053237915039, + "learning_rate": 9.985559998556e-06, + "loss": 0.2017, + "step": 14447 + }, + { + "epoch": 6.716875871687587, + "grad_norm": 7.203287124633789, + "learning_rate": 9.985558998555901e-06, + "loss": 0.2586, + "step": 14448 + }, + { + "epoch": 6.717340771734078, + "grad_norm": 5.566648006439209, + "learning_rate": 9.9855579985558e-06, + "loss": 0.0835, + "step": 14449 + }, + { + "epoch": 6.717805671780567, + "grad_norm": 7.056594371795654, + "learning_rate": 9.985556998555701e-06, + "loss": 0.1947, + "step": 14450 + }, + { + "epoch": 6.718270571827057, + "grad_norm": 4.758766174316406, + "learning_rate": 9.9855559985556e-06, + "loss": 0.1531, + "step": 14451 + }, + { + "epoch": 6.718735471873547, + "grad_norm": 5.607349395751953, + "learning_rate": 9.9855549985555e-06, + "loss": 0.2804, + "step": 14452 + }, + { + "epoch": 6.719200371920037, + "grad_norm": 9.565279006958008, + "learning_rate": 9.985553998555401e-06, + "loss": 0.4521, + "step": 14453 + }, + { + "epoch": 6.7196652719665275, + "grad_norm": 9.197789192199707, + "learning_rate": 9.9855529985553e-06, + "loss": 0.3096, + "step": 14454 + }, + { + "epoch": 6.720130172013017, + "grad_norm": 4.316072463989258, + "learning_rate": 9.9855519985552e-06, + "loss": 0.1267, + "step": 14455 + }, + { + "epoch": 6.720595072059507, + "grad_norm": 5.488943576812744, + "learning_rate": 9.9855509985551e-06, + "loss": 0.1179, + "step": 14456 + }, + { + "epoch": 6.721059972105997, + "grad_norm": 4.984157562255859, + "learning_rate": 9.985549998555e-06, + "loss": 0.1437, + "step": 14457 + }, + { + "epoch": 6.721524872152488, + "grad_norm": 3.3362622261047363, + "learning_rate": 9.985548998554901e-06, + "loss": 0.0619, + "step": 14458 + }, + { + "epoch": 6.721989772198977, + "grad_norm": 3.796816825866699, + "learning_rate": 9.985547998554802e-06, + "loss": 0.1098, + "step": 14459 + }, + { + "epoch": 6.722454672245467, + "grad_norm": 6.792807102203369, + "learning_rate": 9.9855469985547e-06, + "loss": 0.3477, + "step": 14460 + }, + { + "epoch": 6.722919572291957, + "grad_norm": 6.781996250152588, + "learning_rate": 9.985545998554601e-06, + "loss": 0.2611, + "step": 14461 + }, + { + "epoch": 6.723384472338447, + "grad_norm": 7.049630165100098, + "learning_rate": 9.9855449985545e-06, + "loss": 0.3126, + "step": 14462 + }, + { + "epoch": 6.7238493723849375, + "grad_norm": 4.695311546325684, + "learning_rate": 9.985543998554401e-06, + "loss": 0.1133, + "step": 14463 + }, + { + "epoch": 6.724314272431427, + "grad_norm": 3.127681255340576, + "learning_rate": 9.9855429985543e-06, + "loss": 0.0676, + "step": 14464 + }, + { + "epoch": 6.724779172477917, + "grad_norm": 4.569309234619141, + "learning_rate": 9.9855419985542e-06, + "loss": 0.0928, + "step": 14465 + }, + { + "epoch": 6.725244072524407, + "grad_norm": 6.826639175415039, + "learning_rate": 9.985540998554101e-06, + "loss": 0.1496, + "step": 14466 + }, + { + "epoch": 6.725708972570898, + "grad_norm": 4.786266326904297, + "learning_rate": 9.985539998554e-06, + "loss": 0.2223, + "step": 14467 + }, + { + "epoch": 6.726173872617387, + "grad_norm": 6.829389572143555, + "learning_rate": 9.9855389985539e-06, + "loss": 0.2569, + "step": 14468 + }, + { + "epoch": 6.726638772663877, + "grad_norm": 6.619804859161377, + "learning_rate": 9.9855379985538e-06, + "loss": 0.2468, + "step": 14469 + }, + { + "epoch": 6.727103672710367, + "grad_norm": 6.331316947937012, + "learning_rate": 9.9855369985537e-06, + "loss": 0.1965, + "step": 14470 + }, + { + "epoch": 6.727568572756857, + "grad_norm": 4.6209492683410645, + "learning_rate": 9.985535998553601e-06, + "loss": 0.1356, + "step": 14471 + }, + { + "epoch": 6.7280334728033475, + "grad_norm": 7.275901794433594, + "learning_rate": 9.985534998553502e-06, + "loss": 0.2302, + "step": 14472 + }, + { + "epoch": 6.728498372849837, + "grad_norm": 8.972024917602539, + "learning_rate": 9.9855339985534e-06, + "loss": 0.1669, + "step": 14473 + }, + { + "epoch": 6.728963272896327, + "grad_norm": 5.650127410888672, + "learning_rate": 9.985532998553301e-06, + "loss": 0.2667, + "step": 14474 + }, + { + "epoch": 6.729428172942817, + "grad_norm": 5.37312126159668, + "learning_rate": 9.9855319985532e-06, + "loss": 0.2366, + "step": 14475 + }, + { + "epoch": 6.729893072989308, + "grad_norm": 3.5767822265625, + "learning_rate": 9.9855309985531e-06, + "loss": 0.107, + "step": 14476 + }, + { + "epoch": 6.730357973035797, + "grad_norm": 4.861913681030273, + "learning_rate": 9.985529998553e-06, + "loss": 0.1612, + "step": 14477 + }, + { + "epoch": 6.730822873082287, + "grad_norm": 5.223037242889404, + "learning_rate": 9.9855289985529e-06, + "loss": 0.2056, + "step": 14478 + }, + { + "epoch": 6.731287773128777, + "grad_norm": 6.783889293670654, + "learning_rate": 9.985527998552801e-06, + "loss": 0.2132, + "step": 14479 + }, + { + "epoch": 6.731752673175267, + "grad_norm": 4.040096282958984, + "learning_rate": 9.9855269985527e-06, + "loss": 0.0594, + "step": 14480 + }, + { + "epoch": 6.7322175732217575, + "grad_norm": 3.2663896083831787, + "learning_rate": 9.985525998552601e-06, + "loss": 0.0746, + "step": 14481 + }, + { + "epoch": 6.732682473268247, + "grad_norm": 5.428460121154785, + "learning_rate": 9.9855249985525e-06, + "loss": 0.268, + "step": 14482 + }, + { + "epoch": 6.733147373314737, + "grad_norm": 8.366296768188477, + "learning_rate": 9.9855239985524e-06, + "loss": 0.3754, + "step": 14483 + }, + { + "epoch": 6.733612273361228, + "grad_norm": 3.6064629554748535, + "learning_rate": 9.985522998552301e-06, + "loss": 0.0654, + "step": 14484 + }, + { + "epoch": 6.734077173407718, + "grad_norm": 3.0589494705200195, + "learning_rate": 9.985521998552202e-06, + "loss": 0.0744, + "step": 14485 + }, + { + "epoch": 6.734542073454207, + "grad_norm": 4.764289379119873, + "learning_rate": 9.9855209985521e-06, + "loss": 0.1332, + "step": 14486 + }, + { + "epoch": 6.735006973500697, + "grad_norm": 4.815712928771973, + "learning_rate": 9.985519998552e-06, + "loss": 0.1006, + "step": 14487 + }, + { + "epoch": 6.735471873547187, + "grad_norm": 4.990792751312256, + "learning_rate": 9.9855189985519e-06, + "loss": 0.1552, + "step": 14488 + }, + { + "epoch": 6.735936773593678, + "grad_norm": 8.281792640686035, + "learning_rate": 9.985517998551801e-06, + "loss": 0.1675, + "step": 14489 + }, + { + "epoch": 6.7364016736401675, + "grad_norm": 6.942577362060547, + "learning_rate": 9.9855169985517e-06, + "loss": 0.1113, + "step": 14490 + }, + { + "epoch": 6.736866573686657, + "grad_norm": 3.9362919330596924, + "learning_rate": 9.9855159985516e-06, + "loss": 0.1092, + "step": 14491 + }, + { + "epoch": 6.737331473733147, + "grad_norm": 4.859588146209717, + "learning_rate": 9.985514998551501e-06, + "loss": 0.2088, + "step": 14492 + }, + { + "epoch": 6.737796373779638, + "grad_norm": 3.204098701477051, + "learning_rate": 9.9855139985514e-06, + "loss": 0.0737, + "step": 14493 + }, + { + "epoch": 6.738261273826128, + "grad_norm": 5.820461273193359, + "learning_rate": 9.985512998551301e-06, + "loss": 0.1578, + "step": 14494 + }, + { + "epoch": 6.738726173872617, + "grad_norm": 11.662631034851074, + "learning_rate": 9.9855119985512e-06, + "loss": 0.3547, + "step": 14495 + }, + { + "epoch": 6.739191073919107, + "grad_norm": 7.843786239624023, + "learning_rate": 9.9855109985511e-06, + "loss": 0.1766, + "step": 14496 + }, + { + "epoch": 6.739655973965597, + "grad_norm": 5.2469353675842285, + "learning_rate": 9.985509998551001e-06, + "loss": 0.1398, + "step": 14497 + }, + { + "epoch": 6.740120874012088, + "grad_norm": 7.3070149421691895, + "learning_rate": 9.9855089985509e-06, + "loss": 0.3519, + "step": 14498 + }, + { + "epoch": 6.7405857740585775, + "grad_norm": 6.9383931159973145, + "learning_rate": 9.985507998550801e-06, + "loss": 0.2138, + "step": 14499 + }, + { + "epoch": 6.741050674105067, + "grad_norm": 5.739604473114014, + "learning_rate": 9.9855069985507e-06, + "loss": 0.181, + "step": 14500 + }, + { + "epoch": 6.741515574151557, + "grad_norm": 5.03860330581665, + "learning_rate": 9.9855059985506e-06, + "loss": 0.1981, + "step": 14501 + }, + { + "epoch": 6.741980474198048, + "grad_norm": 8.94556713104248, + "learning_rate": 9.985504998550501e-06, + "loss": 0.1336, + "step": 14502 + }, + { + "epoch": 6.742445374244538, + "grad_norm": 7.908596992492676, + "learning_rate": 9.9855039985504e-06, + "loss": 0.3708, + "step": 14503 + }, + { + "epoch": 6.742910274291027, + "grad_norm": 7.8120317459106445, + "learning_rate": 9.9855029985503e-06, + "loss": 0.3341, + "step": 14504 + }, + { + "epoch": 6.743375174337517, + "grad_norm": 6.200285911560059, + "learning_rate": 9.985501998550201e-06, + "loss": 0.1288, + "step": 14505 + }, + { + "epoch": 6.743840074384007, + "grad_norm": 4.3600993156433105, + "learning_rate": 9.9855009985501e-06, + "loss": 0.1411, + "step": 14506 + }, + { + "epoch": 6.744304974430498, + "grad_norm": 7.7783403396606445, + "learning_rate": 9.985499998550001e-06, + "loss": 0.2918, + "step": 14507 + }, + { + "epoch": 6.7447698744769875, + "grad_norm": 4.976774215698242, + "learning_rate": 9.9854989985499e-06, + "loss": 0.1371, + "step": 14508 + }, + { + "epoch": 6.745234774523477, + "grad_norm": 3.791604518890381, + "learning_rate": 9.9854979985498e-06, + "loss": 0.1132, + "step": 14509 + }, + { + "epoch": 6.745699674569967, + "grad_norm": 5.290558338165283, + "learning_rate": 9.985496998549701e-06, + "loss": 0.1907, + "step": 14510 + }, + { + "epoch": 6.746164574616458, + "grad_norm": 5.554874420166016, + "learning_rate": 9.9854959985496e-06, + "loss": 0.3268, + "step": 14511 + }, + { + "epoch": 6.746629474662948, + "grad_norm": 4.532668590545654, + "learning_rate": 9.985494998549501e-06, + "loss": 0.1384, + "step": 14512 + }, + { + "epoch": 6.747094374709437, + "grad_norm": 7.560028076171875, + "learning_rate": 9.9854939985494e-06, + "loss": 0.3571, + "step": 14513 + }, + { + "epoch": 6.747559274755927, + "grad_norm": 2.362729072570801, + "learning_rate": 9.9854929985493e-06, + "loss": 0.0318, + "step": 14514 + }, + { + "epoch": 6.748024174802417, + "grad_norm": 7.500811576843262, + "learning_rate": 9.985491998549201e-06, + "loss": 0.4368, + "step": 14515 + }, + { + "epoch": 6.748489074848908, + "grad_norm": 4.978443622589111, + "learning_rate": 9.985490998549102e-06, + "loss": 0.1545, + "step": 14516 + }, + { + "epoch": 6.7489539748953975, + "grad_norm": 3.1003806591033936, + "learning_rate": 9.985489998549001e-06, + "loss": 0.1038, + "step": 14517 + }, + { + "epoch": 6.749418874941887, + "grad_norm": 5.4179487228393555, + "learning_rate": 9.985488998548902e-06, + "loss": 0.3171, + "step": 14518 + }, + { + "epoch": 6.749883774988378, + "grad_norm": 4.6943559646606445, + "learning_rate": 9.9854879985488e-06, + "loss": 0.1961, + "step": 14519 + }, + { + "epoch": 6.750348675034868, + "grad_norm": 4.220245361328125, + "learning_rate": 9.9854869985487e-06, + "loss": 0.0933, + "step": 14520 + }, + { + "epoch": 6.750813575081358, + "grad_norm": 4.53518533706665, + "learning_rate": 9.9854859985486e-06, + "loss": 0.1017, + "step": 14521 + }, + { + "epoch": 6.751278475127847, + "grad_norm": 4.435178756713867, + "learning_rate": 9.9854849985485e-06, + "loss": 0.1189, + "step": 14522 + }, + { + "epoch": 6.751743375174337, + "grad_norm": 10.376934051513672, + "learning_rate": 9.985483998548401e-06, + "loss": 0.3218, + "step": 14523 + }, + { + "epoch": 6.752208275220828, + "grad_norm": 5.143715858459473, + "learning_rate": 9.9854829985483e-06, + "loss": 0.1852, + "step": 14524 + }, + { + "epoch": 6.752673175267318, + "grad_norm": 5.4227294921875, + "learning_rate": 9.985481998548201e-06, + "loss": 0.221, + "step": 14525 + }, + { + "epoch": 6.7531380753138075, + "grad_norm": 4.059508800506592, + "learning_rate": 9.9854809985481e-06, + "loss": 0.0951, + "step": 14526 + }, + { + "epoch": 6.753602975360297, + "grad_norm": 5.23541259765625, + "learning_rate": 9.985479998548e-06, + "loss": 0.145, + "step": 14527 + }, + { + "epoch": 6.754067875406788, + "grad_norm": 3.294731378555298, + "learning_rate": 9.985478998547901e-06, + "loss": 0.0745, + "step": 14528 + }, + { + "epoch": 6.754532775453278, + "grad_norm": 9.806403160095215, + "learning_rate": 9.985477998547802e-06, + "loss": 0.2463, + "step": 14529 + }, + { + "epoch": 6.754997675499768, + "grad_norm": 11.821406364440918, + "learning_rate": 9.985476998547701e-06, + "loss": 0.1385, + "step": 14530 + }, + { + "epoch": 6.755462575546257, + "grad_norm": 12.431443214416504, + "learning_rate": 9.9854759985476e-06, + "loss": 0.3816, + "step": 14531 + }, + { + "epoch": 6.755927475592747, + "grad_norm": 11.9142427444458, + "learning_rate": 9.9854749985475e-06, + "loss": 0.2846, + "step": 14532 + }, + { + "epoch": 6.756392375639238, + "grad_norm": 5.6291937828063965, + "learning_rate": 9.9854739985474e-06, + "loss": 0.2302, + "step": 14533 + }, + { + "epoch": 6.756857275685728, + "grad_norm": 4.805056571960449, + "learning_rate": 9.9854729985473e-06, + "loss": 0.1035, + "step": 14534 + }, + { + "epoch": 6.7573221757322175, + "grad_norm": 6.051329612731934, + "learning_rate": 9.985471998547201e-06, + "loss": 0.1663, + "step": 14535 + }, + { + "epoch": 6.757787075778707, + "grad_norm": 6.5253071784973145, + "learning_rate": 9.985470998547102e-06, + "loss": 0.2013, + "step": 14536 + }, + { + "epoch": 6.758251975825198, + "grad_norm": 2.855884552001953, + "learning_rate": 9.985469998547e-06, + "loss": 0.0742, + "step": 14537 + }, + { + "epoch": 6.758716875871688, + "grad_norm": 5.0489420890808105, + "learning_rate": 9.985468998546901e-06, + "loss": 0.1372, + "step": 14538 + }, + { + "epoch": 6.759181775918178, + "grad_norm": 6.521661281585693, + "learning_rate": 9.9854679985468e-06, + "loss": 0.294, + "step": 14539 + }, + { + "epoch": 6.759646675964667, + "grad_norm": 6.1151227951049805, + "learning_rate": 9.9854669985467e-06, + "loss": 0.2019, + "step": 14540 + }, + { + "epoch": 6.760111576011157, + "grad_norm": 5.288907051086426, + "learning_rate": 9.985465998546601e-06, + "loss": 0.2109, + "step": 14541 + }, + { + "epoch": 6.760576476057648, + "grad_norm": 3.504913806915283, + "learning_rate": 9.9854649985465e-06, + "loss": 0.0551, + "step": 14542 + }, + { + "epoch": 6.761041376104138, + "grad_norm": 6.582364082336426, + "learning_rate": 9.985463998546401e-06, + "loss": 0.1957, + "step": 14543 + }, + { + "epoch": 6.7615062761506275, + "grad_norm": 9.473295211791992, + "learning_rate": 9.9854629985463e-06, + "loss": 0.3816, + "step": 14544 + }, + { + "epoch": 6.761971176197117, + "grad_norm": 5.317539691925049, + "learning_rate": 9.9854619985462e-06, + "loss": 0.1847, + "step": 14545 + }, + { + "epoch": 6.762436076243608, + "grad_norm": 4.586456298828125, + "learning_rate": 9.9854609985461e-06, + "loss": 0.2278, + "step": 14546 + }, + { + "epoch": 6.762900976290098, + "grad_norm": 11.902985572814941, + "learning_rate": 9.985459998546e-06, + "loss": 0.2617, + "step": 14547 + }, + { + "epoch": 6.763365876336588, + "grad_norm": 4.039616107940674, + "learning_rate": 9.985458998545901e-06, + "loss": 0.0971, + "step": 14548 + }, + { + "epoch": 6.763830776383077, + "grad_norm": 5.46242094039917, + "learning_rate": 9.985457998545802e-06, + "loss": 0.1239, + "step": 14549 + }, + { + "epoch": 6.764295676429567, + "grad_norm": 5.021065711975098, + "learning_rate": 9.9854569985457e-06, + "loss": 0.2051, + "step": 14550 + }, + { + "epoch": 6.764760576476058, + "grad_norm": 6.765015602111816, + "learning_rate": 9.985455998545601e-06, + "loss": 0.2463, + "step": 14551 + }, + { + "epoch": 6.765225476522548, + "grad_norm": 5.552060604095459, + "learning_rate": 9.9854549985455e-06, + "loss": 0.1398, + "step": 14552 + }, + { + "epoch": 6.7656903765690375, + "grad_norm": 4.8330464363098145, + "learning_rate": 9.985453998545401e-06, + "loss": 0.0861, + "step": 14553 + }, + { + "epoch": 6.766155276615527, + "grad_norm": 4.25297737121582, + "learning_rate": 9.985452998545302e-06, + "loss": 0.0488, + "step": 14554 + }, + { + "epoch": 6.766620176662018, + "grad_norm": 5.332136631011963, + "learning_rate": 9.9854519985452e-06, + "loss": 0.1549, + "step": 14555 + }, + { + "epoch": 6.767085076708508, + "grad_norm": 4.7655839920043945, + "learning_rate": 9.985450998545101e-06, + "loss": 0.1474, + "step": 14556 + }, + { + "epoch": 6.767549976754998, + "grad_norm": 6.587204456329346, + "learning_rate": 9.985449998545e-06, + "loss": 0.2155, + "step": 14557 + }, + { + "epoch": 6.768014876801487, + "grad_norm": 5.285032749176025, + "learning_rate": 9.9854489985449e-06, + "loss": 0.1899, + "step": 14558 + }, + { + "epoch": 6.768479776847977, + "grad_norm": 5.847560405731201, + "learning_rate": 9.9854479985448e-06, + "loss": 0.224, + "step": 14559 + }, + { + "epoch": 6.768944676894468, + "grad_norm": 6.563594818115234, + "learning_rate": 9.9854469985447e-06, + "loss": 0.1827, + "step": 14560 + }, + { + "epoch": 6.769409576940958, + "grad_norm": 3.038588047027588, + "learning_rate": 9.985445998544601e-06, + "loss": 0.0792, + "step": 14561 + }, + { + "epoch": 6.7698744769874475, + "grad_norm": 5.291452407836914, + "learning_rate": 9.985444998544502e-06, + "loss": 0.2453, + "step": 14562 + }, + { + "epoch": 6.770339377033938, + "grad_norm": 2.9117348194122314, + "learning_rate": 9.9854439985444e-06, + "loss": 0.0571, + "step": 14563 + }, + { + "epoch": 6.770804277080428, + "grad_norm": 6.850531578063965, + "learning_rate": 9.9854429985443e-06, + "loss": 0.2351, + "step": 14564 + }, + { + "epoch": 6.771269177126918, + "grad_norm": 4.868594169616699, + "learning_rate": 9.9854419985442e-06, + "loss": 0.2223, + "step": 14565 + }, + { + "epoch": 6.771734077173408, + "grad_norm": 5.656191825866699, + "learning_rate": 9.985440998544101e-06, + "loss": 0.2147, + "step": 14566 + }, + { + "epoch": 6.772198977219897, + "grad_norm": 4.892880916595459, + "learning_rate": 9.985439998544002e-06, + "loss": 0.1664, + "step": 14567 + }, + { + "epoch": 6.772663877266388, + "grad_norm": 3.454465627670288, + "learning_rate": 9.9854389985439e-06, + "loss": 0.0561, + "step": 14568 + }, + { + "epoch": 6.773128777312878, + "grad_norm": 3.946833372116089, + "learning_rate": 9.985437998543801e-06, + "loss": 0.1165, + "step": 14569 + }, + { + "epoch": 6.773593677359368, + "grad_norm": 5.438302516937256, + "learning_rate": 9.9854369985437e-06, + "loss": 0.1374, + "step": 14570 + }, + { + "epoch": 6.7740585774058575, + "grad_norm": 7.054573059082031, + "learning_rate": 9.9854359985436e-06, + "loss": 0.2652, + "step": 14571 + }, + { + "epoch": 6.774523477452348, + "grad_norm": 4.779032230377197, + "learning_rate": 9.9854349985435e-06, + "loss": 0.1652, + "step": 14572 + }, + { + "epoch": 6.774988377498838, + "grad_norm": 6.343433380126953, + "learning_rate": 9.9854339985434e-06, + "loss": 0.1699, + "step": 14573 + }, + { + "epoch": 6.775453277545328, + "grad_norm": 3.926574945449829, + "learning_rate": 9.985432998543301e-06, + "loss": 0.0964, + "step": 14574 + }, + { + "epoch": 6.775918177591818, + "grad_norm": 4.3489274978637695, + "learning_rate": 9.9854319985432e-06, + "loss": 0.0937, + "step": 14575 + }, + { + "epoch": 6.776383077638307, + "grad_norm": 5.720210552215576, + "learning_rate": 9.9854309985431e-06, + "loss": 0.2366, + "step": 14576 + }, + { + "epoch": 6.776847977684798, + "grad_norm": 6.174887180328369, + "learning_rate": 9.985429998543e-06, + "loss": 0.4354, + "step": 14577 + }, + { + "epoch": 6.777312877731288, + "grad_norm": 3.375723123550415, + "learning_rate": 9.9854289985429e-06, + "loss": 0.1101, + "step": 14578 + }, + { + "epoch": 6.777777777777778, + "grad_norm": 3.4032084941864014, + "learning_rate": 9.985427998542801e-06, + "loss": 0.1234, + "step": 14579 + }, + { + "epoch": 6.7782426778242675, + "grad_norm": 4.848309516906738, + "learning_rate": 9.985426998542702e-06, + "loss": 0.1205, + "step": 14580 + }, + { + "epoch": 6.778707577870758, + "grad_norm": 5.46317195892334, + "learning_rate": 9.9854259985426e-06, + "loss": 0.2014, + "step": 14581 + }, + { + "epoch": 6.779172477917248, + "grad_norm": 5.6849284172058105, + "learning_rate": 9.985424998542501e-06, + "loss": 0.1651, + "step": 14582 + }, + { + "epoch": 6.779637377963738, + "grad_norm": 4.8638410568237305, + "learning_rate": 9.9854239985424e-06, + "loss": 0.172, + "step": 14583 + }, + { + "epoch": 6.780102278010228, + "grad_norm": 4.85115385055542, + "learning_rate": 9.985422998542301e-06, + "loss": 0.1852, + "step": 14584 + }, + { + "epoch": 6.780567178056717, + "grad_norm": 13.080490112304688, + "learning_rate": 9.985421998542202e-06, + "loss": 0.0364, + "step": 14585 + }, + { + "epoch": 6.781032078103208, + "grad_norm": 5.76513147354126, + "learning_rate": 9.9854209985421e-06, + "loss": 0.2441, + "step": 14586 + }, + { + "epoch": 6.781496978149698, + "grad_norm": 4.072534561157227, + "learning_rate": 9.985419998542001e-06, + "loss": 0.128, + "step": 14587 + }, + { + "epoch": 6.781961878196188, + "grad_norm": 5.0885701179504395, + "learning_rate": 9.9854189985419e-06, + "loss": 0.2359, + "step": 14588 + }, + { + "epoch": 6.7824267782426775, + "grad_norm": 6.652768135070801, + "learning_rate": 9.9854179985418e-06, + "loss": 0.2723, + "step": 14589 + }, + { + "epoch": 6.782891678289168, + "grad_norm": 3.990841865539551, + "learning_rate": 9.9854169985417e-06, + "loss": 0.0863, + "step": 14590 + }, + { + "epoch": 6.783356578335658, + "grad_norm": 7.023630142211914, + "learning_rate": 9.9854159985416e-06, + "loss": 0.2217, + "step": 14591 + }, + { + "epoch": 6.783821478382148, + "grad_norm": 7.6902360916137695, + "learning_rate": 9.985414998541501e-06, + "loss": 0.3387, + "step": 14592 + }, + { + "epoch": 6.784286378428638, + "grad_norm": 3.740166664123535, + "learning_rate": 9.985413998541402e-06, + "loss": 0.1134, + "step": 14593 + }, + { + "epoch": 6.784751278475127, + "grad_norm": 6.584697723388672, + "learning_rate": 9.9854129985413e-06, + "loss": 0.1703, + "step": 14594 + }, + { + "epoch": 6.785216178521618, + "grad_norm": 6.569856643676758, + "learning_rate": 9.9854119985412e-06, + "loss": 0.2391, + "step": 14595 + }, + { + "epoch": 6.785681078568108, + "grad_norm": 6.054235935211182, + "learning_rate": 9.9854109985411e-06, + "loss": 0.2512, + "step": 14596 + }, + { + "epoch": 6.786145978614598, + "grad_norm": 7.330780029296875, + "learning_rate": 9.985409998541001e-06, + "loss": 0.2341, + "step": 14597 + }, + { + "epoch": 6.786610878661088, + "grad_norm": 5.962178707122803, + "learning_rate": 9.985408998540902e-06, + "loss": 0.22, + "step": 14598 + }, + { + "epoch": 6.787075778707578, + "grad_norm": 1.5701358318328857, + "learning_rate": 9.9854079985408e-06, + "loss": 0.0246, + "step": 14599 + }, + { + "epoch": 6.787540678754068, + "grad_norm": 6.985813617706299, + "learning_rate": 9.985406998540701e-06, + "loss": 0.2725, + "step": 14600 + }, + { + "epoch": 6.788005578800558, + "grad_norm": 3.0753302574157715, + "learning_rate": 9.9854059985406e-06, + "loss": 0.1157, + "step": 14601 + }, + { + "epoch": 6.788470478847048, + "grad_norm": 5.337082862854004, + "learning_rate": 9.985404998540501e-06, + "loss": 0.1221, + "step": 14602 + }, + { + "epoch": 6.788935378893538, + "grad_norm": 4.001583576202393, + "learning_rate": 9.9854039985404e-06, + "loss": 0.085, + "step": 14603 + }, + { + "epoch": 6.789400278940028, + "grad_norm": 9.530570030212402, + "learning_rate": 9.9854029985403e-06, + "loss": 0.3116, + "step": 14604 + }, + { + "epoch": 6.789865178986518, + "grad_norm": 5.3342976570129395, + "learning_rate": 9.985401998540201e-06, + "loss": 0.151, + "step": 14605 + }, + { + "epoch": 6.790330079033008, + "grad_norm": 9.48365592956543, + "learning_rate": 9.9854009985401e-06, + "loss": 0.2182, + "step": 14606 + }, + { + "epoch": 6.790794979079498, + "grad_norm": 6.303954124450684, + "learning_rate": 9.98539999854e-06, + "loss": 0.1223, + "step": 14607 + }, + { + "epoch": 6.791259879125988, + "grad_norm": 7.28068733215332, + "learning_rate": 9.9853989985399e-06, + "loss": 0.383, + "step": 14608 + }, + { + "epoch": 6.791724779172478, + "grad_norm": 4.613163948059082, + "learning_rate": 9.9853979985398e-06, + "loss": 0.1593, + "step": 14609 + }, + { + "epoch": 6.792189679218968, + "grad_norm": 5.405152320861816, + "learning_rate": 9.985396998539701e-06, + "loss": 0.1226, + "step": 14610 + }, + { + "epoch": 6.792654579265458, + "grad_norm": 6.795519828796387, + "learning_rate": 9.985395998539602e-06, + "loss": 0.3446, + "step": 14611 + }, + { + "epoch": 6.793119479311948, + "grad_norm": 14.352334976196289, + "learning_rate": 9.9853949985395e-06, + "loss": 0.1404, + "step": 14612 + }, + { + "epoch": 6.793584379358438, + "grad_norm": 10.800283432006836, + "learning_rate": 9.985393998539401e-06, + "loss": 0.6368, + "step": 14613 + }, + { + "epoch": 6.794049279404928, + "grad_norm": 3.9808309078216553, + "learning_rate": 9.9853929985393e-06, + "loss": 0.0903, + "step": 14614 + }, + { + "epoch": 6.794514179451418, + "grad_norm": 6.9940924644470215, + "learning_rate": 9.985391998539201e-06, + "loss": 0.2221, + "step": 14615 + }, + { + "epoch": 6.794979079497908, + "grad_norm": 5.3274149894714355, + "learning_rate": 9.9853909985391e-06, + "loss": 0.1644, + "step": 14616 + }, + { + "epoch": 6.795443979544398, + "grad_norm": 5.238703727722168, + "learning_rate": 9.985389998539e-06, + "loss": 0.1732, + "step": 14617 + }, + { + "epoch": 6.795908879590888, + "grad_norm": 4.31707239151001, + "learning_rate": 9.985388998538901e-06, + "loss": 0.0962, + "step": 14618 + }, + { + "epoch": 6.796373779637378, + "grad_norm": 9.653294563293457, + "learning_rate": 9.9853879985388e-06, + "loss": 0.2509, + "step": 14619 + }, + { + "epoch": 6.796838679683868, + "grad_norm": 5.101099491119385, + "learning_rate": 9.985386998538701e-06, + "loss": 0.2881, + "step": 14620 + }, + { + "epoch": 6.797303579730358, + "grad_norm": 5.913433074951172, + "learning_rate": 9.9853859985386e-06, + "loss": 0.1973, + "step": 14621 + }, + { + "epoch": 6.797768479776848, + "grad_norm": 7.761765956878662, + "learning_rate": 9.9853849985385e-06, + "loss": 0.33, + "step": 14622 + }, + { + "epoch": 6.798233379823338, + "grad_norm": 4.717674732208252, + "learning_rate": 9.985383998538401e-06, + "loss": 0.0902, + "step": 14623 + }, + { + "epoch": 6.798698279869828, + "grad_norm": 4.042498588562012, + "learning_rate": 9.985382998538302e-06, + "loss": 0.0854, + "step": 14624 + }, + { + "epoch": 6.799163179916318, + "grad_norm": 4.808681011199951, + "learning_rate": 9.9853819985382e-06, + "loss": 0.1405, + "step": 14625 + }, + { + "epoch": 6.799628079962808, + "grad_norm": 4.214759349822998, + "learning_rate": 9.985380998538101e-06, + "loss": 0.164, + "step": 14626 + }, + { + "epoch": 6.800092980009298, + "grad_norm": 5.421505928039551, + "learning_rate": 9.985379998538e-06, + "loss": 0.1697, + "step": 14627 + }, + { + "epoch": 6.800557880055788, + "grad_norm": 8.115562438964844, + "learning_rate": 9.9853789985379e-06, + "loss": 0.1494, + "step": 14628 + }, + { + "epoch": 6.801022780102278, + "grad_norm": 3.3288533687591553, + "learning_rate": 9.9853779985378e-06, + "loss": 0.0646, + "step": 14629 + }, + { + "epoch": 6.801487680148768, + "grad_norm": 4.265213966369629, + "learning_rate": 9.9853769985377e-06, + "loss": 0.1099, + "step": 14630 + }, + { + "epoch": 6.801952580195258, + "grad_norm": 3.3431758880615234, + "learning_rate": 9.985375998537601e-06, + "loss": 0.0311, + "step": 14631 + }, + { + "epoch": 6.802417480241748, + "grad_norm": 3.7483937740325928, + "learning_rate": 9.9853749985375e-06, + "loss": 0.0861, + "step": 14632 + }, + { + "epoch": 6.802882380288238, + "grad_norm": 4.79652214050293, + "learning_rate": 9.985373998537401e-06, + "loss": 0.1774, + "step": 14633 + }, + { + "epoch": 6.803347280334728, + "grad_norm": 2.0964057445526123, + "learning_rate": 9.9853729985373e-06, + "loss": 0.0382, + "step": 14634 + }, + { + "epoch": 6.803812180381218, + "grad_norm": 6.035717010498047, + "learning_rate": 9.9853719985372e-06, + "loss": 0.1466, + "step": 14635 + }, + { + "epoch": 6.804277080427708, + "grad_norm": 5.247200965881348, + "learning_rate": 9.985370998537101e-06, + "loss": 0.1262, + "step": 14636 + }, + { + "epoch": 6.804741980474198, + "grad_norm": 3.7359776496887207, + "learning_rate": 9.985369998537002e-06, + "loss": 0.0991, + "step": 14637 + }, + { + "epoch": 6.805206880520688, + "grad_norm": 4.516944885253906, + "learning_rate": 9.985368998536901e-06, + "loss": 0.1186, + "step": 14638 + }, + { + "epoch": 6.805671780567178, + "grad_norm": 6.924589157104492, + "learning_rate": 9.9853679985368e-06, + "loss": 0.3133, + "step": 14639 + }, + { + "epoch": 6.806136680613668, + "grad_norm": 6.311101913452148, + "learning_rate": 9.9853669985367e-06, + "loss": 0.1487, + "step": 14640 + }, + { + "epoch": 6.806601580660158, + "grad_norm": 4.112150192260742, + "learning_rate": 9.9853659985366e-06, + "loss": 0.1325, + "step": 14641 + }, + { + "epoch": 6.8070664807066485, + "grad_norm": 5.794697284698486, + "learning_rate": 9.9853649985365e-06, + "loss": 0.184, + "step": 14642 + }, + { + "epoch": 6.807531380753138, + "grad_norm": 3.244602918624878, + "learning_rate": 9.9853639985364e-06, + "loss": 0.0627, + "step": 14643 + }, + { + "epoch": 6.807996280799628, + "grad_norm": 4.246188640594482, + "learning_rate": 9.985362998536301e-06, + "loss": 0.0899, + "step": 14644 + }, + { + "epoch": 6.808461180846118, + "grad_norm": 6.088785171508789, + "learning_rate": 9.9853619985362e-06, + "loss": 0.1449, + "step": 14645 + }, + { + "epoch": 6.808926080892608, + "grad_norm": 6.495599746704102, + "learning_rate": 9.985360998536101e-06, + "loss": 0.2456, + "step": 14646 + }, + { + "epoch": 6.8093909809390984, + "grad_norm": 7.954503536224365, + "learning_rate": 9.985359998536e-06, + "loss": 0.3256, + "step": 14647 + }, + { + "epoch": 6.809855880985588, + "grad_norm": 8.136600494384766, + "learning_rate": 9.9853589985359e-06, + "loss": 0.1537, + "step": 14648 + }, + { + "epoch": 6.810320781032078, + "grad_norm": 6.017627716064453, + "learning_rate": 9.985357998535801e-06, + "loss": 0.175, + "step": 14649 + }, + { + "epoch": 6.810785681078568, + "grad_norm": 3.8395583629608154, + "learning_rate": 9.9853569985357e-06, + "loss": 0.0697, + "step": 14650 + }, + { + "epoch": 6.8112505811250585, + "grad_norm": 5.268765926361084, + "learning_rate": 9.985355998535601e-06, + "loss": 0.1745, + "step": 14651 + }, + { + "epoch": 6.811715481171548, + "grad_norm": 2.8283143043518066, + "learning_rate": 9.9853549985355e-06, + "loss": 0.0388, + "step": 14652 + }, + { + "epoch": 6.812180381218038, + "grad_norm": 8.476456642150879, + "learning_rate": 9.9853539985354e-06, + "loss": 0.3585, + "step": 14653 + }, + { + "epoch": 6.812645281264528, + "grad_norm": 4.687866687774658, + "learning_rate": 9.985352998535301e-06, + "loss": 0.1835, + "step": 14654 + }, + { + "epoch": 6.813110181311018, + "grad_norm": 5.724179744720459, + "learning_rate": 9.9853519985352e-06, + "loss": 0.1583, + "step": 14655 + }, + { + "epoch": 6.8135750813575084, + "grad_norm": 5.804684162139893, + "learning_rate": 9.9853509985351e-06, + "loss": 0.1532, + "step": 14656 + }, + { + "epoch": 6.814039981403998, + "grad_norm": 7.583367347717285, + "learning_rate": 9.985349998535001e-06, + "loss": 0.1877, + "step": 14657 + }, + { + "epoch": 6.814504881450488, + "grad_norm": 4.402253150939941, + "learning_rate": 9.9853489985349e-06, + "loss": 0.1036, + "step": 14658 + }, + { + "epoch": 6.814969781496978, + "grad_norm": 3.046232223510742, + "learning_rate": 9.985347998534801e-06, + "loss": 0.0504, + "step": 14659 + }, + { + "epoch": 6.8154346815434685, + "grad_norm": 6.237385272979736, + "learning_rate": 9.9853469985347e-06, + "loss": 0.1743, + "step": 14660 + }, + { + "epoch": 6.815899581589958, + "grad_norm": 5.249796390533447, + "learning_rate": 9.9853459985346e-06, + "loss": 0.1192, + "step": 14661 + }, + { + "epoch": 6.816364481636448, + "grad_norm": 5.75394344329834, + "learning_rate": 9.985344998534501e-06, + "loss": 0.1877, + "step": 14662 + }, + { + "epoch": 6.816829381682938, + "grad_norm": 6.219442844390869, + "learning_rate": 9.9853439985344e-06, + "loss": 0.0948, + "step": 14663 + }, + { + "epoch": 6.817294281729428, + "grad_norm": 4.813907146453857, + "learning_rate": 9.985342998534301e-06, + "loss": 0.0505, + "step": 14664 + }, + { + "epoch": 6.817759181775918, + "grad_norm": 5.505586624145508, + "learning_rate": 9.9853419985342e-06, + "loss": 0.1593, + "step": 14665 + }, + { + "epoch": 6.818224081822408, + "grad_norm": 5.955582141876221, + "learning_rate": 9.9853409985341e-06, + "loss": 0.0828, + "step": 14666 + }, + { + "epoch": 6.818688981868898, + "grad_norm": 4.75998067855835, + "learning_rate": 9.985339998534001e-06, + "loss": 0.1447, + "step": 14667 + }, + { + "epoch": 6.819153881915388, + "grad_norm": 6.317476272583008, + "learning_rate": 9.9853389985339e-06, + "loss": 0.1044, + "step": 14668 + }, + { + "epoch": 6.8196187819618785, + "grad_norm": 10.430379867553711, + "learning_rate": 9.985337998533801e-06, + "loss": 0.3126, + "step": 14669 + }, + { + "epoch": 6.820083682008368, + "grad_norm": 4.0626044273376465, + "learning_rate": 9.985336998533702e-06, + "loss": 0.119, + "step": 14670 + }, + { + "epoch": 6.820548582054858, + "grad_norm": 7.3177313804626465, + "learning_rate": 9.9853359985336e-06, + "loss": 0.3088, + "step": 14671 + }, + { + "epoch": 6.821013482101348, + "grad_norm": 5.984055519104004, + "learning_rate": 9.9853349985335e-06, + "loss": 0.1431, + "step": 14672 + }, + { + "epoch": 6.821478382147838, + "grad_norm": 5.228395462036133, + "learning_rate": 9.9853339985334e-06, + "loss": 0.1614, + "step": 14673 + }, + { + "epoch": 6.821943282194328, + "grad_norm": 6.822410583496094, + "learning_rate": 9.9853329985333e-06, + "loss": 0.0683, + "step": 14674 + }, + { + "epoch": 6.822408182240818, + "grad_norm": 5.884701251983643, + "learning_rate": 9.985331998533201e-06, + "loss": 0.1307, + "step": 14675 + }, + { + "epoch": 6.822873082287308, + "grad_norm": 3.3830950260162354, + "learning_rate": 9.9853309985331e-06, + "loss": 0.0724, + "step": 14676 + }, + { + "epoch": 6.823337982333798, + "grad_norm": 4.624907493591309, + "learning_rate": 9.985329998533001e-06, + "loss": 0.1006, + "step": 14677 + }, + { + "epoch": 6.8238028823802885, + "grad_norm": 4.887964248657227, + "learning_rate": 9.9853289985329e-06, + "loss": 0.1152, + "step": 14678 + }, + { + "epoch": 6.824267782426778, + "grad_norm": 5.669438362121582, + "learning_rate": 9.9853279985328e-06, + "loss": 0.1415, + "step": 14679 + }, + { + "epoch": 6.824732682473268, + "grad_norm": 9.2637300491333, + "learning_rate": 9.985326998532701e-06, + "loss": 0.2023, + "step": 14680 + }, + { + "epoch": 6.825197582519758, + "grad_norm": 5.18449592590332, + "learning_rate": 9.985325998532602e-06, + "loss": 0.2083, + "step": 14681 + }, + { + "epoch": 6.825662482566249, + "grad_norm": 7.362532138824463, + "learning_rate": 9.985324998532501e-06, + "loss": 0.2462, + "step": 14682 + }, + { + "epoch": 6.826127382612738, + "grad_norm": 6.314716339111328, + "learning_rate": 9.9853239985324e-06, + "loss": 0.2059, + "step": 14683 + }, + { + "epoch": 6.826592282659228, + "grad_norm": 4.466240882873535, + "learning_rate": 9.9853229985323e-06, + "loss": 0.0717, + "step": 14684 + }, + { + "epoch": 6.827057182705718, + "grad_norm": 5.303974628448486, + "learning_rate": 9.9853219985322e-06, + "loss": 0.0874, + "step": 14685 + }, + { + "epoch": 6.827522082752209, + "grad_norm": 9.074934005737305, + "learning_rate": 9.9853209985321e-06, + "loss": 0.4477, + "step": 14686 + }, + { + "epoch": 6.8279869827986985, + "grad_norm": 7.2992048263549805, + "learning_rate": 9.985319998532001e-06, + "loss": 0.3543, + "step": 14687 + }, + { + "epoch": 6.828451882845188, + "grad_norm": 4.190518379211426, + "learning_rate": 9.985318998531902e-06, + "loss": 0.0556, + "step": 14688 + }, + { + "epoch": 6.828916782891678, + "grad_norm": 20.391319274902344, + "learning_rate": 9.9853179985318e-06, + "loss": 1.3237, + "step": 14689 + }, + { + "epoch": 6.829381682938168, + "grad_norm": 5.17393159866333, + "learning_rate": 9.985316998531701e-06, + "loss": 0.1547, + "step": 14690 + }, + { + "epoch": 6.829846582984659, + "grad_norm": 3.596353054046631, + "learning_rate": 9.9853159985316e-06, + "loss": 0.1355, + "step": 14691 + }, + { + "epoch": 6.830311483031148, + "grad_norm": 5.6845526695251465, + "learning_rate": 9.9853149985315e-06, + "loss": 0.2634, + "step": 14692 + }, + { + "epoch": 6.830776383077638, + "grad_norm": 9.389294624328613, + "learning_rate": 9.985313998531401e-06, + "loss": 0.211, + "step": 14693 + }, + { + "epoch": 6.831241283124128, + "grad_norm": 5.606850624084473, + "learning_rate": 9.9853129985313e-06, + "loss": 0.1451, + "step": 14694 + }, + { + "epoch": 6.831706183170619, + "grad_norm": 5.905896186828613, + "learning_rate": 9.985311998531201e-06, + "loss": 0.228, + "step": 14695 + }, + { + "epoch": 6.8321710832171085, + "grad_norm": 4.693964004516602, + "learning_rate": 9.9853109985311e-06, + "loss": 0.1418, + "step": 14696 + }, + { + "epoch": 6.832635983263598, + "grad_norm": 4.53559684753418, + "learning_rate": 9.985309998531e-06, + "loss": 0.2018, + "step": 14697 + }, + { + "epoch": 6.833100883310088, + "grad_norm": 5.014327526092529, + "learning_rate": 9.9853089985309e-06, + "loss": 0.226, + "step": 14698 + }, + { + "epoch": 6.833565783356578, + "grad_norm": 5.927753925323486, + "learning_rate": 9.9853079985308e-06, + "loss": 0.2719, + "step": 14699 + }, + { + "epoch": 6.834030683403069, + "grad_norm": 6.9585652351379395, + "learning_rate": 9.985306998530701e-06, + "loss": 0.1787, + "step": 14700 + }, + { + "epoch": 6.834495583449558, + "grad_norm": 5.168231010437012, + "learning_rate": 9.985305998530602e-06, + "loss": 0.1876, + "step": 14701 + }, + { + "epoch": 6.834960483496048, + "grad_norm": 3.8555076122283936, + "learning_rate": 9.9853049985305e-06, + "loss": 0.0471, + "step": 14702 + }, + { + "epoch": 6.835425383542538, + "grad_norm": 6.273199558258057, + "learning_rate": 9.985303998530401e-06, + "loss": 0.2274, + "step": 14703 + }, + { + "epoch": 6.835890283589029, + "grad_norm": 6.159404277801514, + "learning_rate": 9.9853029985303e-06, + "loss": 0.1688, + "step": 14704 + }, + { + "epoch": 6.8363551836355185, + "grad_norm": 3.9885034561157227, + "learning_rate": 9.985301998530201e-06, + "loss": 0.073, + "step": 14705 + }, + { + "epoch": 6.836820083682008, + "grad_norm": 5.861137866973877, + "learning_rate": 9.985300998530102e-06, + "loss": 0.1673, + "step": 14706 + }, + { + "epoch": 6.837284983728498, + "grad_norm": 6.253775119781494, + "learning_rate": 9.98529999853e-06, + "loss": 0.2069, + "step": 14707 + }, + { + "epoch": 6.837749883774988, + "grad_norm": 4.521875858306885, + "learning_rate": 9.985298998529901e-06, + "loss": 0.0651, + "step": 14708 + }, + { + "epoch": 6.838214783821479, + "grad_norm": 4.53515625, + "learning_rate": 9.9852979985298e-06, + "loss": 0.1352, + "step": 14709 + }, + { + "epoch": 6.838679683867968, + "grad_norm": 4.539947032928467, + "learning_rate": 9.9852969985297e-06, + "loss": 0.2503, + "step": 14710 + }, + { + "epoch": 6.839144583914458, + "grad_norm": 8.914709091186523, + "learning_rate": 9.9852959985296e-06, + "loss": 0.5188, + "step": 14711 + }, + { + "epoch": 6.839609483960948, + "grad_norm": 5.288710594177246, + "learning_rate": 9.9852949985295e-06, + "loss": 0.2414, + "step": 14712 + }, + { + "epoch": 6.840074384007439, + "grad_norm": 4.69602632522583, + "learning_rate": 9.985293998529401e-06, + "loss": 0.1653, + "step": 14713 + }, + { + "epoch": 6.8405392840539285, + "grad_norm": 3.3828375339508057, + "learning_rate": 9.9852929985293e-06, + "loss": 0.0871, + "step": 14714 + }, + { + "epoch": 6.841004184100418, + "grad_norm": 5.8810505867004395, + "learning_rate": 9.9852919985292e-06, + "loss": 0.2227, + "step": 14715 + }, + { + "epoch": 6.841469084146908, + "grad_norm": 4.577090263366699, + "learning_rate": 9.9852909985291e-06, + "loss": 0.1344, + "step": 14716 + }, + { + "epoch": 6.841933984193398, + "grad_norm": 5.903160095214844, + "learning_rate": 9.985289998529e-06, + "loss": 0.2675, + "step": 14717 + }, + { + "epoch": 6.842398884239889, + "grad_norm": 10.732754707336426, + "learning_rate": 9.985288998528901e-06, + "loss": 0.3518, + "step": 14718 + }, + { + "epoch": 6.842863784286378, + "grad_norm": 3.965167999267578, + "learning_rate": 9.985287998528802e-06, + "loss": 0.1568, + "step": 14719 + }, + { + "epoch": 6.843328684332868, + "grad_norm": 6.111323833465576, + "learning_rate": 9.9852869985287e-06, + "loss": 0.3344, + "step": 14720 + }, + { + "epoch": 6.843793584379359, + "grad_norm": 6.900325298309326, + "learning_rate": 9.985285998528601e-06, + "loss": 0.2619, + "step": 14721 + }, + { + "epoch": 6.844258484425849, + "grad_norm": 3.6108415126800537, + "learning_rate": 9.9852849985285e-06, + "loss": 0.1564, + "step": 14722 + }, + { + "epoch": 6.8447233844723385, + "grad_norm": 6.627835750579834, + "learning_rate": 9.985283998528401e-06, + "loss": 0.1346, + "step": 14723 + }, + { + "epoch": 6.845188284518828, + "grad_norm": 4.457483768463135, + "learning_rate": 9.9852829985283e-06, + "loss": 0.1141, + "step": 14724 + }, + { + "epoch": 6.845653184565318, + "grad_norm": 6.550507545471191, + "learning_rate": 9.9852819985282e-06, + "loss": 0.1921, + "step": 14725 + }, + { + "epoch": 6.846118084611809, + "grad_norm": 5.525444507598877, + "learning_rate": 9.985280998528101e-06, + "loss": 0.0715, + "step": 14726 + }, + { + "epoch": 6.846582984658299, + "grad_norm": 5.893231391906738, + "learning_rate": 9.985279998528e-06, + "loss": 0.2543, + "step": 14727 + }, + { + "epoch": 6.847047884704788, + "grad_norm": 6.848794937133789, + "learning_rate": 9.9852789985279e-06, + "loss": 0.3767, + "step": 14728 + }, + { + "epoch": 6.847512784751278, + "grad_norm": 7.153419017791748, + "learning_rate": 9.9852779985278e-06, + "loss": 0.2461, + "step": 14729 + }, + { + "epoch": 6.847977684797769, + "grad_norm": 7.466519355773926, + "learning_rate": 9.9852769985277e-06, + "loss": 0.157, + "step": 14730 + }, + { + "epoch": 6.848442584844259, + "grad_norm": 6.960805892944336, + "learning_rate": 9.985275998527601e-06, + "loss": 0.2515, + "step": 14731 + }, + { + "epoch": 6.8489074848907485, + "grad_norm": 2.8975772857666016, + "learning_rate": 9.985274998527502e-06, + "loss": 0.0428, + "step": 14732 + }, + { + "epoch": 6.849372384937238, + "grad_norm": 6.555139541625977, + "learning_rate": 9.9852739985274e-06, + "loss": 0.2118, + "step": 14733 + }, + { + "epoch": 6.849837284983728, + "grad_norm": 4.926163196563721, + "learning_rate": 9.985272998527301e-06, + "loss": 0.1514, + "step": 14734 + }, + { + "epoch": 6.850302185030219, + "grad_norm": 4.875600337982178, + "learning_rate": 9.9852719985272e-06, + "loss": 0.1506, + "step": 14735 + }, + { + "epoch": 6.850767085076709, + "grad_norm": 3.2325985431671143, + "learning_rate": 9.985270998527101e-06, + "loss": 0.1115, + "step": 14736 + }, + { + "epoch": 6.851231985123198, + "grad_norm": 3.5616509914398193, + "learning_rate": 9.985269998527e-06, + "loss": 0.0654, + "step": 14737 + }, + { + "epoch": 6.851696885169688, + "grad_norm": 11.713199615478516, + "learning_rate": 9.9852689985269e-06, + "loss": 0.2723, + "step": 14738 + }, + { + "epoch": 6.852161785216179, + "grad_norm": 5.317234039306641, + "learning_rate": 9.985267998526801e-06, + "loss": 0.1598, + "step": 14739 + }, + { + "epoch": 6.852626685262669, + "grad_norm": 3.9845056533813477, + "learning_rate": 9.9852669985267e-06, + "loss": 0.1182, + "step": 14740 + }, + { + "epoch": 6.8530915853091585, + "grad_norm": 6.084488868713379, + "learning_rate": 9.9852659985266e-06, + "loss": 0.3203, + "step": 14741 + }, + { + "epoch": 6.853556485355648, + "grad_norm": 5.005555152893066, + "learning_rate": 9.9852649985265e-06, + "loss": 0.1205, + "step": 14742 + }, + { + "epoch": 6.854021385402138, + "grad_norm": 5.050784587860107, + "learning_rate": 9.9852639985264e-06, + "loss": 0.2284, + "step": 14743 + }, + { + "epoch": 6.854486285448629, + "grad_norm": 6.659544467926025, + "learning_rate": 9.985262998526301e-06, + "loss": 0.215, + "step": 14744 + }, + { + "epoch": 6.854951185495119, + "grad_norm": 10.514801025390625, + "learning_rate": 9.985261998526202e-06, + "loss": 0.1891, + "step": 14745 + }, + { + "epoch": 6.855416085541608, + "grad_norm": 4.616992950439453, + "learning_rate": 9.9852609985261e-06, + "loss": 0.1033, + "step": 14746 + }, + { + "epoch": 6.855880985588098, + "grad_norm": 3.1589272022247314, + "learning_rate": 9.985259998526e-06, + "loss": 0.0814, + "step": 14747 + }, + { + "epoch": 6.856345885634589, + "grad_norm": 6.791149139404297, + "learning_rate": 9.9852589985259e-06, + "loss": 0.3551, + "step": 14748 + }, + { + "epoch": 6.856810785681079, + "grad_norm": 7.57690954208374, + "learning_rate": 9.985257998525801e-06, + "loss": 0.2196, + "step": 14749 + }, + { + "epoch": 6.8572756857275685, + "grad_norm": 8.372518539428711, + "learning_rate": 9.985256998525702e-06, + "loss": 0.2941, + "step": 14750 + }, + { + "epoch": 6.857740585774058, + "grad_norm": 5.4578070640563965, + "learning_rate": 9.9852559985256e-06, + "loss": 0.1344, + "step": 14751 + }, + { + "epoch": 6.858205485820548, + "grad_norm": 7.4329633712768555, + "learning_rate": 9.985254998525501e-06, + "loss": 0.2286, + "step": 14752 + }, + { + "epoch": 6.858670385867039, + "grad_norm": 5.990217208862305, + "learning_rate": 9.9852539985254e-06, + "loss": 0.2598, + "step": 14753 + }, + { + "epoch": 6.859135285913529, + "grad_norm": 4.236191272735596, + "learning_rate": 9.985252998525301e-06, + "loss": 0.0797, + "step": 14754 + }, + { + "epoch": 6.859600185960018, + "grad_norm": 8.882636070251465, + "learning_rate": 9.9852519985252e-06, + "loss": 0.2177, + "step": 14755 + }, + { + "epoch": 6.860065086006508, + "grad_norm": 3.911259651184082, + "learning_rate": 9.9852509985251e-06, + "loss": 0.0823, + "step": 14756 + }, + { + "epoch": 6.860529986052999, + "grad_norm": 4.3677659034729, + "learning_rate": 9.985249998525001e-06, + "loss": 0.0859, + "step": 14757 + }, + { + "epoch": 6.860994886099489, + "grad_norm": 5.603241920471191, + "learning_rate": 9.9852489985249e-06, + "loss": 0.1048, + "step": 14758 + }, + { + "epoch": 6.8614597861459785, + "grad_norm": 4.000874996185303, + "learning_rate": 9.9852479985248e-06, + "loss": 0.1762, + "step": 14759 + }, + { + "epoch": 6.861924686192468, + "grad_norm": 4.789954662322998, + "learning_rate": 9.9852469985247e-06, + "loss": 0.1992, + "step": 14760 + }, + { + "epoch": 6.862389586238959, + "grad_norm": 5.72076940536499, + "learning_rate": 9.9852459985246e-06, + "loss": 0.227, + "step": 14761 + }, + { + "epoch": 6.862854486285449, + "grad_norm": 6.788334369659424, + "learning_rate": 9.985244998524501e-06, + "loss": 0.2708, + "step": 14762 + }, + { + "epoch": 6.863319386331939, + "grad_norm": 4.0596604347229, + "learning_rate": 9.985243998524402e-06, + "loss": 0.1141, + "step": 14763 + }, + { + "epoch": 6.863784286378428, + "grad_norm": 6.668142318725586, + "learning_rate": 9.9852429985243e-06, + "loss": 0.1335, + "step": 14764 + }, + { + "epoch": 6.864249186424919, + "grad_norm": 6.560819149017334, + "learning_rate": 9.985241998524201e-06, + "loss": 0.3518, + "step": 14765 + }, + { + "epoch": 6.864714086471409, + "grad_norm": 3.858323335647583, + "learning_rate": 9.9852409985241e-06, + "loss": 0.1216, + "step": 14766 + }, + { + "epoch": 6.865178986517899, + "grad_norm": 11.43232250213623, + "learning_rate": 9.985239998524001e-06, + "loss": 0.262, + "step": 14767 + }, + { + "epoch": 6.8656438865643885, + "grad_norm": 3.8069424629211426, + "learning_rate": 9.9852389985239e-06, + "loss": 0.0896, + "step": 14768 + }, + { + "epoch": 6.866108786610878, + "grad_norm": 5.756965160369873, + "learning_rate": 9.9852379985238e-06, + "loss": 0.2098, + "step": 14769 + }, + { + "epoch": 6.866573686657369, + "grad_norm": 4.28585958480835, + "learning_rate": 9.985236998523701e-06, + "loss": 0.1109, + "step": 14770 + }, + { + "epoch": 6.867038586703859, + "grad_norm": 3.8062820434570312, + "learning_rate": 9.9852359985236e-06, + "loss": 0.0813, + "step": 14771 + }, + { + "epoch": 6.867503486750349, + "grad_norm": 7.030083656311035, + "learning_rate": 9.985234998523501e-06, + "loss": 0.2847, + "step": 14772 + }, + { + "epoch": 6.867968386796838, + "grad_norm": 4.437774658203125, + "learning_rate": 9.9852339985234e-06, + "loss": 0.1294, + "step": 14773 + }, + { + "epoch": 6.868433286843329, + "grad_norm": 3.875201940536499, + "learning_rate": 9.9852329985233e-06, + "loss": 0.1004, + "step": 14774 + }, + { + "epoch": 6.868898186889819, + "grad_norm": 6.4523820877075195, + "learning_rate": 9.985231998523201e-06, + "loss": 0.3641, + "step": 14775 + }, + { + "epoch": 6.869363086936309, + "grad_norm": 7.229584693908691, + "learning_rate": 9.985230998523102e-06, + "loss": 0.2366, + "step": 14776 + }, + { + "epoch": 6.8698279869827985, + "grad_norm": 4.20968770980835, + "learning_rate": 9.985229998523e-06, + "loss": 0.1366, + "step": 14777 + }, + { + "epoch": 6.870292887029288, + "grad_norm": 6.155418395996094, + "learning_rate": 9.985228998522901e-06, + "loss": 0.2904, + "step": 14778 + }, + { + "epoch": 6.870757787075779, + "grad_norm": 6.094714641571045, + "learning_rate": 9.9852279985228e-06, + "loss": 0.2165, + "step": 14779 + }, + { + "epoch": 6.871222687122269, + "grad_norm": 4.217211723327637, + "learning_rate": 9.9852269985227e-06, + "loss": 0.1021, + "step": 14780 + }, + { + "epoch": 6.871687587168759, + "grad_norm": 2.50280499458313, + "learning_rate": 9.9852259985226e-06, + "loss": 0.0347, + "step": 14781 + }, + { + "epoch": 6.872152487215248, + "grad_norm": 5.241019248962402, + "learning_rate": 9.9852249985225e-06, + "loss": 0.1223, + "step": 14782 + }, + { + "epoch": 6.872617387261739, + "grad_norm": 6.203040599822998, + "learning_rate": 9.985223998522401e-06, + "loss": 0.2179, + "step": 14783 + }, + { + "epoch": 6.873082287308229, + "grad_norm": 4.8691182136535645, + "learning_rate": 9.9852229985223e-06, + "loss": 0.2029, + "step": 14784 + }, + { + "epoch": 6.873547187354719, + "grad_norm": 3.6809184551239014, + "learning_rate": 9.985221998522201e-06, + "loss": 0.0812, + "step": 14785 + }, + { + "epoch": 6.8740120874012085, + "grad_norm": 6.838489532470703, + "learning_rate": 9.9852209985221e-06, + "loss": 0.4026, + "step": 14786 + }, + { + "epoch": 6.874476987447698, + "grad_norm": 10.31623363494873, + "learning_rate": 9.985219998522e-06, + "loss": 0.5501, + "step": 14787 + }, + { + "epoch": 6.874941887494189, + "grad_norm": 5.64394474029541, + "learning_rate": 9.985218998521901e-06, + "loss": 0.2883, + "step": 14788 + }, + { + "epoch": 6.875406787540679, + "grad_norm": 6.417457580566406, + "learning_rate": 9.985217998521802e-06, + "loss": 0.2583, + "step": 14789 + }, + { + "epoch": 6.875871687587169, + "grad_norm": 3.622337579727173, + "learning_rate": 9.985216998521701e-06, + "loss": 0.1086, + "step": 14790 + }, + { + "epoch": 6.876336587633658, + "grad_norm": 4.100744247436523, + "learning_rate": 9.9852159985216e-06, + "loss": 0.0682, + "step": 14791 + }, + { + "epoch": 6.876801487680149, + "grad_norm": 4.1672234535217285, + "learning_rate": 9.9852149985215e-06, + "loss": 0.1488, + "step": 14792 + }, + { + "epoch": 6.877266387726639, + "grad_norm": 7.282739639282227, + "learning_rate": 9.9852139985214e-06, + "loss": 0.2043, + "step": 14793 + }, + { + "epoch": 6.877731287773129, + "grad_norm": 5.058342456817627, + "learning_rate": 9.9852129985213e-06, + "loss": 0.2131, + "step": 14794 + }, + { + "epoch": 6.8781961878196185, + "grad_norm": 2.305824041366577, + "learning_rate": 9.9852119985212e-06, + "loss": 0.0444, + "step": 14795 + }, + { + "epoch": 6.878661087866108, + "grad_norm": 7.594357967376709, + "learning_rate": 9.985210998521101e-06, + "loss": 0.2085, + "step": 14796 + }, + { + "epoch": 6.879125987912599, + "grad_norm": 4.7306694984436035, + "learning_rate": 9.985209998521e-06, + "loss": 0.141, + "step": 14797 + }, + { + "epoch": 6.879590887959089, + "grad_norm": 4.616827487945557, + "learning_rate": 9.985208998520901e-06, + "loss": 0.2176, + "step": 14798 + }, + { + "epoch": 6.880055788005579, + "grad_norm": 6.336207389831543, + "learning_rate": 9.9852079985208e-06, + "loss": 0.2452, + "step": 14799 + }, + { + "epoch": 6.880520688052069, + "grad_norm": 4.6315693855285645, + "learning_rate": 9.9852069985207e-06, + "loss": 0.2171, + "step": 14800 + }, + { + "epoch": 6.880985588098559, + "grad_norm": 2.8225035667419434, + "learning_rate": 9.985205998520601e-06, + "loss": 0.0631, + "step": 14801 + }, + { + "epoch": 6.881450488145049, + "grad_norm": 4.363080024719238, + "learning_rate": 9.9852049985205e-06, + "loss": 0.0852, + "step": 14802 + }, + { + "epoch": 6.881915388191539, + "grad_norm": 4.238395690917969, + "learning_rate": 9.985203998520401e-06, + "loss": 0.0648, + "step": 14803 + }, + { + "epoch": 6.8823802882380285, + "grad_norm": 4.98714017868042, + "learning_rate": 9.9852029985203e-06, + "loss": 0.1637, + "step": 14804 + }, + { + "epoch": 6.882845188284519, + "grad_norm": 6.253007888793945, + "learning_rate": 9.9852019985202e-06, + "loss": 0.275, + "step": 14805 + }, + { + "epoch": 6.883310088331009, + "grad_norm": 7.232720375061035, + "learning_rate": 9.9852009985201e-06, + "loss": 0.3115, + "step": 14806 + }, + { + "epoch": 6.883774988377499, + "grad_norm": 8.19123363494873, + "learning_rate": 9.98519999852e-06, + "loss": 0.2243, + "step": 14807 + }, + { + "epoch": 6.884239888423989, + "grad_norm": 3.554705858230591, + "learning_rate": 9.985198998519901e-06, + "loss": 0.0852, + "step": 14808 + }, + { + "epoch": 6.884704788470479, + "grad_norm": 4.837335586547852, + "learning_rate": 9.985197998519802e-06, + "loss": 0.2579, + "step": 14809 + }, + { + "epoch": 6.885169688516969, + "grad_norm": 3.1663479804992676, + "learning_rate": 9.9851969985197e-06, + "loss": 0.0609, + "step": 14810 + }, + { + "epoch": 6.885634588563459, + "grad_norm": 4.673016548156738, + "learning_rate": 9.985195998519601e-06, + "loss": 0.1731, + "step": 14811 + }, + { + "epoch": 6.886099488609949, + "grad_norm": 7.9106855392456055, + "learning_rate": 9.9851949985195e-06, + "loss": 0.2251, + "step": 14812 + }, + { + "epoch": 6.8865643886564385, + "grad_norm": 8.044633865356445, + "learning_rate": 9.9851939985194e-06, + "loss": 0.1789, + "step": 14813 + }, + { + "epoch": 6.887029288702929, + "grad_norm": 10.398838996887207, + "learning_rate": 9.985192998519301e-06, + "loss": 0.247, + "step": 14814 + }, + { + "epoch": 6.887494188749419, + "grad_norm": 4.405643463134766, + "learning_rate": 9.9851919985192e-06, + "loss": 0.126, + "step": 14815 + }, + { + "epoch": 6.887959088795909, + "grad_norm": 4.763040542602539, + "learning_rate": 9.985190998519101e-06, + "loss": 0.1896, + "step": 14816 + }, + { + "epoch": 6.888423988842399, + "grad_norm": 5.052464962005615, + "learning_rate": 9.985189998519e-06, + "loss": 0.1779, + "step": 14817 + }, + { + "epoch": 6.888888888888889, + "grad_norm": 9.76246166229248, + "learning_rate": 9.9851889985189e-06, + "loss": 0.2653, + "step": 14818 + }, + { + "epoch": 6.889353788935379, + "grad_norm": 8.084080696105957, + "learning_rate": 9.985187998518801e-06, + "loss": 0.3314, + "step": 14819 + }, + { + "epoch": 6.889818688981869, + "grad_norm": 5.961977958679199, + "learning_rate": 9.9851869985187e-06, + "loss": 0.2212, + "step": 14820 + }, + { + "epoch": 6.890283589028359, + "grad_norm": 9.729846954345703, + "learning_rate": 9.985185998518601e-06, + "loss": 0.1911, + "step": 14821 + }, + { + "epoch": 6.8907484890748485, + "grad_norm": 5.968767166137695, + "learning_rate": 9.985184998518502e-06, + "loss": 0.2282, + "step": 14822 + }, + { + "epoch": 6.891213389121339, + "grad_norm": 4.010174751281738, + "learning_rate": 9.9851839985184e-06, + "loss": 0.1539, + "step": 14823 + }, + { + "epoch": 6.891678289167829, + "grad_norm": 6.288170337677002, + "learning_rate": 9.9851829985183e-06, + "loss": 0.2937, + "step": 14824 + }, + { + "epoch": 6.892143189214319, + "grad_norm": 6.047830581665039, + "learning_rate": 9.9851819985182e-06, + "loss": 0.1564, + "step": 14825 + }, + { + "epoch": 6.892608089260809, + "grad_norm": 4.606053829193115, + "learning_rate": 9.985180998518101e-06, + "loss": 0.1038, + "step": 14826 + }, + { + "epoch": 6.893072989307299, + "grad_norm": 6.052285194396973, + "learning_rate": 9.985179998518002e-06, + "loss": 0.3145, + "step": 14827 + }, + { + "epoch": 6.893537889353789, + "grad_norm": 7.444308280944824, + "learning_rate": 9.9851789985179e-06, + "loss": 0.212, + "step": 14828 + }, + { + "epoch": 6.894002789400279, + "grad_norm": 5.083962440490723, + "learning_rate": 9.985177998517801e-06, + "loss": 0.2393, + "step": 14829 + }, + { + "epoch": 6.894467689446769, + "grad_norm": 1.0254958868026733, + "learning_rate": 9.9851769985177e-06, + "loss": 0.0143, + "step": 14830 + }, + { + "epoch": 6.8949325894932585, + "grad_norm": 8.53764533996582, + "learning_rate": 9.9851759985176e-06, + "loss": 0.1535, + "step": 14831 + }, + { + "epoch": 6.895397489539749, + "grad_norm": 4.480588436126709, + "learning_rate": 9.985174998517501e-06, + "loss": 0.2124, + "step": 14832 + }, + { + "epoch": 6.895862389586239, + "grad_norm": 6.88731050491333, + "learning_rate": 9.9851739985174e-06, + "loss": 0.3919, + "step": 14833 + }, + { + "epoch": 6.896327289632729, + "grad_norm": 7.967892169952393, + "learning_rate": 9.985172998517301e-06, + "loss": 0.3223, + "step": 14834 + }, + { + "epoch": 6.896792189679219, + "grad_norm": 6.239243030548096, + "learning_rate": 9.9851719985172e-06, + "loss": 0.2542, + "step": 14835 + }, + { + "epoch": 6.897257089725709, + "grad_norm": 5.033742427825928, + "learning_rate": 9.9851709985171e-06, + "loss": 0.1841, + "step": 14836 + }, + { + "epoch": 6.897721989772199, + "grad_norm": 4.607324600219727, + "learning_rate": 9.985169998517e-06, + "loss": 0.1637, + "step": 14837 + }, + { + "epoch": 6.898186889818689, + "grad_norm": 6.301019191741943, + "learning_rate": 9.9851689985169e-06, + "loss": 0.2568, + "step": 14838 + }, + { + "epoch": 6.898651789865179, + "grad_norm": 3.514328718185425, + "learning_rate": 9.985167998516801e-06, + "loss": 0.0597, + "step": 14839 + }, + { + "epoch": 6.899116689911669, + "grad_norm": 3.6737136840820312, + "learning_rate": 9.985166998516702e-06, + "loss": 0.1118, + "step": 14840 + }, + { + "epoch": 6.899581589958159, + "grad_norm": 6.248421669006348, + "learning_rate": 9.9851659985166e-06, + "loss": 0.2458, + "step": 14841 + }, + { + "epoch": 6.900046490004649, + "grad_norm": 8.234061241149902, + "learning_rate": 9.985164998516501e-06, + "loss": 0.4434, + "step": 14842 + }, + { + "epoch": 6.900511390051139, + "grad_norm": 5.316529750823975, + "learning_rate": 9.9851639985164e-06, + "loss": 0.2555, + "step": 14843 + }, + { + "epoch": 6.9009762900976295, + "grad_norm": 5.137557506561279, + "learning_rate": 9.9851629985163e-06, + "loss": 0.0634, + "step": 14844 + }, + { + "epoch": 6.901441190144119, + "grad_norm": 7.7244791984558105, + "learning_rate": 9.985161998516201e-06, + "loss": 0.4175, + "step": 14845 + }, + { + "epoch": 6.901906090190609, + "grad_norm": 4.399478912353516, + "learning_rate": 9.9851609985161e-06, + "loss": 0.133, + "step": 14846 + }, + { + "epoch": 6.902370990237099, + "grad_norm": 7.1342010498046875, + "learning_rate": 9.985159998516001e-06, + "loss": 0.2762, + "step": 14847 + }, + { + "epoch": 6.902835890283589, + "grad_norm": 7.597680568695068, + "learning_rate": 9.9851589985159e-06, + "loss": 0.2209, + "step": 14848 + }, + { + "epoch": 6.903300790330079, + "grad_norm": 7.081957817077637, + "learning_rate": 9.9851579985158e-06, + "loss": 0.3458, + "step": 14849 + }, + { + "epoch": 6.903765690376569, + "grad_norm": 2.602861166000366, + "learning_rate": 9.9851569985157e-06, + "loss": 0.0456, + "step": 14850 + }, + { + "epoch": 6.904230590423059, + "grad_norm": 6.990931987762451, + "learning_rate": 9.9851559985156e-06, + "loss": 0.3739, + "step": 14851 + }, + { + "epoch": 6.904695490469549, + "grad_norm": 4.890588283538818, + "learning_rate": 9.985154998515501e-06, + "loss": 0.1516, + "step": 14852 + }, + { + "epoch": 6.9051603905160395, + "grad_norm": 6.548762798309326, + "learning_rate": 9.985153998515402e-06, + "loss": 0.2748, + "step": 14853 + }, + { + "epoch": 6.905625290562529, + "grad_norm": 2.9307632446289062, + "learning_rate": 9.9851529985153e-06, + "loss": 0.0962, + "step": 14854 + }, + { + "epoch": 6.906090190609019, + "grad_norm": 6.503091335296631, + "learning_rate": 9.9851519985152e-06, + "loss": 0.2938, + "step": 14855 + }, + { + "epoch": 6.906555090655509, + "grad_norm": 7.092919826507568, + "learning_rate": 9.9851509985151e-06, + "loss": 0.1312, + "step": 14856 + }, + { + "epoch": 6.907019990701999, + "grad_norm": 3.4600329399108887, + "learning_rate": 9.985149998515001e-06, + "loss": 0.0843, + "step": 14857 + }, + { + "epoch": 6.907484890748489, + "grad_norm": 6.664603233337402, + "learning_rate": 9.985148998514902e-06, + "loss": 0.2661, + "step": 14858 + }, + { + "epoch": 6.907949790794979, + "grad_norm": 5.20229959487915, + "learning_rate": 9.9851479985148e-06, + "loss": 0.1402, + "step": 14859 + }, + { + "epoch": 6.908414690841469, + "grad_norm": 8.447542190551758, + "learning_rate": 9.985146998514701e-06, + "loss": 0.3352, + "step": 14860 + }, + { + "epoch": 6.908879590887959, + "grad_norm": 8.552774429321289, + "learning_rate": 9.9851459985146e-06, + "loss": 0.3604, + "step": 14861 + }, + { + "epoch": 6.9093444909344495, + "grad_norm": 2.8620593547821045, + "learning_rate": 9.9851449985145e-06, + "loss": 0.0818, + "step": 14862 + }, + { + "epoch": 6.909809390980939, + "grad_norm": 6.529703617095947, + "learning_rate": 9.9851439985144e-06, + "loss": 0.102, + "step": 14863 + }, + { + "epoch": 6.910274291027429, + "grad_norm": 5.551860332489014, + "learning_rate": 9.9851429985143e-06, + "loss": 0.2055, + "step": 14864 + }, + { + "epoch": 6.910739191073919, + "grad_norm": 4.44672155380249, + "learning_rate": 9.985141998514201e-06, + "loss": 0.1501, + "step": 14865 + }, + { + "epoch": 6.911204091120409, + "grad_norm": 5.610680103302002, + "learning_rate": 9.9851409985141e-06, + "loss": 0.1751, + "step": 14866 + }, + { + "epoch": 6.911668991166899, + "grad_norm": 6.828463077545166, + "learning_rate": 9.985139998514e-06, + "loss": 0.2314, + "step": 14867 + }, + { + "epoch": 6.912133891213389, + "grad_norm": 7.677336692810059, + "learning_rate": 9.9851389985139e-06, + "loss": 0.4524, + "step": 14868 + }, + { + "epoch": 6.912598791259879, + "grad_norm": 4.020958423614502, + "learning_rate": 9.9851379985138e-06, + "loss": 0.1128, + "step": 14869 + }, + { + "epoch": 6.913063691306369, + "grad_norm": 7.323768615722656, + "learning_rate": 9.985136998513701e-06, + "loss": 0.3456, + "step": 14870 + }, + { + "epoch": 6.9135285913528595, + "grad_norm": 10.488662719726562, + "learning_rate": 9.985135998513602e-06, + "loss": 0.2598, + "step": 14871 + }, + { + "epoch": 6.913993491399349, + "grad_norm": 7.128018379211426, + "learning_rate": 9.9851349985135e-06, + "loss": 0.2655, + "step": 14872 + }, + { + "epoch": 6.914458391445839, + "grad_norm": 5.147561550140381, + "learning_rate": 9.985133998513401e-06, + "loss": 0.114, + "step": 14873 + }, + { + "epoch": 6.914923291492329, + "grad_norm": 4.783700466156006, + "learning_rate": 9.9851329985133e-06, + "loss": 0.1851, + "step": 14874 + }, + { + "epoch": 6.915388191538819, + "grad_norm": 7.396942615509033, + "learning_rate": 9.985131998513201e-06, + "loss": 0.2106, + "step": 14875 + }, + { + "epoch": 6.915853091585309, + "grad_norm": 5.435577392578125, + "learning_rate": 9.9851309985131e-06, + "loss": 0.2126, + "step": 14876 + }, + { + "epoch": 6.916317991631799, + "grad_norm": 6.107428073883057, + "learning_rate": 9.985129998513e-06, + "loss": 0.3145, + "step": 14877 + }, + { + "epoch": 6.916782891678289, + "grad_norm": 6.00291633605957, + "learning_rate": 9.985128998512901e-06, + "loss": 0.2183, + "step": 14878 + }, + { + "epoch": 6.91724779172478, + "grad_norm": 7.130781650543213, + "learning_rate": 9.9851279985128e-06, + "loss": 0.1473, + "step": 14879 + }, + { + "epoch": 6.9177126917712695, + "grad_norm": 5.295953273773193, + "learning_rate": 9.9851269985127e-06, + "loss": 0.1828, + "step": 14880 + }, + { + "epoch": 6.918177591817759, + "grad_norm": 5.422684669494629, + "learning_rate": 9.9851259985126e-06, + "loss": 0.1676, + "step": 14881 + }, + { + "epoch": 6.918642491864249, + "grad_norm": 4.98032283782959, + "learning_rate": 9.9851249985125e-06, + "loss": 0.1882, + "step": 14882 + }, + { + "epoch": 6.919107391910739, + "grad_norm": 4.241208553314209, + "learning_rate": 9.985123998512401e-06, + "loss": 0.1207, + "step": 14883 + }, + { + "epoch": 6.91957229195723, + "grad_norm": 4.520400047302246, + "learning_rate": 9.985122998512302e-06, + "loss": 0.1003, + "step": 14884 + }, + { + "epoch": 6.920037192003719, + "grad_norm": 6.519408226013184, + "learning_rate": 9.9851219985122e-06, + "loss": 0.3236, + "step": 14885 + }, + { + "epoch": 6.920502092050209, + "grad_norm": 7.177246570587158, + "learning_rate": 9.985120998512101e-06, + "loss": 0.3562, + "step": 14886 + }, + { + "epoch": 6.920966992096699, + "grad_norm": 5.497581958770752, + "learning_rate": 9.985119998512e-06, + "loss": 0.2735, + "step": 14887 + }, + { + "epoch": 6.92143189214319, + "grad_norm": 5.5780181884765625, + "learning_rate": 9.985118998511901e-06, + "loss": 0.2121, + "step": 14888 + }, + { + "epoch": 6.9218967921896795, + "grad_norm": 7.228084087371826, + "learning_rate": 9.9851179985118e-06, + "loss": 0.2936, + "step": 14889 + }, + { + "epoch": 6.922361692236169, + "grad_norm": 4.176124095916748, + "learning_rate": 9.9851169985117e-06, + "loss": 0.1267, + "step": 14890 + }, + { + "epoch": 6.922826592282659, + "grad_norm": 5.442586898803711, + "learning_rate": 9.985115998511601e-06, + "loss": 0.2164, + "step": 14891 + }, + { + "epoch": 6.923291492329149, + "grad_norm": 3.7338979244232178, + "learning_rate": 9.9851149985115e-06, + "loss": 0.0898, + "step": 14892 + }, + { + "epoch": 6.92375639237564, + "grad_norm": 4.898582935333252, + "learning_rate": 9.985113998511401e-06, + "loss": 0.1937, + "step": 14893 + }, + { + "epoch": 6.924221292422129, + "grad_norm": 6.843324184417725, + "learning_rate": 9.9851129985113e-06, + "loss": 0.3424, + "step": 14894 + }, + { + "epoch": 6.924686192468619, + "grad_norm": 6.18842077255249, + "learning_rate": 9.9851119985112e-06, + "loss": 0.126, + "step": 14895 + }, + { + "epoch": 6.925151092515109, + "grad_norm": 2.89723801612854, + "learning_rate": 9.985110998511101e-06, + "loss": 0.0836, + "step": 14896 + }, + { + "epoch": 6.9256159925616, + "grad_norm": 6.139056205749512, + "learning_rate": 9.985109998511002e-06, + "loss": 0.3008, + "step": 14897 + }, + { + "epoch": 6.9260808926080895, + "grad_norm": 3.542360544204712, + "learning_rate": 9.9851089985109e-06, + "loss": 0.0795, + "step": 14898 + }, + { + "epoch": 6.926545792654579, + "grad_norm": 8.64622688293457, + "learning_rate": 9.9851079985108e-06, + "loss": 0.3122, + "step": 14899 + }, + { + "epoch": 6.927010692701069, + "grad_norm": 4.854213237762451, + "learning_rate": 9.9851069985107e-06, + "loss": 0.1025, + "step": 14900 + }, + { + "epoch": 6.927475592747559, + "grad_norm": 6.294001579284668, + "learning_rate": 9.985105998510601e-06, + "loss": 0.0923, + "step": 14901 + }, + { + "epoch": 6.9279404927940496, + "grad_norm": 4.548885345458984, + "learning_rate": 9.9851049985105e-06, + "loss": 0.1642, + "step": 14902 + }, + { + "epoch": 6.928405392840539, + "grad_norm": 5.112637042999268, + "learning_rate": 9.9851039985104e-06, + "loss": 0.1683, + "step": 14903 + }, + { + "epoch": 6.928870292887029, + "grad_norm": 7.5898356437683105, + "learning_rate": 9.985102998510301e-06, + "loss": 0.4158, + "step": 14904 + }, + { + "epoch": 6.929335192933519, + "grad_norm": 5.618527412414551, + "learning_rate": 9.9851019985102e-06, + "loss": 0.1747, + "step": 14905 + }, + { + "epoch": 6.92980009298001, + "grad_norm": 5.622377395629883, + "learning_rate": 9.985100998510101e-06, + "loss": 0.2026, + "step": 14906 + }, + { + "epoch": 6.9302649930264995, + "grad_norm": 5.44632625579834, + "learning_rate": 9.98509999851e-06, + "loss": 0.1848, + "step": 14907 + }, + { + "epoch": 6.930729893072989, + "grad_norm": 6.100516319274902, + "learning_rate": 9.9850989985099e-06, + "loss": 0.1728, + "step": 14908 + }, + { + "epoch": 6.931194793119479, + "grad_norm": 2.575998544692993, + "learning_rate": 9.985097998509801e-06, + "loss": 0.0902, + "step": 14909 + }, + { + "epoch": 6.931659693165969, + "grad_norm": 4.44918966293335, + "learning_rate": 9.9850969985097e-06, + "loss": 0.1056, + "step": 14910 + }, + { + "epoch": 6.9321245932124596, + "grad_norm": 5.051644802093506, + "learning_rate": 9.985095998509601e-06, + "loss": 0.2118, + "step": 14911 + }, + { + "epoch": 6.932589493258949, + "grad_norm": 3.4602627754211426, + "learning_rate": 9.9850949985095e-06, + "loss": 0.0685, + "step": 14912 + }, + { + "epoch": 6.933054393305439, + "grad_norm": 6.297729015350342, + "learning_rate": 9.9850939985094e-06, + "loss": 0.2352, + "step": 14913 + }, + { + "epoch": 6.933519293351929, + "grad_norm": 3.694105863571167, + "learning_rate": 9.985092998509301e-06, + "loss": 0.091, + "step": 14914 + }, + { + "epoch": 6.93398419339842, + "grad_norm": 4.081076622009277, + "learning_rate": 9.985091998509202e-06, + "loss": 0.1332, + "step": 14915 + }, + { + "epoch": 6.9344490934449095, + "grad_norm": 4.670361518859863, + "learning_rate": 9.9850909985091e-06, + "loss": 0.0917, + "step": 14916 + }, + { + "epoch": 6.934913993491399, + "grad_norm": 6.991526126861572, + "learning_rate": 9.985089998509001e-06, + "loss": 0.2906, + "step": 14917 + }, + { + "epoch": 6.935378893537889, + "grad_norm": 5.991152286529541, + "learning_rate": 9.9850889985089e-06, + "loss": 0.1114, + "step": 14918 + }, + { + "epoch": 6.93584379358438, + "grad_norm": 5.83331823348999, + "learning_rate": 9.985087998508801e-06, + "loss": 0.1681, + "step": 14919 + }, + { + "epoch": 6.9363086936308695, + "grad_norm": 6.564290523529053, + "learning_rate": 9.9850869985087e-06, + "loss": 0.2155, + "step": 14920 + }, + { + "epoch": 6.936773593677359, + "grad_norm": 4.351553916931152, + "learning_rate": 9.9850859985086e-06, + "loss": 0.1143, + "step": 14921 + }, + { + "epoch": 6.937238493723849, + "grad_norm": 7.415225982666016, + "learning_rate": 9.985084998508501e-06, + "loss": 0.3323, + "step": 14922 + }, + { + "epoch": 6.93770339377034, + "grad_norm": 6.803149223327637, + "learning_rate": 9.9850839985084e-06, + "loss": 0.1557, + "step": 14923 + }, + { + "epoch": 6.93816829381683, + "grad_norm": 5.151736259460449, + "learning_rate": 9.985082998508301e-06, + "loss": 0.2259, + "step": 14924 + }, + { + "epoch": 6.9386331938633194, + "grad_norm": 7.806595802307129, + "learning_rate": 9.9850819985082e-06, + "loss": 0.1316, + "step": 14925 + }, + { + "epoch": 6.939098093909809, + "grad_norm": 3.2359812259674072, + "learning_rate": 9.9850809985081e-06, + "loss": 0.0802, + "step": 14926 + }, + { + "epoch": 6.939562993956299, + "grad_norm": 7.137385368347168, + "learning_rate": 9.985079998508001e-06, + "loss": 0.4001, + "step": 14927 + }, + { + "epoch": 6.94002789400279, + "grad_norm": 6.106534957885742, + "learning_rate": 9.985078998507902e-06, + "loss": 0.2074, + "step": 14928 + }, + { + "epoch": 6.9404927940492795, + "grad_norm": 4.262767314910889, + "learning_rate": 9.9850779985078e-06, + "loss": 0.2222, + "step": 14929 + }, + { + "epoch": 6.940957694095769, + "grad_norm": 4.551526069641113, + "learning_rate": 9.985076998507701e-06, + "loss": 0.1596, + "step": 14930 + }, + { + "epoch": 6.941422594142259, + "grad_norm": 3.7853844165802, + "learning_rate": 9.9850759985076e-06, + "loss": 0.0938, + "step": 14931 + }, + { + "epoch": 6.94188749418875, + "grad_norm": 10.28852653503418, + "learning_rate": 9.9850749985075e-06, + "loss": 0.2386, + "step": 14932 + }, + { + "epoch": 6.94235239423524, + "grad_norm": 4.819042682647705, + "learning_rate": 9.9850739985074e-06, + "loss": 0.1789, + "step": 14933 + }, + { + "epoch": 6.9428172942817294, + "grad_norm": 4.8790411949157715, + "learning_rate": 9.9850729985073e-06, + "loss": 0.1461, + "step": 14934 + }, + { + "epoch": 6.943282194328219, + "grad_norm": 2.997812509536743, + "learning_rate": 9.985071998507201e-06, + "loss": 0.0498, + "step": 14935 + }, + { + "epoch": 6.943747094374709, + "grad_norm": 5.511777877807617, + "learning_rate": 9.9850709985071e-06, + "loss": 0.3314, + "step": 14936 + }, + { + "epoch": 6.9442119944212, + "grad_norm": 8.153706550598145, + "learning_rate": 9.985069998507001e-06, + "loss": 0.3703, + "step": 14937 + }, + { + "epoch": 6.9446768944676895, + "grad_norm": 4.494490146636963, + "learning_rate": 9.9850689985069e-06, + "loss": 0.1145, + "step": 14938 + }, + { + "epoch": 6.945141794514179, + "grad_norm": 5.636542797088623, + "learning_rate": 9.9850679985068e-06, + "loss": 0.0828, + "step": 14939 + }, + { + "epoch": 6.945606694560669, + "grad_norm": 3.9270122051239014, + "learning_rate": 9.985066998506701e-06, + "loss": 0.1691, + "step": 14940 + }, + { + "epoch": 6.94607159460716, + "grad_norm": 7.577341556549072, + "learning_rate": 9.985065998506602e-06, + "loss": 0.2691, + "step": 14941 + }, + { + "epoch": 6.94653649465365, + "grad_norm": 6.1104207038879395, + "learning_rate": 9.985064998506501e-06, + "loss": 0.3045, + "step": 14942 + }, + { + "epoch": 6.947001394700139, + "grad_norm": 7.691005706787109, + "learning_rate": 9.9850639985064e-06, + "loss": 0.229, + "step": 14943 + }, + { + "epoch": 6.947466294746629, + "grad_norm": 4.972574710845947, + "learning_rate": 9.9850629985063e-06, + "loss": 0.2212, + "step": 14944 + }, + { + "epoch": 6.947931194793119, + "grad_norm": 4.894063472747803, + "learning_rate": 9.9850619985062e-06, + "loss": 0.113, + "step": 14945 + }, + { + "epoch": 6.94839609483961, + "grad_norm": 5.260997295379639, + "learning_rate": 9.9850609985061e-06, + "loss": 0.1623, + "step": 14946 + }, + { + "epoch": 6.9488609948860995, + "grad_norm": 6.836026668548584, + "learning_rate": 9.985059998506e-06, + "loss": 0.2525, + "step": 14947 + }, + { + "epoch": 6.949325894932589, + "grad_norm": 5.268182277679443, + "learning_rate": 9.985058998505901e-06, + "loss": 0.1635, + "step": 14948 + }, + { + "epoch": 6.949790794979079, + "grad_norm": 5.2691731452941895, + "learning_rate": 9.9850579985058e-06, + "loss": 0.2976, + "step": 14949 + }, + { + "epoch": 6.95025569502557, + "grad_norm": 25.178401947021484, + "learning_rate": 9.985056998505701e-06, + "loss": 0.2966, + "step": 14950 + }, + { + "epoch": 6.95072059507206, + "grad_norm": 4.842479705810547, + "learning_rate": 9.9850559985056e-06, + "loss": 0.1293, + "step": 14951 + }, + { + "epoch": 6.951185495118549, + "grad_norm": 8.896674156188965, + "learning_rate": 9.9850549985055e-06, + "loss": 0.1874, + "step": 14952 + }, + { + "epoch": 6.951650395165039, + "grad_norm": 3.848992109298706, + "learning_rate": 9.985053998505401e-06, + "loss": 0.0688, + "step": 14953 + }, + { + "epoch": 6.952115295211529, + "grad_norm": 5.344127655029297, + "learning_rate": 9.9850529985053e-06, + "loss": 0.1503, + "step": 14954 + }, + { + "epoch": 6.95258019525802, + "grad_norm": 5.450689792633057, + "learning_rate": 9.985051998505201e-06, + "loss": 0.1686, + "step": 14955 + }, + { + "epoch": 6.9530450953045095, + "grad_norm": 8.04057502746582, + "learning_rate": 9.9850509985051e-06, + "loss": 0.2543, + "step": 14956 + }, + { + "epoch": 6.953509995350999, + "grad_norm": 5.003421783447266, + "learning_rate": 9.985049998505e-06, + "loss": 0.2323, + "step": 14957 + }, + { + "epoch": 6.95397489539749, + "grad_norm": 8.617466926574707, + "learning_rate": 9.9850489985049e-06, + "loss": 0.3075, + "step": 14958 + }, + { + "epoch": 6.95443979544398, + "grad_norm": 5.459234714508057, + "learning_rate": 9.9850479985048e-06, + "loss": 0.1626, + "step": 14959 + }, + { + "epoch": 6.95490469549047, + "grad_norm": 10.05202865600586, + "learning_rate": 9.985046998504701e-06, + "loss": 0.1872, + "step": 14960 + }, + { + "epoch": 6.955369595536959, + "grad_norm": 3.8845503330230713, + "learning_rate": 9.985045998504602e-06, + "loss": 0.1022, + "step": 14961 + }, + { + "epoch": 6.955834495583449, + "grad_norm": 7.803828716278076, + "learning_rate": 9.9850449985045e-06, + "loss": 0.2275, + "step": 14962 + }, + { + "epoch": 6.95629939562994, + "grad_norm": 5.8797383308410645, + "learning_rate": 9.985043998504401e-06, + "loss": 0.1979, + "step": 14963 + }, + { + "epoch": 6.95676429567643, + "grad_norm": 5.265753746032715, + "learning_rate": 9.9850429985043e-06, + "loss": 0.1241, + "step": 14964 + }, + { + "epoch": 6.9572291957229195, + "grad_norm": 4.039416790008545, + "learning_rate": 9.9850419985042e-06, + "loss": 0.0637, + "step": 14965 + }, + { + "epoch": 6.957694095769409, + "grad_norm": 3.938345432281494, + "learning_rate": 9.985040998504101e-06, + "loss": 0.1261, + "step": 14966 + }, + { + "epoch": 6.9581589958159, + "grad_norm": 5.359302520751953, + "learning_rate": 9.985039998504e-06, + "loss": 0.2013, + "step": 14967 + }, + { + "epoch": 6.95862389586239, + "grad_norm": 5.615991115570068, + "learning_rate": 9.985038998503901e-06, + "loss": 0.2634, + "step": 14968 + }, + { + "epoch": 6.95908879590888, + "grad_norm": 3.4823057651519775, + "learning_rate": 9.9850379985038e-06, + "loss": 0.1047, + "step": 14969 + }, + { + "epoch": 6.959553695955369, + "grad_norm": 3.881791353225708, + "learning_rate": 9.9850369985037e-06, + "loss": 0.095, + "step": 14970 + }, + { + "epoch": 6.960018596001859, + "grad_norm": 7.555379390716553, + "learning_rate": 9.9850359985036e-06, + "loss": 0.2394, + "step": 14971 + }, + { + "epoch": 6.96048349604835, + "grad_norm": 7.858299732208252, + "learning_rate": 9.9850349985035e-06, + "loss": 0.2592, + "step": 14972 + }, + { + "epoch": 6.96094839609484, + "grad_norm": 7.708330154418945, + "learning_rate": 9.985033998503401e-06, + "loss": 0.4161, + "step": 14973 + }, + { + "epoch": 6.9614132961413295, + "grad_norm": 5.105140686035156, + "learning_rate": 9.985032998503302e-06, + "loss": 0.1495, + "step": 14974 + }, + { + "epoch": 6.961878196187819, + "grad_norm": 5.163631916046143, + "learning_rate": 9.9850319985032e-06, + "loss": 0.1601, + "step": 14975 + }, + { + "epoch": 6.96234309623431, + "grad_norm": 3.7970762252807617, + "learning_rate": 9.9850309985031e-06, + "loss": 0.0897, + "step": 14976 + }, + { + "epoch": 6.9628079962808, + "grad_norm": 5.343723297119141, + "learning_rate": 9.985029998503e-06, + "loss": 0.195, + "step": 14977 + }, + { + "epoch": 6.96327289632729, + "grad_norm": 4.461476802825928, + "learning_rate": 9.985028998502901e-06, + "loss": 0.0809, + "step": 14978 + }, + { + "epoch": 6.963737796373779, + "grad_norm": 5.206148147583008, + "learning_rate": 9.985027998502802e-06, + "loss": 0.1915, + "step": 14979 + }, + { + "epoch": 6.964202696420269, + "grad_norm": 6.22092342376709, + "learning_rate": 9.9850269985027e-06, + "loss": 0.2301, + "step": 14980 + }, + { + "epoch": 6.96466759646676, + "grad_norm": 6.111321449279785, + "learning_rate": 9.985025998502601e-06, + "loss": 0.2012, + "step": 14981 + }, + { + "epoch": 6.96513249651325, + "grad_norm": 6.003852367401123, + "learning_rate": 9.9850249985025e-06, + "loss": 0.1759, + "step": 14982 + }, + { + "epoch": 6.9655973965597395, + "grad_norm": 6.984818935394287, + "learning_rate": 9.9850239985024e-06, + "loss": 0.3268, + "step": 14983 + }, + { + "epoch": 6.966062296606229, + "grad_norm": 5.658280849456787, + "learning_rate": 9.985022998502301e-06, + "loss": 0.2246, + "step": 14984 + }, + { + "epoch": 6.96652719665272, + "grad_norm": 6.359549522399902, + "learning_rate": 9.9850219985022e-06, + "loss": 0.1627, + "step": 14985 + }, + { + "epoch": 6.96699209669921, + "grad_norm": 3.262873649597168, + "learning_rate": 9.985020998502101e-06, + "loss": 0.0944, + "step": 14986 + }, + { + "epoch": 6.9674569967457, + "grad_norm": 3.758517265319824, + "learning_rate": 9.985019998502e-06, + "loss": 0.091, + "step": 14987 + }, + { + "epoch": 6.967921896792189, + "grad_norm": 6.133572101593018, + "learning_rate": 9.9850189985019e-06, + "loss": 0.1358, + "step": 14988 + }, + { + "epoch": 6.968386796838679, + "grad_norm": 6.307690143585205, + "learning_rate": 9.9850179985018e-06, + "loss": 0.1273, + "step": 14989 + }, + { + "epoch": 6.96885169688517, + "grad_norm": 5.237015247344971, + "learning_rate": 9.9850169985017e-06, + "loss": 0.1608, + "step": 14990 + }, + { + "epoch": 6.96931659693166, + "grad_norm": 6.121021747589111, + "learning_rate": 9.985015998501601e-06, + "loss": 0.1435, + "step": 14991 + }, + { + "epoch": 6.9697814969781495, + "grad_norm": 7.6070051193237305, + "learning_rate": 9.985014998501502e-06, + "loss": 0.3724, + "step": 14992 + }, + { + "epoch": 6.970246397024639, + "grad_norm": 5.126426696777344, + "learning_rate": 9.9850139985014e-06, + "loss": 0.1957, + "step": 14993 + }, + { + "epoch": 6.97071129707113, + "grad_norm": 3.8821113109588623, + "learning_rate": 9.985012998501301e-06, + "loss": 0.0574, + "step": 14994 + }, + { + "epoch": 6.97117619711762, + "grad_norm": 5.689839839935303, + "learning_rate": 9.9850119985012e-06, + "loss": 0.1583, + "step": 14995 + }, + { + "epoch": 6.97164109716411, + "grad_norm": 5.956048011779785, + "learning_rate": 9.985010998501101e-06, + "loss": 0.3751, + "step": 14996 + }, + { + "epoch": 6.972105997210599, + "grad_norm": 8.365638732910156, + "learning_rate": 9.985009998501002e-06, + "loss": 0.3499, + "step": 14997 + }, + { + "epoch": 6.972570897257089, + "grad_norm": 4.513110160827637, + "learning_rate": 9.9850089985009e-06, + "loss": 0.1108, + "step": 14998 + }, + { + "epoch": 6.97303579730358, + "grad_norm": 5.796096324920654, + "learning_rate": 9.985007998500801e-06, + "loss": 0.2411, + "step": 14999 + }, + { + "epoch": 6.97350069735007, + "grad_norm": 4.941866874694824, + "learning_rate": 9.9850069985007e-06, + "loss": 0.2078, + "step": 15000 + }, + { + "epoch": 6.9739655973965595, + "grad_norm": 4.453806400299072, + "learning_rate": 9.9850059985006e-06, + "loss": 0.0637, + "step": 15001 + }, + { + "epoch": 6.97443049744305, + "grad_norm": 2.692145347595215, + "learning_rate": 9.9850049985005e-06, + "loss": 0.0466, + "step": 15002 + }, + { + "epoch": 6.97489539748954, + "grad_norm": 4.645771503448486, + "learning_rate": 9.9850039985004e-06, + "loss": 0.1795, + "step": 15003 + }, + { + "epoch": 6.97536029753603, + "grad_norm": 5.213980197906494, + "learning_rate": 9.985002998500301e-06, + "loss": 0.164, + "step": 15004 + }, + { + "epoch": 6.97582519758252, + "grad_norm": 6.28004789352417, + "learning_rate": 9.985001998500202e-06, + "loss": 0.205, + "step": 15005 + }, + { + "epoch": 6.976290097629009, + "grad_norm": 7.149617671966553, + "learning_rate": 9.9850009985001e-06, + "loss": 0.3259, + "step": 15006 + }, + { + "epoch": 6.9767549976755, + "grad_norm": 2.948094129562378, + "learning_rate": 9.9849999985e-06, + "loss": 0.0284, + "step": 15007 + }, + { + "epoch": 6.97721989772199, + "grad_norm": 7.035444259643555, + "learning_rate": 9.9849989984999e-06, + "loss": 0.3358, + "step": 15008 + }, + { + "epoch": 6.97768479776848, + "grad_norm": 2.985551118850708, + "learning_rate": 9.984997998499801e-06, + "loss": 0.0552, + "step": 15009 + }, + { + "epoch": 6.9781496978149695, + "grad_norm": 6.061812400817871, + "learning_rate": 9.984996998499702e-06, + "loss": 0.1933, + "step": 15010 + }, + { + "epoch": 6.97861459786146, + "grad_norm": 4.966233253479004, + "learning_rate": 9.9849959984996e-06, + "loss": 0.1834, + "step": 15011 + }, + { + "epoch": 6.97907949790795, + "grad_norm": 6.371131896972656, + "learning_rate": 9.984994998499501e-06, + "loss": 0.2316, + "step": 15012 + }, + { + "epoch": 6.97954439795444, + "grad_norm": 2.7380869388580322, + "learning_rate": 9.9849939984994e-06, + "loss": 0.067, + "step": 15013 + }, + { + "epoch": 6.98000929800093, + "grad_norm": 4.20405387878418, + "learning_rate": 9.984992998499301e-06, + "loss": 0.0913, + "step": 15014 + }, + { + "epoch": 6.980474198047419, + "grad_norm": 3.2685914039611816, + "learning_rate": 9.9849919984992e-06, + "loss": 0.0688, + "step": 15015 + }, + { + "epoch": 6.98093909809391, + "grad_norm": 6.2823805809021, + "learning_rate": 9.9849909984991e-06, + "loss": 0.1499, + "step": 15016 + }, + { + "epoch": 6.9814039981404, + "grad_norm": 7.760384559631348, + "learning_rate": 9.984989998499001e-06, + "loss": 0.291, + "step": 15017 + }, + { + "epoch": 6.98186889818689, + "grad_norm": 4.224205017089844, + "learning_rate": 9.9849889984989e-06, + "loss": 0.0924, + "step": 15018 + }, + { + "epoch": 6.9823337982333795, + "grad_norm": 3.8096299171447754, + "learning_rate": 9.9849879984988e-06, + "loss": 0.0808, + "step": 15019 + }, + { + "epoch": 6.98279869827987, + "grad_norm": 10.758525848388672, + "learning_rate": 9.9849869984987e-06, + "loss": 0.331, + "step": 15020 + }, + { + "epoch": 6.98326359832636, + "grad_norm": 4.7001118659973145, + "learning_rate": 9.9849859984986e-06, + "loss": 0.0523, + "step": 15021 + }, + { + "epoch": 6.98372849837285, + "grad_norm": 3.9660451412200928, + "learning_rate": 9.984984998498501e-06, + "loss": 0.086, + "step": 15022 + }, + { + "epoch": 6.98419339841934, + "grad_norm": 7.35451078414917, + "learning_rate": 9.984983998498402e-06, + "loss": 0.2345, + "step": 15023 + }, + { + "epoch": 6.984658298465829, + "grad_norm": 7.396295547485352, + "learning_rate": 9.9849829984983e-06, + "loss": 0.2707, + "step": 15024 + }, + { + "epoch": 6.98512319851232, + "grad_norm": 4.664092540740967, + "learning_rate": 9.984981998498201e-06, + "loss": 0.1046, + "step": 15025 + }, + { + "epoch": 6.98558809855881, + "grad_norm": 8.045344352722168, + "learning_rate": 9.9849809984981e-06, + "loss": 0.2047, + "step": 15026 + }, + { + "epoch": 6.9860529986053, + "grad_norm": 5.560776710510254, + "learning_rate": 9.984979998498001e-06, + "loss": 0.1343, + "step": 15027 + }, + { + "epoch": 6.9865178986517895, + "grad_norm": 5.001460075378418, + "learning_rate": 9.9849789984979e-06, + "loss": 0.1681, + "step": 15028 + }, + { + "epoch": 6.98698279869828, + "grad_norm": 7.9483723640441895, + "learning_rate": 9.9849779984978e-06, + "loss": 0.1894, + "step": 15029 + }, + { + "epoch": 6.98744769874477, + "grad_norm": 4.175883769989014, + "learning_rate": 9.984976998497701e-06, + "loss": 0.0705, + "step": 15030 + }, + { + "epoch": 6.98791259879126, + "grad_norm": 4.842620372772217, + "learning_rate": 9.9849759984976e-06, + "loss": 0.1167, + "step": 15031 + }, + { + "epoch": 6.98837749883775, + "grad_norm": 7.156294822692871, + "learning_rate": 9.9849749984975e-06, + "loss": 0.2357, + "step": 15032 + }, + { + "epoch": 6.988842398884239, + "grad_norm": 6.646973133087158, + "learning_rate": 9.9849739984974e-06, + "loss": 0.0727, + "step": 15033 + }, + { + "epoch": 6.98930729893073, + "grad_norm": 5.791356563568115, + "learning_rate": 9.9849729984973e-06, + "loss": 0.1966, + "step": 15034 + }, + { + "epoch": 6.98977219897722, + "grad_norm": 7.455392837524414, + "learning_rate": 9.984971998497201e-06, + "loss": 0.2318, + "step": 15035 + }, + { + "epoch": 6.99023709902371, + "grad_norm": 8.957836151123047, + "learning_rate": 9.984970998497102e-06, + "loss": 0.1445, + "step": 15036 + }, + { + "epoch": 6.9907019990702, + "grad_norm": 4.812228679656982, + "learning_rate": 9.984969998497e-06, + "loss": 0.1028, + "step": 15037 + }, + { + "epoch": 6.99116689911669, + "grad_norm": 7.884998798370361, + "learning_rate": 9.984968998496901e-06, + "loss": 0.2836, + "step": 15038 + }, + { + "epoch": 6.99163179916318, + "grad_norm": 6.009337902069092, + "learning_rate": 9.9849679984968e-06, + "loss": 0.2957, + "step": 15039 + }, + { + "epoch": 6.99209669920967, + "grad_norm": 4.618271350860596, + "learning_rate": 9.9849669984967e-06, + "loss": 0.0941, + "step": 15040 + }, + { + "epoch": 6.99256159925616, + "grad_norm": 2.768420696258545, + "learning_rate": 9.9849659984966e-06, + "loss": 0.086, + "step": 15041 + }, + { + "epoch": 6.99302649930265, + "grad_norm": 5.201353549957275, + "learning_rate": 9.9849649984965e-06, + "loss": 0.1076, + "step": 15042 + }, + { + "epoch": 6.99349139934914, + "grad_norm": 4.522775650024414, + "learning_rate": 9.984963998496401e-06, + "loss": 0.0922, + "step": 15043 + }, + { + "epoch": 6.99395629939563, + "grad_norm": 5.4008355140686035, + "learning_rate": 9.9849629984963e-06, + "loss": 0.3337, + "step": 15044 + }, + { + "epoch": 6.99442119944212, + "grad_norm": 8.387330055236816, + "learning_rate": 9.984961998496201e-06, + "loss": 0.2289, + "step": 15045 + }, + { + "epoch": 6.99488609948861, + "grad_norm": 3.840718984603882, + "learning_rate": 9.9849609984961e-06, + "loss": 0.0808, + "step": 15046 + }, + { + "epoch": 6.9953509995351, + "grad_norm": 5.620176792144775, + "learning_rate": 9.984959998496e-06, + "loss": 0.113, + "step": 15047 + }, + { + "epoch": 6.99581589958159, + "grad_norm": 7.240722179412842, + "learning_rate": 9.984958998495901e-06, + "loss": 0.2037, + "step": 15048 + }, + { + "epoch": 6.99628079962808, + "grad_norm": 8.001570701599121, + "learning_rate": 9.984957998495802e-06, + "loss": 0.1103, + "step": 15049 + }, + { + "epoch": 6.99674569967457, + "grad_norm": 1.124372959136963, + "learning_rate": 9.9849569984957e-06, + "loss": 0.0157, + "step": 15050 + }, + { + "epoch": 6.99721059972106, + "grad_norm": 4.098062992095947, + "learning_rate": 9.9849559984956e-06, + "loss": 0.0896, + "step": 15051 + }, + { + "epoch": 6.99767549976755, + "grad_norm": 5.089489936828613, + "learning_rate": 9.9849549984955e-06, + "loss": 0.0404, + "step": 15052 + }, + { + "epoch": 6.99814039981404, + "grad_norm": 5.723287582397461, + "learning_rate": 9.984953998495401e-06, + "loss": 0.1319, + "step": 15053 + }, + { + "epoch": 6.99860529986053, + "grad_norm": 3.568081855773926, + "learning_rate": 9.9849529984953e-06, + "loss": 0.0664, + "step": 15054 + }, + { + "epoch": 6.99907019990702, + "grad_norm": 5.267127990722656, + "learning_rate": 9.9849519984952e-06, + "loss": 0.0959, + "step": 15055 + }, + { + "epoch": 6.99953509995351, + "grad_norm": 4.961790561676025, + "learning_rate": 9.984950998495101e-06, + "loss": 0.1669, + "step": 15056 + }, + { + "epoch": 7.0, + "grad_norm": 6.184147834777832, + "learning_rate": 9.984949998495e-06, + "loss": 0.1344, + "step": 15057 + }, + { + "epoch": 7.00046490004649, + "grad_norm": 2.1234724521636963, + "learning_rate": 9.984948998494901e-06, + "loss": 0.0186, + "step": 15058 + }, + { + "epoch": 7.00092980009298, + "grad_norm": 9.620197296142578, + "learning_rate": 9.9849479984948e-06, + "loss": 0.1463, + "step": 15059 + }, + { + "epoch": 7.00139470013947, + "grad_norm": 1.7374687194824219, + "learning_rate": 9.9849469984947e-06, + "loss": 0.0553, + "step": 15060 + }, + { + "epoch": 7.00185960018596, + "grad_norm": 4.855591773986816, + "learning_rate": 9.984945998494601e-06, + "loss": 0.1347, + "step": 15061 + }, + { + "epoch": 7.00232450023245, + "grad_norm": 4.050594329833984, + "learning_rate": 9.9849449984945e-06, + "loss": 0.1103, + "step": 15062 + }, + { + "epoch": 7.00278940027894, + "grad_norm": 3.5586090087890625, + "learning_rate": 9.984943998494401e-06, + "loss": 0.0969, + "step": 15063 + }, + { + "epoch": 7.00325430032543, + "grad_norm": 4.135092735290527, + "learning_rate": 9.9849429984943e-06, + "loss": 0.0843, + "step": 15064 + }, + { + "epoch": 7.00371920037192, + "grad_norm": 3.847898006439209, + "learning_rate": 9.9849419984942e-06, + "loss": 0.1271, + "step": 15065 + }, + { + "epoch": 7.00418410041841, + "grad_norm": 6.327724933624268, + "learning_rate": 9.984940998494101e-06, + "loss": 0.2691, + "step": 15066 + }, + { + "epoch": 7.0046490004649, + "grad_norm": 3.241726875305176, + "learning_rate": 9.984939998494e-06, + "loss": 0.0712, + "step": 15067 + }, + { + "epoch": 7.0051139005113905, + "grad_norm": 4.477230548858643, + "learning_rate": 9.9849389984939e-06, + "loss": 0.09, + "step": 15068 + }, + { + "epoch": 7.00557880055788, + "grad_norm": 3.2789571285247803, + "learning_rate": 9.984937998493801e-06, + "loss": 0.0785, + "step": 15069 + }, + { + "epoch": 7.00604370060437, + "grad_norm": 7.7469706535339355, + "learning_rate": 9.9849369984937e-06, + "loss": 0.1753, + "step": 15070 + }, + { + "epoch": 7.00650860065086, + "grad_norm": 6.690537929534912, + "learning_rate": 9.984935998493601e-06, + "loss": 0.2549, + "step": 15071 + }, + { + "epoch": 7.00697350069735, + "grad_norm": 6.949761390686035, + "learning_rate": 9.9849349984935e-06, + "loss": 0.2236, + "step": 15072 + }, + { + "epoch": 7.00743840074384, + "grad_norm": 5.054314136505127, + "learning_rate": 9.9849339984934e-06, + "loss": 0.1179, + "step": 15073 + }, + { + "epoch": 7.00790330079033, + "grad_norm": 4.1311116218566895, + "learning_rate": 9.984932998493301e-06, + "loss": 0.1139, + "step": 15074 + }, + { + "epoch": 7.00836820083682, + "grad_norm": 7.571715354919434, + "learning_rate": 9.9849319984932e-06, + "loss": 0.1694, + "step": 15075 + }, + { + "epoch": 7.00883310088331, + "grad_norm": 3.6934661865234375, + "learning_rate": 9.984930998493101e-06, + "loss": 0.1046, + "step": 15076 + }, + { + "epoch": 7.0092980009298005, + "grad_norm": 3.2497568130493164, + "learning_rate": 9.984929998493e-06, + "loss": 0.0599, + "step": 15077 + }, + { + "epoch": 7.00976290097629, + "grad_norm": 4.781828880310059, + "learning_rate": 9.9849289984929e-06, + "loss": 0.1136, + "step": 15078 + }, + { + "epoch": 7.01022780102278, + "grad_norm": 2.5042667388916016, + "learning_rate": 9.984927998492801e-06, + "loss": 0.0626, + "step": 15079 + }, + { + "epoch": 7.01069270106927, + "grad_norm": 4.501817226409912, + "learning_rate": 9.984926998492702e-06, + "loss": 0.1518, + "step": 15080 + }, + { + "epoch": 7.01115760111576, + "grad_norm": 2.4397568702697754, + "learning_rate": 9.984925998492601e-06, + "loss": 0.0517, + "step": 15081 + }, + { + "epoch": 7.01162250116225, + "grad_norm": 2.871791362762451, + "learning_rate": 9.984924998492502e-06, + "loss": 0.0519, + "step": 15082 + }, + { + "epoch": 7.01208740120874, + "grad_norm": 7.692547798156738, + "learning_rate": 9.9849239984924e-06, + "loss": 0.3082, + "step": 15083 + }, + { + "epoch": 7.01255230125523, + "grad_norm": 3.8233399391174316, + "learning_rate": 9.9849229984923e-06, + "loss": 0.1109, + "step": 15084 + }, + { + "epoch": 7.01301720130172, + "grad_norm": 4.0563483238220215, + "learning_rate": 9.9849219984922e-06, + "loss": 0.0617, + "step": 15085 + }, + { + "epoch": 7.0134821013482105, + "grad_norm": 6.262454509735107, + "learning_rate": 9.9849209984921e-06, + "loss": 0.217, + "step": 15086 + }, + { + "epoch": 7.0139470013947, + "grad_norm": 3.739910364151001, + "learning_rate": 9.984919998492001e-06, + "loss": 0.0621, + "step": 15087 + }, + { + "epoch": 7.01441190144119, + "grad_norm": 8.659883499145508, + "learning_rate": 9.9849189984919e-06, + "loss": 0.1969, + "step": 15088 + }, + { + "epoch": 7.01487680148768, + "grad_norm": 3.5512144565582275, + "learning_rate": 9.984917998491801e-06, + "loss": 0.0839, + "step": 15089 + }, + { + "epoch": 7.015341701534171, + "grad_norm": 4.3690972328186035, + "learning_rate": 9.9849169984917e-06, + "loss": 0.0993, + "step": 15090 + }, + { + "epoch": 7.01580660158066, + "grad_norm": 5.694579601287842, + "learning_rate": 9.9849159984916e-06, + "loss": 0.1202, + "step": 15091 + }, + { + "epoch": 7.01627150162715, + "grad_norm": 4.236285209655762, + "learning_rate": 9.984914998491501e-06, + "loss": 0.1035, + "step": 15092 + }, + { + "epoch": 7.01673640167364, + "grad_norm": 3.7834975719451904, + "learning_rate": 9.984913998491402e-06, + "loss": 0.0881, + "step": 15093 + }, + { + "epoch": 7.01720130172013, + "grad_norm": 6.945192813873291, + "learning_rate": 9.984912998491301e-06, + "loss": 0.1673, + "step": 15094 + }, + { + "epoch": 7.0176662017666205, + "grad_norm": 18.447982788085938, + "learning_rate": 9.9849119984912e-06, + "loss": 0.1514, + "step": 15095 + }, + { + "epoch": 7.01813110181311, + "grad_norm": 5.477786540985107, + "learning_rate": 9.9849109984911e-06, + "loss": 0.0994, + "step": 15096 + }, + { + "epoch": 7.0185960018596, + "grad_norm": 6.171140670776367, + "learning_rate": 9.984909998491e-06, + "loss": 0.1885, + "step": 15097 + }, + { + "epoch": 7.01906090190609, + "grad_norm": 5.768144130706787, + "learning_rate": 9.9849089984909e-06, + "loss": 0.1138, + "step": 15098 + }, + { + "epoch": 7.019525801952581, + "grad_norm": 6.76637601852417, + "learning_rate": 9.984907998490801e-06, + "loss": 0.2328, + "step": 15099 + }, + { + "epoch": 7.01999070199907, + "grad_norm": 4.670990943908691, + "learning_rate": 9.984906998490702e-06, + "loss": 0.1271, + "step": 15100 + }, + { + "epoch": 7.02045560204556, + "grad_norm": 3.8813788890838623, + "learning_rate": 9.9849059984906e-06, + "loss": 0.1155, + "step": 15101 + }, + { + "epoch": 7.02092050209205, + "grad_norm": 4.603567600250244, + "learning_rate": 9.984904998490501e-06, + "loss": 0.1731, + "step": 15102 + }, + { + "epoch": 7.02138540213854, + "grad_norm": 6.495639324188232, + "learning_rate": 9.9849039984904e-06, + "loss": 0.2781, + "step": 15103 + }, + { + "epoch": 7.0218503021850305, + "grad_norm": 5.307614326477051, + "learning_rate": 9.9849029984903e-06, + "loss": 0.1405, + "step": 15104 + }, + { + "epoch": 7.02231520223152, + "grad_norm": 3.9885947704315186, + "learning_rate": 9.984901998490201e-06, + "loss": 0.0871, + "step": 15105 + }, + { + "epoch": 7.02278010227801, + "grad_norm": 5.69614315032959, + "learning_rate": 9.9849009984901e-06, + "loss": 0.2049, + "step": 15106 + }, + { + "epoch": 7.0232450023245, + "grad_norm": 2.830786943435669, + "learning_rate": 9.984899998490001e-06, + "loss": 0.0534, + "step": 15107 + }, + { + "epoch": 7.023709902370991, + "grad_norm": 5.44850492477417, + "learning_rate": 9.9848989984899e-06, + "loss": 0.3048, + "step": 15108 + }, + { + "epoch": 7.02417480241748, + "grad_norm": 5.131076812744141, + "learning_rate": 9.9848979984898e-06, + "loss": 0.18, + "step": 15109 + }, + { + "epoch": 7.02463970246397, + "grad_norm": 5.80018424987793, + "learning_rate": 9.9848969984897e-06, + "loss": 0.1674, + "step": 15110 + }, + { + "epoch": 7.02510460251046, + "grad_norm": 3.3676187992095947, + "learning_rate": 9.9848959984896e-06, + "loss": 0.0693, + "step": 15111 + }, + { + "epoch": 7.025569502556951, + "grad_norm": 3.4456939697265625, + "learning_rate": 9.984894998489501e-06, + "loss": 0.0547, + "step": 15112 + }, + { + "epoch": 7.0260344026034405, + "grad_norm": 7.781586170196533, + "learning_rate": 9.984893998489402e-06, + "loss": 0.3886, + "step": 15113 + }, + { + "epoch": 7.02649930264993, + "grad_norm": 1.8719149827957153, + "learning_rate": 9.9848929984893e-06, + "loss": 0.029, + "step": 15114 + }, + { + "epoch": 7.02696420269642, + "grad_norm": 4.394139766693115, + "learning_rate": 9.984891998489201e-06, + "loss": 0.1338, + "step": 15115 + }, + { + "epoch": 7.02742910274291, + "grad_norm": 2.297044515609741, + "learning_rate": 9.9848909984891e-06, + "loss": 0.033, + "step": 15116 + }, + { + "epoch": 7.027894002789401, + "grad_norm": 2.4153943061828613, + "learning_rate": 9.984889998489e-06, + "loss": 0.0315, + "step": 15117 + }, + { + "epoch": 7.02835890283589, + "grad_norm": 5.656598091125488, + "learning_rate": 9.984888998488901e-06, + "loss": 0.1355, + "step": 15118 + }, + { + "epoch": 7.02882380288238, + "grad_norm": 6.032445430755615, + "learning_rate": 9.9848879984888e-06, + "loss": 0.1893, + "step": 15119 + }, + { + "epoch": 7.02928870292887, + "grad_norm": 3.927124500274658, + "learning_rate": 9.984886998488701e-06, + "loss": 0.0624, + "step": 15120 + }, + { + "epoch": 7.029753602975361, + "grad_norm": 4.434049606323242, + "learning_rate": 9.9848859984886e-06, + "loss": 0.1207, + "step": 15121 + }, + { + "epoch": 7.0302185030218505, + "grad_norm": 7.134222030639648, + "learning_rate": 9.9848849984885e-06, + "loss": 0.3218, + "step": 15122 + }, + { + "epoch": 7.03068340306834, + "grad_norm": 3.148205280303955, + "learning_rate": 9.9848839984884e-06, + "loss": 0.0564, + "step": 15123 + }, + { + "epoch": 7.03114830311483, + "grad_norm": 5.982363700866699, + "learning_rate": 9.9848829984883e-06, + "loss": 0.1133, + "step": 15124 + }, + { + "epoch": 7.03161320316132, + "grad_norm": 1.6008154153823853, + "learning_rate": 9.984881998488201e-06, + "loss": 0.0272, + "step": 15125 + }, + { + "epoch": 7.032078103207811, + "grad_norm": 9.798150062561035, + "learning_rate": 9.984880998488102e-06, + "loss": 0.2106, + "step": 15126 + }, + { + "epoch": 7.0325430032543, + "grad_norm": 3.896315097808838, + "learning_rate": 9.984879998488e-06, + "loss": 0.0682, + "step": 15127 + }, + { + "epoch": 7.03300790330079, + "grad_norm": 1.7531987428665161, + "learning_rate": 9.9848789984879e-06, + "loss": 0.0203, + "step": 15128 + }, + { + "epoch": 7.03347280334728, + "grad_norm": 1.4592359066009521, + "learning_rate": 9.9848779984878e-06, + "loss": 0.0243, + "step": 15129 + }, + { + "epoch": 7.033937703393771, + "grad_norm": 5.797436237335205, + "learning_rate": 9.984876998487701e-06, + "loss": 0.1309, + "step": 15130 + }, + { + "epoch": 7.0344026034402605, + "grad_norm": 3.3103630542755127, + "learning_rate": 9.984875998487602e-06, + "loss": 0.0401, + "step": 15131 + }, + { + "epoch": 7.03486750348675, + "grad_norm": 6.032726764678955, + "learning_rate": 9.9848749984875e-06, + "loss": 0.1749, + "step": 15132 + }, + { + "epoch": 7.03533240353324, + "grad_norm": 5.879261016845703, + "learning_rate": 9.984873998487401e-06, + "loss": 0.067, + "step": 15133 + }, + { + "epoch": 7.035797303579731, + "grad_norm": 2.9709911346435547, + "learning_rate": 9.9848729984873e-06, + "loss": 0.0519, + "step": 15134 + }, + { + "epoch": 7.036262203626221, + "grad_norm": 4.993862152099609, + "learning_rate": 9.9848719984872e-06, + "loss": 0.1342, + "step": 15135 + }, + { + "epoch": 7.03672710367271, + "grad_norm": 3.301018238067627, + "learning_rate": 9.9848709984871e-06, + "loss": 0.065, + "step": 15136 + }, + { + "epoch": 7.0371920037192, + "grad_norm": 4.623376369476318, + "learning_rate": 9.984869998487e-06, + "loss": 0.1151, + "step": 15137 + }, + { + "epoch": 7.03765690376569, + "grad_norm": 5.306723117828369, + "learning_rate": 9.984868998486901e-06, + "loss": 0.1769, + "step": 15138 + }, + { + "epoch": 7.038121803812181, + "grad_norm": 7.019335746765137, + "learning_rate": 9.9848679984868e-06, + "loss": 0.1561, + "step": 15139 + }, + { + "epoch": 7.0385867038586705, + "grad_norm": 4.935449600219727, + "learning_rate": 9.9848669984867e-06, + "loss": 0.1456, + "step": 15140 + }, + { + "epoch": 7.03905160390516, + "grad_norm": 5.2246832847595215, + "learning_rate": 9.9848659984866e-06, + "loss": 0.1104, + "step": 15141 + }, + { + "epoch": 7.03951650395165, + "grad_norm": 6.967220306396484, + "learning_rate": 9.9848649984865e-06, + "loss": 0.2185, + "step": 15142 + }, + { + "epoch": 7.039981403998141, + "grad_norm": 4.499705791473389, + "learning_rate": 9.984863998486401e-06, + "loss": 0.1079, + "step": 15143 + }, + { + "epoch": 7.040446304044631, + "grad_norm": 4.320911407470703, + "learning_rate": 9.984862998486302e-06, + "loss": 0.1674, + "step": 15144 + }, + { + "epoch": 7.04091120409112, + "grad_norm": 7.407101631164551, + "learning_rate": 9.9848619984862e-06, + "loss": 0.1108, + "step": 15145 + }, + { + "epoch": 7.04137610413761, + "grad_norm": 5.912492275238037, + "learning_rate": 9.984860998486101e-06, + "loss": 0.2607, + "step": 15146 + }, + { + "epoch": 7.0418410041841, + "grad_norm": 5.905815124511719, + "learning_rate": 9.984859998486e-06, + "loss": 0.1823, + "step": 15147 + }, + { + "epoch": 7.042305904230591, + "grad_norm": 4.997721195220947, + "learning_rate": 9.984858998485901e-06, + "loss": 0.148, + "step": 15148 + }, + { + "epoch": 7.0427708042770805, + "grad_norm": 4.3465471267700195, + "learning_rate": 9.984857998485802e-06, + "loss": 0.1392, + "step": 15149 + }, + { + "epoch": 7.04323570432357, + "grad_norm": 4.604761600494385, + "learning_rate": 9.9848569984857e-06, + "loss": 0.099, + "step": 15150 + }, + { + "epoch": 7.04370060437006, + "grad_norm": 4.728597640991211, + "learning_rate": 9.984855998485601e-06, + "loss": 0.144, + "step": 15151 + }, + { + "epoch": 7.044165504416551, + "grad_norm": 5.038106441497803, + "learning_rate": 9.9848549984855e-06, + "loss": 0.1048, + "step": 15152 + }, + { + "epoch": 7.044630404463041, + "grad_norm": 6.260807514190674, + "learning_rate": 9.9848539984854e-06, + "loss": 0.1585, + "step": 15153 + }, + { + "epoch": 7.04509530450953, + "grad_norm": 5.054345607757568, + "learning_rate": 9.9848529984853e-06, + "loss": 0.0857, + "step": 15154 + }, + { + "epoch": 7.04556020455602, + "grad_norm": 4.086819171905518, + "learning_rate": 9.9848519984852e-06, + "loss": 0.0656, + "step": 15155 + }, + { + "epoch": 7.046025104602511, + "grad_norm": 2.695497989654541, + "learning_rate": 9.984850998485101e-06, + "loss": 0.0493, + "step": 15156 + }, + { + "epoch": 7.046490004649001, + "grad_norm": 8.407620429992676, + "learning_rate": 9.984849998485002e-06, + "loss": 0.1401, + "step": 15157 + }, + { + "epoch": 7.0469549046954905, + "grad_norm": 5.094388008117676, + "learning_rate": 9.9848489984849e-06, + "loss": 0.0724, + "step": 15158 + }, + { + "epoch": 7.04741980474198, + "grad_norm": 7.473015785217285, + "learning_rate": 9.9848479984848e-06, + "loss": 0.2558, + "step": 15159 + }, + { + "epoch": 7.04788470478847, + "grad_norm": 6.463756561279297, + "learning_rate": 9.9848469984847e-06, + "loss": 0.1421, + "step": 15160 + }, + { + "epoch": 7.048349604834961, + "grad_norm": 4.162755012512207, + "learning_rate": 9.984845998484601e-06, + "loss": 0.179, + "step": 15161 + }, + { + "epoch": 7.048814504881451, + "grad_norm": 1.989263653755188, + "learning_rate": 9.984844998484502e-06, + "loss": 0.0242, + "step": 15162 + }, + { + "epoch": 7.04927940492794, + "grad_norm": 2.4505198001861572, + "learning_rate": 9.9848439984844e-06, + "loss": 0.0432, + "step": 15163 + }, + { + "epoch": 7.04974430497443, + "grad_norm": 3.3077774047851562, + "learning_rate": 9.984842998484301e-06, + "loss": 0.1397, + "step": 15164 + }, + { + "epoch": 7.050209205020921, + "grad_norm": 5.867178916931152, + "learning_rate": 9.9848419984842e-06, + "loss": 0.2835, + "step": 15165 + }, + { + "epoch": 7.050674105067411, + "grad_norm": 7.634551048278809, + "learning_rate": 9.984840998484101e-06, + "loss": 0.2798, + "step": 15166 + }, + { + "epoch": 7.0511390051139005, + "grad_norm": 3.062424421310425, + "learning_rate": 9.984839998484e-06, + "loss": 0.059, + "step": 15167 + }, + { + "epoch": 7.05160390516039, + "grad_norm": 4.385312080383301, + "learning_rate": 9.9848389984839e-06, + "loss": 0.0704, + "step": 15168 + }, + { + "epoch": 7.052068805206881, + "grad_norm": 4.450347900390625, + "learning_rate": 9.984837998483801e-06, + "loss": 0.1381, + "step": 15169 + }, + { + "epoch": 7.052533705253371, + "grad_norm": 4.05956506729126, + "learning_rate": 9.9848369984837e-06, + "loss": 0.1049, + "step": 15170 + }, + { + "epoch": 7.052998605299861, + "grad_norm": 1.8716363906860352, + "learning_rate": 9.9848359984836e-06, + "loss": 0.0214, + "step": 15171 + }, + { + "epoch": 7.05346350534635, + "grad_norm": 5.0835347175598145, + "learning_rate": 9.9848349984835e-06, + "loss": 0.1295, + "step": 15172 + }, + { + "epoch": 7.05392840539284, + "grad_norm": 1.9151802062988281, + "learning_rate": 9.9848339984834e-06, + "loss": 0.0313, + "step": 15173 + }, + { + "epoch": 7.054393305439331, + "grad_norm": 5.4015021324157715, + "learning_rate": 9.984832998483301e-06, + "loss": 0.081, + "step": 15174 + }, + { + "epoch": 7.054858205485821, + "grad_norm": 3.605656623840332, + "learning_rate": 9.984831998483202e-06, + "loss": 0.0887, + "step": 15175 + }, + { + "epoch": 7.0553231055323105, + "grad_norm": 5.8553571701049805, + "learning_rate": 9.9848309984831e-06, + "loss": 0.0684, + "step": 15176 + }, + { + "epoch": 7.0557880055788, + "grad_norm": 2.3059308528900146, + "learning_rate": 9.984829998483001e-06, + "loss": 0.0235, + "step": 15177 + }, + { + "epoch": 7.056252905625291, + "grad_norm": 8.996681213378906, + "learning_rate": 9.9848289984829e-06, + "loss": 0.2275, + "step": 15178 + }, + { + "epoch": 7.056717805671781, + "grad_norm": 6.401684761047363, + "learning_rate": 9.984827998482801e-06, + "loss": 0.1745, + "step": 15179 + }, + { + "epoch": 7.0571827057182706, + "grad_norm": 3.7237324714660645, + "learning_rate": 9.9848269984827e-06, + "loss": 0.0921, + "step": 15180 + }, + { + "epoch": 7.05764760576476, + "grad_norm": 8.714098930358887, + "learning_rate": 9.9848259984826e-06, + "loss": 0.2679, + "step": 15181 + }, + { + "epoch": 7.05811250581125, + "grad_norm": 9.535400390625, + "learning_rate": 9.984824998482501e-06, + "loss": 0.2203, + "step": 15182 + }, + { + "epoch": 7.058577405857741, + "grad_norm": 5.388853073120117, + "learning_rate": 9.9848239984824e-06, + "loss": 0.1847, + "step": 15183 + }, + { + "epoch": 7.059042305904231, + "grad_norm": 2.5789670944213867, + "learning_rate": 9.984822998482301e-06, + "loss": 0.0487, + "step": 15184 + }, + { + "epoch": 7.0595072059507205, + "grad_norm": 6.9602251052856445, + "learning_rate": 9.9848219984822e-06, + "loss": 0.1431, + "step": 15185 + }, + { + "epoch": 7.05997210599721, + "grad_norm": 6.713922023773193, + "learning_rate": 9.9848209984821e-06, + "loss": 0.162, + "step": 15186 + }, + { + "epoch": 7.060437006043701, + "grad_norm": 3.141590118408203, + "learning_rate": 9.984819998482001e-06, + "loss": 0.0933, + "step": 15187 + }, + { + "epoch": 7.060901906090191, + "grad_norm": 4.964783191680908, + "learning_rate": 9.984818998481902e-06, + "loss": 0.1235, + "step": 15188 + }, + { + "epoch": 7.0613668061366806, + "grad_norm": 4.809986591339111, + "learning_rate": 9.9848179984818e-06, + "loss": 0.0996, + "step": 15189 + }, + { + "epoch": 7.06183170618317, + "grad_norm": 4.063107013702393, + "learning_rate": 9.984816998481701e-06, + "loss": 0.1488, + "step": 15190 + }, + { + "epoch": 7.062296606229661, + "grad_norm": 5.816638946533203, + "learning_rate": 9.9848159984816e-06, + "loss": 0.2158, + "step": 15191 + }, + { + "epoch": 7.062761506276151, + "grad_norm": 6.6648454666137695, + "learning_rate": 9.9848149984815e-06, + "loss": 0.1737, + "step": 15192 + }, + { + "epoch": 7.063226406322641, + "grad_norm": 2.0903728008270264, + "learning_rate": 9.9848139984814e-06, + "loss": 0.0323, + "step": 15193 + }, + { + "epoch": 7.0636913063691305, + "grad_norm": 20.83365821838379, + "learning_rate": 9.9848129984813e-06, + "loss": 0.1703, + "step": 15194 + }, + { + "epoch": 7.06415620641562, + "grad_norm": 3.349720001220703, + "learning_rate": 9.984811998481201e-06, + "loss": 0.0683, + "step": 15195 + }, + { + "epoch": 7.064621106462111, + "grad_norm": 6.847404479980469, + "learning_rate": 9.9848109984811e-06, + "loss": 0.1576, + "step": 15196 + }, + { + "epoch": 7.065086006508601, + "grad_norm": 5.848873615264893, + "learning_rate": 9.984809998481001e-06, + "loss": 0.0903, + "step": 15197 + }, + { + "epoch": 7.0655509065550905, + "grad_norm": 4.069166660308838, + "learning_rate": 9.9848089984809e-06, + "loss": 0.1041, + "step": 15198 + }, + { + "epoch": 7.06601580660158, + "grad_norm": 7.119555473327637, + "learning_rate": 9.9848079984808e-06, + "loss": 0.2316, + "step": 15199 + }, + { + "epoch": 7.066480706648071, + "grad_norm": 4.299427032470703, + "learning_rate": 9.984806998480701e-06, + "loss": 0.0823, + "step": 15200 + }, + { + "epoch": 7.066945606694561, + "grad_norm": 4.727161884307861, + "learning_rate": 9.984805998480602e-06, + "loss": 0.1017, + "step": 15201 + }, + { + "epoch": 7.067410506741051, + "grad_norm": 11.769806861877441, + "learning_rate": 9.9848049984805e-06, + "loss": 0.3679, + "step": 15202 + }, + { + "epoch": 7.0678754067875404, + "grad_norm": 4.0123796463012695, + "learning_rate": 9.9848039984804e-06, + "loss": 0.0473, + "step": 15203 + }, + { + "epoch": 7.06834030683403, + "grad_norm": 4.0537614822387695, + "learning_rate": 9.9848029984803e-06, + "loss": 0.0886, + "step": 15204 + }, + { + "epoch": 7.068805206880521, + "grad_norm": 4.310972213745117, + "learning_rate": 9.9848019984802e-06, + "loss": 0.1053, + "step": 15205 + }, + { + "epoch": 7.069270106927011, + "grad_norm": 5.682461261749268, + "learning_rate": 9.9848009984801e-06, + "loss": 0.2135, + "step": 15206 + }, + { + "epoch": 7.0697350069735005, + "grad_norm": 10.506637573242188, + "learning_rate": 9.98479999848e-06, + "loss": 0.2246, + "step": 15207 + }, + { + "epoch": 7.07019990701999, + "grad_norm": 8.308219909667969, + "learning_rate": 9.984798998479901e-06, + "loss": 0.0466, + "step": 15208 + }, + { + "epoch": 7.070664807066481, + "grad_norm": 11.000743865966797, + "learning_rate": 9.9847979984798e-06, + "loss": 0.1792, + "step": 15209 + }, + { + "epoch": 7.071129707112971, + "grad_norm": 8.638607025146484, + "learning_rate": 9.984796998479701e-06, + "loss": 0.1175, + "step": 15210 + }, + { + "epoch": 7.071594607159461, + "grad_norm": 7.010780334472656, + "learning_rate": 9.9847959984796e-06, + "loss": 0.4212, + "step": 15211 + }, + { + "epoch": 7.0720595072059504, + "grad_norm": 7.821456432342529, + "learning_rate": 9.9847949984795e-06, + "loss": 0.2936, + "step": 15212 + }, + { + "epoch": 7.072524407252441, + "grad_norm": 5.402672290802002, + "learning_rate": 9.984793998479401e-06, + "loss": 0.183, + "step": 15213 + }, + { + "epoch": 7.072989307298931, + "grad_norm": 6.663821220397949, + "learning_rate": 9.9847929984793e-06, + "loss": 0.0912, + "step": 15214 + }, + { + "epoch": 7.073454207345421, + "grad_norm": 4.943943977355957, + "learning_rate": 9.984791998479201e-06, + "loss": 0.0874, + "step": 15215 + }, + { + "epoch": 7.0739191073919105, + "grad_norm": 4.555756092071533, + "learning_rate": 9.9847909984791e-06, + "loss": 0.0928, + "step": 15216 + }, + { + "epoch": 7.0743840074384, + "grad_norm": 4.181107044219971, + "learning_rate": 9.984789998479e-06, + "loss": 0.0858, + "step": 15217 + }, + { + "epoch": 7.074848907484891, + "grad_norm": 3.0004546642303467, + "learning_rate": 9.984788998478901e-06, + "loss": 0.0398, + "step": 15218 + }, + { + "epoch": 7.075313807531381, + "grad_norm": 4.180063724517822, + "learning_rate": 9.9847879984788e-06, + "loss": 0.1032, + "step": 15219 + }, + { + "epoch": 7.075778707577871, + "grad_norm": 4.498480796813965, + "learning_rate": 9.9847869984787e-06, + "loss": 0.0801, + "step": 15220 + }, + { + "epoch": 7.07624360762436, + "grad_norm": 5.362067699432373, + "learning_rate": 9.984785998478601e-06, + "loss": 0.1678, + "step": 15221 + }, + { + "epoch": 7.076708507670851, + "grad_norm": 6.878396987915039, + "learning_rate": 9.9847849984785e-06, + "loss": 0.2136, + "step": 15222 + }, + { + "epoch": 7.077173407717341, + "grad_norm": 4.003575325012207, + "learning_rate": 9.984783998478401e-06, + "loss": 0.075, + "step": 15223 + }, + { + "epoch": 7.077638307763831, + "grad_norm": 4.091615200042725, + "learning_rate": 9.9847829984783e-06, + "loss": 0.067, + "step": 15224 + }, + { + "epoch": 7.0781032078103205, + "grad_norm": 4.01171875, + "learning_rate": 9.9847819984782e-06, + "loss": 0.0674, + "step": 15225 + }, + { + "epoch": 7.07856810785681, + "grad_norm": 6.214353084564209, + "learning_rate": 9.984780998478101e-06, + "loss": 0.1831, + "step": 15226 + }, + { + "epoch": 7.079033007903301, + "grad_norm": 10.403491973876953, + "learning_rate": 9.984779998478e-06, + "loss": 0.2118, + "step": 15227 + }, + { + "epoch": 7.079497907949791, + "grad_norm": 5.059696674346924, + "learning_rate": 9.984778998477901e-06, + "loss": 0.1193, + "step": 15228 + }, + { + "epoch": 7.079962807996281, + "grad_norm": 3.297433853149414, + "learning_rate": 9.9847779984778e-06, + "loss": 0.0769, + "step": 15229 + }, + { + "epoch": 7.08042770804277, + "grad_norm": 3.649963855743408, + "learning_rate": 9.9847769984777e-06, + "loss": 0.0684, + "step": 15230 + }, + { + "epoch": 7.080892608089261, + "grad_norm": 6.168162822723389, + "learning_rate": 9.984775998477601e-06, + "loss": 0.1949, + "step": 15231 + }, + { + "epoch": 7.081357508135751, + "grad_norm": 5.121642112731934, + "learning_rate": 9.9847749984775e-06, + "loss": 0.1745, + "step": 15232 + }, + { + "epoch": 7.081822408182241, + "grad_norm": 6.4058942794799805, + "learning_rate": 9.984773998477401e-06, + "loss": 0.2254, + "step": 15233 + }, + { + "epoch": 7.0822873082287305, + "grad_norm": 4.78471040725708, + "learning_rate": 9.984772998477302e-06, + "loss": 0.0336, + "step": 15234 + }, + { + "epoch": 7.082752208275221, + "grad_norm": 3.5127668380737305, + "learning_rate": 9.9847719984772e-06, + "loss": 0.0589, + "step": 15235 + }, + { + "epoch": 7.083217108321711, + "grad_norm": 2.223191976547241, + "learning_rate": 9.9847709984771e-06, + "loss": 0.018, + "step": 15236 + }, + { + "epoch": 7.083682008368201, + "grad_norm": 3.385605573654175, + "learning_rate": 9.984769998477e-06, + "loss": 0.0672, + "step": 15237 + }, + { + "epoch": 7.084146908414691, + "grad_norm": 3.3843884468078613, + "learning_rate": 9.9847689984769e-06, + "loss": 0.0595, + "step": 15238 + }, + { + "epoch": 7.08461180846118, + "grad_norm": 4.43869686126709, + "learning_rate": 9.984767998476801e-06, + "loss": 0.0748, + "step": 15239 + }, + { + "epoch": 7.085076708507671, + "grad_norm": 2.3924379348754883, + "learning_rate": 9.9847669984767e-06, + "loss": 0.0519, + "step": 15240 + }, + { + "epoch": 7.085541608554161, + "grad_norm": 4.0632004737854, + "learning_rate": 9.984765998476601e-06, + "loss": 0.1193, + "step": 15241 + }, + { + "epoch": 7.086006508600651, + "grad_norm": 5.163794994354248, + "learning_rate": 9.9847649984765e-06, + "loss": 0.1081, + "step": 15242 + }, + { + "epoch": 7.0864714086471405, + "grad_norm": 7.792972564697266, + "learning_rate": 9.9847639984764e-06, + "loss": 0.2842, + "step": 15243 + }, + { + "epoch": 7.086936308693631, + "grad_norm": 3.9043185710906982, + "learning_rate": 9.984762998476301e-06, + "loss": 0.0421, + "step": 15244 + }, + { + "epoch": 7.087401208740121, + "grad_norm": 7.056829452514648, + "learning_rate": 9.984761998476202e-06, + "loss": 0.2984, + "step": 15245 + }, + { + "epoch": 7.087866108786611, + "grad_norm": 2.1289174556732178, + "learning_rate": 9.984760998476101e-06, + "loss": 0.033, + "step": 15246 + }, + { + "epoch": 7.088331008833101, + "grad_norm": 5.842343330383301, + "learning_rate": 9.984759998476e-06, + "loss": 0.1433, + "step": 15247 + }, + { + "epoch": 7.088795908879591, + "grad_norm": 2.910512924194336, + "learning_rate": 9.9847589984759e-06, + "loss": 0.0809, + "step": 15248 + }, + { + "epoch": 7.089260808926081, + "grad_norm": 4.545862674713135, + "learning_rate": 9.9847579984758e-06, + "loss": 0.0831, + "step": 15249 + }, + { + "epoch": 7.089725708972571, + "grad_norm": 7.898166656494141, + "learning_rate": 9.9847569984757e-06, + "loss": 0.1381, + "step": 15250 + }, + { + "epoch": 7.090190609019061, + "grad_norm": 3.946974515914917, + "learning_rate": 9.984755998475601e-06, + "loss": 0.0904, + "step": 15251 + }, + { + "epoch": 7.0906555090655505, + "grad_norm": 5.84512996673584, + "learning_rate": 9.984754998475502e-06, + "loss": 0.141, + "step": 15252 + }, + { + "epoch": 7.091120409112041, + "grad_norm": 5.658949851989746, + "learning_rate": 9.9847539984754e-06, + "loss": 0.2537, + "step": 15253 + }, + { + "epoch": 7.091585309158531, + "grad_norm": 5.683025360107422, + "learning_rate": 9.984752998475301e-06, + "loss": 0.0864, + "step": 15254 + }, + { + "epoch": 7.092050209205021, + "grad_norm": 8.083486557006836, + "learning_rate": 9.9847519984752e-06, + "loss": 0.1511, + "step": 15255 + }, + { + "epoch": 7.092515109251511, + "grad_norm": 4.509425163269043, + "learning_rate": 9.9847509984751e-06, + "loss": 0.0639, + "step": 15256 + }, + { + "epoch": 7.092980009298001, + "grad_norm": 4.123547554016113, + "learning_rate": 9.984749998475001e-06, + "loss": 0.1238, + "step": 15257 + }, + { + "epoch": 7.093444909344491, + "grad_norm": 4.694133758544922, + "learning_rate": 9.9847489984749e-06, + "loss": 0.128, + "step": 15258 + }, + { + "epoch": 7.093909809390981, + "grad_norm": 2.6067287921905518, + "learning_rate": 9.984747998474801e-06, + "loss": 0.0314, + "step": 15259 + }, + { + "epoch": 7.094374709437471, + "grad_norm": 4.657804489135742, + "learning_rate": 9.9847469984747e-06, + "loss": 0.1446, + "step": 15260 + }, + { + "epoch": 7.0948396094839605, + "grad_norm": 4.8183488845825195, + "learning_rate": 9.9847459984746e-06, + "loss": 0.0929, + "step": 15261 + }, + { + "epoch": 7.095304509530451, + "grad_norm": 2.8655202388763428, + "learning_rate": 9.9847449984745e-06, + "loss": 0.0361, + "step": 15262 + }, + { + "epoch": 7.095769409576941, + "grad_norm": 3.109792470932007, + "learning_rate": 9.9847439984744e-06, + "loss": 0.0515, + "step": 15263 + }, + { + "epoch": 7.096234309623431, + "grad_norm": 4.599428176879883, + "learning_rate": 9.984742998474301e-06, + "loss": 0.1149, + "step": 15264 + }, + { + "epoch": 7.096699209669921, + "grad_norm": 4.765151023864746, + "learning_rate": 9.984741998474202e-06, + "loss": 0.1, + "step": 15265 + }, + { + "epoch": 7.097164109716411, + "grad_norm": 2.628976583480835, + "learning_rate": 9.9847409984741e-06, + "loss": 0.0381, + "step": 15266 + }, + { + "epoch": 7.097629009762901, + "grad_norm": 4.6153717041015625, + "learning_rate": 9.984739998474001e-06, + "loss": 0.1264, + "step": 15267 + }, + { + "epoch": 7.098093909809391, + "grad_norm": 6.359836101531982, + "learning_rate": 9.9847389984739e-06, + "loss": 0.2398, + "step": 15268 + }, + { + "epoch": 7.098558809855881, + "grad_norm": 4.128632068634033, + "learning_rate": 9.984737998473801e-06, + "loss": 0.0588, + "step": 15269 + }, + { + "epoch": 7.099023709902371, + "grad_norm": 4.438507556915283, + "learning_rate": 9.984736998473702e-06, + "loss": 0.0902, + "step": 15270 + }, + { + "epoch": 7.099488609948861, + "grad_norm": 8.365885734558105, + "learning_rate": 9.9847359984736e-06, + "loss": 0.2412, + "step": 15271 + }, + { + "epoch": 7.099953509995351, + "grad_norm": 3.731052875518799, + "learning_rate": 9.984734998473501e-06, + "loss": 0.087, + "step": 15272 + }, + { + "epoch": 7.100418410041841, + "grad_norm": 6.223088264465332, + "learning_rate": 9.9847339984734e-06, + "loss": 0.1458, + "step": 15273 + }, + { + "epoch": 7.100883310088331, + "grad_norm": 2.4209585189819336, + "learning_rate": 9.9847329984733e-06, + "loss": 0.0332, + "step": 15274 + }, + { + "epoch": 7.101348210134821, + "grad_norm": 5.439640998840332, + "learning_rate": 9.9847319984732e-06, + "loss": 0.1137, + "step": 15275 + }, + { + "epoch": 7.101813110181311, + "grad_norm": 4.256399154663086, + "learning_rate": 9.9847309984731e-06, + "loss": 0.1154, + "step": 15276 + }, + { + "epoch": 7.102278010227801, + "grad_norm": 4.079805374145508, + "learning_rate": 9.984729998473001e-06, + "loss": 0.0825, + "step": 15277 + }, + { + "epoch": 7.102742910274291, + "grad_norm": 7.792973041534424, + "learning_rate": 9.9847289984729e-06, + "loss": 0.2721, + "step": 15278 + }, + { + "epoch": 7.103207810320781, + "grad_norm": 3.2899587154388428, + "learning_rate": 9.9847279984728e-06, + "loss": 0.0664, + "step": 15279 + }, + { + "epoch": 7.103672710367271, + "grad_norm": 3.118016481399536, + "learning_rate": 9.9847269984727e-06, + "loss": 0.0847, + "step": 15280 + }, + { + "epoch": 7.104137610413761, + "grad_norm": 5.612105846405029, + "learning_rate": 9.9847259984726e-06, + "loss": 0.1294, + "step": 15281 + }, + { + "epoch": 7.104602510460251, + "grad_norm": 2.61818528175354, + "learning_rate": 9.984724998472501e-06, + "loss": 0.0419, + "step": 15282 + }, + { + "epoch": 7.105067410506741, + "grad_norm": 5.091592311859131, + "learning_rate": 9.984723998472402e-06, + "loss": 0.0971, + "step": 15283 + }, + { + "epoch": 7.105532310553231, + "grad_norm": 4.3064703941345215, + "learning_rate": 9.9847229984723e-06, + "loss": 0.1018, + "step": 15284 + }, + { + "epoch": 7.105997210599721, + "grad_norm": 7.07798433303833, + "learning_rate": 9.984721998472201e-06, + "loss": 0.1035, + "step": 15285 + }, + { + "epoch": 7.106462110646211, + "grad_norm": 6.037516117095947, + "learning_rate": 9.9847209984721e-06, + "loss": 0.0865, + "step": 15286 + }, + { + "epoch": 7.106927010692701, + "grad_norm": 4.859367370605469, + "learning_rate": 9.984719998472001e-06, + "loss": 0.1185, + "step": 15287 + }, + { + "epoch": 7.107391910739191, + "grad_norm": 6.23228645324707, + "learning_rate": 9.9847189984719e-06, + "loss": 0.295, + "step": 15288 + }, + { + "epoch": 7.107856810785681, + "grad_norm": 5.289906024932861, + "learning_rate": 9.9847179984718e-06, + "loss": 0.138, + "step": 15289 + }, + { + "epoch": 7.108321710832171, + "grad_norm": 6.709517478942871, + "learning_rate": 9.984716998471701e-06, + "loss": 0.1593, + "step": 15290 + }, + { + "epoch": 7.108786610878661, + "grad_norm": 4.0082783699035645, + "learning_rate": 9.9847159984716e-06, + "loss": 0.0705, + "step": 15291 + }, + { + "epoch": 7.1092515109251515, + "grad_norm": 4.381811141967773, + "learning_rate": 9.9847149984715e-06, + "loss": 0.0489, + "step": 15292 + }, + { + "epoch": 7.109716410971641, + "grad_norm": 4.513078212738037, + "learning_rate": 9.9847139984714e-06, + "loss": 0.0954, + "step": 15293 + }, + { + "epoch": 7.110181311018131, + "grad_norm": 4.251951694488525, + "learning_rate": 9.9847129984713e-06, + "loss": 0.0534, + "step": 15294 + }, + { + "epoch": 7.110646211064621, + "grad_norm": 5.355739593505859, + "learning_rate": 9.984711998471201e-06, + "loss": 0.1171, + "step": 15295 + }, + { + "epoch": 7.111111111111111, + "grad_norm": 7.077868461608887, + "learning_rate": 9.984710998471102e-06, + "loss": 0.1885, + "step": 15296 + }, + { + "epoch": 7.111576011157601, + "grad_norm": 3.74743914604187, + "learning_rate": 9.984709998471e-06, + "loss": 0.0882, + "step": 15297 + }, + { + "epoch": 7.112040911204091, + "grad_norm": 4.716919898986816, + "learning_rate": 9.984708998470901e-06, + "loss": 0.0997, + "step": 15298 + }, + { + "epoch": 7.112505811250581, + "grad_norm": 6.422705173492432, + "learning_rate": 9.9847079984708e-06, + "loss": 0.26, + "step": 15299 + }, + { + "epoch": 7.112970711297071, + "grad_norm": 2.884556770324707, + "learning_rate": 9.984706998470701e-06, + "loss": 0.0458, + "step": 15300 + }, + { + "epoch": 7.1134356113435615, + "grad_norm": 4.4274444580078125, + "learning_rate": 9.9847059984706e-06, + "loss": 0.1094, + "step": 15301 + }, + { + "epoch": 7.113900511390051, + "grad_norm": 3.792747735977173, + "learning_rate": 9.9847049984705e-06, + "loss": 0.0655, + "step": 15302 + }, + { + "epoch": 7.114365411436541, + "grad_norm": 6.1721038818359375, + "learning_rate": 9.984703998470401e-06, + "loss": 0.1777, + "step": 15303 + }, + { + "epoch": 7.114830311483031, + "grad_norm": 5.669679164886475, + "learning_rate": 9.9847029984703e-06, + "loss": 0.1726, + "step": 15304 + }, + { + "epoch": 7.115295211529521, + "grad_norm": 5.744257926940918, + "learning_rate": 9.9847019984702e-06, + "loss": 0.1444, + "step": 15305 + }, + { + "epoch": 7.115760111576011, + "grad_norm": 6.817150115966797, + "learning_rate": 9.9847009984701e-06, + "loss": 0.2249, + "step": 15306 + }, + { + "epoch": 7.116225011622501, + "grad_norm": 7.298929691314697, + "learning_rate": 9.98469999847e-06, + "loss": 0.1107, + "step": 15307 + }, + { + "epoch": 7.116689911668991, + "grad_norm": 3.9255058765411377, + "learning_rate": 9.984698998469901e-06, + "loss": 0.088, + "step": 15308 + }, + { + "epoch": 7.117154811715481, + "grad_norm": 3.6033859252929688, + "learning_rate": 9.984697998469802e-06, + "loss": 0.0653, + "step": 15309 + }, + { + "epoch": 7.1176197117619715, + "grad_norm": 9.006181716918945, + "learning_rate": 9.9846969984697e-06, + "loss": 0.3079, + "step": 15310 + }, + { + "epoch": 7.118084611808461, + "grad_norm": 5.569603443145752, + "learning_rate": 9.9846959984696e-06, + "loss": 0.1351, + "step": 15311 + }, + { + "epoch": 7.118549511854951, + "grad_norm": 6.655263423919678, + "learning_rate": 9.9846949984695e-06, + "loss": 0.1528, + "step": 15312 + }, + { + "epoch": 7.119014411901441, + "grad_norm": 2.1290042400360107, + "learning_rate": 9.984693998469401e-06, + "loss": 0.0279, + "step": 15313 + }, + { + "epoch": 7.119479311947932, + "grad_norm": 5.4520649909973145, + "learning_rate": 9.984692998469302e-06, + "loss": 0.1554, + "step": 15314 + }, + { + "epoch": 7.119944211994421, + "grad_norm": 4.90014123916626, + "learning_rate": 9.9846919984692e-06, + "loss": 0.1384, + "step": 15315 + }, + { + "epoch": 7.120409112040911, + "grad_norm": 6.752671241760254, + "learning_rate": 9.984690998469101e-06, + "loss": 0.1762, + "step": 15316 + }, + { + "epoch": 7.120874012087401, + "grad_norm": 4.844414710998535, + "learning_rate": 9.984689998469e-06, + "loss": 0.1524, + "step": 15317 + }, + { + "epoch": 7.121338912133891, + "grad_norm": 1.9921010732650757, + "learning_rate": 9.984688998468901e-06, + "loss": 0.0495, + "step": 15318 + }, + { + "epoch": 7.1218038121803815, + "grad_norm": 7.730679035186768, + "learning_rate": 9.9846879984688e-06, + "loss": 0.1709, + "step": 15319 + }, + { + "epoch": 7.122268712226871, + "grad_norm": 4.168414115905762, + "learning_rate": 9.9846869984687e-06, + "loss": 0.1053, + "step": 15320 + }, + { + "epoch": 7.122733612273361, + "grad_norm": 5.01591682434082, + "learning_rate": 9.984685998468601e-06, + "loss": 0.0829, + "step": 15321 + }, + { + "epoch": 7.123198512319851, + "grad_norm": 5.86016321182251, + "learning_rate": 9.9846849984685e-06, + "loss": 0.1402, + "step": 15322 + }, + { + "epoch": 7.123663412366342, + "grad_norm": 6.689244270324707, + "learning_rate": 9.9846839984684e-06, + "loss": 0.2101, + "step": 15323 + }, + { + "epoch": 7.124128312412831, + "grad_norm": 5.738551139831543, + "learning_rate": 9.9846829984683e-06, + "loss": 0.1837, + "step": 15324 + }, + { + "epoch": 7.124593212459321, + "grad_norm": 7.923681735992432, + "learning_rate": 9.9846819984682e-06, + "loss": 0.2988, + "step": 15325 + }, + { + "epoch": 7.125058112505811, + "grad_norm": 7.27900505065918, + "learning_rate": 9.984680998468101e-06, + "loss": 0.0986, + "step": 15326 + }, + { + "epoch": 7.125523012552302, + "grad_norm": 4.7258453369140625, + "learning_rate": 9.984679998468002e-06, + "loss": 0.0876, + "step": 15327 + }, + { + "epoch": 7.1259879125987915, + "grad_norm": 4.162836074829102, + "learning_rate": 9.9846789984679e-06, + "loss": 0.1561, + "step": 15328 + }, + { + "epoch": 7.126452812645281, + "grad_norm": 2.47049880027771, + "learning_rate": 9.984677998467801e-06, + "loss": 0.0338, + "step": 15329 + }, + { + "epoch": 7.126917712691771, + "grad_norm": 3.596132278442383, + "learning_rate": 9.9846769984677e-06, + "loss": 0.0981, + "step": 15330 + }, + { + "epoch": 7.127382612738261, + "grad_norm": 2.8030662536621094, + "learning_rate": 9.984675998467601e-06, + "loss": 0.0442, + "step": 15331 + }, + { + "epoch": 7.127847512784752, + "grad_norm": 7.583187103271484, + "learning_rate": 9.9846749984675e-06, + "loss": 0.2518, + "step": 15332 + }, + { + "epoch": 7.128312412831241, + "grad_norm": 9.865689277648926, + "learning_rate": 9.9846739984674e-06, + "loss": 0.2703, + "step": 15333 + }, + { + "epoch": 7.128777312877731, + "grad_norm": 5.0626044273376465, + "learning_rate": 9.984672998467301e-06, + "loss": 0.1224, + "step": 15334 + }, + { + "epoch": 7.129242212924221, + "grad_norm": 8.861686706542969, + "learning_rate": 9.9846719984672e-06, + "loss": 0.3294, + "step": 15335 + }, + { + "epoch": 7.129707112970712, + "grad_norm": 4.258528232574463, + "learning_rate": 9.984670998467101e-06, + "loss": 0.1006, + "step": 15336 + }, + { + "epoch": 7.1301720130172015, + "grad_norm": 4.969771385192871, + "learning_rate": 9.984669998467e-06, + "loss": 0.1506, + "step": 15337 + }, + { + "epoch": 7.130636913063691, + "grad_norm": 4.098114013671875, + "learning_rate": 9.9846689984669e-06, + "loss": 0.1086, + "step": 15338 + }, + { + "epoch": 7.131101813110181, + "grad_norm": 4.387730121612549, + "learning_rate": 9.984667998466801e-06, + "loss": 0.1156, + "step": 15339 + }, + { + "epoch": 7.131566713156671, + "grad_norm": 5.167040824890137, + "learning_rate": 9.984666998466702e-06, + "loss": 0.1271, + "step": 15340 + }, + { + "epoch": 7.132031613203162, + "grad_norm": 2.772338628768921, + "learning_rate": 9.9846659984666e-06, + "loss": 0.0413, + "step": 15341 + }, + { + "epoch": 7.132496513249651, + "grad_norm": 3.6358673572540283, + "learning_rate": 9.984664998466501e-06, + "loss": 0.0787, + "step": 15342 + }, + { + "epoch": 7.132961413296141, + "grad_norm": 4.535645961761475, + "learning_rate": 9.9846639984664e-06, + "loss": 0.0849, + "step": 15343 + }, + { + "epoch": 7.133426313342631, + "grad_norm": 2.9874722957611084, + "learning_rate": 9.9846629984663e-06, + "loss": 0.0481, + "step": 15344 + }, + { + "epoch": 7.133891213389122, + "grad_norm": 7.436214923858643, + "learning_rate": 9.9846619984662e-06, + "loss": 0.1796, + "step": 15345 + }, + { + "epoch": 7.1343561134356115, + "grad_norm": 3.941774845123291, + "learning_rate": 9.9846609984661e-06, + "loss": 0.074, + "step": 15346 + }, + { + "epoch": 7.134821013482101, + "grad_norm": 4.6973676681518555, + "learning_rate": 9.984659998466001e-06, + "loss": 0.18, + "step": 15347 + }, + { + "epoch": 7.135285913528591, + "grad_norm": 6.166204452514648, + "learning_rate": 9.9846589984659e-06, + "loss": 0.149, + "step": 15348 + }, + { + "epoch": 7.135750813575082, + "grad_norm": 3.8224055767059326, + "learning_rate": 9.984657998465801e-06, + "loss": 0.0343, + "step": 15349 + }, + { + "epoch": 7.136215713621572, + "grad_norm": 3.416102647781372, + "learning_rate": 9.9846569984657e-06, + "loss": 0.0734, + "step": 15350 + }, + { + "epoch": 7.136680613668061, + "grad_norm": 4.651432514190674, + "learning_rate": 9.9846559984656e-06, + "loss": 0.0863, + "step": 15351 + }, + { + "epoch": 7.137145513714551, + "grad_norm": 4.43466854095459, + "learning_rate": 9.984654998465501e-06, + "loss": 0.0794, + "step": 15352 + }, + { + "epoch": 7.137610413761041, + "grad_norm": 4.979189872741699, + "learning_rate": 9.984653998465402e-06, + "loss": 0.0634, + "step": 15353 + }, + { + "epoch": 7.138075313807532, + "grad_norm": 4.34775972366333, + "learning_rate": 9.984652998465301e-06, + "loss": 0.077, + "step": 15354 + }, + { + "epoch": 7.1385402138540215, + "grad_norm": 2.9478766918182373, + "learning_rate": 9.9846519984652e-06, + "loss": 0.0666, + "step": 15355 + }, + { + "epoch": 7.139005113900511, + "grad_norm": 4.862457752227783, + "learning_rate": 9.9846509984651e-06, + "loss": 0.0984, + "step": 15356 + }, + { + "epoch": 7.139470013947001, + "grad_norm": 2.9980366230010986, + "learning_rate": 9.984649998465e-06, + "loss": 0.0516, + "step": 15357 + }, + { + "epoch": 7.139934913993492, + "grad_norm": 4.151941299438477, + "learning_rate": 9.9846489984649e-06, + "loss": 0.082, + "step": 15358 + }, + { + "epoch": 7.140399814039982, + "grad_norm": 4.309471607208252, + "learning_rate": 9.9846479984648e-06, + "loss": 0.09, + "step": 15359 + }, + { + "epoch": 7.140864714086471, + "grad_norm": 4.839651584625244, + "learning_rate": 9.984646998464701e-06, + "loss": 0.128, + "step": 15360 + }, + { + "epoch": 7.141329614132961, + "grad_norm": 4.428020000457764, + "learning_rate": 9.9846459984646e-06, + "loss": 0.0959, + "step": 15361 + }, + { + "epoch": 7.141794514179451, + "grad_norm": 4.371557235717773, + "learning_rate": 9.984644998464501e-06, + "loss": 0.0761, + "step": 15362 + }, + { + "epoch": 7.142259414225942, + "grad_norm": 3.8619871139526367, + "learning_rate": 9.9846439984644e-06, + "loss": 0.0699, + "step": 15363 + }, + { + "epoch": 7.1427243142724315, + "grad_norm": 4.3616042137146, + "learning_rate": 9.9846429984643e-06, + "loss": 0.1478, + "step": 15364 + }, + { + "epoch": 7.143189214318921, + "grad_norm": 5.400933265686035, + "learning_rate": 9.984641998464201e-06, + "loss": 0.178, + "step": 15365 + }, + { + "epoch": 7.143654114365411, + "grad_norm": 4.571890354156494, + "learning_rate": 9.9846409984641e-06, + "loss": 0.1075, + "step": 15366 + }, + { + "epoch": 7.144119014411902, + "grad_norm": 6.904974460601807, + "learning_rate": 9.984639998464001e-06, + "loss": 0.0623, + "step": 15367 + }, + { + "epoch": 7.144583914458392, + "grad_norm": 6.134594440460205, + "learning_rate": 9.9846389984639e-06, + "loss": 0.1405, + "step": 15368 + }, + { + "epoch": 7.145048814504881, + "grad_norm": 4.155548572540283, + "learning_rate": 9.9846379984638e-06, + "loss": 0.0876, + "step": 15369 + }, + { + "epoch": 7.145513714551371, + "grad_norm": 4.592036247253418, + "learning_rate": 9.9846369984637e-06, + "loss": 0.1346, + "step": 15370 + }, + { + "epoch": 7.145978614597862, + "grad_norm": 6.231808662414551, + "learning_rate": 9.9846359984636e-06, + "loss": 0.1757, + "step": 15371 + }, + { + "epoch": 7.146443514644352, + "grad_norm": 8.933492660522461, + "learning_rate": 9.984634998463501e-06, + "loss": 0.1454, + "step": 15372 + }, + { + "epoch": 7.1469084146908415, + "grad_norm": 6.94960355758667, + "learning_rate": 9.984633998463402e-06, + "loss": 0.2306, + "step": 15373 + }, + { + "epoch": 7.147373314737331, + "grad_norm": 2.0290210247039795, + "learning_rate": 9.9846329984633e-06, + "loss": 0.0255, + "step": 15374 + }, + { + "epoch": 7.147838214783821, + "grad_norm": 12.061819076538086, + "learning_rate": 9.984631998463201e-06, + "loss": 0.1384, + "step": 15375 + }, + { + "epoch": 7.148303114830312, + "grad_norm": 4.273094177246094, + "learning_rate": 9.9846309984631e-06, + "loss": 0.0818, + "step": 15376 + }, + { + "epoch": 7.148768014876802, + "grad_norm": 9.16814136505127, + "learning_rate": 9.984629998463e-06, + "loss": 0.2122, + "step": 15377 + }, + { + "epoch": 7.149232914923291, + "grad_norm": 5.044422626495361, + "learning_rate": 9.984628998462901e-06, + "loss": 0.1088, + "step": 15378 + }, + { + "epoch": 7.149697814969781, + "grad_norm": 5.311323165893555, + "learning_rate": 9.9846279984628e-06, + "loss": 0.114, + "step": 15379 + }, + { + "epoch": 7.150162715016272, + "grad_norm": 5.929371356964111, + "learning_rate": 9.984626998462701e-06, + "loss": 0.163, + "step": 15380 + }, + { + "epoch": 7.150627615062762, + "grad_norm": 6.324291706085205, + "learning_rate": 9.9846259984626e-06, + "loss": 0.1287, + "step": 15381 + }, + { + "epoch": 7.1510925151092515, + "grad_norm": 4.068136215209961, + "learning_rate": 9.9846249984625e-06, + "loss": 0.0845, + "step": 15382 + }, + { + "epoch": 7.151557415155741, + "grad_norm": 4.003767490386963, + "learning_rate": 9.984623998462401e-06, + "loss": 0.0681, + "step": 15383 + }, + { + "epoch": 7.152022315202231, + "grad_norm": 5.943828105926514, + "learning_rate": 9.9846229984623e-06, + "loss": 0.1264, + "step": 15384 + }, + { + "epoch": 7.152487215248722, + "grad_norm": 4.505175590515137, + "learning_rate": 9.984621998462201e-06, + "loss": 0.0886, + "step": 15385 + }, + { + "epoch": 7.152952115295212, + "grad_norm": 5.39245080947876, + "learning_rate": 9.984620998462102e-06, + "loss": 0.1566, + "step": 15386 + }, + { + "epoch": 7.153417015341701, + "grad_norm": 4.221739768981934, + "learning_rate": 9.984619998462e-06, + "loss": 0.1005, + "step": 15387 + }, + { + "epoch": 7.153881915388191, + "grad_norm": 5.776895523071289, + "learning_rate": 9.9846189984619e-06, + "loss": 0.1546, + "step": 15388 + }, + { + "epoch": 7.154346815434682, + "grad_norm": 1.818247675895691, + "learning_rate": 9.9846179984618e-06, + "loss": 0.0254, + "step": 15389 + }, + { + "epoch": 7.154811715481172, + "grad_norm": 3.8092660903930664, + "learning_rate": 9.9846169984617e-06, + "loss": 0.0597, + "step": 15390 + }, + { + "epoch": 7.1552766155276615, + "grad_norm": 4.161937713623047, + "learning_rate": 9.984615998461601e-06, + "loss": 0.1169, + "step": 15391 + }, + { + "epoch": 7.155741515574151, + "grad_norm": 5.01603889465332, + "learning_rate": 9.9846149984615e-06, + "loss": 0.1232, + "step": 15392 + }, + { + "epoch": 7.156206415620642, + "grad_norm": 3.2562806606292725, + "learning_rate": 9.984613998461401e-06, + "loss": 0.0651, + "step": 15393 + }, + { + "epoch": 7.156671315667132, + "grad_norm": 8.190557479858398, + "learning_rate": 9.9846129984613e-06, + "loss": 0.219, + "step": 15394 + }, + { + "epoch": 7.157136215713622, + "grad_norm": 5.060513496398926, + "learning_rate": 9.9846119984612e-06, + "loss": 0.0871, + "step": 15395 + }, + { + "epoch": 7.157601115760111, + "grad_norm": 5.151711940765381, + "learning_rate": 9.984610998461101e-06, + "loss": 0.1518, + "step": 15396 + }, + { + "epoch": 7.158066015806601, + "grad_norm": 6.024079322814941, + "learning_rate": 9.984609998461e-06, + "loss": 0.2335, + "step": 15397 + }, + { + "epoch": 7.158530915853092, + "grad_norm": 3.5487585067749023, + "learning_rate": 9.984608998460901e-06, + "loss": 0.064, + "step": 15398 + }, + { + "epoch": 7.158995815899582, + "grad_norm": 4.4359331130981445, + "learning_rate": 9.9846079984608e-06, + "loss": 0.1691, + "step": 15399 + }, + { + "epoch": 7.1594607159460715, + "grad_norm": 2.9615418910980225, + "learning_rate": 9.9846069984607e-06, + "loss": 0.0497, + "step": 15400 + }, + { + "epoch": 7.159925615992561, + "grad_norm": 6.033602237701416, + "learning_rate": 9.9846059984606e-06, + "loss": 0.1328, + "step": 15401 + }, + { + "epoch": 7.160390516039052, + "grad_norm": 4.826061248779297, + "learning_rate": 9.9846049984605e-06, + "loss": 0.138, + "step": 15402 + }, + { + "epoch": 7.160855416085542, + "grad_norm": 3.377183675765991, + "learning_rate": 9.984603998460401e-06, + "loss": 0.0715, + "step": 15403 + }, + { + "epoch": 7.161320316132032, + "grad_norm": 2.925387144088745, + "learning_rate": 9.984602998460302e-06, + "loss": 0.0532, + "step": 15404 + }, + { + "epoch": 7.161785216178521, + "grad_norm": 8.32076644897461, + "learning_rate": 9.9846019984602e-06, + "loss": 0.1981, + "step": 15405 + }, + { + "epoch": 7.162250116225012, + "grad_norm": 6.337951183319092, + "learning_rate": 9.984600998460101e-06, + "loss": 0.1294, + "step": 15406 + }, + { + "epoch": 7.162715016271502, + "grad_norm": 3.2857730388641357, + "learning_rate": 9.98459999846e-06, + "loss": 0.074, + "step": 15407 + }, + { + "epoch": 7.163179916317992, + "grad_norm": 4.77269983291626, + "learning_rate": 9.9845989984599e-06, + "loss": 0.1213, + "step": 15408 + }, + { + "epoch": 7.1636448163644815, + "grad_norm": 4.934322834014893, + "learning_rate": 9.984597998459801e-06, + "loss": 0.0711, + "step": 15409 + }, + { + "epoch": 7.164109716410971, + "grad_norm": 8.135581970214844, + "learning_rate": 9.9845969984597e-06, + "loss": 0.3458, + "step": 15410 + }, + { + "epoch": 7.164574616457462, + "grad_norm": 4.395566940307617, + "learning_rate": 9.984595998459601e-06, + "loss": 0.0823, + "step": 15411 + }, + { + "epoch": 7.165039516503952, + "grad_norm": 7.356789588928223, + "learning_rate": 9.9845949984595e-06, + "loss": 0.1877, + "step": 15412 + }, + { + "epoch": 7.165504416550442, + "grad_norm": 5.625191688537598, + "learning_rate": 9.9845939984594e-06, + "loss": 0.1543, + "step": 15413 + }, + { + "epoch": 7.165969316596931, + "grad_norm": 4.34071683883667, + "learning_rate": 9.9845929984593e-06, + "loss": 0.1296, + "step": 15414 + }, + { + "epoch": 7.166434216643422, + "grad_norm": 1.719420313835144, + "learning_rate": 9.9845919984592e-06, + "loss": 0.0163, + "step": 15415 + }, + { + "epoch": 7.166899116689912, + "grad_norm": 7.532530307769775, + "learning_rate": 9.984590998459101e-06, + "loss": 0.2186, + "step": 15416 + }, + { + "epoch": 7.167364016736402, + "grad_norm": 4.471784591674805, + "learning_rate": 9.984589998459002e-06, + "loss": 0.1302, + "step": 15417 + }, + { + "epoch": 7.1678289167828915, + "grad_norm": 5.331892490386963, + "learning_rate": 9.9845889984589e-06, + "loss": 0.1562, + "step": 15418 + }, + { + "epoch": 7.168293816829381, + "grad_norm": 4.88864278793335, + "learning_rate": 9.9845879984588e-06, + "loss": 0.1047, + "step": 15419 + }, + { + "epoch": 7.168758716875872, + "grad_norm": 6.172667980194092, + "learning_rate": 9.9845869984587e-06, + "loss": 0.121, + "step": 15420 + }, + { + "epoch": 7.169223616922362, + "grad_norm": 5.729132652282715, + "learning_rate": 9.984585998458601e-06, + "loss": 0.0928, + "step": 15421 + }, + { + "epoch": 7.169688516968852, + "grad_norm": 2.6333167552948, + "learning_rate": 9.984584998458502e-06, + "loss": 0.0635, + "step": 15422 + }, + { + "epoch": 7.170153417015341, + "grad_norm": 4.942734241485596, + "learning_rate": 9.9845839984584e-06, + "loss": 0.1675, + "step": 15423 + }, + { + "epoch": 7.170618317061832, + "grad_norm": 3.2913811206817627, + "learning_rate": 9.984582998458301e-06, + "loss": 0.061, + "step": 15424 + }, + { + "epoch": 7.171083217108322, + "grad_norm": 4.876601696014404, + "learning_rate": 9.9845819984582e-06, + "loss": 0.1056, + "step": 15425 + }, + { + "epoch": 7.171548117154812, + "grad_norm": 7.850440979003906, + "learning_rate": 9.9845809984581e-06, + "loss": 0.1033, + "step": 15426 + }, + { + "epoch": 7.1720130172013015, + "grad_norm": 6.119693279266357, + "learning_rate": 9.984579998458e-06, + "loss": 0.1539, + "step": 15427 + }, + { + "epoch": 7.172477917247791, + "grad_norm": 6.375845909118652, + "learning_rate": 9.9845789984579e-06, + "loss": 0.3164, + "step": 15428 + }, + { + "epoch": 7.172942817294282, + "grad_norm": 3.4448108673095703, + "learning_rate": 9.984577998457801e-06, + "loss": 0.0452, + "step": 15429 + }, + { + "epoch": 7.173407717340772, + "grad_norm": 6.515023708343506, + "learning_rate": 9.9845769984577e-06, + "loss": 0.2184, + "step": 15430 + }, + { + "epoch": 7.173872617387262, + "grad_norm": 5.174365043640137, + "learning_rate": 9.9845759984576e-06, + "loss": 0.0939, + "step": 15431 + }, + { + "epoch": 7.174337517433751, + "grad_norm": 9.594884872436523, + "learning_rate": 9.9845749984575e-06, + "loss": 0.2744, + "step": 15432 + }, + { + "epoch": 7.174802417480242, + "grad_norm": 4.229883670806885, + "learning_rate": 9.9845739984574e-06, + "loss": 0.0998, + "step": 15433 + }, + { + "epoch": 7.175267317526732, + "grad_norm": 5.870479583740234, + "learning_rate": 9.984572998457301e-06, + "loss": 0.2169, + "step": 15434 + }, + { + "epoch": 7.175732217573222, + "grad_norm": 5.170496940612793, + "learning_rate": 9.984571998457202e-06, + "loss": 0.1157, + "step": 15435 + }, + { + "epoch": 7.1761971176197115, + "grad_norm": 2.4922380447387695, + "learning_rate": 9.9845709984571e-06, + "loss": 0.0173, + "step": 15436 + }, + { + "epoch": 7.176662017666202, + "grad_norm": 3.3046014308929443, + "learning_rate": 9.984569998457001e-06, + "loss": 0.039, + "step": 15437 + }, + { + "epoch": 7.177126917712692, + "grad_norm": 6.078765869140625, + "learning_rate": 9.9845689984569e-06, + "loss": 0.2345, + "step": 15438 + }, + { + "epoch": 7.177591817759182, + "grad_norm": 5.1396589279174805, + "learning_rate": 9.984567998456801e-06, + "loss": 0.1553, + "step": 15439 + }, + { + "epoch": 7.178056717805672, + "grad_norm": 3.5450096130371094, + "learning_rate": 9.9845669984567e-06, + "loss": 0.0935, + "step": 15440 + }, + { + "epoch": 7.178521617852161, + "grad_norm": 4.248461723327637, + "learning_rate": 9.9845659984566e-06, + "loss": 0.057, + "step": 15441 + }, + { + "epoch": 7.178986517898652, + "grad_norm": 4.33626127243042, + "learning_rate": 9.984564998456501e-06, + "loss": 0.1144, + "step": 15442 + }, + { + "epoch": 7.179451417945142, + "grad_norm": 3.798935890197754, + "learning_rate": 9.9845639984564e-06, + "loss": 0.0752, + "step": 15443 + }, + { + "epoch": 7.179916317991632, + "grad_norm": 4.832057476043701, + "learning_rate": 9.9845629984563e-06, + "loss": 0.1834, + "step": 15444 + }, + { + "epoch": 7.1803812180381215, + "grad_norm": 5.359344482421875, + "learning_rate": 9.9845619984562e-06, + "loss": 0.2176, + "step": 15445 + }, + { + "epoch": 7.180846118084612, + "grad_norm": 3.703026056289673, + "learning_rate": 9.9845609984561e-06, + "loss": 0.0772, + "step": 15446 + }, + { + "epoch": 7.181311018131102, + "grad_norm": 4.794456958770752, + "learning_rate": 9.984559998456001e-06, + "loss": 0.0884, + "step": 15447 + }, + { + "epoch": 7.181775918177592, + "grad_norm": 0.9116426706314087, + "learning_rate": 9.984558998455902e-06, + "loss": 0.007, + "step": 15448 + }, + { + "epoch": 7.182240818224082, + "grad_norm": 8.641687393188477, + "learning_rate": 9.9845579984558e-06, + "loss": 0.1188, + "step": 15449 + }, + { + "epoch": 7.182705718270572, + "grad_norm": 3.598287343978882, + "learning_rate": 9.984556998455701e-06, + "loss": 0.0752, + "step": 15450 + }, + { + "epoch": 7.183170618317062, + "grad_norm": 6.196580410003662, + "learning_rate": 9.9845559984556e-06, + "loss": 0.272, + "step": 15451 + }, + { + "epoch": 7.183635518363552, + "grad_norm": 3.8117196559906006, + "learning_rate": 9.984554998455501e-06, + "loss": 0.064, + "step": 15452 + }, + { + "epoch": 7.184100418410042, + "grad_norm": 5.441165924072266, + "learning_rate": 9.9845539984554e-06, + "loss": 0.1885, + "step": 15453 + }, + { + "epoch": 7.1845653184565315, + "grad_norm": 3.5740725994110107, + "learning_rate": 9.9845529984553e-06, + "loss": 0.0819, + "step": 15454 + }, + { + "epoch": 7.185030218503022, + "grad_norm": 5.207942485809326, + "learning_rate": 9.984551998455201e-06, + "loss": 0.1031, + "step": 15455 + }, + { + "epoch": 7.185495118549512, + "grad_norm": 3.431018829345703, + "learning_rate": 9.9845509984551e-06, + "loss": 0.0991, + "step": 15456 + }, + { + "epoch": 7.185960018596002, + "grad_norm": 4.06334924697876, + "learning_rate": 9.984549998455001e-06, + "loss": 0.1021, + "step": 15457 + }, + { + "epoch": 7.1864249186424916, + "grad_norm": 0.897130012512207, + "learning_rate": 9.9845489984549e-06, + "loss": 0.0153, + "step": 15458 + }, + { + "epoch": 7.186889818688982, + "grad_norm": 6.737115383148193, + "learning_rate": 9.9845479984548e-06, + "loss": 0.2157, + "step": 15459 + }, + { + "epoch": 7.187354718735472, + "grad_norm": 4.102445125579834, + "learning_rate": 9.984546998454701e-06, + "loss": 0.0842, + "step": 15460 + }, + { + "epoch": 7.187819618781962, + "grad_norm": 6.03772497177124, + "learning_rate": 9.984545998454602e-06, + "loss": 0.1783, + "step": 15461 + }, + { + "epoch": 7.188284518828452, + "grad_norm": 8.813692092895508, + "learning_rate": 9.9845449984545e-06, + "loss": 0.2851, + "step": 15462 + }, + { + "epoch": 7.1887494188749415, + "grad_norm": 3.7908124923706055, + "learning_rate": 9.9845439984544e-06, + "loss": 0.082, + "step": 15463 + }, + { + "epoch": 7.189214318921432, + "grad_norm": 2.649527072906494, + "learning_rate": 9.9845429984543e-06, + "loss": 0.0611, + "step": 15464 + }, + { + "epoch": 7.189679218967922, + "grad_norm": 5.259807109832764, + "learning_rate": 9.984541998454201e-06, + "loss": 0.1224, + "step": 15465 + }, + { + "epoch": 7.190144119014412, + "grad_norm": 6.077775478363037, + "learning_rate": 9.9845409984541e-06, + "loss": 0.162, + "step": 15466 + }, + { + "epoch": 7.1906090190609016, + "grad_norm": 7.015239238739014, + "learning_rate": 9.984539998454e-06, + "loss": 0.2242, + "step": 15467 + }, + { + "epoch": 7.191073919107392, + "grad_norm": 7.954788684844971, + "learning_rate": 9.984538998453901e-06, + "loss": 0.2812, + "step": 15468 + }, + { + "epoch": 7.191538819153882, + "grad_norm": 4.465973854064941, + "learning_rate": 9.9845379984538e-06, + "loss": 0.0864, + "step": 15469 + }, + { + "epoch": 7.192003719200372, + "grad_norm": 2.022247314453125, + "learning_rate": 9.984536998453701e-06, + "loss": 0.0232, + "step": 15470 + }, + { + "epoch": 7.192468619246862, + "grad_norm": 4.116738319396973, + "learning_rate": 9.9845359984536e-06, + "loss": 0.1377, + "step": 15471 + }, + { + "epoch": 7.192933519293352, + "grad_norm": 5.473301410675049, + "learning_rate": 9.9845349984535e-06, + "loss": 0.1447, + "step": 15472 + }, + { + "epoch": 7.193398419339842, + "grad_norm": 6.665370464324951, + "learning_rate": 9.984533998453401e-06, + "loss": 0.1328, + "step": 15473 + }, + { + "epoch": 7.193863319386332, + "grad_norm": 5.189499378204346, + "learning_rate": 9.9845329984533e-06, + "loss": 0.1775, + "step": 15474 + }, + { + "epoch": 7.194328219432822, + "grad_norm": 4.222759246826172, + "learning_rate": 9.984531998453201e-06, + "loss": 0.0973, + "step": 15475 + }, + { + "epoch": 7.1947931194793115, + "grad_norm": 2.7588210105895996, + "learning_rate": 9.9845309984531e-06, + "loss": 0.0469, + "step": 15476 + }, + { + "epoch": 7.195258019525802, + "grad_norm": 5.529178619384766, + "learning_rate": 9.984529998453e-06, + "loss": 0.1113, + "step": 15477 + }, + { + "epoch": 7.195722919572292, + "grad_norm": 6.323272705078125, + "learning_rate": 9.984528998452901e-06, + "loss": 0.1646, + "step": 15478 + }, + { + "epoch": 7.196187819618782, + "grad_norm": 5.318271636962891, + "learning_rate": 9.984527998452802e-06, + "loss": 0.229, + "step": 15479 + }, + { + "epoch": 7.196652719665272, + "grad_norm": 7.609912872314453, + "learning_rate": 9.9845269984527e-06, + "loss": 0.2236, + "step": 15480 + }, + { + "epoch": 7.197117619711762, + "grad_norm": 5.906622886657715, + "learning_rate": 9.984525998452601e-06, + "loss": 0.1095, + "step": 15481 + }, + { + "epoch": 7.197582519758252, + "grad_norm": 7.890691757202148, + "learning_rate": 9.9845249984525e-06, + "loss": 0.3748, + "step": 15482 + }, + { + "epoch": 7.198047419804742, + "grad_norm": 3.28051495552063, + "learning_rate": 9.984523998452401e-06, + "loss": 0.1022, + "step": 15483 + }, + { + "epoch": 7.198512319851232, + "grad_norm": 5.333343029022217, + "learning_rate": 9.9845229984523e-06, + "loss": 0.1219, + "step": 15484 + }, + { + "epoch": 7.198977219897722, + "grad_norm": 7.5274457931518555, + "learning_rate": 9.9845219984522e-06, + "loss": 0.3069, + "step": 15485 + }, + { + "epoch": 7.199442119944212, + "grad_norm": 3.790426254272461, + "learning_rate": 9.984520998452101e-06, + "loss": 0.0617, + "step": 15486 + }, + { + "epoch": 7.199907019990702, + "grad_norm": 4.983832836151123, + "learning_rate": 9.984519998452e-06, + "loss": 0.193, + "step": 15487 + }, + { + "epoch": 7.200371920037192, + "grad_norm": 1.294163703918457, + "learning_rate": 9.984518998451901e-06, + "loss": 0.0224, + "step": 15488 + }, + { + "epoch": 7.200836820083682, + "grad_norm": 2.932295799255371, + "learning_rate": 9.9845179984518e-06, + "loss": 0.0573, + "step": 15489 + }, + { + "epoch": 7.201301720130172, + "grad_norm": 4.866278648376465, + "learning_rate": 9.9845169984517e-06, + "loss": 0.1649, + "step": 15490 + }, + { + "epoch": 7.201766620176662, + "grad_norm": 1.8086235523223877, + "learning_rate": 9.984515998451601e-06, + "loss": 0.0287, + "step": 15491 + }, + { + "epoch": 7.202231520223152, + "grad_norm": 4.218417167663574, + "learning_rate": 9.984514998451502e-06, + "loss": 0.1021, + "step": 15492 + }, + { + "epoch": 7.202696420269642, + "grad_norm": 4.9890336990356445, + "learning_rate": 9.9845139984514e-06, + "loss": 0.1256, + "step": 15493 + }, + { + "epoch": 7.203161320316132, + "grad_norm": 1.825061321258545, + "learning_rate": 9.984512998451301e-06, + "loss": 0.0191, + "step": 15494 + }, + { + "epoch": 7.203626220362622, + "grad_norm": 9.458599090576172, + "learning_rate": 9.9845119984512e-06, + "loss": 0.4006, + "step": 15495 + }, + { + "epoch": 7.204091120409112, + "grad_norm": 3.7949304580688477, + "learning_rate": 9.9845109984511e-06, + "loss": 0.0934, + "step": 15496 + }, + { + "epoch": 7.204556020455602, + "grad_norm": 4.583148002624512, + "learning_rate": 9.984509998451e-06, + "loss": 0.0991, + "step": 15497 + }, + { + "epoch": 7.205020920502092, + "grad_norm": 4.686832427978516, + "learning_rate": 9.9845089984509e-06, + "loss": 0.077, + "step": 15498 + }, + { + "epoch": 7.205485820548582, + "grad_norm": 4.881364822387695, + "learning_rate": 9.984507998450801e-06, + "loss": 0.0912, + "step": 15499 + }, + { + "epoch": 7.205950720595072, + "grad_norm": 3.24359130859375, + "learning_rate": 9.9845069984507e-06, + "loss": 0.0542, + "step": 15500 + }, + { + "epoch": 7.206415620641562, + "grad_norm": 6.701648235321045, + "learning_rate": 9.984505998450601e-06, + "loss": 0.297, + "step": 15501 + }, + { + "epoch": 7.206880520688052, + "grad_norm": 2.98069167137146, + "learning_rate": 9.9845049984505e-06, + "loss": 0.0712, + "step": 15502 + }, + { + "epoch": 7.207345420734542, + "grad_norm": 7.0677337646484375, + "learning_rate": 9.9845039984504e-06, + "loss": 0.1382, + "step": 15503 + }, + { + "epoch": 7.207810320781032, + "grad_norm": 5.849700450897217, + "learning_rate": 9.984502998450301e-06, + "loss": 0.1418, + "step": 15504 + }, + { + "epoch": 7.208275220827522, + "grad_norm": 4.884572982788086, + "learning_rate": 9.984501998450202e-06, + "loss": 0.0985, + "step": 15505 + }, + { + "epoch": 7.208740120874012, + "grad_norm": 6.021245956420898, + "learning_rate": 9.984500998450101e-06, + "loss": 0.2112, + "step": 15506 + }, + { + "epoch": 7.209205020920502, + "grad_norm": 3.4283719062805176, + "learning_rate": 9.98449999845e-06, + "loss": 0.0533, + "step": 15507 + }, + { + "epoch": 7.209669920966992, + "grad_norm": 3.9000039100646973, + "learning_rate": 9.9844989984499e-06, + "loss": 0.0977, + "step": 15508 + }, + { + "epoch": 7.210134821013482, + "grad_norm": 7.516404628753662, + "learning_rate": 9.9844979984498e-06, + "loss": 0.2395, + "step": 15509 + }, + { + "epoch": 7.210599721059972, + "grad_norm": 2.082440137863159, + "learning_rate": 9.9844969984497e-06, + "loss": 0.0228, + "step": 15510 + }, + { + "epoch": 7.211064621106462, + "grad_norm": 6.934351444244385, + "learning_rate": 9.9844959984496e-06, + "loss": 0.2211, + "step": 15511 + }, + { + "epoch": 7.211529521152952, + "grad_norm": 6.680822372436523, + "learning_rate": 9.984494998449501e-06, + "loss": 0.2414, + "step": 15512 + }, + { + "epoch": 7.211994421199442, + "grad_norm": 4.846813201904297, + "learning_rate": 9.9844939984494e-06, + "loss": 0.1424, + "step": 15513 + }, + { + "epoch": 7.212459321245932, + "grad_norm": 4.827147006988525, + "learning_rate": 9.984492998449301e-06, + "loss": 0.1503, + "step": 15514 + }, + { + "epoch": 7.212924221292422, + "grad_norm": 4.400742530822754, + "learning_rate": 9.9844919984492e-06, + "loss": 0.1459, + "step": 15515 + }, + { + "epoch": 7.2133891213389125, + "grad_norm": 6.027887344360352, + "learning_rate": 9.9844909984491e-06, + "loss": 0.228, + "step": 15516 + }, + { + "epoch": 7.213854021385402, + "grad_norm": 3.111422538757324, + "learning_rate": 9.984489998449001e-06, + "loss": 0.0486, + "step": 15517 + }, + { + "epoch": 7.214318921431892, + "grad_norm": 4.059264659881592, + "learning_rate": 9.9844889984489e-06, + "loss": 0.1293, + "step": 15518 + }, + { + "epoch": 7.214783821478382, + "grad_norm": 4.850531578063965, + "learning_rate": 9.984487998448801e-06, + "loss": 0.1884, + "step": 15519 + }, + { + "epoch": 7.215248721524872, + "grad_norm": 5.317748069763184, + "learning_rate": 9.9844869984487e-06, + "loss": 0.1396, + "step": 15520 + }, + { + "epoch": 7.215713621571362, + "grad_norm": 4.454763412475586, + "learning_rate": 9.9844859984486e-06, + "loss": 0.1143, + "step": 15521 + }, + { + "epoch": 7.216178521617852, + "grad_norm": 4.163949012756348, + "learning_rate": 9.9844849984485e-06, + "loss": 0.0741, + "step": 15522 + }, + { + "epoch": 7.216643421664342, + "grad_norm": 3.3892226219177246, + "learning_rate": 9.9844839984484e-06, + "loss": 0.0881, + "step": 15523 + }, + { + "epoch": 7.217108321710832, + "grad_norm": 3.8140509128570557, + "learning_rate": 9.984482998448301e-06, + "loss": 0.0785, + "step": 15524 + }, + { + "epoch": 7.2175732217573225, + "grad_norm": 8.711514472961426, + "learning_rate": 9.984481998448202e-06, + "loss": 0.2557, + "step": 15525 + }, + { + "epoch": 7.218038121803812, + "grad_norm": 3.8076744079589844, + "learning_rate": 9.9844809984481e-06, + "loss": 0.1363, + "step": 15526 + }, + { + "epoch": 7.218503021850302, + "grad_norm": 7.404539585113525, + "learning_rate": 9.984479998448001e-06, + "loss": 0.2666, + "step": 15527 + }, + { + "epoch": 7.218967921896792, + "grad_norm": 4.729537487030029, + "learning_rate": 9.9844789984479e-06, + "loss": 0.1274, + "step": 15528 + }, + { + "epoch": 7.219432821943283, + "grad_norm": 3.736799478530884, + "learning_rate": 9.9844779984478e-06, + "loss": 0.0807, + "step": 15529 + }, + { + "epoch": 7.219897721989772, + "grad_norm": 6.095840930938721, + "learning_rate": 9.984476998447701e-06, + "loss": 0.1562, + "step": 15530 + }, + { + "epoch": 7.220362622036262, + "grad_norm": 3.6312968730926514, + "learning_rate": 9.9844759984476e-06, + "loss": 0.0521, + "step": 15531 + }, + { + "epoch": 7.220827522082752, + "grad_norm": 4.205573081970215, + "learning_rate": 9.984474998447501e-06, + "loss": 0.1093, + "step": 15532 + }, + { + "epoch": 7.221292422129242, + "grad_norm": 6.076830863952637, + "learning_rate": 9.9844739984474e-06, + "loss": 0.2245, + "step": 15533 + }, + { + "epoch": 7.2217573221757325, + "grad_norm": 2.876344919204712, + "learning_rate": 9.9844729984473e-06, + "loss": 0.0541, + "step": 15534 + }, + { + "epoch": 7.222222222222222, + "grad_norm": 3.4015164375305176, + "learning_rate": 9.9844719984472e-06, + "loss": 0.0636, + "step": 15535 + }, + { + "epoch": 7.222687122268712, + "grad_norm": 5.133989334106445, + "learning_rate": 9.9844709984471e-06, + "loss": 0.0347, + "step": 15536 + }, + { + "epoch": 7.223152022315202, + "grad_norm": 7.27198600769043, + "learning_rate": 9.984469998447001e-06, + "loss": 0.1676, + "step": 15537 + }, + { + "epoch": 7.223616922361693, + "grad_norm": 4.759926795959473, + "learning_rate": 9.984468998446902e-06, + "loss": 0.1479, + "step": 15538 + }, + { + "epoch": 7.224081822408182, + "grad_norm": 6.158658504486084, + "learning_rate": 9.9844679984468e-06, + "loss": 0.1463, + "step": 15539 + }, + { + "epoch": 7.224546722454672, + "grad_norm": 2.8739869594573975, + "learning_rate": 9.9844669984467e-06, + "loss": 0.0421, + "step": 15540 + }, + { + "epoch": 7.225011622501162, + "grad_norm": 8.691327095031738, + "learning_rate": 9.9844659984466e-06, + "loss": 0.2267, + "step": 15541 + }, + { + "epoch": 7.225476522547652, + "grad_norm": 5.106838703155518, + "learning_rate": 9.984464998446501e-06, + "loss": 0.1293, + "step": 15542 + }, + { + "epoch": 7.2259414225941425, + "grad_norm": 6.410107612609863, + "learning_rate": 9.984463998446402e-06, + "loss": 0.2589, + "step": 15543 + }, + { + "epoch": 7.226406322640632, + "grad_norm": 5.2020440101623535, + "learning_rate": 9.9844629984463e-06, + "loss": 0.2113, + "step": 15544 + }, + { + "epoch": 7.226871222687122, + "grad_norm": 10.174201011657715, + "learning_rate": 9.984461998446201e-06, + "loss": 0.1093, + "step": 15545 + }, + { + "epoch": 7.227336122733612, + "grad_norm": 7.178328990936279, + "learning_rate": 9.9844609984461e-06, + "loss": 0.1281, + "step": 15546 + }, + { + "epoch": 7.227801022780103, + "grad_norm": 4.6762003898620605, + "learning_rate": 9.984459998446e-06, + "loss": 0.1683, + "step": 15547 + }, + { + "epoch": 7.228265922826592, + "grad_norm": 3.507305383682251, + "learning_rate": 9.984458998445901e-06, + "loss": 0.0529, + "step": 15548 + }, + { + "epoch": 7.228730822873082, + "grad_norm": 3.786440372467041, + "learning_rate": 9.9844579984458e-06, + "loss": 0.0877, + "step": 15549 + }, + { + "epoch": 7.229195722919572, + "grad_norm": 6.669147491455078, + "learning_rate": 9.984456998445701e-06, + "loss": 0.2214, + "step": 15550 + }, + { + "epoch": 7.229660622966063, + "grad_norm": 5.952110767364502, + "learning_rate": 9.9844559984456e-06, + "loss": 0.2106, + "step": 15551 + }, + { + "epoch": 7.2301255230125525, + "grad_norm": 5.5183000564575195, + "learning_rate": 9.9844549984455e-06, + "loss": 0.1933, + "step": 15552 + }, + { + "epoch": 7.230590423059042, + "grad_norm": 5.153769493103027, + "learning_rate": 9.9844539984454e-06, + "loss": 0.1258, + "step": 15553 + }, + { + "epoch": 7.231055323105532, + "grad_norm": 3.8677778244018555, + "learning_rate": 9.9844529984453e-06, + "loss": 0.0982, + "step": 15554 + }, + { + "epoch": 7.231520223152022, + "grad_norm": 4.414563179016113, + "learning_rate": 9.984451998445201e-06, + "loss": 0.0937, + "step": 15555 + }, + { + "epoch": 7.231985123198513, + "grad_norm": 5.268930435180664, + "learning_rate": 9.984450998445102e-06, + "loss": 0.1339, + "step": 15556 + }, + { + "epoch": 7.232450023245002, + "grad_norm": 4.758829116821289, + "learning_rate": 9.984449998445e-06, + "loss": 0.1871, + "step": 15557 + }, + { + "epoch": 7.232914923291492, + "grad_norm": 2.886350154876709, + "learning_rate": 9.984448998444901e-06, + "loss": 0.0686, + "step": 15558 + }, + { + "epoch": 7.233379823337982, + "grad_norm": 2.8057706356048584, + "learning_rate": 9.9844479984448e-06, + "loss": 0.049, + "step": 15559 + }, + { + "epoch": 7.233844723384473, + "grad_norm": 3.4971165657043457, + "learning_rate": 9.984446998444701e-06, + "loss": 0.0865, + "step": 15560 + }, + { + "epoch": 7.2343096234309625, + "grad_norm": 8.62649917602539, + "learning_rate": 9.984445998444602e-06, + "loss": 0.3198, + "step": 15561 + }, + { + "epoch": 7.234774523477452, + "grad_norm": 4.945692539215088, + "learning_rate": 9.9844449984445e-06, + "loss": 0.1202, + "step": 15562 + }, + { + "epoch": 7.235239423523942, + "grad_norm": 1.9587796926498413, + "learning_rate": 9.984443998444401e-06, + "loss": 0.0386, + "step": 15563 + }, + { + "epoch": 7.235704323570433, + "grad_norm": 5.945891857147217, + "learning_rate": 9.9844429984443e-06, + "loss": 0.1675, + "step": 15564 + }, + { + "epoch": 7.236169223616923, + "grad_norm": 3.4269216060638428, + "learning_rate": 9.9844419984442e-06, + "loss": 0.0829, + "step": 15565 + }, + { + "epoch": 7.236634123663412, + "grad_norm": 5.737220287322998, + "learning_rate": 9.9844409984441e-06, + "loss": 0.1317, + "step": 15566 + }, + { + "epoch": 7.237099023709902, + "grad_norm": 3.844163656234741, + "learning_rate": 9.984439998444e-06, + "loss": 0.0617, + "step": 15567 + }, + { + "epoch": 7.237563923756392, + "grad_norm": 7.281937122344971, + "learning_rate": 9.984438998443901e-06, + "loss": 0.1585, + "step": 15568 + }, + { + "epoch": 7.238028823802883, + "grad_norm": 7.196857929229736, + "learning_rate": 9.984437998443802e-06, + "loss": 0.1135, + "step": 15569 + }, + { + "epoch": 7.2384937238493725, + "grad_norm": 7.3251190185546875, + "learning_rate": 9.9844369984437e-06, + "loss": 0.2254, + "step": 15570 + }, + { + "epoch": 7.238958623895862, + "grad_norm": 3.9857630729675293, + "learning_rate": 9.9844359984436e-06, + "loss": 0.0441, + "step": 15571 + }, + { + "epoch": 7.239423523942352, + "grad_norm": 4.526794910430908, + "learning_rate": 9.9844349984435e-06, + "loss": 0.04, + "step": 15572 + }, + { + "epoch": 7.239888423988843, + "grad_norm": 8.245064735412598, + "learning_rate": 9.984433998443401e-06, + "loss": 0.222, + "step": 15573 + }, + { + "epoch": 7.240353324035333, + "grad_norm": 5.251161575317383, + "learning_rate": 9.984432998443302e-06, + "loss": 0.093, + "step": 15574 + }, + { + "epoch": 7.240818224081822, + "grad_norm": 5.800543785095215, + "learning_rate": 9.9844319984432e-06, + "loss": 0.1065, + "step": 15575 + }, + { + "epoch": 7.241283124128312, + "grad_norm": 3.70171856880188, + "learning_rate": 9.984430998443101e-06, + "loss": 0.0264, + "step": 15576 + }, + { + "epoch": 7.241748024174802, + "grad_norm": 6.1206955909729, + "learning_rate": 9.984429998443e-06, + "loss": 0.1342, + "step": 15577 + }, + { + "epoch": 7.242212924221293, + "grad_norm": 6.574766635894775, + "learning_rate": 9.9844289984429e-06, + "loss": 0.1111, + "step": 15578 + }, + { + "epoch": 7.2426778242677825, + "grad_norm": 9.429868698120117, + "learning_rate": 9.9844279984428e-06, + "loss": 0.0455, + "step": 15579 + }, + { + "epoch": 7.243142724314272, + "grad_norm": 4.960545539855957, + "learning_rate": 9.9844269984427e-06, + "loss": 0.0714, + "step": 15580 + }, + { + "epoch": 7.243607624360762, + "grad_norm": 9.339349746704102, + "learning_rate": 9.984425998442601e-06, + "loss": 0.2931, + "step": 15581 + }, + { + "epoch": 7.244072524407253, + "grad_norm": 5.332241058349609, + "learning_rate": 9.9844249984425e-06, + "loss": 0.1485, + "step": 15582 + }, + { + "epoch": 7.244537424453743, + "grad_norm": 3.691237449645996, + "learning_rate": 9.9844239984424e-06, + "loss": 0.0761, + "step": 15583 + }, + { + "epoch": 7.245002324500232, + "grad_norm": 7.178295612335205, + "learning_rate": 9.9844229984423e-06, + "loss": 0.2394, + "step": 15584 + }, + { + "epoch": 7.245467224546722, + "grad_norm": 2.3324460983276367, + "learning_rate": 9.9844219984422e-06, + "loss": 0.0371, + "step": 15585 + }, + { + "epoch": 7.245932124593212, + "grad_norm": 5.4067816734313965, + "learning_rate": 9.984420998442101e-06, + "loss": 0.1594, + "step": 15586 + }, + { + "epoch": 7.246397024639703, + "grad_norm": 3.7265310287475586, + "learning_rate": 9.984419998442002e-06, + "loss": 0.0854, + "step": 15587 + }, + { + "epoch": 7.2468619246861925, + "grad_norm": 5.624577522277832, + "learning_rate": 9.9844189984419e-06, + "loss": 0.1505, + "step": 15588 + }, + { + "epoch": 7.247326824732682, + "grad_norm": 5.726665019989014, + "learning_rate": 9.984417998441801e-06, + "loss": 0.2051, + "step": 15589 + }, + { + "epoch": 7.247791724779172, + "grad_norm": 5.286290168762207, + "learning_rate": 9.9844169984417e-06, + "loss": 0.1018, + "step": 15590 + }, + { + "epoch": 7.248256624825663, + "grad_norm": 5.56626558303833, + "learning_rate": 9.984415998441601e-06, + "loss": 0.1202, + "step": 15591 + }, + { + "epoch": 7.248721524872153, + "grad_norm": 5.991386890411377, + "learning_rate": 9.9844149984415e-06, + "loss": 0.1165, + "step": 15592 + }, + { + "epoch": 7.249186424918642, + "grad_norm": 2.426146984100342, + "learning_rate": 9.9844139984414e-06, + "loss": 0.0327, + "step": 15593 + }, + { + "epoch": 7.249651324965132, + "grad_norm": 5.607985496520996, + "learning_rate": 9.984412998441301e-06, + "loss": 0.0887, + "step": 15594 + }, + { + "epoch": 7.250116225011623, + "grad_norm": 4.590752124786377, + "learning_rate": 9.9844119984412e-06, + "loss": 0.1213, + "step": 15595 + }, + { + "epoch": 7.250581125058113, + "grad_norm": 4.148460865020752, + "learning_rate": 9.9844109984411e-06, + "loss": 0.0694, + "step": 15596 + }, + { + "epoch": 7.2510460251046025, + "grad_norm": 3.8971846103668213, + "learning_rate": 9.984409998441e-06, + "loss": 0.0563, + "step": 15597 + }, + { + "epoch": 7.251510925151092, + "grad_norm": 2.999979257583618, + "learning_rate": 9.9844089984409e-06, + "loss": 0.0382, + "step": 15598 + }, + { + "epoch": 7.251975825197582, + "grad_norm": 9.527471542358398, + "learning_rate": 9.984407998440801e-06, + "loss": 0.5746, + "step": 15599 + }, + { + "epoch": 7.252440725244073, + "grad_norm": 5.3685784339904785, + "learning_rate": 9.984406998440702e-06, + "loss": 0.1935, + "step": 15600 + }, + { + "epoch": 7.252905625290563, + "grad_norm": 4.132340431213379, + "learning_rate": 9.9844059984406e-06, + "loss": 0.0699, + "step": 15601 + }, + { + "epoch": 7.253370525337052, + "grad_norm": 6.254533767700195, + "learning_rate": 9.984404998440501e-06, + "loss": 0.1223, + "step": 15602 + }, + { + "epoch": 7.253835425383542, + "grad_norm": 4.529016494750977, + "learning_rate": 9.9844039984404e-06, + "loss": 0.0956, + "step": 15603 + }, + { + "epoch": 7.254300325430033, + "grad_norm": 4.033874988555908, + "learning_rate": 9.9844029984403e-06, + "loss": 0.0848, + "step": 15604 + }, + { + "epoch": 7.254765225476523, + "grad_norm": 3.9492316246032715, + "learning_rate": 9.9844019984402e-06, + "loss": 0.0807, + "step": 15605 + }, + { + "epoch": 7.2552301255230125, + "grad_norm": 4.823835849761963, + "learning_rate": 9.9844009984401e-06, + "loss": 0.1317, + "step": 15606 + }, + { + "epoch": 7.255695025569502, + "grad_norm": 1.7349345684051514, + "learning_rate": 9.984399998440001e-06, + "loss": 0.0268, + "step": 15607 + }, + { + "epoch": 7.256159925615993, + "grad_norm": 5.69976282119751, + "learning_rate": 9.9843989984399e-06, + "loss": 0.2333, + "step": 15608 + }, + { + "epoch": 7.256624825662483, + "grad_norm": 3.5875630378723145, + "learning_rate": 9.984397998439801e-06, + "loss": 0.0527, + "step": 15609 + }, + { + "epoch": 7.257089725708973, + "grad_norm": 2.7933526039123535, + "learning_rate": 9.9843969984397e-06, + "loss": 0.0434, + "step": 15610 + }, + { + "epoch": 7.257554625755462, + "grad_norm": 5.1351399421691895, + "learning_rate": 9.9843959984396e-06, + "loss": 0.1013, + "step": 15611 + }, + { + "epoch": 7.258019525801952, + "grad_norm": 2.064189910888672, + "learning_rate": 9.984394998439501e-06, + "loss": 0.0303, + "step": 15612 + }, + { + "epoch": 7.258484425848443, + "grad_norm": 5.878558158874512, + "learning_rate": 9.984393998439402e-06, + "loss": 0.1734, + "step": 15613 + }, + { + "epoch": 7.258949325894933, + "grad_norm": 7.25774621963501, + "learning_rate": 9.9843929984393e-06, + "loss": 0.1728, + "step": 15614 + }, + { + "epoch": 7.2594142259414225, + "grad_norm": 10.61683464050293, + "learning_rate": 9.9843919984392e-06, + "loss": 0.1975, + "step": 15615 + }, + { + "epoch": 7.259879125987912, + "grad_norm": 4.081056594848633, + "learning_rate": 9.9843909984391e-06, + "loss": 0.0595, + "step": 15616 + }, + { + "epoch": 7.260344026034403, + "grad_norm": 5.990633964538574, + "learning_rate": 9.984389998439001e-06, + "loss": 0.0808, + "step": 15617 + }, + { + "epoch": 7.260808926080893, + "grad_norm": 1.8385018110275269, + "learning_rate": 9.9843889984389e-06, + "loss": 0.0311, + "step": 15618 + }, + { + "epoch": 7.261273826127383, + "grad_norm": 3.928614616394043, + "learning_rate": 9.9843879984388e-06, + "loss": 0.1152, + "step": 15619 + }, + { + "epoch": 7.261738726173872, + "grad_norm": 5.440615653991699, + "learning_rate": 9.984386998438701e-06, + "loss": 0.0845, + "step": 15620 + }, + { + "epoch": 7.262203626220362, + "grad_norm": 2.9908652305603027, + "learning_rate": 9.9843859984386e-06, + "loss": 0.0625, + "step": 15621 + }, + { + "epoch": 7.262668526266853, + "grad_norm": 2.9099056720733643, + "learning_rate": 9.984384998438501e-06, + "loss": 0.0685, + "step": 15622 + }, + { + "epoch": 7.263133426313343, + "grad_norm": 3.1497154235839844, + "learning_rate": 9.9843839984384e-06, + "loss": 0.0459, + "step": 15623 + }, + { + "epoch": 7.2635983263598325, + "grad_norm": 5.854000568389893, + "learning_rate": 9.9843829984383e-06, + "loss": 0.2207, + "step": 15624 + }, + { + "epoch": 7.264063226406322, + "grad_norm": 5.6924967765808105, + "learning_rate": 9.984381998438201e-06, + "loss": 0.1768, + "step": 15625 + }, + { + "epoch": 7.264528126452813, + "grad_norm": 5.076308250427246, + "learning_rate": 9.9843809984381e-06, + "loss": 0.1114, + "step": 15626 + }, + { + "epoch": 7.264993026499303, + "grad_norm": 11.74567699432373, + "learning_rate": 9.984379998438001e-06, + "loss": 0.1462, + "step": 15627 + }, + { + "epoch": 7.265457926545793, + "grad_norm": 6.55424165725708, + "learning_rate": 9.9843789984379e-06, + "loss": 0.1526, + "step": 15628 + }, + { + "epoch": 7.265922826592282, + "grad_norm": 2.6548426151275635, + "learning_rate": 9.9843779984378e-06, + "loss": 0.0739, + "step": 15629 + }, + { + "epoch": 7.266387726638772, + "grad_norm": 2.465117931365967, + "learning_rate": 9.984376998437701e-06, + "loss": 0.046, + "step": 15630 + }, + { + "epoch": 7.266852626685263, + "grad_norm": 3.3591980934143066, + "learning_rate": 9.9843759984376e-06, + "loss": 0.0358, + "step": 15631 + }, + { + "epoch": 7.267317526731753, + "grad_norm": 5.868247985839844, + "learning_rate": 9.9843749984375e-06, + "loss": 0.1814, + "step": 15632 + }, + { + "epoch": 7.2677824267782425, + "grad_norm": 7.610766410827637, + "learning_rate": 9.984373998437401e-06, + "loss": 0.3364, + "step": 15633 + }, + { + "epoch": 7.268247326824732, + "grad_norm": 4.0215654373168945, + "learning_rate": 9.9843729984373e-06, + "loss": 0.1378, + "step": 15634 + }, + { + "epoch": 7.268712226871223, + "grad_norm": 9.886689186096191, + "learning_rate": 9.984371998437201e-06, + "loss": 0.2322, + "step": 15635 + }, + { + "epoch": 7.269177126917713, + "grad_norm": 3.4057135581970215, + "learning_rate": 9.9843709984371e-06, + "loss": 0.0571, + "step": 15636 + }, + { + "epoch": 7.269642026964203, + "grad_norm": 4.719091415405273, + "learning_rate": 9.984369998437e-06, + "loss": 0.0725, + "step": 15637 + }, + { + "epoch": 7.270106927010692, + "grad_norm": 6.735379695892334, + "learning_rate": 9.984368998436901e-06, + "loss": 0.2193, + "step": 15638 + }, + { + "epoch": 7.270571827057183, + "grad_norm": 12.158515930175781, + "learning_rate": 9.9843679984368e-06, + "loss": 0.1945, + "step": 15639 + }, + { + "epoch": 7.271036727103673, + "grad_norm": 5.369568347930908, + "learning_rate": 9.984366998436701e-06, + "loss": 0.2204, + "step": 15640 + }, + { + "epoch": 7.271501627150163, + "grad_norm": 5.813526630401611, + "learning_rate": 9.9843659984366e-06, + "loss": 0.1612, + "step": 15641 + }, + { + "epoch": 7.2719665271966525, + "grad_norm": 4.05733585357666, + "learning_rate": 9.9843649984365e-06, + "loss": 0.0618, + "step": 15642 + }, + { + "epoch": 7.272431427243143, + "grad_norm": 5.007638454437256, + "learning_rate": 9.984363998436401e-06, + "loss": 0.2013, + "step": 15643 + }, + { + "epoch": 7.272896327289633, + "grad_norm": 7.681268215179443, + "learning_rate": 9.984362998436302e-06, + "loss": 0.2446, + "step": 15644 + }, + { + "epoch": 7.273361227336123, + "grad_norm": 4.579587459564209, + "learning_rate": 9.984361998436201e-06, + "loss": 0.0983, + "step": 15645 + }, + { + "epoch": 7.273826127382613, + "grad_norm": 3.7040655612945557, + "learning_rate": 9.984360998436102e-06, + "loss": 0.0799, + "step": 15646 + }, + { + "epoch": 7.274291027429102, + "grad_norm": 4.093618869781494, + "learning_rate": 9.984359998436e-06, + "loss": 0.0794, + "step": 15647 + }, + { + "epoch": 7.274755927475593, + "grad_norm": 4.634353160858154, + "learning_rate": 9.9843589984359e-06, + "loss": 0.0789, + "step": 15648 + }, + { + "epoch": 7.275220827522083, + "grad_norm": 3.6339142322540283, + "learning_rate": 9.9843579984358e-06, + "loss": 0.1428, + "step": 15649 + }, + { + "epoch": 7.275685727568573, + "grad_norm": 4.561561584472656, + "learning_rate": 9.9843569984357e-06, + "loss": 0.1707, + "step": 15650 + }, + { + "epoch": 7.2761506276150625, + "grad_norm": 10.839714050292969, + "learning_rate": 9.984355998435601e-06, + "loss": 0.1294, + "step": 15651 + }, + { + "epoch": 7.276615527661553, + "grad_norm": 5.6264328956604, + "learning_rate": 9.9843549984355e-06, + "loss": 0.0882, + "step": 15652 + }, + { + "epoch": 7.277080427708043, + "grad_norm": 5.402394771575928, + "learning_rate": 9.984353998435401e-06, + "loss": 0.1459, + "step": 15653 + }, + { + "epoch": 7.277545327754533, + "grad_norm": 2.419858932495117, + "learning_rate": 9.9843529984353e-06, + "loss": 0.0422, + "step": 15654 + }, + { + "epoch": 7.278010227801023, + "grad_norm": 2.6051957607269287, + "learning_rate": 9.9843519984352e-06, + "loss": 0.0286, + "step": 15655 + }, + { + "epoch": 7.278475127847512, + "grad_norm": 5.05732536315918, + "learning_rate": 9.984350998435101e-06, + "loss": 0.1089, + "step": 15656 + }, + { + "epoch": 7.278940027894003, + "grad_norm": 5.781986713409424, + "learning_rate": 9.984349998435002e-06, + "loss": 0.1341, + "step": 15657 + }, + { + "epoch": 7.279404927940493, + "grad_norm": 4.425254821777344, + "learning_rate": 9.984348998434901e-06, + "loss": 0.1508, + "step": 15658 + }, + { + "epoch": 7.279869827986983, + "grad_norm": 2.3977513313293457, + "learning_rate": 9.9843479984348e-06, + "loss": 0.0399, + "step": 15659 + }, + { + "epoch": 7.2803347280334725, + "grad_norm": 4.050374984741211, + "learning_rate": 9.9843469984347e-06, + "loss": 0.0799, + "step": 15660 + }, + { + "epoch": 7.280799628079963, + "grad_norm": 7.1396894454956055, + "learning_rate": 9.9843459984346e-06, + "loss": 0.106, + "step": 15661 + }, + { + "epoch": 7.281264528126453, + "grad_norm": 3.423435688018799, + "learning_rate": 9.9843449984345e-06, + "loss": 0.0549, + "step": 15662 + }, + { + "epoch": 7.281729428172943, + "grad_norm": 5.3777360916137695, + "learning_rate": 9.9843439984344e-06, + "loss": 0.1782, + "step": 15663 + }, + { + "epoch": 7.282194328219433, + "grad_norm": 4.734966278076172, + "learning_rate": 9.984342998434302e-06, + "loss": 0.1481, + "step": 15664 + }, + { + "epoch": 7.282659228265922, + "grad_norm": 4.256595134735107, + "learning_rate": 9.9843419984342e-06, + "loss": 0.1552, + "step": 15665 + }, + { + "epoch": 7.283124128312413, + "grad_norm": 5.519396781921387, + "learning_rate": 9.984340998434101e-06, + "loss": 0.0888, + "step": 15666 + }, + { + "epoch": 7.283589028358903, + "grad_norm": 6.3139448165893555, + "learning_rate": 9.984339998434e-06, + "loss": 0.1073, + "step": 15667 + }, + { + "epoch": 7.284053928405393, + "grad_norm": 7.21607780456543, + "learning_rate": 9.9843389984339e-06, + "loss": 0.1815, + "step": 15668 + }, + { + "epoch": 7.2845188284518825, + "grad_norm": 5.13456392288208, + "learning_rate": 9.984337998433801e-06, + "loss": 0.137, + "step": 15669 + }, + { + "epoch": 7.284983728498373, + "grad_norm": 3.225271224975586, + "learning_rate": 9.9843369984337e-06, + "loss": 0.0464, + "step": 15670 + }, + { + "epoch": 7.285448628544863, + "grad_norm": 9.447509765625, + "learning_rate": 9.984335998433601e-06, + "loss": 0.2482, + "step": 15671 + }, + { + "epoch": 7.285913528591353, + "grad_norm": 6.0129828453063965, + "learning_rate": 9.9843349984335e-06, + "loss": 0.103, + "step": 15672 + }, + { + "epoch": 7.286378428637843, + "grad_norm": 2.8860037326812744, + "learning_rate": 9.9843339984334e-06, + "loss": 0.0612, + "step": 15673 + }, + { + "epoch": 7.286843328684333, + "grad_norm": 5.422711372375488, + "learning_rate": 9.9843329984333e-06, + "loss": 0.1643, + "step": 15674 + }, + { + "epoch": 7.287308228730823, + "grad_norm": 6.454352855682373, + "learning_rate": 9.9843319984332e-06, + "loss": 0.201, + "step": 15675 + }, + { + "epoch": 7.287773128777313, + "grad_norm": 6.57066011428833, + "learning_rate": 9.984330998433101e-06, + "loss": 0.0939, + "step": 15676 + }, + { + "epoch": 7.288238028823803, + "grad_norm": 1.958606243133545, + "learning_rate": 9.984329998433002e-06, + "loss": 0.0261, + "step": 15677 + }, + { + "epoch": 7.2887029288702925, + "grad_norm": 6.632012367248535, + "learning_rate": 9.9843289984329e-06, + "loss": 0.2399, + "step": 15678 + }, + { + "epoch": 7.289167828916783, + "grad_norm": 8.043855667114258, + "learning_rate": 9.984327998432801e-06, + "loss": 0.1631, + "step": 15679 + }, + { + "epoch": 7.289632728963273, + "grad_norm": 2.680222272872925, + "learning_rate": 9.9843269984327e-06, + "loss": 0.0498, + "step": 15680 + }, + { + "epoch": 7.290097629009763, + "grad_norm": 6.144375801086426, + "learning_rate": 9.9843259984326e-06, + "loss": 0.1797, + "step": 15681 + }, + { + "epoch": 7.290562529056253, + "grad_norm": 5.646681785583496, + "learning_rate": 9.984324998432501e-06, + "loss": 0.1161, + "step": 15682 + }, + { + "epoch": 7.291027429102743, + "grad_norm": 3.1842188835144043, + "learning_rate": 9.9843239984324e-06, + "loss": 0.0633, + "step": 15683 + }, + { + "epoch": 7.291492329149233, + "grad_norm": 3.3845198154449463, + "learning_rate": 9.984322998432301e-06, + "loss": 0.0527, + "step": 15684 + }, + { + "epoch": 7.291957229195723, + "grad_norm": 4.234490871429443, + "learning_rate": 9.9843219984322e-06, + "loss": 0.1135, + "step": 15685 + }, + { + "epoch": 7.292422129242213, + "grad_norm": 7.084415912628174, + "learning_rate": 9.9843209984321e-06, + "loss": 0.129, + "step": 15686 + }, + { + "epoch": 7.292887029288703, + "grad_norm": 5.770450115203857, + "learning_rate": 9.984319998432e-06, + "loss": 0.1342, + "step": 15687 + }, + { + "epoch": 7.293351929335193, + "grad_norm": 3.080634832382202, + "learning_rate": 9.9843189984319e-06, + "loss": 0.053, + "step": 15688 + }, + { + "epoch": 7.293816829381683, + "grad_norm": 7.277524948120117, + "learning_rate": 9.984317998431801e-06, + "loss": 0.1472, + "step": 15689 + }, + { + "epoch": 7.294281729428173, + "grad_norm": 4.942774772644043, + "learning_rate": 9.984316998431702e-06, + "loss": 0.1253, + "step": 15690 + }, + { + "epoch": 7.294746629474663, + "grad_norm": 15.045439720153809, + "learning_rate": 9.9843159984316e-06, + "loss": 0.4058, + "step": 15691 + }, + { + "epoch": 7.295211529521153, + "grad_norm": 7.735775947570801, + "learning_rate": 9.9843149984315e-06, + "loss": 0.2476, + "step": 15692 + }, + { + "epoch": 7.295676429567643, + "grad_norm": 5.3777594566345215, + "learning_rate": 9.9843139984314e-06, + "loss": 0.0856, + "step": 15693 + }, + { + "epoch": 7.296141329614133, + "grad_norm": 5.1455607414245605, + "learning_rate": 9.984312998431301e-06, + "loss": 0.143, + "step": 15694 + }, + { + "epoch": 7.296606229660623, + "grad_norm": 8.772788047790527, + "learning_rate": 9.984311998431202e-06, + "loss": 0.2069, + "step": 15695 + }, + { + "epoch": 7.297071129707113, + "grad_norm": 5.279211044311523, + "learning_rate": 9.9843109984311e-06, + "loss": 0.125, + "step": 15696 + }, + { + "epoch": 7.297536029753603, + "grad_norm": 4.787064552307129, + "learning_rate": 9.984309998431001e-06, + "loss": 0.1466, + "step": 15697 + }, + { + "epoch": 7.298000929800093, + "grad_norm": 4.453113079071045, + "learning_rate": 9.9843089984309e-06, + "loss": 0.0845, + "step": 15698 + }, + { + "epoch": 7.298465829846583, + "grad_norm": 3.692913770675659, + "learning_rate": 9.9843079984308e-06, + "loss": 0.1053, + "step": 15699 + }, + { + "epoch": 7.298930729893073, + "grad_norm": 13.4612455368042, + "learning_rate": 9.9843069984307e-06, + "loss": 0.1555, + "step": 15700 + }, + { + "epoch": 7.299395629939563, + "grad_norm": 5.313905715942383, + "learning_rate": 9.9843059984306e-06, + "loss": 0.0965, + "step": 15701 + }, + { + "epoch": 7.299860529986053, + "grad_norm": 2.0032010078430176, + "learning_rate": 9.984304998430501e-06, + "loss": 0.0292, + "step": 15702 + }, + { + "epoch": 7.300325430032543, + "grad_norm": 4.593167304992676, + "learning_rate": 9.9843039984304e-06, + "loss": 0.1515, + "step": 15703 + }, + { + "epoch": 7.300790330079033, + "grad_norm": 7.8297038078308105, + "learning_rate": 9.9843029984303e-06, + "loss": 0.19, + "step": 15704 + }, + { + "epoch": 7.301255230125523, + "grad_norm": 4.42351770401001, + "learning_rate": 9.9843019984302e-06, + "loss": 0.0628, + "step": 15705 + }, + { + "epoch": 7.301720130172013, + "grad_norm": 6.338405132293701, + "learning_rate": 9.9843009984301e-06, + "loss": 0.1647, + "step": 15706 + }, + { + "epoch": 7.302185030218503, + "grad_norm": 6.9178385734558105, + "learning_rate": 9.984299998430001e-06, + "loss": 0.1842, + "step": 15707 + }, + { + "epoch": 7.302649930264993, + "grad_norm": 2.9270942211151123, + "learning_rate": 9.984298998429902e-06, + "loss": 0.0546, + "step": 15708 + }, + { + "epoch": 7.303114830311483, + "grad_norm": 2.50747013092041, + "learning_rate": 9.9842979984298e-06, + "loss": 0.0676, + "step": 15709 + }, + { + "epoch": 7.303579730357973, + "grad_norm": 5.407525062561035, + "learning_rate": 9.984296998429701e-06, + "loss": 0.1293, + "step": 15710 + }, + { + "epoch": 7.304044630404463, + "grad_norm": 6.231155872344971, + "learning_rate": 9.9842959984296e-06, + "loss": 0.1425, + "step": 15711 + }, + { + "epoch": 7.304509530450953, + "grad_norm": 7.81311559677124, + "learning_rate": 9.984294998429501e-06, + "loss": 0.1982, + "step": 15712 + }, + { + "epoch": 7.304974430497443, + "grad_norm": 14.599933624267578, + "learning_rate": 9.984293998429402e-06, + "loss": 0.193, + "step": 15713 + }, + { + "epoch": 7.305439330543933, + "grad_norm": 3.9617760181427, + "learning_rate": 9.9842929984293e-06, + "loss": 0.0688, + "step": 15714 + }, + { + "epoch": 7.305904230590423, + "grad_norm": 4.222359657287598, + "learning_rate": 9.984291998429201e-06, + "loss": 0.0764, + "step": 15715 + }, + { + "epoch": 7.306369130636913, + "grad_norm": 3.447540760040283, + "learning_rate": 9.9842909984291e-06, + "loss": 0.0524, + "step": 15716 + }, + { + "epoch": 7.306834030683403, + "grad_norm": 8.382781982421875, + "learning_rate": 9.984289998429e-06, + "loss": 0.1426, + "step": 15717 + }, + { + "epoch": 7.3072989307298934, + "grad_norm": 3.2637174129486084, + "learning_rate": 9.9842889984289e-06, + "loss": 0.0675, + "step": 15718 + }, + { + "epoch": 7.307763830776383, + "grad_norm": 8.484274864196777, + "learning_rate": 9.9842879984288e-06, + "loss": 0.1964, + "step": 15719 + }, + { + "epoch": 7.308228730822873, + "grad_norm": 5.280543327331543, + "learning_rate": 9.984286998428701e-06, + "loss": 0.0827, + "step": 15720 + }, + { + "epoch": 7.308693630869363, + "grad_norm": 4.808394908905029, + "learning_rate": 9.984285998428602e-06, + "loss": 0.1167, + "step": 15721 + }, + { + "epoch": 7.3091585309158535, + "grad_norm": 6.618891716003418, + "learning_rate": 9.9842849984285e-06, + "loss": 0.2761, + "step": 15722 + }, + { + "epoch": 7.309623430962343, + "grad_norm": 4.153401851654053, + "learning_rate": 9.9842839984284e-06, + "loss": 0.0793, + "step": 15723 + }, + { + "epoch": 7.310088331008833, + "grad_norm": 3.7983644008636475, + "learning_rate": 9.9842829984283e-06, + "loss": 0.0654, + "step": 15724 + }, + { + "epoch": 7.310553231055323, + "grad_norm": 10.67743968963623, + "learning_rate": 9.984281998428201e-06, + "loss": 0.2064, + "step": 15725 + }, + { + "epoch": 7.311018131101813, + "grad_norm": 3.7330939769744873, + "learning_rate": 9.984280998428102e-06, + "loss": 0.0631, + "step": 15726 + }, + { + "epoch": 7.3114830311483034, + "grad_norm": 6.810614585876465, + "learning_rate": 9.984279998428e-06, + "loss": 0.251, + "step": 15727 + }, + { + "epoch": 7.311947931194793, + "grad_norm": 7.189486503601074, + "learning_rate": 9.984278998427901e-06, + "loss": 0.2158, + "step": 15728 + }, + { + "epoch": 7.312412831241283, + "grad_norm": 14.158758163452148, + "learning_rate": 9.9842779984278e-06, + "loss": 0.1049, + "step": 15729 + }, + { + "epoch": 7.312877731287773, + "grad_norm": 1.614987850189209, + "learning_rate": 9.984276998427701e-06, + "loss": 0.0233, + "step": 15730 + }, + { + "epoch": 7.3133426313342635, + "grad_norm": 4.6151041984558105, + "learning_rate": 9.9842759984276e-06, + "loss": 0.11, + "step": 15731 + }, + { + "epoch": 7.313807531380753, + "grad_norm": 4.092373371124268, + "learning_rate": 9.9842749984275e-06, + "loss": 0.0498, + "step": 15732 + }, + { + "epoch": 7.314272431427243, + "grad_norm": 4.19836950302124, + "learning_rate": 9.984273998427401e-06, + "loss": 0.0844, + "step": 15733 + }, + { + "epoch": 7.314737331473733, + "grad_norm": 5.355379581451416, + "learning_rate": 9.9842729984273e-06, + "loss": 0.1078, + "step": 15734 + }, + { + "epoch": 7.315202231520223, + "grad_norm": 10.749079704284668, + "learning_rate": 9.9842719984272e-06, + "loss": 0.2086, + "step": 15735 + }, + { + "epoch": 7.315667131566713, + "grad_norm": 9.197213172912598, + "learning_rate": 9.9842709984271e-06, + "loss": 0.3384, + "step": 15736 + }, + { + "epoch": 7.316132031613203, + "grad_norm": 3.7748098373413086, + "learning_rate": 9.984269998427e-06, + "loss": 0.101, + "step": 15737 + }, + { + "epoch": 7.316596931659693, + "grad_norm": 5.557522296905518, + "learning_rate": 9.984268998426901e-06, + "loss": 0.102, + "step": 15738 + }, + { + "epoch": 7.317061831706183, + "grad_norm": 5.8030524253845215, + "learning_rate": 9.984267998426802e-06, + "loss": 0.1386, + "step": 15739 + }, + { + "epoch": 7.3175267317526735, + "grad_norm": 4.92950963973999, + "learning_rate": 9.9842669984267e-06, + "loss": 0.1218, + "step": 15740 + }, + { + "epoch": 7.317991631799163, + "grad_norm": 11.05687141418457, + "learning_rate": 9.984265998426601e-06, + "loss": 0.2051, + "step": 15741 + }, + { + "epoch": 7.318456531845653, + "grad_norm": 6.65558385848999, + "learning_rate": 9.9842649984265e-06, + "loss": 0.266, + "step": 15742 + }, + { + "epoch": 7.318921431892143, + "grad_norm": 3.4773995876312256, + "learning_rate": 9.984263998426401e-06, + "loss": 0.0531, + "step": 15743 + }, + { + "epoch": 7.319386331938633, + "grad_norm": 5.287954807281494, + "learning_rate": 9.9842629984263e-06, + "loss": 0.0881, + "step": 15744 + }, + { + "epoch": 7.319851231985123, + "grad_norm": 4.012388229370117, + "learning_rate": 9.9842619984262e-06, + "loss": 0.0645, + "step": 15745 + }, + { + "epoch": 7.320316132031613, + "grad_norm": 4.130242824554443, + "learning_rate": 9.984260998426101e-06, + "loss": 0.0543, + "step": 15746 + }, + { + "epoch": 7.320781032078103, + "grad_norm": 4.800066947937012, + "learning_rate": 9.984259998426e-06, + "loss": 0.1621, + "step": 15747 + }, + { + "epoch": 7.321245932124593, + "grad_norm": 4.245713710784912, + "learning_rate": 9.984258998425901e-06, + "loss": 0.0593, + "step": 15748 + }, + { + "epoch": 7.3217108321710835, + "grad_norm": 6.259105205535889, + "learning_rate": 9.9842579984258e-06, + "loss": 0.2066, + "step": 15749 + }, + { + "epoch": 7.322175732217573, + "grad_norm": 4.410625457763672, + "learning_rate": 9.9842569984257e-06, + "loss": 0.0724, + "step": 15750 + }, + { + "epoch": 7.322640632264063, + "grad_norm": 7.135679244995117, + "learning_rate": 9.984255998425601e-06, + "loss": 0.1438, + "step": 15751 + }, + { + "epoch": 7.323105532310553, + "grad_norm": 6.9652791023254395, + "learning_rate": 9.984254998425502e-06, + "loss": 0.1494, + "step": 15752 + }, + { + "epoch": 7.323570432357044, + "grad_norm": 6.7039337158203125, + "learning_rate": 9.9842539984254e-06, + "loss": 0.1516, + "step": 15753 + }, + { + "epoch": 7.324035332403533, + "grad_norm": 7.827138423919678, + "learning_rate": 9.984252998425301e-06, + "loss": 0.1592, + "step": 15754 + }, + { + "epoch": 7.324500232450023, + "grad_norm": 7.242160320281982, + "learning_rate": 9.9842519984252e-06, + "loss": 0.1841, + "step": 15755 + }, + { + "epoch": 7.324965132496513, + "grad_norm": 3.070909023284912, + "learning_rate": 9.9842509984251e-06, + "loss": 0.0643, + "step": 15756 + }, + { + "epoch": 7.325430032543003, + "grad_norm": 2.9582386016845703, + "learning_rate": 9.984249998425e-06, + "loss": 0.0808, + "step": 15757 + }, + { + "epoch": 7.3258949325894935, + "grad_norm": 2.9808478355407715, + "learning_rate": 9.9842489984249e-06, + "loss": 0.0405, + "step": 15758 + }, + { + "epoch": 7.326359832635983, + "grad_norm": 1.7111576795578003, + "learning_rate": 9.984247998424801e-06, + "loss": 0.0306, + "step": 15759 + }, + { + "epoch": 7.326824732682473, + "grad_norm": 7.629878044128418, + "learning_rate": 9.9842469984247e-06, + "loss": 0.3201, + "step": 15760 + }, + { + "epoch": 7.327289632728963, + "grad_norm": 4.866611957550049, + "learning_rate": 9.984245998424601e-06, + "loss": 0.0874, + "step": 15761 + }, + { + "epoch": 7.327754532775454, + "grad_norm": 4.849971771240234, + "learning_rate": 9.9842449984245e-06, + "loss": 0.132, + "step": 15762 + }, + { + "epoch": 7.328219432821943, + "grad_norm": 6.825715065002441, + "learning_rate": 9.9842439984244e-06, + "loss": 0.3481, + "step": 15763 + }, + { + "epoch": 7.328684332868433, + "grad_norm": 2.567624092102051, + "learning_rate": 9.984242998424301e-06, + "loss": 0.0141, + "step": 15764 + }, + { + "epoch": 7.329149232914923, + "grad_norm": 3.6940348148345947, + "learning_rate": 9.984241998424202e-06, + "loss": 0.0516, + "step": 15765 + }, + { + "epoch": 7.329614132961414, + "grad_norm": 8.190972328186035, + "learning_rate": 9.9842409984241e-06, + "loss": 0.1584, + "step": 15766 + }, + { + "epoch": 7.3300790330079035, + "grad_norm": 14.683136940002441, + "learning_rate": 9.984239998424e-06, + "loss": 0.1115, + "step": 15767 + }, + { + "epoch": 7.330543933054393, + "grad_norm": 4.275065898895264, + "learning_rate": 9.9842389984239e-06, + "loss": 0.0896, + "step": 15768 + }, + { + "epoch": 7.331008833100883, + "grad_norm": 6.697376251220703, + "learning_rate": 9.9842379984238e-06, + "loss": 0.2629, + "step": 15769 + }, + { + "epoch": 7.331473733147373, + "grad_norm": 5.206666946411133, + "learning_rate": 9.9842369984237e-06, + "loss": 0.1231, + "step": 15770 + }, + { + "epoch": 7.331938633193864, + "grad_norm": 4.2078962326049805, + "learning_rate": 9.9842359984236e-06, + "loss": 0.1001, + "step": 15771 + }, + { + "epoch": 7.332403533240353, + "grad_norm": 8.077452659606934, + "learning_rate": 9.984234998423501e-06, + "loss": 0.2001, + "step": 15772 + }, + { + "epoch": 7.332868433286843, + "grad_norm": 2.7622809410095215, + "learning_rate": 9.9842339984234e-06, + "loss": 0.0407, + "step": 15773 + }, + { + "epoch": 7.333333333333333, + "grad_norm": 6.329754829406738, + "learning_rate": 9.984232998423301e-06, + "loss": 0.2397, + "step": 15774 + }, + { + "epoch": 7.333798233379824, + "grad_norm": 7.495077610015869, + "learning_rate": 9.9842319984232e-06, + "loss": 0.187, + "step": 15775 + }, + { + "epoch": 7.3342631334263135, + "grad_norm": 4.1526618003845215, + "learning_rate": 9.9842309984231e-06, + "loss": 0.0798, + "step": 15776 + }, + { + "epoch": 7.334728033472803, + "grad_norm": 6.588744163513184, + "learning_rate": 9.984229998423001e-06, + "loss": 0.1785, + "step": 15777 + }, + { + "epoch": 7.335192933519293, + "grad_norm": 7.244582653045654, + "learning_rate": 9.9842289984229e-06, + "loss": 0.0748, + "step": 15778 + }, + { + "epoch": 7.335657833565783, + "grad_norm": 7.955648899078369, + "learning_rate": 9.984227998422801e-06, + "loss": 0.2725, + "step": 15779 + }, + { + "epoch": 7.336122733612274, + "grad_norm": 6.5585174560546875, + "learning_rate": 9.9842269984227e-06, + "loss": 0.219, + "step": 15780 + }, + { + "epoch": 7.336587633658763, + "grad_norm": 6.3328728675842285, + "learning_rate": 9.9842259984226e-06, + "loss": 0.1213, + "step": 15781 + }, + { + "epoch": 7.337052533705253, + "grad_norm": 7.814780235290527, + "learning_rate": 9.984224998422501e-06, + "loss": 0.1485, + "step": 15782 + }, + { + "epoch": 7.337517433751743, + "grad_norm": 2.360529661178589, + "learning_rate": 9.9842239984224e-06, + "loss": 0.0239, + "step": 15783 + }, + { + "epoch": 7.337982333798234, + "grad_norm": 7.994659423828125, + "learning_rate": 9.9842229984223e-06, + "loss": 0.3173, + "step": 15784 + }, + { + "epoch": 7.3384472338447235, + "grad_norm": 6.088622093200684, + "learning_rate": 9.984221998422201e-06, + "loss": 0.1439, + "step": 15785 + }, + { + "epoch": 7.338912133891213, + "grad_norm": 6.1828532218933105, + "learning_rate": 9.9842209984221e-06, + "loss": 0.1503, + "step": 15786 + }, + { + "epoch": 7.339377033937703, + "grad_norm": 6.111666679382324, + "learning_rate": 9.984219998422001e-06, + "loss": 0.2139, + "step": 15787 + }, + { + "epoch": 7.339841933984193, + "grad_norm": 3.948333263397217, + "learning_rate": 9.9842189984219e-06, + "loss": 0.0631, + "step": 15788 + }, + { + "epoch": 7.340306834030684, + "grad_norm": 5.89371919631958, + "learning_rate": 9.9842179984218e-06, + "loss": 0.2903, + "step": 15789 + }, + { + "epoch": 7.340771734077173, + "grad_norm": 4.054599761962891, + "learning_rate": 9.984216998421701e-06, + "loss": 0.0779, + "step": 15790 + }, + { + "epoch": 7.341236634123663, + "grad_norm": 8.760066986083984, + "learning_rate": 9.9842159984216e-06, + "loss": 0.3558, + "step": 15791 + }, + { + "epoch": 7.341701534170153, + "grad_norm": 14.96368408203125, + "learning_rate": 9.984214998421501e-06, + "loss": 0.0669, + "step": 15792 + }, + { + "epoch": 7.342166434216644, + "grad_norm": 3.8832345008850098, + "learning_rate": 9.9842139984214e-06, + "loss": 0.0633, + "step": 15793 + }, + { + "epoch": 7.3426313342631335, + "grad_norm": 4.444070339202881, + "learning_rate": 9.9842129984213e-06, + "loss": 0.1548, + "step": 15794 + }, + { + "epoch": 7.343096234309623, + "grad_norm": 4.330501556396484, + "learning_rate": 9.984211998421201e-06, + "loss": 0.0608, + "step": 15795 + }, + { + "epoch": 7.343561134356113, + "grad_norm": 4.413910865783691, + "learning_rate": 9.9842109984211e-06, + "loss": 0.0877, + "step": 15796 + }, + { + "epoch": 7.344026034402604, + "grad_norm": 3.5009372234344482, + "learning_rate": 9.984209998421001e-06, + "loss": 0.0597, + "step": 15797 + }, + { + "epoch": 7.344490934449094, + "grad_norm": 4.429891586303711, + "learning_rate": 9.984208998420902e-06, + "loss": 0.1233, + "step": 15798 + }, + { + "epoch": 7.344955834495583, + "grad_norm": 9.509303092956543, + "learning_rate": 9.9842079984208e-06, + "loss": 0.3019, + "step": 15799 + }, + { + "epoch": 7.345420734542073, + "grad_norm": 4.275379657745361, + "learning_rate": 9.9842069984207e-06, + "loss": 0.1489, + "step": 15800 + }, + { + "epoch": 7.345885634588564, + "grad_norm": 6.123692035675049, + "learning_rate": 9.9842059984206e-06, + "loss": 0.2463, + "step": 15801 + }, + { + "epoch": 7.346350534635054, + "grad_norm": 7.64351224899292, + "learning_rate": 9.9842049984205e-06, + "loss": 0.279, + "step": 15802 + }, + { + "epoch": 7.3468154346815435, + "grad_norm": 6.830073833465576, + "learning_rate": 9.984203998420401e-06, + "loss": 0.2345, + "step": 15803 + }, + { + "epoch": 7.347280334728033, + "grad_norm": 4.822045803070068, + "learning_rate": 9.9842029984203e-06, + "loss": 0.1206, + "step": 15804 + }, + { + "epoch": 7.347745234774523, + "grad_norm": 7.481160640716553, + "learning_rate": 9.984201998420201e-06, + "loss": 0.1647, + "step": 15805 + }, + { + "epoch": 7.348210134821014, + "grad_norm": 4.60433292388916, + "learning_rate": 9.9842009984201e-06, + "loss": 0.0847, + "step": 15806 + }, + { + "epoch": 7.348675034867504, + "grad_norm": 4.5565595626831055, + "learning_rate": 9.98419999842e-06, + "loss": 0.1808, + "step": 15807 + }, + { + "epoch": 7.349139934913993, + "grad_norm": 6.970898628234863, + "learning_rate": 9.984198998419901e-06, + "loss": 0.1213, + "step": 15808 + }, + { + "epoch": 7.349604834960483, + "grad_norm": 7.230710029602051, + "learning_rate": 9.984197998419802e-06, + "loss": 0.2599, + "step": 15809 + }, + { + "epoch": 7.350069735006974, + "grad_norm": 6.322710037231445, + "learning_rate": 9.984196998419701e-06, + "loss": 0.1231, + "step": 15810 + }, + { + "epoch": 7.350534635053464, + "grad_norm": 4.716108798980713, + "learning_rate": 9.9841959984196e-06, + "loss": 0.128, + "step": 15811 + }, + { + "epoch": 7.3509995350999535, + "grad_norm": 4.379919528961182, + "learning_rate": 9.9841949984195e-06, + "loss": 0.0891, + "step": 15812 + }, + { + "epoch": 7.351464435146443, + "grad_norm": 4.317760944366455, + "learning_rate": 9.9841939984194e-06, + "loss": 0.0827, + "step": 15813 + }, + { + "epoch": 7.351929335192933, + "grad_norm": 7.041932106018066, + "learning_rate": 9.9841929984193e-06, + "loss": 0.2652, + "step": 15814 + }, + { + "epoch": 7.352394235239424, + "grad_norm": 3.7971975803375244, + "learning_rate": 9.984191998419201e-06, + "loss": 0.0854, + "step": 15815 + }, + { + "epoch": 7.352859135285914, + "grad_norm": 4.626598834991455, + "learning_rate": 9.984190998419102e-06, + "loss": 0.1168, + "step": 15816 + }, + { + "epoch": 7.353324035332403, + "grad_norm": 5.453516960144043, + "learning_rate": 9.984189998419e-06, + "loss": 0.1647, + "step": 15817 + }, + { + "epoch": 7.353788935378893, + "grad_norm": 3.0657267570495605, + "learning_rate": 9.984188998418901e-06, + "loss": 0.0427, + "step": 15818 + }, + { + "epoch": 7.354253835425384, + "grad_norm": 2.88704776763916, + "learning_rate": 9.9841879984188e-06, + "loss": 0.0449, + "step": 15819 + }, + { + "epoch": 7.354718735471874, + "grad_norm": 4.601713180541992, + "learning_rate": 9.9841869984187e-06, + "loss": 0.1281, + "step": 15820 + }, + { + "epoch": 7.3551836355183635, + "grad_norm": 6.153409957885742, + "learning_rate": 9.984185998418601e-06, + "loss": 0.1436, + "step": 15821 + }, + { + "epoch": 7.355648535564853, + "grad_norm": 4.024639129638672, + "learning_rate": 9.9841849984185e-06, + "loss": 0.1052, + "step": 15822 + }, + { + "epoch": 7.356113435611343, + "grad_norm": 6.576153755187988, + "learning_rate": 9.984183998418401e-06, + "loss": 0.2437, + "step": 15823 + }, + { + "epoch": 7.356578335657834, + "grad_norm": 5.166604995727539, + "learning_rate": 9.9841829984183e-06, + "loss": 0.1149, + "step": 15824 + }, + { + "epoch": 7.357043235704324, + "grad_norm": 4.681131839752197, + "learning_rate": 9.9841819984182e-06, + "loss": 0.0978, + "step": 15825 + }, + { + "epoch": 7.357508135750813, + "grad_norm": 10.465364456176758, + "learning_rate": 9.9841809984181e-06, + "loss": 0.3536, + "step": 15826 + }, + { + "epoch": 7.357973035797303, + "grad_norm": 3.6109282970428467, + "learning_rate": 9.984179998418e-06, + "loss": 0.0882, + "step": 15827 + }, + { + "epoch": 7.358437935843794, + "grad_norm": 4.376956939697266, + "learning_rate": 9.984178998417901e-06, + "loss": 0.0806, + "step": 15828 + }, + { + "epoch": 7.358902835890284, + "grad_norm": 3.5811665058135986, + "learning_rate": 9.984177998417802e-06, + "loss": 0.0801, + "step": 15829 + }, + { + "epoch": 7.3593677359367735, + "grad_norm": 2.684987783432007, + "learning_rate": 9.9841769984177e-06, + "loss": 0.0567, + "step": 15830 + }, + { + "epoch": 7.359832635983263, + "grad_norm": 7.079136848449707, + "learning_rate": 9.984175998417601e-06, + "loss": 0.3411, + "step": 15831 + }, + { + "epoch": 7.360297536029754, + "grad_norm": 6.511275291442871, + "learning_rate": 9.9841749984175e-06, + "loss": 0.1565, + "step": 15832 + }, + { + "epoch": 7.360762436076244, + "grad_norm": 3.799204111099243, + "learning_rate": 9.984173998417401e-06, + "loss": 0.0839, + "step": 15833 + }, + { + "epoch": 7.361227336122734, + "grad_norm": 5.582200527191162, + "learning_rate": 9.984172998417302e-06, + "loss": 0.1032, + "step": 15834 + }, + { + "epoch": 7.361692236169223, + "grad_norm": 5.399205207824707, + "learning_rate": 9.9841719984172e-06, + "loss": 0.0436, + "step": 15835 + }, + { + "epoch": 7.362157136215713, + "grad_norm": 4.003139019012451, + "learning_rate": 9.984170998417101e-06, + "loss": 0.0571, + "step": 15836 + }, + { + "epoch": 7.362622036262204, + "grad_norm": 3.7511415481567383, + "learning_rate": 9.984169998417e-06, + "loss": 0.0509, + "step": 15837 + }, + { + "epoch": 7.363086936308694, + "grad_norm": 8.761034965515137, + "learning_rate": 9.9841689984169e-06, + "loss": 0.1813, + "step": 15838 + }, + { + "epoch": 7.3635518363551835, + "grad_norm": 3.0734004974365234, + "learning_rate": 9.9841679984168e-06, + "loss": 0.0389, + "step": 15839 + }, + { + "epoch": 7.364016736401673, + "grad_norm": 5.567323684692383, + "learning_rate": 9.9841669984167e-06, + "loss": 0.2329, + "step": 15840 + }, + { + "epoch": 7.364481636448164, + "grad_norm": 3.7454423904418945, + "learning_rate": 9.984165998416601e-06, + "loss": 0.0406, + "step": 15841 + }, + { + "epoch": 7.364946536494654, + "grad_norm": 2.8895576000213623, + "learning_rate": 9.9841649984165e-06, + "loss": 0.0295, + "step": 15842 + }, + { + "epoch": 7.365411436541144, + "grad_norm": 5.934459209442139, + "learning_rate": 9.9841639984164e-06, + "loss": 0.196, + "step": 15843 + }, + { + "epoch": 7.365876336587633, + "grad_norm": 3.388829469680786, + "learning_rate": 9.9841629984163e-06, + "loss": 0.0572, + "step": 15844 + }, + { + "epoch": 7.366341236634124, + "grad_norm": 4.03132963180542, + "learning_rate": 9.9841619984162e-06, + "loss": 0.0985, + "step": 15845 + }, + { + "epoch": 7.366806136680614, + "grad_norm": 3.0985324382781982, + "learning_rate": 9.984160998416101e-06, + "loss": 0.0749, + "step": 15846 + }, + { + "epoch": 7.367271036727104, + "grad_norm": 5.697320938110352, + "learning_rate": 9.984159998416002e-06, + "loss": 0.1475, + "step": 15847 + }, + { + "epoch": 7.3677359367735935, + "grad_norm": 4.871754169464111, + "learning_rate": 9.9841589984159e-06, + "loss": 0.097, + "step": 15848 + }, + { + "epoch": 7.368200836820083, + "grad_norm": 6.284806251525879, + "learning_rate": 9.984157998415801e-06, + "loss": 0.1786, + "step": 15849 + }, + { + "epoch": 7.368665736866574, + "grad_norm": 5.293858528137207, + "learning_rate": 9.9841569984157e-06, + "loss": 0.1088, + "step": 15850 + }, + { + "epoch": 7.369130636913064, + "grad_norm": 5.3066792488098145, + "learning_rate": 9.9841559984156e-06, + "loss": 0.1599, + "step": 15851 + }, + { + "epoch": 7.369595536959554, + "grad_norm": 3.9574806690216064, + "learning_rate": 9.9841549984155e-06, + "loss": 0.0701, + "step": 15852 + }, + { + "epoch": 7.370060437006043, + "grad_norm": 5.350255966186523, + "learning_rate": 9.9841539984154e-06, + "loss": 0.1295, + "step": 15853 + }, + { + "epoch": 7.370525337052534, + "grad_norm": 10.93124771118164, + "learning_rate": 9.984152998415301e-06, + "loss": 0.1615, + "step": 15854 + }, + { + "epoch": 7.370990237099024, + "grad_norm": 6.368402004241943, + "learning_rate": 9.9841519984152e-06, + "loss": 0.3016, + "step": 15855 + }, + { + "epoch": 7.371455137145514, + "grad_norm": 9.091700553894043, + "learning_rate": 9.9841509984151e-06, + "loss": 0.3583, + "step": 15856 + }, + { + "epoch": 7.3719200371920035, + "grad_norm": 6.626374244689941, + "learning_rate": 9.984149998415e-06, + "loss": 0.1668, + "step": 15857 + }, + { + "epoch": 7.372384937238493, + "grad_norm": 4.5522236824035645, + "learning_rate": 9.9841489984149e-06, + "loss": 0.0728, + "step": 15858 + }, + { + "epoch": 7.372849837284984, + "grad_norm": 9.368706703186035, + "learning_rate": 9.984147998414801e-06, + "loss": 0.2492, + "step": 15859 + }, + { + "epoch": 7.373314737331474, + "grad_norm": 4.874064922332764, + "learning_rate": 9.984146998414702e-06, + "loss": 0.1236, + "step": 15860 + }, + { + "epoch": 7.373779637377964, + "grad_norm": 5.885942459106445, + "learning_rate": 9.9841459984146e-06, + "loss": 0.1716, + "step": 15861 + }, + { + "epoch": 7.374244537424453, + "grad_norm": 7.814970016479492, + "learning_rate": 9.984144998414501e-06, + "loss": 0.1865, + "step": 15862 + }, + { + "epoch": 7.374709437470944, + "grad_norm": 4.455880165100098, + "learning_rate": 9.9841439984144e-06, + "loss": 0.124, + "step": 15863 + }, + { + "epoch": 7.375174337517434, + "grad_norm": 6.505396842956543, + "learning_rate": 9.984142998414301e-06, + "loss": 0.2507, + "step": 15864 + }, + { + "epoch": 7.375639237563924, + "grad_norm": 5.498602390289307, + "learning_rate": 9.9841419984142e-06, + "loss": 0.181, + "step": 15865 + }, + { + "epoch": 7.3761041376104135, + "grad_norm": 3.8272030353546143, + "learning_rate": 9.9841409984141e-06, + "loss": 0.1057, + "step": 15866 + }, + { + "epoch": 7.376569037656903, + "grad_norm": 6.148056983947754, + "learning_rate": 9.984139998414001e-06, + "loss": 0.2351, + "step": 15867 + }, + { + "epoch": 7.377033937703394, + "grad_norm": 6.760238170623779, + "learning_rate": 9.9841389984139e-06, + "loss": 0.1274, + "step": 15868 + }, + { + "epoch": 7.377498837749884, + "grad_norm": 5.492923259735107, + "learning_rate": 9.9841379984138e-06, + "loss": 0.1084, + "step": 15869 + }, + { + "epoch": 7.377963737796374, + "grad_norm": 6.225037097930908, + "learning_rate": 9.9841369984137e-06, + "loss": 0.1679, + "step": 15870 + }, + { + "epoch": 7.378428637842863, + "grad_norm": 6.781972408294678, + "learning_rate": 9.9841359984136e-06, + "loss": 0.2574, + "step": 15871 + }, + { + "epoch": 7.378893537889354, + "grad_norm": 5.449754238128662, + "learning_rate": 9.984134998413501e-06, + "loss": 0.1483, + "step": 15872 + }, + { + "epoch": 7.379358437935844, + "grad_norm": 10.813925743103027, + "learning_rate": 9.984133998413402e-06, + "loss": 0.5603, + "step": 15873 + }, + { + "epoch": 7.379823337982334, + "grad_norm": 5.9476799964904785, + "learning_rate": 9.9841329984133e-06, + "loss": 0.1359, + "step": 15874 + }, + { + "epoch": 7.3802882380288235, + "grad_norm": 6.876668930053711, + "learning_rate": 9.9841319984132e-06, + "loss": 0.2019, + "step": 15875 + }, + { + "epoch": 7.380753138075314, + "grad_norm": 6.5211567878723145, + "learning_rate": 9.9841309984131e-06, + "loss": 0.1097, + "step": 15876 + }, + { + "epoch": 7.381218038121804, + "grad_norm": 4.849381923675537, + "learning_rate": 9.984129998413001e-06, + "loss": 0.1419, + "step": 15877 + }, + { + "epoch": 7.381682938168294, + "grad_norm": 3.662217617034912, + "learning_rate": 9.984128998412902e-06, + "loss": 0.1192, + "step": 15878 + }, + { + "epoch": 7.382147838214784, + "grad_norm": 6.310523509979248, + "learning_rate": 9.9841279984128e-06, + "loss": 0.253, + "step": 15879 + }, + { + "epoch": 7.382612738261274, + "grad_norm": 5.907644748687744, + "learning_rate": 9.984126998412701e-06, + "loss": 0.0961, + "step": 15880 + }, + { + "epoch": 7.383077638307764, + "grad_norm": 5.205209255218506, + "learning_rate": 9.9841259984126e-06, + "loss": 0.2563, + "step": 15881 + }, + { + "epoch": 7.383542538354254, + "grad_norm": 3.0702836513519287, + "learning_rate": 9.984124998412501e-06, + "loss": 0.0712, + "step": 15882 + }, + { + "epoch": 7.384007438400744, + "grad_norm": 3.631381034851074, + "learning_rate": 9.9841239984124e-06, + "loss": 0.0598, + "step": 15883 + }, + { + "epoch": 7.3844723384472335, + "grad_norm": 6.458792209625244, + "learning_rate": 9.9841229984123e-06, + "loss": 0.1278, + "step": 15884 + }, + { + "epoch": 7.384937238493724, + "grad_norm": 6.494484901428223, + "learning_rate": 9.984121998412201e-06, + "loss": 0.1991, + "step": 15885 + }, + { + "epoch": 7.385402138540214, + "grad_norm": 8.61258316040039, + "learning_rate": 9.9841209984121e-06, + "loss": 0.1853, + "step": 15886 + }, + { + "epoch": 7.385867038586704, + "grad_norm": 4.258261203765869, + "learning_rate": 9.984119998412e-06, + "loss": 0.068, + "step": 15887 + }, + { + "epoch": 7.386331938633194, + "grad_norm": 3.616197347640991, + "learning_rate": 9.9841189984119e-06, + "loss": 0.0845, + "step": 15888 + }, + { + "epoch": 7.386796838679684, + "grad_norm": 8.48568344116211, + "learning_rate": 9.9841179984118e-06, + "loss": 0.0769, + "step": 15889 + }, + { + "epoch": 7.387261738726174, + "grad_norm": 3.3762054443359375, + "learning_rate": 9.984116998411701e-06, + "loss": 0.0612, + "step": 15890 + }, + { + "epoch": 7.387726638772664, + "grad_norm": 3.670931100845337, + "learning_rate": 9.984115998411602e-06, + "loss": 0.0754, + "step": 15891 + }, + { + "epoch": 7.388191538819154, + "grad_norm": 4.673495769500732, + "learning_rate": 9.9841149984115e-06, + "loss": 0.1523, + "step": 15892 + }, + { + "epoch": 7.3886564388656435, + "grad_norm": 16.685997009277344, + "learning_rate": 9.984113998411401e-06, + "loss": 0.1452, + "step": 15893 + }, + { + "epoch": 7.389121338912134, + "grad_norm": 3.7364888191223145, + "learning_rate": 9.9841129984113e-06, + "loss": 0.0611, + "step": 15894 + }, + { + "epoch": 7.389586238958624, + "grad_norm": 3.5747299194335938, + "learning_rate": 9.984111998411201e-06, + "loss": 0.0564, + "step": 15895 + }, + { + "epoch": 7.390051139005114, + "grad_norm": 4.407820224761963, + "learning_rate": 9.9841109984111e-06, + "loss": 0.1157, + "step": 15896 + }, + { + "epoch": 7.390516039051604, + "grad_norm": 5.695867538452148, + "learning_rate": 9.984109998411e-06, + "loss": 0.1628, + "step": 15897 + }, + { + "epoch": 7.390980939098094, + "grad_norm": 4.398006439208984, + "learning_rate": 9.984108998410901e-06, + "loss": 0.1396, + "step": 15898 + }, + { + "epoch": 7.391445839144584, + "grad_norm": 2.660364866256714, + "learning_rate": 9.9841079984108e-06, + "loss": 0.0488, + "step": 15899 + }, + { + "epoch": 7.391910739191074, + "grad_norm": 3.6004176139831543, + "learning_rate": 9.984106998410701e-06, + "loss": 0.0619, + "step": 15900 + }, + { + "epoch": 7.392375639237564, + "grad_norm": 3.639857053756714, + "learning_rate": 9.9841059984106e-06, + "loss": 0.0556, + "step": 15901 + }, + { + "epoch": 7.3928405392840535, + "grad_norm": 4.045792579650879, + "learning_rate": 9.9841049984105e-06, + "loss": 0.0751, + "step": 15902 + }, + { + "epoch": 7.393305439330544, + "grad_norm": 4.957617282867432, + "learning_rate": 9.984103998410401e-06, + "loss": 0.1188, + "step": 15903 + }, + { + "epoch": 7.393770339377034, + "grad_norm": 2.4138081073760986, + "learning_rate": 9.984102998410302e-06, + "loss": 0.0476, + "step": 15904 + }, + { + "epoch": 7.394235239423524, + "grad_norm": 5.379878044128418, + "learning_rate": 9.9841019984102e-06, + "loss": 0.184, + "step": 15905 + }, + { + "epoch": 7.394700139470014, + "grad_norm": 3.1051812171936035, + "learning_rate": 9.984100998410101e-06, + "loss": 0.0654, + "step": 15906 + }, + { + "epoch": 7.395165039516504, + "grad_norm": 4.835105895996094, + "learning_rate": 9.98409999841e-06, + "loss": 0.172, + "step": 15907 + }, + { + "epoch": 7.395629939562994, + "grad_norm": 7.613716125488281, + "learning_rate": 9.9840989984099e-06, + "loss": 0.3123, + "step": 15908 + }, + { + "epoch": 7.396094839609484, + "grad_norm": 5.4798264503479, + "learning_rate": 9.9840979984098e-06, + "loss": 0.1996, + "step": 15909 + }, + { + "epoch": 7.396559739655974, + "grad_norm": 5.60930061340332, + "learning_rate": 9.9840969984097e-06, + "loss": 0.1607, + "step": 15910 + }, + { + "epoch": 7.397024639702464, + "grad_norm": 2.3272886276245117, + "learning_rate": 9.984095998409601e-06, + "loss": 0.0342, + "step": 15911 + }, + { + "epoch": 7.397489539748954, + "grad_norm": 5.146169662475586, + "learning_rate": 9.9840949984095e-06, + "loss": 0.0784, + "step": 15912 + }, + { + "epoch": 7.397954439795444, + "grad_norm": 6.426748752593994, + "learning_rate": 9.984093998409401e-06, + "loss": 0.1973, + "step": 15913 + }, + { + "epoch": 7.398419339841934, + "grad_norm": 6.920488357543945, + "learning_rate": 9.9840929984093e-06, + "loss": 0.244, + "step": 15914 + }, + { + "epoch": 7.398884239888424, + "grad_norm": 4.295564651489258, + "learning_rate": 9.9840919984092e-06, + "loss": 0.1353, + "step": 15915 + }, + { + "epoch": 7.399349139934914, + "grad_norm": 5.740469455718994, + "learning_rate": 9.984090998409101e-06, + "loss": 0.1197, + "step": 15916 + }, + { + "epoch": 7.399814039981404, + "grad_norm": 6.552061557769775, + "learning_rate": 9.984089998409002e-06, + "loss": 0.2189, + "step": 15917 + }, + { + "epoch": 7.400278940027894, + "grad_norm": 5.1318159103393555, + "learning_rate": 9.984088998408901e-06, + "loss": 0.1707, + "step": 15918 + }, + { + "epoch": 7.400743840074384, + "grad_norm": 5.005618572235107, + "learning_rate": 9.9840879984088e-06, + "loss": 0.1695, + "step": 15919 + }, + { + "epoch": 7.401208740120874, + "grad_norm": 7.787482261657715, + "learning_rate": 9.9840869984087e-06, + "loss": 0.2705, + "step": 15920 + }, + { + "epoch": 7.401673640167364, + "grad_norm": 3.5700557231903076, + "learning_rate": 9.9840859984086e-06, + "loss": 0.0626, + "step": 15921 + }, + { + "epoch": 7.402138540213854, + "grad_norm": 10.002717971801758, + "learning_rate": 9.9840849984085e-06, + "loss": 0.0688, + "step": 15922 + }, + { + "epoch": 7.402603440260344, + "grad_norm": 6.368231296539307, + "learning_rate": 9.9840839984084e-06, + "loss": 0.0376, + "step": 15923 + }, + { + "epoch": 7.4030683403068345, + "grad_norm": 7.179866313934326, + "learning_rate": 9.984082998408301e-06, + "loss": 0.1762, + "step": 15924 + }, + { + "epoch": 7.403533240353324, + "grad_norm": 3.1356112957000732, + "learning_rate": 9.9840819984082e-06, + "loss": 0.0642, + "step": 15925 + }, + { + "epoch": 7.403998140399814, + "grad_norm": 4.95284366607666, + "learning_rate": 9.984080998408101e-06, + "loss": 0.1389, + "step": 15926 + }, + { + "epoch": 7.404463040446304, + "grad_norm": 3.068439483642578, + "learning_rate": 9.984079998408e-06, + "loss": 0.0423, + "step": 15927 + }, + { + "epoch": 7.404927940492794, + "grad_norm": 3.292632579803467, + "learning_rate": 9.9840789984079e-06, + "loss": 0.0789, + "step": 15928 + }, + { + "epoch": 7.405392840539284, + "grad_norm": 6.244712829589844, + "learning_rate": 9.984077998407801e-06, + "loss": 0.179, + "step": 15929 + }, + { + "epoch": 7.405857740585774, + "grad_norm": 3.559668779373169, + "learning_rate": 9.9840769984077e-06, + "loss": 0.0563, + "step": 15930 + }, + { + "epoch": 7.406322640632264, + "grad_norm": 7.357878684997559, + "learning_rate": 9.984075998407601e-06, + "loss": 0.1278, + "step": 15931 + }, + { + "epoch": 7.406787540678754, + "grad_norm": 4.5857253074646, + "learning_rate": 9.9840749984075e-06, + "loss": 0.17, + "step": 15932 + }, + { + "epoch": 7.4072524407252445, + "grad_norm": 5.922474384307861, + "learning_rate": 9.9840739984074e-06, + "loss": 0.2022, + "step": 15933 + }, + { + "epoch": 7.407717340771734, + "grad_norm": 6.540398120880127, + "learning_rate": 9.9840729984073e-06, + "loss": 0.2606, + "step": 15934 + }, + { + "epoch": 7.408182240818224, + "grad_norm": 6.74242639541626, + "learning_rate": 9.9840719984072e-06, + "loss": 0.2255, + "step": 15935 + }, + { + "epoch": 7.408647140864714, + "grad_norm": 4.229296684265137, + "learning_rate": 9.984070998407101e-06, + "loss": 0.0751, + "step": 15936 + }, + { + "epoch": 7.409112040911204, + "grad_norm": 6.0074310302734375, + "learning_rate": 9.984069998407002e-06, + "loss": 0.1314, + "step": 15937 + }, + { + "epoch": 7.409576940957694, + "grad_norm": 4.1820502281188965, + "learning_rate": 9.9840689984069e-06, + "loss": 0.11, + "step": 15938 + }, + { + "epoch": 7.410041841004184, + "grad_norm": 4.465377330780029, + "learning_rate": 9.984067998406801e-06, + "loss": 0.1016, + "step": 15939 + }, + { + "epoch": 7.410506741050674, + "grad_norm": 2.2326834201812744, + "learning_rate": 9.9840669984067e-06, + "loss": 0.0381, + "step": 15940 + }, + { + "epoch": 7.410971641097164, + "grad_norm": 6.036127090454102, + "learning_rate": 9.9840659984066e-06, + "loss": 0.1591, + "step": 15941 + }, + { + "epoch": 7.4114365411436545, + "grad_norm": 2.805645704269409, + "learning_rate": 9.984064998406501e-06, + "loss": 0.052, + "step": 15942 + }, + { + "epoch": 7.411901441190144, + "grad_norm": 6.492580413818359, + "learning_rate": 9.9840639984064e-06, + "loss": 0.228, + "step": 15943 + }, + { + "epoch": 7.412366341236634, + "grad_norm": 5.54447078704834, + "learning_rate": 9.984062998406301e-06, + "loss": 0.2044, + "step": 15944 + }, + { + "epoch": 7.412831241283124, + "grad_norm": 4.201224327087402, + "learning_rate": 9.9840619984062e-06, + "loss": 0.098, + "step": 15945 + }, + { + "epoch": 7.413296141329614, + "grad_norm": 5.909546375274658, + "learning_rate": 9.9840609984061e-06, + "loss": 0.2533, + "step": 15946 + }, + { + "epoch": 7.413761041376104, + "grad_norm": 6.849898338317871, + "learning_rate": 9.984059998406001e-06, + "loss": 0.1702, + "step": 15947 + }, + { + "epoch": 7.414225941422594, + "grad_norm": 3.4125075340270996, + "learning_rate": 9.9840589984059e-06, + "loss": 0.0559, + "step": 15948 + }, + { + "epoch": 7.414690841469084, + "grad_norm": 7.240363597869873, + "learning_rate": 9.984057998405801e-06, + "loss": 0.1831, + "step": 15949 + }, + { + "epoch": 7.415155741515574, + "grad_norm": 5.778329372406006, + "learning_rate": 9.984056998405702e-06, + "loss": 0.1645, + "step": 15950 + }, + { + "epoch": 7.4156206415620645, + "grad_norm": 5.254227638244629, + "learning_rate": 9.9840559984056e-06, + "loss": 0.1164, + "step": 15951 + }, + { + "epoch": 7.416085541608554, + "grad_norm": 3.952907085418701, + "learning_rate": 9.9840549984055e-06, + "loss": 0.0914, + "step": 15952 + }, + { + "epoch": 7.416550441655044, + "grad_norm": 6.5576958656311035, + "learning_rate": 9.9840539984054e-06, + "loss": 0.1317, + "step": 15953 + }, + { + "epoch": 7.417015341701534, + "grad_norm": 5.089023590087891, + "learning_rate": 9.9840529984053e-06, + "loss": 0.1636, + "step": 15954 + }, + { + "epoch": 7.417480241748025, + "grad_norm": 4.667247295379639, + "learning_rate": 9.984051998405201e-06, + "loss": 0.0552, + "step": 15955 + }, + { + "epoch": 7.417945141794514, + "grad_norm": 8.581334114074707, + "learning_rate": 9.9840509984051e-06, + "loss": 0.2999, + "step": 15956 + }, + { + "epoch": 7.418410041841004, + "grad_norm": 6.13039493560791, + "learning_rate": 9.984049998405001e-06, + "loss": 0.2247, + "step": 15957 + }, + { + "epoch": 7.418874941887494, + "grad_norm": 7.838290214538574, + "learning_rate": 9.9840489984049e-06, + "loss": 0.2918, + "step": 15958 + }, + { + "epoch": 7.419339841933985, + "grad_norm": 6.46225118637085, + "learning_rate": 9.9840479984048e-06, + "loss": 0.3081, + "step": 15959 + }, + { + "epoch": 7.4198047419804745, + "grad_norm": 3.1297640800476074, + "learning_rate": 9.984046998404701e-06, + "loss": 0.0478, + "step": 15960 + }, + { + "epoch": 7.420269642026964, + "grad_norm": 6.383517265319824, + "learning_rate": 9.984045998404602e-06, + "loss": 0.2106, + "step": 15961 + }, + { + "epoch": 7.420734542073454, + "grad_norm": 6.593760013580322, + "learning_rate": 9.984044998404501e-06, + "loss": 0.0973, + "step": 15962 + }, + { + "epoch": 7.421199442119944, + "grad_norm": 7.402289867401123, + "learning_rate": 9.9840439984044e-06, + "loss": 0.088, + "step": 15963 + }, + { + "epoch": 7.421664342166435, + "grad_norm": 2.4314541816711426, + "learning_rate": 9.9840429984043e-06, + "loss": 0.0336, + "step": 15964 + }, + { + "epoch": 7.422129242212924, + "grad_norm": 7.017064094543457, + "learning_rate": 9.9840419984042e-06, + "loss": 0.1865, + "step": 15965 + }, + { + "epoch": 7.422594142259414, + "grad_norm": 4.997385501861572, + "learning_rate": 9.9840409984041e-06, + "loss": 0.0948, + "step": 15966 + }, + { + "epoch": 7.423059042305904, + "grad_norm": 3.957892894744873, + "learning_rate": 9.984039998404001e-06, + "loss": 0.1007, + "step": 15967 + }, + { + "epoch": 7.423523942352395, + "grad_norm": 5.103984832763672, + "learning_rate": 9.984038998403902e-06, + "loss": 0.1707, + "step": 15968 + }, + { + "epoch": 7.4239888423988845, + "grad_norm": 3.8183579444885254, + "learning_rate": 9.9840379984038e-06, + "loss": 0.1285, + "step": 15969 + }, + { + "epoch": 7.424453742445374, + "grad_norm": 4.05864953994751, + "learning_rate": 9.984036998403701e-06, + "loss": 0.0912, + "step": 15970 + }, + { + "epoch": 7.424918642491864, + "grad_norm": 6.632140159606934, + "learning_rate": 9.9840359984036e-06, + "loss": 0.0996, + "step": 15971 + }, + { + "epoch": 7.425383542538354, + "grad_norm": 7.980469226837158, + "learning_rate": 9.9840349984035e-06, + "loss": 0.3034, + "step": 15972 + }, + { + "epoch": 7.4258484425848446, + "grad_norm": 6.998808860778809, + "learning_rate": 9.984033998403401e-06, + "loss": 0.2285, + "step": 15973 + }, + { + "epoch": 7.426313342631334, + "grad_norm": 4.4499335289001465, + "learning_rate": 9.9840329984033e-06, + "loss": 0.0905, + "step": 15974 + }, + { + "epoch": 7.426778242677824, + "grad_norm": 1.244540810585022, + "learning_rate": 9.984031998403201e-06, + "loss": 0.017, + "step": 15975 + }, + { + "epoch": 7.427243142724314, + "grad_norm": 5.5781450271606445, + "learning_rate": 9.9840309984031e-06, + "loss": 0.0924, + "step": 15976 + }, + { + "epoch": 7.427708042770805, + "grad_norm": 7.093169689178467, + "learning_rate": 9.984029998403e-06, + "loss": 0.2597, + "step": 15977 + }, + { + "epoch": 7.4281729428172945, + "grad_norm": 3.588444948196411, + "learning_rate": 9.9840289984029e-06, + "loss": 0.0526, + "step": 15978 + }, + { + "epoch": 7.428637842863784, + "grad_norm": 3.1172211170196533, + "learning_rate": 9.9840279984028e-06, + "loss": 0.0716, + "step": 15979 + }, + { + "epoch": 7.429102742910274, + "grad_norm": 6.39795446395874, + "learning_rate": 9.984026998402701e-06, + "loss": 0.2204, + "step": 15980 + }, + { + "epoch": 7.429567642956764, + "grad_norm": 7.483311653137207, + "learning_rate": 9.984025998402602e-06, + "loss": 0.3035, + "step": 15981 + }, + { + "epoch": 7.4300325430032546, + "grad_norm": 8.222458839416504, + "learning_rate": 9.9840249984025e-06, + "loss": 0.1679, + "step": 15982 + }, + { + "epoch": 7.430497443049744, + "grad_norm": 2.552863359451294, + "learning_rate": 9.9840239984024e-06, + "loss": 0.0348, + "step": 15983 + }, + { + "epoch": 7.430962343096234, + "grad_norm": 9.358975410461426, + "learning_rate": 9.9840229984023e-06, + "loss": 0.3391, + "step": 15984 + }, + { + "epoch": 7.431427243142724, + "grad_norm": 11.431628227233887, + "learning_rate": 9.984021998402201e-06, + "loss": 0.3358, + "step": 15985 + }, + { + "epoch": 7.431892143189215, + "grad_norm": 6.21508264541626, + "learning_rate": 9.984020998402102e-06, + "loss": 0.1982, + "step": 15986 + }, + { + "epoch": 7.4323570432357045, + "grad_norm": 7.217568874359131, + "learning_rate": 9.984019998402e-06, + "loss": 0.2055, + "step": 15987 + }, + { + "epoch": 7.432821943282194, + "grad_norm": 2.8625388145446777, + "learning_rate": 9.984018998401901e-06, + "loss": 0.043, + "step": 15988 + }, + { + "epoch": 7.433286843328684, + "grad_norm": 9.826711654663086, + "learning_rate": 9.9840179984018e-06, + "loss": 0.2127, + "step": 15989 + }, + { + "epoch": 7.433751743375175, + "grad_norm": 5.991591453552246, + "learning_rate": 9.9840169984017e-06, + "loss": 0.179, + "step": 15990 + }, + { + "epoch": 7.4342166434216645, + "grad_norm": 6.237965106964111, + "learning_rate": 9.9840159984016e-06, + "loss": 0.2462, + "step": 15991 + }, + { + "epoch": 7.434681543468154, + "grad_norm": 2.7518579959869385, + "learning_rate": 9.9840149984015e-06, + "loss": 0.0581, + "step": 15992 + }, + { + "epoch": 7.435146443514644, + "grad_norm": 4.689074516296387, + "learning_rate": 9.984013998401401e-06, + "loss": 0.0813, + "step": 15993 + }, + { + "epoch": 7.435611343561134, + "grad_norm": 4.461926460266113, + "learning_rate": 9.9840129984013e-06, + "loss": 0.0837, + "step": 15994 + }, + { + "epoch": 7.436076243607625, + "grad_norm": 5.080416202545166, + "learning_rate": 9.9840119984012e-06, + "loss": 0.0861, + "step": 15995 + }, + { + "epoch": 7.4365411436541144, + "grad_norm": 3.2793612480163574, + "learning_rate": 9.9840109984011e-06, + "loss": 0.073, + "step": 15996 + }, + { + "epoch": 7.437006043700604, + "grad_norm": 9.953462600708008, + "learning_rate": 9.984009998401e-06, + "loss": 0.2791, + "step": 15997 + }, + { + "epoch": 7.437470943747094, + "grad_norm": 5.041614532470703, + "learning_rate": 9.984008998400901e-06, + "loss": 0.1394, + "step": 15998 + }, + { + "epoch": 7.437935843793585, + "grad_norm": 5.16010856628418, + "learning_rate": 9.984007998400802e-06, + "loss": 0.1376, + "step": 15999 + }, + { + "epoch": 7.4384007438400745, + "grad_norm": 7.122220516204834, + "learning_rate": 9.9840069984007e-06, + "loss": 0.2273, + "step": 16000 + }, + { + "epoch": 7.438865643886564, + "grad_norm": 5.677630424499512, + "learning_rate": 9.984005998400601e-06, + "loss": 0.1972, + "step": 16001 + }, + { + "epoch": 7.439330543933054, + "grad_norm": 3.093611717224121, + "learning_rate": 9.9840049984005e-06, + "loss": 0.062, + "step": 16002 + }, + { + "epoch": 7.439795443979545, + "grad_norm": 5.979808330535889, + "learning_rate": 9.984003998400401e-06, + "loss": 0.1354, + "step": 16003 + }, + { + "epoch": 7.440260344026035, + "grad_norm": 4.806941986083984, + "learning_rate": 9.9840029984003e-06, + "loss": 0.1367, + "step": 16004 + }, + { + "epoch": 7.4407252440725244, + "grad_norm": 6.315893650054932, + "learning_rate": 9.9840019984002e-06, + "loss": 0.1393, + "step": 16005 + }, + { + "epoch": 7.441190144119014, + "grad_norm": 14.294254302978516, + "learning_rate": 9.984000998400101e-06, + "loss": 0.451, + "step": 16006 + }, + { + "epoch": 7.441655044165504, + "grad_norm": 4.852384090423584, + "learning_rate": 9.9839999984e-06, + "loss": 0.0804, + "step": 16007 + }, + { + "epoch": 7.442119944211995, + "grad_norm": 5.391429901123047, + "learning_rate": 9.9839989983999e-06, + "loss": 0.1733, + "step": 16008 + }, + { + "epoch": 7.4425848442584845, + "grad_norm": 1.9100911617279053, + "learning_rate": 9.9839979983998e-06, + "loss": 0.0201, + "step": 16009 + }, + { + "epoch": 7.443049744304974, + "grad_norm": 4.6758222579956055, + "learning_rate": 9.9839969983997e-06, + "loss": 0.0787, + "step": 16010 + }, + { + "epoch": 7.443514644351464, + "grad_norm": 5.209042072296143, + "learning_rate": 9.983995998399601e-06, + "loss": 0.1184, + "step": 16011 + }, + { + "epoch": 7.443979544397955, + "grad_norm": 5.594644546508789, + "learning_rate": 9.983994998399502e-06, + "loss": 0.1648, + "step": 16012 + }, + { + "epoch": 7.444444444444445, + "grad_norm": 5.653451919555664, + "learning_rate": 9.9839939983994e-06, + "loss": 0.1015, + "step": 16013 + }, + { + "epoch": 7.4449093444909344, + "grad_norm": 9.492862701416016, + "learning_rate": 9.983992998399301e-06, + "loss": 0.2542, + "step": 16014 + }, + { + "epoch": 7.445374244537424, + "grad_norm": 7.694924354553223, + "learning_rate": 9.9839919983992e-06, + "loss": 0.1783, + "step": 16015 + }, + { + "epoch": 7.445839144583914, + "grad_norm": 5.727797508239746, + "learning_rate": 9.983990998399101e-06, + "loss": 0.0798, + "step": 16016 + }, + { + "epoch": 7.446304044630405, + "grad_norm": 6.044077396392822, + "learning_rate": 9.983989998399e-06, + "loss": 0.1797, + "step": 16017 + }, + { + "epoch": 7.4467689446768945, + "grad_norm": 11.757973670959473, + "learning_rate": 9.9839889983989e-06, + "loss": 0.0866, + "step": 16018 + }, + { + "epoch": 7.447233844723384, + "grad_norm": 7.822447299957275, + "learning_rate": 9.983987998398801e-06, + "loss": 0.1884, + "step": 16019 + }, + { + "epoch": 7.447698744769874, + "grad_norm": 6.099127292633057, + "learning_rate": 9.9839869983987e-06, + "loss": 0.1205, + "step": 16020 + }, + { + "epoch": 7.448163644816365, + "grad_norm": 6.704605579376221, + "learning_rate": 9.983985998398601e-06, + "loss": 0.2169, + "step": 16021 + }, + { + "epoch": 7.448628544862855, + "grad_norm": 6.952407360076904, + "learning_rate": 9.9839849983985e-06, + "loss": 0.1399, + "step": 16022 + }, + { + "epoch": 7.449093444909344, + "grad_norm": 4.519223213195801, + "learning_rate": 9.9839839983984e-06, + "loss": 0.1595, + "step": 16023 + }, + { + "epoch": 7.449558344955834, + "grad_norm": 5.966975212097168, + "learning_rate": 9.983982998398301e-06, + "loss": 0.1209, + "step": 16024 + }, + { + "epoch": 7.450023245002324, + "grad_norm": 3.987229347229004, + "learning_rate": 9.983981998398202e-06, + "loss": 0.0974, + "step": 16025 + }, + { + "epoch": 7.450488145048815, + "grad_norm": 8.011350631713867, + "learning_rate": 9.9839809983981e-06, + "loss": 0.3383, + "step": 16026 + }, + { + "epoch": 7.4509530450953045, + "grad_norm": 4.2519612312316895, + "learning_rate": 9.983979998398e-06, + "loss": 0.0935, + "step": 16027 + }, + { + "epoch": 7.451417945141794, + "grad_norm": 5.291379451751709, + "learning_rate": 9.9839789983979e-06, + "loss": 0.1473, + "step": 16028 + }, + { + "epoch": 7.451882845188284, + "grad_norm": 4.650259971618652, + "learning_rate": 9.983977998397801e-06, + "loss": 0.1553, + "step": 16029 + }, + { + "epoch": 7.452347745234775, + "grad_norm": 8.887043952941895, + "learning_rate": 9.983976998397702e-06, + "loss": 0.2965, + "step": 16030 + }, + { + "epoch": 7.452812645281265, + "grad_norm": 10.000426292419434, + "learning_rate": 9.9839759983976e-06, + "loss": 0.3403, + "step": 16031 + }, + { + "epoch": 7.453277545327754, + "grad_norm": 2.3250980377197266, + "learning_rate": 9.983974998397501e-06, + "loss": 0.0371, + "step": 16032 + }, + { + "epoch": 7.453742445374244, + "grad_norm": 5.5355753898620605, + "learning_rate": 9.9839739983974e-06, + "loss": 0.2252, + "step": 16033 + }, + { + "epoch": 7.454207345420735, + "grad_norm": 4.034704208374023, + "learning_rate": 9.983972998397301e-06, + "loss": 0.0851, + "step": 16034 + }, + { + "epoch": 7.454672245467225, + "grad_norm": 4.302779197692871, + "learning_rate": 9.9839719983972e-06, + "loss": 0.0874, + "step": 16035 + }, + { + "epoch": 7.4551371455137145, + "grad_norm": 3.070568084716797, + "learning_rate": 9.9839709983971e-06, + "loss": 0.0419, + "step": 16036 + }, + { + "epoch": 7.455602045560204, + "grad_norm": 3.346186876296997, + "learning_rate": 9.983969998397001e-06, + "loss": 0.0501, + "step": 16037 + }, + { + "epoch": 7.456066945606695, + "grad_norm": 4.629876136779785, + "learning_rate": 9.9839689983969e-06, + "loss": 0.0942, + "step": 16038 + }, + { + "epoch": 7.456531845653185, + "grad_norm": 7.682029724121094, + "learning_rate": 9.9839679983968e-06, + "loss": 0.2431, + "step": 16039 + }, + { + "epoch": 7.456996745699675, + "grad_norm": 5.6125359535217285, + "learning_rate": 9.9839669983967e-06, + "loss": 0.1524, + "step": 16040 + }, + { + "epoch": 7.457461645746164, + "grad_norm": 7.716052532196045, + "learning_rate": 9.9839659983966e-06, + "loss": 0.2569, + "step": 16041 + }, + { + "epoch": 7.457926545792654, + "grad_norm": 7.266421794891357, + "learning_rate": 9.983964998396501e-06, + "loss": 0.2424, + "step": 16042 + }, + { + "epoch": 7.458391445839145, + "grad_norm": 6.294659614562988, + "learning_rate": 9.983963998396402e-06, + "loss": 0.1459, + "step": 16043 + }, + { + "epoch": 7.458856345885635, + "grad_norm": 6.7052178382873535, + "learning_rate": 9.9839629983963e-06, + "loss": 0.0747, + "step": 16044 + }, + { + "epoch": 7.4593212459321245, + "grad_norm": 4.625628471374512, + "learning_rate": 9.983961998396201e-06, + "loss": 0.1149, + "step": 16045 + }, + { + "epoch": 7.459786145978614, + "grad_norm": 4.111689567565918, + "learning_rate": 9.9839609983961e-06, + "loss": 0.1225, + "step": 16046 + }, + { + "epoch": 7.460251046025105, + "grad_norm": 2.068606376647949, + "learning_rate": 9.983959998396001e-06, + "loss": 0.0182, + "step": 16047 + }, + { + "epoch": 7.460715946071595, + "grad_norm": 5.169097423553467, + "learning_rate": 9.9839589983959e-06, + "loss": 0.1606, + "step": 16048 + }, + { + "epoch": 7.461180846118085, + "grad_norm": 4.343909740447998, + "learning_rate": 9.9839579983958e-06, + "loss": 0.0604, + "step": 16049 + }, + { + "epoch": 7.461645746164574, + "grad_norm": 5.338029384613037, + "learning_rate": 9.983956998395701e-06, + "loss": 0.1896, + "step": 16050 + }, + { + "epoch": 7.462110646211064, + "grad_norm": 5.284976482391357, + "learning_rate": 9.9839559983956e-06, + "loss": 0.0954, + "step": 16051 + }, + { + "epoch": 7.462575546257555, + "grad_norm": 3.279179573059082, + "learning_rate": 9.983954998395501e-06, + "loss": 0.0555, + "step": 16052 + }, + { + "epoch": 7.463040446304045, + "grad_norm": 4.284603118896484, + "learning_rate": 9.9839539983954e-06, + "loss": 0.1117, + "step": 16053 + }, + { + "epoch": 7.4635053463505345, + "grad_norm": 6.147878646850586, + "learning_rate": 9.9839529983953e-06, + "loss": 0.1205, + "step": 16054 + }, + { + "epoch": 7.463970246397024, + "grad_norm": 5.849339485168457, + "learning_rate": 9.983951998395201e-06, + "loss": 0.1677, + "step": 16055 + }, + { + "epoch": 7.464435146443515, + "grad_norm": 3.8992443084716797, + "learning_rate": 9.983950998395102e-06, + "loss": 0.0538, + "step": 16056 + }, + { + "epoch": 7.464900046490005, + "grad_norm": 5.617926597595215, + "learning_rate": 9.983949998395e-06, + "loss": 0.1966, + "step": 16057 + }, + { + "epoch": 7.465364946536495, + "grad_norm": 10.894519805908203, + "learning_rate": 9.983948998394901e-06, + "loss": 0.2122, + "step": 16058 + }, + { + "epoch": 7.465829846582984, + "grad_norm": 5.8491034507751465, + "learning_rate": 9.9839479983948e-06, + "loss": 0.1512, + "step": 16059 + }, + { + "epoch": 7.466294746629474, + "grad_norm": 4.690001010894775, + "learning_rate": 9.9839469983947e-06, + "loss": 0.1009, + "step": 16060 + }, + { + "epoch": 7.466759646675965, + "grad_norm": 6.954822540283203, + "learning_rate": 9.9839459983946e-06, + "loss": 0.1605, + "step": 16061 + }, + { + "epoch": 7.467224546722455, + "grad_norm": 4.151301383972168, + "learning_rate": 9.9839449983945e-06, + "loss": 0.0529, + "step": 16062 + }, + { + "epoch": 7.4676894467689445, + "grad_norm": 4.938438415527344, + "learning_rate": 9.983943998394401e-06, + "loss": 0.1979, + "step": 16063 + }, + { + "epoch": 7.468154346815434, + "grad_norm": 3.983121633529663, + "learning_rate": 9.9839429983943e-06, + "loss": 0.0855, + "step": 16064 + }, + { + "epoch": 7.468619246861925, + "grad_norm": 2.351175308227539, + "learning_rate": 9.983941998394201e-06, + "loss": 0.033, + "step": 16065 + }, + { + "epoch": 7.469084146908415, + "grad_norm": 5.969276428222656, + "learning_rate": 9.9839409983941e-06, + "loss": 0.1829, + "step": 16066 + }, + { + "epoch": 7.469549046954905, + "grad_norm": 4.092048168182373, + "learning_rate": 9.983939998394e-06, + "loss": 0.0945, + "step": 16067 + }, + { + "epoch": 7.470013947001394, + "grad_norm": 3.3193447589874268, + "learning_rate": 9.983938998393901e-06, + "loss": 0.0426, + "step": 16068 + }, + { + "epoch": 7.470478847047885, + "grad_norm": 6.358748912811279, + "learning_rate": 9.983937998393802e-06, + "loss": 0.1954, + "step": 16069 + }, + { + "epoch": 7.470943747094375, + "grad_norm": 5.2975640296936035, + "learning_rate": 9.983936998393701e-06, + "loss": 0.129, + "step": 16070 + }, + { + "epoch": 7.471408647140865, + "grad_norm": 5.0865068435668945, + "learning_rate": 9.9839359983936e-06, + "loss": 0.148, + "step": 16071 + }, + { + "epoch": 7.4718735471873545, + "grad_norm": 5.68256139755249, + "learning_rate": 9.9839349983935e-06, + "loss": 0.2037, + "step": 16072 + }, + { + "epoch": 7.472338447233844, + "grad_norm": 3.7517240047454834, + "learning_rate": 9.9839339983934e-06, + "loss": 0.0687, + "step": 16073 + }, + { + "epoch": 7.472803347280335, + "grad_norm": 8.278694152832031, + "learning_rate": 9.9839329983933e-06, + "loss": 0.2087, + "step": 16074 + }, + { + "epoch": 7.473268247326825, + "grad_norm": 3.4983134269714355, + "learning_rate": 9.9839319983932e-06, + "loss": 0.0416, + "step": 16075 + }, + { + "epoch": 7.473733147373315, + "grad_norm": 6.368930816650391, + "learning_rate": 9.983930998393101e-06, + "loss": 0.2321, + "step": 16076 + }, + { + "epoch": 7.474198047419804, + "grad_norm": 6.454529762268066, + "learning_rate": 9.983929998393e-06, + "loss": 0.1523, + "step": 16077 + }, + { + "epoch": 7.474662947466295, + "grad_norm": 7.119606971740723, + "learning_rate": 9.983928998392901e-06, + "loss": 0.1291, + "step": 16078 + }, + { + "epoch": 7.475127847512785, + "grad_norm": 4.735374927520752, + "learning_rate": 9.9839279983928e-06, + "loss": 0.1292, + "step": 16079 + }, + { + "epoch": 7.475592747559275, + "grad_norm": 5.052401542663574, + "learning_rate": 9.9839269983927e-06, + "loss": 0.0907, + "step": 16080 + }, + { + "epoch": 7.4760576476057645, + "grad_norm": 10.486730575561523, + "learning_rate": 9.983925998392601e-06, + "loss": 0.1713, + "step": 16081 + }, + { + "epoch": 7.476522547652255, + "grad_norm": 7.7849297523498535, + "learning_rate": 9.9839249983925e-06, + "loss": 0.14, + "step": 16082 + }, + { + "epoch": 7.476987447698745, + "grad_norm": 4.457630157470703, + "learning_rate": 9.983923998392401e-06, + "loss": 0.0845, + "step": 16083 + }, + { + "epoch": 7.477452347745235, + "grad_norm": 8.26363468170166, + "learning_rate": 9.9839229983923e-06, + "loss": 0.0808, + "step": 16084 + }, + { + "epoch": 7.477917247791725, + "grad_norm": 5.0902886390686035, + "learning_rate": 9.9839219983922e-06, + "loss": 0.1597, + "step": 16085 + }, + { + "epoch": 7.478382147838214, + "grad_norm": 4.361978054046631, + "learning_rate": 9.9839209983921e-06, + "loss": 0.0709, + "step": 16086 + }, + { + "epoch": 7.478847047884705, + "grad_norm": 4.797949314117432, + "learning_rate": 9.983919998392e-06, + "loss": 0.1269, + "step": 16087 + }, + { + "epoch": 7.479311947931195, + "grad_norm": 5.877004146575928, + "learning_rate": 9.983918998391901e-06, + "loss": 0.1931, + "step": 16088 + }, + { + "epoch": 7.479776847977685, + "grad_norm": 3.2798516750335693, + "learning_rate": 9.983917998391802e-06, + "loss": 0.0461, + "step": 16089 + }, + { + "epoch": 7.4802417480241745, + "grad_norm": 4.520254611968994, + "learning_rate": 9.9839169983917e-06, + "loss": 0.0608, + "step": 16090 + }, + { + "epoch": 7.480706648070665, + "grad_norm": 4.154953479766846, + "learning_rate": 9.983915998391601e-06, + "loss": 0.09, + "step": 16091 + }, + { + "epoch": 7.481171548117155, + "grad_norm": 5.32565450668335, + "learning_rate": 9.9839149983915e-06, + "loss": 0.1699, + "step": 16092 + }, + { + "epoch": 7.481636448163645, + "grad_norm": 2.659149408340454, + "learning_rate": 9.9839139983914e-06, + "loss": 0.0304, + "step": 16093 + }, + { + "epoch": 7.482101348210135, + "grad_norm": 4.937232494354248, + "learning_rate": 9.983912998391301e-06, + "loss": 0.0419, + "step": 16094 + }, + { + "epoch": 7.482566248256624, + "grad_norm": 4.1256537437438965, + "learning_rate": 9.9839119983912e-06, + "loss": 0.1229, + "step": 16095 + }, + { + "epoch": 7.483031148303115, + "grad_norm": 6.562041282653809, + "learning_rate": 9.983910998391101e-06, + "loss": 0.202, + "step": 16096 + }, + { + "epoch": 7.483496048349605, + "grad_norm": 9.648113250732422, + "learning_rate": 9.983909998391e-06, + "loss": 0.3306, + "step": 16097 + }, + { + "epoch": 7.483960948396095, + "grad_norm": 4.835995197296143, + "learning_rate": 9.9839089983909e-06, + "loss": 0.1334, + "step": 16098 + }, + { + "epoch": 7.4844258484425845, + "grad_norm": 3.8769733905792236, + "learning_rate": 9.9839079983908e-06, + "loss": 0.0463, + "step": 16099 + }, + { + "epoch": 7.484890748489075, + "grad_norm": 5.07838773727417, + "learning_rate": 9.9839069983907e-06, + "loss": 0.0773, + "step": 16100 + }, + { + "epoch": 7.485355648535565, + "grad_norm": 5.584734916687012, + "learning_rate": 9.983905998390601e-06, + "loss": 0.1477, + "step": 16101 + }, + { + "epoch": 7.485820548582055, + "grad_norm": 3.597825527191162, + "learning_rate": 9.983904998390502e-06, + "loss": 0.0471, + "step": 16102 + }, + { + "epoch": 7.486285448628545, + "grad_norm": 3.0115578174591064, + "learning_rate": 9.9839039983904e-06, + "loss": 0.0617, + "step": 16103 + }, + { + "epoch": 7.486750348675034, + "grad_norm": 4.666576385498047, + "learning_rate": 9.9839029983903e-06, + "loss": 0.0806, + "step": 16104 + }, + { + "epoch": 7.487215248721525, + "grad_norm": 5.392650604248047, + "learning_rate": 9.9839019983902e-06, + "loss": 0.1149, + "step": 16105 + }, + { + "epoch": 7.487680148768015, + "grad_norm": 5.507225036621094, + "learning_rate": 9.983900998390101e-06, + "loss": 0.2352, + "step": 16106 + }, + { + "epoch": 7.488145048814505, + "grad_norm": 3.759397268295288, + "learning_rate": 9.983899998390002e-06, + "loss": 0.0662, + "step": 16107 + }, + { + "epoch": 7.4886099488609945, + "grad_norm": 6.062546730041504, + "learning_rate": 9.9838989983899e-06, + "loss": 0.2005, + "step": 16108 + }, + { + "epoch": 7.489074848907485, + "grad_norm": 7.925382137298584, + "learning_rate": 9.983897998389801e-06, + "loss": 0.2621, + "step": 16109 + }, + { + "epoch": 7.489539748953975, + "grad_norm": 5.898969650268555, + "learning_rate": 9.9838969983897e-06, + "loss": 0.236, + "step": 16110 + }, + { + "epoch": 7.490004649000465, + "grad_norm": 2.6192002296447754, + "learning_rate": 9.9838959983896e-06, + "loss": 0.0386, + "step": 16111 + }, + { + "epoch": 7.490469549046955, + "grad_norm": 3.8839428424835205, + "learning_rate": 9.983894998389501e-06, + "loss": 0.0548, + "step": 16112 + }, + { + "epoch": 7.490934449093445, + "grad_norm": 4.704314708709717, + "learning_rate": 9.9838939983894e-06, + "loss": 0.0993, + "step": 16113 + }, + { + "epoch": 7.491399349139935, + "grad_norm": 6.334939956665039, + "learning_rate": 9.983892998389301e-06, + "loss": 0.2026, + "step": 16114 + }, + { + "epoch": 7.491864249186425, + "grad_norm": 2.489079475402832, + "learning_rate": 9.9838919983892e-06, + "loss": 0.0377, + "step": 16115 + }, + { + "epoch": 7.492329149232915, + "grad_norm": 9.103285789489746, + "learning_rate": 9.9838909983891e-06, + "loss": 0.1156, + "step": 16116 + }, + { + "epoch": 7.492794049279405, + "grad_norm": 4.694563388824463, + "learning_rate": 9.983889998389e-06, + "loss": 0.1438, + "step": 16117 + }, + { + "epoch": 7.493258949325895, + "grad_norm": 3.1961543560028076, + "learning_rate": 9.9838889983889e-06, + "loss": 0.0711, + "step": 16118 + }, + { + "epoch": 7.493723849372385, + "grad_norm": 2.6613686084747314, + "learning_rate": 9.983887998388801e-06, + "loss": 0.0365, + "step": 16119 + }, + { + "epoch": 7.494188749418875, + "grad_norm": 4.404818534851074, + "learning_rate": 9.983886998388702e-06, + "loss": 0.1224, + "step": 16120 + }, + { + "epoch": 7.494653649465365, + "grad_norm": 4.373404026031494, + "learning_rate": 9.9838859983886e-06, + "loss": 0.1186, + "step": 16121 + }, + { + "epoch": 7.495118549511855, + "grad_norm": 3.8664586544036865, + "learning_rate": 9.983884998388501e-06, + "loss": 0.1371, + "step": 16122 + }, + { + "epoch": 7.495583449558345, + "grad_norm": 5.3000617027282715, + "learning_rate": 9.9838839983884e-06, + "loss": 0.1364, + "step": 16123 + }, + { + "epoch": 7.496048349604835, + "grad_norm": 5.684783458709717, + "learning_rate": 9.9838829983883e-06, + "loss": 0.0967, + "step": 16124 + }, + { + "epoch": 7.496513249651325, + "grad_norm": 4.492739200592041, + "learning_rate": 9.983881998388202e-06, + "loss": 0.1115, + "step": 16125 + }, + { + "epoch": 7.496978149697815, + "grad_norm": 7.155323028564453, + "learning_rate": 9.9838809983881e-06, + "loss": 0.178, + "step": 16126 + }, + { + "epoch": 7.497443049744305, + "grad_norm": 1.2940673828125, + "learning_rate": 9.983879998388001e-06, + "loss": 0.0148, + "step": 16127 + }, + { + "epoch": 7.497907949790795, + "grad_norm": 6.186997413635254, + "learning_rate": 9.9838789983879e-06, + "loss": 0.1505, + "step": 16128 + }, + { + "epoch": 7.498372849837285, + "grad_norm": 5.591780185699463, + "learning_rate": 9.9838779983878e-06, + "loss": 0.2021, + "step": 16129 + }, + { + "epoch": 7.498837749883775, + "grad_norm": 4.318623065948486, + "learning_rate": 9.9838769983877e-06, + "loss": 0.1041, + "step": 16130 + }, + { + "epoch": 7.499302649930265, + "grad_norm": 5.298221111297607, + "learning_rate": 9.9838759983876e-06, + "loss": 0.1416, + "step": 16131 + }, + { + "epoch": 7.499767549976755, + "grad_norm": 5.796093940734863, + "learning_rate": 9.983874998387501e-06, + "loss": 0.1761, + "step": 16132 + }, + { + "epoch": 7.500232450023245, + "grad_norm": 4.543456554412842, + "learning_rate": 9.983873998387402e-06, + "loss": 0.0901, + "step": 16133 + }, + { + "epoch": 7.500697350069735, + "grad_norm": Infinity, + "learning_rate": 9.983873998387402e-06, + "loss": 0.1136, + "step": 16134 + }, + { + "epoch": 7.501162250116225, + "grad_norm": 3.4487392902374268, + "learning_rate": 9.9838729983873e-06, + "loss": 0.0982, + "step": 16135 + }, + { + "epoch": 7.501627150162715, + "grad_norm": 4.779450416564941, + "learning_rate": 9.9838719983872e-06, + "loss": 0.1022, + "step": 16136 + }, + { + "epoch": 7.502092050209205, + "grad_norm": 5.4083333015441895, + "learning_rate": 9.9838709983871e-06, + "loss": 0.1538, + "step": 16137 + }, + { + "epoch": 7.502556950255695, + "grad_norm": 5.527165412902832, + "learning_rate": 9.983869998387001e-06, + "loss": 0.0569, + "step": 16138 + }, + { + "epoch": 7.503021850302185, + "grad_norm": 3.9736900329589844, + "learning_rate": 9.983868998386902e-06, + "loss": 0.0544, + "step": 16139 + }, + { + "epoch": 7.503486750348675, + "grad_norm": 9.829278945922852, + "learning_rate": 9.9838679983868e-06, + "loss": 0.3782, + "step": 16140 + }, + { + "epoch": 7.503951650395165, + "grad_norm": 7.395517349243164, + "learning_rate": 9.983866998386701e-06, + "loss": 0.333, + "step": 16141 + }, + { + "epoch": 7.504416550441655, + "grad_norm": 5.7729315757751465, + "learning_rate": 9.9838659983866e-06, + "loss": 0.135, + "step": 16142 + }, + { + "epoch": 7.504881450488145, + "grad_norm": 5.858462333679199, + "learning_rate": 9.9838649983865e-06, + "loss": 0.155, + "step": 16143 + }, + { + "epoch": 7.505346350534635, + "grad_norm": 4.238173484802246, + "learning_rate": 9.9838639983864e-06, + "loss": 0.0981, + "step": 16144 + }, + { + "epoch": 7.505811250581125, + "grad_norm": 7.351012229919434, + "learning_rate": 9.9838629983863e-06, + "loss": 0.2438, + "step": 16145 + }, + { + "epoch": 7.506276150627615, + "grad_norm": 5.820895671844482, + "learning_rate": 9.983861998386201e-06, + "loss": 0.1958, + "step": 16146 + }, + { + "epoch": 7.506741050674105, + "grad_norm": 4.664760589599609, + "learning_rate": 9.9838609983861e-06, + "loss": 0.1429, + "step": 16147 + }, + { + "epoch": 7.507205950720595, + "grad_norm": 3.901496410369873, + "learning_rate": 9.983859998386e-06, + "loss": 0.0865, + "step": 16148 + }, + { + "epoch": 7.507670850767085, + "grad_norm": 5.483670234680176, + "learning_rate": 9.9838589983859e-06, + "loss": 0.2049, + "step": 16149 + }, + { + "epoch": 7.508135750813575, + "grad_norm": 2.5648488998413086, + "learning_rate": 9.9838579983858e-06, + "loss": 0.0255, + "step": 16150 + }, + { + "epoch": 7.508600650860065, + "grad_norm": 5.802597999572754, + "learning_rate": 9.983856998385701e-06, + "loss": 0.0508, + "step": 16151 + }, + { + "epoch": 7.509065550906556, + "grad_norm": 3.101743698120117, + "learning_rate": 9.983855998385602e-06, + "loss": 0.0478, + "step": 16152 + }, + { + "epoch": 7.509530450953045, + "grad_norm": 6.675249099731445, + "learning_rate": 9.9838549983855e-06, + "loss": 0.2347, + "step": 16153 + }, + { + "epoch": 7.509995350999535, + "grad_norm": 5.3565778732299805, + "learning_rate": 9.983853998385401e-06, + "loss": 0.1304, + "step": 16154 + }, + { + "epoch": 7.510460251046025, + "grad_norm": 5.550121784210205, + "learning_rate": 9.9838529983853e-06, + "loss": 0.1601, + "step": 16155 + }, + { + "epoch": 7.510925151092515, + "grad_norm": 4.250685691833496, + "learning_rate": 9.983851998385201e-06, + "loss": 0.0864, + "step": 16156 + }, + { + "epoch": 7.5113900511390055, + "grad_norm": 6.029175281524658, + "learning_rate": 9.9838509983851e-06, + "loss": 0.0708, + "step": 16157 + }, + { + "epoch": 7.511854951185495, + "grad_norm": 5.966298580169678, + "learning_rate": 9.983849998385e-06, + "loss": 0.1708, + "step": 16158 + }, + { + "epoch": 7.512319851231985, + "grad_norm": 6.676338195800781, + "learning_rate": 9.983848998384901e-06, + "loss": 0.2045, + "step": 16159 + }, + { + "epoch": 7.512784751278475, + "grad_norm": 1.8288085460662842, + "learning_rate": 9.9838479983848e-06, + "loss": 0.0188, + "step": 16160 + }, + { + "epoch": 7.513249651324966, + "grad_norm": 3.98657488822937, + "learning_rate": 9.9838469983847e-06, + "loss": 0.0769, + "step": 16161 + }, + { + "epoch": 7.513714551371455, + "grad_norm": 4.044292449951172, + "learning_rate": 9.9838459983846e-06, + "loss": 0.071, + "step": 16162 + }, + { + "epoch": 7.514179451417945, + "grad_norm": 6.637384414672852, + "learning_rate": 9.9838449983845e-06, + "loss": 0.2252, + "step": 16163 + }, + { + "epoch": 7.514644351464435, + "grad_norm": 7.428062438964844, + "learning_rate": 9.983843998384401e-06, + "loss": 0.2416, + "step": 16164 + }, + { + "epoch": 7.515109251510925, + "grad_norm": 7.263986110687256, + "learning_rate": 9.983842998384302e-06, + "loss": 0.2856, + "step": 16165 + }, + { + "epoch": 7.5155741515574155, + "grad_norm": 4.414648532867432, + "learning_rate": 9.9838419983842e-06, + "loss": 0.0826, + "step": 16166 + }, + { + "epoch": 7.516039051603905, + "grad_norm": 3.272324562072754, + "learning_rate": 9.983840998384101e-06, + "loss": 0.0464, + "step": 16167 + }, + { + "epoch": 7.516503951650395, + "grad_norm": 2.2348313331604004, + "learning_rate": 9.983839998384e-06, + "loss": 0.0309, + "step": 16168 + }, + { + "epoch": 7.516968851696885, + "grad_norm": 5.7428765296936035, + "learning_rate": 9.9838389983839e-06, + "loss": 0.2501, + "step": 16169 + }, + { + "epoch": 7.517433751743376, + "grad_norm": 8.482782363891602, + "learning_rate": 9.9838379983838e-06, + "loss": 0.2629, + "step": 16170 + }, + { + "epoch": 7.517898651789865, + "grad_norm": 5.298795700073242, + "learning_rate": 9.9838369983837e-06, + "loss": 0.1411, + "step": 16171 + }, + { + "epoch": 7.518363551836355, + "grad_norm": 6.519645690917969, + "learning_rate": 9.983835998383601e-06, + "loss": 0.1311, + "step": 16172 + }, + { + "epoch": 7.518828451882845, + "grad_norm": 3.908754587173462, + "learning_rate": 9.9838349983835e-06, + "loss": 0.0991, + "step": 16173 + }, + { + "epoch": 7.519293351929335, + "grad_norm": 5.835596561431885, + "learning_rate": 9.983833998383401e-06, + "loss": 0.1761, + "step": 16174 + }, + { + "epoch": 7.5197582519758255, + "grad_norm": 2.8763229846954346, + "learning_rate": 9.9838329983833e-06, + "loss": 0.0315, + "step": 16175 + }, + { + "epoch": 7.520223152022315, + "grad_norm": 5.091221332550049, + "learning_rate": 9.9838319983832e-06, + "loss": 0.1779, + "step": 16176 + }, + { + "epoch": 7.520688052068805, + "grad_norm": 4.024824142456055, + "learning_rate": 9.983830998383101e-06, + "loss": 0.0595, + "step": 16177 + }, + { + "epoch": 7.521152952115295, + "grad_norm": 9.446917533874512, + "learning_rate": 9.983829998383002e-06, + "loss": 0.1482, + "step": 16178 + }, + { + "epoch": 7.521617852161786, + "grad_norm": 7.955252170562744, + "learning_rate": 9.9838289983829e-06, + "loss": 0.2715, + "step": 16179 + }, + { + "epoch": 7.522082752208275, + "grad_norm": 7.807054042816162, + "learning_rate": 9.9838279983828e-06, + "loss": 0.2757, + "step": 16180 + }, + { + "epoch": 7.522547652254765, + "grad_norm": 3.9464051723480225, + "learning_rate": 9.9838269983827e-06, + "loss": 0.0925, + "step": 16181 + }, + { + "epoch": 7.523012552301255, + "grad_norm": 5.01908540725708, + "learning_rate": 9.983825998382601e-06, + "loss": 0.1019, + "step": 16182 + }, + { + "epoch": 7.523477452347745, + "grad_norm": 2.736201524734497, + "learning_rate": 9.9838249983825e-06, + "loss": 0.0632, + "step": 16183 + }, + { + "epoch": 7.5239423523942355, + "grad_norm": 5.7511138916015625, + "learning_rate": 9.9838239983824e-06, + "loss": 0.1338, + "step": 16184 + }, + { + "epoch": 7.524407252440725, + "grad_norm": 4.901595592498779, + "learning_rate": 9.983822998382301e-06, + "loss": 0.095, + "step": 16185 + }, + { + "epoch": 7.524872152487215, + "grad_norm": 4.944225788116455, + "learning_rate": 9.9838219983822e-06, + "loss": 0.1081, + "step": 16186 + }, + { + "epoch": 7.525337052533705, + "grad_norm": 6.2475104331970215, + "learning_rate": 9.983820998382101e-06, + "loss": 0.1569, + "step": 16187 + }, + { + "epoch": 7.525801952580196, + "grad_norm": 3.421513795852661, + "learning_rate": 9.983819998382e-06, + "loss": 0.0483, + "step": 16188 + }, + { + "epoch": 7.526266852626685, + "grad_norm": 2.8974287509918213, + "learning_rate": 9.9838189983819e-06, + "loss": 0.0662, + "step": 16189 + }, + { + "epoch": 7.526731752673175, + "grad_norm": 6.278094291687012, + "learning_rate": 9.983817998381801e-06, + "loss": 0.1251, + "step": 16190 + }, + { + "epoch": 7.527196652719665, + "grad_norm": 7.63355016708374, + "learning_rate": 9.9838169983817e-06, + "loss": 0.2007, + "step": 16191 + }, + { + "epoch": 7.527661552766155, + "grad_norm": 4.349850654602051, + "learning_rate": 9.983815998381601e-06, + "loss": 0.1, + "step": 16192 + }, + { + "epoch": 7.5281264528126455, + "grad_norm": 9.072832107543945, + "learning_rate": 9.9838149983815e-06, + "loss": 0.2964, + "step": 16193 + }, + { + "epoch": 7.528591352859135, + "grad_norm": 5.436173439025879, + "learning_rate": 9.9838139983814e-06, + "loss": 0.1831, + "step": 16194 + }, + { + "epoch": 7.529056252905625, + "grad_norm": 9.451641082763672, + "learning_rate": 9.983812998381301e-06, + "loss": 0.1549, + "step": 16195 + }, + { + "epoch": 7.529521152952116, + "grad_norm": 16.922250747680664, + "learning_rate": 9.983811998381202e-06, + "loss": 0.9101, + "step": 16196 + }, + { + "epoch": 7.529986052998606, + "grad_norm": 4.585453033447266, + "learning_rate": 9.9838109983811e-06, + "loss": 0.0805, + "step": 16197 + }, + { + "epoch": 7.530450953045095, + "grad_norm": 5.45883321762085, + "learning_rate": 9.983809998381001e-06, + "loss": 0.1756, + "step": 16198 + }, + { + "epoch": 7.530915853091585, + "grad_norm": 3.687472343444824, + "learning_rate": 9.9838089983809e-06, + "loss": 0.074, + "step": 16199 + }, + { + "epoch": 7.531380753138075, + "grad_norm": 5.305160045623779, + "learning_rate": 9.983807998380801e-06, + "loss": 0.1309, + "step": 16200 + }, + { + "epoch": 7.531845653184566, + "grad_norm": 4.8095383644104, + "learning_rate": 9.9838069983807e-06, + "loss": 0.1265, + "step": 16201 + }, + { + "epoch": 7.5323105532310555, + "grad_norm": 6.498210430145264, + "learning_rate": 9.9838059983806e-06, + "loss": 0.3798, + "step": 16202 + }, + { + "epoch": 7.532775453277545, + "grad_norm": 5.424859046936035, + "learning_rate": 9.983804998380501e-06, + "loss": 0.0864, + "step": 16203 + }, + { + "epoch": 7.533240353324035, + "grad_norm": 8.388383865356445, + "learning_rate": 9.9838039983804e-06, + "loss": 0.4529, + "step": 16204 + }, + { + "epoch": 7.533705253370526, + "grad_norm": 5.101362228393555, + "learning_rate": 9.983802998380301e-06, + "loss": 0.1227, + "step": 16205 + }, + { + "epoch": 7.534170153417016, + "grad_norm": 7.543402671813965, + "learning_rate": 9.9838019983802e-06, + "loss": 0.2321, + "step": 16206 + }, + { + "epoch": 7.534635053463505, + "grad_norm": 4.9092020988464355, + "learning_rate": 9.9838009983801e-06, + "loss": 0.1733, + "step": 16207 + }, + { + "epoch": 7.535099953509995, + "grad_norm": 8.066542625427246, + "learning_rate": 9.983799998380001e-06, + "loss": 0.0503, + "step": 16208 + }, + { + "epoch": 7.535564853556485, + "grad_norm": 18.078176498413086, + "learning_rate": 9.983798998379902e-06, + "loss": 0.3257, + "step": 16209 + }, + { + "epoch": 7.536029753602976, + "grad_norm": 1.1623510122299194, + "learning_rate": 9.983797998379801e-06, + "loss": 0.0104, + "step": 16210 + }, + { + "epoch": 7.5364946536494655, + "grad_norm": 5.684873104095459, + "learning_rate": 9.983796998379702e-06, + "loss": 0.2431, + "step": 16211 + }, + { + "epoch": 7.536959553695955, + "grad_norm": 5.759188175201416, + "learning_rate": 9.9837959983796e-06, + "loss": 0.1787, + "step": 16212 + }, + { + "epoch": 7.537424453742445, + "grad_norm": 8.262840270996094, + "learning_rate": 9.9837949983795e-06, + "loss": 0.2793, + "step": 16213 + }, + { + "epoch": 7.537889353788936, + "grad_norm": 3.1506142616271973, + "learning_rate": 9.9837939983794e-06, + "loss": 0.0547, + "step": 16214 + }, + { + "epoch": 7.538354253835426, + "grad_norm": 7.25428581237793, + "learning_rate": 9.9837929983793e-06, + "loss": 0.1672, + "step": 16215 + }, + { + "epoch": 7.538819153881915, + "grad_norm": 11.675622940063477, + "learning_rate": 9.983791998379201e-06, + "loss": 0.1862, + "step": 16216 + }, + { + "epoch": 7.539284053928405, + "grad_norm": 4.998941421508789, + "learning_rate": 9.9837909983791e-06, + "loss": 0.0899, + "step": 16217 + }, + { + "epoch": 7.539748953974895, + "grad_norm": 8.003083229064941, + "learning_rate": 9.983789998379001e-06, + "loss": 0.2692, + "step": 16218 + }, + { + "epoch": 7.540213854021386, + "grad_norm": 4.6418375968933105, + "learning_rate": 9.9837889983789e-06, + "loss": 0.1055, + "step": 16219 + }, + { + "epoch": 7.5406787540678755, + "grad_norm": 5.584416389465332, + "learning_rate": 9.9837879983788e-06, + "loss": 0.1485, + "step": 16220 + }, + { + "epoch": 7.541143654114365, + "grad_norm": 8.54051685333252, + "learning_rate": 9.983786998378701e-06, + "loss": 0.2301, + "step": 16221 + }, + { + "epoch": 7.541608554160855, + "grad_norm": 2.8744194507598877, + "learning_rate": 9.983785998378602e-06, + "loss": 0.0293, + "step": 16222 + }, + { + "epoch": 7.542073454207346, + "grad_norm": 3.0283265113830566, + "learning_rate": 9.983784998378501e-06, + "loss": 0.0424, + "step": 16223 + }, + { + "epoch": 7.542538354253836, + "grad_norm": 5.199313163757324, + "learning_rate": 9.9837839983784e-06, + "loss": 0.1021, + "step": 16224 + }, + { + "epoch": 7.543003254300325, + "grad_norm": 13.047661781311035, + "learning_rate": 9.9837829983783e-06, + "loss": 0.2575, + "step": 16225 + }, + { + "epoch": 7.543468154346815, + "grad_norm": 5.5659308433532715, + "learning_rate": 9.9837819983782e-06, + "loss": 0.1589, + "step": 16226 + }, + { + "epoch": 7.543933054393305, + "grad_norm": 2.958235740661621, + "learning_rate": 9.9837809983781e-06, + "loss": 0.0882, + "step": 16227 + }, + { + "epoch": 7.544397954439796, + "grad_norm": 5.061203956604004, + "learning_rate": 9.983779998378e-06, + "loss": 0.1898, + "step": 16228 + }, + { + "epoch": 7.5448628544862855, + "grad_norm": 2.8448402881622314, + "learning_rate": 9.983778998377901e-06, + "loss": 0.0492, + "step": 16229 + }, + { + "epoch": 7.545327754532775, + "grad_norm": 1.711064100265503, + "learning_rate": 9.9837779983778e-06, + "loss": 0.0221, + "step": 16230 + }, + { + "epoch": 7.545792654579266, + "grad_norm": 2.3622994422912598, + "learning_rate": 9.983776998377701e-06, + "loss": 0.0345, + "step": 16231 + }, + { + "epoch": 7.546257554625756, + "grad_norm": 5.049036502838135, + "learning_rate": 9.9837759983776e-06, + "loss": 0.1446, + "step": 16232 + }, + { + "epoch": 7.546722454672246, + "grad_norm": 6.517838954925537, + "learning_rate": 9.9837749983775e-06, + "loss": 0.199, + "step": 16233 + }, + { + "epoch": 7.547187354718735, + "grad_norm": 5.40577507019043, + "learning_rate": 9.983773998377401e-06, + "loss": 0.0794, + "step": 16234 + }, + { + "epoch": 7.547652254765225, + "grad_norm": 4.521163463592529, + "learning_rate": 9.9837729983773e-06, + "loss": 0.1346, + "step": 16235 + }, + { + "epoch": 7.548117154811716, + "grad_norm": 5.01801061630249, + "learning_rate": 9.983771998377201e-06, + "loss": 0.1411, + "step": 16236 + }, + { + "epoch": 7.548582054858206, + "grad_norm": 7.444967269897461, + "learning_rate": 9.9837709983771e-06, + "loss": 0.2847, + "step": 16237 + }, + { + "epoch": 7.5490469549046955, + "grad_norm": 7.0883612632751465, + "learning_rate": 9.983769998377e-06, + "loss": 0.375, + "step": 16238 + }, + { + "epoch": 7.549511854951185, + "grad_norm": 6.477017402648926, + "learning_rate": 9.9837689983769e-06, + "loss": 0.1706, + "step": 16239 + }, + { + "epoch": 7.549976754997676, + "grad_norm": 6.359900951385498, + "learning_rate": 9.9837679983768e-06, + "loss": 0.1966, + "step": 16240 + }, + { + "epoch": 7.550441655044166, + "grad_norm": 4.749046802520752, + "learning_rate": 9.983766998376701e-06, + "loss": 0.1472, + "step": 16241 + }, + { + "epoch": 7.550906555090656, + "grad_norm": 6.441154479980469, + "learning_rate": 9.983765998376602e-06, + "loss": 0.1351, + "step": 16242 + }, + { + "epoch": 7.551371455137145, + "grad_norm": 7.652149677276611, + "learning_rate": 9.9837649983765e-06, + "loss": 0.1122, + "step": 16243 + }, + { + "epoch": 7.551836355183635, + "grad_norm": 7.084753513336182, + "learning_rate": 9.983763998376401e-06, + "loss": 0.3549, + "step": 16244 + }, + { + "epoch": 7.552301255230126, + "grad_norm": 5.039938926696777, + "learning_rate": 9.9837629983763e-06, + "loss": 0.1593, + "step": 16245 + }, + { + "epoch": 7.552766155276616, + "grad_norm": 3.4187514781951904, + "learning_rate": 9.9837619983762e-06, + "loss": 0.0478, + "step": 16246 + }, + { + "epoch": 7.5532310553231055, + "grad_norm": 7.129341125488281, + "learning_rate": 9.983760998376101e-06, + "loss": 0.1349, + "step": 16247 + }, + { + "epoch": 7.553695955369595, + "grad_norm": 5.785648345947266, + "learning_rate": 9.983759998376e-06, + "loss": 0.1128, + "step": 16248 + }, + { + "epoch": 7.554160855416086, + "grad_norm": 4.8859543800354, + "learning_rate": 9.983758998375901e-06, + "loss": 0.1197, + "step": 16249 + }, + { + "epoch": 7.554625755462576, + "grad_norm": 5.549897193908691, + "learning_rate": 9.9837579983758e-06, + "loss": 0.1166, + "step": 16250 + }, + { + "epoch": 7.5550906555090656, + "grad_norm": 3.9923791885375977, + "learning_rate": 9.9837569983757e-06, + "loss": 0.0825, + "step": 16251 + }, + { + "epoch": 7.555555555555555, + "grad_norm": 2.257319211959839, + "learning_rate": 9.9837559983756e-06, + "loss": 0.0349, + "step": 16252 + }, + { + "epoch": 7.556020455602045, + "grad_norm": 9.20543384552002, + "learning_rate": 9.9837549983755e-06, + "loss": 0.2413, + "step": 16253 + }, + { + "epoch": 7.556485355648536, + "grad_norm": 9.438223838806152, + "learning_rate": 9.983753998375401e-06, + "loss": 0.1782, + "step": 16254 + }, + { + "epoch": 7.556950255695026, + "grad_norm": 4.5092363357543945, + "learning_rate": 9.983752998375302e-06, + "loss": 0.105, + "step": 16255 + }, + { + "epoch": 7.5574151557415155, + "grad_norm": 3.904146671295166, + "learning_rate": 9.9837519983752e-06, + "loss": 0.0758, + "step": 16256 + }, + { + "epoch": 7.557880055788005, + "grad_norm": 5.262264728546143, + "learning_rate": 9.9837509983751e-06, + "loss": 0.1515, + "step": 16257 + }, + { + "epoch": 7.558344955834496, + "grad_norm": 2.9004862308502197, + "learning_rate": 9.983749998375e-06, + "loss": 0.036, + "step": 16258 + }, + { + "epoch": 7.558809855880986, + "grad_norm": 8.41970157623291, + "learning_rate": 9.983748998374901e-06, + "loss": 0.0792, + "step": 16259 + }, + { + "epoch": 7.5592747559274756, + "grad_norm": 4.717407703399658, + "learning_rate": 9.983747998374802e-06, + "loss": 0.1005, + "step": 16260 + }, + { + "epoch": 7.559739655973965, + "grad_norm": 5.832355499267578, + "learning_rate": 9.9837469983747e-06, + "loss": 0.0933, + "step": 16261 + }, + { + "epoch": 7.560204556020455, + "grad_norm": 7.582959175109863, + "learning_rate": 9.983745998374601e-06, + "loss": 0.3372, + "step": 16262 + }, + { + "epoch": 7.560669456066946, + "grad_norm": 4.7715301513671875, + "learning_rate": 9.9837449983745e-06, + "loss": 0.0611, + "step": 16263 + }, + { + "epoch": 7.561134356113436, + "grad_norm": 4.514388084411621, + "learning_rate": 9.9837439983744e-06, + "loss": 0.0757, + "step": 16264 + }, + { + "epoch": 7.5615992561599255, + "grad_norm": 5.8966827392578125, + "learning_rate": 9.983742998374301e-06, + "loss": 0.1416, + "step": 16265 + }, + { + "epoch": 7.562064156206415, + "grad_norm": 4.254491806030273, + "learning_rate": 9.9837419983742e-06, + "loss": 0.0899, + "step": 16266 + }, + { + "epoch": 7.562529056252906, + "grad_norm": 5.617269515991211, + "learning_rate": 9.983740998374101e-06, + "loss": 0.15, + "step": 16267 + }, + { + "epoch": 7.562993956299396, + "grad_norm": 4.52694845199585, + "learning_rate": 9.983739998374e-06, + "loss": 0.1225, + "step": 16268 + }, + { + "epoch": 7.5634588563458856, + "grad_norm": 3.336127281188965, + "learning_rate": 9.9837389983739e-06, + "loss": 0.0809, + "step": 16269 + }, + { + "epoch": 7.563923756392375, + "grad_norm": 6.064873695373535, + "learning_rate": 9.9837379983738e-06, + "loss": 0.1181, + "step": 16270 + }, + { + "epoch": 7.564388656438865, + "grad_norm": 4.06451416015625, + "learning_rate": 9.9837369983737e-06, + "loss": 0.0737, + "step": 16271 + }, + { + "epoch": 7.564853556485356, + "grad_norm": 7.327414512634277, + "learning_rate": 9.983735998373601e-06, + "loss": 0.315, + "step": 16272 + }, + { + "epoch": 7.565318456531846, + "grad_norm": 6.904183387756348, + "learning_rate": 9.983734998373502e-06, + "loss": 0.2205, + "step": 16273 + }, + { + "epoch": 7.5657833565783355, + "grad_norm": 5.5166544914245605, + "learning_rate": 9.9837339983734e-06, + "loss": 0.1949, + "step": 16274 + }, + { + "epoch": 7.566248256624826, + "grad_norm": 5.557531356811523, + "learning_rate": 9.983732998373301e-06, + "loss": 0.1012, + "step": 16275 + }, + { + "epoch": 7.566713156671316, + "grad_norm": 5.93145751953125, + "learning_rate": 9.9837319983732e-06, + "loss": 0.07, + "step": 16276 + }, + { + "epoch": 7.567178056717806, + "grad_norm": 5.396517753601074, + "learning_rate": 9.983730998373101e-06, + "loss": 0.1095, + "step": 16277 + }, + { + "epoch": 7.5676429567642955, + "grad_norm": 5.603152275085449, + "learning_rate": 9.983729998373002e-06, + "loss": 0.1303, + "step": 16278 + }, + { + "epoch": 7.568107856810785, + "grad_norm": 6.105605125427246, + "learning_rate": 9.9837289983729e-06, + "loss": 0.1385, + "step": 16279 + }, + { + "epoch": 7.568572756857276, + "grad_norm": 5.754682540893555, + "learning_rate": 9.983727998372801e-06, + "loss": 0.1292, + "step": 16280 + }, + { + "epoch": 7.569037656903766, + "grad_norm": 5.739464282989502, + "learning_rate": 9.9837269983727e-06, + "loss": 0.1748, + "step": 16281 + }, + { + "epoch": 7.569502556950256, + "grad_norm": 6.181435585021973, + "learning_rate": 9.9837259983726e-06, + "loss": 0.1612, + "step": 16282 + }, + { + "epoch": 7.5699674569967454, + "grad_norm": 3.9821174144744873, + "learning_rate": 9.9837249983725e-06, + "loss": 0.1107, + "step": 16283 + }, + { + "epoch": 7.570432357043236, + "grad_norm": 4.786909580230713, + "learning_rate": 9.9837239983724e-06, + "loss": 0.0386, + "step": 16284 + }, + { + "epoch": 7.570897257089726, + "grad_norm": 4.80936336517334, + "learning_rate": 9.983722998372301e-06, + "loss": 0.0912, + "step": 16285 + }, + { + "epoch": 7.571362157136216, + "grad_norm": 7.96278715133667, + "learning_rate": 9.983721998372202e-06, + "loss": 0.1272, + "step": 16286 + }, + { + "epoch": 7.5718270571827055, + "grad_norm": 7.273451805114746, + "learning_rate": 9.9837209983721e-06, + "loss": 0.0393, + "step": 16287 + }, + { + "epoch": 7.572291957229195, + "grad_norm": 4.512131690979004, + "learning_rate": 9.983719998372e-06, + "loss": 0.0808, + "step": 16288 + }, + { + "epoch": 7.572756857275686, + "grad_norm": 4.577925682067871, + "learning_rate": 9.9837189983719e-06, + "loss": 0.1117, + "step": 16289 + }, + { + "epoch": 7.573221757322176, + "grad_norm": 6.944225311279297, + "learning_rate": 9.983717998371801e-06, + "loss": 0.1602, + "step": 16290 + }, + { + "epoch": 7.573686657368666, + "grad_norm": 9.071982383728027, + "learning_rate": 9.983716998371702e-06, + "loss": 0.4192, + "step": 16291 + }, + { + "epoch": 7.5741515574151554, + "grad_norm": 4.990314483642578, + "learning_rate": 9.9837159983716e-06, + "loss": 0.1814, + "step": 16292 + }, + { + "epoch": 7.574616457461646, + "grad_norm": 5.168087959289551, + "learning_rate": 9.983714998371501e-06, + "loss": 0.1086, + "step": 16293 + }, + { + "epoch": 7.575081357508136, + "grad_norm": 7.816757678985596, + "learning_rate": 9.9837139983714e-06, + "loss": 0.1695, + "step": 16294 + }, + { + "epoch": 7.575546257554626, + "grad_norm": 4.926120281219482, + "learning_rate": 9.983712998371301e-06, + "loss": 0.1083, + "step": 16295 + }, + { + "epoch": 7.5760111576011155, + "grad_norm": 6.011855602264404, + "learning_rate": 9.9837119983712e-06, + "loss": 0.2161, + "step": 16296 + }, + { + "epoch": 7.576476057647605, + "grad_norm": 6.514577865600586, + "learning_rate": 9.9837109983711e-06, + "loss": 0.3041, + "step": 16297 + }, + { + "epoch": 7.576940957694096, + "grad_norm": 1.8156609535217285, + "learning_rate": 9.983709998371001e-06, + "loss": 0.0194, + "step": 16298 + }, + { + "epoch": 7.577405857740586, + "grad_norm": 5.181534767150879, + "learning_rate": 9.9837089983709e-06, + "loss": 0.1517, + "step": 16299 + }, + { + "epoch": 7.577870757787076, + "grad_norm": 7.45181131362915, + "learning_rate": 9.9837079983708e-06, + "loss": 0.2138, + "step": 16300 + }, + { + "epoch": 7.578335657833565, + "grad_norm": 3.9846789836883545, + "learning_rate": 9.9837069983707e-06, + "loss": 0.0529, + "step": 16301 + }, + { + "epoch": 7.578800557880056, + "grad_norm": 4.716061592102051, + "learning_rate": 9.9837059983706e-06, + "loss": 0.0864, + "step": 16302 + }, + { + "epoch": 7.579265457926546, + "grad_norm": 5.3301215171813965, + "learning_rate": 9.983704998370501e-06, + "loss": 0.1595, + "step": 16303 + }, + { + "epoch": 7.579730357973036, + "grad_norm": 9.167200088500977, + "learning_rate": 9.983703998370402e-06, + "loss": 0.1595, + "step": 16304 + }, + { + "epoch": 7.5801952580195255, + "grad_norm": 6.038614749908447, + "learning_rate": 9.9837029983703e-06, + "loss": 0.1551, + "step": 16305 + }, + { + "epoch": 7.580660158066015, + "grad_norm": 4.678929805755615, + "learning_rate": 9.983701998370201e-06, + "loss": 0.0859, + "step": 16306 + }, + { + "epoch": 7.581125058112506, + "grad_norm": 2.0510780811309814, + "learning_rate": 9.9837009983701e-06, + "loss": 0.0348, + "step": 16307 + }, + { + "epoch": 7.581589958158996, + "grad_norm": 5.882523059844971, + "learning_rate": 9.983699998370001e-06, + "loss": 0.1605, + "step": 16308 + }, + { + "epoch": 7.582054858205486, + "grad_norm": 8.43363094329834, + "learning_rate": 9.9836989983699e-06, + "loss": 0.3619, + "step": 16309 + }, + { + "epoch": 7.582519758251975, + "grad_norm": 4.160376071929932, + "learning_rate": 9.9836979983698e-06, + "loss": 0.1009, + "step": 16310 + }, + { + "epoch": 7.582984658298466, + "grad_norm": 2.936068534851074, + "learning_rate": 9.983696998369701e-06, + "loss": 0.0659, + "step": 16311 + }, + { + "epoch": 7.583449558344956, + "grad_norm": 6.567098617553711, + "learning_rate": 9.9836959983696e-06, + "loss": 0.0925, + "step": 16312 + }, + { + "epoch": 7.583914458391446, + "grad_norm": 5.770237445831299, + "learning_rate": 9.9836949983695e-06, + "loss": 0.1333, + "step": 16313 + }, + { + "epoch": 7.5843793584379355, + "grad_norm": 6.505758285522461, + "learning_rate": 9.9836939983694e-06, + "loss": 0.2079, + "step": 16314 + }, + { + "epoch": 7.584844258484426, + "grad_norm": 6.935832977294922, + "learning_rate": 9.9836929983693e-06, + "loss": 0.1403, + "step": 16315 + }, + { + "epoch": 7.585309158530916, + "grad_norm": 3.844438314437866, + "learning_rate": 9.983691998369201e-06, + "loss": 0.0619, + "step": 16316 + }, + { + "epoch": 7.585774058577406, + "grad_norm": 5.541999816894531, + "learning_rate": 9.983690998369102e-06, + "loss": 0.1839, + "step": 16317 + }, + { + "epoch": 7.586238958623896, + "grad_norm": 7.846805095672607, + "learning_rate": 9.983689998369e-06, + "loss": 0.2127, + "step": 16318 + }, + { + "epoch": 7.586703858670386, + "grad_norm": 3.5536065101623535, + "learning_rate": 9.983688998368901e-06, + "loss": 0.0369, + "step": 16319 + }, + { + "epoch": 7.587168758716876, + "grad_norm": 3.4562768936157227, + "learning_rate": 9.9836879983688e-06, + "loss": 0.0514, + "step": 16320 + }, + { + "epoch": 7.587633658763366, + "grad_norm": 2.9766178131103516, + "learning_rate": 9.9836869983687e-06, + "loss": 0.0329, + "step": 16321 + }, + { + "epoch": 7.588098558809856, + "grad_norm": 5.761383533477783, + "learning_rate": 9.9836859983686e-06, + "loss": 0.2661, + "step": 16322 + }, + { + "epoch": 7.5885634588563455, + "grad_norm": 8.878092765808105, + "learning_rate": 9.9836849983685e-06, + "loss": 0.2712, + "step": 16323 + }, + { + "epoch": 7.589028358902836, + "grad_norm": 6.334620475769043, + "learning_rate": 9.983683998368401e-06, + "loss": 0.1578, + "step": 16324 + }, + { + "epoch": 7.589493258949326, + "grad_norm": 4.009832382202148, + "learning_rate": 9.9836829983683e-06, + "loss": 0.0623, + "step": 16325 + }, + { + "epoch": 7.589958158995816, + "grad_norm": 9.014543533325195, + "learning_rate": 9.983681998368201e-06, + "loss": 0.2063, + "step": 16326 + }, + { + "epoch": 7.590423059042306, + "grad_norm": 4.912768840789795, + "learning_rate": 9.9836809983681e-06, + "loss": 0.0862, + "step": 16327 + }, + { + "epoch": 7.590887959088796, + "grad_norm": 3.0828816890716553, + "learning_rate": 9.983679998368e-06, + "loss": 0.0633, + "step": 16328 + }, + { + "epoch": 7.591352859135286, + "grad_norm": 3.658604621887207, + "learning_rate": 9.983678998367901e-06, + "loss": 0.0713, + "step": 16329 + }, + { + "epoch": 7.591817759181776, + "grad_norm": 3.7273333072662354, + "learning_rate": 9.983677998367802e-06, + "loss": 0.0639, + "step": 16330 + }, + { + "epoch": 7.592282659228266, + "grad_norm": 5.252938270568848, + "learning_rate": 9.9836769983677e-06, + "loss": 0.1697, + "step": 16331 + }, + { + "epoch": 7.5927475592747555, + "grad_norm": 3.1530067920684814, + "learning_rate": 9.9836759983676e-06, + "loss": 0.0704, + "step": 16332 + }, + { + "epoch": 7.593212459321246, + "grad_norm": 3.651705265045166, + "learning_rate": 9.9836749983675e-06, + "loss": 0.1036, + "step": 16333 + }, + { + "epoch": 7.593677359367736, + "grad_norm": 5.047473430633545, + "learning_rate": 9.9836739983674e-06, + "loss": 0.0844, + "step": 16334 + }, + { + "epoch": 7.594142259414226, + "grad_norm": 3.760570526123047, + "learning_rate": 9.9836729983673e-06, + "loss": 0.092, + "step": 16335 + }, + { + "epoch": 7.594607159460716, + "grad_norm": 6.8869805335998535, + "learning_rate": 9.9836719983672e-06, + "loss": 0.2378, + "step": 16336 + }, + { + "epoch": 7.595072059507206, + "grad_norm": 3.043471097946167, + "learning_rate": 9.983670998367101e-06, + "loss": 0.0349, + "step": 16337 + }, + { + "epoch": 7.595536959553696, + "grad_norm": 3.7765579223632812, + "learning_rate": 9.983669998367e-06, + "loss": 0.1139, + "step": 16338 + }, + { + "epoch": 7.596001859600186, + "grad_norm": 4.444372177124023, + "learning_rate": 9.983668998366901e-06, + "loss": 0.1115, + "step": 16339 + }, + { + "epoch": 7.596466759646676, + "grad_norm": 3.788562059402466, + "learning_rate": 9.9836679983668e-06, + "loss": 0.0576, + "step": 16340 + }, + { + "epoch": 7.5969316596931655, + "grad_norm": 6.819499969482422, + "learning_rate": 9.9836669983667e-06, + "loss": 0.0988, + "step": 16341 + }, + { + "epoch": 7.597396559739656, + "grad_norm": 5.074859619140625, + "learning_rate": 9.983665998366601e-06, + "loss": 0.1717, + "step": 16342 + }, + { + "epoch": 7.597861459786146, + "grad_norm": 5.67247200012207, + "learning_rate": 9.9836649983665e-06, + "loss": 0.167, + "step": 16343 + }, + { + "epoch": 7.598326359832636, + "grad_norm": 4.024726867675781, + "learning_rate": 9.983663998366401e-06, + "loss": 0.0977, + "step": 16344 + }, + { + "epoch": 7.598791259879126, + "grad_norm": 3.835775375366211, + "learning_rate": 9.9836629983663e-06, + "loss": 0.0707, + "step": 16345 + }, + { + "epoch": 7.599256159925616, + "grad_norm": 1.2046570777893066, + "learning_rate": 9.9836619983662e-06, + "loss": 0.0132, + "step": 16346 + }, + { + "epoch": 7.599721059972106, + "grad_norm": 5.670752048492432, + "learning_rate": 9.983660998366101e-06, + "loss": 0.1545, + "step": 16347 + }, + { + "epoch": 7.600185960018596, + "grad_norm": 7.064894676208496, + "learning_rate": 9.983659998366e-06, + "loss": 0.1838, + "step": 16348 + }, + { + "epoch": 7.600650860065086, + "grad_norm": 3.365431547164917, + "learning_rate": 9.9836589983659e-06, + "loss": 0.0389, + "step": 16349 + }, + { + "epoch": 7.6011157601115755, + "grad_norm": 5.670388698577881, + "learning_rate": 9.983657998365801e-06, + "loss": 0.144, + "step": 16350 + }, + { + "epoch": 7.601580660158066, + "grad_norm": 6.219570636749268, + "learning_rate": 9.9836569983657e-06, + "loss": 0.1256, + "step": 16351 + }, + { + "epoch": 7.602045560204556, + "grad_norm": 3.7343316078186035, + "learning_rate": 9.983655998365601e-06, + "loss": 0.0448, + "step": 16352 + }, + { + "epoch": 7.602510460251046, + "grad_norm": 5.483452796936035, + "learning_rate": 9.9836549983655e-06, + "loss": 0.2226, + "step": 16353 + }, + { + "epoch": 7.6029753602975365, + "grad_norm": 4.140985012054443, + "learning_rate": 9.9836539983654e-06, + "loss": 0.0768, + "step": 16354 + }, + { + "epoch": 7.603440260344026, + "grad_norm": 5.560600280761719, + "learning_rate": 9.983652998365301e-06, + "loss": 0.0782, + "step": 16355 + }, + { + "epoch": 7.603905160390516, + "grad_norm": 6.313736915588379, + "learning_rate": 9.9836519983652e-06, + "loss": 0.1336, + "step": 16356 + }, + { + "epoch": 7.604370060437006, + "grad_norm": 10.274394035339355, + "learning_rate": 9.983650998365101e-06, + "loss": 0.4134, + "step": 16357 + }, + { + "epoch": 7.604834960483496, + "grad_norm": 5.419198513031006, + "learning_rate": 9.983649998365e-06, + "loss": 0.0745, + "step": 16358 + }, + { + "epoch": 7.605299860529986, + "grad_norm": 5.511902809143066, + "learning_rate": 9.9836489983649e-06, + "loss": 0.1436, + "step": 16359 + }, + { + "epoch": 7.605764760576476, + "grad_norm": 5.157042503356934, + "learning_rate": 9.983647998364801e-06, + "loss": 0.0725, + "step": 16360 + }, + { + "epoch": 7.606229660622966, + "grad_norm": 5.784443378448486, + "learning_rate": 9.983646998364702e-06, + "loss": 0.1812, + "step": 16361 + }, + { + "epoch": 7.606694560669456, + "grad_norm": 6.983968257904053, + "learning_rate": 9.983645998364601e-06, + "loss": 0.2337, + "step": 16362 + }, + { + "epoch": 7.6071594607159465, + "grad_norm": 6.5941290855407715, + "learning_rate": 9.983644998364502e-06, + "loss": 0.2061, + "step": 16363 + }, + { + "epoch": 7.607624360762436, + "grad_norm": 7.383811950683594, + "learning_rate": 9.9836439983644e-06, + "loss": 0.146, + "step": 16364 + }, + { + "epoch": 7.608089260808926, + "grad_norm": 4.030415058135986, + "learning_rate": 9.9836429983643e-06, + "loss": 0.0775, + "step": 16365 + }, + { + "epoch": 7.608554160855416, + "grad_norm": 3.5042762756347656, + "learning_rate": 9.9836419983642e-06, + "loss": 0.0769, + "step": 16366 + }, + { + "epoch": 7.609019060901906, + "grad_norm": 6.989298343658447, + "learning_rate": 9.9836409983641e-06, + "loss": 0.1902, + "step": 16367 + }, + { + "epoch": 7.609483960948396, + "grad_norm": 5.2466721534729, + "learning_rate": 9.983639998364001e-06, + "loss": 0.114, + "step": 16368 + }, + { + "epoch": 7.609948860994886, + "grad_norm": 7.7459917068481445, + "learning_rate": 9.9836389983639e-06, + "loss": 0.2368, + "step": 16369 + }, + { + "epoch": 7.610413761041376, + "grad_norm": 2.038618326187134, + "learning_rate": 9.983637998363801e-06, + "loss": 0.0323, + "step": 16370 + }, + { + "epoch": 7.610878661087866, + "grad_norm": 4.887950897216797, + "learning_rate": 9.9836369983637e-06, + "loss": 0.1497, + "step": 16371 + }, + { + "epoch": 7.6113435611343565, + "grad_norm": 8.122472763061523, + "learning_rate": 9.9836359983636e-06, + "loss": 0.3019, + "step": 16372 + }, + { + "epoch": 7.611808461180846, + "grad_norm": 3.5762441158294678, + "learning_rate": 9.983634998363501e-06, + "loss": 0.0708, + "step": 16373 + }, + { + "epoch": 7.612273361227336, + "grad_norm": 8.21401596069336, + "learning_rate": 9.983633998363402e-06, + "loss": 0.2568, + "step": 16374 + }, + { + "epoch": 7.612738261273826, + "grad_norm": 5.1269354820251465, + "learning_rate": 9.983632998363301e-06, + "loss": 0.1564, + "step": 16375 + }, + { + "epoch": 7.613203161320316, + "grad_norm": 5.526599407196045, + "learning_rate": 9.9836319983632e-06, + "loss": 0.1733, + "step": 16376 + }, + { + "epoch": 7.613668061366806, + "grad_norm": 3.9333319664001465, + "learning_rate": 9.9836309983631e-06, + "loss": 0.0507, + "step": 16377 + }, + { + "epoch": 7.614132961413296, + "grad_norm": 4.153718948364258, + "learning_rate": 9.983629998363e-06, + "loss": 0.0991, + "step": 16378 + }, + { + "epoch": 7.614597861459786, + "grad_norm": 5.036649703979492, + "learning_rate": 9.9836289983629e-06, + "loss": 0.1543, + "step": 16379 + }, + { + "epoch": 7.615062761506276, + "grad_norm": 5.926989555358887, + "learning_rate": 9.983627998362801e-06, + "loss": 0.1313, + "step": 16380 + }, + { + "epoch": 7.6155276615527665, + "grad_norm": 5.876142501831055, + "learning_rate": 9.983626998362702e-06, + "loss": 0.1044, + "step": 16381 + }, + { + "epoch": 7.615992561599256, + "grad_norm": 6.966691970825195, + "learning_rate": 9.9836259983626e-06, + "loss": 0.3153, + "step": 16382 + }, + { + "epoch": 7.616457461645746, + "grad_norm": 4.5843353271484375, + "learning_rate": 9.983624998362501e-06, + "loss": 0.0641, + "step": 16383 + }, + { + "epoch": 7.616922361692236, + "grad_norm": 4.84965705871582, + "learning_rate": 9.9836239983624e-06, + "loss": 0.1822, + "step": 16384 + }, + { + "epoch": 7.617387261738726, + "grad_norm": 4.739159107208252, + "learning_rate": 9.9836229983623e-06, + "loss": 0.0839, + "step": 16385 + }, + { + "epoch": 7.617852161785216, + "grad_norm": 4.488406181335449, + "learning_rate": 9.983621998362201e-06, + "loss": 0.0819, + "step": 16386 + }, + { + "epoch": 7.618317061831706, + "grad_norm": 4.762197017669678, + "learning_rate": 9.9836209983621e-06, + "loss": 0.1106, + "step": 16387 + }, + { + "epoch": 7.618781961878196, + "grad_norm": 5.929954528808594, + "learning_rate": 9.983619998362001e-06, + "loss": 0.1094, + "step": 16388 + }, + { + "epoch": 7.619246861924686, + "grad_norm": 5.624749660491943, + "learning_rate": 9.9836189983619e-06, + "loss": 0.2349, + "step": 16389 + }, + { + "epoch": 7.6197117619711765, + "grad_norm": 7.38154411315918, + "learning_rate": 9.9836179983618e-06, + "loss": 0.1437, + "step": 16390 + }, + { + "epoch": 7.620176662017666, + "grad_norm": 4.535342216491699, + "learning_rate": 9.9836169983617e-06, + "loss": 0.1064, + "step": 16391 + }, + { + "epoch": 7.620641562064156, + "grad_norm": 4.767557621002197, + "learning_rate": 9.9836159983616e-06, + "loss": 0.2096, + "step": 16392 + }, + { + "epoch": 7.621106462110646, + "grad_norm": 4.443418025970459, + "learning_rate": 9.983614998361501e-06, + "loss": 0.0798, + "step": 16393 + }, + { + "epoch": 7.621571362157137, + "grad_norm": 7.433175086975098, + "learning_rate": 9.983613998361402e-06, + "loss": 0.0855, + "step": 16394 + }, + { + "epoch": 7.622036262203626, + "grad_norm": 6.8624467849731445, + "learning_rate": 9.9836129983613e-06, + "loss": 0.1677, + "step": 16395 + }, + { + "epoch": 7.622501162250116, + "grad_norm": 5.010804653167725, + "learning_rate": 9.983611998361201e-06, + "loss": 0.1206, + "step": 16396 + }, + { + "epoch": 7.622966062296606, + "grad_norm": 4.545330047607422, + "learning_rate": 9.9836109983611e-06, + "loss": 0.0836, + "step": 16397 + }, + { + "epoch": 7.623430962343097, + "grad_norm": 7.028510570526123, + "learning_rate": 9.983609998361001e-06, + "loss": 0.2284, + "step": 16398 + }, + { + "epoch": 7.6238958623895865, + "grad_norm": 5.381852626800537, + "learning_rate": 9.983608998360902e-06, + "loss": 0.1574, + "step": 16399 + }, + { + "epoch": 7.624360762436076, + "grad_norm": 4.689038276672363, + "learning_rate": 9.9836079983608e-06, + "loss": 0.1384, + "step": 16400 + }, + { + "epoch": 7.624825662482566, + "grad_norm": 3.9648172855377197, + "learning_rate": 9.983606998360701e-06, + "loss": 0.0732, + "step": 16401 + }, + { + "epoch": 7.625290562529056, + "grad_norm": 8.908502578735352, + "learning_rate": 9.9836059983606e-06, + "loss": 0.3158, + "step": 16402 + }, + { + "epoch": 7.625755462575547, + "grad_norm": 5.056369304656982, + "learning_rate": 9.9836049983605e-06, + "loss": 0.1449, + "step": 16403 + }, + { + "epoch": 7.626220362622036, + "grad_norm": 7.845433712005615, + "learning_rate": 9.9836039983604e-06, + "loss": 0.142, + "step": 16404 + }, + { + "epoch": 7.626685262668526, + "grad_norm": 5.859767436981201, + "learning_rate": 9.9836029983603e-06, + "loss": 0.0995, + "step": 16405 + }, + { + "epoch": 7.627150162715016, + "grad_norm": 5.493223667144775, + "learning_rate": 9.983601998360201e-06, + "loss": 0.0677, + "step": 16406 + }, + { + "epoch": 7.627615062761507, + "grad_norm": 4.49765157699585, + "learning_rate": 9.9836009983601e-06, + "loss": 0.1163, + "step": 16407 + }, + { + "epoch": 7.6280799628079965, + "grad_norm": 4.705824375152588, + "learning_rate": 9.98359999836e-06, + "loss": 0.081, + "step": 16408 + }, + { + "epoch": 7.628544862854486, + "grad_norm": 4.0934247970581055, + "learning_rate": 9.9835989983599e-06, + "loss": 0.075, + "step": 16409 + }, + { + "epoch": 7.629009762900976, + "grad_norm": 5.508029460906982, + "learning_rate": 9.9835979983598e-06, + "loss": 0.1113, + "step": 16410 + }, + { + "epoch": 7.629474662947466, + "grad_norm": 3.5582947731018066, + "learning_rate": 9.983596998359701e-06, + "loss": 0.0573, + "step": 16411 + }, + { + "epoch": 7.629939562993957, + "grad_norm": 3.9317123889923096, + "learning_rate": 9.983595998359602e-06, + "loss": 0.0871, + "step": 16412 + }, + { + "epoch": 7.630404463040446, + "grad_norm": 9.740958213806152, + "learning_rate": 9.9835949983595e-06, + "loss": 0.2042, + "step": 16413 + }, + { + "epoch": 7.630869363086936, + "grad_norm": 4.2509074211120605, + "learning_rate": 9.983593998359401e-06, + "loss": 0.0806, + "step": 16414 + }, + { + "epoch": 7.631334263133426, + "grad_norm": 5.324028491973877, + "learning_rate": 9.9835929983593e-06, + "loss": 0.1207, + "step": 16415 + }, + { + "epoch": 7.631799163179917, + "grad_norm": 5.112666130065918, + "learning_rate": 9.9835919983592e-06, + "loss": 0.1548, + "step": 16416 + }, + { + "epoch": 7.6322640632264065, + "grad_norm": 3.9060487747192383, + "learning_rate": 9.9835909983591e-06, + "loss": 0.0643, + "step": 16417 + }, + { + "epoch": 7.632728963272896, + "grad_norm": 6.565614700317383, + "learning_rate": 9.983589998359e-06, + "loss": 0.1514, + "step": 16418 + }, + { + "epoch": 7.633193863319386, + "grad_norm": 8.020625114440918, + "learning_rate": 9.983588998358901e-06, + "loss": 0.1557, + "step": 16419 + }, + { + "epoch": 7.633658763365876, + "grad_norm": 5.42484712600708, + "learning_rate": 9.9835879983588e-06, + "loss": 0.134, + "step": 16420 + }, + { + "epoch": 7.634123663412367, + "grad_norm": 3.161017656326294, + "learning_rate": 9.9835869983587e-06, + "loss": 0.0427, + "step": 16421 + }, + { + "epoch": 7.634588563458856, + "grad_norm": 3.9840598106384277, + "learning_rate": 9.9835859983586e-06, + "loss": 0.1293, + "step": 16422 + }, + { + "epoch": 7.635053463505346, + "grad_norm": 8.290270805358887, + "learning_rate": 9.9835849983585e-06, + "loss": 0.1941, + "step": 16423 + }, + { + "epoch": 7.635518363551836, + "grad_norm": 13.067657470703125, + "learning_rate": 9.983583998358401e-06, + "loss": 0.2714, + "step": 16424 + }, + { + "epoch": 7.635983263598327, + "grad_norm": 1.676960825920105, + "learning_rate": 9.983582998358302e-06, + "loss": 0.0187, + "step": 16425 + }, + { + "epoch": 7.6364481636448165, + "grad_norm": 5.431075096130371, + "learning_rate": 9.9835819983582e-06, + "loss": 0.1668, + "step": 16426 + }, + { + "epoch": 7.636913063691306, + "grad_norm": 1.0545680522918701, + "learning_rate": 9.983580998358101e-06, + "loss": 0.0085, + "step": 16427 + }, + { + "epoch": 7.637377963737796, + "grad_norm": 3.011000871658325, + "learning_rate": 9.983579998358e-06, + "loss": 0.0356, + "step": 16428 + }, + { + "epoch": 7.637842863784286, + "grad_norm": 5.662883758544922, + "learning_rate": 9.983578998357901e-06, + "loss": 0.1627, + "step": 16429 + }, + { + "epoch": 7.638307763830777, + "grad_norm": 4.789484024047852, + "learning_rate": 9.983577998357802e-06, + "loss": 0.0759, + "step": 16430 + }, + { + "epoch": 7.638772663877266, + "grad_norm": 4.45576286315918, + "learning_rate": 9.9835769983577e-06, + "loss": 0.1256, + "step": 16431 + }, + { + "epoch": 7.639237563923756, + "grad_norm": 4.129849910736084, + "learning_rate": 9.983575998357601e-06, + "loss": 0.0654, + "step": 16432 + }, + { + "epoch": 7.639702463970247, + "grad_norm": 6.412408351898193, + "learning_rate": 9.9835749983575e-06, + "loss": 0.1556, + "step": 16433 + }, + { + "epoch": 7.640167364016737, + "grad_norm": 5.4616169929504395, + "learning_rate": 9.9835739983574e-06, + "loss": 0.1182, + "step": 16434 + }, + { + "epoch": 7.6406322640632265, + "grad_norm": 5.569368362426758, + "learning_rate": 9.9835729983573e-06, + "loss": 0.2298, + "step": 16435 + }, + { + "epoch": 7.641097164109716, + "grad_norm": 4.878380298614502, + "learning_rate": 9.9835719983572e-06, + "loss": 0.1319, + "step": 16436 + }, + { + "epoch": 7.641562064156206, + "grad_norm": 6.575472354888916, + "learning_rate": 9.983570998357101e-06, + "loss": 0.2398, + "step": 16437 + }, + { + "epoch": 7.642026964202697, + "grad_norm": 4.691313743591309, + "learning_rate": 9.983569998357002e-06, + "loss": 0.085, + "step": 16438 + }, + { + "epoch": 7.642491864249187, + "grad_norm": 4.628913879394531, + "learning_rate": 9.9835689983569e-06, + "loss": 0.1117, + "step": 16439 + }, + { + "epoch": 7.642956764295676, + "grad_norm": 5.030702114105225, + "learning_rate": 9.9835679983568e-06, + "loss": 0.1394, + "step": 16440 + }, + { + "epoch": 7.643421664342166, + "grad_norm": 3.5139565467834473, + "learning_rate": 9.9835669983567e-06, + "loss": 0.084, + "step": 16441 + }, + { + "epoch": 7.643886564388657, + "grad_norm": 7.093101978302002, + "learning_rate": 9.983565998356601e-06, + "loss": 0.3166, + "step": 16442 + }, + { + "epoch": 7.644351464435147, + "grad_norm": 3.0615358352661133, + "learning_rate": 9.983564998356502e-06, + "loss": 0.0559, + "step": 16443 + }, + { + "epoch": 7.6448163644816365, + "grad_norm": 2.1667981147766113, + "learning_rate": 9.9835639983564e-06, + "loss": 0.0378, + "step": 16444 + }, + { + "epoch": 7.645281264528126, + "grad_norm": 5.535441875457764, + "learning_rate": 9.983562998356301e-06, + "loss": 0.1781, + "step": 16445 + }, + { + "epoch": 7.645746164574616, + "grad_norm": 4.432277202606201, + "learning_rate": 9.9835619983562e-06, + "loss": 0.144, + "step": 16446 + }, + { + "epoch": 7.646211064621107, + "grad_norm": 5.5778727531433105, + "learning_rate": 9.983560998356101e-06, + "loss": 0.177, + "step": 16447 + }, + { + "epoch": 7.646675964667597, + "grad_norm": 5.277877330780029, + "learning_rate": 9.983559998356e-06, + "loss": 0.1215, + "step": 16448 + }, + { + "epoch": 7.647140864714086, + "grad_norm": 9.19054126739502, + "learning_rate": 9.9835589983559e-06, + "loss": 0.3386, + "step": 16449 + }, + { + "epoch": 7.647605764760576, + "grad_norm": 2.150320291519165, + "learning_rate": 9.983557998355801e-06, + "loss": 0.0275, + "step": 16450 + }, + { + "epoch": 7.648070664807067, + "grad_norm": 8.416337013244629, + "learning_rate": 9.9835569983557e-06, + "loss": 0.3011, + "step": 16451 + }, + { + "epoch": 7.648535564853557, + "grad_norm": 5.164627552032471, + "learning_rate": 9.9835559983556e-06, + "loss": 0.124, + "step": 16452 + }, + { + "epoch": 7.6490004649000465, + "grad_norm": 4.076944828033447, + "learning_rate": 9.9835549983555e-06, + "loss": 0.0738, + "step": 16453 + }, + { + "epoch": 7.649465364946536, + "grad_norm": 5.970311164855957, + "learning_rate": 9.9835539983554e-06, + "loss": 0.1634, + "step": 16454 + }, + { + "epoch": 7.649930264993026, + "grad_norm": 7.097362995147705, + "learning_rate": 9.983552998355301e-06, + "loss": 0.1853, + "step": 16455 + }, + { + "epoch": 7.650395165039517, + "grad_norm": 3.802321672439575, + "learning_rate": 9.983551998355202e-06, + "loss": 0.0585, + "step": 16456 + }, + { + "epoch": 7.650860065086007, + "grad_norm": 6.241832256317139, + "learning_rate": 9.9835509983551e-06, + "loss": 0.1703, + "step": 16457 + }, + { + "epoch": 7.651324965132496, + "grad_norm": 4.151679992675781, + "learning_rate": 9.983549998355001e-06, + "loss": 0.0965, + "step": 16458 + }, + { + "epoch": 7.651789865178986, + "grad_norm": 2.8658957481384277, + "learning_rate": 9.9835489983549e-06, + "loss": 0.0349, + "step": 16459 + }, + { + "epoch": 7.652254765225477, + "grad_norm": 4.173716068267822, + "learning_rate": 9.983547998354801e-06, + "loss": 0.0693, + "step": 16460 + }, + { + "epoch": 7.652719665271967, + "grad_norm": 6.660766124725342, + "learning_rate": 9.9835469983547e-06, + "loss": 0.1721, + "step": 16461 + }, + { + "epoch": 7.6531845653184565, + "grad_norm": 5.719512462615967, + "learning_rate": 9.9835459983546e-06, + "loss": 0.1689, + "step": 16462 + }, + { + "epoch": 7.653649465364946, + "grad_norm": 5.653817176818848, + "learning_rate": 9.983544998354501e-06, + "loss": 0.1661, + "step": 16463 + }, + { + "epoch": 7.654114365411436, + "grad_norm": 6.217688083648682, + "learning_rate": 9.9835439983544e-06, + "loss": 0.0331, + "step": 16464 + }, + { + "epoch": 7.654579265457927, + "grad_norm": 4.246175289154053, + "learning_rate": 9.983542998354301e-06, + "loss": 0.0564, + "step": 16465 + }, + { + "epoch": 7.655044165504417, + "grad_norm": 7.061737060546875, + "learning_rate": 9.9835419983542e-06, + "loss": 0.2168, + "step": 16466 + }, + { + "epoch": 7.655509065550906, + "grad_norm": 9.011391639709473, + "learning_rate": 9.9835409983541e-06, + "loss": 0.4015, + "step": 16467 + }, + { + "epoch": 7.655973965597396, + "grad_norm": 7.942361354827881, + "learning_rate": 9.983539998354001e-06, + "loss": 0.1945, + "step": 16468 + }, + { + "epoch": 7.656438865643887, + "grad_norm": 5.659554481506348, + "learning_rate": 9.983538998353902e-06, + "loss": 0.1343, + "step": 16469 + }, + { + "epoch": 7.656903765690377, + "grad_norm": 1.782370686531067, + "learning_rate": 9.9835379983538e-06, + "loss": 0.019, + "step": 16470 + }, + { + "epoch": 7.6573686657368665, + "grad_norm": 8.086562156677246, + "learning_rate": 9.983536998353701e-06, + "loss": 0.1241, + "step": 16471 + }, + { + "epoch": 7.657833565783356, + "grad_norm": 5.153304576873779, + "learning_rate": 9.9835359983536e-06, + "loss": 0.2113, + "step": 16472 + }, + { + "epoch": 7.658298465829847, + "grad_norm": 7.635302543640137, + "learning_rate": 9.9835349983535e-06, + "loss": 0.1852, + "step": 16473 + }, + { + "epoch": 7.658763365876337, + "grad_norm": 2.341829299926758, + "learning_rate": 9.9835339983534e-06, + "loss": 0.0367, + "step": 16474 + }, + { + "epoch": 7.659228265922827, + "grad_norm": 2.996622085571289, + "learning_rate": 9.9835329983533e-06, + "loss": 0.0503, + "step": 16475 + }, + { + "epoch": 7.659693165969316, + "grad_norm": 6.665425777435303, + "learning_rate": 9.983531998353201e-06, + "loss": 0.2283, + "step": 16476 + }, + { + "epoch": 7.660158066015807, + "grad_norm": 3.6789119243621826, + "learning_rate": 9.9835309983531e-06, + "loss": 0.0687, + "step": 16477 + }, + { + "epoch": 7.660622966062297, + "grad_norm": 5.658687114715576, + "learning_rate": 9.983529998353001e-06, + "loss": 0.1385, + "step": 16478 + }, + { + "epoch": 7.661087866108787, + "grad_norm": 4.323816299438477, + "learning_rate": 9.9835289983529e-06, + "loss": 0.0948, + "step": 16479 + }, + { + "epoch": 7.6615527661552765, + "grad_norm": 4.5194010734558105, + "learning_rate": 9.9835279983528e-06, + "loss": 0.1949, + "step": 16480 + }, + { + "epoch": 7.662017666201766, + "grad_norm": 5.4978108406066895, + "learning_rate": 9.983526998352701e-06, + "loss": 0.1352, + "step": 16481 + }, + { + "epoch": 7.662482566248257, + "grad_norm": 4.849621295928955, + "learning_rate": 9.983525998352602e-06, + "loss": 0.0877, + "step": 16482 + }, + { + "epoch": 7.662947466294747, + "grad_norm": 3.1793441772460938, + "learning_rate": 9.983524998352501e-06, + "loss": 0.0582, + "step": 16483 + }, + { + "epoch": 7.663412366341237, + "grad_norm": 6.954376220703125, + "learning_rate": 9.9835239983524e-06, + "loss": 0.1046, + "step": 16484 + }, + { + "epoch": 7.663877266387726, + "grad_norm": 5.239450931549072, + "learning_rate": 9.9835229983523e-06, + "loss": 0.1098, + "step": 16485 + }, + { + "epoch": 7.664342166434217, + "grad_norm": 3.0405144691467285, + "learning_rate": 9.9835219983522e-06, + "loss": 0.0446, + "step": 16486 + }, + { + "epoch": 7.664807066480707, + "grad_norm": 4.364569187164307, + "learning_rate": 9.9835209983521e-06, + "loss": 0.0453, + "step": 16487 + }, + { + "epoch": 7.665271966527197, + "grad_norm": 5.704826354980469, + "learning_rate": 9.983519998352e-06, + "loss": 0.1202, + "step": 16488 + }, + { + "epoch": 7.6657368665736865, + "grad_norm": 7.062305927276611, + "learning_rate": 9.983518998351901e-06, + "loss": 0.2676, + "step": 16489 + }, + { + "epoch": 7.666201766620176, + "grad_norm": 8.258732795715332, + "learning_rate": 9.9835179983518e-06, + "loss": 0.2325, + "step": 16490 + }, + { + "epoch": 7.666666666666667, + "grad_norm": 6.770398139953613, + "learning_rate": 9.983516998351701e-06, + "loss": 0.0542, + "step": 16491 + }, + { + "epoch": 7.667131566713157, + "grad_norm": 4.844921588897705, + "learning_rate": 9.9835159983516e-06, + "loss": 0.1144, + "step": 16492 + }, + { + "epoch": 7.667596466759647, + "grad_norm": 5.073827266693115, + "learning_rate": 9.9835149983515e-06, + "loss": 0.0818, + "step": 16493 + }, + { + "epoch": 7.668061366806136, + "grad_norm": 5.913788795471191, + "learning_rate": 9.983513998351401e-06, + "loss": 0.119, + "step": 16494 + }, + { + "epoch": 7.668526266852627, + "grad_norm": 2.9692118167877197, + "learning_rate": 9.9835129983513e-06, + "loss": 0.0463, + "step": 16495 + }, + { + "epoch": 7.668991166899117, + "grad_norm": 6.275150775909424, + "learning_rate": 9.983511998351201e-06, + "loss": 0.1062, + "step": 16496 + }, + { + "epoch": 7.669456066945607, + "grad_norm": 8.405518531799316, + "learning_rate": 9.9835109983511e-06, + "loss": 0.3799, + "step": 16497 + }, + { + "epoch": 7.6699209669920965, + "grad_norm": 18.799453735351562, + "learning_rate": 9.983509998351e-06, + "loss": 1.162, + "step": 16498 + }, + { + "epoch": 7.670385867038586, + "grad_norm": 5.097768306732178, + "learning_rate": 9.983508998350901e-06, + "loss": 0.1037, + "step": 16499 + }, + { + "epoch": 7.670850767085077, + "grad_norm": 8.13638973236084, + "learning_rate": 9.9835079983508e-06, + "loss": 0.3997, + "step": 16500 + }, + { + "epoch": 7.671315667131567, + "grad_norm": 9.638580322265625, + "learning_rate": 9.9835069983507e-06, + "loss": 0.2788, + "step": 16501 + }, + { + "epoch": 7.671780567178057, + "grad_norm": 3.848860740661621, + "learning_rate": 9.983505998350601e-06, + "loss": 0.0907, + "step": 16502 + }, + { + "epoch": 7.672245467224546, + "grad_norm": 3.8033132553100586, + "learning_rate": 9.9835049983505e-06, + "loss": 0.081, + "step": 16503 + }, + { + "epoch": 7.672710367271037, + "grad_norm": 4.163052082061768, + "learning_rate": 9.983503998350401e-06, + "loss": 0.0973, + "step": 16504 + }, + { + "epoch": 7.673175267317527, + "grad_norm": 4.703883647918701, + "learning_rate": 9.9835029983503e-06, + "loss": 0.1541, + "step": 16505 + }, + { + "epoch": 7.673640167364017, + "grad_norm": 7.422368049621582, + "learning_rate": 9.9835019983502e-06, + "loss": 0.2135, + "step": 16506 + }, + { + "epoch": 7.6741050674105065, + "grad_norm": 5.918975353240967, + "learning_rate": 9.983500998350101e-06, + "loss": 0.1412, + "step": 16507 + }, + { + "epoch": 7.674569967456996, + "grad_norm": 3.7076056003570557, + "learning_rate": 9.98349999835e-06, + "loss": 0.0611, + "step": 16508 + }, + { + "epoch": 7.675034867503487, + "grad_norm": 4.724909782409668, + "learning_rate": 9.983498998349901e-06, + "loss": 0.1774, + "step": 16509 + }, + { + "epoch": 7.675499767549977, + "grad_norm": 5.635777950286865, + "learning_rate": 9.9834979983498e-06, + "loss": 0.1381, + "step": 16510 + }, + { + "epoch": 7.675964667596467, + "grad_norm": 4.225979328155518, + "learning_rate": 9.9834969983497e-06, + "loss": 0.0952, + "step": 16511 + }, + { + "epoch": 7.676429567642957, + "grad_norm": 10.299232482910156, + "learning_rate": 9.983495998349601e-06, + "loss": 0.1142, + "step": 16512 + }, + { + "epoch": 7.676894467689447, + "grad_norm": 5.485869884490967, + "learning_rate": 9.9834949983495e-06, + "loss": 0.1203, + "step": 16513 + }, + { + "epoch": 7.677359367735937, + "grad_norm": 2.4123575687408447, + "learning_rate": 9.983493998349401e-06, + "loss": 0.0298, + "step": 16514 + }, + { + "epoch": 7.677824267782427, + "grad_norm": 3.312546491622925, + "learning_rate": 9.983492998349302e-06, + "loss": 0.0597, + "step": 16515 + }, + { + "epoch": 7.6782891678289165, + "grad_norm": 2.302509069442749, + "learning_rate": 9.9834919983492e-06, + "loss": 0.0308, + "step": 16516 + }, + { + "epoch": 7.678754067875407, + "grad_norm": 8.03331184387207, + "learning_rate": 9.9834909983491e-06, + "loss": 0.1546, + "step": 16517 + }, + { + "epoch": 7.679218967921897, + "grad_norm": 3.682020664215088, + "learning_rate": 9.983489998349e-06, + "loss": 0.0851, + "step": 16518 + }, + { + "epoch": 7.679683867968387, + "grad_norm": 4.310915470123291, + "learning_rate": 9.9834889983489e-06, + "loss": 0.143, + "step": 16519 + }, + { + "epoch": 7.680148768014877, + "grad_norm": 5.539740085601807, + "learning_rate": 9.983487998348801e-06, + "loss": 0.1702, + "step": 16520 + }, + { + "epoch": 7.680613668061367, + "grad_norm": 5.387061595916748, + "learning_rate": 9.9834869983487e-06, + "loss": 0.1675, + "step": 16521 + }, + { + "epoch": 7.681078568107857, + "grad_norm": 4.676170349121094, + "learning_rate": 9.983485998348601e-06, + "loss": 0.0851, + "step": 16522 + }, + { + "epoch": 7.681543468154347, + "grad_norm": 3.404623508453369, + "learning_rate": 9.9834849983485e-06, + "loss": 0.0523, + "step": 16523 + }, + { + "epoch": 7.682008368200837, + "grad_norm": 5.737625598907471, + "learning_rate": 9.9834839983484e-06, + "loss": 0.1195, + "step": 16524 + }, + { + "epoch": 7.6824732682473265, + "grad_norm": 1.0271124839782715, + "learning_rate": 9.983482998348301e-06, + "loss": 0.011, + "step": 16525 + }, + { + "epoch": 7.682938168293817, + "grad_norm": 8.168030738830566, + "learning_rate": 9.983481998348202e-06, + "loss": 0.2281, + "step": 16526 + }, + { + "epoch": 7.683403068340307, + "grad_norm": 6.095114707946777, + "learning_rate": 9.983480998348101e-06, + "loss": 0.1598, + "step": 16527 + }, + { + "epoch": 7.683867968386797, + "grad_norm": 9.354116439819336, + "learning_rate": 9.983479998348e-06, + "loss": 0.2605, + "step": 16528 + }, + { + "epoch": 7.684332868433287, + "grad_norm": 5.999555587768555, + "learning_rate": 9.9834789983479e-06, + "loss": 0.1293, + "step": 16529 + }, + { + "epoch": 7.684797768479777, + "grad_norm": 5.325755596160889, + "learning_rate": 9.9834779983478e-06, + "loss": 0.1711, + "step": 16530 + }, + { + "epoch": 7.685262668526267, + "grad_norm": 5.686607360839844, + "learning_rate": 9.9834769983477e-06, + "loss": 0.096, + "step": 16531 + }, + { + "epoch": 7.685727568572757, + "grad_norm": 6.587772369384766, + "learning_rate": 9.983475998347601e-06, + "loss": 0.2178, + "step": 16532 + }, + { + "epoch": 7.686192468619247, + "grad_norm": 4.461658954620361, + "learning_rate": 9.983474998347502e-06, + "loss": 0.0677, + "step": 16533 + }, + { + "epoch": 7.6866573686657365, + "grad_norm": 2.2447009086608887, + "learning_rate": 9.9834739983474e-06, + "loss": 0.0339, + "step": 16534 + }, + { + "epoch": 7.687122268712227, + "grad_norm": 8.505370140075684, + "learning_rate": 9.983472998347301e-06, + "loss": 0.2129, + "step": 16535 + }, + { + "epoch": 7.687587168758717, + "grad_norm": 1.7905100584030151, + "learning_rate": 9.9834719983472e-06, + "loss": 0.0237, + "step": 16536 + }, + { + "epoch": 7.688052068805207, + "grad_norm": 3.8661370277404785, + "learning_rate": 9.9834709983471e-06, + "loss": 0.0593, + "step": 16537 + }, + { + "epoch": 7.6885169688516966, + "grad_norm": 5.071096420288086, + "learning_rate": 9.983469998347001e-06, + "loss": 0.0966, + "step": 16538 + }, + { + "epoch": 7.688981868898187, + "grad_norm": 10.19166374206543, + "learning_rate": 9.9834689983469e-06, + "loss": 0.1441, + "step": 16539 + }, + { + "epoch": 7.689446768944677, + "grad_norm": 4.940647125244141, + "learning_rate": 9.983467998346801e-06, + "loss": 0.0681, + "step": 16540 + }, + { + "epoch": 7.689911668991167, + "grad_norm": 8.00688648223877, + "learning_rate": 9.9834669983467e-06, + "loss": 0.1046, + "step": 16541 + }, + { + "epoch": 7.690376569037657, + "grad_norm": 7.295387268066406, + "learning_rate": 9.9834659983466e-06, + "loss": 0.2885, + "step": 16542 + }, + { + "epoch": 7.6908414690841465, + "grad_norm": 3.0421979427337646, + "learning_rate": 9.9834649983465e-06, + "loss": 0.0314, + "step": 16543 + }, + { + "epoch": 7.691306369130637, + "grad_norm": 4.393687725067139, + "learning_rate": 9.9834639983464e-06, + "loss": 0.0874, + "step": 16544 + }, + { + "epoch": 7.691771269177127, + "grad_norm": 4.43922758102417, + "learning_rate": 9.983462998346301e-06, + "loss": 0.0937, + "step": 16545 + }, + { + "epoch": 7.692236169223617, + "grad_norm": 6.669992446899414, + "learning_rate": 9.983461998346202e-06, + "loss": 0.2065, + "step": 16546 + }, + { + "epoch": 7.6927010692701066, + "grad_norm": 6.166790008544922, + "learning_rate": 9.9834609983461e-06, + "loss": 0.1764, + "step": 16547 + }, + { + "epoch": 7.693165969316597, + "grad_norm": 9.952125549316406, + "learning_rate": 9.983459998346e-06, + "loss": 0.3213, + "step": 16548 + }, + { + "epoch": 7.693630869363087, + "grad_norm": 4.808615207672119, + "learning_rate": 9.9834589983459e-06, + "loss": 0.1306, + "step": 16549 + }, + { + "epoch": 7.694095769409577, + "grad_norm": 3.9469261169433594, + "learning_rate": 9.983457998345801e-06, + "loss": 0.0775, + "step": 16550 + }, + { + "epoch": 7.694560669456067, + "grad_norm": 7.1896514892578125, + "learning_rate": 9.983456998345702e-06, + "loss": 0.231, + "step": 16551 + }, + { + "epoch": 7.695025569502557, + "grad_norm": 5.676793575286865, + "learning_rate": 9.9834559983456e-06, + "loss": 0.1181, + "step": 16552 + }, + { + "epoch": 7.695490469549047, + "grad_norm": 4.509378433227539, + "learning_rate": 9.983454998345501e-06, + "loss": 0.1236, + "step": 16553 + }, + { + "epoch": 7.695955369595537, + "grad_norm": 7.8958821296691895, + "learning_rate": 9.9834539983454e-06, + "loss": 0.1679, + "step": 16554 + }, + { + "epoch": 7.696420269642027, + "grad_norm": 5.301951885223389, + "learning_rate": 9.9834529983453e-06, + "loss": 0.1618, + "step": 16555 + }, + { + "epoch": 7.696885169688517, + "grad_norm": 8.405692100524902, + "learning_rate": 9.9834519983452e-06, + "loss": 0.1945, + "step": 16556 + }, + { + "epoch": 7.697350069735007, + "grad_norm": 7.031067371368408, + "learning_rate": 9.9834509983451e-06, + "loss": 0.1996, + "step": 16557 + }, + { + "epoch": 7.697814969781497, + "grad_norm": 4.553946495056152, + "learning_rate": 9.983449998345001e-06, + "loss": 0.1089, + "step": 16558 + }, + { + "epoch": 7.698279869827987, + "grad_norm": 4.614922523498535, + "learning_rate": 9.9834489983449e-06, + "loss": 0.0871, + "step": 16559 + }, + { + "epoch": 7.698744769874477, + "grad_norm": 5.592398643493652, + "learning_rate": 9.9834479983448e-06, + "loss": 0.116, + "step": 16560 + }, + { + "epoch": 7.699209669920967, + "grad_norm": 4.614721775054932, + "learning_rate": 9.9834469983447e-06, + "loss": 0.1562, + "step": 16561 + }, + { + "epoch": 7.699674569967457, + "grad_norm": 6.996487617492676, + "learning_rate": 9.9834459983446e-06, + "loss": 0.1716, + "step": 16562 + }, + { + "epoch": 7.700139470013947, + "grad_norm": 2.6562914848327637, + "learning_rate": 9.983444998344501e-06, + "loss": 0.0281, + "step": 16563 + }, + { + "epoch": 7.700604370060437, + "grad_norm": 5.091876029968262, + "learning_rate": 9.983443998344402e-06, + "loss": 0.0954, + "step": 16564 + }, + { + "epoch": 7.701069270106927, + "grad_norm": 5.362982273101807, + "learning_rate": 9.9834429983443e-06, + "loss": 0.1839, + "step": 16565 + }, + { + "epoch": 7.701534170153417, + "grad_norm": 3.2958076000213623, + "learning_rate": 9.983441998344201e-06, + "loss": 0.0628, + "step": 16566 + }, + { + "epoch": 7.701999070199907, + "grad_norm": 5.475606918334961, + "learning_rate": 9.9834409983441e-06, + "loss": 0.2451, + "step": 16567 + }, + { + "epoch": 7.702463970246397, + "grad_norm": 6.236153602600098, + "learning_rate": 9.983439998344001e-06, + "loss": 0.2229, + "step": 16568 + }, + { + "epoch": 7.702928870292887, + "grad_norm": 4.791405200958252, + "learning_rate": 9.9834389983439e-06, + "loss": 0.1071, + "step": 16569 + }, + { + "epoch": 7.703393770339377, + "grad_norm": 8.314714431762695, + "learning_rate": 9.9834379983438e-06, + "loss": 0.1488, + "step": 16570 + }, + { + "epoch": 7.703858670385867, + "grad_norm": 4.505756855010986, + "learning_rate": 9.983436998343701e-06, + "loss": 0.1254, + "step": 16571 + }, + { + "epoch": 7.704323570432357, + "grad_norm": 4.016557693481445, + "learning_rate": 9.9834359983436e-06, + "loss": 0.0909, + "step": 16572 + }, + { + "epoch": 7.704788470478847, + "grad_norm": 10.892852783203125, + "learning_rate": 9.9834349983435e-06, + "loss": 0.1478, + "step": 16573 + }, + { + "epoch": 7.705253370525337, + "grad_norm": 6.575251579284668, + "learning_rate": 9.9834339983434e-06, + "loss": 0.2167, + "step": 16574 + }, + { + "epoch": 7.705718270571827, + "grad_norm": 8.646410942077637, + "learning_rate": 9.9834329983433e-06, + "loss": 0.128, + "step": 16575 + }, + { + "epoch": 7.706183170618317, + "grad_norm": 9.528339385986328, + "learning_rate": 9.983431998343201e-06, + "loss": 0.2569, + "step": 16576 + }, + { + "epoch": 7.706648070664807, + "grad_norm": 3.2088358402252197, + "learning_rate": 9.983430998343102e-06, + "loss": 0.0832, + "step": 16577 + }, + { + "epoch": 7.707112970711297, + "grad_norm": 5.466684341430664, + "learning_rate": 9.983429998343e-06, + "loss": 0.1286, + "step": 16578 + }, + { + "epoch": 7.707577870757787, + "grad_norm": 4.743404388427734, + "learning_rate": 9.983428998342901e-06, + "loss": 0.201, + "step": 16579 + }, + { + "epoch": 7.708042770804277, + "grad_norm": 6.510566234588623, + "learning_rate": 9.9834279983428e-06, + "loss": 0.2578, + "step": 16580 + }, + { + "epoch": 7.708507670850767, + "grad_norm": 6.241790771484375, + "learning_rate": 9.983426998342701e-06, + "loss": 0.2119, + "step": 16581 + }, + { + "epoch": 7.708972570897257, + "grad_norm": 3.7133967876434326, + "learning_rate": 9.9834259983426e-06, + "loss": 0.0468, + "step": 16582 + }, + { + "epoch": 7.709437470943747, + "grad_norm": 5.0219407081604, + "learning_rate": 9.9834249983425e-06, + "loss": 0.1612, + "step": 16583 + }, + { + "epoch": 7.709902370990237, + "grad_norm": 9.127747535705566, + "learning_rate": 9.983423998342401e-06, + "loss": 0.2567, + "step": 16584 + }, + { + "epoch": 7.710367271036727, + "grad_norm": 6.176699638366699, + "learning_rate": 9.9834229983423e-06, + "loss": 0.0341, + "step": 16585 + }, + { + "epoch": 7.710832171083217, + "grad_norm": 7.514473915100098, + "learning_rate": 9.9834219983422e-06, + "loss": 0.1781, + "step": 16586 + }, + { + "epoch": 7.711297071129707, + "grad_norm": 6.5823445320129395, + "learning_rate": 9.9834209983421e-06, + "loss": 0.1714, + "step": 16587 + }, + { + "epoch": 7.711761971176197, + "grad_norm": 7.315545558929443, + "learning_rate": 9.983419998342e-06, + "loss": 0.3401, + "step": 16588 + }, + { + "epoch": 7.712226871222687, + "grad_norm": 9.302821159362793, + "learning_rate": 9.983418998341901e-06, + "loss": 0.2625, + "step": 16589 + }, + { + "epoch": 7.712691771269177, + "grad_norm": 5.587922096252441, + "learning_rate": 9.983417998341802e-06, + "loss": 0.1677, + "step": 16590 + }, + { + "epoch": 7.713156671315668, + "grad_norm": 6.0341033935546875, + "learning_rate": 9.9834169983417e-06, + "loss": 0.1824, + "step": 16591 + }, + { + "epoch": 7.713621571362157, + "grad_norm": 6.733758926391602, + "learning_rate": 9.9834159983416e-06, + "loss": 0.1895, + "step": 16592 + }, + { + "epoch": 7.714086471408647, + "grad_norm": 4.323891639709473, + "learning_rate": 9.9834149983415e-06, + "loss": 0.1361, + "step": 16593 + }, + { + "epoch": 7.714551371455137, + "grad_norm": 5.049529075622559, + "learning_rate": 9.983413998341401e-06, + "loss": 0.1664, + "step": 16594 + }, + { + "epoch": 7.715016271501627, + "grad_norm": 1.422396183013916, + "learning_rate": 9.983412998341302e-06, + "loss": 0.0136, + "step": 16595 + }, + { + "epoch": 7.7154811715481175, + "grad_norm": 3.882473945617676, + "learning_rate": 9.9834119983412e-06, + "loss": 0.1051, + "step": 16596 + }, + { + "epoch": 7.715946071594607, + "grad_norm": 4.632081031799316, + "learning_rate": 9.983410998341101e-06, + "loss": 0.146, + "step": 16597 + }, + { + "epoch": 7.716410971641097, + "grad_norm": 8.567970275878906, + "learning_rate": 9.983409998341e-06, + "loss": 0.2859, + "step": 16598 + }, + { + "epoch": 7.716875871687587, + "grad_norm": 3.878368616104126, + "learning_rate": 9.983408998340901e-06, + "loss": 0.0792, + "step": 16599 + }, + { + "epoch": 7.717340771734078, + "grad_norm": 6.762191295623779, + "learning_rate": 9.9834079983408e-06, + "loss": 0.1303, + "step": 16600 + }, + { + "epoch": 7.717805671780567, + "grad_norm": 4.705954074859619, + "learning_rate": 9.9834069983407e-06, + "loss": 0.0972, + "step": 16601 + }, + { + "epoch": 7.718270571827057, + "grad_norm": 6.394546031951904, + "learning_rate": 9.983405998340601e-06, + "loss": 0.1169, + "step": 16602 + }, + { + "epoch": 7.718735471873547, + "grad_norm": 6.6194634437561035, + "learning_rate": 9.9834049983405e-06, + "loss": 0.2446, + "step": 16603 + }, + { + "epoch": 7.719200371920037, + "grad_norm": 7.171077251434326, + "learning_rate": 9.9834039983404e-06, + "loss": 0.0685, + "step": 16604 + }, + { + "epoch": 7.7196652719665275, + "grad_norm": 4.800981521606445, + "learning_rate": 9.9834029983403e-06, + "loss": 0.1313, + "step": 16605 + }, + { + "epoch": 7.720130172013017, + "grad_norm": 3.829420804977417, + "learning_rate": 9.9834019983402e-06, + "loss": 0.0855, + "step": 16606 + }, + { + "epoch": 7.720595072059507, + "grad_norm": 6.8621931076049805, + "learning_rate": 9.983400998340101e-06, + "loss": 0.2568, + "step": 16607 + }, + { + "epoch": 7.721059972105997, + "grad_norm": 8.068737030029297, + "learning_rate": 9.983399998340002e-06, + "loss": 0.1744, + "step": 16608 + }, + { + "epoch": 7.721524872152488, + "grad_norm": 3.2550389766693115, + "learning_rate": 9.9833989983399e-06, + "loss": 0.0647, + "step": 16609 + }, + { + "epoch": 7.721989772198977, + "grad_norm": 7.8975749015808105, + "learning_rate": 9.983397998339801e-06, + "loss": 0.2541, + "step": 16610 + }, + { + "epoch": 7.722454672245467, + "grad_norm": 6.178246974945068, + "learning_rate": 9.9833969983397e-06, + "loss": 0.2044, + "step": 16611 + }, + { + "epoch": 7.722919572291957, + "grad_norm": 4.758695602416992, + "learning_rate": 9.983395998339601e-06, + "loss": 0.157, + "step": 16612 + }, + { + "epoch": 7.723384472338447, + "grad_norm": 6.292357921600342, + "learning_rate": 9.9833949983395e-06, + "loss": 0.1749, + "step": 16613 + }, + { + "epoch": 7.7238493723849375, + "grad_norm": 10.930806159973145, + "learning_rate": 9.9833939983394e-06, + "loss": 0.2225, + "step": 16614 + }, + { + "epoch": 7.724314272431427, + "grad_norm": 3.494532823562622, + "learning_rate": 9.983392998339301e-06, + "loss": 0.0862, + "step": 16615 + }, + { + "epoch": 7.724779172477917, + "grad_norm": 6.49605131149292, + "learning_rate": 9.9833919983392e-06, + "loss": 0.0877, + "step": 16616 + }, + { + "epoch": 7.725244072524407, + "grad_norm": 6.979753017425537, + "learning_rate": 9.983390998339101e-06, + "loss": 0.4762, + "step": 16617 + }, + { + "epoch": 7.725708972570898, + "grad_norm": 8.562959671020508, + "learning_rate": 9.983389998339e-06, + "loss": 0.1734, + "step": 16618 + }, + { + "epoch": 7.726173872617387, + "grad_norm": 4.390889644622803, + "learning_rate": 9.9833889983389e-06, + "loss": 0.1111, + "step": 16619 + }, + { + "epoch": 7.726638772663877, + "grad_norm": 4.019209861755371, + "learning_rate": 9.983387998338801e-06, + "loss": 0.1011, + "step": 16620 + }, + { + "epoch": 7.727103672710367, + "grad_norm": 5.309836387634277, + "learning_rate": 9.983386998338702e-06, + "loss": 0.1171, + "step": 16621 + }, + { + "epoch": 7.727568572756857, + "grad_norm": 5.896092891693115, + "learning_rate": 9.9833859983386e-06, + "loss": 0.1748, + "step": 16622 + }, + { + "epoch": 7.7280334728033475, + "grad_norm": 4.122189044952393, + "learning_rate": 9.983384998338501e-06, + "loss": 0.0671, + "step": 16623 + }, + { + "epoch": 7.728498372849837, + "grad_norm": 5.490752696990967, + "learning_rate": 9.9833839983384e-06, + "loss": 0.0816, + "step": 16624 + }, + { + "epoch": 7.728963272896327, + "grad_norm": 6.496742248535156, + "learning_rate": 9.9833829983383e-06, + "loss": 0.1001, + "step": 16625 + }, + { + "epoch": 7.729428172942817, + "grad_norm": 3.27844500541687, + "learning_rate": 9.9833819983382e-06, + "loss": 0.0486, + "step": 16626 + }, + { + "epoch": 7.729893072989308, + "grad_norm": 2.20426082611084, + "learning_rate": 9.9833809983381e-06, + "loss": 0.0419, + "step": 16627 + }, + { + "epoch": 7.730357973035797, + "grad_norm": 5.999841213226318, + "learning_rate": 9.983379998338001e-06, + "loss": 0.2828, + "step": 16628 + }, + { + "epoch": 7.730822873082287, + "grad_norm": 2.8933119773864746, + "learning_rate": 9.9833789983379e-06, + "loss": 0.0451, + "step": 16629 + }, + { + "epoch": 7.731287773128777, + "grad_norm": 5.173616886138916, + "learning_rate": 9.983377998337801e-06, + "loss": 0.1495, + "step": 16630 + }, + { + "epoch": 7.731752673175267, + "grad_norm": 10.712733268737793, + "learning_rate": 9.9833769983377e-06, + "loss": 0.1397, + "step": 16631 + }, + { + "epoch": 7.7322175732217575, + "grad_norm": 7.26064395904541, + "learning_rate": 9.9833759983376e-06, + "loss": 0.2322, + "step": 16632 + }, + { + "epoch": 7.732682473268247, + "grad_norm": 8.971123695373535, + "learning_rate": 9.983374998337501e-06, + "loss": 0.3362, + "step": 16633 + }, + { + "epoch": 7.733147373314737, + "grad_norm": 5.3510894775390625, + "learning_rate": 9.983373998337402e-06, + "loss": 0.1018, + "step": 16634 + }, + { + "epoch": 7.733612273361228, + "grad_norm": 1.841725468635559, + "learning_rate": 9.983372998337301e-06, + "loss": 0.0261, + "step": 16635 + }, + { + "epoch": 7.734077173407718, + "grad_norm": 4.663609504699707, + "learning_rate": 9.9833719983372e-06, + "loss": 0.1054, + "step": 16636 + }, + { + "epoch": 7.734542073454207, + "grad_norm": 5.561977863311768, + "learning_rate": 9.9833709983371e-06, + "loss": 0.1652, + "step": 16637 + }, + { + "epoch": 7.735006973500697, + "grad_norm": 4.63226842880249, + "learning_rate": 9.983369998337e-06, + "loss": 0.0884, + "step": 16638 + }, + { + "epoch": 7.735471873547187, + "grad_norm": 5.377200126647949, + "learning_rate": 9.9833689983369e-06, + "loss": 0.1447, + "step": 16639 + }, + { + "epoch": 7.735936773593678, + "grad_norm": 7.461029529571533, + "learning_rate": 9.9833679983368e-06, + "loss": 0.1521, + "step": 16640 + }, + { + "epoch": 7.7364016736401675, + "grad_norm": 5.093911170959473, + "learning_rate": 9.983366998336701e-06, + "loss": 0.1726, + "step": 16641 + }, + { + "epoch": 7.736866573686657, + "grad_norm": 2.7166213989257812, + "learning_rate": 9.9833659983366e-06, + "loss": 0.0239, + "step": 16642 + }, + { + "epoch": 7.737331473733147, + "grad_norm": 5.247936248779297, + "learning_rate": 9.983364998336501e-06, + "loss": 0.1625, + "step": 16643 + }, + { + "epoch": 7.737796373779638, + "grad_norm": 6.235334396362305, + "learning_rate": 9.9833639983364e-06, + "loss": 0.2344, + "step": 16644 + }, + { + "epoch": 7.738261273826128, + "grad_norm": 10.909557342529297, + "learning_rate": 9.9833629983363e-06, + "loss": 0.4444, + "step": 16645 + }, + { + "epoch": 7.738726173872617, + "grad_norm": 3.065917730331421, + "learning_rate": 9.983361998336201e-06, + "loss": 0.0671, + "step": 16646 + }, + { + "epoch": 7.739191073919107, + "grad_norm": 6.405457496643066, + "learning_rate": 9.9833609983361e-06, + "loss": 0.326, + "step": 16647 + }, + { + "epoch": 7.739655973965597, + "grad_norm": 5.302619457244873, + "learning_rate": 9.983359998336001e-06, + "loss": 0.1487, + "step": 16648 + }, + { + "epoch": 7.740120874012088, + "grad_norm": 4.980654239654541, + "learning_rate": 9.9833589983359e-06, + "loss": 0.1028, + "step": 16649 + }, + { + "epoch": 7.7405857740585775, + "grad_norm": 4.819366931915283, + "learning_rate": 9.9833579983358e-06, + "loss": 0.1199, + "step": 16650 + }, + { + "epoch": 7.741050674105067, + "grad_norm": 4.492312908172607, + "learning_rate": 9.9833569983357e-06, + "loss": 0.1019, + "step": 16651 + }, + { + "epoch": 7.741515574151557, + "grad_norm": 6.896704196929932, + "learning_rate": 9.9833559983356e-06, + "loss": 0.0819, + "step": 16652 + }, + { + "epoch": 7.741980474198048, + "grad_norm": 4.241611957550049, + "learning_rate": 9.983354998335501e-06, + "loss": 0.0954, + "step": 16653 + }, + { + "epoch": 7.742445374244538, + "grad_norm": 4.7095627784729, + "learning_rate": 9.983353998335402e-06, + "loss": 0.0993, + "step": 16654 + }, + { + "epoch": 7.742910274291027, + "grad_norm": 5.867119789123535, + "learning_rate": 9.9833529983353e-06, + "loss": 0.2083, + "step": 16655 + }, + { + "epoch": 7.743375174337517, + "grad_norm": 19.595117568969727, + "learning_rate": 9.983351998335201e-06, + "loss": 0.2286, + "step": 16656 + }, + { + "epoch": 7.743840074384007, + "grad_norm": 5.321532249450684, + "learning_rate": 9.9833509983351e-06, + "loss": 0.1886, + "step": 16657 + }, + { + "epoch": 7.744304974430498, + "grad_norm": 1.6418207883834839, + "learning_rate": 9.983349998335e-06, + "loss": 0.0254, + "step": 16658 + }, + { + "epoch": 7.7447698744769875, + "grad_norm": 7.433474540710449, + "learning_rate": 9.983348998334901e-06, + "loss": 0.2181, + "step": 16659 + }, + { + "epoch": 7.745234774523477, + "grad_norm": 4.368890762329102, + "learning_rate": 9.9833479983348e-06, + "loss": 0.1424, + "step": 16660 + }, + { + "epoch": 7.745699674569967, + "grad_norm": 3.7451672554016113, + "learning_rate": 9.983346998334701e-06, + "loss": 0.078, + "step": 16661 + }, + { + "epoch": 7.746164574616458, + "grad_norm": 3.3492543697357178, + "learning_rate": 9.9833459983346e-06, + "loss": 0.0446, + "step": 16662 + }, + { + "epoch": 7.746629474662948, + "grad_norm": 5.30425500869751, + "learning_rate": 9.9833449983345e-06, + "loss": 0.1225, + "step": 16663 + }, + { + "epoch": 7.747094374709437, + "grad_norm": 6.430002212524414, + "learning_rate": 9.983343998334401e-06, + "loss": 0.2835, + "step": 16664 + }, + { + "epoch": 7.747559274755927, + "grad_norm": 2.592344045639038, + "learning_rate": 9.9833429983343e-06, + "loss": 0.0544, + "step": 16665 + }, + { + "epoch": 7.748024174802417, + "grad_norm": 4.4481048583984375, + "learning_rate": 9.983341998334201e-06, + "loss": 0.1101, + "step": 16666 + }, + { + "epoch": 7.748489074848908, + "grad_norm": 3.3864195346832275, + "learning_rate": 9.983340998334102e-06, + "loss": 0.0754, + "step": 16667 + }, + { + "epoch": 7.7489539748953975, + "grad_norm": 6.153584957122803, + "learning_rate": 9.983339998334e-06, + "loss": 0.1071, + "step": 16668 + }, + { + "epoch": 7.749418874941887, + "grad_norm": 6.918482780456543, + "learning_rate": 9.9833389983339e-06, + "loss": 0.0798, + "step": 16669 + }, + { + "epoch": 7.749883774988378, + "grad_norm": 3.635867118835449, + "learning_rate": 9.9833379983338e-06, + "loss": 0.0627, + "step": 16670 + }, + { + "epoch": 7.750348675034868, + "grad_norm": 5.276127338409424, + "learning_rate": 9.983336998333701e-06, + "loss": 0.1825, + "step": 16671 + }, + { + "epoch": 7.750813575081358, + "grad_norm": 3.2315878868103027, + "learning_rate": 9.983335998333602e-06, + "loss": 0.0709, + "step": 16672 + }, + { + "epoch": 7.751278475127847, + "grad_norm": 4.792520523071289, + "learning_rate": 9.9833349983335e-06, + "loss": 0.1105, + "step": 16673 + }, + { + "epoch": 7.751743375174337, + "grad_norm": 4.623516082763672, + "learning_rate": 9.983333998333401e-06, + "loss": 0.084, + "step": 16674 + }, + { + "epoch": 7.752208275220828, + "grad_norm": 14.814967155456543, + "learning_rate": 9.9833329983333e-06, + "loss": 0.2792, + "step": 16675 + }, + { + "epoch": 7.752673175267318, + "grad_norm": 8.471436500549316, + "learning_rate": 9.9833319983332e-06, + "loss": 0.2214, + "step": 16676 + }, + { + "epoch": 7.7531380753138075, + "grad_norm": 3.597403049468994, + "learning_rate": 9.983330998333101e-06, + "loss": 0.1001, + "step": 16677 + }, + { + "epoch": 7.753602975360297, + "grad_norm": 5.912128925323486, + "learning_rate": 9.983329998333e-06, + "loss": 0.1541, + "step": 16678 + }, + { + "epoch": 7.754067875406788, + "grad_norm": 3.3163537979125977, + "learning_rate": 9.983328998332901e-06, + "loss": 0.0382, + "step": 16679 + }, + { + "epoch": 7.754532775453278, + "grad_norm": 3.025494337081909, + "learning_rate": 9.9833279983328e-06, + "loss": 0.0473, + "step": 16680 + }, + { + "epoch": 7.754997675499768, + "grad_norm": 6.654849529266357, + "learning_rate": 9.9833269983327e-06, + "loss": 0.1216, + "step": 16681 + }, + { + "epoch": 7.755462575546257, + "grad_norm": 7.237128257751465, + "learning_rate": 9.9833259983326e-06, + "loss": 0.1419, + "step": 16682 + }, + { + "epoch": 7.755927475592747, + "grad_norm": 2.451910972595215, + "learning_rate": 9.9833249983325e-06, + "loss": 0.0228, + "step": 16683 + }, + { + "epoch": 7.756392375639238, + "grad_norm": 7.145055770874023, + "learning_rate": 9.983323998332401e-06, + "loss": 0.2983, + "step": 16684 + }, + { + "epoch": 7.756857275685728, + "grad_norm": 2.46240496635437, + "learning_rate": 9.983322998332302e-06, + "loss": 0.0393, + "step": 16685 + }, + { + "epoch": 7.7573221757322175, + "grad_norm": 3.5200037956237793, + "learning_rate": 9.9833219983322e-06, + "loss": 0.0553, + "step": 16686 + }, + { + "epoch": 7.757787075778707, + "grad_norm": 6.195530891418457, + "learning_rate": 9.983320998332101e-06, + "loss": 0.1429, + "step": 16687 + }, + { + "epoch": 7.758251975825198, + "grad_norm": 6.144162178039551, + "learning_rate": 9.983319998332e-06, + "loss": 0.1416, + "step": 16688 + }, + { + "epoch": 7.758716875871688, + "grad_norm": 6.207293510437012, + "learning_rate": 9.9833189983319e-06, + "loss": 0.093, + "step": 16689 + }, + { + "epoch": 7.759181775918178, + "grad_norm": 5.482795715332031, + "learning_rate": 9.983317998331801e-06, + "loss": 0.1357, + "step": 16690 + }, + { + "epoch": 7.759646675964667, + "grad_norm": 6.7945966720581055, + "learning_rate": 9.9833169983317e-06, + "loss": 0.2561, + "step": 16691 + }, + { + "epoch": 7.760111576011157, + "grad_norm": 4.765639781951904, + "learning_rate": 9.983315998331601e-06, + "loss": 0.0849, + "step": 16692 + }, + { + "epoch": 7.760576476057648, + "grad_norm": 8.303194046020508, + "learning_rate": 9.9833149983315e-06, + "loss": 0.2314, + "step": 16693 + }, + { + "epoch": 7.761041376104138, + "grad_norm": 5.158012390136719, + "learning_rate": 9.9833139983314e-06, + "loss": 0.1206, + "step": 16694 + }, + { + "epoch": 7.7615062761506275, + "grad_norm": 4.898734092712402, + "learning_rate": 9.9833129983313e-06, + "loss": 0.1312, + "step": 16695 + }, + { + "epoch": 7.761971176197117, + "grad_norm": 7.215723037719727, + "learning_rate": 9.9833119983312e-06, + "loss": 0.2596, + "step": 16696 + }, + { + "epoch": 7.762436076243608, + "grad_norm": 2.9927120208740234, + "learning_rate": 9.983310998331101e-06, + "loss": 0.0489, + "step": 16697 + }, + { + "epoch": 7.762900976290098, + "grad_norm": 3.9998648166656494, + "learning_rate": 9.983309998331002e-06, + "loss": 0.0263, + "step": 16698 + }, + { + "epoch": 7.763365876336588, + "grad_norm": 8.618398666381836, + "learning_rate": 9.9833089983309e-06, + "loss": 0.3954, + "step": 16699 + }, + { + "epoch": 7.763830776383077, + "grad_norm": 6.412743091583252, + "learning_rate": 9.9833079983308e-06, + "loss": 0.1298, + "step": 16700 + }, + { + "epoch": 7.764295676429567, + "grad_norm": 4.381031036376953, + "learning_rate": 9.9833069983307e-06, + "loss": 0.1072, + "step": 16701 + }, + { + "epoch": 7.764760576476058, + "grad_norm": 7.3436479568481445, + "learning_rate": 9.983305998330601e-06, + "loss": 0.2577, + "step": 16702 + }, + { + "epoch": 7.765225476522548, + "grad_norm": 4.568084239959717, + "learning_rate": 9.983304998330502e-06, + "loss": 0.0751, + "step": 16703 + }, + { + "epoch": 7.7656903765690375, + "grad_norm": 3.6302452087402344, + "learning_rate": 9.9833039983304e-06, + "loss": 0.1377, + "step": 16704 + }, + { + "epoch": 7.766155276615527, + "grad_norm": 5.371520042419434, + "learning_rate": 9.983302998330301e-06, + "loss": 0.2175, + "step": 16705 + }, + { + "epoch": 7.766620176662018, + "grad_norm": 6.418495178222656, + "learning_rate": 9.9833019983302e-06, + "loss": 0.1903, + "step": 16706 + }, + { + "epoch": 7.767085076708508, + "grad_norm": 9.904314041137695, + "learning_rate": 9.9833009983301e-06, + "loss": 0.0417, + "step": 16707 + }, + { + "epoch": 7.767549976754998, + "grad_norm": 5.182551860809326, + "learning_rate": 9.98329999833e-06, + "loss": 0.1245, + "step": 16708 + }, + { + "epoch": 7.768014876801487, + "grad_norm": 4.86821174621582, + "learning_rate": 9.9832989983299e-06, + "loss": 0.1171, + "step": 16709 + }, + { + "epoch": 7.768479776847977, + "grad_norm": 5.506555557250977, + "learning_rate": 9.983297998329801e-06, + "loss": 0.1187, + "step": 16710 + }, + { + "epoch": 7.768944676894468, + "grad_norm": 1.1566046476364136, + "learning_rate": 9.9832969983297e-06, + "loss": 0.0135, + "step": 16711 + }, + { + "epoch": 7.769409576940958, + "grad_norm": 2.8235950469970703, + "learning_rate": 9.9832959983296e-06, + "loss": 0.0359, + "step": 16712 + }, + { + "epoch": 7.7698744769874475, + "grad_norm": 4.287241458892822, + "learning_rate": 9.9832949983295e-06, + "loss": 0.0924, + "step": 16713 + }, + { + "epoch": 7.770339377033938, + "grad_norm": 2.256837844848633, + "learning_rate": 9.9832939983294e-06, + "loss": 0.0291, + "step": 16714 + }, + { + "epoch": 7.770804277080428, + "grad_norm": 6.49566650390625, + "learning_rate": 9.983292998329301e-06, + "loss": 0.1548, + "step": 16715 + }, + { + "epoch": 7.771269177126918, + "grad_norm": 5.620742321014404, + "learning_rate": 9.983291998329202e-06, + "loss": 0.107, + "step": 16716 + }, + { + "epoch": 7.771734077173408, + "grad_norm": 1.0752580165863037, + "learning_rate": 9.9832909983291e-06, + "loss": 0.0128, + "step": 16717 + }, + { + "epoch": 7.772198977219897, + "grad_norm": 4.278528690338135, + "learning_rate": 9.983289998329001e-06, + "loss": 0.0911, + "step": 16718 + }, + { + "epoch": 7.772663877266388, + "grad_norm": 7.9590630531311035, + "learning_rate": 9.9832889983289e-06, + "loss": 0.1511, + "step": 16719 + }, + { + "epoch": 7.773128777312878, + "grad_norm": 7.4571990966796875, + "learning_rate": 9.983287998328801e-06, + "loss": 0.1904, + "step": 16720 + }, + { + "epoch": 7.773593677359368, + "grad_norm": 5.093347072601318, + "learning_rate": 9.9832869983287e-06, + "loss": 0.2089, + "step": 16721 + }, + { + "epoch": 7.7740585774058575, + "grad_norm": 6.793184280395508, + "learning_rate": 9.9832859983286e-06, + "loss": 0.2289, + "step": 16722 + }, + { + "epoch": 7.774523477452348, + "grad_norm": 5.819273948669434, + "learning_rate": 9.983284998328501e-06, + "loss": 0.1802, + "step": 16723 + }, + { + "epoch": 7.774988377498838, + "grad_norm": 4.867458343505859, + "learning_rate": 9.9832839983284e-06, + "loss": 0.1675, + "step": 16724 + }, + { + "epoch": 7.775453277545328, + "grad_norm": 5.459412097930908, + "learning_rate": 9.9832829983283e-06, + "loss": 0.1157, + "step": 16725 + }, + { + "epoch": 7.775918177591818, + "grad_norm": 3.5589613914489746, + "learning_rate": 9.9832819983282e-06, + "loss": 0.1213, + "step": 16726 + }, + { + "epoch": 7.776383077638307, + "grad_norm": 0.7213277816772461, + "learning_rate": 9.9832809983281e-06, + "loss": 0.0067, + "step": 16727 + }, + { + "epoch": 7.776847977684798, + "grad_norm": 3.5598697662353516, + "learning_rate": 9.983279998328001e-06, + "loss": 0.0532, + "step": 16728 + }, + { + "epoch": 7.777312877731288, + "grad_norm": 3.9790613651275635, + "learning_rate": 9.983278998327902e-06, + "loss": 0.0603, + "step": 16729 + }, + { + "epoch": 7.777777777777778, + "grad_norm": 8.879632949829102, + "learning_rate": 9.9832779983278e-06, + "loss": 0.1735, + "step": 16730 + }, + { + "epoch": 7.7782426778242675, + "grad_norm": 1.2473515272140503, + "learning_rate": 9.983276998327701e-06, + "loss": 0.0156, + "step": 16731 + }, + { + "epoch": 7.778707577870758, + "grad_norm": 4.3523454666137695, + "learning_rate": 9.9832759983276e-06, + "loss": 0.1002, + "step": 16732 + }, + { + "epoch": 7.779172477917248, + "grad_norm": 6.434270858764648, + "learning_rate": 9.983274998327501e-06, + "loss": 0.2999, + "step": 16733 + }, + { + "epoch": 7.779637377963738, + "grad_norm": 6.318775653839111, + "learning_rate": 9.9832739983274e-06, + "loss": 0.1226, + "step": 16734 + }, + { + "epoch": 7.780102278010228, + "grad_norm": 6.603014945983887, + "learning_rate": 9.9832729983273e-06, + "loss": 0.1128, + "step": 16735 + }, + { + "epoch": 7.780567178056717, + "grad_norm": 6.794249057769775, + "learning_rate": 9.983271998327201e-06, + "loss": 0.0984, + "step": 16736 + }, + { + "epoch": 7.781032078103208, + "grad_norm": 5.106141567230225, + "learning_rate": 9.9832709983271e-06, + "loss": 0.116, + "step": 16737 + }, + { + "epoch": 7.781496978149698, + "grad_norm": 6.044084548950195, + "learning_rate": 9.983269998327001e-06, + "loss": 0.1874, + "step": 16738 + }, + { + "epoch": 7.781961878196188, + "grad_norm": 3.9204134941101074, + "learning_rate": 9.9832689983269e-06, + "loss": 0.0878, + "step": 16739 + }, + { + "epoch": 7.7824267782426775, + "grad_norm": 8.005260467529297, + "learning_rate": 9.9832679983268e-06, + "loss": 0.2549, + "step": 16740 + }, + { + "epoch": 7.782891678289168, + "grad_norm": 3.7232167720794678, + "learning_rate": 9.983266998326701e-06, + "loss": 0.0813, + "step": 16741 + }, + { + "epoch": 7.783356578335658, + "grad_norm": 4.78034782409668, + "learning_rate": 9.983265998326602e-06, + "loss": 0.1349, + "step": 16742 + }, + { + "epoch": 7.783821478382148, + "grad_norm": 5.213376998901367, + "learning_rate": 9.9832649983265e-06, + "loss": 0.1148, + "step": 16743 + }, + { + "epoch": 7.784286378428638, + "grad_norm": 5.308063507080078, + "learning_rate": 9.9832639983264e-06, + "loss": 0.0858, + "step": 16744 + }, + { + "epoch": 7.784751278475127, + "grad_norm": 6.138131141662598, + "learning_rate": 9.9832629983263e-06, + "loss": 0.1907, + "step": 16745 + }, + { + "epoch": 7.785216178521618, + "grad_norm": 8.585651397705078, + "learning_rate": 9.983261998326201e-06, + "loss": 0.3755, + "step": 16746 + }, + { + "epoch": 7.785681078568108, + "grad_norm": 5.009734630584717, + "learning_rate": 9.9832609983261e-06, + "loss": 0.1203, + "step": 16747 + }, + { + "epoch": 7.786145978614598, + "grad_norm": 4.224968433380127, + "learning_rate": 9.983259998326e-06, + "loss": 0.0476, + "step": 16748 + }, + { + "epoch": 7.786610878661088, + "grad_norm": 6.636036396026611, + "learning_rate": 9.983258998325901e-06, + "loss": 0.3181, + "step": 16749 + }, + { + "epoch": 7.787075778707578, + "grad_norm": 6.530735969543457, + "learning_rate": 9.9832579983258e-06, + "loss": 0.0975, + "step": 16750 + }, + { + "epoch": 7.787540678754068, + "grad_norm": 5.689839839935303, + "learning_rate": 9.983256998325701e-06, + "loss": 0.1897, + "step": 16751 + }, + { + "epoch": 7.788005578800558, + "grad_norm": 3.3385794162750244, + "learning_rate": 9.9832559983256e-06, + "loss": 0.0607, + "step": 16752 + }, + { + "epoch": 7.788470478847048, + "grad_norm": 3.0541322231292725, + "learning_rate": 9.9832549983255e-06, + "loss": 0.0547, + "step": 16753 + }, + { + "epoch": 7.788935378893538, + "grad_norm": 4.870643138885498, + "learning_rate": 9.983253998325401e-06, + "loss": 0.1186, + "step": 16754 + }, + { + "epoch": 7.789400278940028, + "grad_norm": 16.58753776550293, + "learning_rate": 9.9832529983253e-06, + "loss": 0.2235, + "step": 16755 + }, + { + "epoch": 7.789865178986518, + "grad_norm": 3.4222679138183594, + "learning_rate": 9.983251998325201e-06, + "loss": 0.0532, + "step": 16756 + }, + { + "epoch": 7.790330079033008, + "grad_norm": 4.592341899871826, + "learning_rate": 9.9832509983251e-06, + "loss": 0.1677, + "step": 16757 + }, + { + "epoch": 7.790794979079498, + "grad_norm": 6.695774555206299, + "learning_rate": 9.983249998325e-06, + "loss": 0.2558, + "step": 16758 + }, + { + "epoch": 7.791259879125988, + "grad_norm": 2.144155979156494, + "learning_rate": 9.983248998324901e-06, + "loss": 0.0361, + "step": 16759 + }, + { + "epoch": 7.791724779172478, + "grad_norm": 5.330380916595459, + "learning_rate": 9.983247998324802e-06, + "loss": 0.1438, + "step": 16760 + }, + { + "epoch": 7.792189679218968, + "grad_norm": 6.730685234069824, + "learning_rate": 9.9832469983247e-06, + "loss": 0.1562, + "step": 16761 + }, + { + "epoch": 7.792654579265458, + "grad_norm": 2.6357321739196777, + "learning_rate": 9.983245998324601e-06, + "loss": 0.043, + "step": 16762 + }, + { + "epoch": 7.793119479311948, + "grad_norm": 4.211103916168213, + "learning_rate": 9.9832449983245e-06, + "loss": 0.0556, + "step": 16763 + }, + { + "epoch": 7.793584379358438, + "grad_norm": 7.213900566101074, + "learning_rate": 9.983243998324401e-06, + "loss": 0.2766, + "step": 16764 + }, + { + "epoch": 7.794049279404928, + "grad_norm": 7.536220550537109, + "learning_rate": 9.9832429983243e-06, + "loss": 0.2243, + "step": 16765 + }, + { + "epoch": 7.794514179451418, + "grad_norm": 1.1791493892669678, + "learning_rate": 9.9832419983242e-06, + "loss": 0.0101, + "step": 16766 + }, + { + "epoch": 7.794979079497908, + "grad_norm": 6.001810550689697, + "learning_rate": 9.983240998324101e-06, + "loss": 0.1246, + "step": 16767 + }, + { + "epoch": 7.795443979544398, + "grad_norm": 3.4396185874938965, + "learning_rate": 9.983239998324e-06, + "loss": 0.0726, + "step": 16768 + }, + { + "epoch": 7.795908879590888, + "grad_norm": 7.007732391357422, + "learning_rate": 9.983238998323901e-06, + "loss": 0.2011, + "step": 16769 + }, + { + "epoch": 7.796373779637378, + "grad_norm": 4.824277877807617, + "learning_rate": 9.9832379983238e-06, + "loss": 0.0842, + "step": 16770 + }, + { + "epoch": 7.796838679683868, + "grad_norm": 6.1945672035217285, + "learning_rate": 9.9832369983237e-06, + "loss": 0.1749, + "step": 16771 + }, + { + "epoch": 7.797303579730358, + "grad_norm": 6.105825901031494, + "learning_rate": 9.983235998323601e-06, + "loss": 0.1859, + "step": 16772 + }, + { + "epoch": 7.797768479776848, + "grad_norm": 5.245434284210205, + "learning_rate": 9.983234998323502e-06, + "loss": 0.1163, + "step": 16773 + }, + { + "epoch": 7.798233379823338, + "grad_norm": 4.315037727355957, + "learning_rate": 9.9832339983234e-06, + "loss": 0.1524, + "step": 16774 + }, + { + "epoch": 7.798698279869828, + "grad_norm": 9.012761116027832, + "learning_rate": 9.983232998323302e-06, + "loss": 0.2287, + "step": 16775 + }, + { + "epoch": 7.799163179916318, + "grad_norm": 3.1884348392486572, + "learning_rate": 9.9832319983232e-06, + "loss": 0.0577, + "step": 16776 + }, + { + "epoch": 7.799628079962808, + "grad_norm": 3.385108709335327, + "learning_rate": 9.9832309983231e-06, + "loss": 0.0455, + "step": 16777 + }, + { + "epoch": 7.800092980009298, + "grad_norm": 7.121001243591309, + "learning_rate": 9.983229998323e-06, + "loss": 0.1926, + "step": 16778 + }, + { + "epoch": 7.800557880055788, + "grad_norm": 7.083696365356445, + "learning_rate": 9.9832289983229e-06, + "loss": 0.1367, + "step": 16779 + }, + { + "epoch": 7.801022780102278, + "grad_norm": 7.0843915939331055, + "learning_rate": 9.983227998322801e-06, + "loss": 0.183, + "step": 16780 + }, + { + "epoch": 7.801487680148768, + "grad_norm": 8.203320503234863, + "learning_rate": 9.9832269983227e-06, + "loss": 0.2407, + "step": 16781 + }, + { + "epoch": 7.801952580195258, + "grad_norm": 6.571549415588379, + "learning_rate": 9.983225998322601e-06, + "loss": 0.1471, + "step": 16782 + }, + { + "epoch": 7.802417480241748, + "grad_norm": 4.627574920654297, + "learning_rate": 9.9832249983225e-06, + "loss": 0.1631, + "step": 16783 + }, + { + "epoch": 7.802882380288238, + "grad_norm": 6.973902225494385, + "learning_rate": 9.9832239983224e-06, + "loss": 0.2419, + "step": 16784 + }, + { + "epoch": 7.803347280334728, + "grad_norm": 4.49553918838501, + "learning_rate": 9.983222998322301e-06, + "loss": 0.0774, + "step": 16785 + }, + { + "epoch": 7.803812180381218, + "grad_norm": 5.942849159240723, + "learning_rate": 9.983221998322202e-06, + "loss": 0.1697, + "step": 16786 + }, + { + "epoch": 7.804277080427708, + "grad_norm": 6.596025466918945, + "learning_rate": 9.983220998322101e-06, + "loss": 0.1499, + "step": 16787 + }, + { + "epoch": 7.804741980474198, + "grad_norm": 5.1338043212890625, + "learning_rate": 9.983219998322e-06, + "loss": 0.1387, + "step": 16788 + }, + { + "epoch": 7.805206880520688, + "grad_norm": 4.898691654205322, + "learning_rate": 9.9832189983219e-06, + "loss": 0.1153, + "step": 16789 + }, + { + "epoch": 7.805671780567178, + "grad_norm": 4.111758708953857, + "learning_rate": 9.9832179983218e-06, + "loss": 0.0835, + "step": 16790 + }, + { + "epoch": 7.806136680613668, + "grad_norm": 3.4951977729797363, + "learning_rate": 9.9832169983217e-06, + "loss": 0.0736, + "step": 16791 + }, + { + "epoch": 7.806601580660158, + "grad_norm": 4.678302764892578, + "learning_rate": 9.9832159983216e-06, + "loss": 0.1273, + "step": 16792 + }, + { + "epoch": 7.8070664807066485, + "grad_norm": 7.464382648468018, + "learning_rate": 9.983214998321501e-06, + "loss": 0.2183, + "step": 16793 + }, + { + "epoch": 7.807531380753138, + "grad_norm": 3.9251279830932617, + "learning_rate": 9.9832139983214e-06, + "loss": 0.083, + "step": 16794 + }, + { + "epoch": 7.807996280799628, + "grad_norm": 3.2044689655303955, + "learning_rate": 9.983212998321301e-06, + "loss": 0.0515, + "step": 16795 + }, + { + "epoch": 7.808461180846118, + "grad_norm": 5.383207321166992, + "learning_rate": 9.9832119983212e-06, + "loss": 0.1245, + "step": 16796 + }, + { + "epoch": 7.808926080892608, + "grad_norm": 6.00999116897583, + "learning_rate": 9.9832109983211e-06, + "loss": 0.2616, + "step": 16797 + }, + { + "epoch": 7.8093909809390984, + "grad_norm": 5.327507972717285, + "learning_rate": 9.983209998321001e-06, + "loss": 0.1191, + "step": 16798 + }, + { + "epoch": 7.809855880985588, + "grad_norm": 3.467571258544922, + "learning_rate": 9.9832089983209e-06, + "loss": 0.0388, + "step": 16799 + }, + { + "epoch": 7.810320781032078, + "grad_norm": 3.469841480255127, + "learning_rate": 9.983207998320801e-06, + "loss": 0.1095, + "step": 16800 + }, + { + "epoch": 7.810785681078568, + "grad_norm": 6.1583333015441895, + "learning_rate": 9.9832069983207e-06, + "loss": 0.1745, + "step": 16801 + }, + { + "epoch": 7.8112505811250585, + "grad_norm": 5.793198108673096, + "learning_rate": 9.9832059983206e-06, + "loss": 0.1472, + "step": 16802 + }, + { + "epoch": 7.811715481171548, + "grad_norm": 6.0775628089904785, + "learning_rate": 9.9832049983205e-06, + "loss": 0.1161, + "step": 16803 + }, + { + "epoch": 7.812180381218038, + "grad_norm": 3.8068275451660156, + "learning_rate": 9.9832039983204e-06, + "loss": 0.0891, + "step": 16804 + }, + { + "epoch": 7.812645281264528, + "grad_norm": 2.838752031326294, + "learning_rate": 9.983202998320301e-06, + "loss": 0.0785, + "step": 16805 + }, + { + "epoch": 7.813110181311018, + "grad_norm": 5.205294609069824, + "learning_rate": 9.983201998320202e-06, + "loss": 0.1297, + "step": 16806 + }, + { + "epoch": 7.8135750813575084, + "grad_norm": 7.595271110534668, + "learning_rate": 9.9832009983201e-06, + "loss": 0.2996, + "step": 16807 + }, + { + "epoch": 7.814039981403998, + "grad_norm": 2.7366461753845215, + "learning_rate": 9.983199998320001e-06, + "loss": 0.0483, + "step": 16808 + }, + { + "epoch": 7.814504881450488, + "grad_norm": 6.178081035614014, + "learning_rate": 9.9831989983199e-06, + "loss": 0.1944, + "step": 16809 + }, + { + "epoch": 7.814969781496978, + "grad_norm": 3.7785608768463135, + "learning_rate": 9.9831979983198e-06, + "loss": 0.062, + "step": 16810 + }, + { + "epoch": 7.8154346815434685, + "grad_norm": 4.844801425933838, + "learning_rate": 9.983196998319701e-06, + "loss": 0.1607, + "step": 16811 + }, + { + "epoch": 7.815899581589958, + "grad_norm": 6.3298749923706055, + "learning_rate": 9.9831959983196e-06, + "loss": 0.1385, + "step": 16812 + }, + { + "epoch": 7.816364481636448, + "grad_norm": 4.376326560974121, + "learning_rate": 9.983194998319501e-06, + "loss": 0.0912, + "step": 16813 + }, + { + "epoch": 7.816829381682938, + "grad_norm": 6.004374027252197, + "learning_rate": 9.9831939983194e-06, + "loss": 0.202, + "step": 16814 + }, + { + "epoch": 7.817294281729428, + "grad_norm": 8.754592895507812, + "learning_rate": 9.9831929983193e-06, + "loss": 0.1792, + "step": 16815 + }, + { + "epoch": 7.817759181775918, + "grad_norm": 2.8646888732910156, + "learning_rate": 9.9831919983192e-06, + "loss": 0.0586, + "step": 16816 + }, + { + "epoch": 7.818224081822408, + "grad_norm": 5.8368706703186035, + "learning_rate": 9.9831909983191e-06, + "loss": 0.1892, + "step": 16817 + }, + { + "epoch": 7.818688981868898, + "grad_norm": 6.241809368133545, + "learning_rate": 9.983189998319001e-06, + "loss": 0.247, + "step": 16818 + }, + { + "epoch": 7.819153881915388, + "grad_norm": 7.391965866088867, + "learning_rate": 9.983188998318902e-06, + "loss": 0.2047, + "step": 16819 + }, + { + "epoch": 7.8196187819618785, + "grad_norm": 3.314934253692627, + "learning_rate": 9.9831879983188e-06, + "loss": 0.0649, + "step": 16820 + }, + { + "epoch": 7.820083682008368, + "grad_norm": 5.548309803009033, + "learning_rate": 9.9831869983187e-06, + "loss": 0.2002, + "step": 16821 + }, + { + "epoch": 7.820548582054858, + "grad_norm": 2.7893974781036377, + "learning_rate": 9.9831859983186e-06, + "loss": 0.0383, + "step": 16822 + }, + { + "epoch": 7.821013482101348, + "grad_norm": 5.209325313568115, + "learning_rate": 9.983184998318501e-06, + "loss": 0.1188, + "step": 16823 + }, + { + "epoch": 7.821478382147838, + "grad_norm": 4.028379917144775, + "learning_rate": 9.983183998318402e-06, + "loss": 0.0722, + "step": 16824 + }, + { + "epoch": 7.821943282194328, + "grad_norm": 5.829727649688721, + "learning_rate": 9.9831829983183e-06, + "loss": 0.1487, + "step": 16825 + }, + { + "epoch": 7.822408182240818, + "grad_norm": 3.867302417755127, + "learning_rate": 9.983181998318201e-06, + "loss": 0.0784, + "step": 16826 + }, + { + "epoch": 7.822873082287308, + "grad_norm": 4.650292873382568, + "learning_rate": 9.9831809983181e-06, + "loss": 0.1368, + "step": 16827 + }, + { + "epoch": 7.823337982333798, + "grad_norm": 5.177801132202148, + "learning_rate": 9.983179998318e-06, + "loss": 0.0924, + "step": 16828 + }, + { + "epoch": 7.8238028823802885, + "grad_norm": 3.529116153717041, + "learning_rate": 9.983178998317901e-06, + "loss": 0.0809, + "step": 16829 + }, + { + "epoch": 7.824267782426778, + "grad_norm": 4.465368747711182, + "learning_rate": 9.9831779983178e-06, + "loss": 0.0985, + "step": 16830 + }, + { + "epoch": 7.824732682473268, + "grad_norm": 3.437483310699463, + "learning_rate": 9.983176998317701e-06, + "loss": 0.0556, + "step": 16831 + }, + { + "epoch": 7.825197582519758, + "grad_norm": 5.7991251945495605, + "learning_rate": 9.9831759983176e-06, + "loss": 0.0929, + "step": 16832 + }, + { + "epoch": 7.825662482566249, + "grad_norm": 6.350208759307861, + "learning_rate": 9.9831749983175e-06, + "loss": 0.0529, + "step": 16833 + }, + { + "epoch": 7.826127382612738, + "grad_norm": 5.5471577644348145, + "learning_rate": 9.9831739983174e-06, + "loss": 0.1178, + "step": 16834 + }, + { + "epoch": 7.826592282659228, + "grad_norm": 10.44972038269043, + "learning_rate": 9.9831729983173e-06, + "loss": 0.104, + "step": 16835 + }, + { + "epoch": 7.827057182705718, + "grad_norm": 2.0436737537384033, + "learning_rate": 9.983171998317201e-06, + "loss": 0.0266, + "step": 16836 + }, + { + "epoch": 7.827522082752209, + "grad_norm": 5.657005310058594, + "learning_rate": 9.983170998317102e-06, + "loss": 0.1915, + "step": 16837 + }, + { + "epoch": 7.8279869827986985, + "grad_norm": 5.472520351409912, + "learning_rate": 9.983169998317e-06, + "loss": 0.1868, + "step": 16838 + }, + { + "epoch": 7.828451882845188, + "grad_norm": 7.392200469970703, + "learning_rate": 9.983168998316901e-06, + "loss": 0.1298, + "step": 16839 + }, + { + "epoch": 7.828916782891678, + "grad_norm": 11.17013168334961, + "learning_rate": 9.9831679983168e-06, + "loss": 0.5772, + "step": 16840 + }, + { + "epoch": 7.829381682938168, + "grad_norm": 4.519496917724609, + "learning_rate": 9.983166998316701e-06, + "loss": 0.0833, + "step": 16841 + }, + { + "epoch": 7.829846582984659, + "grad_norm": 8.446268081665039, + "learning_rate": 9.983165998316602e-06, + "loss": 0.2479, + "step": 16842 + }, + { + "epoch": 7.830311483031148, + "grad_norm": 4.7723283767700195, + "learning_rate": 9.9831649983165e-06, + "loss": 0.0457, + "step": 16843 + }, + { + "epoch": 7.830776383077638, + "grad_norm": 4.482721328735352, + "learning_rate": 9.983163998316401e-06, + "loss": 0.0567, + "step": 16844 + }, + { + "epoch": 7.831241283124128, + "grad_norm": 4.036398887634277, + "learning_rate": 9.9831629983163e-06, + "loss": 0.0434, + "step": 16845 + }, + { + "epoch": 7.831706183170619, + "grad_norm": 4.9146318435668945, + "learning_rate": 9.9831619983162e-06, + "loss": 0.1301, + "step": 16846 + }, + { + "epoch": 7.8321710832171085, + "grad_norm": 5.96647834777832, + "learning_rate": 9.9831609983161e-06, + "loss": 0.2944, + "step": 16847 + }, + { + "epoch": 7.832635983263598, + "grad_norm": 4.436991214752197, + "learning_rate": 9.983159998316e-06, + "loss": 0.102, + "step": 16848 + }, + { + "epoch": 7.833100883310088, + "grad_norm": 5.356497287750244, + "learning_rate": 9.983158998315901e-06, + "loss": 0.1363, + "step": 16849 + }, + { + "epoch": 7.833565783356578, + "grad_norm": 5.967051982879639, + "learning_rate": 9.983157998315802e-06, + "loss": 0.1508, + "step": 16850 + }, + { + "epoch": 7.834030683403069, + "grad_norm": 4.9242682456970215, + "learning_rate": 9.9831569983157e-06, + "loss": 0.1281, + "step": 16851 + }, + { + "epoch": 7.834495583449558, + "grad_norm": 5.809433460235596, + "learning_rate": 9.9831559983156e-06, + "loss": 0.2291, + "step": 16852 + }, + { + "epoch": 7.834960483496048, + "grad_norm": 11.776203155517578, + "learning_rate": 9.9831549983155e-06, + "loss": 0.321, + "step": 16853 + }, + { + "epoch": 7.835425383542538, + "grad_norm": 5.540592670440674, + "learning_rate": 9.983153998315401e-06, + "loss": 0.1043, + "step": 16854 + }, + { + "epoch": 7.835890283589029, + "grad_norm": 5.469420433044434, + "learning_rate": 9.983152998315302e-06, + "loss": 0.15, + "step": 16855 + }, + { + "epoch": 7.8363551836355185, + "grad_norm": 8.056719779968262, + "learning_rate": 9.9831519983152e-06, + "loss": 0.1835, + "step": 16856 + }, + { + "epoch": 7.836820083682008, + "grad_norm": 2.302988052368164, + "learning_rate": 9.983150998315101e-06, + "loss": 0.032, + "step": 16857 + }, + { + "epoch": 7.837284983728498, + "grad_norm": 4.059962749481201, + "learning_rate": 9.983149998315e-06, + "loss": 0.0842, + "step": 16858 + }, + { + "epoch": 7.837749883774988, + "grad_norm": 2.2398648262023926, + "learning_rate": 9.983148998314901e-06, + "loss": 0.024, + "step": 16859 + }, + { + "epoch": 7.838214783821479, + "grad_norm": 3.841125011444092, + "learning_rate": 9.9831479983148e-06, + "loss": 0.0932, + "step": 16860 + }, + { + "epoch": 7.838679683867968, + "grad_norm": 8.610307693481445, + "learning_rate": 9.9831469983147e-06, + "loss": 0.1011, + "step": 16861 + }, + { + "epoch": 7.839144583914458, + "grad_norm": 7.500833034515381, + "learning_rate": 9.983145998314601e-06, + "loss": 0.2347, + "step": 16862 + }, + { + "epoch": 7.839609483960948, + "grad_norm": 3.62001371383667, + "learning_rate": 9.9831449983145e-06, + "loss": 0.0909, + "step": 16863 + }, + { + "epoch": 7.840074384007439, + "grad_norm": 4.592639446258545, + "learning_rate": 9.9831439983144e-06, + "loss": 0.1261, + "step": 16864 + }, + { + "epoch": 7.8405392840539285, + "grad_norm": 6.985036849975586, + "learning_rate": 9.9831429983143e-06, + "loss": 0.3449, + "step": 16865 + }, + { + "epoch": 7.841004184100418, + "grad_norm": 5.210597515106201, + "learning_rate": 9.9831419983142e-06, + "loss": 0.1531, + "step": 16866 + }, + { + "epoch": 7.841469084146908, + "grad_norm": 3.6005446910858154, + "learning_rate": 9.983140998314101e-06, + "loss": 0.0671, + "step": 16867 + }, + { + "epoch": 7.841933984193398, + "grad_norm": 4.208439826965332, + "learning_rate": 9.983139998314002e-06, + "loss": 0.0674, + "step": 16868 + }, + { + "epoch": 7.842398884239889, + "grad_norm": 7.991547584533691, + "learning_rate": 9.9831389983139e-06, + "loss": 0.1673, + "step": 16869 + }, + { + "epoch": 7.842863784286378, + "grad_norm": 4.67088508605957, + "learning_rate": 9.983137998313801e-06, + "loss": 0.1424, + "step": 16870 + }, + { + "epoch": 7.843328684332868, + "grad_norm": 3.085190534591675, + "learning_rate": 9.9831369983137e-06, + "loss": 0.0462, + "step": 16871 + }, + { + "epoch": 7.843793584379359, + "grad_norm": 9.742870330810547, + "learning_rate": 9.983135998313601e-06, + "loss": 0.3208, + "step": 16872 + }, + { + "epoch": 7.844258484425849, + "grad_norm": 6.3465046882629395, + "learning_rate": 9.9831349983135e-06, + "loss": 0.2563, + "step": 16873 + }, + { + "epoch": 7.8447233844723385, + "grad_norm": 4.114963531494141, + "learning_rate": 9.9831339983134e-06, + "loss": 0.0937, + "step": 16874 + }, + { + "epoch": 7.845188284518828, + "grad_norm": 4.974564075469971, + "learning_rate": 9.983132998313301e-06, + "loss": 0.1083, + "step": 16875 + }, + { + "epoch": 7.845653184565318, + "grad_norm": 6.564191818237305, + "learning_rate": 9.9831319983132e-06, + "loss": 0.1964, + "step": 16876 + }, + { + "epoch": 7.846118084611809, + "grad_norm": 6.013476848602295, + "learning_rate": 9.9831309983131e-06, + "loss": 0.1873, + "step": 16877 + }, + { + "epoch": 7.846582984658299, + "grad_norm": 2.6830575466156006, + "learning_rate": 9.983129998313e-06, + "loss": 0.0322, + "step": 16878 + }, + { + "epoch": 7.847047884704788, + "grad_norm": 8.2115478515625, + "learning_rate": 9.9831289983129e-06, + "loss": 0.1398, + "step": 16879 + }, + { + "epoch": 7.847512784751278, + "grad_norm": 3.0587077140808105, + "learning_rate": 9.983127998312801e-06, + "loss": 0.0604, + "step": 16880 + }, + { + "epoch": 7.847977684797769, + "grad_norm": 10.738572120666504, + "learning_rate": 9.983126998312702e-06, + "loss": 0.1847, + "step": 16881 + }, + { + "epoch": 7.848442584844259, + "grad_norm": 5.532493591308594, + "learning_rate": 9.9831259983126e-06, + "loss": 0.1338, + "step": 16882 + }, + { + "epoch": 7.8489074848907485, + "grad_norm": 6.081152439117432, + "learning_rate": 9.983124998312501e-06, + "loss": 0.2807, + "step": 16883 + }, + { + "epoch": 7.849372384937238, + "grad_norm": 11.01656436920166, + "learning_rate": 9.9831239983124e-06, + "loss": 0.2381, + "step": 16884 + }, + { + "epoch": 7.849837284983728, + "grad_norm": 2.5618085861206055, + "learning_rate": 9.9831229983123e-06, + "loss": 0.039, + "step": 16885 + }, + { + "epoch": 7.850302185030219, + "grad_norm": 7.754560470581055, + "learning_rate": 9.9831219983122e-06, + "loss": 0.2691, + "step": 16886 + }, + { + "epoch": 7.850767085076709, + "grad_norm": 5.6348700523376465, + "learning_rate": 9.9831209983121e-06, + "loss": 0.1364, + "step": 16887 + }, + { + "epoch": 7.851231985123198, + "grad_norm": 4.949256420135498, + "learning_rate": 9.983119998312001e-06, + "loss": 0.1582, + "step": 16888 + }, + { + "epoch": 7.851696885169688, + "grad_norm": 5.749995231628418, + "learning_rate": 9.9831189983119e-06, + "loss": 0.1494, + "step": 16889 + }, + { + "epoch": 7.852161785216179, + "grad_norm": 3.5202300548553467, + "learning_rate": 9.983117998311801e-06, + "loss": 0.0638, + "step": 16890 + }, + { + "epoch": 7.852626685262669, + "grad_norm": 4.58402681350708, + "learning_rate": 9.9831169983117e-06, + "loss": 0.0529, + "step": 16891 + }, + { + "epoch": 7.8530915853091585, + "grad_norm": 7.705761909484863, + "learning_rate": 9.9831159983116e-06, + "loss": 0.0979, + "step": 16892 + }, + { + "epoch": 7.853556485355648, + "grad_norm": 1.4029866456985474, + "learning_rate": 9.983114998311501e-06, + "loss": 0.018, + "step": 16893 + }, + { + "epoch": 7.854021385402138, + "grad_norm": 4.669768810272217, + "learning_rate": 9.983113998311402e-06, + "loss": 0.1097, + "step": 16894 + }, + { + "epoch": 7.854486285448629, + "grad_norm": 7.704535961151123, + "learning_rate": 9.9831129983113e-06, + "loss": 0.3276, + "step": 16895 + }, + { + "epoch": 7.854951185495119, + "grad_norm": 5.7241435050964355, + "learning_rate": 9.9831119983112e-06, + "loss": 0.2466, + "step": 16896 + }, + { + "epoch": 7.855416085541608, + "grad_norm": 5.933099746704102, + "learning_rate": 9.9831109983111e-06, + "loss": 0.1125, + "step": 16897 + }, + { + "epoch": 7.855880985588098, + "grad_norm": 6.555121421813965, + "learning_rate": 9.983109998311001e-06, + "loss": 0.2211, + "step": 16898 + }, + { + "epoch": 7.856345885634589, + "grad_norm": 8.230982780456543, + "learning_rate": 9.9831089983109e-06, + "loss": 0.3679, + "step": 16899 + }, + { + "epoch": 7.856810785681079, + "grad_norm": 5.963613986968994, + "learning_rate": 9.9831079983108e-06, + "loss": 0.1513, + "step": 16900 + }, + { + "epoch": 7.8572756857275685, + "grad_norm": 5.71273946762085, + "learning_rate": 9.983106998310701e-06, + "loss": 0.1712, + "step": 16901 + }, + { + "epoch": 7.857740585774058, + "grad_norm": 4.471088886260986, + "learning_rate": 9.9831059983106e-06, + "loss": 0.0814, + "step": 16902 + }, + { + "epoch": 7.858205485820548, + "grad_norm": 11.1463041305542, + "learning_rate": 9.983104998310501e-06, + "loss": 0.1645, + "step": 16903 + }, + { + "epoch": 7.858670385867039, + "grad_norm": 4.216073989868164, + "learning_rate": 9.9831039983104e-06, + "loss": 0.0894, + "step": 16904 + }, + { + "epoch": 7.859135285913529, + "grad_norm": 6.066062927246094, + "learning_rate": 9.9831029983103e-06, + "loss": 0.1652, + "step": 16905 + }, + { + "epoch": 7.859600185960018, + "grad_norm": 4.196122646331787, + "learning_rate": 9.983101998310201e-06, + "loss": 0.1144, + "step": 16906 + }, + { + "epoch": 7.860065086006508, + "grad_norm": 4.344267845153809, + "learning_rate": 9.9831009983101e-06, + "loss": 0.1038, + "step": 16907 + }, + { + "epoch": 7.860529986052999, + "grad_norm": 7.864559173583984, + "learning_rate": 9.983099998310001e-06, + "loss": 0.3197, + "step": 16908 + }, + { + "epoch": 7.860994886099489, + "grad_norm": 3.1473336219787598, + "learning_rate": 9.9830989983099e-06, + "loss": 0.0513, + "step": 16909 + }, + { + "epoch": 7.8614597861459785, + "grad_norm": 6.065067768096924, + "learning_rate": 9.9830979983098e-06, + "loss": 0.1342, + "step": 16910 + }, + { + "epoch": 7.861924686192468, + "grad_norm": 3.920687198638916, + "learning_rate": 9.983096998309701e-06, + "loss": 0.0948, + "step": 16911 + }, + { + "epoch": 7.862389586238959, + "grad_norm": 4.95185661315918, + "learning_rate": 9.9830959983096e-06, + "loss": 0.0672, + "step": 16912 + }, + { + "epoch": 7.862854486285449, + "grad_norm": 7.672780990600586, + "learning_rate": 9.9830949983095e-06, + "loss": 0.3197, + "step": 16913 + }, + { + "epoch": 7.863319386331939, + "grad_norm": 5.4136962890625, + "learning_rate": 9.983093998309401e-06, + "loss": 0.0826, + "step": 16914 + }, + { + "epoch": 7.863784286378428, + "grad_norm": 2.876195192337036, + "learning_rate": 9.9830929983093e-06, + "loss": 0.0559, + "step": 16915 + }, + { + "epoch": 7.864249186424919, + "grad_norm": 5.271787166595459, + "learning_rate": 9.983091998309201e-06, + "loss": 0.1492, + "step": 16916 + }, + { + "epoch": 7.864714086471409, + "grad_norm": 5.95695161819458, + "learning_rate": 9.9830909983091e-06, + "loss": 0.1536, + "step": 16917 + }, + { + "epoch": 7.865178986517899, + "grad_norm": 9.0623779296875, + "learning_rate": 9.983089998309e-06, + "loss": 0.2384, + "step": 16918 + }, + { + "epoch": 7.8656438865643885, + "grad_norm": 11.235021591186523, + "learning_rate": 9.983088998308901e-06, + "loss": 0.282, + "step": 16919 + }, + { + "epoch": 7.866108786610878, + "grad_norm": 4.583868026733398, + "learning_rate": 9.9830879983088e-06, + "loss": 0.0615, + "step": 16920 + }, + { + "epoch": 7.866573686657369, + "grad_norm": 7.043489456176758, + "learning_rate": 9.983086998308701e-06, + "loss": 0.1425, + "step": 16921 + }, + { + "epoch": 7.867038586703859, + "grad_norm": 2.8826663494110107, + "learning_rate": 9.9830859983086e-06, + "loss": 0.0485, + "step": 16922 + }, + { + "epoch": 7.867503486750349, + "grad_norm": 3.4988927841186523, + "learning_rate": 9.9830849983085e-06, + "loss": 0.0848, + "step": 16923 + }, + { + "epoch": 7.867968386796838, + "grad_norm": 1.7362279891967773, + "learning_rate": 9.983083998308401e-06, + "loss": 0.0146, + "step": 16924 + }, + { + "epoch": 7.868433286843329, + "grad_norm": 3.645909547805786, + "learning_rate": 9.983082998308302e-06, + "loss": 0.0661, + "step": 16925 + }, + { + "epoch": 7.868898186889819, + "grad_norm": 5.450438022613525, + "learning_rate": 9.983081998308201e-06, + "loss": 0.1165, + "step": 16926 + }, + { + "epoch": 7.869363086936309, + "grad_norm": 6.151578903198242, + "learning_rate": 9.983080998308102e-06, + "loss": 0.1044, + "step": 16927 + }, + { + "epoch": 7.8698279869827985, + "grad_norm": 6.455636501312256, + "learning_rate": 9.983079998308e-06, + "loss": 0.2098, + "step": 16928 + }, + { + "epoch": 7.870292887029288, + "grad_norm": 7.059702396392822, + "learning_rate": 9.9830789983079e-06, + "loss": 0.2425, + "step": 16929 + }, + { + "epoch": 7.870757787075779, + "grad_norm": 6.999938488006592, + "learning_rate": 9.9830779983078e-06, + "loss": 0.2501, + "step": 16930 + }, + { + "epoch": 7.871222687122269, + "grad_norm": 5.769277572631836, + "learning_rate": 9.9830769983077e-06, + "loss": 0.1495, + "step": 16931 + }, + { + "epoch": 7.871687587168759, + "grad_norm": 5.571402072906494, + "learning_rate": 9.983075998307601e-06, + "loss": 0.1471, + "step": 16932 + }, + { + "epoch": 7.872152487215248, + "grad_norm": 3.080353260040283, + "learning_rate": 9.9830749983075e-06, + "loss": 0.0517, + "step": 16933 + }, + { + "epoch": 7.872617387261739, + "grad_norm": 5.507781982421875, + "learning_rate": 9.983073998307401e-06, + "loss": 0.1413, + "step": 16934 + }, + { + "epoch": 7.873082287308229, + "grad_norm": 5.5049309730529785, + "learning_rate": 9.9830729983073e-06, + "loss": 0.1237, + "step": 16935 + }, + { + "epoch": 7.873547187354719, + "grad_norm": 4.065103054046631, + "learning_rate": 9.9830719983072e-06, + "loss": 0.0303, + "step": 16936 + }, + { + "epoch": 7.8740120874012085, + "grad_norm": 9.59154224395752, + "learning_rate": 9.983070998307101e-06, + "loss": 0.19, + "step": 16937 + }, + { + "epoch": 7.874476987447698, + "grad_norm": 7.064264297485352, + "learning_rate": 9.983069998307002e-06, + "loss": 0.2584, + "step": 16938 + }, + { + "epoch": 7.874941887494189, + "grad_norm": 5.233162879943848, + "learning_rate": 9.983068998306901e-06, + "loss": 0.1211, + "step": 16939 + }, + { + "epoch": 7.875406787540679, + "grad_norm": 6.177163600921631, + "learning_rate": 9.9830679983068e-06, + "loss": 0.1147, + "step": 16940 + }, + { + "epoch": 7.875871687587169, + "grad_norm": 7.712098121643066, + "learning_rate": 9.9830669983067e-06, + "loss": 0.2745, + "step": 16941 + }, + { + "epoch": 7.876336587633658, + "grad_norm": 4.965764999389648, + "learning_rate": 9.9830659983066e-06, + "loss": 0.097, + "step": 16942 + }, + { + "epoch": 7.876801487680149, + "grad_norm": 7.098116874694824, + "learning_rate": 9.9830649983065e-06, + "loss": 0.1761, + "step": 16943 + }, + { + "epoch": 7.877266387726639, + "grad_norm": 5.835248947143555, + "learning_rate": 9.983063998306401e-06, + "loss": 0.1938, + "step": 16944 + }, + { + "epoch": 7.877731287773129, + "grad_norm": 4.302311897277832, + "learning_rate": 9.983062998306302e-06, + "loss": 0.1611, + "step": 16945 + }, + { + "epoch": 7.8781961878196185, + "grad_norm": 2.534224510192871, + "learning_rate": 9.9830619983062e-06, + "loss": 0.0414, + "step": 16946 + }, + { + "epoch": 7.878661087866108, + "grad_norm": 2.406312942504883, + "learning_rate": 9.983060998306101e-06, + "loss": 0.042, + "step": 16947 + }, + { + "epoch": 7.879125987912599, + "grad_norm": 4.725657939910889, + "learning_rate": 9.983059998306e-06, + "loss": 0.1058, + "step": 16948 + }, + { + "epoch": 7.879590887959089, + "grad_norm": 7.910056114196777, + "learning_rate": 9.9830589983059e-06, + "loss": 0.2289, + "step": 16949 + }, + { + "epoch": 7.880055788005579, + "grad_norm": 4.624119758605957, + "learning_rate": 9.983057998305801e-06, + "loss": 0.1077, + "step": 16950 + }, + { + "epoch": 7.880520688052069, + "grad_norm": 6.283924102783203, + "learning_rate": 9.9830569983057e-06, + "loss": 0.1052, + "step": 16951 + }, + { + "epoch": 7.880985588098559, + "grad_norm": 4.673515319824219, + "learning_rate": 9.983055998305601e-06, + "loss": 0.0763, + "step": 16952 + }, + { + "epoch": 7.881450488145049, + "grad_norm": 6.14653205871582, + "learning_rate": 9.9830549983055e-06, + "loss": 0.0738, + "step": 16953 + }, + { + "epoch": 7.881915388191539, + "grad_norm": 5.8623151779174805, + "learning_rate": 9.9830539983054e-06, + "loss": 0.136, + "step": 16954 + }, + { + "epoch": 7.8823802882380285, + "grad_norm": 5.644513130187988, + "learning_rate": 9.9830529983053e-06, + "loss": 0.1218, + "step": 16955 + }, + { + "epoch": 7.882845188284519, + "grad_norm": 12.385984420776367, + "learning_rate": 9.9830519983052e-06, + "loss": 0.2361, + "step": 16956 + }, + { + "epoch": 7.883310088331009, + "grad_norm": 5.298112869262695, + "learning_rate": 9.983050998305101e-06, + "loss": 0.1022, + "step": 16957 + }, + { + "epoch": 7.883774988377499, + "grad_norm": 2.7048885822296143, + "learning_rate": 9.983049998305002e-06, + "loss": 0.0306, + "step": 16958 + }, + { + "epoch": 7.884239888423989, + "grad_norm": 6.4124755859375, + "learning_rate": 9.9830489983049e-06, + "loss": 0.2427, + "step": 16959 + }, + { + "epoch": 7.884704788470479, + "grad_norm": 9.745798110961914, + "learning_rate": 9.983047998304801e-06, + "loss": 0.1486, + "step": 16960 + }, + { + "epoch": 7.885169688516969, + "grad_norm": 6.27370548248291, + "learning_rate": 9.9830469983047e-06, + "loss": 0.1028, + "step": 16961 + }, + { + "epoch": 7.885634588563459, + "grad_norm": 3.802499294281006, + "learning_rate": 9.9830459983046e-06, + "loss": 0.0632, + "step": 16962 + }, + { + "epoch": 7.886099488609949, + "grad_norm": 4.532344818115234, + "learning_rate": 9.983044998304501e-06, + "loss": 0.1002, + "step": 16963 + }, + { + "epoch": 7.8865643886564385, + "grad_norm": 6.626570701599121, + "learning_rate": 9.9830439983044e-06, + "loss": 0.3297, + "step": 16964 + }, + { + "epoch": 7.887029288702929, + "grad_norm": 5.206122398376465, + "learning_rate": 9.983042998304301e-06, + "loss": 0.1384, + "step": 16965 + }, + { + "epoch": 7.887494188749419, + "grad_norm": 5.1117167472839355, + "learning_rate": 9.9830419983042e-06, + "loss": 0.1402, + "step": 16966 + }, + { + "epoch": 7.887959088795909, + "grad_norm": 6.334923267364502, + "learning_rate": 9.9830409983041e-06, + "loss": 0.1152, + "step": 16967 + }, + { + "epoch": 7.888423988842399, + "grad_norm": 7.590034008026123, + "learning_rate": 9.983039998304e-06, + "loss": 0.2283, + "step": 16968 + }, + { + "epoch": 7.888888888888889, + "grad_norm": 4.394374370574951, + "learning_rate": 9.9830389983039e-06, + "loss": 0.0961, + "step": 16969 + }, + { + "epoch": 7.889353788935379, + "grad_norm": 5.964049816131592, + "learning_rate": 9.983037998303801e-06, + "loss": 0.125, + "step": 16970 + }, + { + "epoch": 7.889818688981869, + "grad_norm": 10.34508228302002, + "learning_rate": 9.9830369983037e-06, + "loss": 0.2518, + "step": 16971 + }, + { + "epoch": 7.890283589028359, + "grad_norm": 7.676820278167725, + "learning_rate": 9.9830359983036e-06, + "loss": 0.1737, + "step": 16972 + }, + { + "epoch": 7.8907484890748485, + "grad_norm": 5.7401442527771, + "learning_rate": 9.9830349983035e-06, + "loss": 0.1588, + "step": 16973 + }, + { + "epoch": 7.891213389121339, + "grad_norm": 7.786920547485352, + "learning_rate": 9.9830339983034e-06, + "loss": 0.2885, + "step": 16974 + }, + { + "epoch": 7.891678289167829, + "grad_norm": 5.648083209991455, + "learning_rate": 9.983032998303301e-06, + "loss": 0.1015, + "step": 16975 + }, + { + "epoch": 7.892143189214319, + "grad_norm": 8.591354370117188, + "learning_rate": 9.983031998303202e-06, + "loss": 0.2625, + "step": 16976 + }, + { + "epoch": 7.892608089260809, + "grad_norm": 6.006316184997559, + "learning_rate": 9.9830309983031e-06, + "loss": 0.2578, + "step": 16977 + }, + { + "epoch": 7.893072989307299, + "grad_norm": 7.069599628448486, + "learning_rate": 9.983029998303001e-06, + "loss": 0.1848, + "step": 16978 + }, + { + "epoch": 7.893537889353789, + "grad_norm": 4.861171722412109, + "learning_rate": 9.9830289983029e-06, + "loss": 0.0529, + "step": 16979 + }, + { + "epoch": 7.894002789400279, + "grad_norm": 6.188619136810303, + "learning_rate": 9.9830279983028e-06, + "loss": 0.1966, + "step": 16980 + }, + { + "epoch": 7.894467689446769, + "grad_norm": 4.200199604034424, + "learning_rate": 9.9830269983027e-06, + "loss": 0.0914, + "step": 16981 + }, + { + "epoch": 7.8949325894932585, + "grad_norm": 6.188166618347168, + "learning_rate": 9.9830259983026e-06, + "loss": 0.152, + "step": 16982 + }, + { + "epoch": 7.895397489539749, + "grad_norm": 3.1105399131774902, + "learning_rate": 9.983024998302501e-06, + "loss": 0.0467, + "step": 16983 + }, + { + "epoch": 7.895862389586239, + "grad_norm": 6.509103775024414, + "learning_rate": 9.9830239983024e-06, + "loss": 0.1913, + "step": 16984 + }, + { + "epoch": 7.896327289632729, + "grad_norm": 4.409621238708496, + "learning_rate": 9.9830229983023e-06, + "loss": 0.1179, + "step": 16985 + }, + { + "epoch": 7.896792189679219, + "grad_norm": 6.254547119140625, + "learning_rate": 9.9830219983022e-06, + "loss": 0.2378, + "step": 16986 + }, + { + "epoch": 7.897257089725709, + "grad_norm": 2.3362479209899902, + "learning_rate": 9.9830209983021e-06, + "loss": 0.0449, + "step": 16987 + }, + { + "epoch": 7.897721989772199, + "grad_norm": 6.896369457244873, + "learning_rate": 9.983019998302001e-06, + "loss": 0.1826, + "step": 16988 + }, + { + "epoch": 7.898186889818689, + "grad_norm": 3.996856451034546, + "learning_rate": 9.983018998301902e-06, + "loss": 0.0461, + "step": 16989 + }, + { + "epoch": 7.898651789865179, + "grad_norm": 4.726360321044922, + "learning_rate": 9.9830179983018e-06, + "loss": 0.1464, + "step": 16990 + }, + { + "epoch": 7.899116689911669, + "grad_norm": 8.439376831054688, + "learning_rate": 9.983016998301701e-06, + "loss": 0.2331, + "step": 16991 + }, + { + "epoch": 7.899581589958159, + "grad_norm": 5.688134670257568, + "learning_rate": 9.9830159983016e-06, + "loss": 0.1402, + "step": 16992 + }, + { + "epoch": 7.900046490004649, + "grad_norm": 9.29540729522705, + "learning_rate": 9.983014998301501e-06, + "loss": 0.3979, + "step": 16993 + }, + { + "epoch": 7.900511390051139, + "grad_norm": 4.081732273101807, + "learning_rate": 9.983013998301402e-06, + "loss": 0.0575, + "step": 16994 + }, + { + "epoch": 7.9009762900976295, + "grad_norm": 3.79237699508667, + "learning_rate": 9.9830129983013e-06, + "loss": 0.1044, + "step": 16995 + }, + { + "epoch": 7.901441190144119, + "grad_norm": 6.324005126953125, + "learning_rate": 9.983011998301201e-06, + "loss": 0.1597, + "step": 16996 + }, + { + "epoch": 7.901906090190609, + "grad_norm": 5.6679511070251465, + "learning_rate": 9.9830109983011e-06, + "loss": 0.2223, + "step": 16997 + }, + { + "epoch": 7.902370990237099, + "grad_norm": 7.035114765167236, + "learning_rate": 9.983009998301e-06, + "loss": 0.2384, + "step": 16998 + }, + { + "epoch": 7.902835890283589, + "grad_norm": 3.423736095428467, + "learning_rate": 9.9830089983009e-06, + "loss": 0.0638, + "step": 16999 + }, + { + "epoch": 7.903300790330079, + "grad_norm": 7.550312519073486, + "learning_rate": 9.9830079983008e-06, + "loss": 0.2225, + "step": 17000 + }, + { + "epoch": 7.903765690376569, + "grad_norm": 3.2531256675720215, + "learning_rate": 9.983006998300701e-06, + "loss": 0.0712, + "step": 17001 + }, + { + "epoch": 7.904230590423059, + "grad_norm": 5.364672660827637, + "learning_rate": 9.983005998300602e-06, + "loss": 0.1574, + "step": 17002 + }, + { + "epoch": 7.904695490469549, + "grad_norm": 8.448248863220215, + "learning_rate": 9.9830049983005e-06, + "loss": 0.4037, + "step": 17003 + }, + { + "epoch": 7.9051603905160395, + "grad_norm": 4.954360008239746, + "learning_rate": 9.9830039983004e-06, + "loss": 0.105, + "step": 17004 + }, + { + "epoch": 7.905625290562529, + "grad_norm": 3.983513593673706, + "learning_rate": 9.9830029983003e-06, + "loss": 0.0743, + "step": 17005 + }, + { + "epoch": 7.906090190609019, + "grad_norm": 4.821080684661865, + "learning_rate": 9.983001998300201e-06, + "loss": 0.1587, + "step": 17006 + }, + { + "epoch": 7.906555090655509, + "grad_norm": 1.9159659147262573, + "learning_rate": 9.983000998300102e-06, + "loss": 0.0337, + "step": 17007 + }, + { + "epoch": 7.907019990701999, + "grad_norm": 7.346183776855469, + "learning_rate": 9.9829999983e-06, + "loss": 0.0903, + "step": 17008 + }, + { + "epoch": 7.907484890748489, + "grad_norm": 9.907234191894531, + "learning_rate": 9.982998998299901e-06, + "loss": 0.1134, + "step": 17009 + }, + { + "epoch": 7.907949790794979, + "grad_norm": 5.040562629699707, + "learning_rate": 9.9829979982998e-06, + "loss": 0.2101, + "step": 17010 + }, + { + "epoch": 7.908414690841469, + "grad_norm": 9.20637321472168, + "learning_rate": 9.982996998299701e-06, + "loss": 0.17, + "step": 17011 + }, + { + "epoch": 7.908879590887959, + "grad_norm": 8.16837215423584, + "learning_rate": 9.9829959982996e-06, + "loss": 0.3769, + "step": 17012 + }, + { + "epoch": 7.9093444909344495, + "grad_norm": 3.9017632007598877, + "learning_rate": 9.9829949982995e-06, + "loss": 0.0763, + "step": 17013 + }, + { + "epoch": 7.909809390980939, + "grad_norm": 4.951605319976807, + "learning_rate": 9.982993998299401e-06, + "loss": 0.1353, + "step": 17014 + }, + { + "epoch": 7.910274291027429, + "grad_norm": 3.5071165561676025, + "learning_rate": 9.9829929982993e-06, + "loss": 0.0672, + "step": 17015 + }, + { + "epoch": 7.910739191073919, + "grad_norm": 3.721470355987549, + "learning_rate": 9.9829919982992e-06, + "loss": 0.1026, + "step": 17016 + }, + { + "epoch": 7.911204091120409, + "grad_norm": 3.8491222858428955, + "learning_rate": 9.9829909982991e-06, + "loss": 0.0988, + "step": 17017 + }, + { + "epoch": 7.911668991166899, + "grad_norm": 5.354891300201416, + "learning_rate": 9.982989998299e-06, + "loss": 0.1392, + "step": 17018 + }, + { + "epoch": 7.912133891213389, + "grad_norm": 3.2193315029144287, + "learning_rate": 9.982988998298901e-06, + "loss": 0.0558, + "step": 17019 + }, + { + "epoch": 7.912598791259879, + "grad_norm": 9.198299407958984, + "learning_rate": 9.982987998298802e-06, + "loss": 0.1044, + "step": 17020 + }, + { + "epoch": 7.913063691306369, + "grad_norm": 3.3815431594848633, + "learning_rate": 9.9829869982987e-06, + "loss": 0.0733, + "step": 17021 + }, + { + "epoch": 7.9135285913528595, + "grad_norm": 6.603413105010986, + "learning_rate": 9.982985998298601e-06, + "loss": 0.1433, + "step": 17022 + }, + { + "epoch": 7.913993491399349, + "grad_norm": 1.483778715133667, + "learning_rate": 9.9829849982985e-06, + "loss": 0.0087, + "step": 17023 + }, + { + "epoch": 7.914458391445839, + "grad_norm": 9.35127067565918, + "learning_rate": 9.982983998298401e-06, + "loss": 0.1576, + "step": 17024 + }, + { + "epoch": 7.914923291492329, + "grad_norm": 2.68390154838562, + "learning_rate": 9.9829829982983e-06, + "loss": 0.0328, + "step": 17025 + }, + { + "epoch": 7.915388191538819, + "grad_norm": 2.9436769485473633, + "learning_rate": 9.9829819982982e-06, + "loss": 0.0412, + "step": 17026 + }, + { + "epoch": 7.915853091585309, + "grad_norm": 5.343438625335693, + "learning_rate": 9.982980998298101e-06, + "loss": 0.1596, + "step": 17027 + }, + { + "epoch": 7.916317991631799, + "grad_norm": 9.718710899353027, + "learning_rate": 9.982979998298e-06, + "loss": 0.2638, + "step": 17028 + }, + { + "epoch": 7.916782891678289, + "grad_norm": 3.0511245727539062, + "learning_rate": 9.982978998297901e-06, + "loss": 0.0929, + "step": 17029 + }, + { + "epoch": 7.91724779172478, + "grad_norm": 6.132394313812256, + "learning_rate": 9.9829779982978e-06, + "loss": 0.1884, + "step": 17030 + }, + { + "epoch": 7.9177126917712695, + "grad_norm": 1.9183841943740845, + "learning_rate": 9.9829769982977e-06, + "loss": 0.0117, + "step": 17031 + }, + { + "epoch": 7.918177591817759, + "grad_norm": 6.097613334655762, + "learning_rate": 9.982975998297601e-06, + "loss": 0.132, + "step": 17032 + }, + { + "epoch": 7.918642491864249, + "grad_norm": 6.080667972564697, + "learning_rate": 9.982974998297502e-06, + "loss": 0.2185, + "step": 17033 + }, + { + "epoch": 7.919107391910739, + "grad_norm": 8.0725679397583, + "learning_rate": 9.9829739982974e-06, + "loss": 0.1345, + "step": 17034 + }, + { + "epoch": 7.91957229195723, + "grad_norm": 4.105856418609619, + "learning_rate": 9.982972998297301e-06, + "loss": 0.1156, + "step": 17035 + }, + { + "epoch": 7.920037192003719, + "grad_norm": 7.854269504547119, + "learning_rate": 9.9829719982972e-06, + "loss": 0.2038, + "step": 17036 + }, + { + "epoch": 7.920502092050209, + "grad_norm": 4.137773036956787, + "learning_rate": 9.9829709982971e-06, + "loss": 0.0707, + "step": 17037 + }, + { + "epoch": 7.920966992096699, + "grad_norm": 6.171111106872559, + "learning_rate": 9.982969998297e-06, + "loss": 0.1308, + "step": 17038 + }, + { + "epoch": 7.92143189214319, + "grad_norm": 7.65816068649292, + "learning_rate": 9.9829689982969e-06, + "loss": 0.2744, + "step": 17039 + }, + { + "epoch": 7.9218967921896795, + "grad_norm": 3.8925886154174805, + "learning_rate": 9.982967998296801e-06, + "loss": 0.0774, + "step": 17040 + }, + { + "epoch": 7.922361692236169, + "grad_norm": 4.847796440124512, + "learning_rate": 9.9829669982967e-06, + "loss": 0.1599, + "step": 17041 + }, + { + "epoch": 7.922826592282659, + "grad_norm": 4.408819675445557, + "learning_rate": 9.982965998296601e-06, + "loss": 0.0847, + "step": 17042 + }, + { + "epoch": 7.923291492329149, + "grad_norm": 4.621715068817139, + "learning_rate": 9.9829649982965e-06, + "loss": 0.1137, + "step": 17043 + }, + { + "epoch": 7.92375639237564, + "grad_norm": 3.6632537841796875, + "learning_rate": 9.9829639982964e-06, + "loss": 0.0854, + "step": 17044 + }, + { + "epoch": 7.924221292422129, + "grad_norm": 4.083292484283447, + "learning_rate": 9.982962998296301e-06, + "loss": 0.0467, + "step": 17045 + }, + { + "epoch": 7.924686192468619, + "grad_norm": 4.245035648345947, + "learning_rate": 9.982961998296202e-06, + "loss": 0.0889, + "step": 17046 + }, + { + "epoch": 7.925151092515109, + "grad_norm": 4.905300140380859, + "learning_rate": 9.9829609982961e-06, + "loss": 0.1978, + "step": 17047 + }, + { + "epoch": 7.9256159925616, + "grad_norm": 4.438144683837891, + "learning_rate": 9.982959998296e-06, + "loss": 0.0762, + "step": 17048 + }, + { + "epoch": 7.9260808926080895, + "grad_norm": 6.805122375488281, + "learning_rate": 9.9829589982959e-06, + "loss": 0.0961, + "step": 17049 + }, + { + "epoch": 7.926545792654579, + "grad_norm": 5.435309886932373, + "learning_rate": 9.9829579982958e-06, + "loss": 0.0987, + "step": 17050 + }, + { + "epoch": 7.927010692701069, + "grad_norm": 4.778933525085449, + "learning_rate": 9.9829569982957e-06, + "loss": 0.1293, + "step": 17051 + }, + { + "epoch": 7.927475592747559, + "grad_norm": 6.775227069854736, + "learning_rate": 9.9829559982956e-06, + "loss": 0.3505, + "step": 17052 + }, + { + "epoch": 7.9279404927940496, + "grad_norm": 6.544523239135742, + "learning_rate": 9.982954998295501e-06, + "loss": 0.303, + "step": 17053 + }, + { + "epoch": 7.928405392840539, + "grad_norm": 2.039541244506836, + "learning_rate": 9.9829539982954e-06, + "loss": 0.0292, + "step": 17054 + }, + { + "epoch": 7.928870292887029, + "grad_norm": 4.779314994812012, + "learning_rate": 9.982952998295301e-06, + "loss": 0.0944, + "step": 17055 + }, + { + "epoch": 7.929335192933519, + "grad_norm": 4.150566577911377, + "learning_rate": 9.9829519982952e-06, + "loss": 0.1444, + "step": 17056 + }, + { + "epoch": 7.92980009298001, + "grad_norm": 5.321030616760254, + "learning_rate": 9.9829509982951e-06, + "loss": 0.1012, + "step": 17057 + }, + { + "epoch": 7.9302649930264995, + "grad_norm": 5.708653450012207, + "learning_rate": 9.982949998295001e-06, + "loss": 0.1363, + "step": 17058 + }, + { + "epoch": 7.930729893072989, + "grad_norm": 2.7741904258728027, + "learning_rate": 9.9829489982949e-06, + "loss": 0.0424, + "step": 17059 + }, + { + "epoch": 7.931194793119479, + "grad_norm": 4.5670166015625, + "learning_rate": 9.982947998294801e-06, + "loss": 0.0924, + "step": 17060 + }, + { + "epoch": 7.931659693165969, + "grad_norm": 6.092253684997559, + "learning_rate": 9.9829469982947e-06, + "loss": 0.1328, + "step": 17061 + }, + { + "epoch": 7.9321245932124596, + "grad_norm": 6.867648124694824, + "learning_rate": 9.9829459982946e-06, + "loss": 0.2517, + "step": 17062 + }, + { + "epoch": 7.932589493258949, + "grad_norm": 2.4940860271453857, + "learning_rate": 9.982944998294501e-06, + "loss": 0.0274, + "step": 17063 + }, + { + "epoch": 7.933054393305439, + "grad_norm": 2.642094612121582, + "learning_rate": 9.9829439982944e-06, + "loss": 0.0397, + "step": 17064 + }, + { + "epoch": 7.933519293351929, + "grad_norm": 5.656230449676514, + "learning_rate": 9.9829429982943e-06, + "loss": 0.1277, + "step": 17065 + }, + { + "epoch": 7.93398419339842, + "grad_norm": 4.887856960296631, + "learning_rate": 9.982941998294201e-06, + "loss": 0.1693, + "step": 17066 + }, + { + "epoch": 7.9344490934449095, + "grad_norm": 5.2305192947387695, + "learning_rate": 9.9829409982941e-06, + "loss": 0.1379, + "step": 17067 + }, + { + "epoch": 7.934913993491399, + "grad_norm": 5.501399517059326, + "learning_rate": 9.982939998294001e-06, + "loss": 0.1149, + "step": 17068 + }, + { + "epoch": 7.935378893537889, + "grad_norm": 7.100464344024658, + "learning_rate": 9.9829389982939e-06, + "loss": 0.1672, + "step": 17069 + }, + { + "epoch": 7.93584379358438, + "grad_norm": 6.312687397003174, + "learning_rate": 9.9829379982938e-06, + "loss": 0.223, + "step": 17070 + }, + { + "epoch": 7.9363086936308695, + "grad_norm": 5.18148136138916, + "learning_rate": 9.982936998293701e-06, + "loss": 0.1473, + "step": 17071 + }, + { + "epoch": 7.936773593677359, + "grad_norm": 10.32750415802002, + "learning_rate": 9.9829359982936e-06, + "loss": 0.3438, + "step": 17072 + }, + { + "epoch": 7.937238493723849, + "grad_norm": 8.080711364746094, + "learning_rate": 9.982934998293501e-06, + "loss": 0.2967, + "step": 17073 + }, + { + "epoch": 7.93770339377034, + "grad_norm": 5.19196081161499, + "learning_rate": 9.9829339982934e-06, + "loss": 0.1559, + "step": 17074 + }, + { + "epoch": 7.93816829381683, + "grad_norm": 7.178285598754883, + "learning_rate": 9.9829329982933e-06, + "loss": 0.1718, + "step": 17075 + }, + { + "epoch": 7.9386331938633194, + "grad_norm": 5.092031955718994, + "learning_rate": 9.982931998293201e-06, + "loss": 0.1223, + "step": 17076 + }, + { + "epoch": 7.939098093909809, + "grad_norm": 6.269043445587158, + "learning_rate": 9.9829309982931e-06, + "loss": 0.2125, + "step": 17077 + }, + { + "epoch": 7.939562993956299, + "grad_norm": 4.4937872886657715, + "learning_rate": 9.982929998293001e-06, + "loss": 0.0709, + "step": 17078 + }, + { + "epoch": 7.94002789400279, + "grad_norm": 7.328379154205322, + "learning_rate": 9.982928998292902e-06, + "loss": 0.4047, + "step": 17079 + }, + { + "epoch": 7.9404927940492795, + "grad_norm": 11.16726303100586, + "learning_rate": 9.9829279982928e-06, + "loss": 0.2028, + "step": 17080 + }, + { + "epoch": 7.940957694095769, + "grad_norm": 9.6583251953125, + "learning_rate": 9.9829269982927e-06, + "loss": 0.185, + "step": 17081 + }, + { + "epoch": 7.941422594142259, + "grad_norm": 6.567773818969727, + "learning_rate": 9.9829259982926e-06, + "loss": 0.2613, + "step": 17082 + }, + { + "epoch": 7.94188749418875, + "grad_norm": 5.661920070648193, + "learning_rate": 9.9829249982925e-06, + "loss": 0.1541, + "step": 17083 + }, + { + "epoch": 7.94235239423524, + "grad_norm": 7.822536945343018, + "learning_rate": 9.982923998292401e-06, + "loss": 0.2373, + "step": 17084 + }, + { + "epoch": 7.9428172942817294, + "grad_norm": 5.488408088684082, + "learning_rate": 9.9829229982923e-06, + "loss": 0.1856, + "step": 17085 + }, + { + "epoch": 7.943282194328219, + "grad_norm": 7.838588714599609, + "learning_rate": 9.982921998292201e-06, + "loss": 0.1934, + "step": 17086 + }, + { + "epoch": 7.943747094374709, + "grad_norm": 6.099390506744385, + "learning_rate": 9.9829209982921e-06, + "loss": 0.1232, + "step": 17087 + }, + { + "epoch": 7.9442119944212, + "grad_norm": 6.675954818725586, + "learning_rate": 9.982919998292e-06, + "loss": 0.1693, + "step": 17088 + }, + { + "epoch": 7.9446768944676895, + "grad_norm": 5.528357982635498, + "learning_rate": 9.982918998291901e-06, + "loss": 0.147, + "step": 17089 + }, + { + "epoch": 7.945141794514179, + "grad_norm": 7.4640350341796875, + "learning_rate": 9.982917998291802e-06, + "loss": 0.3164, + "step": 17090 + }, + { + "epoch": 7.945606694560669, + "grad_norm": 5.384706020355225, + "learning_rate": 9.982916998291701e-06, + "loss": 0.1323, + "step": 17091 + }, + { + "epoch": 7.94607159460716, + "grad_norm": 4.226874828338623, + "learning_rate": 9.9829159982916e-06, + "loss": 0.0705, + "step": 17092 + }, + { + "epoch": 7.94653649465365, + "grad_norm": 8.676591873168945, + "learning_rate": 9.9829149982915e-06, + "loss": 0.1799, + "step": 17093 + }, + { + "epoch": 7.947001394700139, + "grad_norm": 4.040536880493164, + "learning_rate": 9.9829139982914e-06, + "loss": 0.098, + "step": 17094 + }, + { + "epoch": 7.947466294746629, + "grad_norm": 5.357146739959717, + "learning_rate": 9.9829129982913e-06, + "loss": 0.1171, + "step": 17095 + }, + { + "epoch": 7.947931194793119, + "grad_norm": 5.909775257110596, + "learning_rate": 9.982911998291201e-06, + "loss": 0.1531, + "step": 17096 + }, + { + "epoch": 7.94839609483961, + "grad_norm": 9.055764198303223, + "learning_rate": 9.982910998291102e-06, + "loss": 0.2968, + "step": 17097 + }, + { + "epoch": 7.9488609948860995, + "grad_norm": 5.902835369110107, + "learning_rate": 9.982909998291e-06, + "loss": 0.157, + "step": 17098 + }, + { + "epoch": 7.949325894932589, + "grad_norm": 5.162913799285889, + "learning_rate": 9.982908998290901e-06, + "loss": 0.1032, + "step": 17099 + }, + { + "epoch": 7.949790794979079, + "grad_norm": 6.80171012878418, + "learning_rate": 9.9829079982908e-06, + "loss": 0.1108, + "step": 17100 + }, + { + "epoch": 7.95025569502557, + "grad_norm": 5.26826810836792, + "learning_rate": 9.9829069982907e-06, + "loss": 0.167, + "step": 17101 + }, + { + "epoch": 7.95072059507206, + "grad_norm": 6.73712158203125, + "learning_rate": 9.982905998290601e-06, + "loss": 0.2459, + "step": 17102 + }, + { + "epoch": 7.951185495118549, + "grad_norm": 8.543946266174316, + "learning_rate": 9.9829049982905e-06, + "loss": 0.1236, + "step": 17103 + }, + { + "epoch": 7.951650395165039, + "grad_norm": 4.86175537109375, + "learning_rate": 9.982903998290401e-06, + "loss": 0.1117, + "step": 17104 + }, + { + "epoch": 7.952115295211529, + "grad_norm": 3.79032039642334, + "learning_rate": 9.9829029982903e-06, + "loss": 0.0522, + "step": 17105 + }, + { + "epoch": 7.95258019525802, + "grad_norm": 7.8864521980285645, + "learning_rate": 9.9829019982902e-06, + "loss": 0.1827, + "step": 17106 + }, + { + "epoch": 7.9530450953045095, + "grad_norm": 4.4288763999938965, + "learning_rate": 9.9829009982901e-06, + "loss": 0.1225, + "step": 17107 + }, + { + "epoch": 7.953509995350999, + "grad_norm": 6.094046592712402, + "learning_rate": 9.98289999829e-06, + "loss": 0.2232, + "step": 17108 + }, + { + "epoch": 7.95397489539749, + "grad_norm": 6.565385341644287, + "learning_rate": 9.982898998289901e-06, + "loss": 0.1503, + "step": 17109 + }, + { + "epoch": 7.95443979544398, + "grad_norm": 7.549901008605957, + "learning_rate": 9.982897998289802e-06, + "loss": 0.174, + "step": 17110 + }, + { + "epoch": 7.95490469549047, + "grad_norm": 3.0282278060913086, + "learning_rate": 9.9828969982897e-06, + "loss": 0.0411, + "step": 17111 + }, + { + "epoch": 7.955369595536959, + "grad_norm": 6.449915409088135, + "learning_rate": 9.9828959982896e-06, + "loss": 0.1902, + "step": 17112 + }, + { + "epoch": 7.955834495583449, + "grad_norm": 5.048168659210205, + "learning_rate": 9.9828949982895e-06, + "loss": 0.1049, + "step": 17113 + }, + { + "epoch": 7.95629939562994, + "grad_norm": 4.377949237823486, + "learning_rate": 9.982893998289401e-06, + "loss": 0.082, + "step": 17114 + }, + { + "epoch": 7.95676429567643, + "grad_norm": 3.4267799854278564, + "learning_rate": 9.982892998289302e-06, + "loss": 0.0902, + "step": 17115 + }, + { + "epoch": 7.9572291957229195, + "grad_norm": 8.555951118469238, + "learning_rate": 9.9828919982892e-06, + "loss": 0.3715, + "step": 17116 + }, + { + "epoch": 7.957694095769409, + "grad_norm": 4.0648603439331055, + "learning_rate": 9.982890998289101e-06, + "loss": 0.1175, + "step": 17117 + }, + { + "epoch": 7.9581589958159, + "grad_norm": 5.335607528686523, + "learning_rate": 9.982889998289e-06, + "loss": 0.0623, + "step": 17118 + }, + { + "epoch": 7.95862389586239, + "grad_norm": 6.0707244873046875, + "learning_rate": 9.9828889982889e-06, + "loss": 0.1914, + "step": 17119 + }, + { + "epoch": 7.95908879590888, + "grad_norm": 3.5650815963745117, + "learning_rate": 9.9828879982888e-06, + "loss": 0.094, + "step": 17120 + }, + { + "epoch": 7.959553695955369, + "grad_norm": 4.0033860206604, + "learning_rate": 9.9828869982887e-06, + "loss": 0.0689, + "step": 17121 + }, + { + "epoch": 7.960018596001859, + "grad_norm": 5.000399589538574, + "learning_rate": 9.982885998288601e-06, + "loss": 0.1013, + "step": 17122 + }, + { + "epoch": 7.96048349604835, + "grad_norm": 4.7268476486206055, + "learning_rate": 9.9828849982885e-06, + "loss": 0.1549, + "step": 17123 + }, + { + "epoch": 7.96094839609484, + "grad_norm": 11.71833610534668, + "learning_rate": 9.9828839982884e-06, + "loss": 0.1336, + "step": 17124 + }, + { + "epoch": 7.9614132961413295, + "grad_norm": 2.9480981826782227, + "learning_rate": 9.9828829982883e-06, + "loss": 0.0464, + "step": 17125 + }, + { + "epoch": 7.961878196187819, + "grad_norm": 5.239193916320801, + "learning_rate": 9.9828819982882e-06, + "loss": 0.2184, + "step": 17126 + }, + { + "epoch": 7.96234309623431, + "grad_norm": 5.270391941070557, + "learning_rate": 9.982880998288101e-06, + "loss": 0.0958, + "step": 17127 + }, + { + "epoch": 7.9628079962808, + "grad_norm": 3.3091318607330322, + "learning_rate": 9.982879998288002e-06, + "loss": 0.052, + "step": 17128 + }, + { + "epoch": 7.96327289632729, + "grad_norm": 4.223939418792725, + "learning_rate": 9.9828789982879e-06, + "loss": 0.083, + "step": 17129 + }, + { + "epoch": 7.963737796373779, + "grad_norm": 3.531236171722412, + "learning_rate": 9.982877998287801e-06, + "loss": 0.0978, + "step": 17130 + }, + { + "epoch": 7.964202696420269, + "grad_norm": 8.450483322143555, + "learning_rate": 9.9828769982877e-06, + "loss": 0.246, + "step": 17131 + }, + { + "epoch": 7.96466759646676, + "grad_norm": 5.078461647033691, + "learning_rate": 9.982875998287601e-06, + "loss": 0.1269, + "step": 17132 + }, + { + "epoch": 7.96513249651325, + "grad_norm": 3.8584225177764893, + "learning_rate": 9.9828749982875e-06, + "loss": 0.0564, + "step": 17133 + }, + { + "epoch": 7.9655973965597395, + "grad_norm": 4.273465156555176, + "learning_rate": 9.9828739982874e-06, + "loss": 0.0942, + "step": 17134 + }, + { + "epoch": 7.966062296606229, + "grad_norm": 5.555017471313477, + "learning_rate": 9.982872998287301e-06, + "loss": 0.174, + "step": 17135 + }, + { + "epoch": 7.96652719665272, + "grad_norm": 5.607870578765869, + "learning_rate": 9.9828719982872e-06, + "loss": 0.1542, + "step": 17136 + }, + { + "epoch": 7.96699209669921, + "grad_norm": 2.418548822402954, + "learning_rate": 9.9828709982871e-06, + "loss": 0.0248, + "step": 17137 + }, + { + "epoch": 7.9674569967457, + "grad_norm": 5.8860368728637695, + "learning_rate": 9.982869998287e-06, + "loss": 0.232, + "step": 17138 + }, + { + "epoch": 7.967921896792189, + "grad_norm": 2.2966792583465576, + "learning_rate": 9.9828689982869e-06, + "loss": 0.0371, + "step": 17139 + }, + { + "epoch": 7.968386796838679, + "grad_norm": 6.133965492248535, + "learning_rate": 9.982867998286801e-06, + "loss": 0.1205, + "step": 17140 + }, + { + "epoch": 7.96885169688517, + "grad_norm": 6.153031349182129, + "learning_rate": 9.982866998286702e-06, + "loss": 0.1568, + "step": 17141 + }, + { + "epoch": 7.96931659693166, + "grad_norm": 4.449592113494873, + "learning_rate": 9.9828659982866e-06, + "loss": 0.1048, + "step": 17142 + }, + { + "epoch": 7.9697814969781495, + "grad_norm": 2.8847286701202393, + "learning_rate": 9.982864998286501e-06, + "loss": 0.0753, + "step": 17143 + }, + { + "epoch": 7.970246397024639, + "grad_norm": 9.146797180175781, + "learning_rate": 9.9828639982864e-06, + "loss": 0.2744, + "step": 17144 + }, + { + "epoch": 7.97071129707113, + "grad_norm": 2.900456666946411, + "learning_rate": 9.982862998286301e-06, + "loss": 0.0333, + "step": 17145 + }, + { + "epoch": 7.97117619711762, + "grad_norm": 5.515718460083008, + "learning_rate": 9.9828619982862e-06, + "loss": 0.1126, + "step": 17146 + }, + { + "epoch": 7.97164109716411, + "grad_norm": 6.876318454742432, + "learning_rate": 9.9828609982861e-06, + "loss": 0.1537, + "step": 17147 + }, + { + "epoch": 7.972105997210599, + "grad_norm": 6.70517110824585, + "learning_rate": 9.982859998286001e-06, + "loss": 0.1195, + "step": 17148 + }, + { + "epoch": 7.972570897257089, + "grad_norm": 8.178765296936035, + "learning_rate": 9.9828589982859e-06, + "loss": 0.2676, + "step": 17149 + }, + { + "epoch": 7.97303579730358, + "grad_norm": 5.668246269226074, + "learning_rate": 9.9828579982858e-06, + "loss": 0.158, + "step": 17150 + }, + { + "epoch": 7.97350069735007, + "grad_norm": 7.402012348175049, + "learning_rate": 9.9828569982857e-06, + "loss": 0.186, + "step": 17151 + }, + { + "epoch": 7.9739655973965595, + "grad_norm": 2.1893157958984375, + "learning_rate": 9.9828559982856e-06, + "loss": 0.0266, + "step": 17152 + }, + { + "epoch": 7.97443049744305, + "grad_norm": 3.9361586570739746, + "learning_rate": 9.982854998285501e-06, + "loss": 0.0827, + "step": 17153 + }, + { + "epoch": 7.97489539748954, + "grad_norm": 4.742891788482666, + "learning_rate": 9.982853998285402e-06, + "loss": 0.0834, + "step": 17154 + }, + { + "epoch": 7.97536029753603, + "grad_norm": 4.734408855438232, + "learning_rate": 9.9828529982853e-06, + "loss": 0.1232, + "step": 17155 + }, + { + "epoch": 7.97582519758252, + "grad_norm": 6.024378776550293, + "learning_rate": 9.9828519982852e-06, + "loss": 0.213, + "step": 17156 + }, + { + "epoch": 7.976290097629009, + "grad_norm": 7.550882339477539, + "learning_rate": 9.9828509982851e-06, + "loss": 0.2811, + "step": 17157 + }, + { + "epoch": 7.9767549976755, + "grad_norm": 3.4020638465881348, + "learning_rate": 9.982849998285001e-06, + "loss": 0.0442, + "step": 17158 + }, + { + "epoch": 7.97721989772199, + "grad_norm": 2.64357590675354, + "learning_rate": 9.982848998284902e-06, + "loss": 0.0395, + "step": 17159 + }, + { + "epoch": 7.97768479776848, + "grad_norm": 4.834405899047852, + "learning_rate": 9.9828479982848e-06, + "loss": 0.0992, + "step": 17160 + }, + { + "epoch": 7.9781496978149695, + "grad_norm": 6.375398635864258, + "learning_rate": 9.982846998284701e-06, + "loss": 0.1379, + "step": 17161 + }, + { + "epoch": 7.97861459786146, + "grad_norm": 5.793851852416992, + "learning_rate": 9.9828459982846e-06, + "loss": 0.0751, + "step": 17162 + }, + { + "epoch": 7.97907949790795, + "grad_norm": 4.806524753570557, + "learning_rate": 9.982844998284501e-06, + "loss": 0.0713, + "step": 17163 + }, + { + "epoch": 7.97954439795444, + "grad_norm": 6.674041271209717, + "learning_rate": 9.9828439982844e-06, + "loss": 0.1413, + "step": 17164 + }, + { + "epoch": 7.98000929800093, + "grad_norm": 8.344890594482422, + "learning_rate": 9.9828429982843e-06, + "loss": 0.409, + "step": 17165 + }, + { + "epoch": 7.980474198047419, + "grad_norm": 3.8435544967651367, + "learning_rate": 9.982841998284201e-06, + "loss": 0.0631, + "step": 17166 + }, + { + "epoch": 7.98093909809391, + "grad_norm": 5.482803821563721, + "learning_rate": 9.9828409982841e-06, + "loss": 0.1139, + "step": 17167 + }, + { + "epoch": 7.9814039981404, + "grad_norm": 1.6286615133285522, + "learning_rate": 9.982839998284e-06, + "loss": 0.0249, + "step": 17168 + }, + { + "epoch": 7.98186889818689, + "grad_norm": 5.152353286743164, + "learning_rate": 9.9828389982839e-06, + "loss": 0.1109, + "step": 17169 + }, + { + "epoch": 7.9823337982333795, + "grad_norm": 6.787710666656494, + "learning_rate": 9.9828379982838e-06, + "loss": 0.3484, + "step": 17170 + }, + { + "epoch": 7.98279869827987, + "grad_norm": 5.091573238372803, + "learning_rate": 9.982836998283701e-06, + "loss": 0.1991, + "step": 17171 + }, + { + "epoch": 7.98326359832636, + "grad_norm": 7.5573601722717285, + "learning_rate": 9.982835998283602e-06, + "loss": 0.2342, + "step": 17172 + }, + { + "epoch": 7.98372849837285, + "grad_norm": 4.304774284362793, + "learning_rate": 9.9828349982835e-06, + "loss": 0.1052, + "step": 17173 + }, + { + "epoch": 7.98419339841934, + "grad_norm": 3.643050193786621, + "learning_rate": 9.982833998283401e-06, + "loss": 0.0495, + "step": 17174 + }, + { + "epoch": 7.984658298465829, + "grad_norm": 8.396964073181152, + "learning_rate": 9.9828329982833e-06, + "loss": 0.2922, + "step": 17175 + }, + { + "epoch": 7.98512319851232, + "grad_norm": 6.911900997161865, + "learning_rate": 9.982831998283201e-06, + "loss": 0.226, + "step": 17176 + }, + { + "epoch": 7.98558809855881, + "grad_norm": 8.115405082702637, + "learning_rate": 9.9828309982831e-06, + "loss": 0.2763, + "step": 17177 + }, + { + "epoch": 7.9860529986053, + "grad_norm": 2.236365556716919, + "learning_rate": 9.982829998283e-06, + "loss": 0.0247, + "step": 17178 + }, + { + "epoch": 7.9865178986517895, + "grad_norm": 9.858132362365723, + "learning_rate": 9.982828998282901e-06, + "loss": 0.2721, + "step": 17179 + }, + { + "epoch": 7.98698279869828, + "grad_norm": 5.7804365158081055, + "learning_rate": 9.9828279982828e-06, + "loss": 0.1573, + "step": 17180 + }, + { + "epoch": 7.98744769874477, + "grad_norm": 13.458011627197266, + "learning_rate": 9.982826998282701e-06, + "loss": 0.086, + "step": 17181 + }, + { + "epoch": 7.98791259879126, + "grad_norm": 5.320595741271973, + "learning_rate": 9.9828259982826e-06, + "loss": 0.0788, + "step": 17182 + }, + { + "epoch": 7.98837749883775, + "grad_norm": 10.809183120727539, + "learning_rate": 9.9828249982825e-06, + "loss": 0.4107, + "step": 17183 + }, + { + "epoch": 7.988842398884239, + "grad_norm": 4.108271598815918, + "learning_rate": 9.982823998282401e-06, + "loss": 0.0493, + "step": 17184 + }, + { + "epoch": 7.98930729893073, + "grad_norm": 6.278789043426514, + "learning_rate": 9.982822998282302e-06, + "loss": 0.11, + "step": 17185 + }, + { + "epoch": 7.98977219897722, + "grad_norm": 8.920787811279297, + "learning_rate": 9.9828219982822e-06, + "loss": 0.2142, + "step": 17186 + }, + { + "epoch": 7.99023709902371, + "grad_norm": 4.4999470710754395, + "learning_rate": 9.982820998282101e-06, + "loss": 0.1451, + "step": 17187 + }, + { + "epoch": 7.9907019990702, + "grad_norm": 6.052845001220703, + "learning_rate": 9.982819998282e-06, + "loss": 0.2285, + "step": 17188 + }, + { + "epoch": 7.99116689911669, + "grad_norm": 5.3376145362854, + "learning_rate": 9.9828189982819e-06, + "loss": 0.1026, + "step": 17189 + }, + { + "epoch": 7.99163179916318, + "grad_norm": 9.778023719787598, + "learning_rate": 9.9828179982818e-06, + "loss": 0.5049, + "step": 17190 + }, + { + "epoch": 7.99209669920967, + "grad_norm": 4.5963287353515625, + "learning_rate": 9.9828169982817e-06, + "loss": 0.0798, + "step": 17191 + }, + { + "epoch": 7.99256159925616, + "grad_norm": 5.978337287902832, + "learning_rate": 9.982815998281601e-06, + "loss": 0.0848, + "step": 17192 + }, + { + "epoch": 7.99302649930265, + "grad_norm": 3.889465570449829, + "learning_rate": 9.9828149982815e-06, + "loss": 0.0698, + "step": 17193 + }, + { + "epoch": 7.99349139934914, + "grad_norm": 4.764575004577637, + "learning_rate": 9.982813998281401e-06, + "loss": 0.0933, + "step": 17194 + }, + { + "epoch": 7.99395629939563, + "grad_norm": 4.009669303894043, + "learning_rate": 9.9828129982813e-06, + "loss": 0.1292, + "step": 17195 + }, + { + "epoch": 7.99442119944212, + "grad_norm": 6.34096622467041, + "learning_rate": 9.9828119982812e-06, + "loss": 0.2406, + "step": 17196 + }, + { + "epoch": 7.99488609948861, + "grad_norm": 3.0922555923461914, + "learning_rate": 9.982810998281101e-06, + "loss": 0.0512, + "step": 17197 + }, + { + "epoch": 7.9953509995351, + "grad_norm": 4.4187140464782715, + "learning_rate": 9.982809998281002e-06, + "loss": 0.049, + "step": 17198 + }, + { + "epoch": 7.99581589958159, + "grad_norm": 3.0000154972076416, + "learning_rate": 9.982808998280901e-06, + "loss": 0.0454, + "step": 17199 + }, + { + "epoch": 7.99628079962808, + "grad_norm": 4.347187042236328, + "learning_rate": 9.9828079982808e-06, + "loss": 0.1015, + "step": 17200 + }, + { + "epoch": 7.99674569967457, + "grad_norm": 5.9308905601501465, + "learning_rate": 9.9828069982807e-06, + "loss": 0.1536, + "step": 17201 + }, + { + "epoch": 7.99721059972106, + "grad_norm": 5.789645195007324, + "learning_rate": 9.9828059982806e-06, + "loss": 0.1711, + "step": 17202 + }, + { + "epoch": 7.99767549976755, + "grad_norm": 4.680807590484619, + "learning_rate": 9.9828049982805e-06, + "loss": 0.1287, + "step": 17203 + }, + { + "epoch": 7.99814039981404, + "grad_norm": 5.075160503387451, + "learning_rate": 9.9828039982804e-06, + "loss": 0.208, + "step": 17204 + }, + { + "epoch": 7.99860529986053, + "grad_norm": 5.33853006362915, + "learning_rate": 9.982802998280301e-06, + "loss": 0.1093, + "step": 17205 + }, + { + "epoch": 7.99907019990702, + "grad_norm": 3.8530914783477783, + "learning_rate": 9.9828019982802e-06, + "loss": 0.0816, + "step": 17206 + }, + { + "epoch": 7.99953509995351, + "grad_norm": 7.453943729400635, + "learning_rate": 9.982800998280101e-06, + "loss": 0.1655, + "step": 17207 + }, + { + "epoch": 8.0, + "grad_norm": 7.226437091827393, + "learning_rate": 9.98279999828e-06, + "loss": 0.1513, + "step": 17208 + }, + { + "epoch": 8.00046490004649, + "grad_norm": 4.599994659423828, + "learning_rate": 9.9827989982799e-06, + "loss": 0.0506, + "step": 17209 + }, + { + "epoch": 8.00092980009298, + "grad_norm": 4.6764817237854, + "learning_rate": 9.982797998279801e-06, + "loss": 0.0905, + "step": 17210 + }, + { + "epoch": 8.00139470013947, + "grad_norm": 3.5479490756988525, + "learning_rate": 9.9827969982797e-06, + "loss": 0.0658, + "step": 17211 + }, + { + "epoch": 8.00185960018596, + "grad_norm": 5.68881893157959, + "learning_rate": 9.982795998279601e-06, + "loss": 0.0971, + "step": 17212 + }, + { + "epoch": 8.00232450023245, + "grad_norm": 3.951150894165039, + "learning_rate": 9.9827949982795e-06, + "loss": 0.0927, + "step": 17213 + }, + { + "epoch": 8.00278940027894, + "grad_norm": 3.4449880123138428, + "learning_rate": 9.9827939982794e-06, + "loss": 0.0486, + "step": 17214 + }, + { + "epoch": 8.00325430032543, + "grad_norm": 6.445145130157471, + "learning_rate": 9.9827929982793e-06, + "loss": 0.1296, + "step": 17215 + }, + { + "epoch": 8.00371920037192, + "grad_norm": 4.165614128112793, + "learning_rate": 9.9827919982792e-06, + "loss": 0.1217, + "step": 17216 + }, + { + "epoch": 8.00418410041841, + "grad_norm": 4.343128681182861, + "learning_rate": 9.982790998279101e-06, + "loss": 0.076, + "step": 17217 + }, + { + "epoch": 8.0046490004649, + "grad_norm": 3.925166606903076, + "learning_rate": 9.982789998279002e-06, + "loss": 0.092, + "step": 17218 + }, + { + "epoch": 8.00511390051139, + "grad_norm": 0.738274872303009, + "learning_rate": 9.9827889982789e-06, + "loss": 0.0071, + "step": 17219 + }, + { + "epoch": 8.00557880055788, + "grad_norm": 2.4972550868988037, + "learning_rate": 9.982787998278801e-06, + "loss": 0.0282, + "step": 17220 + }, + { + "epoch": 8.00604370060437, + "grad_norm": 4.2161760330200195, + "learning_rate": 9.9827869982787e-06, + "loss": 0.1084, + "step": 17221 + }, + { + "epoch": 8.00650860065086, + "grad_norm": 3.2239065170288086, + "learning_rate": 9.9827859982786e-06, + "loss": 0.031, + "step": 17222 + }, + { + "epoch": 8.00697350069735, + "grad_norm": 6.019814968109131, + "learning_rate": 9.982784998278501e-06, + "loss": 0.1158, + "step": 17223 + }, + { + "epoch": 8.00743840074384, + "grad_norm": 5.291481018066406, + "learning_rate": 9.9827839982784e-06, + "loss": 0.1331, + "step": 17224 + }, + { + "epoch": 8.00790330079033, + "grad_norm": 4.597778797149658, + "learning_rate": 9.982782998278301e-06, + "loss": 0.1514, + "step": 17225 + }, + { + "epoch": 8.00836820083682, + "grad_norm": 5.237550258636475, + "learning_rate": 9.9827819982782e-06, + "loss": 0.068, + "step": 17226 + }, + { + "epoch": 8.00883310088331, + "grad_norm": 7.027193069458008, + "learning_rate": 9.9827809982781e-06, + "loss": 0.2371, + "step": 17227 + }, + { + "epoch": 8.0092980009298, + "grad_norm": 6.127621650695801, + "learning_rate": 9.982779998278001e-06, + "loss": 0.107, + "step": 17228 + }, + { + "epoch": 8.00976290097629, + "grad_norm": 4.5120930671691895, + "learning_rate": 9.9827789982779e-06, + "loss": 0.0967, + "step": 17229 + }, + { + "epoch": 8.010227801022781, + "grad_norm": 5.438161373138428, + "learning_rate": 9.982777998277801e-06, + "loss": 0.1543, + "step": 17230 + }, + { + "epoch": 8.01069270106927, + "grad_norm": 5.018343925476074, + "learning_rate": 9.982776998277702e-06, + "loss": 0.0902, + "step": 17231 + }, + { + "epoch": 8.01115760111576, + "grad_norm": 3.0512845516204834, + "learning_rate": 9.9827759982776e-06, + "loss": 0.0611, + "step": 17232 + }, + { + "epoch": 8.01162250116225, + "grad_norm": 4.755585670471191, + "learning_rate": 9.9827749982775e-06, + "loss": 0.1032, + "step": 17233 + }, + { + "epoch": 8.01208740120874, + "grad_norm": 1.7646592855453491, + "learning_rate": 9.9827739982774e-06, + "loss": 0.0196, + "step": 17234 + }, + { + "epoch": 8.01255230125523, + "grad_norm": 2.6239588260650635, + "learning_rate": 9.9827729982773e-06, + "loss": 0.0545, + "step": 17235 + }, + { + "epoch": 8.01301720130172, + "grad_norm": 4.041385650634766, + "learning_rate": 9.982771998277202e-06, + "loss": 0.0623, + "step": 17236 + }, + { + "epoch": 8.01348210134821, + "grad_norm": 1.763500452041626, + "learning_rate": 9.9827709982771e-06, + "loss": 0.0205, + "step": 17237 + }, + { + "epoch": 8.0139470013947, + "grad_norm": 2.3858039379119873, + "learning_rate": 9.982769998277001e-06, + "loss": 0.0217, + "step": 17238 + }, + { + "epoch": 8.014411901441191, + "grad_norm": 6.199762344360352, + "learning_rate": 9.9827689982769e-06, + "loss": 0.1277, + "step": 17239 + }, + { + "epoch": 8.01487680148768, + "grad_norm": 5.205715179443359, + "learning_rate": 9.9827679982768e-06, + "loss": 0.0441, + "step": 17240 + }, + { + "epoch": 8.01534170153417, + "grad_norm": 5.823945999145508, + "learning_rate": 9.982766998276701e-06, + "loss": 0.3073, + "step": 17241 + }, + { + "epoch": 8.01580660158066, + "grad_norm": 2.096449613571167, + "learning_rate": 9.9827659982766e-06, + "loss": 0.032, + "step": 17242 + }, + { + "epoch": 8.01627150162715, + "grad_norm": 6.213005542755127, + "learning_rate": 9.982764998276501e-06, + "loss": 0.1147, + "step": 17243 + }, + { + "epoch": 8.01673640167364, + "grad_norm": 6.081392288208008, + "learning_rate": 9.9827639982764e-06, + "loss": 0.13, + "step": 17244 + }, + { + "epoch": 8.01720130172013, + "grad_norm": 1.3508027791976929, + "learning_rate": 9.9827629982763e-06, + "loss": 0.0168, + "step": 17245 + }, + { + "epoch": 8.01766620176662, + "grad_norm": 4.399350643157959, + "learning_rate": 9.9827619982762e-06, + "loss": 0.0913, + "step": 17246 + }, + { + "epoch": 8.01813110181311, + "grad_norm": 7.302082538604736, + "learning_rate": 9.9827609982761e-06, + "loss": 0.0465, + "step": 17247 + }, + { + "epoch": 8.018596001859601, + "grad_norm": 6.3941874504089355, + "learning_rate": 9.982759998276001e-06, + "loss": 0.1296, + "step": 17248 + }, + { + "epoch": 8.01906090190609, + "grad_norm": 5.605107307434082, + "learning_rate": 9.982758998275902e-06, + "loss": 0.1373, + "step": 17249 + }, + { + "epoch": 8.01952580195258, + "grad_norm": 8.041790962219238, + "learning_rate": 9.9827579982758e-06, + "loss": 0.1215, + "step": 17250 + }, + { + "epoch": 8.01999070199907, + "grad_norm": 3.6521100997924805, + "learning_rate": 9.982756998275701e-06, + "loss": 0.0878, + "step": 17251 + }, + { + "epoch": 8.02045560204556, + "grad_norm": 2.302353858947754, + "learning_rate": 9.9827559982756e-06, + "loss": 0.0237, + "step": 17252 + }, + { + "epoch": 8.02092050209205, + "grad_norm": 6.139383792877197, + "learning_rate": 9.9827549982755e-06, + "loss": 0.0793, + "step": 17253 + }, + { + "epoch": 8.02138540213854, + "grad_norm": 6.14135217666626, + "learning_rate": 9.982753998275401e-06, + "loss": 0.0721, + "step": 17254 + }, + { + "epoch": 8.02185030218503, + "grad_norm": 3.728982925415039, + "learning_rate": 9.9827529982753e-06, + "loss": 0.0761, + "step": 17255 + }, + { + "epoch": 8.02231520223152, + "grad_norm": 5.0399065017700195, + "learning_rate": 9.982751998275201e-06, + "loss": 0.0921, + "step": 17256 + }, + { + "epoch": 8.022780102278011, + "grad_norm": 1.5391560792922974, + "learning_rate": 9.9827509982751e-06, + "loss": 0.0224, + "step": 17257 + }, + { + "epoch": 8.0232450023245, + "grad_norm": 4.456899642944336, + "learning_rate": 9.982749998275e-06, + "loss": 0.1099, + "step": 17258 + }, + { + "epoch": 8.02370990237099, + "grad_norm": 5.166774749755859, + "learning_rate": 9.9827489982749e-06, + "loss": 0.088, + "step": 17259 + }, + { + "epoch": 8.02417480241748, + "grad_norm": 4.014657020568848, + "learning_rate": 9.9827479982748e-06, + "loss": 0.0994, + "step": 17260 + }, + { + "epoch": 8.02463970246397, + "grad_norm": 13.751716613769531, + "learning_rate": 9.982746998274701e-06, + "loss": 0.1617, + "step": 17261 + }, + { + "epoch": 8.02510460251046, + "grad_norm": 2.328310251235962, + "learning_rate": 9.982745998274602e-06, + "loss": 0.0426, + "step": 17262 + }, + { + "epoch": 8.02556950255695, + "grad_norm": 4.117763042449951, + "learning_rate": 9.9827449982745e-06, + "loss": 0.0932, + "step": 17263 + }, + { + "epoch": 8.02603440260344, + "grad_norm": 3.3658225536346436, + "learning_rate": 9.9827439982744e-06, + "loss": 0.0436, + "step": 17264 + }, + { + "epoch": 8.02649930264993, + "grad_norm": 5.674056529998779, + "learning_rate": 9.9827429982743e-06, + "loss": 0.1144, + "step": 17265 + }, + { + "epoch": 8.026964202696421, + "grad_norm": 7.406521320343018, + "learning_rate": 9.982741998274201e-06, + "loss": 0.118, + "step": 17266 + }, + { + "epoch": 8.02742910274291, + "grad_norm": 5.91575813293457, + "learning_rate": 9.982740998274102e-06, + "loss": 0.1324, + "step": 17267 + }, + { + "epoch": 8.0278940027894, + "grad_norm": 4.802196979522705, + "learning_rate": 9.982739998274e-06, + "loss": 0.1482, + "step": 17268 + }, + { + "epoch": 8.02835890283589, + "grad_norm": 4.442540168762207, + "learning_rate": 9.982738998273901e-06, + "loss": 0.0785, + "step": 17269 + }, + { + "epoch": 8.02882380288238, + "grad_norm": 8.637701988220215, + "learning_rate": 9.9827379982738e-06, + "loss": 0.0649, + "step": 17270 + }, + { + "epoch": 8.02928870292887, + "grad_norm": 5.08730936050415, + "learning_rate": 9.9827369982737e-06, + "loss": 0.1168, + "step": 17271 + }, + { + "epoch": 8.02975360297536, + "grad_norm": 4.794325351715088, + "learning_rate": 9.9827359982736e-06, + "loss": 0.0831, + "step": 17272 + }, + { + "epoch": 8.03021850302185, + "grad_norm": 2.65596866607666, + "learning_rate": 9.9827349982735e-06, + "loss": 0.0221, + "step": 17273 + }, + { + "epoch": 8.030683403068341, + "grad_norm": 8.383350372314453, + "learning_rate": 9.982733998273401e-06, + "loss": 0.2306, + "step": 17274 + }, + { + "epoch": 8.031148303114831, + "grad_norm": 4.5658063888549805, + "learning_rate": 9.9827329982733e-06, + "loss": 0.0886, + "step": 17275 + }, + { + "epoch": 8.03161320316132, + "grad_norm": 7.00847053527832, + "learning_rate": 9.9827319982732e-06, + "loss": 0.1389, + "step": 17276 + }, + { + "epoch": 8.03207810320781, + "grad_norm": 5.994534015655518, + "learning_rate": 9.9827309982731e-06, + "loss": 0.1799, + "step": 17277 + }, + { + "epoch": 8.0325430032543, + "grad_norm": 3.773066759109497, + "learning_rate": 9.982729998273e-06, + "loss": 0.0626, + "step": 17278 + }, + { + "epoch": 8.03300790330079, + "grad_norm": 6.703011512756348, + "learning_rate": 9.982728998272901e-06, + "loss": 0.1199, + "step": 17279 + }, + { + "epoch": 8.03347280334728, + "grad_norm": 1.218569040298462, + "learning_rate": 9.982727998272802e-06, + "loss": 0.0126, + "step": 17280 + }, + { + "epoch": 8.03393770339377, + "grad_norm": 4.2066521644592285, + "learning_rate": 9.9827269982727e-06, + "loss": 0.0961, + "step": 17281 + }, + { + "epoch": 8.03440260344026, + "grad_norm": 4.194593906402588, + "learning_rate": 9.982725998272601e-06, + "loss": 0.043, + "step": 17282 + }, + { + "epoch": 8.034867503486751, + "grad_norm": 3.550299644470215, + "learning_rate": 9.9827249982725e-06, + "loss": 0.0514, + "step": 17283 + }, + { + "epoch": 8.035332403533241, + "grad_norm": 3.5638058185577393, + "learning_rate": 9.982723998272401e-06, + "loss": 0.0793, + "step": 17284 + }, + { + "epoch": 8.03579730357973, + "grad_norm": 1.5878922939300537, + "learning_rate": 9.9827229982723e-06, + "loss": 0.014, + "step": 17285 + }, + { + "epoch": 8.03626220362622, + "grad_norm": 3.0320074558258057, + "learning_rate": 9.9827219982722e-06, + "loss": 0.0347, + "step": 17286 + }, + { + "epoch": 8.03672710367271, + "grad_norm": 5.09207820892334, + "learning_rate": 9.982720998272101e-06, + "loss": 0.1418, + "step": 17287 + }, + { + "epoch": 8.0371920037192, + "grad_norm": 4.997923374176025, + "learning_rate": 9.982719998272e-06, + "loss": 0.1105, + "step": 17288 + }, + { + "epoch": 8.03765690376569, + "grad_norm": 5.415613174438477, + "learning_rate": 9.9827189982719e-06, + "loss": 0.0942, + "step": 17289 + }, + { + "epoch": 8.03812180381218, + "grad_norm": 7.28861141204834, + "learning_rate": 9.9827179982718e-06, + "loss": 0.1795, + "step": 17290 + }, + { + "epoch": 8.03858670385867, + "grad_norm": 7.744186878204346, + "learning_rate": 9.9827169982717e-06, + "loss": 0.3062, + "step": 17291 + }, + { + "epoch": 8.039051603905161, + "grad_norm": 6.470304489135742, + "learning_rate": 9.982715998271601e-06, + "loss": 0.1228, + "step": 17292 + }, + { + "epoch": 8.039516503951651, + "grad_norm": 6.442198753356934, + "learning_rate": 9.982714998271502e-06, + "loss": 0.1526, + "step": 17293 + }, + { + "epoch": 8.03998140399814, + "grad_norm": 2.478766679763794, + "learning_rate": 9.9827139982714e-06, + "loss": 0.039, + "step": 17294 + }, + { + "epoch": 8.04044630404463, + "grad_norm": 5.15245246887207, + "learning_rate": 9.982712998271301e-06, + "loss": 0.0828, + "step": 17295 + }, + { + "epoch": 8.04091120409112, + "grad_norm": 6.741162300109863, + "learning_rate": 9.9827119982712e-06, + "loss": 0.1808, + "step": 17296 + }, + { + "epoch": 8.04137610413761, + "grad_norm": 6.768281936645508, + "learning_rate": 9.982710998271101e-06, + "loss": 0.1674, + "step": 17297 + }, + { + "epoch": 8.0418410041841, + "grad_norm": 3.4766712188720703, + "learning_rate": 9.982709998271e-06, + "loss": 0.0566, + "step": 17298 + }, + { + "epoch": 8.04230590423059, + "grad_norm": 7.58317232131958, + "learning_rate": 9.9827089982709e-06, + "loss": 0.174, + "step": 17299 + }, + { + "epoch": 8.04277080427708, + "grad_norm": 0.5014830827713013, + "learning_rate": 9.982707998270801e-06, + "loss": 0.0059, + "step": 17300 + }, + { + "epoch": 8.043235704323571, + "grad_norm": 1.8994388580322266, + "learning_rate": 9.9827069982707e-06, + "loss": 0.0109, + "step": 17301 + }, + { + "epoch": 8.043700604370061, + "grad_norm": 2.9897546768188477, + "learning_rate": 9.982705998270601e-06, + "loss": 0.0451, + "step": 17302 + }, + { + "epoch": 8.04416550441655, + "grad_norm": 5.456529140472412, + "learning_rate": 9.9827049982705e-06, + "loss": 0.1461, + "step": 17303 + }, + { + "epoch": 8.04463040446304, + "grad_norm": 6.215733051300049, + "learning_rate": 9.9827039982704e-06, + "loss": 0.11, + "step": 17304 + }, + { + "epoch": 8.04509530450953, + "grad_norm": 3.71588134765625, + "learning_rate": 9.982702998270301e-06, + "loss": 0.0484, + "step": 17305 + }, + { + "epoch": 8.04556020455602, + "grad_norm": 3.36576509475708, + "learning_rate": 9.982701998270202e-06, + "loss": 0.0674, + "step": 17306 + }, + { + "epoch": 8.04602510460251, + "grad_norm": 5.145989418029785, + "learning_rate": 9.9827009982701e-06, + "loss": 0.1027, + "step": 17307 + }, + { + "epoch": 8.046490004649, + "grad_norm": 6.369197845458984, + "learning_rate": 9.98269999827e-06, + "loss": 0.3058, + "step": 17308 + }, + { + "epoch": 8.046954904695491, + "grad_norm": 3.858208417892456, + "learning_rate": 9.9826989982699e-06, + "loss": 0.0958, + "step": 17309 + }, + { + "epoch": 8.047419804741981, + "grad_norm": 4.1378960609436035, + "learning_rate": 9.982697998269801e-06, + "loss": 0.0708, + "step": 17310 + }, + { + "epoch": 8.047884704788471, + "grad_norm": 6.414917469024658, + "learning_rate": 9.9826969982697e-06, + "loss": 0.125, + "step": 17311 + }, + { + "epoch": 8.04834960483496, + "grad_norm": 7.501294136047363, + "learning_rate": 9.9826959982696e-06, + "loss": 0.1617, + "step": 17312 + }, + { + "epoch": 8.04881450488145, + "grad_norm": 5.981869697570801, + "learning_rate": 9.982694998269501e-06, + "loss": 0.214, + "step": 17313 + }, + { + "epoch": 8.04927940492794, + "grad_norm": 8.735809326171875, + "learning_rate": 9.9826939982694e-06, + "loss": 0.2417, + "step": 17314 + }, + { + "epoch": 8.04974430497443, + "grad_norm": 7.354002475738525, + "learning_rate": 9.982692998269301e-06, + "loss": 0.266, + "step": 17315 + }, + { + "epoch": 8.05020920502092, + "grad_norm": 7.696069240570068, + "learning_rate": 9.9826919982692e-06, + "loss": 0.2649, + "step": 17316 + }, + { + "epoch": 8.05067410506741, + "grad_norm": 6.029652118682861, + "learning_rate": 9.9826909982691e-06, + "loss": 0.1727, + "step": 17317 + }, + { + "epoch": 8.051139005113901, + "grad_norm": 3.6523356437683105, + "learning_rate": 9.982689998269001e-06, + "loss": 0.048, + "step": 17318 + }, + { + "epoch": 8.051603905160391, + "grad_norm": 8.011567115783691, + "learning_rate": 9.9826889982689e-06, + "loss": 0.2638, + "step": 17319 + }, + { + "epoch": 8.052068805206881, + "grad_norm": 3.9084227085113525, + "learning_rate": 9.982687998268801e-06, + "loss": 0.1001, + "step": 17320 + }, + { + "epoch": 8.05253370525337, + "grad_norm": 1.7715706825256348, + "learning_rate": 9.9826869982687e-06, + "loss": 0.0129, + "step": 17321 + }, + { + "epoch": 8.05299860529986, + "grad_norm": 4.668197154998779, + "learning_rate": 9.9826859982686e-06, + "loss": 0.1252, + "step": 17322 + }, + { + "epoch": 8.05346350534635, + "grad_norm": 3.800002098083496, + "learning_rate": 9.982684998268501e-06, + "loss": 0.0574, + "step": 17323 + }, + { + "epoch": 8.05392840539284, + "grad_norm": 3.3634605407714844, + "learning_rate": 9.982683998268402e-06, + "loss": 0.0486, + "step": 17324 + }, + { + "epoch": 8.05439330543933, + "grad_norm": 3.7713944911956787, + "learning_rate": 9.9826829982683e-06, + "loss": 0.0759, + "step": 17325 + }, + { + "epoch": 8.05485820548582, + "grad_norm": 2.7314183712005615, + "learning_rate": 9.982681998268201e-06, + "loss": 0.031, + "step": 17326 + }, + { + "epoch": 8.055323105532311, + "grad_norm": 6.823123455047607, + "learning_rate": 9.9826809982681e-06, + "loss": 0.0881, + "step": 17327 + }, + { + "epoch": 8.055788005578801, + "grad_norm": 2.930992841720581, + "learning_rate": 9.982679998268001e-06, + "loss": 0.0302, + "step": 17328 + }, + { + "epoch": 8.056252905625291, + "grad_norm": 3.408376932144165, + "learning_rate": 9.9826789982679e-06, + "loss": 0.0394, + "step": 17329 + }, + { + "epoch": 8.05671780567178, + "grad_norm": 8.117321014404297, + "learning_rate": 9.9826779982678e-06, + "loss": 0.1654, + "step": 17330 + }, + { + "epoch": 8.05718270571827, + "grad_norm": 6.030336380004883, + "learning_rate": 9.982676998267701e-06, + "loss": 0.1374, + "step": 17331 + }, + { + "epoch": 8.05764760576476, + "grad_norm": 3.4710731506347656, + "learning_rate": 9.9826759982676e-06, + "loss": 0.069, + "step": 17332 + }, + { + "epoch": 8.05811250581125, + "grad_norm": 4.021346569061279, + "learning_rate": 9.982674998267501e-06, + "loss": 0.0862, + "step": 17333 + }, + { + "epoch": 8.05857740585774, + "grad_norm": 6.134240627288818, + "learning_rate": 9.9826739982674e-06, + "loss": 0.0821, + "step": 17334 + }, + { + "epoch": 8.05904230590423, + "grad_norm": 6.382781505584717, + "learning_rate": 9.9826729982673e-06, + "loss": 0.1209, + "step": 17335 + }, + { + "epoch": 8.059507205950721, + "grad_norm": 4.371177673339844, + "learning_rate": 9.982671998267201e-06, + "loss": 0.079, + "step": 17336 + }, + { + "epoch": 8.059972105997211, + "grad_norm": 5.45206356048584, + "learning_rate": 9.982670998267102e-06, + "loss": 0.1204, + "step": 17337 + }, + { + "epoch": 8.060437006043701, + "grad_norm": 6.234546184539795, + "learning_rate": 9.982669998267e-06, + "loss": 0.1041, + "step": 17338 + }, + { + "epoch": 8.06090190609019, + "grad_norm": 3.366853713989258, + "learning_rate": 9.982668998266901e-06, + "loss": 0.0525, + "step": 17339 + }, + { + "epoch": 8.06136680613668, + "grad_norm": 6.082070350646973, + "learning_rate": 9.9826679982668e-06, + "loss": 0.0563, + "step": 17340 + }, + { + "epoch": 8.06183170618317, + "grad_norm": 4.305606842041016, + "learning_rate": 9.9826669982667e-06, + "loss": 0.1001, + "step": 17341 + }, + { + "epoch": 8.06229660622966, + "grad_norm": 3.9752843379974365, + "learning_rate": 9.9826659982666e-06, + "loss": 0.0649, + "step": 17342 + }, + { + "epoch": 8.06276150627615, + "grad_norm": 2.770524501800537, + "learning_rate": 9.9826649982665e-06, + "loss": 0.0484, + "step": 17343 + }, + { + "epoch": 8.06322640632264, + "grad_norm": 3.8078596591949463, + "learning_rate": 9.982663998266401e-06, + "loss": 0.0591, + "step": 17344 + }, + { + "epoch": 8.063691306369131, + "grad_norm": 3.5035979747772217, + "learning_rate": 9.9826629982663e-06, + "loss": 0.0307, + "step": 17345 + }, + { + "epoch": 8.064156206415621, + "grad_norm": 8.788095474243164, + "learning_rate": 9.982661998266201e-06, + "loss": 0.1121, + "step": 17346 + }, + { + "epoch": 8.064621106462111, + "grad_norm": 10.413520812988281, + "learning_rate": 9.9826609982661e-06, + "loss": 0.1318, + "step": 17347 + }, + { + "epoch": 8.0650860065086, + "grad_norm": 5.265239715576172, + "learning_rate": 9.982659998266e-06, + "loss": 0.2026, + "step": 17348 + }, + { + "epoch": 8.06555090655509, + "grad_norm": 4.168956279754639, + "learning_rate": 9.982658998265901e-06, + "loss": 0.0722, + "step": 17349 + }, + { + "epoch": 8.06601580660158, + "grad_norm": 8.230481147766113, + "learning_rate": 9.982657998265802e-06, + "loss": 0.0758, + "step": 17350 + }, + { + "epoch": 8.06648070664807, + "grad_norm": 7.275238513946533, + "learning_rate": 9.982656998265701e-06, + "loss": 0.2478, + "step": 17351 + }, + { + "epoch": 8.06694560669456, + "grad_norm": 5.447935104370117, + "learning_rate": 9.9826559982656e-06, + "loss": 0.1523, + "step": 17352 + }, + { + "epoch": 8.067410506741052, + "grad_norm": 6.705043792724609, + "learning_rate": 9.9826549982655e-06, + "loss": 0.1646, + "step": 17353 + }, + { + "epoch": 8.067875406787541, + "grad_norm": 5.666625022888184, + "learning_rate": 9.9826539982654e-06, + "loss": 0.099, + "step": 17354 + }, + { + "epoch": 8.068340306834031, + "grad_norm": 4.006077766418457, + "learning_rate": 9.9826529982653e-06, + "loss": 0.071, + "step": 17355 + }, + { + "epoch": 8.068805206880521, + "grad_norm": 3.0783581733703613, + "learning_rate": 9.9826519982652e-06, + "loss": 0.0407, + "step": 17356 + }, + { + "epoch": 8.06927010692701, + "grad_norm": 6.334716796875, + "learning_rate": 9.982650998265101e-06, + "loss": 0.2684, + "step": 17357 + }, + { + "epoch": 8.0697350069735, + "grad_norm": 1.784978985786438, + "learning_rate": 9.982649998265e-06, + "loss": 0.0257, + "step": 17358 + }, + { + "epoch": 8.07019990701999, + "grad_norm": 4.043915271759033, + "learning_rate": 9.982648998264901e-06, + "loss": 0.0601, + "step": 17359 + }, + { + "epoch": 8.07066480706648, + "grad_norm": 3.1650214195251465, + "learning_rate": 9.9826479982648e-06, + "loss": 0.0409, + "step": 17360 + }, + { + "epoch": 8.07112970711297, + "grad_norm": 5.245437145233154, + "learning_rate": 9.9826469982647e-06, + "loss": 0.1065, + "step": 17361 + }, + { + "epoch": 8.071594607159462, + "grad_norm": 3.4644854068756104, + "learning_rate": 9.982645998264601e-06, + "loss": 0.074, + "step": 17362 + }, + { + "epoch": 8.072059507205951, + "grad_norm": 3.996966600418091, + "learning_rate": 9.9826449982645e-06, + "loss": 0.0891, + "step": 17363 + }, + { + "epoch": 8.072524407252441, + "grad_norm": 1.4962050914764404, + "learning_rate": 9.982643998264401e-06, + "loss": 0.0192, + "step": 17364 + }, + { + "epoch": 8.072989307298931, + "grad_norm": 10.46806526184082, + "learning_rate": 9.9826429982643e-06, + "loss": 0.1632, + "step": 17365 + }, + { + "epoch": 8.07345420734542, + "grad_norm": 3.330873966217041, + "learning_rate": 9.9826419982642e-06, + "loss": 0.0371, + "step": 17366 + }, + { + "epoch": 8.07391910739191, + "grad_norm": 12.968681335449219, + "learning_rate": 9.9826409982641e-06, + "loss": 0.1255, + "step": 17367 + }, + { + "epoch": 8.0743840074384, + "grad_norm": 2.0583643913269043, + "learning_rate": 9.982639998264e-06, + "loss": 0.0287, + "step": 17368 + }, + { + "epoch": 8.07484890748489, + "grad_norm": 1.5699007511138916, + "learning_rate": 9.982638998263901e-06, + "loss": 0.017, + "step": 17369 + }, + { + "epoch": 8.07531380753138, + "grad_norm": 5.187437534332275, + "learning_rate": 9.982637998263802e-06, + "loss": 0.0822, + "step": 17370 + }, + { + "epoch": 8.075778707577872, + "grad_norm": 6.345186710357666, + "learning_rate": 9.9826369982637e-06, + "loss": 0.2456, + "step": 17371 + }, + { + "epoch": 8.076243607624361, + "grad_norm": 6.642428874969482, + "learning_rate": 9.982635998263601e-06, + "loss": 0.1643, + "step": 17372 + }, + { + "epoch": 8.076708507670851, + "grad_norm": 5.258322715759277, + "learning_rate": 9.9826349982635e-06, + "loss": 0.0537, + "step": 17373 + }, + { + "epoch": 8.077173407717341, + "grad_norm": 2.523890495300293, + "learning_rate": 9.9826339982634e-06, + "loss": 0.0477, + "step": 17374 + }, + { + "epoch": 8.07763830776383, + "grad_norm": 10.905364036560059, + "learning_rate": 9.982632998263301e-06, + "loss": 0.172, + "step": 17375 + }, + { + "epoch": 8.07810320781032, + "grad_norm": 8.376220703125, + "learning_rate": 9.9826319982632e-06, + "loss": 0.2229, + "step": 17376 + }, + { + "epoch": 8.07856810785681, + "grad_norm": 5.353899955749512, + "learning_rate": 9.982630998263101e-06, + "loss": 0.1162, + "step": 17377 + }, + { + "epoch": 8.0790330079033, + "grad_norm": 3.490412473678589, + "learning_rate": 9.982629998263e-06, + "loss": 0.0518, + "step": 17378 + }, + { + "epoch": 8.07949790794979, + "grad_norm": 4.220127582550049, + "learning_rate": 9.9826289982629e-06, + "loss": 0.0807, + "step": 17379 + }, + { + "epoch": 8.079962807996282, + "grad_norm": 5.22272253036499, + "learning_rate": 9.9826279982628e-06, + "loss": 0.0956, + "step": 17380 + }, + { + "epoch": 8.080427708042771, + "grad_norm": 4.1869707107543945, + "learning_rate": 9.9826269982627e-06, + "loss": 0.1356, + "step": 17381 + }, + { + "epoch": 8.080892608089261, + "grad_norm": 2.8321595191955566, + "learning_rate": 9.982625998262601e-06, + "loss": 0.0281, + "step": 17382 + }, + { + "epoch": 8.081357508135751, + "grad_norm": 7.291059494018555, + "learning_rate": 9.9826249982625e-06, + "loss": 0.1566, + "step": 17383 + }, + { + "epoch": 8.08182240818224, + "grad_norm": 3.998401641845703, + "learning_rate": 9.9826239982624e-06, + "loss": 0.0741, + "step": 17384 + }, + { + "epoch": 8.08228730822873, + "grad_norm": 7.939416408538818, + "learning_rate": 9.9826229982623e-06, + "loss": 0.1789, + "step": 17385 + }, + { + "epoch": 8.08275220827522, + "grad_norm": 3.5208590030670166, + "learning_rate": 9.9826219982622e-06, + "loss": 0.0401, + "step": 17386 + }, + { + "epoch": 8.08321710832171, + "grad_norm": 7.34268856048584, + "learning_rate": 9.982620998262101e-06, + "loss": 0.1631, + "step": 17387 + }, + { + "epoch": 8.0836820083682, + "grad_norm": 2.538182497024536, + "learning_rate": 9.982619998262002e-06, + "loss": 0.0375, + "step": 17388 + }, + { + "epoch": 8.084146908414692, + "grad_norm": 4.441656589508057, + "learning_rate": 9.9826189982619e-06, + "loss": 0.0964, + "step": 17389 + }, + { + "epoch": 8.084611808461181, + "grad_norm": 5.557715892791748, + "learning_rate": 9.982617998261801e-06, + "loss": 0.122, + "step": 17390 + }, + { + "epoch": 8.085076708507671, + "grad_norm": 4.57404088973999, + "learning_rate": 9.9826169982617e-06, + "loss": 0.0919, + "step": 17391 + }, + { + "epoch": 8.085541608554161, + "grad_norm": 4.859737873077393, + "learning_rate": 9.9826159982616e-06, + "loss": 0.089, + "step": 17392 + }, + { + "epoch": 8.08600650860065, + "grad_norm": 6.79885721206665, + "learning_rate": 9.982614998261501e-06, + "loss": 0.1717, + "step": 17393 + }, + { + "epoch": 8.08647140864714, + "grad_norm": 4.852291107177734, + "learning_rate": 9.9826139982614e-06, + "loss": 0.126, + "step": 17394 + }, + { + "epoch": 8.08693630869363, + "grad_norm": 4.316200256347656, + "learning_rate": 9.982612998261301e-06, + "loss": 0.0623, + "step": 17395 + }, + { + "epoch": 8.08740120874012, + "grad_norm": 4.333755970001221, + "learning_rate": 9.9826119982612e-06, + "loss": 0.0839, + "step": 17396 + }, + { + "epoch": 8.087866108786612, + "grad_norm": 2.021785259246826, + "learning_rate": 9.9826109982611e-06, + "loss": 0.0233, + "step": 17397 + }, + { + "epoch": 8.088331008833102, + "grad_norm": 2.1240646839141846, + "learning_rate": 9.982609998261e-06, + "loss": 0.0289, + "step": 17398 + }, + { + "epoch": 8.088795908879591, + "grad_norm": 4.2625041007995605, + "learning_rate": 9.9826089982609e-06, + "loss": 0.0945, + "step": 17399 + }, + { + "epoch": 8.089260808926081, + "grad_norm": 1.3833461999893188, + "learning_rate": 9.982607998260801e-06, + "loss": 0.0153, + "step": 17400 + }, + { + "epoch": 8.089725708972571, + "grad_norm": 2.7339131832122803, + "learning_rate": 9.982606998260702e-06, + "loss": 0.0457, + "step": 17401 + }, + { + "epoch": 8.09019060901906, + "grad_norm": 4.270131587982178, + "learning_rate": 9.9826059982606e-06, + "loss": 0.1081, + "step": 17402 + }, + { + "epoch": 8.09065550906555, + "grad_norm": 5.615660190582275, + "learning_rate": 9.982604998260501e-06, + "loss": 0.1705, + "step": 17403 + }, + { + "epoch": 8.09112040911204, + "grad_norm": 5.345175266265869, + "learning_rate": 9.9826039982604e-06, + "loss": 0.1144, + "step": 17404 + }, + { + "epoch": 8.09158530915853, + "grad_norm": 2.268524169921875, + "learning_rate": 9.982602998260301e-06, + "loss": 0.029, + "step": 17405 + }, + { + "epoch": 8.092050209205022, + "grad_norm": 4.915775299072266, + "learning_rate": 9.982601998260202e-06, + "loss": 0.1063, + "step": 17406 + }, + { + "epoch": 8.092515109251512, + "grad_norm": 3.3343005180358887, + "learning_rate": 9.9826009982601e-06, + "loss": 0.0492, + "step": 17407 + }, + { + "epoch": 8.092980009298001, + "grad_norm": 4.619029998779297, + "learning_rate": 9.982599998260001e-06, + "loss": 0.0858, + "step": 17408 + }, + { + "epoch": 8.093444909344491, + "grad_norm": 4.461773872375488, + "learning_rate": 9.9825989982599e-06, + "loss": 0.0879, + "step": 17409 + }, + { + "epoch": 8.093909809390981, + "grad_norm": 2.4730045795440674, + "learning_rate": 9.9825979982598e-06, + "loss": 0.033, + "step": 17410 + }, + { + "epoch": 8.09437470943747, + "grad_norm": 1.0963689088821411, + "learning_rate": 9.9825969982597e-06, + "loss": 0.0085, + "step": 17411 + }, + { + "epoch": 8.09483960948396, + "grad_norm": 3.1590871810913086, + "learning_rate": 9.9825959982596e-06, + "loss": 0.0267, + "step": 17412 + }, + { + "epoch": 8.09530450953045, + "grad_norm": 3.0961837768554688, + "learning_rate": 9.982594998259501e-06, + "loss": 0.0538, + "step": 17413 + }, + { + "epoch": 8.09576940957694, + "grad_norm": 5.933335781097412, + "learning_rate": 9.982593998259402e-06, + "loss": 0.0987, + "step": 17414 + }, + { + "epoch": 8.096234309623432, + "grad_norm": 2.4486474990844727, + "learning_rate": 9.9825929982593e-06, + "loss": 0.0189, + "step": 17415 + }, + { + "epoch": 8.096699209669922, + "grad_norm": 6.31741189956665, + "learning_rate": 9.9825919982592e-06, + "loss": 0.1546, + "step": 17416 + }, + { + "epoch": 8.097164109716411, + "grad_norm": 6.132631301879883, + "learning_rate": 9.9825909982591e-06, + "loss": 0.1431, + "step": 17417 + }, + { + "epoch": 8.097629009762901, + "grad_norm": 5.82487154006958, + "learning_rate": 9.982589998259001e-06, + "loss": 0.1212, + "step": 17418 + }, + { + "epoch": 8.098093909809391, + "grad_norm": 7.821682929992676, + "learning_rate": 9.982588998258902e-06, + "loss": 0.268, + "step": 17419 + }, + { + "epoch": 8.09855880985588, + "grad_norm": 4.58340311050415, + "learning_rate": 9.9825879982588e-06, + "loss": 0.1001, + "step": 17420 + }, + { + "epoch": 8.09902370990237, + "grad_norm": 4.818793773651123, + "learning_rate": 9.982586998258701e-06, + "loss": 0.1591, + "step": 17421 + }, + { + "epoch": 8.09948860994886, + "grad_norm": 7.247520446777344, + "learning_rate": 9.9825859982586e-06, + "loss": 0.1575, + "step": 17422 + }, + { + "epoch": 8.09995350999535, + "grad_norm": 2.967663526535034, + "learning_rate": 9.9825849982585e-06, + "loss": 0.0463, + "step": 17423 + }, + { + "epoch": 8.100418410041842, + "grad_norm": 3.738433361053467, + "learning_rate": 9.9825839982584e-06, + "loss": 0.0669, + "step": 17424 + }, + { + "epoch": 8.100883310088332, + "grad_norm": 4.666871547698975, + "learning_rate": 9.9825829982583e-06, + "loss": 0.0616, + "step": 17425 + }, + { + "epoch": 8.101348210134821, + "grad_norm": 4.3806471824646, + "learning_rate": 9.982581998258201e-06, + "loss": 0.0926, + "step": 17426 + }, + { + "epoch": 8.101813110181311, + "grad_norm": 4.474429607391357, + "learning_rate": 9.9825809982581e-06, + "loss": 0.0652, + "step": 17427 + }, + { + "epoch": 8.102278010227801, + "grad_norm": 6.641543865203857, + "learning_rate": 9.982579998258e-06, + "loss": 0.1177, + "step": 17428 + }, + { + "epoch": 8.10274291027429, + "grad_norm": 5.784114837646484, + "learning_rate": 9.9825789982579e-06, + "loss": 0.095, + "step": 17429 + }, + { + "epoch": 8.10320781032078, + "grad_norm": 6.59024715423584, + "learning_rate": 9.9825779982578e-06, + "loss": 0.3023, + "step": 17430 + }, + { + "epoch": 8.10367271036727, + "grad_norm": 2.692060947418213, + "learning_rate": 9.982576998257701e-06, + "loss": 0.0459, + "step": 17431 + }, + { + "epoch": 8.104137610413762, + "grad_norm": 5.440814018249512, + "learning_rate": 9.982575998257602e-06, + "loss": 0.0622, + "step": 17432 + }, + { + "epoch": 8.104602510460252, + "grad_norm": 7.400440692901611, + "learning_rate": 9.9825749982575e-06, + "loss": 0.1242, + "step": 17433 + }, + { + "epoch": 8.105067410506742, + "grad_norm": 5.646861553192139, + "learning_rate": 9.982573998257401e-06, + "loss": 0.1258, + "step": 17434 + }, + { + "epoch": 8.105532310553231, + "grad_norm": 4.232062339782715, + "learning_rate": 9.9825729982573e-06, + "loss": 0.0341, + "step": 17435 + }, + { + "epoch": 8.105997210599721, + "grad_norm": 5.6338653564453125, + "learning_rate": 9.982571998257201e-06, + "loss": 0.1501, + "step": 17436 + }, + { + "epoch": 8.106462110646211, + "grad_norm": 4.0224714279174805, + "learning_rate": 9.9825709982571e-06, + "loss": 0.0758, + "step": 17437 + }, + { + "epoch": 8.1069270106927, + "grad_norm": 4.459452152252197, + "learning_rate": 9.982569998257e-06, + "loss": 0.0663, + "step": 17438 + }, + { + "epoch": 8.10739191073919, + "grad_norm": 4.267306804656982, + "learning_rate": 9.982568998256901e-06, + "loss": 0.0794, + "step": 17439 + }, + { + "epoch": 8.10785681078568, + "grad_norm": 5.711405277252197, + "learning_rate": 9.9825679982568e-06, + "loss": 0.1683, + "step": 17440 + }, + { + "epoch": 8.108321710832172, + "grad_norm": 4.097346305847168, + "learning_rate": 9.9825669982567e-06, + "loss": 0.047, + "step": 17441 + }, + { + "epoch": 8.108786610878662, + "grad_norm": 5.304488182067871, + "learning_rate": 9.9825659982566e-06, + "loss": 0.1189, + "step": 17442 + }, + { + "epoch": 8.109251510925152, + "grad_norm": 4.067605018615723, + "learning_rate": 9.9825649982565e-06, + "loss": 0.08, + "step": 17443 + }, + { + "epoch": 8.109716410971641, + "grad_norm": 6.346184730529785, + "learning_rate": 9.982563998256401e-06, + "loss": 0.1896, + "step": 17444 + }, + { + "epoch": 8.110181311018131, + "grad_norm": 5.068058490753174, + "learning_rate": 9.982562998256302e-06, + "loss": 0.1085, + "step": 17445 + }, + { + "epoch": 8.110646211064621, + "grad_norm": 5.509054660797119, + "learning_rate": 9.9825619982562e-06, + "loss": 0.1719, + "step": 17446 + }, + { + "epoch": 8.11111111111111, + "grad_norm": 5.316082954406738, + "learning_rate": 9.982560998256101e-06, + "loss": 0.1111, + "step": 17447 + }, + { + "epoch": 8.1115760111576, + "grad_norm": 2.9632632732391357, + "learning_rate": 9.982559998256e-06, + "loss": 0.0261, + "step": 17448 + }, + { + "epoch": 8.11204091120409, + "grad_norm": 4.9163737297058105, + "learning_rate": 9.9825589982559e-06, + "loss": 0.0981, + "step": 17449 + }, + { + "epoch": 8.112505811250582, + "grad_norm": 4.667916774749756, + "learning_rate": 9.9825579982558e-06, + "loss": 0.0526, + "step": 17450 + }, + { + "epoch": 8.112970711297072, + "grad_norm": 4.716217041015625, + "learning_rate": 9.9825569982557e-06, + "loss": 0.1003, + "step": 17451 + }, + { + "epoch": 8.113435611343562, + "grad_norm": 5.688877105712891, + "learning_rate": 9.982555998255601e-06, + "loss": 0.1791, + "step": 17452 + }, + { + "epoch": 8.113900511390051, + "grad_norm": 2.8463711738586426, + "learning_rate": 9.9825549982555e-06, + "loss": 0.0414, + "step": 17453 + }, + { + "epoch": 8.114365411436541, + "grad_norm": 4.337066173553467, + "learning_rate": 9.982553998255401e-06, + "loss": 0.0905, + "step": 17454 + }, + { + "epoch": 8.114830311483031, + "grad_norm": 4.348216533660889, + "learning_rate": 9.9825529982553e-06, + "loss": 0.0963, + "step": 17455 + }, + { + "epoch": 8.11529521152952, + "grad_norm": 6.501049041748047, + "learning_rate": 9.9825519982552e-06, + "loss": 0.0288, + "step": 17456 + }, + { + "epoch": 8.11576011157601, + "grad_norm": 1.649725079536438, + "learning_rate": 9.982550998255101e-06, + "loss": 0.0161, + "step": 17457 + }, + { + "epoch": 8.1162250116225, + "grad_norm": 4.134016990661621, + "learning_rate": 9.982549998255002e-06, + "loss": 0.0605, + "step": 17458 + }, + { + "epoch": 8.116689911668992, + "grad_norm": 6.766760349273682, + "learning_rate": 9.9825489982549e-06, + "loss": 0.184, + "step": 17459 + }, + { + "epoch": 8.117154811715482, + "grad_norm": 3.528815984725952, + "learning_rate": 9.9825479982548e-06, + "loss": 0.1265, + "step": 17460 + }, + { + "epoch": 8.117619711761972, + "grad_norm": 3.876443386077881, + "learning_rate": 9.9825469982547e-06, + "loss": 0.0543, + "step": 17461 + }, + { + "epoch": 8.118084611808461, + "grad_norm": 3.3466668128967285, + "learning_rate": 9.982545998254601e-06, + "loss": 0.0779, + "step": 17462 + }, + { + "epoch": 8.118549511854951, + "grad_norm": 5.69216775894165, + "learning_rate": 9.9825449982545e-06, + "loss": 0.1311, + "step": 17463 + }, + { + "epoch": 8.119014411901441, + "grad_norm": 2.6879677772521973, + "learning_rate": 9.9825439982544e-06, + "loss": 0.0437, + "step": 17464 + }, + { + "epoch": 8.11947931194793, + "grad_norm": 2.015501022338867, + "learning_rate": 9.982542998254301e-06, + "loss": 0.0253, + "step": 17465 + }, + { + "epoch": 8.11994421199442, + "grad_norm": 5.43511438369751, + "learning_rate": 9.9825419982542e-06, + "loss": 0.0991, + "step": 17466 + }, + { + "epoch": 8.12040911204091, + "grad_norm": 9.768088340759277, + "learning_rate": 9.982540998254101e-06, + "loss": 0.3293, + "step": 17467 + }, + { + "epoch": 8.120874012087402, + "grad_norm": 2.6280438899993896, + "learning_rate": 9.982539998254e-06, + "loss": 0.0351, + "step": 17468 + }, + { + "epoch": 8.121338912133892, + "grad_norm": 10.470372200012207, + "learning_rate": 9.9825389982539e-06, + "loss": 0.1808, + "step": 17469 + }, + { + "epoch": 8.121803812180381, + "grad_norm": 6.603893756866455, + "learning_rate": 9.982537998253801e-06, + "loss": 0.1976, + "step": 17470 + }, + { + "epoch": 8.122268712226871, + "grad_norm": 5.217596054077148, + "learning_rate": 9.9825369982537e-06, + "loss": 0.0523, + "step": 17471 + }, + { + "epoch": 8.122733612273361, + "grad_norm": 5.5923004150390625, + "learning_rate": 9.982535998253601e-06, + "loss": 0.0795, + "step": 17472 + }, + { + "epoch": 8.123198512319851, + "grad_norm": 4.229358196258545, + "learning_rate": 9.9825349982535e-06, + "loss": 0.1117, + "step": 17473 + }, + { + "epoch": 8.12366341236634, + "grad_norm": 3.741001605987549, + "learning_rate": 9.9825339982534e-06, + "loss": 0.0798, + "step": 17474 + }, + { + "epoch": 8.12412831241283, + "grad_norm": 4.753225326538086, + "learning_rate": 9.982532998253301e-06, + "loss": 0.0805, + "step": 17475 + }, + { + "epoch": 8.124593212459322, + "grad_norm": 1.8924992084503174, + "learning_rate": 9.9825319982532e-06, + "loss": 0.0252, + "step": 17476 + }, + { + "epoch": 8.125058112505812, + "grad_norm": 5.564459323883057, + "learning_rate": 9.9825309982531e-06, + "loss": 0.129, + "step": 17477 + }, + { + "epoch": 8.125523012552302, + "grad_norm": 3.2372593879699707, + "learning_rate": 9.982529998253001e-06, + "loss": 0.0335, + "step": 17478 + }, + { + "epoch": 8.125987912598791, + "grad_norm": 6.580048084259033, + "learning_rate": 9.9825289982529e-06, + "loss": 0.0512, + "step": 17479 + }, + { + "epoch": 8.126452812645281, + "grad_norm": 2.6483123302459717, + "learning_rate": 9.982527998252801e-06, + "loss": 0.0528, + "step": 17480 + }, + { + "epoch": 8.126917712691771, + "grad_norm": 5.25838565826416, + "learning_rate": 9.9825269982527e-06, + "loss": 0.1414, + "step": 17481 + }, + { + "epoch": 8.127382612738261, + "grad_norm": 5.289873123168945, + "learning_rate": 9.9825259982526e-06, + "loss": 0.17, + "step": 17482 + }, + { + "epoch": 8.12784751278475, + "grad_norm": 1.447879433631897, + "learning_rate": 9.982524998252501e-06, + "loss": 0.0172, + "step": 17483 + }, + { + "epoch": 8.12831241283124, + "grad_norm": 3.513963460922241, + "learning_rate": 9.9825239982524e-06, + "loss": 0.0594, + "step": 17484 + }, + { + "epoch": 8.128777312877732, + "grad_norm": 3.5936777591705322, + "learning_rate": 9.982522998252301e-06, + "loss": 0.0274, + "step": 17485 + }, + { + "epoch": 8.129242212924222, + "grad_norm": 4.54525089263916, + "learning_rate": 9.9825219982522e-06, + "loss": 0.0494, + "step": 17486 + }, + { + "epoch": 8.129707112970712, + "grad_norm": 4.3359551429748535, + "learning_rate": 9.9825209982521e-06, + "loss": 0.1323, + "step": 17487 + }, + { + "epoch": 8.130172013017201, + "grad_norm": 7.190768718719482, + "learning_rate": 9.982519998252001e-06, + "loss": 0.1689, + "step": 17488 + }, + { + "epoch": 8.130636913063691, + "grad_norm": 9.299630165100098, + "learning_rate": 9.982518998251902e-06, + "loss": 0.034, + "step": 17489 + }, + { + "epoch": 8.131101813110181, + "grad_norm": 6.913590431213379, + "learning_rate": 9.982517998251801e-06, + "loss": 0.0981, + "step": 17490 + }, + { + "epoch": 8.131566713156671, + "grad_norm": 4.306560516357422, + "learning_rate": 9.982516998251702e-06, + "loss": 0.1071, + "step": 17491 + }, + { + "epoch": 8.13203161320316, + "grad_norm": 2.874582290649414, + "learning_rate": 9.9825159982516e-06, + "loss": 0.031, + "step": 17492 + }, + { + "epoch": 8.13249651324965, + "grad_norm": 4.26581335067749, + "learning_rate": 9.9825149982515e-06, + "loss": 0.0628, + "step": 17493 + }, + { + "epoch": 8.132961413296142, + "grad_norm": 5.794704914093018, + "learning_rate": 9.9825139982514e-06, + "loss": 0.1527, + "step": 17494 + }, + { + "epoch": 8.133426313342632, + "grad_norm": 15.060302734375, + "learning_rate": 9.9825129982513e-06, + "loss": 0.1779, + "step": 17495 + }, + { + "epoch": 8.133891213389122, + "grad_norm": 2.1085972785949707, + "learning_rate": 9.982511998251201e-06, + "loss": 0.0242, + "step": 17496 + }, + { + "epoch": 8.134356113435611, + "grad_norm": 0.9495651125907898, + "learning_rate": 9.9825109982511e-06, + "loss": 0.01, + "step": 17497 + }, + { + "epoch": 8.134821013482101, + "grad_norm": 3.6943600177764893, + "learning_rate": 9.982509998251001e-06, + "loss": 0.1063, + "step": 17498 + }, + { + "epoch": 8.135285913528591, + "grad_norm": 1.0603582859039307, + "learning_rate": 9.9825089982509e-06, + "loss": 0.0143, + "step": 17499 + }, + { + "epoch": 8.135750813575081, + "grad_norm": 5.086562633514404, + "learning_rate": 9.9825079982508e-06, + "loss": 0.114, + "step": 17500 + }, + { + "epoch": 8.13621571362157, + "grad_norm": 6.402583122253418, + "learning_rate": 9.982506998250701e-06, + "loss": 0.0624, + "step": 17501 + }, + { + "epoch": 8.13668061366806, + "grad_norm": 4.241105079650879, + "learning_rate": 9.982505998250602e-06, + "loss": 0.0677, + "step": 17502 + }, + { + "epoch": 8.137145513714552, + "grad_norm": 4.712679862976074, + "learning_rate": 9.982504998250501e-06, + "loss": 0.0952, + "step": 17503 + }, + { + "epoch": 8.137610413761042, + "grad_norm": 4.153275489807129, + "learning_rate": 9.9825039982504e-06, + "loss": 0.0345, + "step": 17504 + }, + { + "epoch": 8.138075313807532, + "grad_norm": 5.448049545288086, + "learning_rate": 9.9825029982503e-06, + "loss": 0.0695, + "step": 17505 + }, + { + "epoch": 8.138540213854021, + "grad_norm": 4.802433490753174, + "learning_rate": 9.9825019982502e-06, + "loss": 0.1098, + "step": 17506 + }, + { + "epoch": 8.139005113900511, + "grad_norm": 6.137930393218994, + "learning_rate": 9.9825009982501e-06, + "loss": 0.1642, + "step": 17507 + }, + { + "epoch": 8.139470013947001, + "grad_norm": 4.480652809143066, + "learning_rate": 9.98249999825e-06, + "loss": 0.0977, + "step": 17508 + }, + { + "epoch": 8.139934913993491, + "grad_norm": 3.553954839706421, + "learning_rate": 9.982498998249902e-06, + "loss": 0.0651, + "step": 17509 + }, + { + "epoch": 8.14039981403998, + "grad_norm": 7.940351486206055, + "learning_rate": 9.9824979982498e-06, + "loss": 0.1328, + "step": 17510 + }, + { + "epoch": 8.140864714086472, + "grad_norm": 7.3293938636779785, + "learning_rate": 9.982496998249701e-06, + "loss": 0.1851, + "step": 17511 + }, + { + "epoch": 8.141329614132962, + "grad_norm": 3.258396625518799, + "learning_rate": 9.9824959982496e-06, + "loss": 0.0506, + "step": 17512 + }, + { + "epoch": 8.141794514179452, + "grad_norm": 4.8124823570251465, + "learning_rate": 9.9824949982495e-06, + "loss": 0.0304, + "step": 17513 + }, + { + "epoch": 8.142259414225942, + "grad_norm": 1.2728052139282227, + "learning_rate": 9.982493998249401e-06, + "loss": 0.0151, + "step": 17514 + }, + { + "epoch": 8.142724314272431, + "grad_norm": 3.9539473056793213, + "learning_rate": 9.9824929982493e-06, + "loss": 0.0309, + "step": 17515 + }, + { + "epoch": 8.143189214318921, + "grad_norm": 8.361883163452148, + "learning_rate": 9.982491998249201e-06, + "loss": 0.1056, + "step": 17516 + }, + { + "epoch": 8.143654114365411, + "grad_norm": 6.543632984161377, + "learning_rate": 9.9824909982491e-06, + "loss": 0.1587, + "step": 17517 + }, + { + "epoch": 8.144119014411901, + "grad_norm": 8.38416576385498, + "learning_rate": 9.982489998249e-06, + "loss": 0.2431, + "step": 17518 + }, + { + "epoch": 8.14458391445839, + "grad_norm": 5.284557342529297, + "learning_rate": 9.9824889982489e-06, + "loss": 0.139, + "step": 17519 + }, + { + "epoch": 8.145048814504882, + "grad_norm": 4.008060932159424, + "learning_rate": 9.9824879982488e-06, + "loss": 0.0531, + "step": 17520 + }, + { + "epoch": 8.145513714551372, + "grad_norm": 3.399735927581787, + "learning_rate": 9.982486998248701e-06, + "loss": 0.0692, + "step": 17521 + }, + { + "epoch": 8.145978614597862, + "grad_norm": 3.733438491821289, + "learning_rate": 9.982485998248602e-06, + "loss": 0.0635, + "step": 17522 + }, + { + "epoch": 8.146443514644352, + "grad_norm": 3.1135401725769043, + "learning_rate": 9.9824849982485e-06, + "loss": 0.0629, + "step": 17523 + }, + { + "epoch": 8.146908414690841, + "grad_norm": 2.0009360313415527, + "learning_rate": 9.9824839982484e-06, + "loss": 0.024, + "step": 17524 + }, + { + "epoch": 8.147373314737331, + "grad_norm": 5.305828094482422, + "learning_rate": 9.9824829982483e-06, + "loss": 0.1335, + "step": 17525 + }, + { + "epoch": 8.147838214783821, + "grad_norm": 3.3889896869659424, + "learning_rate": 9.9824819982482e-06, + "loss": 0.0419, + "step": 17526 + }, + { + "epoch": 8.148303114830311, + "grad_norm": 5.774170875549316, + "learning_rate": 9.982480998248101e-06, + "loss": 0.0654, + "step": 17527 + }, + { + "epoch": 8.1487680148768, + "grad_norm": 5.410150051116943, + "learning_rate": 9.982479998248e-06, + "loss": 0.1046, + "step": 17528 + }, + { + "epoch": 8.149232914923292, + "grad_norm": 5.131313800811768, + "learning_rate": 9.982478998247901e-06, + "loss": 0.109, + "step": 17529 + }, + { + "epoch": 8.149697814969782, + "grad_norm": 4.308051109313965, + "learning_rate": 9.9824779982478e-06, + "loss": 0.0516, + "step": 17530 + }, + { + "epoch": 8.150162715016272, + "grad_norm": 3.1311821937561035, + "learning_rate": 9.9824769982477e-06, + "loss": 0.0419, + "step": 17531 + }, + { + "epoch": 8.150627615062762, + "grad_norm": 4.39079475402832, + "learning_rate": 9.9824759982476e-06, + "loss": 0.0684, + "step": 17532 + }, + { + "epoch": 8.151092515109251, + "grad_norm": 2.7463066577911377, + "learning_rate": 9.9824749982475e-06, + "loss": 0.0547, + "step": 17533 + }, + { + "epoch": 8.151557415155741, + "grad_norm": 3.6149487495422363, + "learning_rate": 9.982473998247401e-06, + "loss": 0.0451, + "step": 17534 + }, + { + "epoch": 8.152022315202231, + "grad_norm": 4.949427604675293, + "learning_rate": 9.9824729982473e-06, + "loss": 0.1013, + "step": 17535 + }, + { + "epoch": 8.15248721524872, + "grad_norm": 1.8631796836853027, + "learning_rate": 9.9824719982472e-06, + "loss": 0.0278, + "step": 17536 + }, + { + "epoch": 8.15295211529521, + "grad_norm": 3.927741527557373, + "learning_rate": 9.9824709982471e-06, + "loss": 0.0868, + "step": 17537 + }, + { + "epoch": 8.153417015341702, + "grad_norm": 6.896612167358398, + "learning_rate": 9.982469998247e-06, + "loss": 0.2199, + "step": 17538 + }, + { + "epoch": 8.153881915388192, + "grad_norm": 1.1830849647521973, + "learning_rate": 9.982468998246901e-06, + "loss": 0.0116, + "step": 17539 + }, + { + "epoch": 8.154346815434682, + "grad_norm": 2.8086423873901367, + "learning_rate": 9.982467998246802e-06, + "loss": 0.0647, + "step": 17540 + }, + { + "epoch": 8.154811715481172, + "grad_norm": 3.8487188816070557, + "learning_rate": 9.9824669982467e-06, + "loss": 0.0643, + "step": 17541 + }, + { + "epoch": 8.155276615527661, + "grad_norm": 7.0121307373046875, + "learning_rate": 9.982465998246601e-06, + "loss": 0.0771, + "step": 17542 + }, + { + "epoch": 8.155741515574151, + "grad_norm": 6.209547996520996, + "learning_rate": 9.9824649982465e-06, + "loss": 0.1148, + "step": 17543 + }, + { + "epoch": 8.156206415620641, + "grad_norm": 6.190002918243408, + "learning_rate": 9.9824639982464e-06, + "loss": 0.1155, + "step": 17544 + }, + { + "epoch": 8.15667131566713, + "grad_norm": 2.8192079067230225, + "learning_rate": 9.9824629982463e-06, + "loss": 0.0278, + "step": 17545 + }, + { + "epoch": 8.15713621571362, + "grad_norm": 2.7239253520965576, + "learning_rate": 9.9824619982462e-06, + "loss": 0.0254, + "step": 17546 + }, + { + "epoch": 8.157601115760112, + "grad_norm": 3.42573881149292, + "learning_rate": 9.982460998246101e-06, + "loss": 0.0269, + "step": 17547 + }, + { + "epoch": 8.158066015806602, + "grad_norm": 5.877243518829346, + "learning_rate": 9.982459998246e-06, + "loss": 0.1509, + "step": 17548 + }, + { + "epoch": 8.158530915853092, + "grad_norm": 4.666221618652344, + "learning_rate": 9.9824589982459e-06, + "loss": 0.1353, + "step": 17549 + }, + { + "epoch": 8.158995815899582, + "grad_norm": 2.8309085369110107, + "learning_rate": 9.9824579982458e-06, + "loss": 0.0483, + "step": 17550 + }, + { + "epoch": 8.159460715946071, + "grad_norm": 3.3196303844451904, + "learning_rate": 9.9824569982457e-06, + "loss": 0.0367, + "step": 17551 + }, + { + "epoch": 8.159925615992561, + "grad_norm": 4.036142826080322, + "learning_rate": 9.982455998245601e-06, + "loss": 0.0916, + "step": 17552 + }, + { + "epoch": 8.160390516039051, + "grad_norm": 3.2837395668029785, + "learning_rate": 9.982454998245502e-06, + "loss": 0.0554, + "step": 17553 + }, + { + "epoch": 8.16085541608554, + "grad_norm": 4.631749153137207, + "learning_rate": 9.9824539982454e-06, + "loss": 0.1005, + "step": 17554 + }, + { + "epoch": 8.161320316132032, + "grad_norm": 5.269659042358398, + "learning_rate": 9.982452998245301e-06, + "loss": 0.0356, + "step": 17555 + }, + { + "epoch": 8.161785216178522, + "grad_norm": 5.71989107131958, + "learning_rate": 9.9824519982452e-06, + "loss": 0.0272, + "step": 17556 + }, + { + "epoch": 8.162250116225012, + "grad_norm": 6.114136219024658, + "learning_rate": 9.982450998245101e-06, + "loss": 0.0701, + "step": 17557 + }, + { + "epoch": 8.162715016271502, + "grad_norm": 6.785975456237793, + "learning_rate": 9.982449998245002e-06, + "loss": 0.1892, + "step": 17558 + }, + { + "epoch": 8.163179916317992, + "grad_norm": 5.000910758972168, + "learning_rate": 9.9824489982449e-06, + "loss": 0.0756, + "step": 17559 + }, + { + "epoch": 8.163644816364481, + "grad_norm": 4.653187274932861, + "learning_rate": 9.982447998244801e-06, + "loss": 0.0841, + "step": 17560 + }, + { + "epoch": 8.164109716410971, + "grad_norm": 5.289646148681641, + "learning_rate": 9.9824469982447e-06, + "loss": 0.0635, + "step": 17561 + }, + { + "epoch": 8.164574616457461, + "grad_norm": 6.374571800231934, + "learning_rate": 9.9824459982446e-06, + "loss": 0.1737, + "step": 17562 + }, + { + "epoch": 8.16503951650395, + "grad_norm": 10.27802848815918, + "learning_rate": 9.9824449982445e-06, + "loss": 0.2328, + "step": 17563 + }, + { + "epoch": 8.165504416550442, + "grad_norm": 1.813828945159912, + "learning_rate": 9.9824439982444e-06, + "loss": 0.0163, + "step": 17564 + }, + { + "epoch": 8.165969316596932, + "grad_norm": 5.96809720993042, + "learning_rate": 9.982442998244301e-06, + "loss": 0.1653, + "step": 17565 + }, + { + "epoch": 8.166434216643422, + "grad_norm": 3.813498020172119, + "learning_rate": 9.982441998244202e-06, + "loss": 0.0563, + "step": 17566 + }, + { + "epoch": 8.166899116689912, + "grad_norm": 4.6298747062683105, + "learning_rate": 9.9824409982441e-06, + "loss": 0.0837, + "step": 17567 + }, + { + "epoch": 8.167364016736402, + "grad_norm": 6.7848334312438965, + "learning_rate": 9.982439998244e-06, + "loss": 0.163, + "step": 17568 + }, + { + "epoch": 8.167828916782891, + "grad_norm": 5.734289169311523, + "learning_rate": 9.9824389982439e-06, + "loss": 0.1189, + "step": 17569 + }, + { + "epoch": 8.168293816829381, + "grad_norm": 6.495918273925781, + "learning_rate": 9.982437998243801e-06, + "loss": 0.0991, + "step": 17570 + }, + { + "epoch": 8.168758716875871, + "grad_norm": 3.323789596557617, + "learning_rate": 9.982436998243702e-06, + "loss": 0.0692, + "step": 17571 + }, + { + "epoch": 8.16922361692236, + "grad_norm": 6.616654396057129, + "learning_rate": 9.9824359982436e-06, + "loss": 0.1711, + "step": 17572 + }, + { + "epoch": 8.169688516968852, + "grad_norm": 1.3470548391342163, + "learning_rate": 9.982434998243501e-06, + "loss": 0.0138, + "step": 17573 + }, + { + "epoch": 8.170153417015342, + "grad_norm": 4.659143924713135, + "learning_rate": 9.9824339982434e-06, + "loss": 0.0803, + "step": 17574 + }, + { + "epoch": 8.170618317061832, + "grad_norm": 6.245935916900635, + "learning_rate": 9.982432998243301e-06, + "loss": 0.094, + "step": 17575 + }, + { + "epoch": 8.171083217108322, + "grad_norm": 3.7624197006225586, + "learning_rate": 9.9824319982432e-06, + "loss": 0.0508, + "step": 17576 + }, + { + "epoch": 8.171548117154812, + "grad_norm": 4.508901596069336, + "learning_rate": 9.9824309982431e-06, + "loss": 0.0511, + "step": 17577 + }, + { + "epoch": 8.172013017201301, + "grad_norm": 4.353636741638184, + "learning_rate": 9.982429998243001e-06, + "loss": 0.0644, + "step": 17578 + }, + { + "epoch": 8.172477917247791, + "grad_norm": 4.050585746765137, + "learning_rate": 9.9824289982429e-06, + "loss": 0.0457, + "step": 17579 + }, + { + "epoch": 8.172942817294281, + "grad_norm": 3.5037527084350586, + "learning_rate": 9.9824279982428e-06, + "loss": 0.0664, + "step": 17580 + }, + { + "epoch": 8.17340771734077, + "grad_norm": 3.7225914001464844, + "learning_rate": 9.9824269982427e-06, + "loss": 0.0695, + "step": 17581 + }, + { + "epoch": 8.173872617387262, + "grad_norm": 5.006959915161133, + "learning_rate": 9.9824259982426e-06, + "loss": 0.107, + "step": 17582 + }, + { + "epoch": 8.174337517433752, + "grad_norm": 4.108951091766357, + "learning_rate": 9.982424998242501e-06, + "loss": 0.1005, + "step": 17583 + }, + { + "epoch": 8.174802417480242, + "grad_norm": 9.919323921203613, + "learning_rate": 9.982423998242402e-06, + "loss": 0.6312, + "step": 17584 + }, + { + "epoch": 8.175267317526732, + "grad_norm": 2.7610182762145996, + "learning_rate": 9.9824229982423e-06, + "loss": 0.044, + "step": 17585 + }, + { + "epoch": 8.175732217573222, + "grad_norm": 5.443370819091797, + "learning_rate": 9.982421998242201e-06, + "loss": 0.1431, + "step": 17586 + }, + { + "epoch": 8.176197117619711, + "grad_norm": 3.9998373985290527, + "learning_rate": 9.9824209982421e-06, + "loss": 0.0571, + "step": 17587 + }, + { + "epoch": 8.176662017666201, + "grad_norm": 2.9547102451324463, + "learning_rate": 9.982419998242001e-06, + "loss": 0.0575, + "step": 17588 + }, + { + "epoch": 8.177126917712691, + "grad_norm": 3.3661916255950928, + "learning_rate": 9.9824189982419e-06, + "loss": 0.055, + "step": 17589 + }, + { + "epoch": 8.177591817759183, + "grad_norm": 4.76170015335083, + "learning_rate": 9.9824179982418e-06, + "loss": 0.0913, + "step": 17590 + }, + { + "epoch": 8.178056717805672, + "grad_norm": 5.386596202850342, + "learning_rate": 9.982416998241701e-06, + "loss": 0.1995, + "step": 17591 + }, + { + "epoch": 8.178521617852162, + "grad_norm": 7.129925727844238, + "learning_rate": 9.9824159982416e-06, + "loss": 0.1575, + "step": 17592 + }, + { + "epoch": 8.178986517898652, + "grad_norm": 8.631330490112305, + "learning_rate": 9.982414998241501e-06, + "loss": 0.1854, + "step": 17593 + }, + { + "epoch": 8.179451417945142, + "grad_norm": 4.75363826751709, + "learning_rate": 9.9824139982414e-06, + "loss": 0.0887, + "step": 17594 + }, + { + "epoch": 8.179916317991632, + "grad_norm": 5.871790409088135, + "learning_rate": 9.9824129982413e-06, + "loss": 0.0624, + "step": 17595 + }, + { + "epoch": 8.180381218038121, + "grad_norm": 4.663618564605713, + "learning_rate": 9.982411998241201e-06, + "loss": 0.1295, + "step": 17596 + }, + { + "epoch": 8.180846118084611, + "grad_norm": 5.086233139038086, + "learning_rate": 9.982410998241102e-06, + "loss": 0.1378, + "step": 17597 + }, + { + "epoch": 8.181311018131101, + "grad_norm": 5.94913911819458, + "learning_rate": 9.982409998241e-06, + "loss": 0.0634, + "step": 17598 + }, + { + "epoch": 8.181775918177593, + "grad_norm": 6.581818103790283, + "learning_rate": 9.982408998240901e-06, + "loss": 0.259, + "step": 17599 + }, + { + "epoch": 8.182240818224082, + "grad_norm": 3.650136947631836, + "learning_rate": 9.9824079982408e-06, + "loss": 0.0563, + "step": 17600 + }, + { + "epoch": 8.182705718270572, + "grad_norm": 9.130135536193848, + "learning_rate": 9.9824069982407e-06, + "loss": 0.2166, + "step": 17601 + }, + { + "epoch": 8.183170618317062, + "grad_norm": 4.934665679931641, + "learning_rate": 9.9824059982406e-06, + "loss": 0.1001, + "step": 17602 + }, + { + "epoch": 8.183635518363552, + "grad_norm": 5.6486663818359375, + "learning_rate": 9.9824049982405e-06, + "loss": 0.0685, + "step": 17603 + }, + { + "epoch": 8.184100418410042, + "grad_norm": 1.538563847541809, + "learning_rate": 9.982403998240401e-06, + "loss": 0.019, + "step": 17604 + }, + { + "epoch": 8.184565318456531, + "grad_norm": 4.581446170806885, + "learning_rate": 9.9824029982403e-06, + "loss": 0.1303, + "step": 17605 + }, + { + "epoch": 8.185030218503021, + "grad_norm": 5.657509803771973, + "learning_rate": 9.982401998240201e-06, + "loss": 0.0794, + "step": 17606 + }, + { + "epoch": 8.185495118549511, + "grad_norm": 7.289535999298096, + "learning_rate": 9.9824009982401e-06, + "loss": 0.1455, + "step": 17607 + }, + { + "epoch": 8.185960018596003, + "grad_norm": 7.288778781890869, + "learning_rate": 9.98239999824e-06, + "loss": 0.2401, + "step": 17608 + }, + { + "epoch": 8.186424918642492, + "grad_norm": 2.78556489944458, + "learning_rate": 9.982398998239901e-06, + "loss": 0.0355, + "step": 17609 + }, + { + "epoch": 8.186889818688982, + "grad_norm": 3.711716890335083, + "learning_rate": 9.982397998239802e-06, + "loss": 0.0349, + "step": 17610 + }, + { + "epoch": 8.187354718735472, + "grad_norm": 1.3072396516799927, + "learning_rate": 9.9823969982397e-06, + "loss": 0.0101, + "step": 17611 + }, + { + "epoch": 8.187819618781962, + "grad_norm": 5.594305038452148, + "learning_rate": 9.9823959982396e-06, + "loss": 0.0834, + "step": 17612 + }, + { + "epoch": 8.188284518828452, + "grad_norm": 5.768432140350342, + "learning_rate": 9.9823949982395e-06, + "loss": 0.0989, + "step": 17613 + }, + { + "epoch": 8.188749418874941, + "grad_norm": 3.7947442531585693, + "learning_rate": 9.9823939982394e-06, + "loss": 0.0777, + "step": 17614 + }, + { + "epoch": 8.189214318921431, + "grad_norm": 6.485776901245117, + "learning_rate": 9.9823929982393e-06, + "loss": 0.1942, + "step": 17615 + }, + { + "epoch": 8.189679218967921, + "grad_norm": 3.6533641815185547, + "learning_rate": 9.9823919982392e-06, + "loss": 0.0392, + "step": 17616 + }, + { + "epoch": 8.190144119014413, + "grad_norm": 5.562355995178223, + "learning_rate": 9.982390998239101e-06, + "loss": 0.0399, + "step": 17617 + }, + { + "epoch": 8.190609019060902, + "grad_norm": 4.272220134735107, + "learning_rate": 9.982389998239e-06, + "loss": 0.0474, + "step": 17618 + }, + { + "epoch": 8.191073919107392, + "grad_norm": 4.155586242675781, + "learning_rate": 9.982388998238901e-06, + "loss": 0.1392, + "step": 17619 + }, + { + "epoch": 8.191538819153882, + "grad_norm": 8.203791618347168, + "learning_rate": 9.9823879982388e-06, + "loss": 0.0624, + "step": 17620 + }, + { + "epoch": 8.192003719200372, + "grad_norm": 4.205268859863281, + "learning_rate": 9.9823869982387e-06, + "loss": 0.0543, + "step": 17621 + }, + { + "epoch": 8.192468619246862, + "grad_norm": 4.967653751373291, + "learning_rate": 9.982385998238601e-06, + "loss": 0.1036, + "step": 17622 + }, + { + "epoch": 8.192933519293351, + "grad_norm": 3.961521863937378, + "learning_rate": 9.9823849982385e-06, + "loss": 0.1122, + "step": 17623 + }, + { + "epoch": 8.193398419339841, + "grad_norm": 7.7505340576171875, + "learning_rate": 9.982383998238401e-06, + "loss": 0.164, + "step": 17624 + }, + { + "epoch": 8.193863319386331, + "grad_norm": 3.5624756813049316, + "learning_rate": 9.9823829982383e-06, + "loss": 0.0474, + "step": 17625 + }, + { + "epoch": 8.194328219432823, + "grad_norm": 5.482219219207764, + "learning_rate": 9.9823819982382e-06, + "loss": 0.1605, + "step": 17626 + }, + { + "epoch": 8.194793119479312, + "grad_norm": 6.802191257476807, + "learning_rate": 9.982380998238101e-06, + "loss": 0.2344, + "step": 17627 + }, + { + "epoch": 8.195258019525802, + "grad_norm": 4.4646453857421875, + "learning_rate": 9.982379998238e-06, + "loss": 0.0851, + "step": 17628 + }, + { + "epoch": 8.195722919572292, + "grad_norm": 6.172651290893555, + "learning_rate": 9.9823789982379e-06, + "loss": 0.0964, + "step": 17629 + }, + { + "epoch": 8.196187819618782, + "grad_norm": 6.171413421630859, + "learning_rate": 9.982377998237801e-06, + "loss": 0.0934, + "step": 17630 + }, + { + "epoch": 8.196652719665272, + "grad_norm": 8.842432975769043, + "learning_rate": 9.9823769982377e-06, + "loss": 0.197, + "step": 17631 + }, + { + "epoch": 8.197117619711761, + "grad_norm": 2.0575449466705322, + "learning_rate": 9.982375998237601e-06, + "loss": 0.0391, + "step": 17632 + }, + { + "epoch": 8.197582519758251, + "grad_norm": 1.744868516921997, + "learning_rate": 9.9823749982375e-06, + "loss": 0.0199, + "step": 17633 + }, + { + "epoch": 8.198047419804743, + "grad_norm": 1.7105751037597656, + "learning_rate": 9.9823739982374e-06, + "loss": 0.018, + "step": 17634 + }, + { + "epoch": 8.198512319851233, + "grad_norm": 5.692076683044434, + "learning_rate": 9.982372998237301e-06, + "loss": 0.1055, + "step": 17635 + }, + { + "epoch": 8.198977219897722, + "grad_norm": 4.773223876953125, + "learning_rate": 9.9823719982372e-06, + "loss": 0.0624, + "step": 17636 + }, + { + "epoch": 8.199442119944212, + "grad_norm": 4.431015491485596, + "learning_rate": 9.982370998237101e-06, + "loss": 0.1315, + "step": 17637 + }, + { + "epoch": 8.199907019990702, + "grad_norm": 3.467609167098999, + "learning_rate": 9.982369998237e-06, + "loss": 0.047, + "step": 17638 + }, + { + "epoch": 8.200371920037192, + "grad_norm": 9.532483100891113, + "learning_rate": 9.9823689982369e-06, + "loss": 0.3146, + "step": 17639 + }, + { + "epoch": 8.200836820083682, + "grad_norm": 4.45769739151001, + "learning_rate": 9.982367998236801e-06, + "loss": 0.0942, + "step": 17640 + }, + { + "epoch": 8.201301720130171, + "grad_norm": 4.301830768585205, + "learning_rate": 9.9823669982367e-06, + "loss": 0.0852, + "step": 17641 + }, + { + "epoch": 8.201766620176661, + "grad_norm": 6.3707733154296875, + "learning_rate": 9.982365998236601e-06, + "loss": 0.103, + "step": 17642 + }, + { + "epoch": 8.202231520223153, + "grad_norm": 4.6540327072143555, + "learning_rate": 9.982364998236502e-06, + "loss": 0.1013, + "step": 17643 + }, + { + "epoch": 8.202696420269643, + "grad_norm": 4.435519218444824, + "learning_rate": 9.9823639982364e-06, + "loss": 0.1193, + "step": 17644 + }, + { + "epoch": 8.203161320316132, + "grad_norm": 6.129276752471924, + "learning_rate": 9.9823629982363e-06, + "loss": 0.0799, + "step": 17645 + }, + { + "epoch": 8.203626220362622, + "grad_norm": 5.1529951095581055, + "learning_rate": 9.9823619982362e-06, + "loss": 0.124, + "step": 17646 + }, + { + "epoch": 8.204091120409112, + "grad_norm": 4.056743144989014, + "learning_rate": 9.9823609982361e-06, + "loss": 0.0907, + "step": 17647 + }, + { + "epoch": 8.204556020455602, + "grad_norm": 3.52050518989563, + "learning_rate": 9.982359998236001e-06, + "loss": 0.0425, + "step": 17648 + }, + { + "epoch": 8.205020920502092, + "grad_norm": 4.935365200042725, + "learning_rate": 9.9823589982359e-06, + "loss": 0.0943, + "step": 17649 + }, + { + "epoch": 8.205485820548581, + "grad_norm": 5.24942684173584, + "learning_rate": 9.982357998235801e-06, + "loss": 0.0822, + "step": 17650 + }, + { + "epoch": 8.205950720595071, + "grad_norm": 5.125123023986816, + "learning_rate": 9.9823569982357e-06, + "loss": 0.0945, + "step": 17651 + }, + { + "epoch": 8.206415620641563, + "grad_norm": 1.048885464668274, + "learning_rate": 9.9823559982356e-06, + "loss": 0.0099, + "step": 17652 + }, + { + "epoch": 8.206880520688053, + "grad_norm": 4.901048183441162, + "learning_rate": 9.982354998235501e-06, + "loss": 0.1101, + "step": 17653 + }, + { + "epoch": 8.207345420734542, + "grad_norm": 3.8495099544525146, + "learning_rate": 9.982353998235402e-06, + "loss": 0.092, + "step": 17654 + }, + { + "epoch": 8.207810320781032, + "grad_norm": 5.856617450714111, + "learning_rate": 9.982352998235301e-06, + "loss": 0.1143, + "step": 17655 + }, + { + "epoch": 8.208275220827522, + "grad_norm": 7.034377574920654, + "learning_rate": 9.9823519982352e-06, + "loss": 0.1997, + "step": 17656 + }, + { + "epoch": 8.208740120874012, + "grad_norm": 2.3411900997161865, + "learning_rate": 9.9823509982351e-06, + "loss": 0.0263, + "step": 17657 + }, + { + "epoch": 8.209205020920502, + "grad_norm": 6.8007493019104, + "learning_rate": 9.982349998235e-06, + "loss": 0.111, + "step": 17658 + }, + { + "epoch": 8.209669920966991, + "grad_norm": 4.238467693328857, + "learning_rate": 9.9823489982349e-06, + "loss": 0.0551, + "step": 17659 + }, + { + "epoch": 8.210134821013481, + "grad_norm": 3.7438511848449707, + "learning_rate": 9.982347998234801e-06, + "loss": 0.0739, + "step": 17660 + }, + { + "epoch": 8.210599721059973, + "grad_norm": 6.266748905181885, + "learning_rate": 9.982346998234702e-06, + "loss": 0.1514, + "step": 17661 + }, + { + "epoch": 8.211064621106463, + "grad_norm": 3.7543673515319824, + "learning_rate": 9.9823459982346e-06, + "loss": 0.0953, + "step": 17662 + }, + { + "epoch": 8.211529521152952, + "grad_norm": 5.800658226013184, + "learning_rate": 9.982344998234501e-06, + "loss": 0.1134, + "step": 17663 + }, + { + "epoch": 8.211994421199442, + "grad_norm": 4.618396282196045, + "learning_rate": 9.9823439982344e-06, + "loss": 0.0961, + "step": 17664 + }, + { + "epoch": 8.212459321245932, + "grad_norm": 4.2535223960876465, + "learning_rate": 9.9823429982343e-06, + "loss": 0.0718, + "step": 17665 + }, + { + "epoch": 8.212924221292422, + "grad_norm": 4.291769504547119, + "learning_rate": 9.982341998234201e-06, + "loss": 0.0659, + "step": 17666 + }, + { + "epoch": 8.213389121338912, + "grad_norm": 2.363668918609619, + "learning_rate": 9.9823409982341e-06, + "loss": 0.028, + "step": 17667 + }, + { + "epoch": 8.213854021385401, + "grad_norm": 5.70021390914917, + "learning_rate": 9.982339998234001e-06, + "loss": 0.0547, + "step": 17668 + }, + { + "epoch": 8.214318921431893, + "grad_norm": 3.5149950981140137, + "learning_rate": 9.9823389982339e-06, + "loss": 0.0566, + "step": 17669 + }, + { + "epoch": 8.214783821478383, + "grad_norm": 4.561690807342529, + "learning_rate": 9.9823379982338e-06, + "loss": 0.0808, + "step": 17670 + }, + { + "epoch": 8.215248721524873, + "grad_norm": 6.187793254852295, + "learning_rate": 9.9823369982337e-06, + "loss": 0.1201, + "step": 17671 + }, + { + "epoch": 8.215713621571362, + "grad_norm": 3.705101728439331, + "learning_rate": 9.9823359982336e-06, + "loss": 0.072, + "step": 17672 + }, + { + "epoch": 8.216178521617852, + "grad_norm": 4.978965759277344, + "learning_rate": 9.982334998233501e-06, + "loss": 0.0566, + "step": 17673 + }, + { + "epoch": 8.216643421664342, + "grad_norm": 5.268990993499756, + "learning_rate": 9.982333998233402e-06, + "loss": 0.0941, + "step": 17674 + }, + { + "epoch": 8.217108321710832, + "grad_norm": 4.659386157989502, + "learning_rate": 9.9823329982333e-06, + "loss": 0.0889, + "step": 17675 + }, + { + "epoch": 8.217573221757322, + "grad_norm": 5.348933219909668, + "learning_rate": 9.9823319982332e-06, + "loss": 0.1033, + "step": 17676 + }, + { + "epoch": 8.218038121803811, + "grad_norm": 3.9068961143493652, + "learning_rate": 9.9823309982331e-06, + "loss": 0.0664, + "step": 17677 + }, + { + "epoch": 8.218503021850303, + "grad_norm": 2.5201573371887207, + "learning_rate": 9.982329998233001e-06, + "loss": 0.0352, + "step": 17678 + }, + { + "epoch": 8.218967921896793, + "grad_norm": 6.708131313323975, + "learning_rate": 9.982328998232902e-06, + "loss": 0.1258, + "step": 17679 + }, + { + "epoch": 8.219432821943283, + "grad_norm": 3.806952476501465, + "learning_rate": 9.9823279982328e-06, + "loss": 0.0429, + "step": 17680 + }, + { + "epoch": 8.219897721989772, + "grad_norm": 3.863675117492676, + "learning_rate": 9.982326998232701e-06, + "loss": 0.0785, + "step": 17681 + }, + { + "epoch": 8.220362622036262, + "grad_norm": 4.919745445251465, + "learning_rate": 9.9823259982326e-06, + "loss": 0.1117, + "step": 17682 + }, + { + "epoch": 8.220827522082752, + "grad_norm": 8.226219177246094, + "learning_rate": 9.9823249982325e-06, + "loss": 0.2356, + "step": 17683 + }, + { + "epoch": 8.221292422129242, + "grad_norm": 2.2611865997314453, + "learning_rate": 9.9823239982324e-06, + "loss": 0.0242, + "step": 17684 + }, + { + "epoch": 8.221757322175732, + "grad_norm": 4.8436384201049805, + "learning_rate": 9.9823229982323e-06, + "loss": 0.0931, + "step": 17685 + }, + { + "epoch": 8.222222222222221, + "grad_norm": 3.8470652103424072, + "learning_rate": 9.982321998232201e-06, + "loss": 0.0616, + "step": 17686 + }, + { + "epoch": 8.222687122268713, + "grad_norm": 4.4857354164123535, + "learning_rate": 9.9823209982321e-06, + "loss": 0.0309, + "step": 17687 + }, + { + "epoch": 8.223152022315203, + "grad_norm": 1.8801143169403076, + "learning_rate": 9.982319998232e-06, + "loss": 0.0138, + "step": 17688 + }, + { + "epoch": 8.223616922361693, + "grad_norm": 7.315993785858154, + "learning_rate": 9.9823189982319e-06, + "loss": 0.1646, + "step": 17689 + }, + { + "epoch": 8.224081822408182, + "grad_norm": 4.763540744781494, + "learning_rate": 9.9823179982318e-06, + "loss": 0.115, + "step": 17690 + }, + { + "epoch": 8.224546722454672, + "grad_norm": 4.598278045654297, + "learning_rate": 9.982316998231701e-06, + "loss": 0.0425, + "step": 17691 + }, + { + "epoch": 8.225011622501162, + "grad_norm": 6.582831382751465, + "learning_rate": 9.982315998231602e-06, + "loss": 0.2212, + "step": 17692 + }, + { + "epoch": 8.225476522547652, + "grad_norm": 4.931220531463623, + "learning_rate": 9.9823149982315e-06, + "loss": 0.1147, + "step": 17693 + }, + { + "epoch": 8.225941422594142, + "grad_norm": 3.747471570968628, + "learning_rate": 9.982313998231401e-06, + "loss": 0.0532, + "step": 17694 + }, + { + "epoch": 8.226406322640631, + "grad_norm": 5.99900484085083, + "learning_rate": 9.9823129982313e-06, + "loss": 0.1017, + "step": 17695 + }, + { + "epoch": 8.226871222687123, + "grad_norm": 4.470707893371582, + "learning_rate": 9.9823119982312e-06, + "loss": 0.1007, + "step": 17696 + }, + { + "epoch": 8.227336122733613, + "grad_norm": 7.104070663452148, + "learning_rate": 9.9823109982311e-06, + "loss": 0.2188, + "step": 17697 + }, + { + "epoch": 8.227801022780103, + "grad_norm": 3.796193838119507, + "learning_rate": 9.982309998231e-06, + "loss": 0.1022, + "step": 17698 + }, + { + "epoch": 8.228265922826592, + "grad_norm": 8.282791137695312, + "learning_rate": 9.982308998230901e-06, + "loss": 0.2977, + "step": 17699 + }, + { + "epoch": 8.228730822873082, + "grad_norm": 6.0851898193359375, + "learning_rate": 9.9823079982308e-06, + "loss": 0.1692, + "step": 17700 + }, + { + "epoch": 8.229195722919572, + "grad_norm": 5.122959136962891, + "learning_rate": 9.9823069982307e-06, + "loss": 0.0908, + "step": 17701 + }, + { + "epoch": 8.229660622966062, + "grad_norm": 3.806959629058838, + "learning_rate": 9.9823059982306e-06, + "loss": 0.0456, + "step": 17702 + }, + { + "epoch": 8.230125523012552, + "grad_norm": 6.48719596862793, + "learning_rate": 9.9823049982305e-06, + "loss": 0.2278, + "step": 17703 + }, + { + "epoch": 8.230590423059041, + "grad_norm": 4.347138404846191, + "learning_rate": 9.982303998230401e-06, + "loss": 0.1001, + "step": 17704 + }, + { + "epoch": 8.231055323105533, + "grad_norm": 3.66064453125, + "learning_rate": 9.982302998230302e-06, + "loss": 0.0629, + "step": 17705 + }, + { + "epoch": 8.231520223152023, + "grad_norm": 4.929072856903076, + "learning_rate": 9.9823019982302e-06, + "loss": 0.1211, + "step": 17706 + }, + { + "epoch": 8.231985123198513, + "grad_norm": 7.738246440887451, + "learning_rate": 9.982300998230101e-06, + "loss": 0.2469, + "step": 17707 + }, + { + "epoch": 8.232450023245002, + "grad_norm": 2.493696928024292, + "learning_rate": 9.98229999823e-06, + "loss": 0.0205, + "step": 17708 + }, + { + "epoch": 8.232914923291492, + "grad_norm": 4.192878246307373, + "learning_rate": 9.982298998229901e-06, + "loss": 0.0436, + "step": 17709 + }, + { + "epoch": 8.233379823337982, + "grad_norm": 4.1995415687561035, + "learning_rate": 9.9822979982298e-06, + "loss": 0.0458, + "step": 17710 + }, + { + "epoch": 8.233844723384472, + "grad_norm": 5.62082052230835, + "learning_rate": 9.9822969982297e-06, + "loss": 0.096, + "step": 17711 + }, + { + "epoch": 8.234309623430962, + "grad_norm": 9.58464527130127, + "learning_rate": 9.982295998229601e-06, + "loss": 0.1937, + "step": 17712 + }, + { + "epoch": 8.234774523477453, + "grad_norm": 11.183013916015625, + "learning_rate": 9.9822949982295e-06, + "loss": 0.1675, + "step": 17713 + }, + { + "epoch": 8.235239423523943, + "grad_norm": 3.635967254638672, + "learning_rate": 9.9822939982294e-06, + "loss": 0.0545, + "step": 17714 + }, + { + "epoch": 8.235704323570433, + "grad_norm": 6.5181565284729, + "learning_rate": 9.9822929982293e-06, + "loss": 0.123, + "step": 17715 + }, + { + "epoch": 8.236169223616923, + "grad_norm": 4.774085521697998, + "learning_rate": 9.9822919982292e-06, + "loss": 0.0745, + "step": 17716 + }, + { + "epoch": 8.236634123663412, + "grad_norm": 5.960671424865723, + "learning_rate": 9.982290998229101e-06, + "loss": 0.1196, + "step": 17717 + }, + { + "epoch": 8.237099023709902, + "grad_norm": 4.743346691131592, + "learning_rate": 9.982289998229002e-06, + "loss": 0.1226, + "step": 17718 + }, + { + "epoch": 8.237563923756392, + "grad_norm": 3.5880730152130127, + "learning_rate": 9.9822889982289e-06, + "loss": 0.0381, + "step": 17719 + }, + { + "epoch": 8.238028823802882, + "grad_norm": 2.313593626022339, + "learning_rate": 9.9822879982288e-06, + "loss": 0.0318, + "step": 17720 + }, + { + "epoch": 8.238493723849372, + "grad_norm": 6.398428440093994, + "learning_rate": 9.9822869982287e-06, + "loss": 0.1369, + "step": 17721 + }, + { + "epoch": 8.238958623895863, + "grad_norm": 4.406614780426025, + "learning_rate": 9.982285998228601e-06, + "loss": 0.0772, + "step": 17722 + }, + { + "epoch": 8.239423523942353, + "grad_norm": 8.142279624938965, + "learning_rate": 9.982284998228502e-06, + "loss": 0.3311, + "step": 17723 + }, + { + "epoch": 8.239888423988843, + "grad_norm": 4.038975715637207, + "learning_rate": 9.9822839982284e-06, + "loss": 0.079, + "step": 17724 + }, + { + "epoch": 8.240353324035333, + "grad_norm": 6.496581554412842, + "learning_rate": 9.982282998228301e-06, + "loss": 0.2258, + "step": 17725 + }, + { + "epoch": 8.240818224081822, + "grad_norm": 1.533023715019226, + "learning_rate": 9.9822819982282e-06, + "loss": 0.0158, + "step": 17726 + }, + { + "epoch": 8.241283124128312, + "grad_norm": 2.5248825550079346, + "learning_rate": 9.982280998228101e-06, + "loss": 0.0361, + "step": 17727 + }, + { + "epoch": 8.241748024174802, + "grad_norm": 4.4437761306762695, + "learning_rate": 9.982279998228e-06, + "loss": 0.0695, + "step": 17728 + }, + { + "epoch": 8.242212924221292, + "grad_norm": 4.335240840911865, + "learning_rate": 9.9822789982279e-06, + "loss": 0.0831, + "step": 17729 + }, + { + "epoch": 8.242677824267782, + "grad_norm": 3.278318166732788, + "learning_rate": 9.982277998227801e-06, + "loss": 0.059, + "step": 17730 + }, + { + "epoch": 8.243142724314273, + "grad_norm": 3.13838529586792, + "learning_rate": 9.9822769982277e-06, + "loss": 0.0615, + "step": 17731 + }, + { + "epoch": 8.243607624360763, + "grad_norm": 9.765877723693848, + "learning_rate": 9.9822759982276e-06, + "loss": 0.1588, + "step": 17732 + }, + { + "epoch": 8.244072524407253, + "grad_norm": 5.0769453048706055, + "learning_rate": 9.9822749982275e-06, + "loss": 0.1265, + "step": 17733 + }, + { + "epoch": 8.244537424453743, + "grad_norm": 7.329440116882324, + "learning_rate": 9.9822739982274e-06, + "loss": 0.161, + "step": 17734 + }, + { + "epoch": 8.245002324500232, + "grad_norm": 4.663931846618652, + "learning_rate": 9.982272998227301e-06, + "loss": 0.0893, + "step": 17735 + }, + { + "epoch": 8.245467224546722, + "grad_norm": 6.424941062927246, + "learning_rate": 9.982271998227202e-06, + "loss": 0.1414, + "step": 17736 + }, + { + "epoch": 8.245932124593212, + "grad_norm": 2.2429306507110596, + "learning_rate": 9.9822709982271e-06, + "loss": 0.0335, + "step": 17737 + }, + { + "epoch": 8.246397024639702, + "grad_norm": 2.851632833480835, + "learning_rate": 9.982269998227001e-06, + "loss": 0.0346, + "step": 17738 + }, + { + "epoch": 8.246861924686192, + "grad_norm": 4.322358131408691, + "learning_rate": 9.9822689982269e-06, + "loss": 0.0679, + "step": 17739 + }, + { + "epoch": 8.247326824732683, + "grad_norm": 7.955330848693848, + "learning_rate": 9.982267998226801e-06, + "loss": 0.1982, + "step": 17740 + }, + { + "epoch": 8.247791724779173, + "grad_norm": 11.960497856140137, + "learning_rate": 9.9822669982267e-06, + "loss": 0.1795, + "step": 17741 + }, + { + "epoch": 8.248256624825663, + "grad_norm": 3.2268106937408447, + "learning_rate": 9.9822659982266e-06, + "loss": 0.0571, + "step": 17742 + }, + { + "epoch": 8.248721524872153, + "grad_norm": 6.636539936065674, + "learning_rate": 9.982264998226501e-06, + "loss": 0.1453, + "step": 17743 + }, + { + "epoch": 8.249186424918642, + "grad_norm": 2.113534450531006, + "learning_rate": 9.9822639982264e-06, + "loss": 0.0221, + "step": 17744 + }, + { + "epoch": 8.249651324965132, + "grad_norm": 4.445288181304932, + "learning_rate": 9.982262998226301e-06, + "loss": 0.0775, + "step": 17745 + }, + { + "epoch": 8.250116225011622, + "grad_norm": 8.00126838684082, + "learning_rate": 9.9822619982262e-06, + "loss": 0.2447, + "step": 17746 + }, + { + "epoch": 8.250581125058112, + "grad_norm": 8.978398323059082, + "learning_rate": 9.9822609982261e-06, + "loss": 0.2321, + "step": 17747 + }, + { + "epoch": 8.251046025104603, + "grad_norm": 6.505034923553467, + "learning_rate": 9.982259998226001e-06, + "loss": 0.1506, + "step": 17748 + }, + { + "epoch": 8.251510925151093, + "grad_norm": 4.710115909576416, + "learning_rate": 9.982258998225902e-06, + "loss": 0.1018, + "step": 17749 + }, + { + "epoch": 8.251975825197583, + "grad_norm": 2.9271552562713623, + "learning_rate": 9.9822579982258e-06, + "loss": 0.0312, + "step": 17750 + }, + { + "epoch": 8.252440725244073, + "grad_norm": 7.192147731781006, + "learning_rate": 9.982256998225701e-06, + "loss": 0.1672, + "step": 17751 + }, + { + "epoch": 8.252905625290563, + "grad_norm": 6.809051513671875, + "learning_rate": 9.9822559982256e-06, + "loss": 0.2135, + "step": 17752 + }, + { + "epoch": 8.253370525337052, + "grad_norm": 9.197710990905762, + "learning_rate": 9.9822549982255e-06, + "loss": 0.351, + "step": 17753 + }, + { + "epoch": 8.253835425383542, + "grad_norm": 2.431414842605591, + "learning_rate": 9.9822539982254e-06, + "loss": 0.0425, + "step": 17754 + }, + { + "epoch": 8.254300325430032, + "grad_norm": 7.510818958282471, + "learning_rate": 9.9822529982253e-06, + "loss": 0.1092, + "step": 17755 + }, + { + "epoch": 8.254765225476522, + "grad_norm": 10.976608276367188, + "learning_rate": 9.982251998225201e-06, + "loss": 0.1786, + "step": 17756 + }, + { + "epoch": 8.255230125523013, + "grad_norm": 3.846691846847534, + "learning_rate": 9.9822509982251e-06, + "loss": 0.0358, + "step": 17757 + }, + { + "epoch": 8.255695025569503, + "grad_norm": 3.701014280319214, + "learning_rate": 9.982249998225001e-06, + "loss": 0.0398, + "step": 17758 + }, + { + "epoch": 8.256159925615993, + "grad_norm": 9.165349006652832, + "learning_rate": 9.9822489982249e-06, + "loss": 0.1105, + "step": 17759 + }, + { + "epoch": 8.256624825662483, + "grad_norm": 12.378952026367188, + "learning_rate": 9.9822479982248e-06, + "loss": 0.4906, + "step": 17760 + }, + { + "epoch": 8.257089725708973, + "grad_norm": 6.333646774291992, + "learning_rate": 9.982246998224701e-06, + "loss": 0.2162, + "step": 17761 + }, + { + "epoch": 8.257554625755462, + "grad_norm": 4.336416244506836, + "learning_rate": 9.982245998224602e-06, + "loss": 0.1149, + "step": 17762 + }, + { + "epoch": 8.258019525801952, + "grad_norm": 6.90402889251709, + "learning_rate": 9.982244998224501e-06, + "loss": 0.1288, + "step": 17763 + }, + { + "epoch": 8.258484425848442, + "grad_norm": 5.221933841705322, + "learning_rate": 9.9822439982244e-06, + "loss": 0.1415, + "step": 17764 + }, + { + "epoch": 8.258949325894932, + "grad_norm": 6.054269313812256, + "learning_rate": 9.9822429982243e-06, + "loss": 0.1197, + "step": 17765 + }, + { + "epoch": 8.259414225941423, + "grad_norm": 7.186636447906494, + "learning_rate": 9.9822419982242e-06, + "loss": 0.2177, + "step": 17766 + }, + { + "epoch": 8.259879125987913, + "grad_norm": 2.1120517253875732, + "learning_rate": 9.9822409982241e-06, + "loss": 0.0287, + "step": 17767 + }, + { + "epoch": 8.260344026034403, + "grad_norm": 5.683676719665527, + "learning_rate": 9.982239998224e-06, + "loss": 0.1611, + "step": 17768 + }, + { + "epoch": 8.260808926080893, + "grad_norm": 6.329883575439453, + "learning_rate": 9.982238998223901e-06, + "loss": 0.0919, + "step": 17769 + }, + { + "epoch": 8.261273826127383, + "grad_norm": 7.819310665130615, + "learning_rate": 9.9822379982238e-06, + "loss": 0.1574, + "step": 17770 + }, + { + "epoch": 8.261738726173872, + "grad_norm": 3.0676653385162354, + "learning_rate": 9.982236998223701e-06, + "loss": 0.0223, + "step": 17771 + }, + { + "epoch": 8.262203626220362, + "grad_norm": 4.864523410797119, + "learning_rate": 9.9822359982236e-06, + "loss": 0.0739, + "step": 17772 + }, + { + "epoch": 8.262668526266852, + "grad_norm": 2.9623630046844482, + "learning_rate": 9.9822349982235e-06, + "loss": 0.0213, + "step": 17773 + }, + { + "epoch": 8.263133426313342, + "grad_norm": 5.610918998718262, + "learning_rate": 9.982233998223401e-06, + "loss": 0.1089, + "step": 17774 + }, + { + "epoch": 8.263598326359833, + "grad_norm": 5.185115814208984, + "learning_rate": 9.9822329982233e-06, + "loss": 0.1067, + "step": 17775 + }, + { + "epoch": 8.264063226406323, + "grad_norm": 3.764747381210327, + "learning_rate": 9.982231998223201e-06, + "loss": 0.1073, + "step": 17776 + }, + { + "epoch": 8.264528126452813, + "grad_norm": 7.67011022567749, + "learning_rate": 9.9822309982231e-06, + "loss": 0.1738, + "step": 17777 + }, + { + "epoch": 8.264993026499303, + "grad_norm": 5.717376708984375, + "learning_rate": 9.982229998223e-06, + "loss": 0.1285, + "step": 17778 + }, + { + "epoch": 8.265457926545793, + "grad_norm": 5.131432056427002, + "learning_rate": 9.9822289982229e-06, + "loss": 0.1044, + "step": 17779 + }, + { + "epoch": 8.265922826592282, + "grad_norm": 4.268527507781982, + "learning_rate": 9.9822279982228e-06, + "loss": 0.0963, + "step": 17780 + }, + { + "epoch": 8.266387726638772, + "grad_norm": 4.479511737823486, + "learning_rate": 9.982226998222701e-06, + "loss": 0.0794, + "step": 17781 + }, + { + "epoch": 8.266852626685262, + "grad_norm": 3.751326560974121, + "learning_rate": 9.982225998222602e-06, + "loss": 0.0717, + "step": 17782 + }, + { + "epoch": 8.267317526731752, + "grad_norm": 5.555661678314209, + "learning_rate": 9.9822249982225e-06, + "loss": 0.1519, + "step": 17783 + }, + { + "epoch": 8.267782426778243, + "grad_norm": 4.863856792449951, + "learning_rate": 9.982223998222401e-06, + "loss": 0.0706, + "step": 17784 + }, + { + "epoch": 8.268247326824733, + "grad_norm": 4.041388511657715, + "learning_rate": 9.9822229982223e-06, + "loss": 0.1213, + "step": 17785 + }, + { + "epoch": 8.268712226871223, + "grad_norm": 5.280613422393799, + "learning_rate": 9.9822219982222e-06, + "loss": 0.1237, + "step": 17786 + }, + { + "epoch": 8.269177126917713, + "grad_norm": 8.776387214660645, + "learning_rate": 9.982220998222101e-06, + "loss": 0.2095, + "step": 17787 + }, + { + "epoch": 8.269642026964203, + "grad_norm": 5.040329933166504, + "learning_rate": 9.982219998222e-06, + "loss": 0.1002, + "step": 17788 + }, + { + "epoch": 8.270106927010692, + "grad_norm": 3.711643934249878, + "learning_rate": 9.982218998221901e-06, + "loss": 0.0389, + "step": 17789 + }, + { + "epoch": 8.270571827057182, + "grad_norm": 7.611664772033691, + "learning_rate": 9.9822179982218e-06, + "loss": 0.1084, + "step": 17790 + }, + { + "epoch": 8.271036727103672, + "grad_norm": 2.7928249835968018, + "learning_rate": 9.9822169982217e-06, + "loss": 0.0629, + "step": 17791 + }, + { + "epoch": 8.271501627150164, + "grad_norm": 3.2647082805633545, + "learning_rate": 9.982215998221601e-06, + "loss": 0.0604, + "step": 17792 + }, + { + "epoch": 8.271966527196653, + "grad_norm": 4.7888264656066895, + "learning_rate": 9.9822149982215e-06, + "loss": 0.0667, + "step": 17793 + }, + { + "epoch": 8.272431427243143, + "grad_norm": 8.619832038879395, + "learning_rate": 9.982213998221401e-06, + "loss": 0.2102, + "step": 17794 + }, + { + "epoch": 8.272896327289633, + "grad_norm": 5.7838454246521, + "learning_rate": 9.982212998221302e-06, + "loss": 0.0851, + "step": 17795 + }, + { + "epoch": 8.273361227336123, + "grad_norm": 6.648274898529053, + "learning_rate": 9.9822119982212e-06, + "loss": 0.1405, + "step": 17796 + }, + { + "epoch": 8.273826127382613, + "grad_norm": 6.089762210845947, + "learning_rate": 9.9822109982211e-06, + "loss": 0.1216, + "step": 17797 + }, + { + "epoch": 8.274291027429102, + "grad_norm": 3.7563047409057617, + "learning_rate": 9.982209998221e-06, + "loss": 0.0506, + "step": 17798 + }, + { + "epoch": 8.274755927475592, + "grad_norm": 5.564242839813232, + "learning_rate": 9.9822089982209e-06, + "loss": 0.0837, + "step": 17799 + }, + { + "epoch": 8.275220827522082, + "grad_norm": 5.000063419342041, + "learning_rate": 9.982207998220801e-06, + "loss": 0.104, + "step": 17800 + }, + { + "epoch": 8.275685727568574, + "grad_norm": 4.2738189697265625, + "learning_rate": 9.9822069982207e-06, + "loss": 0.0847, + "step": 17801 + }, + { + "epoch": 8.276150627615063, + "grad_norm": 4.2171125411987305, + "learning_rate": 9.982205998220601e-06, + "loss": 0.1018, + "step": 17802 + }, + { + "epoch": 8.276615527661553, + "grad_norm": 5.629530429840088, + "learning_rate": 9.9822049982205e-06, + "loss": 0.1046, + "step": 17803 + }, + { + "epoch": 8.277080427708043, + "grad_norm": 3.079834222793579, + "learning_rate": 9.9822039982204e-06, + "loss": 0.0372, + "step": 17804 + }, + { + "epoch": 8.277545327754533, + "grad_norm": 3.6398732662200928, + "learning_rate": 9.982202998220301e-06, + "loss": 0.0497, + "step": 17805 + }, + { + "epoch": 8.278010227801023, + "grad_norm": 8.684796333312988, + "learning_rate": 9.9822019982202e-06, + "loss": 0.1384, + "step": 17806 + }, + { + "epoch": 8.278475127847512, + "grad_norm": 2.8861887454986572, + "learning_rate": 9.982200998220101e-06, + "loss": 0.0151, + "step": 17807 + }, + { + "epoch": 8.278940027894002, + "grad_norm": 5.2140583992004395, + "learning_rate": 9.98219999822e-06, + "loss": 0.1425, + "step": 17808 + }, + { + "epoch": 8.279404927940492, + "grad_norm": 3.5779478549957275, + "learning_rate": 9.9821989982199e-06, + "loss": 0.0756, + "step": 17809 + }, + { + "epoch": 8.279869827986984, + "grad_norm": 4.449139595031738, + "learning_rate": 9.9821979982198e-06, + "loss": 0.045, + "step": 17810 + }, + { + "epoch": 8.280334728033473, + "grad_norm": 5.20997953414917, + "learning_rate": 9.9821969982197e-06, + "loss": 0.1581, + "step": 17811 + }, + { + "epoch": 8.280799628079963, + "grad_norm": 5.402681827545166, + "learning_rate": 9.982195998219601e-06, + "loss": 0.1485, + "step": 17812 + }, + { + "epoch": 8.281264528126453, + "grad_norm": 5.088736057281494, + "learning_rate": 9.982194998219502e-06, + "loss": 0.087, + "step": 17813 + }, + { + "epoch": 8.281729428172943, + "grad_norm": 5.275140285491943, + "learning_rate": 9.9821939982194e-06, + "loss": 0.1661, + "step": 17814 + }, + { + "epoch": 8.282194328219433, + "grad_norm": 13.015751838684082, + "learning_rate": 9.982192998219301e-06, + "loss": 0.7803, + "step": 17815 + }, + { + "epoch": 8.282659228265922, + "grad_norm": 2.7482516765594482, + "learning_rate": 9.9821919982192e-06, + "loss": 0.0399, + "step": 17816 + }, + { + "epoch": 8.283124128312412, + "grad_norm": 3.918818235397339, + "learning_rate": 9.9821909982191e-06, + "loss": 0.067, + "step": 17817 + }, + { + "epoch": 8.283589028358902, + "grad_norm": 7.1581220626831055, + "learning_rate": 9.982189998219001e-06, + "loss": 0.1375, + "step": 17818 + }, + { + "epoch": 8.284053928405394, + "grad_norm": 5.882970809936523, + "learning_rate": 9.9821889982189e-06, + "loss": 0.1629, + "step": 17819 + }, + { + "epoch": 8.284518828451883, + "grad_norm": 5.48150110244751, + "learning_rate": 9.982187998218801e-06, + "loss": 0.1306, + "step": 17820 + }, + { + "epoch": 8.284983728498373, + "grad_norm": 4.167657852172852, + "learning_rate": 9.9821869982187e-06, + "loss": 0.0682, + "step": 17821 + }, + { + "epoch": 8.285448628544863, + "grad_norm": 1.9510705471038818, + "learning_rate": 9.9821859982186e-06, + "loss": 0.0173, + "step": 17822 + }, + { + "epoch": 8.285913528591353, + "grad_norm": 6.991206169128418, + "learning_rate": 9.9821849982185e-06, + "loss": 0.2047, + "step": 17823 + }, + { + "epoch": 8.286378428637843, + "grad_norm": 4.560690402984619, + "learning_rate": 9.9821839982184e-06, + "loss": 0.056, + "step": 17824 + }, + { + "epoch": 8.286843328684332, + "grad_norm": 5.97938346862793, + "learning_rate": 9.982182998218301e-06, + "loss": 0.1105, + "step": 17825 + }, + { + "epoch": 8.287308228730822, + "grad_norm": 4.575105667114258, + "learning_rate": 9.982181998218202e-06, + "loss": 0.0767, + "step": 17826 + }, + { + "epoch": 8.287773128777314, + "grad_norm": 7.889183044433594, + "learning_rate": 9.9821809982181e-06, + "loss": 0.1277, + "step": 17827 + }, + { + "epoch": 8.288238028823804, + "grad_norm": 5.249627590179443, + "learning_rate": 9.982179998218e-06, + "loss": 0.2023, + "step": 17828 + }, + { + "epoch": 8.288702928870293, + "grad_norm": 4.793777942657471, + "learning_rate": 9.9821789982179e-06, + "loss": 0.1079, + "step": 17829 + }, + { + "epoch": 8.289167828916783, + "grad_norm": 5.294475078582764, + "learning_rate": 9.982177998217801e-06, + "loss": 0.1215, + "step": 17830 + }, + { + "epoch": 8.289632728963273, + "grad_norm": 6.664316177368164, + "learning_rate": 9.982176998217702e-06, + "loss": 0.1613, + "step": 17831 + }, + { + "epoch": 8.290097629009763, + "grad_norm": 7.949835777282715, + "learning_rate": 9.9821759982176e-06, + "loss": 0.1327, + "step": 17832 + }, + { + "epoch": 8.290562529056253, + "grad_norm": 1.839188814163208, + "learning_rate": 9.982174998217501e-06, + "loss": 0.0156, + "step": 17833 + }, + { + "epoch": 8.291027429102742, + "grad_norm": 3.734124183654785, + "learning_rate": 9.9821739982174e-06, + "loss": 0.0288, + "step": 17834 + }, + { + "epoch": 8.291492329149232, + "grad_norm": 4.814345359802246, + "learning_rate": 9.9821729982173e-06, + "loss": 0.145, + "step": 17835 + }, + { + "epoch": 8.291957229195724, + "grad_norm": 2.7336089611053467, + "learning_rate": 9.9821719982172e-06, + "loss": 0.0486, + "step": 17836 + }, + { + "epoch": 8.292422129242214, + "grad_norm": 4.406726360321045, + "learning_rate": 9.9821709982171e-06, + "loss": 0.0797, + "step": 17837 + }, + { + "epoch": 8.292887029288703, + "grad_norm": 4.40635347366333, + "learning_rate": 9.982169998217001e-06, + "loss": 0.0614, + "step": 17838 + }, + { + "epoch": 8.293351929335193, + "grad_norm": 9.343017578125, + "learning_rate": 9.9821689982169e-06, + "loss": 0.0911, + "step": 17839 + }, + { + "epoch": 8.293816829381683, + "grad_norm": 3.3038127422332764, + "learning_rate": 9.9821679982168e-06, + "loss": 0.0349, + "step": 17840 + }, + { + "epoch": 8.294281729428173, + "grad_norm": 3.512666702270508, + "learning_rate": 9.9821669982167e-06, + "loss": 0.033, + "step": 17841 + }, + { + "epoch": 8.294746629474663, + "grad_norm": 5.655304908752441, + "learning_rate": 9.9821659982166e-06, + "loss": 0.0677, + "step": 17842 + }, + { + "epoch": 8.295211529521152, + "grad_norm": 5.332894325256348, + "learning_rate": 9.982164998216501e-06, + "loss": 0.1149, + "step": 17843 + }, + { + "epoch": 8.295676429567642, + "grad_norm": 6.184900283813477, + "learning_rate": 9.982163998216402e-06, + "loss": 0.1273, + "step": 17844 + }, + { + "epoch": 8.296141329614134, + "grad_norm": 2.9939119815826416, + "learning_rate": 9.9821629982163e-06, + "loss": 0.0502, + "step": 17845 + }, + { + "epoch": 8.296606229660624, + "grad_norm": 3.276055335998535, + "learning_rate": 9.982161998216201e-06, + "loss": 0.0474, + "step": 17846 + }, + { + "epoch": 8.297071129707113, + "grad_norm": 16.687671661376953, + "learning_rate": 9.9821609982161e-06, + "loss": 0.2338, + "step": 17847 + }, + { + "epoch": 8.297536029753603, + "grad_norm": 5.950453758239746, + "learning_rate": 9.982159998216001e-06, + "loss": 0.1029, + "step": 17848 + }, + { + "epoch": 8.298000929800093, + "grad_norm": 4.0284342765808105, + "learning_rate": 9.9821589982159e-06, + "loss": 0.0946, + "step": 17849 + }, + { + "epoch": 8.298465829846583, + "grad_norm": 8.994490623474121, + "learning_rate": 9.9821579982158e-06, + "loss": 0.1097, + "step": 17850 + }, + { + "epoch": 8.298930729893073, + "grad_norm": 6.1322221755981445, + "learning_rate": 9.982156998215701e-06, + "loss": 0.0689, + "step": 17851 + }, + { + "epoch": 8.299395629939562, + "grad_norm": 5.390318393707275, + "learning_rate": 9.9821559982156e-06, + "loss": 0.0868, + "step": 17852 + }, + { + "epoch": 8.299860529986052, + "grad_norm": 10.283746719360352, + "learning_rate": 9.9821549982155e-06, + "loss": 0.1344, + "step": 17853 + }, + { + "epoch": 8.300325430032544, + "grad_norm": 3.7523510456085205, + "learning_rate": 9.9821539982154e-06, + "loss": 0.0498, + "step": 17854 + }, + { + "epoch": 8.300790330079034, + "grad_norm": 8.072051048278809, + "learning_rate": 9.9821529982153e-06, + "loss": 0.1024, + "step": 17855 + }, + { + "epoch": 8.301255230125523, + "grad_norm": 3.4335289001464844, + "learning_rate": 9.982151998215201e-06, + "loss": 0.0525, + "step": 17856 + }, + { + "epoch": 8.301720130172013, + "grad_norm": 5.697702407836914, + "learning_rate": 9.982150998215102e-06, + "loss": 0.1419, + "step": 17857 + }, + { + "epoch": 8.302185030218503, + "grad_norm": 4.115478992462158, + "learning_rate": 9.982149998215e-06, + "loss": 0.0742, + "step": 17858 + }, + { + "epoch": 8.302649930264993, + "grad_norm": 4.746278285980225, + "learning_rate": 9.982148998214901e-06, + "loss": 0.075, + "step": 17859 + }, + { + "epoch": 8.303114830311483, + "grad_norm": 4.411182880401611, + "learning_rate": 9.9821479982148e-06, + "loss": 0.101, + "step": 17860 + }, + { + "epoch": 8.303579730357972, + "grad_norm": 1.7835873365402222, + "learning_rate": 9.982146998214701e-06, + "loss": 0.0148, + "step": 17861 + }, + { + "epoch": 8.304044630404462, + "grad_norm": 3.071432113647461, + "learning_rate": 9.9821459982146e-06, + "loss": 0.0402, + "step": 17862 + }, + { + "epoch": 8.304509530450954, + "grad_norm": 8.870949745178223, + "learning_rate": 9.9821449982145e-06, + "loss": 0.1634, + "step": 17863 + }, + { + "epoch": 8.304974430497444, + "grad_norm": 4.300585746765137, + "learning_rate": 9.982143998214401e-06, + "loss": 0.0602, + "step": 17864 + }, + { + "epoch": 8.305439330543933, + "grad_norm": 5.0650315284729, + "learning_rate": 9.9821429982143e-06, + "loss": 0.072, + "step": 17865 + }, + { + "epoch": 8.305904230590423, + "grad_norm": 6.066627025604248, + "learning_rate": 9.982141998214201e-06, + "loss": 0.0787, + "step": 17866 + }, + { + "epoch": 8.306369130636913, + "grad_norm": 10.673133850097656, + "learning_rate": 9.9821409982141e-06, + "loss": 0.118, + "step": 17867 + }, + { + "epoch": 8.306834030683403, + "grad_norm": 9.809279441833496, + "learning_rate": 9.982139998214e-06, + "loss": 0.2767, + "step": 17868 + }, + { + "epoch": 8.307298930729893, + "grad_norm": 8.01166820526123, + "learning_rate": 9.982138998213901e-06, + "loss": 0.2117, + "step": 17869 + }, + { + "epoch": 8.307763830776382, + "grad_norm": 3.7896816730499268, + "learning_rate": 9.982137998213802e-06, + "loss": 0.0863, + "step": 17870 + }, + { + "epoch": 8.308228730822874, + "grad_norm": 5.442124843597412, + "learning_rate": 9.9821369982137e-06, + "loss": 0.1404, + "step": 17871 + }, + { + "epoch": 8.308693630869364, + "grad_norm": 6.339612007141113, + "learning_rate": 9.9821359982136e-06, + "loss": 0.1207, + "step": 17872 + }, + { + "epoch": 8.309158530915854, + "grad_norm": 3.336059808731079, + "learning_rate": 9.9821349982135e-06, + "loss": 0.0524, + "step": 17873 + }, + { + "epoch": 8.309623430962343, + "grad_norm": 7.930651664733887, + "learning_rate": 9.982133998213401e-06, + "loss": 0.3108, + "step": 17874 + }, + { + "epoch": 8.310088331008833, + "grad_norm": 6.059904098510742, + "learning_rate": 9.9821329982133e-06, + "loss": 0.1781, + "step": 17875 + }, + { + "epoch": 8.310553231055323, + "grad_norm": 1.8994834423065186, + "learning_rate": 9.9821319982132e-06, + "loss": 0.019, + "step": 17876 + }, + { + "epoch": 8.311018131101813, + "grad_norm": 5.3300886154174805, + "learning_rate": 9.982130998213101e-06, + "loss": 0.1147, + "step": 17877 + }, + { + "epoch": 8.311483031148303, + "grad_norm": 4.491698741912842, + "learning_rate": 9.982129998213e-06, + "loss": 0.0906, + "step": 17878 + }, + { + "epoch": 8.311947931194792, + "grad_norm": 3.050779104232788, + "learning_rate": 9.982128998212901e-06, + "loss": 0.0499, + "step": 17879 + }, + { + "epoch": 8.312412831241284, + "grad_norm": 2.5742502212524414, + "learning_rate": 9.9821279982128e-06, + "loss": 0.0325, + "step": 17880 + }, + { + "epoch": 8.312877731287774, + "grad_norm": 4.436281681060791, + "learning_rate": 9.9821269982127e-06, + "loss": 0.1046, + "step": 17881 + }, + { + "epoch": 8.313342631334264, + "grad_norm": 7.063749313354492, + "learning_rate": 9.982125998212601e-06, + "loss": 0.1858, + "step": 17882 + }, + { + "epoch": 8.313807531380753, + "grad_norm": 8.535898208618164, + "learning_rate": 9.9821249982125e-06, + "loss": 0.0959, + "step": 17883 + }, + { + "epoch": 8.314272431427243, + "grad_norm": 5.353615760803223, + "learning_rate": 9.9821239982124e-06, + "loss": 0.1083, + "step": 17884 + }, + { + "epoch": 8.314737331473733, + "grad_norm": 7.890034198760986, + "learning_rate": 9.9821229982123e-06, + "loss": 0.1651, + "step": 17885 + }, + { + "epoch": 8.315202231520223, + "grad_norm": 6.0805344581604, + "learning_rate": 9.9821219982122e-06, + "loss": 0.1932, + "step": 17886 + }, + { + "epoch": 8.315667131566713, + "grad_norm": 4.250282287597656, + "learning_rate": 9.982120998212101e-06, + "loss": 0.0986, + "step": 17887 + }, + { + "epoch": 8.316132031613202, + "grad_norm": 5.634453773498535, + "learning_rate": 9.982119998212002e-06, + "loss": 0.1336, + "step": 17888 + }, + { + "epoch": 8.316596931659694, + "grad_norm": 7.814250469207764, + "learning_rate": 9.9821189982119e-06, + "loss": 0.1885, + "step": 17889 + }, + { + "epoch": 8.317061831706184, + "grad_norm": 6.712172508239746, + "learning_rate": 9.982117998211801e-06, + "loss": 0.1776, + "step": 17890 + }, + { + "epoch": 8.317526731752674, + "grad_norm": 4.140498638153076, + "learning_rate": 9.9821169982117e-06, + "loss": 0.0598, + "step": 17891 + }, + { + "epoch": 8.317991631799163, + "grad_norm": 6.7156782150268555, + "learning_rate": 9.982115998211601e-06, + "loss": 0.2675, + "step": 17892 + }, + { + "epoch": 8.318456531845653, + "grad_norm": 7.434933662414551, + "learning_rate": 9.9821149982115e-06, + "loss": 0.1922, + "step": 17893 + }, + { + "epoch": 8.318921431892143, + "grad_norm": 6.379201889038086, + "learning_rate": 9.9821139982114e-06, + "loss": 0.0835, + "step": 17894 + }, + { + "epoch": 8.319386331938633, + "grad_norm": 2.7640438079833984, + "learning_rate": 9.982112998211301e-06, + "loss": 0.0293, + "step": 17895 + }, + { + "epoch": 8.319851231985123, + "grad_norm": 6.581839561462402, + "learning_rate": 9.9821119982112e-06, + "loss": 0.1735, + "step": 17896 + }, + { + "epoch": 8.320316132031612, + "grad_norm": 6.977762699127197, + "learning_rate": 9.982110998211101e-06, + "loss": 0.1412, + "step": 17897 + }, + { + "epoch": 8.320781032078104, + "grad_norm": 3.7122373580932617, + "learning_rate": 9.982109998211e-06, + "loss": 0.0716, + "step": 17898 + }, + { + "epoch": 8.321245932124594, + "grad_norm": 4.280832767486572, + "learning_rate": 9.9821089982109e-06, + "loss": 0.0798, + "step": 17899 + }, + { + "epoch": 8.321710832171084, + "grad_norm": 10.165199279785156, + "learning_rate": 9.982107998210801e-06, + "loss": 0.2018, + "step": 17900 + }, + { + "epoch": 8.322175732217573, + "grad_norm": 4.58420991897583, + "learning_rate": 9.982106998210702e-06, + "loss": 0.0598, + "step": 17901 + }, + { + "epoch": 8.322640632264063, + "grad_norm": 4.869275093078613, + "learning_rate": 9.9821059982106e-06, + "loss": 0.0791, + "step": 17902 + }, + { + "epoch": 8.323105532310553, + "grad_norm": 6.440901756286621, + "learning_rate": 9.982104998210501e-06, + "loss": 0.0905, + "step": 17903 + }, + { + "epoch": 8.323570432357043, + "grad_norm": 4.624668598175049, + "learning_rate": 9.9821039982104e-06, + "loss": 0.1393, + "step": 17904 + }, + { + "epoch": 8.324035332403533, + "grad_norm": 1.890991449356079, + "learning_rate": 9.9821029982103e-06, + "loss": 0.0385, + "step": 17905 + }, + { + "epoch": 8.324500232450024, + "grad_norm": 4.604469299316406, + "learning_rate": 9.9821019982102e-06, + "loss": 0.0813, + "step": 17906 + }, + { + "epoch": 8.324965132496514, + "grad_norm": 4.570897579193115, + "learning_rate": 9.9821009982101e-06, + "loss": 0.0415, + "step": 17907 + }, + { + "epoch": 8.325430032543004, + "grad_norm": 2.7506110668182373, + "learning_rate": 9.982099998210001e-06, + "loss": 0.0396, + "step": 17908 + }, + { + "epoch": 8.325894932589494, + "grad_norm": 3.8445727825164795, + "learning_rate": 9.9820989982099e-06, + "loss": 0.0399, + "step": 17909 + }, + { + "epoch": 8.326359832635983, + "grad_norm": 7.256392002105713, + "learning_rate": 9.982097998209801e-06, + "loss": 0.2027, + "step": 17910 + }, + { + "epoch": 8.326824732682473, + "grad_norm": 7.737934112548828, + "learning_rate": 9.9820969982097e-06, + "loss": 0.1112, + "step": 17911 + }, + { + "epoch": 8.327289632728963, + "grad_norm": 3.641742467880249, + "learning_rate": 9.9820959982096e-06, + "loss": 0.0489, + "step": 17912 + }, + { + "epoch": 8.327754532775453, + "grad_norm": 5.1949639320373535, + "learning_rate": 9.982094998209501e-06, + "loss": 0.085, + "step": 17913 + }, + { + "epoch": 8.328219432821943, + "grad_norm": 4.633061408996582, + "learning_rate": 9.982093998209402e-06, + "loss": 0.0662, + "step": 17914 + }, + { + "epoch": 8.328684332868434, + "grad_norm": 4.039157390594482, + "learning_rate": 9.982092998209301e-06, + "loss": 0.0744, + "step": 17915 + }, + { + "epoch": 8.329149232914924, + "grad_norm": 3.412539482116699, + "learning_rate": 9.9820919982092e-06, + "loss": 0.0513, + "step": 17916 + }, + { + "epoch": 8.329614132961414, + "grad_norm": 3.5206308364868164, + "learning_rate": 9.9820909982091e-06, + "loss": 0.0632, + "step": 17917 + }, + { + "epoch": 8.330079033007904, + "grad_norm": 12.350250244140625, + "learning_rate": 9.982089998209e-06, + "loss": 0.047, + "step": 17918 + }, + { + "epoch": 8.330543933054393, + "grad_norm": 4.54279899597168, + "learning_rate": 9.9820889982089e-06, + "loss": 0.1212, + "step": 17919 + }, + { + "epoch": 8.331008833100883, + "grad_norm": 7.377416133880615, + "learning_rate": 9.9820879982088e-06, + "loss": 0.2472, + "step": 17920 + }, + { + "epoch": 8.331473733147373, + "grad_norm": 5.640183448791504, + "learning_rate": 9.982086998208701e-06, + "loss": 0.1426, + "step": 17921 + }, + { + "epoch": 8.331938633193863, + "grad_norm": 2.5904910564422607, + "learning_rate": 9.9820859982086e-06, + "loss": 0.0399, + "step": 17922 + }, + { + "epoch": 8.332403533240353, + "grad_norm": 6.261828899383545, + "learning_rate": 9.982084998208501e-06, + "loss": 0.1994, + "step": 17923 + }, + { + "epoch": 8.332868433286844, + "grad_norm": 6.329906463623047, + "learning_rate": 9.9820839982084e-06, + "loss": 0.1046, + "step": 17924 + }, + { + "epoch": 8.333333333333334, + "grad_norm": 7.515004634857178, + "learning_rate": 9.9820829982083e-06, + "loss": 0.1819, + "step": 17925 + }, + { + "epoch": 8.333798233379824, + "grad_norm": 2.530937671661377, + "learning_rate": 9.982081998208201e-06, + "loss": 0.035, + "step": 17926 + }, + { + "epoch": 8.334263133426314, + "grad_norm": 5.334035396575928, + "learning_rate": 9.9820809982081e-06, + "loss": 0.1021, + "step": 17927 + }, + { + "epoch": 8.334728033472803, + "grad_norm": 2.4470415115356445, + "learning_rate": 9.982079998208001e-06, + "loss": 0.0301, + "step": 17928 + }, + { + "epoch": 8.335192933519293, + "grad_norm": 4.942739963531494, + "learning_rate": 9.9820789982079e-06, + "loss": 0.0483, + "step": 17929 + }, + { + "epoch": 8.335657833565783, + "grad_norm": 5.763735771179199, + "learning_rate": 9.9820779982078e-06, + "loss": 0.0793, + "step": 17930 + }, + { + "epoch": 8.336122733612273, + "grad_norm": 4.6430840492248535, + "learning_rate": 9.9820769982077e-06, + "loss": 0.1067, + "step": 17931 + }, + { + "epoch": 8.336587633658763, + "grad_norm": 1.9204503297805786, + "learning_rate": 9.9820759982076e-06, + "loss": 0.0282, + "step": 17932 + }, + { + "epoch": 8.337052533705254, + "grad_norm": 4.8315253257751465, + "learning_rate": 9.982074998207501e-06, + "loss": 0.1201, + "step": 17933 + }, + { + "epoch": 8.337517433751744, + "grad_norm": 5.370585918426514, + "learning_rate": 9.982073998207402e-06, + "loss": 0.0909, + "step": 17934 + }, + { + "epoch": 8.337982333798234, + "grad_norm": 4.320861339569092, + "learning_rate": 9.9820729982073e-06, + "loss": 0.1077, + "step": 17935 + }, + { + "epoch": 8.338447233844724, + "grad_norm": 4.23004150390625, + "learning_rate": 9.982071998207201e-06, + "loss": 0.1007, + "step": 17936 + }, + { + "epoch": 8.338912133891213, + "grad_norm": 6.955672740936279, + "learning_rate": 9.9820709982071e-06, + "loss": 0.1824, + "step": 17937 + }, + { + "epoch": 8.339377033937703, + "grad_norm": 9.145277976989746, + "learning_rate": 9.982069998207e-06, + "loss": 0.1626, + "step": 17938 + }, + { + "epoch": 8.339841933984193, + "grad_norm": 6.927888870239258, + "learning_rate": 9.982068998206901e-06, + "loss": 0.167, + "step": 17939 + }, + { + "epoch": 8.340306834030683, + "grad_norm": 5.791351318359375, + "learning_rate": 9.9820679982068e-06, + "loss": 0.2163, + "step": 17940 + }, + { + "epoch": 8.340771734077173, + "grad_norm": 6.714727401733398, + "learning_rate": 9.982066998206701e-06, + "loss": 0.2134, + "step": 17941 + }, + { + "epoch": 8.341236634123664, + "grad_norm": 7.59462308883667, + "learning_rate": 9.9820659982066e-06, + "loss": 0.256, + "step": 17942 + }, + { + "epoch": 8.341701534170154, + "grad_norm": 5.787370204925537, + "learning_rate": 9.9820649982065e-06, + "loss": 0.1763, + "step": 17943 + }, + { + "epoch": 8.342166434216644, + "grad_norm": 5.593390464782715, + "learning_rate": 9.9820639982064e-06, + "loss": 0.1565, + "step": 17944 + }, + { + "epoch": 8.342631334263134, + "grad_norm": 4.519514560699463, + "learning_rate": 9.9820629982063e-06, + "loss": 0.0835, + "step": 17945 + }, + { + "epoch": 8.343096234309623, + "grad_norm": 6.6828484535217285, + "learning_rate": 9.982061998206201e-06, + "loss": 0.1216, + "step": 17946 + }, + { + "epoch": 8.343561134356113, + "grad_norm": 2.0995938777923584, + "learning_rate": 9.9820609982061e-06, + "loss": 0.0251, + "step": 17947 + }, + { + "epoch": 8.344026034402603, + "grad_norm": 2.6778013706207275, + "learning_rate": 9.982059998206e-06, + "loss": 0.0356, + "step": 17948 + }, + { + "epoch": 8.344490934449093, + "grad_norm": 3.2174904346466064, + "learning_rate": 9.9820589982059e-06, + "loss": 0.0386, + "step": 17949 + }, + { + "epoch": 8.344955834495583, + "grad_norm": 7.437867164611816, + "learning_rate": 9.9820579982058e-06, + "loss": 0.1652, + "step": 17950 + }, + { + "epoch": 8.345420734542074, + "grad_norm": 5.306356430053711, + "learning_rate": 9.982056998205701e-06, + "loss": 0.0654, + "step": 17951 + }, + { + "epoch": 8.345885634588564, + "grad_norm": 4.877562046051025, + "learning_rate": 9.982055998205602e-06, + "loss": 0.1721, + "step": 17952 + }, + { + "epoch": 8.346350534635054, + "grad_norm": 2.8176283836364746, + "learning_rate": 9.9820549982055e-06, + "loss": 0.0478, + "step": 17953 + }, + { + "epoch": 8.346815434681544, + "grad_norm": 4.105783939361572, + "learning_rate": 9.982053998205401e-06, + "loss": 0.0638, + "step": 17954 + }, + { + "epoch": 8.347280334728033, + "grad_norm": 3.6355559825897217, + "learning_rate": 9.9820529982053e-06, + "loss": 0.0503, + "step": 17955 + }, + { + "epoch": 8.347745234774523, + "grad_norm": 6.302964210510254, + "learning_rate": 9.9820519982052e-06, + "loss": 0.1891, + "step": 17956 + }, + { + "epoch": 8.348210134821013, + "grad_norm": 6.373645305633545, + "learning_rate": 9.982050998205101e-06, + "loss": 0.1317, + "step": 17957 + }, + { + "epoch": 8.348675034867503, + "grad_norm": 5.042356491088867, + "learning_rate": 9.982049998205e-06, + "loss": 0.1021, + "step": 17958 + }, + { + "epoch": 8.349139934913994, + "grad_norm": 2.460587739944458, + "learning_rate": 9.982048998204901e-06, + "loss": 0.044, + "step": 17959 + }, + { + "epoch": 8.349604834960484, + "grad_norm": 3.444770336151123, + "learning_rate": 9.9820479982048e-06, + "loss": 0.0614, + "step": 17960 + }, + { + "epoch": 8.350069735006974, + "grad_norm": 4.697824001312256, + "learning_rate": 9.9820469982047e-06, + "loss": 0.1036, + "step": 17961 + }, + { + "epoch": 8.350534635053464, + "grad_norm": 5.8426971435546875, + "learning_rate": 9.9820459982046e-06, + "loss": 0.1576, + "step": 17962 + }, + { + "epoch": 8.350999535099954, + "grad_norm": 7.400531768798828, + "learning_rate": 9.9820449982045e-06, + "loss": 0.1389, + "step": 17963 + }, + { + "epoch": 8.351464435146443, + "grad_norm": 7.3853535652160645, + "learning_rate": 9.982043998204401e-06, + "loss": 0.1607, + "step": 17964 + }, + { + "epoch": 8.351929335192933, + "grad_norm": 3.948387384414673, + "learning_rate": 9.982042998204302e-06, + "loss": 0.0617, + "step": 17965 + }, + { + "epoch": 8.352394235239423, + "grad_norm": 7.599775791168213, + "learning_rate": 9.9820419982042e-06, + "loss": 0.1559, + "step": 17966 + }, + { + "epoch": 8.352859135285913, + "grad_norm": 5.734162330627441, + "learning_rate": 9.982040998204101e-06, + "loss": 0.1451, + "step": 17967 + }, + { + "epoch": 8.353324035332404, + "grad_norm": 3.907158374786377, + "learning_rate": 9.982039998204e-06, + "loss": 0.0734, + "step": 17968 + }, + { + "epoch": 8.353788935378894, + "grad_norm": 6.831973552703857, + "learning_rate": 9.982038998203901e-06, + "loss": 0.3124, + "step": 17969 + }, + { + "epoch": 8.354253835425384, + "grad_norm": 5.47298526763916, + "learning_rate": 9.982037998203802e-06, + "loss": 0.1286, + "step": 17970 + }, + { + "epoch": 8.354718735471874, + "grad_norm": 5.925434112548828, + "learning_rate": 9.9820369982037e-06, + "loss": 0.1587, + "step": 17971 + }, + { + "epoch": 8.355183635518364, + "grad_norm": 1.777367115020752, + "learning_rate": 9.982035998203601e-06, + "loss": 0.0159, + "step": 17972 + }, + { + "epoch": 8.355648535564853, + "grad_norm": 6.568455696105957, + "learning_rate": 9.9820349982035e-06, + "loss": 0.1656, + "step": 17973 + }, + { + "epoch": 8.356113435611343, + "grad_norm": 5.2778778076171875, + "learning_rate": 9.9820339982034e-06, + "loss": 0.1481, + "step": 17974 + }, + { + "epoch": 8.356578335657833, + "grad_norm": 8.021553039550781, + "learning_rate": 9.9820329982033e-06, + "loss": 0.2288, + "step": 17975 + }, + { + "epoch": 8.357043235704323, + "grad_norm": 2.7155985832214355, + "learning_rate": 9.9820319982032e-06, + "loss": 0.0361, + "step": 17976 + }, + { + "epoch": 8.357508135750814, + "grad_norm": 3.4716081619262695, + "learning_rate": 9.982030998203101e-06, + "loss": 0.0451, + "step": 17977 + }, + { + "epoch": 8.357973035797304, + "grad_norm": 8.01447868347168, + "learning_rate": 9.982029998203002e-06, + "loss": 0.1051, + "step": 17978 + }, + { + "epoch": 8.358437935843794, + "grad_norm": 5.952157497406006, + "learning_rate": 9.9820289982029e-06, + "loss": 0.0935, + "step": 17979 + }, + { + "epoch": 8.358902835890284, + "grad_norm": 6.4093098640441895, + "learning_rate": 9.9820279982028e-06, + "loss": 0.1989, + "step": 17980 + }, + { + "epoch": 8.359367735936774, + "grad_norm": 2.748955011367798, + "learning_rate": 9.9820269982027e-06, + "loss": 0.0446, + "step": 17981 + }, + { + "epoch": 8.359832635983263, + "grad_norm": 7.646358966827393, + "learning_rate": 9.982025998202601e-06, + "loss": 0.068, + "step": 17982 + }, + { + "epoch": 8.360297536029753, + "grad_norm": 7.641932964324951, + "learning_rate": 9.982024998202502e-06, + "loss": 0.194, + "step": 17983 + }, + { + "epoch": 8.360762436076243, + "grad_norm": 4.529084205627441, + "learning_rate": 9.9820239982024e-06, + "loss": 0.1389, + "step": 17984 + }, + { + "epoch": 8.361227336122734, + "grad_norm": 5.316307067871094, + "learning_rate": 9.982022998202301e-06, + "loss": 0.0764, + "step": 17985 + }, + { + "epoch": 8.361692236169224, + "grad_norm": 5.1574320793151855, + "learning_rate": 9.9820219982022e-06, + "loss": 0.0875, + "step": 17986 + }, + { + "epoch": 8.362157136215714, + "grad_norm": 5.034584999084473, + "learning_rate": 9.9820209982021e-06, + "loss": 0.0993, + "step": 17987 + }, + { + "epoch": 8.362622036262204, + "grad_norm": 5.672256946563721, + "learning_rate": 9.982019998202e-06, + "loss": 0.1433, + "step": 17988 + }, + { + "epoch": 8.363086936308694, + "grad_norm": 6.529961585998535, + "learning_rate": 9.9820189982019e-06, + "loss": 0.1506, + "step": 17989 + }, + { + "epoch": 8.363551836355184, + "grad_norm": 5.404284477233887, + "learning_rate": 9.982017998201801e-06, + "loss": 0.0891, + "step": 17990 + }, + { + "epoch": 8.364016736401673, + "grad_norm": 9.710466384887695, + "learning_rate": 9.9820169982017e-06, + "loss": 0.0551, + "step": 17991 + }, + { + "epoch": 8.364481636448163, + "grad_norm": 6.97504997253418, + "learning_rate": 9.9820159982016e-06, + "loss": 0.1812, + "step": 17992 + }, + { + "epoch": 8.364946536494653, + "grad_norm": 6.555444717407227, + "learning_rate": 9.9820149982015e-06, + "loss": 0.1034, + "step": 17993 + }, + { + "epoch": 8.365411436541144, + "grad_norm": 3.2423439025878906, + "learning_rate": 9.9820139982014e-06, + "loss": 0.0621, + "step": 17994 + }, + { + "epoch": 8.365876336587634, + "grad_norm": 5.645316123962402, + "learning_rate": 9.982012998201301e-06, + "loss": 0.1699, + "step": 17995 + }, + { + "epoch": 8.366341236634124, + "grad_norm": 5.604521751403809, + "learning_rate": 9.982011998201202e-06, + "loss": 0.0736, + "step": 17996 + }, + { + "epoch": 8.366806136680614, + "grad_norm": 4.0115861892700195, + "learning_rate": 9.9820109982011e-06, + "loss": 0.0534, + "step": 17997 + }, + { + "epoch": 8.367271036727104, + "grad_norm": 3.641183376312256, + "learning_rate": 9.982009998201001e-06, + "loss": 0.1081, + "step": 17998 + }, + { + "epoch": 8.367735936773594, + "grad_norm": 5.219658851623535, + "learning_rate": 9.9820089982009e-06, + "loss": 0.0938, + "step": 17999 + }, + { + "epoch": 8.368200836820083, + "grad_norm": 2.7791748046875, + "learning_rate": 9.982007998200801e-06, + "loss": 0.0474, + "step": 18000 + }, + { + "epoch": 8.368665736866573, + "grad_norm": 2.1940033435821533, + "learning_rate": 9.9820069982007e-06, + "loss": 0.0195, + "step": 18001 + }, + { + "epoch": 8.369130636913063, + "grad_norm": 4.545984268188477, + "learning_rate": 9.9820059982006e-06, + "loss": 0.0593, + "step": 18002 + }, + { + "epoch": 8.369595536959554, + "grad_norm": 3.21734356880188, + "learning_rate": 9.982004998200501e-06, + "loss": 0.0626, + "step": 18003 + }, + { + "epoch": 8.370060437006044, + "grad_norm": 5.728819370269775, + "learning_rate": 9.9820039982004e-06, + "loss": 0.1471, + "step": 18004 + }, + { + "epoch": 8.370525337052534, + "grad_norm": 7.027886390686035, + "learning_rate": 9.9820029982003e-06, + "loss": 0.0809, + "step": 18005 + }, + { + "epoch": 8.370990237099024, + "grad_norm": 3.9660277366638184, + "learning_rate": 9.9820019982002e-06, + "loss": 0.106, + "step": 18006 + }, + { + "epoch": 8.371455137145514, + "grad_norm": 4.982625484466553, + "learning_rate": 9.9820009982001e-06, + "loss": 0.1501, + "step": 18007 + }, + { + "epoch": 8.371920037192004, + "grad_norm": 2.4259719848632812, + "learning_rate": 9.981999998200001e-06, + "loss": 0.0269, + "step": 18008 + }, + { + "epoch": 8.372384937238493, + "grad_norm": 10.142107963562012, + "learning_rate": 9.981998998199902e-06, + "loss": 0.1522, + "step": 18009 + }, + { + "epoch": 8.372849837284983, + "grad_norm": 6.765808582305908, + "learning_rate": 9.9819979981998e-06, + "loss": 0.177, + "step": 18010 + }, + { + "epoch": 8.373314737331473, + "grad_norm": 4.686108589172363, + "learning_rate": 9.981996998199701e-06, + "loss": 0.1261, + "step": 18011 + }, + { + "epoch": 8.373779637377964, + "grad_norm": 8.250368118286133, + "learning_rate": 9.9819959981996e-06, + "loss": 0.1042, + "step": 18012 + }, + { + "epoch": 8.374244537424454, + "grad_norm": 5.408681869506836, + "learning_rate": 9.9819949981995e-06, + "loss": 0.0433, + "step": 18013 + }, + { + "epoch": 8.374709437470944, + "grad_norm": 2.969316005706787, + "learning_rate": 9.9819939981994e-06, + "loss": 0.0278, + "step": 18014 + }, + { + "epoch": 8.375174337517434, + "grad_norm": 4.90857458114624, + "learning_rate": 9.9819929981993e-06, + "loss": 0.1083, + "step": 18015 + }, + { + "epoch": 8.375639237563924, + "grad_norm": 5.480133533477783, + "learning_rate": 9.981991998199201e-06, + "loss": 0.0632, + "step": 18016 + }, + { + "epoch": 8.376104137610414, + "grad_norm": 3.9783689975738525, + "learning_rate": 9.9819909981991e-06, + "loss": 0.0445, + "step": 18017 + }, + { + "epoch": 8.376569037656903, + "grad_norm": 6.549678325653076, + "learning_rate": 9.981989998199001e-06, + "loss": 0.2596, + "step": 18018 + }, + { + "epoch": 8.377033937703393, + "grad_norm": 7.3195881843566895, + "learning_rate": 9.9819889981989e-06, + "loss": 0.262, + "step": 18019 + }, + { + "epoch": 8.377498837749883, + "grad_norm": 5.561860084533691, + "learning_rate": 9.9819879981988e-06, + "loss": 0.0725, + "step": 18020 + }, + { + "epoch": 8.377963737796374, + "grad_norm": 4.793115139007568, + "learning_rate": 9.981986998198701e-06, + "loss": 0.1404, + "step": 18021 + }, + { + "epoch": 8.378428637842864, + "grad_norm": 5.592365264892578, + "learning_rate": 9.981985998198602e-06, + "loss": 0.182, + "step": 18022 + }, + { + "epoch": 8.378893537889354, + "grad_norm": 4.214451313018799, + "learning_rate": 9.9819849981985e-06, + "loss": 0.0659, + "step": 18023 + }, + { + "epoch": 8.379358437935844, + "grad_norm": 4.749155044555664, + "learning_rate": 9.9819839981984e-06, + "loss": 0.0484, + "step": 18024 + }, + { + "epoch": 8.379823337982334, + "grad_norm": 4.176814556121826, + "learning_rate": 9.9819829981983e-06, + "loss": 0.0777, + "step": 18025 + }, + { + "epoch": 8.380288238028823, + "grad_norm": 3.9590554237365723, + "learning_rate": 9.981981998198201e-06, + "loss": 0.0616, + "step": 18026 + }, + { + "epoch": 8.380753138075313, + "grad_norm": 5.659102916717529, + "learning_rate": 9.9819809981981e-06, + "loss": 0.1521, + "step": 18027 + }, + { + "epoch": 8.381218038121803, + "grad_norm": 2.972399950027466, + "learning_rate": 9.981979998198e-06, + "loss": 0.0312, + "step": 18028 + }, + { + "epoch": 8.381682938168293, + "grad_norm": 5.9008469581604, + "learning_rate": 9.981978998197901e-06, + "loss": 0.0955, + "step": 18029 + }, + { + "epoch": 8.382147838214784, + "grad_norm": 4.615843296051025, + "learning_rate": 9.9819779981978e-06, + "loss": 0.0925, + "step": 18030 + }, + { + "epoch": 8.382612738261274, + "grad_norm": 2.1812405586242676, + "learning_rate": 9.981976998197701e-06, + "loss": 0.0203, + "step": 18031 + }, + { + "epoch": 8.383077638307764, + "grad_norm": 3.1888785362243652, + "learning_rate": 9.9819759981976e-06, + "loss": 0.06, + "step": 18032 + }, + { + "epoch": 8.383542538354254, + "grad_norm": 3.1032633781433105, + "learning_rate": 9.9819749981975e-06, + "loss": 0.0413, + "step": 18033 + }, + { + "epoch": 8.384007438400744, + "grad_norm": 5.4230499267578125, + "learning_rate": 9.981973998197401e-06, + "loss": 0.1609, + "step": 18034 + }, + { + "epoch": 8.384472338447233, + "grad_norm": 5.354151725769043, + "learning_rate": 9.9819729981973e-06, + "loss": 0.1139, + "step": 18035 + }, + { + "epoch": 8.384937238493723, + "grad_norm": 3.266555070877075, + "learning_rate": 9.981971998197201e-06, + "loss": 0.0597, + "step": 18036 + }, + { + "epoch": 8.385402138540213, + "grad_norm": 4.908863544464111, + "learning_rate": 9.9819709981971e-06, + "loss": 0.0574, + "step": 18037 + }, + { + "epoch": 8.385867038586705, + "grad_norm": 3.9089717864990234, + "learning_rate": 9.981969998197e-06, + "loss": 0.0812, + "step": 18038 + }, + { + "epoch": 8.386331938633194, + "grad_norm": 12.665962219238281, + "learning_rate": 9.981968998196901e-06, + "loss": 0.1677, + "step": 18039 + }, + { + "epoch": 8.386796838679684, + "grad_norm": 1.3825432062149048, + "learning_rate": 9.9819679981968e-06, + "loss": 0.0181, + "step": 18040 + }, + { + "epoch": 8.387261738726174, + "grad_norm": 3.867724895477295, + "learning_rate": 9.9819669981967e-06, + "loss": 0.0865, + "step": 18041 + }, + { + "epoch": 8.387726638772664, + "grad_norm": 5.419615745544434, + "learning_rate": 9.981965998196601e-06, + "loss": 0.1098, + "step": 18042 + }, + { + "epoch": 8.388191538819154, + "grad_norm": 6.6312055587768555, + "learning_rate": 9.9819649981965e-06, + "loss": 0.1394, + "step": 18043 + }, + { + "epoch": 8.388656438865643, + "grad_norm": 14.492377281188965, + "learning_rate": 9.981963998196401e-06, + "loss": 0.1193, + "step": 18044 + }, + { + "epoch": 8.389121338912133, + "grad_norm": 4.755809783935547, + "learning_rate": 9.9819629981963e-06, + "loss": 0.1464, + "step": 18045 + }, + { + "epoch": 8.389586238958623, + "grad_norm": 4.024484157562256, + "learning_rate": 9.9819619981962e-06, + "loss": 0.054, + "step": 18046 + }, + { + "epoch": 8.390051139005115, + "grad_norm": 7.6448774337768555, + "learning_rate": 9.981960998196101e-06, + "loss": 0.1759, + "step": 18047 + }, + { + "epoch": 8.390516039051604, + "grad_norm": 9.400723457336426, + "learning_rate": 9.981959998196e-06, + "loss": 0.1525, + "step": 18048 + }, + { + "epoch": 8.390980939098094, + "grad_norm": 6.112632751464844, + "learning_rate": 9.981958998195901e-06, + "loss": 0.1728, + "step": 18049 + }, + { + "epoch": 8.391445839144584, + "grad_norm": 1.568663477897644, + "learning_rate": 9.9819579981958e-06, + "loss": 0.018, + "step": 18050 + }, + { + "epoch": 8.391910739191074, + "grad_norm": 2.952526807785034, + "learning_rate": 9.9819569981957e-06, + "loss": 0.0515, + "step": 18051 + }, + { + "epoch": 8.392375639237564, + "grad_norm": 3.9089527130126953, + "learning_rate": 9.981955998195601e-06, + "loss": 0.0776, + "step": 18052 + }, + { + "epoch": 8.392840539284053, + "grad_norm": 3.9199154376983643, + "learning_rate": 9.981954998195502e-06, + "loss": 0.0535, + "step": 18053 + }, + { + "epoch": 8.393305439330543, + "grad_norm": 2.444357395172119, + "learning_rate": 9.981953998195401e-06, + "loss": 0.0234, + "step": 18054 + }, + { + "epoch": 8.393770339377033, + "grad_norm": 8.6087007522583, + "learning_rate": 9.981952998195302e-06, + "loss": 0.1352, + "step": 18055 + }, + { + "epoch": 8.394235239423525, + "grad_norm": 4.0398969650268555, + "learning_rate": 9.9819519981952e-06, + "loss": 0.0722, + "step": 18056 + }, + { + "epoch": 8.394700139470014, + "grad_norm": 5.246220111846924, + "learning_rate": 9.9819509981951e-06, + "loss": 0.07, + "step": 18057 + }, + { + "epoch": 8.395165039516504, + "grad_norm": 3.285977602005005, + "learning_rate": 9.981949998195e-06, + "loss": 0.048, + "step": 18058 + }, + { + "epoch": 8.395629939562994, + "grad_norm": 5.640044212341309, + "learning_rate": 9.9819489981949e-06, + "loss": 0.0908, + "step": 18059 + }, + { + "epoch": 8.396094839609484, + "grad_norm": 2.8601455688476562, + "learning_rate": 9.981947998194801e-06, + "loss": 0.0456, + "step": 18060 + }, + { + "epoch": 8.396559739655974, + "grad_norm": 8.197646141052246, + "learning_rate": 9.9819469981947e-06, + "loss": 0.2384, + "step": 18061 + }, + { + "epoch": 8.397024639702463, + "grad_norm": 4.526236534118652, + "learning_rate": 9.981945998194601e-06, + "loss": 0.0917, + "step": 18062 + }, + { + "epoch": 8.397489539748953, + "grad_norm": 5.329880714416504, + "learning_rate": 9.9819449981945e-06, + "loss": 0.1115, + "step": 18063 + }, + { + "epoch": 8.397954439795445, + "grad_norm": 4.784928798675537, + "learning_rate": 9.9819439981944e-06, + "loss": 0.0476, + "step": 18064 + }, + { + "epoch": 8.398419339841935, + "grad_norm": 5.526580333709717, + "learning_rate": 9.981942998194301e-06, + "loss": 0.1342, + "step": 18065 + }, + { + "epoch": 8.398884239888424, + "grad_norm": 6.603405475616455, + "learning_rate": 9.981941998194202e-06, + "loss": 0.1556, + "step": 18066 + }, + { + "epoch": 8.399349139934914, + "grad_norm": 4.53122091293335, + "learning_rate": 9.981940998194101e-06, + "loss": 0.0689, + "step": 18067 + }, + { + "epoch": 8.399814039981404, + "grad_norm": 4.51025390625, + "learning_rate": 9.981939998194e-06, + "loss": 0.0586, + "step": 18068 + }, + { + "epoch": 8.400278940027894, + "grad_norm": 14.38235092163086, + "learning_rate": 9.9819389981939e-06, + "loss": 0.1077, + "step": 18069 + }, + { + "epoch": 8.400743840074384, + "grad_norm": 4.076505184173584, + "learning_rate": 9.9819379981938e-06, + "loss": 0.0904, + "step": 18070 + }, + { + "epoch": 8.401208740120873, + "grad_norm": 6.105038642883301, + "learning_rate": 9.9819369981937e-06, + "loss": 0.1098, + "step": 18071 + }, + { + "epoch": 8.401673640167363, + "grad_norm": 6.280511856079102, + "learning_rate": 9.9819359981936e-06, + "loss": 0.0886, + "step": 18072 + }, + { + "epoch": 8.402138540213855, + "grad_norm": 6.281982898712158, + "learning_rate": 9.981934998193501e-06, + "loss": 0.1441, + "step": 18073 + }, + { + "epoch": 8.402603440260345, + "grad_norm": 4.132071018218994, + "learning_rate": 9.9819339981934e-06, + "loss": 0.0286, + "step": 18074 + }, + { + "epoch": 8.403068340306834, + "grad_norm": 6.221817970275879, + "learning_rate": 9.981932998193301e-06, + "loss": 0.1863, + "step": 18075 + }, + { + "epoch": 8.403533240353324, + "grad_norm": 5.161725997924805, + "learning_rate": 9.9819319981932e-06, + "loss": 0.1003, + "step": 18076 + }, + { + "epoch": 8.403998140399814, + "grad_norm": 1.5484658479690552, + "learning_rate": 9.9819309981931e-06, + "loss": 0.019, + "step": 18077 + }, + { + "epoch": 8.404463040446304, + "grad_norm": 5.874402046203613, + "learning_rate": 9.981929998193001e-06, + "loss": 0.1311, + "step": 18078 + }, + { + "epoch": 8.404927940492794, + "grad_norm": 4.87841272354126, + "learning_rate": 9.9819289981929e-06, + "loss": 0.0762, + "step": 18079 + }, + { + "epoch": 8.405392840539283, + "grad_norm": 7.594050884246826, + "learning_rate": 9.981927998192801e-06, + "loss": 0.1424, + "step": 18080 + }, + { + "epoch": 8.405857740585773, + "grad_norm": 3.19722843170166, + "learning_rate": 9.9819269981927e-06, + "loss": 0.0446, + "step": 18081 + }, + { + "epoch": 8.406322640632265, + "grad_norm": 6.499549388885498, + "learning_rate": 9.9819259981926e-06, + "loss": 0.13, + "step": 18082 + }, + { + "epoch": 8.406787540678755, + "grad_norm": 4.156444549560547, + "learning_rate": 9.9819249981925e-06, + "loss": 0.0551, + "step": 18083 + }, + { + "epoch": 8.407252440725244, + "grad_norm": 3.5719757080078125, + "learning_rate": 9.9819239981924e-06, + "loss": 0.0492, + "step": 18084 + }, + { + "epoch": 8.407717340771734, + "grad_norm": 4.642876148223877, + "learning_rate": 9.981922998192301e-06, + "loss": 0.0756, + "step": 18085 + }, + { + "epoch": 8.408182240818224, + "grad_norm": 5.560910224914551, + "learning_rate": 9.981921998192202e-06, + "loss": 0.1551, + "step": 18086 + }, + { + "epoch": 8.408647140864714, + "grad_norm": 5.30568265914917, + "learning_rate": 9.9819209981921e-06, + "loss": 0.1184, + "step": 18087 + }, + { + "epoch": 8.409112040911204, + "grad_norm": 4.267660140991211, + "learning_rate": 9.981919998192e-06, + "loss": 0.0531, + "step": 18088 + }, + { + "epoch": 8.409576940957693, + "grad_norm": 6.162161350250244, + "learning_rate": 9.9819189981919e-06, + "loss": 0.1093, + "step": 18089 + }, + { + "epoch": 8.410041841004183, + "grad_norm": 8.50989818572998, + "learning_rate": 9.9819179981918e-06, + "loss": 0.1188, + "step": 18090 + }, + { + "epoch": 8.410506741050675, + "grad_norm": 7.044781684875488, + "learning_rate": 9.981916998191701e-06, + "loss": 0.283, + "step": 18091 + }, + { + "epoch": 8.410971641097165, + "grad_norm": 7.298428058624268, + "learning_rate": 9.9819159981916e-06, + "loss": 0.221, + "step": 18092 + }, + { + "epoch": 8.411436541143654, + "grad_norm": 4.415063381195068, + "learning_rate": 9.981914998191501e-06, + "loss": 0.0658, + "step": 18093 + }, + { + "epoch": 8.411901441190144, + "grad_norm": 4.0336833000183105, + "learning_rate": 9.9819139981914e-06, + "loss": 0.0847, + "step": 18094 + }, + { + "epoch": 8.412366341236634, + "grad_norm": 3.1693646907806396, + "learning_rate": 9.9819129981913e-06, + "loss": 0.0474, + "step": 18095 + }, + { + "epoch": 8.412831241283124, + "grad_norm": 3.967191696166992, + "learning_rate": 9.9819119981912e-06, + "loss": 0.0538, + "step": 18096 + }, + { + "epoch": 8.413296141329614, + "grad_norm": 5.394720554351807, + "learning_rate": 9.9819109981911e-06, + "loss": 0.0346, + "step": 18097 + }, + { + "epoch": 8.413761041376103, + "grad_norm": 2.857097864151001, + "learning_rate": 9.981909998191001e-06, + "loss": 0.037, + "step": 18098 + }, + { + "epoch": 8.414225941422593, + "grad_norm": 3.453214645385742, + "learning_rate": 9.9819089981909e-06, + "loss": 0.0566, + "step": 18099 + }, + { + "epoch": 8.414690841469085, + "grad_norm": 4.392127990722656, + "learning_rate": 9.9819079981908e-06, + "loss": 0.135, + "step": 18100 + }, + { + "epoch": 8.415155741515575, + "grad_norm": 3.037492036819458, + "learning_rate": 9.9819069981907e-06, + "loss": 0.0363, + "step": 18101 + }, + { + "epoch": 8.415620641562064, + "grad_norm": 7.529816627502441, + "learning_rate": 9.9819059981906e-06, + "loss": 0.1616, + "step": 18102 + }, + { + "epoch": 8.416085541608554, + "grad_norm": 8.900471687316895, + "learning_rate": 9.981904998190501e-06, + "loss": 0.2287, + "step": 18103 + }, + { + "epoch": 8.416550441655044, + "grad_norm": 8.585740089416504, + "learning_rate": 9.981903998190402e-06, + "loss": 0.2485, + "step": 18104 + }, + { + "epoch": 8.417015341701534, + "grad_norm": 5.602853775024414, + "learning_rate": 9.9819029981903e-06, + "loss": 0.0365, + "step": 18105 + }, + { + "epoch": 8.417480241748024, + "grad_norm": 3.7429897785186768, + "learning_rate": 9.981901998190201e-06, + "loss": 0.08, + "step": 18106 + }, + { + "epoch": 8.417945141794513, + "grad_norm": 6.400432109832764, + "learning_rate": 9.9819009981901e-06, + "loss": 0.1201, + "step": 18107 + }, + { + "epoch": 8.418410041841003, + "grad_norm": 6.501263618469238, + "learning_rate": 9.98189999819e-06, + "loss": 0.2935, + "step": 18108 + }, + { + "epoch": 8.418874941887495, + "grad_norm": 4.974551677703857, + "learning_rate": 9.9818989981899e-06, + "loss": 0.1108, + "step": 18109 + }, + { + "epoch": 8.419339841933985, + "grad_norm": 7.278332710266113, + "learning_rate": 9.9818979981898e-06, + "loss": 0.136, + "step": 18110 + }, + { + "epoch": 8.419804741980474, + "grad_norm": 5.346593379974365, + "learning_rate": 9.981896998189701e-06, + "loss": 0.0498, + "step": 18111 + }, + { + "epoch": 8.420269642026964, + "grad_norm": 1.1914705038070679, + "learning_rate": 9.9818959981896e-06, + "loss": 0.008, + "step": 18112 + }, + { + "epoch": 8.420734542073454, + "grad_norm": 4.985928058624268, + "learning_rate": 9.9818949981895e-06, + "loss": 0.0918, + "step": 18113 + }, + { + "epoch": 8.421199442119944, + "grad_norm": 5.89387845993042, + "learning_rate": 9.9818939981894e-06, + "loss": 0.1642, + "step": 18114 + }, + { + "epoch": 8.421664342166434, + "grad_norm": 5.100769519805908, + "learning_rate": 9.9818929981893e-06, + "loss": 0.0875, + "step": 18115 + }, + { + "epoch": 8.422129242212923, + "grad_norm": 8.2694673538208, + "learning_rate": 9.981891998189201e-06, + "loss": 0.0946, + "step": 18116 + }, + { + "epoch": 8.422594142259415, + "grad_norm": 4.626702308654785, + "learning_rate": 9.981890998189102e-06, + "loss": 0.0733, + "step": 18117 + }, + { + "epoch": 8.423059042305905, + "grad_norm": 5.373080253601074, + "learning_rate": 9.981889998189e-06, + "loss": 0.0839, + "step": 18118 + }, + { + "epoch": 8.423523942352395, + "grad_norm": 9.479846000671387, + "learning_rate": 9.981888998188901e-06, + "loss": 0.304, + "step": 18119 + }, + { + "epoch": 8.423988842398884, + "grad_norm": 4.937399864196777, + "learning_rate": 9.9818879981888e-06, + "loss": 0.0867, + "step": 18120 + }, + { + "epoch": 8.424453742445374, + "grad_norm": 4.6154608726501465, + "learning_rate": 9.981886998188701e-06, + "loss": 0.078, + "step": 18121 + }, + { + "epoch": 8.424918642491864, + "grad_norm": 7.863913536071777, + "learning_rate": 9.981885998188602e-06, + "loss": 0.1287, + "step": 18122 + }, + { + "epoch": 8.425383542538354, + "grad_norm": 3.51483416557312, + "learning_rate": 9.9818849981885e-06, + "loss": 0.0775, + "step": 18123 + }, + { + "epoch": 8.425848442584844, + "grad_norm": 3.30969500541687, + "learning_rate": 9.981883998188401e-06, + "loss": 0.0252, + "step": 18124 + }, + { + "epoch": 8.426313342631333, + "grad_norm": 5.005225658416748, + "learning_rate": 9.9818829981883e-06, + "loss": 0.0773, + "step": 18125 + }, + { + "epoch": 8.426778242677825, + "grad_norm": 2.102898120880127, + "learning_rate": 9.9818819981882e-06, + "loss": 0.0214, + "step": 18126 + }, + { + "epoch": 8.427243142724315, + "grad_norm": 3.8309125900268555, + "learning_rate": 9.9818809981881e-06, + "loss": 0.0416, + "step": 18127 + }, + { + "epoch": 8.427708042770805, + "grad_norm": 4.367971897125244, + "learning_rate": 9.981879998188e-06, + "loss": 0.0641, + "step": 18128 + }, + { + "epoch": 8.428172942817294, + "grad_norm": 3.8912429809570312, + "learning_rate": 9.981878998187901e-06, + "loss": 0.0741, + "step": 18129 + }, + { + "epoch": 8.428637842863784, + "grad_norm": 8.038507461547852, + "learning_rate": 9.981877998187802e-06, + "loss": 0.2127, + "step": 18130 + }, + { + "epoch": 8.429102742910274, + "grad_norm": 1.6903128623962402, + "learning_rate": 9.9818769981877e-06, + "loss": 0.0179, + "step": 18131 + }, + { + "epoch": 8.429567642956764, + "grad_norm": 9.525324821472168, + "learning_rate": 9.9818759981876e-06, + "loss": 0.3145, + "step": 18132 + }, + { + "epoch": 8.430032543003254, + "grad_norm": 9.606398582458496, + "learning_rate": 9.9818749981875e-06, + "loss": 0.195, + "step": 18133 + }, + { + "epoch": 8.430497443049743, + "grad_norm": 7.78630256652832, + "learning_rate": 9.981873998187401e-06, + "loss": 0.1546, + "step": 18134 + }, + { + "epoch": 8.430962343096235, + "grad_norm": 6.759040355682373, + "learning_rate": 9.981872998187302e-06, + "loss": 0.3114, + "step": 18135 + }, + { + "epoch": 8.431427243142725, + "grad_norm": 2.7012743949890137, + "learning_rate": 9.9818719981872e-06, + "loss": 0.0583, + "step": 18136 + }, + { + "epoch": 8.431892143189215, + "grad_norm": 5.603825569152832, + "learning_rate": 9.981870998187101e-06, + "loss": 0.1456, + "step": 18137 + }, + { + "epoch": 8.432357043235704, + "grad_norm": 8.572587966918945, + "learning_rate": 9.981869998187e-06, + "loss": 0.2042, + "step": 18138 + }, + { + "epoch": 8.432821943282194, + "grad_norm": 5.9296770095825195, + "learning_rate": 9.981868998186901e-06, + "loss": 0.0792, + "step": 18139 + }, + { + "epoch": 8.433286843328684, + "grad_norm": 2.2343294620513916, + "learning_rate": 9.9818679981868e-06, + "loss": 0.0278, + "step": 18140 + }, + { + "epoch": 8.433751743375174, + "grad_norm": 10.052166938781738, + "learning_rate": 9.9818669981867e-06, + "loss": 0.174, + "step": 18141 + }, + { + "epoch": 8.434216643421664, + "grad_norm": 3.307678461074829, + "learning_rate": 9.981865998186601e-06, + "loss": 0.0537, + "step": 18142 + }, + { + "epoch": 8.434681543468155, + "grad_norm": 4.372839450836182, + "learning_rate": 9.9818649981865e-06, + "loss": 0.034, + "step": 18143 + }, + { + "epoch": 8.435146443514645, + "grad_norm": 8.070226669311523, + "learning_rate": 9.9818639981864e-06, + "loss": 0.3262, + "step": 18144 + }, + { + "epoch": 8.435611343561135, + "grad_norm": 3.830005407333374, + "learning_rate": 9.9818629981863e-06, + "loss": 0.0976, + "step": 18145 + }, + { + "epoch": 8.436076243607625, + "grad_norm": 3.5818042755126953, + "learning_rate": 9.9818619981862e-06, + "loss": 0.0503, + "step": 18146 + }, + { + "epoch": 8.436541143654114, + "grad_norm": 4.748481273651123, + "learning_rate": 9.981860998186101e-06, + "loss": 0.1456, + "step": 18147 + }, + { + "epoch": 8.437006043700604, + "grad_norm": 3.32167649269104, + "learning_rate": 9.981859998186002e-06, + "loss": 0.0586, + "step": 18148 + }, + { + "epoch": 8.437470943747094, + "grad_norm": 3.9286603927612305, + "learning_rate": 9.9818589981859e-06, + "loss": 0.0467, + "step": 18149 + }, + { + "epoch": 8.437935843793584, + "grad_norm": 4.526516914367676, + "learning_rate": 9.981857998185801e-06, + "loss": 0.0607, + "step": 18150 + }, + { + "epoch": 8.438400743840074, + "grad_norm": 2.023956060409546, + "learning_rate": 9.9818569981857e-06, + "loss": 0.0186, + "step": 18151 + }, + { + "epoch": 8.438865643886565, + "grad_norm": 5.507802963256836, + "learning_rate": 9.981855998185601e-06, + "loss": 0.1392, + "step": 18152 + }, + { + "epoch": 8.439330543933055, + "grad_norm": 5.491048336029053, + "learning_rate": 9.9818549981855e-06, + "loss": 0.1074, + "step": 18153 + }, + { + "epoch": 8.439795443979545, + "grad_norm": 5.6846113204956055, + "learning_rate": 9.9818539981854e-06, + "loss": 0.1159, + "step": 18154 + }, + { + "epoch": 8.440260344026035, + "grad_norm": 5.379668235778809, + "learning_rate": 9.981852998185301e-06, + "loss": 0.0517, + "step": 18155 + }, + { + "epoch": 8.440725244072524, + "grad_norm": 5.191702365875244, + "learning_rate": 9.9818519981852e-06, + "loss": 0.1971, + "step": 18156 + }, + { + "epoch": 8.441190144119014, + "grad_norm": 4.878514289855957, + "learning_rate": 9.9818509981851e-06, + "loss": 0.0954, + "step": 18157 + }, + { + "epoch": 8.441655044165504, + "grad_norm": 6.760409832000732, + "learning_rate": 9.981849998185e-06, + "loss": 0.1679, + "step": 18158 + }, + { + "epoch": 8.442119944211994, + "grad_norm": 4.119014739990234, + "learning_rate": 9.9818489981849e-06, + "loss": 0.0662, + "step": 18159 + }, + { + "epoch": 8.442584844258484, + "grad_norm": 7.125461101531982, + "learning_rate": 9.981847998184801e-06, + "loss": 0.0594, + "step": 18160 + }, + { + "epoch": 8.443049744304975, + "grad_norm": 3.5705811977386475, + "learning_rate": 9.981846998184702e-06, + "loss": 0.0334, + "step": 18161 + }, + { + "epoch": 8.443514644351465, + "grad_norm": 4.444820880889893, + "learning_rate": 9.9818459981846e-06, + "loss": 0.0837, + "step": 18162 + }, + { + "epoch": 8.443979544397955, + "grad_norm": 7.018889427185059, + "learning_rate": 9.981844998184501e-06, + "loss": 0.1251, + "step": 18163 + }, + { + "epoch": 8.444444444444445, + "grad_norm": 5.099968433380127, + "learning_rate": 9.9818439981844e-06, + "loss": 0.1173, + "step": 18164 + }, + { + "epoch": 8.444909344490934, + "grad_norm": 7.185643196105957, + "learning_rate": 9.9818429981843e-06, + "loss": 0.2507, + "step": 18165 + }, + { + "epoch": 8.445374244537424, + "grad_norm": 4.736110210418701, + "learning_rate": 9.9818419981842e-06, + "loss": 0.0958, + "step": 18166 + }, + { + "epoch": 8.445839144583914, + "grad_norm": 5.032177448272705, + "learning_rate": 9.9818409981841e-06, + "loss": 0.1053, + "step": 18167 + }, + { + "epoch": 8.446304044630404, + "grad_norm": 7.435056209564209, + "learning_rate": 9.981839998184001e-06, + "loss": 0.2166, + "step": 18168 + }, + { + "epoch": 8.446768944676894, + "grad_norm": 5.769066333770752, + "learning_rate": 9.9818389981839e-06, + "loss": 0.1229, + "step": 18169 + }, + { + "epoch": 8.447233844723385, + "grad_norm": 6.012243270874023, + "learning_rate": 9.981837998183801e-06, + "loss": 0.1661, + "step": 18170 + }, + { + "epoch": 8.447698744769875, + "grad_norm": 6.591946125030518, + "learning_rate": 9.9818369981837e-06, + "loss": 0.0964, + "step": 18171 + }, + { + "epoch": 8.448163644816365, + "grad_norm": 4.328782558441162, + "learning_rate": 9.9818359981836e-06, + "loss": 0.0756, + "step": 18172 + }, + { + "epoch": 8.448628544862855, + "grad_norm": 6.940502166748047, + "learning_rate": 9.981834998183501e-06, + "loss": 0.1699, + "step": 18173 + }, + { + "epoch": 8.449093444909344, + "grad_norm": 4.715513706207275, + "learning_rate": 9.981833998183402e-06, + "loss": 0.0549, + "step": 18174 + }, + { + "epoch": 8.449558344955834, + "grad_norm": 8.486019134521484, + "learning_rate": 9.9818329981833e-06, + "loss": 0.0836, + "step": 18175 + }, + { + "epoch": 8.450023245002324, + "grad_norm": 6.7402663230896, + "learning_rate": 9.9818319981832e-06, + "loss": 0.2578, + "step": 18176 + }, + { + "epoch": 8.450488145048814, + "grad_norm": 4.472352027893066, + "learning_rate": 9.9818309981831e-06, + "loss": 0.0308, + "step": 18177 + }, + { + "epoch": 8.450953045095304, + "grad_norm": 4.152373790740967, + "learning_rate": 9.981829998183e-06, + "loss": 0.0897, + "step": 18178 + }, + { + "epoch": 8.451417945141795, + "grad_norm": 6.402257919311523, + "learning_rate": 9.9818289981829e-06, + "loss": 0.1425, + "step": 18179 + }, + { + "epoch": 8.451882845188285, + "grad_norm": 4.808570861816406, + "learning_rate": 9.9818279981828e-06, + "loss": 0.0841, + "step": 18180 + }, + { + "epoch": 8.452347745234775, + "grad_norm": 10.373984336853027, + "learning_rate": 9.981826998182701e-06, + "loss": 0.1007, + "step": 18181 + }, + { + "epoch": 8.452812645281265, + "grad_norm": 4.422021865844727, + "learning_rate": 9.9818259981826e-06, + "loss": 0.0701, + "step": 18182 + }, + { + "epoch": 8.453277545327754, + "grad_norm": 3.8262569904327393, + "learning_rate": 9.981824998182501e-06, + "loss": 0.0961, + "step": 18183 + }, + { + "epoch": 8.453742445374244, + "grad_norm": 4.653397083282471, + "learning_rate": 9.9818239981824e-06, + "loss": 0.079, + "step": 18184 + }, + { + "epoch": 8.454207345420734, + "grad_norm": 2.593088388442993, + "learning_rate": 9.9818229981823e-06, + "loss": 0.054, + "step": 18185 + }, + { + "epoch": 8.454672245467224, + "grad_norm": 7.224581241607666, + "learning_rate": 9.981821998182201e-06, + "loss": 0.1086, + "step": 18186 + }, + { + "epoch": 8.455137145513714, + "grad_norm": 3.769646644592285, + "learning_rate": 9.9818209981821e-06, + "loss": 0.0554, + "step": 18187 + }, + { + "epoch": 8.455602045560205, + "grad_norm": 5.056246757507324, + "learning_rate": 9.981819998182001e-06, + "loss": 0.1431, + "step": 18188 + }, + { + "epoch": 8.456066945606695, + "grad_norm": 14.11413288116455, + "learning_rate": 9.9818189981819e-06, + "loss": 0.4289, + "step": 18189 + }, + { + "epoch": 8.456531845653185, + "grad_norm": 4.493190288543701, + "learning_rate": 9.9818179981818e-06, + "loss": 0.0956, + "step": 18190 + }, + { + "epoch": 8.456996745699675, + "grad_norm": 4.604454517364502, + "learning_rate": 9.981816998181701e-06, + "loss": 0.1637, + "step": 18191 + }, + { + "epoch": 8.457461645746164, + "grad_norm": 4.506017684936523, + "learning_rate": 9.9818159981816e-06, + "loss": 0.133, + "step": 18192 + }, + { + "epoch": 8.457926545792654, + "grad_norm": 2.521341562271118, + "learning_rate": 9.9818149981815e-06, + "loss": 0.0245, + "step": 18193 + }, + { + "epoch": 8.458391445839144, + "grad_norm": 5.6525187492370605, + "learning_rate": 9.981813998181401e-06, + "loss": 0.115, + "step": 18194 + }, + { + "epoch": 8.458856345885634, + "grad_norm": 5.355027675628662, + "learning_rate": 9.9818129981813e-06, + "loss": 0.118, + "step": 18195 + }, + { + "epoch": 8.459321245932125, + "grad_norm": 7.116545677185059, + "learning_rate": 9.981811998181201e-06, + "loss": 0.1619, + "step": 18196 + }, + { + "epoch": 8.459786145978615, + "grad_norm": 7.284702777862549, + "learning_rate": 9.9818109981811e-06, + "loss": 0.1802, + "step": 18197 + }, + { + "epoch": 8.460251046025105, + "grad_norm": 6.286590576171875, + "learning_rate": 9.981809998181e-06, + "loss": 0.1328, + "step": 18198 + }, + { + "epoch": 8.460715946071595, + "grad_norm": 6.4299092292785645, + "learning_rate": 9.981808998180901e-06, + "loss": 0.1676, + "step": 18199 + }, + { + "epoch": 8.461180846118085, + "grad_norm": 8.775668144226074, + "learning_rate": 9.9818079981808e-06, + "loss": 0.203, + "step": 18200 + }, + { + "epoch": 8.461645746164574, + "grad_norm": 5.808422565460205, + "learning_rate": 9.981806998180701e-06, + "loss": 0.1055, + "step": 18201 + }, + { + "epoch": 8.462110646211064, + "grad_norm": 3.599274158477783, + "learning_rate": 9.9818059981806e-06, + "loss": 0.0758, + "step": 18202 + }, + { + "epoch": 8.462575546257554, + "grad_norm": 4.926008701324463, + "learning_rate": 9.9818049981805e-06, + "loss": 0.0777, + "step": 18203 + }, + { + "epoch": 8.463040446304044, + "grad_norm": 5.323861122131348, + "learning_rate": 9.981803998180401e-06, + "loss": 0.1218, + "step": 18204 + }, + { + "epoch": 8.463505346350535, + "grad_norm": 5.438888072967529, + "learning_rate": 9.9818029981803e-06, + "loss": 0.117, + "step": 18205 + }, + { + "epoch": 8.463970246397025, + "grad_norm": 5.8433027267456055, + "learning_rate": 9.981801998180201e-06, + "loss": 0.1642, + "step": 18206 + }, + { + "epoch": 8.464435146443515, + "grad_norm": 3.672678232192993, + "learning_rate": 9.981800998180102e-06, + "loss": 0.0357, + "step": 18207 + }, + { + "epoch": 8.464900046490005, + "grad_norm": 5.972898483276367, + "learning_rate": 9.98179999818e-06, + "loss": 0.1467, + "step": 18208 + }, + { + "epoch": 8.465364946536495, + "grad_norm": 4.486330986022949, + "learning_rate": 9.9817989981799e-06, + "loss": 0.0555, + "step": 18209 + }, + { + "epoch": 8.465829846582984, + "grad_norm": 5.597640037536621, + "learning_rate": 9.9817979981798e-06, + "loss": 0.136, + "step": 18210 + }, + { + "epoch": 8.466294746629474, + "grad_norm": 4.748443126678467, + "learning_rate": 9.9817969981797e-06, + "loss": 0.1079, + "step": 18211 + }, + { + "epoch": 8.466759646675964, + "grad_norm": 5.231080055236816, + "learning_rate": 9.981795998179601e-06, + "loss": 0.1747, + "step": 18212 + }, + { + "epoch": 8.467224546722454, + "grad_norm": 6.989020347595215, + "learning_rate": 9.9817949981795e-06, + "loss": 0.2319, + "step": 18213 + }, + { + "epoch": 8.467689446768945, + "grad_norm": 4.881045341491699, + "learning_rate": 9.981793998179401e-06, + "loss": 0.1117, + "step": 18214 + }, + { + "epoch": 8.468154346815435, + "grad_norm": 7.202824115753174, + "learning_rate": 9.9817929981793e-06, + "loss": 0.2366, + "step": 18215 + }, + { + "epoch": 8.468619246861925, + "grad_norm": 1.9766910076141357, + "learning_rate": 9.9817919981792e-06, + "loss": 0.0183, + "step": 18216 + }, + { + "epoch": 8.469084146908415, + "grad_norm": 8.240848541259766, + "learning_rate": 9.981790998179101e-06, + "loss": 0.1556, + "step": 18217 + }, + { + "epoch": 8.469549046954905, + "grad_norm": 2.8528363704681396, + "learning_rate": 9.981789998179002e-06, + "loss": 0.0438, + "step": 18218 + }, + { + "epoch": 8.470013947001394, + "grad_norm": 4.904710292816162, + "learning_rate": 9.981788998178901e-06, + "loss": 0.1044, + "step": 18219 + }, + { + "epoch": 8.470478847047884, + "grad_norm": 5.282170295715332, + "learning_rate": 9.9817879981788e-06, + "loss": 0.1743, + "step": 18220 + }, + { + "epoch": 8.470943747094374, + "grad_norm": 9.328033447265625, + "learning_rate": 9.9817869981787e-06, + "loss": 0.0383, + "step": 18221 + }, + { + "epoch": 8.471408647140866, + "grad_norm": 3.69378924369812, + "learning_rate": 9.9817859981786e-06, + "loss": 0.0556, + "step": 18222 + }, + { + "epoch": 8.471873547187355, + "grad_norm": 6.409320831298828, + "learning_rate": 9.9817849981785e-06, + "loss": 0.0987, + "step": 18223 + }, + { + "epoch": 8.472338447233845, + "grad_norm": 6.761849403381348, + "learning_rate": 9.981783998178401e-06, + "loss": 0.1044, + "step": 18224 + }, + { + "epoch": 8.472803347280335, + "grad_norm": 6.373859405517578, + "learning_rate": 9.981782998178302e-06, + "loss": 0.1346, + "step": 18225 + }, + { + "epoch": 8.473268247326825, + "grad_norm": 9.943288803100586, + "learning_rate": 9.9817819981782e-06, + "loss": 0.2585, + "step": 18226 + }, + { + "epoch": 8.473733147373315, + "grad_norm": 4.878108501434326, + "learning_rate": 9.981780998178101e-06, + "loss": 0.0442, + "step": 18227 + }, + { + "epoch": 8.474198047419804, + "grad_norm": 5.540174961090088, + "learning_rate": 9.981779998178e-06, + "loss": 0.1498, + "step": 18228 + }, + { + "epoch": 8.474662947466294, + "grad_norm": 5.2048749923706055, + "learning_rate": 9.9817789981779e-06, + "loss": 0.1027, + "step": 18229 + }, + { + "epoch": 8.475127847512784, + "grad_norm": 4.804382801055908, + "learning_rate": 9.981777998177801e-06, + "loss": 0.0896, + "step": 18230 + }, + { + "epoch": 8.475592747559276, + "grad_norm": 5.897213935852051, + "learning_rate": 9.9817769981777e-06, + "loss": 0.1647, + "step": 18231 + }, + { + "epoch": 8.476057647605765, + "grad_norm": 5.82843017578125, + "learning_rate": 9.981775998177601e-06, + "loss": 0.1329, + "step": 18232 + }, + { + "epoch": 8.476522547652255, + "grad_norm": 8.313645362854004, + "learning_rate": 9.9817749981775e-06, + "loss": 0.0941, + "step": 18233 + }, + { + "epoch": 8.476987447698745, + "grad_norm": 4.409976959228516, + "learning_rate": 9.9817739981774e-06, + "loss": 0.0704, + "step": 18234 + }, + { + "epoch": 8.477452347745235, + "grad_norm": 3.0466649532318115, + "learning_rate": 9.9817729981773e-06, + "loss": 0.0369, + "step": 18235 + }, + { + "epoch": 8.477917247791725, + "grad_norm": 5.502004146575928, + "learning_rate": 9.9817719981772e-06, + "loss": 0.0957, + "step": 18236 + }, + { + "epoch": 8.478382147838214, + "grad_norm": 7.048101902008057, + "learning_rate": 9.981770998177101e-06, + "loss": 0.1779, + "step": 18237 + }, + { + "epoch": 8.478847047884704, + "grad_norm": 6.715924263000488, + "learning_rate": 9.981769998177002e-06, + "loss": 0.179, + "step": 18238 + }, + { + "epoch": 8.479311947931194, + "grad_norm": 2.0290610790252686, + "learning_rate": 9.9817689981769e-06, + "loss": 0.0148, + "step": 18239 + }, + { + "epoch": 8.479776847977686, + "grad_norm": 5.022523403167725, + "learning_rate": 9.9817679981768e-06, + "loss": 0.0447, + "step": 18240 + }, + { + "epoch": 8.480241748024175, + "grad_norm": 7.458309173583984, + "learning_rate": 9.9817669981767e-06, + "loss": 0.1623, + "step": 18241 + }, + { + "epoch": 8.480706648070665, + "grad_norm": 6.965104103088379, + "learning_rate": 9.981765998176601e-06, + "loss": 0.3329, + "step": 18242 + }, + { + "epoch": 8.481171548117155, + "grad_norm": 7.062830448150635, + "learning_rate": 9.981764998176502e-06, + "loss": 0.1447, + "step": 18243 + }, + { + "epoch": 8.481636448163645, + "grad_norm": 4.861510276794434, + "learning_rate": 9.9817639981764e-06, + "loss": 0.1015, + "step": 18244 + }, + { + "epoch": 8.482101348210135, + "grad_norm": 7.412729263305664, + "learning_rate": 9.981762998176301e-06, + "loss": 0.1497, + "step": 18245 + }, + { + "epoch": 8.482566248256624, + "grad_norm": 3.0721018314361572, + "learning_rate": 9.9817619981762e-06, + "loss": 0.0343, + "step": 18246 + }, + { + "epoch": 8.483031148303114, + "grad_norm": 5.886314392089844, + "learning_rate": 9.9817609981761e-06, + "loss": 0.1243, + "step": 18247 + }, + { + "epoch": 8.483496048349604, + "grad_norm": 7.63297176361084, + "learning_rate": 9.981759998176e-06, + "loss": 0.2215, + "step": 18248 + }, + { + "epoch": 8.483960948396096, + "grad_norm": 5.513291358947754, + "learning_rate": 9.9817589981759e-06, + "loss": 0.0904, + "step": 18249 + }, + { + "epoch": 8.484425848442585, + "grad_norm": 9.970572471618652, + "learning_rate": 9.981757998175801e-06, + "loss": 0.3424, + "step": 18250 + }, + { + "epoch": 8.484890748489075, + "grad_norm": 4.074650287628174, + "learning_rate": 9.9817569981757e-06, + "loss": 0.0682, + "step": 18251 + }, + { + "epoch": 8.485355648535565, + "grad_norm": 8.85068416595459, + "learning_rate": 9.9817559981756e-06, + "loss": 0.1981, + "step": 18252 + }, + { + "epoch": 8.485820548582055, + "grad_norm": 4.329159736633301, + "learning_rate": 9.9817549981755e-06, + "loss": 0.0583, + "step": 18253 + }, + { + "epoch": 8.486285448628545, + "grad_norm": 5.770695209503174, + "learning_rate": 9.9817539981754e-06, + "loss": 0.0728, + "step": 18254 + }, + { + "epoch": 8.486750348675034, + "grad_norm": 7.14862585067749, + "learning_rate": 9.981752998175301e-06, + "loss": 0.1663, + "step": 18255 + }, + { + "epoch": 8.487215248721524, + "grad_norm": 7.735523223876953, + "learning_rate": 9.981751998175202e-06, + "loss": 0.1053, + "step": 18256 + }, + { + "epoch": 8.487680148768014, + "grad_norm": 5.283644676208496, + "learning_rate": 9.9817509981751e-06, + "loss": 0.1009, + "step": 18257 + }, + { + "epoch": 8.488145048814506, + "grad_norm": 5.802675724029541, + "learning_rate": 9.981749998175001e-06, + "loss": 0.1059, + "step": 18258 + }, + { + "epoch": 8.488609948860995, + "grad_norm": 6.4933762550354, + "learning_rate": 9.9817489981749e-06, + "loss": 0.1098, + "step": 18259 + }, + { + "epoch": 8.489074848907485, + "grad_norm": 3.4907732009887695, + "learning_rate": 9.9817479981748e-06, + "loss": 0.0404, + "step": 18260 + }, + { + "epoch": 8.489539748953975, + "grad_norm": 5.602140426635742, + "learning_rate": 9.9817469981747e-06, + "loss": 0.108, + "step": 18261 + }, + { + "epoch": 8.490004649000465, + "grad_norm": 6.603521823883057, + "learning_rate": 9.9817459981746e-06, + "loss": 0.1466, + "step": 18262 + }, + { + "epoch": 8.490469549046955, + "grad_norm": 6.509703159332275, + "learning_rate": 9.981744998174501e-06, + "loss": 0.1952, + "step": 18263 + }, + { + "epoch": 8.490934449093444, + "grad_norm": 3.486064910888672, + "learning_rate": 9.9817439981744e-06, + "loss": 0.0512, + "step": 18264 + }, + { + "epoch": 8.491399349139934, + "grad_norm": 7.452874183654785, + "learning_rate": 9.9817429981743e-06, + "loss": 0.2346, + "step": 18265 + }, + { + "epoch": 8.491864249186424, + "grad_norm": 5.145008563995361, + "learning_rate": 9.9817419981742e-06, + "loss": 0.1163, + "step": 18266 + }, + { + "epoch": 8.492329149232916, + "grad_norm": 9.719053268432617, + "learning_rate": 9.9817409981741e-06, + "loss": 0.2541, + "step": 18267 + }, + { + "epoch": 8.492794049279405, + "grad_norm": 4.104516983032227, + "learning_rate": 9.981739998174001e-06, + "loss": 0.0735, + "step": 18268 + }, + { + "epoch": 8.493258949325895, + "grad_norm": 5.949637413024902, + "learning_rate": 9.981738998173902e-06, + "loss": 0.2373, + "step": 18269 + }, + { + "epoch": 8.493723849372385, + "grad_norm": 5.514326572418213, + "learning_rate": 9.9817379981738e-06, + "loss": 0.1002, + "step": 18270 + }, + { + "epoch": 8.494188749418875, + "grad_norm": 6.420703411102295, + "learning_rate": 9.981736998173701e-06, + "loss": 0.1361, + "step": 18271 + }, + { + "epoch": 8.494653649465365, + "grad_norm": 5.4007887840271, + "learning_rate": 9.9817359981736e-06, + "loss": 0.1573, + "step": 18272 + }, + { + "epoch": 8.495118549511854, + "grad_norm": 6.089663505554199, + "learning_rate": 9.981734998173501e-06, + "loss": 0.0684, + "step": 18273 + }, + { + "epoch": 8.495583449558344, + "grad_norm": 8.090139389038086, + "learning_rate": 9.9817339981734e-06, + "loss": 0.2208, + "step": 18274 + }, + { + "epoch": 8.496048349604836, + "grad_norm": 2.033578872680664, + "learning_rate": 9.9817329981733e-06, + "loss": 0.0207, + "step": 18275 + }, + { + "epoch": 8.496513249651326, + "grad_norm": 4.706499099731445, + "learning_rate": 9.981731998173201e-06, + "loss": 0.0934, + "step": 18276 + }, + { + "epoch": 8.496978149697815, + "grad_norm": 9.931955337524414, + "learning_rate": 9.9817309981731e-06, + "loss": 0.1043, + "step": 18277 + }, + { + "epoch": 8.497443049744305, + "grad_norm": 3.2581429481506348, + "learning_rate": 9.981729998173e-06, + "loss": 0.0752, + "step": 18278 + }, + { + "epoch": 8.497907949790795, + "grad_norm": 4.070776462554932, + "learning_rate": 9.9817289981729e-06, + "loss": 0.0595, + "step": 18279 + }, + { + "epoch": 8.498372849837285, + "grad_norm": 11.780572891235352, + "learning_rate": 9.9817279981728e-06, + "loss": 0.2114, + "step": 18280 + }, + { + "epoch": 8.498837749883775, + "grad_norm": 4.838888168334961, + "learning_rate": 9.981726998172701e-06, + "loss": 0.1686, + "step": 18281 + }, + { + "epoch": 8.499302649930264, + "grad_norm": 4.597367286682129, + "learning_rate": 9.981725998172602e-06, + "loss": 0.0577, + "step": 18282 + }, + { + "epoch": 8.499767549976754, + "grad_norm": 6.69289493560791, + "learning_rate": 9.9817249981725e-06, + "loss": 0.1038, + "step": 18283 + }, + { + "epoch": 8.500232450023246, + "grad_norm": 10.616536140441895, + "learning_rate": 9.9817239981724e-06, + "loss": 0.1555, + "step": 18284 + }, + { + "epoch": 8.500697350069736, + "grad_norm": 6.551665306091309, + "learning_rate": 9.9817229981723e-06, + "loss": 0.0985, + "step": 18285 + }, + { + "epoch": 8.501162250116225, + "grad_norm": 10.5630521774292, + "learning_rate": 9.981721998172201e-06, + "loss": 0.2349, + "step": 18286 + }, + { + "epoch": 8.501627150162715, + "grad_norm": 6.109292984008789, + "learning_rate": 9.981720998172102e-06, + "loss": 0.2927, + "step": 18287 + }, + { + "epoch": 8.502092050209205, + "grad_norm": 5.679097652435303, + "learning_rate": 9.981719998172e-06, + "loss": 0.1521, + "step": 18288 + }, + { + "epoch": 8.502556950255695, + "grad_norm": 4.133053302764893, + "learning_rate": 9.981718998171901e-06, + "loss": 0.0559, + "step": 18289 + }, + { + "epoch": 8.503021850302185, + "grad_norm": 4.2868523597717285, + "learning_rate": 9.9817179981718e-06, + "loss": 0.1076, + "step": 18290 + }, + { + "epoch": 8.503486750348674, + "grad_norm": 6.58754825592041, + "learning_rate": 9.981716998171701e-06, + "loss": 0.1081, + "step": 18291 + }, + { + "epoch": 8.503951650395164, + "grad_norm": 6.397489070892334, + "learning_rate": 9.9817159981716e-06, + "loss": 0.1557, + "step": 18292 + }, + { + "epoch": 8.504416550441656, + "grad_norm": 1.843836784362793, + "learning_rate": 9.9817149981715e-06, + "loss": 0.024, + "step": 18293 + }, + { + "epoch": 8.504881450488146, + "grad_norm": 7.562657356262207, + "learning_rate": 9.981713998171401e-06, + "loss": 0.1782, + "step": 18294 + }, + { + "epoch": 8.505346350534635, + "grad_norm": 7.665367126464844, + "learning_rate": 9.9817129981713e-06, + "loss": 0.0805, + "step": 18295 + }, + { + "epoch": 8.505811250581125, + "grad_norm": 8.960116386413574, + "learning_rate": 9.9817119981712e-06, + "loss": 0.2419, + "step": 18296 + }, + { + "epoch": 8.506276150627615, + "grad_norm": 5.071922302246094, + "learning_rate": 9.9817109981711e-06, + "loss": 0.1207, + "step": 18297 + }, + { + "epoch": 8.506741050674105, + "grad_norm": 6.976100444793701, + "learning_rate": 9.981709998171e-06, + "loss": 0.1063, + "step": 18298 + }, + { + "epoch": 8.507205950720595, + "grad_norm": 3.2791848182678223, + "learning_rate": 9.981708998170901e-06, + "loss": 0.0441, + "step": 18299 + }, + { + "epoch": 8.507670850767084, + "grad_norm": 7.448563098907471, + "learning_rate": 9.981707998170802e-06, + "loss": 0.1445, + "step": 18300 + }, + { + "epoch": 8.508135750813576, + "grad_norm": 2.9322381019592285, + "learning_rate": 9.9817069981707e-06, + "loss": 0.0431, + "step": 18301 + }, + { + "epoch": 8.508600650860066, + "grad_norm": 6.076827049255371, + "learning_rate": 9.981705998170601e-06, + "loss": 0.1394, + "step": 18302 + }, + { + "epoch": 8.509065550906556, + "grad_norm": 6.470766067504883, + "learning_rate": 9.9817049981705e-06, + "loss": 0.1397, + "step": 18303 + }, + { + "epoch": 8.509530450953045, + "grad_norm": 2.1704354286193848, + "learning_rate": 9.981703998170401e-06, + "loss": 0.0358, + "step": 18304 + }, + { + "epoch": 8.509995350999535, + "grad_norm": 5.33257532119751, + "learning_rate": 9.9817029981703e-06, + "loss": 0.1653, + "step": 18305 + }, + { + "epoch": 8.510460251046025, + "grad_norm": 5.849205493927002, + "learning_rate": 9.9817019981702e-06, + "loss": 0.1023, + "step": 18306 + }, + { + "epoch": 8.510925151092515, + "grad_norm": 6.518138408660889, + "learning_rate": 9.981700998170101e-06, + "loss": 0.0673, + "step": 18307 + }, + { + "epoch": 8.511390051139005, + "grad_norm": 2.4908652305603027, + "learning_rate": 9.98169999817e-06, + "loss": 0.0508, + "step": 18308 + }, + { + "epoch": 8.511854951185494, + "grad_norm": 5.541572570800781, + "learning_rate": 9.981698998169901e-06, + "loss": 0.1753, + "step": 18309 + }, + { + "epoch": 8.512319851231986, + "grad_norm": 4.1494460105896, + "learning_rate": 9.9816979981698e-06, + "loss": 0.0545, + "step": 18310 + }, + { + "epoch": 8.512784751278476, + "grad_norm": 6.653372287750244, + "learning_rate": 9.9816969981697e-06, + "loss": 0.1772, + "step": 18311 + }, + { + "epoch": 8.513249651324966, + "grad_norm": 6.737540245056152, + "learning_rate": 9.981695998169601e-06, + "loss": 0.1109, + "step": 18312 + }, + { + "epoch": 8.513714551371455, + "grad_norm": 4.733560562133789, + "learning_rate": 9.981694998169502e-06, + "loss": 0.0754, + "step": 18313 + }, + { + "epoch": 8.514179451417945, + "grad_norm": 4.4684343338012695, + "learning_rate": 9.9816939981694e-06, + "loss": 0.0449, + "step": 18314 + }, + { + "epoch": 8.514644351464435, + "grad_norm": 4.814915657043457, + "learning_rate": 9.981692998169301e-06, + "loss": 0.0767, + "step": 18315 + }, + { + "epoch": 8.515109251510925, + "grad_norm": 2.8778879642486572, + "learning_rate": 9.9816919981692e-06, + "loss": 0.0309, + "step": 18316 + }, + { + "epoch": 8.515574151557415, + "grad_norm": 2.347590923309326, + "learning_rate": 9.9816909981691e-06, + "loss": 0.0267, + "step": 18317 + }, + { + "epoch": 8.516039051603904, + "grad_norm": 7.9817657470703125, + "learning_rate": 9.981689998169e-06, + "loss": 0.155, + "step": 18318 + }, + { + "epoch": 8.516503951650396, + "grad_norm": 2.3974649906158447, + "learning_rate": 9.9816889981689e-06, + "loss": 0.0345, + "step": 18319 + }, + { + "epoch": 8.516968851696886, + "grad_norm": 3.707263231277466, + "learning_rate": 9.981687998168801e-06, + "loss": 0.0718, + "step": 18320 + }, + { + "epoch": 8.517433751743376, + "grad_norm": 5.900111198425293, + "learning_rate": 9.9816869981687e-06, + "loss": 0.1332, + "step": 18321 + }, + { + "epoch": 8.517898651789865, + "grad_norm": 3.6297144889831543, + "learning_rate": 9.981685998168601e-06, + "loss": 0.0694, + "step": 18322 + }, + { + "epoch": 8.518363551836355, + "grad_norm": 3.7411818504333496, + "learning_rate": 9.9816849981685e-06, + "loss": 0.0982, + "step": 18323 + }, + { + "epoch": 8.518828451882845, + "grad_norm": 7.170544147491455, + "learning_rate": 9.9816839981684e-06, + "loss": 0.1144, + "step": 18324 + }, + { + "epoch": 8.519293351929335, + "grad_norm": 4.304017066955566, + "learning_rate": 9.981682998168301e-06, + "loss": 0.0944, + "step": 18325 + }, + { + "epoch": 8.519758251975825, + "grad_norm": 4.017855167388916, + "learning_rate": 9.981681998168202e-06, + "loss": 0.0655, + "step": 18326 + }, + { + "epoch": 8.520223152022314, + "grad_norm": 1.2769581079483032, + "learning_rate": 9.981680998168101e-06, + "loss": 0.0134, + "step": 18327 + }, + { + "epoch": 8.520688052068806, + "grad_norm": 5.143224716186523, + "learning_rate": 9.981679998168e-06, + "loss": 0.1107, + "step": 18328 + }, + { + "epoch": 8.521152952115296, + "grad_norm": 5.909348964691162, + "learning_rate": 9.9816789981679e-06, + "loss": 0.1955, + "step": 18329 + }, + { + "epoch": 8.521617852161786, + "grad_norm": 5.5986552238464355, + "learning_rate": 9.9816779981678e-06, + "loss": 0.1388, + "step": 18330 + }, + { + "epoch": 8.522082752208275, + "grad_norm": 4.655524253845215, + "learning_rate": 9.9816769981677e-06, + "loss": 0.11, + "step": 18331 + }, + { + "epoch": 8.522547652254765, + "grad_norm": 2.992804527282715, + "learning_rate": 9.9816759981676e-06, + "loss": 0.047, + "step": 18332 + }, + { + "epoch": 8.523012552301255, + "grad_norm": 1.1472773551940918, + "learning_rate": 9.981674998167501e-06, + "loss": 0.0121, + "step": 18333 + }, + { + "epoch": 8.523477452347745, + "grad_norm": 4.341512203216553, + "learning_rate": 9.9816739981674e-06, + "loss": 0.055, + "step": 18334 + }, + { + "epoch": 8.523942352394235, + "grad_norm": 5.83893346786499, + "learning_rate": 9.981672998167301e-06, + "loss": 0.1475, + "step": 18335 + }, + { + "epoch": 8.524407252440724, + "grad_norm": 2.650989532470703, + "learning_rate": 9.9816719981672e-06, + "loss": 0.0551, + "step": 18336 + }, + { + "epoch": 8.524872152487216, + "grad_norm": 4.75358772277832, + "learning_rate": 9.9816709981671e-06, + "loss": 0.0554, + "step": 18337 + }, + { + "epoch": 8.525337052533706, + "grad_norm": 7.453856468200684, + "learning_rate": 9.981669998167001e-06, + "loss": 0.1814, + "step": 18338 + }, + { + "epoch": 8.525801952580196, + "grad_norm": 5.316797733306885, + "learning_rate": 9.9816689981669e-06, + "loss": 0.1215, + "step": 18339 + }, + { + "epoch": 8.526266852626685, + "grad_norm": 8.352880477905273, + "learning_rate": 9.981667998166801e-06, + "loss": 0.2796, + "step": 18340 + }, + { + "epoch": 8.526731752673175, + "grad_norm": 10.547172546386719, + "learning_rate": 9.9816669981667e-06, + "loss": 0.1214, + "step": 18341 + }, + { + "epoch": 8.527196652719665, + "grad_norm": 4.028879642486572, + "learning_rate": 9.9816659981666e-06, + "loss": 0.0634, + "step": 18342 + }, + { + "epoch": 8.527661552766155, + "grad_norm": 6.1483635902404785, + "learning_rate": 9.9816649981665e-06, + "loss": 0.1727, + "step": 18343 + }, + { + "epoch": 8.528126452812645, + "grad_norm": 9.332514762878418, + "learning_rate": 9.9816639981664e-06, + "loss": 0.2434, + "step": 18344 + }, + { + "epoch": 8.528591352859134, + "grad_norm": 3.473792552947998, + "learning_rate": 9.9816629981663e-06, + "loss": 0.0416, + "step": 18345 + }, + { + "epoch": 8.529056252905626, + "grad_norm": 4.787380695343018, + "learning_rate": 9.981661998166201e-06, + "loss": 0.1032, + "step": 18346 + }, + { + "epoch": 8.529521152952116, + "grad_norm": 5.501521110534668, + "learning_rate": 9.9816609981661e-06, + "loss": 0.1029, + "step": 18347 + }, + { + "epoch": 8.529986052998606, + "grad_norm": 11.716405868530273, + "learning_rate": 9.981659998166001e-06, + "loss": 0.295, + "step": 18348 + }, + { + "epoch": 8.530450953045095, + "grad_norm": 3.451927661895752, + "learning_rate": 9.9816589981659e-06, + "loss": 0.0595, + "step": 18349 + }, + { + "epoch": 8.530915853091585, + "grad_norm": 4.066868305206299, + "learning_rate": 9.9816579981658e-06, + "loss": 0.1462, + "step": 18350 + }, + { + "epoch": 8.531380753138075, + "grad_norm": 3.065133571624756, + "learning_rate": 9.981656998165701e-06, + "loss": 0.0592, + "step": 18351 + }, + { + "epoch": 8.531845653184565, + "grad_norm": 8.078618049621582, + "learning_rate": 9.9816559981656e-06, + "loss": 0.1741, + "step": 18352 + }, + { + "epoch": 8.532310553231055, + "grad_norm": 5.381713390350342, + "learning_rate": 9.981654998165501e-06, + "loss": 0.0701, + "step": 18353 + }, + { + "epoch": 8.532775453277544, + "grad_norm": 5.898880481719971, + "learning_rate": 9.9816539981654e-06, + "loss": 0.1405, + "step": 18354 + }, + { + "epoch": 8.533240353324036, + "grad_norm": 1.3827131986618042, + "learning_rate": 9.9816529981653e-06, + "loss": 0.014, + "step": 18355 + }, + { + "epoch": 8.533705253370526, + "grad_norm": 3.5521960258483887, + "learning_rate": 9.981651998165201e-06, + "loss": 0.0319, + "step": 18356 + }, + { + "epoch": 8.534170153417016, + "grad_norm": 5.475245475769043, + "learning_rate": 9.9816509981651e-06, + "loss": 0.1429, + "step": 18357 + }, + { + "epoch": 8.534635053463505, + "grad_norm": 6.610136032104492, + "learning_rate": 9.981649998165001e-06, + "loss": 0.0532, + "step": 18358 + }, + { + "epoch": 8.535099953509995, + "grad_norm": 4.7395429611206055, + "learning_rate": 9.981648998164902e-06, + "loss": 0.1028, + "step": 18359 + }, + { + "epoch": 8.535564853556485, + "grad_norm": 5.658166885375977, + "learning_rate": 9.9816479981648e-06, + "loss": 0.053, + "step": 18360 + }, + { + "epoch": 8.536029753602975, + "grad_norm": 6.881280899047852, + "learning_rate": 9.9816469981647e-06, + "loss": 0.1329, + "step": 18361 + }, + { + "epoch": 8.536494653649465, + "grad_norm": 3.986701011657715, + "learning_rate": 9.9816459981646e-06, + "loss": 0.0906, + "step": 18362 + }, + { + "epoch": 8.536959553695956, + "grad_norm": 3.3750438690185547, + "learning_rate": 9.9816449981645e-06, + "loss": 0.0635, + "step": 18363 + }, + { + "epoch": 8.537424453742446, + "grad_norm": 4.562286853790283, + "learning_rate": 9.981643998164401e-06, + "loss": 0.0746, + "step": 18364 + }, + { + "epoch": 8.537889353788936, + "grad_norm": 8.56846809387207, + "learning_rate": 9.9816429981643e-06, + "loss": 0.0514, + "step": 18365 + }, + { + "epoch": 8.538354253835426, + "grad_norm": 7.306209564208984, + "learning_rate": 9.981641998164201e-06, + "loss": 0.17, + "step": 18366 + }, + { + "epoch": 8.538819153881915, + "grad_norm": 7.854860782623291, + "learning_rate": 9.9816409981641e-06, + "loss": 0.1632, + "step": 18367 + }, + { + "epoch": 8.539284053928405, + "grad_norm": 3.338139772415161, + "learning_rate": 9.981639998164e-06, + "loss": 0.0501, + "step": 18368 + }, + { + "epoch": 8.539748953974895, + "grad_norm": 7.089101791381836, + "learning_rate": 9.981638998163901e-06, + "loss": 0.1239, + "step": 18369 + }, + { + "epoch": 8.540213854021385, + "grad_norm": 4.9315009117126465, + "learning_rate": 9.9816379981638e-06, + "loss": 0.0955, + "step": 18370 + }, + { + "epoch": 8.540678754067875, + "grad_norm": 5.0988078117370605, + "learning_rate": 9.981636998163701e-06, + "loss": 0.0817, + "step": 18371 + }, + { + "epoch": 8.541143654114366, + "grad_norm": 4.392897129058838, + "learning_rate": 9.9816359981636e-06, + "loss": 0.0681, + "step": 18372 + }, + { + "epoch": 8.541608554160856, + "grad_norm": 4.779259204864502, + "learning_rate": 9.9816349981635e-06, + "loss": 0.0885, + "step": 18373 + }, + { + "epoch": 8.542073454207346, + "grad_norm": 5.195240020751953, + "learning_rate": 9.9816339981634e-06, + "loss": 0.079, + "step": 18374 + }, + { + "epoch": 8.542538354253836, + "grad_norm": 5.817032337188721, + "learning_rate": 9.9816329981633e-06, + "loss": 0.1146, + "step": 18375 + }, + { + "epoch": 8.543003254300325, + "grad_norm": 2.8927180767059326, + "learning_rate": 9.981631998163201e-06, + "loss": 0.0548, + "step": 18376 + }, + { + "epoch": 8.543468154346815, + "grad_norm": 2.9891557693481445, + "learning_rate": 9.981630998163102e-06, + "loss": 0.0382, + "step": 18377 + }, + { + "epoch": 8.543933054393305, + "grad_norm": 8.597478866577148, + "learning_rate": 9.981629998163e-06, + "loss": 0.1915, + "step": 18378 + }, + { + "epoch": 8.544397954439795, + "grad_norm": 6.175235748291016, + "learning_rate": 9.981628998162901e-06, + "loss": 0.179, + "step": 18379 + }, + { + "epoch": 8.544862854486286, + "grad_norm": 5.626120090484619, + "learning_rate": 9.9816279981628e-06, + "loss": 0.0291, + "step": 18380 + }, + { + "epoch": 8.545327754532776, + "grad_norm": 8.913616180419922, + "learning_rate": 9.9816269981627e-06, + "loss": 0.1262, + "step": 18381 + }, + { + "epoch": 8.545792654579266, + "grad_norm": 1.5708372592926025, + "learning_rate": 9.981625998162601e-06, + "loss": 0.0131, + "step": 18382 + }, + { + "epoch": 8.546257554625756, + "grad_norm": 7.034671306610107, + "learning_rate": 9.9816249981625e-06, + "loss": 0.1894, + "step": 18383 + }, + { + "epoch": 8.546722454672246, + "grad_norm": 5.363525390625, + "learning_rate": 9.981623998162401e-06, + "loss": 0.1219, + "step": 18384 + }, + { + "epoch": 8.547187354718735, + "grad_norm": 4.186769485473633, + "learning_rate": 9.9816229981623e-06, + "loss": 0.0629, + "step": 18385 + }, + { + "epoch": 8.547652254765225, + "grad_norm": 6.248907089233398, + "learning_rate": 9.9816219981622e-06, + "loss": 0.0627, + "step": 18386 + }, + { + "epoch": 8.548117154811715, + "grad_norm": 4.972081661224365, + "learning_rate": 9.9816209981621e-06, + "loss": 0.0768, + "step": 18387 + }, + { + "epoch": 8.548582054858205, + "grad_norm": 7.711709976196289, + "learning_rate": 9.981619998162e-06, + "loss": 0.1605, + "step": 18388 + }, + { + "epoch": 8.549046954904696, + "grad_norm": 9.859403610229492, + "learning_rate": 9.981618998161901e-06, + "loss": 0.2699, + "step": 18389 + }, + { + "epoch": 8.549511854951186, + "grad_norm": 4.608705997467041, + "learning_rate": 9.981617998161802e-06, + "loss": 0.0674, + "step": 18390 + }, + { + "epoch": 8.549976754997676, + "grad_norm": 6.416714668273926, + "learning_rate": 9.9816169981617e-06, + "loss": 0.1081, + "step": 18391 + }, + { + "epoch": 8.550441655044166, + "grad_norm": 3.7911245822906494, + "learning_rate": 9.9816159981616e-06, + "loss": 0.0535, + "step": 18392 + }, + { + "epoch": 8.550906555090656, + "grad_norm": 4.3751139640808105, + "learning_rate": 9.9816149981615e-06, + "loss": 0.0756, + "step": 18393 + }, + { + "epoch": 8.551371455137145, + "grad_norm": 8.91894245147705, + "learning_rate": 9.981613998161401e-06, + "loss": 0.14, + "step": 18394 + }, + { + "epoch": 8.551836355183635, + "grad_norm": 12.455385208129883, + "learning_rate": 9.981612998161302e-06, + "loss": 0.38, + "step": 18395 + }, + { + "epoch": 8.552301255230125, + "grad_norm": 5.106831073760986, + "learning_rate": 9.9816119981612e-06, + "loss": 0.1532, + "step": 18396 + }, + { + "epoch": 8.552766155276615, + "grad_norm": 4.056753635406494, + "learning_rate": 9.981610998161101e-06, + "loss": 0.0653, + "step": 18397 + }, + { + "epoch": 8.553231055323106, + "grad_norm": 2.4537718296051025, + "learning_rate": 9.981609998161e-06, + "loss": 0.0265, + "step": 18398 + }, + { + "epoch": 8.553695955369596, + "grad_norm": 9.946398735046387, + "learning_rate": 9.9816089981609e-06, + "loss": 0.3812, + "step": 18399 + }, + { + "epoch": 8.554160855416086, + "grad_norm": 7.669124126434326, + "learning_rate": 9.9816079981608e-06, + "loss": 0.1653, + "step": 18400 + }, + { + "epoch": 8.554625755462576, + "grad_norm": 3.117581844329834, + "learning_rate": 9.9816069981607e-06, + "loss": 0.0392, + "step": 18401 + }, + { + "epoch": 8.555090655509066, + "grad_norm": 4.942415237426758, + "learning_rate": 9.981605998160601e-06, + "loss": 0.0708, + "step": 18402 + }, + { + "epoch": 8.555555555555555, + "grad_norm": 8.507476806640625, + "learning_rate": 9.9816049981605e-06, + "loss": 0.1297, + "step": 18403 + }, + { + "epoch": 8.556020455602045, + "grad_norm": 11.173848152160645, + "learning_rate": 9.9816039981604e-06, + "loss": 0.2524, + "step": 18404 + }, + { + "epoch": 8.556485355648535, + "grad_norm": 4.898244857788086, + "learning_rate": 9.9816029981603e-06, + "loss": 0.087, + "step": 18405 + }, + { + "epoch": 8.556950255695025, + "grad_norm": 6.647433757781982, + "learning_rate": 9.9816019981602e-06, + "loss": 0.1659, + "step": 18406 + }, + { + "epoch": 8.557415155741516, + "grad_norm": 7.559918403625488, + "learning_rate": 9.981600998160101e-06, + "loss": 0.109, + "step": 18407 + }, + { + "epoch": 8.557880055788006, + "grad_norm": 5.370911121368408, + "learning_rate": 9.981599998160002e-06, + "loss": 0.1264, + "step": 18408 + }, + { + "epoch": 8.558344955834496, + "grad_norm": 4.631765365600586, + "learning_rate": 9.9815989981599e-06, + "loss": 0.091, + "step": 18409 + }, + { + "epoch": 8.558809855880986, + "grad_norm": 7.4659037590026855, + "learning_rate": 9.981597998159801e-06, + "loss": 0.0785, + "step": 18410 + }, + { + "epoch": 8.559274755927476, + "grad_norm": 1.1388370990753174, + "learning_rate": 9.9815969981597e-06, + "loss": 0.0119, + "step": 18411 + }, + { + "epoch": 8.559739655973965, + "grad_norm": 1.5792292356491089, + "learning_rate": 9.981595998159601e-06, + "loss": 0.0198, + "step": 18412 + }, + { + "epoch": 8.560204556020455, + "grad_norm": 4.148149490356445, + "learning_rate": 9.9815949981595e-06, + "loss": 0.0398, + "step": 18413 + }, + { + "epoch": 8.560669456066945, + "grad_norm": 3.739480495452881, + "learning_rate": 9.9815939981594e-06, + "loss": 0.0581, + "step": 18414 + }, + { + "epoch": 8.561134356113435, + "grad_norm": 5.317870140075684, + "learning_rate": 9.981592998159301e-06, + "loss": 0.1449, + "step": 18415 + }, + { + "epoch": 8.561599256159926, + "grad_norm": 7.46881103515625, + "learning_rate": 9.9815919981592e-06, + "loss": 0.0996, + "step": 18416 + }, + { + "epoch": 8.562064156206416, + "grad_norm": 3.827315330505371, + "learning_rate": 9.9815909981591e-06, + "loss": 0.111, + "step": 18417 + }, + { + "epoch": 8.562529056252906, + "grad_norm": 3.275200366973877, + "learning_rate": 9.981589998159e-06, + "loss": 0.0321, + "step": 18418 + }, + { + "epoch": 8.562993956299396, + "grad_norm": 3.4383649826049805, + "learning_rate": 9.9815889981589e-06, + "loss": 0.0511, + "step": 18419 + }, + { + "epoch": 8.563458856345886, + "grad_norm": 5.49456262588501, + "learning_rate": 9.981587998158801e-06, + "loss": 0.1346, + "step": 18420 + }, + { + "epoch": 8.563923756392375, + "grad_norm": 1.6053261756896973, + "learning_rate": 9.981586998158702e-06, + "loss": 0.0166, + "step": 18421 + }, + { + "epoch": 8.564388656438865, + "grad_norm": 3.1400201320648193, + "learning_rate": 9.9815859981586e-06, + "loss": 0.0333, + "step": 18422 + }, + { + "epoch": 8.564853556485355, + "grad_norm": 3.8896217346191406, + "learning_rate": 9.981584998158501e-06, + "loss": 0.0484, + "step": 18423 + }, + { + "epoch": 8.565318456531845, + "grad_norm": 10.637519836425781, + "learning_rate": 9.9815839981584e-06, + "loss": 0.2775, + "step": 18424 + }, + { + "epoch": 8.565783356578336, + "grad_norm": 5.7278828620910645, + "learning_rate": 9.981582998158301e-06, + "loss": 0.1668, + "step": 18425 + }, + { + "epoch": 8.566248256624826, + "grad_norm": 4.238749027252197, + "learning_rate": 9.9815819981582e-06, + "loss": 0.0652, + "step": 18426 + }, + { + "epoch": 8.566713156671316, + "grad_norm": 8.356518745422363, + "learning_rate": 9.9815809981581e-06, + "loss": 0.1852, + "step": 18427 + }, + { + "epoch": 8.567178056717806, + "grad_norm": 5.97860050201416, + "learning_rate": 9.981579998158001e-06, + "loss": 0.1907, + "step": 18428 + }, + { + "epoch": 8.567642956764296, + "grad_norm": 5.342854976654053, + "learning_rate": 9.9815789981579e-06, + "loss": 0.0884, + "step": 18429 + }, + { + "epoch": 8.568107856810785, + "grad_norm": 3.001239776611328, + "learning_rate": 9.981577998157801e-06, + "loss": 0.0547, + "step": 18430 + }, + { + "epoch": 8.568572756857275, + "grad_norm": 5.548428058624268, + "learning_rate": 9.9815769981577e-06, + "loss": 0.0734, + "step": 18431 + }, + { + "epoch": 8.569037656903765, + "grad_norm": 5.889673709869385, + "learning_rate": 9.9815759981576e-06, + "loss": 0.1245, + "step": 18432 + }, + { + "epoch": 8.569502556950255, + "grad_norm": 3.171781301498413, + "learning_rate": 9.981574998157501e-06, + "loss": 0.0352, + "step": 18433 + }, + { + "epoch": 8.569967456996746, + "grad_norm": 4.9187517166137695, + "learning_rate": 9.981573998157402e-06, + "loss": 0.0548, + "step": 18434 + }, + { + "epoch": 8.570432357043236, + "grad_norm": 4.0727338790893555, + "learning_rate": 9.9815729981573e-06, + "loss": 0.0746, + "step": 18435 + }, + { + "epoch": 8.570897257089726, + "grad_norm": 6.691031455993652, + "learning_rate": 9.9815719981572e-06, + "loss": 0.074, + "step": 18436 + }, + { + "epoch": 8.571362157136216, + "grad_norm": 10.85780143737793, + "learning_rate": 9.9815709981571e-06, + "loss": 0.0776, + "step": 18437 + }, + { + "epoch": 8.571827057182706, + "grad_norm": 6.077392101287842, + "learning_rate": 9.981569998157001e-06, + "loss": 0.0826, + "step": 18438 + }, + { + "epoch": 8.572291957229195, + "grad_norm": 2.5673444271087646, + "learning_rate": 9.9815689981569e-06, + "loss": 0.0272, + "step": 18439 + }, + { + "epoch": 8.572756857275685, + "grad_norm": 5.4550018310546875, + "learning_rate": 9.9815679981568e-06, + "loss": 0.1444, + "step": 18440 + }, + { + "epoch": 8.573221757322175, + "grad_norm": 5.13348388671875, + "learning_rate": 9.981566998156701e-06, + "loss": 0.0695, + "step": 18441 + }, + { + "epoch": 8.573686657368667, + "grad_norm": 3.6320972442626953, + "learning_rate": 9.9815659981566e-06, + "loss": 0.0274, + "step": 18442 + }, + { + "epoch": 8.574151557415156, + "grad_norm": 8.266119956970215, + "learning_rate": 9.981564998156501e-06, + "loss": 0.2465, + "step": 18443 + }, + { + "epoch": 8.574616457461646, + "grad_norm": 8.53861141204834, + "learning_rate": 9.9815639981564e-06, + "loss": 0.2294, + "step": 18444 + }, + { + "epoch": 8.575081357508136, + "grad_norm": 5.412029266357422, + "learning_rate": 9.9815629981563e-06, + "loss": 0.093, + "step": 18445 + }, + { + "epoch": 8.575546257554626, + "grad_norm": 4.671779155731201, + "learning_rate": 9.981561998156201e-06, + "loss": 0.0672, + "step": 18446 + }, + { + "epoch": 8.576011157601116, + "grad_norm": 6.697450637817383, + "learning_rate": 9.9815609981561e-06, + "loss": 0.1571, + "step": 18447 + }, + { + "epoch": 8.576476057647605, + "grad_norm": 4.055670261383057, + "learning_rate": 9.981559998156e-06, + "loss": 0.0851, + "step": 18448 + }, + { + "epoch": 8.576940957694095, + "grad_norm": 5.883598804473877, + "learning_rate": 9.9815589981559e-06, + "loss": 0.115, + "step": 18449 + }, + { + "epoch": 8.577405857740585, + "grad_norm": 6.679086208343506, + "learning_rate": 9.9815579981558e-06, + "loss": 0.1645, + "step": 18450 + }, + { + "epoch": 8.577870757787077, + "grad_norm": 7.169626712799072, + "learning_rate": 9.981556998155701e-06, + "loss": 0.0953, + "step": 18451 + }, + { + "epoch": 8.578335657833566, + "grad_norm": 4.243823528289795, + "learning_rate": 9.981555998155602e-06, + "loss": 0.0425, + "step": 18452 + }, + { + "epoch": 8.578800557880056, + "grad_norm": 4.154604911804199, + "learning_rate": 9.9815549981555e-06, + "loss": 0.072, + "step": 18453 + }, + { + "epoch": 8.579265457926546, + "grad_norm": 3.6249518394470215, + "learning_rate": 9.981553998155401e-06, + "loss": 0.0598, + "step": 18454 + }, + { + "epoch": 8.579730357973036, + "grad_norm": 3.329247236251831, + "learning_rate": 9.9815529981553e-06, + "loss": 0.1008, + "step": 18455 + }, + { + "epoch": 8.580195258019526, + "grad_norm": 10.463624000549316, + "learning_rate": 9.981551998155201e-06, + "loss": 0.105, + "step": 18456 + }, + { + "epoch": 8.580660158066015, + "grad_norm": 2.6827268600463867, + "learning_rate": 9.9815509981551e-06, + "loss": 0.0188, + "step": 18457 + }, + { + "epoch": 8.581125058112505, + "grad_norm": 7.480957508087158, + "learning_rate": 9.981549998155e-06, + "loss": 0.1603, + "step": 18458 + }, + { + "epoch": 8.581589958158997, + "grad_norm": 5.509185791015625, + "learning_rate": 9.981548998154901e-06, + "loss": 0.0972, + "step": 18459 + }, + { + "epoch": 8.582054858205487, + "grad_norm": 4.4019646644592285, + "learning_rate": 9.9815479981548e-06, + "loss": 0.0959, + "step": 18460 + }, + { + "epoch": 8.582519758251976, + "grad_norm": 6.7055768966674805, + "learning_rate": 9.981546998154701e-06, + "loss": 0.1308, + "step": 18461 + }, + { + "epoch": 8.582984658298466, + "grad_norm": 5.762819766998291, + "learning_rate": 9.9815459981546e-06, + "loss": 0.1124, + "step": 18462 + }, + { + "epoch": 8.583449558344956, + "grad_norm": 4.001942157745361, + "learning_rate": 9.9815449981545e-06, + "loss": 0.0547, + "step": 18463 + }, + { + "epoch": 8.583914458391446, + "grad_norm": 7.157456874847412, + "learning_rate": 9.981543998154401e-06, + "loss": 0.2534, + "step": 18464 + }, + { + "epoch": 8.584379358437936, + "grad_norm": 3.500824451446533, + "learning_rate": 9.981542998154302e-06, + "loss": 0.0447, + "step": 18465 + }, + { + "epoch": 8.584844258484425, + "grad_norm": 3.8703389167785645, + "learning_rate": 9.9815419981542e-06, + "loss": 0.0721, + "step": 18466 + }, + { + "epoch": 8.585309158530915, + "grad_norm": 4.612654209136963, + "learning_rate": 9.981540998154101e-06, + "loss": 0.0742, + "step": 18467 + }, + { + "epoch": 8.585774058577407, + "grad_norm": 2.0968618392944336, + "learning_rate": 9.981539998154e-06, + "loss": 0.0235, + "step": 18468 + }, + { + "epoch": 8.586238958623897, + "grad_norm": 4.9380388259887695, + "learning_rate": 9.9815389981539e-06, + "loss": 0.113, + "step": 18469 + }, + { + "epoch": 8.586703858670386, + "grad_norm": 2.995671033859253, + "learning_rate": 9.9815379981538e-06, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 8.587168758716876, + "grad_norm": 3.4001731872558594, + "learning_rate": 9.9815369981537e-06, + "loss": 0.0678, + "step": 18471 + }, + { + "epoch": 8.587633658763366, + "grad_norm": 4.212493896484375, + "learning_rate": 9.981535998153601e-06, + "loss": 0.1589, + "step": 18472 + }, + { + "epoch": 8.588098558809856, + "grad_norm": 3.716846227645874, + "learning_rate": 9.9815349981535e-06, + "loss": 0.0346, + "step": 18473 + }, + { + "epoch": 8.588563458856346, + "grad_norm": 2.6354331970214844, + "learning_rate": 9.981533998153401e-06, + "loss": 0.0277, + "step": 18474 + }, + { + "epoch": 8.589028358902835, + "grad_norm": 5.369649887084961, + "learning_rate": 9.9815329981533e-06, + "loss": 0.0896, + "step": 18475 + }, + { + "epoch": 8.589493258949325, + "grad_norm": 6.182034492492676, + "learning_rate": 9.9815319981532e-06, + "loss": 0.1528, + "step": 18476 + }, + { + "epoch": 8.589958158995817, + "grad_norm": 3.296297311782837, + "learning_rate": 9.981530998153101e-06, + "loss": 0.0479, + "step": 18477 + }, + { + "epoch": 8.590423059042307, + "grad_norm": 8.896265029907227, + "learning_rate": 9.981529998153002e-06, + "loss": 0.1805, + "step": 18478 + }, + { + "epoch": 8.590887959088796, + "grad_norm": 7.050494194030762, + "learning_rate": 9.981528998152901e-06, + "loss": 0.2674, + "step": 18479 + }, + { + "epoch": 8.591352859135286, + "grad_norm": 5.186097145080566, + "learning_rate": 9.9815279981528e-06, + "loss": 0.0862, + "step": 18480 + }, + { + "epoch": 8.591817759181776, + "grad_norm": 9.465965270996094, + "learning_rate": 9.9815269981527e-06, + "loss": 0.1873, + "step": 18481 + }, + { + "epoch": 8.592282659228266, + "grad_norm": 6.075200080871582, + "learning_rate": 9.9815259981526e-06, + "loss": 0.1297, + "step": 18482 + }, + { + "epoch": 8.592747559274756, + "grad_norm": 2.928109645843506, + "learning_rate": 9.9815249981525e-06, + "loss": 0.0347, + "step": 18483 + }, + { + "epoch": 8.593212459321245, + "grad_norm": 4.448753356933594, + "learning_rate": 9.9815239981524e-06, + "loss": 0.0861, + "step": 18484 + }, + { + "epoch": 8.593677359367735, + "grad_norm": 3.534334897994995, + "learning_rate": 9.981522998152301e-06, + "loss": 0.0605, + "step": 18485 + }, + { + "epoch": 8.594142259414227, + "grad_norm": 4.885871887207031, + "learning_rate": 9.9815219981522e-06, + "loss": 0.1098, + "step": 18486 + }, + { + "epoch": 8.594607159460717, + "grad_norm": 4.757744312286377, + "learning_rate": 9.981520998152101e-06, + "loss": 0.0604, + "step": 18487 + }, + { + "epoch": 8.595072059507206, + "grad_norm": 6.008694171905518, + "learning_rate": 9.981519998152e-06, + "loss": 0.1282, + "step": 18488 + }, + { + "epoch": 8.595536959553696, + "grad_norm": 2.2021749019622803, + "learning_rate": 9.9815189981519e-06, + "loss": 0.0258, + "step": 18489 + }, + { + "epoch": 8.596001859600186, + "grad_norm": 13.44180679321289, + "learning_rate": 9.981517998151801e-06, + "loss": 0.1627, + "step": 18490 + }, + { + "epoch": 8.596466759646676, + "grad_norm": 4.993373870849609, + "learning_rate": 9.9815169981517e-06, + "loss": 0.1433, + "step": 18491 + }, + { + "epoch": 8.596931659693166, + "grad_norm": 5.5704169273376465, + "learning_rate": 9.981515998151601e-06, + "loss": 0.1389, + "step": 18492 + }, + { + "epoch": 8.597396559739655, + "grad_norm": 5.803818702697754, + "learning_rate": 9.9815149981515e-06, + "loss": 0.1256, + "step": 18493 + }, + { + "epoch": 8.597861459786145, + "grad_norm": 5.857314586639404, + "learning_rate": 9.9815139981514e-06, + "loss": 0.1735, + "step": 18494 + }, + { + "epoch": 8.598326359832637, + "grad_norm": 4.61329984664917, + "learning_rate": 9.9815129981513e-06, + "loss": 0.1169, + "step": 18495 + }, + { + "epoch": 8.598791259879127, + "grad_norm": 7.709939956665039, + "learning_rate": 9.9815119981512e-06, + "loss": 0.0524, + "step": 18496 + }, + { + "epoch": 8.599256159925616, + "grad_norm": 8.003787994384766, + "learning_rate": 9.981510998151101e-06, + "loss": 0.1584, + "step": 18497 + }, + { + "epoch": 8.599721059972106, + "grad_norm": 3.6748313903808594, + "learning_rate": 9.981509998151002e-06, + "loss": 0.0413, + "step": 18498 + }, + { + "epoch": 8.600185960018596, + "grad_norm": 4.290585041046143, + "learning_rate": 9.9815089981509e-06, + "loss": 0.0622, + "step": 18499 + }, + { + "epoch": 8.600650860065086, + "grad_norm": 3.0583479404449463, + "learning_rate": 9.981507998150801e-06, + "loss": 0.0462, + "step": 18500 + }, + { + "epoch": 8.601115760111576, + "grad_norm": 2.4037623405456543, + "learning_rate": 9.9815069981507e-06, + "loss": 0.0319, + "step": 18501 + }, + { + "epoch": 8.601580660158065, + "grad_norm": 11.152566909790039, + "learning_rate": 9.9815059981506e-06, + "loss": 0.2632, + "step": 18502 + }, + { + "epoch": 8.602045560204555, + "grad_norm": 1.8780608177185059, + "learning_rate": 9.981504998150501e-06, + "loss": 0.0177, + "step": 18503 + }, + { + "epoch": 8.602510460251047, + "grad_norm": 3.6309125423431396, + "learning_rate": 9.9815039981504e-06, + "loss": 0.046, + "step": 18504 + }, + { + "epoch": 8.602975360297537, + "grad_norm": 2.1023380756378174, + "learning_rate": 9.981502998150301e-06, + "loss": 0.0204, + "step": 18505 + }, + { + "epoch": 8.603440260344026, + "grad_norm": 4.836330413818359, + "learning_rate": 9.9815019981502e-06, + "loss": 0.1072, + "step": 18506 + }, + { + "epoch": 8.603905160390516, + "grad_norm": 4.436007976531982, + "learning_rate": 9.9815009981501e-06, + "loss": 0.101, + "step": 18507 + }, + { + "epoch": 8.604370060437006, + "grad_norm": 3.0137031078338623, + "learning_rate": 9.98149999815e-06, + "loss": 0.0499, + "step": 18508 + }, + { + "epoch": 8.604834960483496, + "grad_norm": 7.124214172363281, + "learning_rate": 9.9814989981499e-06, + "loss": 0.1315, + "step": 18509 + }, + { + "epoch": 8.605299860529986, + "grad_norm": 11.634737968444824, + "learning_rate": 9.981497998149801e-06, + "loss": 0.2299, + "step": 18510 + }, + { + "epoch": 8.605764760576475, + "grad_norm": 3.497814655303955, + "learning_rate": 9.9814969981497e-06, + "loss": 0.0554, + "step": 18511 + }, + { + "epoch": 8.606229660622965, + "grad_norm": 6.8773956298828125, + "learning_rate": 9.9814959981496e-06, + "loss": 0.1753, + "step": 18512 + }, + { + "epoch": 8.606694560669457, + "grad_norm": 5.187995433807373, + "learning_rate": 9.9814949981495e-06, + "loss": 0.111, + "step": 18513 + }, + { + "epoch": 8.607159460715947, + "grad_norm": 6.658773899078369, + "learning_rate": 9.9814939981494e-06, + "loss": 0.27, + "step": 18514 + }, + { + "epoch": 8.607624360762436, + "grad_norm": 6.812741279602051, + "learning_rate": 9.981492998149301e-06, + "loss": 0.2397, + "step": 18515 + }, + { + "epoch": 8.608089260808926, + "grad_norm": 5.480581760406494, + "learning_rate": 9.981491998149202e-06, + "loss": 0.1399, + "step": 18516 + }, + { + "epoch": 8.608554160855416, + "grad_norm": 4.857003688812256, + "learning_rate": 9.9814909981491e-06, + "loss": 0.0935, + "step": 18517 + }, + { + "epoch": 8.609019060901906, + "grad_norm": 3.9440605640411377, + "learning_rate": 9.981489998149001e-06, + "loss": 0.0567, + "step": 18518 + }, + { + "epoch": 8.609483960948396, + "grad_norm": 4.515775680541992, + "learning_rate": 9.9814889981489e-06, + "loss": 0.1549, + "step": 18519 + }, + { + "epoch": 8.609948860994885, + "grad_norm": 4.630257606506348, + "learning_rate": 9.9814879981488e-06, + "loss": 0.0525, + "step": 18520 + }, + { + "epoch": 8.610413761041377, + "grad_norm": 4.2440996170043945, + "learning_rate": 9.981486998148701e-06, + "loss": 0.1017, + "step": 18521 + }, + { + "epoch": 8.610878661087867, + "grad_norm": 2.3533670902252197, + "learning_rate": 9.9814859981486e-06, + "loss": 0.0257, + "step": 18522 + }, + { + "epoch": 8.611343561134357, + "grad_norm": 3.8560404777526855, + "learning_rate": 9.981484998148501e-06, + "loss": 0.0731, + "step": 18523 + }, + { + "epoch": 8.611808461180846, + "grad_norm": 3.1787421703338623, + "learning_rate": 9.9814839981484e-06, + "loss": 0.0404, + "step": 18524 + }, + { + "epoch": 8.612273361227336, + "grad_norm": 3.5140483379364014, + "learning_rate": 9.9814829981483e-06, + "loss": 0.0585, + "step": 18525 + }, + { + "epoch": 8.612738261273826, + "grad_norm": 5.956086158752441, + "learning_rate": 9.9814819981482e-06, + "loss": 0.1226, + "step": 18526 + }, + { + "epoch": 8.613203161320316, + "grad_norm": 4.852027893066406, + "learning_rate": 9.9814809981481e-06, + "loss": 0.1032, + "step": 18527 + }, + { + "epoch": 8.613668061366806, + "grad_norm": 8.648270606994629, + "learning_rate": 9.981479998148001e-06, + "loss": 0.0872, + "step": 18528 + }, + { + "epoch": 8.614132961413295, + "grad_norm": 5.3344244956970215, + "learning_rate": 9.981478998147902e-06, + "loss": 0.1566, + "step": 18529 + }, + { + "epoch": 8.614597861459787, + "grad_norm": 6.2592692375183105, + "learning_rate": 9.9814779981478e-06, + "loss": 0.1188, + "step": 18530 + }, + { + "epoch": 8.615062761506277, + "grad_norm": 6.066409587860107, + "learning_rate": 9.981476998147701e-06, + "loss": 0.1163, + "step": 18531 + }, + { + "epoch": 8.615527661552767, + "grad_norm": 1.9642119407653809, + "learning_rate": 9.9814759981476e-06, + "loss": 0.0195, + "step": 18532 + }, + { + "epoch": 8.615992561599256, + "grad_norm": 2.9104361534118652, + "learning_rate": 9.9814749981475e-06, + "loss": 0.0372, + "step": 18533 + }, + { + "epoch": 8.616457461645746, + "grad_norm": 4.46312141418457, + "learning_rate": 9.981473998147401e-06, + "loss": 0.105, + "step": 18534 + }, + { + "epoch": 8.616922361692236, + "grad_norm": 4.874323844909668, + "learning_rate": 9.9814729981473e-06, + "loss": 0.0717, + "step": 18535 + }, + { + "epoch": 8.617387261738726, + "grad_norm": 14.658380508422852, + "learning_rate": 9.981471998147201e-06, + "loss": 0.2022, + "step": 18536 + }, + { + "epoch": 8.617852161785216, + "grad_norm": 5.536159038543701, + "learning_rate": 9.9814709981471e-06, + "loss": 0.1187, + "step": 18537 + }, + { + "epoch": 8.618317061831707, + "grad_norm": 15.959223747253418, + "learning_rate": 9.981469998147e-06, + "loss": 0.2855, + "step": 18538 + }, + { + "epoch": 8.618781961878197, + "grad_norm": 3.330204963684082, + "learning_rate": 9.9814689981469e-06, + "loss": 0.0466, + "step": 18539 + }, + { + "epoch": 8.619246861924687, + "grad_norm": 7.082759380340576, + "learning_rate": 9.9814679981468e-06, + "loss": 0.2635, + "step": 18540 + }, + { + "epoch": 8.619711761971177, + "grad_norm": 5.930813789367676, + "learning_rate": 9.981466998146701e-06, + "loss": 0.1411, + "step": 18541 + }, + { + "epoch": 8.620176662017666, + "grad_norm": 3.7585790157318115, + "learning_rate": 9.981465998146602e-06, + "loss": 0.0494, + "step": 18542 + }, + { + "epoch": 8.620641562064156, + "grad_norm": 7.272674560546875, + "learning_rate": 9.9814649981465e-06, + "loss": 0.2139, + "step": 18543 + }, + { + "epoch": 8.621106462110646, + "grad_norm": 3.579364776611328, + "learning_rate": 9.9814639981464e-06, + "loss": 0.0585, + "step": 18544 + }, + { + "epoch": 8.621571362157136, + "grad_norm": 4.097227096557617, + "learning_rate": 9.9814629981463e-06, + "loss": 0.038, + "step": 18545 + }, + { + "epoch": 8.622036262203626, + "grad_norm": 8.019017219543457, + "learning_rate": 9.981461998146201e-06, + "loss": 0.2451, + "step": 18546 + }, + { + "epoch": 8.622501162250117, + "grad_norm": 1.9838138818740845, + "learning_rate": 9.981460998146102e-06, + "loss": 0.0219, + "step": 18547 + }, + { + "epoch": 8.622966062296607, + "grad_norm": 3.2408549785614014, + "learning_rate": 9.981459998146e-06, + "loss": 0.0489, + "step": 18548 + }, + { + "epoch": 8.623430962343097, + "grad_norm": 6.4204535484313965, + "learning_rate": 9.981458998145901e-06, + "loss": 0.0854, + "step": 18549 + }, + { + "epoch": 8.623895862389586, + "grad_norm": 3.79215931892395, + "learning_rate": 9.9814579981458e-06, + "loss": 0.0613, + "step": 18550 + }, + { + "epoch": 8.624360762436076, + "grad_norm": 2.7831125259399414, + "learning_rate": 9.9814569981457e-06, + "loss": 0.0343, + "step": 18551 + }, + { + "epoch": 8.624825662482566, + "grad_norm": 6.014641284942627, + "learning_rate": 9.9814559981456e-06, + "loss": 0.1093, + "step": 18552 + }, + { + "epoch": 8.625290562529056, + "grad_norm": 4.368762016296387, + "learning_rate": 9.9814549981455e-06, + "loss": 0.0759, + "step": 18553 + }, + { + "epoch": 8.625755462575546, + "grad_norm": 5.4727606773376465, + "learning_rate": 9.981453998145401e-06, + "loss": 0.1206, + "step": 18554 + }, + { + "epoch": 8.626220362622036, + "grad_norm": 5.116207122802734, + "learning_rate": 9.9814529981453e-06, + "loss": 0.0606, + "step": 18555 + }, + { + "epoch": 8.626685262668527, + "grad_norm": 7.140434265136719, + "learning_rate": 9.9814519981452e-06, + "loss": 0.1981, + "step": 18556 + }, + { + "epoch": 8.627150162715017, + "grad_norm": 6.958547592163086, + "learning_rate": 9.9814509981451e-06, + "loss": 0.2071, + "step": 18557 + }, + { + "epoch": 8.627615062761507, + "grad_norm": 8.799503326416016, + "learning_rate": 9.981449998145e-06, + "loss": 0.1533, + "step": 18558 + }, + { + "epoch": 8.628079962807996, + "grad_norm": 2.573162794113159, + "learning_rate": 9.981448998144901e-06, + "loss": 0.0491, + "step": 18559 + }, + { + "epoch": 8.628544862854486, + "grad_norm": 5.878509998321533, + "learning_rate": 9.981447998144802e-06, + "loss": 0.0824, + "step": 18560 + }, + { + "epoch": 8.629009762900976, + "grad_norm": 4.081353187561035, + "learning_rate": 9.9814469981447e-06, + "loss": 0.0596, + "step": 18561 + }, + { + "epoch": 8.629474662947466, + "grad_norm": 5.79559326171875, + "learning_rate": 9.981445998144601e-06, + "loss": 0.1313, + "step": 18562 + }, + { + "epoch": 8.629939562993956, + "grad_norm": 3.760612726211548, + "learning_rate": 9.9814449981445e-06, + "loss": 0.0615, + "step": 18563 + }, + { + "epoch": 8.630404463040446, + "grad_norm": 5.181830406188965, + "learning_rate": 9.981443998144401e-06, + "loss": 0.1271, + "step": 18564 + }, + { + "epoch": 8.630869363086937, + "grad_norm": 10.849916458129883, + "learning_rate": 9.9814429981443e-06, + "loss": 0.1999, + "step": 18565 + }, + { + "epoch": 8.631334263133427, + "grad_norm": 6.880411148071289, + "learning_rate": 9.9814419981442e-06, + "loss": 0.212, + "step": 18566 + }, + { + "epoch": 8.631799163179917, + "grad_norm": 9.689679145812988, + "learning_rate": 9.981440998144101e-06, + "loss": 0.2304, + "step": 18567 + }, + { + "epoch": 8.632264063226406, + "grad_norm": 7.475794792175293, + "learning_rate": 9.981439998144e-06, + "loss": 0.1635, + "step": 18568 + }, + { + "epoch": 8.632728963272896, + "grad_norm": 7.062572002410889, + "learning_rate": 9.9814389981439e-06, + "loss": 0.1723, + "step": 18569 + }, + { + "epoch": 8.633193863319386, + "grad_norm": 4.0524420738220215, + "learning_rate": 9.9814379981438e-06, + "loss": 0.0967, + "step": 18570 + }, + { + "epoch": 8.633658763365876, + "grad_norm": 3.5596723556518555, + "learning_rate": 9.9814369981437e-06, + "loss": 0.0496, + "step": 18571 + }, + { + "epoch": 8.634123663412366, + "grad_norm": 5.369840145111084, + "learning_rate": 9.981435998143601e-06, + "loss": 0.1412, + "step": 18572 + }, + { + "epoch": 8.634588563458856, + "grad_norm": 9.494874954223633, + "learning_rate": 9.981434998143502e-06, + "loss": 0.109, + "step": 18573 + }, + { + "epoch": 8.635053463505347, + "grad_norm": 7.066555023193359, + "learning_rate": 9.9814339981434e-06, + "loss": 0.1871, + "step": 18574 + }, + { + "epoch": 8.635518363551837, + "grad_norm": 4.61909818649292, + "learning_rate": 9.981432998143301e-06, + "loss": 0.0936, + "step": 18575 + }, + { + "epoch": 8.635983263598327, + "grad_norm": 7.159496307373047, + "learning_rate": 9.9814319981432e-06, + "loss": 0.1399, + "step": 18576 + }, + { + "epoch": 8.636448163644816, + "grad_norm": 3.8825809955596924, + "learning_rate": 9.9814309981431e-06, + "loss": 0.1169, + "step": 18577 + }, + { + "epoch": 8.636913063691306, + "grad_norm": 5.592648029327393, + "learning_rate": 9.981429998143e-06, + "loss": 0.1371, + "step": 18578 + }, + { + "epoch": 8.637377963737796, + "grad_norm": 10.03173542022705, + "learning_rate": 9.9814289981429e-06, + "loss": 0.2433, + "step": 18579 + }, + { + "epoch": 8.637842863784286, + "grad_norm": 3.638899803161621, + "learning_rate": 9.981427998142801e-06, + "loss": 0.0461, + "step": 18580 + }, + { + "epoch": 8.638307763830776, + "grad_norm": 2.733053684234619, + "learning_rate": 9.9814269981427e-06, + "loss": 0.0409, + "step": 18581 + }, + { + "epoch": 8.638772663877266, + "grad_norm": 5.638546466827393, + "learning_rate": 9.981425998142601e-06, + "loss": 0.1232, + "step": 18582 + }, + { + "epoch": 8.639237563923757, + "grad_norm": 5.097543239593506, + "learning_rate": 9.9814249981425e-06, + "loss": 0.1062, + "step": 18583 + }, + { + "epoch": 8.639702463970247, + "grad_norm": 0.9936851263046265, + "learning_rate": 9.9814239981424e-06, + "loss": 0.0112, + "step": 18584 + }, + { + "epoch": 8.640167364016737, + "grad_norm": 8.127516746520996, + "learning_rate": 9.981422998142301e-06, + "loss": 0.1892, + "step": 18585 + }, + { + "epoch": 8.640632264063226, + "grad_norm": 6.8167805671691895, + "learning_rate": 9.981421998142202e-06, + "loss": 0.0928, + "step": 18586 + }, + { + "epoch": 8.641097164109716, + "grad_norm": 3.1895666122436523, + "learning_rate": 9.9814209981421e-06, + "loss": 0.0338, + "step": 18587 + }, + { + "epoch": 8.641562064156206, + "grad_norm": 5.2672929763793945, + "learning_rate": 9.981419998142e-06, + "loss": 0.1208, + "step": 18588 + }, + { + "epoch": 8.642026964202696, + "grad_norm": 3.5062038898468018, + "learning_rate": 9.9814189981419e-06, + "loss": 0.0425, + "step": 18589 + }, + { + "epoch": 8.642491864249186, + "grad_norm": 5.942391395568848, + "learning_rate": 9.981417998141801e-06, + "loss": 0.1154, + "step": 18590 + }, + { + "epoch": 8.642956764295675, + "grad_norm": 2.8282010555267334, + "learning_rate": 9.9814169981417e-06, + "loss": 0.0489, + "step": 18591 + }, + { + "epoch": 8.643421664342167, + "grad_norm": 5.770865440368652, + "learning_rate": 9.9814159981416e-06, + "loss": 0.0826, + "step": 18592 + }, + { + "epoch": 8.643886564388657, + "grad_norm": 5.962427616119385, + "learning_rate": 9.981414998141501e-06, + "loss": 0.0509, + "step": 18593 + }, + { + "epoch": 8.644351464435147, + "grad_norm": 7.022154808044434, + "learning_rate": 9.9814139981414e-06, + "loss": 0.2329, + "step": 18594 + }, + { + "epoch": 8.644816364481636, + "grad_norm": 7.107037544250488, + "learning_rate": 9.981412998141301e-06, + "loss": 0.1365, + "step": 18595 + }, + { + "epoch": 8.645281264528126, + "grad_norm": 4.4409871101379395, + "learning_rate": 9.9814119981412e-06, + "loss": 0.0875, + "step": 18596 + }, + { + "epoch": 8.645746164574616, + "grad_norm": 7.392205715179443, + "learning_rate": 9.9814109981411e-06, + "loss": 0.2124, + "step": 18597 + }, + { + "epoch": 8.646211064621106, + "grad_norm": 4.425720691680908, + "learning_rate": 9.981409998141001e-06, + "loss": 0.0692, + "step": 18598 + }, + { + "epoch": 8.646675964667596, + "grad_norm": 6.8748674392700195, + "learning_rate": 9.9814089981409e-06, + "loss": 0.2716, + "step": 18599 + }, + { + "epoch": 8.647140864714087, + "grad_norm": 4.771142959594727, + "learning_rate": 9.981407998140801e-06, + "loss": 0.0586, + "step": 18600 + }, + { + "epoch": 8.647605764760577, + "grad_norm": 6.631175518035889, + "learning_rate": 9.9814069981407e-06, + "loss": 0.1109, + "step": 18601 + }, + { + "epoch": 8.648070664807067, + "grad_norm": 6.935229301452637, + "learning_rate": 9.9814059981406e-06, + "loss": 0.0856, + "step": 18602 + }, + { + "epoch": 8.648535564853557, + "grad_norm": 5.25390100479126, + "learning_rate": 9.981404998140501e-06, + "loss": 0.0535, + "step": 18603 + }, + { + "epoch": 8.649000464900046, + "grad_norm": 2.807810068130493, + "learning_rate": 9.9814039981404e-06, + "loss": 0.0406, + "step": 18604 + }, + { + "epoch": 8.649465364946536, + "grad_norm": 5.2752909660339355, + "learning_rate": 9.9814029981403e-06, + "loss": 0.1229, + "step": 18605 + }, + { + "epoch": 8.649930264993026, + "grad_norm": 3.739370107650757, + "learning_rate": 9.981401998140201e-06, + "loss": 0.0368, + "step": 18606 + }, + { + "epoch": 8.650395165039516, + "grad_norm": 7.25787353515625, + "learning_rate": 9.9814009981401e-06, + "loss": 0.101, + "step": 18607 + }, + { + "epoch": 8.650860065086006, + "grad_norm": 5.688644886016846, + "learning_rate": 9.981399998140001e-06, + "loss": 0.1688, + "step": 18608 + }, + { + "epoch": 8.651324965132497, + "grad_norm": 6.866466045379639, + "learning_rate": 9.9813989981399e-06, + "loss": 0.2099, + "step": 18609 + }, + { + "epoch": 8.651789865178987, + "grad_norm": 9.064371109008789, + "learning_rate": 9.9813979981398e-06, + "loss": 0.1581, + "step": 18610 + }, + { + "epoch": 8.652254765225477, + "grad_norm": 3.9008772373199463, + "learning_rate": 9.981396998139701e-06, + "loss": 0.0515, + "step": 18611 + }, + { + "epoch": 8.652719665271967, + "grad_norm": 4.923440933227539, + "learning_rate": 9.9813959981396e-06, + "loss": 0.1227, + "step": 18612 + }, + { + "epoch": 8.653184565318456, + "grad_norm": 2.5638506412506104, + "learning_rate": 9.981394998139501e-06, + "loss": 0.0311, + "step": 18613 + }, + { + "epoch": 8.653649465364946, + "grad_norm": 9.168126106262207, + "learning_rate": 9.9813939981394e-06, + "loss": 0.2058, + "step": 18614 + }, + { + "epoch": 8.654114365411436, + "grad_norm": 9.099955558776855, + "learning_rate": 9.9813929981393e-06, + "loss": 0.2523, + "step": 18615 + }, + { + "epoch": 8.654579265457926, + "grad_norm": 5.367936611175537, + "learning_rate": 9.981391998139201e-06, + "loss": 0.0893, + "step": 18616 + }, + { + "epoch": 8.655044165504417, + "grad_norm": 4.272452354431152, + "learning_rate": 9.981390998139102e-06, + "loss": 0.0478, + "step": 18617 + }, + { + "epoch": 8.655509065550907, + "grad_norm": 6.438763618469238, + "learning_rate": 9.981389998139e-06, + "loss": 0.1079, + "step": 18618 + }, + { + "epoch": 8.655973965597397, + "grad_norm": 4.633930206298828, + "learning_rate": 9.981388998138902e-06, + "loss": 0.0937, + "step": 18619 + }, + { + "epoch": 8.656438865643887, + "grad_norm": 4.295237064361572, + "learning_rate": 9.9813879981388e-06, + "loss": 0.0725, + "step": 18620 + }, + { + "epoch": 8.656903765690377, + "grad_norm": 4.613938808441162, + "learning_rate": 9.9813869981387e-06, + "loss": 0.0624, + "step": 18621 + }, + { + "epoch": 8.657368665736866, + "grad_norm": 6.113221168518066, + "learning_rate": 9.9813859981386e-06, + "loss": 0.1316, + "step": 18622 + }, + { + "epoch": 8.657833565783356, + "grad_norm": 2.4277751445770264, + "learning_rate": 9.9813849981385e-06, + "loss": 0.0282, + "step": 18623 + }, + { + "epoch": 8.658298465829846, + "grad_norm": 4.87819242477417, + "learning_rate": 9.981383998138401e-06, + "loss": 0.093, + "step": 18624 + }, + { + "epoch": 8.658763365876336, + "grad_norm": 4.105834484100342, + "learning_rate": 9.9813829981383e-06, + "loss": 0.0495, + "step": 18625 + }, + { + "epoch": 8.659228265922827, + "grad_norm": 5.22206974029541, + "learning_rate": 9.981381998138201e-06, + "loss": 0.1175, + "step": 18626 + }, + { + "epoch": 8.659693165969317, + "grad_norm": 3.571047306060791, + "learning_rate": 9.9813809981381e-06, + "loss": 0.0869, + "step": 18627 + }, + { + "epoch": 8.660158066015807, + "grad_norm": 4.036596775054932, + "learning_rate": 9.981379998138e-06, + "loss": 0.0626, + "step": 18628 + }, + { + "epoch": 8.660622966062297, + "grad_norm": 4.156589984893799, + "learning_rate": 9.981378998137901e-06, + "loss": 0.0781, + "step": 18629 + }, + { + "epoch": 8.661087866108787, + "grad_norm": 2.9929513931274414, + "learning_rate": 9.981377998137802e-06, + "loss": 0.0251, + "step": 18630 + }, + { + "epoch": 8.661552766155276, + "grad_norm": 3.5931262969970703, + "learning_rate": 9.981376998137701e-06, + "loss": 0.0292, + "step": 18631 + }, + { + "epoch": 8.662017666201766, + "grad_norm": 7.412452220916748, + "learning_rate": 9.9813759981376e-06, + "loss": 0.1881, + "step": 18632 + }, + { + "epoch": 8.662482566248256, + "grad_norm": 10.111892700195312, + "learning_rate": 9.9813749981375e-06, + "loss": 0.5582, + "step": 18633 + }, + { + "epoch": 8.662947466294746, + "grad_norm": 2.634626865386963, + "learning_rate": 9.9813739981374e-06, + "loss": 0.0349, + "step": 18634 + }, + { + "epoch": 8.663412366341237, + "grad_norm": 4.272809982299805, + "learning_rate": 9.9813729981373e-06, + "loss": 0.1397, + "step": 18635 + }, + { + "epoch": 8.663877266387727, + "grad_norm": 4.0661468505859375, + "learning_rate": 9.9813719981372e-06, + "loss": 0.0439, + "step": 18636 + }, + { + "epoch": 8.664342166434217, + "grad_norm": 6.573347091674805, + "learning_rate": 9.981370998137101e-06, + "loss": 0.0646, + "step": 18637 + }, + { + "epoch": 8.664807066480707, + "grad_norm": 6.145536422729492, + "learning_rate": 9.981369998137e-06, + "loss": 0.1433, + "step": 18638 + }, + { + "epoch": 8.665271966527197, + "grad_norm": 5.923717498779297, + "learning_rate": 9.981368998136901e-06, + "loss": 0.1143, + "step": 18639 + }, + { + "epoch": 8.665736866573686, + "grad_norm": 10.682188987731934, + "learning_rate": 9.9813679981368e-06, + "loss": 0.1132, + "step": 18640 + }, + { + "epoch": 8.666201766620176, + "grad_norm": 5.6853790283203125, + "learning_rate": 9.9813669981367e-06, + "loss": 0.101, + "step": 18641 + }, + { + "epoch": 8.666666666666666, + "grad_norm": 2.4981915950775146, + "learning_rate": 9.981365998136601e-06, + "loss": 0.0227, + "step": 18642 + }, + { + "epoch": 8.667131566713156, + "grad_norm": 4.127025127410889, + "learning_rate": 9.9813649981365e-06, + "loss": 0.1466, + "step": 18643 + }, + { + "epoch": 8.667596466759647, + "grad_norm": 4.182271957397461, + "learning_rate": 9.981363998136401e-06, + "loss": 0.047, + "step": 18644 + }, + { + "epoch": 8.668061366806137, + "grad_norm": 7.453575134277344, + "learning_rate": 9.9813629981363e-06, + "loss": 0.1034, + "step": 18645 + }, + { + "epoch": 8.668526266852627, + "grad_norm": 6.155381679534912, + "learning_rate": 9.9813619981362e-06, + "loss": 0.0952, + "step": 18646 + }, + { + "epoch": 8.668991166899117, + "grad_norm": 4.740445613861084, + "learning_rate": 9.9813609981361e-06, + "loss": 0.1381, + "step": 18647 + }, + { + "epoch": 8.669456066945607, + "grad_norm": 7.988277435302734, + "learning_rate": 9.981359998136e-06, + "loss": 0.2197, + "step": 18648 + }, + { + "epoch": 8.669920966992096, + "grad_norm": 4.737257480621338, + "learning_rate": 9.981358998135901e-06, + "loss": 0.2008, + "step": 18649 + }, + { + "epoch": 8.670385867038586, + "grad_norm": 5.993889331817627, + "learning_rate": 9.981357998135802e-06, + "loss": 0.1707, + "step": 18650 + }, + { + "epoch": 8.670850767085076, + "grad_norm": 6.001937389373779, + "learning_rate": 9.9813569981357e-06, + "loss": 0.0476, + "step": 18651 + }, + { + "epoch": 8.671315667131566, + "grad_norm": 3.7574284076690674, + "learning_rate": 9.9813559981356e-06, + "loss": 0.0739, + "step": 18652 + }, + { + "epoch": 8.671780567178057, + "grad_norm": 4.848381519317627, + "learning_rate": 9.9813549981355e-06, + "loss": 0.1074, + "step": 18653 + }, + { + "epoch": 8.672245467224547, + "grad_norm": 5.038010597229004, + "learning_rate": 9.9813539981354e-06, + "loss": 0.1491, + "step": 18654 + }, + { + "epoch": 8.672710367271037, + "grad_norm": 3.4326703548431396, + "learning_rate": 9.981352998135301e-06, + "loss": 0.0425, + "step": 18655 + }, + { + "epoch": 8.673175267317527, + "grad_norm": 3.607501745223999, + "learning_rate": 9.9813519981352e-06, + "loss": 0.0654, + "step": 18656 + }, + { + "epoch": 8.673640167364017, + "grad_norm": 3.6308465003967285, + "learning_rate": 9.981350998135101e-06, + "loss": 0.0548, + "step": 18657 + }, + { + "epoch": 8.674105067410506, + "grad_norm": 2.680887460708618, + "learning_rate": 9.981349998135e-06, + "loss": 0.0239, + "step": 18658 + }, + { + "epoch": 8.674569967456996, + "grad_norm": 8.905430793762207, + "learning_rate": 9.9813489981349e-06, + "loss": 0.3518, + "step": 18659 + }, + { + "epoch": 8.675034867503486, + "grad_norm": 2.4856553077697754, + "learning_rate": 9.9813479981348e-06, + "loss": 0.0256, + "step": 18660 + }, + { + "epoch": 8.675499767549976, + "grad_norm": 3.322368621826172, + "learning_rate": 9.9813469981347e-06, + "loss": 0.0545, + "step": 18661 + }, + { + "epoch": 8.675964667596467, + "grad_norm": 3.6419429779052734, + "learning_rate": 9.981345998134601e-06, + "loss": 0.0901, + "step": 18662 + }, + { + "epoch": 8.676429567642957, + "grad_norm": 4.726164817810059, + "learning_rate": 9.9813449981345e-06, + "loss": 0.0936, + "step": 18663 + }, + { + "epoch": 8.676894467689447, + "grad_norm": 5.899270534515381, + "learning_rate": 9.9813439981344e-06, + "loss": 0.1223, + "step": 18664 + }, + { + "epoch": 8.677359367735937, + "grad_norm": 3.696089267730713, + "learning_rate": 9.9813429981343e-06, + "loss": 0.0381, + "step": 18665 + }, + { + "epoch": 8.677824267782427, + "grad_norm": 5.454065322875977, + "learning_rate": 9.9813419981342e-06, + "loss": 0.0893, + "step": 18666 + }, + { + "epoch": 8.678289167828916, + "grad_norm": 4.758763790130615, + "learning_rate": 9.981340998134101e-06, + "loss": 0.094, + "step": 18667 + }, + { + "epoch": 8.678754067875406, + "grad_norm": 6.117623329162598, + "learning_rate": 9.981339998134002e-06, + "loss": 0.1655, + "step": 18668 + }, + { + "epoch": 8.679218967921896, + "grad_norm": 4.569734573364258, + "learning_rate": 9.9813389981339e-06, + "loss": 0.0616, + "step": 18669 + }, + { + "epoch": 8.679683867968386, + "grad_norm": 2.982077121734619, + "learning_rate": 9.981337998133801e-06, + "loss": 0.0422, + "step": 18670 + }, + { + "epoch": 8.680148768014877, + "grad_norm": 4.196254253387451, + "learning_rate": 9.9813369981337e-06, + "loss": 0.0935, + "step": 18671 + }, + { + "epoch": 8.680613668061367, + "grad_norm": 2.692507266998291, + "learning_rate": 9.9813359981336e-06, + "loss": 0.0245, + "step": 18672 + }, + { + "epoch": 8.681078568107857, + "grad_norm": 3.1005568504333496, + "learning_rate": 9.9813349981335e-06, + "loss": 0.0492, + "step": 18673 + }, + { + "epoch": 8.681543468154347, + "grad_norm": 3.006361484527588, + "learning_rate": 9.9813339981334e-06, + "loss": 0.0252, + "step": 18674 + }, + { + "epoch": 8.682008368200837, + "grad_norm": 6.502903938293457, + "learning_rate": 9.981332998133301e-06, + "loss": 0.0696, + "step": 18675 + }, + { + "epoch": 8.682473268247326, + "grad_norm": 3.7255773544311523, + "learning_rate": 9.9813319981332e-06, + "loss": 0.0725, + "step": 18676 + }, + { + "epoch": 8.682938168293816, + "grad_norm": 3.18007493019104, + "learning_rate": 9.9813309981331e-06, + "loss": 0.0318, + "step": 18677 + }, + { + "epoch": 8.683403068340306, + "grad_norm": 5.269449234008789, + "learning_rate": 9.981329998133e-06, + "loss": 0.1316, + "step": 18678 + }, + { + "epoch": 8.683867968386798, + "grad_norm": 5.373717308044434, + "learning_rate": 9.9813289981329e-06, + "loss": 0.1473, + "step": 18679 + }, + { + "epoch": 8.684332868433287, + "grad_norm": 4.558984279632568, + "learning_rate": 9.981327998132801e-06, + "loss": 0.1193, + "step": 18680 + }, + { + "epoch": 8.684797768479777, + "grad_norm": 5.1131367683410645, + "learning_rate": 9.981326998132702e-06, + "loss": 0.0829, + "step": 18681 + }, + { + "epoch": 8.685262668526267, + "grad_norm": 7.585017681121826, + "learning_rate": 9.9813259981326e-06, + "loss": 0.1096, + "step": 18682 + }, + { + "epoch": 8.685727568572757, + "grad_norm": 7.19258975982666, + "learning_rate": 9.981324998132501e-06, + "loss": 0.1312, + "step": 18683 + }, + { + "epoch": 8.686192468619247, + "grad_norm": 3.2512824535369873, + "learning_rate": 9.9813239981324e-06, + "loss": 0.0599, + "step": 18684 + }, + { + "epoch": 8.686657368665736, + "grad_norm": 3.483927011489868, + "learning_rate": 9.981322998132301e-06, + "loss": 0.0351, + "step": 18685 + }, + { + "epoch": 8.687122268712226, + "grad_norm": 4.163636207580566, + "learning_rate": 9.981321998132202e-06, + "loss": 0.0479, + "step": 18686 + }, + { + "epoch": 8.687587168758716, + "grad_norm": 4.186587810516357, + "learning_rate": 9.9813209981321e-06, + "loss": 0.0767, + "step": 18687 + }, + { + "epoch": 8.688052068805208, + "grad_norm": 5.7035369873046875, + "learning_rate": 9.981319998132001e-06, + "loss": 0.1658, + "step": 18688 + }, + { + "epoch": 8.688516968851697, + "grad_norm": 6.569637775421143, + "learning_rate": 9.9813189981319e-06, + "loss": 0.1445, + "step": 18689 + }, + { + "epoch": 8.688981868898187, + "grad_norm": 4.285706520080566, + "learning_rate": 9.9813179981318e-06, + "loss": 0.0407, + "step": 18690 + }, + { + "epoch": 8.689446768944677, + "grad_norm": 4.779482841491699, + "learning_rate": 9.9813169981317e-06, + "loss": 0.1006, + "step": 18691 + }, + { + "epoch": 8.689911668991167, + "grad_norm": 8.559532165527344, + "learning_rate": 9.9813159981316e-06, + "loss": 0.1914, + "step": 18692 + }, + { + "epoch": 8.690376569037657, + "grad_norm": 3.870725631713867, + "learning_rate": 9.981314998131501e-06, + "loss": 0.0398, + "step": 18693 + }, + { + "epoch": 8.690841469084146, + "grad_norm": 4.361174583435059, + "learning_rate": 9.981313998131402e-06, + "loss": 0.1312, + "step": 18694 + }, + { + "epoch": 8.691306369130636, + "grad_norm": 5.8312153816223145, + "learning_rate": 9.9813129981313e-06, + "loss": 0.0936, + "step": 18695 + }, + { + "epoch": 8.691771269177128, + "grad_norm": 5.354983329772949, + "learning_rate": 9.9813119981312e-06, + "loss": 0.1003, + "step": 18696 + }, + { + "epoch": 8.692236169223618, + "grad_norm": 4.335209369659424, + "learning_rate": 9.9813109981311e-06, + "loss": 0.0723, + "step": 18697 + }, + { + "epoch": 8.692701069270107, + "grad_norm": 4.586507320404053, + "learning_rate": 9.981309998131001e-06, + "loss": 0.0679, + "step": 18698 + }, + { + "epoch": 8.693165969316597, + "grad_norm": 5.509810447692871, + "learning_rate": 9.981308998130902e-06, + "loss": 0.1069, + "step": 18699 + }, + { + "epoch": 8.693630869363087, + "grad_norm": 5.389026165008545, + "learning_rate": 9.9813079981308e-06, + "loss": 0.1919, + "step": 18700 + }, + { + "epoch": 8.694095769409577, + "grad_norm": 5.517432689666748, + "learning_rate": 9.981306998130701e-06, + "loss": 0.1825, + "step": 18701 + }, + { + "epoch": 8.694560669456067, + "grad_norm": 9.02410888671875, + "learning_rate": 9.9813059981306e-06, + "loss": 0.3001, + "step": 18702 + }, + { + "epoch": 8.695025569502556, + "grad_norm": 5.583376884460449, + "learning_rate": 9.981304998130501e-06, + "loss": 0.1722, + "step": 18703 + }, + { + "epoch": 8.695490469549046, + "grad_norm": 4.275928497314453, + "learning_rate": 9.9813039981304e-06, + "loss": 0.0473, + "step": 18704 + }, + { + "epoch": 8.695955369595538, + "grad_norm": 4.728330135345459, + "learning_rate": 9.9813029981303e-06, + "loss": 0.1579, + "step": 18705 + }, + { + "epoch": 8.696420269642028, + "grad_norm": 1.3612743616104126, + "learning_rate": 9.981301998130201e-06, + "loss": 0.0123, + "step": 18706 + }, + { + "epoch": 8.696885169688517, + "grad_norm": 9.341442108154297, + "learning_rate": 9.9813009981301e-06, + "loss": 0.3811, + "step": 18707 + }, + { + "epoch": 8.697350069735007, + "grad_norm": 4.470688343048096, + "learning_rate": 9.98129999813e-06, + "loss": 0.1293, + "step": 18708 + }, + { + "epoch": 8.697814969781497, + "grad_norm": 5.266422748565674, + "learning_rate": 9.9812989981299e-06, + "loss": 0.136, + "step": 18709 + }, + { + "epoch": 8.698279869827987, + "grad_norm": 4.42458438873291, + "learning_rate": 9.9812979981298e-06, + "loss": 0.0761, + "step": 18710 + }, + { + "epoch": 8.698744769874477, + "grad_norm": 5.999748706817627, + "learning_rate": 9.981296998129701e-06, + "loss": 0.1155, + "step": 18711 + }, + { + "epoch": 8.699209669920966, + "grad_norm": 5.381725788116455, + "learning_rate": 9.981295998129602e-06, + "loss": 0.0946, + "step": 18712 + }, + { + "epoch": 8.699674569967456, + "grad_norm": 8.779162406921387, + "learning_rate": 9.9812949981295e-06, + "loss": 0.2202, + "step": 18713 + }, + { + "epoch": 8.700139470013948, + "grad_norm": 3.850789785385132, + "learning_rate": 9.981293998129401e-06, + "loss": 0.048, + "step": 18714 + }, + { + "epoch": 8.700604370060438, + "grad_norm": 4.9304680824279785, + "learning_rate": 9.9812929981293e-06, + "loss": 0.1227, + "step": 18715 + }, + { + "epoch": 8.701069270106927, + "grad_norm": 3.11333966255188, + "learning_rate": 9.981291998129201e-06, + "loss": 0.0708, + "step": 18716 + }, + { + "epoch": 8.701534170153417, + "grad_norm": 8.863472938537598, + "learning_rate": 9.9812909981291e-06, + "loss": 0.1352, + "step": 18717 + }, + { + "epoch": 8.701999070199907, + "grad_norm": 5.394630432128906, + "learning_rate": 9.981289998129e-06, + "loss": 0.103, + "step": 18718 + }, + { + "epoch": 8.702463970246397, + "grad_norm": 9.184233665466309, + "learning_rate": 9.981288998128901e-06, + "loss": 0.2257, + "step": 18719 + }, + { + "epoch": 8.702928870292887, + "grad_norm": 1.2528223991394043, + "learning_rate": 9.9812879981288e-06, + "loss": 0.0118, + "step": 18720 + }, + { + "epoch": 8.703393770339376, + "grad_norm": 3.4312872886657715, + "learning_rate": 9.9812869981287e-06, + "loss": 0.0408, + "step": 18721 + }, + { + "epoch": 8.703858670385866, + "grad_norm": 3.03606915473938, + "learning_rate": 9.9812859981286e-06, + "loss": 0.0479, + "step": 18722 + }, + { + "epoch": 8.704323570432358, + "grad_norm": 8.093612670898438, + "learning_rate": 9.9812849981285e-06, + "loss": 0.1726, + "step": 18723 + }, + { + "epoch": 8.704788470478848, + "grad_norm": 4.322637557983398, + "learning_rate": 9.981283998128401e-06, + "loss": 0.1383, + "step": 18724 + }, + { + "epoch": 8.705253370525337, + "grad_norm": 8.289573669433594, + "learning_rate": 9.981282998128302e-06, + "loss": 0.1048, + "step": 18725 + }, + { + "epoch": 8.705718270571827, + "grad_norm": 3.516892194747925, + "learning_rate": 9.9812819981282e-06, + "loss": 0.0323, + "step": 18726 + }, + { + "epoch": 8.706183170618317, + "grad_norm": 4.616891384124756, + "learning_rate": 9.981280998128101e-06, + "loss": 0.0897, + "step": 18727 + }, + { + "epoch": 8.706648070664807, + "grad_norm": 4.810426235198975, + "learning_rate": 9.981279998128e-06, + "loss": 0.0506, + "step": 18728 + }, + { + "epoch": 8.707112970711297, + "grad_norm": 0.4328469932079315, + "learning_rate": 9.9812789981279e-06, + "loss": 0.0052, + "step": 18729 + }, + { + "epoch": 8.707577870757786, + "grad_norm": 4.364994525909424, + "learning_rate": 9.9812779981278e-06, + "loss": 0.0676, + "step": 18730 + }, + { + "epoch": 8.708042770804276, + "grad_norm": 3.8566811084747314, + "learning_rate": 9.9812769981277e-06, + "loss": 0.06, + "step": 18731 + }, + { + "epoch": 8.708507670850768, + "grad_norm": 3.9942753314971924, + "learning_rate": 9.981275998127601e-06, + "loss": 0.0608, + "step": 18732 + }, + { + "epoch": 8.708972570897258, + "grad_norm": 5.967846393585205, + "learning_rate": 9.9812749981275e-06, + "loss": 0.1602, + "step": 18733 + }, + { + "epoch": 8.709437470943747, + "grad_norm": 1.432534098625183, + "learning_rate": 9.981273998127401e-06, + "loss": 0.0112, + "step": 18734 + }, + { + "epoch": 8.709902370990237, + "grad_norm": 4.382500648498535, + "learning_rate": 9.9812729981273e-06, + "loss": 0.0674, + "step": 18735 + }, + { + "epoch": 8.710367271036727, + "grad_norm": 3.8286449909210205, + "learning_rate": 9.9812719981272e-06, + "loss": 0.0829, + "step": 18736 + }, + { + "epoch": 8.710832171083217, + "grad_norm": 4.135752201080322, + "learning_rate": 9.981270998127101e-06, + "loss": 0.0579, + "step": 18737 + }, + { + "epoch": 8.711297071129707, + "grad_norm": 7.318904399871826, + "learning_rate": 9.981269998127002e-06, + "loss": 0.1432, + "step": 18738 + }, + { + "epoch": 8.711761971176196, + "grad_norm": 3.02742600440979, + "learning_rate": 9.9812689981269e-06, + "loss": 0.0383, + "step": 18739 + }, + { + "epoch": 8.712226871222686, + "grad_norm": 4.20548677444458, + "learning_rate": 9.9812679981268e-06, + "loss": 0.0622, + "step": 18740 + }, + { + "epoch": 8.712691771269178, + "grad_norm": 8.84782600402832, + "learning_rate": 9.9812669981267e-06, + "loss": 0.3884, + "step": 18741 + }, + { + "epoch": 8.713156671315668, + "grad_norm": 2.801135540008545, + "learning_rate": 9.9812659981266e-06, + "loss": 0.0241, + "step": 18742 + }, + { + "epoch": 8.713621571362157, + "grad_norm": 5.898066997528076, + "learning_rate": 9.9812649981265e-06, + "loss": 0.0851, + "step": 18743 + }, + { + "epoch": 8.714086471408647, + "grad_norm": 3.2510838508605957, + "learning_rate": 9.9812639981264e-06, + "loss": 0.0328, + "step": 18744 + }, + { + "epoch": 8.714551371455137, + "grad_norm": 13.45289134979248, + "learning_rate": 9.981262998126301e-06, + "loss": 0.4391, + "step": 18745 + }, + { + "epoch": 8.715016271501627, + "grad_norm": 1.9399946928024292, + "learning_rate": 9.9812619981262e-06, + "loss": 0.0203, + "step": 18746 + }, + { + "epoch": 8.715481171548117, + "grad_norm": 6.825377941131592, + "learning_rate": 9.981260998126101e-06, + "loss": 0.1309, + "step": 18747 + }, + { + "epoch": 8.715946071594606, + "grad_norm": 4.121970176696777, + "learning_rate": 9.981259998126e-06, + "loss": 0.085, + "step": 18748 + }, + { + "epoch": 8.716410971641096, + "grad_norm": 4.821796894073486, + "learning_rate": 9.9812589981259e-06, + "loss": 0.0776, + "step": 18749 + }, + { + "epoch": 8.716875871687588, + "grad_norm": 4.953238487243652, + "learning_rate": 9.981257998125801e-06, + "loss": 0.0504, + "step": 18750 + }, + { + "epoch": 8.717340771734078, + "grad_norm": 8.664447784423828, + "learning_rate": 9.9812569981257e-06, + "loss": 0.1448, + "step": 18751 + }, + { + "epoch": 8.717805671780567, + "grad_norm": 2.5876693725585938, + "learning_rate": 9.981255998125601e-06, + "loss": 0.0282, + "step": 18752 + }, + { + "epoch": 8.718270571827057, + "grad_norm": 5.467107772827148, + "learning_rate": 9.9812549981255e-06, + "loss": 0.0836, + "step": 18753 + }, + { + "epoch": 8.718735471873547, + "grad_norm": 10.788052558898926, + "learning_rate": 9.9812539981254e-06, + "loss": 0.1349, + "step": 18754 + }, + { + "epoch": 8.719200371920037, + "grad_norm": 4.071966648101807, + "learning_rate": 9.981252998125301e-06, + "loss": 0.0343, + "step": 18755 + }, + { + "epoch": 8.719665271966527, + "grad_norm": 4.687928676605225, + "learning_rate": 9.9812519981252e-06, + "loss": 0.093, + "step": 18756 + }, + { + "epoch": 8.720130172013016, + "grad_norm": 3.481062650680542, + "learning_rate": 9.9812509981251e-06, + "loss": 0.0385, + "step": 18757 + }, + { + "epoch": 8.720595072059508, + "grad_norm": 5.470849990844727, + "learning_rate": 9.981249998125001e-06, + "loss": 0.1153, + "step": 18758 + }, + { + "epoch": 8.721059972105998, + "grad_norm": 6.0237274169921875, + "learning_rate": 9.9812489981249e-06, + "loss": 0.106, + "step": 18759 + }, + { + "epoch": 8.721524872152488, + "grad_norm": 5.626230239868164, + "learning_rate": 9.981247998124801e-06, + "loss": 0.0789, + "step": 18760 + }, + { + "epoch": 8.721989772198977, + "grad_norm": 6.414767265319824, + "learning_rate": 9.9812469981247e-06, + "loss": 0.1114, + "step": 18761 + }, + { + "epoch": 8.722454672245467, + "grad_norm": 6.190150260925293, + "learning_rate": 9.9812459981246e-06, + "loss": 0.1481, + "step": 18762 + }, + { + "epoch": 8.722919572291957, + "grad_norm": 4.68493127822876, + "learning_rate": 9.981244998124501e-06, + "loss": 0.1154, + "step": 18763 + }, + { + "epoch": 8.723384472338447, + "grad_norm": 7.588446140289307, + "learning_rate": 9.9812439981244e-06, + "loss": 0.15, + "step": 18764 + }, + { + "epoch": 8.723849372384937, + "grad_norm": 5.040651798248291, + "learning_rate": 9.981242998124301e-06, + "loss": 0.1195, + "step": 18765 + }, + { + "epoch": 8.724314272431426, + "grad_norm": 8.008138656616211, + "learning_rate": 9.9812419981242e-06, + "loss": 0.1691, + "step": 18766 + }, + { + "epoch": 8.724779172477918, + "grad_norm": 4.7246994972229, + "learning_rate": 9.9812409981241e-06, + "loss": 0.1627, + "step": 18767 + }, + { + "epoch": 8.725244072524408, + "grad_norm": 4.762235641479492, + "learning_rate": 9.981239998124001e-06, + "loss": 0.0955, + "step": 18768 + }, + { + "epoch": 8.725708972570898, + "grad_norm": 3.5776443481445312, + "learning_rate": 9.981238998123902e-06, + "loss": 0.0691, + "step": 18769 + }, + { + "epoch": 8.726173872617387, + "grad_norm": 3.7406859397888184, + "learning_rate": 9.981237998123801e-06, + "loss": 0.0408, + "step": 18770 + }, + { + "epoch": 8.726638772663877, + "grad_norm": 4.838192462921143, + "learning_rate": 9.981236998123702e-06, + "loss": 0.1116, + "step": 18771 + }, + { + "epoch": 8.727103672710367, + "grad_norm": 8.390271186828613, + "learning_rate": 9.9812359981236e-06, + "loss": 0.2899, + "step": 18772 + }, + { + "epoch": 8.727568572756857, + "grad_norm": 6.758922100067139, + "learning_rate": 9.9812349981235e-06, + "loss": 0.1653, + "step": 18773 + }, + { + "epoch": 8.728033472803347, + "grad_norm": 4.500243186950684, + "learning_rate": 9.9812339981234e-06, + "loss": 0.1029, + "step": 18774 + }, + { + "epoch": 8.728498372849838, + "grad_norm": 5.731347560882568, + "learning_rate": 9.9812329981233e-06, + "loss": 0.1264, + "step": 18775 + }, + { + "epoch": 8.728963272896328, + "grad_norm": 5.177783966064453, + "learning_rate": 9.981231998123201e-06, + "loss": 0.1393, + "step": 18776 + }, + { + "epoch": 8.729428172942818, + "grad_norm": 3.120988607406616, + "learning_rate": 9.9812309981231e-06, + "loss": 0.0405, + "step": 18777 + }, + { + "epoch": 8.729893072989308, + "grad_norm": 3.8083510398864746, + "learning_rate": 9.981229998123001e-06, + "loss": 0.0586, + "step": 18778 + }, + { + "epoch": 8.730357973035797, + "grad_norm": 5.446202278137207, + "learning_rate": 9.9812289981229e-06, + "loss": 0.0958, + "step": 18779 + }, + { + "epoch": 8.730822873082287, + "grad_norm": 5.670170307159424, + "learning_rate": 9.9812279981228e-06, + "loss": 0.0627, + "step": 18780 + }, + { + "epoch": 8.731287773128777, + "grad_norm": 3.6915082931518555, + "learning_rate": 9.981226998122701e-06, + "loss": 0.0373, + "step": 18781 + }, + { + "epoch": 8.731752673175267, + "grad_norm": 3.7252488136291504, + "learning_rate": 9.981225998122602e-06, + "loss": 0.0667, + "step": 18782 + }, + { + "epoch": 8.732217573221757, + "grad_norm": 4.95107889175415, + "learning_rate": 9.981224998122501e-06, + "loss": 0.0562, + "step": 18783 + }, + { + "epoch": 8.732682473268248, + "grad_norm": 2.4731791019439697, + "learning_rate": 9.9812239981224e-06, + "loss": 0.0148, + "step": 18784 + }, + { + "epoch": 8.733147373314738, + "grad_norm": 4.480373859405518, + "learning_rate": 9.9812229981223e-06, + "loss": 0.1213, + "step": 18785 + }, + { + "epoch": 8.733612273361228, + "grad_norm": 3.3554201126098633, + "learning_rate": 9.9812219981222e-06, + "loss": 0.0653, + "step": 18786 + }, + { + "epoch": 8.734077173407718, + "grad_norm": 6.134305000305176, + "learning_rate": 9.9812209981221e-06, + "loss": 0.0623, + "step": 18787 + }, + { + "epoch": 8.734542073454207, + "grad_norm": 5.556762218475342, + "learning_rate": 9.981219998122001e-06, + "loss": 0.2265, + "step": 18788 + }, + { + "epoch": 8.735006973500697, + "grad_norm": 3.137349843978882, + "learning_rate": 9.981218998121902e-06, + "loss": 0.038, + "step": 18789 + }, + { + "epoch": 8.735471873547187, + "grad_norm": 4.043558120727539, + "learning_rate": 9.9812179981218e-06, + "loss": 0.1752, + "step": 18790 + }, + { + "epoch": 8.735936773593677, + "grad_norm": 4.265885829925537, + "learning_rate": 9.981216998121701e-06, + "loss": 0.0953, + "step": 18791 + }, + { + "epoch": 8.736401673640167, + "grad_norm": 3.7236039638519287, + "learning_rate": 9.9812159981216e-06, + "loss": 0.1081, + "step": 18792 + }, + { + "epoch": 8.736866573686658, + "grad_norm": 3.399068593978882, + "learning_rate": 9.9812149981215e-06, + "loss": 0.0396, + "step": 18793 + }, + { + "epoch": 8.737331473733148, + "grad_norm": 5.442551136016846, + "learning_rate": 9.981213998121401e-06, + "loss": 0.1641, + "step": 18794 + }, + { + "epoch": 8.737796373779638, + "grad_norm": 8.935752868652344, + "learning_rate": 9.9812129981213e-06, + "loss": 0.2264, + "step": 18795 + }, + { + "epoch": 8.738261273826128, + "grad_norm": 6.585389137268066, + "learning_rate": 9.981211998121201e-06, + "loss": 0.2099, + "step": 18796 + }, + { + "epoch": 8.738726173872617, + "grad_norm": 3.892271041870117, + "learning_rate": 9.9812109981211e-06, + "loss": 0.0565, + "step": 18797 + }, + { + "epoch": 8.739191073919107, + "grad_norm": 4.6679606437683105, + "learning_rate": 9.981209998121e-06, + "loss": 0.0924, + "step": 18798 + }, + { + "epoch": 8.739655973965597, + "grad_norm": 7.857702255249023, + "learning_rate": 9.9812089981209e-06, + "loss": 0.2999, + "step": 18799 + }, + { + "epoch": 8.740120874012087, + "grad_norm": 1.8774622678756714, + "learning_rate": 9.9812079981208e-06, + "loss": 0.0166, + "step": 18800 + }, + { + "epoch": 8.740585774058577, + "grad_norm": 5.353076457977295, + "learning_rate": 9.981206998120701e-06, + "loss": 0.1206, + "step": 18801 + }, + { + "epoch": 8.741050674105068, + "grad_norm": 3.8969316482543945, + "learning_rate": 9.981205998120602e-06, + "loss": 0.0828, + "step": 18802 + }, + { + "epoch": 8.741515574151558, + "grad_norm": 1.841274619102478, + "learning_rate": 9.9812049981205e-06, + "loss": 0.0152, + "step": 18803 + }, + { + "epoch": 8.741980474198048, + "grad_norm": 5.50470495223999, + "learning_rate": 9.9812039981204e-06, + "loss": 0.145, + "step": 18804 + }, + { + "epoch": 8.742445374244538, + "grad_norm": 5.32841682434082, + "learning_rate": 9.9812029981203e-06, + "loss": 0.0898, + "step": 18805 + }, + { + "epoch": 8.742910274291027, + "grad_norm": 3.6826963424682617, + "learning_rate": 9.9812019981202e-06, + "loss": 0.0812, + "step": 18806 + }, + { + "epoch": 8.743375174337517, + "grad_norm": 4.81138277053833, + "learning_rate": 9.981200998120101e-06, + "loss": 0.0792, + "step": 18807 + }, + { + "epoch": 8.743840074384007, + "grad_norm": 4.055922985076904, + "learning_rate": 9.98119999812e-06, + "loss": 0.0795, + "step": 18808 + }, + { + "epoch": 8.744304974430497, + "grad_norm": 8.839437484741211, + "learning_rate": 9.981198998119901e-06, + "loss": 0.253, + "step": 18809 + }, + { + "epoch": 8.744769874476987, + "grad_norm": 6.7708563804626465, + "learning_rate": 9.9811979981198e-06, + "loss": 0.1744, + "step": 18810 + }, + { + "epoch": 8.745234774523478, + "grad_norm": 5.13870096206665, + "learning_rate": 9.9811969981197e-06, + "loss": 0.0986, + "step": 18811 + }, + { + "epoch": 8.745699674569968, + "grad_norm": 11.731287956237793, + "learning_rate": 9.9811959981196e-06, + "loss": 0.1007, + "step": 18812 + }, + { + "epoch": 8.746164574616458, + "grad_norm": 5.026402473449707, + "learning_rate": 9.9811949981195e-06, + "loss": 0.1458, + "step": 18813 + }, + { + "epoch": 8.746629474662948, + "grad_norm": 9.388193130493164, + "learning_rate": 9.981193998119401e-06, + "loss": 0.2671, + "step": 18814 + }, + { + "epoch": 8.747094374709437, + "grad_norm": 4.5225090980529785, + "learning_rate": 9.9811929981193e-06, + "loss": 0.0787, + "step": 18815 + }, + { + "epoch": 8.747559274755927, + "grad_norm": 1.6717530488967896, + "learning_rate": 9.9811919981192e-06, + "loss": 0.0215, + "step": 18816 + }, + { + "epoch": 8.748024174802417, + "grad_norm": 4.191883087158203, + "learning_rate": 9.9811909981191e-06, + "loss": 0.0746, + "step": 18817 + }, + { + "epoch": 8.748489074848907, + "grad_norm": 2.602400779724121, + "learning_rate": 9.981189998119e-06, + "loss": 0.0221, + "step": 18818 + }, + { + "epoch": 8.748953974895397, + "grad_norm": 3.4494566917419434, + "learning_rate": 9.981188998118901e-06, + "loss": 0.1015, + "step": 18819 + }, + { + "epoch": 8.749418874941888, + "grad_norm": 5.5539140701293945, + "learning_rate": 9.981187998118802e-06, + "loss": 0.1105, + "step": 18820 + }, + { + "epoch": 8.749883774988378, + "grad_norm": 9.084471702575684, + "learning_rate": 9.9811869981187e-06, + "loss": 0.1351, + "step": 18821 + }, + { + "epoch": 8.750348675034868, + "grad_norm": 5.215773582458496, + "learning_rate": 9.981185998118601e-06, + "loss": 0.0686, + "step": 18822 + }, + { + "epoch": 8.750813575081358, + "grad_norm": 4.979523658752441, + "learning_rate": 9.9811849981185e-06, + "loss": 0.0745, + "step": 18823 + }, + { + "epoch": 8.751278475127847, + "grad_norm": 7.016881465911865, + "learning_rate": 9.9811839981184e-06, + "loss": 0.1613, + "step": 18824 + }, + { + "epoch": 8.751743375174337, + "grad_norm": 7.357120990753174, + "learning_rate": 9.9811829981183e-06, + "loss": 0.1091, + "step": 18825 + }, + { + "epoch": 8.752208275220827, + "grad_norm": 5.837925434112549, + "learning_rate": 9.9811819981182e-06, + "loss": 0.083, + "step": 18826 + }, + { + "epoch": 8.752673175267317, + "grad_norm": 2.8680684566497803, + "learning_rate": 9.981180998118101e-06, + "loss": 0.0382, + "step": 18827 + }, + { + "epoch": 8.753138075313807, + "grad_norm": 6.255086421966553, + "learning_rate": 9.981179998118e-06, + "loss": 0.1096, + "step": 18828 + }, + { + "epoch": 8.753602975360298, + "grad_norm": 4.101586818695068, + "learning_rate": 9.9811789981179e-06, + "loss": 0.0413, + "step": 18829 + }, + { + "epoch": 8.754067875406788, + "grad_norm": 8.939413070678711, + "learning_rate": 9.9811779981178e-06, + "loss": 0.3233, + "step": 18830 + }, + { + "epoch": 8.754532775453278, + "grad_norm": 5.110038757324219, + "learning_rate": 9.9811769981177e-06, + "loss": 0.1066, + "step": 18831 + }, + { + "epoch": 8.754997675499768, + "grad_norm": 4.816323280334473, + "learning_rate": 9.981175998117601e-06, + "loss": 0.0478, + "step": 18832 + }, + { + "epoch": 8.755462575546257, + "grad_norm": 5.189713001251221, + "learning_rate": 9.981174998117502e-06, + "loss": 0.0654, + "step": 18833 + }, + { + "epoch": 8.755927475592747, + "grad_norm": 4.430155277252197, + "learning_rate": 9.9811739981174e-06, + "loss": 0.0763, + "step": 18834 + }, + { + "epoch": 8.756392375639237, + "grad_norm": 6.496611595153809, + "learning_rate": 9.981172998117301e-06, + "loss": 0.1766, + "step": 18835 + }, + { + "epoch": 8.756857275685727, + "grad_norm": 6.083935260772705, + "learning_rate": 9.9811719981172e-06, + "loss": 0.102, + "step": 18836 + }, + { + "epoch": 8.757322175732218, + "grad_norm": 4.550170421600342, + "learning_rate": 9.981170998117101e-06, + "loss": 0.112, + "step": 18837 + }, + { + "epoch": 8.757787075778708, + "grad_norm": 6.4749064445495605, + "learning_rate": 9.981169998117002e-06, + "loss": 0.2322, + "step": 18838 + }, + { + "epoch": 8.758251975825198, + "grad_norm": 5.161174297332764, + "learning_rate": 9.9811689981169e-06, + "loss": 0.1077, + "step": 18839 + }, + { + "epoch": 8.758716875871688, + "grad_norm": 3.099825382232666, + "learning_rate": 9.981167998116801e-06, + "loss": 0.0332, + "step": 18840 + }, + { + "epoch": 8.759181775918178, + "grad_norm": 5.195369243621826, + "learning_rate": 9.9811669981167e-06, + "loss": 0.1477, + "step": 18841 + }, + { + "epoch": 8.759646675964667, + "grad_norm": 4.250961780548096, + "learning_rate": 9.9811659981166e-06, + "loss": 0.0693, + "step": 18842 + }, + { + "epoch": 8.760111576011157, + "grad_norm": 3.275023937225342, + "learning_rate": 9.9811649981165e-06, + "loss": 0.0423, + "step": 18843 + }, + { + "epoch": 8.760576476057647, + "grad_norm": 5.705878257751465, + "learning_rate": 9.9811639981164e-06, + "loss": 0.0654, + "step": 18844 + }, + { + "epoch": 8.761041376104137, + "grad_norm": 5.5680460929870605, + "learning_rate": 9.981162998116301e-06, + "loss": 0.1445, + "step": 18845 + }, + { + "epoch": 8.761506276150628, + "grad_norm": 3.442540407180786, + "learning_rate": 9.981161998116202e-06, + "loss": 0.0556, + "step": 18846 + }, + { + "epoch": 8.761971176197118, + "grad_norm": 3.2908987998962402, + "learning_rate": 9.9811609981161e-06, + "loss": 0.0356, + "step": 18847 + }, + { + "epoch": 8.762436076243608, + "grad_norm": 2.7635738849639893, + "learning_rate": 9.981159998116e-06, + "loss": 0.0317, + "step": 18848 + }, + { + "epoch": 8.762900976290098, + "grad_norm": 1.7155898809432983, + "learning_rate": 9.9811589981159e-06, + "loss": 0.012, + "step": 18849 + }, + { + "epoch": 8.763365876336588, + "grad_norm": 10.237297058105469, + "learning_rate": 9.981157998115801e-06, + "loss": 0.0581, + "step": 18850 + }, + { + "epoch": 8.763830776383077, + "grad_norm": 5.356700897216797, + "learning_rate": 9.981156998115702e-06, + "loss": 0.103, + "step": 18851 + }, + { + "epoch": 8.764295676429567, + "grad_norm": 11.548722267150879, + "learning_rate": 9.9811559981156e-06, + "loss": 0.3267, + "step": 18852 + }, + { + "epoch": 8.764760576476057, + "grad_norm": 5.763779640197754, + "learning_rate": 9.981154998115501e-06, + "loss": 0.1299, + "step": 18853 + }, + { + "epoch": 8.765225476522549, + "grad_norm": 7.089320659637451, + "learning_rate": 9.9811539981154e-06, + "loss": 0.152, + "step": 18854 + }, + { + "epoch": 8.765690376569038, + "grad_norm": 4.436183452606201, + "learning_rate": 9.981152998115301e-06, + "loss": 0.0629, + "step": 18855 + }, + { + "epoch": 8.766155276615528, + "grad_norm": 4.045369625091553, + "learning_rate": 9.9811519981152e-06, + "loss": 0.0894, + "step": 18856 + }, + { + "epoch": 8.766620176662018, + "grad_norm": 6.513494491577148, + "learning_rate": 9.9811509981151e-06, + "loss": 0.2421, + "step": 18857 + }, + { + "epoch": 8.767085076708508, + "grad_norm": 9.27441120147705, + "learning_rate": 9.981149998115001e-06, + "loss": 0.1743, + "step": 18858 + }, + { + "epoch": 8.767549976754998, + "grad_norm": 7.757122993469238, + "learning_rate": 9.9811489981149e-06, + "loss": 0.1327, + "step": 18859 + }, + { + "epoch": 8.768014876801487, + "grad_norm": 8.276812553405762, + "learning_rate": 9.9811479981148e-06, + "loss": 0.1582, + "step": 18860 + }, + { + "epoch": 8.768479776847977, + "grad_norm": 6.139682292938232, + "learning_rate": 9.9811469981147e-06, + "loss": 0.1915, + "step": 18861 + }, + { + "epoch": 8.768944676894467, + "grad_norm": 6.918138027191162, + "learning_rate": 9.9811459981146e-06, + "loss": 0.0853, + "step": 18862 + }, + { + "epoch": 8.769409576940959, + "grad_norm": 5.363656997680664, + "learning_rate": 9.981144998114501e-06, + "loss": 0.1035, + "step": 18863 + }, + { + "epoch": 8.769874476987448, + "grad_norm": 6.396173477172852, + "learning_rate": 9.981143998114402e-06, + "loss": 0.1257, + "step": 18864 + }, + { + "epoch": 8.770339377033938, + "grad_norm": 2.7537519931793213, + "learning_rate": 9.9811429981143e-06, + "loss": 0.0291, + "step": 18865 + }, + { + "epoch": 8.770804277080428, + "grad_norm": 2.9165279865264893, + "learning_rate": 9.981141998114201e-06, + "loss": 0.0252, + "step": 18866 + }, + { + "epoch": 8.771269177126918, + "grad_norm": 4.496919631958008, + "learning_rate": 9.9811409981141e-06, + "loss": 0.0681, + "step": 18867 + }, + { + "epoch": 8.771734077173408, + "grad_norm": 4.759772300720215, + "learning_rate": 9.981139998114001e-06, + "loss": 0.1041, + "step": 18868 + }, + { + "epoch": 8.772198977219897, + "grad_norm": 3.3201286792755127, + "learning_rate": 9.9811389981139e-06, + "loss": 0.0444, + "step": 18869 + }, + { + "epoch": 8.772663877266387, + "grad_norm": 2.4280436038970947, + "learning_rate": 9.9811379981138e-06, + "loss": 0.021, + "step": 18870 + }, + { + "epoch": 8.773128777312877, + "grad_norm": 4.639920711517334, + "learning_rate": 9.981136998113701e-06, + "loss": 0.0838, + "step": 18871 + }, + { + "epoch": 8.773593677359369, + "grad_norm": 1.3349144458770752, + "learning_rate": 9.9811359981136e-06, + "loss": 0.0111, + "step": 18872 + }, + { + "epoch": 8.774058577405858, + "grad_norm": 3.759114980697632, + "learning_rate": 9.981134998113501e-06, + "loss": 0.0485, + "step": 18873 + }, + { + "epoch": 8.774523477452348, + "grad_norm": 4.911850929260254, + "learning_rate": 9.9811339981134e-06, + "loss": 0.1059, + "step": 18874 + }, + { + "epoch": 8.774988377498838, + "grad_norm": 3.9337685108184814, + "learning_rate": 9.9811329981133e-06, + "loss": 0.0482, + "step": 18875 + }, + { + "epoch": 8.775453277545328, + "grad_norm": 4.594339370727539, + "learning_rate": 9.981131998113201e-06, + "loss": 0.1188, + "step": 18876 + }, + { + "epoch": 8.775918177591818, + "grad_norm": 3.364509105682373, + "learning_rate": 9.981130998113102e-06, + "loss": 0.0545, + "step": 18877 + }, + { + "epoch": 8.776383077638307, + "grad_norm": 3.4154629707336426, + "learning_rate": 9.981129998113e-06, + "loss": 0.0432, + "step": 18878 + }, + { + "epoch": 8.776847977684797, + "grad_norm": 7.467628479003906, + "learning_rate": 9.981128998112901e-06, + "loss": 0.1504, + "step": 18879 + }, + { + "epoch": 8.777312877731287, + "grad_norm": 5.344357013702393, + "learning_rate": 9.9811279981128e-06, + "loss": 0.1613, + "step": 18880 + }, + { + "epoch": 8.777777777777779, + "grad_norm": 6.397130966186523, + "learning_rate": 9.9811269981127e-06, + "loss": 0.2193, + "step": 18881 + }, + { + "epoch": 8.778242677824268, + "grad_norm": 3.817085027694702, + "learning_rate": 9.9811259981126e-06, + "loss": 0.0796, + "step": 18882 + }, + { + "epoch": 8.778707577870758, + "grad_norm": 3.993490695953369, + "learning_rate": 9.9811249981125e-06, + "loss": 0.0928, + "step": 18883 + }, + { + "epoch": 8.779172477917248, + "grad_norm": 5.048955917358398, + "learning_rate": 9.981123998112401e-06, + "loss": 0.0815, + "step": 18884 + }, + { + "epoch": 8.779637377963738, + "grad_norm": 6.9421281814575195, + "learning_rate": 9.9811229981123e-06, + "loss": 0.1387, + "step": 18885 + }, + { + "epoch": 8.780102278010228, + "grad_norm": 11.31933879852295, + "learning_rate": 9.981121998112201e-06, + "loss": 0.0636, + "step": 18886 + }, + { + "epoch": 8.780567178056717, + "grad_norm": 4.481043815612793, + "learning_rate": 9.9811209981121e-06, + "loss": 0.0673, + "step": 18887 + }, + { + "epoch": 8.781032078103207, + "grad_norm": 3.4976279735565186, + "learning_rate": 9.981119998112e-06, + "loss": 0.0406, + "step": 18888 + }, + { + "epoch": 8.781496978149697, + "grad_norm": 7.481789588928223, + "learning_rate": 9.981118998111901e-06, + "loss": 0.1964, + "step": 18889 + }, + { + "epoch": 8.781961878196189, + "grad_norm": 6.698208332061768, + "learning_rate": 9.981117998111802e-06, + "loss": 0.1623, + "step": 18890 + }, + { + "epoch": 8.782426778242678, + "grad_norm": 6.59843111038208, + "learning_rate": 9.981116998111701e-06, + "loss": 0.1062, + "step": 18891 + }, + { + "epoch": 8.782891678289168, + "grad_norm": 2.9930176734924316, + "learning_rate": 9.9811159981116e-06, + "loss": 0.0368, + "step": 18892 + }, + { + "epoch": 8.783356578335658, + "grad_norm": 3.107105016708374, + "learning_rate": 9.9811149981115e-06, + "loss": 0.0554, + "step": 18893 + }, + { + "epoch": 8.783821478382148, + "grad_norm": 6.577373504638672, + "learning_rate": 9.9811139981114e-06, + "loss": 0.2326, + "step": 18894 + }, + { + "epoch": 8.784286378428638, + "grad_norm": 5.334037780761719, + "learning_rate": 9.9811129981113e-06, + "loss": 0.0749, + "step": 18895 + }, + { + "epoch": 8.784751278475127, + "grad_norm": 7.673763751983643, + "learning_rate": 9.9811119981112e-06, + "loss": 0.0826, + "step": 18896 + }, + { + "epoch": 8.785216178521617, + "grad_norm": 4.6057000160217285, + "learning_rate": 9.981110998111101e-06, + "loss": 0.0844, + "step": 18897 + }, + { + "epoch": 8.785681078568107, + "grad_norm": 3.7714569568634033, + "learning_rate": 9.981109998111e-06, + "loss": 0.0386, + "step": 18898 + }, + { + "epoch": 8.786145978614599, + "grad_norm": 4.121251583099365, + "learning_rate": 9.981108998110901e-06, + "loss": 0.0791, + "step": 18899 + }, + { + "epoch": 8.786610878661088, + "grad_norm": 6.734982013702393, + "learning_rate": 9.9811079981108e-06, + "loss": 0.1925, + "step": 18900 + }, + { + "epoch": 8.787075778707578, + "grad_norm": 4.911334991455078, + "learning_rate": 9.9811069981107e-06, + "loss": 0.1267, + "step": 18901 + }, + { + "epoch": 8.787540678754068, + "grad_norm": 4.282605171203613, + "learning_rate": 9.981105998110601e-06, + "loss": 0.0844, + "step": 18902 + }, + { + "epoch": 8.788005578800558, + "grad_norm": 3.432154893875122, + "learning_rate": 9.9811049981105e-06, + "loss": 0.0631, + "step": 18903 + }, + { + "epoch": 8.788470478847048, + "grad_norm": 6.943142414093018, + "learning_rate": 9.981103998110401e-06, + "loss": 0.1339, + "step": 18904 + }, + { + "epoch": 8.788935378893537, + "grad_norm": 12.649923324584961, + "learning_rate": 9.9811029981103e-06, + "loss": 0.2716, + "step": 18905 + }, + { + "epoch": 8.789400278940027, + "grad_norm": 5.538649559020996, + "learning_rate": 9.9811019981102e-06, + "loss": 0.1477, + "step": 18906 + }, + { + "epoch": 8.789865178986517, + "grad_norm": 4.0153703689575195, + "learning_rate": 9.9811009981101e-06, + "loss": 0.0637, + "step": 18907 + }, + { + "epoch": 8.790330079033009, + "grad_norm": 5.551448345184326, + "learning_rate": 9.98109999811e-06, + "loss": 0.1656, + "step": 18908 + }, + { + "epoch": 8.790794979079498, + "grad_norm": 4.283625602722168, + "learning_rate": 9.9810989981099e-06, + "loss": 0.0534, + "step": 18909 + }, + { + "epoch": 8.791259879125988, + "grad_norm": 3.954430341720581, + "learning_rate": 9.981097998109801e-06, + "loss": 0.0769, + "step": 18910 + }, + { + "epoch": 8.791724779172478, + "grad_norm": 4.490479946136475, + "learning_rate": 9.9810969981097e-06, + "loss": 0.1423, + "step": 18911 + }, + { + "epoch": 8.792189679218968, + "grad_norm": 2.3418641090393066, + "learning_rate": 9.981095998109601e-06, + "loss": 0.0341, + "step": 18912 + }, + { + "epoch": 8.792654579265458, + "grad_norm": 4.174225330352783, + "learning_rate": 9.9810949981095e-06, + "loss": 0.0599, + "step": 18913 + }, + { + "epoch": 8.793119479311947, + "grad_norm": 6.886867046356201, + "learning_rate": 9.9810939981094e-06, + "loss": 0.1013, + "step": 18914 + }, + { + "epoch": 8.793584379358437, + "grad_norm": 4.824507713317871, + "learning_rate": 9.981092998109301e-06, + "loss": 0.1664, + "step": 18915 + }, + { + "epoch": 8.794049279404929, + "grad_norm": 3.2650327682495117, + "learning_rate": 9.9810919981092e-06, + "loss": 0.0336, + "step": 18916 + }, + { + "epoch": 8.794514179451419, + "grad_norm": 4.347447395324707, + "learning_rate": 9.981090998109101e-06, + "loss": 0.1232, + "step": 18917 + }, + { + "epoch": 8.794979079497908, + "grad_norm": 2.7662124633789062, + "learning_rate": 9.981089998109e-06, + "loss": 0.0248, + "step": 18918 + }, + { + "epoch": 8.795443979544398, + "grad_norm": 3.0933616161346436, + "learning_rate": 9.9810889981089e-06, + "loss": 0.0499, + "step": 18919 + }, + { + "epoch": 8.795908879590888, + "grad_norm": 7.0821452140808105, + "learning_rate": 9.981087998108801e-06, + "loss": 0.2965, + "step": 18920 + }, + { + "epoch": 8.796373779637378, + "grad_norm": 7.536679744720459, + "learning_rate": 9.9810869981087e-06, + "loss": 0.2091, + "step": 18921 + }, + { + "epoch": 8.796838679683868, + "grad_norm": 5.271977424621582, + "learning_rate": 9.981085998108601e-06, + "loss": 0.0988, + "step": 18922 + }, + { + "epoch": 8.797303579730357, + "grad_norm": 6.317158222198486, + "learning_rate": 9.981084998108502e-06, + "loss": 0.1132, + "step": 18923 + }, + { + "epoch": 8.797768479776847, + "grad_norm": 4.52705717086792, + "learning_rate": 9.9810839981084e-06, + "loss": 0.0519, + "step": 18924 + }, + { + "epoch": 8.798233379823339, + "grad_norm": 8.92803955078125, + "learning_rate": 9.9810829981083e-06, + "loss": 0.1695, + "step": 18925 + }, + { + "epoch": 8.798698279869829, + "grad_norm": 9.297327995300293, + "learning_rate": 9.9810819981082e-06, + "loss": 0.2089, + "step": 18926 + }, + { + "epoch": 8.799163179916318, + "grad_norm": 8.130337715148926, + "learning_rate": 9.9810809981081e-06, + "loss": 0.164, + "step": 18927 + }, + { + "epoch": 8.799628079962808, + "grad_norm": 3.231649398803711, + "learning_rate": 9.981079998108001e-06, + "loss": 0.0566, + "step": 18928 + }, + { + "epoch": 8.800092980009298, + "grad_norm": 4.678538799285889, + "learning_rate": 9.9810789981079e-06, + "loss": 0.0736, + "step": 18929 + }, + { + "epoch": 8.800557880055788, + "grad_norm": 5.501817226409912, + "learning_rate": 9.981077998107801e-06, + "loss": 0.1539, + "step": 18930 + }, + { + "epoch": 8.801022780102278, + "grad_norm": 5.195159435272217, + "learning_rate": 9.9810769981077e-06, + "loss": 0.0669, + "step": 18931 + }, + { + "epoch": 8.801487680148767, + "grad_norm": 10.05361270904541, + "learning_rate": 9.9810759981076e-06, + "loss": 0.0732, + "step": 18932 + }, + { + "epoch": 8.801952580195259, + "grad_norm": 2.4728734493255615, + "learning_rate": 9.981074998107501e-06, + "loss": 0.0284, + "step": 18933 + }, + { + "epoch": 8.802417480241749, + "grad_norm": 3.976738929748535, + "learning_rate": 9.9810739981074e-06, + "loss": 0.0706, + "step": 18934 + }, + { + "epoch": 8.802882380288239, + "grad_norm": 5.686290264129639, + "learning_rate": 9.981072998107301e-06, + "loss": 0.0695, + "step": 18935 + }, + { + "epoch": 8.803347280334728, + "grad_norm": 5.8209404945373535, + "learning_rate": 9.9810719981072e-06, + "loss": 0.1047, + "step": 18936 + }, + { + "epoch": 8.803812180381218, + "grad_norm": 4.780200481414795, + "learning_rate": 9.9810709981071e-06, + "loss": 0.1077, + "step": 18937 + }, + { + "epoch": 8.804277080427708, + "grad_norm": 3.7991974353790283, + "learning_rate": 9.981069998107e-06, + "loss": 0.0736, + "step": 18938 + }, + { + "epoch": 8.804741980474198, + "grad_norm": 2.6413981914520264, + "learning_rate": 9.9810689981069e-06, + "loss": 0.0206, + "step": 18939 + }, + { + "epoch": 8.805206880520688, + "grad_norm": 5.987780570983887, + "learning_rate": 9.981067998106801e-06, + "loss": 0.11, + "step": 18940 + }, + { + "epoch": 8.805671780567177, + "grad_norm": 9.308486938476562, + "learning_rate": 9.981066998106702e-06, + "loss": 0.3238, + "step": 18941 + }, + { + "epoch": 8.806136680613669, + "grad_norm": 4.509479999542236, + "learning_rate": 9.9810659981066e-06, + "loss": 0.097, + "step": 18942 + }, + { + "epoch": 8.806601580660159, + "grad_norm": 8.52534294128418, + "learning_rate": 9.981064998106501e-06, + "loss": 0.1171, + "step": 18943 + }, + { + "epoch": 8.807066480706649, + "grad_norm": 5.123503684997559, + "learning_rate": 9.9810639981064e-06, + "loss": 0.15, + "step": 18944 + }, + { + "epoch": 8.807531380753138, + "grad_norm": 6.5497589111328125, + "learning_rate": 9.9810629981063e-06, + "loss": 0.1592, + "step": 18945 + }, + { + "epoch": 8.807996280799628, + "grad_norm": 6.018131732940674, + "learning_rate": 9.981061998106201e-06, + "loss": 0.075, + "step": 18946 + }, + { + "epoch": 8.808461180846118, + "grad_norm": 6.315598964691162, + "learning_rate": 9.9810609981061e-06, + "loss": 0.1275, + "step": 18947 + }, + { + "epoch": 8.808926080892608, + "grad_norm": 5.491761207580566, + "learning_rate": 9.981059998106001e-06, + "loss": 0.1136, + "step": 18948 + }, + { + "epoch": 8.809390980939098, + "grad_norm": 5.448115825653076, + "learning_rate": 9.9810589981059e-06, + "loss": 0.0531, + "step": 18949 + }, + { + "epoch": 8.809855880985587, + "grad_norm": 4.743195533752441, + "learning_rate": 9.9810579981058e-06, + "loss": 0.0978, + "step": 18950 + }, + { + "epoch": 8.810320781032079, + "grad_norm": 4.883028030395508, + "learning_rate": 9.9810569981057e-06, + "loss": 0.0815, + "step": 18951 + }, + { + "epoch": 8.810785681078569, + "grad_norm": 4.660624027252197, + "learning_rate": 9.9810559981056e-06, + "loss": 0.0676, + "step": 18952 + }, + { + "epoch": 8.811250581125059, + "grad_norm": 8.74268627166748, + "learning_rate": 9.981054998105501e-06, + "loss": 0.1377, + "step": 18953 + }, + { + "epoch": 8.811715481171548, + "grad_norm": 7.2048420906066895, + "learning_rate": 9.981053998105402e-06, + "loss": 0.1144, + "step": 18954 + }, + { + "epoch": 8.812180381218038, + "grad_norm": 3.039578676223755, + "learning_rate": 9.9810529981053e-06, + "loss": 0.0371, + "step": 18955 + }, + { + "epoch": 8.812645281264528, + "grad_norm": 6.431849479675293, + "learning_rate": 9.9810519981052e-06, + "loss": 0.1087, + "step": 18956 + }, + { + "epoch": 8.813110181311018, + "grad_norm": 7.268593788146973, + "learning_rate": 9.9810509981051e-06, + "loss": 0.2115, + "step": 18957 + }, + { + "epoch": 8.813575081357508, + "grad_norm": 3.8345608711242676, + "learning_rate": 9.981049998105001e-06, + "loss": 0.0589, + "step": 18958 + }, + { + "epoch": 8.814039981403997, + "grad_norm": 7.321274757385254, + "learning_rate": 9.981048998104902e-06, + "loss": 0.2542, + "step": 18959 + }, + { + "epoch": 8.814504881450489, + "grad_norm": 4.225701808929443, + "learning_rate": 9.9810479981048e-06, + "loss": 0.063, + "step": 18960 + }, + { + "epoch": 8.814969781496979, + "grad_norm": 1.851575255393982, + "learning_rate": 9.981046998104701e-06, + "loss": 0.0143, + "step": 18961 + }, + { + "epoch": 8.815434681543469, + "grad_norm": 7.042969226837158, + "learning_rate": 9.9810459981046e-06, + "loss": 0.2413, + "step": 18962 + }, + { + "epoch": 8.815899581589958, + "grad_norm": 8.53711223602295, + "learning_rate": 9.9810449981045e-06, + "loss": 0.1563, + "step": 18963 + }, + { + "epoch": 8.816364481636448, + "grad_norm": 4.901017189025879, + "learning_rate": 9.9810439981044e-06, + "loss": 0.1211, + "step": 18964 + }, + { + "epoch": 8.816829381682938, + "grad_norm": 3.234050989151001, + "learning_rate": 9.9810429981043e-06, + "loss": 0.0393, + "step": 18965 + }, + { + "epoch": 8.817294281729428, + "grad_norm": 5.962577819824219, + "learning_rate": 9.981041998104201e-06, + "loss": 0.1338, + "step": 18966 + }, + { + "epoch": 8.817759181775918, + "grad_norm": 4.362493991851807, + "learning_rate": 9.9810409981041e-06, + "loss": 0.0732, + "step": 18967 + }, + { + "epoch": 8.818224081822407, + "grad_norm": 1.745572805404663, + "learning_rate": 9.981039998104e-06, + "loss": 0.0187, + "step": 18968 + }, + { + "epoch": 8.818688981868899, + "grad_norm": 4.716848373413086, + "learning_rate": 9.9810389981039e-06, + "loss": 0.1254, + "step": 18969 + }, + { + "epoch": 8.819153881915389, + "grad_norm": 8.170964241027832, + "learning_rate": 9.9810379981038e-06, + "loss": 0.1978, + "step": 18970 + }, + { + "epoch": 8.819618781961879, + "grad_norm": 6.212221145629883, + "learning_rate": 9.981036998103701e-06, + "loss": 0.0605, + "step": 18971 + }, + { + "epoch": 8.820083682008368, + "grad_norm": 9.490285873413086, + "learning_rate": 9.981035998103602e-06, + "loss": 0.1902, + "step": 18972 + }, + { + "epoch": 8.820548582054858, + "grad_norm": 6.566684722900391, + "learning_rate": 9.9810349981035e-06, + "loss": 0.1907, + "step": 18973 + }, + { + "epoch": 8.821013482101348, + "grad_norm": 2.532851457595825, + "learning_rate": 9.981033998103401e-06, + "loss": 0.0327, + "step": 18974 + }, + { + "epoch": 8.821478382147838, + "grad_norm": 5.918947219848633, + "learning_rate": 9.9810329981033e-06, + "loss": 0.0565, + "step": 18975 + }, + { + "epoch": 8.821943282194328, + "grad_norm": 5.264045238494873, + "learning_rate": 9.981031998103201e-06, + "loss": 0.1318, + "step": 18976 + }, + { + "epoch": 8.822408182240817, + "grad_norm": 5.46170711517334, + "learning_rate": 9.9810309981031e-06, + "loss": 0.1159, + "step": 18977 + }, + { + "epoch": 8.822873082287309, + "grad_norm": 6.744555950164795, + "learning_rate": 9.981029998103e-06, + "loss": 0.2088, + "step": 18978 + }, + { + "epoch": 8.823337982333799, + "grad_norm": 3.731907367706299, + "learning_rate": 9.981028998102901e-06, + "loss": 0.0487, + "step": 18979 + }, + { + "epoch": 8.823802882380289, + "grad_norm": 3.5104782581329346, + "learning_rate": 9.9810279981028e-06, + "loss": 0.077, + "step": 18980 + }, + { + "epoch": 8.824267782426778, + "grad_norm": 5.4689860343933105, + "learning_rate": 9.9810269981027e-06, + "loss": 0.1727, + "step": 18981 + }, + { + "epoch": 8.824732682473268, + "grad_norm": 4.878307342529297, + "learning_rate": 9.9810259981026e-06, + "loss": 0.0806, + "step": 18982 + }, + { + "epoch": 8.825197582519758, + "grad_norm": 3.0218520164489746, + "learning_rate": 9.9810249981025e-06, + "loss": 0.0305, + "step": 18983 + }, + { + "epoch": 8.825662482566248, + "grad_norm": 5.707810401916504, + "learning_rate": 9.981023998102401e-06, + "loss": 0.123, + "step": 18984 + }, + { + "epoch": 8.826127382612738, + "grad_norm": 5.000912189483643, + "learning_rate": 9.981022998102302e-06, + "loss": 0.0874, + "step": 18985 + }, + { + "epoch": 8.826592282659227, + "grad_norm": 5.044889450073242, + "learning_rate": 9.9810219981022e-06, + "loss": 0.0912, + "step": 18986 + }, + { + "epoch": 8.827057182705719, + "grad_norm": 4.753304481506348, + "learning_rate": 9.981020998102101e-06, + "loss": 0.0831, + "step": 18987 + }, + { + "epoch": 8.827522082752209, + "grad_norm": 4.457289695739746, + "learning_rate": 9.981019998102e-06, + "loss": 0.0814, + "step": 18988 + }, + { + "epoch": 8.827986982798699, + "grad_norm": 5.5331501960754395, + "learning_rate": 9.981018998101901e-06, + "loss": 0.0902, + "step": 18989 + }, + { + "epoch": 8.828451882845188, + "grad_norm": 6.200251579284668, + "learning_rate": 9.9810179981018e-06, + "loss": 0.1488, + "step": 18990 + }, + { + "epoch": 8.828916782891678, + "grad_norm": 7.570943355560303, + "learning_rate": 9.9810169981017e-06, + "loss": 0.2945, + "step": 18991 + }, + { + "epoch": 8.829381682938168, + "grad_norm": 2.317094564437866, + "learning_rate": 9.981015998101601e-06, + "loss": 0.0274, + "step": 18992 + }, + { + "epoch": 8.829846582984658, + "grad_norm": 6.218613624572754, + "learning_rate": 9.9810149981015e-06, + "loss": 0.133, + "step": 18993 + }, + { + "epoch": 8.830311483031148, + "grad_norm": 0.578274667263031, + "learning_rate": 9.9810139981014e-06, + "loss": 0.0073, + "step": 18994 + }, + { + "epoch": 8.83077638307764, + "grad_norm": 4.030361652374268, + "learning_rate": 9.9810129981013e-06, + "loss": 0.0854, + "step": 18995 + }, + { + "epoch": 8.831241283124129, + "grad_norm": 6.201639175415039, + "learning_rate": 9.9810119981012e-06, + "loss": 0.1565, + "step": 18996 + }, + { + "epoch": 8.831706183170619, + "grad_norm": 4.592885494232178, + "learning_rate": 9.981010998101101e-06, + "loss": 0.1379, + "step": 18997 + }, + { + "epoch": 8.832171083217109, + "grad_norm": 4.501534938812256, + "learning_rate": 9.981009998101002e-06, + "loss": 0.0913, + "step": 18998 + }, + { + "epoch": 8.832635983263598, + "grad_norm": 6.322637557983398, + "learning_rate": 9.9810089981009e-06, + "loss": 0.112, + "step": 18999 + }, + { + "epoch": 8.833100883310088, + "grad_norm": 1.3037256002426147, + "learning_rate": 9.9810079981008e-06, + "loss": 0.0123, + "step": 19000 + }, + { + "epoch": 8.833565783356578, + "grad_norm": 6.451251983642578, + "learning_rate": 9.9810069981007e-06, + "loss": 0.1249, + "step": 19001 + }, + { + "epoch": 8.834030683403068, + "grad_norm": 6.809831142425537, + "learning_rate": 9.981005998100601e-06, + "loss": 0.1361, + "step": 19002 + }, + { + "epoch": 8.834495583449558, + "grad_norm": 4.043458938598633, + "learning_rate": 9.981004998100502e-06, + "loss": 0.0529, + "step": 19003 + }, + { + "epoch": 8.83496048349605, + "grad_norm": 2.167559862136841, + "learning_rate": 9.9810039981004e-06, + "loss": 0.0338, + "step": 19004 + }, + { + "epoch": 8.835425383542539, + "grad_norm": 3.6537744998931885, + "learning_rate": 9.981002998100301e-06, + "loss": 0.0426, + "step": 19005 + }, + { + "epoch": 8.835890283589029, + "grad_norm": 4.407625675201416, + "learning_rate": 9.9810019981002e-06, + "loss": 0.0531, + "step": 19006 + }, + { + "epoch": 8.836355183635519, + "grad_norm": 3.7385413646698, + "learning_rate": 9.981000998100101e-06, + "loss": 0.0515, + "step": 19007 + }, + { + "epoch": 8.836820083682008, + "grad_norm": 6.225035667419434, + "learning_rate": 9.9809999981e-06, + "loss": 0.1879, + "step": 19008 + }, + { + "epoch": 8.837284983728498, + "grad_norm": 5.3073577880859375, + "learning_rate": 9.9809989980999e-06, + "loss": 0.1171, + "step": 19009 + }, + { + "epoch": 8.837749883774988, + "grad_norm": 2.561302423477173, + "learning_rate": 9.980997998099801e-06, + "loss": 0.031, + "step": 19010 + }, + { + "epoch": 8.838214783821478, + "grad_norm": 3.371187210083008, + "learning_rate": 9.9809969980997e-06, + "loss": 0.0325, + "step": 19011 + }, + { + "epoch": 8.83867968386797, + "grad_norm": 5.14064359664917, + "learning_rate": 9.9809959980996e-06, + "loss": 0.072, + "step": 19012 + }, + { + "epoch": 8.83914458391446, + "grad_norm": 7.94360876083374, + "learning_rate": 9.9809949980995e-06, + "loss": 0.1336, + "step": 19013 + }, + { + "epoch": 8.839609483960949, + "grad_norm": 3.7944931983947754, + "learning_rate": 9.9809939980994e-06, + "loss": 0.0692, + "step": 19014 + }, + { + "epoch": 8.840074384007439, + "grad_norm": 6.550165176391602, + "learning_rate": 9.980992998099301e-06, + "loss": 0.1323, + "step": 19015 + }, + { + "epoch": 8.840539284053929, + "grad_norm": 2.978260040283203, + "learning_rate": 9.980991998099202e-06, + "loss": 0.0569, + "step": 19016 + }, + { + "epoch": 8.841004184100418, + "grad_norm": 10.194070816040039, + "learning_rate": 9.9809909980991e-06, + "loss": 0.0634, + "step": 19017 + }, + { + "epoch": 8.841469084146908, + "grad_norm": 6.543529510498047, + "learning_rate": 9.980989998099001e-06, + "loss": 0.1611, + "step": 19018 + }, + { + "epoch": 8.841933984193398, + "grad_norm": 7.0033793449401855, + "learning_rate": 9.9809889980989e-06, + "loss": 0.1437, + "step": 19019 + }, + { + "epoch": 8.842398884239888, + "grad_norm": 4.77756404876709, + "learning_rate": 9.980987998098801e-06, + "loss": 0.0731, + "step": 19020 + }, + { + "epoch": 8.84286378428638, + "grad_norm": 7.42952299118042, + "learning_rate": 9.9809869980987e-06, + "loss": 0.1453, + "step": 19021 + }, + { + "epoch": 8.84332868433287, + "grad_norm": 5.466451644897461, + "learning_rate": 9.9809859980986e-06, + "loss": 0.1472, + "step": 19022 + }, + { + "epoch": 8.843793584379359, + "grad_norm": 9.006244659423828, + "learning_rate": 9.980984998098501e-06, + "loss": 0.1498, + "step": 19023 + }, + { + "epoch": 8.844258484425849, + "grad_norm": 5.122038841247559, + "learning_rate": 9.9809839980984e-06, + "loss": 0.1153, + "step": 19024 + }, + { + "epoch": 8.844723384472339, + "grad_norm": 2.200157880783081, + "learning_rate": 9.980982998098301e-06, + "loss": 0.0199, + "step": 19025 + }, + { + "epoch": 8.845188284518828, + "grad_norm": 7.36334228515625, + "learning_rate": 9.9809819980982e-06, + "loss": 0.1756, + "step": 19026 + }, + { + "epoch": 8.845653184565318, + "grad_norm": 5.294691562652588, + "learning_rate": 9.9809809980981e-06, + "loss": 0.1277, + "step": 19027 + }, + { + "epoch": 8.846118084611808, + "grad_norm": 4.204486846923828, + "learning_rate": 9.980979998098001e-06, + "loss": 0.0628, + "step": 19028 + }, + { + "epoch": 8.846582984658298, + "grad_norm": 3.6683032512664795, + "learning_rate": 9.980978998097902e-06, + "loss": 0.0306, + "step": 19029 + }, + { + "epoch": 8.84704788470479, + "grad_norm": 7.488461971282959, + "learning_rate": 9.9809779980978e-06, + "loss": 0.0833, + "step": 19030 + }, + { + "epoch": 8.84751278475128, + "grad_norm": 3.3205087184906006, + "learning_rate": 9.980976998097701e-06, + "loss": 0.0433, + "step": 19031 + }, + { + "epoch": 8.847977684797769, + "grad_norm": 5.892043590545654, + "learning_rate": 9.9809759980976e-06, + "loss": 0.0922, + "step": 19032 + }, + { + "epoch": 8.848442584844259, + "grad_norm": 13.5595121383667, + "learning_rate": 9.9809749980975e-06, + "loss": 0.2663, + "step": 19033 + }, + { + "epoch": 8.848907484890749, + "grad_norm": 5.550561904907227, + "learning_rate": 9.9809739980974e-06, + "loss": 0.1128, + "step": 19034 + }, + { + "epoch": 8.849372384937238, + "grad_norm": 5.960772514343262, + "learning_rate": 9.9809729980973e-06, + "loss": 0.1079, + "step": 19035 + }, + { + "epoch": 8.849837284983728, + "grad_norm": 5.963407039642334, + "learning_rate": 9.980971998097201e-06, + "loss": 0.111, + "step": 19036 + }, + { + "epoch": 8.850302185030218, + "grad_norm": 8.864228248596191, + "learning_rate": 9.9809709980971e-06, + "loss": 0.1967, + "step": 19037 + }, + { + "epoch": 8.850767085076708, + "grad_norm": 3.450685977935791, + "learning_rate": 9.980969998097001e-06, + "loss": 0.0658, + "step": 19038 + }, + { + "epoch": 8.8512319851232, + "grad_norm": 7.684843063354492, + "learning_rate": 9.9809689980969e-06, + "loss": 0.1901, + "step": 19039 + }, + { + "epoch": 8.851696885169689, + "grad_norm": 5.945705413818359, + "learning_rate": 9.9809679980968e-06, + "loss": 0.0974, + "step": 19040 + }, + { + "epoch": 8.852161785216179, + "grad_norm": 3.0308728218078613, + "learning_rate": 9.980966998096701e-06, + "loss": 0.0397, + "step": 19041 + }, + { + "epoch": 8.852626685262669, + "grad_norm": 6.99688196182251, + "learning_rate": 9.980965998096602e-06, + "loss": 0.1308, + "step": 19042 + }, + { + "epoch": 8.853091585309159, + "grad_norm": 3.825436592102051, + "learning_rate": 9.980964998096501e-06, + "loss": 0.0339, + "step": 19043 + }, + { + "epoch": 8.853556485355648, + "grad_norm": 5.434325695037842, + "learning_rate": 9.9809639980964e-06, + "loss": 0.0799, + "step": 19044 + }, + { + "epoch": 8.854021385402138, + "grad_norm": 5.675891876220703, + "learning_rate": 9.9809629980963e-06, + "loss": 0.1031, + "step": 19045 + }, + { + "epoch": 8.854486285448628, + "grad_norm": 4.778573989868164, + "learning_rate": 9.9809619980962e-06, + "loss": 0.0992, + "step": 19046 + }, + { + "epoch": 8.854951185495118, + "grad_norm": 3.2257580757141113, + "learning_rate": 9.9809609980961e-06, + "loss": 0.0564, + "step": 19047 + }, + { + "epoch": 8.85541608554161, + "grad_norm": 11.25631332397461, + "learning_rate": 9.980959998096e-06, + "loss": 0.2979, + "step": 19048 + }, + { + "epoch": 8.855880985588099, + "grad_norm": 7.81573486328125, + "learning_rate": 9.980958998095901e-06, + "loss": 0.1064, + "step": 19049 + }, + { + "epoch": 8.856345885634589, + "grad_norm": 5.549238681793213, + "learning_rate": 9.9809579980958e-06, + "loss": 0.0398, + "step": 19050 + }, + { + "epoch": 8.856810785681079, + "grad_norm": 4.8562541007995605, + "learning_rate": 9.980956998095701e-06, + "loss": 0.0243, + "step": 19051 + }, + { + "epoch": 8.857275685727569, + "grad_norm": 6.363263130187988, + "learning_rate": 9.9809559980956e-06, + "loss": 0.1699, + "step": 19052 + }, + { + "epoch": 8.857740585774058, + "grad_norm": 5.282962322235107, + "learning_rate": 9.9809549980955e-06, + "loss": 0.0877, + "step": 19053 + }, + { + "epoch": 8.858205485820548, + "grad_norm": 5.066569805145264, + "learning_rate": 9.980953998095401e-06, + "loss": 0.0898, + "step": 19054 + }, + { + "epoch": 8.858670385867038, + "grad_norm": 3.314807415008545, + "learning_rate": 9.9809529980953e-06, + "loss": 0.0596, + "step": 19055 + }, + { + "epoch": 8.859135285913528, + "grad_norm": 4.410854816436768, + "learning_rate": 9.980951998095201e-06, + "loss": 0.0953, + "step": 19056 + }, + { + "epoch": 8.85960018596002, + "grad_norm": 6.321290016174316, + "learning_rate": 9.9809509980951e-06, + "loss": 0.1197, + "step": 19057 + }, + { + "epoch": 8.860065086006509, + "grad_norm": 6.139907360076904, + "learning_rate": 9.980949998095e-06, + "loss": 0.1193, + "step": 19058 + }, + { + "epoch": 8.860529986052999, + "grad_norm": 5.343510150909424, + "learning_rate": 9.9809489980949e-06, + "loss": 0.0702, + "step": 19059 + }, + { + "epoch": 8.860994886099489, + "grad_norm": 5.007284164428711, + "learning_rate": 9.9809479980948e-06, + "loss": 0.1245, + "step": 19060 + }, + { + "epoch": 8.861459786145979, + "grad_norm": 5.619215965270996, + "learning_rate": 9.980946998094701e-06, + "loss": 0.071, + "step": 19061 + }, + { + "epoch": 8.861924686192468, + "grad_norm": 8.273076057434082, + "learning_rate": 9.980945998094602e-06, + "loss": 0.241, + "step": 19062 + }, + { + "epoch": 8.862389586238958, + "grad_norm": 4.623734951019287, + "learning_rate": 9.9809449980945e-06, + "loss": 0.0595, + "step": 19063 + }, + { + "epoch": 8.862854486285448, + "grad_norm": 5.370913028717041, + "learning_rate": 9.980943998094401e-06, + "loss": 0.1153, + "step": 19064 + }, + { + "epoch": 8.863319386331938, + "grad_norm": 4.583154678344727, + "learning_rate": 9.9809429980943e-06, + "loss": 0.0939, + "step": 19065 + }, + { + "epoch": 8.86378428637843, + "grad_norm": 4.0653276443481445, + "learning_rate": 9.9809419980942e-06, + "loss": 0.055, + "step": 19066 + }, + { + "epoch": 8.864249186424919, + "grad_norm": 5.486716270446777, + "learning_rate": 9.980940998094101e-06, + "loss": 0.0682, + "step": 19067 + }, + { + "epoch": 8.864714086471409, + "grad_norm": 6.682815074920654, + "learning_rate": 9.980939998094e-06, + "loss": 0.1225, + "step": 19068 + }, + { + "epoch": 8.865178986517899, + "grad_norm": 5.9674177169799805, + "learning_rate": 9.980938998093901e-06, + "loss": 0.186, + "step": 19069 + }, + { + "epoch": 8.865643886564389, + "grad_norm": 4.511585235595703, + "learning_rate": 9.9809379980938e-06, + "loss": 0.0764, + "step": 19070 + }, + { + "epoch": 8.866108786610878, + "grad_norm": 4.678345680236816, + "learning_rate": 9.9809369980937e-06, + "loss": 0.0758, + "step": 19071 + }, + { + "epoch": 8.866573686657368, + "grad_norm": 6.310126781463623, + "learning_rate": 9.980935998093601e-06, + "loss": 0.1323, + "step": 19072 + }, + { + "epoch": 8.867038586703858, + "grad_norm": 2.605717182159424, + "learning_rate": 9.9809349980935e-06, + "loss": 0.0577, + "step": 19073 + }, + { + "epoch": 8.86750348675035, + "grad_norm": 6.915827751159668, + "learning_rate": 9.980933998093401e-06, + "loss": 0.1247, + "step": 19074 + }, + { + "epoch": 8.86796838679684, + "grad_norm": 4.696047782897949, + "learning_rate": 9.9809329980933e-06, + "loss": 0.0768, + "step": 19075 + }, + { + "epoch": 8.868433286843329, + "grad_norm": 1.6742016077041626, + "learning_rate": 9.9809319980932e-06, + "loss": 0.0158, + "step": 19076 + }, + { + "epoch": 8.868898186889819, + "grad_norm": 8.384096145629883, + "learning_rate": 9.9809309980931e-06, + "loss": 0.2486, + "step": 19077 + }, + { + "epoch": 8.869363086936309, + "grad_norm": 9.778428077697754, + "learning_rate": 9.980929998093e-06, + "loss": 0.19, + "step": 19078 + }, + { + "epoch": 8.869827986982799, + "grad_norm": 6.837318420410156, + "learning_rate": 9.9809289980929e-06, + "loss": 0.1181, + "step": 19079 + }, + { + "epoch": 8.870292887029288, + "grad_norm": 4.991265296936035, + "learning_rate": 9.980927998092802e-06, + "loss": 0.0832, + "step": 19080 + }, + { + "epoch": 8.870757787075778, + "grad_norm": 5.693905353546143, + "learning_rate": 9.9809269980927e-06, + "loss": 0.1189, + "step": 19081 + }, + { + "epoch": 8.871222687122268, + "grad_norm": 3.8550338745117188, + "learning_rate": 9.980925998092601e-06, + "loss": 0.0213, + "step": 19082 + }, + { + "epoch": 8.87168758716876, + "grad_norm": 7.862631797790527, + "learning_rate": 9.9809249980925e-06, + "loss": 0.1267, + "step": 19083 + }, + { + "epoch": 8.87215248721525, + "grad_norm": 8.013862609863281, + "learning_rate": 9.9809239980924e-06, + "loss": 0.1502, + "step": 19084 + }, + { + "epoch": 8.872617387261739, + "grad_norm": 8.265490531921387, + "learning_rate": 9.980922998092301e-06, + "loss": 0.1996, + "step": 19085 + }, + { + "epoch": 8.873082287308229, + "grad_norm": 4.443179130554199, + "learning_rate": 9.9809219980922e-06, + "loss": 0.0772, + "step": 19086 + }, + { + "epoch": 8.873547187354719, + "grad_norm": 8.408065795898438, + "learning_rate": 9.980920998092101e-06, + "loss": 0.2129, + "step": 19087 + }, + { + "epoch": 8.874012087401209, + "grad_norm": 3.834099054336548, + "learning_rate": 9.980919998092e-06, + "loss": 0.0768, + "step": 19088 + }, + { + "epoch": 8.874476987447698, + "grad_norm": 6.630343437194824, + "learning_rate": 9.9809189980919e-06, + "loss": 0.1409, + "step": 19089 + }, + { + "epoch": 8.874941887494188, + "grad_norm": 9.190391540527344, + "learning_rate": 9.9809179980918e-06, + "loss": 0.2327, + "step": 19090 + }, + { + "epoch": 8.87540678754068, + "grad_norm": 7.871956825256348, + "learning_rate": 9.9809169980917e-06, + "loss": 0.1755, + "step": 19091 + }, + { + "epoch": 8.87587168758717, + "grad_norm": 4.377994537353516, + "learning_rate": 9.980915998091601e-06, + "loss": 0.0612, + "step": 19092 + }, + { + "epoch": 8.87633658763366, + "grad_norm": 4.555298805236816, + "learning_rate": 9.980914998091502e-06, + "loss": 0.0293, + "step": 19093 + }, + { + "epoch": 8.876801487680149, + "grad_norm": 5.193678379058838, + "learning_rate": 9.9809139980914e-06, + "loss": 0.1077, + "step": 19094 + }, + { + "epoch": 8.877266387726639, + "grad_norm": 12.255685806274414, + "learning_rate": 9.980912998091301e-06, + "loss": 0.1297, + "step": 19095 + }, + { + "epoch": 8.877731287773129, + "grad_norm": 5.022604942321777, + "learning_rate": 9.9809119980912e-06, + "loss": 0.0808, + "step": 19096 + }, + { + "epoch": 8.878196187819619, + "grad_norm": 10.110291481018066, + "learning_rate": 9.9809109980911e-06, + "loss": 0.1283, + "step": 19097 + }, + { + "epoch": 8.878661087866108, + "grad_norm": 5.161725997924805, + "learning_rate": 9.980909998091001e-06, + "loss": 0.101, + "step": 19098 + }, + { + "epoch": 8.879125987912598, + "grad_norm": 5.420489311218262, + "learning_rate": 9.9809089980909e-06, + "loss": 0.1383, + "step": 19099 + }, + { + "epoch": 8.87959088795909, + "grad_norm": 3.24035906791687, + "learning_rate": 9.980907998090801e-06, + "loss": 0.0451, + "step": 19100 + }, + { + "epoch": 8.88005578800558, + "grad_norm": 5.102604389190674, + "learning_rate": 9.9809069980907e-06, + "loss": 0.1175, + "step": 19101 + }, + { + "epoch": 8.88052068805207, + "grad_norm": 4.584300518035889, + "learning_rate": 9.9809059980906e-06, + "loss": 0.1155, + "step": 19102 + }, + { + "epoch": 8.880985588098559, + "grad_norm": 10.686517715454102, + "learning_rate": 9.9809049980905e-06, + "loss": 0.2956, + "step": 19103 + }, + { + "epoch": 8.881450488145049, + "grad_norm": 3.8310136795043945, + "learning_rate": 9.9809039980904e-06, + "loss": 0.0735, + "step": 19104 + }, + { + "epoch": 8.881915388191539, + "grad_norm": 7.363473892211914, + "learning_rate": 9.980902998090301e-06, + "loss": 0.1318, + "step": 19105 + }, + { + "epoch": 8.882380288238028, + "grad_norm": 8.329601287841797, + "learning_rate": 9.980901998090202e-06, + "loss": 0.1085, + "step": 19106 + }, + { + "epoch": 8.882845188284518, + "grad_norm": 8.89114761352539, + "learning_rate": 9.9809009980901e-06, + "loss": 0.122, + "step": 19107 + }, + { + "epoch": 8.883310088331008, + "grad_norm": 5.254057884216309, + "learning_rate": 9.98089999809e-06, + "loss": 0.12, + "step": 19108 + }, + { + "epoch": 8.8837749883775, + "grad_norm": 6.043005466461182, + "learning_rate": 9.9808989980899e-06, + "loss": 0.0896, + "step": 19109 + }, + { + "epoch": 8.88423988842399, + "grad_norm": 2.7057788372039795, + "learning_rate": 9.980897998089801e-06, + "loss": 0.0342, + "step": 19110 + }, + { + "epoch": 8.88470478847048, + "grad_norm": 4.448930263519287, + "learning_rate": 9.980896998089702e-06, + "loss": 0.0822, + "step": 19111 + }, + { + "epoch": 8.885169688516969, + "grad_norm": 2.9027528762817383, + "learning_rate": 9.9808959980896e-06, + "loss": 0.0325, + "step": 19112 + }, + { + "epoch": 8.885634588563459, + "grad_norm": 5.410031795501709, + "learning_rate": 9.980894998089501e-06, + "loss": 0.1275, + "step": 19113 + }, + { + "epoch": 8.886099488609949, + "grad_norm": 7.00640869140625, + "learning_rate": 9.9808939980894e-06, + "loss": 0.2626, + "step": 19114 + }, + { + "epoch": 8.886564388656438, + "grad_norm": 2.892514944076538, + "learning_rate": 9.9808929980893e-06, + "loss": 0.0481, + "step": 19115 + }, + { + "epoch": 8.887029288702928, + "grad_norm": 8.214944839477539, + "learning_rate": 9.9808919980892e-06, + "loss": 0.2501, + "step": 19116 + }, + { + "epoch": 8.887494188749418, + "grad_norm": 5.10505485534668, + "learning_rate": 9.9808909980891e-06, + "loss": 0.0721, + "step": 19117 + }, + { + "epoch": 8.88795908879591, + "grad_norm": 5.053567409515381, + "learning_rate": 9.980889998089001e-06, + "loss": 0.1695, + "step": 19118 + }, + { + "epoch": 8.8884239888424, + "grad_norm": 8.118829727172852, + "learning_rate": 9.9808889980889e-06, + "loss": 0.1147, + "step": 19119 + }, + { + "epoch": 8.88888888888889, + "grad_norm": 5.854910850524902, + "learning_rate": 9.9808879980888e-06, + "loss": 0.1916, + "step": 19120 + }, + { + "epoch": 8.889353788935379, + "grad_norm": 3.8398375511169434, + "learning_rate": 9.9808869980887e-06, + "loss": 0.0642, + "step": 19121 + }, + { + "epoch": 8.889818688981869, + "grad_norm": 5.675055980682373, + "learning_rate": 9.9808859980886e-06, + "loss": 0.102, + "step": 19122 + }, + { + "epoch": 8.890283589028359, + "grad_norm": 7.89542818069458, + "learning_rate": 9.980884998088501e-06, + "loss": 0.1803, + "step": 19123 + }, + { + "epoch": 8.890748489074848, + "grad_norm": 5.236927509307861, + "learning_rate": 9.980883998088402e-06, + "loss": 0.0537, + "step": 19124 + }, + { + "epoch": 8.891213389121338, + "grad_norm": 3.1332719326019287, + "learning_rate": 9.9808829980883e-06, + "loss": 0.0397, + "step": 19125 + }, + { + "epoch": 8.891678289167828, + "grad_norm": 7.324311256408691, + "learning_rate": 9.980881998088201e-06, + "loss": 0.2719, + "step": 19126 + }, + { + "epoch": 8.89214318921432, + "grad_norm": 3.5097551345825195, + "learning_rate": 9.9808809980881e-06, + "loss": 0.0294, + "step": 19127 + }, + { + "epoch": 8.89260808926081, + "grad_norm": 4.806396484375, + "learning_rate": 9.980879998088001e-06, + "loss": 0.067, + "step": 19128 + }, + { + "epoch": 8.8930729893073, + "grad_norm": 6.852829933166504, + "learning_rate": 9.9808789980879e-06, + "loss": 0.1302, + "step": 19129 + }, + { + "epoch": 8.893537889353789, + "grad_norm": 7.522727012634277, + "learning_rate": 9.9808779980878e-06, + "loss": 0.1802, + "step": 19130 + }, + { + "epoch": 8.894002789400279, + "grad_norm": 4.847314357757568, + "learning_rate": 9.980876998087701e-06, + "loss": 0.1064, + "step": 19131 + }, + { + "epoch": 8.894467689446769, + "grad_norm": 1.7569128274917603, + "learning_rate": 9.9808759980876e-06, + "loss": 0.0237, + "step": 19132 + }, + { + "epoch": 8.894932589493258, + "grad_norm": 5.9526262283325195, + "learning_rate": 9.9808749980875e-06, + "loss": 0.0821, + "step": 19133 + }, + { + "epoch": 8.895397489539748, + "grad_norm": 4.420551300048828, + "learning_rate": 9.9808739980874e-06, + "loss": 0.0677, + "step": 19134 + }, + { + "epoch": 8.895862389586238, + "grad_norm": 5.152888774871826, + "learning_rate": 9.9808729980873e-06, + "loss": 0.1143, + "step": 19135 + }, + { + "epoch": 8.89632728963273, + "grad_norm": 7.20153284072876, + "learning_rate": 9.980871998087201e-06, + "loss": 0.106, + "step": 19136 + }, + { + "epoch": 8.89679218967922, + "grad_norm": 4.1009931564331055, + "learning_rate": 9.980870998087102e-06, + "loss": 0.0639, + "step": 19137 + }, + { + "epoch": 8.89725708972571, + "grad_norm": 5.178077220916748, + "learning_rate": 9.980869998087e-06, + "loss": 0.1098, + "step": 19138 + }, + { + "epoch": 8.897721989772199, + "grad_norm": 2.0104353427886963, + "learning_rate": 9.980868998086901e-06, + "loss": 0.0237, + "step": 19139 + }, + { + "epoch": 8.898186889818689, + "grad_norm": 5.5816779136657715, + "learning_rate": 9.9808679980868e-06, + "loss": 0.1178, + "step": 19140 + }, + { + "epoch": 8.898651789865179, + "grad_norm": 5.754333019256592, + "learning_rate": 9.9808669980867e-06, + "loss": 0.1038, + "step": 19141 + }, + { + "epoch": 8.899116689911668, + "grad_norm": 4.819518566131592, + "learning_rate": 9.9808659980866e-06, + "loss": 0.0529, + "step": 19142 + }, + { + "epoch": 8.899581589958158, + "grad_norm": 5.872389316558838, + "learning_rate": 9.9808649980865e-06, + "loss": 0.2093, + "step": 19143 + }, + { + "epoch": 8.900046490004648, + "grad_norm": 4.883072376251221, + "learning_rate": 9.980863998086401e-06, + "loss": 0.0904, + "step": 19144 + }, + { + "epoch": 8.90051139005114, + "grad_norm": 2.730029582977295, + "learning_rate": 9.9808629980863e-06, + "loss": 0.0335, + "step": 19145 + }, + { + "epoch": 8.90097629009763, + "grad_norm": 4.488394260406494, + "learning_rate": 9.980861998086201e-06, + "loss": 0.084, + "step": 19146 + }, + { + "epoch": 8.90144119014412, + "grad_norm": 4.278445243835449, + "learning_rate": 9.9808609980861e-06, + "loss": 0.0756, + "step": 19147 + }, + { + "epoch": 8.901906090190609, + "grad_norm": 5.545329570770264, + "learning_rate": 9.980859998086e-06, + "loss": 0.059, + "step": 19148 + }, + { + "epoch": 8.902370990237099, + "grad_norm": 5.5070013999938965, + "learning_rate": 9.980858998085901e-06, + "loss": 0.0798, + "step": 19149 + }, + { + "epoch": 8.902835890283589, + "grad_norm": 3.6465649604797363, + "learning_rate": 9.980857998085802e-06, + "loss": 0.0609, + "step": 19150 + }, + { + "epoch": 8.903300790330078, + "grad_norm": 6.300066947937012, + "learning_rate": 9.9808569980857e-06, + "loss": 0.129, + "step": 19151 + }, + { + "epoch": 8.903765690376568, + "grad_norm": 6.599280834197998, + "learning_rate": 9.9808559980856e-06, + "loss": 0.1726, + "step": 19152 + }, + { + "epoch": 8.90423059042306, + "grad_norm": 3.8215558528900146, + "learning_rate": 9.9808549980855e-06, + "loss": 0.0795, + "step": 19153 + }, + { + "epoch": 8.90469549046955, + "grad_norm": 4.618061065673828, + "learning_rate": 9.980853998085401e-06, + "loss": 0.1272, + "step": 19154 + }, + { + "epoch": 8.90516039051604, + "grad_norm": 5.059680461883545, + "learning_rate": 9.9808529980853e-06, + "loss": 0.1197, + "step": 19155 + }, + { + "epoch": 8.90562529056253, + "grad_norm": 5.050498008728027, + "learning_rate": 9.9808519980852e-06, + "loss": 0.0533, + "step": 19156 + }, + { + "epoch": 8.906090190609019, + "grad_norm": 8.643340110778809, + "learning_rate": 9.980850998085101e-06, + "loss": 0.1872, + "step": 19157 + }, + { + "epoch": 8.906555090655509, + "grad_norm": 8.09397029876709, + "learning_rate": 9.980849998085e-06, + "loss": 0.1416, + "step": 19158 + }, + { + "epoch": 8.907019990701999, + "grad_norm": 4.074454307556152, + "learning_rate": 9.980848998084901e-06, + "loss": 0.0637, + "step": 19159 + }, + { + "epoch": 8.907484890748488, + "grad_norm": 5.038389205932617, + "learning_rate": 9.9808479980848e-06, + "loss": 0.1063, + "step": 19160 + }, + { + "epoch": 8.907949790794978, + "grad_norm": 6.964289665222168, + "learning_rate": 9.9808469980847e-06, + "loss": 0.184, + "step": 19161 + }, + { + "epoch": 8.90841469084147, + "grad_norm": 7.655527591705322, + "learning_rate": 9.980845998084601e-06, + "loss": 0.1235, + "step": 19162 + }, + { + "epoch": 8.90887959088796, + "grad_norm": 1.960817575454712, + "learning_rate": 9.9808449980845e-06, + "loss": 0.0173, + "step": 19163 + }, + { + "epoch": 8.90934449093445, + "grad_norm": 7.215651035308838, + "learning_rate": 9.980843998084401e-06, + "loss": 0.1554, + "step": 19164 + }, + { + "epoch": 8.90980939098094, + "grad_norm": 3.362981081008911, + "learning_rate": 9.9808429980843e-06, + "loss": 0.0538, + "step": 19165 + }, + { + "epoch": 8.910274291027429, + "grad_norm": 5.184852600097656, + "learning_rate": 9.9808419980842e-06, + "loss": 0.1184, + "step": 19166 + }, + { + "epoch": 8.910739191073919, + "grad_norm": 7.7385711669921875, + "learning_rate": 9.980840998084101e-06, + "loss": 0.1844, + "step": 19167 + }, + { + "epoch": 8.911204091120409, + "grad_norm": 6.632650852203369, + "learning_rate": 9.980839998084002e-06, + "loss": 0.0744, + "step": 19168 + }, + { + "epoch": 8.911668991166898, + "grad_norm": 9.840091705322266, + "learning_rate": 9.9808389980839e-06, + "loss": 0.3186, + "step": 19169 + }, + { + "epoch": 8.91213389121339, + "grad_norm": 4.952028751373291, + "learning_rate": 9.980837998083801e-06, + "loss": 0.1356, + "step": 19170 + }, + { + "epoch": 8.91259879125988, + "grad_norm": 7.346377849578857, + "learning_rate": 9.9808369980837e-06, + "loss": 0.2592, + "step": 19171 + }, + { + "epoch": 8.91306369130637, + "grad_norm": 5.953848361968994, + "learning_rate": 9.980835998083601e-06, + "loss": 0.1676, + "step": 19172 + }, + { + "epoch": 8.91352859135286, + "grad_norm": 3.6582112312316895, + "learning_rate": 9.9808349980835e-06, + "loss": 0.0667, + "step": 19173 + }, + { + "epoch": 8.91399349139935, + "grad_norm": 7.571230888366699, + "learning_rate": 9.9808339980834e-06, + "loss": 0.1374, + "step": 19174 + }, + { + "epoch": 8.914458391445839, + "grad_norm": 6.758081912994385, + "learning_rate": 9.980832998083301e-06, + "loss": 0.0927, + "step": 19175 + }, + { + "epoch": 8.914923291492329, + "grad_norm": 4.907578945159912, + "learning_rate": 9.9808319980832e-06, + "loss": 0.0622, + "step": 19176 + }, + { + "epoch": 8.915388191538819, + "grad_norm": 5.210885524749756, + "learning_rate": 9.980830998083101e-06, + "loss": 0.0886, + "step": 19177 + }, + { + "epoch": 8.915853091585308, + "grad_norm": 4.708967685699463, + "learning_rate": 9.980829998083e-06, + "loss": 0.1214, + "step": 19178 + }, + { + "epoch": 8.9163179916318, + "grad_norm": 4.470268726348877, + "learning_rate": 9.9808289980829e-06, + "loss": 0.0932, + "step": 19179 + }, + { + "epoch": 8.91678289167829, + "grad_norm": 4.969753742218018, + "learning_rate": 9.980827998082801e-06, + "loss": 0.1325, + "step": 19180 + }, + { + "epoch": 8.91724779172478, + "grad_norm": 3.230736255645752, + "learning_rate": 9.980826998082702e-06, + "loss": 0.0322, + "step": 19181 + }, + { + "epoch": 8.91771269177127, + "grad_norm": 6.552498817443848, + "learning_rate": 9.9808259980826e-06, + "loss": 0.1756, + "step": 19182 + }, + { + "epoch": 8.91817759181776, + "grad_norm": 4.733869552612305, + "learning_rate": 9.980824998082501e-06, + "loss": 0.0912, + "step": 19183 + }, + { + "epoch": 8.918642491864249, + "grad_norm": 8.459088325500488, + "learning_rate": 9.9808239980824e-06, + "loss": 0.2901, + "step": 19184 + }, + { + "epoch": 8.919107391910739, + "grad_norm": 3.7148854732513428, + "learning_rate": 9.9808229980823e-06, + "loss": 0.0414, + "step": 19185 + }, + { + "epoch": 8.919572291957229, + "grad_norm": 6.815496444702148, + "learning_rate": 9.9808219980822e-06, + "loss": 0.0794, + "step": 19186 + }, + { + "epoch": 8.920037192003718, + "grad_norm": 5.877988815307617, + "learning_rate": 9.9808209980821e-06, + "loss": 0.1958, + "step": 19187 + }, + { + "epoch": 8.92050209205021, + "grad_norm": 5.3011674880981445, + "learning_rate": 9.980819998082001e-06, + "loss": 0.0966, + "step": 19188 + }, + { + "epoch": 8.9209669920967, + "grad_norm": 8.416617393493652, + "learning_rate": 9.9808189980819e-06, + "loss": 0.2078, + "step": 19189 + }, + { + "epoch": 8.92143189214319, + "grad_norm": 3.2416701316833496, + "learning_rate": 9.980817998081801e-06, + "loss": 0.0368, + "step": 19190 + }, + { + "epoch": 8.92189679218968, + "grad_norm": 7.631546974182129, + "learning_rate": 9.9808169980817e-06, + "loss": 0.0921, + "step": 19191 + }, + { + "epoch": 8.92236169223617, + "grad_norm": 3.8655664920806885, + "learning_rate": 9.9808159980816e-06, + "loss": 0.0871, + "step": 19192 + }, + { + "epoch": 8.922826592282659, + "grad_norm": 7.880354881286621, + "learning_rate": 9.980814998081501e-06, + "loss": 0.0992, + "step": 19193 + }, + { + "epoch": 8.923291492329149, + "grad_norm": 4.866986274719238, + "learning_rate": 9.980813998081402e-06, + "loss": 0.0999, + "step": 19194 + }, + { + "epoch": 8.923756392375639, + "grad_norm": 7.177622318267822, + "learning_rate": 9.980812998081301e-06, + "loss": 0.2525, + "step": 19195 + }, + { + "epoch": 8.924221292422128, + "grad_norm": 3.755134344100952, + "learning_rate": 9.9808119980812e-06, + "loss": 0.0812, + "step": 19196 + }, + { + "epoch": 8.92468619246862, + "grad_norm": 7.093413352966309, + "learning_rate": 9.9808109980811e-06, + "loss": 0.2493, + "step": 19197 + }, + { + "epoch": 8.92515109251511, + "grad_norm": 6.5253005027771, + "learning_rate": 9.980809998081e-06, + "loss": 0.1469, + "step": 19198 + }, + { + "epoch": 8.9256159925616, + "grad_norm": 7.065113544464111, + "learning_rate": 9.9808089980809e-06, + "loss": 0.2109, + "step": 19199 + }, + { + "epoch": 8.92608089260809, + "grad_norm": 4.748392581939697, + "learning_rate": 9.9808079980808e-06, + "loss": 0.1117, + "step": 19200 + }, + { + "epoch": 8.92654579265458, + "grad_norm": 0.4442707598209381, + "learning_rate": 9.980806998080701e-06, + "loss": 0.0039, + "step": 19201 + }, + { + "epoch": 8.927010692701069, + "grad_norm": 3.308051586151123, + "learning_rate": 9.9808059980806e-06, + "loss": 0.0472, + "step": 19202 + }, + { + "epoch": 8.927475592747559, + "grad_norm": 9.27401351928711, + "learning_rate": 9.980804998080501e-06, + "loss": 0.2098, + "step": 19203 + }, + { + "epoch": 8.927940492794049, + "grad_norm": 5.347329616546631, + "learning_rate": 9.9808039980804e-06, + "loss": 0.1274, + "step": 19204 + }, + { + "epoch": 8.928405392840538, + "grad_norm": 4.882455348968506, + "learning_rate": 9.9808029980803e-06, + "loss": 0.0231, + "step": 19205 + }, + { + "epoch": 8.92887029288703, + "grad_norm": 4.711799621582031, + "learning_rate": 9.980801998080201e-06, + "loss": 0.1039, + "step": 19206 + }, + { + "epoch": 8.92933519293352, + "grad_norm": 4.761070728302002, + "learning_rate": 9.9808009980801e-06, + "loss": 0.0889, + "step": 19207 + }, + { + "epoch": 8.92980009298001, + "grad_norm": 8.540678024291992, + "learning_rate": 9.980799998080001e-06, + "loss": 0.1591, + "step": 19208 + }, + { + "epoch": 8.9302649930265, + "grad_norm": 5.458542346954346, + "learning_rate": 9.9807989980799e-06, + "loss": 0.1364, + "step": 19209 + }, + { + "epoch": 8.93072989307299, + "grad_norm": 2.4236176013946533, + "learning_rate": 9.9807979980798e-06, + "loss": 0.0309, + "step": 19210 + }, + { + "epoch": 8.931194793119479, + "grad_norm": 11.623369216918945, + "learning_rate": 9.9807969980797e-06, + "loss": 0.2043, + "step": 19211 + }, + { + "epoch": 8.931659693165969, + "grad_norm": 1.7132121324539185, + "learning_rate": 9.9807959980796e-06, + "loss": 0.0244, + "step": 19212 + }, + { + "epoch": 8.932124593212459, + "grad_norm": 3.325592041015625, + "learning_rate": 9.980794998079501e-06, + "loss": 0.0662, + "step": 19213 + }, + { + "epoch": 8.932589493258948, + "grad_norm": 7.184981346130371, + "learning_rate": 9.980793998079402e-06, + "loss": 0.0736, + "step": 19214 + }, + { + "epoch": 8.93305439330544, + "grad_norm": 3.5843698978424072, + "learning_rate": 9.9807929980793e-06, + "loss": 0.0504, + "step": 19215 + }, + { + "epoch": 8.93351929335193, + "grad_norm": 6.6490278244018555, + "learning_rate": 9.9807919980792e-06, + "loss": 0.2046, + "step": 19216 + }, + { + "epoch": 8.93398419339842, + "grad_norm": 6.671572685241699, + "learning_rate": 9.9807909980791e-06, + "loss": 0.1275, + "step": 19217 + }, + { + "epoch": 8.93444909344491, + "grad_norm": 8.051339149475098, + "learning_rate": 9.980789998079e-06, + "loss": 0.127, + "step": 19218 + }, + { + "epoch": 8.9349139934914, + "grad_norm": 9.28948974609375, + "learning_rate": 9.980788998078901e-06, + "loss": 0.2012, + "step": 19219 + }, + { + "epoch": 8.935378893537889, + "grad_norm": 5.662003040313721, + "learning_rate": 9.9807879980788e-06, + "loss": 0.1465, + "step": 19220 + }, + { + "epoch": 8.935843793584379, + "grad_norm": 5.7337541580200195, + "learning_rate": 9.980786998078701e-06, + "loss": 0.1215, + "step": 19221 + }, + { + "epoch": 8.936308693630869, + "grad_norm": 3.097987651824951, + "learning_rate": 9.9807859980786e-06, + "loss": 0.0358, + "step": 19222 + }, + { + "epoch": 8.936773593677358, + "grad_norm": 3.638362407684326, + "learning_rate": 9.9807849980785e-06, + "loss": 0.0655, + "step": 19223 + }, + { + "epoch": 8.93723849372385, + "grad_norm": 5.103986740112305, + "learning_rate": 9.9807839980784e-06, + "loss": 0.149, + "step": 19224 + }, + { + "epoch": 8.93770339377034, + "grad_norm": 2.1373131275177, + "learning_rate": 9.9807829980783e-06, + "loss": 0.0268, + "step": 19225 + }, + { + "epoch": 8.93816829381683, + "grad_norm": 2.06630802154541, + "learning_rate": 9.980781998078201e-06, + "loss": 0.0323, + "step": 19226 + }, + { + "epoch": 8.93863319386332, + "grad_norm": 4.595266819000244, + "learning_rate": 9.9807809980781e-06, + "loss": 0.0676, + "step": 19227 + }, + { + "epoch": 8.93909809390981, + "grad_norm": 6.452381610870361, + "learning_rate": 9.980779998078e-06, + "loss": 0.2241, + "step": 19228 + }, + { + "epoch": 8.939562993956299, + "grad_norm": 4.499588966369629, + "learning_rate": 9.9807789980779e-06, + "loss": 0.0793, + "step": 19229 + }, + { + "epoch": 8.940027894002789, + "grad_norm": 2.356050968170166, + "learning_rate": 9.9807779980778e-06, + "loss": 0.0318, + "step": 19230 + }, + { + "epoch": 8.940492794049279, + "grad_norm": 4.144290447235107, + "learning_rate": 9.980776998077701e-06, + "loss": 0.0499, + "step": 19231 + }, + { + "epoch": 8.94095769409577, + "grad_norm": 8.690827369689941, + "learning_rate": 9.980775998077602e-06, + "loss": 0.2265, + "step": 19232 + }, + { + "epoch": 8.94142259414226, + "grad_norm": 5.832061767578125, + "learning_rate": 9.9807749980775e-06, + "loss": 0.1542, + "step": 19233 + }, + { + "epoch": 8.94188749418875, + "grad_norm": 5.198264122009277, + "learning_rate": 9.980773998077401e-06, + "loss": 0.1278, + "step": 19234 + }, + { + "epoch": 8.94235239423524, + "grad_norm": 3.0079989433288574, + "learning_rate": 9.9807729980773e-06, + "loss": 0.0353, + "step": 19235 + }, + { + "epoch": 8.94281729428173, + "grad_norm": 3.5454940795898438, + "learning_rate": 9.9807719980772e-06, + "loss": 0.0372, + "step": 19236 + }, + { + "epoch": 8.94328219432822, + "grad_norm": 7.24130392074585, + "learning_rate": 9.980770998077101e-06, + "loss": 0.2021, + "step": 19237 + }, + { + "epoch": 8.943747094374709, + "grad_norm": 1.251356601715088, + "learning_rate": 9.980769998077e-06, + "loss": 0.0101, + "step": 19238 + }, + { + "epoch": 8.944211994421199, + "grad_norm": 4.31777286529541, + "learning_rate": 9.980768998076901e-06, + "loss": 0.1119, + "step": 19239 + }, + { + "epoch": 8.944676894467689, + "grad_norm": 3.257079839706421, + "learning_rate": 9.9807679980768e-06, + "loss": 0.095, + "step": 19240 + }, + { + "epoch": 8.94514179451418, + "grad_norm": 5.204816818237305, + "learning_rate": 9.9807669980767e-06, + "loss": 0.1208, + "step": 19241 + }, + { + "epoch": 8.94560669456067, + "grad_norm": 3.1332175731658936, + "learning_rate": 9.9807659980766e-06, + "loss": 0.0598, + "step": 19242 + }, + { + "epoch": 8.94607159460716, + "grad_norm": 6.4920854568481445, + "learning_rate": 9.9807649980765e-06, + "loss": 0.2073, + "step": 19243 + }, + { + "epoch": 8.94653649465365, + "grad_norm": 2.122370719909668, + "learning_rate": 9.980763998076401e-06, + "loss": 0.0233, + "step": 19244 + }, + { + "epoch": 8.94700139470014, + "grad_norm": 11.390811920166016, + "learning_rate": 9.980762998076302e-06, + "loss": 0.1902, + "step": 19245 + }, + { + "epoch": 8.94746629474663, + "grad_norm": 3.409482717514038, + "learning_rate": 9.9807619980762e-06, + "loss": 0.0452, + "step": 19246 + }, + { + "epoch": 8.947931194793119, + "grad_norm": 7.567502498626709, + "learning_rate": 9.980760998076101e-06, + "loss": 0.2691, + "step": 19247 + }, + { + "epoch": 8.948396094839609, + "grad_norm": 2.305649995803833, + "learning_rate": 9.980759998076e-06, + "loss": 0.0314, + "step": 19248 + }, + { + "epoch": 8.9488609948861, + "grad_norm": 2.30107045173645, + "learning_rate": 9.980758998075901e-06, + "loss": 0.0257, + "step": 19249 + }, + { + "epoch": 8.94932589493259, + "grad_norm": 4.774809837341309, + "learning_rate": 9.980757998075802e-06, + "loss": 0.0756, + "step": 19250 + }, + { + "epoch": 8.94979079497908, + "grad_norm": 8.86432933807373, + "learning_rate": 9.9807569980757e-06, + "loss": 0.2344, + "step": 19251 + }, + { + "epoch": 8.95025569502557, + "grad_norm": 6.116054534912109, + "learning_rate": 9.980755998075601e-06, + "loss": 0.1492, + "step": 19252 + }, + { + "epoch": 8.95072059507206, + "grad_norm": 4.438571929931641, + "learning_rate": 9.9807549980755e-06, + "loss": 0.0852, + "step": 19253 + }, + { + "epoch": 8.95118549511855, + "grad_norm": 3.7388241291046143, + "learning_rate": 9.9807539980754e-06, + "loss": 0.0599, + "step": 19254 + }, + { + "epoch": 8.95165039516504, + "grad_norm": 4.673884391784668, + "learning_rate": 9.9807529980753e-06, + "loss": 0.0445, + "step": 19255 + }, + { + "epoch": 8.952115295211529, + "grad_norm": 8.49714183807373, + "learning_rate": 9.9807519980752e-06, + "loss": 0.2468, + "step": 19256 + }, + { + "epoch": 8.952580195258019, + "grad_norm": 4.246619701385498, + "learning_rate": 9.980750998075101e-06, + "loss": 0.0817, + "step": 19257 + }, + { + "epoch": 8.95304509530451, + "grad_norm": 5.262194633483887, + "learning_rate": 9.980749998075002e-06, + "loss": 0.0546, + "step": 19258 + }, + { + "epoch": 8.953509995351, + "grad_norm": 9.790513038635254, + "learning_rate": 9.9807489980749e-06, + "loss": 0.1041, + "step": 19259 + }, + { + "epoch": 8.95397489539749, + "grad_norm": 6.441954135894775, + "learning_rate": 9.9807479980748e-06, + "loss": 0.1001, + "step": 19260 + }, + { + "epoch": 8.95443979544398, + "grad_norm": 5.68648099899292, + "learning_rate": 9.9807469980747e-06, + "loss": 0.1062, + "step": 19261 + }, + { + "epoch": 8.95490469549047, + "grad_norm": 5.875322341918945, + "learning_rate": 9.980745998074601e-06, + "loss": 0.1071, + "step": 19262 + }, + { + "epoch": 8.95536959553696, + "grad_norm": 4.942568778991699, + "learning_rate": 9.980744998074502e-06, + "loss": 0.1063, + "step": 19263 + }, + { + "epoch": 8.95583449558345, + "grad_norm": 5.644944667816162, + "learning_rate": 9.9807439980744e-06, + "loss": 0.1714, + "step": 19264 + }, + { + "epoch": 8.956299395629939, + "grad_norm": 4.314201354980469, + "learning_rate": 9.980742998074301e-06, + "loss": 0.0495, + "step": 19265 + }, + { + "epoch": 8.956764295676429, + "grad_norm": 6.0954999923706055, + "learning_rate": 9.9807419980742e-06, + "loss": 0.08, + "step": 19266 + }, + { + "epoch": 8.95722919572292, + "grad_norm": 6.116894245147705, + "learning_rate": 9.9807409980741e-06, + "loss": 0.0718, + "step": 19267 + }, + { + "epoch": 8.95769409576941, + "grad_norm": 6.381427764892578, + "learning_rate": 9.980739998074e-06, + "loss": 0.1433, + "step": 19268 + }, + { + "epoch": 8.9581589958159, + "grad_norm": 4.452476978302002, + "learning_rate": 9.9807389980739e-06, + "loss": 0.0609, + "step": 19269 + }, + { + "epoch": 8.95862389586239, + "grad_norm": 7.29417610168457, + "learning_rate": 9.980737998073801e-06, + "loss": 0.1126, + "step": 19270 + }, + { + "epoch": 8.95908879590888, + "grad_norm": 3.447618246078491, + "learning_rate": 9.9807369980737e-06, + "loss": 0.0544, + "step": 19271 + }, + { + "epoch": 8.95955369595537, + "grad_norm": 8.451088905334473, + "learning_rate": 9.9807359980736e-06, + "loss": 0.0927, + "step": 19272 + }, + { + "epoch": 8.96001859600186, + "grad_norm": 3.326556921005249, + "learning_rate": 9.9807349980735e-06, + "loss": 0.0769, + "step": 19273 + }, + { + "epoch": 8.960483496048349, + "grad_norm": 4.755763530731201, + "learning_rate": 9.9807339980734e-06, + "loss": 0.1222, + "step": 19274 + }, + { + "epoch": 8.960948396094839, + "grad_norm": 4.421482086181641, + "learning_rate": 9.980732998073301e-06, + "loss": 0.0606, + "step": 19275 + }, + { + "epoch": 8.96141329614133, + "grad_norm": 5.442867279052734, + "learning_rate": 9.980731998073202e-06, + "loss": 0.093, + "step": 19276 + }, + { + "epoch": 8.96187819618782, + "grad_norm": 6.845210552215576, + "learning_rate": 9.9807309980731e-06, + "loss": 0.1059, + "step": 19277 + }, + { + "epoch": 8.96234309623431, + "grad_norm": 5.491179466247559, + "learning_rate": 9.980729998073001e-06, + "loss": 0.1154, + "step": 19278 + }, + { + "epoch": 8.9628079962808, + "grad_norm": 1.1970874071121216, + "learning_rate": 9.9807289980729e-06, + "loss": 0.0135, + "step": 19279 + }, + { + "epoch": 8.96327289632729, + "grad_norm": 4.744162082672119, + "learning_rate": 9.980727998072801e-06, + "loss": 0.0912, + "step": 19280 + }, + { + "epoch": 8.96373779637378, + "grad_norm": 4.660313606262207, + "learning_rate": 9.9807269980727e-06, + "loss": 0.0588, + "step": 19281 + }, + { + "epoch": 8.96420269642027, + "grad_norm": 3.304286241531372, + "learning_rate": 9.9807259980726e-06, + "loss": 0.0517, + "step": 19282 + }, + { + "epoch": 8.964667596466759, + "grad_norm": 4.589828014373779, + "learning_rate": 9.980724998072501e-06, + "loss": 0.0567, + "step": 19283 + }, + { + "epoch": 8.965132496513249, + "grad_norm": 4.715961456298828, + "learning_rate": 9.9807239980724e-06, + "loss": 0.0829, + "step": 19284 + }, + { + "epoch": 8.96559739655974, + "grad_norm": 8.875388145446777, + "learning_rate": 9.9807229980723e-06, + "loss": 0.1202, + "step": 19285 + }, + { + "epoch": 8.96606229660623, + "grad_norm": 3.5659446716308594, + "learning_rate": 9.9807219980722e-06, + "loss": 0.034, + "step": 19286 + }, + { + "epoch": 8.96652719665272, + "grad_norm": 5.629777908325195, + "learning_rate": 9.9807209980721e-06, + "loss": 0.1027, + "step": 19287 + }, + { + "epoch": 8.96699209669921, + "grad_norm": 7.353590965270996, + "learning_rate": 9.980719998072001e-06, + "loss": 0.2174, + "step": 19288 + }, + { + "epoch": 8.9674569967457, + "grad_norm": 11.787047386169434, + "learning_rate": 9.980718998071902e-06, + "loss": 0.1986, + "step": 19289 + }, + { + "epoch": 8.96792189679219, + "grad_norm": 4.775202751159668, + "learning_rate": 9.9807179980718e-06, + "loss": 0.1145, + "step": 19290 + }, + { + "epoch": 8.96838679683868, + "grad_norm": 5.708282947540283, + "learning_rate": 9.980716998071701e-06, + "loss": 0.0899, + "step": 19291 + }, + { + "epoch": 8.968851696885169, + "grad_norm": 4.166515350341797, + "learning_rate": 9.9807159980716e-06, + "loss": 0.0905, + "step": 19292 + }, + { + "epoch": 8.969316596931659, + "grad_norm": 7.110849857330322, + "learning_rate": 9.9807149980715e-06, + "loss": 0.2361, + "step": 19293 + }, + { + "epoch": 8.96978149697815, + "grad_norm": 8.32750415802002, + "learning_rate": 9.9807139980714e-06, + "loss": 0.104, + "step": 19294 + }, + { + "epoch": 8.97024639702464, + "grad_norm": 4.892879486083984, + "learning_rate": 9.9807129980713e-06, + "loss": 0.0682, + "step": 19295 + }, + { + "epoch": 8.97071129707113, + "grad_norm": 6.360534191131592, + "learning_rate": 9.980711998071201e-06, + "loss": 0.1268, + "step": 19296 + }, + { + "epoch": 8.97117619711762, + "grad_norm": 3.6400082111358643, + "learning_rate": 9.9807109980711e-06, + "loss": 0.1046, + "step": 19297 + }, + { + "epoch": 8.97164109716411, + "grad_norm": 6.2062458992004395, + "learning_rate": 9.980709998071001e-06, + "loss": 0.1129, + "step": 19298 + }, + { + "epoch": 8.9721059972106, + "grad_norm": 4.17375373840332, + "learning_rate": 9.9807089980709e-06, + "loss": 0.0492, + "step": 19299 + }, + { + "epoch": 8.97257089725709, + "grad_norm": 4.922286510467529, + "learning_rate": 9.9807079980708e-06, + "loss": 0.0754, + "step": 19300 + }, + { + "epoch": 8.973035797303579, + "grad_norm": 6.747145652770996, + "learning_rate": 9.980706998070701e-06, + "loss": 0.1922, + "step": 19301 + }, + { + "epoch": 8.973500697350069, + "grad_norm": 7.635819911956787, + "learning_rate": 9.980705998070602e-06, + "loss": 0.2669, + "step": 19302 + }, + { + "epoch": 8.97396559739656, + "grad_norm": 7.1151885986328125, + "learning_rate": 9.9807049980705e-06, + "loss": 0.114, + "step": 19303 + }, + { + "epoch": 8.97443049744305, + "grad_norm": 6.760034561157227, + "learning_rate": 9.9807039980704e-06, + "loss": 0.1433, + "step": 19304 + }, + { + "epoch": 8.97489539748954, + "grad_norm": 7.283315658569336, + "learning_rate": 9.9807029980703e-06, + "loss": 0.3213, + "step": 19305 + }, + { + "epoch": 8.97536029753603, + "grad_norm": 6.181377410888672, + "learning_rate": 9.980701998070201e-06, + "loss": 0.1088, + "step": 19306 + }, + { + "epoch": 8.97582519758252, + "grad_norm": 4.151980400085449, + "learning_rate": 9.9807009980701e-06, + "loss": 0.0695, + "step": 19307 + }, + { + "epoch": 8.97629009762901, + "grad_norm": 8.275857925415039, + "learning_rate": 9.98069999807e-06, + "loss": 0.1013, + "step": 19308 + }, + { + "epoch": 8.9767549976755, + "grad_norm": 4.436487674713135, + "learning_rate": 9.980698998069901e-06, + "loss": 0.1095, + "step": 19309 + }, + { + "epoch": 8.977219897721989, + "grad_norm": 7.539225101470947, + "learning_rate": 9.9806979980698e-06, + "loss": 0.2488, + "step": 19310 + }, + { + "epoch": 8.97768479776848, + "grad_norm": 4.270263671875, + "learning_rate": 9.980696998069701e-06, + "loss": 0.121, + "step": 19311 + }, + { + "epoch": 8.97814969781497, + "grad_norm": 5.1999077796936035, + "learning_rate": 9.9806959980696e-06, + "loss": 0.1209, + "step": 19312 + }, + { + "epoch": 8.97861459786146, + "grad_norm": 7.064218997955322, + "learning_rate": 9.9806949980695e-06, + "loss": 0.1204, + "step": 19313 + }, + { + "epoch": 8.97907949790795, + "grad_norm": 5.934605598449707, + "learning_rate": 9.980693998069401e-06, + "loss": 0.1228, + "step": 19314 + }, + { + "epoch": 8.97954439795444, + "grad_norm": 9.5391263961792, + "learning_rate": 9.9806929980693e-06, + "loss": 0.2287, + "step": 19315 + }, + { + "epoch": 8.98000929800093, + "grad_norm": 4.320652961730957, + "learning_rate": 9.980691998069201e-06, + "loss": 0.0766, + "step": 19316 + }, + { + "epoch": 8.98047419804742, + "grad_norm": 6.737740993499756, + "learning_rate": 9.9806909980691e-06, + "loss": 0.147, + "step": 19317 + }, + { + "epoch": 8.98093909809391, + "grad_norm": 3.9192631244659424, + "learning_rate": 9.980689998069e-06, + "loss": 0.0547, + "step": 19318 + }, + { + "epoch": 8.981403998140399, + "grad_norm": 6.424857139587402, + "learning_rate": 9.980688998068901e-06, + "loss": 0.1124, + "step": 19319 + }, + { + "epoch": 8.98186889818689, + "grad_norm": 7.731219291687012, + "learning_rate": 9.9806879980688e-06, + "loss": 0.2751, + "step": 19320 + }, + { + "epoch": 8.98233379823338, + "grad_norm": 3.343348503112793, + "learning_rate": 9.9806869980687e-06, + "loss": 0.0528, + "step": 19321 + }, + { + "epoch": 8.98279869827987, + "grad_norm": 4.932203769683838, + "learning_rate": 9.980685998068601e-06, + "loss": 0.1001, + "step": 19322 + }, + { + "epoch": 8.98326359832636, + "grad_norm": 6.987316608428955, + "learning_rate": 9.9806849980685e-06, + "loss": 0.2057, + "step": 19323 + }, + { + "epoch": 8.98372849837285, + "grad_norm": 12.280414581298828, + "learning_rate": 9.980683998068401e-06, + "loss": 0.0616, + "step": 19324 + }, + { + "epoch": 8.98419339841934, + "grad_norm": 2.242920398712158, + "learning_rate": 9.9806829980683e-06, + "loss": 0.0227, + "step": 19325 + }, + { + "epoch": 8.98465829846583, + "grad_norm": 12.290583610534668, + "learning_rate": 9.9806819980682e-06, + "loss": 0.1658, + "step": 19326 + }, + { + "epoch": 8.98512319851232, + "grad_norm": 7.696313381195068, + "learning_rate": 9.980680998068101e-06, + "loss": 0.1508, + "step": 19327 + }, + { + "epoch": 8.98558809855881, + "grad_norm": 5.661344051361084, + "learning_rate": 9.980679998068e-06, + "loss": 0.1085, + "step": 19328 + }, + { + "epoch": 8.9860529986053, + "grad_norm": 6.569555282592773, + "learning_rate": 9.980678998067901e-06, + "loss": 0.1854, + "step": 19329 + }, + { + "epoch": 8.98651789865179, + "grad_norm": 3.900102138519287, + "learning_rate": 9.9806779980678e-06, + "loss": 0.0742, + "step": 19330 + }, + { + "epoch": 8.98698279869828, + "grad_norm": 6.282718181610107, + "learning_rate": 9.9806769980677e-06, + "loss": 0.0656, + "step": 19331 + }, + { + "epoch": 8.98744769874477, + "grad_norm": 5.314126491546631, + "learning_rate": 9.980675998067601e-06, + "loss": 0.2391, + "step": 19332 + }, + { + "epoch": 8.98791259879126, + "grad_norm": 3.5197384357452393, + "learning_rate": 9.980674998067502e-06, + "loss": 0.0578, + "step": 19333 + }, + { + "epoch": 8.98837749883775, + "grad_norm": 8.236445426940918, + "learning_rate": 9.980673998067401e-06, + "loss": 0.1358, + "step": 19334 + }, + { + "epoch": 8.98884239888424, + "grad_norm": 6.490804672241211, + "learning_rate": 9.980672998067302e-06, + "loss": 0.155, + "step": 19335 + }, + { + "epoch": 8.98930729893073, + "grad_norm": 3.857269048690796, + "learning_rate": 9.9806719980672e-06, + "loss": 0.054, + "step": 19336 + }, + { + "epoch": 8.98977219897722, + "grad_norm": 10.840707778930664, + "learning_rate": 9.9806709980671e-06, + "loss": 0.1517, + "step": 19337 + }, + { + "epoch": 8.99023709902371, + "grad_norm": 3.273552656173706, + "learning_rate": 9.980669998067e-06, + "loss": 0.0344, + "step": 19338 + }, + { + "epoch": 8.9907019990702, + "grad_norm": 4.503715991973877, + "learning_rate": 9.9806689980669e-06, + "loss": 0.044, + "step": 19339 + }, + { + "epoch": 8.99116689911669, + "grad_norm": 2.4596288204193115, + "learning_rate": 9.980667998066801e-06, + "loss": 0.0414, + "step": 19340 + }, + { + "epoch": 8.99163179916318, + "grad_norm": 2.808023691177368, + "learning_rate": 9.9806669980667e-06, + "loss": 0.0428, + "step": 19341 + }, + { + "epoch": 8.99209669920967, + "grad_norm": 4.4762492179870605, + "learning_rate": 9.980665998066601e-06, + "loss": 0.0943, + "step": 19342 + }, + { + "epoch": 8.99256159925616, + "grad_norm": 4.798800468444824, + "learning_rate": 9.9806649980665e-06, + "loss": 0.1336, + "step": 19343 + }, + { + "epoch": 8.99302649930265, + "grad_norm": 4.428250789642334, + "learning_rate": 9.9806639980664e-06, + "loss": 0.0997, + "step": 19344 + }, + { + "epoch": 8.99349139934914, + "grad_norm": 4.4531755447387695, + "learning_rate": 9.980662998066301e-06, + "loss": 0.0704, + "step": 19345 + }, + { + "epoch": 8.99395629939563, + "grad_norm": 6.849560737609863, + "learning_rate": 9.980661998066202e-06, + "loss": 0.1865, + "step": 19346 + }, + { + "epoch": 8.99442119944212, + "grad_norm": 6.320769786834717, + "learning_rate": 9.980660998066101e-06, + "loss": 0.1789, + "step": 19347 + }, + { + "epoch": 8.99488609948861, + "grad_norm": 4.1251726150512695, + "learning_rate": 9.980659998066e-06, + "loss": 0.0526, + "step": 19348 + }, + { + "epoch": 8.9953509995351, + "grad_norm": 3.174421787261963, + "learning_rate": 9.9806589980659e-06, + "loss": 0.0551, + "step": 19349 + }, + { + "epoch": 8.99581589958159, + "grad_norm": 4.818490028381348, + "learning_rate": 9.9806579980658e-06, + "loss": 0.0968, + "step": 19350 + }, + { + "epoch": 8.99628079962808, + "grad_norm": 4.478824138641357, + "learning_rate": 9.9806569980657e-06, + "loss": 0.0615, + "step": 19351 + }, + { + "epoch": 8.99674569967457, + "grad_norm": 8.548190116882324, + "learning_rate": 9.980655998065601e-06, + "loss": 0.2501, + "step": 19352 + }, + { + "epoch": 8.99721059972106, + "grad_norm": 3.734307050704956, + "learning_rate": 9.980654998065502e-06, + "loss": 0.0577, + "step": 19353 + }, + { + "epoch": 8.99767549976755, + "grad_norm": 8.518980979919434, + "learning_rate": 9.9806539980654e-06, + "loss": 0.1923, + "step": 19354 + }, + { + "epoch": 8.99814039981404, + "grad_norm": 5.0239057540893555, + "learning_rate": 9.980652998065301e-06, + "loss": 0.086, + "step": 19355 + }, + { + "epoch": 8.99860529986053, + "grad_norm": 11.298757553100586, + "learning_rate": 9.9806519980652e-06, + "loss": 0.0663, + "step": 19356 + }, + { + "epoch": 8.99907019990702, + "grad_norm": 9.068446159362793, + "learning_rate": 9.9806509980651e-06, + "loss": 0.2843, + "step": 19357 + }, + { + "epoch": 8.99953509995351, + "grad_norm": 4.303317070007324, + "learning_rate": 9.980649998065001e-06, + "loss": 0.0714, + "step": 19358 + }, + { + "epoch": 9.0, + "grad_norm": 13.22787857055664, + "learning_rate": 9.9806489980649e-06, + "loss": 0.1778, + "step": 19359 + }, + { + "epoch": 9.00046490004649, + "grad_norm": 4.820374011993408, + "learning_rate": 9.980647998064801e-06, + "loss": 0.0637, + "step": 19360 + }, + { + "epoch": 9.00092980009298, + "grad_norm": 5.055737018585205, + "learning_rate": 9.9806469980647e-06, + "loss": 0.0849, + "step": 19361 + }, + { + "epoch": 9.00139470013947, + "grad_norm": 2.6436121463775635, + "learning_rate": 9.9806459980646e-06, + "loss": 0.0434, + "step": 19362 + }, + { + "epoch": 9.00185960018596, + "grad_norm": 4.001456260681152, + "learning_rate": 9.9806449980645e-06, + "loss": 0.0226, + "step": 19363 + }, + { + "epoch": 9.00232450023245, + "grad_norm": 5.976827144622803, + "learning_rate": 9.9806439980644e-06, + "loss": 0.1518, + "step": 19364 + }, + { + "epoch": 9.00278940027894, + "grad_norm": 5.598483562469482, + "learning_rate": 9.980642998064301e-06, + "loss": 0.0638, + "step": 19365 + }, + { + "epoch": 9.00325430032543, + "grad_norm": 5.497462272644043, + "learning_rate": 9.980641998064202e-06, + "loss": 0.1169, + "step": 19366 + }, + { + "epoch": 9.00371920037192, + "grad_norm": 5.31728458404541, + "learning_rate": 9.9806409980641e-06, + "loss": 0.0766, + "step": 19367 + }, + { + "epoch": 9.00418410041841, + "grad_norm": 6.907545566558838, + "learning_rate": 9.980639998064e-06, + "loss": 0.1369, + "step": 19368 + }, + { + "epoch": 9.0046490004649, + "grad_norm": 2.5340347290039062, + "learning_rate": 9.9806389980639e-06, + "loss": 0.0284, + "step": 19369 + }, + { + "epoch": 9.00511390051139, + "grad_norm": 2.817528009414673, + "learning_rate": 9.9806379980638e-06, + "loss": 0.0374, + "step": 19370 + }, + { + "epoch": 9.00557880055788, + "grad_norm": 1.8651169538497925, + "learning_rate": 9.980636998063701e-06, + "loss": 0.016, + "step": 19371 + }, + { + "epoch": 9.00604370060437, + "grad_norm": 3.1163222789764404, + "learning_rate": 9.9806359980636e-06, + "loss": 0.0325, + "step": 19372 + }, + { + "epoch": 9.00650860065086, + "grad_norm": 5.458584785461426, + "learning_rate": 9.980634998063501e-06, + "loss": 0.0567, + "step": 19373 + }, + { + "epoch": 9.00697350069735, + "grad_norm": 3.872800350189209, + "learning_rate": 9.9806339980634e-06, + "loss": 0.0669, + "step": 19374 + }, + { + "epoch": 9.00743840074384, + "grad_norm": 6.562297344207764, + "learning_rate": 9.9806329980633e-06, + "loss": 0.1176, + "step": 19375 + }, + { + "epoch": 9.00790330079033, + "grad_norm": 3.138989210128784, + "learning_rate": 9.9806319980632e-06, + "loss": 0.0435, + "step": 19376 + }, + { + "epoch": 9.00836820083682, + "grad_norm": 9.708102226257324, + "learning_rate": 9.9806309980631e-06, + "loss": 0.1785, + "step": 19377 + }, + { + "epoch": 9.00883310088331, + "grad_norm": 8.337137222290039, + "learning_rate": 9.980629998063001e-06, + "loss": 0.3595, + "step": 19378 + }, + { + "epoch": 9.0092980009298, + "grad_norm": 10.40737247467041, + "learning_rate": 9.9806289980629e-06, + "loss": 0.1113, + "step": 19379 + }, + { + "epoch": 9.00976290097629, + "grad_norm": 4.377594947814941, + "learning_rate": 9.9806279980628e-06, + "loss": 0.0863, + "step": 19380 + }, + { + "epoch": 9.010227801022781, + "grad_norm": 5.363729476928711, + "learning_rate": 9.9806269980627e-06, + "loss": 0.0692, + "step": 19381 + }, + { + "epoch": 9.01069270106927, + "grad_norm": 4.796966075897217, + "learning_rate": 9.9806259980626e-06, + "loss": 0.0898, + "step": 19382 + }, + { + "epoch": 9.01115760111576, + "grad_norm": 2.0724639892578125, + "learning_rate": 9.980624998062501e-06, + "loss": 0.02, + "step": 19383 + }, + { + "epoch": 9.01162250116225, + "grad_norm": 5.087323188781738, + "learning_rate": 9.980623998062402e-06, + "loss": 0.0539, + "step": 19384 + }, + { + "epoch": 9.01208740120874, + "grad_norm": 3.762726306915283, + "learning_rate": 9.9806229980623e-06, + "loss": 0.076, + "step": 19385 + }, + { + "epoch": 9.01255230125523, + "grad_norm": 0.8150131702423096, + "learning_rate": 9.980621998062201e-06, + "loss": 0.0106, + "step": 19386 + }, + { + "epoch": 9.01301720130172, + "grad_norm": 4.8060150146484375, + "learning_rate": 9.9806209980621e-06, + "loss": 0.0508, + "step": 19387 + }, + { + "epoch": 9.01348210134821, + "grad_norm": 1.95049250125885, + "learning_rate": 9.980619998062e-06, + "loss": 0.0108, + "step": 19388 + }, + { + "epoch": 9.0139470013947, + "grad_norm": 5.3045172691345215, + "learning_rate": 9.9806189980619e-06, + "loss": 0.0994, + "step": 19389 + }, + { + "epoch": 9.014411901441191, + "grad_norm": 2.4338183403015137, + "learning_rate": 9.9806179980618e-06, + "loss": 0.0235, + "step": 19390 + }, + { + "epoch": 9.01487680148768, + "grad_norm": 5.895768642425537, + "learning_rate": 9.980616998061701e-06, + "loss": 0.1357, + "step": 19391 + }, + { + "epoch": 9.01534170153417, + "grad_norm": 2.7281272411346436, + "learning_rate": 9.9806159980616e-06, + "loss": 0.0405, + "step": 19392 + }, + { + "epoch": 9.01580660158066, + "grad_norm": 3.0162761211395264, + "learning_rate": 9.9806149980615e-06, + "loss": 0.0201, + "step": 19393 + }, + { + "epoch": 9.01627150162715, + "grad_norm": 2.846644639968872, + "learning_rate": 9.9806139980614e-06, + "loss": 0.0478, + "step": 19394 + }, + { + "epoch": 9.01673640167364, + "grad_norm": 2.689761161804199, + "learning_rate": 9.9806129980613e-06, + "loss": 0.0335, + "step": 19395 + }, + { + "epoch": 9.01720130172013, + "grad_norm": 7.757426738739014, + "learning_rate": 9.980611998061201e-06, + "loss": 0.2411, + "step": 19396 + }, + { + "epoch": 9.01766620176662, + "grad_norm": 4.4523444175720215, + "learning_rate": 9.980610998061102e-06, + "loss": 0.0868, + "step": 19397 + }, + { + "epoch": 9.01813110181311, + "grad_norm": 2.786903142929077, + "learning_rate": 9.980609998061e-06, + "loss": 0.0188, + "step": 19398 + }, + { + "epoch": 9.018596001859601, + "grad_norm": 8.029206275939941, + "learning_rate": 9.980608998060901e-06, + "loss": 0.1015, + "step": 19399 + }, + { + "epoch": 9.01906090190609, + "grad_norm": 0.45063135027885437, + "learning_rate": 9.9806079980608e-06, + "loss": 0.004, + "step": 19400 + }, + { + "epoch": 9.01952580195258, + "grad_norm": 1.3446041345596313, + "learning_rate": 9.980606998060701e-06, + "loss": 0.0149, + "step": 19401 + }, + { + "epoch": 9.01999070199907, + "grad_norm": 2.447730541229248, + "learning_rate": 9.980605998060602e-06, + "loss": 0.0203, + "step": 19402 + }, + { + "epoch": 9.02045560204556, + "grad_norm": 5.381226062774658, + "learning_rate": 9.9806049980605e-06, + "loss": 0.1423, + "step": 19403 + }, + { + "epoch": 9.02092050209205, + "grad_norm": 3.8829050064086914, + "learning_rate": 9.980603998060401e-06, + "loss": 0.0654, + "step": 19404 + }, + { + "epoch": 9.02138540213854, + "grad_norm": 2.0669147968292236, + "learning_rate": 9.9806029980603e-06, + "loss": 0.0222, + "step": 19405 + }, + { + "epoch": 9.02185030218503, + "grad_norm": 3.882063150405884, + "learning_rate": 9.9806019980602e-06, + "loss": 0.048, + "step": 19406 + }, + { + "epoch": 9.02231520223152, + "grad_norm": 8.728221893310547, + "learning_rate": 9.9806009980601e-06, + "loss": 0.2233, + "step": 19407 + }, + { + "epoch": 9.022780102278011, + "grad_norm": 2.831544876098633, + "learning_rate": 9.98059999806e-06, + "loss": 0.0371, + "step": 19408 + }, + { + "epoch": 9.0232450023245, + "grad_norm": 6.034128665924072, + "learning_rate": 9.980598998059901e-06, + "loss": 0.2108, + "step": 19409 + }, + { + "epoch": 9.02370990237099, + "grad_norm": 3.4900007247924805, + "learning_rate": 9.980597998059802e-06, + "loss": 0.0307, + "step": 19410 + }, + { + "epoch": 9.02417480241748, + "grad_norm": 4.917760372161865, + "learning_rate": 9.9805969980597e-06, + "loss": 0.074, + "step": 19411 + }, + { + "epoch": 9.02463970246397, + "grad_norm": 2.2844161987304688, + "learning_rate": 9.9805959980596e-06, + "loss": 0.0257, + "step": 19412 + }, + { + "epoch": 9.02510460251046, + "grad_norm": 1.0415456295013428, + "learning_rate": 9.9805949980595e-06, + "loss": 0.0078, + "step": 19413 + }, + { + "epoch": 9.02556950255695, + "grad_norm": 4.6095147132873535, + "learning_rate": 9.980593998059401e-06, + "loss": 0.0587, + "step": 19414 + }, + { + "epoch": 9.02603440260344, + "grad_norm": 5.94534158706665, + "learning_rate": 9.980592998059302e-06, + "loss": 0.0963, + "step": 19415 + }, + { + "epoch": 9.02649930264993, + "grad_norm": 6.800459861755371, + "learning_rate": 9.9805919980592e-06, + "loss": 0.1285, + "step": 19416 + }, + { + "epoch": 9.026964202696421, + "grad_norm": 3.0547640323638916, + "learning_rate": 9.980590998059101e-06, + "loss": 0.0338, + "step": 19417 + }, + { + "epoch": 9.02742910274291, + "grad_norm": 4.1349778175354, + "learning_rate": 9.980589998059e-06, + "loss": 0.0295, + "step": 19418 + }, + { + "epoch": 9.0278940027894, + "grad_norm": 6.2978386878967285, + "learning_rate": 9.980588998058901e-06, + "loss": 0.1338, + "step": 19419 + }, + { + "epoch": 9.02835890283589, + "grad_norm": 1.8869997262954712, + "learning_rate": 9.9805879980588e-06, + "loss": 0.0174, + "step": 19420 + }, + { + "epoch": 9.02882380288238, + "grad_norm": 2.0648787021636963, + "learning_rate": 9.9805869980587e-06, + "loss": 0.0332, + "step": 19421 + }, + { + "epoch": 9.02928870292887, + "grad_norm": 3.7310149669647217, + "learning_rate": 9.980585998058601e-06, + "loss": 0.0479, + "step": 19422 + }, + { + "epoch": 9.02975360297536, + "grad_norm": 3.2256131172180176, + "learning_rate": 9.9805849980585e-06, + "loss": 0.055, + "step": 19423 + }, + { + "epoch": 9.03021850302185, + "grad_norm": 3.6805098056793213, + "learning_rate": 9.9805839980584e-06, + "loss": 0.0687, + "step": 19424 + }, + { + "epoch": 9.030683403068341, + "grad_norm": 2.5558278560638428, + "learning_rate": 9.9805829980583e-06, + "loss": 0.0381, + "step": 19425 + }, + { + "epoch": 9.031148303114831, + "grad_norm": 4.708807468414307, + "learning_rate": 9.9805819980582e-06, + "loss": 0.0945, + "step": 19426 + }, + { + "epoch": 9.03161320316132, + "grad_norm": 5.167888641357422, + "learning_rate": 9.980580998058101e-06, + "loss": 0.1121, + "step": 19427 + }, + { + "epoch": 9.03207810320781, + "grad_norm": 2.6754629611968994, + "learning_rate": 9.980579998058002e-06, + "loss": 0.0364, + "step": 19428 + }, + { + "epoch": 9.0325430032543, + "grad_norm": 2.459336519241333, + "learning_rate": 9.9805789980579e-06, + "loss": 0.0229, + "step": 19429 + }, + { + "epoch": 9.03300790330079, + "grad_norm": 3.84295392036438, + "learning_rate": 9.980577998057801e-06, + "loss": 0.0475, + "step": 19430 + }, + { + "epoch": 9.03347280334728, + "grad_norm": 3.818521499633789, + "learning_rate": 9.9805769980577e-06, + "loss": 0.1315, + "step": 19431 + }, + { + "epoch": 9.03393770339377, + "grad_norm": 2.5150132179260254, + "learning_rate": 9.980575998057601e-06, + "loss": 0.0315, + "step": 19432 + }, + { + "epoch": 9.03440260344026, + "grad_norm": 6.052791595458984, + "learning_rate": 9.9805749980575e-06, + "loss": 0.1412, + "step": 19433 + }, + { + "epoch": 9.034867503486751, + "grad_norm": 6.6494526863098145, + "learning_rate": 9.9805739980574e-06, + "loss": 0.0999, + "step": 19434 + }, + { + "epoch": 9.035332403533241, + "grad_norm": 2.737929582595825, + "learning_rate": 9.980572998057301e-06, + "loss": 0.0259, + "step": 19435 + }, + { + "epoch": 9.03579730357973, + "grad_norm": 8.04685115814209, + "learning_rate": 9.9805719980572e-06, + "loss": 0.1116, + "step": 19436 + }, + { + "epoch": 9.03626220362622, + "grad_norm": 5.362517356872559, + "learning_rate": 9.980570998057101e-06, + "loss": 0.0681, + "step": 19437 + }, + { + "epoch": 9.03672710367271, + "grad_norm": 5.812959671020508, + "learning_rate": 9.980569998057e-06, + "loss": 0.0777, + "step": 19438 + }, + { + "epoch": 9.0371920037192, + "grad_norm": 6.134552478790283, + "learning_rate": 9.9805689980569e-06, + "loss": 0.1139, + "step": 19439 + }, + { + "epoch": 9.03765690376569, + "grad_norm": 4.687517166137695, + "learning_rate": 9.980567998056801e-06, + "loss": 0.0748, + "step": 19440 + }, + { + "epoch": 9.03812180381218, + "grad_norm": 2.7160191535949707, + "learning_rate": 9.980566998056702e-06, + "loss": 0.0312, + "step": 19441 + }, + { + "epoch": 9.03858670385867, + "grad_norm": 4.211343765258789, + "learning_rate": 9.9805659980566e-06, + "loss": 0.1084, + "step": 19442 + }, + { + "epoch": 9.039051603905161, + "grad_norm": 6.5854082107543945, + "learning_rate": 9.980564998056501e-06, + "loss": 0.1026, + "step": 19443 + }, + { + "epoch": 9.039516503951651, + "grad_norm": 5.38449239730835, + "learning_rate": 9.9805639980564e-06, + "loss": 0.1432, + "step": 19444 + }, + { + "epoch": 9.03998140399814, + "grad_norm": 3.3426854610443115, + "learning_rate": 9.9805629980563e-06, + "loss": 0.066, + "step": 19445 + }, + { + "epoch": 9.04044630404463, + "grad_norm": 5.539044380187988, + "learning_rate": 9.9805619980562e-06, + "loss": 0.0926, + "step": 19446 + }, + { + "epoch": 9.04091120409112, + "grad_norm": 3.6681528091430664, + "learning_rate": 9.9805609980561e-06, + "loss": 0.0647, + "step": 19447 + }, + { + "epoch": 9.04137610413761, + "grad_norm": 5.856487274169922, + "learning_rate": 9.980559998056001e-06, + "loss": 0.1063, + "step": 19448 + }, + { + "epoch": 9.0418410041841, + "grad_norm": 4.847934246063232, + "learning_rate": 9.9805589980559e-06, + "loss": 0.0869, + "step": 19449 + }, + { + "epoch": 9.04230590423059, + "grad_norm": 7.712080955505371, + "learning_rate": 9.980557998055801e-06, + "loss": 0.1997, + "step": 19450 + }, + { + "epoch": 9.04277080427708, + "grad_norm": 3.1514456272125244, + "learning_rate": 9.9805569980557e-06, + "loss": 0.0215, + "step": 19451 + }, + { + "epoch": 9.043235704323571, + "grad_norm": 7.331395149230957, + "learning_rate": 9.9805559980556e-06, + "loss": 0.1265, + "step": 19452 + }, + { + "epoch": 9.043700604370061, + "grad_norm": 1.4511637687683105, + "learning_rate": 9.980554998055501e-06, + "loss": 0.0182, + "step": 19453 + }, + { + "epoch": 9.04416550441655, + "grad_norm": 7.625069618225098, + "learning_rate": 9.980553998055402e-06, + "loss": 0.1435, + "step": 19454 + }, + { + "epoch": 9.04463040446304, + "grad_norm": 2.716573715209961, + "learning_rate": 9.9805529980553e-06, + "loss": 0.0293, + "step": 19455 + }, + { + "epoch": 9.04509530450953, + "grad_norm": 5.6555328369140625, + "learning_rate": 9.9805519980552e-06, + "loss": 0.1499, + "step": 19456 + }, + { + "epoch": 9.04556020455602, + "grad_norm": 3.8171310424804688, + "learning_rate": 9.9805509980551e-06, + "loss": 0.0202, + "step": 19457 + }, + { + "epoch": 9.04602510460251, + "grad_norm": 3.8810911178588867, + "learning_rate": 9.980549998055e-06, + "loss": 0.0434, + "step": 19458 + }, + { + "epoch": 9.046490004649, + "grad_norm": 11.585091590881348, + "learning_rate": 9.9805489980549e-06, + "loss": 0.211, + "step": 19459 + }, + { + "epoch": 9.046954904695491, + "grad_norm": 4.171889305114746, + "learning_rate": 9.9805479980548e-06, + "loss": 0.1027, + "step": 19460 + }, + { + "epoch": 9.047419804741981, + "grad_norm": 9.374889373779297, + "learning_rate": 9.980546998054701e-06, + "loss": 0.101, + "step": 19461 + }, + { + "epoch": 9.047884704788471, + "grad_norm": 6.60037899017334, + "learning_rate": 9.9805459980546e-06, + "loss": 0.1257, + "step": 19462 + }, + { + "epoch": 9.04834960483496, + "grad_norm": 5.0959391593933105, + "learning_rate": 9.980544998054501e-06, + "loss": 0.1272, + "step": 19463 + }, + { + "epoch": 9.04881450488145, + "grad_norm": 2.5182931423187256, + "learning_rate": 9.9805439980544e-06, + "loss": 0.0258, + "step": 19464 + }, + { + "epoch": 9.04927940492794, + "grad_norm": 5.94230318069458, + "learning_rate": 9.9805429980543e-06, + "loss": 0.0567, + "step": 19465 + }, + { + "epoch": 9.04974430497443, + "grad_norm": 10.404760360717773, + "learning_rate": 9.980541998054201e-06, + "loss": 0.104, + "step": 19466 + }, + { + "epoch": 9.05020920502092, + "grad_norm": 8.80420970916748, + "learning_rate": 9.9805409980541e-06, + "loss": 0.0618, + "step": 19467 + }, + { + "epoch": 9.05067410506741, + "grad_norm": 6.334624290466309, + "learning_rate": 9.980539998054001e-06, + "loss": 0.0883, + "step": 19468 + }, + { + "epoch": 9.051139005113901, + "grad_norm": 2.973869562149048, + "learning_rate": 9.9805389980539e-06, + "loss": 0.0417, + "step": 19469 + }, + { + "epoch": 9.051603905160391, + "grad_norm": 2.407372236251831, + "learning_rate": 9.9805379980538e-06, + "loss": 0.0382, + "step": 19470 + }, + { + "epoch": 9.052068805206881, + "grad_norm": 4.381226539611816, + "learning_rate": 9.980536998053701e-06, + "loss": 0.0709, + "step": 19471 + }, + { + "epoch": 9.05253370525337, + "grad_norm": 3.2410998344421387, + "learning_rate": 9.9805359980536e-06, + "loss": 0.0255, + "step": 19472 + }, + { + "epoch": 9.05299860529986, + "grad_norm": 6.517363548278809, + "learning_rate": 9.9805349980535e-06, + "loss": 0.2525, + "step": 19473 + }, + { + "epoch": 9.05346350534635, + "grad_norm": 2.9068830013275146, + "learning_rate": 9.980533998053401e-06, + "loss": 0.047, + "step": 19474 + }, + { + "epoch": 9.05392840539284, + "grad_norm": 6.6577887535095215, + "learning_rate": 9.9805329980533e-06, + "loss": 0.0785, + "step": 19475 + }, + { + "epoch": 9.05439330543933, + "grad_norm": 6.541938304901123, + "learning_rate": 9.980531998053201e-06, + "loss": 0.1546, + "step": 19476 + }, + { + "epoch": 9.05485820548582, + "grad_norm": 2.6760241985321045, + "learning_rate": 9.9805309980531e-06, + "loss": 0.0192, + "step": 19477 + }, + { + "epoch": 9.055323105532311, + "grad_norm": 3.576256513595581, + "learning_rate": 9.980529998053e-06, + "loss": 0.0706, + "step": 19478 + }, + { + "epoch": 9.055788005578801, + "grad_norm": 3.7497286796569824, + "learning_rate": 9.980528998052901e-06, + "loss": 0.0738, + "step": 19479 + }, + { + "epoch": 9.056252905625291, + "grad_norm": 4.115678310394287, + "learning_rate": 9.9805279980528e-06, + "loss": 0.0714, + "step": 19480 + }, + { + "epoch": 9.05671780567178, + "grad_norm": 4.434406757354736, + "learning_rate": 9.980526998052701e-06, + "loss": 0.0602, + "step": 19481 + }, + { + "epoch": 9.05718270571827, + "grad_norm": 2.187208652496338, + "learning_rate": 9.9805259980526e-06, + "loss": 0.0256, + "step": 19482 + }, + { + "epoch": 9.05764760576476, + "grad_norm": 3.644698143005371, + "learning_rate": 9.9805249980525e-06, + "loss": 0.0609, + "step": 19483 + }, + { + "epoch": 9.05811250581125, + "grad_norm": 8.074810028076172, + "learning_rate": 9.980523998052401e-06, + "loss": 0.1456, + "step": 19484 + }, + { + "epoch": 9.05857740585774, + "grad_norm": 6.77803897857666, + "learning_rate": 9.9805229980523e-06, + "loss": 0.1778, + "step": 19485 + }, + { + "epoch": 9.05904230590423, + "grad_norm": 3.398413896560669, + "learning_rate": 9.980521998052201e-06, + "loss": 0.0334, + "step": 19486 + }, + { + "epoch": 9.059507205950721, + "grad_norm": 1.8391309976577759, + "learning_rate": 9.980520998052102e-06, + "loss": 0.0126, + "step": 19487 + }, + { + "epoch": 9.059972105997211, + "grad_norm": 7.204632759094238, + "learning_rate": 9.980519998052e-06, + "loss": 0.1044, + "step": 19488 + }, + { + "epoch": 9.060437006043701, + "grad_norm": 4.385707378387451, + "learning_rate": 9.9805189980519e-06, + "loss": 0.0377, + "step": 19489 + }, + { + "epoch": 9.06090190609019, + "grad_norm": 3.381537437438965, + "learning_rate": 9.9805179980518e-06, + "loss": 0.055, + "step": 19490 + }, + { + "epoch": 9.06136680613668, + "grad_norm": 4.133063793182373, + "learning_rate": 9.9805169980517e-06, + "loss": 0.0732, + "step": 19491 + }, + { + "epoch": 9.06183170618317, + "grad_norm": 17.856077194213867, + "learning_rate": 9.980515998051601e-06, + "loss": 0.1254, + "step": 19492 + }, + { + "epoch": 9.06229660622966, + "grad_norm": 7.344166278839111, + "learning_rate": 9.9805149980515e-06, + "loss": 0.1467, + "step": 19493 + }, + { + "epoch": 9.06276150627615, + "grad_norm": 5.016892910003662, + "learning_rate": 9.980513998051401e-06, + "loss": 0.0386, + "step": 19494 + }, + { + "epoch": 9.06322640632264, + "grad_norm": 4.693424224853516, + "learning_rate": 9.9805129980513e-06, + "loss": 0.0641, + "step": 19495 + }, + { + "epoch": 9.063691306369131, + "grad_norm": 3.394594430923462, + "learning_rate": 9.9805119980512e-06, + "loss": 0.07, + "step": 19496 + }, + { + "epoch": 9.064156206415621, + "grad_norm": 3.286571502685547, + "learning_rate": 9.980510998051101e-06, + "loss": 0.0337, + "step": 19497 + }, + { + "epoch": 9.064621106462111, + "grad_norm": 3.514357328414917, + "learning_rate": 9.980509998051e-06, + "loss": 0.0481, + "step": 19498 + }, + { + "epoch": 9.0650860065086, + "grad_norm": 4.627538681030273, + "learning_rate": 9.980508998050901e-06, + "loss": 0.0981, + "step": 19499 + }, + { + "epoch": 9.06555090655509, + "grad_norm": 9.476366996765137, + "learning_rate": 9.9805079980508e-06, + "loss": 0.1862, + "step": 19500 + }, + { + "epoch": 9.06601580660158, + "grad_norm": 4.627943515777588, + "learning_rate": 9.9805069980507e-06, + "loss": 0.0947, + "step": 19501 + }, + { + "epoch": 9.06648070664807, + "grad_norm": 3.405298948287964, + "learning_rate": 9.9805059980506e-06, + "loss": 0.0445, + "step": 19502 + }, + { + "epoch": 9.06694560669456, + "grad_norm": 4.380162239074707, + "learning_rate": 9.9805049980505e-06, + "loss": 0.0636, + "step": 19503 + }, + { + "epoch": 9.067410506741052, + "grad_norm": 1.1476236581802368, + "learning_rate": 9.980503998050401e-06, + "loss": 0.0083, + "step": 19504 + }, + { + "epoch": 9.067875406787541, + "grad_norm": 1.6772693395614624, + "learning_rate": 9.980502998050302e-06, + "loss": 0.0206, + "step": 19505 + }, + { + "epoch": 9.068340306834031, + "grad_norm": 4.158245086669922, + "learning_rate": 9.9805019980502e-06, + "loss": 0.0827, + "step": 19506 + }, + { + "epoch": 9.068805206880521, + "grad_norm": 3.1577045917510986, + "learning_rate": 9.980500998050101e-06, + "loss": 0.0476, + "step": 19507 + }, + { + "epoch": 9.06927010692701, + "grad_norm": 6.105777740478516, + "learning_rate": 9.98049999805e-06, + "loss": 0.1126, + "step": 19508 + }, + { + "epoch": 9.0697350069735, + "grad_norm": 3.229966402053833, + "learning_rate": 9.9804989980499e-06, + "loss": 0.0965, + "step": 19509 + }, + { + "epoch": 9.07019990701999, + "grad_norm": 5.0193376541137695, + "learning_rate": 9.980497998049801e-06, + "loss": 0.0821, + "step": 19510 + }, + { + "epoch": 9.07066480706648, + "grad_norm": 5.397851467132568, + "learning_rate": 9.9804969980497e-06, + "loss": 0.0348, + "step": 19511 + }, + { + "epoch": 9.07112970711297, + "grad_norm": 1.3476495742797852, + "learning_rate": 9.980495998049601e-06, + "loss": 0.0176, + "step": 19512 + }, + { + "epoch": 9.071594607159462, + "grad_norm": 5.976221561431885, + "learning_rate": 9.9804949980495e-06, + "loss": 0.0656, + "step": 19513 + }, + { + "epoch": 9.072059507205951, + "grad_norm": 4.012216091156006, + "learning_rate": 9.9804939980494e-06, + "loss": 0.0592, + "step": 19514 + }, + { + "epoch": 9.072524407252441, + "grad_norm": 5.750731468200684, + "learning_rate": 9.9804929980493e-06, + "loss": 0.1369, + "step": 19515 + }, + { + "epoch": 9.072989307298931, + "grad_norm": 2.2822988033294678, + "learning_rate": 9.9804919980492e-06, + "loss": 0.0244, + "step": 19516 + }, + { + "epoch": 9.07345420734542, + "grad_norm": 7.292328357696533, + "learning_rate": 9.980490998049101e-06, + "loss": 0.1466, + "step": 19517 + }, + { + "epoch": 9.07391910739191, + "grad_norm": 4.320943832397461, + "learning_rate": 9.980489998049002e-06, + "loss": 0.0788, + "step": 19518 + }, + { + "epoch": 9.0743840074384, + "grad_norm": 5.444861888885498, + "learning_rate": 9.9804889980489e-06, + "loss": 0.1182, + "step": 19519 + }, + { + "epoch": 9.07484890748489, + "grad_norm": 4.220223426818848, + "learning_rate": 9.9804879980488e-06, + "loss": 0.0354, + "step": 19520 + }, + { + "epoch": 9.07531380753138, + "grad_norm": 4.890497207641602, + "learning_rate": 9.9804869980487e-06, + "loss": 0.0682, + "step": 19521 + }, + { + "epoch": 9.075778707577872, + "grad_norm": 3.2399120330810547, + "learning_rate": 9.980485998048601e-06, + "loss": 0.0492, + "step": 19522 + }, + { + "epoch": 9.076243607624361, + "grad_norm": 4.477719306945801, + "learning_rate": 9.980484998048502e-06, + "loss": 0.1123, + "step": 19523 + }, + { + "epoch": 9.076708507670851, + "grad_norm": 6.2652740478515625, + "learning_rate": 9.9804839980484e-06, + "loss": 0.1005, + "step": 19524 + }, + { + "epoch": 9.077173407717341, + "grad_norm": 5.613887310028076, + "learning_rate": 9.980482998048301e-06, + "loss": 0.1, + "step": 19525 + }, + { + "epoch": 9.07763830776383, + "grad_norm": 2.4461545944213867, + "learning_rate": 9.9804819980482e-06, + "loss": 0.0249, + "step": 19526 + }, + { + "epoch": 9.07810320781032, + "grad_norm": 3.944875478744507, + "learning_rate": 9.9804809980481e-06, + "loss": 0.0706, + "step": 19527 + }, + { + "epoch": 9.07856810785681, + "grad_norm": 3.299961566925049, + "learning_rate": 9.980479998048e-06, + "loss": 0.0552, + "step": 19528 + }, + { + "epoch": 9.0790330079033, + "grad_norm": 5.776665210723877, + "learning_rate": 9.9804789980479e-06, + "loss": 0.1359, + "step": 19529 + }, + { + "epoch": 9.07949790794979, + "grad_norm": 3.941065788269043, + "learning_rate": 9.980477998047801e-06, + "loss": 0.0516, + "step": 19530 + }, + { + "epoch": 9.079962807996282, + "grad_norm": 5.081822395324707, + "learning_rate": 9.9804769980477e-06, + "loss": 0.1024, + "step": 19531 + }, + { + "epoch": 9.080427708042771, + "grad_norm": 4.307568550109863, + "learning_rate": 9.9804759980476e-06, + "loss": 0.0451, + "step": 19532 + }, + { + "epoch": 9.080892608089261, + "grad_norm": 4.882907390594482, + "learning_rate": 9.9804749980475e-06, + "loss": 0.0558, + "step": 19533 + }, + { + "epoch": 9.081357508135751, + "grad_norm": 1.6471034288406372, + "learning_rate": 9.9804739980474e-06, + "loss": 0.0096, + "step": 19534 + }, + { + "epoch": 9.08182240818224, + "grad_norm": 7.200564861297607, + "learning_rate": 9.980472998047301e-06, + "loss": 0.0704, + "step": 19535 + }, + { + "epoch": 9.08228730822873, + "grad_norm": 5.8668975830078125, + "learning_rate": 9.980471998047202e-06, + "loss": 0.1101, + "step": 19536 + }, + { + "epoch": 9.08275220827522, + "grad_norm": 3.604886770248413, + "learning_rate": 9.9804709980471e-06, + "loss": 0.0766, + "step": 19537 + }, + { + "epoch": 9.08321710832171, + "grad_norm": 5.870755672454834, + "learning_rate": 9.980469998047001e-06, + "loss": 0.0997, + "step": 19538 + }, + { + "epoch": 9.0836820083682, + "grad_norm": 3.9147603511810303, + "learning_rate": 9.9804689980469e-06, + "loss": 0.0362, + "step": 19539 + }, + { + "epoch": 9.084146908414692, + "grad_norm": 5.729898452758789, + "learning_rate": 9.9804679980468e-06, + "loss": 0.0988, + "step": 19540 + }, + { + "epoch": 9.084611808461181, + "grad_norm": 6.9881205558776855, + "learning_rate": 9.9804669980467e-06, + "loss": 0.1226, + "step": 19541 + }, + { + "epoch": 9.085076708507671, + "grad_norm": 1.9587830305099487, + "learning_rate": 9.9804659980466e-06, + "loss": 0.0189, + "step": 19542 + }, + { + "epoch": 9.085541608554161, + "grad_norm": 3.6567471027374268, + "learning_rate": 9.980464998046501e-06, + "loss": 0.065, + "step": 19543 + }, + { + "epoch": 9.08600650860065, + "grad_norm": 4.147896766662598, + "learning_rate": 9.9804639980464e-06, + "loss": 0.1246, + "step": 19544 + }, + { + "epoch": 9.08647140864714, + "grad_norm": 4.948250770568848, + "learning_rate": 9.9804629980463e-06, + "loss": 0.1269, + "step": 19545 + }, + { + "epoch": 9.08693630869363, + "grad_norm": 2.307079553604126, + "learning_rate": 9.9804619980462e-06, + "loss": 0.0265, + "step": 19546 + }, + { + "epoch": 9.08740120874012, + "grad_norm": 1.983894944190979, + "learning_rate": 9.9804609980461e-06, + "loss": 0.0107, + "step": 19547 + }, + { + "epoch": 9.087866108786612, + "grad_norm": 5.324400901794434, + "learning_rate": 9.980459998046001e-06, + "loss": 0.1305, + "step": 19548 + }, + { + "epoch": 9.088331008833102, + "grad_norm": 1.7434529066085815, + "learning_rate": 9.980458998045902e-06, + "loss": 0.0095, + "step": 19549 + }, + { + "epoch": 9.088795908879591, + "grad_norm": 6.759410381317139, + "learning_rate": 9.9804579980458e-06, + "loss": 0.1528, + "step": 19550 + }, + { + "epoch": 9.089260808926081, + "grad_norm": 3.4307138919830322, + "learning_rate": 9.980456998045701e-06, + "loss": 0.0239, + "step": 19551 + }, + { + "epoch": 9.089725708972571, + "grad_norm": 11.18510913848877, + "learning_rate": 9.9804559980456e-06, + "loss": 0.0805, + "step": 19552 + }, + { + "epoch": 9.09019060901906, + "grad_norm": 5.935883522033691, + "learning_rate": 9.980454998045501e-06, + "loss": 0.1783, + "step": 19553 + }, + { + "epoch": 9.09065550906555, + "grad_norm": 1.091504693031311, + "learning_rate": 9.9804539980454e-06, + "loss": 0.0107, + "step": 19554 + }, + { + "epoch": 9.09112040911204, + "grad_norm": 9.361979484558105, + "learning_rate": 9.9804529980453e-06, + "loss": 0.2146, + "step": 19555 + }, + { + "epoch": 9.09158530915853, + "grad_norm": 5.379507064819336, + "learning_rate": 9.980451998045201e-06, + "loss": 0.1209, + "step": 19556 + }, + { + "epoch": 9.092050209205022, + "grad_norm": 6.158921718597412, + "learning_rate": 9.9804509980451e-06, + "loss": 0.063, + "step": 19557 + }, + { + "epoch": 9.092515109251512, + "grad_norm": 5.732098579406738, + "learning_rate": 9.980449998045e-06, + "loss": 0.0662, + "step": 19558 + }, + { + "epoch": 9.092980009298001, + "grad_norm": 9.326252937316895, + "learning_rate": 9.9804489980449e-06, + "loss": 0.1109, + "step": 19559 + }, + { + "epoch": 9.093444909344491, + "grad_norm": 6.924086093902588, + "learning_rate": 9.9804479980448e-06, + "loss": 0.2374, + "step": 19560 + }, + { + "epoch": 9.093909809390981, + "grad_norm": 4.168252944946289, + "learning_rate": 9.980446998044701e-06, + "loss": 0.0781, + "step": 19561 + }, + { + "epoch": 9.09437470943747, + "grad_norm": 4.643226146697998, + "learning_rate": 9.980445998044602e-06, + "loss": 0.0562, + "step": 19562 + }, + { + "epoch": 9.09483960948396, + "grad_norm": 5.3811354637146, + "learning_rate": 9.9804449980445e-06, + "loss": 0.1036, + "step": 19563 + }, + { + "epoch": 9.09530450953045, + "grad_norm": 4.401469707489014, + "learning_rate": 9.9804439980444e-06, + "loss": 0.0796, + "step": 19564 + }, + { + "epoch": 9.09576940957694, + "grad_norm": 4.178922176361084, + "learning_rate": 9.9804429980443e-06, + "loss": 0.0568, + "step": 19565 + }, + { + "epoch": 9.096234309623432, + "grad_norm": 7.010619640350342, + "learning_rate": 9.980441998044201e-06, + "loss": 0.1273, + "step": 19566 + }, + { + "epoch": 9.096699209669922, + "grad_norm": 3.3470990657806396, + "learning_rate": 9.980440998044102e-06, + "loss": 0.034, + "step": 19567 + }, + { + "epoch": 9.097164109716411, + "grad_norm": 3.445756196975708, + "learning_rate": 9.980439998044e-06, + "loss": 0.045, + "step": 19568 + }, + { + "epoch": 9.097629009762901, + "grad_norm": 8.006006240844727, + "learning_rate": 9.980438998043901e-06, + "loss": 0.3238, + "step": 19569 + }, + { + "epoch": 9.098093909809391, + "grad_norm": 5.54623556137085, + "learning_rate": 9.9804379980438e-06, + "loss": 0.1398, + "step": 19570 + }, + { + "epoch": 9.09855880985588, + "grad_norm": 5.808078289031982, + "learning_rate": 9.980436998043701e-06, + "loss": 0.1284, + "step": 19571 + }, + { + "epoch": 9.09902370990237, + "grad_norm": 6.230276107788086, + "learning_rate": 9.9804359980436e-06, + "loss": 0.1212, + "step": 19572 + }, + { + "epoch": 9.09948860994886, + "grad_norm": 4.614293098449707, + "learning_rate": 9.9804349980435e-06, + "loss": 0.101, + "step": 19573 + }, + { + "epoch": 9.09995350999535, + "grad_norm": 4.9729533195495605, + "learning_rate": 9.980433998043401e-06, + "loss": 0.0941, + "step": 19574 + }, + { + "epoch": 9.100418410041842, + "grad_norm": 2.311836004257202, + "learning_rate": 9.9804329980433e-06, + "loss": 0.0312, + "step": 19575 + }, + { + "epoch": 9.100883310088332, + "grad_norm": 7.461214542388916, + "learning_rate": 9.9804319980432e-06, + "loss": 0.1412, + "step": 19576 + }, + { + "epoch": 9.101348210134821, + "grad_norm": 0.5810695886611938, + "learning_rate": 9.9804309980431e-06, + "loss": 0.0046, + "step": 19577 + }, + { + "epoch": 9.101813110181311, + "grad_norm": 2.3248064517974854, + "learning_rate": 9.980429998043e-06, + "loss": 0.0309, + "step": 19578 + }, + { + "epoch": 9.102278010227801, + "grad_norm": 7.290635585784912, + "learning_rate": 9.980428998042901e-06, + "loss": 0.1723, + "step": 19579 + }, + { + "epoch": 9.10274291027429, + "grad_norm": 4.385679244995117, + "learning_rate": 9.980427998042802e-06, + "loss": 0.1127, + "step": 19580 + }, + { + "epoch": 9.10320781032078, + "grad_norm": 3.393531322479248, + "learning_rate": 9.9804269980427e-06, + "loss": 0.0587, + "step": 19581 + }, + { + "epoch": 9.10367271036727, + "grad_norm": 4.859616756439209, + "learning_rate": 9.980425998042601e-06, + "loss": 0.1405, + "step": 19582 + }, + { + "epoch": 9.104137610413762, + "grad_norm": 6.157377243041992, + "learning_rate": 9.9804249980425e-06, + "loss": 0.1229, + "step": 19583 + }, + { + "epoch": 9.104602510460252, + "grad_norm": 6.353137493133545, + "learning_rate": 9.980423998042401e-06, + "loss": 0.0974, + "step": 19584 + }, + { + "epoch": 9.105067410506742, + "grad_norm": 4.529702186584473, + "learning_rate": 9.9804229980423e-06, + "loss": 0.0718, + "step": 19585 + }, + { + "epoch": 9.105532310553231, + "grad_norm": 6.979612827301025, + "learning_rate": 9.9804219980422e-06, + "loss": 0.1684, + "step": 19586 + }, + { + "epoch": 9.105997210599721, + "grad_norm": 3.6126291751861572, + "learning_rate": 9.980420998042101e-06, + "loss": 0.0521, + "step": 19587 + }, + { + "epoch": 9.106462110646211, + "grad_norm": 1.8418318033218384, + "learning_rate": 9.980419998042e-06, + "loss": 0.0247, + "step": 19588 + }, + { + "epoch": 9.1069270106927, + "grad_norm": 4.383051872253418, + "learning_rate": 9.980418998041901e-06, + "loss": 0.1283, + "step": 19589 + }, + { + "epoch": 9.10739191073919, + "grad_norm": 5.1410603523254395, + "learning_rate": 9.9804179980418e-06, + "loss": 0.0735, + "step": 19590 + }, + { + "epoch": 9.10785681078568, + "grad_norm": 2.7754626274108887, + "learning_rate": 9.9804169980417e-06, + "loss": 0.0245, + "step": 19591 + }, + { + "epoch": 9.108321710832172, + "grad_norm": 2.1590168476104736, + "learning_rate": 9.980415998041601e-06, + "loss": 0.0258, + "step": 19592 + }, + { + "epoch": 9.108786610878662, + "grad_norm": 3.9437246322631836, + "learning_rate": 9.980414998041502e-06, + "loss": 0.0524, + "step": 19593 + }, + { + "epoch": 9.109251510925152, + "grad_norm": 5.8748908042907715, + "learning_rate": 9.9804139980414e-06, + "loss": 0.1145, + "step": 19594 + }, + { + "epoch": 9.109716410971641, + "grad_norm": 4.8168463706970215, + "learning_rate": 9.980412998041301e-06, + "loss": 0.0614, + "step": 19595 + }, + { + "epoch": 9.110181311018131, + "grad_norm": 4.604533672332764, + "learning_rate": 9.9804119980412e-06, + "loss": 0.0444, + "step": 19596 + }, + { + "epoch": 9.110646211064621, + "grad_norm": 2.5744597911834717, + "learning_rate": 9.9804109980411e-06, + "loss": 0.0431, + "step": 19597 + }, + { + "epoch": 9.11111111111111, + "grad_norm": 4.92704439163208, + "learning_rate": 9.980409998041e-06, + "loss": 0.0836, + "step": 19598 + }, + { + "epoch": 9.1115760111576, + "grad_norm": 4.819174766540527, + "learning_rate": 9.9804089980409e-06, + "loss": 0.1506, + "step": 19599 + }, + { + "epoch": 9.11204091120409, + "grad_norm": 5.117305278778076, + "learning_rate": 9.980407998040801e-06, + "loss": 0.0555, + "step": 19600 + }, + { + "epoch": 9.112505811250582, + "grad_norm": 6.931251049041748, + "learning_rate": 9.9804069980407e-06, + "loss": 0.0617, + "step": 19601 + }, + { + "epoch": 9.112970711297072, + "grad_norm": 1.669329285621643, + "learning_rate": 9.980405998040601e-06, + "loss": 0.0176, + "step": 19602 + }, + { + "epoch": 9.113435611343562, + "grad_norm": 10.647485733032227, + "learning_rate": 9.9804049980405e-06, + "loss": 0.2901, + "step": 19603 + }, + { + "epoch": 9.113900511390051, + "grad_norm": 5.803891181945801, + "learning_rate": 9.9804039980404e-06, + "loss": 0.069, + "step": 19604 + }, + { + "epoch": 9.114365411436541, + "grad_norm": 2.0200226306915283, + "learning_rate": 9.980402998040301e-06, + "loss": 0.0129, + "step": 19605 + }, + { + "epoch": 9.114830311483031, + "grad_norm": 1.8656606674194336, + "learning_rate": 9.980401998040202e-06, + "loss": 0.0195, + "step": 19606 + }, + { + "epoch": 9.11529521152952, + "grad_norm": 6.276775360107422, + "learning_rate": 9.980400998040101e-06, + "loss": 0.1361, + "step": 19607 + }, + { + "epoch": 9.11576011157601, + "grad_norm": 8.383203506469727, + "learning_rate": 9.98039999804e-06, + "loss": 0.1522, + "step": 19608 + }, + { + "epoch": 9.1162250116225, + "grad_norm": 6.148672580718994, + "learning_rate": 9.9803989980399e-06, + "loss": 0.1634, + "step": 19609 + }, + { + "epoch": 9.116689911668992, + "grad_norm": 8.13753604888916, + "learning_rate": 9.9803979980398e-06, + "loss": 0.1992, + "step": 19610 + }, + { + "epoch": 9.117154811715482, + "grad_norm": 3.8916773796081543, + "learning_rate": 9.9803969980397e-06, + "loss": 0.0582, + "step": 19611 + }, + { + "epoch": 9.117619711761972, + "grad_norm": 6.221802711486816, + "learning_rate": 9.9803959980396e-06, + "loss": 0.1353, + "step": 19612 + }, + { + "epoch": 9.118084611808461, + "grad_norm": 4.362820148468018, + "learning_rate": 9.980394998039501e-06, + "loss": 0.0412, + "step": 19613 + }, + { + "epoch": 9.118549511854951, + "grad_norm": 5.708911895751953, + "learning_rate": 9.9803939980394e-06, + "loss": 0.0947, + "step": 19614 + }, + { + "epoch": 9.119014411901441, + "grad_norm": 4.836276054382324, + "learning_rate": 9.980392998039301e-06, + "loss": 0.0634, + "step": 19615 + }, + { + "epoch": 9.11947931194793, + "grad_norm": 6.212825298309326, + "learning_rate": 9.9803919980392e-06, + "loss": 0.1251, + "step": 19616 + }, + { + "epoch": 9.11994421199442, + "grad_norm": 7.710290431976318, + "learning_rate": 9.9803909980391e-06, + "loss": 0.1867, + "step": 19617 + }, + { + "epoch": 9.12040911204091, + "grad_norm": 7.6474714279174805, + "learning_rate": 9.980389998039001e-06, + "loss": 0.1762, + "step": 19618 + }, + { + "epoch": 9.120874012087402, + "grad_norm": 6.101356029510498, + "learning_rate": 9.9803889980389e-06, + "loss": 0.0393, + "step": 19619 + }, + { + "epoch": 9.121338912133892, + "grad_norm": 3.6634747982025146, + "learning_rate": 9.980387998038801e-06, + "loss": 0.0523, + "step": 19620 + }, + { + "epoch": 9.121803812180381, + "grad_norm": 4.286801338195801, + "learning_rate": 9.9803869980387e-06, + "loss": 0.0766, + "step": 19621 + }, + { + "epoch": 9.122268712226871, + "grad_norm": 6.027868270874023, + "learning_rate": 9.9803859980386e-06, + "loss": 0.1134, + "step": 19622 + }, + { + "epoch": 9.122733612273361, + "grad_norm": 1.619787335395813, + "learning_rate": 9.9803849980385e-06, + "loss": 0.0124, + "step": 19623 + }, + { + "epoch": 9.123198512319851, + "grad_norm": 3.3032472133636475, + "learning_rate": 9.9803839980384e-06, + "loss": 0.0536, + "step": 19624 + }, + { + "epoch": 9.12366341236634, + "grad_norm": 4.8542962074279785, + "learning_rate": 9.980382998038301e-06, + "loss": 0.0594, + "step": 19625 + }, + { + "epoch": 9.12412831241283, + "grad_norm": 8.23999309539795, + "learning_rate": 9.980381998038202e-06, + "loss": 0.1399, + "step": 19626 + }, + { + "epoch": 9.124593212459322, + "grad_norm": 6.3857622146606445, + "learning_rate": 9.9803809980381e-06, + "loss": 0.1313, + "step": 19627 + }, + { + "epoch": 9.125058112505812, + "grad_norm": 6.858147621154785, + "learning_rate": 9.980379998038001e-06, + "loss": 0.1071, + "step": 19628 + }, + { + "epoch": 9.125523012552302, + "grad_norm": 2.5555624961853027, + "learning_rate": 9.9803789980379e-06, + "loss": 0.0201, + "step": 19629 + }, + { + "epoch": 9.125987912598791, + "grad_norm": 7.688173770904541, + "learning_rate": 9.9803779980378e-06, + "loss": 0.1548, + "step": 19630 + }, + { + "epoch": 9.126452812645281, + "grad_norm": 5.309413909912109, + "learning_rate": 9.980376998037701e-06, + "loss": 0.0406, + "step": 19631 + }, + { + "epoch": 9.126917712691771, + "grad_norm": 6.094500541687012, + "learning_rate": 9.9803759980376e-06, + "loss": 0.0829, + "step": 19632 + }, + { + "epoch": 9.127382612738261, + "grad_norm": 5.280954360961914, + "learning_rate": 9.980374998037501e-06, + "loss": 0.0733, + "step": 19633 + }, + { + "epoch": 9.12784751278475, + "grad_norm": 2.885283946990967, + "learning_rate": 9.9803739980374e-06, + "loss": 0.0372, + "step": 19634 + }, + { + "epoch": 9.12831241283124, + "grad_norm": 2.611208915710449, + "learning_rate": 9.9803729980373e-06, + "loss": 0.0313, + "step": 19635 + }, + { + "epoch": 9.128777312877732, + "grad_norm": 6.308938503265381, + "learning_rate": 9.980371998037201e-06, + "loss": 0.107, + "step": 19636 + }, + { + "epoch": 9.129242212924222, + "grad_norm": 4.311603546142578, + "learning_rate": 9.9803709980371e-06, + "loss": 0.0708, + "step": 19637 + }, + { + "epoch": 9.129707112970712, + "grad_norm": 1.6805620193481445, + "learning_rate": 9.980369998037001e-06, + "loss": 0.0161, + "step": 19638 + }, + { + "epoch": 9.130172013017201, + "grad_norm": 3.571849822998047, + "learning_rate": 9.9803689980369e-06, + "loss": 0.0502, + "step": 19639 + }, + { + "epoch": 9.130636913063691, + "grad_norm": 5.020267009735107, + "learning_rate": 9.9803679980368e-06, + "loss": 0.1521, + "step": 19640 + }, + { + "epoch": 9.131101813110181, + "grad_norm": 4.369233131408691, + "learning_rate": 9.9803669980367e-06, + "loss": 0.0487, + "step": 19641 + }, + { + "epoch": 9.131566713156671, + "grad_norm": 6.755733013153076, + "learning_rate": 9.9803659980366e-06, + "loss": 0.0559, + "step": 19642 + }, + { + "epoch": 9.13203161320316, + "grad_norm": 15.690688133239746, + "learning_rate": 9.9803649980365e-06, + "loss": 0.1272, + "step": 19643 + }, + { + "epoch": 9.13249651324965, + "grad_norm": 8.921313285827637, + "learning_rate": 9.980363998036401e-06, + "loss": 0.1724, + "step": 19644 + }, + { + "epoch": 9.132961413296142, + "grad_norm": 1.9023771286010742, + "learning_rate": 9.9803629980363e-06, + "loss": 0.0138, + "step": 19645 + }, + { + "epoch": 9.133426313342632, + "grad_norm": 3.2848451137542725, + "learning_rate": 9.980361998036201e-06, + "loss": 0.0247, + "step": 19646 + }, + { + "epoch": 9.133891213389122, + "grad_norm": 4.753841876983643, + "learning_rate": 9.9803609980361e-06, + "loss": 0.1109, + "step": 19647 + }, + { + "epoch": 9.134356113435611, + "grad_norm": 5.71034574508667, + "learning_rate": 9.980359998036e-06, + "loss": 0.1049, + "step": 19648 + }, + { + "epoch": 9.134821013482101, + "grad_norm": 7.21094274520874, + "learning_rate": 9.980358998035901e-06, + "loss": 0.1044, + "step": 19649 + }, + { + "epoch": 9.135285913528591, + "grad_norm": 3.359149694442749, + "learning_rate": 9.9803579980358e-06, + "loss": 0.0222, + "step": 19650 + }, + { + "epoch": 9.135750813575081, + "grad_norm": 3.9786975383758545, + "learning_rate": 9.980356998035701e-06, + "loss": 0.034, + "step": 19651 + }, + { + "epoch": 9.13621571362157, + "grad_norm": 4.982693195343018, + "learning_rate": 9.9803559980356e-06, + "loss": 0.1001, + "step": 19652 + }, + { + "epoch": 9.13668061366806, + "grad_norm": 3.9645535945892334, + "learning_rate": 9.9803549980355e-06, + "loss": 0.1067, + "step": 19653 + }, + { + "epoch": 9.137145513714552, + "grad_norm": 5.552795886993408, + "learning_rate": 9.9803539980354e-06, + "loss": 0.1359, + "step": 19654 + }, + { + "epoch": 9.137610413761042, + "grad_norm": 7.436444282531738, + "learning_rate": 9.9803529980353e-06, + "loss": 0.1206, + "step": 19655 + }, + { + "epoch": 9.138075313807532, + "grad_norm": 5.549548149108887, + "learning_rate": 9.980351998035201e-06, + "loss": 0.1175, + "step": 19656 + }, + { + "epoch": 9.138540213854021, + "grad_norm": 6.598421096801758, + "learning_rate": 9.980350998035102e-06, + "loss": 0.1407, + "step": 19657 + }, + { + "epoch": 9.139005113900511, + "grad_norm": 0.7337662577629089, + "learning_rate": 9.980349998035e-06, + "loss": 0.0041, + "step": 19658 + }, + { + "epoch": 9.139470013947001, + "grad_norm": 3.6434526443481445, + "learning_rate": 9.980348998034901e-06, + "loss": 0.0576, + "step": 19659 + }, + { + "epoch": 9.139934913993491, + "grad_norm": 4.067992687225342, + "learning_rate": 9.9803479980348e-06, + "loss": 0.0701, + "step": 19660 + }, + { + "epoch": 9.14039981403998, + "grad_norm": 4.521480083465576, + "learning_rate": 9.9803469980347e-06, + "loss": 0.066, + "step": 19661 + }, + { + "epoch": 9.140864714086472, + "grad_norm": 4.184060096740723, + "learning_rate": 9.980345998034601e-06, + "loss": 0.1132, + "step": 19662 + }, + { + "epoch": 9.141329614132962, + "grad_norm": 4.758752346038818, + "learning_rate": 9.9803449980345e-06, + "loss": 0.0635, + "step": 19663 + }, + { + "epoch": 9.141794514179452, + "grad_norm": 2.166752815246582, + "learning_rate": 9.980343998034401e-06, + "loss": 0.0277, + "step": 19664 + }, + { + "epoch": 9.142259414225942, + "grad_norm": 3.195472240447998, + "learning_rate": 9.9803429980343e-06, + "loss": 0.0417, + "step": 19665 + }, + { + "epoch": 9.142724314272431, + "grad_norm": 4.0213541984558105, + "learning_rate": 9.9803419980342e-06, + "loss": 0.0861, + "step": 19666 + }, + { + "epoch": 9.143189214318921, + "grad_norm": 1.7412705421447754, + "learning_rate": 9.9803409980341e-06, + "loss": 0.0165, + "step": 19667 + }, + { + "epoch": 9.143654114365411, + "grad_norm": 4.968581199645996, + "learning_rate": 9.980339998034e-06, + "loss": 0.1306, + "step": 19668 + }, + { + "epoch": 9.144119014411901, + "grad_norm": 6.407374858856201, + "learning_rate": 9.980338998033901e-06, + "loss": 0.0846, + "step": 19669 + }, + { + "epoch": 9.14458391445839, + "grad_norm": 3.9476091861724854, + "learning_rate": 9.980337998033802e-06, + "loss": 0.0494, + "step": 19670 + }, + { + "epoch": 9.145048814504882, + "grad_norm": 3.917557716369629, + "learning_rate": 9.9803369980337e-06, + "loss": 0.0555, + "step": 19671 + }, + { + "epoch": 9.145513714551372, + "grad_norm": 4.904943943023682, + "learning_rate": 9.9803359980336e-06, + "loss": 0.0568, + "step": 19672 + }, + { + "epoch": 9.145978614597862, + "grad_norm": 3.6198432445526123, + "learning_rate": 9.9803349980335e-06, + "loss": 0.0564, + "step": 19673 + }, + { + "epoch": 9.146443514644352, + "grad_norm": 2.6160078048706055, + "learning_rate": 9.980333998033401e-06, + "loss": 0.0292, + "step": 19674 + }, + { + "epoch": 9.146908414690841, + "grad_norm": 5.004144668579102, + "learning_rate": 9.980332998033302e-06, + "loss": 0.0819, + "step": 19675 + }, + { + "epoch": 9.147373314737331, + "grad_norm": 4.330502033233643, + "learning_rate": 9.9803319980332e-06, + "loss": 0.056, + "step": 19676 + }, + { + "epoch": 9.147838214783821, + "grad_norm": 9.113480567932129, + "learning_rate": 9.980330998033101e-06, + "loss": 0.0615, + "step": 19677 + }, + { + "epoch": 9.148303114830311, + "grad_norm": 4.975891590118408, + "learning_rate": 9.980329998033e-06, + "loss": 0.0937, + "step": 19678 + }, + { + "epoch": 9.1487680148768, + "grad_norm": 2.4473395347595215, + "learning_rate": 9.9803289980329e-06, + "loss": 0.0373, + "step": 19679 + }, + { + "epoch": 9.149232914923292, + "grad_norm": 5.625922203063965, + "learning_rate": 9.9803279980328e-06, + "loss": 0.1119, + "step": 19680 + }, + { + "epoch": 9.149697814969782, + "grad_norm": 4.276725769042969, + "learning_rate": 9.9803269980327e-06, + "loss": 0.0423, + "step": 19681 + }, + { + "epoch": 9.150162715016272, + "grad_norm": 5.720322132110596, + "learning_rate": 9.980325998032601e-06, + "loss": 0.0972, + "step": 19682 + }, + { + "epoch": 9.150627615062762, + "grad_norm": 4.3732194900512695, + "learning_rate": 9.9803249980325e-06, + "loss": 0.0669, + "step": 19683 + }, + { + "epoch": 9.151092515109251, + "grad_norm": 2.5960581302642822, + "learning_rate": 9.9803239980324e-06, + "loss": 0.0238, + "step": 19684 + }, + { + "epoch": 9.151557415155741, + "grad_norm": 6.006083965301514, + "learning_rate": 9.9803229980323e-06, + "loss": 0.1838, + "step": 19685 + }, + { + "epoch": 9.152022315202231, + "grad_norm": 2.689563512802124, + "learning_rate": 9.9803219980322e-06, + "loss": 0.0215, + "step": 19686 + }, + { + "epoch": 9.15248721524872, + "grad_norm": 2.7281646728515625, + "learning_rate": 9.980320998032101e-06, + "loss": 0.0179, + "step": 19687 + }, + { + "epoch": 9.15295211529521, + "grad_norm": 5.839321613311768, + "learning_rate": 9.980319998032002e-06, + "loss": 0.0868, + "step": 19688 + }, + { + "epoch": 9.153417015341702, + "grad_norm": 5.786305904388428, + "learning_rate": 9.9803189980319e-06, + "loss": 0.1248, + "step": 19689 + }, + { + "epoch": 9.153881915388192, + "grad_norm": 6.391387939453125, + "learning_rate": 9.980317998031801e-06, + "loss": 0.1225, + "step": 19690 + }, + { + "epoch": 9.154346815434682, + "grad_norm": 5.352688789367676, + "learning_rate": 9.9803169980317e-06, + "loss": 0.0674, + "step": 19691 + }, + { + "epoch": 9.154811715481172, + "grad_norm": 3.63661789894104, + "learning_rate": 9.980315998031601e-06, + "loss": 0.0386, + "step": 19692 + }, + { + "epoch": 9.155276615527661, + "grad_norm": 7.193209171295166, + "learning_rate": 9.9803149980315e-06, + "loss": 0.1622, + "step": 19693 + }, + { + "epoch": 9.155741515574151, + "grad_norm": 4.983187675476074, + "learning_rate": 9.9803139980314e-06, + "loss": 0.1541, + "step": 19694 + }, + { + "epoch": 9.156206415620641, + "grad_norm": 2.780731439590454, + "learning_rate": 9.980312998031301e-06, + "loss": 0.0204, + "step": 19695 + }, + { + "epoch": 9.15667131566713, + "grad_norm": 6.2913408279418945, + "learning_rate": 9.9803119980312e-06, + "loss": 0.0865, + "step": 19696 + }, + { + "epoch": 9.15713621571362, + "grad_norm": 4.358664035797119, + "learning_rate": 9.9803109980311e-06, + "loss": 0.0776, + "step": 19697 + }, + { + "epoch": 9.157601115760112, + "grad_norm": 4.878098487854004, + "learning_rate": 9.980309998031e-06, + "loss": 0.0539, + "step": 19698 + }, + { + "epoch": 9.158066015806602, + "grad_norm": 3.6754748821258545, + "learning_rate": 9.9803089980309e-06, + "loss": 0.0574, + "step": 19699 + }, + { + "epoch": 9.158530915853092, + "grad_norm": 5.114940643310547, + "learning_rate": 9.980307998030801e-06, + "loss": 0.0756, + "step": 19700 + }, + { + "epoch": 9.158995815899582, + "grad_norm": 4.043065071105957, + "learning_rate": 9.980306998030702e-06, + "loss": 0.0852, + "step": 19701 + }, + { + "epoch": 9.159460715946071, + "grad_norm": 8.017292022705078, + "learning_rate": 9.9803059980306e-06, + "loss": 0.0325, + "step": 19702 + }, + { + "epoch": 9.159925615992561, + "grad_norm": 3.0356228351593018, + "learning_rate": 9.980304998030501e-06, + "loss": 0.0264, + "step": 19703 + }, + { + "epoch": 9.160390516039051, + "grad_norm": 2.484598398208618, + "learning_rate": 9.9803039980304e-06, + "loss": 0.0277, + "step": 19704 + }, + { + "epoch": 9.16085541608554, + "grad_norm": 4.539327144622803, + "learning_rate": 9.980302998030301e-06, + "loss": 0.0797, + "step": 19705 + }, + { + "epoch": 9.161320316132032, + "grad_norm": 4.547685146331787, + "learning_rate": 9.9803019980302e-06, + "loss": 0.0683, + "step": 19706 + }, + { + "epoch": 9.161785216178522, + "grad_norm": 2.2318618297576904, + "learning_rate": 9.9803009980301e-06, + "loss": 0.0195, + "step": 19707 + }, + { + "epoch": 9.162250116225012, + "grad_norm": 9.036251068115234, + "learning_rate": 9.980299998030001e-06, + "loss": 0.2903, + "step": 19708 + }, + { + "epoch": 9.162715016271502, + "grad_norm": 10.958828926086426, + "learning_rate": 9.9802989980299e-06, + "loss": 0.28, + "step": 19709 + }, + { + "epoch": 9.163179916317992, + "grad_norm": 2.5610456466674805, + "learning_rate": 9.980297998029801e-06, + "loss": 0.0219, + "step": 19710 + }, + { + "epoch": 9.163644816364481, + "grad_norm": 4.64229154586792, + "learning_rate": 9.9802969980297e-06, + "loss": 0.053, + "step": 19711 + }, + { + "epoch": 9.164109716410971, + "grad_norm": 5.651462078094482, + "learning_rate": 9.9802959980296e-06, + "loss": 0.0883, + "step": 19712 + }, + { + "epoch": 9.164574616457461, + "grad_norm": 3.9684653282165527, + "learning_rate": 9.980294998029501e-06, + "loss": 0.0614, + "step": 19713 + }, + { + "epoch": 9.16503951650395, + "grad_norm": 5.621563911437988, + "learning_rate": 9.980293998029402e-06, + "loss": 0.0954, + "step": 19714 + }, + { + "epoch": 9.165504416550442, + "grad_norm": 3.2461345195770264, + "learning_rate": 9.9802929980293e-06, + "loss": 0.0311, + "step": 19715 + }, + { + "epoch": 9.165969316596932, + "grad_norm": 4.679722309112549, + "learning_rate": 9.9802919980292e-06, + "loss": 0.0685, + "step": 19716 + }, + { + "epoch": 9.166434216643422, + "grad_norm": 4.257512092590332, + "learning_rate": 9.9802909980291e-06, + "loss": 0.0692, + "step": 19717 + }, + { + "epoch": 9.166899116689912, + "grad_norm": 10.694561004638672, + "learning_rate": 9.980289998029001e-06, + "loss": 0.2282, + "step": 19718 + }, + { + "epoch": 9.167364016736402, + "grad_norm": 4.536645889282227, + "learning_rate": 9.9802889980289e-06, + "loss": 0.0967, + "step": 19719 + }, + { + "epoch": 9.167828916782891, + "grad_norm": 4.2319488525390625, + "learning_rate": 9.9802879980288e-06, + "loss": 0.0667, + "step": 19720 + }, + { + "epoch": 9.168293816829381, + "grad_norm": 4.885654449462891, + "learning_rate": 9.980286998028701e-06, + "loss": 0.1114, + "step": 19721 + }, + { + "epoch": 9.168758716875871, + "grad_norm": 3.947205066680908, + "learning_rate": 9.9802859980286e-06, + "loss": 0.0711, + "step": 19722 + }, + { + "epoch": 9.16922361692236, + "grad_norm": 4.467769622802734, + "learning_rate": 9.980284998028501e-06, + "loss": 0.0947, + "step": 19723 + }, + { + "epoch": 9.169688516968852, + "grad_norm": 2.5561845302581787, + "learning_rate": 9.9802839980284e-06, + "loss": 0.0215, + "step": 19724 + }, + { + "epoch": 9.170153417015342, + "grad_norm": 3.1278977394104004, + "learning_rate": 9.9802829980283e-06, + "loss": 0.0625, + "step": 19725 + }, + { + "epoch": 9.170618317061832, + "grad_norm": 7.404210567474365, + "learning_rate": 9.980281998028201e-06, + "loss": 0.1837, + "step": 19726 + }, + { + "epoch": 9.171083217108322, + "grad_norm": 5.778981685638428, + "learning_rate": 9.9802809980281e-06, + "loss": 0.1006, + "step": 19727 + }, + { + "epoch": 9.171548117154812, + "grad_norm": 4.6719441413879395, + "learning_rate": 9.980279998028e-06, + "loss": 0.0925, + "step": 19728 + }, + { + "epoch": 9.172013017201301, + "grad_norm": 7.664889812469482, + "learning_rate": 9.9802789980279e-06, + "loss": 0.1762, + "step": 19729 + }, + { + "epoch": 9.172477917247791, + "grad_norm": 7.569040775299072, + "learning_rate": 9.9802779980278e-06, + "loss": 0.1509, + "step": 19730 + }, + { + "epoch": 9.172942817294281, + "grad_norm": 1.3333444595336914, + "learning_rate": 9.980276998027701e-06, + "loss": 0.0126, + "step": 19731 + }, + { + "epoch": 9.17340771734077, + "grad_norm": 2.976560354232788, + "learning_rate": 9.980275998027602e-06, + "loss": 0.0274, + "step": 19732 + }, + { + "epoch": 9.173872617387262, + "grad_norm": 5.33558988571167, + "learning_rate": 9.9802749980275e-06, + "loss": 0.0983, + "step": 19733 + }, + { + "epoch": 9.174337517433752, + "grad_norm": 6.484309673309326, + "learning_rate": 9.980273998027401e-06, + "loss": 0.1636, + "step": 19734 + }, + { + "epoch": 9.174802417480242, + "grad_norm": 4.2361860275268555, + "learning_rate": 9.9802729980273e-06, + "loss": 0.0723, + "step": 19735 + }, + { + "epoch": 9.175267317526732, + "grad_norm": 2.949084758758545, + "learning_rate": 9.980271998027201e-06, + "loss": 0.0514, + "step": 19736 + }, + { + "epoch": 9.175732217573222, + "grad_norm": 7.022644996643066, + "learning_rate": 9.9802709980271e-06, + "loss": 0.1923, + "step": 19737 + }, + { + "epoch": 9.176197117619711, + "grad_norm": 2.9701426029205322, + "learning_rate": 9.980269998027e-06, + "loss": 0.049, + "step": 19738 + }, + { + "epoch": 9.176662017666201, + "grad_norm": 4.730632305145264, + "learning_rate": 9.980268998026901e-06, + "loss": 0.0758, + "step": 19739 + }, + { + "epoch": 9.177126917712691, + "grad_norm": 3.7619616985321045, + "learning_rate": 9.9802679980268e-06, + "loss": 0.0466, + "step": 19740 + }, + { + "epoch": 9.177591817759183, + "grad_norm": 3.201040029525757, + "learning_rate": 9.980266998026701e-06, + "loss": 0.047, + "step": 19741 + }, + { + "epoch": 9.178056717805672, + "grad_norm": 2.4528090953826904, + "learning_rate": 9.9802659980266e-06, + "loss": 0.0387, + "step": 19742 + }, + { + "epoch": 9.178521617852162, + "grad_norm": 5.6172051429748535, + "learning_rate": 9.9802649980265e-06, + "loss": 0.0558, + "step": 19743 + }, + { + "epoch": 9.178986517898652, + "grad_norm": 4.094752788543701, + "learning_rate": 9.980263998026401e-06, + "loss": 0.0344, + "step": 19744 + }, + { + "epoch": 9.179451417945142, + "grad_norm": 5.916399955749512, + "learning_rate": 9.980262998026302e-06, + "loss": 0.1104, + "step": 19745 + }, + { + "epoch": 9.179916317991632, + "grad_norm": 6.310266017913818, + "learning_rate": 9.9802619980262e-06, + "loss": 0.1541, + "step": 19746 + }, + { + "epoch": 9.180381218038121, + "grad_norm": 6.845948696136475, + "learning_rate": 9.980260998026101e-06, + "loss": 0.1072, + "step": 19747 + }, + { + "epoch": 9.180846118084611, + "grad_norm": 5.9742937088012695, + "learning_rate": 9.980259998026e-06, + "loss": 0.1048, + "step": 19748 + }, + { + "epoch": 9.181311018131101, + "grad_norm": 9.29474925994873, + "learning_rate": 9.9802589980259e-06, + "loss": 0.1862, + "step": 19749 + }, + { + "epoch": 9.181775918177593, + "grad_norm": 7.204781532287598, + "learning_rate": 9.9802579980258e-06, + "loss": 0.1476, + "step": 19750 + }, + { + "epoch": 9.182240818224082, + "grad_norm": 4.971271514892578, + "learning_rate": 9.9802569980257e-06, + "loss": 0.0877, + "step": 19751 + }, + { + "epoch": 9.182705718270572, + "grad_norm": 2.30549955368042, + "learning_rate": 9.980255998025601e-06, + "loss": 0.0257, + "step": 19752 + }, + { + "epoch": 9.183170618317062, + "grad_norm": 2.6852753162384033, + "learning_rate": 9.9802549980255e-06, + "loss": 0.0445, + "step": 19753 + }, + { + "epoch": 9.183635518363552, + "grad_norm": 2.56319522857666, + "learning_rate": 9.980253998025401e-06, + "loss": 0.0315, + "step": 19754 + }, + { + "epoch": 9.184100418410042, + "grad_norm": 1.0372469425201416, + "learning_rate": 9.9802529980253e-06, + "loss": 0.0127, + "step": 19755 + }, + { + "epoch": 9.184565318456531, + "grad_norm": 1.2844319343566895, + "learning_rate": 9.9802519980252e-06, + "loss": 0.0181, + "step": 19756 + }, + { + "epoch": 9.185030218503021, + "grad_norm": 4.964871406555176, + "learning_rate": 9.980250998025101e-06, + "loss": 0.0542, + "step": 19757 + }, + { + "epoch": 9.185495118549511, + "grad_norm": 4.5952348709106445, + "learning_rate": 9.980249998025002e-06, + "loss": 0.0931, + "step": 19758 + }, + { + "epoch": 9.185960018596003, + "grad_norm": 5.039710521697998, + "learning_rate": 9.980248998024901e-06, + "loss": 0.0866, + "step": 19759 + }, + { + "epoch": 9.186424918642492, + "grad_norm": 6.202964782714844, + "learning_rate": 9.9802479980248e-06, + "loss": 0.0853, + "step": 19760 + }, + { + "epoch": 9.186889818688982, + "grad_norm": 7.402314186096191, + "learning_rate": 9.9802469980247e-06, + "loss": 0.1359, + "step": 19761 + }, + { + "epoch": 9.187354718735472, + "grad_norm": 5.764531135559082, + "learning_rate": 9.9802459980246e-06, + "loss": 0.073, + "step": 19762 + }, + { + "epoch": 9.187819618781962, + "grad_norm": 5.200450897216797, + "learning_rate": 9.9802449980245e-06, + "loss": 0.0615, + "step": 19763 + }, + { + "epoch": 9.188284518828452, + "grad_norm": 6.3146796226501465, + "learning_rate": 9.9802439980244e-06, + "loss": 0.177, + "step": 19764 + }, + { + "epoch": 9.188749418874941, + "grad_norm": 2.8627371788024902, + "learning_rate": 9.980242998024301e-06, + "loss": 0.0335, + "step": 19765 + }, + { + "epoch": 9.189214318921431, + "grad_norm": 3.2180256843566895, + "learning_rate": 9.9802419980242e-06, + "loss": 0.0619, + "step": 19766 + }, + { + "epoch": 9.189679218967921, + "grad_norm": 2.0278689861297607, + "learning_rate": 9.980240998024101e-06, + "loss": 0.0224, + "step": 19767 + }, + { + "epoch": 9.190144119014413, + "grad_norm": 1.1001856327056885, + "learning_rate": 9.980239998024e-06, + "loss": 0.0109, + "step": 19768 + }, + { + "epoch": 9.190609019060902, + "grad_norm": 2.665989398956299, + "learning_rate": 9.9802389980239e-06, + "loss": 0.0766, + "step": 19769 + }, + { + "epoch": 9.191073919107392, + "grad_norm": 5.532627582550049, + "learning_rate": 9.980237998023801e-06, + "loss": 0.0962, + "step": 19770 + }, + { + "epoch": 9.191538819153882, + "grad_norm": 8.467866897583008, + "learning_rate": 9.9802369980237e-06, + "loss": 0.116, + "step": 19771 + }, + { + "epoch": 9.192003719200372, + "grad_norm": 7.609797477722168, + "learning_rate": 9.980235998023601e-06, + "loss": 0.1183, + "step": 19772 + }, + { + "epoch": 9.192468619246862, + "grad_norm": 2.8545007705688477, + "learning_rate": 9.9802349980235e-06, + "loss": 0.0395, + "step": 19773 + }, + { + "epoch": 9.192933519293351, + "grad_norm": 1.937428593635559, + "learning_rate": 9.9802339980234e-06, + "loss": 0.0172, + "step": 19774 + }, + { + "epoch": 9.193398419339841, + "grad_norm": 4.928020477294922, + "learning_rate": 9.9802329980233e-06, + "loss": 0.0593, + "step": 19775 + }, + { + "epoch": 9.193863319386331, + "grad_norm": 6.0133442878723145, + "learning_rate": 9.9802319980232e-06, + "loss": 0.1188, + "step": 19776 + }, + { + "epoch": 9.194328219432823, + "grad_norm": 7.741015434265137, + "learning_rate": 9.980230998023101e-06, + "loss": 0.1211, + "step": 19777 + }, + { + "epoch": 9.194793119479312, + "grad_norm": 5.235556602478027, + "learning_rate": 9.980229998023002e-06, + "loss": 0.0872, + "step": 19778 + }, + { + "epoch": 9.195258019525802, + "grad_norm": 4.678212642669678, + "learning_rate": 9.9802289980229e-06, + "loss": 0.0927, + "step": 19779 + }, + { + "epoch": 9.195722919572292, + "grad_norm": 7.657685279846191, + "learning_rate": 9.9802279980228e-06, + "loss": 0.1837, + "step": 19780 + }, + { + "epoch": 9.196187819618782, + "grad_norm": 2.760552406311035, + "learning_rate": 9.9802269980227e-06, + "loss": 0.0243, + "step": 19781 + }, + { + "epoch": 9.196652719665272, + "grad_norm": 7.3985795974731445, + "learning_rate": 9.9802259980226e-06, + "loss": 0.2238, + "step": 19782 + }, + { + "epoch": 9.197117619711761, + "grad_norm": 1.4704530239105225, + "learning_rate": 9.980224998022501e-06, + "loss": 0.0163, + "step": 19783 + }, + { + "epoch": 9.197582519758251, + "grad_norm": 6.4956374168396, + "learning_rate": 9.9802239980224e-06, + "loss": 0.1603, + "step": 19784 + }, + { + "epoch": 9.198047419804743, + "grad_norm": 4.123260498046875, + "learning_rate": 9.980222998022301e-06, + "loss": 0.0658, + "step": 19785 + }, + { + "epoch": 9.198512319851233, + "grad_norm": 5.671250820159912, + "learning_rate": 9.9802219980222e-06, + "loss": 0.2097, + "step": 19786 + }, + { + "epoch": 9.198977219897722, + "grad_norm": 6.609057903289795, + "learning_rate": 9.9802209980221e-06, + "loss": 0.1102, + "step": 19787 + }, + { + "epoch": 9.199442119944212, + "grad_norm": 4.062347412109375, + "learning_rate": 9.980219998022e-06, + "loss": 0.0889, + "step": 19788 + }, + { + "epoch": 9.199907019990702, + "grad_norm": 4.365978717803955, + "learning_rate": 9.9802189980219e-06, + "loss": 0.1077, + "step": 19789 + }, + { + "epoch": 9.200371920037192, + "grad_norm": 3.1811375617980957, + "learning_rate": 9.980217998021801e-06, + "loss": 0.0496, + "step": 19790 + }, + { + "epoch": 9.200836820083682, + "grad_norm": 4.349180221557617, + "learning_rate": 9.9802169980217e-06, + "loss": 0.0618, + "step": 19791 + }, + { + "epoch": 9.201301720130171, + "grad_norm": 10.762395858764648, + "learning_rate": 9.9802159980216e-06, + "loss": 0.3698, + "step": 19792 + }, + { + "epoch": 9.201766620176661, + "grad_norm": 3.596529722213745, + "learning_rate": 9.9802149980215e-06, + "loss": 0.0762, + "step": 19793 + }, + { + "epoch": 9.202231520223153, + "grad_norm": 5.166291236877441, + "learning_rate": 9.9802139980214e-06, + "loss": 0.0509, + "step": 19794 + }, + { + "epoch": 9.202696420269643, + "grad_norm": 4.559540271759033, + "learning_rate": 9.980212998021301e-06, + "loss": 0.0826, + "step": 19795 + }, + { + "epoch": 9.203161320316132, + "grad_norm": 4.072833061218262, + "learning_rate": 9.980211998021202e-06, + "loss": 0.057, + "step": 19796 + }, + { + "epoch": 9.203626220362622, + "grad_norm": 10.591850280761719, + "learning_rate": 9.9802109980211e-06, + "loss": 0.0926, + "step": 19797 + }, + { + "epoch": 9.204091120409112, + "grad_norm": 2.2017760276794434, + "learning_rate": 9.980209998021001e-06, + "loss": 0.0217, + "step": 19798 + }, + { + "epoch": 9.204556020455602, + "grad_norm": 4.329676151275635, + "learning_rate": 9.9802089980209e-06, + "loss": 0.0521, + "step": 19799 + }, + { + "epoch": 9.205020920502092, + "grad_norm": 4.100191593170166, + "learning_rate": 9.9802079980208e-06, + "loss": 0.0456, + "step": 19800 + }, + { + "epoch": 9.205485820548581, + "grad_norm": 7.644526481628418, + "learning_rate": 9.980206998020701e-06, + "loss": 0.1189, + "step": 19801 + }, + { + "epoch": 9.205950720595071, + "grad_norm": 4.000312805175781, + "learning_rate": 9.9802059980206e-06, + "loss": 0.078, + "step": 19802 + }, + { + "epoch": 9.206415620641563, + "grad_norm": 3.874046802520752, + "learning_rate": 9.980204998020501e-06, + "loss": 0.0523, + "step": 19803 + }, + { + "epoch": 9.206880520688053, + "grad_norm": 1.422473669052124, + "learning_rate": 9.9802039980204e-06, + "loss": 0.0065, + "step": 19804 + }, + { + "epoch": 9.207345420734542, + "grad_norm": 4.876908779144287, + "learning_rate": 9.9802029980203e-06, + "loss": 0.0742, + "step": 19805 + }, + { + "epoch": 9.207810320781032, + "grad_norm": 4.255858898162842, + "learning_rate": 9.9802019980202e-06, + "loss": 0.043, + "step": 19806 + }, + { + "epoch": 9.208275220827522, + "grad_norm": 6.419351100921631, + "learning_rate": 9.9802009980201e-06, + "loss": 0.0976, + "step": 19807 + }, + { + "epoch": 9.208740120874012, + "grad_norm": 5.758520603179932, + "learning_rate": 9.980199998020001e-06, + "loss": 0.1678, + "step": 19808 + }, + { + "epoch": 9.209205020920502, + "grad_norm": 4.390885353088379, + "learning_rate": 9.980198998019902e-06, + "loss": 0.071, + "step": 19809 + }, + { + "epoch": 9.209669920966991, + "grad_norm": 3.222071409225464, + "learning_rate": 9.9801979980198e-06, + "loss": 0.0271, + "step": 19810 + }, + { + "epoch": 9.210134821013481, + "grad_norm": 4.804018497467041, + "learning_rate": 9.980196998019701e-06, + "loss": 0.0591, + "step": 19811 + }, + { + "epoch": 9.210599721059973, + "grad_norm": 4.27669620513916, + "learning_rate": 9.9801959980196e-06, + "loss": 0.0593, + "step": 19812 + }, + { + "epoch": 9.211064621106463, + "grad_norm": 5.781990051269531, + "learning_rate": 9.980194998019501e-06, + "loss": 0.1466, + "step": 19813 + }, + { + "epoch": 9.211529521152952, + "grad_norm": 6.111528396606445, + "learning_rate": 9.980193998019402e-06, + "loss": 0.1028, + "step": 19814 + }, + { + "epoch": 9.211994421199442, + "grad_norm": 2.258791923522949, + "learning_rate": 9.9801929980193e-06, + "loss": 0.0196, + "step": 19815 + }, + { + "epoch": 9.212459321245932, + "grad_norm": 2.6525633335113525, + "learning_rate": 9.980191998019201e-06, + "loss": 0.018, + "step": 19816 + }, + { + "epoch": 9.212924221292422, + "grad_norm": 1.4921183586120605, + "learning_rate": 9.9801909980191e-06, + "loss": 0.0077, + "step": 19817 + }, + { + "epoch": 9.213389121338912, + "grad_norm": 3.748300075531006, + "learning_rate": 9.980189998019e-06, + "loss": 0.0448, + "step": 19818 + }, + { + "epoch": 9.213854021385401, + "grad_norm": 1.8541284799575806, + "learning_rate": 9.9801889980189e-06, + "loss": 0.015, + "step": 19819 + }, + { + "epoch": 9.214318921431893, + "grad_norm": 9.512819290161133, + "learning_rate": 9.9801879980188e-06, + "loss": 0.2438, + "step": 19820 + }, + { + "epoch": 9.214783821478383, + "grad_norm": 0.7403483986854553, + "learning_rate": 9.980186998018701e-06, + "loss": 0.0062, + "step": 19821 + }, + { + "epoch": 9.215248721524873, + "grad_norm": 2.498670816421509, + "learning_rate": 9.980185998018602e-06, + "loss": 0.027, + "step": 19822 + }, + { + "epoch": 9.215713621571362, + "grad_norm": 2.8224663734436035, + "learning_rate": 9.9801849980185e-06, + "loss": 0.0196, + "step": 19823 + }, + { + "epoch": 9.216178521617852, + "grad_norm": 3.1627697944641113, + "learning_rate": 9.9801839980184e-06, + "loss": 0.0401, + "step": 19824 + }, + { + "epoch": 9.216643421664342, + "grad_norm": 2.915070056915283, + "learning_rate": 9.9801829980183e-06, + "loss": 0.0495, + "step": 19825 + }, + { + "epoch": 9.217108321710832, + "grad_norm": 4.268050193786621, + "learning_rate": 9.980181998018201e-06, + "loss": 0.0548, + "step": 19826 + }, + { + "epoch": 9.217573221757322, + "grad_norm": 5.671670436859131, + "learning_rate": 9.980180998018102e-06, + "loss": 0.0556, + "step": 19827 + }, + { + "epoch": 9.218038121803811, + "grad_norm": 3.9283039569854736, + "learning_rate": 9.980179998018e-06, + "loss": 0.0857, + "step": 19828 + }, + { + "epoch": 9.218503021850303, + "grad_norm": 2.540313482284546, + "learning_rate": 9.980178998017901e-06, + "loss": 0.0289, + "step": 19829 + }, + { + "epoch": 9.218967921896793, + "grad_norm": 3.033569812774658, + "learning_rate": 9.9801779980178e-06, + "loss": 0.03, + "step": 19830 + }, + { + "epoch": 9.219432821943283, + "grad_norm": 2.9338479042053223, + "learning_rate": 9.9801769980177e-06, + "loss": 0.0593, + "step": 19831 + }, + { + "epoch": 9.219897721989772, + "grad_norm": 5.726809978485107, + "learning_rate": 9.9801759980176e-06, + "loss": 0.1073, + "step": 19832 + }, + { + "epoch": 9.220362622036262, + "grad_norm": 4.652692794799805, + "learning_rate": 9.9801749980175e-06, + "loss": 0.0545, + "step": 19833 + }, + { + "epoch": 9.220827522082752, + "grad_norm": 2.750666856765747, + "learning_rate": 9.980173998017401e-06, + "loss": 0.0391, + "step": 19834 + }, + { + "epoch": 9.221292422129242, + "grad_norm": 4.264527797698975, + "learning_rate": 9.9801729980173e-06, + "loss": 0.061, + "step": 19835 + }, + { + "epoch": 9.221757322175732, + "grad_norm": 4.053568363189697, + "learning_rate": 9.9801719980172e-06, + "loss": 0.0458, + "step": 19836 + }, + { + "epoch": 9.222222222222221, + "grad_norm": 5.402443885803223, + "learning_rate": 9.9801709980171e-06, + "loss": 0.0562, + "step": 19837 + }, + { + "epoch": 9.222687122268713, + "grad_norm": 4.592064380645752, + "learning_rate": 9.980169998017e-06, + "loss": 0.0839, + "step": 19838 + }, + { + "epoch": 9.223152022315203, + "grad_norm": 5.842435836791992, + "learning_rate": 9.980168998016901e-06, + "loss": 0.0888, + "step": 19839 + }, + { + "epoch": 9.223616922361693, + "grad_norm": 5.51036262512207, + "learning_rate": 9.980167998016802e-06, + "loss": 0.1303, + "step": 19840 + }, + { + "epoch": 9.224081822408182, + "grad_norm": 6.1248555183410645, + "learning_rate": 9.9801669980167e-06, + "loss": 0.1371, + "step": 19841 + }, + { + "epoch": 9.224546722454672, + "grad_norm": 5.737387657165527, + "learning_rate": 9.980165998016601e-06, + "loss": 0.1209, + "step": 19842 + }, + { + "epoch": 9.225011622501162, + "grad_norm": 5.097543716430664, + "learning_rate": 9.9801649980165e-06, + "loss": 0.054, + "step": 19843 + }, + { + "epoch": 9.225476522547652, + "grad_norm": 7.170145511627197, + "learning_rate": 9.980163998016401e-06, + "loss": 0.1378, + "step": 19844 + }, + { + "epoch": 9.225941422594142, + "grad_norm": 3.506894826889038, + "learning_rate": 9.9801629980163e-06, + "loss": 0.0472, + "step": 19845 + }, + { + "epoch": 9.226406322640631, + "grad_norm": 5.808078765869141, + "learning_rate": 9.9801619980162e-06, + "loss": 0.0808, + "step": 19846 + }, + { + "epoch": 9.226871222687123, + "grad_norm": 4.309179782867432, + "learning_rate": 9.980160998016101e-06, + "loss": 0.0607, + "step": 19847 + }, + { + "epoch": 9.227336122733613, + "grad_norm": 3.102825164794922, + "learning_rate": 9.980159998016e-06, + "loss": 0.0515, + "step": 19848 + }, + { + "epoch": 9.227801022780103, + "grad_norm": 2.8030800819396973, + "learning_rate": 9.9801589980159e-06, + "loss": 0.053, + "step": 19849 + }, + { + "epoch": 9.228265922826592, + "grad_norm": 4.302008152008057, + "learning_rate": 9.9801579980158e-06, + "loss": 0.0305, + "step": 19850 + }, + { + "epoch": 9.228730822873082, + "grad_norm": 6.917381763458252, + "learning_rate": 9.9801569980157e-06, + "loss": 0.1413, + "step": 19851 + }, + { + "epoch": 9.229195722919572, + "grad_norm": 5.428366184234619, + "learning_rate": 9.980155998015601e-06, + "loss": 0.04, + "step": 19852 + }, + { + "epoch": 9.229660622966062, + "grad_norm": 6.543982028961182, + "learning_rate": 9.980154998015502e-06, + "loss": 0.0712, + "step": 19853 + }, + { + "epoch": 9.230125523012552, + "grad_norm": 12.570164680480957, + "learning_rate": 9.9801539980154e-06, + "loss": 0.3489, + "step": 19854 + }, + { + "epoch": 9.230590423059041, + "grad_norm": 5.64621639251709, + "learning_rate": 9.980152998015301e-06, + "loss": 0.0692, + "step": 19855 + }, + { + "epoch": 9.231055323105533, + "grad_norm": 1.9913568496704102, + "learning_rate": 9.9801519980152e-06, + "loss": 0.0165, + "step": 19856 + }, + { + "epoch": 9.231520223152023, + "grad_norm": 6.70337438583374, + "learning_rate": 9.9801509980151e-06, + "loss": 0.094, + "step": 19857 + }, + { + "epoch": 9.231985123198513, + "grad_norm": 1.7261172533035278, + "learning_rate": 9.980149998015e-06, + "loss": 0.013, + "step": 19858 + }, + { + "epoch": 9.232450023245002, + "grad_norm": 14.533502578735352, + "learning_rate": 9.9801489980149e-06, + "loss": 0.3221, + "step": 19859 + }, + { + "epoch": 9.232914923291492, + "grad_norm": 4.171534061431885, + "learning_rate": 9.980147998014801e-06, + "loss": 0.0558, + "step": 19860 + }, + { + "epoch": 9.233379823337982, + "grad_norm": 8.895230293273926, + "learning_rate": 9.9801469980147e-06, + "loss": 0.2252, + "step": 19861 + }, + { + "epoch": 9.233844723384472, + "grad_norm": 10.358428001403809, + "learning_rate": 9.980145998014601e-06, + "loss": 0.182, + "step": 19862 + }, + { + "epoch": 9.234309623430962, + "grad_norm": 3.7904903888702393, + "learning_rate": 9.9801449980145e-06, + "loss": 0.055, + "step": 19863 + }, + { + "epoch": 9.234774523477453, + "grad_norm": 3.1060726642608643, + "learning_rate": 9.9801439980144e-06, + "loss": 0.0347, + "step": 19864 + }, + { + "epoch": 9.235239423523943, + "grad_norm": 4.485091686248779, + "learning_rate": 9.980142998014301e-06, + "loss": 0.0723, + "step": 19865 + }, + { + "epoch": 9.235704323570433, + "grad_norm": 4.400397777557373, + "learning_rate": 9.980141998014202e-06, + "loss": 0.0519, + "step": 19866 + }, + { + "epoch": 9.236169223616923, + "grad_norm": 9.217208862304688, + "learning_rate": 9.9801409980141e-06, + "loss": 0.2011, + "step": 19867 + }, + { + "epoch": 9.236634123663412, + "grad_norm": 5.3390021324157715, + "learning_rate": 9.980139998014e-06, + "loss": 0.0912, + "step": 19868 + }, + { + "epoch": 9.237099023709902, + "grad_norm": 3.523218870162964, + "learning_rate": 9.9801389980139e-06, + "loss": 0.0507, + "step": 19869 + }, + { + "epoch": 9.237563923756392, + "grad_norm": 4.822748184204102, + "learning_rate": 9.980137998013801e-06, + "loss": 0.1049, + "step": 19870 + }, + { + "epoch": 9.238028823802882, + "grad_norm": 4.343299388885498, + "learning_rate": 9.9801369980137e-06, + "loss": 0.0359, + "step": 19871 + }, + { + "epoch": 9.238493723849372, + "grad_norm": 7.065001010894775, + "learning_rate": 9.9801359980136e-06, + "loss": 0.1088, + "step": 19872 + }, + { + "epoch": 9.238958623895863, + "grad_norm": 8.568747520446777, + "learning_rate": 9.980134998013501e-06, + "loss": 0.1526, + "step": 19873 + }, + { + "epoch": 9.239423523942353, + "grad_norm": 2.8488876819610596, + "learning_rate": 9.9801339980134e-06, + "loss": 0.0456, + "step": 19874 + }, + { + "epoch": 9.239888423988843, + "grad_norm": 5.3842668533325195, + "learning_rate": 9.980132998013301e-06, + "loss": 0.095, + "step": 19875 + }, + { + "epoch": 9.240353324035333, + "grad_norm": 7.327824592590332, + "learning_rate": 9.9801319980132e-06, + "loss": 0.2366, + "step": 19876 + }, + { + "epoch": 9.240818224081822, + "grad_norm": 4.020023822784424, + "learning_rate": 9.9801309980131e-06, + "loss": 0.0446, + "step": 19877 + }, + { + "epoch": 9.241283124128312, + "grad_norm": 5.967428684234619, + "learning_rate": 9.980129998013001e-06, + "loss": 0.1199, + "step": 19878 + }, + { + "epoch": 9.241748024174802, + "grad_norm": 6.557086944580078, + "learning_rate": 9.9801289980129e-06, + "loss": 0.11, + "step": 19879 + }, + { + "epoch": 9.242212924221292, + "grad_norm": 4.386063098907471, + "learning_rate": 9.980127998012801e-06, + "loss": 0.0658, + "step": 19880 + }, + { + "epoch": 9.242677824267782, + "grad_norm": 1.7639057636260986, + "learning_rate": 9.9801269980127e-06, + "loss": 0.0183, + "step": 19881 + }, + { + "epoch": 9.243142724314273, + "grad_norm": 5.565736770629883, + "learning_rate": 9.9801259980126e-06, + "loss": 0.1218, + "step": 19882 + }, + { + "epoch": 9.243607624360763, + "grad_norm": 4.8275275230407715, + "learning_rate": 9.980124998012501e-06, + "loss": 0.0387, + "step": 19883 + }, + { + "epoch": 9.244072524407253, + "grad_norm": 3.1022074222564697, + "learning_rate": 9.9801239980124e-06, + "loss": 0.0489, + "step": 19884 + }, + { + "epoch": 9.244537424453743, + "grad_norm": 6.989318370819092, + "learning_rate": 9.9801229980123e-06, + "loss": 0.1277, + "step": 19885 + }, + { + "epoch": 9.245002324500232, + "grad_norm": 6.744251251220703, + "learning_rate": 9.980121998012201e-06, + "loss": 0.0897, + "step": 19886 + }, + { + "epoch": 9.245467224546722, + "grad_norm": 7.581141948699951, + "learning_rate": 9.9801209980121e-06, + "loss": 0.1047, + "step": 19887 + }, + { + "epoch": 9.245932124593212, + "grad_norm": 1.9723427295684814, + "learning_rate": 9.980119998012001e-06, + "loss": 0.0224, + "step": 19888 + }, + { + "epoch": 9.246397024639702, + "grad_norm": 7.90012788772583, + "learning_rate": 9.9801189980119e-06, + "loss": 0.116, + "step": 19889 + }, + { + "epoch": 9.246861924686192, + "grad_norm": 3.693441390991211, + "learning_rate": 9.9801179980118e-06, + "loss": 0.0711, + "step": 19890 + }, + { + "epoch": 9.247326824732683, + "grad_norm": 8.1412353515625, + "learning_rate": 9.980116998011701e-06, + "loss": 0.1225, + "step": 19891 + }, + { + "epoch": 9.247791724779173, + "grad_norm": 2.0396618843078613, + "learning_rate": 9.9801159980116e-06, + "loss": 0.024, + "step": 19892 + }, + { + "epoch": 9.248256624825663, + "grad_norm": 9.953493118286133, + "learning_rate": 9.980114998011501e-06, + "loss": 0.2673, + "step": 19893 + }, + { + "epoch": 9.248721524872153, + "grad_norm": 6.561968803405762, + "learning_rate": 9.9801139980114e-06, + "loss": 0.1449, + "step": 19894 + }, + { + "epoch": 9.249186424918642, + "grad_norm": 4.729427337646484, + "learning_rate": 9.9801129980113e-06, + "loss": 0.1131, + "step": 19895 + }, + { + "epoch": 9.249651324965132, + "grad_norm": 5.4744181632995605, + "learning_rate": 9.980111998011201e-06, + "loss": 0.1083, + "step": 19896 + }, + { + "epoch": 9.250116225011622, + "grad_norm": 2.476330518722534, + "learning_rate": 9.980110998011102e-06, + "loss": 0.0432, + "step": 19897 + }, + { + "epoch": 9.250581125058112, + "grad_norm": 3.3045735359191895, + "learning_rate": 9.980109998011001e-06, + "loss": 0.0665, + "step": 19898 + }, + { + "epoch": 9.251046025104603, + "grad_norm": 6.470767974853516, + "learning_rate": 9.980108998010902e-06, + "loss": 0.1411, + "step": 19899 + }, + { + "epoch": 9.251510925151093, + "grad_norm": 5.660082817077637, + "learning_rate": 9.9801079980108e-06, + "loss": 0.1224, + "step": 19900 + }, + { + "epoch": 9.251975825197583, + "grad_norm": 5.186901569366455, + "learning_rate": 9.9801069980107e-06, + "loss": 0.0511, + "step": 19901 + }, + { + "epoch": 9.252440725244073, + "grad_norm": 4.949059963226318, + "learning_rate": 9.9801059980106e-06, + "loss": 0.1292, + "step": 19902 + }, + { + "epoch": 9.252905625290563, + "grad_norm": 5.156977653503418, + "learning_rate": 9.9801049980105e-06, + "loss": 0.0914, + "step": 19903 + }, + { + "epoch": 9.253370525337052, + "grad_norm": 3.9045302867889404, + "learning_rate": 9.980103998010401e-06, + "loss": 0.0461, + "step": 19904 + }, + { + "epoch": 9.253835425383542, + "grad_norm": 5.366934776306152, + "learning_rate": 9.9801029980103e-06, + "loss": 0.0643, + "step": 19905 + }, + { + "epoch": 9.254300325430032, + "grad_norm": 5.504922866821289, + "learning_rate": 9.980101998010201e-06, + "loss": 0.0776, + "step": 19906 + }, + { + "epoch": 9.254765225476522, + "grad_norm": 5.432289123535156, + "learning_rate": 9.9801009980101e-06, + "loss": 0.0761, + "step": 19907 + }, + { + "epoch": 9.255230125523013, + "grad_norm": 5.07008171081543, + "learning_rate": 9.98009999801e-06, + "loss": 0.113, + "step": 19908 + }, + { + "epoch": 9.255695025569503, + "grad_norm": 6.2863287925720215, + "learning_rate": 9.980098998009901e-06, + "loss": 0.0906, + "step": 19909 + }, + { + "epoch": 9.256159925615993, + "grad_norm": 13.842157363891602, + "learning_rate": 9.980097998009802e-06, + "loss": 0.1136, + "step": 19910 + }, + { + "epoch": 9.256624825662483, + "grad_norm": 4.409624099731445, + "learning_rate": 9.980096998009701e-06, + "loss": 0.0796, + "step": 19911 + }, + { + "epoch": 9.257089725708973, + "grad_norm": 2.93796706199646, + "learning_rate": 9.9800959980096e-06, + "loss": 0.0321, + "step": 19912 + }, + { + "epoch": 9.257554625755462, + "grad_norm": 9.500297546386719, + "learning_rate": 9.9800949980095e-06, + "loss": 0.0964, + "step": 19913 + }, + { + "epoch": 9.258019525801952, + "grad_norm": 4.209245681762695, + "learning_rate": 9.9800939980094e-06, + "loss": 0.0879, + "step": 19914 + }, + { + "epoch": 9.258484425848442, + "grad_norm": 3.5243821144104004, + "learning_rate": 9.9800929980093e-06, + "loss": 0.0601, + "step": 19915 + }, + { + "epoch": 9.258949325894932, + "grad_norm": 6.784682750701904, + "learning_rate": 9.9800919980092e-06, + "loss": 0.1781, + "step": 19916 + }, + { + "epoch": 9.259414225941423, + "grad_norm": 2.4177989959716797, + "learning_rate": 9.980090998009101e-06, + "loss": 0.025, + "step": 19917 + }, + { + "epoch": 9.259879125987913, + "grad_norm": 2.1505141258239746, + "learning_rate": 9.980089998009e-06, + "loss": 0.0177, + "step": 19918 + }, + { + "epoch": 9.260344026034403, + "grad_norm": 2.946497678756714, + "learning_rate": 9.980088998008901e-06, + "loss": 0.0343, + "step": 19919 + }, + { + "epoch": 9.260808926080893, + "grad_norm": 6.531228542327881, + "learning_rate": 9.9800879980088e-06, + "loss": 0.1568, + "step": 19920 + }, + { + "epoch": 9.261273826127383, + "grad_norm": 11.120790481567383, + "learning_rate": 9.9800869980087e-06, + "loss": 0.1024, + "step": 19921 + }, + { + "epoch": 9.261738726173872, + "grad_norm": 6.781667232513428, + "learning_rate": 9.980085998008601e-06, + "loss": 0.2109, + "step": 19922 + }, + { + "epoch": 9.262203626220362, + "grad_norm": 5.5499749183654785, + "learning_rate": 9.9800849980085e-06, + "loss": 0.0989, + "step": 19923 + }, + { + "epoch": 9.262668526266852, + "grad_norm": 7.5677170753479, + "learning_rate": 9.980083998008401e-06, + "loss": 0.1461, + "step": 19924 + }, + { + "epoch": 9.263133426313342, + "grad_norm": 6.196533203125, + "learning_rate": 9.9800829980083e-06, + "loss": 0.0715, + "step": 19925 + }, + { + "epoch": 9.263598326359833, + "grad_norm": 5.356788635253906, + "learning_rate": 9.9800819980082e-06, + "loss": 0.1281, + "step": 19926 + }, + { + "epoch": 9.264063226406323, + "grad_norm": 7.015480041503906, + "learning_rate": 9.9800809980081e-06, + "loss": 0.1516, + "step": 19927 + }, + { + "epoch": 9.264528126452813, + "grad_norm": 3.8780317306518555, + "learning_rate": 9.980079998008e-06, + "loss": 0.0734, + "step": 19928 + }, + { + "epoch": 9.264993026499303, + "grad_norm": 3.978452682495117, + "learning_rate": 9.980078998007901e-06, + "loss": 0.0534, + "step": 19929 + }, + { + "epoch": 9.265457926545793, + "grad_norm": 6.569169044494629, + "learning_rate": 9.980077998007802e-06, + "loss": 0.1263, + "step": 19930 + }, + { + "epoch": 9.265922826592282, + "grad_norm": 4.521809101104736, + "learning_rate": 9.9800769980077e-06, + "loss": 0.0664, + "step": 19931 + }, + { + "epoch": 9.266387726638772, + "grad_norm": 3.9810214042663574, + "learning_rate": 9.9800759980076e-06, + "loss": 0.0552, + "step": 19932 + }, + { + "epoch": 9.266852626685262, + "grad_norm": 6.895422458648682, + "learning_rate": 9.9800749980075e-06, + "loss": 0.1003, + "step": 19933 + }, + { + "epoch": 9.267317526731752, + "grad_norm": 5.575645923614502, + "learning_rate": 9.9800739980074e-06, + "loss": 0.1071, + "step": 19934 + }, + { + "epoch": 9.267782426778243, + "grad_norm": 4.6161065101623535, + "learning_rate": 9.980072998007301e-06, + "loss": 0.1297, + "step": 19935 + }, + { + "epoch": 9.268247326824733, + "grad_norm": 4.059526443481445, + "learning_rate": 9.9800719980072e-06, + "loss": 0.0694, + "step": 19936 + }, + { + "epoch": 9.268712226871223, + "grad_norm": 2.991072177886963, + "learning_rate": 9.980070998007101e-06, + "loss": 0.024, + "step": 19937 + }, + { + "epoch": 9.269177126917713, + "grad_norm": 1.9325348138809204, + "learning_rate": 9.980069998007e-06, + "loss": 0.0198, + "step": 19938 + }, + { + "epoch": 9.269642026964203, + "grad_norm": 6.934972286224365, + "learning_rate": 9.9800689980069e-06, + "loss": 0.1316, + "step": 19939 + }, + { + "epoch": 9.270106927010692, + "grad_norm": 5.323563575744629, + "learning_rate": 9.9800679980068e-06, + "loss": 0.1218, + "step": 19940 + }, + { + "epoch": 9.270571827057182, + "grad_norm": 5.559248447418213, + "learning_rate": 9.9800669980067e-06, + "loss": 0.0989, + "step": 19941 + }, + { + "epoch": 9.271036727103672, + "grad_norm": 3.332705020904541, + "learning_rate": 9.980065998006601e-06, + "loss": 0.0464, + "step": 19942 + }, + { + "epoch": 9.271501627150164, + "grad_norm": 2.6514089107513428, + "learning_rate": 9.9800649980065e-06, + "loss": 0.0227, + "step": 19943 + }, + { + "epoch": 9.271966527196653, + "grad_norm": 5.152764320373535, + "learning_rate": 9.9800639980064e-06, + "loss": 0.0435, + "step": 19944 + }, + { + "epoch": 9.272431427243143, + "grad_norm": 5.026638507843018, + "learning_rate": 9.9800629980063e-06, + "loss": 0.0576, + "step": 19945 + }, + { + "epoch": 9.272896327289633, + "grad_norm": 6.49506139755249, + "learning_rate": 9.9800619980062e-06, + "loss": 0.1091, + "step": 19946 + }, + { + "epoch": 9.273361227336123, + "grad_norm": 5.999063014984131, + "learning_rate": 9.980060998006101e-06, + "loss": 0.1513, + "step": 19947 + }, + { + "epoch": 9.273826127382613, + "grad_norm": 7.083592891693115, + "learning_rate": 9.980059998006002e-06, + "loss": 0.0885, + "step": 19948 + }, + { + "epoch": 9.274291027429102, + "grad_norm": 6.2013773918151855, + "learning_rate": 9.9800589980059e-06, + "loss": 0.1578, + "step": 19949 + }, + { + "epoch": 9.274755927475592, + "grad_norm": 1.6273235082626343, + "learning_rate": 9.980057998005801e-06, + "loss": 0.0128, + "step": 19950 + }, + { + "epoch": 9.275220827522082, + "grad_norm": 3.9821438789367676, + "learning_rate": 9.9800569980057e-06, + "loss": 0.0712, + "step": 19951 + }, + { + "epoch": 9.275685727568574, + "grad_norm": 6.482639789581299, + "learning_rate": 9.9800559980056e-06, + "loss": 0.1764, + "step": 19952 + }, + { + "epoch": 9.276150627615063, + "grad_norm": 6.16172981262207, + "learning_rate": 9.9800549980055e-06, + "loss": 0.2296, + "step": 19953 + }, + { + "epoch": 9.276615527661553, + "grad_norm": 6.666070461273193, + "learning_rate": 9.9800539980054e-06, + "loss": 0.1173, + "step": 19954 + }, + { + "epoch": 9.277080427708043, + "grad_norm": 7.350200176239014, + "learning_rate": 9.980052998005301e-06, + "loss": 0.2201, + "step": 19955 + }, + { + "epoch": 9.277545327754533, + "grad_norm": 8.109159469604492, + "learning_rate": 9.9800519980052e-06, + "loss": 0.1433, + "step": 19956 + }, + { + "epoch": 9.278010227801023, + "grad_norm": 4.425460338592529, + "learning_rate": 9.9800509980051e-06, + "loss": 0.0769, + "step": 19957 + }, + { + "epoch": 9.278475127847512, + "grad_norm": 4.926674842834473, + "learning_rate": 9.980049998005e-06, + "loss": 0.0585, + "step": 19958 + }, + { + "epoch": 9.278940027894002, + "grad_norm": 3.4286599159240723, + "learning_rate": 9.9800489980049e-06, + "loss": 0.0399, + "step": 19959 + }, + { + "epoch": 9.279404927940492, + "grad_norm": 3.9428763389587402, + "learning_rate": 9.980047998004801e-06, + "loss": 0.0701, + "step": 19960 + }, + { + "epoch": 9.279869827986984, + "grad_norm": 4.210141658782959, + "learning_rate": 9.980046998004702e-06, + "loss": 0.0355, + "step": 19961 + }, + { + "epoch": 9.280334728033473, + "grad_norm": 3.022876024246216, + "learning_rate": 9.9800459980046e-06, + "loss": 0.045, + "step": 19962 + }, + { + "epoch": 9.280799628079963, + "grad_norm": 3.764474630355835, + "learning_rate": 9.980044998004501e-06, + "loss": 0.0688, + "step": 19963 + }, + { + "epoch": 9.281264528126453, + "grad_norm": 4.6825270652771, + "learning_rate": 9.9800439980044e-06, + "loss": 0.0589, + "step": 19964 + }, + { + "epoch": 9.281729428172943, + "grad_norm": 6.802711486816406, + "learning_rate": 9.980042998004301e-06, + "loss": 0.1367, + "step": 19965 + }, + { + "epoch": 9.282194328219433, + "grad_norm": 6.697854042053223, + "learning_rate": 9.980041998004202e-06, + "loss": 0.1109, + "step": 19966 + }, + { + "epoch": 9.282659228265922, + "grad_norm": 3.399604320526123, + "learning_rate": 9.9800409980041e-06, + "loss": 0.0379, + "step": 19967 + }, + { + "epoch": 9.283124128312412, + "grad_norm": 2.340824842453003, + "learning_rate": 9.980039998004001e-06, + "loss": 0.0258, + "step": 19968 + }, + { + "epoch": 9.283589028358902, + "grad_norm": 4.770471096038818, + "learning_rate": 9.9800389980039e-06, + "loss": 0.0725, + "step": 19969 + }, + { + "epoch": 9.284053928405394, + "grad_norm": 3.64345645904541, + "learning_rate": 9.9800379980038e-06, + "loss": 0.0693, + "step": 19970 + }, + { + "epoch": 9.284518828451883, + "grad_norm": 2.7288694381713867, + "learning_rate": 9.9800369980037e-06, + "loss": 0.036, + "step": 19971 + }, + { + "epoch": 9.284983728498373, + "grad_norm": 4.703575134277344, + "learning_rate": 9.9800359980036e-06, + "loss": 0.1074, + "step": 19972 + }, + { + "epoch": 9.285448628544863, + "grad_norm": 6.319881916046143, + "learning_rate": 9.980034998003501e-06, + "loss": 0.1179, + "step": 19973 + }, + { + "epoch": 9.285913528591353, + "grad_norm": 5.35110330581665, + "learning_rate": 9.980033998003402e-06, + "loss": 0.0612, + "step": 19974 + }, + { + "epoch": 9.286378428637843, + "grad_norm": 4.832724571228027, + "learning_rate": 9.9800329980033e-06, + "loss": 0.0409, + "step": 19975 + }, + { + "epoch": 9.286843328684332, + "grad_norm": 9.395851135253906, + "learning_rate": 9.9800319980032e-06, + "loss": 0.2249, + "step": 19976 + }, + { + "epoch": 9.287308228730822, + "grad_norm": 3.96671724319458, + "learning_rate": 9.9800309980031e-06, + "loss": 0.0635, + "step": 19977 + }, + { + "epoch": 9.287773128777314, + "grad_norm": 3.682513952255249, + "learning_rate": 9.980029998003001e-06, + "loss": 0.0462, + "step": 19978 + }, + { + "epoch": 9.288238028823804, + "grad_norm": 1.8554720878601074, + "learning_rate": 9.980028998002902e-06, + "loss": 0.0179, + "step": 19979 + }, + { + "epoch": 9.288702928870293, + "grad_norm": 3.7912778854370117, + "learning_rate": 9.9800279980028e-06, + "loss": 0.0735, + "step": 19980 + }, + { + "epoch": 9.289167828916783, + "grad_norm": 1.3347340822219849, + "learning_rate": 9.980026998002701e-06, + "loss": 0.0098, + "step": 19981 + }, + { + "epoch": 9.289632728963273, + "grad_norm": 3.264338254928589, + "learning_rate": 9.9800259980026e-06, + "loss": 0.0339, + "step": 19982 + }, + { + "epoch": 9.290097629009763, + "grad_norm": 1.6894896030426025, + "learning_rate": 9.980024998002501e-06, + "loss": 0.018, + "step": 19983 + }, + { + "epoch": 9.290562529056253, + "grad_norm": 14.440014839172363, + "learning_rate": 9.9800239980024e-06, + "loss": 0.1556, + "step": 19984 + }, + { + "epoch": 9.291027429102742, + "grad_norm": 1.9526562690734863, + "learning_rate": 9.9800229980023e-06, + "loss": 0.0245, + "step": 19985 + }, + { + "epoch": 9.291492329149232, + "grad_norm": 5.263566970825195, + "learning_rate": 9.980021998002201e-06, + "loss": 0.0662, + "step": 19986 + }, + { + "epoch": 9.291957229195724, + "grad_norm": 5.099310398101807, + "learning_rate": 9.9800209980021e-06, + "loss": 0.0432, + "step": 19987 + }, + { + "epoch": 9.292422129242214, + "grad_norm": 6.611384391784668, + "learning_rate": 9.980019998002e-06, + "loss": 0.019, + "step": 19988 + }, + { + "epoch": 9.292887029288703, + "grad_norm": 3.789045810699463, + "learning_rate": 9.9800189980019e-06, + "loss": 0.0498, + "step": 19989 + }, + { + "epoch": 9.293351929335193, + "grad_norm": 4.162758827209473, + "learning_rate": 9.9800179980018e-06, + "loss": 0.1074, + "step": 19990 + }, + { + "epoch": 9.293816829381683, + "grad_norm": 6.280987739562988, + "learning_rate": 9.980016998001701e-06, + "loss": 0.0687, + "step": 19991 + }, + { + "epoch": 9.294281729428173, + "grad_norm": 3.61657452583313, + "learning_rate": 9.980015998001602e-06, + "loss": 0.0655, + "step": 19992 + }, + { + "epoch": 9.294746629474663, + "grad_norm": 4.032726287841797, + "learning_rate": 9.9800149980015e-06, + "loss": 0.0491, + "step": 19993 + }, + { + "epoch": 9.295211529521152, + "grad_norm": 5.924907207489014, + "learning_rate": 9.980013998001401e-06, + "loss": 0.1277, + "step": 19994 + }, + { + "epoch": 9.295676429567642, + "grad_norm": 6.794639587402344, + "learning_rate": 9.9800129980013e-06, + "loss": 0.1868, + "step": 19995 + }, + { + "epoch": 9.296141329614134, + "grad_norm": 6.902149677276611, + "learning_rate": 9.980011998001201e-06, + "loss": 0.1598, + "step": 19996 + }, + { + "epoch": 9.296606229660624, + "grad_norm": 5.974283218383789, + "learning_rate": 9.9800109980011e-06, + "loss": 0.0461, + "step": 19997 + }, + { + "epoch": 9.297071129707113, + "grad_norm": 3.0679285526275635, + "learning_rate": 9.980009998001e-06, + "loss": 0.0423, + "step": 19998 + }, + { + "epoch": 9.297536029753603, + "grad_norm": 4.5452494621276855, + "learning_rate": 9.980008998000901e-06, + "loss": 0.0574, + "step": 19999 + }, + { + "epoch": 9.298000929800093, + "grad_norm": 6.976518630981445, + "learning_rate": 9.9800079980008e-06, + "loss": 0.1, + "step": 20000 + }, + { + "epoch": 9.298465829846583, + "grad_norm": 5.236626625061035, + "learning_rate": 9.9800069980007e-06, + "loss": 0.0448, + "step": 20001 + }, + { + "epoch": 9.298930729893073, + "grad_norm": 5.976956367492676, + "learning_rate": 9.9800059980006e-06, + "loss": 0.1537, + "step": 20002 + }, + { + "epoch": 9.299395629939562, + "grad_norm": 5.826699256896973, + "learning_rate": 9.9800049980005e-06, + "loss": 0.1201, + "step": 20003 + }, + { + "epoch": 9.299860529986052, + "grad_norm": 7.125094413757324, + "learning_rate": 9.980003998000401e-06, + "loss": 0.1533, + "step": 20004 + }, + { + "epoch": 9.300325430032544, + "grad_norm": 3.5424556732177734, + "learning_rate": 9.980002998000302e-06, + "loss": 0.0704, + "step": 20005 + }, + { + "epoch": 9.300790330079034, + "grad_norm": 5.0614423751831055, + "learning_rate": 9.9800019980002e-06, + "loss": 0.0668, + "step": 20006 + }, + { + "epoch": 9.301255230125523, + "grad_norm": 3.09491229057312, + "learning_rate": 9.980000998000101e-06, + "loss": 0.0474, + "step": 20007 + }, + { + "epoch": 9.301720130172013, + "grad_norm": 5.934370517730713, + "learning_rate": 9.979999998e-06, + "loss": 0.0704, + "step": 20008 + }, + { + "epoch": 9.302185030218503, + "grad_norm": 5.716413974761963, + "learning_rate": 9.9799989979999e-06, + "loss": 0.1398, + "step": 20009 + }, + { + "epoch": 9.302649930264993, + "grad_norm": 10.351634979248047, + "learning_rate": 9.9799979979998e-06, + "loss": 0.3522, + "step": 20010 + }, + { + "epoch": 9.303114830311483, + "grad_norm": 7.74646520614624, + "learning_rate": 9.9799969979997e-06, + "loss": 0.1496, + "step": 20011 + }, + { + "epoch": 9.303579730357972, + "grad_norm": 2.81477952003479, + "learning_rate": 9.979995997999601e-06, + "loss": 0.0453, + "step": 20012 + }, + { + "epoch": 9.304044630404462, + "grad_norm": 4.736658573150635, + "learning_rate": 9.9799949979995e-06, + "loss": 0.0613, + "step": 20013 + }, + { + "epoch": 9.304509530450954, + "grad_norm": 3.656521797180176, + "learning_rate": 9.979993997999401e-06, + "loss": 0.0197, + "step": 20014 + }, + { + "epoch": 9.304974430497444, + "grad_norm": 4.090147495269775, + "learning_rate": 9.9799929979993e-06, + "loss": 0.0954, + "step": 20015 + }, + { + "epoch": 9.305439330543933, + "grad_norm": 6.349493026733398, + "learning_rate": 9.9799919979992e-06, + "loss": 0.1925, + "step": 20016 + }, + { + "epoch": 9.305904230590423, + "grad_norm": 6.661462306976318, + "learning_rate": 9.979990997999101e-06, + "loss": 0.1748, + "step": 20017 + }, + { + "epoch": 9.306369130636913, + "grad_norm": 4.958289623260498, + "learning_rate": 9.979989997999002e-06, + "loss": 0.1238, + "step": 20018 + }, + { + "epoch": 9.306834030683403, + "grad_norm": 5.827499866485596, + "learning_rate": 9.9799889979989e-06, + "loss": 0.1258, + "step": 20019 + }, + { + "epoch": 9.307298930729893, + "grad_norm": 5.3883233070373535, + "learning_rate": 9.9799879979988e-06, + "loss": 0.094, + "step": 20020 + }, + { + "epoch": 9.307763830776382, + "grad_norm": 5.26501989364624, + "learning_rate": 9.9799869979987e-06, + "loss": 0.0974, + "step": 20021 + }, + { + "epoch": 9.308228730822874, + "grad_norm": 4.242022514343262, + "learning_rate": 9.9799859979986e-06, + "loss": 0.0527, + "step": 20022 + }, + { + "epoch": 9.308693630869364, + "grad_norm": 4.22589111328125, + "learning_rate": 9.9799849979985e-06, + "loss": 0.1017, + "step": 20023 + }, + { + "epoch": 9.309158530915854, + "grad_norm": 3.4067611694335938, + "learning_rate": 9.9799839979984e-06, + "loss": 0.0338, + "step": 20024 + }, + { + "epoch": 9.309623430962343, + "grad_norm": 6.317617416381836, + "learning_rate": 9.979982997998301e-06, + "loss": 0.1305, + "step": 20025 + }, + { + "epoch": 9.310088331008833, + "grad_norm": 5.233623027801514, + "learning_rate": 9.9799819979982e-06, + "loss": 0.0537, + "step": 20026 + }, + { + "epoch": 9.310553231055323, + "grad_norm": 7.059096813201904, + "learning_rate": 9.979980997998101e-06, + "loss": 0.125, + "step": 20027 + }, + { + "epoch": 9.311018131101813, + "grad_norm": 8.312010765075684, + "learning_rate": 9.979979997998e-06, + "loss": 0.2538, + "step": 20028 + }, + { + "epoch": 9.311483031148303, + "grad_norm": 5.802382469177246, + "learning_rate": 9.9799789979979e-06, + "loss": 0.1056, + "step": 20029 + }, + { + "epoch": 9.311947931194792, + "grad_norm": 7.452005863189697, + "learning_rate": 9.979977997997801e-06, + "loss": 0.0895, + "step": 20030 + }, + { + "epoch": 9.312412831241284, + "grad_norm": 4.395304203033447, + "learning_rate": 9.9799769979977e-06, + "loss": 0.0605, + "step": 20031 + }, + { + "epoch": 9.312877731287774, + "grad_norm": 6.261448860168457, + "learning_rate": 9.979975997997601e-06, + "loss": 0.1041, + "step": 20032 + }, + { + "epoch": 9.313342631334264, + "grad_norm": 2.77960205078125, + "learning_rate": 9.9799749979975e-06, + "loss": 0.0397, + "step": 20033 + }, + { + "epoch": 9.313807531380753, + "grad_norm": 4.6134033203125, + "learning_rate": 9.9799739979974e-06, + "loss": 0.0337, + "step": 20034 + }, + { + "epoch": 9.314272431427243, + "grad_norm": 7.378767490386963, + "learning_rate": 9.979972997997301e-06, + "loss": 0.1009, + "step": 20035 + }, + { + "epoch": 9.314737331473733, + "grad_norm": 7.126227855682373, + "learning_rate": 9.9799719979972e-06, + "loss": 0.1704, + "step": 20036 + }, + { + "epoch": 9.315202231520223, + "grad_norm": 2.065372943878174, + "learning_rate": 9.9799709979971e-06, + "loss": 0.0194, + "step": 20037 + }, + { + "epoch": 9.315667131566713, + "grad_norm": 3.6151950359344482, + "learning_rate": 9.979969997997001e-06, + "loss": 0.0509, + "step": 20038 + }, + { + "epoch": 9.316132031613202, + "grad_norm": 4.689030170440674, + "learning_rate": 9.9799689979969e-06, + "loss": 0.0915, + "step": 20039 + }, + { + "epoch": 9.316596931659694, + "grad_norm": 6.9592061042785645, + "learning_rate": 9.979967997996801e-06, + "loss": 0.0869, + "step": 20040 + }, + { + "epoch": 9.317061831706184, + "grad_norm": 5.293700695037842, + "learning_rate": 9.9799669979967e-06, + "loss": 0.1102, + "step": 20041 + }, + { + "epoch": 9.317526731752674, + "grad_norm": 6.640824794769287, + "learning_rate": 9.9799659979966e-06, + "loss": 0.1218, + "step": 20042 + }, + { + "epoch": 9.317991631799163, + "grad_norm": 2.021589994430542, + "learning_rate": 9.979964997996501e-06, + "loss": 0.0301, + "step": 20043 + }, + { + "epoch": 9.318456531845653, + "grad_norm": 5.96488094329834, + "learning_rate": 9.9799639979964e-06, + "loss": 0.1232, + "step": 20044 + }, + { + "epoch": 9.318921431892143, + "grad_norm": 3.3929996490478516, + "learning_rate": 9.979962997996301e-06, + "loss": 0.0739, + "step": 20045 + }, + { + "epoch": 9.319386331938633, + "grad_norm": 6.283334732055664, + "learning_rate": 9.9799619979962e-06, + "loss": 0.1549, + "step": 20046 + }, + { + "epoch": 9.319851231985123, + "grad_norm": 3.6122570037841797, + "learning_rate": 9.9799609979961e-06, + "loss": 0.0612, + "step": 20047 + }, + { + "epoch": 9.320316132031612, + "grad_norm": 5.628449440002441, + "learning_rate": 9.979959997996001e-06, + "loss": 0.0799, + "step": 20048 + }, + { + "epoch": 9.320781032078104, + "grad_norm": 1.709032654762268, + "learning_rate": 9.9799589979959e-06, + "loss": 0.0123, + "step": 20049 + }, + { + "epoch": 9.321245932124594, + "grad_norm": 3.83565616607666, + "learning_rate": 9.979957997995801e-06, + "loss": 0.0737, + "step": 20050 + }, + { + "epoch": 9.321710832171084, + "grad_norm": 3.276895046234131, + "learning_rate": 9.979956997995702e-06, + "loss": 0.0282, + "step": 20051 + }, + { + "epoch": 9.322175732217573, + "grad_norm": 3.996354818344116, + "learning_rate": 9.9799559979956e-06, + "loss": 0.0447, + "step": 20052 + }, + { + "epoch": 9.322640632264063, + "grad_norm": 2.8213491439819336, + "learning_rate": 9.9799549979955e-06, + "loss": 0.03, + "step": 20053 + }, + { + "epoch": 9.323105532310553, + "grad_norm": 3.43715238571167, + "learning_rate": 9.9799539979954e-06, + "loss": 0.0357, + "step": 20054 + }, + { + "epoch": 9.323570432357043, + "grad_norm": 6.126453876495361, + "learning_rate": 9.9799529979953e-06, + "loss": 0.1365, + "step": 20055 + }, + { + "epoch": 9.324035332403533, + "grad_norm": 6.150527000427246, + "learning_rate": 9.979951997995201e-06, + "loss": 0.0563, + "step": 20056 + }, + { + "epoch": 9.324500232450024, + "grad_norm": 6.421667575836182, + "learning_rate": 9.9799509979951e-06, + "loss": 0.1421, + "step": 20057 + }, + { + "epoch": 9.324965132496514, + "grad_norm": 5.8247809410095215, + "learning_rate": 9.979949997995001e-06, + "loss": 0.131, + "step": 20058 + }, + { + "epoch": 9.325430032543004, + "grad_norm": 6.079697608947754, + "learning_rate": 9.9799489979949e-06, + "loss": 0.1102, + "step": 20059 + }, + { + "epoch": 9.325894932589494, + "grad_norm": 4.090320110321045, + "learning_rate": 9.9799479979948e-06, + "loss": 0.0608, + "step": 20060 + }, + { + "epoch": 9.326359832635983, + "grad_norm": 5.1386942863464355, + "learning_rate": 9.979946997994701e-06, + "loss": 0.0435, + "step": 20061 + }, + { + "epoch": 9.326824732682473, + "grad_norm": 4.740931034088135, + "learning_rate": 9.9799459979946e-06, + "loss": 0.0567, + "step": 20062 + }, + { + "epoch": 9.327289632728963, + "grad_norm": 5.41981315612793, + "learning_rate": 9.979944997994501e-06, + "loss": 0.0998, + "step": 20063 + }, + { + "epoch": 9.327754532775453, + "grad_norm": 6.266648292541504, + "learning_rate": 9.9799439979944e-06, + "loss": 0.1289, + "step": 20064 + }, + { + "epoch": 9.328219432821943, + "grad_norm": 2.4502758979797363, + "learning_rate": 9.9799429979943e-06, + "loss": 0.0342, + "step": 20065 + }, + { + "epoch": 9.328684332868434, + "grad_norm": 7.222737789154053, + "learning_rate": 9.9799419979942e-06, + "loss": 0.1054, + "step": 20066 + }, + { + "epoch": 9.329149232914924, + "grad_norm": 4.347359657287598, + "learning_rate": 9.9799409979941e-06, + "loss": 0.0448, + "step": 20067 + }, + { + "epoch": 9.329614132961414, + "grad_norm": 8.695165634155273, + "learning_rate": 9.979939997994001e-06, + "loss": 0.137, + "step": 20068 + }, + { + "epoch": 9.330079033007904, + "grad_norm": 6.8986077308654785, + "learning_rate": 9.979938997993902e-06, + "loss": 0.1416, + "step": 20069 + }, + { + "epoch": 9.330543933054393, + "grad_norm": 3.758026361465454, + "learning_rate": 9.9799379979938e-06, + "loss": 0.036, + "step": 20070 + }, + { + "epoch": 9.331008833100883, + "grad_norm": 5.771519184112549, + "learning_rate": 9.979936997993701e-06, + "loss": 0.0881, + "step": 20071 + }, + { + "epoch": 9.331473733147373, + "grad_norm": 5.765999794006348, + "learning_rate": 9.9799359979936e-06, + "loss": 0.0543, + "step": 20072 + }, + { + "epoch": 9.331938633193863, + "grad_norm": 4.18481969833374, + "learning_rate": 9.9799349979935e-06, + "loss": 0.0882, + "step": 20073 + }, + { + "epoch": 9.332403533240353, + "grad_norm": 2.9842400550842285, + "learning_rate": 9.979933997993401e-06, + "loss": 0.0503, + "step": 20074 + }, + { + "epoch": 9.332868433286844, + "grad_norm": 4.265804767608643, + "learning_rate": 9.9799329979933e-06, + "loss": 0.0434, + "step": 20075 + }, + { + "epoch": 9.333333333333334, + "grad_norm": 5.844236850738525, + "learning_rate": 9.979931997993201e-06, + "loss": 0.0959, + "step": 20076 + }, + { + "epoch": 9.333798233379824, + "grad_norm": 7.332011699676514, + "learning_rate": 9.9799309979931e-06, + "loss": 0.1985, + "step": 20077 + }, + { + "epoch": 9.334263133426314, + "grad_norm": 3.063608407974243, + "learning_rate": 9.979929997993e-06, + "loss": 0.0381, + "step": 20078 + }, + { + "epoch": 9.334728033472803, + "grad_norm": 2.2091453075408936, + "learning_rate": 9.9799289979929e-06, + "loss": 0.0313, + "step": 20079 + }, + { + "epoch": 9.335192933519293, + "grad_norm": 2.0762441158294678, + "learning_rate": 9.9799279979928e-06, + "loss": 0.0226, + "step": 20080 + }, + { + "epoch": 9.335657833565783, + "grad_norm": 3.6784067153930664, + "learning_rate": 9.979926997992701e-06, + "loss": 0.0543, + "step": 20081 + }, + { + "epoch": 9.336122733612273, + "grad_norm": 8.825428009033203, + "learning_rate": 9.979925997992602e-06, + "loss": 0.1122, + "step": 20082 + }, + { + "epoch": 9.336587633658763, + "grad_norm": 4.403773307800293, + "learning_rate": 9.9799249979925e-06, + "loss": 0.064, + "step": 20083 + }, + { + "epoch": 9.337052533705254, + "grad_norm": 3.963092565536499, + "learning_rate": 9.9799239979924e-06, + "loss": 0.0222, + "step": 20084 + }, + { + "epoch": 9.337517433751744, + "grad_norm": 6.0442214012146, + "learning_rate": 9.9799229979923e-06, + "loss": 0.0747, + "step": 20085 + }, + { + "epoch": 9.337982333798234, + "grad_norm": 3.8486328125, + "learning_rate": 9.979921997992201e-06, + "loss": 0.0866, + "step": 20086 + }, + { + "epoch": 9.338447233844724, + "grad_norm": 2.5220112800598145, + "learning_rate": 9.979920997992102e-06, + "loss": 0.0406, + "step": 20087 + }, + { + "epoch": 9.338912133891213, + "grad_norm": 5.929510593414307, + "learning_rate": 9.979919997992e-06, + "loss": 0.1214, + "step": 20088 + }, + { + "epoch": 9.339377033937703, + "grad_norm": 7.592121124267578, + "learning_rate": 9.979918997991901e-06, + "loss": 0.1094, + "step": 20089 + }, + { + "epoch": 9.339841933984193, + "grad_norm": 5.5914692878723145, + "learning_rate": 9.9799179979918e-06, + "loss": 0.0638, + "step": 20090 + }, + { + "epoch": 9.340306834030683, + "grad_norm": 5.32194185256958, + "learning_rate": 9.9799169979917e-06, + "loss": 0.0775, + "step": 20091 + }, + { + "epoch": 9.340771734077173, + "grad_norm": 4.502990245819092, + "learning_rate": 9.9799159979916e-06, + "loss": 0.0865, + "step": 20092 + }, + { + "epoch": 9.341236634123664, + "grad_norm": 2.172048807144165, + "learning_rate": 9.9799149979915e-06, + "loss": 0.0228, + "step": 20093 + }, + { + "epoch": 9.341701534170154, + "grad_norm": 1.5548356771469116, + "learning_rate": 9.979913997991401e-06, + "loss": 0.0145, + "step": 20094 + }, + { + "epoch": 9.342166434216644, + "grad_norm": 3.996406078338623, + "learning_rate": 9.9799129979913e-06, + "loss": 0.0394, + "step": 20095 + }, + { + "epoch": 9.342631334263134, + "grad_norm": 3.6829593181610107, + "learning_rate": 9.9799119979912e-06, + "loss": 0.0382, + "step": 20096 + }, + { + "epoch": 9.343096234309623, + "grad_norm": 6.879603862762451, + "learning_rate": 9.9799109979911e-06, + "loss": 0.1123, + "step": 20097 + }, + { + "epoch": 9.343561134356113, + "grad_norm": 11.221919059753418, + "learning_rate": 9.979909997991e-06, + "loss": 0.0895, + "step": 20098 + }, + { + "epoch": 9.344026034402603, + "grad_norm": 5.214844226837158, + "learning_rate": 9.979908997990901e-06, + "loss": 0.1262, + "step": 20099 + }, + { + "epoch": 9.344490934449093, + "grad_norm": 6.216524124145508, + "learning_rate": 9.979907997990802e-06, + "loss": 0.0623, + "step": 20100 + }, + { + "epoch": 9.344955834495583, + "grad_norm": 3.0267093181610107, + "learning_rate": 9.9799069979907e-06, + "loss": 0.0357, + "step": 20101 + }, + { + "epoch": 9.345420734542074, + "grad_norm": 3.6551270484924316, + "learning_rate": 9.979905997990601e-06, + "loss": 0.075, + "step": 20102 + }, + { + "epoch": 9.345885634588564, + "grad_norm": 3.534489631652832, + "learning_rate": 9.9799049979905e-06, + "loss": 0.0817, + "step": 20103 + }, + { + "epoch": 9.346350534635054, + "grad_norm": 8.333149909973145, + "learning_rate": 9.9799039979904e-06, + "loss": 0.1952, + "step": 20104 + }, + { + "epoch": 9.346815434681544, + "grad_norm": 7.210169315338135, + "learning_rate": 9.9799029979903e-06, + "loss": 0.1388, + "step": 20105 + }, + { + "epoch": 9.347280334728033, + "grad_norm": 4.44244384765625, + "learning_rate": 9.9799019979902e-06, + "loss": 0.0298, + "step": 20106 + }, + { + "epoch": 9.347745234774523, + "grad_norm": 4.940907001495361, + "learning_rate": 9.979900997990101e-06, + "loss": 0.1277, + "step": 20107 + }, + { + "epoch": 9.348210134821013, + "grad_norm": 4.353408336639404, + "learning_rate": 9.97989999799e-06, + "loss": 0.0397, + "step": 20108 + }, + { + "epoch": 9.348675034867503, + "grad_norm": 6.846587181091309, + "learning_rate": 9.9798989979899e-06, + "loss": 0.1681, + "step": 20109 + }, + { + "epoch": 9.349139934913994, + "grad_norm": 4.699791431427002, + "learning_rate": 9.9798979979898e-06, + "loss": 0.0314, + "step": 20110 + }, + { + "epoch": 9.349604834960484, + "grad_norm": 3.8868181705474854, + "learning_rate": 9.9798969979897e-06, + "loss": 0.0619, + "step": 20111 + }, + { + "epoch": 9.350069735006974, + "grad_norm": 5.385898113250732, + "learning_rate": 9.979895997989601e-06, + "loss": 0.0917, + "step": 20112 + }, + { + "epoch": 9.350534635053464, + "grad_norm": 2.3391385078430176, + "learning_rate": 9.979894997989502e-06, + "loss": 0.0372, + "step": 20113 + }, + { + "epoch": 9.350999535099954, + "grad_norm": 4.969048023223877, + "learning_rate": 9.9798939979894e-06, + "loss": 0.0875, + "step": 20114 + }, + { + "epoch": 9.351464435146443, + "grad_norm": 6.091274738311768, + "learning_rate": 9.979892997989301e-06, + "loss": 0.139, + "step": 20115 + }, + { + "epoch": 9.351929335192933, + "grad_norm": 7.837427616119385, + "learning_rate": 9.9798919979892e-06, + "loss": 0.0653, + "step": 20116 + }, + { + "epoch": 9.352394235239423, + "grad_norm": 5.834654331207275, + "learning_rate": 9.979890997989101e-06, + "loss": 0.0713, + "step": 20117 + }, + { + "epoch": 9.352859135285913, + "grad_norm": 6.567295074462891, + "learning_rate": 9.979889997989e-06, + "loss": 0.1068, + "step": 20118 + }, + { + "epoch": 9.353324035332404, + "grad_norm": 4.197392463684082, + "learning_rate": 9.9798889979889e-06, + "loss": 0.0527, + "step": 20119 + }, + { + "epoch": 9.353788935378894, + "grad_norm": 7.671231746673584, + "learning_rate": 9.979887997988801e-06, + "loss": 0.0948, + "step": 20120 + }, + { + "epoch": 9.354253835425384, + "grad_norm": 5.935216426849365, + "learning_rate": 9.9798869979887e-06, + "loss": 0.1231, + "step": 20121 + }, + { + "epoch": 9.354718735471874, + "grad_norm": 4.529912948608398, + "learning_rate": 9.9798859979886e-06, + "loss": 0.0504, + "step": 20122 + }, + { + "epoch": 9.355183635518364, + "grad_norm": 5.292999267578125, + "learning_rate": 9.9798849979885e-06, + "loss": 0.0736, + "step": 20123 + }, + { + "epoch": 9.355648535564853, + "grad_norm": 5.840792655944824, + "learning_rate": 9.9798839979884e-06, + "loss": 0.0918, + "step": 20124 + }, + { + "epoch": 9.356113435611343, + "grad_norm": 8.331795692443848, + "learning_rate": 9.979882997988301e-06, + "loss": 0.1902, + "step": 20125 + }, + { + "epoch": 9.356578335657833, + "grad_norm": 1.8446249961853027, + "learning_rate": 9.979881997988202e-06, + "loss": 0.0253, + "step": 20126 + }, + { + "epoch": 9.357043235704323, + "grad_norm": 3.5657143592834473, + "learning_rate": 9.9798809979881e-06, + "loss": 0.0702, + "step": 20127 + }, + { + "epoch": 9.357508135750814, + "grad_norm": 9.780330657958984, + "learning_rate": 9.979879997988e-06, + "loss": 0.2533, + "step": 20128 + }, + { + "epoch": 9.357973035797304, + "grad_norm": 5.1128153800964355, + "learning_rate": 9.9798789979879e-06, + "loss": 0.0859, + "step": 20129 + }, + { + "epoch": 9.358437935843794, + "grad_norm": 2.2348439693450928, + "learning_rate": 9.979877997987801e-06, + "loss": 0.0334, + "step": 20130 + }, + { + "epoch": 9.358902835890284, + "grad_norm": 6.444388389587402, + "learning_rate": 9.979876997987702e-06, + "loss": 0.1071, + "step": 20131 + }, + { + "epoch": 9.359367735936774, + "grad_norm": 2.371715784072876, + "learning_rate": 9.9798759979876e-06, + "loss": 0.017, + "step": 20132 + }, + { + "epoch": 9.359832635983263, + "grad_norm": 5.361291885375977, + "learning_rate": 9.979874997987501e-06, + "loss": 0.0986, + "step": 20133 + }, + { + "epoch": 9.360297536029753, + "grad_norm": 3.916482925415039, + "learning_rate": 9.9798739979874e-06, + "loss": 0.0815, + "step": 20134 + }, + { + "epoch": 9.360762436076243, + "grad_norm": 4.495006084442139, + "learning_rate": 9.979872997987301e-06, + "loss": 0.0726, + "step": 20135 + }, + { + "epoch": 9.361227336122734, + "grad_norm": 3.7200119495391846, + "learning_rate": 9.9798719979872e-06, + "loss": 0.0677, + "step": 20136 + }, + { + "epoch": 9.361692236169224, + "grad_norm": 4.254377841949463, + "learning_rate": 9.9798709979871e-06, + "loss": 0.1201, + "step": 20137 + }, + { + "epoch": 9.362157136215714, + "grad_norm": 6.020198345184326, + "learning_rate": 9.979869997987001e-06, + "loss": 0.1734, + "step": 20138 + }, + { + "epoch": 9.362622036262204, + "grad_norm": 4.031200408935547, + "learning_rate": 9.9798689979869e-06, + "loss": 0.046, + "step": 20139 + }, + { + "epoch": 9.363086936308694, + "grad_norm": 4.020484924316406, + "learning_rate": 9.9798679979868e-06, + "loss": 0.046, + "step": 20140 + }, + { + "epoch": 9.363551836355184, + "grad_norm": 3.790379762649536, + "learning_rate": 9.9798669979867e-06, + "loss": 0.0558, + "step": 20141 + }, + { + "epoch": 9.364016736401673, + "grad_norm": 2.779360771179199, + "learning_rate": 9.9798659979866e-06, + "loss": 0.0606, + "step": 20142 + }, + { + "epoch": 9.364481636448163, + "grad_norm": 4.661181926727295, + "learning_rate": 9.979864997986501e-06, + "loss": 0.06, + "step": 20143 + }, + { + "epoch": 9.364946536494653, + "grad_norm": 4.268700122833252, + "learning_rate": 9.979863997986402e-06, + "loss": 0.1111, + "step": 20144 + }, + { + "epoch": 9.365411436541144, + "grad_norm": Infinity, + "learning_rate": 9.979863997986402e-06, + "loss": 0.0249, + "step": 20145 + }, + { + "epoch": 9.365876336587634, + "grad_norm": 4.369718551635742, + "learning_rate": 9.9798629979863e-06, + "loss": 0.0794, + "step": 20146 + }, + { + "epoch": 9.366341236634124, + "grad_norm": 4.232224464416504, + "learning_rate": 9.979861997986201e-06, + "loss": 0.041, + "step": 20147 + }, + { + "epoch": 9.366806136680614, + "grad_norm": 3.642569065093994, + "learning_rate": 9.9798609979861e-06, + "loss": 0.0741, + "step": 20148 + }, + { + "epoch": 9.367271036727104, + "grad_norm": 3.8011081218719482, + "learning_rate": 9.979859997986001e-06, + "loss": 0.0545, + "step": 20149 + }, + { + "epoch": 9.367735936773594, + "grad_norm": 7.958180904388428, + "learning_rate": 9.9798589979859e-06, + "loss": 0.1659, + "step": 20150 + }, + { + "epoch": 9.368200836820083, + "grad_norm": 4.8563032150268555, + "learning_rate": 9.9798579979858e-06, + "loss": 0.072, + "step": 20151 + }, + { + "epoch": 9.368665736866573, + "grad_norm": 4.391470432281494, + "learning_rate": 9.979856997985701e-06, + "loss": 0.109, + "step": 20152 + }, + { + "epoch": 9.369130636913063, + "grad_norm": 5.170679569244385, + "learning_rate": 9.9798559979856e-06, + "loss": 0.0886, + "step": 20153 + }, + { + "epoch": 9.369595536959554, + "grad_norm": 8.679354667663574, + "learning_rate": 9.979854997985501e-06, + "loss": 0.1825, + "step": 20154 + }, + { + "epoch": 9.370060437006044, + "grad_norm": 7.343753337860107, + "learning_rate": 9.9798539979854e-06, + "loss": 0.1637, + "step": 20155 + }, + { + "epoch": 9.370525337052534, + "grad_norm": 6.188129901885986, + "learning_rate": 9.9798529979853e-06, + "loss": 0.1239, + "step": 20156 + }, + { + "epoch": 9.370990237099024, + "grad_norm": 6.866942882537842, + "learning_rate": 9.979851997985201e-06, + "loss": 0.1471, + "step": 20157 + }, + { + "epoch": 9.371455137145514, + "grad_norm": 7.60009765625, + "learning_rate": 9.979850997985102e-06, + "loss": 0.1353, + "step": 20158 + }, + { + "epoch": 9.371920037192004, + "grad_norm": 4.30719518661499, + "learning_rate": 9.979849997985e-06, + "loss": 0.0702, + "step": 20159 + }, + { + "epoch": 9.372384937238493, + "grad_norm": 3.7619991302490234, + "learning_rate": 9.979848997984901e-06, + "loss": 0.0513, + "step": 20160 + }, + { + "epoch": 9.372849837284983, + "grad_norm": 2.2757883071899414, + "learning_rate": 9.9798479979848e-06, + "loss": 0.0206, + "step": 20161 + }, + { + "epoch": 9.373314737331473, + "grad_norm": 6.299409866333008, + "learning_rate": 9.9798469979847e-06, + "loss": 0.0972, + "step": 20162 + }, + { + "epoch": 9.373779637377964, + "grad_norm": 5.518924236297607, + "learning_rate": 9.9798459979846e-06, + "loss": 0.101, + "step": 20163 + }, + { + "epoch": 9.374244537424454, + "grad_norm": 4.121440887451172, + "learning_rate": 9.9798449979845e-06, + "loss": 0.0622, + "step": 20164 + }, + { + "epoch": 9.374709437470944, + "grad_norm": 5.537661552429199, + "learning_rate": 9.979843997984401e-06, + "loss": 0.1054, + "step": 20165 + }, + { + "epoch": 9.375174337517434, + "grad_norm": 7.304667949676514, + "learning_rate": 9.9798429979843e-06, + "loss": 0.1415, + "step": 20166 + }, + { + "epoch": 9.375639237563924, + "grad_norm": 6.873434066772461, + "learning_rate": 9.979841997984201e-06, + "loss": 0.0511, + "step": 20167 + }, + { + "epoch": 9.376104137610414, + "grad_norm": 5.6041693687438965, + "learning_rate": 9.9798409979841e-06, + "loss": 0.1038, + "step": 20168 + }, + { + "epoch": 9.376569037656903, + "grad_norm": 1.2218284606933594, + "learning_rate": 9.979839997984e-06, + "loss": 0.0117, + "step": 20169 + }, + { + "epoch": 9.377033937703393, + "grad_norm": 8.239218711853027, + "learning_rate": 9.979838997983901e-06, + "loss": 0.2033, + "step": 20170 + }, + { + "epoch": 9.377498837749883, + "grad_norm": 6.6976118087768555, + "learning_rate": 9.979837997983802e-06, + "loss": 0.1562, + "step": 20171 + }, + { + "epoch": 9.377963737796374, + "grad_norm": 5.457742214202881, + "learning_rate": 9.979836997983701e-06, + "loss": 0.0673, + "step": 20172 + }, + { + "epoch": 9.378428637842864, + "grad_norm": 6.013774871826172, + "learning_rate": 9.9798359979836e-06, + "loss": 0.0239, + "step": 20173 + }, + { + "epoch": 9.378893537889354, + "grad_norm": 2.738332509994507, + "learning_rate": 9.9798349979835e-06, + "loss": 0.032, + "step": 20174 + }, + { + "epoch": 9.379358437935844, + "grad_norm": 3.337146520614624, + "learning_rate": 9.9798339979834e-06, + "loss": 0.0343, + "step": 20175 + }, + { + "epoch": 9.379823337982334, + "grad_norm": 5.40869665145874, + "learning_rate": 9.9798329979833e-06, + "loss": 0.1039, + "step": 20176 + }, + { + "epoch": 9.380288238028823, + "grad_norm": 4.03650426864624, + "learning_rate": 9.9798319979832e-06, + "loss": 0.0587, + "step": 20177 + }, + { + "epoch": 9.380753138075313, + "grad_norm": 5.495820999145508, + "learning_rate": 9.979830997983101e-06, + "loss": 0.1066, + "step": 20178 + }, + { + "epoch": 9.381218038121803, + "grad_norm": 6.704146862030029, + "learning_rate": 9.979829997983e-06, + "loss": 0.1229, + "step": 20179 + }, + { + "epoch": 9.381682938168293, + "grad_norm": 4.864040374755859, + "learning_rate": 9.979828997982901e-06, + "loss": 0.1254, + "step": 20180 + }, + { + "epoch": 9.382147838214784, + "grad_norm": 7.740162372589111, + "learning_rate": 9.9798279979828e-06, + "loss": 0.3319, + "step": 20181 + }, + { + "epoch": 9.382612738261274, + "grad_norm": 5.140173435211182, + "learning_rate": 9.9798269979827e-06, + "loss": 0.1027, + "step": 20182 + }, + { + "epoch": 9.383077638307764, + "grad_norm": 3.8670990467071533, + "learning_rate": 9.979825997982601e-06, + "loss": 0.0415, + "step": 20183 + }, + { + "epoch": 9.383542538354254, + "grad_norm": 3.6251823902130127, + "learning_rate": 9.9798249979825e-06, + "loss": 0.0338, + "step": 20184 + }, + { + "epoch": 9.384007438400744, + "grad_norm": 4.240868091583252, + "learning_rate": 9.979823997982401e-06, + "loss": 0.0643, + "step": 20185 + }, + { + "epoch": 9.384472338447233, + "grad_norm": 4.147505760192871, + "learning_rate": 9.9798229979823e-06, + "loss": 0.0772, + "step": 20186 + }, + { + "epoch": 9.384937238493723, + "grad_norm": 2.726257085800171, + "learning_rate": 9.9798219979822e-06, + "loss": 0.0211, + "step": 20187 + }, + { + "epoch": 9.385402138540213, + "grad_norm": 1.9204777479171753, + "learning_rate": 9.9798209979821e-06, + "loss": 0.0238, + "step": 20188 + }, + { + "epoch": 9.385867038586705, + "grad_norm": 0.8189762830734253, + "learning_rate": 9.979819997982e-06, + "loss": 0.0075, + "step": 20189 + }, + { + "epoch": 9.386331938633194, + "grad_norm": 3.417762517929077, + "learning_rate": 9.9798189979819e-06, + "loss": 0.071, + "step": 20190 + }, + { + "epoch": 9.386796838679684, + "grad_norm": 5.572618007659912, + "learning_rate": 9.979817997981801e-06, + "loss": 0.0759, + "step": 20191 + }, + { + "epoch": 9.387261738726174, + "grad_norm": 7.183154106140137, + "learning_rate": 9.9798169979817e-06, + "loss": 0.1945, + "step": 20192 + }, + { + "epoch": 9.387726638772664, + "grad_norm": 3.9124598503112793, + "learning_rate": 9.979815997981601e-06, + "loss": 0.0505, + "step": 20193 + }, + { + "epoch": 9.388191538819154, + "grad_norm": 3.6685123443603516, + "learning_rate": 9.9798149979815e-06, + "loss": 0.0365, + "step": 20194 + }, + { + "epoch": 9.388656438865643, + "grad_norm": 6.205410957336426, + "learning_rate": 9.9798139979814e-06, + "loss": 0.1347, + "step": 20195 + }, + { + "epoch": 9.389121338912133, + "grad_norm": 6.079772472381592, + "learning_rate": 9.979812997981301e-06, + "loss": 0.2152, + "step": 20196 + }, + { + "epoch": 9.389586238958623, + "grad_norm": 3.984081506729126, + "learning_rate": 9.9798119979812e-06, + "loss": 0.0818, + "step": 20197 + }, + { + "epoch": 9.390051139005115, + "grad_norm": 3.316812515258789, + "learning_rate": 9.979810997981101e-06, + "loss": 0.08, + "step": 20198 + }, + { + "epoch": 9.390516039051604, + "grad_norm": 5.660280227661133, + "learning_rate": 9.979809997981e-06, + "loss": 0.0482, + "step": 20199 + }, + { + "epoch": 9.390980939098094, + "grad_norm": 5.999267578125, + "learning_rate": 9.9798089979809e-06, + "loss": 0.0513, + "step": 20200 + }, + { + "epoch": 9.391445839144584, + "grad_norm": 3.678046941757202, + "learning_rate": 9.979807997980801e-06, + "loss": 0.0617, + "step": 20201 + }, + { + "epoch": 9.391910739191074, + "grad_norm": 4.765512466430664, + "learning_rate": 9.9798069979807e-06, + "loss": 0.0799, + "step": 20202 + }, + { + "epoch": 9.392375639237564, + "grad_norm": 4.83400821685791, + "learning_rate": 9.979805997980601e-06, + "loss": 0.0698, + "step": 20203 + }, + { + "epoch": 9.392840539284053, + "grad_norm": 3.105853319168091, + "learning_rate": 9.9798049979805e-06, + "loss": 0.0295, + "step": 20204 + }, + { + "epoch": 9.393305439330543, + "grad_norm": 4.158448219299316, + "learning_rate": 9.9798039979804e-06, + "loss": 0.0544, + "step": 20205 + }, + { + "epoch": 9.393770339377033, + "grad_norm": 5.089951515197754, + "learning_rate": 9.9798029979803e-06, + "loss": 0.0864, + "step": 20206 + }, + { + "epoch": 9.394235239423525, + "grad_norm": 9.699658393859863, + "learning_rate": 9.9798019979802e-06, + "loss": 0.146, + "step": 20207 + }, + { + "epoch": 9.394700139470014, + "grad_norm": 7.345245838165283, + "learning_rate": 9.9798009979801e-06, + "loss": 0.0698, + "step": 20208 + }, + { + "epoch": 9.395165039516504, + "grad_norm": 4.729261875152588, + "learning_rate": 9.979799997980001e-06, + "loss": 0.072, + "step": 20209 + }, + { + "epoch": 9.395629939562994, + "grad_norm": 4.3026604652404785, + "learning_rate": 9.9797989979799e-06, + "loss": 0.0549, + "step": 20210 + }, + { + "epoch": 9.396094839609484, + "grad_norm": 3.317540168762207, + "learning_rate": 9.979797997979801e-06, + "loss": 0.065, + "step": 20211 + }, + { + "epoch": 9.396559739655974, + "grad_norm": 4.351119518280029, + "learning_rate": 9.9797969979797e-06, + "loss": 0.0941, + "step": 20212 + }, + { + "epoch": 9.397024639702463, + "grad_norm": 5.26643180847168, + "learning_rate": 9.9797959979796e-06, + "loss": 0.0462, + "step": 20213 + }, + { + "epoch": 9.397489539748953, + "grad_norm": 4.548062801361084, + "learning_rate": 9.979794997979501e-06, + "loss": 0.0653, + "step": 20214 + }, + { + "epoch": 9.397954439795445, + "grad_norm": 2.7090890407562256, + "learning_rate": 9.9797939979794e-06, + "loss": 0.0281, + "step": 20215 + }, + { + "epoch": 9.398419339841935, + "grad_norm": 4.344911098480225, + "learning_rate": 9.979792997979301e-06, + "loss": 0.0255, + "step": 20216 + }, + { + "epoch": 9.398884239888424, + "grad_norm": 1.8936306238174438, + "learning_rate": 9.9797919979792e-06, + "loss": 0.0172, + "step": 20217 + }, + { + "epoch": 9.399349139934914, + "grad_norm": 6.688055515289307, + "learning_rate": 9.9797909979791e-06, + "loss": 0.2574, + "step": 20218 + }, + { + "epoch": 9.399814039981404, + "grad_norm": 6.056885719299316, + "learning_rate": 9.979789997979e-06, + "loss": 0.1365, + "step": 20219 + }, + { + "epoch": 9.400278940027894, + "grad_norm": 4.111739635467529, + "learning_rate": 9.9797889979789e-06, + "loss": 0.0783, + "step": 20220 + }, + { + "epoch": 9.400743840074384, + "grad_norm": 6.603180885314941, + "learning_rate": 9.979787997978801e-06, + "loss": 0.0682, + "step": 20221 + }, + { + "epoch": 9.401208740120873, + "grad_norm": 10.708637237548828, + "learning_rate": 9.979786997978702e-06, + "loss": 0.1379, + "step": 20222 + }, + { + "epoch": 9.401673640167363, + "grad_norm": 8.490116119384766, + "learning_rate": 9.9797859979786e-06, + "loss": 0.1273, + "step": 20223 + }, + { + "epoch": 9.402138540213855, + "grad_norm": 4.267584323883057, + "learning_rate": 9.979784997978501e-06, + "loss": 0.0623, + "step": 20224 + }, + { + "epoch": 9.402603440260345, + "grad_norm": 2.33935809135437, + "learning_rate": 9.9797839979784e-06, + "loss": 0.0318, + "step": 20225 + }, + { + "epoch": 9.403068340306834, + "grad_norm": 1.643907904624939, + "learning_rate": 9.9797829979783e-06, + "loss": 0.0167, + "step": 20226 + }, + { + "epoch": 9.403533240353324, + "grad_norm": 4.7358927726745605, + "learning_rate": 9.979781997978201e-06, + "loss": 0.0667, + "step": 20227 + }, + { + "epoch": 9.403998140399814, + "grad_norm": 1.8729324340820312, + "learning_rate": 9.9797809979781e-06, + "loss": 0.0125, + "step": 20228 + }, + { + "epoch": 9.404463040446304, + "grad_norm": 7.308242321014404, + "learning_rate": 9.979779997978001e-06, + "loss": 0.1653, + "step": 20229 + }, + { + "epoch": 9.404927940492794, + "grad_norm": 6.292740345001221, + "learning_rate": 9.9797789979779e-06, + "loss": 0.191, + "step": 20230 + }, + { + "epoch": 9.405392840539283, + "grad_norm": 4.100765228271484, + "learning_rate": 9.9797779979778e-06, + "loss": 0.0546, + "step": 20231 + }, + { + "epoch": 9.405857740585773, + "grad_norm": 5.541072368621826, + "learning_rate": 9.9797769979777e-06, + "loss": 0.0619, + "step": 20232 + }, + { + "epoch": 9.406322640632265, + "grad_norm": 3.600606918334961, + "learning_rate": 9.9797759979776e-06, + "loss": 0.0381, + "step": 20233 + }, + { + "epoch": 9.406787540678755, + "grad_norm": 7.725646018981934, + "learning_rate": 9.979774997977501e-06, + "loss": 0.2186, + "step": 20234 + }, + { + "epoch": 9.407252440725244, + "grad_norm": 1.5743913650512695, + "learning_rate": 9.979773997977402e-06, + "loss": 0.0133, + "step": 20235 + }, + { + "epoch": 9.407717340771734, + "grad_norm": 4.4710774421691895, + "learning_rate": 9.9797729979773e-06, + "loss": 0.0943, + "step": 20236 + }, + { + "epoch": 9.408182240818224, + "grad_norm": 3.0718345642089844, + "learning_rate": 9.9797719979772e-06, + "loss": 0.0283, + "step": 20237 + }, + { + "epoch": 9.408647140864714, + "grad_norm": 6.6077470779418945, + "learning_rate": 9.9797709979771e-06, + "loss": 0.0945, + "step": 20238 + }, + { + "epoch": 9.409112040911204, + "grad_norm": 2.905893564224243, + "learning_rate": 9.979769997977001e-06, + "loss": 0.0674, + "step": 20239 + }, + { + "epoch": 9.409576940957693, + "grad_norm": 6.662088871002197, + "learning_rate": 9.979768997976902e-06, + "loss": 0.1143, + "step": 20240 + }, + { + "epoch": 9.410041841004183, + "grad_norm": 6.825048923492432, + "learning_rate": 9.9797679979768e-06, + "loss": 0.0974, + "step": 20241 + }, + { + "epoch": 9.410506741050675, + "grad_norm": 3.022554636001587, + "learning_rate": 9.979766997976701e-06, + "loss": 0.0637, + "step": 20242 + }, + { + "epoch": 9.410971641097165, + "grad_norm": 5.941887855529785, + "learning_rate": 9.9797659979766e-06, + "loss": 0.1087, + "step": 20243 + }, + { + "epoch": 9.411436541143654, + "grad_norm": 3.635981559753418, + "learning_rate": 9.9797649979765e-06, + "loss": 0.0604, + "step": 20244 + }, + { + "epoch": 9.411901441190144, + "grad_norm": 4.389438629150391, + "learning_rate": 9.9797639979764e-06, + "loss": 0.0784, + "step": 20245 + }, + { + "epoch": 9.412366341236634, + "grad_norm": 5.09426736831665, + "learning_rate": 9.9797629979763e-06, + "loss": 0.0725, + "step": 20246 + }, + { + "epoch": 9.412831241283124, + "grad_norm": 3.0723259449005127, + "learning_rate": 9.979761997976201e-06, + "loss": 0.0221, + "step": 20247 + }, + { + "epoch": 9.413296141329614, + "grad_norm": 5.711605548858643, + "learning_rate": 9.9797609979761e-06, + "loss": 0.086, + "step": 20248 + }, + { + "epoch": 9.413761041376103, + "grad_norm": 3.361074686050415, + "learning_rate": 9.979759997976e-06, + "loss": 0.0392, + "step": 20249 + }, + { + "epoch": 9.414225941422593, + "grad_norm": 4.088284492492676, + "learning_rate": 9.9797589979759e-06, + "loss": 0.0575, + "step": 20250 + }, + { + "epoch": 9.414690841469085, + "grad_norm": 4.373762607574463, + "learning_rate": 9.9797579979758e-06, + "loss": 0.1079, + "step": 20251 + }, + { + "epoch": 9.415155741515575, + "grad_norm": 4.668537616729736, + "learning_rate": 9.979756997975701e-06, + "loss": 0.0855, + "step": 20252 + }, + { + "epoch": 9.415620641562064, + "grad_norm": 5.442968368530273, + "learning_rate": 9.979755997975602e-06, + "loss": 0.0933, + "step": 20253 + }, + { + "epoch": 9.416085541608554, + "grad_norm": 4.527110576629639, + "learning_rate": 9.9797549979755e-06, + "loss": 0.0472, + "step": 20254 + }, + { + "epoch": 9.416550441655044, + "grad_norm": 2.0782439708709717, + "learning_rate": 9.979753997975401e-06, + "loss": 0.02, + "step": 20255 + }, + { + "epoch": 9.417015341701534, + "grad_norm": 5.849191665649414, + "learning_rate": 9.9797529979753e-06, + "loss": 0.1477, + "step": 20256 + }, + { + "epoch": 9.417480241748024, + "grad_norm": 15.796207427978516, + "learning_rate": 9.979751997975201e-06, + "loss": 0.0788, + "step": 20257 + }, + { + "epoch": 9.417945141794513, + "grad_norm": 3.468893051147461, + "learning_rate": 9.9797509979751e-06, + "loss": 0.0408, + "step": 20258 + }, + { + "epoch": 9.418410041841003, + "grad_norm": 6.866053581237793, + "learning_rate": 9.979749997975e-06, + "loss": 0.0981, + "step": 20259 + }, + { + "epoch": 9.418874941887495, + "grad_norm": 5.885951995849609, + "learning_rate": 9.979748997974901e-06, + "loss": 0.0615, + "step": 20260 + }, + { + "epoch": 9.419339841933985, + "grad_norm": 4.918031692504883, + "learning_rate": 9.9797479979748e-06, + "loss": 0.0675, + "step": 20261 + }, + { + "epoch": 9.419804741980474, + "grad_norm": 3.5044989585876465, + "learning_rate": 9.9797469979747e-06, + "loss": 0.0484, + "step": 20262 + }, + { + "epoch": 9.420269642026964, + "grad_norm": 5.551517963409424, + "learning_rate": 9.9797459979746e-06, + "loss": 0.0886, + "step": 20263 + }, + { + "epoch": 9.420734542073454, + "grad_norm": 7.045402526855469, + "learning_rate": 9.9797449979745e-06, + "loss": 0.1639, + "step": 20264 + }, + { + "epoch": 9.421199442119944, + "grad_norm": 1.8068372011184692, + "learning_rate": 9.979743997974401e-06, + "loss": 0.0094, + "step": 20265 + }, + { + "epoch": 9.421664342166434, + "grad_norm": 6.194275379180908, + "learning_rate": 9.979742997974302e-06, + "loss": 0.1037, + "step": 20266 + }, + { + "epoch": 9.422129242212923, + "grad_norm": 4.72861909866333, + "learning_rate": 9.9797419979742e-06, + "loss": 0.0724, + "step": 20267 + }, + { + "epoch": 9.422594142259415, + "grad_norm": 6.7860589027404785, + "learning_rate": 9.979740997974101e-06, + "loss": 0.1131, + "step": 20268 + }, + { + "epoch": 9.423059042305905, + "grad_norm": 3.113643169403076, + "learning_rate": 9.979739997974e-06, + "loss": 0.0467, + "step": 20269 + }, + { + "epoch": 9.423523942352395, + "grad_norm": 3.969038248062134, + "learning_rate": 9.979738997973901e-06, + "loss": 0.051, + "step": 20270 + }, + { + "epoch": 9.423988842398884, + "grad_norm": 4.637860298156738, + "learning_rate": 9.9797379979738e-06, + "loss": 0.0495, + "step": 20271 + }, + { + "epoch": 9.424453742445374, + "grad_norm": 3.746760606765747, + "learning_rate": 9.9797369979737e-06, + "loss": 0.0812, + "step": 20272 + }, + { + "epoch": 9.424918642491864, + "grad_norm": 2.108286142349243, + "learning_rate": 9.979735997973601e-06, + "loss": 0.0195, + "step": 20273 + }, + { + "epoch": 9.425383542538354, + "grad_norm": 4.948553562164307, + "learning_rate": 9.9797349979735e-06, + "loss": 0.0823, + "step": 20274 + }, + { + "epoch": 9.425848442584844, + "grad_norm": 8.102704048156738, + "learning_rate": 9.979733997973401e-06, + "loss": 0.1658, + "step": 20275 + }, + { + "epoch": 9.426313342631333, + "grad_norm": 1.7542287111282349, + "learning_rate": 9.9797329979733e-06, + "loss": 0.0154, + "step": 20276 + }, + { + "epoch": 9.426778242677825, + "grad_norm": 4.916921138763428, + "learning_rate": 9.9797319979732e-06, + "loss": 0.0998, + "step": 20277 + }, + { + "epoch": 9.427243142724315, + "grad_norm": 1.5503814220428467, + "learning_rate": 9.979730997973101e-06, + "loss": 0.0163, + "step": 20278 + }, + { + "epoch": 9.427708042770805, + "grad_norm": 5.101861000061035, + "learning_rate": 9.979729997973002e-06, + "loss": 0.0801, + "step": 20279 + }, + { + "epoch": 9.428172942817294, + "grad_norm": 3.9999587535858154, + "learning_rate": 9.9797289979729e-06, + "loss": 0.064, + "step": 20280 + }, + { + "epoch": 9.428637842863784, + "grad_norm": 7.240051746368408, + "learning_rate": 9.9797279979728e-06, + "loss": 0.2221, + "step": 20281 + }, + { + "epoch": 9.429102742910274, + "grad_norm": 4.7408127784729, + "learning_rate": 9.9797269979727e-06, + "loss": 0.0953, + "step": 20282 + }, + { + "epoch": 9.429567642956764, + "grad_norm": 7.419839382171631, + "learning_rate": 9.979725997972601e-06, + "loss": 0.1963, + "step": 20283 + }, + { + "epoch": 9.430032543003254, + "grad_norm": 5.8472394943237305, + "learning_rate": 9.9797249979725e-06, + "loss": 0.0909, + "step": 20284 + }, + { + "epoch": 9.430497443049743, + "grad_norm": 5.902915000915527, + "learning_rate": 9.9797239979724e-06, + "loss": 0.1111, + "step": 20285 + }, + { + "epoch": 9.430962343096235, + "grad_norm": 8.219051361083984, + "learning_rate": 9.979722997972301e-06, + "loss": 0.1639, + "step": 20286 + }, + { + "epoch": 9.431427243142725, + "grad_norm": 2.040008306503296, + "learning_rate": 9.9797219979722e-06, + "loss": 0.021, + "step": 20287 + }, + { + "epoch": 9.431892143189215, + "grad_norm": 4.484203338623047, + "learning_rate": 9.979720997972101e-06, + "loss": 0.0723, + "step": 20288 + }, + { + "epoch": 9.432357043235704, + "grad_norm": 3.9426019191741943, + "learning_rate": 9.979719997972e-06, + "loss": 0.0472, + "step": 20289 + }, + { + "epoch": 9.432821943282194, + "grad_norm": 3.9028754234313965, + "learning_rate": 9.9797189979719e-06, + "loss": 0.0691, + "step": 20290 + }, + { + "epoch": 9.433286843328684, + "grad_norm": 5.6300811767578125, + "learning_rate": 9.979717997971801e-06, + "loss": 0.0884, + "step": 20291 + }, + { + "epoch": 9.433751743375174, + "grad_norm": 5.868207931518555, + "learning_rate": 9.9797169979717e-06, + "loss": 0.0993, + "step": 20292 + }, + { + "epoch": 9.434216643421664, + "grad_norm": 1.7205661535263062, + "learning_rate": 9.9797159979716e-06, + "loss": 0.0177, + "step": 20293 + }, + { + "epoch": 9.434681543468155, + "grad_norm": 5.0996294021606445, + "learning_rate": 9.9797149979715e-06, + "loss": 0.0997, + "step": 20294 + }, + { + "epoch": 9.435146443514645, + "grad_norm": 4.373600482940674, + "learning_rate": 9.9797139979714e-06, + "loss": 0.114, + "step": 20295 + }, + { + "epoch": 9.435611343561135, + "grad_norm": 6.567439556121826, + "learning_rate": 9.979712997971301e-06, + "loss": 0.1492, + "step": 20296 + }, + { + "epoch": 9.436076243607625, + "grad_norm": 4.095936298370361, + "learning_rate": 9.979711997971202e-06, + "loss": 0.0648, + "step": 20297 + }, + { + "epoch": 9.436541143654114, + "grad_norm": 3.2719311714172363, + "learning_rate": 9.9797109979711e-06, + "loss": 0.0435, + "step": 20298 + }, + { + "epoch": 9.437006043700604, + "grad_norm": 5.6200666427612305, + "learning_rate": 9.979709997971001e-06, + "loss": 0.1184, + "step": 20299 + }, + { + "epoch": 9.437470943747094, + "grad_norm": 3.4151546955108643, + "learning_rate": 9.9797089979709e-06, + "loss": 0.0386, + "step": 20300 + }, + { + "epoch": 9.437935843793584, + "grad_norm": 3.394785165786743, + "learning_rate": 9.979707997970801e-06, + "loss": 0.0491, + "step": 20301 + }, + { + "epoch": 9.438400743840074, + "grad_norm": 6.326014995574951, + "learning_rate": 9.9797069979707e-06, + "loss": 0.1186, + "step": 20302 + }, + { + "epoch": 9.438865643886565, + "grad_norm": 3.679064989089966, + "learning_rate": 9.9797059979706e-06, + "loss": 0.0586, + "step": 20303 + }, + { + "epoch": 9.439330543933055, + "grad_norm": 6.3820576667785645, + "learning_rate": 9.979704997970501e-06, + "loss": 0.1199, + "step": 20304 + }, + { + "epoch": 9.439795443979545, + "grad_norm": 9.754608154296875, + "learning_rate": 9.9797039979704e-06, + "loss": 0.1066, + "step": 20305 + }, + { + "epoch": 9.440260344026035, + "grad_norm": 4.577972412109375, + "learning_rate": 9.979702997970301e-06, + "loss": 0.1147, + "step": 20306 + }, + { + "epoch": 9.440725244072524, + "grad_norm": 6.437473773956299, + "learning_rate": 9.9797019979702e-06, + "loss": 0.1749, + "step": 20307 + }, + { + "epoch": 9.441190144119014, + "grad_norm": 5.384864330291748, + "learning_rate": 9.9797009979701e-06, + "loss": 0.0987, + "step": 20308 + }, + { + "epoch": 9.441655044165504, + "grad_norm": 7.321316719055176, + "learning_rate": 9.979699997970001e-06, + "loss": 0.1079, + "step": 20309 + }, + { + "epoch": 9.442119944211994, + "grad_norm": 4.202036380767822, + "learning_rate": 9.979698997969902e-06, + "loss": 0.0509, + "step": 20310 + }, + { + "epoch": 9.442584844258484, + "grad_norm": 3.735231637954712, + "learning_rate": 9.9796979979698e-06, + "loss": 0.0434, + "step": 20311 + }, + { + "epoch": 9.443049744304975, + "grad_norm": 10.443071365356445, + "learning_rate": 9.979696997969701e-06, + "loss": 0.2345, + "step": 20312 + }, + { + "epoch": 9.443514644351465, + "grad_norm": 3.4838624000549316, + "learning_rate": 9.9796959979696e-06, + "loss": 0.0424, + "step": 20313 + }, + { + "epoch": 9.443979544397955, + "grad_norm": 3.2631449699401855, + "learning_rate": 9.9796949979695e-06, + "loss": 0.0384, + "step": 20314 + }, + { + "epoch": 9.444444444444445, + "grad_norm": 0.9040576815605164, + "learning_rate": 9.9796939979694e-06, + "loss": 0.0101, + "step": 20315 + }, + { + "epoch": 9.444909344490934, + "grad_norm": 6.881992340087891, + "learning_rate": 9.9796929979693e-06, + "loss": 0.125, + "step": 20316 + }, + { + "epoch": 9.445374244537424, + "grad_norm": 8.86229133605957, + "learning_rate": 9.979691997969201e-06, + "loss": 0.2533, + "step": 20317 + }, + { + "epoch": 9.445839144583914, + "grad_norm": 6.113405227661133, + "learning_rate": 9.9796909979691e-06, + "loss": 0.0946, + "step": 20318 + }, + { + "epoch": 9.446304044630404, + "grad_norm": 8.724205017089844, + "learning_rate": 9.979689997969001e-06, + "loss": 0.3679, + "step": 20319 + }, + { + "epoch": 9.446768944676894, + "grad_norm": 2.908048629760742, + "learning_rate": 9.9796889979689e-06, + "loss": 0.0215, + "step": 20320 + }, + { + "epoch": 9.447233844723385, + "grad_norm": 2.695458173751831, + "learning_rate": 9.9796879979688e-06, + "loss": 0.0282, + "step": 20321 + }, + { + "epoch": 9.447698744769875, + "grad_norm": 3.9298641681671143, + "learning_rate": 9.979686997968701e-06, + "loss": 0.0582, + "step": 20322 + }, + { + "epoch": 9.448163644816365, + "grad_norm": 3.236158609390259, + "learning_rate": 9.979685997968602e-06, + "loss": 0.0535, + "step": 20323 + }, + { + "epoch": 9.448628544862855, + "grad_norm": 4.741156101226807, + "learning_rate": 9.979684997968501e-06, + "loss": 0.0618, + "step": 20324 + }, + { + "epoch": 9.449093444909344, + "grad_norm": 7.223018646240234, + "learning_rate": 9.9796839979684e-06, + "loss": 0.1733, + "step": 20325 + }, + { + "epoch": 9.449558344955834, + "grad_norm": 6.833465576171875, + "learning_rate": 9.9796829979683e-06, + "loss": 0.0974, + "step": 20326 + }, + { + "epoch": 9.450023245002324, + "grad_norm": 3.0766780376434326, + "learning_rate": 9.9796819979682e-06, + "loss": 0.0514, + "step": 20327 + }, + { + "epoch": 9.450488145048814, + "grad_norm": 6.657310485839844, + "learning_rate": 9.9796809979681e-06, + "loss": 0.1135, + "step": 20328 + }, + { + "epoch": 9.450953045095304, + "grad_norm": 6.334453105926514, + "learning_rate": 9.979679997968e-06, + "loss": 0.0817, + "step": 20329 + }, + { + "epoch": 9.451417945141795, + "grad_norm": 5.015437126159668, + "learning_rate": 9.979678997967901e-06, + "loss": 0.0757, + "step": 20330 + }, + { + "epoch": 9.451882845188285, + "grad_norm": 6.7583160400390625, + "learning_rate": 9.9796779979678e-06, + "loss": 0.1233, + "step": 20331 + }, + { + "epoch": 9.452347745234775, + "grad_norm": 5.0479254722595215, + "learning_rate": 9.979676997967701e-06, + "loss": 0.0819, + "step": 20332 + }, + { + "epoch": 9.452812645281265, + "grad_norm": 9.436470985412598, + "learning_rate": 9.9796759979676e-06, + "loss": 0.1886, + "step": 20333 + }, + { + "epoch": 9.453277545327754, + "grad_norm": 7.003776550292969, + "learning_rate": 9.9796749979675e-06, + "loss": 0.1203, + "step": 20334 + }, + { + "epoch": 9.453742445374244, + "grad_norm": 1.950793981552124, + "learning_rate": 9.979673997967401e-06, + "loss": 0.011, + "step": 20335 + }, + { + "epoch": 9.454207345420734, + "grad_norm": 1.7699558734893799, + "learning_rate": 9.9796729979673e-06, + "loss": 0.0281, + "step": 20336 + }, + { + "epoch": 9.454672245467224, + "grad_norm": 4.609472751617432, + "learning_rate": 9.979671997967201e-06, + "loss": 0.0487, + "step": 20337 + }, + { + "epoch": 9.455137145513714, + "grad_norm": 5.5475592613220215, + "learning_rate": 9.9796709979671e-06, + "loss": 0.133, + "step": 20338 + }, + { + "epoch": 9.455602045560205, + "grad_norm": 4.21429967880249, + "learning_rate": 9.979669997967e-06, + "loss": 0.0632, + "step": 20339 + }, + { + "epoch": 9.456066945606695, + "grad_norm": 4.170689105987549, + "learning_rate": 9.9796689979669e-06, + "loss": 0.0675, + "step": 20340 + }, + { + "epoch": 9.456531845653185, + "grad_norm": 5.639909744262695, + "learning_rate": 9.9796679979668e-06, + "loss": 0.0775, + "step": 20341 + }, + { + "epoch": 9.456996745699675, + "grad_norm": 3.6921308040618896, + "learning_rate": 9.979666997966701e-06, + "loss": 0.0551, + "step": 20342 + }, + { + "epoch": 9.457461645746164, + "grad_norm": 3.724360466003418, + "learning_rate": 9.979665997966602e-06, + "loss": 0.0649, + "step": 20343 + }, + { + "epoch": 9.457926545792654, + "grad_norm": 2.2755560874938965, + "learning_rate": 9.9796649979665e-06, + "loss": 0.0293, + "step": 20344 + }, + { + "epoch": 9.458391445839144, + "grad_norm": Infinity, + "learning_rate": 9.9796649979665e-06, + "loss": 0.1225, + "step": 20345 + }, + { + "epoch": 9.458856345885634, + "grad_norm": 4.517966270446777, + "learning_rate": 9.9796639979664e-06, + "loss": 0.0821, + "step": 20346 + }, + { + "epoch": 9.459321245932125, + "grad_norm": 4.373382091522217, + "learning_rate": 9.9796629979663e-06, + "loss": 0.0712, + "step": 20347 + }, + { + "epoch": 9.459786145978615, + "grad_norm": 3.850635051727295, + "learning_rate": 9.9796619979662e-06, + "loss": 0.0381, + "step": 20348 + }, + { + "epoch": 9.460251046025105, + "grad_norm": 3.7568342685699463, + "learning_rate": 9.979660997966101e-06, + "loss": 0.0682, + "step": 20349 + }, + { + "epoch": 9.460715946071595, + "grad_norm": 7.449536323547363, + "learning_rate": 9.979659997966e-06, + "loss": 0.1908, + "step": 20350 + }, + { + "epoch": 9.461180846118085, + "grad_norm": 11.11320972442627, + "learning_rate": 9.979658997965901e-06, + "loss": 0.2224, + "step": 20351 + }, + { + "epoch": 9.461645746164574, + "grad_norm": 3.6387295722961426, + "learning_rate": 9.9796579979658e-06, + "loss": 0.0638, + "step": 20352 + }, + { + "epoch": 9.462110646211064, + "grad_norm": 8.583084106445312, + "learning_rate": 9.9796569979657e-06, + "loss": 0.2587, + "step": 20353 + }, + { + "epoch": 9.462575546257554, + "grad_norm": 4.411869525909424, + "learning_rate": 9.9796559979656e-06, + "loss": 0.0673, + "step": 20354 + }, + { + "epoch": 9.463040446304044, + "grad_norm": 6.296634674072266, + "learning_rate": 9.9796549979655e-06, + "loss": 0.0875, + "step": 20355 + }, + { + "epoch": 9.463505346350535, + "grad_norm": 3.616497755050659, + "learning_rate": 9.979653997965401e-06, + "loss": 0.0791, + "step": 20356 + }, + { + "epoch": 9.463970246397025, + "grad_norm": 8.29781436920166, + "learning_rate": 9.9796529979653e-06, + "loss": 0.1983, + "step": 20357 + }, + { + "epoch": 9.464435146443515, + "grad_norm": 3.714695930480957, + "learning_rate": 9.9796519979652e-06, + "loss": 0.0312, + "step": 20358 + }, + { + "epoch": 9.464900046490005, + "grad_norm": 4.677492141723633, + "learning_rate": 9.9796509979651e-06, + "loss": 0.0891, + "step": 20359 + }, + { + "epoch": 9.465364946536495, + "grad_norm": 9.529094696044922, + "learning_rate": 9.979649997965e-06, + "loss": 0.1629, + "step": 20360 + }, + { + "epoch": 9.465829846582984, + "grad_norm": 3.9355645179748535, + "learning_rate": 9.979648997964901e-06, + "loss": 0.0576, + "step": 20361 + }, + { + "epoch": 9.466294746629474, + "grad_norm": 6.116519451141357, + "learning_rate": 9.979647997964802e-06, + "loss": 0.0897, + "step": 20362 + }, + { + "epoch": 9.466759646675964, + "grad_norm": 4.906045913696289, + "learning_rate": 9.9796469979647e-06, + "loss": 0.0874, + "step": 20363 + }, + { + "epoch": 9.467224546722454, + "grad_norm": 4.010497570037842, + "learning_rate": 9.979645997964601e-06, + "loss": 0.1132, + "step": 20364 + }, + { + "epoch": 9.467689446768945, + "grad_norm": 3.6910552978515625, + "learning_rate": 9.9796449979645e-06, + "loss": 0.0541, + "step": 20365 + }, + { + "epoch": 9.468154346815435, + "grad_norm": 4.64079475402832, + "learning_rate": 9.9796439979644e-06, + "loss": 0.0779, + "step": 20366 + }, + { + "epoch": 9.468619246861925, + "grad_norm": 7.491764068603516, + "learning_rate": 9.979642997964301e-06, + "loss": 0.175, + "step": 20367 + }, + { + "epoch": 9.469084146908415, + "grad_norm": 4.329891204833984, + "learning_rate": 9.9796419979642e-06, + "loss": 0.0328, + "step": 20368 + }, + { + "epoch": 9.469549046954905, + "grad_norm": 8.058801651000977, + "learning_rate": 9.979640997964101e-06, + "loss": 0.0858, + "step": 20369 + }, + { + "epoch": 9.470013947001394, + "grad_norm": 3.4173271656036377, + "learning_rate": 9.979639997964e-06, + "loss": 0.0454, + "step": 20370 + }, + { + "epoch": 9.470478847047884, + "grad_norm": 5.4795403480529785, + "learning_rate": 9.9796389979639e-06, + "loss": 0.1256, + "step": 20371 + }, + { + "epoch": 9.470943747094374, + "grad_norm": 4.465985298156738, + "learning_rate": 9.9796379979638e-06, + "loss": 0.0683, + "step": 20372 + }, + { + "epoch": 9.471408647140866, + "grad_norm": 3.4469492435455322, + "learning_rate": 9.9796369979637e-06, + "loss": 0.0381, + "step": 20373 + }, + { + "epoch": 9.471873547187355, + "grad_norm": 4.52316951751709, + "learning_rate": 9.979635997963601e-06, + "loss": 0.0832, + "step": 20374 + }, + { + "epoch": 9.472338447233845, + "grad_norm": 4.751290798187256, + "learning_rate": 9.979634997963502e-06, + "loss": 0.0673, + "step": 20375 + }, + { + "epoch": 9.472803347280335, + "grad_norm": 1.8758907318115234, + "learning_rate": 9.9796339979634e-06, + "loss": 0.0238, + "step": 20376 + }, + { + "epoch": 9.473268247326825, + "grad_norm": 2.3337721824645996, + "learning_rate": 9.979632997963301e-06, + "loss": 0.0441, + "step": 20377 + }, + { + "epoch": 9.473733147373315, + "grad_norm": 2.7834999561309814, + "learning_rate": 9.9796319979632e-06, + "loss": 0.021, + "step": 20378 + }, + { + "epoch": 9.474198047419804, + "grad_norm": 3.2293589115142822, + "learning_rate": 9.9796309979631e-06, + "loss": 0.0326, + "step": 20379 + }, + { + "epoch": 9.474662947466294, + "grad_norm": 5.455332279205322, + "learning_rate": 9.979629997963001e-06, + "loss": 0.1144, + "step": 20380 + }, + { + "epoch": 9.475127847512784, + "grad_norm": 3.444382667541504, + "learning_rate": 9.9796289979629e-06, + "loss": 0.0601, + "step": 20381 + }, + { + "epoch": 9.475592747559276, + "grad_norm": 5.1238579750061035, + "learning_rate": 9.979627997962801e-06, + "loss": 0.1074, + "step": 20382 + }, + { + "epoch": 9.476057647605765, + "grad_norm": 7.120719909667969, + "learning_rate": 9.9796269979627e-06, + "loss": 0.1675, + "step": 20383 + }, + { + "epoch": 9.476522547652255, + "grad_norm": 3.725196123123169, + "learning_rate": 9.9796259979626e-06, + "loss": 0.0706, + "step": 20384 + }, + { + "epoch": 9.476987447698745, + "grad_norm": 3.703836679458618, + "learning_rate": 9.9796249979625e-06, + "loss": 0.0676, + "step": 20385 + }, + { + "epoch": 9.477452347745235, + "grad_norm": 6.3135881423950195, + "learning_rate": 9.9796239979624e-06, + "loss": 0.1707, + "step": 20386 + }, + { + "epoch": 9.477917247791725, + "grad_norm": 5.325204372406006, + "learning_rate": 9.979622997962301e-06, + "loss": 0.073, + "step": 20387 + }, + { + "epoch": 9.478382147838214, + "grad_norm": 10.194528579711914, + "learning_rate": 9.979621997962202e-06, + "loss": 0.2335, + "step": 20388 + }, + { + "epoch": 9.478847047884704, + "grad_norm": 4.942804336547852, + "learning_rate": 9.9796209979621e-06, + "loss": 0.0829, + "step": 20389 + }, + { + "epoch": 9.479311947931194, + "grad_norm": 4.4607014656066895, + "learning_rate": 9.979619997962e-06, + "loss": 0.0352, + "step": 20390 + }, + { + "epoch": 9.479776847977686, + "grad_norm": 3.662569284439087, + "learning_rate": 9.9796189979619e-06, + "loss": 0.0504, + "step": 20391 + }, + { + "epoch": 9.480241748024175, + "grad_norm": 5.406985759735107, + "learning_rate": 9.979617997961801e-06, + "loss": 0.0515, + "step": 20392 + }, + { + "epoch": 9.480706648070665, + "grad_norm": 5.898081302642822, + "learning_rate": 9.979616997961702e-06, + "loss": 0.1005, + "step": 20393 + }, + { + "epoch": 9.481171548117155, + "grad_norm": 4.361453056335449, + "learning_rate": 9.9796159979616e-06, + "loss": 0.0435, + "step": 20394 + }, + { + "epoch": 9.481636448163645, + "grad_norm": 11.749906539916992, + "learning_rate": 9.979614997961501e-06, + "loss": 0.1073, + "step": 20395 + }, + { + "epoch": 9.482101348210135, + "grad_norm": 4.770657062530518, + "learning_rate": 9.9796139979614e-06, + "loss": 0.1001, + "step": 20396 + }, + { + "epoch": 9.482566248256624, + "grad_norm": 5.383543491363525, + "learning_rate": 9.9796129979613e-06, + "loss": 0.0586, + "step": 20397 + }, + { + "epoch": 9.483031148303114, + "grad_norm": 2.75618577003479, + "learning_rate": 9.9796119979612e-06, + "loss": 0.0201, + "step": 20398 + }, + { + "epoch": 9.483496048349604, + "grad_norm": 1.710109829902649, + "learning_rate": 9.9796109979611e-06, + "loss": 0.0158, + "step": 20399 + }, + { + "epoch": 9.483960948396096, + "grad_norm": 2.92044734954834, + "learning_rate": 9.979609997961001e-06, + "loss": 0.0553, + "step": 20400 + }, + { + "epoch": 9.484425848442585, + "grad_norm": 5.242796897888184, + "learning_rate": 9.9796089979609e-06, + "loss": 0.1394, + "step": 20401 + }, + { + "epoch": 9.484890748489075, + "grad_norm": 13.10559368133545, + "learning_rate": 9.9796079979608e-06, + "loss": 0.2111, + "step": 20402 + }, + { + "epoch": 9.485355648535565, + "grad_norm": 4.011772632598877, + "learning_rate": 9.9796069979607e-06, + "loss": 0.0223, + "step": 20403 + }, + { + "epoch": 9.485820548582055, + "grad_norm": 6.682239055633545, + "learning_rate": 9.9796059979606e-06, + "loss": 0.0825, + "step": 20404 + }, + { + "epoch": 9.486285448628545, + "grad_norm": 7.684099197387695, + "learning_rate": 9.979604997960501e-06, + "loss": 0.1282, + "step": 20405 + }, + { + "epoch": 9.486750348675034, + "grad_norm": 6.665822982788086, + "learning_rate": 9.979603997960402e-06, + "loss": 0.1345, + "step": 20406 + }, + { + "epoch": 9.487215248721524, + "grad_norm": 4.507498741149902, + "learning_rate": 9.9796029979603e-06, + "loss": 0.0801, + "step": 20407 + }, + { + "epoch": 9.487680148768014, + "grad_norm": 10.964547157287598, + "learning_rate": 9.979601997960201e-06, + "loss": 0.1412, + "step": 20408 + }, + { + "epoch": 9.488145048814506, + "grad_norm": 5.5967020988464355, + "learning_rate": 9.9796009979601e-06, + "loss": 0.1089, + "step": 20409 + }, + { + "epoch": 9.488609948860995, + "grad_norm": 1.2703858613967896, + "learning_rate": 9.979599997960001e-06, + "loss": 0.0129, + "step": 20410 + }, + { + "epoch": 9.489074848907485, + "grad_norm": 4.9492268562316895, + "learning_rate": 9.9795989979599e-06, + "loss": 0.0761, + "step": 20411 + }, + { + "epoch": 9.489539748953975, + "grad_norm": 2.8565030097961426, + "learning_rate": 9.9795979979598e-06, + "loss": 0.0244, + "step": 20412 + }, + { + "epoch": 9.490004649000465, + "grad_norm": 5.104780673980713, + "learning_rate": 9.979596997959701e-06, + "loss": 0.0266, + "step": 20413 + }, + { + "epoch": 9.490469549046955, + "grad_norm": 6.926635265350342, + "learning_rate": 9.9795959979596e-06, + "loss": 0.1597, + "step": 20414 + }, + { + "epoch": 9.490934449093444, + "grad_norm": 4.430788993835449, + "learning_rate": 9.9795949979595e-06, + "loss": 0.0851, + "step": 20415 + }, + { + "epoch": 9.491399349139934, + "grad_norm": 9.617772102355957, + "learning_rate": 9.9795939979594e-06, + "loss": 0.2139, + "step": 20416 + }, + { + "epoch": 9.491864249186424, + "grad_norm": 4.382528305053711, + "learning_rate": 9.9795929979593e-06, + "loss": 0.0809, + "step": 20417 + }, + { + "epoch": 9.492329149232916, + "grad_norm": 2.983247756958008, + "learning_rate": 9.979591997959201e-06, + "loss": 0.0566, + "step": 20418 + }, + { + "epoch": 9.492794049279405, + "grad_norm": 3.265531539916992, + "learning_rate": 9.979590997959102e-06, + "loss": 0.0446, + "step": 20419 + }, + { + "epoch": 9.493258949325895, + "grad_norm": 3.3500680923461914, + "learning_rate": 9.979589997959e-06, + "loss": 0.0341, + "step": 20420 + }, + { + "epoch": 9.493723849372385, + "grad_norm": 11.985013961791992, + "learning_rate": 9.979588997958901e-06, + "loss": 0.2587, + "step": 20421 + }, + { + "epoch": 9.494188749418875, + "grad_norm": 6.462430953979492, + "learning_rate": 9.9795879979588e-06, + "loss": 0.1098, + "step": 20422 + }, + { + "epoch": 9.494653649465365, + "grad_norm": 5.013908863067627, + "learning_rate": 9.9795869979587e-06, + "loss": 0.0566, + "step": 20423 + }, + { + "epoch": 9.495118549511854, + "grad_norm": 4.407286167144775, + "learning_rate": 9.9795859979586e-06, + "loss": 0.0811, + "step": 20424 + }, + { + "epoch": 9.495583449558344, + "grad_norm": 5.599973678588867, + "learning_rate": 9.9795849979585e-06, + "loss": 0.0943, + "step": 20425 + }, + { + "epoch": 9.496048349604836, + "grad_norm": 6.220315456390381, + "learning_rate": 9.979583997958401e-06, + "loss": 0.1096, + "step": 20426 + }, + { + "epoch": 9.496513249651326, + "grad_norm": 3.056818723678589, + "learning_rate": 9.9795829979583e-06, + "loss": 0.0296, + "step": 20427 + }, + { + "epoch": 9.496978149697815, + "grad_norm": 3.9594383239746094, + "learning_rate": 9.979581997958201e-06, + "loss": 0.0652, + "step": 20428 + }, + { + "epoch": 9.497443049744305, + "grad_norm": 2.2525885105133057, + "learning_rate": 9.9795809979581e-06, + "loss": 0.0238, + "step": 20429 + }, + { + "epoch": 9.497907949790795, + "grad_norm": 3.9418818950653076, + "learning_rate": 9.979579997958e-06, + "loss": 0.032, + "step": 20430 + }, + { + "epoch": 9.498372849837285, + "grad_norm": 3.978489875793457, + "learning_rate": 9.979578997957901e-06, + "loss": 0.0686, + "step": 20431 + }, + { + "epoch": 9.498837749883775, + "grad_norm": 2.3262827396392822, + "learning_rate": 9.979577997957802e-06, + "loss": 0.0159, + "step": 20432 + }, + { + "epoch": 9.499302649930264, + "grad_norm": 4.7220778465271, + "learning_rate": 9.9795769979577e-06, + "loss": 0.0836, + "step": 20433 + }, + { + "epoch": 9.499767549976754, + "grad_norm": 3.9522058963775635, + "learning_rate": 9.9795759979576e-06, + "loss": 0.0715, + "step": 20434 + }, + { + "epoch": 9.500232450023246, + "grad_norm": 5.399675369262695, + "learning_rate": 9.9795749979575e-06, + "loss": 0.0956, + "step": 20435 + }, + { + "epoch": 9.500697350069736, + "grad_norm": 5.586780071258545, + "learning_rate": 9.979573997957401e-06, + "loss": 0.0651, + "step": 20436 + }, + { + "epoch": 9.501162250116225, + "grad_norm": 5.183099269866943, + "learning_rate": 9.9795729979573e-06, + "loss": 0.1101, + "step": 20437 + }, + { + "epoch": 9.501627150162715, + "grad_norm": 7.169881343841553, + "learning_rate": 9.9795719979572e-06, + "loss": 0.1419, + "step": 20438 + }, + { + "epoch": 9.502092050209205, + "grad_norm": 5.661850452423096, + "learning_rate": 9.979570997957101e-06, + "loss": 0.096, + "step": 20439 + }, + { + "epoch": 9.502556950255695, + "grad_norm": 2.8665175437927246, + "learning_rate": 9.979569997957e-06, + "loss": 0.0577, + "step": 20440 + }, + { + "epoch": 9.503021850302185, + "grad_norm": 3.8270208835601807, + "learning_rate": 9.979568997956901e-06, + "loss": 0.0432, + "step": 20441 + }, + { + "epoch": 9.503486750348674, + "grad_norm": 6.436462879180908, + "learning_rate": 9.9795679979568e-06, + "loss": 0.0746, + "step": 20442 + }, + { + "epoch": 9.503951650395164, + "grad_norm": 2.794004440307617, + "learning_rate": 9.9795669979567e-06, + "loss": 0.0636, + "step": 20443 + }, + { + "epoch": 9.504416550441656, + "grad_norm": 3.6043055057525635, + "learning_rate": 9.979565997956601e-06, + "loss": 0.0724, + "step": 20444 + }, + { + "epoch": 9.504881450488146, + "grad_norm": 3.8442118167877197, + "learning_rate": 9.9795649979565e-06, + "loss": 0.0508, + "step": 20445 + }, + { + "epoch": 9.505346350534635, + "grad_norm": 4.8944244384765625, + "learning_rate": 9.979563997956401e-06, + "loss": 0.0804, + "step": 20446 + }, + { + "epoch": 9.505811250581125, + "grad_norm": 6.394342422485352, + "learning_rate": 9.9795629979563e-06, + "loss": 0.1508, + "step": 20447 + }, + { + "epoch": 9.506276150627615, + "grad_norm": 7.309157848358154, + "learning_rate": 9.9795619979562e-06, + "loss": 0.1627, + "step": 20448 + }, + { + "epoch": 9.506741050674105, + "grad_norm": 3.168649911880493, + "learning_rate": 9.979560997956101e-06, + "loss": 0.0421, + "step": 20449 + }, + { + "epoch": 9.507205950720595, + "grad_norm": 4.498730659484863, + "learning_rate": 9.979559997956e-06, + "loss": 0.0551, + "step": 20450 + }, + { + "epoch": 9.507670850767084, + "grad_norm": 4.934947967529297, + "learning_rate": 9.9795589979559e-06, + "loss": 0.0905, + "step": 20451 + }, + { + "epoch": 9.508135750813576, + "grad_norm": 6.783345699310303, + "learning_rate": 9.979557997955801e-06, + "loss": 0.1273, + "step": 20452 + }, + { + "epoch": 9.508600650860066, + "grad_norm": 1.8104578256607056, + "learning_rate": 9.9795569979557e-06, + "loss": 0.0188, + "step": 20453 + }, + { + "epoch": 9.509065550906556, + "grad_norm": 7.642745018005371, + "learning_rate": 9.979555997955601e-06, + "loss": 0.1076, + "step": 20454 + }, + { + "epoch": 9.509530450953045, + "grad_norm": 5.875036239624023, + "learning_rate": 9.9795549979555e-06, + "loss": 0.1045, + "step": 20455 + }, + { + "epoch": 9.509995350999535, + "grad_norm": 5.491837978363037, + "learning_rate": 9.9795539979554e-06, + "loss": 0.0519, + "step": 20456 + }, + { + "epoch": 9.510460251046025, + "grad_norm": 5.841312408447266, + "learning_rate": 9.979552997955301e-06, + "loss": 0.0584, + "step": 20457 + }, + { + "epoch": 9.510925151092515, + "grad_norm": 2.8122291564941406, + "learning_rate": 9.9795519979552e-06, + "loss": 0.0385, + "step": 20458 + }, + { + "epoch": 9.511390051139005, + "grad_norm": 5.1160759925842285, + "learning_rate": 9.979550997955101e-06, + "loss": 0.0852, + "step": 20459 + }, + { + "epoch": 9.511854951185494, + "grad_norm": 3.775761842727661, + "learning_rate": 9.979549997955e-06, + "loss": 0.054, + "step": 20460 + }, + { + "epoch": 9.512319851231986, + "grad_norm": 5.889350891113281, + "learning_rate": 9.9795489979549e-06, + "loss": 0.1045, + "step": 20461 + }, + { + "epoch": 9.512784751278476, + "grad_norm": 5.280941486358643, + "learning_rate": 9.979547997954801e-06, + "loss": 0.0486, + "step": 20462 + }, + { + "epoch": 9.513249651324966, + "grad_norm": 6.197245121002197, + "learning_rate": 9.979546997954702e-06, + "loss": 0.0923, + "step": 20463 + }, + { + "epoch": 9.513714551371455, + "grad_norm": 3.651949167251587, + "learning_rate": 9.9795459979546e-06, + "loss": 0.0531, + "step": 20464 + }, + { + "epoch": 9.514179451417945, + "grad_norm": 5.542698860168457, + "learning_rate": 9.979544997954502e-06, + "loss": 0.1283, + "step": 20465 + }, + { + "epoch": 9.514644351464435, + "grad_norm": 3.299382448196411, + "learning_rate": 9.9795439979544e-06, + "loss": 0.0369, + "step": 20466 + }, + { + "epoch": 9.515109251510925, + "grad_norm": 3.9607796669006348, + "learning_rate": 9.9795429979543e-06, + "loss": 0.0438, + "step": 20467 + }, + { + "epoch": 9.515574151557415, + "grad_norm": 5.493736267089844, + "learning_rate": 9.9795419979542e-06, + "loss": 0.1041, + "step": 20468 + }, + { + "epoch": 9.516039051603904, + "grad_norm": 5.3571367263793945, + "learning_rate": 9.9795409979541e-06, + "loss": 0.0833, + "step": 20469 + }, + { + "epoch": 9.516503951650396, + "grad_norm": 11.616218566894531, + "learning_rate": 9.979539997954001e-06, + "loss": 0.2403, + "step": 20470 + }, + { + "epoch": 9.516968851696886, + "grad_norm": 3.582010269165039, + "learning_rate": 9.9795389979539e-06, + "loss": 0.0329, + "step": 20471 + }, + { + "epoch": 9.517433751743376, + "grad_norm": 10.4932861328125, + "learning_rate": 9.979537997953801e-06, + "loss": 0.1651, + "step": 20472 + }, + { + "epoch": 9.517898651789865, + "grad_norm": 4.925248622894287, + "learning_rate": 9.9795369979537e-06, + "loss": 0.0856, + "step": 20473 + }, + { + "epoch": 9.518363551836355, + "grad_norm": 4.178468704223633, + "learning_rate": 9.9795359979536e-06, + "loss": 0.0582, + "step": 20474 + }, + { + "epoch": 9.518828451882845, + "grad_norm": 2.848010778427124, + "learning_rate": 9.979534997953501e-06, + "loss": 0.0362, + "step": 20475 + }, + { + "epoch": 9.519293351929335, + "grad_norm": 6.040157794952393, + "learning_rate": 9.979533997953402e-06, + "loss": 0.1255, + "step": 20476 + }, + { + "epoch": 9.519758251975825, + "grad_norm": 4.346401214599609, + "learning_rate": 9.979532997953301e-06, + "loss": 0.0526, + "step": 20477 + }, + { + "epoch": 9.520223152022314, + "grad_norm": 7.596485614776611, + "learning_rate": 9.9795319979532e-06, + "loss": 0.1889, + "step": 20478 + }, + { + "epoch": 9.520688052068806, + "grad_norm": 2.9272241592407227, + "learning_rate": 9.9795309979531e-06, + "loss": 0.0232, + "step": 20479 + }, + { + "epoch": 9.521152952115296, + "grad_norm": 3.6571860313415527, + "learning_rate": 9.979529997953e-06, + "loss": 0.0458, + "step": 20480 + }, + { + "epoch": 9.521617852161786, + "grad_norm": 6.093339443206787, + "learning_rate": 9.9795289979529e-06, + "loss": 0.0416, + "step": 20481 + }, + { + "epoch": 9.522082752208275, + "grad_norm": 5.178135871887207, + "learning_rate": 9.9795279979528e-06, + "loss": 0.0534, + "step": 20482 + }, + { + "epoch": 9.522547652254765, + "grad_norm": 3.768052816390991, + "learning_rate": 9.979526997952701e-06, + "loss": 0.0625, + "step": 20483 + }, + { + "epoch": 9.523012552301255, + "grad_norm": 4.656585693359375, + "learning_rate": 9.9795259979526e-06, + "loss": 0.0833, + "step": 20484 + }, + { + "epoch": 9.523477452347745, + "grad_norm": 3.9633002281188965, + "learning_rate": 9.979524997952501e-06, + "loss": 0.083, + "step": 20485 + }, + { + "epoch": 9.523942352394235, + "grad_norm": 5.362423896789551, + "learning_rate": 9.9795239979524e-06, + "loss": 0.0432, + "step": 20486 + }, + { + "epoch": 9.524407252440724, + "grad_norm": 7.020168781280518, + "learning_rate": 9.9795229979523e-06, + "loss": 0.1129, + "step": 20487 + }, + { + "epoch": 9.524872152487216, + "grad_norm": 6.56282377243042, + "learning_rate": 9.979521997952201e-06, + "loss": 0.161, + "step": 20488 + }, + { + "epoch": 9.525337052533706, + "grad_norm": 3.627819538116455, + "learning_rate": 9.9795209979521e-06, + "loss": 0.0416, + "step": 20489 + }, + { + "epoch": 9.525801952580196, + "grad_norm": 7.079065799713135, + "learning_rate": 9.979519997952001e-06, + "loss": 0.1548, + "step": 20490 + }, + { + "epoch": 9.526266852626685, + "grad_norm": 2.8600757122039795, + "learning_rate": 9.9795189979519e-06, + "loss": 0.0247, + "step": 20491 + }, + { + "epoch": 9.526731752673175, + "grad_norm": 4.32633113861084, + "learning_rate": 9.9795179979518e-06, + "loss": 0.0818, + "step": 20492 + }, + { + "epoch": 9.527196652719665, + "grad_norm": 7.86049222946167, + "learning_rate": 9.9795169979517e-06, + "loss": 0.1709, + "step": 20493 + }, + { + "epoch": 9.527661552766155, + "grad_norm": 12.726987838745117, + "learning_rate": 9.9795159979516e-06, + "loss": 0.2502, + "step": 20494 + }, + { + "epoch": 9.528126452812645, + "grad_norm": 3.2955496311187744, + "learning_rate": 9.979514997951501e-06, + "loss": 0.0495, + "step": 20495 + }, + { + "epoch": 9.528591352859134, + "grad_norm": 2.7613704204559326, + "learning_rate": 9.979513997951402e-06, + "loss": 0.0207, + "step": 20496 + }, + { + "epoch": 9.529056252905626, + "grad_norm": 1.9056552648544312, + "learning_rate": 9.9795129979513e-06, + "loss": 0.0143, + "step": 20497 + }, + { + "epoch": 9.529521152952116, + "grad_norm": 3.0338151454925537, + "learning_rate": 9.9795119979512e-06, + "loss": 0.0521, + "step": 20498 + }, + { + "epoch": 9.529986052998606, + "grad_norm": 7.185194492340088, + "learning_rate": 9.9795109979511e-06, + "loss": 0.1548, + "step": 20499 + }, + { + "epoch": 9.530450953045095, + "grad_norm": 5.634655475616455, + "learning_rate": 9.979509997951e-06, + "loss": 0.0575, + "step": 20500 + }, + { + "epoch": 9.530915853091585, + "grad_norm": 5.572357177734375, + "learning_rate": 9.979508997950901e-06, + "loss": 0.0836, + "step": 20501 + }, + { + "epoch": 9.531380753138075, + "grad_norm": 5.257533073425293, + "learning_rate": 9.9795079979508e-06, + "loss": 0.159, + "step": 20502 + }, + { + "epoch": 9.531845653184565, + "grad_norm": 2.2038660049438477, + "learning_rate": 9.979506997950701e-06, + "loss": 0.0129, + "step": 20503 + }, + { + "epoch": 9.532310553231055, + "grad_norm": 2.0863699913024902, + "learning_rate": 9.9795059979506e-06, + "loss": 0.0225, + "step": 20504 + }, + { + "epoch": 9.532775453277544, + "grad_norm": 6.725516319274902, + "learning_rate": 9.9795049979505e-06, + "loss": 0.0729, + "step": 20505 + }, + { + "epoch": 9.533240353324036, + "grad_norm": 6.7443037033081055, + "learning_rate": 9.9795039979504e-06, + "loss": 0.1068, + "step": 20506 + }, + { + "epoch": 9.533705253370526, + "grad_norm": 6.137807846069336, + "learning_rate": 9.9795029979503e-06, + "loss": 0.2023, + "step": 20507 + }, + { + "epoch": 9.534170153417016, + "grad_norm": 2.380615711212158, + "learning_rate": 9.979501997950201e-06, + "loss": 0.0289, + "step": 20508 + }, + { + "epoch": 9.534635053463505, + "grad_norm": 1.6497437953948975, + "learning_rate": 9.9795009979501e-06, + "loss": 0.0175, + "step": 20509 + }, + { + "epoch": 9.535099953509995, + "grad_norm": 3.4593610763549805, + "learning_rate": 9.97949999795e-06, + "loss": 0.0595, + "step": 20510 + }, + { + "epoch": 9.535564853556485, + "grad_norm": 1.306562066078186, + "learning_rate": 9.9794989979499e-06, + "loss": 0.0146, + "step": 20511 + }, + { + "epoch": 9.536029753602975, + "grad_norm": 3.25449538230896, + "learning_rate": 9.9794979979498e-06, + "loss": 0.0291, + "step": 20512 + }, + { + "epoch": 9.536494653649465, + "grad_norm": 1.7841219902038574, + "learning_rate": 9.979496997949701e-06, + "loss": 0.02, + "step": 20513 + }, + { + "epoch": 9.536959553695956, + "grad_norm": 4.0286736488342285, + "learning_rate": 9.979495997949602e-06, + "loss": 0.0753, + "step": 20514 + }, + { + "epoch": 9.537424453742446, + "grad_norm": 4.62645149230957, + "learning_rate": 9.9794949979495e-06, + "loss": 0.0769, + "step": 20515 + }, + { + "epoch": 9.537889353788936, + "grad_norm": 6.550863742828369, + "learning_rate": 9.979493997949401e-06, + "loss": 0.0981, + "step": 20516 + }, + { + "epoch": 9.538354253835426, + "grad_norm": 5.733936786651611, + "learning_rate": 9.9794929979493e-06, + "loss": 0.13, + "step": 20517 + }, + { + "epoch": 9.538819153881915, + "grad_norm": 6.745248794555664, + "learning_rate": 9.9794919979492e-06, + "loss": 0.2023, + "step": 20518 + }, + { + "epoch": 9.539284053928405, + "grad_norm": 0.9565764665603638, + "learning_rate": 9.9794909979491e-06, + "loss": 0.0102, + "step": 20519 + }, + { + "epoch": 9.539748953974895, + "grad_norm": 1.5577906370162964, + "learning_rate": 9.979489997949e-06, + "loss": 0.012, + "step": 20520 + }, + { + "epoch": 9.540213854021385, + "grad_norm": 4.346102714538574, + "learning_rate": 9.979488997948901e-06, + "loss": 0.1112, + "step": 20521 + }, + { + "epoch": 9.540678754067875, + "grad_norm": 4.886192798614502, + "learning_rate": 9.9794879979488e-06, + "loss": 0.091, + "step": 20522 + }, + { + "epoch": 9.541143654114366, + "grad_norm": 4.415836334228516, + "learning_rate": 9.9794869979487e-06, + "loss": 0.0484, + "step": 20523 + }, + { + "epoch": 9.541608554160856, + "grad_norm": 12.597683906555176, + "learning_rate": 9.9794859979486e-06, + "loss": 0.1356, + "step": 20524 + }, + { + "epoch": 9.542073454207346, + "grad_norm": 6.689177989959717, + "learning_rate": 9.9794849979485e-06, + "loss": 0.0698, + "step": 20525 + }, + { + "epoch": 9.542538354253836, + "grad_norm": 7.937089443206787, + "learning_rate": 9.979483997948401e-06, + "loss": 0.2519, + "step": 20526 + }, + { + "epoch": 9.543003254300325, + "grad_norm": 7.006215572357178, + "learning_rate": 9.979482997948302e-06, + "loss": 0.1288, + "step": 20527 + }, + { + "epoch": 9.543468154346815, + "grad_norm": 8.83655071258545, + "learning_rate": 9.9794819979482e-06, + "loss": 0.1237, + "step": 20528 + }, + { + "epoch": 9.543933054393305, + "grad_norm": 0.9383996725082397, + "learning_rate": 9.979480997948101e-06, + "loss": 0.0069, + "step": 20529 + }, + { + "epoch": 9.544397954439795, + "grad_norm": 3.2222068309783936, + "learning_rate": 9.979479997948e-06, + "loss": 0.0235, + "step": 20530 + }, + { + "epoch": 9.544862854486286, + "grad_norm": 5.22493839263916, + "learning_rate": 9.979478997947901e-06, + "loss": 0.0959, + "step": 20531 + }, + { + "epoch": 9.545327754532776, + "grad_norm": 5.225654125213623, + "learning_rate": 9.979477997947802e-06, + "loss": 0.0526, + "step": 20532 + }, + { + "epoch": 9.545792654579266, + "grad_norm": 4.864455223083496, + "learning_rate": 9.9794769979477e-06, + "loss": 0.0807, + "step": 20533 + }, + { + "epoch": 9.546257554625756, + "grad_norm": 4.45124626159668, + "learning_rate": 9.979475997947601e-06, + "loss": 0.0462, + "step": 20534 + }, + { + "epoch": 9.546722454672246, + "grad_norm": 1.8675843477249146, + "learning_rate": 9.9794749979475e-06, + "loss": 0.0338, + "step": 20535 + }, + { + "epoch": 9.547187354718735, + "grad_norm": 11.311800003051758, + "learning_rate": 9.9794739979474e-06, + "loss": 0.1981, + "step": 20536 + }, + { + "epoch": 9.547652254765225, + "grad_norm": 8.500733375549316, + "learning_rate": 9.9794729979473e-06, + "loss": 0.143, + "step": 20537 + }, + { + "epoch": 9.548117154811715, + "grad_norm": 5.040336608886719, + "learning_rate": 9.9794719979472e-06, + "loss": 0.0622, + "step": 20538 + }, + { + "epoch": 9.548582054858205, + "grad_norm": 2.727766990661621, + "learning_rate": 9.979470997947101e-06, + "loss": 0.0367, + "step": 20539 + }, + { + "epoch": 9.549046954904696, + "grad_norm": 6.97744607925415, + "learning_rate": 9.979469997947002e-06, + "loss": 0.186, + "step": 20540 + }, + { + "epoch": 9.549511854951186, + "grad_norm": 5.5474042892456055, + "learning_rate": 9.9794689979469e-06, + "loss": 0.1341, + "step": 20541 + }, + { + "epoch": 9.549976754997676, + "grad_norm": 4.940154075622559, + "learning_rate": 9.9794679979468e-06, + "loss": 0.1224, + "step": 20542 + }, + { + "epoch": 9.550441655044166, + "grad_norm": 3.5230047702789307, + "learning_rate": 9.9794669979467e-06, + "loss": 0.0462, + "step": 20543 + }, + { + "epoch": 9.550906555090656, + "grad_norm": 3.2856643199920654, + "learning_rate": 9.979465997946601e-06, + "loss": 0.0436, + "step": 20544 + }, + { + "epoch": 9.551371455137145, + "grad_norm": 8.624545097351074, + "learning_rate": 9.979464997946502e-06, + "loss": 0.0827, + "step": 20545 + }, + { + "epoch": 9.551836355183635, + "grad_norm": 6.771326065063477, + "learning_rate": 9.9794639979464e-06, + "loss": 0.0975, + "step": 20546 + }, + { + "epoch": 9.552301255230125, + "grad_norm": 3.0802321434020996, + "learning_rate": 9.979462997946301e-06, + "loss": 0.031, + "step": 20547 + }, + { + "epoch": 9.552766155276615, + "grad_norm": 5.08535623550415, + "learning_rate": 9.9794619979462e-06, + "loss": 0.0548, + "step": 20548 + }, + { + "epoch": 9.553231055323106, + "grad_norm": 7.574275016784668, + "learning_rate": 9.979460997946101e-06, + "loss": 0.145, + "step": 20549 + }, + { + "epoch": 9.553695955369596, + "grad_norm": 4.332849502563477, + "learning_rate": 9.979459997946e-06, + "loss": 0.0712, + "step": 20550 + }, + { + "epoch": 9.554160855416086, + "grad_norm": 3.7803173065185547, + "learning_rate": 9.9794589979459e-06, + "loss": 0.0588, + "step": 20551 + }, + { + "epoch": 9.554625755462576, + "grad_norm": 4.832295894622803, + "learning_rate": 9.979457997945801e-06, + "loss": 0.061, + "step": 20552 + }, + { + "epoch": 9.555090655509066, + "grad_norm": 8.035684585571289, + "learning_rate": 9.9794569979457e-06, + "loss": 0.2543, + "step": 20553 + }, + { + "epoch": 9.555555555555555, + "grad_norm": 5.098377227783203, + "learning_rate": 9.9794559979456e-06, + "loss": 0.064, + "step": 20554 + }, + { + "epoch": 9.556020455602045, + "grad_norm": 2.7503786087036133, + "learning_rate": 9.9794549979455e-06, + "loss": 0.0504, + "step": 20555 + }, + { + "epoch": 9.556485355648535, + "grad_norm": 5.105254650115967, + "learning_rate": 9.9794539979454e-06, + "loss": 0.0779, + "step": 20556 + }, + { + "epoch": 9.556950255695025, + "grad_norm": 2.3206114768981934, + "learning_rate": 9.979452997945301e-06, + "loss": 0.0264, + "step": 20557 + }, + { + "epoch": 9.557415155741516, + "grad_norm": 5.846179485321045, + "learning_rate": 9.979451997945202e-06, + "loss": 0.098, + "step": 20558 + }, + { + "epoch": 9.557880055788006, + "grad_norm": 5.2567057609558105, + "learning_rate": 9.9794509979451e-06, + "loss": 0.0662, + "step": 20559 + }, + { + "epoch": 9.558344955834496, + "grad_norm": 2.4056832790374756, + "learning_rate": 9.979449997945001e-06, + "loss": 0.0259, + "step": 20560 + }, + { + "epoch": 9.558809855880986, + "grad_norm": 5.330965518951416, + "learning_rate": 9.9794489979449e-06, + "loss": 0.1489, + "step": 20561 + }, + { + "epoch": 9.559274755927476, + "grad_norm": 4.059876918792725, + "learning_rate": 9.979447997944801e-06, + "loss": 0.0541, + "step": 20562 + }, + { + "epoch": 9.559739655973965, + "grad_norm": 7.534268379211426, + "learning_rate": 9.9794469979447e-06, + "loss": 0.1285, + "step": 20563 + }, + { + "epoch": 9.560204556020455, + "grad_norm": 9.513983726501465, + "learning_rate": 9.9794459979446e-06, + "loss": 0.1208, + "step": 20564 + }, + { + "epoch": 9.560669456066945, + "grad_norm": 3.1815590858459473, + "learning_rate": 9.979444997944501e-06, + "loss": 0.0335, + "step": 20565 + }, + { + "epoch": 9.561134356113435, + "grad_norm": 4.964013576507568, + "learning_rate": 9.9794439979444e-06, + "loss": 0.0621, + "step": 20566 + }, + { + "epoch": 9.561599256159926, + "grad_norm": 5.923755168914795, + "learning_rate": 9.9794429979443e-06, + "loss": 0.1368, + "step": 20567 + }, + { + "epoch": 9.562064156206416, + "grad_norm": 3.347550630569458, + "learning_rate": 9.9794419979442e-06, + "loss": 0.037, + "step": 20568 + }, + { + "epoch": 9.562529056252906, + "grad_norm": 4.764091491699219, + "learning_rate": 9.9794409979441e-06, + "loss": 0.0446, + "step": 20569 + }, + { + "epoch": 9.562993956299396, + "grad_norm": 3.0713298320770264, + "learning_rate": 9.979439997944001e-06, + "loss": 0.0377, + "step": 20570 + }, + { + "epoch": 9.563458856345886, + "grad_norm": 4.192438125610352, + "learning_rate": 9.979438997943902e-06, + "loss": 0.0652, + "step": 20571 + }, + { + "epoch": 9.563923756392375, + "grad_norm": 3.175269842147827, + "learning_rate": 9.9794379979438e-06, + "loss": 0.0247, + "step": 20572 + }, + { + "epoch": 9.564388656438865, + "grad_norm": 4.955915451049805, + "learning_rate": 9.979436997943701e-06, + "loss": 0.0691, + "step": 20573 + }, + { + "epoch": 9.564853556485355, + "grad_norm": 2.3242247104644775, + "learning_rate": 9.9794359979436e-06, + "loss": 0.0274, + "step": 20574 + }, + { + "epoch": 9.565318456531845, + "grad_norm": 6.944746017456055, + "learning_rate": 9.9794349979435e-06, + "loss": 0.1738, + "step": 20575 + }, + { + "epoch": 9.565783356578336, + "grad_norm": 1.5854462385177612, + "learning_rate": 9.9794339979434e-06, + "loss": 0.0147, + "step": 20576 + }, + { + "epoch": 9.566248256624826, + "grad_norm": 3.8323984146118164, + "learning_rate": 9.9794329979433e-06, + "loss": 0.0464, + "step": 20577 + }, + { + "epoch": 9.566713156671316, + "grad_norm": 5.4579081535339355, + "learning_rate": 9.979431997943201e-06, + "loss": 0.0554, + "step": 20578 + }, + { + "epoch": 9.567178056717806, + "grad_norm": 5.543150901794434, + "learning_rate": 9.9794309979431e-06, + "loss": 0.0575, + "step": 20579 + }, + { + "epoch": 9.567642956764296, + "grad_norm": 3.1966254711151123, + "learning_rate": 9.979429997943001e-06, + "loss": 0.0491, + "step": 20580 + }, + { + "epoch": 9.568107856810785, + "grad_norm": 2.4861948490142822, + "learning_rate": 9.9794289979429e-06, + "loss": 0.0274, + "step": 20581 + }, + { + "epoch": 9.568572756857275, + "grad_norm": 5.759581565856934, + "learning_rate": 9.9794279979428e-06, + "loss": 0.1362, + "step": 20582 + }, + { + "epoch": 9.569037656903765, + "grad_norm": 4.734350204467773, + "learning_rate": 9.979426997942701e-06, + "loss": 0.0866, + "step": 20583 + }, + { + "epoch": 9.569502556950255, + "grad_norm": 4.641755104064941, + "learning_rate": 9.979425997942602e-06, + "loss": 0.0586, + "step": 20584 + }, + { + "epoch": 9.569967456996746, + "grad_norm": 4.9946417808532715, + "learning_rate": 9.9794249979425e-06, + "loss": 0.078, + "step": 20585 + }, + { + "epoch": 9.570432357043236, + "grad_norm": 5.881008625030518, + "learning_rate": 9.9794239979424e-06, + "loss": 0.1044, + "step": 20586 + }, + { + "epoch": 9.570897257089726, + "grad_norm": 3.901582717895508, + "learning_rate": 9.9794229979423e-06, + "loss": 0.0714, + "step": 20587 + }, + { + "epoch": 9.571362157136216, + "grad_norm": 9.291779518127441, + "learning_rate": 9.9794219979422e-06, + "loss": 0.2567, + "step": 20588 + }, + { + "epoch": 9.571827057182706, + "grad_norm": 11.664359092712402, + "learning_rate": 9.9794209979421e-06, + "loss": 0.0828, + "step": 20589 + }, + { + "epoch": 9.572291957229195, + "grad_norm": 5.543436527252197, + "learning_rate": 9.979419997942e-06, + "loss": 0.0714, + "step": 20590 + }, + { + "epoch": 9.572756857275685, + "grad_norm": 4.292616844177246, + "learning_rate": 9.979418997941901e-06, + "loss": 0.0535, + "step": 20591 + }, + { + "epoch": 9.573221757322175, + "grad_norm": 5.180859088897705, + "learning_rate": 9.9794179979418e-06, + "loss": 0.0475, + "step": 20592 + }, + { + "epoch": 9.573686657368667, + "grad_norm": 1.4641152620315552, + "learning_rate": 9.979416997941701e-06, + "loss": 0.0081, + "step": 20593 + }, + { + "epoch": 9.574151557415156, + "grad_norm": 4.264845371246338, + "learning_rate": 9.9794159979416e-06, + "loss": 0.0785, + "step": 20594 + }, + { + "epoch": 9.574616457461646, + "grad_norm": 6.1849493980407715, + "learning_rate": 9.9794149979415e-06, + "loss": 0.0984, + "step": 20595 + }, + { + "epoch": 9.575081357508136, + "grad_norm": 4.199512004852295, + "learning_rate": 9.979413997941401e-06, + "loss": 0.0468, + "step": 20596 + }, + { + "epoch": 9.575546257554626, + "grad_norm": 7.549309253692627, + "learning_rate": 9.9794129979413e-06, + "loss": 0.1341, + "step": 20597 + }, + { + "epoch": 9.576011157601116, + "grad_norm": 7.556981563568115, + "learning_rate": 9.979411997941201e-06, + "loss": 0.1503, + "step": 20598 + }, + { + "epoch": 9.576476057647605, + "grad_norm": 7.7489166259765625, + "learning_rate": 9.9794109979411e-06, + "loss": 0.1523, + "step": 20599 + }, + { + "epoch": 9.576940957694095, + "grad_norm": 5.615400314331055, + "learning_rate": 9.979409997941e-06, + "loss": 0.0991, + "step": 20600 + }, + { + "epoch": 9.577405857740585, + "grad_norm": 8.173605918884277, + "learning_rate": 9.979408997940901e-06, + "loss": 0.2848, + "step": 20601 + }, + { + "epoch": 9.577870757787077, + "grad_norm": 3.8608596324920654, + "learning_rate": 9.9794079979408e-06, + "loss": 0.0503, + "step": 20602 + }, + { + "epoch": 9.578335657833566, + "grad_norm": 5.068922996520996, + "learning_rate": 9.9794069979407e-06, + "loss": 0.09, + "step": 20603 + }, + { + "epoch": 9.578800557880056, + "grad_norm": 5.949594974517822, + "learning_rate": 9.979405997940601e-06, + "loss": 0.1509, + "step": 20604 + }, + { + "epoch": 9.579265457926546, + "grad_norm": 3.449892997741699, + "learning_rate": 9.9794049979405e-06, + "loss": 0.0319, + "step": 20605 + }, + { + "epoch": 9.579730357973036, + "grad_norm": 8.316824913024902, + "learning_rate": 9.979403997940401e-06, + "loss": 0.1352, + "step": 20606 + }, + { + "epoch": 9.580195258019526, + "grad_norm": 3.737591028213501, + "learning_rate": 9.9794029979403e-06, + "loss": 0.0555, + "step": 20607 + }, + { + "epoch": 9.580660158066015, + "grad_norm": 5.5707807540893555, + "learning_rate": 9.9794019979402e-06, + "loss": 0.1108, + "step": 20608 + }, + { + "epoch": 9.581125058112505, + "grad_norm": 3.9364891052246094, + "learning_rate": 9.979400997940101e-06, + "loss": 0.0684, + "step": 20609 + }, + { + "epoch": 9.581589958158997, + "grad_norm": 2.2677221298217773, + "learning_rate": 9.97939999794e-06, + "loss": 0.0235, + "step": 20610 + }, + { + "epoch": 9.582054858205487, + "grad_norm": 8.696162223815918, + "learning_rate": 9.979398997939901e-06, + "loss": 0.0615, + "step": 20611 + }, + { + "epoch": 9.582519758251976, + "grad_norm": 2.9066431522369385, + "learning_rate": 9.9793979979398e-06, + "loss": 0.0349, + "step": 20612 + }, + { + "epoch": 9.582984658298466, + "grad_norm": 4.393418788909912, + "learning_rate": 9.9793969979397e-06, + "loss": 0.0402, + "step": 20613 + }, + { + "epoch": 9.583449558344956, + "grad_norm": 2.1062803268432617, + "learning_rate": 9.979395997939601e-06, + "loss": 0.0259, + "step": 20614 + }, + { + "epoch": 9.583914458391446, + "grad_norm": 11.048276901245117, + "learning_rate": 9.9793949979395e-06, + "loss": 0.1059, + "step": 20615 + }, + { + "epoch": 9.584379358437936, + "grad_norm": 2.1133873462677, + "learning_rate": 9.979393997939401e-06, + "loss": 0.0412, + "step": 20616 + }, + { + "epoch": 9.584844258484425, + "grad_norm": 5.038806438446045, + "learning_rate": 9.979392997939302e-06, + "loss": 0.0886, + "step": 20617 + }, + { + "epoch": 9.585309158530915, + "grad_norm": 1.9455746412277222, + "learning_rate": 9.9793919979392e-06, + "loss": 0.0219, + "step": 20618 + }, + { + "epoch": 9.585774058577407, + "grad_norm": 3.2187747955322266, + "learning_rate": 9.9793909979391e-06, + "loss": 0.0396, + "step": 20619 + }, + { + "epoch": 9.586238958623897, + "grad_norm": 6.843984603881836, + "learning_rate": 9.979389997939e-06, + "loss": 0.1391, + "step": 20620 + }, + { + "epoch": 9.586703858670386, + "grad_norm": 7.365050792694092, + "learning_rate": 9.9793889979389e-06, + "loss": 0.1349, + "step": 20621 + }, + { + "epoch": 9.587168758716876, + "grad_norm": 4.017827033996582, + "learning_rate": 9.979387997938801e-06, + "loss": 0.0579, + "step": 20622 + }, + { + "epoch": 9.587633658763366, + "grad_norm": 3.7648656368255615, + "learning_rate": 9.9793869979387e-06, + "loss": 0.0496, + "step": 20623 + }, + { + "epoch": 9.588098558809856, + "grad_norm": 6.62831974029541, + "learning_rate": 9.979385997938601e-06, + "loss": 0.1298, + "step": 20624 + }, + { + "epoch": 9.588563458856346, + "grad_norm": 5.541071891784668, + "learning_rate": 9.9793849979385e-06, + "loss": 0.1187, + "step": 20625 + }, + { + "epoch": 9.589028358902835, + "grad_norm": 6.2743120193481445, + "learning_rate": 9.9793839979384e-06, + "loss": 0.1917, + "step": 20626 + }, + { + "epoch": 9.589493258949325, + "grad_norm": 6.076845645904541, + "learning_rate": 9.979382997938301e-06, + "loss": 0.142, + "step": 20627 + }, + { + "epoch": 9.589958158995817, + "grad_norm": 4.002054691314697, + "learning_rate": 9.9793819979382e-06, + "loss": 0.0579, + "step": 20628 + }, + { + "epoch": 9.590423059042307, + "grad_norm": 4.847428321838379, + "learning_rate": 9.979380997938101e-06, + "loss": 0.0815, + "step": 20629 + }, + { + "epoch": 9.590887959088796, + "grad_norm": 5.421841144561768, + "learning_rate": 9.979379997938e-06, + "loss": 0.035, + "step": 20630 + }, + { + "epoch": 9.591352859135286, + "grad_norm": 1.7620859146118164, + "learning_rate": 9.9793789979379e-06, + "loss": 0.0151, + "step": 20631 + }, + { + "epoch": 9.591817759181776, + "grad_norm": 4.034934997558594, + "learning_rate": 9.9793779979378e-06, + "loss": 0.0429, + "step": 20632 + }, + { + "epoch": 9.592282659228266, + "grad_norm": 4.004919052124023, + "learning_rate": 9.9793769979377e-06, + "loss": 0.0731, + "step": 20633 + }, + { + "epoch": 9.592747559274756, + "grad_norm": 2.575517177581787, + "learning_rate": 9.979375997937601e-06, + "loss": 0.0323, + "step": 20634 + }, + { + "epoch": 9.593212459321245, + "grad_norm": 5.445545673370361, + "learning_rate": 9.979374997937502e-06, + "loss": 0.0604, + "step": 20635 + }, + { + "epoch": 9.593677359367735, + "grad_norm": 8.616621971130371, + "learning_rate": 9.9793739979374e-06, + "loss": 0.14, + "step": 20636 + }, + { + "epoch": 9.594142259414227, + "grad_norm": 9.991573333740234, + "learning_rate": 9.979372997937301e-06, + "loss": 0.2066, + "step": 20637 + }, + { + "epoch": 9.594607159460717, + "grad_norm": 3.6268069744110107, + "learning_rate": 9.9793719979372e-06, + "loss": 0.0342, + "step": 20638 + }, + { + "epoch": 9.595072059507206, + "grad_norm": 6.534521579742432, + "learning_rate": 9.9793709979371e-06, + "loss": 0.0912, + "step": 20639 + }, + { + "epoch": 9.595536959553696, + "grad_norm": 4.697718620300293, + "learning_rate": 9.979369997937001e-06, + "loss": 0.075, + "step": 20640 + }, + { + "epoch": 9.596001859600186, + "grad_norm": 6.361972808837891, + "learning_rate": 9.9793689979369e-06, + "loss": 0.0905, + "step": 20641 + }, + { + "epoch": 9.596466759646676, + "grad_norm": 2.855814218521118, + "learning_rate": 9.979367997936801e-06, + "loss": 0.0261, + "step": 20642 + }, + { + "epoch": 9.596931659693166, + "grad_norm": 3.2444396018981934, + "learning_rate": 9.9793669979367e-06, + "loss": 0.0371, + "step": 20643 + }, + { + "epoch": 9.597396559739655, + "grad_norm": 4.530874252319336, + "learning_rate": 9.9793659979366e-06, + "loss": 0.0447, + "step": 20644 + }, + { + "epoch": 9.597861459786145, + "grad_norm": 3.8857309818267822, + "learning_rate": 9.9793649979365e-06, + "loss": 0.0518, + "step": 20645 + }, + { + "epoch": 9.598326359832637, + "grad_norm": 4.795536994934082, + "learning_rate": 9.9793639979364e-06, + "loss": 0.0378, + "step": 20646 + }, + { + "epoch": 9.598791259879127, + "grad_norm": 4.572007656097412, + "learning_rate": 9.979362997936301e-06, + "loss": 0.0605, + "step": 20647 + }, + { + "epoch": 9.599256159925616, + "grad_norm": 6.666210651397705, + "learning_rate": 9.979361997936202e-06, + "loss": 0.1291, + "step": 20648 + }, + { + "epoch": 9.599721059972106, + "grad_norm": 3.6050164699554443, + "learning_rate": 9.9793609979361e-06, + "loss": 0.037, + "step": 20649 + }, + { + "epoch": 9.600185960018596, + "grad_norm": 6.093271255493164, + "learning_rate": 9.979359997936e-06, + "loss": 0.141, + "step": 20650 + }, + { + "epoch": 9.600650860065086, + "grad_norm": 5.80775785446167, + "learning_rate": 9.9793589979359e-06, + "loss": 0.108, + "step": 20651 + }, + { + "epoch": 9.601115760111576, + "grad_norm": 2.011775016784668, + "learning_rate": 9.9793579979358e-06, + "loss": 0.0179, + "step": 20652 + }, + { + "epoch": 9.601580660158065, + "grad_norm": 2.2585244178771973, + "learning_rate": 9.979356997935701e-06, + "loss": 0.0126, + "step": 20653 + }, + { + "epoch": 9.602045560204555, + "grad_norm": 0.9392969012260437, + "learning_rate": 9.9793559979356e-06, + "loss": 0.0072, + "step": 20654 + }, + { + "epoch": 9.602510460251047, + "grad_norm": 4.461960315704346, + "learning_rate": 9.979354997935501e-06, + "loss": 0.0563, + "step": 20655 + }, + { + "epoch": 9.602975360297537, + "grad_norm": 3.6255125999450684, + "learning_rate": 9.9793539979354e-06, + "loss": 0.0255, + "step": 20656 + }, + { + "epoch": 9.603440260344026, + "grad_norm": 4.412420749664307, + "learning_rate": 9.9793529979353e-06, + "loss": 0.0519, + "step": 20657 + }, + { + "epoch": 9.603905160390516, + "grad_norm": 3.6820805072784424, + "learning_rate": 9.9793519979352e-06, + "loss": 0.0619, + "step": 20658 + }, + { + "epoch": 9.604370060437006, + "grad_norm": 7.884833335876465, + "learning_rate": 9.9793509979351e-06, + "loss": 0.1061, + "step": 20659 + }, + { + "epoch": 9.604834960483496, + "grad_norm": 6.031933307647705, + "learning_rate": 9.979349997935001e-06, + "loss": 0.129, + "step": 20660 + }, + { + "epoch": 9.605299860529986, + "grad_norm": 7.337238788604736, + "learning_rate": 9.9793489979349e-06, + "loss": 0.1508, + "step": 20661 + }, + { + "epoch": 9.605764760576475, + "grad_norm": 4.356525897979736, + "learning_rate": 9.9793479979348e-06, + "loss": 0.0679, + "step": 20662 + }, + { + "epoch": 9.606229660622965, + "grad_norm": 5.070032596588135, + "learning_rate": 9.9793469979347e-06, + "loss": 0.0879, + "step": 20663 + }, + { + "epoch": 9.606694560669457, + "grad_norm": 4.284824371337891, + "learning_rate": 9.9793459979346e-06, + "loss": 0.043, + "step": 20664 + }, + { + "epoch": 9.607159460715947, + "grad_norm": 5.388258934020996, + "learning_rate": 9.979344997934501e-06, + "loss": 0.0579, + "step": 20665 + }, + { + "epoch": 9.607624360762436, + "grad_norm": 4.205082893371582, + "learning_rate": 9.979343997934402e-06, + "loss": 0.0519, + "step": 20666 + }, + { + "epoch": 9.608089260808926, + "grad_norm": 3.095858335494995, + "learning_rate": 9.9793429979343e-06, + "loss": 0.0223, + "step": 20667 + }, + { + "epoch": 9.608554160855416, + "grad_norm": 5.027270317077637, + "learning_rate": 9.979341997934201e-06, + "loss": 0.1017, + "step": 20668 + }, + { + "epoch": 9.609019060901906, + "grad_norm": 5.081347942352295, + "learning_rate": 9.9793409979341e-06, + "loss": 0.1044, + "step": 20669 + }, + { + "epoch": 9.609483960948396, + "grad_norm": 9.017362594604492, + "learning_rate": 9.979339997934e-06, + "loss": 0.1323, + "step": 20670 + }, + { + "epoch": 9.609948860994885, + "grad_norm": 3.938788652420044, + "learning_rate": 9.9793389979339e-06, + "loss": 0.0537, + "step": 20671 + }, + { + "epoch": 9.610413761041377, + "grad_norm": 5.406467437744141, + "learning_rate": 9.9793379979338e-06, + "loss": 0.0474, + "step": 20672 + }, + { + "epoch": 9.610878661087867, + "grad_norm": 5.642330646514893, + "learning_rate": 9.979336997933701e-06, + "loss": 0.0621, + "step": 20673 + }, + { + "epoch": 9.611343561134357, + "grad_norm": 5.166611194610596, + "learning_rate": 9.9793359979336e-06, + "loss": 0.0942, + "step": 20674 + }, + { + "epoch": 9.611808461180846, + "grad_norm": 6.108211994171143, + "learning_rate": 9.9793349979335e-06, + "loss": 0.1683, + "step": 20675 + }, + { + "epoch": 9.612273361227336, + "grad_norm": 0.6562756896018982, + "learning_rate": 9.9793339979334e-06, + "loss": 0.0059, + "step": 20676 + }, + { + "epoch": 9.612738261273826, + "grad_norm": 3.784452438354492, + "learning_rate": 9.9793329979333e-06, + "loss": 0.0686, + "step": 20677 + }, + { + "epoch": 9.613203161320316, + "grad_norm": 4.983392715454102, + "learning_rate": 9.979331997933201e-06, + "loss": 0.0683, + "step": 20678 + }, + { + "epoch": 9.613668061366806, + "grad_norm": 5.375589370727539, + "learning_rate": 9.979330997933102e-06, + "loss": 0.0653, + "step": 20679 + }, + { + "epoch": 9.614132961413295, + "grad_norm": 4.310689449310303, + "learning_rate": 9.979329997933e-06, + "loss": 0.0888, + "step": 20680 + }, + { + "epoch": 9.614597861459787, + "grad_norm": 5.820858001708984, + "learning_rate": 9.979328997932901e-06, + "loss": 0.0899, + "step": 20681 + }, + { + "epoch": 9.615062761506277, + "grad_norm": 6.73173189163208, + "learning_rate": 9.9793279979328e-06, + "loss": 0.086, + "step": 20682 + }, + { + "epoch": 9.615527661552767, + "grad_norm": 7.0945048332214355, + "learning_rate": 9.979326997932701e-06, + "loss": 0.1105, + "step": 20683 + }, + { + "epoch": 9.615992561599256, + "grad_norm": 18.5567626953125, + "learning_rate": 9.9793259979326e-06, + "loss": 0.9512, + "step": 20684 + }, + { + "epoch": 9.616457461645746, + "grad_norm": 4.9650421142578125, + "learning_rate": 9.9793249979325e-06, + "loss": 0.0823, + "step": 20685 + }, + { + "epoch": 9.616922361692236, + "grad_norm": 5.223337173461914, + "learning_rate": 9.979323997932401e-06, + "loss": 0.1055, + "step": 20686 + }, + { + "epoch": 9.617387261738726, + "grad_norm": 4.992422103881836, + "learning_rate": 9.9793229979323e-06, + "loss": 0.1144, + "step": 20687 + }, + { + "epoch": 9.617852161785216, + "grad_norm": 2.5113365650177, + "learning_rate": 9.9793219979322e-06, + "loss": 0.036, + "step": 20688 + }, + { + "epoch": 9.618317061831707, + "grad_norm": 6.927366733551025, + "learning_rate": 9.9793209979321e-06, + "loss": 0.1646, + "step": 20689 + }, + { + "epoch": 9.618781961878197, + "grad_norm": 6.749845027923584, + "learning_rate": 9.979319997932e-06, + "loss": 0.1174, + "step": 20690 + }, + { + "epoch": 9.619246861924687, + "grad_norm": 3.8116366863250732, + "learning_rate": 9.979318997931901e-06, + "loss": 0.0522, + "step": 20691 + }, + { + "epoch": 9.619711761971177, + "grad_norm": 6.3159894943237305, + "learning_rate": 9.979317997931802e-06, + "loss": 0.1187, + "step": 20692 + }, + { + "epoch": 9.620176662017666, + "grad_norm": 7.048880577087402, + "learning_rate": 9.9793169979317e-06, + "loss": 0.1574, + "step": 20693 + }, + { + "epoch": 9.620641562064156, + "grad_norm": 2.3968396186828613, + "learning_rate": 9.9793159979316e-06, + "loss": 0.0213, + "step": 20694 + }, + { + "epoch": 9.621106462110646, + "grad_norm": 5.2670063972473145, + "learning_rate": 9.9793149979315e-06, + "loss": 0.1209, + "step": 20695 + }, + { + "epoch": 9.621571362157136, + "grad_norm": 4.814769744873047, + "learning_rate": 9.979313997931401e-06, + "loss": 0.0649, + "step": 20696 + }, + { + "epoch": 9.622036262203626, + "grad_norm": 2.807331085205078, + "learning_rate": 9.979312997931302e-06, + "loss": 0.0282, + "step": 20697 + }, + { + "epoch": 9.622501162250117, + "grad_norm": 6.444860458374023, + "learning_rate": 9.9793119979312e-06, + "loss": 0.0641, + "step": 20698 + }, + { + "epoch": 9.622966062296607, + "grad_norm": 3.010237455368042, + "learning_rate": 9.979310997931101e-06, + "loss": 0.0416, + "step": 20699 + }, + { + "epoch": 9.623430962343097, + "grad_norm": 7.0552802085876465, + "learning_rate": 9.979309997931e-06, + "loss": 0.1805, + "step": 20700 + }, + { + "epoch": 9.623895862389586, + "grad_norm": 7.730175971984863, + "learning_rate": 9.979308997930901e-06, + "loss": 0.1898, + "step": 20701 + }, + { + "epoch": 9.624360762436076, + "grad_norm": 11.176474571228027, + "learning_rate": 9.9793079979308e-06, + "loss": 0.2817, + "step": 20702 + }, + { + "epoch": 9.624825662482566, + "grad_norm": 2.8435277938842773, + "learning_rate": 9.9793069979307e-06, + "loss": 0.0188, + "step": 20703 + }, + { + "epoch": 9.625290562529056, + "grad_norm": 4.279185771942139, + "learning_rate": 9.979305997930601e-06, + "loss": 0.0654, + "step": 20704 + }, + { + "epoch": 9.625755462575546, + "grad_norm": 8.460854530334473, + "learning_rate": 9.9793049979305e-06, + "loss": 0.2628, + "step": 20705 + }, + { + "epoch": 9.626220362622036, + "grad_norm": 9.617337226867676, + "learning_rate": 9.9793039979304e-06, + "loss": 0.3277, + "step": 20706 + }, + { + "epoch": 9.626685262668527, + "grad_norm": 0.6839496493339539, + "learning_rate": 9.9793029979303e-06, + "loss": 0.0062, + "step": 20707 + }, + { + "epoch": 9.627150162715017, + "grad_norm": 4.732081413269043, + "learning_rate": 9.9793019979302e-06, + "loss": 0.0931, + "step": 20708 + }, + { + "epoch": 9.627615062761507, + "grad_norm": 6.977292060852051, + "learning_rate": 9.979300997930101e-06, + "loss": 0.0499, + "step": 20709 + }, + { + "epoch": 9.628079962807996, + "grad_norm": 5.988259315490723, + "learning_rate": 9.979299997930002e-06, + "loss": 0.1134, + "step": 20710 + }, + { + "epoch": 9.628544862854486, + "grad_norm": 6.737338542938232, + "learning_rate": 9.9792989979299e-06, + "loss": 0.1433, + "step": 20711 + }, + { + "epoch": 9.629009762900976, + "grad_norm": 5.860995769500732, + "learning_rate": 9.979297997929801e-06, + "loss": 0.0982, + "step": 20712 + }, + { + "epoch": 9.629474662947466, + "grad_norm": 4.194870471954346, + "learning_rate": 9.9792969979297e-06, + "loss": 0.0905, + "step": 20713 + }, + { + "epoch": 9.629939562993956, + "grad_norm": 4.692971706390381, + "learning_rate": 9.979295997929601e-06, + "loss": 0.0785, + "step": 20714 + }, + { + "epoch": 9.630404463040446, + "grad_norm": 2.6446802616119385, + "learning_rate": 9.9792949979295e-06, + "loss": 0.029, + "step": 20715 + }, + { + "epoch": 9.630869363086937, + "grad_norm": 7.927891254425049, + "learning_rate": 9.9792939979294e-06, + "loss": 0.2268, + "step": 20716 + }, + { + "epoch": 9.631334263133427, + "grad_norm": 4.992774486541748, + "learning_rate": 9.979292997929301e-06, + "loss": 0.1727, + "step": 20717 + }, + { + "epoch": 9.631799163179917, + "grad_norm": 4.803395748138428, + "learning_rate": 9.9792919979292e-06, + "loss": 0.0634, + "step": 20718 + }, + { + "epoch": 9.632264063226406, + "grad_norm": 5.755645751953125, + "learning_rate": 9.979290997929101e-06, + "loss": 0.1193, + "step": 20719 + }, + { + "epoch": 9.632728963272896, + "grad_norm": 2.69214129447937, + "learning_rate": 9.979289997929e-06, + "loss": 0.034, + "step": 20720 + }, + { + "epoch": 9.633193863319386, + "grad_norm": 2.5991404056549072, + "learning_rate": 9.9792889979289e-06, + "loss": 0.0185, + "step": 20721 + }, + { + "epoch": 9.633658763365876, + "grad_norm": 6.1385908126831055, + "learning_rate": 9.979287997928801e-06, + "loss": 0.1247, + "step": 20722 + }, + { + "epoch": 9.634123663412366, + "grad_norm": 5.504177570343018, + "learning_rate": 9.979286997928702e-06, + "loss": 0.0688, + "step": 20723 + }, + { + "epoch": 9.634588563458856, + "grad_norm": 4.257135391235352, + "learning_rate": 9.9792859979286e-06, + "loss": 0.0482, + "step": 20724 + }, + { + "epoch": 9.635053463505347, + "grad_norm": 5.376725196838379, + "learning_rate": 9.979284997928501e-06, + "loss": 0.1118, + "step": 20725 + }, + { + "epoch": 9.635518363551837, + "grad_norm": 5.377027988433838, + "learning_rate": 9.9792839979284e-06, + "loss": 0.0466, + "step": 20726 + }, + { + "epoch": 9.635983263598327, + "grad_norm": 4.635354995727539, + "learning_rate": 9.9792829979283e-06, + "loss": 0.162, + "step": 20727 + }, + { + "epoch": 9.636448163644816, + "grad_norm": 5.23508882522583, + "learning_rate": 9.9792819979282e-06, + "loss": 0.0851, + "step": 20728 + }, + { + "epoch": 9.636913063691306, + "grad_norm": 4.870846271514893, + "learning_rate": 9.9792809979281e-06, + "loss": 0.0679, + "step": 20729 + }, + { + "epoch": 9.637377963737796, + "grad_norm": 5.356911659240723, + "learning_rate": 9.979279997928001e-06, + "loss": 0.103, + "step": 20730 + }, + { + "epoch": 9.637842863784286, + "grad_norm": 3.54013729095459, + "learning_rate": 9.9792789979279e-06, + "loss": 0.0548, + "step": 20731 + }, + { + "epoch": 9.638307763830776, + "grad_norm": 5.968760967254639, + "learning_rate": 9.979277997927801e-06, + "loss": 0.161, + "step": 20732 + }, + { + "epoch": 9.638772663877266, + "grad_norm": 6.980673789978027, + "learning_rate": 9.9792769979277e-06, + "loss": 0.1648, + "step": 20733 + }, + { + "epoch": 9.639237563923757, + "grad_norm": 3.964219093322754, + "learning_rate": 9.9792759979276e-06, + "loss": 0.0464, + "step": 20734 + }, + { + "epoch": 9.639702463970247, + "grad_norm": 4.446634769439697, + "learning_rate": 9.979274997927501e-06, + "loss": 0.1181, + "step": 20735 + }, + { + "epoch": 9.640167364016737, + "grad_norm": 4.440715789794922, + "learning_rate": 9.979273997927402e-06, + "loss": 0.0909, + "step": 20736 + }, + { + "epoch": 9.640632264063226, + "grad_norm": 5.392864227294922, + "learning_rate": 9.979272997927301e-06, + "loss": 0.0884, + "step": 20737 + }, + { + "epoch": 9.641097164109716, + "grad_norm": 3.1855220794677734, + "learning_rate": 9.9792719979272e-06, + "loss": 0.0214, + "step": 20738 + }, + { + "epoch": 9.641562064156206, + "grad_norm": 7.407906532287598, + "learning_rate": 9.9792709979271e-06, + "loss": 0.154, + "step": 20739 + }, + { + "epoch": 9.642026964202696, + "grad_norm": 4.441220283508301, + "learning_rate": 9.979269997927e-06, + "loss": 0.073, + "step": 20740 + }, + { + "epoch": 9.642491864249186, + "grad_norm": 3.054582357406616, + "learning_rate": 9.9792689979269e-06, + "loss": 0.0342, + "step": 20741 + }, + { + "epoch": 9.642956764295675, + "grad_norm": 3.3791630268096924, + "learning_rate": 9.9792679979268e-06, + "loss": 0.0429, + "step": 20742 + }, + { + "epoch": 9.643421664342167, + "grad_norm": 5.090372085571289, + "learning_rate": 9.979266997926701e-06, + "loss": 0.059, + "step": 20743 + }, + { + "epoch": 9.643886564388657, + "grad_norm": 4.199649333953857, + "learning_rate": 9.9792659979266e-06, + "loss": 0.0422, + "step": 20744 + }, + { + "epoch": 9.644351464435147, + "grad_norm": 5.486248970031738, + "learning_rate": 9.979264997926501e-06, + "loss": 0.0828, + "step": 20745 + }, + { + "epoch": 9.644816364481636, + "grad_norm": 2.367936134338379, + "learning_rate": 9.9792639979264e-06, + "loss": 0.0233, + "step": 20746 + }, + { + "epoch": 9.645281264528126, + "grad_norm": 5.401052951812744, + "learning_rate": 9.9792629979263e-06, + "loss": 0.1364, + "step": 20747 + }, + { + "epoch": 9.645746164574616, + "grad_norm": 3.0315067768096924, + "learning_rate": 9.979261997926201e-06, + "loss": 0.0361, + "step": 20748 + }, + { + "epoch": 9.646211064621106, + "grad_norm": 6.206063747406006, + "learning_rate": 9.9792609979261e-06, + "loss": 0.1786, + "step": 20749 + }, + { + "epoch": 9.646675964667596, + "grad_norm": 6.604933261871338, + "learning_rate": 9.979259997926001e-06, + "loss": 0.3086, + "step": 20750 + }, + { + "epoch": 9.647140864714087, + "grad_norm": 8.032413482666016, + "learning_rate": 9.9792589979259e-06, + "loss": 0.1207, + "step": 20751 + }, + { + "epoch": 9.647605764760577, + "grad_norm": 7.926710605621338, + "learning_rate": 9.9792579979258e-06, + "loss": 0.0709, + "step": 20752 + }, + { + "epoch": 9.648070664807067, + "grad_norm": 3.1636714935302734, + "learning_rate": 9.9792569979257e-06, + "loss": 0.0425, + "step": 20753 + }, + { + "epoch": 9.648535564853557, + "grad_norm": 9.648697853088379, + "learning_rate": 9.9792559979256e-06, + "loss": 0.216, + "step": 20754 + }, + { + "epoch": 9.649000464900046, + "grad_norm": 5.560064315795898, + "learning_rate": 9.9792549979255e-06, + "loss": 0.0814, + "step": 20755 + }, + { + "epoch": 9.649465364946536, + "grad_norm": 4.970492362976074, + "learning_rate": 9.979253997925401e-06, + "loss": 0.0643, + "step": 20756 + }, + { + "epoch": 9.649930264993026, + "grad_norm": 4.857167720794678, + "learning_rate": 9.9792529979253e-06, + "loss": 0.0798, + "step": 20757 + }, + { + "epoch": 9.650395165039516, + "grad_norm": 4.428241729736328, + "learning_rate": 9.979251997925201e-06, + "loss": 0.0385, + "step": 20758 + }, + { + "epoch": 9.650860065086006, + "grad_norm": 5.8683929443359375, + "learning_rate": 9.9792509979251e-06, + "loss": 0.1454, + "step": 20759 + }, + { + "epoch": 9.651324965132497, + "grad_norm": 3.93887996673584, + "learning_rate": 9.979249997925e-06, + "loss": 0.058, + "step": 20760 + }, + { + "epoch": 9.651789865178987, + "grad_norm": 4.603644847869873, + "learning_rate": 9.979248997924901e-06, + "loss": 0.0861, + "step": 20761 + }, + { + "epoch": 9.652254765225477, + "grad_norm": 5.428476333618164, + "learning_rate": 9.9792479979248e-06, + "loss": 0.0753, + "step": 20762 + }, + { + "epoch": 9.652719665271967, + "grad_norm": 5.316623210906982, + "learning_rate": 9.979246997924701e-06, + "loss": 0.0702, + "step": 20763 + }, + { + "epoch": 9.653184565318456, + "grad_norm": 7.0459136962890625, + "learning_rate": 9.9792459979246e-06, + "loss": 0.1471, + "step": 20764 + }, + { + "epoch": 9.653649465364946, + "grad_norm": 9.551876068115234, + "learning_rate": 9.9792449979245e-06, + "loss": 0.2784, + "step": 20765 + }, + { + "epoch": 9.654114365411436, + "grad_norm": 4.73303747177124, + "learning_rate": 9.979243997924401e-06, + "loss": 0.0825, + "step": 20766 + }, + { + "epoch": 9.654579265457926, + "grad_norm": 8.23072338104248, + "learning_rate": 9.9792429979243e-06, + "loss": 0.0875, + "step": 20767 + }, + { + "epoch": 9.655044165504417, + "grad_norm": 6.506600856781006, + "learning_rate": 9.979241997924201e-06, + "loss": 0.0958, + "step": 20768 + }, + { + "epoch": 9.655509065550907, + "grad_norm": 5.335646152496338, + "learning_rate": 9.9792409979241e-06, + "loss": 0.1522, + "step": 20769 + }, + { + "epoch": 9.655973965597397, + "grad_norm": 3.0348329544067383, + "learning_rate": 9.979239997924e-06, + "loss": 0.0284, + "step": 20770 + }, + { + "epoch": 9.656438865643887, + "grad_norm": 6.122946262359619, + "learning_rate": 9.9792389979239e-06, + "loss": 0.0744, + "step": 20771 + }, + { + "epoch": 9.656903765690377, + "grad_norm": 1.4782286882400513, + "learning_rate": 9.9792379979238e-06, + "loss": 0.0116, + "step": 20772 + }, + { + "epoch": 9.657368665736866, + "grad_norm": 9.684699058532715, + "learning_rate": 9.9792369979237e-06, + "loss": 0.2757, + "step": 20773 + }, + { + "epoch": 9.657833565783356, + "grad_norm": 3.181609869003296, + "learning_rate": 9.979235997923601e-06, + "loss": 0.0441, + "step": 20774 + }, + { + "epoch": 9.658298465829846, + "grad_norm": 4.074978351593018, + "learning_rate": 9.9792349979235e-06, + "loss": 0.0618, + "step": 20775 + }, + { + "epoch": 9.658763365876336, + "grad_norm": 5.259549617767334, + "learning_rate": 9.979233997923401e-06, + "loss": 0.1083, + "step": 20776 + }, + { + "epoch": 9.659228265922827, + "grad_norm": 6.528714179992676, + "learning_rate": 9.9792329979233e-06, + "loss": 0.1178, + "step": 20777 + }, + { + "epoch": 9.659693165969317, + "grad_norm": 4.681138515472412, + "learning_rate": 9.9792319979232e-06, + "loss": 0.1109, + "step": 20778 + }, + { + "epoch": 9.660158066015807, + "grad_norm": 1.85646653175354, + "learning_rate": 9.979230997923101e-06, + "loss": 0.0156, + "step": 20779 + }, + { + "epoch": 9.660622966062297, + "grad_norm": 3.6499226093292236, + "learning_rate": 9.979229997923e-06, + "loss": 0.0515, + "step": 20780 + }, + { + "epoch": 9.661087866108787, + "grad_norm": 4.680148124694824, + "learning_rate": 9.979228997922901e-06, + "loss": 0.0462, + "step": 20781 + }, + { + "epoch": 9.661552766155276, + "grad_norm": 4.802727699279785, + "learning_rate": 9.9792279979228e-06, + "loss": 0.0955, + "step": 20782 + }, + { + "epoch": 9.662017666201766, + "grad_norm": 4.463298320770264, + "learning_rate": 9.9792269979227e-06, + "loss": 0.0639, + "step": 20783 + }, + { + "epoch": 9.662482566248256, + "grad_norm": 5.128036975860596, + "learning_rate": 9.9792259979226e-06, + "loss": 0.0981, + "step": 20784 + }, + { + "epoch": 9.662947466294746, + "grad_norm": 7.03201961517334, + "learning_rate": 9.9792249979225e-06, + "loss": 0.1197, + "step": 20785 + }, + { + "epoch": 9.663412366341237, + "grad_norm": 5.449002742767334, + "learning_rate": 9.979223997922401e-06, + "loss": 0.1265, + "step": 20786 + }, + { + "epoch": 9.663877266387727, + "grad_norm": 6.610769271850586, + "learning_rate": 9.979222997922302e-06, + "loss": 0.0847, + "step": 20787 + }, + { + "epoch": 9.664342166434217, + "grad_norm": 6.0402302742004395, + "learning_rate": 9.9792219979222e-06, + "loss": 0.1109, + "step": 20788 + }, + { + "epoch": 9.664807066480707, + "grad_norm": 2.371309757232666, + "learning_rate": 9.979220997922101e-06, + "loss": 0.0212, + "step": 20789 + }, + { + "epoch": 9.665271966527197, + "grad_norm": 8.277714729309082, + "learning_rate": 9.979219997922e-06, + "loss": 0.1892, + "step": 20790 + }, + { + "epoch": 9.665736866573686, + "grad_norm": 6.1480393409729, + "learning_rate": 9.9792189979219e-06, + "loss": 0.075, + "step": 20791 + }, + { + "epoch": 9.666201766620176, + "grad_norm": 2.842412233352661, + "learning_rate": 9.979217997921801e-06, + "loss": 0.0243, + "step": 20792 + }, + { + "epoch": 9.666666666666666, + "grad_norm": 5.707180500030518, + "learning_rate": 9.9792169979217e-06, + "loss": 0.1065, + "step": 20793 + }, + { + "epoch": 9.667131566713156, + "grad_norm": 5.537342548370361, + "learning_rate": 9.979215997921601e-06, + "loss": 0.066, + "step": 20794 + }, + { + "epoch": 9.667596466759647, + "grad_norm": 8.452800750732422, + "learning_rate": 9.9792149979215e-06, + "loss": 0.1468, + "step": 20795 + }, + { + "epoch": 9.668061366806137, + "grad_norm": 10.274909019470215, + "learning_rate": 9.9792139979214e-06, + "loss": 0.1634, + "step": 20796 + }, + { + "epoch": 9.668526266852627, + "grad_norm": 6.672461986541748, + "learning_rate": 9.9792129979213e-06, + "loss": 0.1633, + "step": 20797 + }, + { + "epoch": 9.668991166899117, + "grad_norm": 10.18407154083252, + "learning_rate": 9.9792119979212e-06, + "loss": 0.2455, + "step": 20798 + }, + { + "epoch": 9.669456066945607, + "grad_norm": 5.362537384033203, + "learning_rate": 9.979210997921101e-06, + "loss": 0.1126, + "step": 20799 + }, + { + "epoch": 9.669920966992096, + "grad_norm": 7.768566131591797, + "learning_rate": 9.979209997921002e-06, + "loss": 0.2038, + "step": 20800 + }, + { + "epoch": 9.670385867038586, + "grad_norm": 4.014588356018066, + "learning_rate": 9.9792089979209e-06, + "loss": 0.1059, + "step": 20801 + }, + { + "epoch": 9.670850767085076, + "grad_norm": 5.934579372406006, + "learning_rate": 9.9792079979208e-06, + "loss": 0.1232, + "step": 20802 + }, + { + "epoch": 9.671315667131566, + "grad_norm": 2.8670108318328857, + "learning_rate": 9.9792069979207e-06, + "loss": 0.0484, + "step": 20803 + }, + { + "epoch": 9.671780567178057, + "grad_norm": 4.6232686042785645, + "learning_rate": 9.979205997920601e-06, + "loss": 0.0569, + "step": 20804 + }, + { + "epoch": 9.672245467224547, + "grad_norm": 5.354732513427734, + "learning_rate": 9.979204997920502e-06, + "loss": 0.1028, + "step": 20805 + }, + { + "epoch": 9.672710367271037, + "grad_norm": 8.658384323120117, + "learning_rate": 9.9792039979204e-06, + "loss": 0.2662, + "step": 20806 + }, + { + "epoch": 9.673175267317527, + "grad_norm": 8.495555877685547, + "learning_rate": 9.979202997920301e-06, + "loss": 0.2538, + "step": 20807 + }, + { + "epoch": 9.673640167364017, + "grad_norm": 7.978750705718994, + "learning_rate": 9.9792019979202e-06, + "loss": 0.1237, + "step": 20808 + }, + { + "epoch": 9.674105067410506, + "grad_norm": 3.6491496562957764, + "learning_rate": 9.9792009979201e-06, + "loss": 0.0783, + "step": 20809 + }, + { + "epoch": 9.674569967456996, + "grad_norm": 4.6150803565979, + "learning_rate": 9.97919999792e-06, + "loss": 0.0562, + "step": 20810 + }, + { + "epoch": 9.675034867503486, + "grad_norm": 4.762225151062012, + "learning_rate": 9.9791989979199e-06, + "loss": 0.0318, + "step": 20811 + }, + { + "epoch": 9.675499767549976, + "grad_norm": 5.567391872406006, + "learning_rate": 9.979197997919801e-06, + "loss": 0.0867, + "step": 20812 + }, + { + "epoch": 9.675964667596467, + "grad_norm": 2.0444529056549072, + "learning_rate": 9.9791969979197e-06, + "loss": 0.0187, + "step": 20813 + }, + { + "epoch": 9.676429567642957, + "grad_norm": 3.6637682914733887, + "learning_rate": 9.9791959979196e-06, + "loss": 0.0602, + "step": 20814 + }, + { + "epoch": 9.676894467689447, + "grad_norm": 3.5741705894470215, + "learning_rate": 9.9791949979195e-06, + "loss": 0.0532, + "step": 20815 + }, + { + "epoch": 9.677359367735937, + "grad_norm": 0.8639993071556091, + "learning_rate": 9.9791939979194e-06, + "loss": 0.0089, + "step": 20816 + }, + { + "epoch": 9.677824267782427, + "grad_norm": 3.5905539989471436, + "learning_rate": 9.979192997919301e-06, + "loss": 0.0728, + "step": 20817 + }, + { + "epoch": 9.678289167828916, + "grad_norm": 10.807279586791992, + "learning_rate": 9.979191997919202e-06, + "loss": 0.2774, + "step": 20818 + }, + { + "epoch": 9.678754067875406, + "grad_norm": 8.534805297851562, + "learning_rate": 9.9791909979191e-06, + "loss": 0.032, + "step": 20819 + }, + { + "epoch": 9.679218967921896, + "grad_norm": 6.581783294677734, + "learning_rate": 9.979189997919001e-06, + "loss": 0.0689, + "step": 20820 + }, + { + "epoch": 9.679683867968386, + "grad_norm": 3.931942939758301, + "learning_rate": 9.9791889979189e-06, + "loss": 0.0608, + "step": 20821 + }, + { + "epoch": 9.680148768014877, + "grad_norm": 3.893609046936035, + "learning_rate": 9.979187997918801e-06, + "loss": 0.0522, + "step": 20822 + }, + { + "epoch": 9.680613668061367, + "grad_norm": 5.287347793579102, + "learning_rate": 9.9791869979187e-06, + "loss": 0.0522, + "step": 20823 + }, + { + "epoch": 9.681078568107857, + "grad_norm": 7.2822065353393555, + "learning_rate": 9.9791859979186e-06, + "loss": 0.1563, + "step": 20824 + }, + { + "epoch": 9.681543468154347, + "grad_norm": 4.461772441864014, + "learning_rate": 9.979184997918501e-06, + "loss": 0.077, + "step": 20825 + }, + { + "epoch": 9.682008368200837, + "grad_norm": 7.95698356628418, + "learning_rate": 9.9791839979184e-06, + "loss": 0.1437, + "step": 20826 + }, + { + "epoch": 9.682473268247326, + "grad_norm": 4.103158473968506, + "learning_rate": 9.9791829979183e-06, + "loss": 0.0328, + "step": 20827 + }, + { + "epoch": 9.682938168293816, + "grad_norm": 3.9864017963409424, + "learning_rate": 9.9791819979182e-06, + "loss": 0.0729, + "step": 20828 + }, + { + "epoch": 9.683403068340306, + "grad_norm": 5.795746326446533, + "learning_rate": 9.9791809979181e-06, + "loss": 0.1282, + "step": 20829 + }, + { + "epoch": 9.683867968386798, + "grad_norm": 4.246518135070801, + "learning_rate": 9.979179997918001e-06, + "loss": 0.0682, + "step": 20830 + }, + { + "epoch": 9.684332868433287, + "grad_norm": 4.628782272338867, + "learning_rate": 9.979178997917902e-06, + "loss": 0.0956, + "step": 20831 + }, + { + "epoch": 9.684797768479777, + "grad_norm": 2.4005913734436035, + "learning_rate": 9.9791779979178e-06, + "loss": 0.0244, + "step": 20832 + }, + { + "epoch": 9.685262668526267, + "grad_norm": 4.546558856964111, + "learning_rate": 9.979176997917701e-06, + "loss": 0.059, + "step": 20833 + }, + { + "epoch": 9.685727568572757, + "grad_norm": 4.7913289070129395, + "learning_rate": 9.9791759979176e-06, + "loss": 0.1274, + "step": 20834 + }, + { + "epoch": 9.686192468619247, + "grad_norm": 5.753893852233887, + "learning_rate": 9.979174997917501e-06, + "loss": 0.148, + "step": 20835 + }, + { + "epoch": 9.686657368665736, + "grad_norm": 5.453549385070801, + "learning_rate": 9.9791739979174e-06, + "loss": 0.0861, + "step": 20836 + }, + { + "epoch": 9.687122268712226, + "grad_norm": 9.389083862304688, + "learning_rate": 9.9791729979173e-06, + "loss": 0.2013, + "step": 20837 + }, + { + "epoch": 9.687587168758716, + "grad_norm": 5.711416244506836, + "learning_rate": 9.979171997917201e-06, + "loss": 0.1219, + "step": 20838 + }, + { + "epoch": 9.688052068805208, + "grad_norm": 4.305307388305664, + "learning_rate": 9.9791709979171e-06, + "loss": 0.0742, + "step": 20839 + }, + { + "epoch": 9.688516968851697, + "grad_norm": 4.6342668533325195, + "learning_rate": 9.979169997917e-06, + "loss": 0.0605, + "step": 20840 + }, + { + "epoch": 9.688981868898187, + "grad_norm": 6.017315864562988, + "learning_rate": 9.9791689979169e-06, + "loss": 0.1075, + "step": 20841 + }, + { + "epoch": 9.689446768944677, + "grad_norm": 1.8084170818328857, + "learning_rate": 9.9791679979168e-06, + "loss": 0.021, + "step": 20842 + }, + { + "epoch": 9.689911668991167, + "grad_norm": 3.7469112873077393, + "learning_rate": 9.979166997916701e-06, + "loss": 0.0722, + "step": 20843 + }, + { + "epoch": 9.690376569037657, + "grad_norm": 4.455735206604004, + "learning_rate": 9.979165997916602e-06, + "loss": 0.0169, + "step": 20844 + }, + { + "epoch": 9.690841469084146, + "grad_norm": 1.823012113571167, + "learning_rate": 9.9791649979165e-06, + "loss": 0.0128, + "step": 20845 + }, + { + "epoch": 9.691306369130636, + "grad_norm": 6.543231010437012, + "learning_rate": 9.9791639979164e-06, + "loss": 0.1147, + "step": 20846 + }, + { + "epoch": 9.691771269177128, + "grad_norm": 7.194512844085693, + "learning_rate": 9.9791629979163e-06, + "loss": 0.0914, + "step": 20847 + }, + { + "epoch": 9.692236169223618, + "grad_norm": 3.855842113494873, + "learning_rate": 9.979161997916201e-06, + "loss": 0.0434, + "step": 20848 + }, + { + "epoch": 9.692701069270107, + "grad_norm": 5.9049530029296875, + "learning_rate": 9.9791609979161e-06, + "loss": 0.1056, + "step": 20849 + }, + { + "epoch": 9.693165969316597, + "grad_norm": 5.1732988357543945, + "learning_rate": 9.979159997916e-06, + "loss": 0.1029, + "step": 20850 + }, + { + "epoch": 9.693630869363087, + "grad_norm": 17.058385848999023, + "learning_rate": 9.979158997915901e-06, + "loss": 0.3063, + "step": 20851 + }, + { + "epoch": 9.694095769409577, + "grad_norm": 3.1118268966674805, + "learning_rate": 9.9791579979158e-06, + "loss": 0.0342, + "step": 20852 + }, + { + "epoch": 9.694560669456067, + "grad_norm": 2.200024366378784, + "learning_rate": 9.979156997915701e-06, + "loss": 0.0297, + "step": 20853 + }, + { + "epoch": 9.695025569502556, + "grad_norm": 2.4374728202819824, + "learning_rate": 9.9791559979156e-06, + "loss": 0.0302, + "step": 20854 + }, + { + "epoch": 9.695490469549046, + "grad_norm": 4.083104133605957, + "learning_rate": 9.9791549979155e-06, + "loss": 0.0897, + "step": 20855 + }, + { + "epoch": 9.695955369595538, + "grad_norm": 3.9424595832824707, + "learning_rate": 9.979153997915401e-06, + "loss": 0.0472, + "step": 20856 + }, + { + "epoch": 9.696420269642028, + "grad_norm": 5.3989949226379395, + "learning_rate": 9.9791529979153e-06, + "loss": 0.1013, + "step": 20857 + }, + { + "epoch": 9.696885169688517, + "grad_norm": 4.67193603515625, + "learning_rate": 9.9791519979152e-06, + "loss": 0.1094, + "step": 20858 + }, + { + "epoch": 9.697350069735007, + "grad_norm": 14.616618156433105, + "learning_rate": 9.9791509979151e-06, + "loss": 0.1376, + "step": 20859 + }, + { + "epoch": 9.697814969781497, + "grad_norm": 6.8689656257629395, + "learning_rate": 9.979149997915e-06, + "loss": 0.1103, + "step": 20860 + }, + { + "epoch": 9.698279869827987, + "grad_norm": 2.38832950592041, + "learning_rate": 9.979148997914901e-06, + "loss": 0.0314, + "step": 20861 + }, + { + "epoch": 9.698744769874477, + "grad_norm": 4.483788013458252, + "learning_rate": 9.979147997914802e-06, + "loss": 0.2223, + "step": 20862 + }, + { + "epoch": 9.699209669920966, + "grad_norm": 4.406039237976074, + "learning_rate": 9.9791469979147e-06, + "loss": 0.0978, + "step": 20863 + }, + { + "epoch": 9.699674569967456, + "grad_norm": 2.782275438308716, + "learning_rate": 9.979145997914601e-06, + "loss": 0.0255, + "step": 20864 + }, + { + "epoch": 9.700139470013948, + "grad_norm": 4.319241523742676, + "learning_rate": 9.9791449979145e-06, + "loss": 0.0432, + "step": 20865 + }, + { + "epoch": 9.700604370060438, + "grad_norm": 2.979834794998169, + "learning_rate": 9.979143997914401e-06, + "loss": 0.0418, + "step": 20866 + }, + { + "epoch": 9.701069270106927, + "grad_norm": 5.628090858459473, + "learning_rate": 9.9791429979143e-06, + "loss": 0.0627, + "step": 20867 + }, + { + "epoch": 9.701534170153417, + "grad_norm": 7.986654281616211, + "learning_rate": 9.9791419979142e-06, + "loss": 0.1241, + "step": 20868 + }, + { + "epoch": 9.701999070199907, + "grad_norm": 4.304292678833008, + "learning_rate": 9.979140997914101e-06, + "loss": 0.0205, + "step": 20869 + }, + { + "epoch": 9.702463970246397, + "grad_norm": 8.826065063476562, + "learning_rate": 9.979139997914e-06, + "loss": 0.2895, + "step": 20870 + }, + { + "epoch": 9.702928870292887, + "grad_norm": 1.706060528755188, + "learning_rate": 9.979138997913901e-06, + "loss": 0.0183, + "step": 20871 + }, + { + "epoch": 9.703393770339376, + "grad_norm": 9.354925155639648, + "learning_rate": 9.9791379979138e-06, + "loss": 0.2302, + "step": 20872 + }, + { + "epoch": 9.703858670385866, + "grad_norm": 8.874237060546875, + "learning_rate": 9.9791369979137e-06, + "loss": 0.0553, + "step": 20873 + }, + { + "epoch": 9.704323570432358, + "grad_norm": 7.856762886047363, + "learning_rate": 9.979135997913601e-06, + "loss": 0.14, + "step": 20874 + }, + { + "epoch": 9.704788470478848, + "grad_norm": 3.3144800662994385, + "learning_rate": 9.979134997913502e-06, + "loss": 0.0305, + "step": 20875 + }, + { + "epoch": 9.705253370525337, + "grad_norm": 1.6330931186676025, + "learning_rate": 9.9791339979134e-06, + "loss": 0.0164, + "step": 20876 + }, + { + "epoch": 9.705718270571827, + "grad_norm": 1.542277455329895, + "learning_rate": 9.979132997913301e-06, + "loss": 0.0083, + "step": 20877 + }, + { + "epoch": 9.706183170618317, + "grad_norm": 7.190940856933594, + "learning_rate": 9.9791319979132e-06, + "loss": 0.2494, + "step": 20878 + }, + { + "epoch": 9.706648070664807, + "grad_norm": 3.4392623901367188, + "learning_rate": 9.9791309979131e-06, + "loss": 0.0463, + "step": 20879 + }, + { + "epoch": 9.707112970711297, + "grad_norm": 4.692934989929199, + "learning_rate": 9.979129997913e-06, + "loss": 0.0609, + "step": 20880 + }, + { + "epoch": 9.707577870757786, + "grad_norm": 3.530553102493286, + "learning_rate": 9.9791289979129e-06, + "loss": 0.0698, + "step": 20881 + }, + { + "epoch": 9.708042770804276, + "grad_norm": 4.899868488311768, + "learning_rate": 9.979127997912801e-06, + "loss": 0.0598, + "step": 20882 + }, + { + "epoch": 9.708507670850768, + "grad_norm": 6.089208602905273, + "learning_rate": 9.9791269979127e-06, + "loss": 0.1349, + "step": 20883 + }, + { + "epoch": 9.708972570897258, + "grad_norm": 4.754397869110107, + "learning_rate": 9.979125997912601e-06, + "loss": 0.0552, + "step": 20884 + }, + { + "epoch": 9.709437470943747, + "grad_norm": 5.200865745544434, + "learning_rate": 9.9791249979125e-06, + "loss": 0.107, + "step": 20885 + }, + { + "epoch": 9.709902370990237, + "grad_norm": 6.009296417236328, + "learning_rate": 9.9791239979124e-06, + "loss": 0.1546, + "step": 20886 + }, + { + "epoch": 9.710367271036727, + "grad_norm": 6.626126289367676, + "learning_rate": 9.979122997912301e-06, + "loss": 0.1812, + "step": 20887 + }, + { + "epoch": 9.710832171083217, + "grad_norm": 4.898620128631592, + "learning_rate": 9.979121997912202e-06, + "loss": 0.0512, + "step": 20888 + }, + { + "epoch": 9.711297071129707, + "grad_norm": 6.88428258895874, + "learning_rate": 9.979120997912101e-06, + "loss": 0.0829, + "step": 20889 + }, + { + "epoch": 9.711761971176196, + "grad_norm": 3.00388240814209, + "learning_rate": 9.979119997912e-06, + "loss": 0.0371, + "step": 20890 + }, + { + "epoch": 9.712226871222686, + "grad_norm": 6.724892616271973, + "learning_rate": 9.9791189979119e-06, + "loss": 0.1236, + "step": 20891 + }, + { + "epoch": 9.712691771269178, + "grad_norm": 6.853402137756348, + "learning_rate": 9.9791179979118e-06, + "loss": 0.0955, + "step": 20892 + }, + { + "epoch": 9.713156671315668, + "grad_norm": 3.2921838760375977, + "learning_rate": 9.9791169979117e-06, + "loss": 0.0302, + "step": 20893 + }, + { + "epoch": 9.713621571362157, + "grad_norm": 2.230675458908081, + "learning_rate": 9.9791159979116e-06, + "loss": 0.0253, + "step": 20894 + }, + { + "epoch": 9.714086471408647, + "grad_norm": 5.135928630828857, + "learning_rate": 9.979114997911501e-06, + "loss": 0.0857, + "step": 20895 + }, + { + "epoch": 9.714551371455137, + "grad_norm": 5.454525947570801, + "learning_rate": 9.9791139979114e-06, + "loss": 0.1224, + "step": 20896 + }, + { + "epoch": 9.715016271501627, + "grad_norm": 4.826363563537598, + "learning_rate": 9.979112997911301e-06, + "loss": 0.1053, + "step": 20897 + }, + { + "epoch": 9.715481171548117, + "grad_norm": 3.3902390003204346, + "learning_rate": 9.9791119979112e-06, + "loss": 0.0354, + "step": 20898 + }, + { + "epoch": 9.715946071594606, + "grad_norm": 2.6806347370147705, + "learning_rate": 9.9791109979111e-06, + "loss": 0.0245, + "step": 20899 + }, + { + "epoch": 9.716410971641096, + "grad_norm": 3.4244635105133057, + "learning_rate": 9.979109997911001e-06, + "loss": 0.0469, + "step": 20900 + }, + { + "epoch": 9.716875871687588, + "grad_norm": 6.643270492553711, + "learning_rate": 9.9791089979109e-06, + "loss": 0.1958, + "step": 20901 + }, + { + "epoch": 9.717340771734078, + "grad_norm": 8.248177528381348, + "learning_rate": 9.979107997910801e-06, + "loss": 0.1271, + "step": 20902 + }, + { + "epoch": 9.717805671780567, + "grad_norm": 3.567962408065796, + "learning_rate": 9.9791069979107e-06, + "loss": 0.0434, + "step": 20903 + }, + { + "epoch": 9.718270571827057, + "grad_norm": 3.651463270187378, + "learning_rate": 9.9791059979106e-06, + "loss": 0.0379, + "step": 20904 + }, + { + "epoch": 9.718735471873547, + "grad_norm": 6.117926120758057, + "learning_rate": 9.9791049979105e-06, + "loss": 0.1158, + "step": 20905 + }, + { + "epoch": 9.719200371920037, + "grad_norm": 6.684386730194092, + "learning_rate": 9.9791039979104e-06, + "loss": 0.099, + "step": 20906 + }, + { + "epoch": 9.719665271966527, + "grad_norm": 7.139975547790527, + "learning_rate": 9.979102997910301e-06, + "loss": 0.071, + "step": 20907 + }, + { + "epoch": 9.720130172013016, + "grad_norm": 5.497507572174072, + "learning_rate": 9.979101997910202e-06, + "loss": 0.1029, + "step": 20908 + }, + { + "epoch": 9.720595072059508, + "grad_norm": 8.060561180114746, + "learning_rate": 9.9791009979101e-06, + "loss": 0.1998, + "step": 20909 + }, + { + "epoch": 9.721059972105998, + "grad_norm": 5.379416465759277, + "learning_rate": 9.97909999791e-06, + "loss": 0.148, + "step": 20910 + }, + { + "epoch": 9.721524872152488, + "grad_norm": 4.833715438842773, + "learning_rate": 9.9790989979099e-06, + "loss": 0.0389, + "step": 20911 + }, + { + "epoch": 9.721989772198977, + "grad_norm": 2.5818912982940674, + "learning_rate": 9.9790979979098e-06, + "loss": 0.0245, + "step": 20912 + }, + { + "epoch": 9.722454672245467, + "grad_norm": 5.95639705657959, + "learning_rate": 9.979096997909701e-06, + "loss": 0.067, + "step": 20913 + }, + { + "epoch": 9.722919572291957, + "grad_norm": 5.717869281768799, + "learning_rate": 9.9790959979096e-06, + "loss": 0.0731, + "step": 20914 + }, + { + "epoch": 9.723384472338447, + "grad_norm": 4.355677127838135, + "learning_rate": 9.979094997909501e-06, + "loss": 0.0503, + "step": 20915 + }, + { + "epoch": 9.723849372384937, + "grad_norm": 8.665843963623047, + "learning_rate": 9.9790939979094e-06, + "loss": 0.1665, + "step": 20916 + }, + { + "epoch": 9.724314272431426, + "grad_norm": 3.2154548168182373, + "learning_rate": 9.9790929979093e-06, + "loss": 0.0368, + "step": 20917 + }, + { + "epoch": 9.724779172477918, + "grad_norm": 6.94340705871582, + "learning_rate": 9.9790919979092e-06, + "loss": 0.2153, + "step": 20918 + }, + { + "epoch": 9.725244072524408, + "grad_norm": 6.831394672393799, + "learning_rate": 9.9790909979091e-06, + "loss": 0.1382, + "step": 20919 + }, + { + "epoch": 9.725708972570898, + "grad_norm": 3.110304117202759, + "learning_rate": 9.979089997909001e-06, + "loss": 0.0531, + "step": 20920 + }, + { + "epoch": 9.726173872617387, + "grad_norm": 5.273243427276611, + "learning_rate": 9.9790889979089e-06, + "loss": 0.0917, + "step": 20921 + }, + { + "epoch": 9.726638772663877, + "grad_norm": 2.037745952606201, + "learning_rate": 9.9790879979088e-06, + "loss": 0.0164, + "step": 20922 + }, + { + "epoch": 9.727103672710367, + "grad_norm": 7.156190395355225, + "learning_rate": 9.9790869979087e-06, + "loss": 0.2284, + "step": 20923 + }, + { + "epoch": 9.727568572756857, + "grad_norm": 4.07783317565918, + "learning_rate": 9.9790859979086e-06, + "loss": 0.045, + "step": 20924 + }, + { + "epoch": 9.728033472803347, + "grad_norm": 3.7626826763153076, + "learning_rate": 9.9790849979085e-06, + "loss": 0.0478, + "step": 20925 + }, + { + "epoch": 9.728498372849838, + "grad_norm": 8.32809066772461, + "learning_rate": 9.979083997908402e-06, + "loss": 0.269, + "step": 20926 + }, + { + "epoch": 9.728963272896328, + "grad_norm": 2.898394823074341, + "learning_rate": 9.9790829979083e-06, + "loss": 0.0301, + "step": 20927 + }, + { + "epoch": 9.729428172942818, + "grad_norm": 2.2107527256011963, + "learning_rate": 9.979081997908201e-06, + "loss": 0.0257, + "step": 20928 + }, + { + "epoch": 9.729893072989308, + "grad_norm": 4.170248508453369, + "learning_rate": 9.9790809979081e-06, + "loss": 0.0814, + "step": 20929 + }, + { + "epoch": 9.730357973035797, + "grad_norm": 6.194123268127441, + "learning_rate": 9.979079997908e-06, + "loss": 0.0763, + "step": 20930 + }, + { + "epoch": 9.730822873082287, + "grad_norm": 5.958895683288574, + "learning_rate": 9.979078997907901e-06, + "loss": 0.1141, + "step": 20931 + }, + { + "epoch": 9.731287773128777, + "grad_norm": 6.20229434967041, + "learning_rate": 9.9790779979078e-06, + "loss": 0.1642, + "step": 20932 + }, + { + "epoch": 9.731752673175267, + "grad_norm": 4.672994613647461, + "learning_rate": 9.979076997907701e-06, + "loss": 0.0485, + "step": 20933 + }, + { + "epoch": 9.732217573221757, + "grad_norm": 3.4684343338012695, + "learning_rate": 9.9790759979076e-06, + "loss": 0.0464, + "step": 20934 + }, + { + "epoch": 9.732682473268248, + "grad_norm": 3.9225800037384033, + "learning_rate": 9.9790749979075e-06, + "loss": 0.0342, + "step": 20935 + }, + { + "epoch": 9.733147373314738, + "grad_norm": 6.917017459869385, + "learning_rate": 9.9790739979074e-06, + "loss": 0.1197, + "step": 20936 + }, + { + "epoch": 9.733612273361228, + "grad_norm": 8.58317756652832, + "learning_rate": 9.9790729979073e-06, + "loss": 0.146, + "step": 20937 + }, + { + "epoch": 9.734077173407718, + "grad_norm": 5.442508220672607, + "learning_rate": 9.979071997907201e-06, + "loss": 0.0939, + "step": 20938 + }, + { + "epoch": 9.734542073454207, + "grad_norm": 3.6461122035980225, + "learning_rate": 9.979070997907102e-06, + "loss": 0.0344, + "step": 20939 + }, + { + "epoch": 9.735006973500697, + "grad_norm": 8.76765251159668, + "learning_rate": 9.979069997907e-06, + "loss": 0.1945, + "step": 20940 + }, + { + "epoch": 9.735471873547187, + "grad_norm": 6.313580513000488, + "learning_rate": 9.979068997906901e-06, + "loss": 0.1377, + "step": 20941 + }, + { + "epoch": 9.735936773593677, + "grad_norm": 4.847533702850342, + "learning_rate": 9.9790679979068e-06, + "loss": 0.1221, + "step": 20942 + }, + { + "epoch": 9.736401673640167, + "grad_norm": 4.676931858062744, + "learning_rate": 9.9790669979067e-06, + "loss": 0.088, + "step": 20943 + }, + { + "epoch": 9.736866573686658, + "grad_norm": 7.493731498718262, + "learning_rate": 9.979065997906601e-06, + "loss": 0.1841, + "step": 20944 + }, + { + "epoch": 9.737331473733148, + "grad_norm": 3.9298934936523438, + "learning_rate": 9.9790649979065e-06, + "loss": 0.0792, + "step": 20945 + }, + { + "epoch": 9.737796373779638, + "grad_norm": 4.667892932891846, + "learning_rate": 9.979063997906401e-06, + "loss": 0.0865, + "step": 20946 + }, + { + "epoch": 9.738261273826128, + "grad_norm": 5.464920520782471, + "learning_rate": 9.9790629979063e-06, + "loss": 0.0711, + "step": 20947 + }, + { + "epoch": 9.738726173872617, + "grad_norm": 3.441189765930176, + "learning_rate": 9.9790619979062e-06, + "loss": 0.047, + "step": 20948 + }, + { + "epoch": 9.739191073919107, + "grad_norm": 5.974975109100342, + "learning_rate": 9.9790609979061e-06, + "loss": 0.1122, + "step": 20949 + }, + { + "epoch": 9.739655973965597, + "grad_norm": 6.737011909484863, + "learning_rate": 9.979059997906e-06, + "loss": 0.1096, + "step": 20950 + }, + { + "epoch": 9.740120874012087, + "grad_norm": 4.749973297119141, + "learning_rate": 9.979058997905901e-06, + "loss": 0.0774, + "step": 20951 + }, + { + "epoch": 9.740585774058577, + "grad_norm": 4.872488498687744, + "learning_rate": 9.979057997905802e-06, + "loss": 0.0508, + "step": 20952 + }, + { + "epoch": 9.741050674105068, + "grad_norm": 2.8773648738861084, + "learning_rate": 9.9790569979057e-06, + "loss": 0.0201, + "step": 20953 + }, + { + "epoch": 9.741515574151558, + "grad_norm": 5.542076110839844, + "learning_rate": 9.9790559979056e-06, + "loss": 0.1141, + "step": 20954 + }, + { + "epoch": 9.741980474198048, + "grad_norm": 6.101250648498535, + "learning_rate": 9.9790549979055e-06, + "loss": 0.1501, + "step": 20955 + }, + { + "epoch": 9.742445374244538, + "grad_norm": 3.0599231719970703, + "learning_rate": 9.979053997905401e-06, + "loss": 0.0401, + "step": 20956 + }, + { + "epoch": 9.742910274291027, + "grad_norm": 7.429521560668945, + "learning_rate": 9.979052997905302e-06, + "loss": 0.0742, + "step": 20957 + }, + { + "epoch": 9.743375174337517, + "grad_norm": 2.999607801437378, + "learning_rate": 9.9790519979052e-06, + "loss": 0.0303, + "step": 20958 + }, + { + "epoch": 9.743840074384007, + "grad_norm": 1.3891843557357788, + "learning_rate": 9.979050997905101e-06, + "loss": 0.0126, + "step": 20959 + }, + { + "epoch": 9.744304974430497, + "grad_norm": 4.275579929351807, + "learning_rate": 9.979049997905e-06, + "loss": 0.0835, + "step": 20960 + }, + { + "epoch": 9.744769874476987, + "grad_norm": 5.6638665199279785, + "learning_rate": 9.9790489979049e-06, + "loss": 0.0985, + "step": 20961 + }, + { + "epoch": 9.745234774523478, + "grad_norm": 5.689069747924805, + "learning_rate": 9.9790479979048e-06, + "loss": 0.1133, + "step": 20962 + }, + { + "epoch": 9.745699674569968, + "grad_norm": 1.843267560005188, + "learning_rate": 9.9790469979047e-06, + "loss": 0.011, + "step": 20963 + }, + { + "epoch": 9.746164574616458, + "grad_norm": 3.1831841468811035, + "learning_rate": 9.979045997904601e-06, + "loss": 0.0529, + "step": 20964 + }, + { + "epoch": 9.746629474662948, + "grad_norm": 7.182419776916504, + "learning_rate": 9.9790449979045e-06, + "loss": 0.165, + "step": 20965 + }, + { + "epoch": 9.747094374709437, + "grad_norm": 9.579412460327148, + "learning_rate": 9.9790439979044e-06, + "loss": 0.0661, + "step": 20966 + }, + { + "epoch": 9.747559274755927, + "grad_norm": 6.2434186935424805, + "learning_rate": 9.9790429979043e-06, + "loss": 0.084, + "step": 20967 + }, + { + "epoch": 9.748024174802417, + "grad_norm": 5.058647632598877, + "learning_rate": 9.9790419979042e-06, + "loss": 0.0742, + "step": 20968 + }, + { + "epoch": 9.748489074848907, + "grad_norm": 5.4871015548706055, + "learning_rate": 9.979040997904101e-06, + "loss": 0.129, + "step": 20969 + }, + { + "epoch": 9.748953974895397, + "grad_norm": 7.8772501945495605, + "learning_rate": 9.979039997904002e-06, + "loss": 0.0965, + "step": 20970 + }, + { + "epoch": 9.749418874941888, + "grad_norm": 4.067402362823486, + "learning_rate": 9.9790389979039e-06, + "loss": 0.0561, + "step": 20971 + }, + { + "epoch": 9.749883774988378, + "grad_norm": 5.685073375701904, + "learning_rate": 9.979037997903801e-06, + "loss": 0.0293, + "step": 20972 + }, + { + "epoch": 9.750348675034868, + "grad_norm": 7.407691478729248, + "learning_rate": 9.9790369979037e-06, + "loss": 0.1931, + "step": 20973 + }, + { + "epoch": 9.750813575081358, + "grad_norm": 6.861842155456543, + "learning_rate": 9.979035997903601e-06, + "loss": 0.1426, + "step": 20974 + }, + { + "epoch": 9.751278475127847, + "grad_norm": 4.032975673675537, + "learning_rate": 9.9790349979035e-06, + "loss": 0.0904, + "step": 20975 + }, + { + "epoch": 9.751743375174337, + "grad_norm": 1.32848060131073, + "learning_rate": 9.9790339979034e-06, + "loss": 0.0142, + "step": 20976 + }, + { + "epoch": 9.752208275220827, + "grad_norm": 3.1685543060302734, + "learning_rate": 9.979032997903301e-06, + "loss": 0.0288, + "step": 20977 + }, + { + "epoch": 9.752673175267317, + "grad_norm": 4.0156989097595215, + "learning_rate": 9.9790319979032e-06, + "loss": 0.029, + "step": 20978 + }, + { + "epoch": 9.753138075313807, + "grad_norm": 6.9524922370910645, + "learning_rate": 9.9790309979031e-06, + "loss": 0.0898, + "step": 20979 + }, + { + "epoch": 9.753602975360298, + "grad_norm": 7.257271766662598, + "learning_rate": 9.979029997903e-06, + "loss": 0.1299, + "step": 20980 + }, + { + "epoch": 9.754067875406788, + "grad_norm": 2.3487162590026855, + "learning_rate": 9.9790289979029e-06, + "loss": 0.0335, + "step": 20981 + }, + { + "epoch": 9.754532775453278, + "grad_norm": 3.811753988265991, + "learning_rate": 9.979027997902801e-06, + "loss": 0.0664, + "step": 20982 + }, + { + "epoch": 9.754997675499768, + "grad_norm": 5.104008674621582, + "learning_rate": 9.979026997902702e-06, + "loss": 0.1227, + "step": 20983 + }, + { + "epoch": 9.755462575546257, + "grad_norm": 1.47683846950531, + "learning_rate": 9.9790259979026e-06, + "loss": 0.0129, + "step": 20984 + }, + { + "epoch": 9.755927475592747, + "grad_norm": 4.857897758483887, + "learning_rate": 9.979024997902501e-06, + "loss": 0.0485, + "step": 20985 + }, + { + "epoch": 9.756392375639237, + "grad_norm": 2.508101224899292, + "learning_rate": 9.9790239979024e-06, + "loss": 0.0283, + "step": 20986 + }, + { + "epoch": 9.756857275685727, + "grad_norm": 4.571440696716309, + "learning_rate": 9.9790229979023e-06, + "loss": 0.0665, + "step": 20987 + }, + { + "epoch": 9.757322175732218, + "grad_norm": 4.207348823547363, + "learning_rate": 9.9790219979022e-06, + "loss": 0.0427, + "step": 20988 + }, + { + "epoch": 9.757787075778708, + "grad_norm": 3.129661798477173, + "learning_rate": 9.9790209979021e-06, + "loss": 0.0466, + "step": 20989 + }, + { + "epoch": 9.758251975825198, + "grad_norm": 3.234773635864258, + "learning_rate": 9.979019997902001e-06, + "loss": 0.0399, + "step": 20990 + }, + { + "epoch": 9.758716875871688, + "grad_norm": 6.2034759521484375, + "learning_rate": 9.9790189979019e-06, + "loss": 0.0986, + "step": 20991 + }, + { + "epoch": 9.759181775918178, + "grad_norm": 3.626760721206665, + "learning_rate": 9.979017997901801e-06, + "loss": 0.0394, + "step": 20992 + }, + { + "epoch": 9.759646675964667, + "grad_norm": 4.49176549911499, + "learning_rate": 9.9790169979017e-06, + "loss": 0.0942, + "step": 20993 + }, + { + "epoch": 9.760111576011157, + "grad_norm": 6.256497859954834, + "learning_rate": 9.9790159979016e-06, + "loss": 0.1781, + "step": 20994 + }, + { + "epoch": 9.760576476057647, + "grad_norm": 4.940200328826904, + "learning_rate": 9.979014997901501e-06, + "loss": 0.0578, + "step": 20995 + }, + { + "epoch": 9.761041376104137, + "grad_norm": 2.9012510776519775, + "learning_rate": 9.979013997901402e-06, + "loss": 0.0198, + "step": 20996 + }, + { + "epoch": 9.761506276150628, + "grad_norm": 6.119524955749512, + "learning_rate": 9.9790129979013e-06, + "loss": 0.1078, + "step": 20997 + }, + { + "epoch": 9.761971176197118, + "grad_norm": 3.664665937423706, + "learning_rate": 9.9790119979012e-06, + "loss": 0.0362, + "step": 20998 + }, + { + "epoch": 9.762436076243608, + "grad_norm": 4.683517932891846, + "learning_rate": 9.9790109979011e-06, + "loss": 0.0501, + "step": 20999 + }, + { + "epoch": 9.762900976290098, + "grad_norm": 9.813730239868164, + "learning_rate": 9.979009997901001e-06, + "loss": 0.2628, + "step": 21000 + }, + { + "epoch": 9.763365876336588, + "grad_norm": 10.972585678100586, + "learning_rate": 9.9790089979009e-06, + "loss": 0.3483, + "step": 21001 + }, + { + "epoch": 9.763830776383077, + "grad_norm": 4.7593278884887695, + "learning_rate": 9.9790079979008e-06, + "loss": 0.0741, + "step": 21002 + }, + { + "epoch": 9.764295676429567, + "grad_norm": 5.730776309967041, + "learning_rate": 9.979006997900701e-06, + "loss": 0.1191, + "step": 21003 + }, + { + "epoch": 9.764760576476057, + "grad_norm": 2.5625202655792236, + "learning_rate": 9.9790059979006e-06, + "loss": 0.0308, + "step": 21004 + }, + { + "epoch": 9.765225476522549, + "grad_norm": 5.77819299697876, + "learning_rate": 9.979004997900501e-06, + "loss": 0.0931, + "step": 21005 + }, + { + "epoch": 9.765690376569038, + "grad_norm": 4.147116184234619, + "learning_rate": 9.9790039979004e-06, + "loss": 0.0501, + "step": 21006 + }, + { + "epoch": 9.766155276615528, + "grad_norm": 4.472330093383789, + "learning_rate": 9.9790029979003e-06, + "loss": 0.068, + "step": 21007 + }, + { + "epoch": 9.766620176662018, + "grad_norm": 3.591118097305298, + "learning_rate": 9.979001997900201e-06, + "loss": 0.0362, + "step": 21008 + }, + { + "epoch": 9.767085076708508, + "grad_norm": 6.348351955413818, + "learning_rate": 9.9790009979001e-06, + "loss": 0.1566, + "step": 21009 + }, + { + "epoch": 9.767549976754998, + "grad_norm": 3.5835981369018555, + "learning_rate": 9.978999997900001e-06, + "loss": 0.0362, + "step": 21010 + }, + { + "epoch": 9.768014876801487, + "grad_norm": 4.205273628234863, + "learning_rate": 9.9789989978999e-06, + "loss": 0.0428, + "step": 21011 + }, + { + "epoch": 9.768479776847977, + "grad_norm": 8.15931510925293, + "learning_rate": 9.9789979978998e-06, + "loss": 0.1293, + "step": 21012 + }, + { + "epoch": 9.768944676894467, + "grad_norm": 3.1522607803344727, + "learning_rate": 9.978996997899701e-06, + "loss": 0.0334, + "step": 21013 + }, + { + "epoch": 9.769409576940959, + "grad_norm": 2.698057174682617, + "learning_rate": 9.9789959978996e-06, + "loss": 0.0384, + "step": 21014 + }, + { + "epoch": 9.769874476987448, + "grad_norm": 5.908634185791016, + "learning_rate": 9.9789949978995e-06, + "loss": 0.0585, + "step": 21015 + }, + { + "epoch": 9.770339377033938, + "grad_norm": 3.9518823623657227, + "learning_rate": 9.978993997899401e-06, + "loss": 0.0321, + "step": 21016 + }, + { + "epoch": 9.770804277080428, + "grad_norm": 5.238132953643799, + "learning_rate": 9.9789929978993e-06, + "loss": 0.0677, + "step": 21017 + }, + { + "epoch": 9.771269177126918, + "grad_norm": 4.205299377441406, + "learning_rate": 9.978991997899201e-06, + "loss": 0.0662, + "step": 21018 + }, + { + "epoch": 9.771734077173408, + "grad_norm": 4.47968053817749, + "learning_rate": 9.9789909978991e-06, + "loss": 0.1009, + "step": 21019 + }, + { + "epoch": 9.772198977219897, + "grad_norm": 5.025128364562988, + "learning_rate": 9.978989997899e-06, + "loss": 0.0679, + "step": 21020 + }, + { + "epoch": 9.772663877266387, + "grad_norm": 1.7503606081008911, + "learning_rate": 9.978988997898901e-06, + "loss": 0.025, + "step": 21021 + }, + { + "epoch": 9.773128777312877, + "grad_norm": 5.704224109649658, + "learning_rate": 9.9789879978988e-06, + "loss": 0.1002, + "step": 21022 + }, + { + "epoch": 9.773593677359369, + "grad_norm": 5.4449334144592285, + "learning_rate": 9.978986997898701e-06, + "loss": 0.0451, + "step": 21023 + }, + { + "epoch": 9.774058577405858, + "grad_norm": 6.690911769866943, + "learning_rate": 9.9789859978986e-06, + "loss": 0.119, + "step": 21024 + }, + { + "epoch": 9.774523477452348, + "grad_norm": 2.0041344165802, + "learning_rate": 9.9789849978985e-06, + "loss": 0.0161, + "step": 21025 + }, + { + "epoch": 9.774988377498838, + "grad_norm": 1.9901336431503296, + "learning_rate": 9.978983997898401e-06, + "loss": 0.016, + "step": 21026 + }, + { + "epoch": 9.775453277545328, + "grad_norm": 7.038299560546875, + "learning_rate": 9.978982997898302e-06, + "loss": 0.1447, + "step": 21027 + }, + { + "epoch": 9.775918177591818, + "grad_norm": 4.663265228271484, + "learning_rate": 9.9789819978982e-06, + "loss": 0.072, + "step": 21028 + }, + { + "epoch": 9.776383077638307, + "grad_norm": 4.519443035125732, + "learning_rate": 9.978980997898101e-06, + "loss": 0.0477, + "step": 21029 + }, + { + "epoch": 9.776847977684797, + "grad_norm": 8.03745174407959, + "learning_rate": 9.978979997898e-06, + "loss": 0.1741, + "step": 21030 + }, + { + "epoch": 9.777312877731287, + "grad_norm": 7.662503719329834, + "learning_rate": 9.9789789978979e-06, + "loss": 0.1619, + "step": 21031 + }, + { + "epoch": 9.777777777777779, + "grad_norm": 4.9413557052612305, + "learning_rate": 9.9789779978978e-06, + "loss": 0.094, + "step": 21032 + }, + { + "epoch": 9.778242677824268, + "grad_norm": 6.501668453216553, + "learning_rate": 9.9789769978977e-06, + "loss": 0.0707, + "step": 21033 + }, + { + "epoch": 9.778707577870758, + "grad_norm": 9.407234191894531, + "learning_rate": 9.978975997897601e-06, + "loss": 0.3112, + "step": 21034 + }, + { + "epoch": 9.779172477917248, + "grad_norm": 5.6422224044799805, + "learning_rate": 9.9789749978975e-06, + "loss": 0.09, + "step": 21035 + }, + { + "epoch": 9.779637377963738, + "grad_norm": 5.169617176055908, + "learning_rate": 9.978973997897401e-06, + "loss": 0.1363, + "step": 21036 + }, + { + "epoch": 9.780102278010228, + "grad_norm": 4.527475833892822, + "learning_rate": 9.9789729978973e-06, + "loss": 0.0544, + "step": 21037 + }, + { + "epoch": 9.780567178056717, + "grad_norm": 8.840164184570312, + "learning_rate": 9.9789719978972e-06, + "loss": 0.1213, + "step": 21038 + }, + { + "epoch": 9.781032078103207, + "grad_norm": 2.482964515686035, + "learning_rate": 9.978970997897101e-06, + "loss": 0.0257, + "step": 21039 + }, + { + "epoch": 9.781496978149697, + "grad_norm": 2.7401278018951416, + "learning_rate": 9.978969997897002e-06, + "loss": 0.0292, + "step": 21040 + }, + { + "epoch": 9.781961878196189, + "grad_norm": 3.173086404800415, + "learning_rate": 9.978968997896901e-06, + "loss": 0.0406, + "step": 21041 + }, + { + "epoch": 9.782426778242678, + "grad_norm": 3.9491777420043945, + "learning_rate": 9.9789679978968e-06, + "loss": 0.043, + "step": 21042 + }, + { + "epoch": 9.782891678289168, + "grad_norm": 5.675025463104248, + "learning_rate": 9.9789669978967e-06, + "loss": 0.0649, + "step": 21043 + }, + { + "epoch": 9.783356578335658, + "grad_norm": 7.2122015953063965, + "learning_rate": 9.9789659978966e-06, + "loss": 0.0613, + "step": 21044 + }, + { + "epoch": 9.783821478382148, + "grad_norm": 4.683250904083252, + "learning_rate": 9.9789649978965e-06, + "loss": 0.0766, + "step": 21045 + }, + { + "epoch": 9.784286378428638, + "grad_norm": 2.600297212600708, + "learning_rate": 9.9789639978964e-06, + "loss": 0.0331, + "step": 21046 + }, + { + "epoch": 9.784751278475127, + "grad_norm": 9.959797859191895, + "learning_rate": 9.978962997896301e-06, + "loss": 0.2495, + "step": 21047 + }, + { + "epoch": 9.785216178521617, + "grad_norm": 4.825876712799072, + "learning_rate": 9.9789619978962e-06, + "loss": 0.0821, + "step": 21048 + }, + { + "epoch": 9.785681078568107, + "grad_norm": 6.327490329742432, + "learning_rate": 9.978960997896101e-06, + "loss": 0.0827, + "step": 21049 + }, + { + "epoch": 9.786145978614599, + "grad_norm": 7.725550174713135, + "learning_rate": 9.978959997896e-06, + "loss": 0.2364, + "step": 21050 + }, + { + "epoch": 9.786610878661088, + "grad_norm": 3.4352641105651855, + "learning_rate": 9.9789589978959e-06, + "loss": 0.0327, + "step": 21051 + }, + { + "epoch": 9.787075778707578, + "grad_norm": 6.6329450607299805, + "learning_rate": 9.978957997895801e-06, + "loss": 0.1232, + "step": 21052 + }, + { + "epoch": 9.787540678754068, + "grad_norm": 7.2249755859375, + "learning_rate": 9.9789569978957e-06, + "loss": 0.1353, + "step": 21053 + }, + { + "epoch": 9.788005578800558, + "grad_norm": 4.718387603759766, + "learning_rate": 9.978955997895601e-06, + "loss": 0.1089, + "step": 21054 + }, + { + "epoch": 9.788470478847048, + "grad_norm": 5.3153910636901855, + "learning_rate": 9.9789549978955e-06, + "loss": 0.1919, + "step": 21055 + }, + { + "epoch": 9.788935378893537, + "grad_norm": 3.15818190574646, + "learning_rate": 9.9789539978954e-06, + "loss": 0.0497, + "step": 21056 + }, + { + "epoch": 9.789400278940027, + "grad_norm": 3.684957265853882, + "learning_rate": 9.9789529978953e-06, + "loss": 0.0455, + "step": 21057 + }, + { + "epoch": 9.789865178986517, + "grad_norm": 4.711713790893555, + "learning_rate": 9.9789519978952e-06, + "loss": 0.059, + "step": 21058 + }, + { + "epoch": 9.790330079033009, + "grad_norm": 5.649055480957031, + "learning_rate": 9.978950997895101e-06, + "loss": 0.0808, + "step": 21059 + }, + { + "epoch": 9.790794979079498, + "grad_norm": 1.9256938695907593, + "learning_rate": 9.978949997895002e-06, + "loss": 0.0186, + "step": 21060 + }, + { + "epoch": 9.791259879125988, + "grad_norm": 3.723775863647461, + "learning_rate": 9.9789489978949e-06, + "loss": 0.0609, + "step": 21061 + }, + { + "epoch": 9.791724779172478, + "grad_norm": 8.681758880615234, + "learning_rate": 9.9789479978948e-06, + "loss": 0.1174, + "step": 21062 + }, + { + "epoch": 9.792189679218968, + "grad_norm": 7.774898529052734, + "learning_rate": 9.9789469978947e-06, + "loss": 0.1676, + "step": 21063 + }, + { + "epoch": 9.792654579265458, + "grad_norm": 5.361347675323486, + "learning_rate": 9.9789459978946e-06, + "loss": 0.0951, + "step": 21064 + }, + { + "epoch": 9.793119479311947, + "grad_norm": 3.0682201385498047, + "learning_rate": 9.978944997894501e-06, + "loss": 0.0593, + "step": 21065 + }, + { + "epoch": 9.793584379358437, + "grad_norm": 5.8766374588012695, + "learning_rate": 9.9789439978944e-06, + "loss": 0.0995, + "step": 21066 + }, + { + "epoch": 9.794049279404929, + "grad_norm": 3.9859189987182617, + "learning_rate": 9.978942997894301e-06, + "loss": 0.0452, + "step": 21067 + }, + { + "epoch": 9.794514179451419, + "grad_norm": 4.532566070556641, + "learning_rate": 9.9789419978942e-06, + "loss": 0.0548, + "step": 21068 + }, + { + "epoch": 9.794979079497908, + "grad_norm": 6.718008995056152, + "learning_rate": 9.9789409978941e-06, + "loss": 0.112, + "step": 21069 + }, + { + "epoch": 9.795443979544398, + "grad_norm": 4.705626487731934, + "learning_rate": 9.978939997894e-06, + "loss": 0.0749, + "step": 21070 + }, + { + "epoch": 9.795908879590888, + "grad_norm": 7.2014570236206055, + "learning_rate": 9.9789389978939e-06, + "loss": 0.1165, + "step": 21071 + }, + { + "epoch": 9.796373779637378, + "grad_norm": 4.061730861663818, + "learning_rate": 9.978937997893801e-06, + "loss": 0.076, + "step": 21072 + }, + { + "epoch": 9.796838679683868, + "grad_norm": 3.4100656509399414, + "learning_rate": 9.9789369978937e-06, + "loss": 0.0367, + "step": 21073 + }, + { + "epoch": 9.797303579730357, + "grad_norm": 5.315697193145752, + "learning_rate": 9.9789359978936e-06, + "loss": 0.0782, + "step": 21074 + }, + { + "epoch": 9.797768479776847, + "grad_norm": 4.5932817459106445, + "learning_rate": 9.9789349978935e-06, + "loss": 0.0719, + "step": 21075 + }, + { + "epoch": 9.798233379823339, + "grad_norm": 4.063678741455078, + "learning_rate": 9.9789339978934e-06, + "loss": 0.0418, + "step": 21076 + }, + { + "epoch": 9.798698279869829, + "grad_norm": 6.368699073791504, + "learning_rate": 9.978932997893301e-06, + "loss": 0.1984, + "step": 21077 + }, + { + "epoch": 9.799163179916318, + "grad_norm": 6.479359149932861, + "learning_rate": 9.978931997893202e-06, + "loss": 0.154, + "step": 21078 + }, + { + "epoch": 9.799628079962808, + "grad_norm": 5.146584510803223, + "learning_rate": 9.9789309978931e-06, + "loss": 0.0558, + "step": 21079 + }, + { + "epoch": 9.800092980009298, + "grad_norm": 6.305038928985596, + "learning_rate": 9.978929997893001e-06, + "loss": 0.1069, + "step": 21080 + }, + { + "epoch": 9.800557880055788, + "grad_norm": 3.2289249897003174, + "learning_rate": 9.9789289978929e-06, + "loss": 0.029, + "step": 21081 + }, + { + "epoch": 9.801022780102278, + "grad_norm": 6.013680934906006, + "learning_rate": 9.9789279978928e-06, + "loss": 0.0786, + "step": 21082 + }, + { + "epoch": 9.801487680148767, + "grad_norm": 6.217981815338135, + "learning_rate": 9.9789269978927e-06, + "loss": 0.1534, + "step": 21083 + }, + { + "epoch": 9.801952580195259, + "grad_norm": 1.8818819522857666, + "learning_rate": 9.9789259978926e-06, + "loss": 0.0164, + "step": 21084 + }, + { + "epoch": 9.802417480241749, + "grad_norm": 5.133860111236572, + "learning_rate": 9.978924997892501e-06, + "loss": 0.04, + "step": 21085 + }, + { + "epoch": 9.802882380288239, + "grad_norm": 6.842164993286133, + "learning_rate": 9.9789239978924e-06, + "loss": 0.1798, + "step": 21086 + }, + { + "epoch": 9.803347280334728, + "grad_norm": 2.2869515419006348, + "learning_rate": 9.9789229978923e-06, + "loss": 0.0335, + "step": 21087 + }, + { + "epoch": 9.803812180381218, + "grad_norm": 3.633521556854248, + "learning_rate": 9.9789219978922e-06, + "loss": 0.0443, + "step": 21088 + }, + { + "epoch": 9.804277080427708, + "grad_norm": 3.1855242252349854, + "learning_rate": 9.9789209978921e-06, + "loss": 0.0395, + "step": 21089 + }, + { + "epoch": 9.804741980474198, + "grad_norm": 2.959108829498291, + "learning_rate": 9.978919997892001e-06, + "loss": 0.0213, + "step": 21090 + }, + { + "epoch": 9.805206880520688, + "grad_norm": 3.7849109172821045, + "learning_rate": 9.978918997891902e-06, + "loss": 0.0273, + "step": 21091 + }, + { + "epoch": 9.805671780567177, + "grad_norm": 6.156929969787598, + "learning_rate": 9.9789179978918e-06, + "loss": 0.1147, + "step": 21092 + }, + { + "epoch": 9.806136680613669, + "grad_norm": 6.241135597229004, + "learning_rate": 9.978916997891701e-06, + "loss": 0.1101, + "step": 21093 + }, + { + "epoch": 9.806601580660159, + "grad_norm": 5.669447898864746, + "learning_rate": 9.9789159978916e-06, + "loss": 0.1872, + "step": 21094 + }, + { + "epoch": 9.807066480706649, + "grad_norm": 9.71066951751709, + "learning_rate": 9.978914997891501e-06, + "loss": 0.1251, + "step": 21095 + }, + { + "epoch": 9.807531380753138, + "grad_norm": 4.295638084411621, + "learning_rate": 9.978913997891402e-06, + "loss": 0.0735, + "step": 21096 + }, + { + "epoch": 9.807996280799628, + "grad_norm": 5.614317417144775, + "learning_rate": 9.9789129978913e-06, + "loss": 0.1148, + "step": 21097 + }, + { + "epoch": 9.808461180846118, + "grad_norm": 1.897843837738037, + "learning_rate": 9.978911997891201e-06, + "loss": 0.022, + "step": 21098 + }, + { + "epoch": 9.808926080892608, + "grad_norm": 4.147624969482422, + "learning_rate": 9.9789109978911e-06, + "loss": 0.0613, + "step": 21099 + }, + { + "epoch": 9.809390980939098, + "grad_norm": 6.742602348327637, + "learning_rate": 9.978909997891e-06, + "loss": 0.1173, + "step": 21100 + }, + { + "epoch": 9.809855880985587, + "grad_norm": 6.288963317871094, + "learning_rate": 9.9789089978909e-06, + "loss": 0.0836, + "step": 21101 + }, + { + "epoch": 9.810320781032079, + "grad_norm": 4.519094467163086, + "learning_rate": 9.9789079978908e-06, + "loss": 0.0685, + "step": 21102 + }, + { + "epoch": 9.810785681078569, + "grad_norm": 6.670717716217041, + "learning_rate": 9.978906997890701e-06, + "loss": 0.0937, + "step": 21103 + }, + { + "epoch": 9.811250581125059, + "grad_norm": 2.6269333362579346, + "learning_rate": 9.978905997890602e-06, + "loss": 0.0622, + "step": 21104 + }, + { + "epoch": 9.811715481171548, + "grad_norm": 5.677857875823975, + "learning_rate": 9.9789049978905e-06, + "loss": 0.1153, + "step": 21105 + }, + { + "epoch": 9.812180381218038, + "grad_norm": 6.156062126159668, + "learning_rate": 9.9789039978904e-06, + "loss": 0.0939, + "step": 21106 + }, + { + "epoch": 9.812645281264528, + "grad_norm": 5.467942714691162, + "learning_rate": 9.9789029978903e-06, + "loss": 0.1209, + "step": 21107 + }, + { + "epoch": 9.813110181311018, + "grad_norm": 2.68790602684021, + "learning_rate": 9.978901997890201e-06, + "loss": 0.0303, + "step": 21108 + }, + { + "epoch": 9.813575081357508, + "grad_norm": 3.34991455078125, + "learning_rate": 9.978900997890102e-06, + "loss": 0.0216, + "step": 21109 + }, + { + "epoch": 9.814039981403997, + "grad_norm": 3.123628854751587, + "learning_rate": 9.97889999789e-06, + "loss": 0.0345, + "step": 21110 + }, + { + "epoch": 9.814504881450489, + "grad_norm": 4.936289310455322, + "learning_rate": 9.978898997889901e-06, + "loss": 0.1156, + "step": 21111 + }, + { + "epoch": 9.814969781496979, + "grad_norm": 3.380723714828491, + "learning_rate": 9.9788979978898e-06, + "loss": 0.0354, + "step": 21112 + }, + { + "epoch": 9.815434681543469, + "grad_norm": 3.5421154499053955, + "learning_rate": 9.9788969978897e-06, + "loss": 0.0346, + "step": 21113 + }, + { + "epoch": 9.815899581589958, + "grad_norm": 2.5582692623138428, + "learning_rate": 9.9788959978896e-06, + "loss": 0.0358, + "step": 21114 + }, + { + "epoch": 9.816364481636448, + "grad_norm": 5.002018928527832, + "learning_rate": 9.9788949978895e-06, + "loss": 0.0801, + "step": 21115 + }, + { + "epoch": 9.816829381682938, + "grad_norm": 4.375988960266113, + "learning_rate": 9.978893997889401e-06, + "loss": 0.0664, + "step": 21116 + }, + { + "epoch": 9.817294281729428, + "grad_norm": 3.4844698905944824, + "learning_rate": 9.9788929978893e-06, + "loss": 0.0438, + "step": 21117 + }, + { + "epoch": 9.817759181775918, + "grad_norm": 5.783759593963623, + "learning_rate": 9.9788919978892e-06, + "loss": 0.1049, + "step": 21118 + }, + { + "epoch": 9.818224081822407, + "grad_norm": 8.155567169189453, + "learning_rate": 9.9788909978891e-06, + "loss": 0.2304, + "step": 21119 + }, + { + "epoch": 9.818688981868899, + "grad_norm": 4.3672895431518555, + "learning_rate": 9.978889997889e-06, + "loss": 0.0394, + "step": 21120 + }, + { + "epoch": 9.819153881915389, + "grad_norm": 4.310476303100586, + "learning_rate": 9.978888997888901e-06, + "loss": 0.0432, + "step": 21121 + }, + { + "epoch": 9.819618781961879, + "grad_norm": 5.821474552154541, + "learning_rate": 9.978887997888802e-06, + "loss": 0.1021, + "step": 21122 + }, + { + "epoch": 9.820083682008368, + "grad_norm": 7.739163875579834, + "learning_rate": 9.9788869978887e-06, + "loss": 0.1362, + "step": 21123 + }, + { + "epoch": 9.820548582054858, + "grad_norm": 8.902420997619629, + "learning_rate": 9.978885997888601e-06, + "loss": 0.3004, + "step": 21124 + }, + { + "epoch": 9.821013482101348, + "grad_norm": 5.86708927154541, + "learning_rate": 9.9788849978885e-06, + "loss": 0.1331, + "step": 21125 + }, + { + "epoch": 9.821478382147838, + "grad_norm": 6.827130317687988, + "learning_rate": 9.978883997888401e-06, + "loss": 0.1452, + "step": 21126 + }, + { + "epoch": 9.821943282194328, + "grad_norm": 5.85690450668335, + "learning_rate": 9.9788829978883e-06, + "loss": 0.0936, + "step": 21127 + }, + { + "epoch": 9.822408182240817, + "grad_norm": 3.6541550159454346, + "learning_rate": 9.9788819978882e-06, + "loss": 0.0634, + "step": 21128 + }, + { + "epoch": 9.822873082287309, + "grad_norm": 9.569011688232422, + "learning_rate": 9.978880997888101e-06, + "loss": 0.1117, + "step": 21129 + }, + { + "epoch": 9.823337982333799, + "grad_norm": 6.965851783752441, + "learning_rate": 9.978879997888e-06, + "loss": 0.1897, + "step": 21130 + }, + { + "epoch": 9.823802882380289, + "grad_norm": 2.759683609008789, + "learning_rate": 9.9788789978879e-06, + "loss": 0.0155, + "step": 21131 + }, + { + "epoch": 9.824267782426778, + "grad_norm": 4.717240333557129, + "learning_rate": 9.9788779978878e-06, + "loss": 0.0993, + "step": 21132 + }, + { + "epoch": 9.824732682473268, + "grad_norm": 5.262267112731934, + "learning_rate": 9.9788769978877e-06, + "loss": 0.0511, + "step": 21133 + }, + { + "epoch": 9.825197582519758, + "grad_norm": 3.431713342666626, + "learning_rate": 9.978875997887601e-06, + "loss": 0.0236, + "step": 21134 + }, + { + "epoch": 9.825662482566248, + "grad_norm": 3.548814296722412, + "learning_rate": 9.978874997887502e-06, + "loss": 0.0308, + "step": 21135 + }, + { + "epoch": 9.826127382612738, + "grad_norm": 3.7396318912506104, + "learning_rate": 9.9788739978874e-06, + "loss": 0.0455, + "step": 21136 + }, + { + "epoch": 9.826592282659227, + "grad_norm": 4.848369598388672, + "learning_rate": 9.978872997887301e-06, + "loss": 0.0472, + "step": 21137 + }, + { + "epoch": 9.827057182705719, + "grad_norm": 5.483652114868164, + "learning_rate": 9.9788719978872e-06, + "loss": 0.0783, + "step": 21138 + }, + { + "epoch": 9.827522082752209, + "grad_norm": 6.122921466827393, + "learning_rate": 9.9788709978871e-06, + "loss": 0.0702, + "step": 21139 + }, + { + "epoch": 9.827986982798699, + "grad_norm": 5.9487504959106445, + "learning_rate": 9.978869997887e-06, + "loss": 0.0731, + "step": 21140 + }, + { + "epoch": 9.828451882845188, + "grad_norm": 6.974423885345459, + "learning_rate": 9.9788689978869e-06, + "loss": 0.1605, + "step": 21141 + }, + { + "epoch": 9.828916782891678, + "grad_norm": 4.167792320251465, + "learning_rate": 9.978867997886801e-06, + "loss": 0.0441, + "step": 21142 + }, + { + "epoch": 9.829381682938168, + "grad_norm": 8.361384391784668, + "learning_rate": 9.9788669978867e-06, + "loss": 0.0747, + "step": 21143 + }, + { + "epoch": 9.829846582984658, + "grad_norm": 4.687295436859131, + "learning_rate": 9.978865997886601e-06, + "loss": 0.0866, + "step": 21144 + }, + { + "epoch": 9.830311483031148, + "grad_norm": 7.442080497741699, + "learning_rate": 9.9788649978865e-06, + "loss": 0.175, + "step": 21145 + }, + { + "epoch": 9.83077638307764, + "grad_norm": 4.069100379943848, + "learning_rate": 9.9788639978864e-06, + "loss": 0.0305, + "step": 21146 + }, + { + "epoch": 9.831241283124129, + "grad_norm": 6.906838893890381, + "learning_rate": 9.978862997886301e-06, + "loss": 0.2129, + "step": 21147 + }, + { + "epoch": 9.831706183170619, + "grad_norm": 7.020123481750488, + "learning_rate": 9.978861997886202e-06, + "loss": 0.2506, + "step": 21148 + }, + { + "epoch": 9.832171083217109, + "grad_norm": 5.762636661529541, + "learning_rate": 9.9788609978861e-06, + "loss": 0.0559, + "step": 21149 + }, + { + "epoch": 9.832635983263598, + "grad_norm": 4.872121334075928, + "learning_rate": 9.978859997886e-06, + "loss": 0.1074, + "step": 21150 + }, + { + "epoch": 9.833100883310088, + "grad_norm": 11.378717422485352, + "learning_rate": 9.9788589978859e-06, + "loss": 0.2407, + "step": 21151 + }, + { + "epoch": 9.833565783356578, + "grad_norm": 6.472724437713623, + "learning_rate": 9.9788579978858e-06, + "loss": 0.125, + "step": 21152 + }, + { + "epoch": 9.834030683403068, + "grad_norm": 7.432314872741699, + "learning_rate": 9.9788569978857e-06, + "loss": 0.1029, + "step": 21153 + }, + { + "epoch": 9.834495583449558, + "grad_norm": 6.138179302215576, + "learning_rate": 9.9788559978856e-06, + "loss": 0.153, + "step": 21154 + }, + { + "epoch": 9.83496048349605, + "grad_norm": 3.7296411991119385, + "learning_rate": 9.978854997885501e-06, + "loss": 0.0724, + "step": 21155 + }, + { + "epoch": 9.835425383542539, + "grad_norm": 3.562004804611206, + "learning_rate": 9.9788539978854e-06, + "loss": 0.0636, + "step": 21156 + }, + { + "epoch": 9.835890283589029, + "grad_norm": 5.627786636352539, + "learning_rate": 9.978852997885301e-06, + "loss": 0.0712, + "step": 21157 + }, + { + "epoch": 9.836355183635519, + "grad_norm": 3.3715808391571045, + "learning_rate": 9.9788519978852e-06, + "loss": 0.0322, + "step": 21158 + }, + { + "epoch": 9.836820083682008, + "grad_norm": 4.349445819854736, + "learning_rate": 9.9788509978851e-06, + "loss": 0.0618, + "step": 21159 + }, + { + "epoch": 9.837284983728498, + "grad_norm": 3.2940163612365723, + "learning_rate": 9.978849997885001e-06, + "loss": 0.0452, + "step": 21160 + }, + { + "epoch": 9.837749883774988, + "grad_norm": 6.075416088104248, + "learning_rate": 9.9788489978849e-06, + "loss": 0.1075, + "step": 21161 + }, + { + "epoch": 9.838214783821478, + "grad_norm": 4.198498249053955, + "learning_rate": 9.978847997884801e-06, + "loss": 0.0842, + "step": 21162 + }, + { + "epoch": 9.83867968386797, + "grad_norm": 1.5443928241729736, + "learning_rate": 9.9788469978847e-06, + "loss": 0.0154, + "step": 21163 + }, + { + "epoch": 9.83914458391446, + "grad_norm": 1.9908527135849, + "learning_rate": 9.9788459978846e-06, + "loss": 0.0193, + "step": 21164 + }, + { + "epoch": 9.839609483960949, + "grad_norm": 6.968294620513916, + "learning_rate": 9.978844997884501e-06, + "loss": 0.2187, + "step": 21165 + }, + { + "epoch": 9.840074384007439, + "grad_norm": 2.6591391563415527, + "learning_rate": 9.9788439978844e-06, + "loss": 0.0392, + "step": 21166 + }, + { + "epoch": 9.840539284053929, + "grad_norm": 6.002838611602783, + "learning_rate": 9.9788429978843e-06, + "loss": 0.1062, + "step": 21167 + }, + { + "epoch": 9.841004184100418, + "grad_norm": 3.157243490219116, + "learning_rate": 9.978841997884201e-06, + "loss": 0.0265, + "step": 21168 + }, + { + "epoch": 9.841469084146908, + "grad_norm": 7.1884002685546875, + "learning_rate": 9.9788409978841e-06, + "loss": 0.1076, + "step": 21169 + }, + { + "epoch": 9.841933984193398, + "grad_norm": 5.086110591888428, + "learning_rate": 9.978839997884001e-06, + "loss": 0.0791, + "step": 21170 + }, + { + "epoch": 9.842398884239888, + "grad_norm": 3.804459571838379, + "learning_rate": 9.9788389978839e-06, + "loss": 0.0586, + "step": 21171 + }, + { + "epoch": 9.84286378428638, + "grad_norm": 5.16815185546875, + "learning_rate": 9.9788379978838e-06, + "loss": 0.0854, + "step": 21172 + }, + { + "epoch": 9.84332868433287, + "grad_norm": 10.713358879089355, + "learning_rate": 9.978836997883701e-06, + "loss": 0.1072, + "step": 21173 + }, + { + "epoch": 9.843793584379359, + "grad_norm": 4.098247528076172, + "learning_rate": 9.9788359978836e-06, + "loss": 0.1065, + "step": 21174 + }, + { + "epoch": 9.844258484425849, + "grad_norm": 1.028538465499878, + "learning_rate": 9.978834997883501e-06, + "loss": 0.0096, + "step": 21175 + }, + { + "epoch": 9.844723384472339, + "grad_norm": 7.499086856842041, + "learning_rate": 9.9788339978834e-06, + "loss": 0.0946, + "step": 21176 + }, + { + "epoch": 9.845188284518828, + "grad_norm": 6.838472366333008, + "learning_rate": 9.9788329978833e-06, + "loss": 0.0974, + "step": 21177 + }, + { + "epoch": 9.845653184565318, + "grad_norm": 4.038396835327148, + "learning_rate": 9.978831997883201e-06, + "loss": 0.0437, + "step": 21178 + }, + { + "epoch": 9.846118084611808, + "grad_norm": 6.131182670593262, + "learning_rate": 9.9788309978831e-06, + "loss": 0.0446, + "step": 21179 + }, + { + "epoch": 9.846582984658298, + "grad_norm": 6.5545525550842285, + "learning_rate": 9.978829997883001e-06, + "loss": 0.121, + "step": 21180 + }, + { + "epoch": 9.84704788470479, + "grad_norm": 2.447727918624878, + "learning_rate": 9.978828997882902e-06, + "loss": 0.0266, + "step": 21181 + }, + { + "epoch": 9.84751278475128, + "grad_norm": 11.262911796569824, + "learning_rate": 9.9788279978828e-06, + "loss": 0.2053, + "step": 21182 + }, + { + "epoch": 9.847977684797769, + "grad_norm": 4.885281562805176, + "learning_rate": 9.9788269978827e-06, + "loss": 0.0721, + "step": 21183 + }, + { + "epoch": 9.848442584844259, + "grad_norm": 6.417872905731201, + "learning_rate": 9.9788259978826e-06, + "loss": 0.111, + "step": 21184 + }, + { + "epoch": 9.848907484890749, + "grad_norm": 4.693007469177246, + "learning_rate": 9.9788249978825e-06, + "loss": 0.0678, + "step": 21185 + }, + { + "epoch": 9.849372384937238, + "grad_norm": 6.037665367126465, + "learning_rate": 9.978823997882401e-06, + "loss": 0.1214, + "step": 21186 + }, + { + "epoch": 9.849837284983728, + "grad_norm": 9.721650123596191, + "learning_rate": 9.9788229978823e-06, + "loss": 0.1956, + "step": 21187 + }, + { + "epoch": 9.850302185030218, + "grad_norm": 7.487035274505615, + "learning_rate": 9.978821997882201e-06, + "loss": 0.19, + "step": 21188 + }, + { + "epoch": 9.850767085076708, + "grad_norm": 3.9841089248657227, + "learning_rate": 9.9788209978821e-06, + "loss": 0.0756, + "step": 21189 + }, + { + "epoch": 9.8512319851232, + "grad_norm": 4.536160945892334, + "learning_rate": 9.978819997882e-06, + "loss": 0.0638, + "step": 21190 + }, + { + "epoch": 9.851696885169689, + "grad_norm": 7.294368267059326, + "learning_rate": 9.978818997881901e-06, + "loss": 0.1577, + "step": 21191 + }, + { + "epoch": 9.852161785216179, + "grad_norm": 2.508735179901123, + "learning_rate": 9.9788179978818e-06, + "loss": 0.0296, + "step": 21192 + }, + { + "epoch": 9.852626685262669, + "grad_norm": 4.861696720123291, + "learning_rate": 9.978816997881701e-06, + "loss": 0.0571, + "step": 21193 + }, + { + "epoch": 9.853091585309159, + "grad_norm": 3.963853597640991, + "learning_rate": 9.9788159978816e-06, + "loss": 0.0338, + "step": 21194 + }, + { + "epoch": 9.853556485355648, + "grad_norm": 3.6595215797424316, + "learning_rate": 9.9788149978815e-06, + "loss": 0.0782, + "step": 21195 + }, + { + "epoch": 9.854021385402138, + "grad_norm": 3.6747658252716064, + "learning_rate": 9.9788139978814e-06, + "loss": 0.0574, + "step": 21196 + }, + { + "epoch": 9.854486285448628, + "grad_norm": 5.1924591064453125, + "learning_rate": 9.9788129978813e-06, + "loss": 0.0824, + "step": 21197 + }, + { + "epoch": 9.854951185495118, + "grad_norm": 4.62623929977417, + "learning_rate": 9.978811997881201e-06, + "loss": 0.1015, + "step": 21198 + }, + { + "epoch": 9.85541608554161, + "grad_norm": 4.180566787719727, + "learning_rate": 9.978810997881102e-06, + "loss": 0.0973, + "step": 21199 + }, + { + "epoch": 9.855880985588099, + "grad_norm": 4.400595664978027, + "learning_rate": 9.978809997881e-06, + "loss": 0.0758, + "step": 21200 + }, + { + "epoch": 9.856345885634589, + "grad_norm": 8.238381385803223, + "learning_rate": 9.978808997880901e-06, + "loss": 0.2127, + "step": 21201 + }, + { + "epoch": 9.856810785681079, + "grad_norm": 4.901917934417725, + "learning_rate": 9.9788079978808e-06, + "loss": 0.0485, + "step": 21202 + }, + { + "epoch": 9.857275685727569, + "grad_norm": 5.935704231262207, + "learning_rate": 9.9788069978807e-06, + "loss": 0.1054, + "step": 21203 + }, + { + "epoch": 9.857740585774058, + "grad_norm": 4.948586463928223, + "learning_rate": 9.978805997880601e-06, + "loss": 0.0862, + "step": 21204 + }, + { + "epoch": 9.858205485820548, + "grad_norm": 4.51542329788208, + "learning_rate": 9.9788049978805e-06, + "loss": 0.075, + "step": 21205 + }, + { + "epoch": 9.858670385867038, + "grad_norm": 1.5936970710754395, + "learning_rate": 9.978803997880401e-06, + "loss": 0.0165, + "step": 21206 + }, + { + "epoch": 9.859135285913528, + "grad_norm": 3.1094706058502197, + "learning_rate": 9.9788029978803e-06, + "loss": 0.0541, + "step": 21207 + }, + { + "epoch": 9.85960018596002, + "grad_norm": 8.471334457397461, + "learning_rate": 9.9788019978802e-06, + "loss": 0.0861, + "step": 21208 + }, + { + "epoch": 9.860065086006509, + "grad_norm": 4.919710636138916, + "learning_rate": 9.9788009978801e-06, + "loss": 0.1247, + "step": 21209 + }, + { + "epoch": 9.860529986052999, + "grad_norm": 3.2065529823303223, + "learning_rate": 9.97879999788e-06, + "loss": 0.0425, + "step": 21210 + }, + { + "epoch": 9.860994886099489, + "grad_norm": 5.541230201721191, + "learning_rate": 9.978798997879901e-06, + "loss": 0.0823, + "step": 21211 + }, + { + "epoch": 9.861459786145979, + "grad_norm": 6.420576572418213, + "learning_rate": 9.978797997879802e-06, + "loss": 0.1118, + "step": 21212 + }, + { + "epoch": 9.861924686192468, + "grad_norm": 2.811800718307495, + "learning_rate": 9.9787969978797e-06, + "loss": 0.0436, + "step": 21213 + }, + { + "epoch": 9.862389586238958, + "grad_norm": 4.997034072875977, + "learning_rate": 9.9787959978796e-06, + "loss": 0.0519, + "step": 21214 + }, + { + "epoch": 9.862854486285448, + "grad_norm": 6.2944865226745605, + "learning_rate": 9.9787949978795e-06, + "loss": 0.1107, + "step": 21215 + }, + { + "epoch": 9.863319386331938, + "grad_norm": 3.72902250289917, + "learning_rate": 9.9787939978794e-06, + "loss": 0.1002, + "step": 21216 + }, + { + "epoch": 9.86378428637843, + "grad_norm": 5.578247547149658, + "learning_rate": 9.978792997879301e-06, + "loss": 0.0635, + "step": 21217 + }, + { + "epoch": 9.864249186424919, + "grad_norm": 2.9444198608398438, + "learning_rate": 9.9787919978792e-06, + "loss": 0.0377, + "step": 21218 + }, + { + "epoch": 9.864714086471409, + "grad_norm": 4.745115280151367, + "learning_rate": 9.978790997879101e-06, + "loss": 0.0864, + "step": 21219 + }, + { + "epoch": 9.865178986517899, + "grad_norm": 6.648590087890625, + "learning_rate": 9.978789997879e-06, + "loss": 0.1932, + "step": 21220 + }, + { + "epoch": 9.865643886564389, + "grad_norm": 4.105435848236084, + "learning_rate": 9.9787889978789e-06, + "loss": 0.0829, + "step": 21221 + }, + { + "epoch": 9.866108786610878, + "grad_norm": 7.855881214141846, + "learning_rate": 9.9787879978788e-06, + "loss": 0.1578, + "step": 21222 + }, + { + "epoch": 9.866573686657368, + "grad_norm": 5.687683582305908, + "learning_rate": 9.9787869978787e-06, + "loss": 0.0848, + "step": 21223 + }, + { + "epoch": 9.867038586703858, + "grad_norm": 8.879483222961426, + "learning_rate": 9.978785997878601e-06, + "loss": 0.1989, + "step": 21224 + }, + { + "epoch": 9.86750348675035, + "grad_norm": 4.301532745361328, + "learning_rate": 9.9787849978785e-06, + "loss": 0.0928, + "step": 21225 + }, + { + "epoch": 9.86796838679684, + "grad_norm": 5.43212890625, + "learning_rate": 9.9787839978784e-06, + "loss": 0.1002, + "step": 21226 + }, + { + "epoch": 9.868433286843329, + "grad_norm": 1.9733219146728516, + "learning_rate": 9.9787829978783e-06, + "loss": 0.0313, + "step": 21227 + }, + { + "epoch": 9.868898186889819, + "grad_norm": 9.736701011657715, + "learning_rate": 9.9787819978782e-06, + "loss": 0.2255, + "step": 21228 + }, + { + "epoch": 9.869363086936309, + "grad_norm": 3.871257781982422, + "learning_rate": 9.978780997878101e-06, + "loss": 0.0552, + "step": 21229 + }, + { + "epoch": 9.869827986982799, + "grad_norm": 5.302865028381348, + "learning_rate": 9.978779997878002e-06, + "loss": 0.0881, + "step": 21230 + }, + { + "epoch": 9.870292887029288, + "grad_norm": 5.287339210510254, + "learning_rate": 9.9787789978779e-06, + "loss": 0.0916, + "step": 21231 + }, + { + "epoch": 9.870757787075778, + "grad_norm": 3.5199668407440186, + "learning_rate": 9.978777997877801e-06, + "loss": 0.0374, + "step": 21232 + }, + { + "epoch": 9.871222687122268, + "grad_norm": 6.415590286254883, + "learning_rate": 9.9787769978777e-06, + "loss": 0.1233, + "step": 21233 + }, + { + "epoch": 9.87168758716876, + "grad_norm": 4.322881698608398, + "learning_rate": 9.9787759978776e-06, + "loss": 0.0396, + "step": 21234 + }, + { + "epoch": 9.87215248721525, + "grad_norm": 3.1186366081237793, + "learning_rate": 9.9787749978775e-06, + "loss": 0.0392, + "step": 21235 + }, + { + "epoch": 9.872617387261739, + "grad_norm": 3.494889736175537, + "learning_rate": 9.9787739978774e-06, + "loss": 0.0428, + "step": 21236 + }, + { + "epoch": 9.873082287308229, + "grad_norm": 6.256826877593994, + "learning_rate": 9.978772997877301e-06, + "loss": 0.1123, + "step": 21237 + }, + { + "epoch": 9.873547187354719, + "grad_norm": 8.13486099243164, + "learning_rate": 9.9787719978772e-06, + "loss": 0.1751, + "step": 21238 + }, + { + "epoch": 9.874012087401209, + "grad_norm": 2.8936963081359863, + "learning_rate": 9.9787709978771e-06, + "loss": 0.0507, + "step": 21239 + }, + { + "epoch": 9.874476987447698, + "grad_norm": 8.36504077911377, + "learning_rate": 9.978769997877e-06, + "loss": 0.2109, + "step": 21240 + }, + { + "epoch": 9.874941887494188, + "grad_norm": 3.157754898071289, + "learning_rate": 9.9787689978769e-06, + "loss": 0.0383, + "step": 21241 + }, + { + "epoch": 9.87540678754068, + "grad_norm": 1.5635887384414673, + "learning_rate": 9.978767997876801e-06, + "loss": 0.0148, + "step": 21242 + }, + { + "epoch": 9.87587168758717, + "grad_norm": 4.5490546226501465, + "learning_rate": 9.978766997876702e-06, + "loss": 0.0912, + "step": 21243 + }, + { + "epoch": 9.87633658763366, + "grad_norm": 2.3589913845062256, + "learning_rate": 9.9787659978766e-06, + "loss": 0.0265, + "step": 21244 + }, + { + "epoch": 9.876801487680149, + "grad_norm": 3.933955430984497, + "learning_rate": 9.978764997876501e-06, + "loss": 0.0338, + "step": 21245 + }, + { + "epoch": 9.877266387726639, + "grad_norm": 2.3778679370880127, + "learning_rate": 9.9787639978764e-06, + "loss": 0.0179, + "step": 21246 + }, + { + "epoch": 9.877731287773129, + "grad_norm": 2.0338895320892334, + "learning_rate": 9.978762997876301e-06, + "loss": 0.0189, + "step": 21247 + }, + { + "epoch": 9.878196187819619, + "grad_norm": 1.4276902675628662, + "learning_rate": 9.9787619978762e-06, + "loss": 0.0136, + "step": 21248 + }, + { + "epoch": 9.878661087866108, + "grad_norm": 4.084539413452148, + "learning_rate": 9.9787609978761e-06, + "loss": 0.0513, + "step": 21249 + }, + { + "epoch": 9.879125987912598, + "grad_norm": 4.575325965881348, + "learning_rate": 9.978759997876001e-06, + "loss": 0.0894, + "step": 21250 + }, + { + "epoch": 9.87959088795909, + "grad_norm": 6.386226177215576, + "learning_rate": 9.9787589978759e-06, + "loss": 0.0814, + "step": 21251 + }, + { + "epoch": 9.88005578800558, + "grad_norm": 4.79241418838501, + "learning_rate": 9.9787579978758e-06, + "loss": 0.0493, + "step": 21252 + }, + { + "epoch": 9.88052068805207, + "grad_norm": 4.028398036956787, + "learning_rate": 9.9787569978757e-06, + "loss": 0.0488, + "step": 21253 + }, + { + "epoch": 9.880985588098559, + "grad_norm": 8.8922700881958, + "learning_rate": 9.9787559978756e-06, + "loss": 0.171, + "step": 21254 + }, + { + "epoch": 9.881450488145049, + "grad_norm": 5.019423007965088, + "learning_rate": 9.978754997875501e-06, + "loss": 0.115, + "step": 21255 + }, + { + "epoch": 9.881915388191539, + "grad_norm": 7.447957992553711, + "learning_rate": 9.978753997875402e-06, + "loss": 0.1593, + "step": 21256 + }, + { + "epoch": 9.882380288238028, + "grad_norm": 3.903390884399414, + "learning_rate": 9.9787529978753e-06, + "loss": 0.0562, + "step": 21257 + }, + { + "epoch": 9.882845188284518, + "grad_norm": 3.7581887245178223, + "learning_rate": 9.9787519978752e-06, + "loss": 0.0413, + "step": 21258 + }, + { + "epoch": 9.883310088331008, + "grad_norm": 6.058261871337891, + "learning_rate": 9.9787509978751e-06, + "loss": 0.1204, + "step": 21259 + }, + { + "epoch": 9.8837749883775, + "grad_norm": 7.175146579742432, + "learning_rate": 9.978749997875001e-06, + "loss": 0.169, + "step": 21260 + }, + { + "epoch": 9.88423988842399, + "grad_norm": 5.578411102294922, + "learning_rate": 9.978748997874902e-06, + "loss": 0.0714, + "step": 21261 + }, + { + "epoch": 9.88470478847048, + "grad_norm": 8.324292182922363, + "learning_rate": 9.9787479978748e-06, + "loss": 0.1496, + "step": 21262 + }, + { + "epoch": 9.885169688516969, + "grad_norm": 4.120300769805908, + "learning_rate": 9.978746997874701e-06, + "loss": 0.0687, + "step": 21263 + }, + { + "epoch": 9.885634588563459, + "grad_norm": 4.127377510070801, + "learning_rate": 9.9787459978746e-06, + "loss": 0.0316, + "step": 21264 + }, + { + "epoch": 9.886099488609949, + "grad_norm": 4.032102584838867, + "learning_rate": 9.978744997874501e-06, + "loss": 0.0545, + "step": 21265 + }, + { + "epoch": 9.886564388656438, + "grad_norm": 1.0522477626800537, + "learning_rate": 9.9787439978744e-06, + "loss": 0.0083, + "step": 21266 + }, + { + "epoch": 9.887029288702928, + "grad_norm": 3.6776487827301025, + "learning_rate": 9.9787429978743e-06, + "loss": 0.0468, + "step": 21267 + }, + { + "epoch": 9.887494188749418, + "grad_norm": 3.8250439167022705, + "learning_rate": 9.978741997874201e-06, + "loss": 0.0801, + "step": 21268 + }, + { + "epoch": 9.88795908879591, + "grad_norm": 11.579242706298828, + "learning_rate": 9.9787409978741e-06, + "loss": 0.2427, + "step": 21269 + }, + { + "epoch": 9.8884239888424, + "grad_norm": 4.946445941925049, + "learning_rate": 9.978739997874e-06, + "loss": 0.0363, + "step": 21270 + }, + { + "epoch": 9.88888888888889, + "grad_norm": 6.416187286376953, + "learning_rate": 9.9787389978739e-06, + "loss": 0.0654, + "step": 21271 + }, + { + "epoch": 9.889353788935379, + "grad_norm": 5.365309238433838, + "learning_rate": 9.9787379978738e-06, + "loss": 0.0957, + "step": 21272 + }, + { + "epoch": 9.889818688981869, + "grad_norm": 6.589112281799316, + "learning_rate": 9.978736997873701e-06, + "loss": 0.2169, + "step": 21273 + }, + { + "epoch": 9.890283589028359, + "grad_norm": 3.161935567855835, + "learning_rate": 9.978735997873602e-06, + "loss": 0.0407, + "step": 21274 + }, + { + "epoch": 9.890748489074848, + "grad_norm": 5.411855697631836, + "learning_rate": 9.9787349978735e-06, + "loss": 0.0983, + "step": 21275 + }, + { + "epoch": 9.891213389121338, + "grad_norm": 5.135408401489258, + "learning_rate": 9.978733997873401e-06, + "loss": 0.0808, + "step": 21276 + }, + { + "epoch": 9.891678289167828, + "grad_norm": 6.638505458831787, + "learning_rate": 9.9787329978733e-06, + "loss": 0.1127, + "step": 21277 + }, + { + "epoch": 9.89214318921432, + "grad_norm": 1.1448816061019897, + "learning_rate": 9.978731997873201e-06, + "loss": 0.0097, + "step": 21278 + }, + { + "epoch": 9.89260808926081, + "grad_norm": 8.092315673828125, + "learning_rate": 9.9787309978731e-06, + "loss": 0.1683, + "step": 21279 + }, + { + "epoch": 9.8930729893073, + "grad_norm": 8.774298667907715, + "learning_rate": 9.978729997873e-06, + "loss": 0.1664, + "step": 21280 + }, + { + "epoch": 9.893537889353789, + "grad_norm": 8.872541427612305, + "learning_rate": 9.978728997872901e-06, + "loss": 0.1214, + "step": 21281 + }, + { + "epoch": 9.894002789400279, + "grad_norm": 4.595640182495117, + "learning_rate": 9.9787279978728e-06, + "loss": 0.0842, + "step": 21282 + }, + { + "epoch": 9.894467689446769, + "grad_norm": 4.550339698791504, + "learning_rate": 9.978726997872701e-06, + "loss": 0.0715, + "step": 21283 + }, + { + "epoch": 9.894932589493258, + "grad_norm": 8.558143615722656, + "learning_rate": 9.9787259978726e-06, + "loss": 0.1606, + "step": 21284 + }, + { + "epoch": 9.895397489539748, + "grad_norm": 2.6392102241516113, + "learning_rate": 9.9787249978725e-06, + "loss": 0.0136, + "step": 21285 + }, + { + "epoch": 9.895862389586238, + "grad_norm": 6.654774188995361, + "learning_rate": 9.978723997872401e-06, + "loss": 0.1656, + "step": 21286 + }, + { + "epoch": 9.89632728963273, + "grad_norm": 7.087530612945557, + "learning_rate": 9.978722997872302e-06, + "loss": 0.1201, + "step": 21287 + }, + { + "epoch": 9.89679218967922, + "grad_norm": 4.744659423828125, + "learning_rate": 9.9787219978722e-06, + "loss": 0.043, + "step": 21288 + }, + { + "epoch": 9.89725708972571, + "grad_norm": 3.430579423904419, + "learning_rate": 9.978720997872101e-06, + "loss": 0.0416, + "step": 21289 + }, + { + "epoch": 9.897721989772199, + "grad_norm": 8.24733829498291, + "learning_rate": 9.978719997872e-06, + "loss": 0.1927, + "step": 21290 + }, + { + "epoch": 9.898186889818689, + "grad_norm": 6.994964599609375, + "learning_rate": 9.9787189978719e-06, + "loss": 0.0565, + "step": 21291 + }, + { + "epoch": 9.898651789865179, + "grad_norm": 5.227811336517334, + "learning_rate": 9.9787179978718e-06, + "loss": 0.1367, + "step": 21292 + }, + { + "epoch": 9.899116689911668, + "grad_norm": 3.2521607875823975, + "learning_rate": 9.9787169978717e-06, + "loss": 0.045, + "step": 21293 + }, + { + "epoch": 9.899581589958158, + "grad_norm": 2.2970595359802246, + "learning_rate": 9.978715997871601e-06, + "loss": 0.0164, + "step": 21294 + }, + { + "epoch": 9.900046490004648, + "grad_norm": 5.659862518310547, + "learning_rate": 9.9787149978715e-06, + "loss": 0.1094, + "step": 21295 + }, + { + "epoch": 9.90051139005114, + "grad_norm": 4.078200817108154, + "learning_rate": 9.978713997871401e-06, + "loss": 0.025, + "step": 21296 + }, + { + "epoch": 9.90097629009763, + "grad_norm": 5.106376647949219, + "learning_rate": 9.9787129978713e-06, + "loss": 0.1048, + "step": 21297 + }, + { + "epoch": 9.90144119014412, + "grad_norm": 1.6235262155532837, + "learning_rate": 9.9787119978712e-06, + "loss": 0.0137, + "step": 21298 + }, + { + "epoch": 9.901906090190609, + "grad_norm": 2.507300853729248, + "learning_rate": 9.978710997871101e-06, + "loss": 0.018, + "step": 21299 + }, + { + "epoch": 9.902370990237099, + "grad_norm": 5.88853120803833, + "learning_rate": 9.978709997871002e-06, + "loss": 0.0838, + "step": 21300 + }, + { + "epoch": 9.902835890283589, + "grad_norm": 2.7285304069519043, + "learning_rate": 9.9787089978709e-06, + "loss": 0.0354, + "step": 21301 + }, + { + "epoch": 9.903300790330078, + "grad_norm": 8.730034828186035, + "learning_rate": 9.9787079978708e-06, + "loss": 0.1548, + "step": 21302 + }, + { + "epoch": 9.903765690376568, + "grad_norm": 8.956676483154297, + "learning_rate": 9.9787069978707e-06, + "loss": 0.144, + "step": 21303 + }, + { + "epoch": 9.90423059042306, + "grad_norm": 8.299332618713379, + "learning_rate": 9.9787059978706e-06, + "loss": 0.1645, + "step": 21304 + }, + { + "epoch": 9.90469549046955, + "grad_norm": 4.598064422607422, + "learning_rate": 9.9787049978705e-06, + "loss": 0.08, + "step": 21305 + }, + { + "epoch": 9.90516039051604, + "grad_norm": 2.7742981910705566, + "learning_rate": 9.9787039978704e-06, + "loss": 0.0289, + "step": 21306 + }, + { + "epoch": 9.90562529056253, + "grad_norm": 3.906398296356201, + "learning_rate": 9.978702997870301e-06, + "loss": 0.044, + "step": 21307 + }, + { + "epoch": 9.906090190609019, + "grad_norm": 8.579534530639648, + "learning_rate": 9.9787019978702e-06, + "loss": 0.2077, + "step": 21308 + }, + { + "epoch": 9.906555090655509, + "grad_norm": 6.0107622146606445, + "learning_rate": 9.978700997870101e-06, + "loss": 0.1011, + "step": 21309 + }, + { + "epoch": 9.907019990701999, + "grad_norm": 4.464015007019043, + "learning_rate": 9.97869999787e-06, + "loss": 0.0684, + "step": 21310 + }, + { + "epoch": 9.907484890748488, + "grad_norm": 2.049064874649048, + "learning_rate": 9.9786989978699e-06, + "loss": 0.0162, + "step": 21311 + }, + { + "epoch": 9.907949790794978, + "grad_norm": 5.932742595672607, + "learning_rate": 9.978697997869801e-06, + "loss": 0.1194, + "step": 21312 + }, + { + "epoch": 9.90841469084147, + "grad_norm": 6.83492374420166, + "learning_rate": 9.9786969978697e-06, + "loss": 0.133, + "step": 21313 + }, + { + "epoch": 9.90887959088796, + "grad_norm": 4.406186580657959, + "learning_rate": 9.978695997869601e-06, + "loss": 0.0376, + "step": 21314 + }, + { + "epoch": 9.90934449093445, + "grad_norm": 5.370458602905273, + "learning_rate": 9.9786949978695e-06, + "loss": 0.1217, + "step": 21315 + }, + { + "epoch": 9.90980939098094, + "grad_norm": 14.565250396728516, + "learning_rate": 9.9786939978694e-06, + "loss": 0.6335, + "step": 21316 + }, + { + "epoch": 9.910274291027429, + "grad_norm": 5.02921724319458, + "learning_rate": 9.9786929978693e-06, + "loss": 0.0619, + "step": 21317 + }, + { + "epoch": 9.910739191073919, + "grad_norm": 5.156048774719238, + "learning_rate": 9.9786919978692e-06, + "loss": 0.1288, + "step": 21318 + }, + { + "epoch": 9.911204091120409, + "grad_norm": 5.516278266906738, + "learning_rate": 9.9786909978691e-06, + "loss": 0.1116, + "step": 21319 + }, + { + "epoch": 9.911668991166898, + "grad_norm": 6.0212836265563965, + "learning_rate": 9.978689997869001e-06, + "loss": 0.0855, + "step": 21320 + }, + { + "epoch": 9.91213389121339, + "grad_norm": 7.313375949859619, + "learning_rate": 9.9786889978689e-06, + "loss": 0.1361, + "step": 21321 + }, + { + "epoch": 9.91259879125988, + "grad_norm": 8.327386856079102, + "learning_rate": 9.978687997868801e-06, + "loss": 0.1111, + "step": 21322 + }, + { + "epoch": 9.91306369130637, + "grad_norm": 5.379411220550537, + "learning_rate": 9.9786869978687e-06, + "loss": 0.0625, + "step": 21323 + }, + { + "epoch": 9.91352859135286, + "grad_norm": 3.3888156414031982, + "learning_rate": 9.9786859978686e-06, + "loss": 0.0373, + "step": 21324 + }, + { + "epoch": 9.91399349139935, + "grad_norm": 11.044313430786133, + "learning_rate": 9.978684997868501e-06, + "loss": 0.3354, + "step": 21325 + }, + { + "epoch": 9.914458391445839, + "grad_norm": 9.508074760437012, + "learning_rate": 9.9786839978684e-06, + "loss": 0.1836, + "step": 21326 + }, + { + "epoch": 9.914923291492329, + "grad_norm": 2.6561238765716553, + "learning_rate": 9.978682997868301e-06, + "loss": 0.0312, + "step": 21327 + }, + { + "epoch": 9.915388191538819, + "grad_norm": 6.9306745529174805, + "learning_rate": 9.9786819978682e-06, + "loss": 0.0893, + "step": 21328 + }, + { + "epoch": 9.915853091585308, + "grad_norm": 6.499237060546875, + "learning_rate": 9.9786809978681e-06, + "loss": 0.1311, + "step": 21329 + }, + { + "epoch": 9.9163179916318, + "grad_norm": 3.0888845920562744, + "learning_rate": 9.978679997868001e-06, + "loss": 0.0433, + "step": 21330 + }, + { + "epoch": 9.91678289167829, + "grad_norm": 4.62021017074585, + "learning_rate": 9.9786789978679e-06, + "loss": 0.1043, + "step": 21331 + }, + { + "epoch": 9.91724779172478, + "grad_norm": 4.259542942047119, + "learning_rate": 9.978677997867801e-06, + "loss": 0.039, + "step": 21332 + }, + { + "epoch": 9.91771269177127, + "grad_norm": 5.292280673980713, + "learning_rate": 9.9786769978677e-06, + "loss": 0.0751, + "step": 21333 + }, + { + "epoch": 9.91817759181776, + "grad_norm": 1.3882412910461426, + "learning_rate": 9.9786759978676e-06, + "loss": 0.0168, + "step": 21334 + }, + { + "epoch": 9.918642491864249, + "grad_norm": 4.526806831359863, + "learning_rate": 9.9786749978675e-06, + "loss": 0.0693, + "step": 21335 + }, + { + "epoch": 9.919107391910739, + "grad_norm": 3.6836283206939697, + "learning_rate": 9.9786739978674e-06, + "loss": 0.061, + "step": 21336 + }, + { + "epoch": 9.919572291957229, + "grad_norm": 5.267058849334717, + "learning_rate": 9.9786729978673e-06, + "loss": 0.0553, + "step": 21337 + }, + { + "epoch": 9.920037192003718, + "grad_norm": 5.466208457946777, + "learning_rate": 9.978671997867201e-06, + "loss": 0.0922, + "step": 21338 + }, + { + "epoch": 9.92050209205021, + "grad_norm": 8.621758460998535, + "learning_rate": 9.9786709978671e-06, + "loss": 0.151, + "step": 21339 + }, + { + "epoch": 9.9209669920967, + "grad_norm": 6.926427364349365, + "learning_rate": 9.978669997867001e-06, + "loss": 0.1073, + "step": 21340 + }, + { + "epoch": 9.92143189214319, + "grad_norm": 3.7140345573425293, + "learning_rate": 9.9786689978669e-06, + "loss": 0.0428, + "step": 21341 + }, + { + "epoch": 9.92189679218968, + "grad_norm": 5.102763652801514, + "learning_rate": 9.9786679978668e-06, + "loss": 0.056, + "step": 21342 + }, + { + "epoch": 9.92236169223617, + "grad_norm": 5.004387855529785, + "learning_rate": 9.978666997866701e-06, + "loss": 0.0597, + "step": 21343 + }, + { + "epoch": 9.922826592282659, + "grad_norm": 5.744021892547607, + "learning_rate": 9.9786659978666e-06, + "loss": 0.1229, + "step": 21344 + }, + { + "epoch": 9.923291492329149, + "grad_norm": 7.9079132080078125, + "learning_rate": 9.978664997866501e-06, + "loss": 0.1819, + "step": 21345 + }, + { + "epoch": 9.923756392375639, + "grad_norm": 3.9289612770080566, + "learning_rate": 9.9786639978664e-06, + "loss": 0.0581, + "step": 21346 + }, + { + "epoch": 9.924221292422128, + "grad_norm": 8.512959480285645, + "learning_rate": 9.9786629978663e-06, + "loss": 0.1843, + "step": 21347 + }, + { + "epoch": 9.92468619246862, + "grad_norm": 4.8896355628967285, + "learning_rate": 9.9786619978662e-06, + "loss": 0.0723, + "step": 21348 + }, + { + "epoch": 9.92515109251511, + "grad_norm": 8.440136909484863, + "learning_rate": 9.9786609978661e-06, + "loss": 0.0761, + "step": 21349 + }, + { + "epoch": 9.9256159925616, + "grad_norm": 4.867632865905762, + "learning_rate": 9.978659997866001e-06, + "loss": 0.0692, + "step": 21350 + }, + { + "epoch": 9.92608089260809, + "grad_norm": 5.334975719451904, + "learning_rate": 9.978658997865902e-06, + "loss": 0.0619, + "step": 21351 + }, + { + "epoch": 9.92654579265458, + "grad_norm": 3.7383875846862793, + "learning_rate": 9.9786579978658e-06, + "loss": 0.0512, + "step": 21352 + }, + { + "epoch": 9.927010692701069, + "grad_norm": 5.367443561553955, + "learning_rate": 9.978656997865701e-06, + "loss": 0.0797, + "step": 21353 + }, + { + "epoch": 9.927475592747559, + "grad_norm": 10.101213455200195, + "learning_rate": 9.9786559978656e-06, + "loss": 0.2884, + "step": 21354 + }, + { + "epoch": 9.927940492794049, + "grad_norm": 5.219034671783447, + "learning_rate": 9.9786549978655e-06, + "loss": 0.0979, + "step": 21355 + }, + { + "epoch": 9.928405392840538, + "grad_norm": 4.415704250335693, + "learning_rate": 9.978653997865401e-06, + "loss": 0.0398, + "step": 21356 + }, + { + "epoch": 9.92887029288703, + "grad_norm": 6.697789192199707, + "learning_rate": 9.9786529978653e-06, + "loss": 0.1208, + "step": 21357 + }, + { + "epoch": 9.92933519293352, + "grad_norm": 5.528148651123047, + "learning_rate": 9.978651997865201e-06, + "loss": 0.1022, + "step": 21358 + }, + { + "epoch": 9.92980009298001, + "grad_norm": 6.417700290679932, + "learning_rate": 9.9786509978651e-06, + "loss": 0.1252, + "step": 21359 + }, + { + "epoch": 9.9302649930265, + "grad_norm": 6.923408508300781, + "learning_rate": 9.978649997865e-06, + "loss": 0.1838, + "step": 21360 + }, + { + "epoch": 9.93072989307299, + "grad_norm": 8.466337203979492, + "learning_rate": 9.9786489978649e-06, + "loss": 0.1696, + "step": 21361 + }, + { + "epoch": 9.931194793119479, + "grad_norm": 3.8305504322052, + "learning_rate": 9.9786479978648e-06, + "loss": 0.0394, + "step": 21362 + }, + { + "epoch": 9.931659693165969, + "grad_norm": 3.0438907146453857, + "learning_rate": 9.978646997864701e-06, + "loss": 0.0362, + "step": 21363 + }, + { + "epoch": 9.932124593212459, + "grad_norm": 1.77503502368927, + "learning_rate": 9.978645997864602e-06, + "loss": 0.0233, + "step": 21364 + }, + { + "epoch": 9.932589493258948, + "grad_norm": 10.422673225402832, + "learning_rate": 9.9786449978645e-06, + "loss": 0.0488, + "step": 21365 + }, + { + "epoch": 9.93305439330544, + "grad_norm": 5.228694438934326, + "learning_rate": 9.9786439978644e-06, + "loss": 0.1277, + "step": 21366 + }, + { + "epoch": 9.93351929335193, + "grad_norm": 5.260870933532715, + "learning_rate": 9.9786429978643e-06, + "loss": 0.1115, + "step": 21367 + }, + { + "epoch": 9.93398419339842, + "grad_norm": 3.018463134765625, + "learning_rate": 9.978641997864201e-06, + "loss": 0.0445, + "step": 21368 + }, + { + "epoch": 9.93444909344491, + "grad_norm": 5.844757080078125, + "learning_rate": 9.978640997864102e-06, + "loss": 0.0506, + "step": 21369 + }, + { + "epoch": 9.9349139934914, + "grad_norm": 7.606634616851807, + "learning_rate": 9.978639997864e-06, + "loss": 0.1281, + "step": 21370 + }, + { + "epoch": 9.935378893537889, + "grad_norm": 3.5400454998016357, + "learning_rate": 9.978638997863901e-06, + "loss": 0.0367, + "step": 21371 + }, + { + "epoch": 9.935843793584379, + "grad_norm": 2.910417079925537, + "learning_rate": 9.9786379978638e-06, + "loss": 0.0396, + "step": 21372 + }, + { + "epoch": 9.936308693630869, + "grad_norm": 4.919981002807617, + "learning_rate": 9.9786369978637e-06, + "loss": 0.041, + "step": 21373 + }, + { + "epoch": 9.936773593677358, + "grad_norm": 2.454437017440796, + "learning_rate": 9.9786359978636e-06, + "loss": 0.0205, + "step": 21374 + }, + { + "epoch": 9.93723849372385, + "grad_norm": 3.198167085647583, + "learning_rate": 9.9786349978635e-06, + "loss": 0.0382, + "step": 21375 + }, + { + "epoch": 9.93770339377034, + "grad_norm": 3.7934539318084717, + "learning_rate": 9.978633997863401e-06, + "loss": 0.0536, + "step": 21376 + }, + { + "epoch": 9.93816829381683, + "grad_norm": 2.2522242069244385, + "learning_rate": 9.9786329978633e-06, + "loss": 0.0185, + "step": 21377 + }, + { + "epoch": 9.93863319386332, + "grad_norm": 6.948036193847656, + "learning_rate": 9.9786319978632e-06, + "loss": 0.1125, + "step": 21378 + }, + { + "epoch": 9.93909809390981, + "grad_norm": 6.741945266723633, + "learning_rate": 9.9786309978631e-06, + "loss": 0.1342, + "step": 21379 + }, + { + "epoch": 9.939562993956299, + "grad_norm": 7.138664245605469, + "learning_rate": 9.978629997863e-06, + "loss": 0.199, + "step": 21380 + }, + { + "epoch": 9.940027894002789, + "grad_norm": 2.1496291160583496, + "learning_rate": 9.978628997862901e-06, + "loss": 0.03, + "step": 21381 + }, + { + "epoch": 9.940492794049279, + "grad_norm": 3.81178879737854, + "learning_rate": 9.978627997862802e-06, + "loss": 0.0386, + "step": 21382 + }, + { + "epoch": 9.94095769409577, + "grad_norm": 2.509502649307251, + "learning_rate": 9.9786269978627e-06, + "loss": 0.0152, + "step": 21383 + }, + { + "epoch": 9.94142259414226, + "grad_norm": 5.089104175567627, + "learning_rate": 9.978625997862601e-06, + "loss": 0.058, + "step": 21384 + }, + { + "epoch": 9.94188749418875, + "grad_norm": 5.2564921379089355, + "learning_rate": 9.9786249978625e-06, + "loss": 0.0927, + "step": 21385 + }, + { + "epoch": 9.94235239423524, + "grad_norm": 6.397154331207275, + "learning_rate": 9.9786239978624e-06, + "loss": 0.1349, + "step": 21386 + }, + { + "epoch": 9.94281729428173, + "grad_norm": 3.612654447555542, + "learning_rate": 9.9786229978623e-06, + "loss": 0.071, + "step": 21387 + }, + { + "epoch": 9.94328219432822, + "grad_norm": 4.336857795715332, + "learning_rate": 9.9786219978622e-06, + "loss": 0.0813, + "step": 21388 + }, + { + "epoch": 9.943747094374709, + "grad_norm": 2.3917500972747803, + "learning_rate": 9.978620997862101e-06, + "loss": 0.0148, + "step": 21389 + }, + { + "epoch": 9.944211994421199, + "grad_norm": 4.960306167602539, + "learning_rate": 9.978619997862e-06, + "loss": 0.0535, + "step": 21390 + }, + { + "epoch": 9.944676894467689, + "grad_norm": 5.405426979064941, + "learning_rate": 9.9786189978619e-06, + "loss": 0.098, + "step": 21391 + }, + { + "epoch": 9.94514179451418, + "grad_norm": 5.47343635559082, + "learning_rate": 9.9786179978618e-06, + "loss": 0.1399, + "step": 21392 + }, + { + "epoch": 9.94560669456067, + "grad_norm": 4.812084197998047, + "learning_rate": 9.9786169978617e-06, + "loss": 0.1419, + "step": 21393 + }, + { + "epoch": 9.94607159460716, + "grad_norm": 5.622859477996826, + "learning_rate": 9.978615997861601e-06, + "loss": 0.0697, + "step": 21394 + }, + { + "epoch": 9.94653649465365, + "grad_norm": 8.435900688171387, + "learning_rate": 9.978614997861502e-06, + "loss": 0.1051, + "step": 21395 + }, + { + "epoch": 9.94700139470014, + "grad_norm": 4.148349285125732, + "learning_rate": 9.9786139978614e-06, + "loss": 0.0667, + "step": 21396 + }, + { + "epoch": 9.94746629474663, + "grad_norm": 3.7053914070129395, + "learning_rate": 9.978612997861301e-06, + "loss": 0.0448, + "step": 21397 + }, + { + "epoch": 9.947931194793119, + "grad_norm": 2.6248555183410645, + "learning_rate": 9.9786119978612e-06, + "loss": 0.0235, + "step": 21398 + }, + { + "epoch": 9.948396094839609, + "grad_norm": 5.520319938659668, + "learning_rate": 9.978610997861101e-06, + "loss": 0.0942, + "step": 21399 + }, + { + "epoch": 9.9488609948861, + "grad_norm": 7.636782646179199, + "learning_rate": 9.978609997861e-06, + "loss": 0.1356, + "step": 21400 + }, + { + "epoch": 9.94932589493259, + "grad_norm": 2.7525899410247803, + "learning_rate": 9.9786089978609e-06, + "loss": 0.0367, + "step": 21401 + }, + { + "epoch": 9.94979079497908, + "grad_norm": 8.479384422302246, + "learning_rate": 9.978607997860801e-06, + "loss": 0.2271, + "step": 21402 + }, + { + "epoch": 9.95025569502557, + "grad_norm": 4.105260848999023, + "learning_rate": 9.9786069978607e-06, + "loss": 0.0657, + "step": 21403 + }, + { + "epoch": 9.95072059507206, + "grad_norm": 5.466829299926758, + "learning_rate": 9.9786059978606e-06, + "loss": 0.0743, + "step": 21404 + }, + { + "epoch": 9.95118549511855, + "grad_norm": 6.465790271759033, + "learning_rate": 9.9786049978605e-06, + "loss": 0.1651, + "step": 21405 + }, + { + "epoch": 9.95165039516504, + "grad_norm": 4.1774582862854, + "learning_rate": 9.9786039978604e-06, + "loss": 0.0766, + "step": 21406 + }, + { + "epoch": 9.952115295211529, + "grad_norm": 4.746606349945068, + "learning_rate": 9.978602997860301e-06, + "loss": 0.0779, + "step": 21407 + }, + { + "epoch": 9.952580195258019, + "grad_norm": 5.241204261779785, + "learning_rate": 9.978601997860202e-06, + "loss": 0.0948, + "step": 21408 + }, + { + "epoch": 9.95304509530451, + "grad_norm": 2.4443767070770264, + "learning_rate": 9.9786009978601e-06, + "loss": 0.0244, + "step": 21409 + }, + { + "epoch": 9.953509995351, + "grad_norm": 4.976795673370361, + "learning_rate": 9.97859999786e-06, + "loss": 0.0327, + "step": 21410 + }, + { + "epoch": 9.95397489539749, + "grad_norm": 0.6433131694793701, + "learning_rate": 9.9785989978599e-06, + "loss": 0.0062, + "step": 21411 + }, + { + "epoch": 9.95443979544398, + "grad_norm": 6.416019916534424, + "learning_rate": 9.978597997859801e-06, + "loss": 0.0891, + "step": 21412 + }, + { + "epoch": 9.95490469549047, + "grad_norm": 6.486450672149658, + "learning_rate": 9.978596997859702e-06, + "loss": 0.1059, + "step": 21413 + }, + { + "epoch": 9.95536959553696, + "grad_norm": 11.087574005126953, + "learning_rate": 9.9785959978596e-06, + "loss": 0.1025, + "step": 21414 + }, + { + "epoch": 9.95583449558345, + "grad_norm": 5.07808256149292, + "learning_rate": 9.978594997859501e-06, + "loss": 0.1039, + "step": 21415 + }, + { + "epoch": 9.956299395629939, + "grad_norm": 2.248992443084717, + "learning_rate": 9.9785939978594e-06, + "loss": 0.0292, + "step": 21416 + }, + { + "epoch": 9.956764295676429, + "grad_norm": 4.8787946701049805, + "learning_rate": 9.978592997859301e-06, + "loss": 0.0878, + "step": 21417 + }, + { + "epoch": 9.95722919572292, + "grad_norm": 1.3375691175460815, + "learning_rate": 9.9785919978592e-06, + "loss": 0.009, + "step": 21418 + }, + { + "epoch": 9.95769409576941, + "grad_norm": 2.00105357170105, + "learning_rate": 9.9785909978591e-06, + "loss": 0.0252, + "step": 21419 + }, + { + "epoch": 9.9581589958159, + "grad_norm": 5.285226821899414, + "learning_rate": 9.978589997859001e-06, + "loss": 0.0761, + "step": 21420 + }, + { + "epoch": 9.95862389586239, + "grad_norm": 4.9158782958984375, + "learning_rate": 9.9785889978589e-06, + "loss": 0.0784, + "step": 21421 + }, + { + "epoch": 9.95908879590888, + "grad_norm": 3.3501715660095215, + "learning_rate": 9.9785879978588e-06, + "loss": 0.032, + "step": 21422 + }, + { + "epoch": 9.95955369595537, + "grad_norm": 13.025369644165039, + "learning_rate": 9.9785869978587e-06, + "loss": 0.1129, + "step": 21423 + }, + { + "epoch": 9.96001859600186, + "grad_norm": 2.4445574283599854, + "learning_rate": 9.9785859978586e-06, + "loss": 0.017, + "step": 21424 + }, + { + "epoch": 9.960483496048349, + "grad_norm": 3.1595754623413086, + "learning_rate": 9.978584997858501e-06, + "loss": 0.0341, + "step": 21425 + }, + { + "epoch": 9.960948396094839, + "grad_norm": 2.9090421199798584, + "learning_rate": 9.978583997858402e-06, + "loss": 0.0159, + "step": 21426 + }, + { + "epoch": 9.96141329614133, + "grad_norm": 6.855541706085205, + "learning_rate": 9.9785829978583e-06, + "loss": 0.1158, + "step": 21427 + }, + { + "epoch": 9.96187819618782, + "grad_norm": 3.778681755065918, + "learning_rate": 9.978581997858201e-06, + "loss": 0.0391, + "step": 21428 + }, + { + "epoch": 9.96234309623431, + "grad_norm": 2.544745445251465, + "learning_rate": 9.9785809978581e-06, + "loss": 0.0252, + "step": 21429 + }, + { + "epoch": 9.9628079962808, + "grad_norm": 5.713349342346191, + "learning_rate": 9.978579997858001e-06, + "loss": 0.1331, + "step": 21430 + }, + { + "epoch": 9.96327289632729, + "grad_norm": 2.751891851425171, + "learning_rate": 9.9785789978579e-06, + "loss": 0.0212, + "step": 21431 + }, + { + "epoch": 9.96373779637378, + "grad_norm": 5.389816761016846, + "learning_rate": 9.9785779978578e-06, + "loss": 0.073, + "step": 21432 + }, + { + "epoch": 9.96420269642027, + "grad_norm": 6.322959899902344, + "learning_rate": 9.978576997857701e-06, + "loss": 0.1182, + "step": 21433 + }, + { + "epoch": 9.964667596466759, + "grad_norm": 2.9460153579711914, + "learning_rate": 9.9785759978576e-06, + "loss": 0.0222, + "step": 21434 + }, + { + "epoch": 9.965132496513249, + "grad_norm": 5.256680488586426, + "learning_rate": 9.978574997857501e-06, + "loss": 0.0601, + "step": 21435 + }, + { + "epoch": 9.96559739655974, + "grad_norm": 3.04154896736145, + "learning_rate": 9.9785739978574e-06, + "loss": 0.0234, + "step": 21436 + }, + { + "epoch": 9.96606229660623, + "grad_norm": 5.680715560913086, + "learning_rate": 9.9785729978573e-06, + "loss": 0.1424, + "step": 21437 + }, + { + "epoch": 9.96652719665272, + "grad_norm": 4.5064568519592285, + "learning_rate": 9.978571997857201e-06, + "loss": 0.0292, + "step": 21438 + }, + { + "epoch": 9.96699209669921, + "grad_norm": 5.937457084655762, + "learning_rate": 9.978570997857102e-06, + "loss": 0.0975, + "step": 21439 + }, + { + "epoch": 9.9674569967457, + "grad_norm": 9.154787063598633, + "learning_rate": 9.978569997857e-06, + "loss": 0.2088, + "step": 21440 + }, + { + "epoch": 9.96792189679219, + "grad_norm": 6.792988300323486, + "learning_rate": 9.978568997856901e-06, + "loss": 0.1426, + "step": 21441 + }, + { + "epoch": 9.96838679683868, + "grad_norm": 6.003104209899902, + "learning_rate": 9.9785679978568e-06, + "loss": 0.1591, + "step": 21442 + }, + { + "epoch": 9.968851696885169, + "grad_norm": 4.3180251121521, + "learning_rate": 9.9785669978567e-06, + "loss": 0.0885, + "step": 21443 + }, + { + "epoch": 9.969316596931659, + "grad_norm": 3.2599644660949707, + "learning_rate": 9.9785659978566e-06, + "loss": 0.0455, + "step": 21444 + }, + { + "epoch": 9.96978149697815, + "grad_norm": 6.3369927406311035, + "learning_rate": 9.9785649978565e-06, + "loss": 0.0872, + "step": 21445 + }, + { + "epoch": 9.97024639702464, + "grad_norm": 5.686713218688965, + "learning_rate": 9.978563997856401e-06, + "loss": 0.0417, + "step": 21446 + }, + { + "epoch": 9.97071129707113, + "grad_norm": 6.885833263397217, + "learning_rate": 9.9785629978563e-06, + "loss": 0.0826, + "step": 21447 + }, + { + "epoch": 9.97117619711762, + "grad_norm": 4.123606204986572, + "learning_rate": 9.978561997856201e-06, + "loss": 0.0485, + "step": 21448 + }, + { + "epoch": 9.97164109716411, + "grad_norm": 6.301368713378906, + "learning_rate": 9.9785609978561e-06, + "loss": 0.1415, + "step": 21449 + }, + { + "epoch": 9.9721059972106, + "grad_norm": 3.622370481491089, + "learning_rate": 9.978559997856e-06, + "loss": 0.0312, + "step": 21450 + }, + { + "epoch": 9.97257089725709, + "grad_norm": 3.501648187637329, + "learning_rate": 9.978558997855901e-06, + "loss": 0.0387, + "step": 21451 + }, + { + "epoch": 9.973035797303579, + "grad_norm": 6.776945114135742, + "learning_rate": 9.978557997855802e-06, + "loss": 0.129, + "step": 21452 + }, + { + "epoch": 9.973500697350069, + "grad_norm": 8.188115119934082, + "learning_rate": 9.978556997855701e-06, + "loss": 0.1886, + "step": 21453 + }, + { + "epoch": 9.97396559739656, + "grad_norm": 9.453211784362793, + "learning_rate": 9.9785559978556e-06, + "loss": 0.1621, + "step": 21454 + }, + { + "epoch": 9.97443049744305, + "grad_norm": 5.60477352142334, + "learning_rate": 9.9785549978555e-06, + "loss": 0.0866, + "step": 21455 + }, + { + "epoch": 9.97489539748954, + "grad_norm": 6.4177470207214355, + "learning_rate": 9.9785539978554e-06, + "loss": 0.0744, + "step": 21456 + }, + { + "epoch": 9.97536029753603, + "grad_norm": 5.069036483764648, + "learning_rate": 9.9785529978553e-06, + "loss": 0.0747, + "step": 21457 + }, + { + "epoch": 9.97582519758252, + "grad_norm": 6.740447044372559, + "learning_rate": 9.9785519978552e-06, + "loss": 0.1137, + "step": 21458 + }, + { + "epoch": 9.97629009762901, + "grad_norm": 5.749432563781738, + "learning_rate": 9.978550997855101e-06, + "loss": 0.0827, + "step": 21459 + }, + { + "epoch": 9.9767549976755, + "grad_norm": 6.41357946395874, + "learning_rate": 9.978549997855e-06, + "loss": 0.1042, + "step": 21460 + }, + { + "epoch": 9.977219897721989, + "grad_norm": 3.1116740703582764, + "learning_rate": 9.978548997854901e-06, + "loss": 0.0281, + "step": 21461 + }, + { + "epoch": 9.97768479776848, + "grad_norm": 6.28027868270874, + "learning_rate": 9.9785479978548e-06, + "loss": 0.065, + "step": 21462 + }, + { + "epoch": 9.97814969781497, + "grad_norm": 4.950558185577393, + "learning_rate": 9.9785469978547e-06, + "loss": 0.1153, + "step": 21463 + }, + { + "epoch": 9.97861459786146, + "grad_norm": 6.3202290534973145, + "learning_rate": 9.978545997854601e-06, + "loss": 0.1239, + "step": 21464 + }, + { + "epoch": 9.97907949790795, + "grad_norm": 3.6101293563842773, + "learning_rate": 9.9785449978545e-06, + "loss": 0.0309, + "step": 21465 + }, + { + "epoch": 9.97954439795444, + "grad_norm": 3.7095019817352295, + "learning_rate": 9.978543997854401e-06, + "loss": 0.0519, + "step": 21466 + }, + { + "epoch": 9.98000929800093, + "grad_norm": 4.138321399688721, + "learning_rate": 9.9785429978543e-06, + "loss": 0.0801, + "step": 21467 + }, + { + "epoch": 9.98047419804742, + "grad_norm": 4.200099945068359, + "learning_rate": 9.9785419978542e-06, + "loss": 0.071, + "step": 21468 + }, + { + "epoch": 9.98093909809391, + "grad_norm": 5.149965763092041, + "learning_rate": 9.9785409978541e-06, + "loss": 0.1598, + "step": 21469 + }, + { + "epoch": 9.981403998140399, + "grad_norm": 2.3452041149139404, + "learning_rate": 9.978539997854e-06, + "loss": 0.0261, + "step": 21470 + }, + { + "epoch": 9.98186889818689, + "grad_norm": 2.9327199459075928, + "learning_rate": 9.978538997853901e-06, + "loss": 0.0407, + "step": 21471 + }, + { + "epoch": 9.98233379823338, + "grad_norm": 2.8478972911834717, + "learning_rate": 9.978537997853802e-06, + "loss": 0.0244, + "step": 21472 + }, + { + "epoch": 9.98279869827987, + "grad_norm": 4.622865676879883, + "learning_rate": 9.9785369978537e-06, + "loss": 0.0502, + "step": 21473 + }, + { + "epoch": 9.98326359832636, + "grad_norm": 6.4842071533203125, + "learning_rate": 9.9785359978536e-06, + "loss": 0.1441, + "step": 21474 + }, + { + "epoch": 9.98372849837285, + "grad_norm": 5.476555824279785, + "learning_rate": 9.9785349978535e-06, + "loss": 0.0629, + "step": 21475 + }, + { + "epoch": 9.98419339841934, + "grad_norm": 4.704211711883545, + "learning_rate": 9.9785339978534e-06, + "loss": 0.0446, + "step": 21476 + }, + { + "epoch": 9.98465829846583, + "grad_norm": 4.796699047088623, + "learning_rate": 9.978532997853301e-06, + "loss": 0.1067, + "step": 21477 + }, + { + "epoch": 9.98512319851232, + "grad_norm": 3.590543508529663, + "learning_rate": 9.9785319978532e-06, + "loss": 0.0407, + "step": 21478 + }, + { + "epoch": 9.98558809855881, + "grad_norm": 5.041845798492432, + "learning_rate": 9.978530997853101e-06, + "loss": 0.0611, + "step": 21479 + }, + { + "epoch": 9.9860529986053, + "grad_norm": 4.138717174530029, + "learning_rate": 9.978529997853e-06, + "loss": 0.0838, + "step": 21480 + }, + { + "epoch": 9.98651789865179, + "grad_norm": 5.777271747589111, + "learning_rate": 9.9785289978529e-06, + "loss": 0.1746, + "step": 21481 + }, + { + "epoch": 9.98698279869828, + "grad_norm": 4.801168441772461, + "learning_rate": 9.9785279978528e-06, + "loss": 0.0532, + "step": 21482 + }, + { + "epoch": 9.98744769874477, + "grad_norm": 3.374068021774292, + "learning_rate": 9.9785269978527e-06, + "loss": 0.0432, + "step": 21483 + }, + { + "epoch": 9.98791259879126, + "grad_norm": 5.732183456420898, + "learning_rate": 9.978525997852601e-06, + "loss": 0.0822, + "step": 21484 + }, + { + "epoch": 9.98837749883775, + "grad_norm": 6.160574913024902, + "learning_rate": 9.9785249978525e-06, + "loss": 0.1345, + "step": 21485 + }, + { + "epoch": 9.98884239888424, + "grad_norm": 3.211620569229126, + "learning_rate": 9.9785239978524e-06, + "loss": 0.0408, + "step": 21486 + }, + { + "epoch": 9.98930729893073, + "grad_norm": 5.7669901847839355, + "learning_rate": 9.9785229978523e-06, + "loss": 0.07, + "step": 21487 + }, + { + "epoch": 9.98977219897722, + "grad_norm": 6.054833889007568, + "learning_rate": 9.9785219978522e-06, + "loss": 0.1094, + "step": 21488 + }, + { + "epoch": 9.99023709902371, + "grad_norm": 1.8190186023712158, + "learning_rate": 9.9785209978521e-06, + "loss": 0.0198, + "step": 21489 + }, + { + "epoch": 9.9907019990702, + "grad_norm": 4.215506076812744, + "learning_rate": 9.978519997852001e-06, + "loss": 0.0444, + "step": 21490 + }, + { + "epoch": 9.99116689911669, + "grad_norm": 7.688579559326172, + "learning_rate": 9.9785189978519e-06, + "loss": 0.1347, + "step": 21491 + }, + { + "epoch": 9.99163179916318, + "grad_norm": 2.513286828994751, + "learning_rate": 9.978517997851801e-06, + "loss": 0.0148, + "step": 21492 + }, + { + "epoch": 9.99209669920967, + "grad_norm": 2.930823802947998, + "learning_rate": 9.9785169978517e-06, + "loss": 0.0353, + "step": 21493 + }, + { + "epoch": 9.99256159925616, + "grad_norm": 3.945732355117798, + "learning_rate": 9.9785159978516e-06, + "loss": 0.0745, + "step": 21494 + }, + { + "epoch": 9.99302649930265, + "grad_norm": 2.5142509937286377, + "learning_rate": 9.978514997851501e-06, + "loss": 0.0257, + "step": 21495 + }, + { + "epoch": 9.99349139934914, + "grad_norm": 6.399588584899902, + "learning_rate": 9.9785139978514e-06, + "loss": 0.1023, + "step": 21496 + }, + { + "epoch": 9.99395629939563, + "grad_norm": 4.420816898345947, + "learning_rate": 9.978512997851301e-06, + "loss": 0.0696, + "step": 21497 + }, + { + "epoch": 9.99442119944212, + "grad_norm": 5.365597724914551, + "learning_rate": 9.9785119978512e-06, + "loss": 0.0941, + "step": 21498 + }, + { + "epoch": 9.99488609948861, + "grad_norm": 7.036377906799316, + "learning_rate": 9.9785109978511e-06, + "loss": 0.1365, + "step": 21499 + }, + { + "epoch": 9.9953509995351, + "grad_norm": 6.458493709564209, + "learning_rate": 9.978509997851e-06, + "loss": 0.1299, + "step": 21500 + }, + { + "epoch": 9.99581589958159, + "grad_norm": 4.675752639770508, + "learning_rate": 9.9785089978509e-06, + "loss": 0.0378, + "step": 21501 + }, + { + "epoch": 9.99628079962808, + "grad_norm": 2.137881278991699, + "learning_rate": 9.978507997850801e-06, + "loss": 0.0172, + "step": 21502 + }, + { + "epoch": 9.99674569967457, + "grad_norm": 6.574464321136475, + "learning_rate": 9.978506997850702e-06, + "loss": 0.0748, + "step": 21503 + }, + { + "epoch": 9.99721059972106, + "grad_norm": 2.1482443809509277, + "learning_rate": 9.9785059978506e-06, + "loss": 0.0073, + "step": 21504 + }, + { + "epoch": 9.99767549976755, + "grad_norm": 5.883246421813965, + "learning_rate": 9.978504997850501e-06, + "loss": 0.0733, + "step": 21505 + }, + { + "epoch": 9.99814039981404, + "grad_norm": 4.712289333343506, + "learning_rate": 9.9785039978504e-06, + "loss": 0.0946, + "step": 21506 + }, + { + "epoch": 9.99860529986053, + "grad_norm": 4.315939903259277, + "learning_rate": 9.9785029978503e-06, + "loss": 0.0613, + "step": 21507 + }, + { + "epoch": 9.99907019990702, + "grad_norm": 2.9091057777404785, + "learning_rate": 9.978501997850201e-06, + "loss": 0.0372, + "step": 21508 + }, + { + "epoch": 9.99953509995351, + "grad_norm": 4.853121757507324, + "learning_rate": 9.9785009978501e-06, + "loss": 0.0686, + "step": 21509 + }, + { + "epoch": 10.0, + "grad_norm": 4.770119667053223, + "learning_rate": 9.978499997850001e-06, + "loss": 0.0779, + "step": 21510 + }, + { + "epoch": 10.00046490004649, + "grad_norm": 4.591954708099365, + "learning_rate": 9.9784989978499e-06, + "loss": 0.0539, + "step": 21511 + }, + { + "epoch": 10.00092980009298, + "grad_norm": 7.672496795654297, + "learning_rate": 9.9784979978498e-06, + "loss": 0.0631, + "step": 21512 + }, + { + "epoch": 10.00139470013947, + "grad_norm": 2.3899412155151367, + "learning_rate": 9.9784969978497e-06, + "loss": 0.0187, + "step": 21513 + }, + { + "epoch": 10.00185960018596, + "grad_norm": 5.8090410232543945, + "learning_rate": 9.9784959978496e-06, + "loss": 0.063, + "step": 21514 + }, + { + "epoch": 10.00232450023245, + "grad_norm": 5.980052947998047, + "learning_rate": 9.978494997849501e-06, + "loss": 0.0792, + "step": 21515 + }, + { + "epoch": 10.00278940027894, + "grad_norm": 6.381651401519775, + "learning_rate": 9.978493997849402e-06, + "loss": 0.0797, + "step": 21516 + }, + { + "epoch": 10.00325430032543, + "grad_norm": 8.73584270477295, + "learning_rate": 9.9784929978493e-06, + "loss": 0.1441, + "step": 21517 + }, + { + "epoch": 10.00371920037192, + "grad_norm": 5.331912040710449, + "learning_rate": 9.9784919978492e-06, + "loss": 0.0891, + "step": 21518 + }, + { + "epoch": 10.00418410041841, + "grad_norm": 4.068398475646973, + "learning_rate": 9.9784909978491e-06, + "loss": 0.0586, + "step": 21519 + }, + { + "epoch": 10.0046490004649, + "grad_norm": 5.436336517333984, + "learning_rate": 9.978489997849001e-06, + "loss": 0.0682, + "step": 21520 + }, + { + "epoch": 10.00511390051139, + "grad_norm": 2.937211036682129, + "learning_rate": 9.978488997848902e-06, + "loss": 0.0268, + "step": 21521 + }, + { + "epoch": 10.00557880055788, + "grad_norm": 9.673370361328125, + "learning_rate": 9.9784879978488e-06, + "loss": 0.1557, + "step": 21522 + }, + { + "epoch": 10.00604370060437, + "grad_norm": 3.908399820327759, + "learning_rate": 9.978486997848701e-06, + "loss": 0.0322, + "step": 21523 + }, + { + "epoch": 10.00650860065086, + "grad_norm": 2.1913721561431885, + "learning_rate": 9.9784859978486e-06, + "loss": 0.0154, + "step": 21524 + }, + { + "epoch": 10.00697350069735, + "grad_norm": 1.260345458984375, + "learning_rate": 9.9784849978485e-06, + "loss": 0.0123, + "step": 21525 + }, + { + "epoch": 10.00743840074384, + "grad_norm": 4.043231010437012, + "learning_rate": 9.9784839978484e-06, + "loss": 0.0942, + "step": 21526 + }, + { + "epoch": 10.00790330079033, + "grad_norm": 5.2406110763549805, + "learning_rate": 9.9784829978483e-06, + "loss": 0.1345, + "step": 21527 + }, + { + "epoch": 10.00836820083682, + "grad_norm": 2.4730122089385986, + "learning_rate": 9.978481997848201e-06, + "loss": 0.0278, + "step": 21528 + }, + { + "epoch": 10.00883310088331, + "grad_norm": 5.31190824508667, + "learning_rate": 9.9784809978481e-06, + "loss": 0.1074, + "step": 21529 + }, + { + "epoch": 10.0092980009298, + "grad_norm": 6.160672187805176, + "learning_rate": 9.978479997848e-06, + "loss": 0.0815, + "step": 21530 + }, + { + "epoch": 10.00976290097629, + "grad_norm": 5.8159661293029785, + "learning_rate": 9.9784789978479e-06, + "loss": 0.0895, + "step": 21531 + }, + { + "epoch": 10.010227801022781, + "grad_norm": 4.667204856872559, + "learning_rate": 9.9784779978478e-06, + "loss": 0.0876, + "step": 21532 + }, + { + "epoch": 10.01069270106927, + "grad_norm": 4.683096885681152, + "learning_rate": 9.978476997847701e-06, + "loss": 0.0585, + "step": 21533 + }, + { + "epoch": 10.01115760111576, + "grad_norm": 3.4009547233581543, + "learning_rate": 9.978475997847602e-06, + "loss": 0.0425, + "step": 21534 + }, + { + "epoch": 10.01162250116225, + "grad_norm": 3.8397624492645264, + "learning_rate": 9.9784749978475e-06, + "loss": 0.0672, + "step": 21535 + }, + { + "epoch": 10.01208740120874, + "grad_norm": 3.485588788986206, + "learning_rate": 9.978473997847401e-06, + "loss": 0.0611, + "step": 21536 + }, + { + "epoch": 10.01255230125523, + "grad_norm": 1.487205147743225, + "learning_rate": 9.9784729978473e-06, + "loss": 0.0146, + "step": 21537 + }, + { + "epoch": 10.01301720130172, + "grad_norm": 4.9782891273498535, + "learning_rate": 9.978471997847201e-06, + "loss": 0.1002, + "step": 21538 + }, + { + "epoch": 10.01348210134821, + "grad_norm": 1.209731101989746, + "learning_rate": 9.9784709978471e-06, + "loss": 0.007, + "step": 21539 + }, + { + "epoch": 10.0139470013947, + "grad_norm": 3.671438694000244, + "learning_rate": 9.978469997847e-06, + "loss": 0.0396, + "step": 21540 + }, + { + "epoch": 10.014411901441191, + "grad_norm": 2.9510855674743652, + "learning_rate": 9.978468997846901e-06, + "loss": 0.0157, + "step": 21541 + }, + { + "epoch": 10.01487680148768, + "grad_norm": 6.415043830871582, + "learning_rate": 9.9784679978468e-06, + "loss": 0.0898, + "step": 21542 + }, + { + "epoch": 10.01534170153417, + "grad_norm": 12.475848197937012, + "learning_rate": 9.9784669978467e-06, + "loss": 0.1913, + "step": 21543 + }, + { + "epoch": 10.01580660158066, + "grad_norm": 8.433676719665527, + "learning_rate": 9.9784659978466e-06, + "loss": 0.172, + "step": 21544 + }, + { + "epoch": 10.01627150162715, + "grad_norm": 4.401336669921875, + "learning_rate": 9.9784649978465e-06, + "loss": 0.0615, + "step": 21545 + }, + { + "epoch": 10.01673640167364, + "grad_norm": 4.755700588226318, + "learning_rate": 9.978463997846401e-06, + "loss": 0.0502, + "step": 21546 + }, + { + "epoch": 10.01720130172013, + "grad_norm": 3.1108193397521973, + "learning_rate": 9.978462997846302e-06, + "loss": 0.0112, + "step": 21547 + }, + { + "epoch": 10.01766620176662, + "grad_norm": 6.217135429382324, + "learning_rate": 9.9784619978462e-06, + "loss": 0.1201, + "step": 21548 + }, + { + "epoch": 10.01813110181311, + "grad_norm": 5.505528926849365, + "learning_rate": 9.978460997846101e-06, + "loss": 0.1054, + "step": 21549 + }, + { + "epoch": 10.018596001859601, + "grad_norm": 3.396167278289795, + "learning_rate": 9.978459997846e-06, + "loss": 0.0733, + "step": 21550 + }, + { + "epoch": 10.01906090190609, + "grad_norm": 2.375661849975586, + "learning_rate": 9.9784589978459e-06, + "loss": 0.0348, + "step": 21551 + }, + { + "epoch": 10.01952580195258, + "grad_norm": 6.574427127838135, + "learning_rate": 9.9784579978458e-06, + "loss": 0.1496, + "step": 21552 + }, + { + "epoch": 10.01999070199907, + "grad_norm": 3.901106595993042, + "learning_rate": 9.9784569978457e-06, + "loss": 0.0467, + "step": 21553 + }, + { + "epoch": 10.02045560204556, + "grad_norm": 3.4327871799468994, + "learning_rate": 9.978455997845601e-06, + "loss": 0.0632, + "step": 21554 + }, + { + "epoch": 10.02092050209205, + "grad_norm": 3.498502492904663, + "learning_rate": 9.9784549978455e-06, + "loss": 0.0456, + "step": 21555 + }, + { + "epoch": 10.02138540213854, + "grad_norm": 9.156435012817383, + "learning_rate": 9.978453997845401e-06, + "loss": 0.0736, + "step": 21556 + }, + { + "epoch": 10.02185030218503, + "grad_norm": 3.5231540203094482, + "learning_rate": 9.9784529978453e-06, + "loss": 0.0355, + "step": 21557 + }, + { + "epoch": 10.02231520223152, + "grad_norm": 3.7314348220825195, + "learning_rate": 9.9784519978452e-06, + "loss": 0.0414, + "step": 21558 + }, + { + "epoch": 10.022780102278011, + "grad_norm": 6.2489190101623535, + "learning_rate": 9.978450997845101e-06, + "loss": 0.0612, + "step": 21559 + }, + { + "epoch": 10.0232450023245, + "grad_norm": 4.481523513793945, + "learning_rate": 9.978449997845002e-06, + "loss": 0.0563, + "step": 21560 + }, + { + "epoch": 10.02370990237099, + "grad_norm": 10.921688079833984, + "learning_rate": 9.9784489978449e-06, + "loss": 0.5512, + "step": 21561 + }, + { + "epoch": 10.02417480241748, + "grad_norm": 4.345961570739746, + "learning_rate": 9.9784479978448e-06, + "loss": 0.0738, + "step": 21562 + }, + { + "epoch": 10.02463970246397, + "grad_norm": 7.710031986236572, + "learning_rate": 9.9784469978447e-06, + "loss": 0.0881, + "step": 21563 + }, + { + "epoch": 10.02510460251046, + "grad_norm": 4.350480556488037, + "learning_rate": 9.978445997844601e-06, + "loss": 0.0743, + "step": 21564 + }, + { + "epoch": 10.02556950255695, + "grad_norm": 4.778614044189453, + "learning_rate": 9.9784449978445e-06, + "loss": 0.1629, + "step": 21565 + }, + { + "epoch": 10.02603440260344, + "grad_norm": 4.676630020141602, + "learning_rate": 9.9784439978444e-06, + "loss": 0.081, + "step": 21566 + }, + { + "epoch": 10.02649930264993, + "grad_norm": 1.7927192449569702, + "learning_rate": 9.978442997844301e-06, + "loss": 0.0203, + "step": 21567 + }, + { + "epoch": 10.026964202696421, + "grad_norm": 3.918492078781128, + "learning_rate": 9.9784419978442e-06, + "loss": 0.0515, + "step": 21568 + }, + { + "epoch": 10.02742910274291, + "grad_norm": 3.1762516498565674, + "learning_rate": 9.978440997844101e-06, + "loss": 0.0482, + "step": 21569 + }, + { + "epoch": 10.0278940027894, + "grad_norm": 3.8540902137756348, + "learning_rate": 9.978439997844e-06, + "loss": 0.0499, + "step": 21570 + }, + { + "epoch": 10.02835890283589, + "grad_norm": 4.929044246673584, + "learning_rate": 9.9784389978439e-06, + "loss": 0.1408, + "step": 21571 + }, + { + "epoch": 10.02882380288238, + "grad_norm": 7.0439348220825195, + "learning_rate": 9.978437997843801e-06, + "loss": 0.144, + "step": 21572 + }, + { + "epoch": 10.02928870292887, + "grad_norm": 1.9639670848846436, + "learning_rate": 9.9784369978437e-06, + "loss": 0.0226, + "step": 21573 + }, + { + "epoch": 10.02975360297536, + "grad_norm": 2.7655282020568848, + "learning_rate": 9.9784359978436e-06, + "loss": 0.0278, + "step": 21574 + }, + { + "epoch": 10.03021850302185, + "grad_norm": 0.5702165961265564, + "learning_rate": 9.9784349978435e-06, + "loss": 0.0045, + "step": 21575 + }, + { + "epoch": 10.030683403068341, + "grad_norm": 5.021869659423828, + "learning_rate": 9.9784339978434e-06, + "loss": 0.0867, + "step": 21576 + }, + { + "epoch": 10.031148303114831, + "grad_norm": 2.866314649581909, + "learning_rate": 9.978432997843301e-06, + "loss": 0.0297, + "step": 21577 + }, + { + "epoch": 10.03161320316132, + "grad_norm": 4.991691589355469, + "learning_rate": 9.978431997843202e-06, + "loss": 0.0415, + "step": 21578 + }, + { + "epoch": 10.03207810320781, + "grad_norm": 2.6441543102264404, + "learning_rate": 9.9784309978431e-06, + "loss": 0.0314, + "step": 21579 + }, + { + "epoch": 10.0325430032543, + "grad_norm": 5.095271110534668, + "learning_rate": 9.978429997843001e-06, + "loss": 0.0529, + "step": 21580 + }, + { + "epoch": 10.03300790330079, + "grad_norm": 2.077333927154541, + "learning_rate": 9.9784289978429e-06, + "loss": 0.0216, + "step": 21581 + }, + { + "epoch": 10.03347280334728, + "grad_norm": 8.590864181518555, + "learning_rate": 9.978427997842801e-06, + "loss": 0.0806, + "step": 21582 + }, + { + "epoch": 10.03393770339377, + "grad_norm": 0.873244047164917, + "learning_rate": 9.9784269978427e-06, + "loss": 0.0104, + "step": 21583 + }, + { + "epoch": 10.03440260344026, + "grad_norm": 2.0610597133636475, + "learning_rate": 9.9784259978426e-06, + "loss": 0.0166, + "step": 21584 + }, + { + "epoch": 10.034867503486751, + "grad_norm": 5.249984264373779, + "learning_rate": 9.978424997842501e-06, + "loss": 0.0864, + "step": 21585 + }, + { + "epoch": 10.035332403533241, + "grad_norm": 4.127464771270752, + "learning_rate": 9.9784239978424e-06, + "loss": 0.0676, + "step": 21586 + }, + { + "epoch": 10.03579730357973, + "grad_norm": 7.047061443328857, + "learning_rate": 9.978422997842301e-06, + "loss": 0.1567, + "step": 21587 + }, + { + "epoch": 10.03626220362622, + "grad_norm": 3.2880046367645264, + "learning_rate": 9.9784219978422e-06, + "loss": 0.0253, + "step": 21588 + }, + { + "epoch": 10.03672710367271, + "grad_norm": 2.17507004737854, + "learning_rate": 9.9784209978421e-06, + "loss": 0.0176, + "step": 21589 + }, + { + "epoch": 10.0371920037192, + "grad_norm": 12.780098915100098, + "learning_rate": 9.978419997842001e-06, + "loss": 0.1182, + "step": 21590 + }, + { + "epoch": 10.03765690376569, + "grad_norm": 2.5505330562591553, + "learning_rate": 9.978418997841902e-06, + "loss": 0.031, + "step": 21591 + }, + { + "epoch": 10.03812180381218, + "grad_norm": 7.083273410797119, + "learning_rate": 9.9784179978418e-06, + "loss": 0.0255, + "step": 21592 + }, + { + "epoch": 10.03858670385867, + "grad_norm": 5.061406135559082, + "learning_rate": 9.978416997841701e-06, + "loss": 0.107, + "step": 21593 + }, + { + "epoch": 10.039051603905161, + "grad_norm": 6.04922342300415, + "learning_rate": 9.9784159978416e-06, + "loss": 0.1432, + "step": 21594 + }, + { + "epoch": 10.039516503951651, + "grad_norm": 5.609328269958496, + "learning_rate": 9.9784149978415e-06, + "loss": 0.0218, + "step": 21595 + }, + { + "epoch": 10.03998140399814, + "grad_norm": 3.0854339599609375, + "learning_rate": 9.9784139978414e-06, + "loss": 0.0306, + "step": 21596 + }, + { + "epoch": 10.04044630404463, + "grad_norm": 3.7452123165130615, + "learning_rate": 9.9784129978413e-06, + "loss": 0.0409, + "step": 21597 + }, + { + "epoch": 10.04091120409112, + "grad_norm": 4.085317134857178, + "learning_rate": 9.978411997841201e-06, + "loss": 0.0539, + "step": 21598 + }, + { + "epoch": 10.04137610413761, + "grad_norm": 2.7981653213500977, + "learning_rate": 9.9784109978411e-06, + "loss": 0.0203, + "step": 21599 + }, + { + "epoch": 10.0418410041841, + "grad_norm": 4.132509231567383, + "learning_rate": 9.978409997841001e-06, + "loss": 0.0699, + "step": 21600 + }, + { + "epoch": 10.04230590423059, + "grad_norm": 4.342376708984375, + "learning_rate": 9.9784089978409e-06, + "loss": 0.0279, + "step": 21601 + }, + { + "epoch": 10.04277080427708, + "grad_norm": 4.554352283477783, + "learning_rate": 9.9784079978408e-06, + "loss": 0.112, + "step": 21602 + }, + { + "epoch": 10.043235704323571, + "grad_norm": 5.996963024139404, + "learning_rate": 9.978406997840701e-06, + "loss": 0.098, + "step": 21603 + }, + { + "epoch": 10.043700604370061, + "grad_norm": 11.974640846252441, + "learning_rate": 9.978405997840602e-06, + "loss": 0.1474, + "step": 21604 + }, + { + "epoch": 10.04416550441655, + "grad_norm": 4.677569389343262, + "learning_rate": 9.978404997840501e-06, + "loss": 0.0631, + "step": 21605 + }, + { + "epoch": 10.04463040446304, + "grad_norm": 1.3368916511535645, + "learning_rate": 9.9784039978404e-06, + "loss": 0.0108, + "step": 21606 + }, + { + "epoch": 10.04509530450953, + "grad_norm": 5.93841552734375, + "learning_rate": 9.9784029978403e-06, + "loss": 0.1, + "step": 21607 + }, + { + "epoch": 10.04556020455602, + "grad_norm": 4.6823554039001465, + "learning_rate": 9.9784019978402e-06, + "loss": 0.0489, + "step": 21608 + }, + { + "epoch": 10.04602510460251, + "grad_norm": 2.877925395965576, + "learning_rate": 9.9784009978401e-06, + "loss": 0.0265, + "step": 21609 + }, + { + "epoch": 10.046490004649, + "grad_norm": 4.792553901672363, + "learning_rate": 9.97839999784e-06, + "loss": 0.0568, + "step": 21610 + }, + { + "epoch": 10.046954904695491, + "grad_norm": 6.067235469818115, + "learning_rate": 9.978398997839901e-06, + "loss": 0.309, + "step": 21611 + }, + { + "epoch": 10.047419804741981, + "grad_norm": 1.2896424531936646, + "learning_rate": 9.9783979978398e-06, + "loss": 0.008, + "step": 21612 + }, + { + "epoch": 10.047884704788471, + "grad_norm": 4.856793403625488, + "learning_rate": 9.978396997839701e-06, + "loss": 0.0526, + "step": 21613 + }, + { + "epoch": 10.04834960483496, + "grad_norm": 1.6777681112289429, + "learning_rate": 9.9783959978396e-06, + "loss": 0.0094, + "step": 21614 + }, + { + "epoch": 10.04881450488145, + "grad_norm": 2.7154970169067383, + "learning_rate": 9.9783949978395e-06, + "loss": 0.0119, + "step": 21615 + }, + { + "epoch": 10.04927940492794, + "grad_norm": 6.8480401039123535, + "learning_rate": 9.978393997839401e-06, + "loss": 0.0778, + "step": 21616 + }, + { + "epoch": 10.04974430497443, + "grad_norm": 6.35722541809082, + "learning_rate": 9.9783929978393e-06, + "loss": 0.142, + "step": 21617 + }, + { + "epoch": 10.05020920502092, + "grad_norm": 1.1648362874984741, + "learning_rate": 9.978391997839201e-06, + "loss": 0.0116, + "step": 21618 + }, + { + "epoch": 10.05067410506741, + "grad_norm": 2.4836392402648926, + "learning_rate": 9.9783909978391e-06, + "loss": 0.0204, + "step": 21619 + }, + { + "epoch": 10.051139005113901, + "grad_norm": 5.585851192474365, + "learning_rate": 9.978389997839e-06, + "loss": 0.1223, + "step": 21620 + }, + { + "epoch": 10.051603905160391, + "grad_norm": 1.3097269535064697, + "learning_rate": 9.9783889978389e-06, + "loss": 0.0114, + "step": 21621 + }, + { + "epoch": 10.052068805206881, + "grad_norm": 1.7251524925231934, + "learning_rate": 9.9783879978388e-06, + "loss": 0.0144, + "step": 21622 + }, + { + "epoch": 10.05253370525337, + "grad_norm": 7.880632400512695, + "learning_rate": 9.978386997838701e-06, + "loss": 0.1431, + "step": 21623 + }, + { + "epoch": 10.05299860529986, + "grad_norm": 4.321040630340576, + "learning_rate": 9.978385997838602e-06, + "loss": 0.0799, + "step": 21624 + }, + { + "epoch": 10.05346350534635, + "grad_norm": 6.1237921714782715, + "learning_rate": 9.9783849978385e-06, + "loss": 0.0922, + "step": 21625 + }, + { + "epoch": 10.05392840539284, + "grad_norm": 3.4658758640289307, + "learning_rate": 9.9783839978384e-06, + "loss": 0.0266, + "step": 21626 + }, + { + "epoch": 10.05439330543933, + "grad_norm": 7.562139511108398, + "learning_rate": 9.9783829978383e-06, + "loss": 0.1998, + "step": 21627 + }, + { + "epoch": 10.05485820548582, + "grad_norm": 7.053346157073975, + "learning_rate": 9.9783819978382e-06, + "loss": 0.0917, + "step": 21628 + }, + { + "epoch": 10.055323105532311, + "grad_norm": 8.179597854614258, + "learning_rate": 9.978380997838101e-06, + "loss": 0.0956, + "step": 21629 + }, + { + "epoch": 10.055788005578801, + "grad_norm": 3.8539280891418457, + "learning_rate": 9.978379997838e-06, + "loss": 0.0428, + "step": 21630 + }, + { + "epoch": 10.056252905625291, + "grad_norm": 2.833906412124634, + "learning_rate": 9.978378997837901e-06, + "loss": 0.0512, + "step": 21631 + }, + { + "epoch": 10.05671780567178, + "grad_norm": 2.6195578575134277, + "learning_rate": 9.9783779978378e-06, + "loss": 0.0321, + "step": 21632 + }, + { + "epoch": 10.05718270571827, + "grad_norm": 5.627190113067627, + "learning_rate": 9.9783769978377e-06, + "loss": 0.1071, + "step": 21633 + }, + { + "epoch": 10.05764760576476, + "grad_norm": 4.520765781402588, + "learning_rate": 9.9783759978376e-06, + "loss": 0.0628, + "step": 21634 + }, + { + "epoch": 10.05811250581125, + "grad_norm": 3.0593748092651367, + "learning_rate": 9.9783749978375e-06, + "loss": 0.0205, + "step": 21635 + }, + { + "epoch": 10.05857740585774, + "grad_norm": 3.1838219165802, + "learning_rate": 9.978373997837401e-06, + "loss": 0.045, + "step": 21636 + }, + { + "epoch": 10.05904230590423, + "grad_norm": 2.792487859725952, + "learning_rate": 9.9783729978373e-06, + "loss": 0.034, + "step": 21637 + }, + { + "epoch": 10.059507205950721, + "grad_norm": 4.705377101898193, + "learning_rate": 9.9783719978372e-06, + "loss": 0.0498, + "step": 21638 + }, + { + "epoch": 10.059972105997211, + "grad_norm": 1.5356032848358154, + "learning_rate": 9.9783709978371e-06, + "loss": 0.0115, + "step": 21639 + }, + { + "epoch": 10.060437006043701, + "grad_norm": 6.6499762535095215, + "learning_rate": 9.978369997837e-06, + "loss": 0.1174, + "step": 21640 + }, + { + "epoch": 10.06090190609019, + "grad_norm": 4.8869829177856445, + "learning_rate": 9.978368997836901e-06, + "loss": 0.069, + "step": 21641 + }, + { + "epoch": 10.06136680613668, + "grad_norm": 3.0225138664245605, + "learning_rate": 9.978367997836802e-06, + "loss": 0.0307, + "step": 21642 + }, + { + "epoch": 10.06183170618317, + "grad_norm": 6.1920552253723145, + "learning_rate": 9.9783669978367e-06, + "loss": 0.0968, + "step": 21643 + }, + { + "epoch": 10.06229660622966, + "grad_norm": 3.874058246612549, + "learning_rate": 9.978365997836601e-06, + "loss": 0.0796, + "step": 21644 + }, + { + "epoch": 10.06276150627615, + "grad_norm": 2.522815465927124, + "learning_rate": 9.9783649978365e-06, + "loss": 0.0299, + "step": 21645 + }, + { + "epoch": 10.06322640632264, + "grad_norm": 6.550243854522705, + "learning_rate": 9.9783639978364e-06, + "loss": 0.1675, + "step": 21646 + }, + { + "epoch": 10.063691306369131, + "grad_norm": 7.586583614349365, + "learning_rate": 9.978362997836301e-06, + "loss": 0.1226, + "step": 21647 + }, + { + "epoch": 10.064156206415621, + "grad_norm": 2.276033401489258, + "learning_rate": 9.9783619978362e-06, + "loss": 0.0196, + "step": 21648 + }, + { + "epoch": 10.064621106462111, + "grad_norm": 0.6462826132774353, + "learning_rate": 9.978360997836101e-06, + "loss": 0.0059, + "step": 21649 + }, + { + "epoch": 10.0650860065086, + "grad_norm": 4.348398685455322, + "learning_rate": 9.978359997836e-06, + "loss": 0.0654, + "step": 21650 + }, + { + "epoch": 10.06555090655509, + "grad_norm": 10.133015632629395, + "learning_rate": 9.9783589978359e-06, + "loss": 0.0571, + "step": 21651 + }, + { + "epoch": 10.06601580660158, + "grad_norm": 1.7917828559875488, + "learning_rate": 9.9783579978358e-06, + "loss": 0.0155, + "step": 21652 + }, + { + "epoch": 10.06648070664807, + "grad_norm": 3.507413387298584, + "learning_rate": 9.9783569978357e-06, + "loss": 0.0297, + "step": 21653 + }, + { + "epoch": 10.06694560669456, + "grad_norm": 3.871029853820801, + "learning_rate": 9.978355997835601e-06, + "loss": 0.0419, + "step": 21654 + }, + { + "epoch": 10.067410506741052, + "grad_norm": 3.808835029602051, + "learning_rate": 9.978354997835502e-06, + "loss": 0.0516, + "step": 21655 + }, + { + "epoch": 10.067875406787541, + "grad_norm": 2.352531671524048, + "learning_rate": 9.9783539978354e-06, + "loss": 0.0207, + "step": 21656 + }, + { + "epoch": 10.068340306834031, + "grad_norm": 4.464922904968262, + "learning_rate": 9.978352997835301e-06, + "loss": 0.066, + "step": 21657 + }, + { + "epoch": 10.068805206880521, + "grad_norm": 2.910907030105591, + "learning_rate": 9.9783519978352e-06, + "loss": 0.0315, + "step": 21658 + }, + { + "epoch": 10.06927010692701, + "grad_norm": 2.725440740585327, + "learning_rate": 9.978350997835101e-06, + "loss": 0.0379, + "step": 21659 + }, + { + "epoch": 10.0697350069735, + "grad_norm": 7.608070373535156, + "learning_rate": 9.978349997835002e-06, + "loss": 0.1421, + "step": 21660 + }, + { + "epoch": 10.07019990701999, + "grad_norm": 2.7209742069244385, + "learning_rate": 9.9783489978349e-06, + "loss": 0.0126, + "step": 21661 + }, + { + "epoch": 10.07066480706648, + "grad_norm": 3.0404975414276123, + "learning_rate": 9.978347997834801e-06, + "loss": 0.0425, + "step": 21662 + }, + { + "epoch": 10.07112970711297, + "grad_norm": 6.866178035736084, + "learning_rate": 9.9783469978347e-06, + "loss": 0.0501, + "step": 21663 + }, + { + "epoch": 10.071594607159462, + "grad_norm": 3.8431203365325928, + "learning_rate": 9.9783459978346e-06, + "loss": 0.026, + "step": 21664 + }, + { + "epoch": 10.072059507205951, + "grad_norm": 5.031919002532959, + "learning_rate": 9.9783449978345e-06, + "loss": 0.0723, + "step": 21665 + }, + { + "epoch": 10.072524407252441, + "grad_norm": 0.9819850325584412, + "learning_rate": 9.9783439978344e-06, + "loss": 0.0082, + "step": 21666 + }, + { + "epoch": 10.072989307298931, + "grad_norm": 3.2494943141937256, + "learning_rate": 9.978342997834301e-06, + "loss": 0.0432, + "step": 21667 + }, + { + "epoch": 10.07345420734542, + "grad_norm": 5.610039234161377, + "learning_rate": 9.978341997834202e-06, + "loss": 0.0694, + "step": 21668 + }, + { + "epoch": 10.07391910739191, + "grad_norm": 3.794473886489868, + "learning_rate": 9.9783409978341e-06, + "loss": 0.0744, + "step": 21669 + }, + { + "epoch": 10.0743840074384, + "grad_norm": 4.027013778686523, + "learning_rate": 9.978339997834e-06, + "loss": 0.0275, + "step": 21670 + }, + { + "epoch": 10.07484890748489, + "grad_norm": 6.4939069747924805, + "learning_rate": 9.9783389978339e-06, + "loss": 0.0881, + "step": 21671 + }, + { + "epoch": 10.07531380753138, + "grad_norm": 6.092995643615723, + "learning_rate": 9.978337997833801e-06, + "loss": 0.1474, + "step": 21672 + }, + { + "epoch": 10.075778707577872, + "grad_norm": 5.483843803405762, + "learning_rate": 9.978336997833702e-06, + "loss": 0.1118, + "step": 21673 + }, + { + "epoch": 10.076243607624361, + "grad_norm": 3.136977195739746, + "learning_rate": 9.9783359978336e-06, + "loss": 0.0343, + "step": 21674 + }, + { + "epoch": 10.076708507670851, + "grad_norm": 7.716118812561035, + "learning_rate": 9.978334997833501e-06, + "loss": 0.0976, + "step": 21675 + }, + { + "epoch": 10.077173407717341, + "grad_norm": 5.868564605712891, + "learning_rate": 9.9783339978334e-06, + "loss": 0.0755, + "step": 21676 + }, + { + "epoch": 10.07763830776383, + "grad_norm": 4.188355445861816, + "learning_rate": 9.9783329978333e-06, + "loss": 0.0617, + "step": 21677 + }, + { + "epoch": 10.07810320781032, + "grad_norm": 3.9984662532806396, + "learning_rate": 9.9783319978332e-06, + "loss": 0.0527, + "step": 21678 + }, + { + "epoch": 10.07856810785681, + "grad_norm": 2.7243664264678955, + "learning_rate": 9.9783309978331e-06, + "loss": 0.0236, + "step": 21679 + }, + { + "epoch": 10.0790330079033, + "grad_norm": 4.611155986785889, + "learning_rate": 9.978329997833001e-06, + "loss": 0.0863, + "step": 21680 + }, + { + "epoch": 10.07949790794979, + "grad_norm": 6.119781017303467, + "learning_rate": 9.9783289978329e-06, + "loss": 0.0692, + "step": 21681 + }, + { + "epoch": 10.079962807996282, + "grad_norm": 4.589627265930176, + "learning_rate": 9.9783279978328e-06, + "loss": 0.1088, + "step": 21682 + }, + { + "epoch": 10.080427708042771, + "grad_norm": 3.9413084983825684, + "learning_rate": 9.9783269978327e-06, + "loss": 0.0387, + "step": 21683 + }, + { + "epoch": 10.080892608089261, + "grad_norm": 4.24876594543457, + "learning_rate": 9.9783259978326e-06, + "loss": 0.0696, + "step": 21684 + }, + { + "epoch": 10.081357508135751, + "grad_norm": 4.243407726287842, + "learning_rate": 9.978324997832501e-06, + "loss": 0.0471, + "step": 21685 + }, + { + "epoch": 10.08182240818224, + "grad_norm": 3.4215071201324463, + "learning_rate": 9.978323997832402e-06, + "loss": 0.0364, + "step": 21686 + }, + { + "epoch": 10.08228730822873, + "grad_norm": 6.510231018066406, + "learning_rate": 9.9783229978323e-06, + "loss": 0.1324, + "step": 21687 + }, + { + "epoch": 10.08275220827522, + "grad_norm": 3.298872947692871, + "learning_rate": 9.978321997832201e-06, + "loss": 0.0303, + "step": 21688 + }, + { + "epoch": 10.08321710832171, + "grad_norm": 4.146771430969238, + "learning_rate": 9.9783209978321e-06, + "loss": 0.041, + "step": 21689 + }, + { + "epoch": 10.0836820083682, + "grad_norm": 3.062066078186035, + "learning_rate": 9.978319997832001e-06, + "loss": 0.0311, + "step": 21690 + }, + { + "epoch": 10.084146908414692, + "grad_norm": 3.7750537395477295, + "learning_rate": 9.9783189978319e-06, + "loss": 0.0703, + "step": 21691 + }, + { + "epoch": 10.084611808461181, + "grad_norm": 7.81186580657959, + "learning_rate": 9.9783179978318e-06, + "loss": 0.1766, + "step": 21692 + }, + { + "epoch": 10.085076708507671, + "grad_norm": 4.423314094543457, + "learning_rate": 9.978316997831701e-06, + "loss": 0.048, + "step": 21693 + }, + { + "epoch": 10.085541608554161, + "grad_norm": 6.70449161529541, + "learning_rate": 9.9783159978316e-06, + "loss": 0.1633, + "step": 21694 + }, + { + "epoch": 10.08600650860065, + "grad_norm": 4.61660099029541, + "learning_rate": 9.9783149978315e-06, + "loss": 0.0538, + "step": 21695 + }, + { + "epoch": 10.08647140864714, + "grad_norm": 4.431118965148926, + "learning_rate": 9.9783139978314e-06, + "loss": 0.0462, + "step": 21696 + }, + { + "epoch": 10.08693630869363, + "grad_norm": 5.308144569396973, + "learning_rate": 9.9783129978313e-06, + "loss": 0.0303, + "step": 21697 + }, + { + "epoch": 10.08740120874012, + "grad_norm": 6.0552191734313965, + "learning_rate": 9.978311997831201e-06, + "loss": 0.0911, + "step": 21698 + }, + { + "epoch": 10.087866108786612, + "grad_norm": 4.014939785003662, + "learning_rate": 9.978310997831102e-06, + "loss": 0.0497, + "step": 21699 + }, + { + "epoch": 10.088331008833102, + "grad_norm": 1.4361069202423096, + "learning_rate": 9.978309997831e-06, + "loss": 0.0121, + "step": 21700 + }, + { + "epoch": 10.088795908879591, + "grad_norm": 3.4659974575042725, + "learning_rate": 9.978308997830901e-06, + "loss": 0.0421, + "step": 21701 + }, + { + "epoch": 10.089260808926081, + "grad_norm": 5.779717445373535, + "learning_rate": 9.9783079978308e-06, + "loss": 0.0704, + "step": 21702 + }, + { + "epoch": 10.089725708972571, + "grad_norm": 5.741199970245361, + "learning_rate": 9.9783069978307e-06, + "loss": 0.0839, + "step": 21703 + }, + { + "epoch": 10.09019060901906, + "grad_norm": 6.605633735656738, + "learning_rate": 9.9783059978306e-06, + "loss": 0.0386, + "step": 21704 + }, + { + "epoch": 10.09065550906555, + "grad_norm": 13.588757514953613, + "learning_rate": 9.9783049978305e-06, + "loss": 0.0987, + "step": 21705 + }, + { + "epoch": 10.09112040911204, + "grad_norm": 5.860360622406006, + "learning_rate": 9.978303997830401e-06, + "loss": 0.0802, + "step": 21706 + }, + { + "epoch": 10.09158530915853, + "grad_norm": 2.432647466659546, + "learning_rate": 9.9783029978303e-06, + "loss": 0.0232, + "step": 21707 + }, + { + "epoch": 10.092050209205022, + "grad_norm": 1.0383987426757812, + "learning_rate": 9.978301997830201e-06, + "loss": 0.0055, + "step": 21708 + }, + { + "epoch": 10.092515109251512, + "grad_norm": 2.4140689373016357, + "learning_rate": 9.9783009978301e-06, + "loss": 0.0326, + "step": 21709 + }, + { + "epoch": 10.092980009298001, + "grad_norm": 4.735745429992676, + "learning_rate": 9.97829999783e-06, + "loss": 0.0919, + "step": 21710 + }, + { + "epoch": 10.093444909344491, + "grad_norm": 6.482223033905029, + "learning_rate": 9.978298997829901e-06, + "loss": 0.0998, + "step": 21711 + }, + { + "epoch": 10.093909809390981, + "grad_norm": 2.9559903144836426, + "learning_rate": 9.978297997829802e-06, + "loss": 0.0341, + "step": 21712 + }, + { + "epoch": 10.09437470943747, + "grad_norm": 6.712924957275391, + "learning_rate": 9.9782969978297e-06, + "loss": 0.16, + "step": 21713 + }, + { + "epoch": 10.09483960948396, + "grad_norm": 4.157490253448486, + "learning_rate": 9.9782959978296e-06, + "loss": 0.0491, + "step": 21714 + }, + { + "epoch": 10.09530450953045, + "grad_norm": 5.8386712074279785, + "learning_rate": 9.9782949978295e-06, + "loss": 0.045, + "step": 21715 + }, + { + "epoch": 10.09576940957694, + "grad_norm": 1.5091371536254883, + "learning_rate": 9.9782939978294e-06, + "loss": 0.0159, + "step": 21716 + }, + { + "epoch": 10.096234309623432, + "grad_norm": 3.149780511856079, + "learning_rate": 9.9782929978293e-06, + "loss": 0.0591, + "step": 21717 + }, + { + "epoch": 10.096699209669922, + "grad_norm": 3.0510385036468506, + "learning_rate": 9.9782919978292e-06, + "loss": 0.0225, + "step": 21718 + }, + { + "epoch": 10.097164109716411, + "grad_norm": 7.767048358917236, + "learning_rate": 9.978290997829101e-06, + "loss": 0.2116, + "step": 21719 + }, + { + "epoch": 10.097629009762901, + "grad_norm": 1.898041009902954, + "learning_rate": 9.978289997829e-06, + "loss": 0.0148, + "step": 21720 + }, + { + "epoch": 10.098093909809391, + "grad_norm": 6.776692867279053, + "learning_rate": 9.978288997828901e-06, + "loss": 0.0381, + "step": 21721 + }, + { + "epoch": 10.09855880985588, + "grad_norm": 5.463781356811523, + "learning_rate": 9.9782879978288e-06, + "loss": 0.0495, + "step": 21722 + }, + { + "epoch": 10.09902370990237, + "grad_norm": 7.016556262969971, + "learning_rate": 9.9782869978287e-06, + "loss": 0.1244, + "step": 21723 + }, + { + "epoch": 10.09948860994886, + "grad_norm": 3.0845909118652344, + "learning_rate": 9.978285997828601e-06, + "loss": 0.0369, + "step": 21724 + }, + { + "epoch": 10.09995350999535, + "grad_norm": 6.13081169128418, + "learning_rate": 9.9782849978285e-06, + "loss": 0.0967, + "step": 21725 + }, + { + "epoch": 10.100418410041842, + "grad_norm": 4.5007147789001465, + "learning_rate": 9.978283997828401e-06, + "loss": 0.0683, + "step": 21726 + }, + { + "epoch": 10.100883310088332, + "grad_norm": 3.510498046875, + "learning_rate": 9.9782829978283e-06, + "loss": 0.0547, + "step": 21727 + }, + { + "epoch": 10.101348210134821, + "grad_norm": 2.057145833969116, + "learning_rate": 9.9782819978282e-06, + "loss": 0.0202, + "step": 21728 + }, + { + "epoch": 10.101813110181311, + "grad_norm": 3.9044013023376465, + "learning_rate": 9.978280997828101e-06, + "loss": 0.0482, + "step": 21729 + }, + { + "epoch": 10.102278010227801, + "grad_norm": 9.489106178283691, + "learning_rate": 9.978279997828e-06, + "loss": 0.0455, + "step": 21730 + }, + { + "epoch": 10.10274291027429, + "grad_norm": 3.3713996410369873, + "learning_rate": 9.9782789978279e-06, + "loss": 0.0371, + "step": 21731 + }, + { + "epoch": 10.10320781032078, + "grad_norm": 3.4898900985717773, + "learning_rate": 9.978277997827801e-06, + "loss": 0.0334, + "step": 21732 + }, + { + "epoch": 10.10367271036727, + "grad_norm": 2.1994740962982178, + "learning_rate": 9.9782769978277e-06, + "loss": 0.0208, + "step": 21733 + }, + { + "epoch": 10.104137610413762, + "grad_norm": 4.823704719543457, + "learning_rate": 9.978275997827601e-06, + "loss": 0.0893, + "step": 21734 + }, + { + "epoch": 10.104602510460252, + "grad_norm": 4.782199859619141, + "learning_rate": 9.9782749978275e-06, + "loss": 0.0607, + "step": 21735 + }, + { + "epoch": 10.105067410506742, + "grad_norm": 5.832539081573486, + "learning_rate": 9.9782739978274e-06, + "loss": 0.0108, + "step": 21736 + }, + { + "epoch": 10.105532310553231, + "grad_norm": 6.2341508865356445, + "learning_rate": 9.978272997827301e-06, + "loss": 0.1182, + "step": 21737 + }, + { + "epoch": 10.105997210599721, + "grad_norm": 5.465315818786621, + "learning_rate": 9.9782719978272e-06, + "loss": 0.1498, + "step": 21738 + }, + { + "epoch": 10.106462110646211, + "grad_norm": 5.916961193084717, + "learning_rate": 9.978270997827101e-06, + "loss": 0.063, + "step": 21739 + }, + { + "epoch": 10.1069270106927, + "grad_norm": 4.33931827545166, + "learning_rate": 9.978269997827e-06, + "loss": 0.0413, + "step": 21740 + }, + { + "epoch": 10.10739191073919, + "grad_norm": 4.6150288581848145, + "learning_rate": 9.9782689978269e-06, + "loss": 0.0716, + "step": 21741 + }, + { + "epoch": 10.10785681078568, + "grad_norm": 3.3757004737854004, + "learning_rate": 9.978267997826801e-06, + "loss": 0.0302, + "step": 21742 + }, + { + "epoch": 10.108321710832172, + "grad_norm": 3.591948986053467, + "learning_rate": 9.978266997826702e-06, + "loss": 0.0466, + "step": 21743 + }, + { + "epoch": 10.108786610878662, + "grad_norm": 3.3370180130004883, + "learning_rate": 9.978265997826601e-06, + "loss": 0.0266, + "step": 21744 + }, + { + "epoch": 10.109251510925152, + "grad_norm": 2.9686899185180664, + "learning_rate": 9.978264997826502e-06, + "loss": 0.0241, + "step": 21745 + }, + { + "epoch": 10.109716410971641, + "grad_norm": 4.856338977813721, + "learning_rate": 9.9782639978264e-06, + "loss": 0.0351, + "step": 21746 + }, + { + "epoch": 10.110181311018131, + "grad_norm": 6.322050094604492, + "learning_rate": 9.9782629978263e-06, + "loss": 0.0558, + "step": 21747 + }, + { + "epoch": 10.110646211064621, + "grad_norm": 5.885735034942627, + "learning_rate": 9.9782619978262e-06, + "loss": 0.1087, + "step": 21748 + }, + { + "epoch": 10.11111111111111, + "grad_norm": 8.164566040039062, + "learning_rate": 9.9782609978261e-06, + "loss": 0.1596, + "step": 21749 + }, + { + "epoch": 10.1115760111576, + "grad_norm": 3.348090171813965, + "learning_rate": 9.978259997826001e-06, + "loss": 0.024, + "step": 21750 + }, + { + "epoch": 10.11204091120409, + "grad_norm": 2.9234683513641357, + "learning_rate": 9.9782589978259e-06, + "loss": 0.022, + "step": 21751 + }, + { + "epoch": 10.112505811250582, + "grad_norm": 2.537107229232788, + "learning_rate": 9.978257997825801e-06, + "loss": 0.0276, + "step": 21752 + }, + { + "epoch": 10.112970711297072, + "grad_norm": 1.6433426141738892, + "learning_rate": 9.9782569978257e-06, + "loss": 0.0092, + "step": 21753 + }, + { + "epoch": 10.113435611343562, + "grad_norm": 3.9189584255218506, + "learning_rate": 9.9782559978256e-06, + "loss": 0.0738, + "step": 21754 + }, + { + "epoch": 10.113900511390051, + "grad_norm": 2.909147024154663, + "learning_rate": 9.978254997825501e-06, + "loss": 0.0431, + "step": 21755 + }, + { + "epoch": 10.114365411436541, + "grad_norm": 1.3750838041305542, + "learning_rate": 9.9782539978254e-06, + "loss": 0.0123, + "step": 21756 + }, + { + "epoch": 10.114830311483031, + "grad_norm": 1.5673623085021973, + "learning_rate": 9.978252997825301e-06, + "loss": 0.0123, + "step": 21757 + }, + { + "epoch": 10.11529521152952, + "grad_norm": 3.7334275245666504, + "learning_rate": 9.9782519978252e-06, + "loss": 0.0509, + "step": 21758 + }, + { + "epoch": 10.11576011157601, + "grad_norm": 1.1001678705215454, + "learning_rate": 9.9782509978251e-06, + "loss": 0.0079, + "step": 21759 + }, + { + "epoch": 10.1162250116225, + "grad_norm": 3.6452691555023193, + "learning_rate": 9.978249997825e-06, + "loss": 0.0472, + "step": 21760 + }, + { + "epoch": 10.116689911668992, + "grad_norm": 9.111324310302734, + "learning_rate": 9.9782489978249e-06, + "loss": 0.1367, + "step": 21761 + }, + { + "epoch": 10.117154811715482, + "grad_norm": 13.059056282043457, + "learning_rate": 9.9782479978248e-06, + "loss": 0.1879, + "step": 21762 + }, + { + "epoch": 10.117619711761972, + "grad_norm": 2.05346417427063, + "learning_rate": 9.978246997824701e-06, + "loss": 0.0198, + "step": 21763 + }, + { + "epoch": 10.118084611808461, + "grad_norm": 4.3699750900268555, + "learning_rate": 9.9782459978246e-06, + "loss": 0.028, + "step": 21764 + }, + { + "epoch": 10.118549511854951, + "grad_norm": 2.915795087814331, + "learning_rate": 9.978244997824501e-06, + "loss": 0.0327, + "step": 21765 + }, + { + "epoch": 10.119014411901441, + "grad_norm": 3.3255012035369873, + "learning_rate": 9.9782439978244e-06, + "loss": 0.0437, + "step": 21766 + }, + { + "epoch": 10.11947931194793, + "grad_norm": 3.2874081134796143, + "learning_rate": 9.9782429978243e-06, + "loss": 0.0362, + "step": 21767 + }, + { + "epoch": 10.11994421199442, + "grad_norm": 5.021100044250488, + "learning_rate": 9.978241997824201e-06, + "loss": 0.0827, + "step": 21768 + }, + { + "epoch": 10.12040911204091, + "grad_norm": 3.703781843185425, + "learning_rate": 9.9782409978241e-06, + "loss": 0.0533, + "step": 21769 + }, + { + "epoch": 10.120874012087402, + "grad_norm": 2.6202633380889893, + "learning_rate": 9.978239997824001e-06, + "loss": 0.0194, + "step": 21770 + }, + { + "epoch": 10.121338912133892, + "grad_norm": 6.185529708862305, + "learning_rate": 9.9782389978239e-06, + "loss": 0.0692, + "step": 21771 + }, + { + "epoch": 10.121803812180381, + "grad_norm": 3.9996087551116943, + "learning_rate": 9.9782379978238e-06, + "loss": 0.0283, + "step": 21772 + }, + { + "epoch": 10.122268712226871, + "grad_norm": 3.3331549167633057, + "learning_rate": 9.9782369978237e-06, + "loss": 0.0189, + "step": 21773 + }, + { + "epoch": 10.122733612273361, + "grad_norm": 3.7021822929382324, + "learning_rate": 9.9782359978236e-06, + "loss": 0.0361, + "step": 21774 + }, + { + "epoch": 10.123198512319851, + "grad_norm": 10.624288558959961, + "learning_rate": 9.978234997823501e-06, + "loss": 0.2633, + "step": 21775 + }, + { + "epoch": 10.12366341236634, + "grad_norm": 3.062901258468628, + "learning_rate": 9.978233997823402e-06, + "loss": 0.043, + "step": 21776 + }, + { + "epoch": 10.12412831241283, + "grad_norm": 2.3171820640563965, + "learning_rate": 9.9782329978233e-06, + "loss": 0.0349, + "step": 21777 + }, + { + "epoch": 10.124593212459322, + "grad_norm": 3.861203670501709, + "learning_rate": 9.9782319978232e-06, + "loss": 0.0427, + "step": 21778 + }, + { + "epoch": 10.125058112505812, + "grad_norm": 4.8277506828308105, + "learning_rate": 9.9782309978231e-06, + "loss": 0.0592, + "step": 21779 + }, + { + "epoch": 10.125523012552302, + "grad_norm": 5.0830254554748535, + "learning_rate": 9.978229997823e-06, + "loss": 0.0731, + "step": 21780 + }, + { + "epoch": 10.125987912598791, + "grad_norm": 4.2596001625061035, + "learning_rate": 9.978228997822901e-06, + "loss": 0.0479, + "step": 21781 + }, + { + "epoch": 10.126452812645281, + "grad_norm": 1.310813069343567, + "learning_rate": 9.9782279978228e-06, + "loss": 0.0085, + "step": 21782 + }, + { + "epoch": 10.126917712691771, + "grad_norm": 8.555523872375488, + "learning_rate": 9.978226997822701e-06, + "loss": 0.2675, + "step": 21783 + }, + { + "epoch": 10.127382612738261, + "grad_norm": 6.179385662078857, + "learning_rate": 9.9782259978226e-06, + "loss": 0.1866, + "step": 21784 + }, + { + "epoch": 10.12784751278475, + "grad_norm": 4.58325719833374, + "learning_rate": 9.9782249978225e-06, + "loss": 0.0779, + "step": 21785 + }, + { + "epoch": 10.12831241283124, + "grad_norm": 7.102210521697998, + "learning_rate": 9.9782239978224e-06, + "loss": 0.1079, + "step": 21786 + }, + { + "epoch": 10.128777312877732, + "grad_norm": 4.433450698852539, + "learning_rate": 9.9782229978223e-06, + "loss": 0.0423, + "step": 21787 + }, + { + "epoch": 10.129242212924222, + "grad_norm": 2.9427907466888428, + "learning_rate": 9.978221997822201e-06, + "loss": 0.0317, + "step": 21788 + }, + { + "epoch": 10.129707112970712, + "grad_norm": 4.408658981323242, + "learning_rate": 9.9782209978221e-06, + "loss": 0.0538, + "step": 21789 + }, + { + "epoch": 10.130172013017201, + "grad_norm": 2.2493693828582764, + "learning_rate": 9.978219997822e-06, + "loss": 0.0198, + "step": 21790 + }, + { + "epoch": 10.130636913063691, + "grad_norm": 5.683980464935303, + "learning_rate": 9.9782189978219e-06, + "loss": 0.0807, + "step": 21791 + }, + { + "epoch": 10.131101813110181, + "grad_norm": 4.459110736846924, + "learning_rate": 9.9782179978218e-06, + "loss": 0.1017, + "step": 21792 + }, + { + "epoch": 10.131566713156671, + "grad_norm": 3.2120046615600586, + "learning_rate": 9.978216997821701e-06, + "loss": 0.0712, + "step": 21793 + }, + { + "epoch": 10.13203161320316, + "grad_norm": 1.9256489276885986, + "learning_rate": 9.978215997821602e-06, + "loss": 0.0178, + "step": 21794 + }, + { + "epoch": 10.13249651324965, + "grad_norm": 3.619614839553833, + "learning_rate": 9.9782149978215e-06, + "loss": 0.0424, + "step": 21795 + }, + { + "epoch": 10.132961413296142, + "grad_norm": 2.821511745452881, + "learning_rate": 9.978213997821401e-06, + "loss": 0.0369, + "step": 21796 + }, + { + "epoch": 10.133426313342632, + "grad_norm": 3.2069034576416016, + "learning_rate": 9.9782129978213e-06, + "loss": 0.0335, + "step": 21797 + }, + { + "epoch": 10.133891213389122, + "grad_norm": 4.2838239669799805, + "learning_rate": 9.9782119978212e-06, + "loss": 0.0719, + "step": 21798 + }, + { + "epoch": 10.134356113435611, + "grad_norm": 5.80745267868042, + "learning_rate": 9.9782109978211e-06, + "loss": 0.08, + "step": 21799 + }, + { + "epoch": 10.134821013482101, + "grad_norm": 5.147892475128174, + "learning_rate": 9.978209997821e-06, + "loss": 0.1042, + "step": 21800 + }, + { + "epoch": 10.135285913528591, + "grad_norm": 2.742410182952881, + "learning_rate": 9.978208997820901e-06, + "loss": 0.0245, + "step": 21801 + }, + { + "epoch": 10.135750813575081, + "grad_norm": 3.523974657058716, + "learning_rate": 9.9782079978208e-06, + "loss": 0.0431, + "step": 21802 + }, + { + "epoch": 10.13621571362157, + "grad_norm": 4.48096227645874, + "learning_rate": 9.9782069978207e-06, + "loss": 0.0399, + "step": 21803 + }, + { + "epoch": 10.13668061366806, + "grad_norm": 2.1112875938415527, + "learning_rate": 9.9782059978206e-06, + "loss": 0.0156, + "step": 21804 + }, + { + "epoch": 10.137145513714552, + "grad_norm": 5.740334510803223, + "learning_rate": 9.9782049978205e-06, + "loss": 0.1271, + "step": 21805 + }, + { + "epoch": 10.137610413761042, + "grad_norm": 2.794325590133667, + "learning_rate": 9.978203997820401e-06, + "loss": 0.0478, + "step": 21806 + }, + { + "epoch": 10.138075313807532, + "grad_norm": 3.607941150665283, + "learning_rate": 9.978202997820302e-06, + "loss": 0.0387, + "step": 21807 + }, + { + "epoch": 10.138540213854021, + "grad_norm": 3.7249562740325928, + "learning_rate": 9.9782019978202e-06, + "loss": 0.0507, + "step": 21808 + }, + { + "epoch": 10.139005113900511, + "grad_norm": 0.8199386596679688, + "learning_rate": 9.978200997820101e-06, + "loss": 0.0082, + "step": 21809 + }, + { + "epoch": 10.139470013947001, + "grad_norm": 7.098401069641113, + "learning_rate": 9.97819999782e-06, + "loss": 0.1846, + "step": 21810 + }, + { + "epoch": 10.139934913993491, + "grad_norm": 5.789369583129883, + "learning_rate": 9.978198997819901e-06, + "loss": 0.1158, + "step": 21811 + }, + { + "epoch": 10.14039981403998, + "grad_norm": 6.3306965827941895, + "learning_rate": 9.978197997819802e-06, + "loss": 0.1727, + "step": 21812 + }, + { + "epoch": 10.140864714086472, + "grad_norm": 9.060405731201172, + "learning_rate": 9.9781969978197e-06, + "loss": 0.1782, + "step": 21813 + }, + { + "epoch": 10.141329614132962, + "grad_norm": 4.63220739364624, + "learning_rate": 9.978195997819601e-06, + "loss": 0.0748, + "step": 21814 + }, + { + "epoch": 10.141794514179452, + "grad_norm": 6.689177513122559, + "learning_rate": 9.9781949978195e-06, + "loss": 0.1553, + "step": 21815 + }, + { + "epoch": 10.142259414225942, + "grad_norm": 2.5359718799591064, + "learning_rate": 9.9781939978194e-06, + "loss": 0.0287, + "step": 21816 + }, + { + "epoch": 10.142724314272431, + "grad_norm": 5.4185357093811035, + "learning_rate": 9.9781929978193e-06, + "loss": 0.0819, + "step": 21817 + }, + { + "epoch": 10.143189214318921, + "grad_norm": 5.521941661834717, + "learning_rate": 9.9781919978192e-06, + "loss": 0.1036, + "step": 21818 + }, + { + "epoch": 10.143654114365411, + "grad_norm": 7.810240745544434, + "learning_rate": 9.978190997819101e-06, + "loss": 0.2361, + "step": 21819 + }, + { + "epoch": 10.144119014411901, + "grad_norm": 7.139264106750488, + "learning_rate": 9.978189997819002e-06, + "loss": 0.0611, + "step": 21820 + }, + { + "epoch": 10.14458391445839, + "grad_norm": 2.9948019981384277, + "learning_rate": 9.9781889978189e-06, + "loss": 0.0229, + "step": 21821 + }, + { + "epoch": 10.145048814504882, + "grad_norm": 4.275428771972656, + "learning_rate": 9.9781879978188e-06, + "loss": 0.0606, + "step": 21822 + }, + { + "epoch": 10.145513714551372, + "grad_norm": 6.153223037719727, + "learning_rate": 9.9781869978187e-06, + "loss": 0.1109, + "step": 21823 + }, + { + "epoch": 10.145978614597862, + "grad_norm": 5.257091522216797, + "learning_rate": 9.978185997818601e-06, + "loss": 0.0931, + "step": 21824 + }, + { + "epoch": 10.146443514644352, + "grad_norm": 4.101988315582275, + "learning_rate": 9.978184997818502e-06, + "loss": 0.0675, + "step": 21825 + }, + { + "epoch": 10.146908414690841, + "grad_norm": 5.5495219230651855, + "learning_rate": 9.9781839978184e-06, + "loss": 0.1431, + "step": 21826 + }, + { + "epoch": 10.147373314737331, + "grad_norm": 5.866602897644043, + "learning_rate": 9.978182997818301e-06, + "loss": 0.1413, + "step": 21827 + }, + { + "epoch": 10.147838214783821, + "grad_norm": 4.397531032562256, + "learning_rate": 9.9781819978182e-06, + "loss": 0.0618, + "step": 21828 + }, + { + "epoch": 10.148303114830311, + "grad_norm": 6.4935994148254395, + "learning_rate": 9.978180997818101e-06, + "loss": 0.0936, + "step": 21829 + }, + { + "epoch": 10.1487680148768, + "grad_norm": 5.755483627319336, + "learning_rate": 9.978179997818e-06, + "loss": 0.0626, + "step": 21830 + }, + { + "epoch": 10.149232914923292, + "grad_norm": 3.2608354091644287, + "learning_rate": 9.9781789978179e-06, + "loss": 0.0321, + "step": 21831 + }, + { + "epoch": 10.149697814969782, + "grad_norm": 5.90763521194458, + "learning_rate": 9.978177997817801e-06, + "loss": 0.1535, + "step": 21832 + }, + { + "epoch": 10.150162715016272, + "grad_norm": 7.487000465393066, + "learning_rate": 9.9781769978177e-06, + "loss": 0.1388, + "step": 21833 + }, + { + "epoch": 10.150627615062762, + "grad_norm": 3.155780792236328, + "learning_rate": 9.9781759978176e-06, + "loss": 0.0288, + "step": 21834 + }, + { + "epoch": 10.151092515109251, + "grad_norm": 3.194007396697998, + "learning_rate": 9.9781749978175e-06, + "loss": 0.0595, + "step": 21835 + }, + { + "epoch": 10.151557415155741, + "grad_norm": 2.7995176315307617, + "learning_rate": 9.9781739978174e-06, + "loss": 0.0245, + "step": 21836 + }, + { + "epoch": 10.152022315202231, + "grad_norm": 2.343411684036255, + "learning_rate": 9.978172997817301e-06, + "loss": 0.0143, + "step": 21837 + }, + { + "epoch": 10.15248721524872, + "grad_norm": 5.295605659484863, + "learning_rate": 9.978171997817202e-06, + "loss": 0.0979, + "step": 21838 + }, + { + "epoch": 10.15295211529521, + "grad_norm": 8.15295696258545, + "learning_rate": 9.9781709978171e-06, + "loss": 0.1641, + "step": 21839 + }, + { + "epoch": 10.153417015341702, + "grad_norm": 9.444670677185059, + "learning_rate": 9.978169997817001e-06, + "loss": 0.2344, + "step": 21840 + }, + { + "epoch": 10.153881915388192, + "grad_norm": 3.2773566246032715, + "learning_rate": 9.9781689978169e-06, + "loss": 0.0985, + "step": 21841 + }, + { + "epoch": 10.154346815434682, + "grad_norm": 3.5589206218719482, + "learning_rate": 9.978167997816801e-06, + "loss": 0.0601, + "step": 21842 + }, + { + "epoch": 10.154811715481172, + "grad_norm": 6.885006427764893, + "learning_rate": 9.9781669978167e-06, + "loss": 0.0399, + "step": 21843 + }, + { + "epoch": 10.155276615527661, + "grad_norm": 5.9461822509765625, + "learning_rate": 9.9781659978166e-06, + "loss": 0.1692, + "step": 21844 + }, + { + "epoch": 10.155741515574151, + "grad_norm": 2.178675889968872, + "learning_rate": 9.978164997816501e-06, + "loss": 0.0176, + "step": 21845 + }, + { + "epoch": 10.156206415620641, + "grad_norm": 5.979598522186279, + "learning_rate": 9.9781639978164e-06, + "loss": 0.0688, + "step": 21846 + }, + { + "epoch": 10.15667131566713, + "grad_norm": 5.1656646728515625, + "learning_rate": 9.9781629978163e-06, + "loss": 0.0749, + "step": 21847 + }, + { + "epoch": 10.15713621571362, + "grad_norm": 3.6103758811950684, + "learning_rate": 9.9781619978162e-06, + "loss": 0.0275, + "step": 21848 + }, + { + "epoch": 10.157601115760112, + "grad_norm": 3.7482352256774902, + "learning_rate": 9.9781609978161e-06, + "loss": 0.0574, + "step": 21849 + }, + { + "epoch": 10.158066015806602, + "grad_norm": 6.559607982635498, + "learning_rate": 9.978159997816001e-06, + "loss": 0.1862, + "step": 21850 + }, + { + "epoch": 10.158530915853092, + "grad_norm": 3.867065668106079, + "learning_rate": 9.978158997815902e-06, + "loss": 0.0561, + "step": 21851 + }, + { + "epoch": 10.158995815899582, + "grad_norm": 4.073245525360107, + "learning_rate": 9.9781579978158e-06, + "loss": 0.0717, + "step": 21852 + }, + { + "epoch": 10.159460715946071, + "grad_norm": 4.5665669441223145, + "learning_rate": 9.978156997815701e-06, + "loss": 0.0667, + "step": 21853 + }, + { + "epoch": 10.159925615992561, + "grad_norm": 5.981362819671631, + "learning_rate": 9.9781559978156e-06, + "loss": 0.0736, + "step": 21854 + }, + { + "epoch": 10.160390516039051, + "grad_norm": 10.998783111572266, + "learning_rate": 9.9781549978155e-06, + "loss": 0.0666, + "step": 21855 + }, + { + "epoch": 10.16085541608554, + "grad_norm": 4.5886969566345215, + "learning_rate": 9.9781539978154e-06, + "loss": 0.0593, + "step": 21856 + }, + { + "epoch": 10.161320316132032, + "grad_norm": 3.095733404159546, + "learning_rate": 9.9781529978153e-06, + "loss": 0.0379, + "step": 21857 + }, + { + "epoch": 10.161785216178522, + "grad_norm": 1.8391833305358887, + "learning_rate": 9.978151997815201e-06, + "loss": 0.0127, + "step": 21858 + }, + { + "epoch": 10.162250116225012, + "grad_norm": 4.812386989593506, + "learning_rate": 9.9781509978151e-06, + "loss": 0.0593, + "step": 21859 + }, + { + "epoch": 10.162715016271502, + "grad_norm": 3.007394313812256, + "learning_rate": 9.978149997815001e-06, + "loss": 0.0345, + "step": 21860 + }, + { + "epoch": 10.163179916317992, + "grad_norm": 4.7474141120910645, + "learning_rate": 9.9781489978149e-06, + "loss": 0.0342, + "step": 21861 + }, + { + "epoch": 10.163644816364481, + "grad_norm": 12.090398788452148, + "learning_rate": 9.9781479978148e-06, + "loss": 0.2054, + "step": 21862 + }, + { + "epoch": 10.164109716410971, + "grad_norm": 5.263986587524414, + "learning_rate": 9.978146997814701e-06, + "loss": 0.1539, + "step": 21863 + }, + { + "epoch": 10.164574616457461, + "grad_norm": 6.629193305969238, + "learning_rate": 9.978145997814602e-06, + "loss": 0.0787, + "step": 21864 + }, + { + "epoch": 10.16503951650395, + "grad_norm": 4.266034126281738, + "learning_rate": 9.9781449978145e-06, + "loss": 0.0707, + "step": 21865 + }, + { + "epoch": 10.165504416550442, + "grad_norm": 3.779310703277588, + "learning_rate": 9.9781439978144e-06, + "loss": 0.0249, + "step": 21866 + }, + { + "epoch": 10.165969316596932, + "grad_norm": 6.089916706085205, + "learning_rate": 9.9781429978143e-06, + "loss": 0.119, + "step": 21867 + }, + { + "epoch": 10.166434216643422, + "grad_norm": 5.134031772613525, + "learning_rate": 9.9781419978142e-06, + "loss": 0.1112, + "step": 21868 + }, + { + "epoch": 10.166899116689912, + "grad_norm": 3.6093878746032715, + "learning_rate": 9.9781409978141e-06, + "loss": 0.0275, + "step": 21869 + }, + { + "epoch": 10.167364016736402, + "grad_norm": 5.609328269958496, + "learning_rate": 9.978139997814e-06, + "loss": 0.0576, + "step": 21870 + }, + { + "epoch": 10.167828916782891, + "grad_norm": 0.7918329834938049, + "learning_rate": 9.978138997813901e-06, + "loss": 0.004, + "step": 21871 + }, + { + "epoch": 10.168293816829381, + "grad_norm": 3.7965457439422607, + "learning_rate": 9.9781379978138e-06, + "loss": 0.0351, + "step": 21872 + }, + { + "epoch": 10.168758716875871, + "grad_norm": 3.188415288925171, + "learning_rate": 9.978136997813701e-06, + "loss": 0.0255, + "step": 21873 + }, + { + "epoch": 10.16922361692236, + "grad_norm": 4.518691062927246, + "learning_rate": 9.9781359978136e-06, + "loss": 0.0348, + "step": 21874 + }, + { + "epoch": 10.169688516968852, + "grad_norm": 3.771186351776123, + "learning_rate": 9.9781349978135e-06, + "loss": 0.0382, + "step": 21875 + }, + { + "epoch": 10.170153417015342, + "grad_norm": 8.838626861572266, + "learning_rate": 9.978133997813401e-06, + "loss": 0.2189, + "step": 21876 + }, + { + "epoch": 10.170618317061832, + "grad_norm": 2.5053303241729736, + "learning_rate": 9.9781329978133e-06, + "loss": 0.0297, + "step": 21877 + }, + { + "epoch": 10.171083217108322, + "grad_norm": 3.9466476440429688, + "learning_rate": 9.978131997813201e-06, + "loss": 0.0428, + "step": 21878 + }, + { + "epoch": 10.171548117154812, + "grad_norm": 10.354439735412598, + "learning_rate": 9.9781309978131e-06, + "loss": 0.1155, + "step": 21879 + }, + { + "epoch": 10.172013017201301, + "grad_norm": 4.402935028076172, + "learning_rate": 9.978129997813e-06, + "loss": 0.053, + "step": 21880 + }, + { + "epoch": 10.172477917247791, + "grad_norm": 4.390358924865723, + "learning_rate": 9.978128997812901e-06, + "loss": 0.0319, + "step": 21881 + }, + { + "epoch": 10.172942817294281, + "grad_norm": 7.324160575866699, + "learning_rate": 9.9781279978128e-06, + "loss": 0.1769, + "step": 21882 + }, + { + "epoch": 10.17340771734077, + "grad_norm": 7.942612171173096, + "learning_rate": 9.9781269978127e-06, + "loss": 0.0593, + "step": 21883 + }, + { + "epoch": 10.173872617387262, + "grad_norm": 7.400448799133301, + "learning_rate": 9.978125997812601e-06, + "loss": 0.1607, + "step": 21884 + }, + { + "epoch": 10.174337517433752, + "grad_norm": 6.313314437866211, + "learning_rate": 9.9781249978125e-06, + "loss": 0.1038, + "step": 21885 + }, + { + "epoch": 10.174802417480242, + "grad_norm": 4.647014141082764, + "learning_rate": 9.9781239978124e-06, + "loss": 0.0335, + "step": 21886 + }, + { + "epoch": 10.175267317526732, + "grad_norm": 4.457285404205322, + "learning_rate": 9.9781229978123e-06, + "loss": 0.0779, + "step": 21887 + }, + { + "epoch": 10.175732217573222, + "grad_norm": 4.733080863952637, + "learning_rate": 9.9781219978122e-06, + "loss": 0.0629, + "step": 21888 + }, + { + "epoch": 10.176197117619711, + "grad_norm": 7.340285778045654, + "learning_rate": 9.978120997812101e-06, + "loss": 0.039, + "step": 21889 + }, + { + "epoch": 10.176662017666201, + "grad_norm": 6.30647087097168, + "learning_rate": 9.978119997812e-06, + "loss": 0.1669, + "step": 21890 + }, + { + "epoch": 10.177126917712691, + "grad_norm": 3.0205602645874023, + "learning_rate": 9.978118997811901e-06, + "loss": 0.0448, + "step": 21891 + }, + { + "epoch": 10.177591817759183, + "grad_norm": 4.9665703773498535, + "learning_rate": 9.9781179978118e-06, + "loss": 0.0611, + "step": 21892 + }, + { + "epoch": 10.178056717805672, + "grad_norm": 3.661881923675537, + "learning_rate": 9.9781169978117e-06, + "loss": 0.0378, + "step": 21893 + }, + { + "epoch": 10.178521617852162, + "grad_norm": 2.4063222408294678, + "learning_rate": 9.978115997811601e-06, + "loss": 0.0243, + "step": 21894 + }, + { + "epoch": 10.178986517898652, + "grad_norm": 6.595726013183594, + "learning_rate": 9.9781149978115e-06, + "loss": 0.1294, + "step": 21895 + }, + { + "epoch": 10.179451417945142, + "grad_norm": 4.954049110412598, + "learning_rate": 9.978113997811401e-06, + "loss": 0.0453, + "step": 21896 + }, + { + "epoch": 10.179916317991632, + "grad_norm": 4.655083656311035, + "learning_rate": 9.9781129978113e-06, + "loss": 0.0686, + "step": 21897 + }, + { + "epoch": 10.180381218038121, + "grad_norm": 4.437933444976807, + "learning_rate": 9.9781119978112e-06, + "loss": 0.0527, + "step": 21898 + }, + { + "epoch": 10.180846118084611, + "grad_norm": 1.533707857131958, + "learning_rate": 9.9781109978111e-06, + "loss": 0.0085, + "step": 21899 + }, + { + "epoch": 10.181311018131101, + "grad_norm": 3.850249767303467, + "learning_rate": 9.978109997811e-06, + "loss": 0.0275, + "step": 21900 + }, + { + "epoch": 10.181775918177593, + "grad_norm": 6.642127513885498, + "learning_rate": 9.9781089978109e-06, + "loss": 0.1118, + "step": 21901 + }, + { + "epoch": 10.182240818224082, + "grad_norm": 5.768955707550049, + "learning_rate": 9.978107997810801e-06, + "loss": 0.0682, + "step": 21902 + }, + { + "epoch": 10.182705718270572, + "grad_norm": 3.4324281215667725, + "learning_rate": 9.9781069978107e-06, + "loss": 0.0513, + "step": 21903 + }, + { + "epoch": 10.183170618317062, + "grad_norm": 9.214234352111816, + "learning_rate": 9.978105997810601e-06, + "loss": 0.2181, + "step": 21904 + }, + { + "epoch": 10.183635518363552, + "grad_norm": 6.863590240478516, + "learning_rate": 9.9781049978105e-06, + "loss": 0.1499, + "step": 21905 + }, + { + "epoch": 10.184100418410042, + "grad_norm": 4.914795398712158, + "learning_rate": 9.9781039978104e-06, + "loss": 0.0769, + "step": 21906 + }, + { + "epoch": 10.184565318456531, + "grad_norm": 2.119683027267456, + "learning_rate": 9.978102997810301e-06, + "loss": 0.0228, + "step": 21907 + }, + { + "epoch": 10.185030218503021, + "grad_norm": 2.8318042755126953, + "learning_rate": 9.9781019978102e-06, + "loss": 0.0205, + "step": 21908 + }, + { + "epoch": 10.185495118549511, + "grad_norm": 4.104216575622559, + "learning_rate": 9.978100997810101e-06, + "loss": 0.0753, + "step": 21909 + }, + { + "epoch": 10.185960018596003, + "grad_norm": 3.7355422973632812, + "learning_rate": 9.97809999781e-06, + "loss": 0.0429, + "step": 21910 + }, + { + "epoch": 10.186424918642492, + "grad_norm": 9.358935356140137, + "learning_rate": 9.9780989978099e-06, + "loss": 0.306, + "step": 21911 + }, + { + "epoch": 10.186889818688982, + "grad_norm": 3.888888120651245, + "learning_rate": 9.9780979978098e-06, + "loss": 0.0559, + "step": 21912 + }, + { + "epoch": 10.187354718735472, + "grad_norm": 4.497946262359619, + "learning_rate": 9.9780969978097e-06, + "loss": 0.038, + "step": 21913 + }, + { + "epoch": 10.187819618781962, + "grad_norm": 2.8942363262176514, + "learning_rate": 9.978095997809601e-06, + "loss": 0.0262, + "step": 21914 + }, + { + "epoch": 10.188284518828452, + "grad_norm": 5.608415126800537, + "learning_rate": 9.978094997809502e-06, + "loss": 0.1227, + "step": 21915 + }, + { + "epoch": 10.188749418874941, + "grad_norm": 4.307423114776611, + "learning_rate": 9.9780939978094e-06, + "loss": 0.039, + "step": 21916 + }, + { + "epoch": 10.189214318921431, + "grad_norm": 6.629819869995117, + "learning_rate": 9.978092997809301e-06, + "loss": 0.108, + "step": 21917 + }, + { + "epoch": 10.189679218967921, + "grad_norm": 0.9395545125007629, + "learning_rate": 9.9780919978092e-06, + "loss": 0.0082, + "step": 21918 + }, + { + "epoch": 10.190144119014413, + "grad_norm": 6.996090412139893, + "learning_rate": 9.9780909978091e-06, + "loss": 0.0568, + "step": 21919 + }, + { + "epoch": 10.190609019060902, + "grad_norm": 5.882148265838623, + "learning_rate": 9.978089997809001e-06, + "loss": 0.0954, + "step": 21920 + }, + { + "epoch": 10.191073919107392, + "grad_norm": 3.3718931674957275, + "learning_rate": 9.9780889978089e-06, + "loss": 0.0456, + "step": 21921 + }, + { + "epoch": 10.191538819153882, + "grad_norm": 6.257007122039795, + "learning_rate": 9.978087997808801e-06, + "loss": 0.1105, + "step": 21922 + }, + { + "epoch": 10.192003719200372, + "grad_norm": 8.510591506958008, + "learning_rate": 9.9780869978087e-06, + "loss": 0.0771, + "step": 21923 + }, + { + "epoch": 10.192468619246862, + "grad_norm": 7.704932689666748, + "learning_rate": 9.9780859978086e-06, + "loss": 0.0967, + "step": 21924 + }, + { + "epoch": 10.192933519293351, + "grad_norm": 6.288543701171875, + "learning_rate": 9.9780849978085e-06, + "loss": 0.0308, + "step": 21925 + }, + { + "epoch": 10.193398419339841, + "grad_norm": 2.7459418773651123, + "learning_rate": 9.9780839978084e-06, + "loss": 0.0223, + "step": 21926 + }, + { + "epoch": 10.193863319386331, + "grad_norm": 4.470290184020996, + "learning_rate": 9.978082997808301e-06, + "loss": 0.0458, + "step": 21927 + }, + { + "epoch": 10.194328219432823, + "grad_norm": 5.685705184936523, + "learning_rate": 9.978081997808202e-06, + "loss": 0.0514, + "step": 21928 + }, + { + "epoch": 10.194793119479312, + "grad_norm": 4.32313346862793, + "learning_rate": 9.9780809978081e-06, + "loss": 0.0447, + "step": 21929 + }, + { + "epoch": 10.195258019525802, + "grad_norm": 4.515953063964844, + "learning_rate": 9.978079997808e-06, + "loss": 0.041, + "step": 21930 + }, + { + "epoch": 10.195722919572292, + "grad_norm": 6.7936553955078125, + "learning_rate": 9.9780789978079e-06, + "loss": 0.1266, + "step": 21931 + }, + { + "epoch": 10.196187819618782, + "grad_norm": 5.468379020690918, + "learning_rate": 9.978077997807801e-06, + "loss": 0.0597, + "step": 21932 + }, + { + "epoch": 10.196652719665272, + "grad_norm": 1.8779467344284058, + "learning_rate": 9.978076997807702e-06, + "loss": 0.011, + "step": 21933 + }, + { + "epoch": 10.197117619711761, + "grad_norm": 6.5066094398498535, + "learning_rate": 9.9780759978076e-06, + "loss": 0.1367, + "step": 21934 + }, + { + "epoch": 10.197582519758251, + "grad_norm": 6.765255451202393, + "learning_rate": 9.978074997807501e-06, + "loss": 0.1029, + "step": 21935 + }, + { + "epoch": 10.198047419804743, + "grad_norm": 4.587615489959717, + "learning_rate": 9.9780739978074e-06, + "loss": 0.0656, + "step": 21936 + }, + { + "epoch": 10.198512319851233, + "grad_norm": 11.531027793884277, + "learning_rate": 9.9780729978073e-06, + "loss": 0.2917, + "step": 21937 + }, + { + "epoch": 10.198977219897722, + "grad_norm": 1.3587992191314697, + "learning_rate": 9.9780719978072e-06, + "loss": 0.0108, + "step": 21938 + }, + { + "epoch": 10.199442119944212, + "grad_norm": 2.555513381958008, + "learning_rate": 9.9780709978071e-06, + "loss": 0.0282, + "step": 21939 + }, + { + "epoch": 10.199907019990702, + "grad_norm": 2.1439406871795654, + "learning_rate": 9.978069997807001e-06, + "loss": 0.0255, + "step": 21940 + }, + { + "epoch": 10.200371920037192, + "grad_norm": 7.207989692687988, + "learning_rate": 9.9780689978069e-06, + "loss": 0.0712, + "step": 21941 + }, + { + "epoch": 10.200836820083682, + "grad_norm": 4.355983257293701, + "learning_rate": 9.9780679978068e-06, + "loss": 0.0675, + "step": 21942 + }, + { + "epoch": 10.201301720130171, + "grad_norm": 7.61398458480835, + "learning_rate": 9.9780669978067e-06, + "loss": 0.1925, + "step": 21943 + }, + { + "epoch": 10.201766620176661, + "grad_norm": 6.100238800048828, + "learning_rate": 9.9780659978066e-06, + "loss": 0.0723, + "step": 21944 + }, + { + "epoch": 10.202231520223153, + "grad_norm": 7.263975143432617, + "learning_rate": 9.978064997806501e-06, + "loss": 0.2103, + "step": 21945 + }, + { + "epoch": 10.202696420269643, + "grad_norm": 5.676934242248535, + "learning_rate": 9.978063997806402e-06, + "loss": 0.0795, + "step": 21946 + }, + { + "epoch": 10.203161320316132, + "grad_norm": 7.006772518157959, + "learning_rate": 9.9780629978063e-06, + "loss": 0.1314, + "step": 21947 + }, + { + "epoch": 10.203626220362622, + "grad_norm": 4.235074043273926, + "learning_rate": 9.978061997806201e-06, + "loss": 0.0926, + "step": 21948 + }, + { + "epoch": 10.204091120409112, + "grad_norm": 4.604354381561279, + "learning_rate": 9.9780609978061e-06, + "loss": 0.0831, + "step": 21949 + }, + { + "epoch": 10.204556020455602, + "grad_norm": 3.3777389526367188, + "learning_rate": 9.978059997806e-06, + "loss": 0.0299, + "step": 21950 + }, + { + "epoch": 10.205020920502092, + "grad_norm": 6.846292495727539, + "learning_rate": 9.9780589978059e-06, + "loss": 0.124, + "step": 21951 + }, + { + "epoch": 10.205485820548581, + "grad_norm": 2.0857291221618652, + "learning_rate": 9.9780579978058e-06, + "loss": 0.0112, + "step": 21952 + }, + { + "epoch": 10.205950720595071, + "grad_norm": 8.584444999694824, + "learning_rate": 9.978056997805701e-06, + "loss": 0.0912, + "step": 21953 + }, + { + "epoch": 10.206415620641563, + "grad_norm": 4.906986713409424, + "learning_rate": 9.9780559978056e-06, + "loss": 0.0742, + "step": 21954 + }, + { + "epoch": 10.206880520688053, + "grad_norm": 4.509481906890869, + "learning_rate": 9.9780549978055e-06, + "loss": 0.0737, + "step": 21955 + }, + { + "epoch": 10.207345420734542, + "grad_norm": 2.822751998901367, + "learning_rate": 9.9780539978054e-06, + "loss": 0.0288, + "step": 21956 + }, + { + "epoch": 10.207810320781032, + "grad_norm": 5.431180000305176, + "learning_rate": 9.9780529978053e-06, + "loss": 0.0984, + "step": 21957 + }, + { + "epoch": 10.208275220827522, + "grad_norm": 4.963393211364746, + "learning_rate": 9.978051997805201e-06, + "loss": 0.0527, + "step": 21958 + }, + { + "epoch": 10.208740120874012, + "grad_norm": 4.260735034942627, + "learning_rate": 9.978050997805102e-06, + "loss": 0.0752, + "step": 21959 + }, + { + "epoch": 10.209205020920502, + "grad_norm": 8.910834312438965, + "learning_rate": 9.978049997805e-06, + "loss": 0.1034, + "step": 21960 + }, + { + "epoch": 10.209669920966991, + "grad_norm": 3.845741033554077, + "learning_rate": 9.978048997804901e-06, + "loss": 0.0419, + "step": 21961 + }, + { + "epoch": 10.210134821013481, + "grad_norm": 4.0767927169799805, + "learning_rate": 9.9780479978048e-06, + "loss": 0.0365, + "step": 21962 + }, + { + "epoch": 10.210599721059973, + "grad_norm": 4.719191551208496, + "learning_rate": 9.978046997804701e-06, + "loss": 0.0409, + "step": 21963 + }, + { + "epoch": 10.211064621106463, + "grad_norm": 2.864154815673828, + "learning_rate": 9.9780459978046e-06, + "loss": 0.0324, + "step": 21964 + }, + { + "epoch": 10.211529521152952, + "grad_norm": 4.3907694816589355, + "learning_rate": 9.9780449978045e-06, + "loss": 0.0427, + "step": 21965 + }, + { + "epoch": 10.211994421199442, + "grad_norm": 3.0758769512176514, + "learning_rate": 9.978043997804401e-06, + "loss": 0.0278, + "step": 21966 + }, + { + "epoch": 10.212459321245932, + "grad_norm": 6.70267915725708, + "learning_rate": 9.9780429978043e-06, + "loss": 0.0983, + "step": 21967 + }, + { + "epoch": 10.212924221292422, + "grad_norm": 5.540199279785156, + "learning_rate": 9.9780419978042e-06, + "loss": 0.0596, + "step": 21968 + }, + { + "epoch": 10.213389121338912, + "grad_norm": 4.188452243804932, + "learning_rate": 9.9780409978041e-06, + "loss": 0.0441, + "step": 21969 + }, + { + "epoch": 10.213854021385401, + "grad_norm": 4.4059247970581055, + "learning_rate": 9.978039997804e-06, + "loss": 0.061, + "step": 21970 + }, + { + "epoch": 10.214318921431893, + "grad_norm": 8.928661346435547, + "learning_rate": 9.978038997803901e-06, + "loss": 0.1548, + "step": 21971 + }, + { + "epoch": 10.214783821478383, + "grad_norm": 5.98408317565918, + "learning_rate": 9.978037997803802e-06, + "loss": 0.0839, + "step": 21972 + }, + { + "epoch": 10.215248721524873, + "grad_norm": 3.455864667892456, + "learning_rate": 9.9780369978037e-06, + "loss": 0.0415, + "step": 21973 + }, + { + "epoch": 10.215713621571362, + "grad_norm": 5.394974231719971, + "learning_rate": 9.9780359978036e-06, + "loss": 0.1391, + "step": 21974 + }, + { + "epoch": 10.216178521617852, + "grad_norm": 1.2221368551254272, + "learning_rate": 9.9780349978035e-06, + "loss": 0.0072, + "step": 21975 + }, + { + "epoch": 10.216643421664342, + "grad_norm": 6.090486526489258, + "learning_rate": 9.978033997803401e-06, + "loss": 0.1091, + "step": 21976 + }, + { + "epoch": 10.217108321710832, + "grad_norm": 4.682647705078125, + "learning_rate": 9.978032997803302e-06, + "loss": 0.0597, + "step": 21977 + }, + { + "epoch": 10.217573221757322, + "grad_norm": 15.437673568725586, + "learning_rate": 9.9780319978032e-06, + "loss": 0.1341, + "step": 21978 + }, + { + "epoch": 10.218038121803811, + "grad_norm": 3.194780111312866, + "learning_rate": 9.978030997803101e-06, + "loss": 0.0478, + "step": 21979 + }, + { + "epoch": 10.218503021850303, + "grad_norm": 5.14716100692749, + "learning_rate": 9.978029997803e-06, + "loss": 0.1346, + "step": 21980 + }, + { + "epoch": 10.218967921896793, + "grad_norm": 3.006394147872925, + "learning_rate": 9.978028997802901e-06, + "loss": 0.0413, + "step": 21981 + }, + { + "epoch": 10.219432821943283, + "grad_norm": 2.3055026531219482, + "learning_rate": 9.9780279978028e-06, + "loss": 0.0146, + "step": 21982 + }, + { + "epoch": 10.219897721989772, + "grad_norm": 3.5130796432495117, + "learning_rate": 9.9780269978027e-06, + "loss": 0.0528, + "step": 21983 + }, + { + "epoch": 10.220362622036262, + "grad_norm": 1.246915578842163, + "learning_rate": 9.978025997802601e-06, + "loss": 0.01, + "step": 21984 + }, + { + "epoch": 10.220827522082752, + "grad_norm": 5.13881254196167, + "learning_rate": 9.9780249978025e-06, + "loss": 0.12, + "step": 21985 + }, + { + "epoch": 10.221292422129242, + "grad_norm": 9.369044303894043, + "learning_rate": 9.9780239978024e-06, + "loss": 0.1105, + "step": 21986 + }, + { + "epoch": 10.221757322175732, + "grad_norm": 1.9798575639724731, + "learning_rate": 9.9780229978023e-06, + "loss": 0.0174, + "step": 21987 + }, + { + "epoch": 10.222222222222221, + "grad_norm": 5.6795973777771, + "learning_rate": 9.9780219978022e-06, + "loss": 0.0903, + "step": 21988 + }, + { + "epoch": 10.222687122268713, + "grad_norm": 5.614607810974121, + "learning_rate": 9.978020997802101e-06, + "loss": 0.0427, + "step": 21989 + }, + { + "epoch": 10.223152022315203, + "grad_norm": 3.7929701805114746, + "learning_rate": 9.978019997802002e-06, + "loss": 0.0432, + "step": 21990 + }, + { + "epoch": 10.223616922361693, + "grad_norm": 3.7905821800231934, + "learning_rate": 9.9780189978019e-06, + "loss": 0.035, + "step": 21991 + }, + { + "epoch": 10.224081822408182, + "grad_norm": 5.002401828765869, + "learning_rate": 9.978017997801801e-06, + "loss": 0.0753, + "step": 21992 + }, + { + "epoch": 10.224546722454672, + "grad_norm": 5.188263416290283, + "learning_rate": 9.9780169978017e-06, + "loss": 0.0502, + "step": 21993 + }, + { + "epoch": 10.225011622501162, + "grad_norm": 3.839585781097412, + "learning_rate": 9.978015997801601e-06, + "loss": 0.0343, + "step": 21994 + }, + { + "epoch": 10.225476522547652, + "grad_norm": 6.143496036529541, + "learning_rate": 9.9780149978015e-06, + "loss": 0.0626, + "step": 21995 + }, + { + "epoch": 10.225941422594142, + "grad_norm": 2.3798017501831055, + "learning_rate": 9.9780139978014e-06, + "loss": 0.0383, + "step": 21996 + }, + { + "epoch": 10.226406322640631, + "grad_norm": 4.572930335998535, + "learning_rate": 9.978012997801301e-06, + "loss": 0.0205, + "step": 21997 + }, + { + "epoch": 10.226871222687123, + "grad_norm": 6.100230693817139, + "learning_rate": 9.9780119978012e-06, + "loss": 0.0472, + "step": 21998 + }, + { + "epoch": 10.227336122733613, + "grad_norm": 4.8619256019592285, + "learning_rate": 9.978010997801101e-06, + "loss": 0.0627, + "step": 21999 + }, + { + "epoch": 10.227801022780103, + "grad_norm": 6.0521368980407715, + "learning_rate": 9.978009997801e-06, + "loss": 0.1166, + "step": 22000 + }, + { + "epoch": 10.228265922826592, + "grad_norm": 4.941682815551758, + "learning_rate": 9.9780089978009e-06, + "loss": 0.1057, + "step": 22001 + }, + { + "epoch": 10.228730822873082, + "grad_norm": 2.727044105529785, + "learning_rate": 9.978007997800801e-06, + "loss": 0.0456, + "step": 22002 + }, + { + "epoch": 10.229195722919572, + "grad_norm": 9.521303176879883, + "learning_rate": 9.978006997800702e-06, + "loss": 0.0775, + "step": 22003 + }, + { + "epoch": 10.229660622966062, + "grad_norm": 6.576513767242432, + "learning_rate": 9.9780059978006e-06, + "loss": 0.0822, + "step": 22004 + }, + { + "epoch": 10.230125523012552, + "grad_norm": 4.4006876945495605, + "learning_rate": 9.978004997800501e-06, + "loss": 0.0868, + "step": 22005 + }, + { + "epoch": 10.230590423059041, + "grad_norm": 6.235368728637695, + "learning_rate": 9.9780039978004e-06, + "loss": 0.1291, + "step": 22006 + }, + { + "epoch": 10.231055323105533, + "grad_norm": 1.4553499221801758, + "learning_rate": 9.9780029978003e-06, + "loss": 0.016, + "step": 22007 + }, + { + "epoch": 10.231520223152023, + "grad_norm": 4.825466156005859, + "learning_rate": 9.9780019978002e-06, + "loss": 0.0616, + "step": 22008 + }, + { + "epoch": 10.231985123198513, + "grad_norm": 1.6345137357711792, + "learning_rate": 9.9780009978001e-06, + "loss": 0.0109, + "step": 22009 + }, + { + "epoch": 10.232450023245002, + "grad_norm": 3.5975329875946045, + "learning_rate": 9.977999997800001e-06, + "loss": 0.0386, + "step": 22010 + }, + { + "epoch": 10.232914923291492, + "grad_norm": 7.61549186706543, + "learning_rate": 9.9779989977999e-06, + "loss": 0.1067, + "step": 22011 + }, + { + "epoch": 10.233379823337982, + "grad_norm": 5.338262557983398, + "learning_rate": 9.977997997799801e-06, + "loss": 0.0724, + "step": 22012 + }, + { + "epoch": 10.233844723384472, + "grad_norm": 3.424412965774536, + "learning_rate": 9.9779969977997e-06, + "loss": 0.0479, + "step": 22013 + }, + { + "epoch": 10.234309623430962, + "grad_norm": 6.108925819396973, + "learning_rate": 9.9779959977996e-06, + "loss": 0.1194, + "step": 22014 + }, + { + "epoch": 10.234774523477453, + "grad_norm": 8.634987831115723, + "learning_rate": 9.977994997799501e-06, + "loss": 0.1236, + "step": 22015 + }, + { + "epoch": 10.235239423523943, + "grad_norm": 4.990159034729004, + "learning_rate": 9.977993997799402e-06, + "loss": 0.0621, + "step": 22016 + }, + { + "epoch": 10.235704323570433, + "grad_norm": 5.996792316436768, + "learning_rate": 9.977992997799301e-06, + "loss": 0.0857, + "step": 22017 + }, + { + "epoch": 10.236169223616923, + "grad_norm": 5.897852897644043, + "learning_rate": 9.9779919977992e-06, + "loss": 0.1368, + "step": 22018 + }, + { + "epoch": 10.236634123663412, + "grad_norm": 3.955646514892578, + "learning_rate": 9.9779909977991e-06, + "loss": 0.0215, + "step": 22019 + }, + { + "epoch": 10.237099023709902, + "grad_norm": 5.846563816070557, + "learning_rate": 9.977989997799e-06, + "loss": 0.1446, + "step": 22020 + }, + { + "epoch": 10.237563923756392, + "grad_norm": 2.923637628555298, + "learning_rate": 9.9779889977989e-06, + "loss": 0.0192, + "step": 22021 + }, + { + "epoch": 10.238028823802882, + "grad_norm": 4.988077163696289, + "learning_rate": 9.9779879977988e-06, + "loss": 0.0733, + "step": 22022 + }, + { + "epoch": 10.238493723849372, + "grad_norm": 2.3921849727630615, + "learning_rate": 9.977986997798701e-06, + "loss": 0.0251, + "step": 22023 + }, + { + "epoch": 10.238958623895863, + "grad_norm": 2.3133015632629395, + "learning_rate": 9.9779859977986e-06, + "loss": 0.0159, + "step": 22024 + }, + { + "epoch": 10.239423523942353, + "grad_norm": 5.093810081481934, + "learning_rate": 9.977984997798501e-06, + "loss": 0.1414, + "step": 22025 + }, + { + "epoch": 10.239888423988843, + "grad_norm": 3.885085105895996, + "learning_rate": 9.9779839977984e-06, + "loss": 0.0408, + "step": 22026 + }, + { + "epoch": 10.240353324035333, + "grad_norm": 5.470977783203125, + "learning_rate": 9.9779829977983e-06, + "loss": 0.055, + "step": 22027 + }, + { + "epoch": 10.240818224081822, + "grad_norm": 5.200605392456055, + "learning_rate": 9.977981997798201e-06, + "loss": 0.069, + "step": 22028 + }, + { + "epoch": 10.241283124128312, + "grad_norm": 7.244374752044678, + "learning_rate": 9.9779809977981e-06, + "loss": 0.1062, + "step": 22029 + }, + { + "epoch": 10.241748024174802, + "grad_norm": 4.869099140167236, + "learning_rate": 9.977979997798001e-06, + "loss": 0.0832, + "step": 22030 + }, + { + "epoch": 10.242212924221292, + "grad_norm": 5.4009690284729, + "learning_rate": 9.9779789977979e-06, + "loss": 0.1146, + "step": 22031 + }, + { + "epoch": 10.242677824267782, + "grad_norm": 4.818307399749756, + "learning_rate": 9.9779779977978e-06, + "loss": 0.1124, + "step": 22032 + }, + { + "epoch": 10.243142724314273, + "grad_norm": 2.9179859161376953, + "learning_rate": 9.9779769977977e-06, + "loss": 0.0238, + "step": 22033 + }, + { + "epoch": 10.243607624360763, + "grad_norm": 6.468710899353027, + "learning_rate": 9.9779759977976e-06, + "loss": 0.1268, + "step": 22034 + }, + { + "epoch": 10.244072524407253, + "grad_norm": 3.4904439449310303, + "learning_rate": 9.9779749977975e-06, + "loss": 0.0315, + "step": 22035 + }, + { + "epoch": 10.244537424453743, + "grad_norm": 5.208137035369873, + "learning_rate": 9.977973997797402e-06, + "loss": 0.0815, + "step": 22036 + }, + { + "epoch": 10.245002324500232, + "grad_norm": 3.5039727687835693, + "learning_rate": 9.9779729977973e-06, + "loss": 0.0651, + "step": 22037 + }, + { + "epoch": 10.245467224546722, + "grad_norm": 1.6271203756332397, + "learning_rate": 9.9779719977972e-06, + "loss": 0.011, + "step": 22038 + }, + { + "epoch": 10.245932124593212, + "grad_norm": 5.222403049468994, + "learning_rate": 9.9779709977971e-06, + "loss": 0.0737, + "step": 22039 + }, + { + "epoch": 10.246397024639702, + "grad_norm": 2.7305822372436523, + "learning_rate": 9.977969997797e-06, + "loss": 0.0331, + "step": 22040 + }, + { + "epoch": 10.246861924686192, + "grad_norm": 5.6896162033081055, + "learning_rate": 9.977968997796901e-06, + "loss": 0.1098, + "step": 22041 + }, + { + "epoch": 10.247326824732683, + "grad_norm": 3.678524971008301, + "learning_rate": 9.9779679977968e-06, + "loss": 0.0566, + "step": 22042 + }, + { + "epoch": 10.247791724779173, + "grad_norm": 3.4716439247131348, + "learning_rate": 9.977966997796701e-06, + "loss": 0.0293, + "step": 22043 + }, + { + "epoch": 10.248256624825663, + "grad_norm": 8.073040962219238, + "learning_rate": 9.9779659977966e-06, + "loss": 0.0392, + "step": 22044 + }, + { + "epoch": 10.248721524872153, + "grad_norm": 2.6100261211395264, + "learning_rate": 9.9779649977965e-06, + "loss": 0.0143, + "step": 22045 + }, + { + "epoch": 10.249186424918642, + "grad_norm": 2.0855836868286133, + "learning_rate": 9.977963997796401e-06, + "loss": 0.0188, + "step": 22046 + }, + { + "epoch": 10.249651324965132, + "grad_norm": 3.2704508304595947, + "learning_rate": 9.9779629977963e-06, + "loss": 0.0394, + "step": 22047 + }, + { + "epoch": 10.250116225011622, + "grad_norm": 9.39721393585205, + "learning_rate": 9.977961997796201e-06, + "loss": 0.137, + "step": 22048 + }, + { + "epoch": 10.250581125058112, + "grad_norm": 6.8466362953186035, + "learning_rate": 9.9779609977961e-06, + "loss": 0.0987, + "step": 22049 + }, + { + "epoch": 10.251046025104603, + "grad_norm": 2.956604480743408, + "learning_rate": 9.977959997796e-06, + "loss": 0.0288, + "step": 22050 + }, + { + "epoch": 10.251510925151093, + "grad_norm": 6.797702789306641, + "learning_rate": 9.9779589977959e-06, + "loss": 0.0964, + "step": 22051 + }, + { + "epoch": 10.251975825197583, + "grad_norm": 4.695347309112549, + "learning_rate": 9.9779579977958e-06, + "loss": 0.0704, + "step": 22052 + }, + { + "epoch": 10.252440725244073, + "grad_norm": 3.6451473236083984, + "learning_rate": 9.9779569977957e-06, + "loss": 0.0333, + "step": 22053 + }, + { + "epoch": 10.252905625290563, + "grad_norm": 5.244359016418457, + "learning_rate": 9.977955997795601e-06, + "loss": 0.1529, + "step": 22054 + }, + { + "epoch": 10.253370525337052, + "grad_norm": 4.540570259094238, + "learning_rate": 9.9779549977955e-06, + "loss": 0.0774, + "step": 22055 + }, + { + "epoch": 10.253835425383542, + "grad_norm": 3.0587854385375977, + "learning_rate": 9.977953997795401e-06, + "loss": 0.0288, + "step": 22056 + }, + { + "epoch": 10.254300325430032, + "grad_norm": 8.178085327148438, + "learning_rate": 9.9779529977953e-06, + "loss": 0.151, + "step": 22057 + }, + { + "epoch": 10.254765225476522, + "grad_norm": 6.653053283691406, + "learning_rate": 9.9779519977952e-06, + "loss": 0.1555, + "step": 22058 + }, + { + "epoch": 10.255230125523013, + "grad_norm": 3.9762847423553467, + "learning_rate": 9.977950997795101e-06, + "loss": 0.0376, + "step": 22059 + }, + { + "epoch": 10.255695025569503, + "grad_norm": 3.4661664962768555, + "learning_rate": 9.977949997795e-06, + "loss": 0.045, + "step": 22060 + }, + { + "epoch": 10.256159925615993, + "grad_norm": 3.33280611038208, + "learning_rate": 9.977948997794901e-06, + "loss": 0.0256, + "step": 22061 + }, + { + "epoch": 10.256624825662483, + "grad_norm": 3.4244866371154785, + "learning_rate": 9.9779479977948e-06, + "loss": 0.0433, + "step": 22062 + }, + { + "epoch": 10.257089725708973, + "grad_norm": 3.430114984512329, + "learning_rate": 9.9779469977947e-06, + "loss": 0.0269, + "step": 22063 + }, + { + "epoch": 10.257554625755462, + "grad_norm": 4.581254482269287, + "learning_rate": 9.9779459977946e-06, + "loss": 0.082, + "step": 22064 + }, + { + "epoch": 10.258019525801952, + "grad_norm": 4.23232364654541, + "learning_rate": 9.9779449977945e-06, + "loss": 0.0943, + "step": 22065 + }, + { + "epoch": 10.258484425848442, + "grad_norm": 2.743908166885376, + "learning_rate": 9.977943997794401e-06, + "loss": 0.0226, + "step": 22066 + }, + { + "epoch": 10.258949325894932, + "grad_norm": 4.153326034545898, + "learning_rate": 9.977942997794302e-06, + "loss": 0.0902, + "step": 22067 + }, + { + "epoch": 10.259414225941423, + "grad_norm": 2.7891626358032227, + "learning_rate": 9.9779419977942e-06, + "loss": 0.0261, + "step": 22068 + }, + { + "epoch": 10.259879125987913, + "grad_norm": 4.895242214202881, + "learning_rate": 9.977940997794101e-06, + "loss": 0.0856, + "step": 22069 + }, + { + "epoch": 10.260344026034403, + "grad_norm": 8.39931869506836, + "learning_rate": 9.977939997794e-06, + "loss": 0.1441, + "step": 22070 + }, + { + "epoch": 10.260808926080893, + "grad_norm": 3.485542058944702, + "learning_rate": 9.9779389977939e-06, + "loss": 0.0622, + "step": 22071 + }, + { + "epoch": 10.261273826127383, + "grad_norm": 3.5482804775238037, + "learning_rate": 9.977937997793801e-06, + "loss": 0.0596, + "step": 22072 + }, + { + "epoch": 10.261738726173872, + "grad_norm": 5.718253135681152, + "learning_rate": 9.9779369977937e-06, + "loss": 0.1071, + "step": 22073 + }, + { + "epoch": 10.262203626220362, + "grad_norm": 5.218810081481934, + "learning_rate": 9.977935997793601e-06, + "loss": 0.125, + "step": 22074 + }, + { + "epoch": 10.262668526266852, + "grad_norm": 2.723463535308838, + "learning_rate": 9.9779349977935e-06, + "loss": 0.0271, + "step": 22075 + }, + { + "epoch": 10.263133426313342, + "grad_norm": 3.733942747116089, + "learning_rate": 9.9779339977934e-06, + "loss": 0.0516, + "step": 22076 + }, + { + "epoch": 10.263598326359833, + "grad_norm": 1.559516429901123, + "learning_rate": 9.9779329977933e-06, + "loss": 0.0147, + "step": 22077 + }, + { + "epoch": 10.264063226406323, + "grad_norm": 4.932368278503418, + "learning_rate": 9.9779319977932e-06, + "loss": 0.0828, + "step": 22078 + }, + { + "epoch": 10.264528126452813, + "grad_norm": 4.024162292480469, + "learning_rate": 9.977930997793101e-06, + "loss": 0.0722, + "step": 22079 + }, + { + "epoch": 10.264993026499303, + "grad_norm": 6.905936241149902, + "learning_rate": 9.977929997793002e-06, + "loss": 0.2359, + "step": 22080 + }, + { + "epoch": 10.265457926545793, + "grad_norm": 3.532520294189453, + "learning_rate": 9.9779289977929e-06, + "loss": 0.0708, + "step": 22081 + }, + { + "epoch": 10.265922826592282, + "grad_norm": 4.658502578735352, + "learning_rate": 9.9779279977928e-06, + "loss": 0.0536, + "step": 22082 + }, + { + "epoch": 10.266387726638772, + "grad_norm": 2.429807186126709, + "learning_rate": 9.9779269977927e-06, + "loss": 0.0302, + "step": 22083 + }, + { + "epoch": 10.266852626685262, + "grad_norm": 4.830756187438965, + "learning_rate": 9.977925997792601e-06, + "loss": 0.0626, + "step": 22084 + }, + { + "epoch": 10.267317526731752, + "grad_norm": 7.19929313659668, + "learning_rate": 9.977924997792502e-06, + "loss": 0.0528, + "step": 22085 + }, + { + "epoch": 10.267782426778243, + "grad_norm": 2.011482000350952, + "learning_rate": 9.9779239977924e-06, + "loss": 0.0176, + "step": 22086 + }, + { + "epoch": 10.268247326824733, + "grad_norm": 2.374558687210083, + "learning_rate": 9.977922997792301e-06, + "loss": 0.0211, + "step": 22087 + }, + { + "epoch": 10.268712226871223, + "grad_norm": 0.5681727528572083, + "learning_rate": 9.9779219977922e-06, + "loss": 0.0047, + "step": 22088 + }, + { + "epoch": 10.269177126917713, + "grad_norm": 5.433372497558594, + "learning_rate": 9.9779209977921e-06, + "loss": 0.062, + "step": 22089 + }, + { + "epoch": 10.269642026964203, + "grad_norm": 5.389928817749023, + "learning_rate": 9.977919997792e-06, + "loss": 0.0761, + "step": 22090 + }, + { + "epoch": 10.270106927010692, + "grad_norm": 4.0769805908203125, + "learning_rate": 9.9779189977919e-06, + "loss": 0.0762, + "step": 22091 + }, + { + "epoch": 10.270571827057182, + "grad_norm": 1.0242042541503906, + "learning_rate": 9.977917997791801e-06, + "loss": 0.0114, + "step": 22092 + }, + { + "epoch": 10.271036727103672, + "grad_norm": 3.270500898361206, + "learning_rate": 9.9779169977917e-06, + "loss": 0.0493, + "step": 22093 + }, + { + "epoch": 10.271501627150164, + "grad_norm": 9.5597505569458, + "learning_rate": 9.9779159977916e-06, + "loss": 0.2458, + "step": 22094 + }, + { + "epoch": 10.271966527196653, + "grad_norm": 8.359156608581543, + "learning_rate": 9.9779149977915e-06, + "loss": 0.0711, + "step": 22095 + }, + { + "epoch": 10.272431427243143, + "grad_norm": 2.8317031860351562, + "learning_rate": 9.9779139977914e-06, + "loss": 0.0337, + "step": 22096 + }, + { + "epoch": 10.272896327289633, + "grad_norm": 7.021770477294922, + "learning_rate": 9.977912997791301e-06, + "loss": 0.0569, + "step": 22097 + }, + { + "epoch": 10.273361227336123, + "grad_norm": 5.274242401123047, + "learning_rate": 9.977911997791202e-06, + "loss": 0.1005, + "step": 22098 + }, + { + "epoch": 10.273826127382613, + "grad_norm": 2.6893157958984375, + "learning_rate": 9.9779109977911e-06, + "loss": 0.0432, + "step": 22099 + }, + { + "epoch": 10.274291027429102, + "grad_norm": 7.771782875061035, + "learning_rate": 9.977909997791001e-06, + "loss": 0.1183, + "step": 22100 + }, + { + "epoch": 10.274755927475592, + "grad_norm": 6.666135311126709, + "learning_rate": 9.9779089977909e-06, + "loss": 0.0473, + "step": 22101 + }, + { + "epoch": 10.275220827522082, + "grad_norm": 2.6541380882263184, + "learning_rate": 9.977907997790801e-06, + "loss": 0.018, + "step": 22102 + }, + { + "epoch": 10.275685727568574, + "grad_norm": 3.050651788711548, + "learning_rate": 9.9779069977907e-06, + "loss": 0.0341, + "step": 22103 + }, + { + "epoch": 10.276150627615063, + "grad_norm": 3.061204195022583, + "learning_rate": 9.9779059977906e-06, + "loss": 0.0443, + "step": 22104 + }, + { + "epoch": 10.276615527661553, + "grad_norm": 4.0869011878967285, + "learning_rate": 9.977904997790501e-06, + "loss": 0.0417, + "step": 22105 + }, + { + "epoch": 10.277080427708043, + "grad_norm": 7.115027904510498, + "learning_rate": 9.9779039977904e-06, + "loss": 0.0825, + "step": 22106 + }, + { + "epoch": 10.277545327754533, + "grad_norm": 2.62343692779541, + "learning_rate": 9.9779029977903e-06, + "loss": 0.0256, + "step": 22107 + }, + { + "epoch": 10.278010227801023, + "grad_norm": 4.1766862869262695, + "learning_rate": 9.9779019977902e-06, + "loss": 0.0505, + "step": 22108 + }, + { + "epoch": 10.278475127847512, + "grad_norm": 5.006431579589844, + "learning_rate": 9.9779009977901e-06, + "loss": 0.0721, + "step": 22109 + }, + { + "epoch": 10.278940027894002, + "grad_norm": 3.4941182136535645, + "learning_rate": 9.977899997790001e-06, + "loss": 0.0376, + "step": 22110 + }, + { + "epoch": 10.279404927940492, + "grad_norm": 5.944061279296875, + "learning_rate": 9.977898997789902e-06, + "loss": 0.0845, + "step": 22111 + }, + { + "epoch": 10.279869827986984, + "grad_norm": 3.4011237621307373, + "learning_rate": 9.9778979977898e-06, + "loss": 0.0221, + "step": 22112 + }, + { + "epoch": 10.280334728033473, + "grad_norm": 4.665401458740234, + "learning_rate": 9.977896997789701e-06, + "loss": 0.0696, + "step": 22113 + }, + { + "epoch": 10.280799628079963, + "grad_norm": 1.7581781148910522, + "learning_rate": 9.9778959977896e-06, + "loss": 0.0192, + "step": 22114 + }, + { + "epoch": 10.281264528126453, + "grad_norm": 4.83698034286499, + "learning_rate": 9.977894997789501e-06, + "loss": 0.0698, + "step": 22115 + }, + { + "epoch": 10.281729428172943, + "grad_norm": 6.119187831878662, + "learning_rate": 9.9778939977894e-06, + "loss": 0.1182, + "step": 22116 + }, + { + "epoch": 10.282194328219433, + "grad_norm": 4.735052108764648, + "learning_rate": 9.9778929977893e-06, + "loss": 0.075, + "step": 22117 + }, + { + "epoch": 10.282659228265922, + "grad_norm": 5.31998348236084, + "learning_rate": 9.977891997789201e-06, + "loss": 0.0827, + "step": 22118 + }, + { + "epoch": 10.283124128312412, + "grad_norm": 5.909711837768555, + "learning_rate": 9.9778909977891e-06, + "loss": 0.0839, + "step": 22119 + }, + { + "epoch": 10.283589028358902, + "grad_norm": 7.069798946380615, + "learning_rate": 9.977889997789001e-06, + "loss": 0.1214, + "step": 22120 + }, + { + "epoch": 10.284053928405394, + "grad_norm": 4.024745941162109, + "learning_rate": 9.9778889977889e-06, + "loss": 0.0731, + "step": 22121 + }, + { + "epoch": 10.284518828451883, + "grad_norm": 2.8709187507629395, + "learning_rate": 9.9778879977888e-06, + "loss": 0.0372, + "step": 22122 + }, + { + "epoch": 10.284983728498373, + "grad_norm": 5.393953800201416, + "learning_rate": 9.977886997788701e-06, + "loss": 0.0751, + "step": 22123 + }, + { + "epoch": 10.285448628544863, + "grad_norm": 4.82908821105957, + "learning_rate": 9.977885997788602e-06, + "loss": 0.0665, + "step": 22124 + }, + { + "epoch": 10.285913528591353, + "grad_norm": 4.928715705871582, + "learning_rate": 9.9778849977885e-06, + "loss": 0.0323, + "step": 22125 + }, + { + "epoch": 10.286378428637843, + "grad_norm": 7.369714260101318, + "learning_rate": 9.9778839977884e-06, + "loss": 0.123, + "step": 22126 + }, + { + "epoch": 10.286843328684332, + "grad_norm": 3.8084893226623535, + "learning_rate": 9.9778829977883e-06, + "loss": 0.0373, + "step": 22127 + }, + { + "epoch": 10.287308228730822, + "grad_norm": 2.8140017986297607, + "learning_rate": 9.977881997788201e-06, + "loss": 0.0395, + "step": 22128 + }, + { + "epoch": 10.287773128777314, + "grad_norm": 5.640336036682129, + "learning_rate": 9.9778809977881e-06, + "loss": 0.0619, + "step": 22129 + }, + { + "epoch": 10.288238028823804, + "grad_norm": 6.570837497711182, + "learning_rate": 9.977879997788e-06, + "loss": 0.1857, + "step": 22130 + }, + { + "epoch": 10.288702928870293, + "grad_norm": 9.460352897644043, + "learning_rate": 9.977878997787901e-06, + "loss": 0.243, + "step": 22131 + }, + { + "epoch": 10.289167828916783, + "grad_norm": 1.4611964225769043, + "learning_rate": 9.9778779977878e-06, + "loss": 0.0084, + "step": 22132 + }, + { + "epoch": 10.289632728963273, + "grad_norm": 7.297816276550293, + "learning_rate": 9.977876997787701e-06, + "loss": 0.1297, + "step": 22133 + }, + { + "epoch": 10.290097629009763, + "grad_norm": 5.299664497375488, + "learning_rate": 9.9778759977876e-06, + "loss": 0.0977, + "step": 22134 + }, + { + "epoch": 10.290562529056253, + "grad_norm": 0.8243436217308044, + "learning_rate": 9.9778749977875e-06, + "loss": 0.0072, + "step": 22135 + }, + { + "epoch": 10.291027429102742, + "grad_norm": 5.721670627593994, + "learning_rate": 9.977873997787401e-06, + "loss": 0.0946, + "step": 22136 + }, + { + "epoch": 10.291492329149232, + "grad_norm": 3.303647518157959, + "learning_rate": 9.9778729977873e-06, + "loss": 0.0436, + "step": 22137 + }, + { + "epoch": 10.291957229195724, + "grad_norm": 12.36313533782959, + "learning_rate": 9.9778719977872e-06, + "loss": 0.0551, + "step": 22138 + }, + { + "epoch": 10.292422129242214, + "grad_norm": 3.294585704803467, + "learning_rate": 9.9778709977871e-06, + "loss": 0.0346, + "step": 22139 + }, + { + "epoch": 10.292887029288703, + "grad_norm": 4.616093158721924, + "learning_rate": 9.977869997787e-06, + "loss": 0.0408, + "step": 22140 + }, + { + "epoch": 10.293351929335193, + "grad_norm": 10.367539405822754, + "learning_rate": 9.977868997786901e-06, + "loss": 0.0982, + "step": 22141 + }, + { + "epoch": 10.293816829381683, + "grad_norm": 4.511806488037109, + "learning_rate": 9.977867997786802e-06, + "loss": 0.0281, + "step": 22142 + }, + { + "epoch": 10.294281729428173, + "grad_norm": 4.600552082061768, + "learning_rate": 9.9778669977867e-06, + "loss": 0.0615, + "step": 22143 + }, + { + "epoch": 10.294746629474663, + "grad_norm": 6.1217427253723145, + "learning_rate": 9.977865997786601e-06, + "loss": 0.097, + "step": 22144 + }, + { + "epoch": 10.295211529521152, + "grad_norm": 3.4699995517730713, + "learning_rate": 9.9778649977865e-06, + "loss": 0.0497, + "step": 22145 + }, + { + "epoch": 10.295676429567642, + "grad_norm": 5.363652229309082, + "learning_rate": 9.977863997786401e-06, + "loss": 0.1568, + "step": 22146 + }, + { + "epoch": 10.296141329614134, + "grad_norm": 3.7733523845672607, + "learning_rate": 9.9778629977863e-06, + "loss": 0.0831, + "step": 22147 + }, + { + "epoch": 10.296606229660624, + "grad_norm": 4.8544182777404785, + "learning_rate": 9.9778619977862e-06, + "loss": 0.0946, + "step": 22148 + }, + { + "epoch": 10.297071129707113, + "grad_norm": 3.3249096870422363, + "learning_rate": 9.977860997786101e-06, + "loss": 0.0279, + "step": 22149 + }, + { + "epoch": 10.297536029753603, + "grad_norm": 5.872792720794678, + "learning_rate": 9.977859997786e-06, + "loss": 0.05, + "step": 22150 + }, + { + "epoch": 10.298000929800093, + "grad_norm": 5.6356000900268555, + "learning_rate": 9.977858997785901e-06, + "loss": 0.1001, + "step": 22151 + }, + { + "epoch": 10.298465829846583, + "grad_norm": 1.6969425678253174, + "learning_rate": 9.9778579977858e-06, + "loss": 0.0165, + "step": 22152 + }, + { + "epoch": 10.298930729893073, + "grad_norm": 4.408225059509277, + "learning_rate": 9.9778569977857e-06, + "loss": 0.0194, + "step": 22153 + }, + { + "epoch": 10.299395629939562, + "grad_norm": 3.2970144748687744, + "learning_rate": 9.977855997785601e-06, + "loss": 0.0311, + "step": 22154 + }, + { + "epoch": 10.299860529986052, + "grad_norm": 2.722443103790283, + "learning_rate": 9.977854997785502e-06, + "loss": 0.0359, + "step": 22155 + }, + { + "epoch": 10.300325430032544, + "grad_norm": 2.0862371921539307, + "learning_rate": 9.9778539977854e-06, + "loss": 0.0123, + "step": 22156 + }, + { + "epoch": 10.300790330079034, + "grad_norm": 4.887209892272949, + "learning_rate": 9.977852997785301e-06, + "loss": 0.0821, + "step": 22157 + }, + { + "epoch": 10.301255230125523, + "grad_norm": 4.474747180938721, + "learning_rate": 9.9778519977852e-06, + "loss": 0.0471, + "step": 22158 + }, + { + "epoch": 10.301720130172013, + "grad_norm": 1.0834851264953613, + "learning_rate": 9.9778509977851e-06, + "loss": 0.0094, + "step": 22159 + }, + { + "epoch": 10.302185030218503, + "grad_norm": 5.015167713165283, + "learning_rate": 9.977849997785e-06, + "loss": 0.0816, + "step": 22160 + }, + { + "epoch": 10.302649930264993, + "grad_norm": 4.545331954956055, + "learning_rate": 9.9778489977849e-06, + "loss": 0.0544, + "step": 22161 + }, + { + "epoch": 10.303114830311483, + "grad_norm": 5.9624409675598145, + "learning_rate": 9.977847997784801e-06, + "loss": 0.0822, + "step": 22162 + }, + { + "epoch": 10.303579730357972, + "grad_norm": 4.796266555786133, + "learning_rate": 9.9778469977847e-06, + "loss": 0.0509, + "step": 22163 + }, + { + "epoch": 10.304044630404462, + "grad_norm": 13.01606559753418, + "learning_rate": 9.977845997784601e-06, + "loss": 0.1156, + "step": 22164 + }, + { + "epoch": 10.304509530450954, + "grad_norm": 3.151500940322876, + "learning_rate": 9.9778449977845e-06, + "loss": 0.036, + "step": 22165 + }, + { + "epoch": 10.304974430497444, + "grad_norm": 8.059282302856445, + "learning_rate": 9.9778439977844e-06, + "loss": 0.1467, + "step": 22166 + }, + { + "epoch": 10.305439330543933, + "grad_norm": 3.7418816089630127, + "learning_rate": 9.977842997784301e-06, + "loss": 0.0407, + "step": 22167 + }, + { + "epoch": 10.305904230590423, + "grad_norm": 2.0223324298858643, + "learning_rate": 9.9778419977842e-06, + "loss": 0.0216, + "step": 22168 + }, + { + "epoch": 10.306369130636913, + "grad_norm": 1.9945614337921143, + "learning_rate": 9.977840997784101e-06, + "loss": 0.0134, + "step": 22169 + }, + { + "epoch": 10.306834030683403, + "grad_norm": 1.3759690523147583, + "learning_rate": 9.977839997784e-06, + "loss": 0.0111, + "step": 22170 + }, + { + "epoch": 10.307298930729893, + "grad_norm": 2.8008763790130615, + "learning_rate": 9.9778389977839e-06, + "loss": 0.0334, + "step": 22171 + }, + { + "epoch": 10.307763830776382, + "grad_norm": 5.738386154174805, + "learning_rate": 9.9778379977838e-06, + "loss": 0.0909, + "step": 22172 + }, + { + "epoch": 10.308228730822874, + "grad_norm": 2.9603028297424316, + "learning_rate": 9.9778369977837e-06, + "loss": 0.0213, + "step": 22173 + }, + { + "epoch": 10.308693630869364, + "grad_norm": 3.8818163871765137, + "learning_rate": 9.9778359977836e-06, + "loss": 0.0644, + "step": 22174 + }, + { + "epoch": 10.309158530915854, + "grad_norm": 7.025509834289551, + "learning_rate": 9.977834997783501e-06, + "loss": 0.0934, + "step": 22175 + }, + { + "epoch": 10.309623430962343, + "grad_norm": 3.800407886505127, + "learning_rate": 9.9778339977834e-06, + "loss": 0.0335, + "step": 22176 + }, + { + "epoch": 10.310088331008833, + "grad_norm": 7.6055097579956055, + "learning_rate": 9.977832997783301e-06, + "loss": 0.1499, + "step": 22177 + }, + { + "epoch": 10.310553231055323, + "grad_norm": 10.600079536437988, + "learning_rate": 9.9778319977832e-06, + "loss": 0.1553, + "step": 22178 + }, + { + "epoch": 10.311018131101813, + "grad_norm": 4.706234931945801, + "learning_rate": 9.9778309977831e-06, + "loss": 0.0636, + "step": 22179 + }, + { + "epoch": 10.311483031148303, + "grad_norm": 4.235970973968506, + "learning_rate": 9.977829997783001e-06, + "loss": 0.0573, + "step": 22180 + }, + { + "epoch": 10.311947931194792, + "grad_norm": 8.169979095458984, + "learning_rate": 9.9778289977829e-06, + "loss": 0.1247, + "step": 22181 + }, + { + "epoch": 10.312412831241284, + "grad_norm": 5.943710803985596, + "learning_rate": 9.977827997782801e-06, + "loss": 0.0883, + "step": 22182 + }, + { + "epoch": 10.312877731287774, + "grad_norm": 4.787886619567871, + "learning_rate": 9.9778269977827e-06, + "loss": 0.0816, + "step": 22183 + }, + { + "epoch": 10.313342631334264, + "grad_norm": 10.999831199645996, + "learning_rate": 9.9778259977826e-06, + "loss": 0.11, + "step": 22184 + }, + { + "epoch": 10.313807531380753, + "grad_norm": 4.858489513397217, + "learning_rate": 9.9778249977825e-06, + "loss": 0.0554, + "step": 22185 + }, + { + "epoch": 10.314272431427243, + "grad_norm": 6.71553373336792, + "learning_rate": 9.9778239977824e-06, + "loss": 0.1206, + "step": 22186 + }, + { + "epoch": 10.314737331473733, + "grad_norm": 5.7695159912109375, + "learning_rate": 9.977822997782301e-06, + "loss": 0.1337, + "step": 22187 + }, + { + "epoch": 10.315202231520223, + "grad_norm": 1.0548254251480103, + "learning_rate": 9.977821997782202e-06, + "loss": 0.0084, + "step": 22188 + }, + { + "epoch": 10.315667131566713, + "grad_norm": 3.082374095916748, + "learning_rate": 9.9778209977821e-06, + "loss": 0.0395, + "step": 22189 + }, + { + "epoch": 10.316132031613202, + "grad_norm": 4.406999111175537, + "learning_rate": 9.977819997782e-06, + "loss": 0.0354, + "step": 22190 + }, + { + "epoch": 10.316596931659694, + "grad_norm": 8.848637580871582, + "learning_rate": 9.9778189977819e-06, + "loss": 0.0672, + "step": 22191 + }, + { + "epoch": 10.317061831706184, + "grad_norm": 4.180872440338135, + "learning_rate": 9.9778179977818e-06, + "loss": 0.0666, + "step": 22192 + }, + { + "epoch": 10.317526731752674, + "grad_norm": 4.418388843536377, + "learning_rate": 9.977816997781701e-06, + "loss": 0.0443, + "step": 22193 + }, + { + "epoch": 10.317991631799163, + "grad_norm": 3.1011877059936523, + "learning_rate": 9.9778159977816e-06, + "loss": 0.0356, + "step": 22194 + }, + { + "epoch": 10.318456531845653, + "grad_norm": 4.174087047576904, + "learning_rate": 9.977814997781501e-06, + "loss": 0.048, + "step": 22195 + }, + { + "epoch": 10.318921431892143, + "grad_norm": 6.303396701812744, + "learning_rate": 9.9778139977814e-06, + "loss": 0.1366, + "step": 22196 + }, + { + "epoch": 10.319386331938633, + "grad_norm": 4.257532596588135, + "learning_rate": 9.9778129977813e-06, + "loss": 0.0486, + "step": 22197 + }, + { + "epoch": 10.319851231985123, + "grad_norm": 5.870065689086914, + "learning_rate": 9.9778119977812e-06, + "loss": 0.1533, + "step": 22198 + }, + { + "epoch": 10.320316132031612, + "grad_norm": 4.198861598968506, + "learning_rate": 9.9778109977811e-06, + "loss": 0.0537, + "step": 22199 + }, + { + "epoch": 10.320781032078104, + "grad_norm": 4.828550815582275, + "learning_rate": 9.977809997781001e-06, + "loss": 0.0826, + "step": 22200 + }, + { + "epoch": 10.321245932124594, + "grad_norm": 5.558284759521484, + "learning_rate": 9.9778089977809e-06, + "loss": 0.0352, + "step": 22201 + }, + { + "epoch": 10.321710832171084, + "grad_norm": 7.123796463012695, + "learning_rate": 9.9778079977808e-06, + "loss": 0.1359, + "step": 22202 + }, + { + "epoch": 10.322175732217573, + "grad_norm": 5.532482147216797, + "learning_rate": 9.9778069977807e-06, + "loss": 0.0679, + "step": 22203 + }, + { + "epoch": 10.322640632264063, + "grad_norm": 5.636883735656738, + "learning_rate": 9.9778059977806e-06, + "loss": 0.0531, + "step": 22204 + }, + { + "epoch": 10.323105532310553, + "grad_norm": 0.8883512020111084, + "learning_rate": 9.977804997780501e-06, + "loss": 0.0051, + "step": 22205 + }, + { + "epoch": 10.323570432357043, + "grad_norm": 2.2861592769622803, + "learning_rate": 9.977803997780402e-06, + "loss": 0.0364, + "step": 22206 + }, + { + "epoch": 10.324035332403533, + "grad_norm": 7.666858673095703, + "learning_rate": 9.9778029977803e-06, + "loss": 0.0644, + "step": 22207 + }, + { + "epoch": 10.324500232450024, + "grad_norm": 2.9350180625915527, + "learning_rate": 9.977801997780201e-06, + "loss": 0.0156, + "step": 22208 + }, + { + "epoch": 10.324965132496514, + "grad_norm": 4.400208950042725, + "learning_rate": 9.9778009977801e-06, + "loss": 0.0678, + "step": 22209 + }, + { + "epoch": 10.325430032543004, + "grad_norm": 1.9046051502227783, + "learning_rate": 9.97779999778e-06, + "loss": 0.0177, + "step": 22210 + }, + { + "epoch": 10.325894932589494, + "grad_norm": 6.523828029632568, + "learning_rate": 9.977798997779901e-06, + "loss": 0.1315, + "step": 22211 + }, + { + "epoch": 10.326359832635983, + "grad_norm": 3.1257362365722656, + "learning_rate": 9.9777979977798e-06, + "loss": 0.0219, + "step": 22212 + }, + { + "epoch": 10.326824732682473, + "grad_norm": 8.78184700012207, + "learning_rate": 9.977796997779701e-06, + "loss": 0.0646, + "step": 22213 + }, + { + "epoch": 10.327289632728963, + "grad_norm": 5.508233547210693, + "learning_rate": 9.9777959977796e-06, + "loss": 0.0825, + "step": 22214 + }, + { + "epoch": 10.327754532775453, + "grad_norm": 9.04448127746582, + "learning_rate": 9.9777949977795e-06, + "loss": 0.1166, + "step": 22215 + }, + { + "epoch": 10.328219432821943, + "grad_norm": 8.652403831481934, + "learning_rate": 9.9777939977794e-06, + "loss": 0.1115, + "step": 22216 + }, + { + "epoch": 10.328684332868434, + "grad_norm": 9.104277610778809, + "learning_rate": 9.9777929977793e-06, + "loss": 0.0745, + "step": 22217 + }, + { + "epoch": 10.329149232914924, + "grad_norm": 4.530177116394043, + "learning_rate": 9.977791997779201e-06, + "loss": 0.0908, + "step": 22218 + }, + { + "epoch": 10.329614132961414, + "grad_norm": 7.245424270629883, + "learning_rate": 9.977790997779102e-06, + "loss": 0.1073, + "step": 22219 + }, + { + "epoch": 10.330079033007904, + "grad_norm": 5.958315849304199, + "learning_rate": 9.977789997779e-06, + "loss": 0.1119, + "step": 22220 + }, + { + "epoch": 10.330543933054393, + "grad_norm": 1.4194740056991577, + "learning_rate": 9.977788997778901e-06, + "loss": 0.0106, + "step": 22221 + }, + { + "epoch": 10.331008833100883, + "grad_norm": 4.0165910720825195, + "learning_rate": 9.9777879977788e-06, + "loss": 0.0218, + "step": 22222 + }, + { + "epoch": 10.331473733147373, + "grad_norm": 7.555340766906738, + "learning_rate": 9.9777869977787e-06, + "loss": 0.1378, + "step": 22223 + }, + { + "epoch": 10.331938633193863, + "grad_norm": 6.661923885345459, + "learning_rate": 9.977785997778601e-06, + "loss": 0.1538, + "step": 22224 + }, + { + "epoch": 10.332403533240353, + "grad_norm": 2.093324661254883, + "learning_rate": 9.9777849977785e-06, + "loss": 0.028, + "step": 22225 + }, + { + "epoch": 10.332868433286844, + "grad_norm": 4.093616008758545, + "learning_rate": 9.977783997778401e-06, + "loss": 0.0315, + "step": 22226 + }, + { + "epoch": 10.333333333333334, + "grad_norm": 3.835380792617798, + "learning_rate": 9.9777829977783e-06, + "loss": 0.0347, + "step": 22227 + }, + { + "epoch": 10.333798233379824, + "grad_norm": 5.524628162384033, + "learning_rate": 9.9777819977782e-06, + "loss": 0.0868, + "step": 22228 + }, + { + "epoch": 10.334263133426314, + "grad_norm": 3.327373504638672, + "learning_rate": 9.9777809977781e-06, + "loss": 0.0513, + "step": 22229 + }, + { + "epoch": 10.334728033472803, + "grad_norm": 4.8235979080200195, + "learning_rate": 9.977779997778e-06, + "loss": 0.1104, + "step": 22230 + }, + { + "epoch": 10.335192933519293, + "grad_norm": 3.690889358520508, + "learning_rate": 9.977778997777901e-06, + "loss": 0.0534, + "step": 22231 + }, + { + "epoch": 10.335657833565783, + "grad_norm": 7.915682792663574, + "learning_rate": 9.977777997777802e-06, + "loss": 0.1698, + "step": 22232 + }, + { + "epoch": 10.336122733612273, + "grad_norm": 4.738674640655518, + "learning_rate": 9.9777769977777e-06, + "loss": 0.1134, + "step": 22233 + }, + { + "epoch": 10.336587633658763, + "grad_norm": 4.996101379394531, + "learning_rate": 9.9777759977776e-06, + "loss": 0.0684, + "step": 22234 + }, + { + "epoch": 10.337052533705254, + "grad_norm": 5.718834400177002, + "learning_rate": 9.9777749977775e-06, + "loss": 0.079, + "step": 22235 + }, + { + "epoch": 10.337517433751744, + "grad_norm": 4.948658466339111, + "learning_rate": 9.977773997777401e-06, + "loss": 0.0811, + "step": 22236 + }, + { + "epoch": 10.337982333798234, + "grad_norm": 3.466773748397827, + "learning_rate": 9.977772997777302e-06, + "loss": 0.0453, + "step": 22237 + }, + { + "epoch": 10.338447233844724, + "grad_norm": 5.592079162597656, + "learning_rate": 9.9777719977772e-06, + "loss": 0.1384, + "step": 22238 + }, + { + "epoch": 10.338912133891213, + "grad_norm": 2.842076539993286, + "learning_rate": 9.977770997777101e-06, + "loss": 0.0251, + "step": 22239 + }, + { + "epoch": 10.339377033937703, + "grad_norm": 1.2417610883712769, + "learning_rate": 9.977769997777e-06, + "loss": 0.0113, + "step": 22240 + }, + { + "epoch": 10.339841933984193, + "grad_norm": 6.175275802612305, + "learning_rate": 9.9777689977769e-06, + "loss": 0.0826, + "step": 22241 + }, + { + "epoch": 10.340306834030683, + "grad_norm": 3.936037540435791, + "learning_rate": 9.9777679977768e-06, + "loss": 0.0559, + "step": 22242 + }, + { + "epoch": 10.340771734077173, + "grad_norm": 1.793667197227478, + "learning_rate": 9.9777669977767e-06, + "loss": 0.0205, + "step": 22243 + }, + { + "epoch": 10.341236634123664, + "grad_norm": 2.9002673625946045, + "learning_rate": 9.977765997776601e-06, + "loss": 0.0269, + "step": 22244 + }, + { + "epoch": 10.341701534170154, + "grad_norm": 3.818676710128784, + "learning_rate": 9.9777649977765e-06, + "loss": 0.0231, + "step": 22245 + }, + { + "epoch": 10.342166434216644, + "grad_norm": 5.392507076263428, + "learning_rate": 9.9777639977764e-06, + "loss": 0.1, + "step": 22246 + }, + { + "epoch": 10.342631334263134, + "grad_norm": 10.804913520812988, + "learning_rate": 9.9777629977763e-06, + "loss": 0.0825, + "step": 22247 + }, + { + "epoch": 10.343096234309623, + "grad_norm": 9.158380508422852, + "learning_rate": 9.9777619977762e-06, + "loss": 0.0943, + "step": 22248 + }, + { + "epoch": 10.343561134356113, + "grad_norm": 6.3036346435546875, + "learning_rate": 9.977760997776101e-06, + "loss": 0.1636, + "step": 22249 + }, + { + "epoch": 10.344026034402603, + "grad_norm": 2.9605190753936768, + "learning_rate": 9.977759997776002e-06, + "loss": 0.0318, + "step": 22250 + }, + { + "epoch": 10.344490934449093, + "grad_norm": 5.023359775543213, + "learning_rate": 9.9777589977759e-06, + "loss": 0.0334, + "step": 22251 + }, + { + "epoch": 10.344955834495583, + "grad_norm": 1.5495473146438599, + "learning_rate": 9.977757997775801e-06, + "loss": 0.009, + "step": 22252 + }, + { + "epoch": 10.345420734542074, + "grad_norm": 5.515279293060303, + "learning_rate": 9.9777569977757e-06, + "loss": 0.0528, + "step": 22253 + }, + { + "epoch": 10.345885634588564, + "grad_norm": 5.756155014038086, + "learning_rate": 9.977755997775601e-06, + "loss": 0.0865, + "step": 22254 + }, + { + "epoch": 10.346350534635054, + "grad_norm": 3.799379348754883, + "learning_rate": 9.9777549977755e-06, + "loss": 0.0475, + "step": 22255 + }, + { + "epoch": 10.346815434681544, + "grad_norm": 5.372202396392822, + "learning_rate": 9.9777539977754e-06, + "loss": 0.061, + "step": 22256 + }, + { + "epoch": 10.347280334728033, + "grad_norm": 7.302497863769531, + "learning_rate": 9.977752997775301e-06, + "loss": 0.2004, + "step": 22257 + }, + { + "epoch": 10.347745234774523, + "grad_norm": 5.504266262054443, + "learning_rate": 9.9777519977752e-06, + "loss": 0.0839, + "step": 22258 + }, + { + "epoch": 10.348210134821013, + "grad_norm": 1.0993800163269043, + "learning_rate": 9.9777509977751e-06, + "loss": 0.0069, + "step": 22259 + }, + { + "epoch": 10.348675034867503, + "grad_norm": 8.821496963500977, + "learning_rate": 9.977749997775e-06, + "loss": 0.1527, + "step": 22260 + }, + { + "epoch": 10.349139934913994, + "grad_norm": 4.63064432144165, + "learning_rate": 9.9777489977749e-06, + "loss": 0.0752, + "step": 22261 + }, + { + "epoch": 10.349604834960484, + "grad_norm": 4.44512939453125, + "learning_rate": 9.977747997774801e-06, + "loss": 0.0692, + "step": 22262 + }, + { + "epoch": 10.350069735006974, + "grad_norm": 4.805850505828857, + "learning_rate": 9.977746997774702e-06, + "loss": 0.0625, + "step": 22263 + }, + { + "epoch": 10.350534635053464, + "grad_norm": 6.0226054191589355, + "learning_rate": 9.9777459977746e-06, + "loss": 0.0572, + "step": 22264 + }, + { + "epoch": 10.350999535099954, + "grad_norm": 3.9213480949401855, + "learning_rate": 9.977744997774501e-06, + "loss": 0.034, + "step": 22265 + }, + { + "epoch": 10.351464435146443, + "grad_norm": 4.527919292449951, + "learning_rate": 9.9777439977744e-06, + "loss": 0.0551, + "step": 22266 + }, + { + "epoch": 10.351929335192933, + "grad_norm": 8.30266284942627, + "learning_rate": 9.9777429977743e-06, + "loss": 0.1999, + "step": 22267 + }, + { + "epoch": 10.352394235239423, + "grad_norm": 4.340137004852295, + "learning_rate": 9.9777419977742e-06, + "loss": 0.0325, + "step": 22268 + }, + { + "epoch": 10.352859135285913, + "grad_norm": 3.962754487991333, + "learning_rate": 9.9777409977741e-06, + "loss": 0.0786, + "step": 22269 + }, + { + "epoch": 10.353324035332404, + "grad_norm": 12.24347972869873, + "learning_rate": 9.977739997774001e-06, + "loss": 0.3009, + "step": 22270 + }, + { + "epoch": 10.353788935378894, + "grad_norm": 4.632816791534424, + "learning_rate": 9.9777389977739e-06, + "loss": 0.0852, + "step": 22271 + }, + { + "epoch": 10.354253835425384, + "grad_norm": 5.691690921783447, + "learning_rate": 9.977737997773801e-06, + "loss": 0.0588, + "step": 22272 + }, + { + "epoch": 10.354718735471874, + "grad_norm": 5.448801040649414, + "learning_rate": 9.9777369977737e-06, + "loss": 0.1057, + "step": 22273 + }, + { + "epoch": 10.355183635518364, + "grad_norm": 9.221336364746094, + "learning_rate": 9.9777359977736e-06, + "loss": 0.1596, + "step": 22274 + }, + { + "epoch": 10.355648535564853, + "grad_norm": 5.008521556854248, + "learning_rate": 9.977734997773501e-06, + "loss": 0.0329, + "step": 22275 + }, + { + "epoch": 10.356113435611343, + "grad_norm": 1.0885732173919678, + "learning_rate": 9.977733997773402e-06, + "loss": 0.0132, + "step": 22276 + }, + { + "epoch": 10.356578335657833, + "grad_norm": 4.633129596710205, + "learning_rate": 9.9777329977733e-06, + "loss": 0.0652, + "step": 22277 + }, + { + "epoch": 10.357043235704323, + "grad_norm": 2.3486664295196533, + "learning_rate": 9.9777319977732e-06, + "loss": 0.021, + "step": 22278 + }, + { + "epoch": 10.357508135750814, + "grad_norm": 5.249231338500977, + "learning_rate": 9.9777309977731e-06, + "loss": 0.0715, + "step": 22279 + }, + { + "epoch": 10.357973035797304, + "grad_norm": 5.475130081176758, + "learning_rate": 9.977729997773001e-06, + "loss": 0.0866, + "step": 22280 + }, + { + "epoch": 10.358437935843794, + "grad_norm": 6.211458683013916, + "learning_rate": 9.9777289977729e-06, + "loss": 0.1452, + "step": 22281 + }, + { + "epoch": 10.358902835890284, + "grad_norm": 3.040444850921631, + "learning_rate": 9.9777279977728e-06, + "loss": 0.0251, + "step": 22282 + }, + { + "epoch": 10.359367735936774, + "grad_norm": 8.723516464233398, + "learning_rate": 9.977726997772701e-06, + "loss": 0.1721, + "step": 22283 + }, + { + "epoch": 10.359832635983263, + "grad_norm": 5.8330607414245605, + "learning_rate": 9.9777259977726e-06, + "loss": 0.1344, + "step": 22284 + }, + { + "epoch": 10.360297536029753, + "grad_norm": 2.4056410789489746, + "learning_rate": 9.977724997772501e-06, + "loss": 0.0219, + "step": 22285 + }, + { + "epoch": 10.360762436076243, + "grad_norm": 3.949376344680786, + "learning_rate": 9.9777239977724e-06, + "loss": 0.0403, + "step": 22286 + }, + { + "epoch": 10.361227336122734, + "grad_norm": 3.848850965499878, + "learning_rate": 9.9777229977723e-06, + "loss": 0.0652, + "step": 22287 + }, + { + "epoch": 10.361692236169224, + "grad_norm": 8.14154052734375, + "learning_rate": 9.977721997772201e-06, + "loss": 0.1569, + "step": 22288 + }, + { + "epoch": 10.362157136215714, + "grad_norm": 2.301568031311035, + "learning_rate": 9.9777209977721e-06, + "loss": 0.0216, + "step": 22289 + }, + { + "epoch": 10.362622036262204, + "grad_norm": 3.598651647567749, + "learning_rate": 9.977719997772001e-06, + "loss": 0.048, + "step": 22290 + }, + { + "epoch": 10.363086936308694, + "grad_norm": 2.0579798221588135, + "learning_rate": 9.9777189977719e-06, + "loss": 0.0248, + "step": 22291 + }, + { + "epoch": 10.363551836355184, + "grad_norm": 4.696677207946777, + "learning_rate": 9.9777179977718e-06, + "loss": 0.0412, + "step": 22292 + }, + { + "epoch": 10.364016736401673, + "grad_norm": 9.261388778686523, + "learning_rate": 9.977716997771701e-06, + "loss": 0.1619, + "step": 22293 + }, + { + "epoch": 10.364481636448163, + "grad_norm": 6.273637294769287, + "learning_rate": 9.9777159977716e-06, + "loss": 0.184, + "step": 22294 + }, + { + "epoch": 10.364946536494653, + "grad_norm": 2.9179141521453857, + "learning_rate": 9.9777149977715e-06, + "loss": 0.0468, + "step": 22295 + }, + { + "epoch": 10.365411436541144, + "grad_norm": 5.166527271270752, + "learning_rate": 9.977713997771401e-06, + "loss": 0.0813, + "step": 22296 + }, + { + "epoch": 10.365876336587634, + "grad_norm": 5.7157158851623535, + "learning_rate": 9.9777129977713e-06, + "loss": 0.0819, + "step": 22297 + }, + { + "epoch": 10.366341236634124, + "grad_norm": 3.3405344486236572, + "learning_rate": 9.977711997771201e-06, + "loss": 0.0302, + "step": 22298 + }, + { + "epoch": 10.366806136680614, + "grad_norm": 4.629886627197266, + "learning_rate": 9.9777109977711e-06, + "loss": 0.086, + "step": 22299 + }, + { + "epoch": 10.367271036727104, + "grad_norm": 4.505906581878662, + "learning_rate": 9.977709997771e-06, + "loss": 0.0689, + "step": 22300 + }, + { + "epoch": 10.367735936773594, + "grad_norm": 2.443352460861206, + "learning_rate": 9.977708997770901e-06, + "loss": 0.0304, + "step": 22301 + }, + { + "epoch": 10.368200836820083, + "grad_norm": 3.9055705070495605, + "learning_rate": 9.9777079977708e-06, + "loss": 0.0902, + "step": 22302 + }, + { + "epoch": 10.368665736866573, + "grad_norm": 4.538833141326904, + "learning_rate": 9.977706997770701e-06, + "loss": 0.048, + "step": 22303 + }, + { + "epoch": 10.369130636913063, + "grad_norm": 1.0907790660858154, + "learning_rate": 9.9777059977706e-06, + "loss": 0.0115, + "step": 22304 + }, + { + "epoch": 10.369595536959554, + "grad_norm": 2.507831335067749, + "learning_rate": 9.9777049977705e-06, + "loss": 0.0303, + "step": 22305 + }, + { + "epoch": 10.370060437006044, + "grad_norm": 5.225559711456299, + "learning_rate": 9.977703997770401e-06, + "loss": 0.1019, + "step": 22306 + }, + { + "epoch": 10.370525337052534, + "grad_norm": 8.142966270446777, + "learning_rate": 9.977702997770302e-06, + "loss": 0.1382, + "step": 22307 + }, + { + "epoch": 10.370990237099024, + "grad_norm": 6.141227722167969, + "learning_rate": 9.9777019977702e-06, + "loss": 0.0861, + "step": 22308 + }, + { + "epoch": 10.371455137145514, + "grad_norm": 1.0411134958267212, + "learning_rate": 9.9777009977701e-06, + "loss": 0.0075, + "step": 22309 + }, + { + "epoch": 10.371920037192004, + "grad_norm": 4.365100860595703, + "learning_rate": 9.97769999777e-06, + "loss": 0.0543, + "step": 22310 + }, + { + "epoch": 10.372384937238493, + "grad_norm": 3.7869107723236084, + "learning_rate": 9.9776989977699e-06, + "loss": 0.0514, + "step": 22311 + }, + { + "epoch": 10.372849837284983, + "grad_norm": 6.832071781158447, + "learning_rate": 9.9776979977698e-06, + "loss": 0.1168, + "step": 22312 + }, + { + "epoch": 10.373314737331473, + "grad_norm": 3.5145862102508545, + "learning_rate": 9.9776969977697e-06, + "loss": 0.0532, + "step": 22313 + }, + { + "epoch": 10.373779637377964, + "grad_norm": 7.068116664886475, + "learning_rate": 9.977695997769601e-06, + "loss": 0.1742, + "step": 22314 + }, + { + "epoch": 10.374244537424454, + "grad_norm": 5.981443405151367, + "learning_rate": 9.9776949977695e-06, + "loss": 0.1091, + "step": 22315 + }, + { + "epoch": 10.374709437470944, + "grad_norm": 6.267818927764893, + "learning_rate": 9.977693997769401e-06, + "loss": 0.0549, + "step": 22316 + }, + { + "epoch": 10.375174337517434, + "grad_norm": 2.4762749671936035, + "learning_rate": 9.9776929977693e-06, + "loss": 0.0317, + "step": 22317 + }, + { + "epoch": 10.375639237563924, + "grad_norm": 1.8181763887405396, + "learning_rate": 9.9776919977692e-06, + "loss": 0.0175, + "step": 22318 + }, + { + "epoch": 10.376104137610414, + "grad_norm": 2.8551340103149414, + "learning_rate": 9.977690997769101e-06, + "loss": 0.0259, + "step": 22319 + }, + { + "epoch": 10.376569037656903, + "grad_norm": 4.855828762054443, + "learning_rate": 9.977689997769e-06, + "loss": 0.0725, + "step": 22320 + }, + { + "epoch": 10.377033937703393, + "grad_norm": 3.4246230125427246, + "learning_rate": 9.977688997768901e-06, + "loss": 0.0304, + "step": 22321 + }, + { + "epoch": 10.377498837749883, + "grad_norm": 5.584584712982178, + "learning_rate": 9.9776879977688e-06, + "loss": 0.1212, + "step": 22322 + }, + { + "epoch": 10.377963737796374, + "grad_norm": 6.649543285369873, + "learning_rate": 9.9776869977687e-06, + "loss": 0.0453, + "step": 22323 + }, + { + "epoch": 10.378428637842864, + "grad_norm": 4.290881633758545, + "learning_rate": 9.9776859977686e-06, + "loss": 0.0583, + "step": 22324 + }, + { + "epoch": 10.378893537889354, + "grad_norm": 3.3351757526397705, + "learning_rate": 9.9776849977685e-06, + "loss": 0.0287, + "step": 22325 + }, + { + "epoch": 10.379358437935844, + "grad_norm": 7.4203267097473145, + "learning_rate": 9.9776839977684e-06, + "loss": 0.1152, + "step": 22326 + }, + { + "epoch": 10.379823337982334, + "grad_norm": 5.408560276031494, + "learning_rate": 9.977682997768301e-06, + "loss": 0.1023, + "step": 22327 + }, + { + "epoch": 10.380288238028823, + "grad_norm": 4.116042137145996, + "learning_rate": 9.9776819977682e-06, + "loss": 0.0565, + "step": 22328 + }, + { + "epoch": 10.380753138075313, + "grad_norm": 1.8299219608306885, + "learning_rate": 9.977680997768101e-06, + "loss": 0.0267, + "step": 22329 + }, + { + "epoch": 10.381218038121803, + "grad_norm": 2.4401044845581055, + "learning_rate": 9.977679997768e-06, + "loss": 0.0213, + "step": 22330 + }, + { + "epoch": 10.381682938168293, + "grad_norm": 5.096356391906738, + "learning_rate": 9.9776789977679e-06, + "loss": 0.0419, + "step": 22331 + }, + { + "epoch": 10.382147838214784, + "grad_norm": 6.4741129875183105, + "learning_rate": 9.977677997767801e-06, + "loss": 0.0788, + "step": 22332 + }, + { + "epoch": 10.382612738261274, + "grad_norm": 6.69590950012207, + "learning_rate": 9.9776769977677e-06, + "loss": 0.0412, + "step": 22333 + }, + { + "epoch": 10.383077638307764, + "grad_norm": 5.557203769683838, + "learning_rate": 9.977675997767601e-06, + "loss": 0.1469, + "step": 22334 + }, + { + "epoch": 10.383542538354254, + "grad_norm": 4.365315914154053, + "learning_rate": 9.9776749977675e-06, + "loss": 0.038, + "step": 22335 + }, + { + "epoch": 10.384007438400744, + "grad_norm": 6.596358776092529, + "learning_rate": 9.9776739977674e-06, + "loss": 0.0924, + "step": 22336 + }, + { + "epoch": 10.384472338447233, + "grad_norm": 3.822293996810913, + "learning_rate": 9.9776729977673e-06, + "loss": 0.0487, + "step": 22337 + }, + { + "epoch": 10.384937238493723, + "grad_norm": 4.5684494972229, + "learning_rate": 9.9776719977672e-06, + "loss": 0.0712, + "step": 22338 + }, + { + "epoch": 10.385402138540213, + "grad_norm": 1.7598603963851929, + "learning_rate": 9.977670997767101e-06, + "loss": 0.0198, + "step": 22339 + }, + { + "epoch": 10.385867038586705, + "grad_norm": 5.757593154907227, + "learning_rate": 9.977669997767002e-06, + "loss": 0.1347, + "step": 22340 + }, + { + "epoch": 10.386331938633194, + "grad_norm": 5.299623012542725, + "learning_rate": 9.9776689977669e-06, + "loss": 0.0687, + "step": 22341 + }, + { + "epoch": 10.386796838679684, + "grad_norm": 4.887171268463135, + "learning_rate": 9.9776679977668e-06, + "loss": 0.0636, + "step": 22342 + }, + { + "epoch": 10.387261738726174, + "grad_norm": 4.003667831420898, + "learning_rate": 9.9776669977667e-06, + "loss": 0.0886, + "step": 22343 + }, + { + "epoch": 10.387726638772664, + "grad_norm": 4.189523696899414, + "learning_rate": 9.9776659977666e-06, + "loss": 0.0646, + "step": 22344 + }, + { + "epoch": 10.388191538819154, + "grad_norm": 4.575928211212158, + "learning_rate": 9.977664997766501e-06, + "loss": 0.0622, + "step": 22345 + }, + { + "epoch": 10.388656438865643, + "grad_norm": 2.190544843673706, + "learning_rate": 9.9776639977664e-06, + "loss": 0.0188, + "step": 22346 + }, + { + "epoch": 10.389121338912133, + "grad_norm": 6.468393325805664, + "learning_rate": 9.977662997766301e-06, + "loss": 0.18, + "step": 22347 + }, + { + "epoch": 10.389586238958623, + "grad_norm": 4.68487024307251, + "learning_rate": 9.9776619977662e-06, + "loss": 0.0429, + "step": 22348 + }, + { + "epoch": 10.390051139005115, + "grad_norm": 6.99391508102417, + "learning_rate": 9.9776609977661e-06, + "loss": 0.0503, + "step": 22349 + }, + { + "epoch": 10.390516039051604, + "grad_norm": 4.6664042472839355, + "learning_rate": 9.977659997766e-06, + "loss": 0.0522, + "step": 22350 + }, + { + "epoch": 10.390980939098094, + "grad_norm": 5.296655178070068, + "learning_rate": 9.9776589977659e-06, + "loss": 0.0546, + "step": 22351 + }, + { + "epoch": 10.391445839144584, + "grad_norm": 3.734527826309204, + "learning_rate": 9.977657997765801e-06, + "loss": 0.0533, + "step": 22352 + }, + { + "epoch": 10.391910739191074, + "grad_norm": 3.922229290008545, + "learning_rate": 9.9776569977657e-06, + "loss": 0.0722, + "step": 22353 + }, + { + "epoch": 10.392375639237564, + "grad_norm": 4.260704040527344, + "learning_rate": 9.9776559977656e-06, + "loss": 0.0421, + "step": 22354 + }, + { + "epoch": 10.392840539284053, + "grad_norm": 2.266775608062744, + "learning_rate": 9.9776549977655e-06, + "loss": 0.023, + "step": 22355 + }, + { + "epoch": 10.393305439330543, + "grad_norm": 2.6120219230651855, + "learning_rate": 9.9776539977654e-06, + "loss": 0.0348, + "step": 22356 + }, + { + "epoch": 10.393770339377033, + "grad_norm": 2.931595802307129, + "learning_rate": 9.977652997765301e-06, + "loss": 0.0385, + "step": 22357 + }, + { + "epoch": 10.394235239423525, + "grad_norm": 6.7314324378967285, + "learning_rate": 9.977651997765202e-06, + "loss": 0.1437, + "step": 22358 + }, + { + "epoch": 10.394700139470014, + "grad_norm": 3.7769501209259033, + "learning_rate": 9.9776509977651e-06, + "loss": 0.0487, + "step": 22359 + }, + { + "epoch": 10.395165039516504, + "grad_norm": 2.6649580001831055, + "learning_rate": 9.977649997765001e-06, + "loss": 0.0213, + "step": 22360 + }, + { + "epoch": 10.395629939562994, + "grad_norm": 5.195986270904541, + "learning_rate": 9.9776489977649e-06, + "loss": 0.104, + "step": 22361 + }, + { + "epoch": 10.396094839609484, + "grad_norm": 8.613069534301758, + "learning_rate": 9.9776479977648e-06, + "loss": 0.123, + "step": 22362 + }, + { + "epoch": 10.396559739655974, + "grad_norm": 7.7784576416015625, + "learning_rate": 9.9776469977647e-06, + "loss": 0.1923, + "step": 22363 + }, + { + "epoch": 10.397024639702463, + "grad_norm": 5.843015670776367, + "learning_rate": 9.9776459977646e-06, + "loss": 0.0536, + "step": 22364 + }, + { + "epoch": 10.397489539748953, + "grad_norm": 7.369753837585449, + "learning_rate": 9.977644997764501e-06, + "loss": 0.1939, + "step": 22365 + }, + { + "epoch": 10.397954439795445, + "grad_norm": 7.432555675506592, + "learning_rate": 9.9776439977644e-06, + "loss": 0.076, + "step": 22366 + }, + { + "epoch": 10.398419339841935, + "grad_norm": 2.3493869304656982, + "learning_rate": 9.9776429977643e-06, + "loss": 0.0237, + "step": 22367 + }, + { + "epoch": 10.398884239888424, + "grad_norm": 7.742487907409668, + "learning_rate": 9.9776419977642e-06, + "loss": 0.1565, + "step": 22368 + }, + { + "epoch": 10.399349139934914, + "grad_norm": 3.0109682083129883, + "learning_rate": 9.9776409977641e-06, + "loss": 0.0259, + "step": 22369 + }, + { + "epoch": 10.399814039981404, + "grad_norm": 3.597538709640503, + "learning_rate": 9.977639997764001e-06, + "loss": 0.0433, + "step": 22370 + }, + { + "epoch": 10.400278940027894, + "grad_norm": 4.674805164337158, + "learning_rate": 9.977638997763902e-06, + "loss": 0.0491, + "step": 22371 + }, + { + "epoch": 10.400743840074384, + "grad_norm": 5.283912658691406, + "learning_rate": 9.9776379977638e-06, + "loss": 0.0674, + "step": 22372 + }, + { + "epoch": 10.401208740120873, + "grad_norm": 6.774599075317383, + "learning_rate": 9.977636997763701e-06, + "loss": 0.127, + "step": 22373 + }, + { + "epoch": 10.401673640167363, + "grad_norm": 3.5678577423095703, + "learning_rate": 9.9776359977636e-06, + "loss": 0.05, + "step": 22374 + }, + { + "epoch": 10.402138540213855, + "grad_norm": 4.530663967132568, + "learning_rate": 9.977634997763501e-06, + "loss": 0.0651, + "step": 22375 + }, + { + "epoch": 10.402603440260345, + "grad_norm": 3.379589557647705, + "learning_rate": 9.977633997763402e-06, + "loss": 0.0381, + "step": 22376 + }, + { + "epoch": 10.403068340306834, + "grad_norm": 3.2793195247650146, + "learning_rate": 9.9776329977633e-06, + "loss": 0.0284, + "step": 22377 + }, + { + "epoch": 10.403533240353324, + "grad_norm": 5.378786087036133, + "learning_rate": 9.977631997763201e-06, + "loss": 0.089, + "step": 22378 + }, + { + "epoch": 10.403998140399814, + "grad_norm": 4.339467525482178, + "learning_rate": 9.9776309977631e-06, + "loss": 0.0664, + "step": 22379 + }, + { + "epoch": 10.404463040446304, + "grad_norm": 5.096521854400635, + "learning_rate": 9.977629997763e-06, + "loss": 0.0782, + "step": 22380 + }, + { + "epoch": 10.404927940492794, + "grad_norm": 6.088672161102295, + "learning_rate": 9.9776289977629e-06, + "loss": 0.047, + "step": 22381 + }, + { + "epoch": 10.405392840539283, + "grad_norm": 1.879982590675354, + "learning_rate": 9.9776279977628e-06, + "loss": 0.019, + "step": 22382 + }, + { + "epoch": 10.405857740585773, + "grad_norm": 2.3841097354888916, + "learning_rate": 9.977626997762701e-06, + "loss": 0.0322, + "step": 22383 + }, + { + "epoch": 10.406322640632265, + "grad_norm": 7.00637149810791, + "learning_rate": 9.977625997762602e-06, + "loss": 0.0649, + "step": 22384 + }, + { + "epoch": 10.406787540678755, + "grad_norm": 5.111064910888672, + "learning_rate": 9.9776249977625e-06, + "loss": 0.0796, + "step": 22385 + }, + { + "epoch": 10.407252440725244, + "grad_norm": 4.183891296386719, + "learning_rate": 9.9776239977624e-06, + "loss": 0.0576, + "step": 22386 + }, + { + "epoch": 10.407717340771734, + "grad_norm": 1.8855246305465698, + "learning_rate": 9.9776229977623e-06, + "loss": 0.0131, + "step": 22387 + }, + { + "epoch": 10.408182240818224, + "grad_norm": 3.6344778537750244, + "learning_rate": 9.977621997762201e-06, + "loss": 0.0406, + "step": 22388 + }, + { + "epoch": 10.408647140864714, + "grad_norm": 3.012744903564453, + "learning_rate": 9.977620997762102e-06, + "loss": 0.0344, + "step": 22389 + }, + { + "epoch": 10.409112040911204, + "grad_norm": 4.915340900421143, + "learning_rate": 9.977619997762e-06, + "loss": 0.0834, + "step": 22390 + }, + { + "epoch": 10.409576940957693, + "grad_norm": 4.074066638946533, + "learning_rate": 9.977618997761901e-06, + "loss": 0.0458, + "step": 22391 + }, + { + "epoch": 10.410041841004183, + "grad_norm": 3.2531375885009766, + "learning_rate": 9.9776179977618e-06, + "loss": 0.04, + "step": 22392 + }, + { + "epoch": 10.410506741050675, + "grad_norm": 7.790425777435303, + "learning_rate": 9.977616997761701e-06, + "loss": 0.1291, + "step": 22393 + }, + { + "epoch": 10.410971641097165, + "grad_norm": 6.1941399574279785, + "learning_rate": 9.9776159977616e-06, + "loss": 0.1475, + "step": 22394 + }, + { + "epoch": 10.411436541143654, + "grad_norm": 3.7519752979278564, + "learning_rate": 9.9776149977615e-06, + "loss": 0.0565, + "step": 22395 + }, + { + "epoch": 10.411901441190144, + "grad_norm": 3.521296501159668, + "learning_rate": 9.977613997761401e-06, + "loss": 0.0311, + "step": 22396 + }, + { + "epoch": 10.412366341236634, + "grad_norm": 8.036038398742676, + "learning_rate": 9.9776129977613e-06, + "loss": 0.1793, + "step": 22397 + }, + { + "epoch": 10.412831241283124, + "grad_norm": 4.8989362716674805, + "learning_rate": 9.9776119977612e-06, + "loss": 0.0593, + "step": 22398 + }, + { + "epoch": 10.413296141329614, + "grad_norm": 6.568061828613281, + "learning_rate": 9.9776109977611e-06, + "loss": 0.1197, + "step": 22399 + }, + { + "epoch": 10.413761041376103, + "grad_norm": 7.456986904144287, + "learning_rate": 9.977609997761e-06, + "loss": 0.1384, + "step": 22400 + }, + { + "epoch": 10.414225941422593, + "grad_norm": 4.832147121429443, + "learning_rate": 9.977608997760901e-06, + "loss": 0.0283, + "step": 22401 + }, + { + "epoch": 10.414690841469085, + "grad_norm": 4.633166313171387, + "learning_rate": 9.977607997760802e-06, + "loss": 0.0672, + "step": 22402 + }, + { + "epoch": 10.415155741515575, + "grad_norm": 0.536836564540863, + "learning_rate": 9.9776069977607e-06, + "loss": 0.0043, + "step": 22403 + }, + { + "epoch": 10.415620641562064, + "grad_norm": 9.793743133544922, + "learning_rate": 9.977605997760601e-06, + "loss": 0.1678, + "step": 22404 + }, + { + "epoch": 10.416085541608554, + "grad_norm": 4.572716236114502, + "learning_rate": 9.9776049977605e-06, + "loss": 0.0298, + "step": 22405 + }, + { + "epoch": 10.416550441655044, + "grad_norm": 4.874861240386963, + "learning_rate": 9.977603997760401e-06, + "loss": 0.0485, + "step": 22406 + }, + { + "epoch": 10.417015341701534, + "grad_norm": 2.1634364128112793, + "learning_rate": 9.9776029977603e-06, + "loss": 0.0183, + "step": 22407 + }, + { + "epoch": 10.417480241748024, + "grad_norm": 3.961289167404175, + "learning_rate": 9.9776019977602e-06, + "loss": 0.0693, + "step": 22408 + }, + { + "epoch": 10.417945141794513, + "grad_norm": 3.096204996109009, + "learning_rate": 9.977600997760101e-06, + "loss": 0.0298, + "step": 22409 + }, + { + "epoch": 10.418410041841003, + "grad_norm": 1.8000414371490479, + "learning_rate": 9.97759999776e-06, + "loss": 0.0154, + "step": 22410 + }, + { + "epoch": 10.418874941887495, + "grad_norm": 5.470310211181641, + "learning_rate": 9.9775989977599e-06, + "loss": 0.045, + "step": 22411 + }, + { + "epoch": 10.419339841933985, + "grad_norm": 2.2757012844085693, + "learning_rate": 9.9775979977598e-06, + "loss": 0.0164, + "step": 22412 + }, + { + "epoch": 10.419804741980474, + "grad_norm": 1.1531857252120972, + "learning_rate": 9.9775969977597e-06, + "loss": 0.0056, + "step": 22413 + }, + { + "epoch": 10.420269642026964, + "grad_norm": 8.987873077392578, + "learning_rate": 9.977595997759601e-06, + "loss": 0.0712, + "step": 22414 + }, + { + "epoch": 10.420734542073454, + "grad_norm": 1.5057971477508545, + "learning_rate": 9.977594997759502e-06, + "loss": 0.0202, + "step": 22415 + }, + { + "epoch": 10.421199442119944, + "grad_norm": 1.0162628889083862, + "learning_rate": 9.9775939977594e-06, + "loss": 0.006, + "step": 22416 + }, + { + "epoch": 10.421664342166434, + "grad_norm": 5.39591121673584, + "learning_rate": 9.977592997759301e-06, + "loss": 0.0572, + "step": 22417 + }, + { + "epoch": 10.422129242212923, + "grad_norm": 5.36463737487793, + "learning_rate": 9.9775919977592e-06, + "loss": 0.0816, + "step": 22418 + }, + { + "epoch": 10.422594142259415, + "grad_norm": 1.878509759902954, + "learning_rate": 9.9775909977591e-06, + "loss": 0.0125, + "step": 22419 + }, + { + "epoch": 10.423059042305905, + "grad_norm": 4.941616535186768, + "learning_rate": 9.977589997759e-06, + "loss": 0.0804, + "step": 22420 + }, + { + "epoch": 10.423523942352395, + "grad_norm": 2.9809603691101074, + "learning_rate": 9.9775889977589e-06, + "loss": 0.0348, + "step": 22421 + }, + { + "epoch": 10.423988842398884, + "grad_norm": 1.9848713874816895, + "learning_rate": 9.977587997758801e-06, + "loss": 0.0211, + "step": 22422 + }, + { + "epoch": 10.424453742445374, + "grad_norm": 2.148273229598999, + "learning_rate": 9.9775869977587e-06, + "loss": 0.0212, + "step": 22423 + }, + { + "epoch": 10.424918642491864, + "grad_norm": 3.7638189792633057, + "learning_rate": 9.977585997758601e-06, + "loss": 0.0353, + "step": 22424 + }, + { + "epoch": 10.425383542538354, + "grad_norm": 2.9947218894958496, + "learning_rate": 9.9775849977585e-06, + "loss": 0.028, + "step": 22425 + }, + { + "epoch": 10.425848442584844, + "grad_norm": 6.229456424713135, + "learning_rate": 9.9775839977584e-06, + "loss": 0.0987, + "step": 22426 + }, + { + "epoch": 10.426313342631333, + "grad_norm": 4.743224620819092, + "learning_rate": 9.977582997758301e-06, + "loss": 0.0763, + "step": 22427 + }, + { + "epoch": 10.426778242677825, + "grad_norm": 3.027041435241699, + "learning_rate": 9.977581997758202e-06, + "loss": 0.0167, + "step": 22428 + }, + { + "epoch": 10.427243142724315, + "grad_norm": 5.200630187988281, + "learning_rate": 9.9775809977581e-06, + "loss": 0.0504, + "step": 22429 + }, + { + "epoch": 10.427708042770805, + "grad_norm": 8.730401039123535, + "learning_rate": 9.977579997758e-06, + "loss": 0.0718, + "step": 22430 + }, + { + "epoch": 10.428172942817294, + "grad_norm": 3.8488070964813232, + "learning_rate": 9.9775789977579e-06, + "loss": 0.0405, + "step": 22431 + }, + { + "epoch": 10.428637842863784, + "grad_norm": 4.1139445304870605, + "learning_rate": 9.9775779977578e-06, + "loss": 0.0306, + "step": 22432 + }, + { + "epoch": 10.429102742910274, + "grad_norm": 3.7842888832092285, + "learning_rate": 9.9775769977577e-06, + "loss": 0.0573, + "step": 22433 + }, + { + "epoch": 10.429567642956764, + "grad_norm": 4.59697151184082, + "learning_rate": 9.9775759977576e-06, + "loss": 0.079, + "step": 22434 + }, + { + "epoch": 10.430032543003254, + "grad_norm": 4.813450813293457, + "learning_rate": 9.977574997757501e-06, + "loss": 0.0331, + "step": 22435 + }, + { + "epoch": 10.430497443049743, + "grad_norm": 1.6999566555023193, + "learning_rate": 9.9775739977574e-06, + "loss": 0.0138, + "step": 22436 + }, + { + "epoch": 10.430962343096235, + "grad_norm": 5.214402675628662, + "learning_rate": 9.977572997757301e-06, + "loss": 0.0828, + "step": 22437 + }, + { + "epoch": 10.431427243142725, + "grad_norm": 9.154720306396484, + "learning_rate": 9.9775719977572e-06, + "loss": 0.1608, + "step": 22438 + }, + { + "epoch": 10.431892143189215, + "grad_norm": 10.750441551208496, + "learning_rate": 9.9775709977571e-06, + "loss": 0.1691, + "step": 22439 + }, + { + "epoch": 10.432357043235704, + "grad_norm": 3.140472412109375, + "learning_rate": 9.977569997757001e-06, + "loss": 0.0461, + "step": 22440 + }, + { + "epoch": 10.432821943282194, + "grad_norm": 4.8721232414245605, + "learning_rate": 9.9775689977569e-06, + "loss": 0.0992, + "step": 22441 + }, + { + "epoch": 10.433286843328684, + "grad_norm": 4.4205756187438965, + "learning_rate": 9.977567997756801e-06, + "loss": 0.0413, + "step": 22442 + }, + { + "epoch": 10.433751743375174, + "grad_norm": 2.5175576210021973, + "learning_rate": 9.9775669977567e-06, + "loss": 0.0179, + "step": 22443 + }, + { + "epoch": 10.434216643421664, + "grad_norm": 1.2041555643081665, + "learning_rate": 9.9775659977566e-06, + "loss": 0.0128, + "step": 22444 + }, + { + "epoch": 10.434681543468155, + "grad_norm": 5.881338596343994, + "learning_rate": 9.977564997756501e-06, + "loss": 0.0674, + "step": 22445 + }, + { + "epoch": 10.435146443514645, + "grad_norm": 5.787472248077393, + "learning_rate": 9.9775639977564e-06, + "loss": 0.0857, + "step": 22446 + }, + { + "epoch": 10.435611343561135, + "grad_norm": 1.1891471147537231, + "learning_rate": 9.9775629977563e-06, + "loss": 0.0086, + "step": 22447 + }, + { + "epoch": 10.436076243607625, + "grad_norm": 19.08816146850586, + "learning_rate": 9.977561997756201e-06, + "loss": 0.252, + "step": 22448 + }, + { + "epoch": 10.436541143654114, + "grad_norm": 7.589054584503174, + "learning_rate": 9.9775609977561e-06, + "loss": 0.1328, + "step": 22449 + }, + { + "epoch": 10.437006043700604, + "grad_norm": 5.57619571685791, + "learning_rate": 9.977559997756e-06, + "loss": 0.0751, + "step": 22450 + }, + { + "epoch": 10.437470943747094, + "grad_norm": Infinity, + "learning_rate": 9.977559997756e-06, + "loss": 0.0489, + "step": 22451 + }, + { + "epoch": 10.437935843793584, + "grad_norm": 8.425768852233887, + "learning_rate": 9.9775589977559e-06, + "loss": 0.1564, + "step": 22452 + }, + { + "epoch": 10.438400743840074, + "grad_norm": 4.856926918029785, + "learning_rate": 9.9775579977558e-06, + "loss": 0.0703, + "step": 22453 + }, + { + "epoch": 10.438865643886565, + "grad_norm": 5.990525245666504, + "learning_rate": 9.977556997755701e-06, + "loss": 0.1008, + "step": 22454 + }, + { + "epoch": 10.439330543933055, + "grad_norm": 4.912545204162598, + "learning_rate": 9.9775559977556e-06, + "loss": 0.0744, + "step": 22455 + }, + { + "epoch": 10.439795443979545, + "grad_norm": 4.257909297943115, + "learning_rate": 9.977554997755501e-06, + "loss": 0.0452, + "step": 22456 + }, + { + "epoch": 10.440260344026035, + "grad_norm": 2.6568007469177246, + "learning_rate": 9.9775539977554e-06, + "loss": 0.0351, + "step": 22457 + }, + { + "epoch": 10.440725244072524, + "grad_norm": 1.8203420639038086, + "learning_rate": 9.9775529977553e-06, + "loss": 0.0231, + "step": 22458 + }, + { + "epoch": 10.441190144119014, + "grad_norm": 3.58554744720459, + "learning_rate": 9.977551997755201e-06, + "loss": 0.0425, + "step": 22459 + }, + { + "epoch": 10.441655044165504, + "grad_norm": 5.356445789337158, + "learning_rate": 9.9775509977551e-06, + "loss": 0.0563, + "step": 22460 + }, + { + "epoch": 10.442119944211994, + "grad_norm": 0.9895733594894409, + "learning_rate": 9.977549997755001e-06, + "loss": 0.0095, + "step": 22461 + }, + { + "epoch": 10.442584844258484, + "grad_norm": 9.85995864868164, + "learning_rate": 9.9775489977549e-06, + "loss": 0.0228, + "step": 22462 + }, + { + "epoch": 10.443049744304975, + "grad_norm": 2.7425525188446045, + "learning_rate": 9.9775479977548e-06, + "loss": 0.0322, + "step": 22463 + }, + { + "epoch": 10.443514644351465, + "grad_norm": 3.007131338119507, + "learning_rate": 9.9775469977547e-06, + "loss": 0.0152, + "step": 22464 + }, + { + "epoch": 10.443979544397955, + "grad_norm": 7.567829132080078, + "learning_rate": 9.9775459977546e-06, + "loss": 0.0756, + "step": 22465 + }, + { + "epoch": 10.444444444444445, + "grad_norm": 3.693922281265259, + "learning_rate": 9.9775449977545e-06, + "loss": 0.0179, + "step": 22466 + }, + { + "epoch": 10.444909344490934, + "grad_norm": 4.3217082023620605, + "learning_rate": 9.977543997754401e-06, + "loss": 0.0583, + "step": 22467 + }, + { + "epoch": 10.445374244537424, + "grad_norm": 2.1679463386535645, + "learning_rate": 9.9775429977543e-06, + "loss": 0.0116, + "step": 22468 + }, + { + "epoch": 10.445839144583914, + "grad_norm": 6.9014410972595215, + "learning_rate": 9.977541997754201e-06, + "loss": 0.1524, + "step": 22469 + }, + { + "epoch": 10.446304044630404, + "grad_norm": 8.672922134399414, + "learning_rate": 9.9775409977541e-06, + "loss": 0.1728, + "step": 22470 + }, + { + "epoch": 10.446768944676894, + "grad_norm": 5.2107648849487305, + "learning_rate": 9.977539997754e-06, + "loss": 0.0849, + "step": 22471 + }, + { + "epoch": 10.447233844723385, + "grad_norm": 3.5272598266601562, + "learning_rate": 9.977538997753901e-06, + "loss": 0.0336, + "step": 22472 + }, + { + "epoch": 10.447698744769875, + "grad_norm": 1.7987139225006104, + "learning_rate": 9.9775379977538e-06, + "loss": 0.0172, + "step": 22473 + }, + { + "epoch": 10.448163644816365, + "grad_norm": 3.6193902492523193, + "learning_rate": 9.977536997753701e-06, + "loss": 0.0214, + "step": 22474 + }, + { + "epoch": 10.448628544862855, + "grad_norm": 2.68581485748291, + "learning_rate": 9.9775359977536e-06, + "loss": 0.014, + "step": 22475 + }, + { + "epoch": 10.449093444909344, + "grad_norm": 3.9824111461639404, + "learning_rate": 9.9775349977535e-06, + "loss": 0.0493, + "step": 22476 + }, + { + "epoch": 10.449558344955834, + "grad_norm": 5.597261428833008, + "learning_rate": 9.9775339977534e-06, + "loss": 0.0951, + "step": 22477 + }, + { + "epoch": 10.450023245002324, + "grad_norm": 6.585482597351074, + "learning_rate": 9.9775329977533e-06, + "loss": 0.106, + "step": 22478 + }, + { + "epoch": 10.450488145048814, + "grad_norm": 6.901199817657471, + "learning_rate": 9.977531997753201e-06, + "loss": 0.0728, + "step": 22479 + }, + { + "epoch": 10.450953045095304, + "grad_norm": 6.347857475280762, + "learning_rate": 9.977530997753102e-06, + "loss": 0.0296, + "step": 22480 + }, + { + "epoch": 10.451417945141795, + "grad_norm": 6.779892444610596, + "learning_rate": 9.977529997753e-06, + "loss": 0.1012, + "step": 22481 + }, + { + "epoch": 10.451882845188285, + "grad_norm": 6.748566627502441, + "learning_rate": 9.977528997752901e-06, + "loss": 0.1281, + "step": 22482 + }, + { + "epoch": 10.452347745234775, + "grad_norm": 5.395850658416748, + "learning_rate": 9.9775279977528e-06, + "loss": 0.0779, + "step": 22483 + }, + { + "epoch": 10.452812645281265, + "grad_norm": 4.020330429077148, + "learning_rate": 9.9775269977527e-06, + "loss": 0.0414, + "step": 22484 + }, + { + "epoch": 10.453277545327754, + "grad_norm": 3.996095895767212, + "learning_rate": 9.977525997752601e-06, + "loss": 0.0566, + "step": 22485 + }, + { + "epoch": 10.453742445374244, + "grad_norm": 8.060283660888672, + "learning_rate": 9.9775249977525e-06, + "loss": 0.0971, + "step": 22486 + }, + { + "epoch": 10.454207345420734, + "grad_norm": 0.769696056842804, + "learning_rate": 9.977523997752401e-06, + "loss": 0.0047, + "step": 22487 + }, + { + "epoch": 10.454672245467224, + "grad_norm": 2.719601631164551, + "learning_rate": 9.9775229977523e-06, + "loss": 0.0413, + "step": 22488 + }, + { + "epoch": 10.455137145513714, + "grad_norm": 4.271012306213379, + "learning_rate": 9.9775219977522e-06, + "loss": 0.0854, + "step": 22489 + }, + { + "epoch": 10.455602045560205, + "grad_norm": 7.28941011428833, + "learning_rate": 9.9775209977521e-06, + "loss": 0.0638, + "step": 22490 + }, + { + "epoch": 10.456066945606695, + "grad_norm": 6.700522422790527, + "learning_rate": 9.977519997752e-06, + "loss": 0.1349, + "step": 22491 + }, + { + "epoch": 10.456531845653185, + "grad_norm": 6.401631832122803, + "learning_rate": 9.977518997751901e-06, + "loss": 0.0801, + "step": 22492 + }, + { + "epoch": 10.456996745699675, + "grad_norm": 1.09213387966156, + "learning_rate": 9.977517997751802e-06, + "loss": 0.0116, + "step": 22493 + }, + { + "epoch": 10.457461645746164, + "grad_norm": 2.2274625301361084, + "learning_rate": 9.9775169977517e-06, + "loss": 0.0213, + "step": 22494 + }, + { + "epoch": 10.457926545792654, + "grad_norm": 6.155235290527344, + "learning_rate": 9.9775159977516e-06, + "loss": 0.0957, + "step": 22495 + }, + { + "epoch": 10.458391445839144, + "grad_norm": 4.268522262573242, + "learning_rate": 9.9775149977515e-06, + "loss": 0.0888, + "step": 22496 + }, + { + "epoch": 10.458856345885634, + "grad_norm": 6.486726760864258, + "learning_rate": 9.9775139977514e-06, + "loss": 0.187, + "step": 22497 + }, + { + "epoch": 10.459321245932125, + "grad_norm": 8.416769027709961, + "learning_rate": 9.977512997751302e-06, + "loss": 0.0445, + "step": 22498 + }, + { + "epoch": 10.459786145978615, + "grad_norm": 4.287550926208496, + "learning_rate": 9.9775119977512e-06, + "loss": 0.0634, + "step": 22499 + }, + { + "epoch": 10.460251046025105, + "grad_norm": 3.950590133666992, + "learning_rate": 9.977510997751101e-06, + "loss": 0.0387, + "step": 22500 + }, + { + "epoch": 10.460715946071595, + "grad_norm": 6.680884838104248, + "learning_rate": 9.977509997751e-06, + "loss": 0.0749, + "step": 22501 + }, + { + "epoch": 10.461180846118085, + "grad_norm": 4.710672855377197, + "learning_rate": 9.9775089977509e-06, + "loss": 0.0733, + "step": 22502 + }, + { + "epoch": 10.461645746164574, + "grad_norm": 2.3376312255859375, + "learning_rate": 9.9775079977508e-06, + "loss": 0.0183, + "step": 22503 + }, + { + "epoch": 10.462110646211064, + "grad_norm": 5.8342814445495605, + "learning_rate": 9.9775069977507e-06, + "loss": 0.0772, + "step": 22504 + }, + { + "epoch": 10.462575546257554, + "grad_norm": 4.700986385345459, + "learning_rate": 9.977505997750601e-06, + "loss": 0.0547, + "step": 22505 + }, + { + "epoch": 10.463040446304044, + "grad_norm": 0.9147709012031555, + "learning_rate": 9.9775049977505e-06, + "loss": 0.0076, + "step": 22506 + }, + { + "epoch": 10.463505346350535, + "grad_norm": 5.941808223724365, + "learning_rate": 9.9775039977504e-06, + "loss": 0.0859, + "step": 22507 + }, + { + "epoch": 10.463970246397025, + "grad_norm": 5.417187213897705, + "learning_rate": 9.9775029977503e-06, + "loss": 0.0728, + "step": 22508 + }, + { + "epoch": 10.464435146443515, + "grad_norm": 5.738011360168457, + "learning_rate": 9.9775019977502e-06, + "loss": 0.0857, + "step": 22509 + }, + { + "epoch": 10.464900046490005, + "grad_norm": 5.605205059051514, + "learning_rate": 9.977500997750101e-06, + "loss": 0.0639, + "step": 22510 + }, + { + "epoch": 10.465364946536495, + "grad_norm": 5.908965110778809, + "learning_rate": 9.977499997750002e-06, + "loss": 0.0794, + "step": 22511 + }, + { + "epoch": 10.465829846582984, + "grad_norm": 3.149024486541748, + "learning_rate": 9.9774989977499e-06, + "loss": 0.0341, + "step": 22512 + }, + { + "epoch": 10.466294746629474, + "grad_norm": 5.284543991088867, + "learning_rate": 9.977497997749801e-06, + "loss": 0.1359, + "step": 22513 + }, + { + "epoch": 10.466759646675964, + "grad_norm": 8.054885864257812, + "learning_rate": 9.9774969977497e-06, + "loss": 0.159, + "step": 22514 + }, + { + "epoch": 10.467224546722454, + "grad_norm": 3.134506940841675, + "learning_rate": 9.9774959977496e-06, + "loss": 0.0281, + "step": 22515 + }, + { + "epoch": 10.467689446768945, + "grad_norm": 2.665994882583618, + "learning_rate": 9.9774949977495e-06, + "loss": 0.0313, + "step": 22516 + }, + { + "epoch": 10.468154346815435, + "grad_norm": 5.399250507354736, + "learning_rate": 9.9774939977494e-06, + "loss": 0.1164, + "step": 22517 + }, + { + "epoch": 10.468619246861925, + "grad_norm": 2.513026475906372, + "learning_rate": 9.977492997749301e-06, + "loss": 0.0231, + "step": 22518 + }, + { + "epoch": 10.469084146908415, + "grad_norm": 2.892930030822754, + "learning_rate": 9.9774919977492e-06, + "loss": 0.0405, + "step": 22519 + }, + { + "epoch": 10.469549046954905, + "grad_norm": 3.117945432662964, + "learning_rate": 9.9774909977491e-06, + "loss": 0.0198, + "step": 22520 + }, + { + "epoch": 10.470013947001394, + "grad_norm": 5.021913528442383, + "learning_rate": 9.977489997749e-06, + "loss": 0.0681, + "step": 22521 + }, + { + "epoch": 10.470478847047884, + "grad_norm": 4.547823429107666, + "learning_rate": 9.9774889977489e-06, + "loss": 0.0448, + "step": 22522 + }, + { + "epoch": 10.470943747094374, + "grad_norm": 7.7589216232299805, + "learning_rate": 9.977487997748801e-06, + "loss": 0.1297, + "step": 22523 + }, + { + "epoch": 10.471408647140866, + "grad_norm": 7.267570495605469, + "learning_rate": 9.977486997748702e-06, + "loss": 0.119, + "step": 22524 + }, + { + "epoch": 10.471873547187355, + "grad_norm": 3.518606424331665, + "learning_rate": 9.9774859977486e-06, + "loss": 0.0289, + "step": 22525 + }, + { + "epoch": 10.472338447233845, + "grad_norm": 12.762240409851074, + "learning_rate": 9.977484997748501e-06, + "loss": 0.2263, + "step": 22526 + }, + { + "epoch": 10.472803347280335, + "grad_norm": 3.4959568977355957, + "learning_rate": 9.9774839977484e-06, + "loss": 0.0452, + "step": 22527 + }, + { + "epoch": 10.473268247326825, + "grad_norm": 8.165901184082031, + "learning_rate": 9.977482997748301e-06, + "loss": 0.0766, + "step": 22528 + }, + { + "epoch": 10.473733147373315, + "grad_norm": 5.779180526733398, + "learning_rate": 9.9774819977482e-06, + "loss": 0.0841, + "step": 22529 + }, + { + "epoch": 10.474198047419804, + "grad_norm": 4.40308952331543, + "learning_rate": 9.9774809977481e-06, + "loss": 0.0407, + "step": 22530 + }, + { + "epoch": 10.474662947466294, + "grad_norm": 4.22599458694458, + "learning_rate": 9.977479997748001e-06, + "loss": 0.0536, + "step": 22531 + }, + { + "epoch": 10.475127847512784, + "grad_norm": 2.3837625980377197, + "learning_rate": 9.9774789977479e-06, + "loss": 0.0218, + "step": 22532 + }, + { + "epoch": 10.475592747559276, + "grad_norm": 7.600063323974609, + "learning_rate": 9.9774779977478e-06, + "loss": 0.116, + "step": 22533 + }, + { + "epoch": 10.476057647605765, + "grad_norm": 7.23827600479126, + "learning_rate": 9.9774769977477e-06, + "loss": 0.1483, + "step": 22534 + }, + { + "epoch": 10.476522547652255, + "grad_norm": 4.608368396759033, + "learning_rate": 9.9774759977476e-06, + "loss": 0.0802, + "step": 22535 + }, + { + "epoch": 10.476987447698745, + "grad_norm": 2.4700984954833984, + "learning_rate": 9.977474997747501e-06, + "loss": 0.039, + "step": 22536 + }, + { + "epoch": 10.477452347745235, + "grad_norm": 3.605957508087158, + "learning_rate": 9.977473997747402e-06, + "loss": 0.0512, + "step": 22537 + }, + { + "epoch": 10.477917247791725, + "grad_norm": 8.11956787109375, + "learning_rate": 9.9774729977473e-06, + "loss": 0.2115, + "step": 22538 + }, + { + "epoch": 10.478382147838214, + "grad_norm": 7.393863677978516, + "learning_rate": 9.9774719977472e-06, + "loss": 0.09, + "step": 22539 + }, + { + "epoch": 10.478847047884704, + "grad_norm": 7.223166465759277, + "learning_rate": 9.9774709977471e-06, + "loss": 0.1498, + "step": 22540 + }, + { + "epoch": 10.479311947931194, + "grad_norm": 3.2551865577697754, + "learning_rate": 9.977469997747001e-06, + "loss": 0.035, + "step": 22541 + }, + { + "epoch": 10.479776847977686, + "grad_norm": 4.4557647705078125, + "learning_rate": 9.977468997746902e-06, + "loss": 0.0428, + "step": 22542 + }, + { + "epoch": 10.480241748024175, + "grad_norm": 0.8896089196205139, + "learning_rate": 9.9774679977468e-06, + "loss": 0.0083, + "step": 22543 + }, + { + "epoch": 10.480706648070665, + "grad_norm": 3.7623157501220703, + "learning_rate": 9.977466997746701e-06, + "loss": 0.0459, + "step": 22544 + }, + { + "epoch": 10.481171548117155, + "grad_norm": 8.268488883972168, + "learning_rate": 9.9774659977466e-06, + "loss": 0.1495, + "step": 22545 + }, + { + "epoch": 10.481636448163645, + "grad_norm": 5.375566482543945, + "learning_rate": 9.977464997746501e-06, + "loss": 0.1121, + "step": 22546 + }, + { + "epoch": 10.482101348210135, + "grad_norm": 4.205035209655762, + "learning_rate": 9.9774639977464e-06, + "loss": 0.0247, + "step": 22547 + }, + { + "epoch": 10.482566248256624, + "grad_norm": 4.001674652099609, + "learning_rate": 9.9774629977463e-06, + "loss": 0.0461, + "step": 22548 + }, + { + "epoch": 10.483031148303114, + "grad_norm": 4.516814231872559, + "learning_rate": 9.977461997746201e-06, + "loss": 0.07, + "step": 22549 + }, + { + "epoch": 10.483496048349604, + "grad_norm": 6.300269603729248, + "learning_rate": 9.9774609977461e-06, + "loss": 0.1255, + "step": 22550 + }, + { + "epoch": 10.483960948396096, + "grad_norm": 3.510601282119751, + "learning_rate": 9.977459997746e-06, + "loss": 0.0452, + "step": 22551 + }, + { + "epoch": 10.484425848442585, + "grad_norm": 5.417283058166504, + "learning_rate": 9.9774589977459e-06, + "loss": 0.0934, + "step": 22552 + }, + { + "epoch": 10.484890748489075, + "grad_norm": 0.9062909483909607, + "learning_rate": 9.9774579977458e-06, + "loss": 0.008, + "step": 22553 + }, + { + "epoch": 10.485355648535565, + "grad_norm": 1.9305698871612549, + "learning_rate": 9.977456997745701e-06, + "loss": 0.0153, + "step": 22554 + }, + { + "epoch": 10.485820548582055, + "grad_norm": 3.793795585632324, + "learning_rate": 9.977455997745602e-06, + "loss": 0.0443, + "step": 22555 + }, + { + "epoch": 10.486285448628545, + "grad_norm": 5.0096940994262695, + "learning_rate": 9.9774549977455e-06, + "loss": 0.0744, + "step": 22556 + }, + { + "epoch": 10.486750348675034, + "grad_norm": 3.884995460510254, + "learning_rate": 9.977453997745401e-06, + "loss": 0.0217, + "step": 22557 + }, + { + "epoch": 10.487215248721524, + "grad_norm": 7.206756114959717, + "learning_rate": 9.9774529977453e-06, + "loss": 0.155, + "step": 22558 + }, + { + "epoch": 10.487680148768014, + "grad_norm": 5.040481090545654, + "learning_rate": 9.977451997745201e-06, + "loss": 0.0626, + "step": 22559 + }, + { + "epoch": 10.488145048814506, + "grad_norm": 5.593771934509277, + "learning_rate": 9.9774509977451e-06, + "loss": 0.0459, + "step": 22560 + }, + { + "epoch": 10.488609948860995, + "grad_norm": 4.236593246459961, + "learning_rate": 9.977449997745e-06, + "loss": 0.0674, + "step": 22561 + }, + { + "epoch": 10.489074848907485, + "grad_norm": 1.6276935338974, + "learning_rate": 9.977448997744901e-06, + "loss": 0.014, + "step": 22562 + }, + { + "epoch": 10.489539748953975, + "grad_norm": 4.897289276123047, + "learning_rate": 9.9774479977448e-06, + "loss": 0.0346, + "step": 22563 + }, + { + "epoch": 10.490004649000465, + "grad_norm": 8.09427547454834, + "learning_rate": 9.977446997744701e-06, + "loss": 0.1271, + "step": 22564 + }, + { + "epoch": 10.490469549046955, + "grad_norm": 4.1962103843688965, + "learning_rate": 9.9774459977446e-06, + "loss": 0.0219, + "step": 22565 + }, + { + "epoch": 10.490934449093444, + "grad_norm": 5.7304840087890625, + "learning_rate": 9.9774449977445e-06, + "loss": 0.0678, + "step": 22566 + }, + { + "epoch": 10.491399349139934, + "grad_norm": 0.8754092454910278, + "learning_rate": 9.977443997744401e-06, + "loss": 0.0064, + "step": 22567 + }, + { + "epoch": 10.491864249186424, + "grad_norm": 2.09407901763916, + "learning_rate": 9.977442997744302e-06, + "loss": 0.0253, + "step": 22568 + }, + { + "epoch": 10.492329149232916, + "grad_norm": 3.6953651905059814, + "learning_rate": 9.9774419977442e-06, + "loss": 0.0359, + "step": 22569 + }, + { + "epoch": 10.492794049279405, + "grad_norm": 3.829041004180908, + "learning_rate": 9.977440997744101e-06, + "loss": 0.0301, + "step": 22570 + }, + { + "epoch": 10.493258949325895, + "grad_norm": 4.701040744781494, + "learning_rate": 9.977439997744e-06, + "loss": 0.0862, + "step": 22571 + }, + { + "epoch": 10.493723849372385, + "grad_norm": 5.8516106605529785, + "learning_rate": 9.9774389977439e-06, + "loss": 0.1085, + "step": 22572 + }, + { + "epoch": 10.494188749418875, + "grad_norm": 5.0458598136901855, + "learning_rate": 9.9774379977438e-06, + "loss": 0.0682, + "step": 22573 + }, + { + "epoch": 10.494653649465365, + "grad_norm": 2.4626476764678955, + "learning_rate": 9.9774369977437e-06, + "loss": 0.0273, + "step": 22574 + }, + { + "epoch": 10.495118549511854, + "grad_norm": 4.273625373840332, + "learning_rate": 9.977435997743601e-06, + "loss": 0.058, + "step": 22575 + }, + { + "epoch": 10.495583449558344, + "grad_norm": 5.853790760040283, + "learning_rate": 9.9774349977435e-06, + "loss": 0.0393, + "step": 22576 + }, + { + "epoch": 10.496048349604836, + "grad_norm": 6.1814165115356445, + "learning_rate": 9.977433997743401e-06, + "loss": 0.1333, + "step": 22577 + }, + { + "epoch": 10.496513249651326, + "grad_norm": 1.185285210609436, + "learning_rate": 9.9774329977433e-06, + "loss": 0.0067, + "step": 22578 + }, + { + "epoch": 10.496978149697815, + "grad_norm": 5.3943376541137695, + "learning_rate": 9.9774319977432e-06, + "loss": 0.0972, + "step": 22579 + }, + { + "epoch": 10.497443049744305, + "grad_norm": 9.546915054321289, + "learning_rate": 9.977430997743101e-06, + "loss": 0.1745, + "step": 22580 + }, + { + "epoch": 10.497907949790795, + "grad_norm": 3.310999870300293, + "learning_rate": 9.977429997743002e-06, + "loss": 0.0333, + "step": 22581 + }, + { + "epoch": 10.498372849837285, + "grad_norm": 3.302562713623047, + "learning_rate": 9.977428997742901e-06, + "loss": 0.0413, + "step": 22582 + }, + { + "epoch": 10.498837749883775, + "grad_norm": 3.852302074432373, + "learning_rate": 9.9774279977428e-06, + "loss": 0.0359, + "step": 22583 + }, + { + "epoch": 10.499302649930264, + "grad_norm": 5.773025989532471, + "learning_rate": 9.9774269977427e-06, + "loss": 0.1058, + "step": 22584 + }, + { + "epoch": 10.499767549976754, + "grad_norm": 3.784820079803467, + "learning_rate": 9.9774259977426e-06, + "loss": 0.0294, + "step": 22585 + }, + { + "epoch": 10.500232450023246, + "grad_norm": 4.467690944671631, + "learning_rate": 9.9774249977425e-06, + "loss": 0.0552, + "step": 22586 + }, + { + "epoch": 10.500697350069736, + "grad_norm": 2.9308743476867676, + "learning_rate": 9.9774239977424e-06, + "loss": 0.0378, + "step": 22587 + }, + { + "epoch": 10.501162250116225, + "grad_norm": 8.056755065917969, + "learning_rate": 9.977422997742301e-06, + "loss": 0.1379, + "step": 22588 + }, + { + "epoch": 10.501627150162715, + "grad_norm": 6.936691761016846, + "learning_rate": 9.9774219977422e-06, + "loss": 0.0733, + "step": 22589 + }, + { + "epoch": 10.502092050209205, + "grad_norm": 3.5219106674194336, + "learning_rate": 9.977420997742101e-06, + "loss": 0.0357, + "step": 22590 + }, + { + "epoch": 10.502556950255695, + "grad_norm": 6.074166297912598, + "learning_rate": 9.977419997742e-06, + "loss": 0.1003, + "step": 22591 + }, + { + "epoch": 10.503021850302185, + "grad_norm": 5.462009429931641, + "learning_rate": 9.9774189977419e-06, + "loss": 0.0969, + "step": 22592 + }, + { + "epoch": 10.503486750348674, + "grad_norm": 9.947415351867676, + "learning_rate": 9.977417997741801e-06, + "loss": 0.272, + "step": 22593 + }, + { + "epoch": 10.503951650395164, + "grad_norm": 6.244166374206543, + "learning_rate": 9.9774169977417e-06, + "loss": 0.0753, + "step": 22594 + }, + { + "epoch": 10.504416550441656, + "grad_norm": 3.8309900760650635, + "learning_rate": 9.977415997741601e-06, + "loss": 0.0481, + "step": 22595 + }, + { + "epoch": 10.504881450488146, + "grad_norm": 5.075826168060303, + "learning_rate": 9.9774149977415e-06, + "loss": 0.0577, + "step": 22596 + }, + { + "epoch": 10.505346350534635, + "grad_norm": 2.74890398979187, + "learning_rate": 9.9774139977414e-06, + "loss": 0.0288, + "step": 22597 + }, + { + "epoch": 10.505811250581125, + "grad_norm": 2.1681389808654785, + "learning_rate": 9.9774129977413e-06, + "loss": 0.0198, + "step": 22598 + }, + { + "epoch": 10.506276150627615, + "grad_norm": 5.891228199005127, + "learning_rate": 9.9774119977412e-06, + "loss": 0.1001, + "step": 22599 + }, + { + "epoch": 10.506741050674105, + "grad_norm": 3.387665271759033, + "learning_rate": 9.9774109977411e-06, + "loss": 0.0443, + "step": 22600 + }, + { + "epoch": 10.507205950720595, + "grad_norm": 0.7829504013061523, + "learning_rate": 9.977409997741001e-06, + "loss": 0.0069, + "step": 22601 + }, + { + "epoch": 10.507670850767084, + "grad_norm": 4.801507949829102, + "learning_rate": 9.9774089977409e-06, + "loss": 0.1223, + "step": 22602 + }, + { + "epoch": 10.508135750813576, + "grad_norm": 8.339144706726074, + "learning_rate": 9.9774079977408e-06, + "loss": 0.0921, + "step": 22603 + }, + { + "epoch": 10.508600650860066, + "grad_norm": 7.056282043457031, + "learning_rate": 9.9774069977407e-06, + "loss": 0.1207, + "step": 22604 + }, + { + "epoch": 10.509065550906556, + "grad_norm": 3.2792654037475586, + "learning_rate": 9.9774059977406e-06, + "loss": 0.0472, + "step": 22605 + }, + { + "epoch": 10.509530450953045, + "grad_norm": 7.222952842712402, + "learning_rate": 9.977404997740501e-06, + "loss": 0.117, + "step": 22606 + }, + { + "epoch": 10.509995350999535, + "grad_norm": 5.214432239532471, + "learning_rate": 9.9774039977404e-06, + "loss": 0.0549, + "step": 22607 + }, + { + "epoch": 10.510460251046025, + "grad_norm": 5.60028600692749, + "learning_rate": 9.977402997740301e-06, + "loss": 0.1114, + "step": 22608 + }, + { + "epoch": 10.510925151092515, + "grad_norm": 3.8403799533843994, + "learning_rate": 9.9774019977402e-06, + "loss": 0.0523, + "step": 22609 + }, + { + "epoch": 10.511390051139005, + "grad_norm": 5.857977867126465, + "learning_rate": 9.9774009977401e-06, + "loss": 0.1282, + "step": 22610 + }, + { + "epoch": 10.511854951185494, + "grad_norm": 2.712517261505127, + "learning_rate": 9.977399997740001e-06, + "loss": 0.0378, + "step": 22611 + }, + { + "epoch": 10.512319851231986, + "grad_norm": 5.235821723937988, + "learning_rate": 9.9773989977399e-06, + "loss": 0.0772, + "step": 22612 + }, + { + "epoch": 10.512784751278476, + "grad_norm": 5.652141571044922, + "learning_rate": 9.977397997739801e-06, + "loss": 0.0994, + "step": 22613 + }, + { + "epoch": 10.513249651324966, + "grad_norm": 7.755229473114014, + "learning_rate": 9.9773969977397e-06, + "loss": 0.1555, + "step": 22614 + }, + { + "epoch": 10.513714551371455, + "grad_norm": 5.531989097595215, + "learning_rate": 9.9773959977396e-06, + "loss": 0.0644, + "step": 22615 + }, + { + "epoch": 10.514179451417945, + "grad_norm": 6.537729740142822, + "learning_rate": 9.9773949977395e-06, + "loss": 0.1422, + "step": 22616 + }, + { + "epoch": 10.514644351464435, + "grad_norm": 4.106064319610596, + "learning_rate": 9.9773939977394e-06, + "loss": 0.0758, + "step": 22617 + }, + { + "epoch": 10.515109251510925, + "grad_norm": 2.851243019104004, + "learning_rate": 9.9773929977393e-06, + "loss": 0.0412, + "step": 22618 + }, + { + "epoch": 10.515574151557415, + "grad_norm": 5.909782409667969, + "learning_rate": 9.977391997739201e-06, + "loss": 0.0945, + "step": 22619 + }, + { + "epoch": 10.516039051603904, + "grad_norm": 2.7110302448272705, + "learning_rate": 9.9773909977391e-06, + "loss": 0.031, + "step": 22620 + }, + { + "epoch": 10.516503951650396, + "grad_norm": 1.9049445390701294, + "learning_rate": 9.977389997739001e-06, + "loss": 0.0153, + "step": 22621 + }, + { + "epoch": 10.516968851696886, + "grad_norm": 4.805277347564697, + "learning_rate": 9.9773889977389e-06, + "loss": 0.0866, + "step": 22622 + }, + { + "epoch": 10.517433751743376, + "grad_norm": 9.939788818359375, + "learning_rate": 9.9773879977388e-06, + "loss": 0.0885, + "step": 22623 + }, + { + "epoch": 10.517898651789865, + "grad_norm": 5.001017093658447, + "learning_rate": 9.977386997738701e-06, + "loss": 0.1061, + "step": 22624 + }, + { + "epoch": 10.518363551836355, + "grad_norm": 5.53001070022583, + "learning_rate": 9.9773859977386e-06, + "loss": 0.0873, + "step": 22625 + }, + { + "epoch": 10.518828451882845, + "grad_norm": 3.594249725341797, + "learning_rate": 9.977384997738501e-06, + "loss": 0.0357, + "step": 22626 + }, + { + "epoch": 10.519293351929335, + "grad_norm": 4.122611045837402, + "learning_rate": 9.9773839977384e-06, + "loss": 0.0483, + "step": 22627 + }, + { + "epoch": 10.519758251975825, + "grad_norm": 8.548665046691895, + "learning_rate": 9.9773829977383e-06, + "loss": 0.1737, + "step": 22628 + }, + { + "epoch": 10.520223152022314, + "grad_norm": 3.8357996940612793, + "learning_rate": 9.9773819977382e-06, + "loss": 0.0639, + "step": 22629 + }, + { + "epoch": 10.520688052068806, + "grad_norm": 5.613067626953125, + "learning_rate": 9.9773809977381e-06, + "loss": 0.0839, + "step": 22630 + }, + { + "epoch": 10.521152952115296, + "grad_norm": 4.600648403167725, + "learning_rate": 9.977379997738001e-06, + "loss": 0.0393, + "step": 22631 + }, + { + "epoch": 10.521617852161786, + "grad_norm": 2.8512425422668457, + "learning_rate": 9.977378997737902e-06, + "loss": 0.0432, + "step": 22632 + }, + { + "epoch": 10.522082752208275, + "grad_norm": 4.2793498039245605, + "learning_rate": 9.9773779977378e-06, + "loss": 0.0751, + "step": 22633 + }, + { + "epoch": 10.522547652254765, + "grad_norm": 4.417285442352295, + "learning_rate": 9.977376997737701e-06, + "loss": 0.0618, + "step": 22634 + }, + { + "epoch": 10.523012552301255, + "grad_norm": 5.889036655426025, + "learning_rate": 9.9773759977376e-06, + "loss": 0.0785, + "step": 22635 + }, + { + "epoch": 10.523477452347745, + "grad_norm": 7.688382625579834, + "learning_rate": 9.9773749977375e-06, + "loss": 0.2099, + "step": 22636 + }, + { + "epoch": 10.523942352394235, + "grad_norm": 5.260536193847656, + "learning_rate": 9.977373997737401e-06, + "loss": 0.087, + "step": 22637 + }, + { + "epoch": 10.524407252440724, + "grad_norm": 5.787991046905518, + "learning_rate": 9.9773729977373e-06, + "loss": 0.0808, + "step": 22638 + }, + { + "epoch": 10.524872152487216, + "grad_norm": 4.708867073059082, + "learning_rate": 9.977371997737201e-06, + "loss": 0.0509, + "step": 22639 + }, + { + "epoch": 10.525337052533706, + "grad_norm": 2.173983335494995, + "learning_rate": 9.9773709977371e-06, + "loss": 0.0156, + "step": 22640 + }, + { + "epoch": 10.525801952580196, + "grad_norm": 3.230203151702881, + "learning_rate": 9.977369997737e-06, + "loss": 0.043, + "step": 22641 + }, + { + "epoch": 10.526266852626685, + "grad_norm": 4.659658908843994, + "learning_rate": 9.9773689977369e-06, + "loss": 0.0659, + "step": 22642 + }, + { + "epoch": 10.526731752673175, + "grad_norm": 2.0968899726867676, + "learning_rate": 9.9773679977368e-06, + "loss": 0.0198, + "step": 22643 + }, + { + "epoch": 10.527196652719665, + "grad_norm": 5.188501834869385, + "learning_rate": 9.977366997736701e-06, + "loss": 0.0889, + "step": 22644 + }, + { + "epoch": 10.527661552766155, + "grad_norm": 5.062763214111328, + "learning_rate": 9.977365997736602e-06, + "loss": 0.1286, + "step": 22645 + }, + { + "epoch": 10.528126452812645, + "grad_norm": 3.2966670989990234, + "learning_rate": 9.9773649977365e-06, + "loss": 0.0255, + "step": 22646 + }, + { + "epoch": 10.528591352859134, + "grad_norm": 6.278491973876953, + "learning_rate": 9.9773639977364e-06, + "loss": 0.0993, + "step": 22647 + }, + { + "epoch": 10.529056252905626, + "grad_norm": 5.272966384887695, + "learning_rate": 9.9773629977363e-06, + "loss": 0.0848, + "step": 22648 + }, + { + "epoch": 10.529521152952116, + "grad_norm": 4.875514030456543, + "learning_rate": 9.977361997736201e-06, + "loss": 0.1554, + "step": 22649 + }, + { + "epoch": 10.529986052998606, + "grad_norm": 3.4155914783477783, + "learning_rate": 9.977360997736102e-06, + "loss": 0.0464, + "step": 22650 + }, + { + "epoch": 10.530450953045095, + "grad_norm": 5.184606552124023, + "learning_rate": 9.977359997736e-06, + "loss": 0.0534, + "step": 22651 + }, + { + "epoch": 10.530915853091585, + "grad_norm": 4.622805118560791, + "learning_rate": 9.977358997735901e-06, + "loss": 0.0644, + "step": 22652 + }, + { + "epoch": 10.531380753138075, + "grad_norm": 2.733741044998169, + "learning_rate": 9.9773579977358e-06, + "loss": 0.0271, + "step": 22653 + }, + { + "epoch": 10.531845653184565, + "grad_norm": 2.319519281387329, + "learning_rate": 9.9773569977357e-06, + "loss": 0.0179, + "step": 22654 + }, + { + "epoch": 10.532310553231055, + "grad_norm": 4.37470006942749, + "learning_rate": 9.9773559977356e-06, + "loss": 0.0562, + "step": 22655 + }, + { + "epoch": 10.532775453277544, + "grad_norm": 5.877681732177734, + "learning_rate": 9.9773549977355e-06, + "loss": 0.0752, + "step": 22656 + }, + { + "epoch": 10.533240353324036, + "grad_norm": 4.343132019042969, + "learning_rate": 9.977353997735401e-06, + "loss": 0.081, + "step": 22657 + }, + { + "epoch": 10.533705253370526, + "grad_norm": 0.83753502368927, + "learning_rate": 9.9773529977353e-06, + "loss": 0.0087, + "step": 22658 + }, + { + "epoch": 10.534170153417016, + "grad_norm": 8.993191719055176, + "learning_rate": 9.9773519977352e-06, + "loss": 0.0519, + "step": 22659 + }, + { + "epoch": 10.534635053463505, + "grad_norm": 3.052598476409912, + "learning_rate": 9.9773509977351e-06, + "loss": 0.0252, + "step": 22660 + }, + { + "epoch": 10.535099953509995, + "grad_norm": 6.201953887939453, + "learning_rate": 9.977349997735e-06, + "loss": 0.1215, + "step": 22661 + }, + { + "epoch": 10.535564853556485, + "grad_norm": 3.4402639865875244, + "learning_rate": 9.977348997734901e-06, + "loss": 0.0249, + "step": 22662 + }, + { + "epoch": 10.536029753602975, + "grad_norm": 5.551884174346924, + "learning_rate": 9.977347997734802e-06, + "loss": 0.1023, + "step": 22663 + }, + { + "epoch": 10.536494653649465, + "grad_norm": 1.8263752460479736, + "learning_rate": 9.9773469977347e-06, + "loss": 0.016, + "step": 22664 + }, + { + "epoch": 10.536959553695956, + "grad_norm": 5.731416702270508, + "learning_rate": 9.977345997734601e-06, + "loss": 0.0592, + "step": 22665 + }, + { + "epoch": 10.537424453742446, + "grad_norm": 7.393606185913086, + "learning_rate": 9.9773449977345e-06, + "loss": 0.1413, + "step": 22666 + }, + { + "epoch": 10.537889353788936, + "grad_norm": 6.492157936096191, + "learning_rate": 9.977343997734401e-06, + "loss": 0.0698, + "step": 22667 + }, + { + "epoch": 10.538354253835426, + "grad_norm": 4.78366231918335, + "learning_rate": 9.9773429977343e-06, + "loss": 0.0734, + "step": 22668 + }, + { + "epoch": 10.538819153881915, + "grad_norm": 8.025284767150879, + "learning_rate": 9.9773419977342e-06, + "loss": 0.1722, + "step": 22669 + }, + { + "epoch": 10.539284053928405, + "grad_norm": 6.490952968597412, + "learning_rate": 9.977340997734101e-06, + "loss": 0.1426, + "step": 22670 + }, + { + "epoch": 10.539748953974895, + "grad_norm": 3.6985936164855957, + "learning_rate": 9.977339997734e-06, + "loss": 0.0564, + "step": 22671 + }, + { + "epoch": 10.540213854021385, + "grad_norm": 5.474651336669922, + "learning_rate": 9.9773389977339e-06, + "loss": 0.0791, + "step": 22672 + }, + { + "epoch": 10.540678754067875, + "grad_norm": 3.2159745693206787, + "learning_rate": 9.9773379977338e-06, + "loss": 0.0365, + "step": 22673 + }, + { + "epoch": 10.541143654114366, + "grad_norm": 2.7570788860321045, + "learning_rate": 9.9773369977337e-06, + "loss": 0.0337, + "step": 22674 + }, + { + "epoch": 10.541608554160856, + "grad_norm": 7.616330623626709, + "learning_rate": 9.977335997733601e-06, + "loss": 0.1277, + "step": 22675 + }, + { + "epoch": 10.542073454207346, + "grad_norm": 7.835608959197998, + "learning_rate": 9.977334997733502e-06, + "loss": 0.2436, + "step": 22676 + }, + { + "epoch": 10.542538354253836, + "grad_norm": 3.91200852394104, + "learning_rate": 9.9773339977334e-06, + "loss": 0.0566, + "step": 22677 + }, + { + "epoch": 10.543003254300325, + "grad_norm": 5.301703929901123, + "learning_rate": 9.977332997733301e-06, + "loss": 0.1559, + "step": 22678 + }, + { + "epoch": 10.543468154346815, + "grad_norm": 10.265331268310547, + "learning_rate": 9.9773319977332e-06, + "loss": 0.04, + "step": 22679 + }, + { + "epoch": 10.543933054393305, + "grad_norm": 7.024227619171143, + "learning_rate": 9.977330997733101e-06, + "loss": 0.1674, + "step": 22680 + }, + { + "epoch": 10.544397954439795, + "grad_norm": 9.341599464416504, + "learning_rate": 9.977329997733e-06, + "loss": 0.2196, + "step": 22681 + }, + { + "epoch": 10.544862854486286, + "grad_norm": 3.0358009338378906, + "learning_rate": 9.9773289977329e-06, + "loss": 0.0401, + "step": 22682 + }, + { + "epoch": 10.545327754532776, + "grad_norm": 8.078397750854492, + "learning_rate": 9.977327997732801e-06, + "loss": 0.2086, + "step": 22683 + }, + { + "epoch": 10.545792654579266, + "grad_norm": 6.238412857055664, + "learning_rate": 9.9773269977327e-06, + "loss": 0.0892, + "step": 22684 + }, + { + "epoch": 10.546257554625756, + "grad_norm": 6.3707990646362305, + "learning_rate": 9.9773259977326e-06, + "loss": 0.0851, + "step": 22685 + }, + { + "epoch": 10.546722454672246, + "grad_norm": 7.366359710693359, + "learning_rate": 9.9773249977325e-06, + "loss": 0.1074, + "step": 22686 + }, + { + "epoch": 10.547187354718735, + "grad_norm": 4.1558122634887695, + "learning_rate": 9.9773239977324e-06, + "loss": 0.0345, + "step": 22687 + }, + { + "epoch": 10.547652254765225, + "grad_norm": 3.9336438179016113, + "learning_rate": 9.977322997732301e-06, + "loss": 0.0284, + "step": 22688 + }, + { + "epoch": 10.548117154811715, + "grad_norm": 5.583680152893066, + "learning_rate": 9.977321997732202e-06, + "loss": 0.1115, + "step": 22689 + }, + { + "epoch": 10.548582054858205, + "grad_norm": 3.3482167720794678, + "learning_rate": 9.9773209977321e-06, + "loss": 0.0314, + "step": 22690 + }, + { + "epoch": 10.549046954904696, + "grad_norm": 4.296785354614258, + "learning_rate": 9.977319997732e-06, + "loss": 0.0421, + "step": 22691 + }, + { + "epoch": 10.549511854951186, + "grad_norm": 7.589565753936768, + "learning_rate": 9.9773189977319e-06, + "loss": 0.1831, + "step": 22692 + }, + { + "epoch": 10.549976754997676, + "grad_norm": 5.705369472503662, + "learning_rate": 9.977317997731801e-06, + "loss": 0.0684, + "step": 22693 + }, + { + "epoch": 10.550441655044166, + "grad_norm": 7.408698558807373, + "learning_rate": 9.9773169977317e-06, + "loss": 0.1592, + "step": 22694 + }, + { + "epoch": 10.550906555090656, + "grad_norm": 4.025921821594238, + "learning_rate": 9.9773159977316e-06, + "loss": 0.0581, + "step": 22695 + }, + { + "epoch": 10.551371455137145, + "grad_norm": 1.4250478744506836, + "learning_rate": 9.977314997731501e-06, + "loss": 0.0145, + "step": 22696 + }, + { + "epoch": 10.551836355183635, + "grad_norm": 1.6550548076629639, + "learning_rate": 9.9773139977314e-06, + "loss": 0.0079, + "step": 22697 + }, + { + "epoch": 10.552301255230125, + "grad_norm": 6.254283905029297, + "learning_rate": 9.977312997731301e-06, + "loss": 0.1459, + "step": 22698 + }, + { + "epoch": 10.552766155276615, + "grad_norm": 6.549433708190918, + "learning_rate": 9.9773119977312e-06, + "loss": 0.0752, + "step": 22699 + }, + { + "epoch": 10.553231055323106, + "grad_norm": 1.3347162008285522, + "learning_rate": 9.9773109977311e-06, + "loss": 0.0069, + "step": 22700 + }, + { + "epoch": 10.553695955369596, + "grad_norm": 3.066255807876587, + "learning_rate": 9.977309997731001e-06, + "loss": 0.0257, + "step": 22701 + }, + { + "epoch": 10.554160855416086, + "grad_norm": 4.722487449645996, + "learning_rate": 9.9773089977309e-06, + "loss": 0.0472, + "step": 22702 + }, + { + "epoch": 10.554625755462576, + "grad_norm": 5.349888801574707, + "learning_rate": 9.9773079977308e-06, + "loss": 0.1451, + "step": 22703 + }, + { + "epoch": 10.555090655509066, + "grad_norm": 0.9821576476097107, + "learning_rate": 9.9773069977307e-06, + "loss": 0.0093, + "step": 22704 + }, + { + "epoch": 10.555555555555555, + "grad_norm": 5.225645542144775, + "learning_rate": 9.9773059977306e-06, + "loss": 0.0656, + "step": 22705 + }, + { + "epoch": 10.556020455602045, + "grad_norm": 3.1931068897247314, + "learning_rate": 9.977304997730501e-06, + "loss": 0.0466, + "step": 22706 + }, + { + "epoch": 10.556485355648535, + "grad_norm": 5.973233222961426, + "learning_rate": 9.977303997730402e-06, + "loss": 0.0631, + "step": 22707 + }, + { + "epoch": 10.556950255695025, + "grad_norm": 2.8232431411743164, + "learning_rate": 9.9773029977303e-06, + "loss": 0.0231, + "step": 22708 + }, + { + "epoch": 10.557415155741516, + "grad_norm": 1.352254033088684, + "learning_rate": 9.977301997730201e-06, + "loss": 0.0115, + "step": 22709 + }, + { + "epoch": 10.557880055788006, + "grad_norm": 3.560117244720459, + "learning_rate": 9.9773009977301e-06, + "loss": 0.0402, + "step": 22710 + }, + { + "epoch": 10.558344955834496, + "grad_norm": 2.1185901165008545, + "learning_rate": 9.977299997730001e-06, + "loss": 0.0184, + "step": 22711 + }, + { + "epoch": 10.558809855880986, + "grad_norm": 3.629287004470825, + "learning_rate": 9.9772989977299e-06, + "loss": 0.0484, + "step": 22712 + }, + { + "epoch": 10.559274755927476, + "grad_norm": 2.925208568572998, + "learning_rate": 9.9772979977298e-06, + "loss": 0.0353, + "step": 22713 + }, + { + "epoch": 10.559739655973965, + "grad_norm": 2.602508783340454, + "learning_rate": 9.977296997729701e-06, + "loss": 0.039, + "step": 22714 + }, + { + "epoch": 10.560204556020455, + "grad_norm": 3.742647171020508, + "learning_rate": 9.9772959977296e-06, + "loss": 0.0316, + "step": 22715 + }, + { + "epoch": 10.560669456066945, + "grad_norm": 3.4971039295196533, + "learning_rate": 9.977294997729501e-06, + "loss": 0.066, + "step": 22716 + }, + { + "epoch": 10.561134356113435, + "grad_norm": 3.671637773513794, + "learning_rate": 9.9772939977294e-06, + "loss": 0.03, + "step": 22717 + }, + { + "epoch": 10.561599256159926, + "grad_norm": 3.105886697769165, + "learning_rate": 9.9772929977293e-06, + "loss": 0.0323, + "step": 22718 + }, + { + "epoch": 10.562064156206416, + "grad_norm": 6.772280216217041, + "learning_rate": 9.977291997729201e-06, + "loss": 0.1011, + "step": 22719 + }, + { + "epoch": 10.562529056252906, + "grad_norm": 3.673902988433838, + "learning_rate": 9.977290997729102e-06, + "loss": 0.0392, + "step": 22720 + }, + { + "epoch": 10.562993956299396, + "grad_norm": 5.689248561859131, + "learning_rate": 9.977289997729e-06, + "loss": 0.0745, + "step": 22721 + }, + { + "epoch": 10.563458856345886, + "grad_norm": 2.7999987602233887, + "learning_rate": 9.977288997728901e-06, + "loss": 0.0245, + "step": 22722 + }, + { + "epoch": 10.563923756392375, + "grad_norm": 4.829177379608154, + "learning_rate": 9.9772879977288e-06, + "loss": 0.0538, + "step": 22723 + }, + { + "epoch": 10.564388656438865, + "grad_norm": 6.091412544250488, + "learning_rate": 9.9772869977287e-06, + "loss": 0.0485, + "step": 22724 + }, + { + "epoch": 10.564853556485355, + "grad_norm": 3.8575921058654785, + "learning_rate": 9.9772859977286e-06, + "loss": 0.0537, + "step": 22725 + }, + { + "epoch": 10.565318456531845, + "grad_norm": 1.5352314710617065, + "learning_rate": 9.9772849977285e-06, + "loss": 0.0165, + "step": 22726 + }, + { + "epoch": 10.565783356578336, + "grad_norm": 4.423789024353027, + "learning_rate": 9.977283997728401e-06, + "loss": 0.037, + "step": 22727 + }, + { + "epoch": 10.566248256624826, + "grad_norm": 2.0269408226013184, + "learning_rate": 9.9772829977283e-06, + "loss": 0.0222, + "step": 22728 + }, + { + "epoch": 10.566713156671316, + "grad_norm": 4.794368743896484, + "learning_rate": 9.977281997728201e-06, + "loss": 0.035, + "step": 22729 + }, + { + "epoch": 10.567178056717806, + "grad_norm": 9.548083305358887, + "learning_rate": 9.9772809977281e-06, + "loss": 0.1197, + "step": 22730 + }, + { + "epoch": 10.567642956764296, + "grad_norm": 5.3526740074157715, + "learning_rate": 9.977279997728e-06, + "loss": 0.1131, + "step": 22731 + }, + { + "epoch": 10.568107856810785, + "grad_norm": 9.663501739501953, + "learning_rate": 9.977278997727901e-06, + "loss": 0.2019, + "step": 22732 + }, + { + "epoch": 10.568572756857275, + "grad_norm": 6.302774429321289, + "learning_rate": 9.9772779977278e-06, + "loss": 0.1577, + "step": 22733 + }, + { + "epoch": 10.569037656903765, + "grad_norm": 5.436728000640869, + "learning_rate": 9.977276997727701e-06, + "loss": 0.1039, + "step": 22734 + }, + { + "epoch": 10.569502556950255, + "grad_norm": 3.6563773155212402, + "learning_rate": 9.9772759977276e-06, + "loss": 0.0509, + "step": 22735 + }, + { + "epoch": 10.569967456996746, + "grad_norm": 4.670798301696777, + "learning_rate": 9.9772749977275e-06, + "loss": 0.0547, + "step": 22736 + }, + { + "epoch": 10.570432357043236, + "grad_norm": 3.906984567642212, + "learning_rate": 9.9772739977274e-06, + "loss": 0.0251, + "step": 22737 + }, + { + "epoch": 10.570897257089726, + "grad_norm": 5.717990398406982, + "learning_rate": 9.9772729977273e-06, + "loss": 0.0537, + "step": 22738 + }, + { + "epoch": 10.571362157136216, + "grad_norm": 1.9139623641967773, + "learning_rate": 9.9772719977272e-06, + "loss": 0.0108, + "step": 22739 + }, + { + "epoch": 10.571827057182706, + "grad_norm": 5.846219062805176, + "learning_rate": 9.977270997727101e-06, + "loss": 0.1241, + "step": 22740 + }, + { + "epoch": 10.572291957229195, + "grad_norm": 7.165853023529053, + "learning_rate": 9.977269997727e-06, + "loss": 0.0904, + "step": 22741 + }, + { + "epoch": 10.572756857275685, + "grad_norm": 5.793680667877197, + "learning_rate": 9.977268997726901e-06, + "loss": 0.1031, + "step": 22742 + }, + { + "epoch": 10.573221757322175, + "grad_norm": 4.236849308013916, + "learning_rate": 9.9772679977268e-06, + "loss": 0.0396, + "step": 22743 + }, + { + "epoch": 10.573686657368667, + "grad_norm": 12.129884719848633, + "learning_rate": 9.9772669977267e-06, + "loss": 0.1927, + "step": 22744 + }, + { + "epoch": 10.574151557415156, + "grad_norm": 5.457406997680664, + "learning_rate": 9.977265997726601e-06, + "loss": 0.0922, + "step": 22745 + }, + { + "epoch": 10.574616457461646, + "grad_norm": 4.2192487716674805, + "learning_rate": 9.9772649977265e-06, + "loss": 0.0579, + "step": 22746 + }, + { + "epoch": 10.575081357508136, + "grad_norm": 6.366212844848633, + "learning_rate": 9.977263997726401e-06, + "loss": 0.0876, + "step": 22747 + }, + { + "epoch": 10.575546257554626, + "grad_norm": 1.4645745754241943, + "learning_rate": 9.9772629977263e-06, + "loss": 0.0148, + "step": 22748 + }, + { + "epoch": 10.576011157601116, + "grad_norm": 4.832663536071777, + "learning_rate": 9.9772619977262e-06, + "loss": 0.0977, + "step": 22749 + }, + { + "epoch": 10.576476057647605, + "grad_norm": 3.3447229862213135, + "learning_rate": 9.9772609977261e-06, + "loss": 0.0439, + "step": 22750 + }, + { + "epoch": 10.576940957694095, + "grad_norm": 3.598313093185425, + "learning_rate": 9.977259997726e-06, + "loss": 0.0446, + "step": 22751 + }, + { + "epoch": 10.577405857740585, + "grad_norm": 5.845542907714844, + "learning_rate": 9.977258997725901e-06, + "loss": 0.0785, + "step": 22752 + }, + { + "epoch": 10.577870757787077, + "grad_norm": 3.093207359313965, + "learning_rate": 9.977257997725802e-06, + "loss": 0.0212, + "step": 22753 + }, + { + "epoch": 10.578335657833566, + "grad_norm": 4.937476634979248, + "learning_rate": 9.9772569977257e-06, + "loss": 0.0869, + "step": 22754 + }, + { + "epoch": 10.578800557880056, + "grad_norm": 6.3532915115356445, + "learning_rate": 9.9772559977256e-06, + "loss": 0.1072, + "step": 22755 + }, + { + "epoch": 10.579265457926546, + "grad_norm": 6.150436878204346, + "learning_rate": 9.9772549977255e-06, + "loss": 0.1138, + "step": 22756 + }, + { + "epoch": 10.579730357973036, + "grad_norm": 6.41099214553833, + "learning_rate": 9.9772539977254e-06, + "loss": 0.0553, + "step": 22757 + }, + { + "epoch": 10.580195258019526, + "grad_norm": 7.412923812866211, + "learning_rate": 9.977252997725301e-06, + "loss": 0.098, + "step": 22758 + }, + { + "epoch": 10.580660158066015, + "grad_norm": 3.7658424377441406, + "learning_rate": 9.9772519977252e-06, + "loss": 0.0369, + "step": 22759 + }, + { + "epoch": 10.581125058112505, + "grad_norm": 4.605867862701416, + "learning_rate": 9.977250997725101e-06, + "loss": 0.0559, + "step": 22760 + }, + { + "epoch": 10.581589958158997, + "grad_norm": 5.696435451507568, + "learning_rate": 9.977249997725e-06, + "loss": 0.0462, + "step": 22761 + }, + { + "epoch": 10.582054858205487, + "grad_norm": 5.436406135559082, + "learning_rate": 9.9772489977249e-06, + "loss": 0.0846, + "step": 22762 + }, + { + "epoch": 10.582519758251976, + "grad_norm": 0.7639496922492981, + "learning_rate": 9.9772479977248e-06, + "loss": 0.005, + "step": 22763 + }, + { + "epoch": 10.582984658298466, + "grad_norm": 6.821920394897461, + "learning_rate": 9.9772469977247e-06, + "loss": 0.1124, + "step": 22764 + }, + { + "epoch": 10.583449558344956, + "grad_norm": 2.5194149017333984, + "learning_rate": 9.977245997724601e-06, + "loss": 0.0269, + "step": 22765 + }, + { + "epoch": 10.583914458391446, + "grad_norm": 2.9577438831329346, + "learning_rate": 9.9772449977245e-06, + "loss": 0.0388, + "step": 22766 + }, + { + "epoch": 10.584379358437936, + "grad_norm": 9.608611106872559, + "learning_rate": 9.9772439977244e-06, + "loss": 0.1681, + "step": 22767 + }, + { + "epoch": 10.584844258484425, + "grad_norm": 8.210862159729004, + "learning_rate": 9.9772429977243e-06, + "loss": 0.1083, + "step": 22768 + }, + { + "epoch": 10.585309158530915, + "grad_norm": 8.496537208557129, + "learning_rate": 9.9772419977242e-06, + "loss": 0.0455, + "step": 22769 + }, + { + "epoch": 10.585774058577407, + "grad_norm": 5.169001579284668, + "learning_rate": 9.9772409977241e-06, + "loss": 0.0947, + "step": 22770 + }, + { + "epoch": 10.586238958623897, + "grad_norm": 5.841897964477539, + "learning_rate": 9.977239997724002e-06, + "loss": 0.0936, + "step": 22771 + }, + { + "epoch": 10.586703858670386, + "grad_norm": 5.905508995056152, + "learning_rate": 9.9772389977239e-06, + "loss": 0.0463, + "step": 22772 + }, + { + "epoch": 10.587168758716876, + "grad_norm": 6.0647664070129395, + "learning_rate": 9.977237997723801e-06, + "loss": 0.0932, + "step": 22773 + }, + { + "epoch": 10.587633658763366, + "grad_norm": 5.506957530975342, + "learning_rate": 9.9772369977237e-06, + "loss": 0.0584, + "step": 22774 + }, + { + "epoch": 10.588098558809856, + "grad_norm": 1.7729405164718628, + "learning_rate": 9.9772359977236e-06, + "loss": 0.0157, + "step": 22775 + }, + { + "epoch": 10.588563458856346, + "grad_norm": 5.237789154052734, + "learning_rate": 9.977234997723501e-06, + "loss": 0.0938, + "step": 22776 + }, + { + "epoch": 10.589028358902835, + "grad_norm": 4.7609333992004395, + "learning_rate": 9.9772339977234e-06, + "loss": 0.0569, + "step": 22777 + }, + { + "epoch": 10.589493258949325, + "grad_norm": 2.629503011703491, + "learning_rate": 9.977232997723301e-06, + "loss": 0.0374, + "step": 22778 + }, + { + "epoch": 10.589958158995817, + "grad_norm": 6.2547221183776855, + "learning_rate": 9.9772319977232e-06, + "loss": 0.1055, + "step": 22779 + }, + { + "epoch": 10.590423059042307, + "grad_norm": 3.205803871154785, + "learning_rate": 9.9772309977231e-06, + "loss": 0.0281, + "step": 22780 + }, + { + "epoch": 10.590887959088796, + "grad_norm": 2.7809908390045166, + "learning_rate": 9.977229997723e-06, + "loss": 0.02, + "step": 22781 + }, + { + "epoch": 10.591352859135286, + "grad_norm": 8.709522247314453, + "learning_rate": 9.9772289977229e-06, + "loss": 0.0514, + "step": 22782 + }, + { + "epoch": 10.591817759181776, + "grad_norm": 2.7202911376953125, + "learning_rate": 9.977227997722801e-06, + "loss": 0.0324, + "step": 22783 + }, + { + "epoch": 10.592282659228266, + "grad_norm": 2.8897199630737305, + "learning_rate": 9.977226997722702e-06, + "loss": 0.0313, + "step": 22784 + }, + { + "epoch": 10.592747559274756, + "grad_norm": 5.2934489250183105, + "learning_rate": 9.9772259977226e-06, + "loss": 0.0994, + "step": 22785 + }, + { + "epoch": 10.593212459321245, + "grad_norm": 6.304070949554443, + "learning_rate": 9.977224997722501e-06, + "loss": 0.0973, + "step": 22786 + }, + { + "epoch": 10.593677359367735, + "grad_norm": 1.8552253246307373, + "learning_rate": 9.9772239977224e-06, + "loss": 0.0164, + "step": 22787 + }, + { + "epoch": 10.594142259414227, + "grad_norm": 2.7735495567321777, + "learning_rate": 9.9772229977223e-06, + "loss": 0.0299, + "step": 22788 + }, + { + "epoch": 10.594607159460717, + "grad_norm": 2.9054317474365234, + "learning_rate": 9.977221997722201e-06, + "loss": 0.0316, + "step": 22789 + }, + { + "epoch": 10.595072059507206, + "grad_norm": 5.416401386260986, + "learning_rate": 9.9772209977221e-06, + "loss": 0.0769, + "step": 22790 + }, + { + "epoch": 10.595536959553696, + "grad_norm": 2.6327264308929443, + "learning_rate": 9.977219997722001e-06, + "loss": 0.0232, + "step": 22791 + }, + { + "epoch": 10.596001859600186, + "grad_norm": 2.339606523513794, + "learning_rate": 9.9772189977219e-06, + "loss": 0.0145, + "step": 22792 + }, + { + "epoch": 10.596466759646676, + "grad_norm": 0.831251859664917, + "learning_rate": 9.9772179977218e-06, + "loss": 0.0058, + "step": 22793 + }, + { + "epoch": 10.596931659693166, + "grad_norm": 8.068071365356445, + "learning_rate": 9.9772169977217e-06, + "loss": 0.0727, + "step": 22794 + }, + { + "epoch": 10.597396559739655, + "grad_norm": 6.75523042678833, + "learning_rate": 9.9772159977216e-06, + "loss": 0.1513, + "step": 22795 + }, + { + "epoch": 10.597861459786145, + "grad_norm": 4.280444622039795, + "learning_rate": 9.977214997721501e-06, + "loss": 0.0531, + "step": 22796 + }, + { + "epoch": 10.598326359832637, + "grad_norm": 3.2349414825439453, + "learning_rate": 9.977213997721402e-06, + "loss": 0.025, + "step": 22797 + }, + { + "epoch": 10.598791259879127, + "grad_norm": 6.522146224975586, + "learning_rate": 9.9772129977213e-06, + "loss": 0.1194, + "step": 22798 + }, + { + "epoch": 10.599256159925616, + "grad_norm": 3.825274705886841, + "learning_rate": 9.9772119977212e-06, + "loss": 0.0452, + "step": 22799 + }, + { + "epoch": 10.599721059972106, + "grad_norm": 7.528687477111816, + "learning_rate": 9.9772109977211e-06, + "loss": 0.1468, + "step": 22800 + }, + { + "epoch": 10.600185960018596, + "grad_norm": 6.3411030769348145, + "learning_rate": 9.977209997721001e-06, + "loss": 0.0464, + "step": 22801 + }, + { + "epoch": 10.600650860065086, + "grad_norm": 2.0481715202331543, + "learning_rate": 9.977208997720902e-06, + "loss": 0.0233, + "step": 22802 + }, + { + "epoch": 10.601115760111576, + "grad_norm": 4.873271942138672, + "learning_rate": 9.9772079977208e-06, + "loss": 0.1126, + "step": 22803 + }, + { + "epoch": 10.601580660158065, + "grad_norm": 3.221909284591675, + "learning_rate": 9.977206997720701e-06, + "loss": 0.0486, + "step": 22804 + }, + { + "epoch": 10.602045560204555, + "grad_norm": 5.068111419677734, + "learning_rate": 9.9772059977206e-06, + "loss": 0.1, + "step": 22805 + }, + { + "epoch": 10.602510460251047, + "grad_norm": 7.656494617462158, + "learning_rate": 9.9772049977205e-06, + "loss": 0.1022, + "step": 22806 + }, + { + "epoch": 10.602975360297537, + "grad_norm": 1.8886609077453613, + "learning_rate": 9.9772039977204e-06, + "loss": 0.0101, + "step": 22807 + }, + { + "epoch": 10.603440260344026, + "grad_norm": 5.377890110015869, + "learning_rate": 9.9772029977203e-06, + "loss": 0.0654, + "step": 22808 + }, + { + "epoch": 10.603905160390516, + "grad_norm": 3.6953084468841553, + "learning_rate": 9.977201997720201e-06, + "loss": 0.0695, + "step": 22809 + }, + { + "epoch": 10.604370060437006, + "grad_norm": 5.591876029968262, + "learning_rate": 9.9772009977201e-06, + "loss": 0.1257, + "step": 22810 + }, + { + "epoch": 10.604834960483496, + "grad_norm": 1.7692022323608398, + "learning_rate": 9.97719999772e-06, + "loss": 0.0146, + "step": 22811 + }, + { + "epoch": 10.605299860529986, + "grad_norm": 4.183372974395752, + "learning_rate": 9.9771989977199e-06, + "loss": 0.0522, + "step": 22812 + }, + { + "epoch": 10.605764760576475, + "grad_norm": 6.502434253692627, + "learning_rate": 9.9771979977198e-06, + "loss": 0.1361, + "step": 22813 + }, + { + "epoch": 10.606229660622965, + "grad_norm": 4.880478858947754, + "learning_rate": 9.977196997719701e-06, + "loss": 0.0616, + "step": 22814 + }, + { + "epoch": 10.606694560669457, + "grad_norm": 6.492509841918945, + "learning_rate": 9.977195997719602e-06, + "loss": 0.1097, + "step": 22815 + }, + { + "epoch": 10.607159460715947, + "grad_norm": 8.061657905578613, + "learning_rate": 9.9771949977195e-06, + "loss": 0.2352, + "step": 22816 + }, + { + "epoch": 10.607624360762436, + "grad_norm": 4.164923191070557, + "learning_rate": 9.977193997719401e-06, + "loss": 0.0304, + "step": 22817 + }, + { + "epoch": 10.608089260808926, + "grad_norm": 5.185154914855957, + "learning_rate": 9.9771929977193e-06, + "loss": 0.0743, + "step": 22818 + }, + { + "epoch": 10.608554160855416, + "grad_norm": 3.2118825912475586, + "learning_rate": 9.977191997719201e-06, + "loss": 0.0445, + "step": 22819 + }, + { + "epoch": 10.609019060901906, + "grad_norm": 1.348143458366394, + "learning_rate": 9.9771909977191e-06, + "loss": 0.0117, + "step": 22820 + }, + { + "epoch": 10.609483960948396, + "grad_norm": 3.279879093170166, + "learning_rate": 9.977189997719e-06, + "loss": 0.031, + "step": 22821 + }, + { + "epoch": 10.609948860994885, + "grad_norm": 3.2897872924804688, + "learning_rate": 9.977188997718901e-06, + "loss": 0.0282, + "step": 22822 + }, + { + "epoch": 10.610413761041377, + "grad_norm": 4.945242881774902, + "learning_rate": 9.9771879977188e-06, + "loss": 0.0671, + "step": 22823 + }, + { + "epoch": 10.610878661087867, + "grad_norm": 5.9875993728637695, + "learning_rate": 9.9771869977187e-06, + "loss": 0.0558, + "step": 22824 + }, + { + "epoch": 10.611343561134357, + "grad_norm": 4.509525775909424, + "learning_rate": 9.9771859977186e-06, + "loss": 0.0554, + "step": 22825 + }, + { + "epoch": 10.611808461180846, + "grad_norm": 3.970306873321533, + "learning_rate": 9.9771849977185e-06, + "loss": 0.0452, + "step": 22826 + }, + { + "epoch": 10.612273361227336, + "grad_norm": 3.414351224899292, + "learning_rate": 9.977183997718401e-06, + "loss": 0.0347, + "step": 22827 + }, + { + "epoch": 10.612738261273826, + "grad_norm": 10.158236503601074, + "learning_rate": 9.977182997718302e-06, + "loss": 0.1565, + "step": 22828 + }, + { + "epoch": 10.613203161320316, + "grad_norm": 4.423817157745361, + "learning_rate": 9.9771819977182e-06, + "loss": 0.0611, + "step": 22829 + }, + { + "epoch": 10.613668061366806, + "grad_norm": 3.7610225677490234, + "learning_rate": 9.977180997718101e-06, + "loss": 0.0329, + "step": 22830 + }, + { + "epoch": 10.614132961413295, + "grad_norm": 3.332693338394165, + "learning_rate": 9.977179997718e-06, + "loss": 0.0357, + "step": 22831 + }, + { + "epoch": 10.614597861459787, + "grad_norm": 6.609009265899658, + "learning_rate": 9.9771789977179e-06, + "loss": 0.1607, + "step": 22832 + }, + { + "epoch": 10.615062761506277, + "grad_norm": 6.423935413360596, + "learning_rate": 9.9771779977178e-06, + "loss": 0.1142, + "step": 22833 + }, + { + "epoch": 10.615527661552767, + "grad_norm": 3.4152262210845947, + "learning_rate": 9.9771769977177e-06, + "loss": 0.0425, + "step": 22834 + }, + { + "epoch": 10.615992561599256, + "grad_norm": 5.165721893310547, + "learning_rate": 9.977175997717601e-06, + "loss": 0.0523, + "step": 22835 + }, + { + "epoch": 10.616457461645746, + "grad_norm": 3.0881237983703613, + "learning_rate": 9.9771749977175e-06, + "loss": 0.0512, + "step": 22836 + }, + { + "epoch": 10.616922361692236, + "grad_norm": 3.4162752628326416, + "learning_rate": 9.977173997717401e-06, + "loss": 0.0385, + "step": 22837 + }, + { + "epoch": 10.617387261738726, + "grad_norm": 10.071002006530762, + "learning_rate": 9.9771729977173e-06, + "loss": 0.3134, + "step": 22838 + }, + { + "epoch": 10.617852161785216, + "grad_norm": 2.0250813961029053, + "learning_rate": 9.9771719977172e-06, + "loss": 0.0148, + "step": 22839 + }, + { + "epoch": 10.618317061831707, + "grad_norm": 4.142244815826416, + "learning_rate": 9.977170997717101e-06, + "loss": 0.0693, + "step": 22840 + }, + { + "epoch": 10.618781961878197, + "grad_norm": 7.1340765953063965, + "learning_rate": 9.977169997717002e-06, + "loss": 0.116, + "step": 22841 + }, + { + "epoch": 10.619246861924687, + "grad_norm": 2.339979887008667, + "learning_rate": 9.9771689977169e-06, + "loss": 0.0367, + "step": 22842 + }, + { + "epoch": 10.619711761971177, + "grad_norm": 8.8374605178833, + "learning_rate": 9.9771679977168e-06, + "loss": 0.1657, + "step": 22843 + }, + { + "epoch": 10.620176662017666, + "grad_norm": 7.674012660980225, + "learning_rate": 9.9771669977167e-06, + "loss": 0.2213, + "step": 22844 + }, + { + "epoch": 10.620641562064156, + "grad_norm": 5.154200553894043, + "learning_rate": 9.977165997716601e-06, + "loss": 0.0734, + "step": 22845 + }, + { + "epoch": 10.621106462110646, + "grad_norm": 3.6106700897216797, + "learning_rate": 9.9771649977165e-06, + "loss": 0.0767, + "step": 22846 + }, + { + "epoch": 10.621571362157136, + "grad_norm": 6.241075038909912, + "learning_rate": 9.9771639977164e-06, + "loss": 0.0746, + "step": 22847 + }, + { + "epoch": 10.622036262203626, + "grad_norm": 7.201013565063477, + "learning_rate": 9.977162997716301e-06, + "loss": 0.0453, + "step": 22848 + }, + { + "epoch": 10.622501162250117, + "grad_norm": 6.167176246643066, + "learning_rate": 9.9771619977162e-06, + "loss": 0.1155, + "step": 22849 + }, + { + "epoch": 10.622966062296607, + "grad_norm": 5.230909824371338, + "learning_rate": 9.977160997716101e-06, + "loss": 0.0964, + "step": 22850 + }, + { + "epoch": 10.623430962343097, + "grad_norm": 6.072365760803223, + "learning_rate": 9.977159997716e-06, + "loss": 0.1405, + "step": 22851 + }, + { + "epoch": 10.623895862389586, + "grad_norm": 3.164971113204956, + "learning_rate": 9.9771589977159e-06, + "loss": 0.0425, + "step": 22852 + }, + { + "epoch": 10.624360762436076, + "grad_norm": 5.046878814697266, + "learning_rate": 9.977157997715801e-06, + "loss": 0.1466, + "step": 22853 + }, + { + "epoch": 10.624825662482566, + "grad_norm": 3.680126667022705, + "learning_rate": 9.9771569977157e-06, + "loss": 0.0567, + "step": 22854 + }, + { + "epoch": 10.625290562529056, + "grad_norm": 4.559530258178711, + "learning_rate": 9.977155997715601e-06, + "loss": 0.0884, + "step": 22855 + }, + { + "epoch": 10.625755462575546, + "grad_norm": 1.178399682044983, + "learning_rate": 9.9771549977155e-06, + "loss": 0.0131, + "step": 22856 + }, + { + "epoch": 10.626220362622036, + "grad_norm": 5.302659511566162, + "learning_rate": 9.9771539977154e-06, + "loss": 0.0972, + "step": 22857 + }, + { + "epoch": 10.626685262668527, + "grad_norm": 3.8500051498413086, + "learning_rate": 9.977152997715301e-06, + "loss": 0.0407, + "step": 22858 + }, + { + "epoch": 10.627150162715017, + "grad_norm": 2.6586227416992188, + "learning_rate": 9.9771519977152e-06, + "loss": 0.0329, + "step": 22859 + }, + { + "epoch": 10.627615062761507, + "grad_norm": 3.79302978515625, + "learning_rate": 9.9771509977151e-06, + "loss": 0.0476, + "step": 22860 + }, + { + "epoch": 10.628079962807996, + "grad_norm": 2.126830577850342, + "learning_rate": 9.977149997715001e-06, + "loss": 0.0248, + "step": 22861 + }, + { + "epoch": 10.628544862854486, + "grad_norm": 4.769711971282959, + "learning_rate": 9.9771489977149e-06, + "loss": 0.046, + "step": 22862 + }, + { + "epoch": 10.629009762900976, + "grad_norm": 3.558504819869995, + "learning_rate": 9.977147997714801e-06, + "loss": 0.0391, + "step": 22863 + }, + { + "epoch": 10.629474662947466, + "grad_norm": 4.286270618438721, + "learning_rate": 9.9771469977147e-06, + "loss": 0.0775, + "step": 22864 + }, + { + "epoch": 10.629939562993956, + "grad_norm": 5.114658832550049, + "learning_rate": 9.9771459977146e-06, + "loss": 0.0422, + "step": 22865 + }, + { + "epoch": 10.630404463040446, + "grad_norm": 3.1274259090423584, + "learning_rate": 9.977144997714501e-06, + "loss": 0.0334, + "step": 22866 + }, + { + "epoch": 10.630869363086937, + "grad_norm": 3.34977650642395, + "learning_rate": 9.9771439977144e-06, + "loss": 0.0379, + "step": 22867 + }, + { + "epoch": 10.631334263133427, + "grad_norm": 6.740087509155273, + "learning_rate": 9.977142997714301e-06, + "loss": 0.0907, + "step": 22868 + }, + { + "epoch": 10.631799163179917, + "grad_norm": 4.0382561683654785, + "learning_rate": 9.9771419977142e-06, + "loss": 0.0328, + "step": 22869 + }, + { + "epoch": 10.632264063226406, + "grad_norm": 6.329133987426758, + "learning_rate": 9.9771409977141e-06, + "loss": 0.0993, + "step": 22870 + }, + { + "epoch": 10.632728963272896, + "grad_norm": 4.803959846496582, + "learning_rate": 9.977139997714001e-06, + "loss": 0.0832, + "step": 22871 + }, + { + "epoch": 10.633193863319386, + "grad_norm": 6.7181477546691895, + "learning_rate": 9.977138997713902e-06, + "loss": 0.1123, + "step": 22872 + }, + { + "epoch": 10.633658763365876, + "grad_norm": 6.605010986328125, + "learning_rate": 9.9771379977138e-06, + "loss": 0.1489, + "step": 22873 + }, + { + "epoch": 10.634123663412366, + "grad_norm": 6.964249134063721, + "learning_rate": 9.9771369977137e-06, + "loss": 0.1438, + "step": 22874 + }, + { + "epoch": 10.634588563458856, + "grad_norm": 3.5701346397399902, + "learning_rate": 9.9771359977136e-06, + "loss": 0.0355, + "step": 22875 + }, + { + "epoch": 10.635053463505347, + "grad_norm": 4.189579963684082, + "learning_rate": 9.9771349977135e-06, + "loss": 0.0533, + "step": 22876 + }, + { + "epoch": 10.635518363551837, + "grad_norm": 4.127927303314209, + "learning_rate": 9.9771339977134e-06, + "loss": 0.0395, + "step": 22877 + }, + { + "epoch": 10.635983263598327, + "grad_norm": 2.003147840499878, + "learning_rate": 9.9771329977133e-06, + "loss": 0.0215, + "step": 22878 + }, + { + "epoch": 10.636448163644816, + "grad_norm": 2.700744152069092, + "learning_rate": 9.977131997713201e-06, + "loss": 0.0323, + "step": 22879 + }, + { + "epoch": 10.636913063691306, + "grad_norm": 2.8378655910491943, + "learning_rate": 9.9771309977131e-06, + "loss": 0.0312, + "step": 22880 + }, + { + "epoch": 10.637377963737796, + "grad_norm": 6.692045211791992, + "learning_rate": 9.977129997713001e-06, + "loss": 0.0504, + "step": 22881 + }, + { + "epoch": 10.637842863784286, + "grad_norm": 1.2660152912139893, + "learning_rate": 9.9771289977129e-06, + "loss": 0.0115, + "step": 22882 + }, + { + "epoch": 10.638307763830776, + "grad_norm": 3.464895725250244, + "learning_rate": 9.9771279977128e-06, + "loss": 0.0386, + "step": 22883 + }, + { + "epoch": 10.638772663877266, + "grad_norm": 3.514958381652832, + "learning_rate": 9.977126997712701e-06, + "loss": 0.0375, + "step": 22884 + }, + { + "epoch": 10.639237563923757, + "grad_norm": 3.2403714656829834, + "learning_rate": 9.9771259977126e-06, + "loss": 0.0361, + "step": 22885 + }, + { + "epoch": 10.639702463970247, + "grad_norm": 4.900225639343262, + "learning_rate": 9.977124997712501e-06, + "loss": 0.1157, + "step": 22886 + }, + { + "epoch": 10.640167364016737, + "grad_norm": 4.888001918792725, + "learning_rate": 9.9771239977124e-06, + "loss": 0.0612, + "step": 22887 + }, + { + "epoch": 10.640632264063226, + "grad_norm": 11.582600593566895, + "learning_rate": 9.9771229977123e-06, + "loss": 0.1227, + "step": 22888 + }, + { + "epoch": 10.641097164109716, + "grad_norm": 2.4365382194519043, + "learning_rate": 9.9771219977122e-06, + "loss": 0.0222, + "step": 22889 + }, + { + "epoch": 10.641562064156206, + "grad_norm": 8.692211151123047, + "learning_rate": 9.9771209977121e-06, + "loss": 0.1214, + "step": 22890 + }, + { + "epoch": 10.642026964202696, + "grad_norm": 5.279953956604004, + "learning_rate": 9.977119997712e-06, + "loss": 0.0681, + "step": 22891 + }, + { + "epoch": 10.642491864249186, + "grad_norm": 6.669711589813232, + "learning_rate": 9.977118997711901e-06, + "loss": 0.1585, + "step": 22892 + }, + { + "epoch": 10.642956764295675, + "grad_norm": 1.4280420541763306, + "learning_rate": 9.9771179977118e-06, + "loss": 0.0181, + "step": 22893 + }, + { + "epoch": 10.643421664342167, + "grad_norm": 5.15358304977417, + "learning_rate": 9.977116997711701e-06, + "loss": 0.0823, + "step": 22894 + }, + { + "epoch": 10.643886564388657, + "grad_norm": 8.290531158447266, + "learning_rate": 9.9771159977116e-06, + "loss": 0.1076, + "step": 22895 + }, + { + "epoch": 10.644351464435147, + "grad_norm": 4.837099552154541, + "learning_rate": 9.9771149977115e-06, + "loss": 0.0574, + "step": 22896 + }, + { + "epoch": 10.644816364481636, + "grad_norm": 4.7859883308410645, + "learning_rate": 9.977113997711401e-06, + "loss": 0.046, + "step": 22897 + }, + { + "epoch": 10.645281264528126, + "grad_norm": 7.122119903564453, + "learning_rate": 9.9771129977113e-06, + "loss": 0.1534, + "step": 22898 + }, + { + "epoch": 10.645746164574616, + "grad_norm": 5.4356689453125, + "learning_rate": 9.977111997711201e-06, + "loss": 0.0727, + "step": 22899 + }, + { + "epoch": 10.646211064621106, + "grad_norm": 2.002969264984131, + "learning_rate": 9.9771109977111e-06, + "loss": 0.0203, + "step": 22900 + }, + { + "epoch": 10.646675964667596, + "grad_norm": 4.67442512512207, + "learning_rate": 9.977109997711e-06, + "loss": 0.0372, + "step": 22901 + }, + { + "epoch": 10.647140864714087, + "grad_norm": 3.7614080905914307, + "learning_rate": 9.9771089977109e-06, + "loss": 0.0391, + "step": 22902 + }, + { + "epoch": 10.647605764760577, + "grad_norm": 1.5462745428085327, + "learning_rate": 9.9771079977108e-06, + "loss": 0.0132, + "step": 22903 + }, + { + "epoch": 10.648070664807067, + "grad_norm": 6.287144660949707, + "learning_rate": 9.977106997710701e-06, + "loss": 0.0618, + "step": 22904 + }, + { + "epoch": 10.648535564853557, + "grad_norm": 5.941581726074219, + "learning_rate": 9.977105997710602e-06, + "loss": 0.0398, + "step": 22905 + }, + { + "epoch": 10.649000464900046, + "grad_norm": 4.846466064453125, + "learning_rate": 9.9771049977105e-06, + "loss": 0.0847, + "step": 22906 + }, + { + "epoch": 10.649465364946536, + "grad_norm": 5.231109619140625, + "learning_rate": 9.9771039977104e-06, + "loss": 0.074, + "step": 22907 + }, + { + "epoch": 10.649930264993026, + "grad_norm": 5.294482231140137, + "learning_rate": 9.9771029977103e-06, + "loss": 0.0515, + "step": 22908 + }, + { + "epoch": 10.650395165039516, + "grad_norm": 4.464739799499512, + "learning_rate": 9.9771019977102e-06, + "loss": 0.0936, + "step": 22909 + }, + { + "epoch": 10.650860065086006, + "grad_norm": 8.710099220275879, + "learning_rate": 9.977100997710101e-06, + "loss": 0.0813, + "step": 22910 + }, + { + "epoch": 10.651324965132497, + "grad_norm": 3.022066116333008, + "learning_rate": 9.97709999771e-06, + "loss": 0.0273, + "step": 22911 + }, + { + "epoch": 10.651789865178987, + "grad_norm": 4.514498710632324, + "learning_rate": 9.977098997709901e-06, + "loss": 0.0666, + "step": 22912 + }, + { + "epoch": 10.652254765225477, + "grad_norm": 4.193394660949707, + "learning_rate": 9.9770979977098e-06, + "loss": 0.087, + "step": 22913 + }, + { + "epoch": 10.652719665271967, + "grad_norm": 6.7686848640441895, + "learning_rate": 9.9770969977097e-06, + "loss": 0.1501, + "step": 22914 + }, + { + "epoch": 10.653184565318456, + "grad_norm": 3.8492774963378906, + "learning_rate": 9.9770959977096e-06, + "loss": 0.0403, + "step": 22915 + }, + { + "epoch": 10.653649465364946, + "grad_norm": 5.525205135345459, + "learning_rate": 9.9770949977095e-06, + "loss": 0.0571, + "step": 22916 + }, + { + "epoch": 10.654114365411436, + "grad_norm": 4.0918707847595215, + "learning_rate": 9.977093997709401e-06, + "loss": 0.0698, + "step": 22917 + }, + { + "epoch": 10.654579265457926, + "grad_norm": 8.449140548706055, + "learning_rate": 9.9770929977093e-06, + "loss": 0.0842, + "step": 22918 + }, + { + "epoch": 10.655044165504417, + "grad_norm": 0.5575900673866272, + "learning_rate": 9.9770919977092e-06, + "loss": 0.0022, + "step": 22919 + }, + { + "epoch": 10.655509065550907, + "grad_norm": 1.7924139499664307, + "learning_rate": 9.9770909977091e-06, + "loss": 0.0187, + "step": 22920 + }, + { + "epoch": 10.655973965597397, + "grad_norm": 5.570494651794434, + "learning_rate": 9.977089997709e-06, + "loss": 0.1374, + "step": 22921 + }, + { + "epoch": 10.656438865643887, + "grad_norm": 2.3898980617523193, + "learning_rate": 9.977088997708901e-06, + "loss": 0.0237, + "step": 22922 + }, + { + "epoch": 10.656903765690377, + "grad_norm": 5.016604900360107, + "learning_rate": 9.977087997708802e-06, + "loss": 0.0934, + "step": 22923 + }, + { + "epoch": 10.657368665736866, + "grad_norm": 6.18584680557251, + "learning_rate": 9.9770869977087e-06, + "loss": 0.139, + "step": 22924 + }, + { + "epoch": 10.657833565783356, + "grad_norm": 4.002445220947266, + "learning_rate": 9.977085997708601e-06, + "loss": 0.0593, + "step": 22925 + }, + { + "epoch": 10.658298465829846, + "grad_norm": 3.2752904891967773, + "learning_rate": 9.9770849977085e-06, + "loss": 0.0257, + "step": 22926 + }, + { + "epoch": 10.658763365876336, + "grad_norm": 7.685663223266602, + "learning_rate": 9.9770839977084e-06, + "loss": 0.2341, + "step": 22927 + }, + { + "epoch": 10.659228265922827, + "grad_norm": 1.497010588645935, + "learning_rate": 9.9770829977083e-06, + "loss": 0.0165, + "step": 22928 + }, + { + "epoch": 10.659693165969317, + "grad_norm": 4.710701942443848, + "learning_rate": 9.9770819977082e-06, + "loss": 0.0886, + "step": 22929 + }, + { + "epoch": 10.660158066015807, + "grad_norm": 5.6206583976745605, + "learning_rate": 9.977080997708101e-06, + "loss": 0.095, + "step": 22930 + }, + { + "epoch": 10.660622966062297, + "grad_norm": 6.153391361236572, + "learning_rate": 9.977079997708e-06, + "loss": 0.1467, + "step": 22931 + }, + { + "epoch": 10.661087866108787, + "grad_norm": 5.820449352264404, + "learning_rate": 9.9770789977079e-06, + "loss": 0.1424, + "step": 22932 + }, + { + "epoch": 10.661552766155276, + "grad_norm": 4.408016204833984, + "learning_rate": 9.9770779977078e-06, + "loss": 0.0486, + "step": 22933 + }, + { + "epoch": 10.662017666201766, + "grad_norm": 5.820084571838379, + "learning_rate": 9.9770769977077e-06, + "loss": 0.0562, + "step": 22934 + }, + { + "epoch": 10.662482566248256, + "grad_norm": 2.8029165267944336, + "learning_rate": 9.977075997707601e-06, + "loss": 0.0374, + "step": 22935 + }, + { + "epoch": 10.662947466294746, + "grad_norm": 7.423979759216309, + "learning_rate": 9.977074997707502e-06, + "loss": 0.1895, + "step": 22936 + }, + { + "epoch": 10.663412366341237, + "grad_norm": 4.534229755401611, + "learning_rate": 9.9770739977074e-06, + "loss": 0.0706, + "step": 22937 + }, + { + "epoch": 10.663877266387727, + "grad_norm": 2.5361642837524414, + "learning_rate": 9.977072997707301e-06, + "loss": 0.0302, + "step": 22938 + }, + { + "epoch": 10.664342166434217, + "grad_norm": 3.332629919052124, + "learning_rate": 9.9770719977072e-06, + "loss": 0.0654, + "step": 22939 + }, + { + "epoch": 10.664807066480707, + "grad_norm": 4.983631134033203, + "learning_rate": 9.977070997707101e-06, + "loss": 0.1103, + "step": 22940 + }, + { + "epoch": 10.665271966527197, + "grad_norm": 2.199678897857666, + "learning_rate": 9.977069997707002e-06, + "loss": 0.019, + "step": 22941 + }, + { + "epoch": 10.665736866573686, + "grad_norm": 4.35161828994751, + "learning_rate": 9.9770689977069e-06, + "loss": 0.0644, + "step": 22942 + }, + { + "epoch": 10.666201766620176, + "grad_norm": 4.72423791885376, + "learning_rate": 9.977067997706801e-06, + "loss": 0.0576, + "step": 22943 + }, + { + "epoch": 10.666666666666666, + "grad_norm": 1.779213786125183, + "learning_rate": 9.9770669977067e-06, + "loss": 0.0171, + "step": 22944 + }, + { + "epoch": 10.667131566713156, + "grad_norm": 1.7975643873214722, + "learning_rate": 9.9770659977066e-06, + "loss": 0.0256, + "step": 22945 + }, + { + "epoch": 10.667596466759647, + "grad_norm": 2.7850613594055176, + "learning_rate": 9.9770649977065e-06, + "loss": 0.0257, + "step": 22946 + }, + { + "epoch": 10.668061366806137, + "grad_norm": 4.87289571762085, + "learning_rate": 9.9770639977064e-06, + "loss": 0.0911, + "step": 22947 + }, + { + "epoch": 10.668526266852627, + "grad_norm": 2.0751802921295166, + "learning_rate": 9.977062997706301e-06, + "loss": 0.014, + "step": 22948 + }, + { + "epoch": 10.668991166899117, + "grad_norm": 4.414037227630615, + "learning_rate": 9.977061997706202e-06, + "loss": 0.0472, + "step": 22949 + }, + { + "epoch": 10.669456066945607, + "grad_norm": 2.9325098991394043, + "learning_rate": 9.9770609977061e-06, + "loss": 0.0233, + "step": 22950 + }, + { + "epoch": 10.669920966992096, + "grad_norm": 4.849343299865723, + "learning_rate": 9.977059997706e-06, + "loss": 0.0749, + "step": 22951 + }, + { + "epoch": 10.670385867038586, + "grad_norm": 6.435194492340088, + "learning_rate": 9.9770589977059e-06, + "loss": 0.1627, + "step": 22952 + }, + { + "epoch": 10.670850767085076, + "grad_norm": 2.6958489418029785, + "learning_rate": 9.977057997705801e-06, + "loss": 0.0281, + "step": 22953 + }, + { + "epoch": 10.671315667131566, + "grad_norm": 4.191440105438232, + "learning_rate": 9.977056997705702e-06, + "loss": 0.0391, + "step": 22954 + }, + { + "epoch": 10.671780567178057, + "grad_norm": 4.747952938079834, + "learning_rate": 9.9770559977056e-06, + "loss": 0.0435, + "step": 22955 + }, + { + "epoch": 10.672245467224547, + "grad_norm": 6.261482238769531, + "learning_rate": 9.977054997705501e-06, + "loss": 0.1268, + "step": 22956 + }, + { + "epoch": 10.672710367271037, + "grad_norm": 5.858582019805908, + "learning_rate": 9.9770539977054e-06, + "loss": 0.1101, + "step": 22957 + }, + { + "epoch": 10.673175267317527, + "grad_norm": 3.4945363998413086, + "learning_rate": 9.9770529977053e-06, + "loss": 0.0453, + "step": 22958 + }, + { + "epoch": 10.673640167364017, + "grad_norm": 11.324457168579102, + "learning_rate": 9.9770519977052e-06, + "loss": 0.3668, + "step": 22959 + }, + { + "epoch": 10.674105067410506, + "grad_norm": 3.993489980697632, + "learning_rate": 9.9770509977051e-06, + "loss": 0.0476, + "step": 22960 + }, + { + "epoch": 10.674569967456996, + "grad_norm": 3.7586491107940674, + "learning_rate": 9.977049997705001e-06, + "loss": 0.0377, + "step": 22961 + }, + { + "epoch": 10.675034867503486, + "grad_norm": 5.892847061157227, + "learning_rate": 9.9770489977049e-06, + "loss": 0.105, + "step": 22962 + }, + { + "epoch": 10.675499767549976, + "grad_norm": 3.299520254135132, + "learning_rate": 9.9770479977048e-06, + "loss": 0.0402, + "step": 22963 + }, + { + "epoch": 10.675964667596467, + "grad_norm": 2.7707455158233643, + "learning_rate": 9.9770469977047e-06, + "loss": 0.0257, + "step": 22964 + }, + { + "epoch": 10.676429567642957, + "grad_norm": 2.981818437576294, + "learning_rate": 9.9770459977046e-06, + "loss": 0.0327, + "step": 22965 + }, + { + "epoch": 10.676894467689447, + "grad_norm": 3.309020519256592, + "learning_rate": 9.977044997704501e-06, + "loss": 0.03, + "step": 22966 + }, + { + "epoch": 10.677359367735937, + "grad_norm": 4.063387393951416, + "learning_rate": 9.977043997704402e-06, + "loss": 0.0446, + "step": 22967 + }, + { + "epoch": 10.677824267782427, + "grad_norm": 4.9782819747924805, + "learning_rate": 9.9770429977043e-06, + "loss": 0.0847, + "step": 22968 + }, + { + "epoch": 10.678289167828916, + "grad_norm": 4.657717227935791, + "learning_rate": 9.977041997704201e-06, + "loss": 0.0602, + "step": 22969 + }, + { + "epoch": 10.678754067875406, + "grad_norm": 3.9044198989868164, + "learning_rate": 9.9770409977041e-06, + "loss": 0.0328, + "step": 22970 + }, + { + "epoch": 10.679218967921896, + "grad_norm": 7.465451240539551, + "learning_rate": 9.977039997704001e-06, + "loss": 0.2305, + "step": 22971 + }, + { + "epoch": 10.679683867968386, + "grad_norm": 2.5534708499908447, + "learning_rate": 9.9770389977039e-06, + "loss": 0.0169, + "step": 22972 + }, + { + "epoch": 10.680148768014877, + "grad_norm": 7.348380088806152, + "learning_rate": 9.9770379977038e-06, + "loss": 0.0953, + "step": 22973 + }, + { + "epoch": 10.680613668061367, + "grad_norm": 5.99804162979126, + "learning_rate": 9.977036997703701e-06, + "loss": 0.0597, + "step": 22974 + }, + { + "epoch": 10.681078568107857, + "grad_norm": 1.2048135995864868, + "learning_rate": 9.9770359977036e-06, + "loss": 0.0147, + "step": 22975 + }, + { + "epoch": 10.681543468154347, + "grad_norm": 4.105268478393555, + "learning_rate": 9.9770349977035e-06, + "loss": 0.0679, + "step": 22976 + }, + { + "epoch": 10.682008368200837, + "grad_norm": 3.802002429962158, + "learning_rate": 9.9770339977034e-06, + "loss": 0.0297, + "step": 22977 + }, + { + "epoch": 10.682473268247326, + "grad_norm": 7.376128673553467, + "learning_rate": 9.9770329977033e-06, + "loss": 0.1734, + "step": 22978 + }, + { + "epoch": 10.682938168293816, + "grad_norm": 4.0174407958984375, + "learning_rate": 9.977031997703201e-06, + "loss": 0.085, + "step": 22979 + }, + { + "epoch": 10.683403068340306, + "grad_norm": 3.672356367111206, + "learning_rate": 9.977030997703102e-06, + "loss": 0.0643, + "step": 22980 + }, + { + "epoch": 10.683867968386798, + "grad_norm": 4.784267902374268, + "learning_rate": 9.977029997703e-06, + "loss": 0.0821, + "step": 22981 + }, + { + "epoch": 10.684332868433287, + "grad_norm": 1.7733834981918335, + "learning_rate": 9.977028997702901e-06, + "loss": 0.0139, + "step": 22982 + }, + { + "epoch": 10.684797768479777, + "grad_norm": 3.85141921043396, + "learning_rate": 9.9770279977028e-06, + "loss": 0.0412, + "step": 22983 + }, + { + "epoch": 10.685262668526267, + "grad_norm": 9.94151496887207, + "learning_rate": 9.9770269977027e-06, + "loss": 0.1495, + "step": 22984 + }, + { + "epoch": 10.685727568572757, + "grad_norm": 5.328567981719971, + "learning_rate": 9.9770259977026e-06, + "loss": 0.1288, + "step": 22985 + }, + { + "epoch": 10.686192468619247, + "grad_norm": 2.944913625717163, + "learning_rate": 9.9770249977025e-06, + "loss": 0.0181, + "step": 22986 + }, + { + "epoch": 10.686657368665736, + "grad_norm": 2.070525884628296, + "learning_rate": 9.977023997702401e-06, + "loss": 0.0214, + "step": 22987 + }, + { + "epoch": 10.687122268712226, + "grad_norm": 0.9698947072029114, + "learning_rate": 9.9770229977023e-06, + "loss": 0.0042, + "step": 22988 + }, + { + "epoch": 10.687587168758716, + "grad_norm": 3.374302864074707, + "learning_rate": 9.977021997702201e-06, + "loss": 0.0199, + "step": 22989 + }, + { + "epoch": 10.688052068805208, + "grad_norm": 5.43383264541626, + "learning_rate": 9.9770209977021e-06, + "loss": 0.0447, + "step": 22990 + }, + { + "epoch": 10.688516968851697, + "grad_norm": 7.305586338043213, + "learning_rate": 9.977019997702e-06, + "loss": 0.141, + "step": 22991 + }, + { + "epoch": 10.688981868898187, + "grad_norm": 3.6099889278411865, + "learning_rate": 9.977018997701901e-06, + "loss": 0.0415, + "step": 22992 + }, + { + "epoch": 10.689446768944677, + "grad_norm": 7.482144832611084, + "learning_rate": 9.977017997701802e-06, + "loss": 0.149, + "step": 22993 + }, + { + "epoch": 10.689911668991167, + "grad_norm": 0.899491012096405, + "learning_rate": 9.9770169977017e-06, + "loss": 0.0081, + "step": 22994 + }, + { + "epoch": 10.690376569037657, + "grad_norm": 3.75240159034729, + "learning_rate": 9.9770159977016e-06, + "loss": 0.0519, + "step": 22995 + }, + { + "epoch": 10.690841469084146, + "grad_norm": 6.942165374755859, + "learning_rate": 9.9770149977015e-06, + "loss": 0.1145, + "step": 22996 + }, + { + "epoch": 10.691306369130636, + "grad_norm": 2.4252490997314453, + "learning_rate": 9.9770139977014e-06, + "loss": 0.0263, + "step": 22997 + }, + { + "epoch": 10.691771269177128, + "grad_norm": 9.094945907592773, + "learning_rate": 9.9770129977013e-06, + "loss": 0.1296, + "step": 22998 + }, + { + "epoch": 10.692236169223618, + "grad_norm": 4.951802730560303, + "learning_rate": 9.9770119977012e-06, + "loss": 0.0624, + "step": 22999 + }, + { + "epoch": 10.692701069270107, + "grad_norm": 4.77438497543335, + "learning_rate": 9.977010997701101e-06, + "loss": 0.0413, + "step": 23000 + }, + { + "epoch": 10.693165969316597, + "grad_norm": 3.384249448776245, + "learning_rate": 9.977009997701e-06, + "loss": 0.0388, + "step": 23001 + }, + { + "epoch": 10.693630869363087, + "grad_norm": 5.826745510101318, + "learning_rate": 9.977008997700901e-06, + "loss": 0.135, + "step": 23002 + }, + { + "epoch": 10.694095769409577, + "grad_norm": 3.1177031993865967, + "learning_rate": 9.9770079977008e-06, + "loss": 0.03, + "step": 23003 + }, + { + "epoch": 10.694560669456067, + "grad_norm": 7.391109466552734, + "learning_rate": 9.9770069977007e-06, + "loss": 0.1394, + "step": 23004 + }, + { + "epoch": 10.695025569502556, + "grad_norm": 10.373927116394043, + "learning_rate": 9.977005997700601e-06, + "loss": 0.1456, + "step": 23005 + }, + { + "epoch": 10.695490469549046, + "grad_norm": 4.297547817230225, + "learning_rate": 9.9770049977005e-06, + "loss": 0.0414, + "step": 23006 + }, + { + "epoch": 10.695955369595538, + "grad_norm": 4.337991237640381, + "learning_rate": 9.977003997700401e-06, + "loss": 0.0796, + "step": 23007 + }, + { + "epoch": 10.696420269642028, + "grad_norm": 7.836141109466553, + "learning_rate": 9.9770029977003e-06, + "loss": 0.1461, + "step": 23008 + }, + { + "epoch": 10.696885169688517, + "grad_norm": 7.640057563781738, + "learning_rate": 9.9770019977002e-06, + "loss": 0.1102, + "step": 23009 + }, + { + "epoch": 10.697350069735007, + "grad_norm": 2.5573205947875977, + "learning_rate": 9.977000997700101e-06, + "loss": 0.0235, + "step": 23010 + }, + { + "epoch": 10.697814969781497, + "grad_norm": 4.808942794799805, + "learning_rate": 9.9769999977e-06, + "loss": 0.1041, + "step": 23011 + }, + { + "epoch": 10.698279869827987, + "grad_norm": 10.181169509887695, + "learning_rate": 9.9769989976999e-06, + "loss": 0.277, + "step": 23012 + }, + { + "epoch": 10.698744769874477, + "grad_norm": 3.927448272705078, + "learning_rate": 9.976997997699801e-06, + "loss": 0.0557, + "step": 23013 + }, + { + "epoch": 10.699209669920966, + "grad_norm": 7.915115833282471, + "learning_rate": 9.9769969976997e-06, + "loss": 0.2019, + "step": 23014 + }, + { + "epoch": 10.699674569967456, + "grad_norm": 3.769453525543213, + "learning_rate": 9.9769959976996e-06, + "loss": 0.0193, + "step": 23015 + }, + { + "epoch": 10.700139470013948, + "grad_norm": 2.6333742141723633, + "learning_rate": 9.9769949976995e-06, + "loss": 0.0228, + "step": 23016 + }, + { + "epoch": 10.700604370060438, + "grad_norm": 5.501925945281982, + "learning_rate": 9.9769939976994e-06, + "loss": 0.1053, + "step": 23017 + }, + { + "epoch": 10.701069270106927, + "grad_norm": 2.212912082672119, + "learning_rate": 9.976992997699301e-06, + "loss": 0.014, + "step": 23018 + }, + { + "epoch": 10.701534170153417, + "grad_norm": 4.037487506866455, + "learning_rate": 9.9769919976992e-06, + "loss": 0.0459, + "step": 23019 + }, + { + "epoch": 10.701999070199907, + "grad_norm": 5.792354583740234, + "learning_rate": 9.976990997699101e-06, + "loss": 0.0954, + "step": 23020 + }, + { + "epoch": 10.702463970246397, + "grad_norm": 11.095924377441406, + "learning_rate": 9.976989997699e-06, + "loss": 0.1548, + "step": 23021 + }, + { + "epoch": 10.702928870292887, + "grad_norm": 6.264240264892578, + "learning_rate": 9.9769889976989e-06, + "loss": 0.08, + "step": 23022 + }, + { + "epoch": 10.703393770339376, + "grad_norm": 3.972616672515869, + "learning_rate": 9.976987997698801e-06, + "loss": 0.0655, + "step": 23023 + }, + { + "epoch": 10.703858670385866, + "grad_norm": 4.397042751312256, + "learning_rate": 9.9769869976987e-06, + "loss": 0.1009, + "step": 23024 + }, + { + "epoch": 10.704323570432358, + "grad_norm": 9.597631454467773, + "learning_rate": 9.976985997698601e-06, + "loss": 0.1268, + "step": 23025 + }, + { + "epoch": 10.704788470478848, + "grad_norm": 4.689866542816162, + "learning_rate": 9.9769849976985e-06, + "loss": 0.0606, + "step": 23026 + }, + { + "epoch": 10.705253370525337, + "grad_norm": 2.9546239376068115, + "learning_rate": 9.9769839976984e-06, + "loss": 0.0212, + "step": 23027 + }, + { + "epoch": 10.705718270571827, + "grad_norm": 7.372409820556641, + "learning_rate": 9.9769829976983e-06, + "loss": 0.0913, + "step": 23028 + }, + { + "epoch": 10.706183170618317, + "grad_norm": 3.7116432189941406, + "learning_rate": 9.9769819976982e-06, + "loss": 0.0436, + "step": 23029 + }, + { + "epoch": 10.706648070664807, + "grad_norm": 7.618277072906494, + "learning_rate": 9.9769809976981e-06, + "loss": 0.132, + "step": 23030 + }, + { + "epoch": 10.707112970711297, + "grad_norm": 2.084212303161621, + "learning_rate": 9.976979997698001e-06, + "loss": 0.0197, + "step": 23031 + }, + { + "epoch": 10.707577870757786, + "grad_norm": 4.870553970336914, + "learning_rate": 9.9769789976979e-06, + "loss": 0.075, + "step": 23032 + }, + { + "epoch": 10.708042770804276, + "grad_norm": 7.422646522521973, + "learning_rate": 9.976977997697801e-06, + "loss": 0.1133, + "step": 23033 + }, + { + "epoch": 10.708507670850768, + "grad_norm": 5.180825710296631, + "learning_rate": 9.9769769976977e-06, + "loss": 0.0684, + "step": 23034 + }, + { + "epoch": 10.708972570897258, + "grad_norm": 5.33065938949585, + "learning_rate": 9.9769759976976e-06, + "loss": 0.0582, + "step": 23035 + }, + { + "epoch": 10.709437470943747, + "grad_norm": 3.517029047012329, + "learning_rate": 9.976974997697501e-06, + "loss": 0.0374, + "step": 23036 + }, + { + "epoch": 10.709902370990237, + "grad_norm": 3.7028160095214844, + "learning_rate": 9.9769739976974e-06, + "loss": 0.0547, + "step": 23037 + }, + { + "epoch": 10.710367271036727, + "grad_norm": 2.748318672180176, + "learning_rate": 9.976972997697301e-06, + "loss": 0.0173, + "step": 23038 + }, + { + "epoch": 10.710832171083217, + "grad_norm": 4.61320686340332, + "learning_rate": 9.9769719976972e-06, + "loss": 0.0997, + "step": 23039 + }, + { + "epoch": 10.711297071129707, + "grad_norm": 5.437068939208984, + "learning_rate": 9.9769709976971e-06, + "loss": 0.0718, + "step": 23040 + }, + { + "epoch": 10.711761971176196, + "grad_norm": 6.530153274536133, + "learning_rate": 9.976969997697e-06, + "loss": 0.1139, + "step": 23041 + }, + { + "epoch": 10.712226871222686, + "grad_norm": 6.936050891876221, + "learning_rate": 9.9769689976969e-06, + "loss": 0.1158, + "step": 23042 + }, + { + "epoch": 10.712691771269178, + "grad_norm": 8.58862018585205, + "learning_rate": 9.976967997696801e-06, + "loss": 0.0923, + "step": 23043 + }, + { + "epoch": 10.713156671315668, + "grad_norm": 8.64881420135498, + "learning_rate": 9.976966997696702e-06, + "loss": 0.1021, + "step": 23044 + }, + { + "epoch": 10.713621571362157, + "grad_norm": 5.141942501068115, + "learning_rate": 9.9769659976966e-06, + "loss": 0.0459, + "step": 23045 + }, + { + "epoch": 10.714086471408647, + "grad_norm": 5.025834560394287, + "learning_rate": 9.976964997696501e-06, + "loss": 0.0882, + "step": 23046 + }, + { + "epoch": 10.714551371455137, + "grad_norm": 5.432950973510742, + "learning_rate": 9.9769639976964e-06, + "loss": 0.0617, + "step": 23047 + }, + { + "epoch": 10.715016271501627, + "grad_norm": 6.127033233642578, + "learning_rate": 9.9769629976963e-06, + "loss": 0.1086, + "step": 23048 + }, + { + "epoch": 10.715481171548117, + "grad_norm": 2.6108574867248535, + "learning_rate": 9.976961997696201e-06, + "loss": 0.0259, + "step": 23049 + }, + { + "epoch": 10.715946071594606, + "grad_norm": 4.814849853515625, + "learning_rate": 9.9769609976961e-06, + "loss": 0.0429, + "step": 23050 + }, + { + "epoch": 10.716410971641096, + "grad_norm": 6.221872806549072, + "learning_rate": 9.976959997696001e-06, + "loss": 0.0992, + "step": 23051 + }, + { + "epoch": 10.716875871687588, + "grad_norm": 4.135679721832275, + "learning_rate": 9.9769589976959e-06, + "loss": 0.044, + "step": 23052 + }, + { + "epoch": 10.717340771734078, + "grad_norm": 1.7661808729171753, + "learning_rate": 9.9769579976958e-06, + "loss": 0.0162, + "step": 23053 + }, + { + "epoch": 10.717805671780567, + "grad_norm": 2.9793314933776855, + "learning_rate": 9.9769569976957e-06, + "loss": 0.039, + "step": 23054 + }, + { + "epoch": 10.718270571827057, + "grad_norm": 4.429332256317139, + "learning_rate": 9.9769559976956e-06, + "loss": 0.0718, + "step": 23055 + }, + { + "epoch": 10.718735471873547, + "grad_norm": 2.8078219890594482, + "learning_rate": 9.976954997695501e-06, + "loss": 0.0316, + "step": 23056 + }, + { + "epoch": 10.719200371920037, + "grad_norm": 4.541801452636719, + "learning_rate": 9.976953997695402e-06, + "loss": 0.1084, + "step": 23057 + }, + { + "epoch": 10.719665271966527, + "grad_norm": 5.86231803894043, + "learning_rate": 9.9769529976953e-06, + "loss": 0.1099, + "step": 23058 + }, + { + "epoch": 10.720130172013016, + "grad_norm": 6.456648826599121, + "learning_rate": 9.9769519976952e-06, + "loss": 0.0889, + "step": 23059 + }, + { + "epoch": 10.720595072059508, + "grad_norm": 3.834651231765747, + "learning_rate": 9.9769509976951e-06, + "loss": 0.0554, + "step": 23060 + }, + { + "epoch": 10.721059972105998, + "grad_norm": 3.191284656524658, + "learning_rate": 9.976949997695e-06, + "loss": 0.0463, + "step": 23061 + }, + { + "epoch": 10.721524872152488, + "grad_norm": 4.705909252166748, + "learning_rate": 9.976948997694901e-06, + "loss": 0.0943, + "step": 23062 + }, + { + "epoch": 10.721989772198977, + "grad_norm": 8.601651191711426, + "learning_rate": 9.9769479976948e-06, + "loss": 0.1675, + "step": 23063 + }, + { + "epoch": 10.722454672245467, + "grad_norm": 6.901068687438965, + "learning_rate": 9.976946997694701e-06, + "loss": 0.0859, + "step": 23064 + }, + { + "epoch": 10.722919572291957, + "grad_norm": 3.3571114540100098, + "learning_rate": 9.9769459976946e-06, + "loss": 0.03, + "step": 23065 + }, + { + "epoch": 10.723384472338447, + "grad_norm": 5.5568413734436035, + "learning_rate": 9.9769449976945e-06, + "loss": 0.0975, + "step": 23066 + }, + { + "epoch": 10.723849372384937, + "grad_norm": 4.879020690917969, + "learning_rate": 9.9769439976944e-06, + "loss": 0.0721, + "step": 23067 + }, + { + "epoch": 10.724314272431426, + "grad_norm": 5.812129974365234, + "learning_rate": 9.9769429976943e-06, + "loss": 0.1419, + "step": 23068 + }, + { + "epoch": 10.724779172477918, + "grad_norm": 7.019697666168213, + "learning_rate": 9.976941997694201e-06, + "loss": 0.0541, + "step": 23069 + }, + { + "epoch": 10.725244072524408, + "grad_norm": 4.586422920227051, + "learning_rate": 9.9769409976941e-06, + "loss": 0.0747, + "step": 23070 + }, + { + "epoch": 10.725708972570898, + "grad_norm": 5.7370524406433105, + "learning_rate": 9.976939997694e-06, + "loss": 0.0945, + "step": 23071 + }, + { + "epoch": 10.726173872617387, + "grad_norm": 5.398159027099609, + "learning_rate": 9.9769389976939e-06, + "loss": 0.0767, + "step": 23072 + }, + { + "epoch": 10.726638772663877, + "grad_norm": 3.848543882369995, + "learning_rate": 9.9769379976938e-06, + "loss": 0.0453, + "step": 23073 + }, + { + "epoch": 10.727103672710367, + "grad_norm": 2.7051045894622803, + "learning_rate": 9.976936997693701e-06, + "loss": 0.0285, + "step": 23074 + }, + { + "epoch": 10.727568572756857, + "grad_norm": 6.155869960784912, + "learning_rate": 9.976935997693602e-06, + "loss": 0.1025, + "step": 23075 + }, + { + "epoch": 10.728033472803347, + "grad_norm": 3.743412733078003, + "learning_rate": 9.9769349976935e-06, + "loss": 0.0267, + "step": 23076 + }, + { + "epoch": 10.728498372849838, + "grad_norm": 7.679838180541992, + "learning_rate": 9.976933997693401e-06, + "loss": 0.2224, + "step": 23077 + }, + { + "epoch": 10.728963272896328, + "grad_norm": 2.276411771774292, + "learning_rate": 9.9769329976933e-06, + "loss": 0.0303, + "step": 23078 + }, + { + "epoch": 10.729428172942818, + "grad_norm": 5.8931803703308105, + "learning_rate": 9.9769319976932e-06, + "loss": 0.1386, + "step": 23079 + }, + { + "epoch": 10.729893072989308, + "grad_norm": 3.34383225440979, + "learning_rate": 9.9769309976931e-06, + "loss": 0.0252, + "step": 23080 + }, + { + "epoch": 10.730357973035797, + "grad_norm": 5.631933212280273, + "learning_rate": 9.976929997693e-06, + "loss": 0.0834, + "step": 23081 + }, + { + "epoch": 10.730822873082287, + "grad_norm": 12.085278511047363, + "learning_rate": 9.976928997692901e-06, + "loss": 0.258, + "step": 23082 + }, + { + "epoch": 10.731287773128777, + "grad_norm": 4.757380485534668, + "learning_rate": 9.9769279976928e-06, + "loss": 0.1235, + "step": 23083 + }, + { + "epoch": 10.731752673175267, + "grad_norm": 1.864064335823059, + "learning_rate": 9.9769269976927e-06, + "loss": 0.0151, + "step": 23084 + }, + { + "epoch": 10.732217573221757, + "grad_norm": 6.404393196105957, + "learning_rate": 9.9769259976926e-06, + "loss": 0.0636, + "step": 23085 + }, + { + "epoch": 10.732682473268248, + "grad_norm": 7.016977787017822, + "learning_rate": 9.9769249976925e-06, + "loss": 0.1479, + "step": 23086 + }, + { + "epoch": 10.733147373314738, + "grad_norm": 7.622035026550293, + "learning_rate": 9.976923997692401e-06, + "loss": 0.2324, + "step": 23087 + }, + { + "epoch": 10.733612273361228, + "grad_norm": 4.902592658996582, + "learning_rate": 9.976922997692302e-06, + "loss": 0.07, + "step": 23088 + }, + { + "epoch": 10.734077173407718, + "grad_norm": 3.7114481925964355, + "learning_rate": 9.9769219976922e-06, + "loss": 0.0392, + "step": 23089 + }, + { + "epoch": 10.734542073454207, + "grad_norm": 5.634212017059326, + "learning_rate": 9.976920997692101e-06, + "loss": 0.1233, + "step": 23090 + }, + { + "epoch": 10.735006973500697, + "grad_norm": 8.414411544799805, + "learning_rate": 9.976919997692e-06, + "loss": 0.2156, + "step": 23091 + }, + { + "epoch": 10.735471873547187, + "grad_norm": 7.03741979598999, + "learning_rate": 9.976918997691901e-06, + "loss": 0.1355, + "step": 23092 + }, + { + "epoch": 10.735936773593677, + "grad_norm": 7.75044059753418, + "learning_rate": 9.9769179976918e-06, + "loss": 0.0678, + "step": 23093 + }, + { + "epoch": 10.736401673640167, + "grad_norm": 6.832518100738525, + "learning_rate": 9.9769169976917e-06, + "loss": 0.102, + "step": 23094 + }, + { + "epoch": 10.736866573686658, + "grad_norm": 6.274731159210205, + "learning_rate": 9.976915997691601e-06, + "loss": 0.1108, + "step": 23095 + }, + { + "epoch": 10.737331473733148, + "grad_norm": 4.388481616973877, + "learning_rate": 9.9769149976915e-06, + "loss": 0.0738, + "step": 23096 + }, + { + "epoch": 10.737796373779638, + "grad_norm": 16.2789249420166, + "learning_rate": 9.9769139976914e-06, + "loss": 0.3869, + "step": 23097 + }, + { + "epoch": 10.738261273826128, + "grad_norm": 2.690732002258301, + "learning_rate": 9.9769129976913e-06, + "loss": 0.0304, + "step": 23098 + }, + { + "epoch": 10.738726173872617, + "grad_norm": 5.522968769073486, + "learning_rate": 9.9769119976912e-06, + "loss": 0.1184, + "step": 23099 + }, + { + "epoch": 10.739191073919107, + "grad_norm": 3.6336557865142822, + "learning_rate": 9.976910997691101e-06, + "loss": 0.042, + "step": 23100 + }, + { + "epoch": 10.739655973965597, + "grad_norm": 6.250870227813721, + "learning_rate": 9.976909997691002e-06, + "loss": 0.0958, + "step": 23101 + }, + { + "epoch": 10.740120874012087, + "grad_norm": 5.731887340545654, + "learning_rate": 9.9769089976909e-06, + "loss": 0.0574, + "step": 23102 + }, + { + "epoch": 10.740585774058577, + "grad_norm": 1.879862904548645, + "learning_rate": 9.9769079976908e-06, + "loss": 0.0138, + "step": 23103 + }, + { + "epoch": 10.741050674105068, + "grad_norm": 1.0361236333847046, + "learning_rate": 9.9769069976907e-06, + "loss": 0.0088, + "step": 23104 + }, + { + "epoch": 10.741515574151558, + "grad_norm": 8.921152114868164, + "learning_rate": 9.976905997690601e-06, + "loss": 0.2683, + "step": 23105 + }, + { + "epoch": 10.741980474198048, + "grad_norm": 3.5472819805145264, + "learning_rate": 9.976904997690502e-06, + "loss": 0.0397, + "step": 23106 + }, + { + "epoch": 10.742445374244538, + "grad_norm": 1.6656339168548584, + "learning_rate": 9.9769039976904e-06, + "loss": 0.017, + "step": 23107 + }, + { + "epoch": 10.742910274291027, + "grad_norm": 1.2588142156600952, + "learning_rate": 9.976902997690301e-06, + "loss": 0.0106, + "step": 23108 + }, + { + "epoch": 10.743375174337517, + "grad_norm": 6.4858622550964355, + "learning_rate": 9.9769019976902e-06, + "loss": 0.1029, + "step": 23109 + }, + { + "epoch": 10.743840074384007, + "grad_norm": 6.428311824798584, + "learning_rate": 9.976900997690101e-06, + "loss": 0.0952, + "step": 23110 + }, + { + "epoch": 10.744304974430497, + "grad_norm": 5.396516799926758, + "learning_rate": 9.97689999769e-06, + "loss": 0.0947, + "step": 23111 + }, + { + "epoch": 10.744769874476987, + "grad_norm": 1.0750874280929565, + "learning_rate": 9.9768989976899e-06, + "loss": 0.0121, + "step": 23112 + }, + { + "epoch": 10.745234774523478, + "grad_norm": 4.326958656311035, + "learning_rate": 9.976897997689801e-06, + "loss": 0.0401, + "step": 23113 + }, + { + "epoch": 10.745699674569968, + "grad_norm": 8.021915435791016, + "learning_rate": 9.9768969976897e-06, + "loss": 0.2028, + "step": 23114 + }, + { + "epoch": 10.746164574616458, + "grad_norm": 4.223660945892334, + "learning_rate": 9.9768959976896e-06, + "loss": 0.0453, + "step": 23115 + }, + { + "epoch": 10.746629474662948, + "grad_norm": 4.926130771636963, + "learning_rate": 9.9768949976895e-06, + "loss": 0.0532, + "step": 23116 + }, + { + "epoch": 10.747094374709437, + "grad_norm": 3.4596331119537354, + "learning_rate": 9.9768939976894e-06, + "loss": 0.0429, + "step": 23117 + }, + { + "epoch": 10.747559274755927, + "grad_norm": 7.560913562774658, + "learning_rate": 9.976892997689301e-06, + "loss": 0.1301, + "step": 23118 + }, + { + "epoch": 10.748024174802417, + "grad_norm": 10.451220512390137, + "learning_rate": 9.976891997689202e-06, + "loss": 0.2037, + "step": 23119 + }, + { + "epoch": 10.748489074848907, + "grad_norm": 5.478310585021973, + "learning_rate": 9.9768909976891e-06, + "loss": 0.1117, + "step": 23120 + }, + { + "epoch": 10.748953974895397, + "grad_norm": 4.300070762634277, + "learning_rate": 9.976889997689001e-06, + "loss": 0.036, + "step": 23121 + }, + { + "epoch": 10.749418874941888, + "grad_norm": 3.315911293029785, + "learning_rate": 9.9768889976889e-06, + "loss": 0.0258, + "step": 23122 + }, + { + "epoch": 10.749883774988378, + "grad_norm": 5.780185222625732, + "learning_rate": 9.976887997688801e-06, + "loss": 0.0623, + "step": 23123 + }, + { + "epoch": 10.750348675034868, + "grad_norm": 7.163794040679932, + "learning_rate": 9.9768869976887e-06, + "loss": 0.056, + "step": 23124 + }, + { + "epoch": 10.750813575081358, + "grad_norm": 6.6648712158203125, + "learning_rate": 9.9768859976886e-06, + "loss": 0.0935, + "step": 23125 + }, + { + "epoch": 10.751278475127847, + "grad_norm": 4.5723371505737305, + "learning_rate": 9.976884997688501e-06, + "loss": 0.0776, + "step": 23126 + }, + { + "epoch": 10.751743375174337, + "grad_norm": 3.403773546218872, + "learning_rate": 9.9768839976884e-06, + "loss": 0.0343, + "step": 23127 + }, + { + "epoch": 10.752208275220827, + "grad_norm": 23.4886417388916, + "learning_rate": 9.976882997688301e-06, + "loss": 0.0676, + "step": 23128 + }, + { + "epoch": 10.752673175267317, + "grad_norm": 5.531357765197754, + "learning_rate": 9.9768819976882e-06, + "loss": 0.0423, + "step": 23129 + }, + { + "epoch": 10.753138075313807, + "grad_norm": 2.6902682781219482, + "learning_rate": 9.9768809976881e-06, + "loss": 0.0321, + "step": 23130 + }, + { + "epoch": 10.753602975360298, + "grad_norm": 9.674668312072754, + "learning_rate": 9.976879997688001e-06, + "loss": 0.1175, + "step": 23131 + }, + { + "epoch": 10.754067875406788, + "grad_norm": 5.505809307098389, + "learning_rate": 9.976878997687902e-06, + "loss": 0.1119, + "step": 23132 + }, + { + "epoch": 10.754532775453278, + "grad_norm": 6.462216854095459, + "learning_rate": 9.9768779976878e-06, + "loss": 0.116, + "step": 23133 + }, + { + "epoch": 10.754997675499768, + "grad_norm": 3.817446231842041, + "learning_rate": 9.976876997687701e-06, + "loss": 0.0551, + "step": 23134 + }, + { + "epoch": 10.755462575546257, + "grad_norm": 9.47375774383545, + "learning_rate": 9.9768759976876e-06, + "loss": 0.1594, + "step": 23135 + }, + { + "epoch": 10.755927475592747, + "grad_norm": 2.435898780822754, + "learning_rate": 9.9768749976875e-06, + "loss": 0.0172, + "step": 23136 + }, + { + "epoch": 10.756392375639237, + "grad_norm": 6.599018096923828, + "learning_rate": 9.9768739976874e-06, + "loss": 0.0874, + "step": 23137 + }, + { + "epoch": 10.756857275685727, + "grad_norm": 0.6492968797683716, + "learning_rate": 9.9768729976873e-06, + "loss": 0.006, + "step": 23138 + }, + { + "epoch": 10.757322175732218, + "grad_norm": 3.6150121688842773, + "learning_rate": 9.976871997687201e-06, + "loss": 0.0361, + "step": 23139 + }, + { + "epoch": 10.757787075778708, + "grad_norm": 4.699492454528809, + "learning_rate": 9.9768709976871e-06, + "loss": 0.0633, + "step": 23140 + }, + { + "epoch": 10.758251975825198, + "grad_norm": 6.037116050720215, + "learning_rate": 9.976869997687001e-06, + "loss": 0.1009, + "step": 23141 + }, + { + "epoch": 10.758716875871688, + "grad_norm": 5.101351261138916, + "learning_rate": 9.9768689976869e-06, + "loss": 0.0317, + "step": 23142 + }, + { + "epoch": 10.759181775918178, + "grad_norm": 6.95809268951416, + "learning_rate": 9.9768679976868e-06, + "loss": 0.2163, + "step": 23143 + }, + { + "epoch": 10.759646675964667, + "grad_norm": 8.109752655029297, + "learning_rate": 9.976866997686701e-06, + "loss": 0.1986, + "step": 23144 + }, + { + "epoch": 10.760111576011157, + "grad_norm": 5.609168529510498, + "learning_rate": 9.976865997686602e-06, + "loss": 0.0579, + "step": 23145 + }, + { + "epoch": 10.760576476057647, + "grad_norm": 2.647589921951294, + "learning_rate": 9.9768649976865e-06, + "loss": 0.0308, + "step": 23146 + }, + { + "epoch": 10.761041376104137, + "grad_norm": 5.712673187255859, + "learning_rate": 9.9768639976864e-06, + "loss": 0.0428, + "step": 23147 + }, + { + "epoch": 10.761506276150628, + "grad_norm": 6.353481769561768, + "learning_rate": 9.9768629976863e-06, + "loss": 0.1533, + "step": 23148 + }, + { + "epoch": 10.761971176197118, + "grad_norm": 3.7915830612182617, + "learning_rate": 9.9768619976862e-06, + "loss": 0.0262, + "step": 23149 + }, + { + "epoch": 10.762436076243608, + "grad_norm": 3.3720309734344482, + "learning_rate": 9.9768609976861e-06, + "loss": 0.0422, + "step": 23150 + }, + { + "epoch": 10.762900976290098, + "grad_norm": 3.432034730911255, + "learning_rate": 9.976859997686e-06, + "loss": 0.0162, + "step": 23151 + }, + { + "epoch": 10.763365876336588, + "grad_norm": 0.6681323647499084, + "learning_rate": 9.976858997685901e-06, + "loss": 0.0084, + "step": 23152 + }, + { + "epoch": 10.763830776383077, + "grad_norm": 5.181787967681885, + "learning_rate": 9.9768579976858e-06, + "loss": 0.0966, + "step": 23153 + }, + { + "epoch": 10.764295676429567, + "grad_norm": 3.097442626953125, + "learning_rate": 9.976856997685701e-06, + "loss": 0.0232, + "step": 23154 + }, + { + "epoch": 10.764760576476057, + "grad_norm": 5.932000637054443, + "learning_rate": 9.9768559976856e-06, + "loss": 0.0859, + "step": 23155 + }, + { + "epoch": 10.765225476522549, + "grad_norm": 9.541723251342773, + "learning_rate": 9.9768549976855e-06, + "loss": 0.0521, + "step": 23156 + }, + { + "epoch": 10.765690376569038, + "grad_norm": 3.500699281692505, + "learning_rate": 9.976853997685401e-06, + "loss": 0.0584, + "step": 23157 + }, + { + "epoch": 10.766155276615528, + "grad_norm": 3.953073740005493, + "learning_rate": 9.9768529976853e-06, + "loss": 0.0629, + "step": 23158 + }, + { + "epoch": 10.766620176662018, + "grad_norm": 7.831437110900879, + "learning_rate": 9.976851997685201e-06, + "loss": 0.1518, + "step": 23159 + }, + { + "epoch": 10.767085076708508, + "grad_norm": 5.519185543060303, + "learning_rate": 9.9768509976851e-06, + "loss": 0.129, + "step": 23160 + }, + { + "epoch": 10.767549976754998, + "grad_norm": 4.5112128257751465, + "learning_rate": 9.976849997685e-06, + "loss": 0.0435, + "step": 23161 + }, + { + "epoch": 10.768014876801487, + "grad_norm": 7.173830509185791, + "learning_rate": 9.9768489976849e-06, + "loss": 0.1151, + "step": 23162 + }, + { + "epoch": 10.768479776847977, + "grad_norm": 4.159936428070068, + "learning_rate": 9.9768479976848e-06, + "loss": 0.0507, + "step": 23163 + }, + { + "epoch": 10.768944676894467, + "grad_norm": 3.7711262702941895, + "learning_rate": 9.9768469976847e-06, + "loss": 0.0312, + "step": 23164 + }, + { + "epoch": 10.769409576940959, + "grad_norm": 6.591448783874512, + "learning_rate": 9.976845997684601e-06, + "loss": 0.0797, + "step": 23165 + }, + { + "epoch": 10.769874476987448, + "grad_norm": 2.5715279579162598, + "learning_rate": 9.9768449976845e-06, + "loss": 0.0245, + "step": 23166 + }, + { + "epoch": 10.770339377033938, + "grad_norm": 5.90723180770874, + "learning_rate": 9.9768439976844e-06, + "loss": 0.1661, + "step": 23167 + }, + { + "epoch": 10.770804277080428, + "grad_norm": 5.891357898712158, + "learning_rate": 9.9768429976843e-06, + "loss": 0.081, + "step": 23168 + }, + { + "epoch": 10.771269177126918, + "grad_norm": 1.7600494623184204, + "learning_rate": 9.9768419976842e-06, + "loss": 0.0151, + "step": 23169 + }, + { + "epoch": 10.771734077173408, + "grad_norm": 4.5518903732299805, + "learning_rate": 9.976840997684101e-06, + "loss": 0.0915, + "step": 23170 + }, + { + "epoch": 10.772198977219897, + "grad_norm": 13.15283489227295, + "learning_rate": 9.976839997684e-06, + "loss": 0.0947, + "step": 23171 + }, + { + "epoch": 10.772663877266387, + "grad_norm": 6.125743389129639, + "learning_rate": 9.976838997683901e-06, + "loss": 0.1146, + "step": 23172 + }, + { + "epoch": 10.773128777312877, + "grad_norm": 1.7171456813812256, + "learning_rate": 9.9768379976838e-06, + "loss": 0.0189, + "step": 23173 + }, + { + "epoch": 10.773593677359369, + "grad_norm": 4.225838661193848, + "learning_rate": 9.9768369976837e-06, + "loss": 0.0561, + "step": 23174 + }, + { + "epoch": 10.774058577405858, + "grad_norm": 3.943876266479492, + "learning_rate": 9.976835997683601e-06, + "loss": 0.0418, + "step": 23175 + }, + { + "epoch": 10.774523477452348, + "grad_norm": 5.1218767166137695, + "learning_rate": 9.9768349976835e-06, + "loss": 0.0691, + "step": 23176 + }, + { + "epoch": 10.774988377498838, + "grad_norm": 5.689609527587891, + "learning_rate": 9.976833997683401e-06, + "loss": 0.0459, + "step": 23177 + }, + { + "epoch": 10.775453277545328, + "grad_norm": 4.310893535614014, + "learning_rate": 9.9768329976833e-06, + "loss": 0.0534, + "step": 23178 + }, + { + "epoch": 10.775918177591818, + "grad_norm": 4.796943664550781, + "learning_rate": 9.9768319976832e-06, + "loss": 0.068, + "step": 23179 + }, + { + "epoch": 10.776383077638307, + "grad_norm": 4.1441826820373535, + "learning_rate": 9.9768309976831e-06, + "loss": 0.0537, + "step": 23180 + }, + { + "epoch": 10.776847977684797, + "grad_norm": 3.9432268142700195, + "learning_rate": 9.976829997683e-06, + "loss": 0.0752, + "step": 23181 + }, + { + "epoch": 10.777312877731287, + "grad_norm": 3.8828063011169434, + "learning_rate": 9.9768289976829e-06, + "loss": 0.0599, + "step": 23182 + }, + { + "epoch": 10.777777777777779, + "grad_norm": 2.4748079776763916, + "learning_rate": 9.976827997682801e-06, + "loss": 0.0287, + "step": 23183 + }, + { + "epoch": 10.778242677824268, + "grad_norm": 14.330582618713379, + "learning_rate": 9.9768269976827e-06, + "loss": 0.2439, + "step": 23184 + }, + { + "epoch": 10.778707577870758, + "grad_norm": 4.3078789710998535, + "learning_rate": 9.976825997682601e-06, + "loss": 0.0791, + "step": 23185 + }, + { + "epoch": 10.779172477917248, + "grad_norm": 3.3264074325561523, + "learning_rate": 9.9768249976825e-06, + "loss": 0.0426, + "step": 23186 + }, + { + "epoch": 10.779637377963738, + "grad_norm": 5.521560192108154, + "learning_rate": 9.9768239976824e-06, + "loss": 0.0664, + "step": 23187 + }, + { + "epoch": 10.780102278010228, + "grad_norm": 3.851329803466797, + "learning_rate": 9.976822997682301e-06, + "loss": 0.0287, + "step": 23188 + }, + { + "epoch": 10.780567178056717, + "grad_norm": 5.860396862030029, + "learning_rate": 9.9768219976822e-06, + "loss": 0.1047, + "step": 23189 + }, + { + "epoch": 10.781032078103207, + "grad_norm": 5.807417869567871, + "learning_rate": 9.976820997682101e-06, + "loss": 0.1009, + "step": 23190 + }, + { + "epoch": 10.781496978149697, + "grad_norm": 3.602170944213867, + "learning_rate": 9.976819997682e-06, + "loss": 0.0403, + "step": 23191 + }, + { + "epoch": 10.781961878196189, + "grad_norm": 3.2564775943756104, + "learning_rate": 9.9768189976819e-06, + "loss": 0.0446, + "step": 23192 + }, + { + "epoch": 10.782426778242678, + "grad_norm": 5.813723564147949, + "learning_rate": 9.9768179976818e-06, + "loss": 0.1308, + "step": 23193 + }, + { + "epoch": 10.782891678289168, + "grad_norm": 5.200113296508789, + "learning_rate": 9.9768169976817e-06, + "loss": 0.0995, + "step": 23194 + }, + { + "epoch": 10.783356578335658, + "grad_norm": 5.624858379364014, + "learning_rate": 9.976815997681601e-06, + "loss": 0.0482, + "step": 23195 + }, + { + "epoch": 10.783821478382148, + "grad_norm": 1.162474274635315, + "learning_rate": 9.976814997681502e-06, + "loss": 0.0117, + "step": 23196 + }, + { + "epoch": 10.784286378428638, + "grad_norm": 6.247503280639648, + "learning_rate": 9.9768139976814e-06, + "loss": 0.0656, + "step": 23197 + }, + { + "epoch": 10.784751278475127, + "grad_norm": 2.8557350635528564, + "learning_rate": 9.976812997681301e-06, + "loss": 0.0311, + "step": 23198 + }, + { + "epoch": 10.785216178521617, + "grad_norm": 9.309121131896973, + "learning_rate": 9.9768119976812e-06, + "loss": 0.1839, + "step": 23199 + }, + { + "epoch": 10.785681078568107, + "grad_norm": 3.573859691619873, + "learning_rate": 9.9768109976811e-06, + "loss": 0.0351, + "step": 23200 + }, + { + "epoch": 10.786145978614599, + "grad_norm": 3.4142699241638184, + "learning_rate": 9.976809997681001e-06, + "loss": 0.0548, + "step": 23201 + }, + { + "epoch": 10.786610878661088, + "grad_norm": 3.2862753868103027, + "learning_rate": 9.9768089976809e-06, + "loss": 0.0393, + "step": 23202 + }, + { + "epoch": 10.787075778707578, + "grad_norm": 4.5840067863464355, + "learning_rate": 9.976807997680801e-06, + "loss": 0.0578, + "step": 23203 + }, + { + "epoch": 10.787540678754068, + "grad_norm": 4.031424522399902, + "learning_rate": 9.9768069976807e-06, + "loss": 0.0681, + "step": 23204 + }, + { + "epoch": 10.788005578800558, + "grad_norm": 6.087038516998291, + "learning_rate": 9.9768059976806e-06, + "loss": 0.0421, + "step": 23205 + }, + { + "epoch": 10.788470478847048, + "grad_norm": 6.004051208496094, + "learning_rate": 9.9768049976805e-06, + "loss": 0.0213, + "step": 23206 + }, + { + "epoch": 10.788935378893537, + "grad_norm": 4.345851898193359, + "learning_rate": 9.9768039976804e-06, + "loss": 0.0872, + "step": 23207 + }, + { + "epoch": 10.789400278940027, + "grad_norm": 9.213043212890625, + "learning_rate": 9.976802997680301e-06, + "loss": 0.0818, + "step": 23208 + }, + { + "epoch": 10.789865178986517, + "grad_norm": 0.7370861172676086, + "learning_rate": 9.976801997680202e-06, + "loss": 0.0041, + "step": 23209 + }, + { + "epoch": 10.790330079033009, + "grad_norm": 5.791417598724365, + "learning_rate": 9.9768009976801e-06, + "loss": 0.0821, + "step": 23210 + }, + { + "epoch": 10.790794979079498, + "grad_norm": 2.983212471008301, + "learning_rate": 9.97679999768e-06, + "loss": 0.0208, + "step": 23211 + }, + { + "epoch": 10.791259879125988, + "grad_norm": 8.924302101135254, + "learning_rate": 9.9767989976799e-06, + "loss": 0.1084, + "step": 23212 + }, + { + "epoch": 10.791724779172478, + "grad_norm": 4.063845157623291, + "learning_rate": 9.976797997679801e-06, + "loss": 0.0679, + "step": 23213 + }, + { + "epoch": 10.792189679218968, + "grad_norm": 11.200243949890137, + "learning_rate": 9.976796997679702e-06, + "loss": 0.1555, + "step": 23214 + }, + { + "epoch": 10.792654579265458, + "grad_norm": 7.648696422576904, + "learning_rate": 9.9767959976796e-06, + "loss": 0.1459, + "step": 23215 + }, + { + "epoch": 10.793119479311947, + "grad_norm": 5.125034809112549, + "learning_rate": 9.976794997679501e-06, + "loss": 0.0631, + "step": 23216 + }, + { + "epoch": 10.793584379358437, + "grad_norm": 2.1114773750305176, + "learning_rate": 9.9767939976794e-06, + "loss": 0.0141, + "step": 23217 + }, + { + "epoch": 10.794049279404929, + "grad_norm": 5.963925361633301, + "learning_rate": 9.9767929976793e-06, + "loss": 0.1416, + "step": 23218 + }, + { + "epoch": 10.794514179451419, + "grad_norm": 10.018622398376465, + "learning_rate": 9.9767919976792e-06, + "loss": 0.1237, + "step": 23219 + }, + { + "epoch": 10.794979079497908, + "grad_norm": 2.357273578643799, + "learning_rate": 9.9767909976791e-06, + "loss": 0.0142, + "step": 23220 + }, + { + "epoch": 10.795443979544398, + "grad_norm": 8.581324577331543, + "learning_rate": 9.976789997679001e-06, + "loss": 0.1134, + "step": 23221 + }, + { + "epoch": 10.795908879590888, + "grad_norm": 5.939477443695068, + "learning_rate": 9.9767889976789e-06, + "loss": 0.0505, + "step": 23222 + }, + { + "epoch": 10.796373779637378, + "grad_norm": 2.4442267417907715, + "learning_rate": 9.9767879976788e-06, + "loss": 0.0202, + "step": 23223 + }, + { + "epoch": 10.796838679683868, + "grad_norm": 4.026083946228027, + "learning_rate": 9.9767869976787e-06, + "loss": 0.0254, + "step": 23224 + }, + { + "epoch": 10.797303579730357, + "grad_norm": 4.395589351654053, + "learning_rate": 9.9767859976786e-06, + "loss": 0.0311, + "step": 23225 + }, + { + "epoch": 10.797768479776847, + "grad_norm": 7.251776695251465, + "learning_rate": 9.976784997678501e-06, + "loss": 0.2007, + "step": 23226 + }, + { + "epoch": 10.798233379823339, + "grad_norm": 4.3406596183776855, + "learning_rate": 9.976783997678402e-06, + "loss": 0.0738, + "step": 23227 + }, + { + "epoch": 10.798698279869829, + "grad_norm": 6.228662490844727, + "learning_rate": 9.9767829976783e-06, + "loss": 0.1081, + "step": 23228 + }, + { + "epoch": 10.799163179916318, + "grad_norm": 1.759524941444397, + "learning_rate": 9.976781997678201e-06, + "loss": 0.0175, + "step": 23229 + }, + { + "epoch": 10.799628079962808, + "grad_norm": 6.749518871307373, + "learning_rate": 9.9767809976781e-06, + "loss": 0.1471, + "step": 23230 + }, + { + "epoch": 10.800092980009298, + "grad_norm": 4.672595500946045, + "learning_rate": 9.976779997678e-06, + "loss": 0.0542, + "step": 23231 + }, + { + "epoch": 10.800557880055788, + "grad_norm": 6.63409948348999, + "learning_rate": 9.9767789976779e-06, + "loss": 0.1528, + "step": 23232 + }, + { + "epoch": 10.801022780102278, + "grad_norm": 3.4858639240264893, + "learning_rate": 9.9767779976778e-06, + "loss": 0.0282, + "step": 23233 + }, + { + "epoch": 10.801487680148767, + "grad_norm": 3.544813632965088, + "learning_rate": 9.976776997677701e-06, + "loss": 0.0227, + "step": 23234 + }, + { + "epoch": 10.801952580195259, + "grad_norm": 6.478381156921387, + "learning_rate": 9.9767759976776e-06, + "loss": 0.0795, + "step": 23235 + }, + { + "epoch": 10.802417480241749, + "grad_norm": 5.267820835113525, + "learning_rate": 9.9767749976775e-06, + "loss": 0.0837, + "step": 23236 + }, + { + "epoch": 10.802882380288239, + "grad_norm": 6.736508369445801, + "learning_rate": 9.9767739976774e-06, + "loss": 0.0983, + "step": 23237 + }, + { + "epoch": 10.803347280334728, + "grad_norm": 7.8207526206970215, + "learning_rate": 9.9767729976773e-06, + "loss": 0.1156, + "step": 23238 + }, + { + "epoch": 10.803812180381218, + "grad_norm": 3.711164712905884, + "learning_rate": 9.976771997677201e-06, + "loss": 0.0407, + "step": 23239 + }, + { + "epoch": 10.804277080427708, + "grad_norm": 12.386990547180176, + "learning_rate": 9.976770997677102e-06, + "loss": 0.2327, + "step": 23240 + }, + { + "epoch": 10.804741980474198, + "grad_norm": 8.328667640686035, + "learning_rate": 9.976769997677e-06, + "loss": 0.1133, + "step": 23241 + }, + { + "epoch": 10.805206880520688, + "grad_norm": 4.511114597320557, + "learning_rate": 9.976768997676901e-06, + "loss": 0.0406, + "step": 23242 + }, + { + "epoch": 10.805671780567177, + "grad_norm": 3.964205026626587, + "learning_rate": 9.9767679976768e-06, + "loss": 0.0623, + "step": 23243 + }, + { + "epoch": 10.806136680613669, + "grad_norm": 6.054858684539795, + "learning_rate": 9.976766997676701e-06, + "loss": 0.0524, + "step": 23244 + }, + { + "epoch": 10.806601580660159, + "grad_norm": 3.1370322704315186, + "learning_rate": 9.9767659976766e-06, + "loss": 0.0288, + "step": 23245 + }, + { + "epoch": 10.807066480706649, + "grad_norm": 4.10054874420166, + "learning_rate": 9.9767649976765e-06, + "loss": 0.0511, + "step": 23246 + }, + { + "epoch": 10.807531380753138, + "grad_norm": 4.15212345123291, + "learning_rate": 9.976763997676401e-06, + "loss": 0.0421, + "step": 23247 + }, + { + "epoch": 10.807996280799628, + "grad_norm": 5.941666603088379, + "learning_rate": 9.9767629976763e-06, + "loss": 0.1066, + "step": 23248 + }, + { + "epoch": 10.808461180846118, + "grad_norm": 6.2891387939453125, + "learning_rate": 9.9767619976762e-06, + "loss": 0.0983, + "step": 23249 + }, + { + "epoch": 10.808926080892608, + "grad_norm": 10.04023265838623, + "learning_rate": 9.9767609976761e-06, + "loss": 0.2576, + "step": 23250 + }, + { + "epoch": 10.809390980939098, + "grad_norm": 7.996504783630371, + "learning_rate": 9.976759997676e-06, + "loss": 0.1396, + "step": 23251 + }, + { + "epoch": 10.809855880985587, + "grad_norm": 5.469558238983154, + "learning_rate": 9.976758997675901e-06, + "loss": 0.0552, + "step": 23252 + }, + { + "epoch": 10.810320781032079, + "grad_norm": 4.82853364944458, + "learning_rate": 9.976757997675802e-06, + "loss": 0.0418, + "step": 23253 + }, + { + "epoch": 10.810785681078569, + "grad_norm": 4.676257610321045, + "learning_rate": 9.9767569976757e-06, + "loss": 0.0794, + "step": 23254 + }, + { + "epoch": 10.811250581125059, + "grad_norm": 2.4540517330169678, + "learning_rate": 9.9767559976756e-06, + "loss": 0.0278, + "step": 23255 + }, + { + "epoch": 10.811715481171548, + "grad_norm": 2.6615326404571533, + "learning_rate": 9.9767549976755e-06, + "loss": 0.0272, + "step": 23256 + }, + { + "epoch": 10.812180381218038, + "grad_norm": 2.171318292617798, + "learning_rate": 9.976753997675401e-06, + "loss": 0.0147, + "step": 23257 + }, + { + "epoch": 10.812645281264528, + "grad_norm": 6.8230485916137695, + "learning_rate": 9.9767529976753e-06, + "loss": 0.1307, + "step": 23258 + }, + { + "epoch": 10.813110181311018, + "grad_norm": 3.3981130123138428, + "learning_rate": 9.9767519976752e-06, + "loss": 0.0426, + "step": 23259 + }, + { + "epoch": 10.813575081357508, + "grad_norm": 2.080488443374634, + "learning_rate": 9.976750997675101e-06, + "loss": 0.0164, + "step": 23260 + }, + { + "epoch": 10.814039981403997, + "grad_norm": 7.1129279136657715, + "learning_rate": 9.976749997675e-06, + "loss": 0.0711, + "step": 23261 + }, + { + "epoch": 10.814504881450489, + "grad_norm": 5.095426559448242, + "learning_rate": 9.976748997674901e-06, + "loss": 0.1321, + "step": 23262 + }, + { + "epoch": 10.814969781496979, + "grad_norm": 1.0331531763076782, + "learning_rate": 9.9767479976748e-06, + "loss": 0.0069, + "step": 23263 + }, + { + "epoch": 10.815434681543469, + "grad_norm": 4.191064357757568, + "learning_rate": 9.9767469976747e-06, + "loss": 0.0621, + "step": 23264 + }, + { + "epoch": 10.815899581589958, + "grad_norm": 4.318072319030762, + "learning_rate": 9.976745997674601e-06, + "loss": 0.0801, + "step": 23265 + }, + { + "epoch": 10.816364481636448, + "grad_norm": 5.000720024108887, + "learning_rate": 9.9767449976745e-06, + "loss": 0.0769, + "step": 23266 + }, + { + "epoch": 10.816829381682938, + "grad_norm": 3.1311800479888916, + "learning_rate": 9.9767439976744e-06, + "loss": 0.0196, + "step": 23267 + }, + { + "epoch": 10.817294281729428, + "grad_norm": 5.610750198364258, + "learning_rate": 9.9767429976743e-06, + "loss": 0.0944, + "step": 23268 + }, + { + "epoch": 10.817759181775918, + "grad_norm": 7.122770309448242, + "learning_rate": 9.9767419976742e-06, + "loss": 0.0881, + "step": 23269 + }, + { + "epoch": 10.818224081822407, + "grad_norm": 2.7304341793060303, + "learning_rate": 9.976740997674101e-06, + "loss": 0.0334, + "step": 23270 + }, + { + "epoch": 10.818688981868899, + "grad_norm": 1.4639155864715576, + "learning_rate": 9.976739997674002e-06, + "loss": 0.0146, + "step": 23271 + }, + { + "epoch": 10.819153881915389, + "grad_norm": 2.7950196266174316, + "learning_rate": 9.9767389976739e-06, + "loss": 0.0216, + "step": 23272 + }, + { + "epoch": 10.819618781961879, + "grad_norm": 3.9795238971710205, + "learning_rate": 9.976737997673801e-06, + "loss": 0.0485, + "step": 23273 + }, + { + "epoch": 10.820083682008368, + "grad_norm": 1.1611350774765015, + "learning_rate": 9.9767369976737e-06, + "loss": 0.0108, + "step": 23274 + }, + { + "epoch": 10.820548582054858, + "grad_norm": 3.1225311756134033, + "learning_rate": 9.976735997673601e-06, + "loss": 0.0325, + "step": 23275 + }, + { + "epoch": 10.821013482101348, + "grad_norm": 1.370409607887268, + "learning_rate": 9.9767349976735e-06, + "loss": 0.0078, + "step": 23276 + }, + { + "epoch": 10.821478382147838, + "grad_norm": 6.323364734649658, + "learning_rate": 9.9767339976734e-06, + "loss": 0.1182, + "step": 23277 + }, + { + "epoch": 10.821943282194328, + "grad_norm": 1.8166364431381226, + "learning_rate": 9.976732997673301e-06, + "loss": 0.0213, + "step": 23278 + }, + { + "epoch": 10.822408182240817, + "grad_norm": 1.3252331018447876, + "learning_rate": 9.9767319976732e-06, + "loss": 0.0102, + "step": 23279 + }, + { + "epoch": 10.822873082287309, + "grad_norm": 7.0439982414245605, + "learning_rate": 9.976730997673101e-06, + "loss": 0.0609, + "step": 23280 + }, + { + "epoch": 10.823337982333799, + "grad_norm": 4.565033912658691, + "learning_rate": 9.976729997673e-06, + "loss": 0.072, + "step": 23281 + }, + { + "epoch": 10.823802882380289, + "grad_norm": 4.537172317504883, + "learning_rate": 9.9767289976729e-06, + "loss": 0.0512, + "step": 23282 + }, + { + "epoch": 10.824267782426778, + "grad_norm": 2.0320229530334473, + "learning_rate": 9.976727997672801e-06, + "loss": 0.0105, + "step": 23283 + }, + { + "epoch": 10.824732682473268, + "grad_norm": 4.430311679840088, + "learning_rate": 9.976726997672702e-06, + "loss": 0.0808, + "step": 23284 + }, + { + "epoch": 10.825197582519758, + "grad_norm": 2.7917306423187256, + "learning_rate": 9.9767259976726e-06, + "loss": 0.0545, + "step": 23285 + }, + { + "epoch": 10.825662482566248, + "grad_norm": 5.637840747833252, + "learning_rate": 9.976724997672501e-06, + "loss": 0.1236, + "step": 23286 + }, + { + "epoch": 10.826127382612738, + "grad_norm": 4.166886329650879, + "learning_rate": 9.9767239976724e-06, + "loss": 0.1055, + "step": 23287 + }, + { + "epoch": 10.826592282659227, + "grad_norm": 7.262294292449951, + "learning_rate": 9.9767229976723e-06, + "loss": 0.105, + "step": 23288 + }, + { + "epoch": 10.827057182705719, + "grad_norm": 4.048748016357422, + "learning_rate": 9.9767219976722e-06, + "loss": 0.079, + "step": 23289 + }, + { + "epoch": 10.827522082752209, + "grad_norm": 6.394311904907227, + "learning_rate": 9.9767209976721e-06, + "loss": 0.1049, + "step": 23290 + }, + { + "epoch": 10.827986982798699, + "grad_norm": 6.222258567810059, + "learning_rate": 9.976719997672001e-06, + "loss": 0.0383, + "step": 23291 + }, + { + "epoch": 10.828451882845188, + "grad_norm": 2.709294557571411, + "learning_rate": 9.9767189976719e-06, + "loss": 0.0242, + "step": 23292 + }, + { + "epoch": 10.828916782891678, + "grad_norm": 5.091104030609131, + "learning_rate": 9.976717997671801e-06, + "loss": 0.0801, + "step": 23293 + }, + { + "epoch": 10.829381682938168, + "grad_norm": 3.460057497024536, + "learning_rate": 9.9767169976717e-06, + "loss": 0.0398, + "step": 23294 + }, + { + "epoch": 10.829846582984658, + "grad_norm": 3.6136059761047363, + "learning_rate": 9.9767159976716e-06, + "loss": 0.0666, + "step": 23295 + }, + { + "epoch": 10.830311483031148, + "grad_norm": 5.792974948883057, + "learning_rate": 9.976714997671501e-06, + "loss": 0.1037, + "step": 23296 + }, + { + "epoch": 10.83077638307764, + "grad_norm": 6.455870151519775, + "learning_rate": 9.9767139976714e-06, + "loss": 0.0993, + "step": 23297 + }, + { + "epoch": 10.831241283124129, + "grad_norm": 5.420674800872803, + "learning_rate": 9.976712997671301e-06, + "loss": 0.1231, + "step": 23298 + }, + { + "epoch": 10.831706183170619, + "grad_norm": 6.476269721984863, + "learning_rate": 9.9767119976712e-06, + "loss": 0.1414, + "step": 23299 + }, + { + "epoch": 10.832171083217109, + "grad_norm": 7.397531032562256, + "learning_rate": 9.9767109976711e-06, + "loss": 0.1375, + "step": 23300 + }, + { + "epoch": 10.832635983263598, + "grad_norm": 1.3144872188568115, + "learning_rate": 9.976709997671e-06, + "loss": 0.0162, + "step": 23301 + }, + { + "epoch": 10.833100883310088, + "grad_norm": 4.906193256378174, + "learning_rate": 9.9767089976709e-06, + "loss": 0.06, + "step": 23302 + }, + { + "epoch": 10.833565783356578, + "grad_norm": 3.4767868518829346, + "learning_rate": 9.9767079976708e-06, + "loss": 0.047, + "step": 23303 + }, + { + "epoch": 10.834030683403068, + "grad_norm": 4.985673904418945, + "learning_rate": 9.976706997670701e-06, + "loss": 0.0338, + "step": 23304 + }, + { + "epoch": 10.834495583449558, + "grad_norm": 0.9240430593490601, + "learning_rate": 9.9767059976706e-06, + "loss": 0.0082, + "step": 23305 + }, + { + "epoch": 10.83496048349605, + "grad_norm": 5.727251052856445, + "learning_rate": 9.976704997670501e-06, + "loss": 0.0549, + "step": 23306 + }, + { + "epoch": 10.835425383542539, + "grad_norm": 4.576471328735352, + "learning_rate": 9.9767039976704e-06, + "loss": 0.0579, + "step": 23307 + }, + { + "epoch": 10.835890283589029, + "grad_norm": 1.0089898109436035, + "learning_rate": 9.9767029976703e-06, + "loss": 0.0073, + "step": 23308 + }, + { + "epoch": 10.836355183635519, + "grad_norm": 3.9423468112945557, + "learning_rate": 9.976701997670201e-06, + "loss": 0.0434, + "step": 23309 + }, + { + "epoch": 10.836820083682008, + "grad_norm": 3.027498483657837, + "learning_rate": 9.9767009976701e-06, + "loss": 0.0197, + "step": 23310 + }, + { + "epoch": 10.837284983728498, + "grad_norm": 9.533075332641602, + "learning_rate": 9.976699997670001e-06, + "loss": 0.2237, + "step": 23311 + }, + { + "epoch": 10.837749883774988, + "grad_norm": 5.5226640701293945, + "learning_rate": 9.9766989976699e-06, + "loss": 0.1009, + "step": 23312 + }, + { + "epoch": 10.838214783821478, + "grad_norm": 6.641122817993164, + "learning_rate": 9.9766979976698e-06, + "loss": 0.0903, + "step": 23313 + }, + { + "epoch": 10.83867968386797, + "grad_norm": 6.690120697021484, + "learning_rate": 9.9766969976697e-06, + "loss": 0.1351, + "step": 23314 + }, + { + "epoch": 10.83914458391446, + "grad_norm": 6.874373435974121, + "learning_rate": 9.9766959976696e-06, + "loss": 0.1692, + "step": 23315 + }, + { + "epoch": 10.839609483960949, + "grad_norm": 3.0030548572540283, + "learning_rate": 9.976694997669501e-06, + "loss": 0.0375, + "step": 23316 + }, + { + "epoch": 10.840074384007439, + "grad_norm": 3.747391939163208, + "learning_rate": 9.976693997669402e-06, + "loss": 0.0385, + "step": 23317 + }, + { + "epoch": 10.840539284053929, + "grad_norm": 5.917975425720215, + "learning_rate": 9.9766929976693e-06, + "loss": 0.0835, + "step": 23318 + }, + { + "epoch": 10.841004184100418, + "grad_norm": 1.790113925933838, + "learning_rate": 9.9766919976692e-06, + "loss": 0.017, + "step": 23319 + }, + { + "epoch": 10.841469084146908, + "grad_norm": 1.9420784711837769, + "learning_rate": 9.9766909976691e-06, + "loss": 0.0144, + "step": 23320 + }, + { + "epoch": 10.841933984193398, + "grad_norm": 9.263314247131348, + "learning_rate": 9.976689997669e-06, + "loss": 0.1426, + "step": 23321 + }, + { + "epoch": 10.842398884239888, + "grad_norm": 7.897594928741455, + "learning_rate": 9.976688997668901e-06, + "loss": 0.154, + "step": 23322 + }, + { + "epoch": 10.84286378428638, + "grad_norm": 3.5062036514282227, + "learning_rate": 9.9766879976688e-06, + "loss": 0.042, + "step": 23323 + }, + { + "epoch": 10.84332868433287, + "grad_norm": 5.218997001647949, + "learning_rate": 9.976686997668701e-06, + "loss": 0.0877, + "step": 23324 + }, + { + "epoch": 10.843793584379359, + "grad_norm": 5.3311004638671875, + "learning_rate": 9.9766859976686e-06, + "loss": 0.1073, + "step": 23325 + }, + { + "epoch": 10.844258484425849, + "grad_norm": 3.312002420425415, + "learning_rate": 9.9766849976685e-06, + "loss": 0.0382, + "step": 23326 + }, + { + "epoch": 10.844723384472339, + "grad_norm": 1.4304006099700928, + "learning_rate": 9.9766839976684e-06, + "loss": 0.0115, + "step": 23327 + }, + { + "epoch": 10.845188284518828, + "grad_norm": 2.961081027984619, + "learning_rate": 9.9766829976683e-06, + "loss": 0.0198, + "step": 23328 + }, + { + "epoch": 10.845653184565318, + "grad_norm": 6.214178085327148, + "learning_rate": 9.976681997668201e-06, + "loss": 0.1402, + "step": 23329 + }, + { + "epoch": 10.846118084611808, + "grad_norm": 7.59570837020874, + "learning_rate": 9.9766809976681e-06, + "loss": 0.1332, + "step": 23330 + }, + { + "epoch": 10.846582984658298, + "grad_norm": 2.919166088104248, + "learning_rate": 9.976679997668e-06, + "loss": 0.0337, + "step": 23331 + }, + { + "epoch": 10.84704788470479, + "grad_norm": 4.112165451049805, + "learning_rate": 9.9766789976679e-06, + "loss": 0.0785, + "step": 23332 + }, + { + "epoch": 10.84751278475128, + "grad_norm": 4.589323997497559, + "learning_rate": 9.9766779976678e-06, + "loss": 0.0918, + "step": 23333 + }, + { + "epoch": 10.847977684797769, + "grad_norm": 2.9609267711639404, + "learning_rate": 9.9766769976677e-06, + "loss": 0.0269, + "step": 23334 + }, + { + "epoch": 10.848442584844259, + "grad_norm": 4.373684883117676, + "learning_rate": 9.976675997667601e-06, + "loss": 0.0861, + "step": 23335 + }, + { + "epoch": 10.848907484890749, + "grad_norm": 5.180375099182129, + "learning_rate": 9.9766749976675e-06, + "loss": 0.084, + "step": 23336 + }, + { + "epoch": 10.849372384937238, + "grad_norm": 4.710533142089844, + "learning_rate": 9.976673997667401e-06, + "loss": 0.0525, + "step": 23337 + }, + { + "epoch": 10.849837284983728, + "grad_norm": 5.849747180938721, + "learning_rate": 9.9766729976673e-06, + "loss": 0.1198, + "step": 23338 + }, + { + "epoch": 10.850302185030218, + "grad_norm": 2.75716233253479, + "learning_rate": 9.9766719976672e-06, + "loss": 0.0123, + "step": 23339 + }, + { + "epoch": 10.850767085076708, + "grad_norm": 1.5169776678085327, + "learning_rate": 9.976670997667101e-06, + "loss": 0.0157, + "step": 23340 + }, + { + "epoch": 10.8512319851232, + "grad_norm": 2.5667266845703125, + "learning_rate": 9.976669997667e-06, + "loss": 0.0261, + "step": 23341 + }, + { + "epoch": 10.851696885169689, + "grad_norm": 6.094749927520752, + "learning_rate": 9.976668997666901e-06, + "loss": 0.0887, + "step": 23342 + }, + { + "epoch": 10.852161785216179, + "grad_norm": 3.5568106174468994, + "learning_rate": 9.9766679976668e-06, + "loss": 0.0482, + "step": 23343 + }, + { + "epoch": 10.852626685262669, + "grad_norm": 3.4958479404449463, + "learning_rate": 9.9766669976667e-06, + "loss": 0.0274, + "step": 23344 + }, + { + "epoch": 10.853091585309159, + "grad_norm": 5.547826290130615, + "learning_rate": 9.9766659976666e-06, + "loss": 0.0305, + "step": 23345 + }, + { + "epoch": 10.853556485355648, + "grad_norm": 3.4095828533172607, + "learning_rate": 9.9766649976665e-06, + "loss": 0.0395, + "step": 23346 + }, + { + "epoch": 10.854021385402138, + "grad_norm": 6.034183025360107, + "learning_rate": 9.976663997666401e-06, + "loss": 0.1564, + "step": 23347 + }, + { + "epoch": 10.854486285448628, + "grad_norm": 5.114605903625488, + "learning_rate": 9.976662997666302e-06, + "loss": 0.0574, + "step": 23348 + }, + { + "epoch": 10.854951185495118, + "grad_norm": 6.1342082023620605, + "learning_rate": 9.9766619976662e-06, + "loss": 0.1754, + "step": 23349 + }, + { + "epoch": 10.85541608554161, + "grad_norm": 6.623970985412598, + "learning_rate": 9.976660997666101e-06, + "loss": 0.1488, + "step": 23350 + }, + { + "epoch": 10.855880985588099, + "grad_norm": 5.6255950927734375, + "learning_rate": 9.976659997666e-06, + "loss": 0.0544, + "step": 23351 + }, + { + "epoch": 10.856345885634589, + "grad_norm": 4.103719711303711, + "learning_rate": 9.9766589976659e-06, + "loss": 0.0378, + "step": 23352 + }, + { + "epoch": 10.856810785681079, + "grad_norm": 7.276285648345947, + "learning_rate": 9.976657997665801e-06, + "loss": 0.1369, + "step": 23353 + }, + { + "epoch": 10.857275685727569, + "grad_norm": 6.062945365905762, + "learning_rate": 9.9766569976657e-06, + "loss": 0.1439, + "step": 23354 + }, + { + "epoch": 10.857740585774058, + "grad_norm": 4.661067485809326, + "learning_rate": 9.976655997665601e-06, + "loss": 0.0785, + "step": 23355 + }, + { + "epoch": 10.858205485820548, + "grad_norm": 4.711824417114258, + "learning_rate": 9.9766549976655e-06, + "loss": 0.0562, + "step": 23356 + }, + { + "epoch": 10.858670385867038, + "grad_norm": 3.6507821083068848, + "learning_rate": 9.9766539976654e-06, + "loss": 0.0404, + "step": 23357 + }, + { + "epoch": 10.859135285913528, + "grad_norm": 3.6901659965515137, + "learning_rate": 9.9766529976653e-06, + "loss": 0.0496, + "step": 23358 + }, + { + "epoch": 10.85960018596002, + "grad_norm": 10.481907844543457, + "learning_rate": 9.9766519976652e-06, + "loss": 0.1333, + "step": 23359 + }, + { + "epoch": 10.860065086006509, + "grad_norm": 4.604835033416748, + "learning_rate": 9.976650997665101e-06, + "loss": 0.0757, + "step": 23360 + }, + { + "epoch": 10.860529986052999, + "grad_norm": 3.4261486530303955, + "learning_rate": 9.976649997665002e-06, + "loss": 0.0479, + "step": 23361 + }, + { + "epoch": 10.860994886099489, + "grad_norm": 7.737485885620117, + "learning_rate": 9.9766489976649e-06, + "loss": 0.0468, + "step": 23362 + }, + { + "epoch": 10.861459786145979, + "grad_norm": 3.0609278678894043, + "learning_rate": 9.9766479976648e-06, + "loss": 0.0317, + "step": 23363 + }, + { + "epoch": 10.861924686192468, + "grad_norm": 5.4966840744018555, + "learning_rate": 9.9766469976647e-06, + "loss": 0.0694, + "step": 23364 + }, + { + "epoch": 10.862389586238958, + "grad_norm": 4.304144382476807, + "learning_rate": 9.976645997664601e-06, + "loss": 0.0604, + "step": 23365 + }, + { + "epoch": 10.862854486285448, + "grad_norm": 7.096161842346191, + "learning_rate": 9.976644997664502e-06, + "loss": 0.1792, + "step": 23366 + }, + { + "epoch": 10.863319386331938, + "grad_norm": 4.429696559906006, + "learning_rate": 9.9766439976644e-06, + "loss": 0.0778, + "step": 23367 + }, + { + "epoch": 10.86378428637843, + "grad_norm": 5.099061965942383, + "learning_rate": 9.976642997664301e-06, + "loss": 0.0682, + "step": 23368 + }, + { + "epoch": 10.864249186424919, + "grad_norm": 3.13828706741333, + "learning_rate": 9.9766419976642e-06, + "loss": 0.0799, + "step": 23369 + }, + { + "epoch": 10.864714086471409, + "grad_norm": 5.7669596672058105, + "learning_rate": 9.9766409976641e-06, + "loss": 0.0759, + "step": 23370 + }, + { + "epoch": 10.865178986517899, + "grad_norm": 4.562223434448242, + "learning_rate": 9.976639997664e-06, + "loss": 0.032, + "step": 23371 + }, + { + "epoch": 10.865643886564389, + "grad_norm": 2.5040595531463623, + "learning_rate": 9.9766389976639e-06, + "loss": 0.0153, + "step": 23372 + }, + { + "epoch": 10.866108786610878, + "grad_norm": 3.7374913692474365, + "learning_rate": 9.976637997663801e-06, + "loss": 0.0399, + "step": 23373 + }, + { + "epoch": 10.866573686657368, + "grad_norm": 2.815378189086914, + "learning_rate": 9.9766369976637e-06, + "loss": 0.0331, + "step": 23374 + }, + { + "epoch": 10.867038586703858, + "grad_norm": 5.297181606292725, + "learning_rate": 9.9766359976636e-06, + "loss": 0.0528, + "step": 23375 + }, + { + "epoch": 10.86750348675035, + "grad_norm": 2.1579339504241943, + "learning_rate": 9.9766349976635e-06, + "loss": 0.0148, + "step": 23376 + }, + { + "epoch": 10.86796838679684, + "grad_norm": 2.7746169567108154, + "learning_rate": 9.9766339976634e-06, + "loss": 0.0241, + "step": 23377 + }, + { + "epoch": 10.868433286843329, + "grad_norm": 3.453883647918701, + "learning_rate": 9.976632997663301e-06, + "loss": 0.0557, + "step": 23378 + }, + { + "epoch": 10.868898186889819, + "grad_norm": 6.782881259918213, + "learning_rate": 9.976631997663202e-06, + "loss": 0.1734, + "step": 23379 + }, + { + "epoch": 10.869363086936309, + "grad_norm": 5.606607437133789, + "learning_rate": 9.9766309976631e-06, + "loss": 0.1103, + "step": 23380 + }, + { + "epoch": 10.869827986982799, + "grad_norm": 3.8692147731781006, + "learning_rate": 9.976629997663001e-06, + "loss": 0.0311, + "step": 23381 + }, + { + "epoch": 10.870292887029288, + "grad_norm": 5.750943183898926, + "learning_rate": 9.9766289976629e-06, + "loss": 0.0801, + "step": 23382 + }, + { + "epoch": 10.870757787075778, + "grad_norm": 3.609280824661255, + "learning_rate": 9.976627997662801e-06, + "loss": 0.0752, + "step": 23383 + }, + { + "epoch": 10.871222687122268, + "grad_norm": 7.535374641418457, + "learning_rate": 9.9766269976627e-06, + "loss": 0.1816, + "step": 23384 + }, + { + "epoch": 10.87168758716876, + "grad_norm": 5.976346969604492, + "learning_rate": 9.9766259976626e-06, + "loss": 0.0427, + "step": 23385 + }, + { + "epoch": 10.87215248721525, + "grad_norm": 3.281334400177002, + "learning_rate": 9.976624997662501e-06, + "loss": 0.0298, + "step": 23386 + }, + { + "epoch": 10.872617387261739, + "grad_norm": 6.404345512390137, + "learning_rate": 9.9766239976624e-06, + "loss": 0.0954, + "step": 23387 + }, + { + "epoch": 10.873082287308229, + "grad_norm": 4.1117072105407715, + "learning_rate": 9.9766229976623e-06, + "loss": 0.0415, + "step": 23388 + }, + { + "epoch": 10.873547187354719, + "grad_norm": 4.702449321746826, + "learning_rate": 9.9766219976622e-06, + "loss": 0.0877, + "step": 23389 + }, + { + "epoch": 10.874012087401209, + "grad_norm": 6.097254276275635, + "learning_rate": 9.9766209976621e-06, + "loss": 0.1145, + "step": 23390 + }, + { + "epoch": 10.874476987447698, + "grad_norm": 6.80827522277832, + "learning_rate": 9.976619997662001e-06, + "loss": 0.0222, + "step": 23391 + }, + { + "epoch": 10.874941887494188, + "grad_norm": 5.893721103668213, + "learning_rate": 9.976618997661902e-06, + "loss": 0.1178, + "step": 23392 + }, + { + "epoch": 10.87540678754068, + "grad_norm": 4.832674980163574, + "learning_rate": 9.9766179976618e-06, + "loss": 0.0713, + "step": 23393 + }, + { + "epoch": 10.87587168758717, + "grad_norm": 4.632593154907227, + "learning_rate": 9.976616997661701e-06, + "loss": 0.1107, + "step": 23394 + }, + { + "epoch": 10.87633658763366, + "grad_norm": 5.238542079925537, + "learning_rate": 9.9766159976616e-06, + "loss": 0.1164, + "step": 23395 + }, + { + "epoch": 10.876801487680149, + "grad_norm": 7.380821704864502, + "learning_rate": 9.9766149976615e-06, + "loss": 0.104, + "step": 23396 + }, + { + "epoch": 10.877266387726639, + "grad_norm": 3.7112154960632324, + "learning_rate": 9.9766139976614e-06, + "loss": 0.045, + "step": 23397 + }, + { + "epoch": 10.877731287773129, + "grad_norm": 4.238590717315674, + "learning_rate": 9.9766129976613e-06, + "loss": 0.0686, + "step": 23398 + }, + { + "epoch": 10.878196187819619, + "grad_norm": 3.038304090499878, + "learning_rate": 9.976611997661201e-06, + "loss": 0.0298, + "step": 23399 + }, + { + "epoch": 10.878661087866108, + "grad_norm": 4.403672695159912, + "learning_rate": 9.9766109976611e-06, + "loss": 0.0665, + "step": 23400 + }, + { + "epoch": 10.879125987912598, + "grad_norm": 4.682732582092285, + "learning_rate": 9.976609997661001e-06, + "loss": 0.0404, + "step": 23401 + }, + { + "epoch": 10.87959088795909, + "grad_norm": 3.1643078327178955, + "learning_rate": 9.9766089976609e-06, + "loss": 0.0297, + "step": 23402 + }, + { + "epoch": 10.88005578800558, + "grad_norm": 7.8006134033203125, + "learning_rate": 9.9766079976608e-06, + "loss": 0.138, + "step": 23403 + }, + { + "epoch": 10.88052068805207, + "grad_norm": 6.820908069610596, + "learning_rate": 9.976606997660701e-06, + "loss": 0.1, + "step": 23404 + }, + { + "epoch": 10.880985588098559, + "grad_norm": 3.2212586402893066, + "learning_rate": 9.976605997660602e-06, + "loss": 0.0275, + "step": 23405 + }, + { + "epoch": 10.881450488145049, + "grad_norm": 11.4873628616333, + "learning_rate": 9.9766049976605e-06, + "loss": 0.2106, + "step": 23406 + }, + { + "epoch": 10.881915388191539, + "grad_norm": 1.780319333076477, + "learning_rate": 9.9766039976604e-06, + "loss": 0.0144, + "step": 23407 + }, + { + "epoch": 10.882380288238028, + "grad_norm": 2.5948591232299805, + "learning_rate": 9.9766029976603e-06, + "loss": 0.0177, + "step": 23408 + }, + { + "epoch": 10.882845188284518, + "grad_norm": 3.8792564868927, + "learning_rate": 9.976601997660201e-06, + "loss": 0.0512, + "step": 23409 + }, + { + "epoch": 10.883310088331008, + "grad_norm": 9.601754188537598, + "learning_rate": 9.9766009976601e-06, + "loss": 0.212, + "step": 23410 + }, + { + "epoch": 10.8837749883775, + "grad_norm": 2.0720646381378174, + "learning_rate": 9.97659999766e-06, + "loss": 0.0174, + "step": 23411 + }, + { + "epoch": 10.88423988842399, + "grad_norm": 5.381400108337402, + "learning_rate": 9.976598997659901e-06, + "loss": 0.125, + "step": 23412 + }, + { + "epoch": 10.88470478847048, + "grad_norm": 4.347452163696289, + "learning_rate": 9.9765979976598e-06, + "loss": 0.0581, + "step": 23413 + }, + { + "epoch": 10.885169688516969, + "grad_norm": 3.8012444972991943, + "learning_rate": 9.976596997659701e-06, + "loss": 0.0335, + "step": 23414 + }, + { + "epoch": 10.885634588563459, + "grad_norm": 4.085598468780518, + "learning_rate": 9.9765959976596e-06, + "loss": 0.0576, + "step": 23415 + }, + { + "epoch": 10.886099488609949, + "grad_norm": 3.7579312324523926, + "learning_rate": 9.9765949976595e-06, + "loss": 0.0421, + "step": 23416 + }, + { + "epoch": 10.886564388656438, + "grad_norm": 4.757864952087402, + "learning_rate": 9.976593997659401e-06, + "loss": 0.0366, + "step": 23417 + }, + { + "epoch": 10.887029288702928, + "grad_norm": 4.093263626098633, + "learning_rate": 9.9765929976593e-06, + "loss": 0.0656, + "step": 23418 + }, + { + "epoch": 10.887494188749418, + "grad_norm": 5.470614910125732, + "learning_rate": 9.9765919976592e-06, + "loss": 0.0374, + "step": 23419 + }, + { + "epoch": 10.88795908879591, + "grad_norm": 3.6825132369995117, + "learning_rate": 9.9765909976591e-06, + "loss": 0.0441, + "step": 23420 + }, + { + "epoch": 10.8884239888424, + "grad_norm": 3.7156453132629395, + "learning_rate": 9.976589997659e-06, + "loss": 0.0467, + "step": 23421 + }, + { + "epoch": 10.88888888888889, + "grad_norm": 4.472161769866943, + "learning_rate": 9.976588997658901e-06, + "loss": 0.0613, + "step": 23422 + }, + { + "epoch": 10.889353788935379, + "grad_norm": 4.928775310516357, + "learning_rate": 9.9765879976588e-06, + "loss": 0.0415, + "step": 23423 + }, + { + "epoch": 10.889818688981869, + "grad_norm": 10.669852256774902, + "learning_rate": 9.9765869976587e-06, + "loss": 0.1329, + "step": 23424 + }, + { + "epoch": 10.890283589028359, + "grad_norm": 6.634924411773682, + "learning_rate": 9.976585997658601e-06, + "loss": 0.097, + "step": 23425 + }, + { + "epoch": 10.890748489074848, + "grad_norm": 8.879136085510254, + "learning_rate": 9.9765849976585e-06, + "loss": 0.0186, + "step": 23426 + }, + { + "epoch": 10.891213389121338, + "grad_norm": 4.104262828826904, + "learning_rate": 9.976583997658401e-06, + "loss": 0.0355, + "step": 23427 + }, + { + "epoch": 10.891678289167828, + "grad_norm": 7.541265487670898, + "learning_rate": 9.9765829976583e-06, + "loss": 0.1667, + "step": 23428 + }, + { + "epoch": 10.89214318921432, + "grad_norm": 2.7396509647369385, + "learning_rate": 9.9765819976582e-06, + "loss": 0.0247, + "step": 23429 + }, + { + "epoch": 10.89260808926081, + "grad_norm": 3.8281874656677246, + "learning_rate": 9.976580997658101e-06, + "loss": 0.0243, + "step": 23430 + }, + { + "epoch": 10.8930729893073, + "grad_norm": 4.617835998535156, + "learning_rate": 9.976579997658e-06, + "loss": 0.0781, + "step": 23431 + }, + { + "epoch": 10.893537889353789, + "grad_norm": 8.263401985168457, + "learning_rate": 9.976578997657901e-06, + "loss": 0.1887, + "step": 23432 + }, + { + "epoch": 10.894002789400279, + "grad_norm": 10.438554763793945, + "learning_rate": 9.9765779976578e-06, + "loss": 0.225, + "step": 23433 + }, + { + "epoch": 10.894467689446769, + "grad_norm": 3.3788888454437256, + "learning_rate": 9.9765769976577e-06, + "loss": 0.0317, + "step": 23434 + }, + { + "epoch": 10.894932589493258, + "grad_norm": 6.083755016326904, + "learning_rate": 9.976575997657601e-06, + "loss": 0.0715, + "step": 23435 + }, + { + "epoch": 10.895397489539748, + "grad_norm": 5.232116222381592, + "learning_rate": 9.976574997657502e-06, + "loss": 0.0756, + "step": 23436 + }, + { + "epoch": 10.895862389586238, + "grad_norm": 4.646817684173584, + "learning_rate": 9.9765739976574e-06, + "loss": 0.0665, + "step": 23437 + }, + { + "epoch": 10.89632728963273, + "grad_norm": 6.531705379486084, + "learning_rate": 9.9765729976573e-06, + "loss": 0.1049, + "step": 23438 + }, + { + "epoch": 10.89679218967922, + "grad_norm": 8.432374000549316, + "learning_rate": 9.9765719976572e-06, + "loss": 0.0733, + "step": 23439 + }, + { + "epoch": 10.89725708972571, + "grad_norm": 5.138523578643799, + "learning_rate": 9.9765709976571e-06, + "loss": 0.0919, + "step": 23440 + }, + { + "epoch": 10.897721989772199, + "grad_norm": 2.357590675354004, + "learning_rate": 9.976569997657e-06, + "loss": 0.0224, + "step": 23441 + }, + { + "epoch": 10.898186889818689, + "grad_norm": 4.862215042114258, + "learning_rate": 9.9765689976569e-06, + "loss": 0.0563, + "step": 23442 + }, + { + "epoch": 10.898651789865179, + "grad_norm": 4.538851737976074, + "learning_rate": 9.976567997656801e-06, + "loss": 0.0759, + "step": 23443 + }, + { + "epoch": 10.899116689911668, + "grad_norm": 6.545502185821533, + "learning_rate": 9.9765669976567e-06, + "loss": 0.0853, + "step": 23444 + }, + { + "epoch": 10.899581589958158, + "grad_norm": 5.123319625854492, + "learning_rate": 9.976565997656601e-06, + "loss": 0.0544, + "step": 23445 + }, + { + "epoch": 10.900046490004648, + "grad_norm": 5.881319522857666, + "learning_rate": 9.9765649976565e-06, + "loss": 0.0444, + "step": 23446 + }, + { + "epoch": 10.90051139005114, + "grad_norm": 7.069537162780762, + "learning_rate": 9.9765639976564e-06, + "loss": 0.1, + "step": 23447 + }, + { + "epoch": 10.90097629009763, + "grad_norm": 10.482901573181152, + "learning_rate": 9.976562997656301e-06, + "loss": 0.0936, + "step": 23448 + }, + { + "epoch": 10.90144119014412, + "grad_norm": 1.1665154695510864, + "learning_rate": 9.9765619976562e-06, + "loss": 0.0151, + "step": 23449 + }, + { + "epoch": 10.901906090190609, + "grad_norm": 3.7905726432800293, + "learning_rate": 9.976560997656101e-06, + "loss": 0.0473, + "step": 23450 + }, + { + "epoch": 10.902370990237099, + "grad_norm": 5.240987777709961, + "learning_rate": 9.976559997656e-06, + "loss": 0.0466, + "step": 23451 + }, + { + "epoch": 10.902835890283589, + "grad_norm": 3.4260261058807373, + "learning_rate": 9.9765589976559e-06, + "loss": 0.0311, + "step": 23452 + }, + { + "epoch": 10.903300790330078, + "grad_norm": 6.303274631500244, + "learning_rate": 9.9765579976558e-06, + "loss": 0.125, + "step": 23453 + }, + { + "epoch": 10.903765690376568, + "grad_norm": 5.639753818511963, + "learning_rate": 9.9765569976557e-06, + "loss": 0.1199, + "step": 23454 + }, + { + "epoch": 10.90423059042306, + "grad_norm": 6.187562942504883, + "learning_rate": 9.9765559976556e-06, + "loss": 0.1177, + "step": 23455 + }, + { + "epoch": 10.90469549046955, + "grad_norm": 7.5352373123168945, + "learning_rate": 9.976554997655501e-06, + "loss": 0.1493, + "step": 23456 + }, + { + "epoch": 10.90516039051604, + "grad_norm": 11.144113540649414, + "learning_rate": 9.9765539976554e-06, + "loss": 0.11, + "step": 23457 + }, + { + "epoch": 10.90562529056253, + "grad_norm": 6.68034553527832, + "learning_rate": 9.976552997655301e-06, + "loss": 0.1117, + "step": 23458 + }, + { + "epoch": 10.906090190609019, + "grad_norm": 7.9107465744018555, + "learning_rate": 9.9765519976552e-06, + "loss": 0.1822, + "step": 23459 + }, + { + "epoch": 10.906555090655509, + "grad_norm": 3.831094980239868, + "learning_rate": 9.9765509976551e-06, + "loss": 0.0283, + "step": 23460 + }, + { + "epoch": 10.907019990701999, + "grad_norm": 6.573317527770996, + "learning_rate": 9.976549997655001e-06, + "loss": 0.1154, + "step": 23461 + }, + { + "epoch": 10.907484890748488, + "grad_norm": 4.113778114318848, + "learning_rate": 9.9765489976549e-06, + "loss": 0.0642, + "step": 23462 + }, + { + "epoch": 10.907949790794978, + "grad_norm": 10.56522274017334, + "learning_rate": 9.976547997654801e-06, + "loss": 0.1475, + "step": 23463 + }, + { + "epoch": 10.90841469084147, + "grad_norm": 6.235517501831055, + "learning_rate": 9.9765469976547e-06, + "loss": 0.0768, + "step": 23464 + }, + { + "epoch": 10.90887959088796, + "grad_norm": 4.2637128829956055, + "learning_rate": 9.9765459976546e-06, + "loss": 0.0593, + "step": 23465 + }, + { + "epoch": 10.90934449093445, + "grad_norm": 5.293846130371094, + "learning_rate": 9.9765449976545e-06, + "loss": 0.0432, + "step": 23466 + }, + { + "epoch": 10.90980939098094, + "grad_norm": 5.51381254196167, + "learning_rate": 9.9765439976544e-06, + "loss": 0.0381, + "step": 23467 + }, + { + "epoch": 10.910274291027429, + "grad_norm": 4.733435153961182, + "learning_rate": 9.976542997654301e-06, + "loss": 0.0538, + "step": 23468 + }, + { + "epoch": 10.910739191073919, + "grad_norm": 1.3746850490570068, + "learning_rate": 9.976541997654202e-06, + "loss": 0.0106, + "step": 23469 + }, + { + "epoch": 10.911204091120409, + "grad_norm": 9.035642623901367, + "learning_rate": 9.9765409976541e-06, + "loss": 0.1108, + "step": 23470 + }, + { + "epoch": 10.911668991166898, + "grad_norm": 5.0794596672058105, + "learning_rate": 9.976539997654e-06, + "loss": 0.0845, + "step": 23471 + }, + { + "epoch": 10.91213389121339, + "grad_norm": 6.393058776855469, + "learning_rate": 9.9765389976539e-06, + "loss": 0.1136, + "step": 23472 + }, + { + "epoch": 10.91259879125988, + "grad_norm": 4.565347671508789, + "learning_rate": 9.9765379976538e-06, + "loss": 0.059, + "step": 23473 + }, + { + "epoch": 10.91306369130637, + "grad_norm": 5.340136528015137, + "learning_rate": 9.976536997653701e-06, + "loss": 0.0496, + "step": 23474 + }, + { + "epoch": 10.91352859135286, + "grad_norm": 4.459986686706543, + "learning_rate": 9.9765359976536e-06, + "loss": 0.0421, + "step": 23475 + }, + { + "epoch": 10.91399349139935, + "grad_norm": 1.782007098197937, + "learning_rate": 9.976534997653501e-06, + "loss": 0.0118, + "step": 23476 + }, + { + "epoch": 10.914458391445839, + "grad_norm": 5.006964206695557, + "learning_rate": 9.9765339976534e-06, + "loss": 0.0473, + "step": 23477 + }, + { + "epoch": 10.914923291492329, + "grad_norm": 3.8264405727386475, + "learning_rate": 9.9765329976533e-06, + "loss": 0.0581, + "step": 23478 + }, + { + "epoch": 10.915388191538819, + "grad_norm": 9.101631164550781, + "learning_rate": 9.9765319976532e-06, + "loss": 0.2449, + "step": 23479 + }, + { + "epoch": 10.915853091585308, + "grad_norm": 4.529843807220459, + "learning_rate": 9.9765309976531e-06, + "loss": 0.0337, + "step": 23480 + }, + { + "epoch": 10.9163179916318, + "grad_norm": 4.730900764465332, + "learning_rate": 9.976529997653001e-06, + "loss": 0.0734, + "step": 23481 + }, + { + "epoch": 10.91678289167829, + "grad_norm": 6.638482093811035, + "learning_rate": 9.9765289976529e-06, + "loss": 0.0641, + "step": 23482 + }, + { + "epoch": 10.91724779172478, + "grad_norm": 11.368840217590332, + "learning_rate": 9.9765279976528e-06, + "loss": 0.1042, + "step": 23483 + }, + { + "epoch": 10.91771269177127, + "grad_norm": 6.622750759124756, + "learning_rate": 9.9765269976527e-06, + "loss": 0.1107, + "step": 23484 + }, + { + "epoch": 10.91817759181776, + "grad_norm": 7.1800761222839355, + "learning_rate": 9.9765259976526e-06, + "loss": 0.135, + "step": 23485 + }, + { + "epoch": 10.918642491864249, + "grad_norm": 5.041162014007568, + "learning_rate": 9.976524997652501e-06, + "loss": 0.0605, + "step": 23486 + }, + { + "epoch": 10.919107391910739, + "grad_norm": 8.984223365783691, + "learning_rate": 9.976523997652402e-06, + "loss": 0.185, + "step": 23487 + }, + { + "epoch": 10.919572291957229, + "grad_norm": 6.222836971282959, + "learning_rate": 9.9765229976523e-06, + "loss": 0.1001, + "step": 23488 + }, + { + "epoch": 10.920037192003718, + "grad_norm": 4.31753396987915, + "learning_rate": 9.976521997652201e-06, + "loss": 0.0501, + "step": 23489 + }, + { + "epoch": 10.92050209205021, + "grad_norm": 3.598996639251709, + "learning_rate": 9.9765209976521e-06, + "loss": 0.0548, + "step": 23490 + }, + { + "epoch": 10.9209669920967, + "grad_norm": 4.886122703552246, + "learning_rate": 9.976519997652e-06, + "loss": 0.0714, + "step": 23491 + }, + { + "epoch": 10.92143189214319, + "grad_norm": 6.38684606552124, + "learning_rate": 9.9765189976519e-06, + "loss": 0.0538, + "step": 23492 + }, + { + "epoch": 10.92189679218968, + "grad_norm": 6.37807559967041, + "learning_rate": 9.9765179976518e-06, + "loss": 0.1143, + "step": 23493 + }, + { + "epoch": 10.92236169223617, + "grad_norm": 6.324448585510254, + "learning_rate": 9.976516997651701e-06, + "loss": 0.0996, + "step": 23494 + }, + { + "epoch": 10.922826592282659, + "grad_norm": 13.199552536010742, + "learning_rate": 9.9765159976516e-06, + "loss": 0.1537, + "step": 23495 + }, + { + "epoch": 10.923291492329149, + "grad_norm": 4.843209743499756, + "learning_rate": 9.9765149976515e-06, + "loss": 0.0583, + "step": 23496 + }, + { + "epoch": 10.923756392375639, + "grad_norm": 9.685622215270996, + "learning_rate": 9.9765139976514e-06, + "loss": 0.2364, + "step": 23497 + }, + { + "epoch": 10.924221292422128, + "grad_norm": 2.404282569885254, + "learning_rate": 9.9765129976513e-06, + "loss": 0.0218, + "step": 23498 + }, + { + "epoch": 10.92468619246862, + "grad_norm": 5.999081134796143, + "learning_rate": 9.976511997651201e-06, + "loss": 0.138, + "step": 23499 + }, + { + "epoch": 10.92515109251511, + "grad_norm": 5.276488304138184, + "learning_rate": 9.976510997651102e-06, + "loss": 0.0824, + "step": 23500 + }, + { + "epoch": 10.9256159925616, + "grad_norm": 4.560415267944336, + "learning_rate": 9.976509997651e-06, + "loss": 0.0708, + "step": 23501 + }, + { + "epoch": 10.92608089260809, + "grad_norm": 3.7484164237976074, + "learning_rate": 9.976508997650901e-06, + "loss": 0.0397, + "step": 23502 + }, + { + "epoch": 10.92654579265458, + "grad_norm": 11.80625057220459, + "learning_rate": 9.9765079976508e-06, + "loss": 0.0893, + "step": 23503 + }, + { + "epoch": 10.927010692701069, + "grad_norm": 2.236313819885254, + "learning_rate": 9.976506997650701e-06, + "loss": 0.025, + "step": 23504 + }, + { + "epoch": 10.927475592747559, + "grad_norm": 1.6909126043319702, + "learning_rate": 9.976505997650602e-06, + "loss": 0.0162, + "step": 23505 + }, + { + "epoch": 10.927940492794049, + "grad_norm": 5.9050750732421875, + "learning_rate": 9.9765049976505e-06, + "loss": 0.1303, + "step": 23506 + }, + { + "epoch": 10.928405392840538, + "grad_norm": 6.854994773864746, + "learning_rate": 9.976503997650401e-06, + "loss": 0.1575, + "step": 23507 + }, + { + "epoch": 10.92887029288703, + "grad_norm": 5.828646659851074, + "learning_rate": 9.9765029976503e-06, + "loss": 0.1557, + "step": 23508 + }, + { + "epoch": 10.92933519293352, + "grad_norm": 3.1218533515930176, + "learning_rate": 9.9765019976502e-06, + "loss": 0.0321, + "step": 23509 + }, + { + "epoch": 10.92980009298001, + "grad_norm": 6.022575378417969, + "learning_rate": 9.9765009976501e-06, + "loss": 0.07, + "step": 23510 + }, + { + "epoch": 10.9302649930265, + "grad_norm": 5.032196521759033, + "learning_rate": 9.97649999765e-06, + "loss": 0.0799, + "step": 23511 + }, + { + "epoch": 10.93072989307299, + "grad_norm": 5.609201908111572, + "learning_rate": 9.976498997649901e-06, + "loss": 0.129, + "step": 23512 + }, + { + "epoch": 10.931194793119479, + "grad_norm": 4.434935569763184, + "learning_rate": 9.976497997649802e-06, + "loss": 0.0349, + "step": 23513 + }, + { + "epoch": 10.931659693165969, + "grad_norm": 6.568381309509277, + "learning_rate": 9.9764969976497e-06, + "loss": 0.1094, + "step": 23514 + }, + { + "epoch": 10.932124593212459, + "grad_norm": 7.271627902984619, + "learning_rate": 9.9764959976496e-06, + "loss": 0.1035, + "step": 23515 + }, + { + "epoch": 10.932589493258948, + "grad_norm": 5.207662105560303, + "learning_rate": 9.9764949976495e-06, + "loss": 0.0911, + "step": 23516 + }, + { + "epoch": 10.93305439330544, + "grad_norm": 5.013957500457764, + "learning_rate": 9.976493997649401e-06, + "loss": 0.1158, + "step": 23517 + }, + { + "epoch": 10.93351929335193, + "grad_norm": 2.114596128463745, + "learning_rate": 9.976492997649302e-06, + "loss": 0.0161, + "step": 23518 + }, + { + "epoch": 10.93398419339842, + "grad_norm": 5.4505295753479, + "learning_rate": 9.9764919976492e-06, + "loss": 0.059, + "step": 23519 + }, + { + "epoch": 10.93444909344491, + "grad_norm": 5.292253017425537, + "learning_rate": 9.976490997649101e-06, + "loss": 0.0631, + "step": 23520 + }, + { + "epoch": 10.9349139934914, + "grad_norm": 3.4990651607513428, + "learning_rate": 9.976489997649e-06, + "loss": 0.05, + "step": 23521 + }, + { + "epoch": 10.935378893537889, + "grad_norm": 3.964167594909668, + "learning_rate": 9.9764889976489e-06, + "loss": 0.0353, + "step": 23522 + }, + { + "epoch": 10.935843793584379, + "grad_norm": 4.939726829528809, + "learning_rate": 9.9764879976488e-06, + "loss": 0.0759, + "step": 23523 + }, + { + "epoch": 10.936308693630869, + "grad_norm": 6.497065544128418, + "learning_rate": 9.9764869976487e-06, + "loss": 0.0791, + "step": 23524 + }, + { + "epoch": 10.936773593677358, + "grad_norm": 7.1444315910339355, + "learning_rate": 9.976485997648601e-06, + "loss": 0.1121, + "step": 23525 + }, + { + "epoch": 10.93723849372385, + "grad_norm": 5.938309669494629, + "learning_rate": 9.9764849976485e-06, + "loss": 0.101, + "step": 23526 + }, + { + "epoch": 10.93770339377034, + "grad_norm": 8.29107666015625, + "learning_rate": 9.9764839976484e-06, + "loss": 0.1872, + "step": 23527 + }, + { + "epoch": 10.93816829381683, + "grad_norm": 5.457608699798584, + "learning_rate": 9.9764829976483e-06, + "loss": 0.1098, + "step": 23528 + }, + { + "epoch": 10.93863319386332, + "grad_norm": 6.772934913635254, + "learning_rate": 9.9764819976482e-06, + "loss": 0.1433, + "step": 23529 + }, + { + "epoch": 10.93909809390981, + "grad_norm": 4.585312366485596, + "learning_rate": 9.976480997648101e-06, + "loss": 0.0294, + "step": 23530 + }, + { + "epoch": 10.939562993956299, + "grad_norm": 5.364400863647461, + "learning_rate": 9.976479997648002e-06, + "loss": 0.093, + "step": 23531 + }, + { + "epoch": 10.940027894002789, + "grad_norm": 0.7931203842163086, + "learning_rate": 9.9764789976479e-06, + "loss": 0.0054, + "step": 23532 + }, + { + "epoch": 10.940492794049279, + "grad_norm": 6.000156402587891, + "learning_rate": 9.976477997647801e-06, + "loss": 0.0806, + "step": 23533 + }, + { + "epoch": 10.94095769409577, + "grad_norm": 5.637688636779785, + "learning_rate": 9.9764769976477e-06, + "loss": 0.1207, + "step": 23534 + }, + { + "epoch": 10.94142259414226, + "grad_norm": 0.49160364270210266, + "learning_rate": 9.976475997647601e-06, + "loss": 0.0054, + "step": 23535 + }, + { + "epoch": 10.94188749418875, + "grad_norm": 3.8201587200164795, + "learning_rate": 9.9764749976475e-06, + "loss": 0.0266, + "step": 23536 + }, + { + "epoch": 10.94235239423524, + "grad_norm": 3.1490848064422607, + "learning_rate": 9.9764739976474e-06, + "loss": 0.0242, + "step": 23537 + }, + { + "epoch": 10.94281729428173, + "grad_norm": 6.711391448974609, + "learning_rate": 9.976472997647301e-06, + "loss": 0.1643, + "step": 23538 + }, + { + "epoch": 10.94328219432822, + "grad_norm": 3.6340487003326416, + "learning_rate": 9.9764719976472e-06, + "loss": 0.0369, + "step": 23539 + }, + { + "epoch": 10.943747094374709, + "grad_norm": 2.0406124591827393, + "learning_rate": 9.9764709976471e-06, + "loss": 0.018, + "step": 23540 + }, + { + "epoch": 10.944211994421199, + "grad_norm": 2.338155746459961, + "learning_rate": 9.976469997647e-06, + "loss": 0.0223, + "step": 23541 + }, + { + "epoch": 10.944676894467689, + "grad_norm": 5.996752738952637, + "learning_rate": 9.9764689976469e-06, + "loss": 0.0732, + "step": 23542 + }, + { + "epoch": 10.94514179451418, + "grad_norm": 5.456514835357666, + "learning_rate": 9.976467997646801e-06, + "loss": 0.1019, + "step": 23543 + }, + { + "epoch": 10.94560669456067, + "grad_norm": 4.61299991607666, + "learning_rate": 9.976466997646702e-06, + "loss": 0.0573, + "step": 23544 + }, + { + "epoch": 10.94607159460716, + "grad_norm": 7.661081790924072, + "learning_rate": 9.9764659976466e-06, + "loss": 0.1233, + "step": 23545 + }, + { + "epoch": 10.94653649465365, + "grad_norm": 4.221716403961182, + "learning_rate": 9.976464997646501e-06, + "loss": 0.0353, + "step": 23546 + }, + { + "epoch": 10.94700139470014, + "grad_norm": 4.651132583618164, + "learning_rate": 9.9764639976464e-06, + "loss": 0.0724, + "step": 23547 + }, + { + "epoch": 10.94746629474663, + "grad_norm": 4.2431321144104, + "learning_rate": 9.9764629976463e-06, + "loss": 0.0564, + "step": 23548 + }, + { + "epoch": 10.947931194793119, + "grad_norm": 6.066456317901611, + "learning_rate": 9.9764619976462e-06, + "loss": 0.0996, + "step": 23549 + }, + { + "epoch": 10.948396094839609, + "grad_norm": 7.564982891082764, + "learning_rate": 9.9764609976461e-06, + "loss": 0.1682, + "step": 23550 + }, + { + "epoch": 10.9488609948861, + "grad_norm": 4.342748165130615, + "learning_rate": 9.976459997646001e-06, + "loss": 0.0752, + "step": 23551 + }, + { + "epoch": 10.94932589493259, + "grad_norm": 3.5358998775482178, + "learning_rate": 9.9764589976459e-06, + "loss": 0.0354, + "step": 23552 + }, + { + "epoch": 10.94979079497908, + "grad_norm": 2.5433413982391357, + "learning_rate": 9.976457997645801e-06, + "loss": 0.0227, + "step": 23553 + }, + { + "epoch": 10.95025569502557, + "grad_norm": 4.168413162231445, + "learning_rate": 9.9764569976457e-06, + "loss": 0.0605, + "step": 23554 + }, + { + "epoch": 10.95072059507206, + "grad_norm": 5.40827751159668, + "learning_rate": 9.9764559976456e-06, + "loss": 0.112, + "step": 23555 + }, + { + "epoch": 10.95118549511855, + "grad_norm": 4.101483345031738, + "learning_rate": 9.976454997645501e-06, + "loss": 0.0529, + "step": 23556 + }, + { + "epoch": 10.95165039516504, + "grad_norm": 8.627839088439941, + "learning_rate": 9.976453997645402e-06, + "loss": 0.1506, + "step": 23557 + }, + { + "epoch": 10.952115295211529, + "grad_norm": 4.857481479644775, + "learning_rate": 9.9764529976453e-06, + "loss": 0.054, + "step": 23558 + }, + { + "epoch": 10.952580195258019, + "grad_norm": 4.179110527038574, + "learning_rate": 9.9764519976452e-06, + "loss": 0.0411, + "step": 23559 + }, + { + "epoch": 10.95304509530451, + "grad_norm": 6.394094944000244, + "learning_rate": 9.9764509976451e-06, + "loss": 0.128, + "step": 23560 + }, + { + "epoch": 10.953509995351, + "grad_norm": 10.38998031616211, + "learning_rate": 9.976449997645e-06, + "loss": 0.1947, + "step": 23561 + }, + { + "epoch": 10.95397489539749, + "grad_norm": 4.863638877868652, + "learning_rate": 9.9764489976449e-06, + "loss": 0.0547, + "step": 23562 + }, + { + "epoch": 10.95443979544398, + "grad_norm": 8.143190383911133, + "learning_rate": 9.9764479976448e-06, + "loss": 0.1873, + "step": 23563 + }, + { + "epoch": 10.95490469549047, + "grad_norm": 3.459043264389038, + "learning_rate": 9.976446997644701e-06, + "loss": 0.0389, + "step": 23564 + }, + { + "epoch": 10.95536959553696, + "grad_norm": 7.903997421264648, + "learning_rate": 9.9764459976446e-06, + "loss": 0.1941, + "step": 23565 + }, + { + "epoch": 10.95583449558345, + "grad_norm": 7.4779052734375, + "learning_rate": 9.976444997644501e-06, + "loss": 0.1832, + "step": 23566 + }, + { + "epoch": 10.956299395629939, + "grad_norm": 7.006007671356201, + "learning_rate": 9.9764439976444e-06, + "loss": 0.1128, + "step": 23567 + }, + { + "epoch": 10.956764295676429, + "grad_norm": 4.7871904373168945, + "learning_rate": 9.9764429976443e-06, + "loss": 0.0807, + "step": 23568 + }, + { + "epoch": 10.95722919572292, + "grad_norm": 5.484988212585449, + "learning_rate": 9.976441997644201e-06, + "loss": 0.1132, + "step": 23569 + }, + { + "epoch": 10.95769409576941, + "grad_norm": 2.569936990737915, + "learning_rate": 9.9764409976441e-06, + "loss": 0.0209, + "step": 23570 + }, + { + "epoch": 10.9581589958159, + "grad_norm": 6.296834468841553, + "learning_rate": 9.976439997644001e-06, + "loss": 0.1224, + "step": 23571 + }, + { + "epoch": 10.95862389586239, + "grad_norm": 5.848362922668457, + "learning_rate": 9.9764389976439e-06, + "loss": 0.0691, + "step": 23572 + }, + { + "epoch": 10.95908879590888, + "grad_norm": 3.589554786682129, + "learning_rate": 9.9764379976438e-06, + "loss": 0.038, + "step": 23573 + }, + { + "epoch": 10.95955369595537, + "grad_norm": 0.7616850137710571, + "learning_rate": 9.976436997643701e-06, + "loss": 0.0064, + "step": 23574 + }, + { + "epoch": 10.96001859600186, + "grad_norm": 3.0820040702819824, + "learning_rate": 9.9764359976436e-06, + "loss": 0.0232, + "step": 23575 + }, + { + "epoch": 10.960483496048349, + "grad_norm": 3.409356117248535, + "learning_rate": 9.9764349976435e-06, + "loss": 0.0272, + "step": 23576 + }, + { + "epoch": 10.960948396094839, + "grad_norm": 4.289454936981201, + "learning_rate": 9.976433997643401e-06, + "loss": 0.0449, + "step": 23577 + }, + { + "epoch": 10.96141329614133, + "grad_norm": 1.6660250425338745, + "learning_rate": 9.9764329976433e-06, + "loss": 0.0171, + "step": 23578 + }, + { + "epoch": 10.96187819618782, + "grad_norm": 4.042313098907471, + "learning_rate": 9.9764319976432e-06, + "loss": 0.069, + "step": 23579 + }, + { + "epoch": 10.96234309623431, + "grad_norm": 6.130354881286621, + "learning_rate": 9.9764309976431e-06, + "loss": 0.1026, + "step": 23580 + }, + { + "epoch": 10.9628079962808, + "grad_norm": 3.431337594985962, + "learning_rate": 9.976429997643e-06, + "loss": 0.0334, + "step": 23581 + }, + { + "epoch": 10.96327289632729, + "grad_norm": 5.107426643371582, + "learning_rate": 9.976428997642901e-06, + "loss": 0.0724, + "step": 23582 + }, + { + "epoch": 10.96373779637378, + "grad_norm": 3.6755714416503906, + "learning_rate": 9.9764279976428e-06, + "loss": 0.0563, + "step": 23583 + }, + { + "epoch": 10.96420269642027, + "grad_norm": 3.7152369022369385, + "learning_rate": 9.976426997642701e-06, + "loss": 0.0496, + "step": 23584 + }, + { + "epoch": 10.964667596466759, + "grad_norm": 3.490419864654541, + "learning_rate": 9.9764259976426e-06, + "loss": 0.0313, + "step": 23585 + }, + { + "epoch": 10.965132496513249, + "grad_norm": 4.306743144989014, + "learning_rate": 9.9764249976425e-06, + "loss": 0.0403, + "step": 23586 + }, + { + "epoch": 10.96559739655974, + "grad_norm": 5.2520527839660645, + "learning_rate": 9.976423997642401e-06, + "loss": 0.071, + "step": 23587 + }, + { + "epoch": 10.96606229660623, + "grad_norm": 3.966583013534546, + "learning_rate": 9.9764229976423e-06, + "loss": 0.029, + "step": 23588 + }, + { + "epoch": 10.96652719665272, + "grad_norm": 5.754523754119873, + "learning_rate": 9.976421997642201e-06, + "loss": 0.0915, + "step": 23589 + }, + { + "epoch": 10.96699209669921, + "grad_norm": 3.3614799976348877, + "learning_rate": 9.9764209976421e-06, + "loss": 0.0313, + "step": 23590 + }, + { + "epoch": 10.9674569967457, + "grad_norm": 6.8594560623168945, + "learning_rate": 9.976419997642e-06, + "loss": 0.1515, + "step": 23591 + }, + { + "epoch": 10.96792189679219, + "grad_norm": 2.8806333541870117, + "learning_rate": 9.9764189976419e-06, + "loss": 0.0266, + "step": 23592 + }, + { + "epoch": 10.96838679683868, + "grad_norm": 1.8185049295425415, + "learning_rate": 9.9764179976418e-06, + "loss": 0.0058, + "step": 23593 + }, + { + "epoch": 10.968851696885169, + "grad_norm": 4.502580642700195, + "learning_rate": 9.9764169976417e-06, + "loss": 0.059, + "step": 23594 + }, + { + "epoch": 10.969316596931659, + "grad_norm": 3.8504786491394043, + "learning_rate": 9.976415997641601e-06, + "loss": 0.0442, + "step": 23595 + }, + { + "epoch": 10.96978149697815, + "grad_norm": 5.825829029083252, + "learning_rate": 9.9764149976415e-06, + "loss": 0.1078, + "step": 23596 + }, + { + "epoch": 10.97024639702464, + "grad_norm": 2.072157144546509, + "learning_rate": 9.976413997641401e-06, + "loss": 0.022, + "step": 23597 + }, + { + "epoch": 10.97071129707113, + "grad_norm": 2.274336814880371, + "learning_rate": 9.9764129976413e-06, + "loss": 0.0237, + "step": 23598 + }, + { + "epoch": 10.97117619711762, + "grad_norm": 2.3404695987701416, + "learning_rate": 9.9764119976412e-06, + "loss": 0.0198, + "step": 23599 + }, + { + "epoch": 10.97164109716411, + "grad_norm": 7.911244869232178, + "learning_rate": 9.976410997641101e-06, + "loss": 0.103, + "step": 23600 + }, + { + "epoch": 10.9721059972106, + "grad_norm": 1.5692394971847534, + "learning_rate": 9.976409997641e-06, + "loss": 0.0106, + "step": 23601 + }, + { + "epoch": 10.97257089725709, + "grad_norm": 5.129601955413818, + "learning_rate": 9.976408997640901e-06, + "loss": 0.0873, + "step": 23602 + }, + { + "epoch": 10.973035797303579, + "grad_norm": 1.8201828002929688, + "learning_rate": 9.9764079976408e-06, + "loss": 0.0135, + "step": 23603 + }, + { + "epoch": 10.973500697350069, + "grad_norm": 7.500277042388916, + "learning_rate": 9.9764069976407e-06, + "loss": 0.0987, + "step": 23604 + }, + { + "epoch": 10.97396559739656, + "grad_norm": 6.24553108215332, + "learning_rate": 9.9764059976406e-06, + "loss": 0.0683, + "step": 23605 + }, + { + "epoch": 10.97443049744305, + "grad_norm": 10.0137357711792, + "learning_rate": 9.9764049976405e-06, + "loss": 0.0489, + "step": 23606 + }, + { + "epoch": 10.97489539748954, + "grad_norm": 7.024380207061768, + "learning_rate": 9.9764039976404e-06, + "loss": 0.165, + "step": 23607 + }, + { + "epoch": 10.97536029753603, + "grad_norm": 5.847963333129883, + "learning_rate": 9.976402997640301e-06, + "loss": 0.1424, + "step": 23608 + }, + { + "epoch": 10.97582519758252, + "grad_norm": 5.858828544616699, + "learning_rate": 9.9764019976402e-06, + "loss": 0.0935, + "step": 23609 + }, + { + "epoch": 10.97629009762901, + "grad_norm": 4.234710216522217, + "learning_rate": 9.976400997640101e-06, + "loss": 0.0345, + "step": 23610 + }, + { + "epoch": 10.9767549976755, + "grad_norm": 7.514207363128662, + "learning_rate": 9.97639999764e-06, + "loss": 0.2611, + "step": 23611 + }, + { + "epoch": 10.977219897721989, + "grad_norm": 5.241878986358643, + "learning_rate": 9.9763989976399e-06, + "loss": 0.0794, + "step": 23612 + }, + { + "epoch": 10.97768479776848, + "grad_norm": 6.456976413726807, + "learning_rate": 9.976397997639801e-06, + "loss": 0.1109, + "step": 23613 + }, + { + "epoch": 10.97814969781497, + "grad_norm": 2.757214069366455, + "learning_rate": 9.9763969976397e-06, + "loss": 0.0225, + "step": 23614 + }, + { + "epoch": 10.97861459786146, + "grad_norm": 4.752225875854492, + "learning_rate": 9.976395997639601e-06, + "loss": 0.0426, + "step": 23615 + }, + { + "epoch": 10.97907949790795, + "grad_norm": 9.151723861694336, + "learning_rate": 9.9763949976395e-06, + "loss": 0.0811, + "step": 23616 + }, + { + "epoch": 10.97954439795444, + "grad_norm": 0.9096987843513489, + "learning_rate": 9.9763939976394e-06, + "loss": 0.0099, + "step": 23617 + }, + { + "epoch": 10.98000929800093, + "grad_norm": 2.0813050270080566, + "learning_rate": 9.9763929976393e-06, + "loss": 0.0213, + "step": 23618 + }, + { + "epoch": 10.98047419804742, + "grad_norm": 7.198451519012451, + "learning_rate": 9.9763919976392e-06, + "loss": 0.0704, + "step": 23619 + }, + { + "epoch": 10.98093909809391, + "grad_norm": 5.786093711853027, + "learning_rate": 9.976390997639101e-06, + "loss": 0.0814, + "step": 23620 + }, + { + "epoch": 10.981403998140399, + "grad_norm": 2.2831220626831055, + "learning_rate": 9.976389997639002e-06, + "loss": 0.0248, + "step": 23621 + }, + { + "epoch": 10.98186889818689, + "grad_norm": 5.191592693328857, + "learning_rate": 9.9763889976389e-06, + "loss": 0.0858, + "step": 23622 + }, + { + "epoch": 10.98233379823338, + "grad_norm": 6.542749404907227, + "learning_rate": 9.9763879976388e-06, + "loss": 0.1207, + "step": 23623 + }, + { + "epoch": 10.98279869827987, + "grad_norm": 3.5006752014160156, + "learning_rate": 9.9763869976387e-06, + "loss": 0.0257, + "step": 23624 + }, + { + "epoch": 10.98326359832636, + "grad_norm": 3.9763917922973633, + "learning_rate": 9.9763859976386e-06, + "loss": 0.034, + "step": 23625 + }, + { + "epoch": 10.98372849837285, + "grad_norm": 6.739395618438721, + "learning_rate": 9.976384997638501e-06, + "loss": 0.1713, + "step": 23626 + }, + { + "epoch": 10.98419339841934, + "grad_norm": 1.8622658252716064, + "learning_rate": 9.9763839976384e-06, + "loss": 0.0106, + "step": 23627 + }, + { + "epoch": 10.98465829846583, + "grad_norm": 4.671448230743408, + "learning_rate": 9.976382997638301e-06, + "loss": 0.0696, + "step": 23628 + }, + { + "epoch": 10.98512319851232, + "grad_norm": 3.942401647567749, + "learning_rate": 9.9763819976382e-06, + "loss": 0.0238, + "step": 23629 + }, + { + "epoch": 10.98558809855881, + "grad_norm": 1.7253080606460571, + "learning_rate": 9.9763809976381e-06, + "loss": 0.0169, + "step": 23630 + }, + { + "epoch": 10.9860529986053, + "grad_norm": 5.7915263175964355, + "learning_rate": 9.976379997638e-06, + "loss": 0.1211, + "step": 23631 + }, + { + "epoch": 10.98651789865179, + "grad_norm": 6.9751386642456055, + "learning_rate": 9.9763789976379e-06, + "loss": 0.1649, + "step": 23632 + }, + { + "epoch": 10.98698279869828, + "grad_norm": 4.468760967254639, + "learning_rate": 9.976377997637801e-06, + "loss": 0.0645, + "step": 23633 + }, + { + "epoch": 10.98744769874477, + "grad_norm": 6.015045166015625, + "learning_rate": 9.9763769976377e-06, + "loss": 0.0668, + "step": 23634 + }, + { + "epoch": 10.98791259879126, + "grad_norm": 6.320126533508301, + "learning_rate": 9.9763759976376e-06, + "loss": 0.0888, + "step": 23635 + }, + { + "epoch": 10.98837749883775, + "grad_norm": 6.356434345245361, + "learning_rate": 9.9763749976375e-06, + "loss": 0.0816, + "step": 23636 + }, + { + "epoch": 10.98884239888424, + "grad_norm": 6.720874786376953, + "learning_rate": 9.9763739976374e-06, + "loss": 0.1071, + "step": 23637 + }, + { + "epoch": 10.98930729893073, + "grad_norm": 11.288761138916016, + "learning_rate": 9.976372997637301e-06, + "loss": 0.2379, + "step": 23638 + }, + { + "epoch": 10.98977219897722, + "grad_norm": 3.745063543319702, + "learning_rate": 9.976371997637202e-06, + "loss": 0.0505, + "step": 23639 + }, + { + "epoch": 10.99023709902371, + "grad_norm": 3.011228084564209, + "learning_rate": 9.9763709976371e-06, + "loss": 0.031, + "step": 23640 + }, + { + "epoch": 10.9907019990702, + "grad_norm": 7.060513019561768, + "learning_rate": 9.976369997637001e-06, + "loss": 0.1379, + "step": 23641 + }, + { + "epoch": 10.99116689911669, + "grad_norm": 9.371021270751953, + "learning_rate": 9.9763689976369e-06, + "loss": 0.1555, + "step": 23642 + }, + { + "epoch": 10.99163179916318, + "grad_norm": 3.8544769287109375, + "learning_rate": 9.9763679976368e-06, + "loss": 0.0595, + "step": 23643 + }, + { + "epoch": 10.99209669920967, + "grad_norm": 5.9126482009887695, + "learning_rate": 9.9763669976367e-06, + "loss": 0.1224, + "step": 23644 + }, + { + "epoch": 10.99256159925616, + "grad_norm": 4.4932732582092285, + "learning_rate": 9.9763659976366e-06, + "loss": 0.0477, + "step": 23645 + }, + { + "epoch": 10.99302649930265, + "grad_norm": 6.516061782836914, + "learning_rate": 9.976364997636501e-06, + "loss": 0.1172, + "step": 23646 + }, + { + "epoch": 10.99349139934914, + "grad_norm": 6.380131244659424, + "learning_rate": 9.9763639976364e-06, + "loss": 0.0382, + "step": 23647 + }, + { + "epoch": 10.99395629939563, + "grad_norm": 4.73848819732666, + "learning_rate": 9.9763629976363e-06, + "loss": 0.042, + "step": 23648 + }, + { + "epoch": 10.99442119944212, + "grad_norm": 1.8118489980697632, + "learning_rate": 9.9763619976362e-06, + "loss": 0.0156, + "step": 23649 + }, + { + "epoch": 10.99488609948861, + "grad_norm": 7.784886360168457, + "learning_rate": 9.9763609976361e-06, + "loss": 0.1346, + "step": 23650 + }, + { + "epoch": 10.9953509995351, + "grad_norm": 4.039823532104492, + "learning_rate": 9.976359997636001e-06, + "loss": 0.061, + "step": 23651 + }, + { + "epoch": 10.99581589958159, + "grad_norm": 7.1000165939331055, + "learning_rate": 9.976358997635902e-06, + "loss": 0.0897, + "step": 23652 + }, + { + "epoch": 10.99628079962808, + "grad_norm": 5.504733562469482, + "learning_rate": 9.9763579976358e-06, + "loss": 0.1475, + "step": 23653 + }, + { + "epoch": 10.99674569967457, + "grad_norm": 4.376578330993652, + "learning_rate": 9.976356997635701e-06, + "loss": 0.0684, + "step": 23654 + }, + { + "epoch": 10.99721059972106, + "grad_norm": 5.293181419372559, + "learning_rate": 9.9763559976356e-06, + "loss": 0.0718, + "step": 23655 + }, + { + "epoch": 10.99767549976755, + "grad_norm": 3.7189300060272217, + "learning_rate": 9.976354997635501e-06, + "loss": 0.0294, + "step": 23656 + }, + { + "epoch": 10.99814039981404, + "grad_norm": 8.44052791595459, + "learning_rate": 9.9763539976354e-06, + "loss": 0.2573, + "step": 23657 + }, + { + "epoch": 10.99860529986053, + "grad_norm": 6.638973712921143, + "learning_rate": 9.9763529976353e-06, + "loss": 0.0423, + "step": 23658 + }, + { + "epoch": 10.99907019990702, + "grad_norm": 4.436564922332764, + "learning_rate": 9.976351997635201e-06, + "loss": 0.0751, + "step": 23659 + }, + { + "epoch": 10.99953509995351, + "grad_norm": 3.552579641342163, + "learning_rate": 9.9763509976351e-06, + "loss": 0.046, + "step": 23660 + }, + { + "epoch": 11.0, + "grad_norm": 14.659565925598145, + "learning_rate": 9.976349997635e-06, + "loss": 0.3776, + "step": 23661 + }, + { + "epoch": 11.00046490004649, + "grad_norm": 2.143428325653076, + "learning_rate": 9.9763489976349e-06, + "loss": 0.0172, + "step": 23662 + }, + { + "epoch": 11.00092980009298, + "grad_norm": 3.886744260787964, + "learning_rate": 9.9763479976348e-06, + "loss": 0.0693, + "step": 23663 + }, + { + "epoch": 11.00139470013947, + "grad_norm": 2.3730103969573975, + "learning_rate": 9.976346997634701e-06, + "loss": 0.0158, + "step": 23664 + }, + { + "epoch": 11.00185960018596, + "grad_norm": 4.298160552978516, + "learning_rate": 9.976345997634602e-06, + "loss": 0.0654, + "step": 23665 + }, + { + "epoch": 11.00232450023245, + "grad_norm": 4.74788236618042, + "learning_rate": 9.9763449976345e-06, + "loss": 0.0764, + "step": 23666 + }, + { + "epoch": 11.00278940027894, + "grad_norm": 3.4988021850585938, + "learning_rate": 9.9763439976344e-06, + "loss": 0.0436, + "step": 23667 + }, + { + "epoch": 11.00325430032543, + "grad_norm": 1.8612656593322754, + "learning_rate": 9.9763429976343e-06, + "loss": 0.0166, + "step": 23668 + }, + { + "epoch": 11.00371920037192, + "grad_norm": 6.018458366394043, + "learning_rate": 9.976341997634201e-06, + "loss": 0.0658, + "step": 23669 + }, + { + "epoch": 11.00418410041841, + "grad_norm": 6.44950008392334, + "learning_rate": 9.976340997634102e-06, + "loss": 0.0907, + "step": 23670 + }, + { + "epoch": 11.0046490004649, + "grad_norm": 8.187679290771484, + "learning_rate": 9.976339997634e-06, + "loss": 0.1618, + "step": 23671 + }, + { + "epoch": 11.00511390051139, + "grad_norm": 6.600465297698975, + "learning_rate": 9.976338997633901e-06, + "loss": 0.1111, + "step": 23672 + }, + { + "epoch": 11.00557880055788, + "grad_norm": 7.643665313720703, + "learning_rate": 9.9763379976338e-06, + "loss": 0.035, + "step": 23673 + }, + { + "epoch": 11.00604370060437, + "grad_norm": 5.099341869354248, + "learning_rate": 9.976336997633701e-06, + "loss": 0.0693, + "step": 23674 + }, + { + "epoch": 11.00650860065086, + "grad_norm": 3.122742176055908, + "learning_rate": 9.9763359976336e-06, + "loss": 0.0382, + "step": 23675 + }, + { + "epoch": 11.00697350069735, + "grad_norm": 2.286689043045044, + "learning_rate": 9.9763349976335e-06, + "loss": 0.0193, + "step": 23676 + }, + { + "epoch": 11.00743840074384, + "grad_norm": 8.253342628479004, + "learning_rate": 9.976333997633401e-06, + "loss": 0.1152, + "step": 23677 + }, + { + "epoch": 11.00790330079033, + "grad_norm": 5.856411933898926, + "learning_rate": 9.9763329976333e-06, + "loss": 0.0825, + "step": 23678 + }, + { + "epoch": 11.00836820083682, + "grad_norm": 2.1172101497650146, + "learning_rate": 9.9763319976332e-06, + "loss": 0.0202, + "step": 23679 + }, + { + "epoch": 11.00883310088331, + "grad_norm": 1.9185233116149902, + "learning_rate": 9.9763309976331e-06, + "loss": 0.0223, + "step": 23680 + }, + { + "epoch": 11.0092980009298, + "grad_norm": 2.0519702434539795, + "learning_rate": 9.976329997633e-06, + "loss": 0.0133, + "step": 23681 + }, + { + "epoch": 11.00976290097629, + "grad_norm": 5.514898777008057, + "learning_rate": 9.976328997632901e-06, + "loss": 0.0966, + "step": 23682 + }, + { + "epoch": 11.010227801022781, + "grad_norm": 3.4591732025146484, + "learning_rate": 9.976327997632802e-06, + "loss": 0.0166, + "step": 23683 + }, + { + "epoch": 11.01069270106927, + "grad_norm": 2.2959144115448, + "learning_rate": 9.9763269976327e-06, + "loss": 0.0169, + "step": 23684 + }, + { + "epoch": 11.01115760111576, + "grad_norm": 4.429865837097168, + "learning_rate": 9.976325997632601e-06, + "loss": 0.0683, + "step": 23685 + }, + { + "epoch": 11.01162250116225, + "grad_norm": 5.821037292480469, + "learning_rate": 9.9763249976325e-06, + "loss": 0.0495, + "step": 23686 + }, + { + "epoch": 11.01208740120874, + "grad_norm": 1.0773422718048096, + "learning_rate": 9.976323997632401e-06, + "loss": 0.0089, + "step": 23687 + }, + { + "epoch": 11.01255230125523, + "grad_norm": 4.9012651443481445, + "learning_rate": 9.9763229976323e-06, + "loss": 0.0613, + "step": 23688 + }, + { + "epoch": 11.01301720130172, + "grad_norm": 4.4320902824401855, + "learning_rate": 9.9763219976322e-06, + "loss": 0.0584, + "step": 23689 + }, + { + "epoch": 11.01348210134821, + "grad_norm": 4.093812465667725, + "learning_rate": 9.976320997632101e-06, + "loss": 0.0471, + "step": 23690 + }, + { + "epoch": 11.0139470013947, + "grad_norm": 12.43770980834961, + "learning_rate": 9.976319997632e-06, + "loss": 0.2402, + "step": 23691 + }, + { + "epoch": 11.014411901441191, + "grad_norm": 5.421019077301025, + "learning_rate": 9.9763189976319e-06, + "loss": 0.0795, + "step": 23692 + }, + { + "epoch": 11.01487680148768, + "grad_norm": 5.575109481811523, + "learning_rate": 9.9763179976318e-06, + "loss": 0.0669, + "step": 23693 + }, + { + "epoch": 11.01534170153417, + "grad_norm": 3.6416053771972656, + "learning_rate": 9.9763169976317e-06, + "loss": 0.035, + "step": 23694 + }, + { + "epoch": 11.01580660158066, + "grad_norm": 5.364793300628662, + "learning_rate": 9.976315997631601e-06, + "loss": 0.0499, + "step": 23695 + }, + { + "epoch": 11.01627150162715, + "grad_norm": 7.897280693054199, + "learning_rate": 9.976314997631502e-06, + "loss": 0.1953, + "step": 23696 + }, + { + "epoch": 11.01673640167364, + "grad_norm": 3.0965914726257324, + "learning_rate": 9.9763139976314e-06, + "loss": 0.0374, + "step": 23697 + }, + { + "epoch": 11.01720130172013, + "grad_norm": 2.518986940383911, + "learning_rate": 9.976312997631301e-06, + "loss": 0.0197, + "step": 23698 + }, + { + "epoch": 11.01766620176662, + "grad_norm": 6.817248344421387, + "learning_rate": 9.9763119976312e-06, + "loss": 0.0901, + "step": 23699 + }, + { + "epoch": 11.01813110181311, + "grad_norm": 2.1560099124908447, + "learning_rate": 9.9763109976311e-06, + "loss": 0.0196, + "step": 23700 + }, + { + "epoch": 11.018596001859601, + "grad_norm": 2.6080429553985596, + "learning_rate": 9.976309997631e-06, + "loss": 0.0234, + "step": 23701 + }, + { + "epoch": 11.01906090190609, + "grad_norm": 2.052403450012207, + "learning_rate": 9.9763089976309e-06, + "loss": 0.0192, + "step": 23702 + }, + { + "epoch": 11.01952580195258, + "grad_norm": 2.38962984085083, + "learning_rate": 9.976307997630801e-06, + "loss": 0.0172, + "step": 23703 + }, + { + "epoch": 11.01999070199907, + "grad_norm": 2.1350479125976562, + "learning_rate": 9.9763069976307e-06, + "loss": 0.0179, + "step": 23704 + }, + { + "epoch": 11.02045560204556, + "grad_norm": 2.337048292160034, + "learning_rate": 9.976305997630601e-06, + "loss": 0.0404, + "step": 23705 + }, + { + "epoch": 11.02092050209205, + "grad_norm": 2.700029134750366, + "learning_rate": 9.9763049976305e-06, + "loss": 0.0276, + "step": 23706 + }, + { + "epoch": 11.02138540213854, + "grad_norm": 2.886369228363037, + "learning_rate": 9.9763039976304e-06, + "loss": 0.031, + "step": 23707 + }, + { + "epoch": 11.02185030218503, + "grad_norm": 2.7586252689361572, + "learning_rate": 9.976302997630301e-06, + "loss": 0.0285, + "step": 23708 + }, + { + "epoch": 11.02231520223152, + "grad_norm": 3.910693883895874, + "learning_rate": 9.976301997630202e-06, + "loss": 0.0544, + "step": 23709 + }, + { + "epoch": 11.022780102278011, + "grad_norm": 7.140404224395752, + "learning_rate": 9.9763009976301e-06, + "loss": 0.189, + "step": 23710 + }, + { + "epoch": 11.0232450023245, + "grad_norm": 4.893030166625977, + "learning_rate": 9.97629999763e-06, + "loss": 0.0634, + "step": 23711 + }, + { + "epoch": 11.02370990237099, + "grad_norm": 1.3308613300323486, + "learning_rate": 9.9762989976299e-06, + "loss": 0.0083, + "step": 23712 + }, + { + "epoch": 11.02417480241748, + "grad_norm": 6.524997711181641, + "learning_rate": 9.9762979976298e-06, + "loss": 0.1629, + "step": 23713 + }, + { + "epoch": 11.02463970246397, + "grad_norm": 2.0446410179138184, + "learning_rate": 9.9762969976297e-06, + "loss": 0.0154, + "step": 23714 + }, + { + "epoch": 11.02510460251046, + "grad_norm": 4.575623512268066, + "learning_rate": 9.9762959976296e-06, + "loss": 0.0593, + "step": 23715 + }, + { + "epoch": 11.02556950255695, + "grad_norm": 3.2743849754333496, + "learning_rate": 9.976294997629501e-06, + "loss": 0.0451, + "step": 23716 + }, + { + "epoch": 11.02603440260344, + "grad_norm": 5.622649669647217, + "learning_rate": 9.9762939976294e-06, + "loss": 0.0499, + "step": 23717 + }, + { + "epoch": 11.02649930264993, + "grad_norm": 3.3976850509643555, + "learning_rate": 9.976292997629301e-06, + "loss": 0.0334, + "step": 23718 + }, + { + "epoch": 11.026964202696421, + "grad_norm": 5.297150611877441, + "learning_rate": 9.9762919976292e-06, + "loss": 0.1192, + "step": 23719 + }, + { + "epoch": 11.02742910274291, + "grad_norm": 4.221357822418213, + "learning_rate": 9.9762909976291e-06, + "loss": 0.0289, + "step": 23720 + }, + { + "epoch": 11.0278940027894, + "grad_norm": 3.7685396671295166, + "learning_rate": 9.976289997629001e-06, + "loss": 0.0385, + "step": 23721 + }, + { + "epoch": 11.02835890283589, + "grad_norm": 1.4069068431854248, + "learning_rate": 9.9762889976289e-06, + "loss": 0.0136, + "step": 23722 + }, + { + "epoch": 11.02882380288238, + "grad_norm": 4.321125507354736, + "learning_rate": 9.976287997628801e-06, + "loss": 0.0524, + "step": 23723 + }, + { + "epoch": 11.02928870292887, + "grad_norm": 3.83821702003479, + "learning_rate": 9.9762869976287e-06, + "loss": 0.0312, + "step": 23724 + }, + { + "epoch": 11.02975360297536, + "grad_norm": 2.7975518703460693, + "learning_rate": 9.9762859976286e-06, + "loss": 0.029, + "step": 23725 + }, + { + "epoch": 11.03021850302185, + "grad_norm": 1.4329569339752197, + "learning_rate": 9.9762849976285e-06, + "loss": 0.0133, + "step": 23726 + }, + { + "epoch": 11.030683403068341, + "grad_norm": 3.531139612197876, + "learning_rate": 9.9762839976284e-06, + "loss": 0.035, + "step": 23727 + }, + { + "epoch": 11.031148303114831, + "grad_norm": 6.101874351501465, + "learning_rate": 9.9762829976283e-06, + "loss": 0.113, + "step": 23728 + }, + { + "epoch": 11.03161320316132, + "grad_norm": 3.393841505050659, + "learning_rate": 9.976281997628201e-06, + "loss": 0.0253, + "step": 23729 + }, + { + "epoch": 11.03207810320781, + "grad_norm": 4.872150421142578, + "learning_rate": 9.9762809976281e-06, + "loss": 0.0751, + "step": 23730 + }, + { + "epoch": 11.0325430032543, + "grad_norm": 1.468367576599121, + "learning_rate": 9.976279997628e-06, + "loss": 0.0113, + "step": 23731 + }, + { + "epoch": 11.03300790330079, + "grad_norm": 4.5570831298828125, + "learning_rate": 9.9762789976279e-06, + "loss": 0.0439, + "step": 23732 + }, + { + "epoch": 11.03347280334728, + "grad_norm": 3.949658155441284, + "learning_rate": 9.9762779976278e-06, + "loss": 0.0365, + "step": 23733 + }, + { + "epoch": 11.03393770339377, + "grad_norm": 3.6755056381225586, + "learning_rate": 9.976276997627701e-06, + "loss": 0.0386, + "step": 23734 + }, + { + "epoch": 11.03440260344026, + "grad_norm": 3.644641637802124, + "learning_rate": 9.9762759976276e-06, + "loss": 0.0342, + "step": 23735 + }, + { + "epoch": 11.034867503486751, + "grad_norm": 4.012454509735107, + "learning_rate": 9.976274997627501e-06, + "loss": 0.0406, + "step": 23736 + }, + { + "epoch": 11.035332403533241, + "grad_norm": 2.5407652854919434, + "learning_rate": 9.9762739976274e-06, + "loss": 0.0255, + "step": 23737 + }, + { + "epoch": 11.03579730357973, + "grad_norm": 6.044159412384033, + "learning_rate": 9.9762729976273e-06, + "loss": 0.0286, + "step": 23738 + }, + { + "epoch": 11.03626220362622, + "grad_norm": 5.884697437286377, + "learning_rate": 9.976271997627201e-06, + "loss": 0.102, + "step": 23739 + }, + { + "epoch": 11.03672710367271, + "grad_norm": 8.883275985717773, + "learning_rate": 9.9762709976271e-06, + "loss": 0.1004, + "step": 23740 + }, + { + "epoch": 11.0371920037192, + "grad_norm": 10.679295539855957, + "learning_rate": 9.976269997627001e-06, + "loss": 0.0334, + "step": 23741 + }, + { + "epoch": 11.03765690376569, + "grad_norm": 3.043168306350708, + "learning_rate": 9.9762689976269e-06, + "loss": 0.0271, + "step": 23742 + }, + { + "epoch": 11.03812180381218, + "grad_norm": 3.470738410949707, + "learning_rate": 9.9762679976268e-06, + "loss": 0.0291, + "step": 23743 + }, + { + "epoch": 11.03858670385867, + "grad_norm": 12.31995677947998, + "learning_rate": 9.9762669976267e-06, + "loss": 0.0665, + "step": 23744 + }, + { + "epoch": 11.039051603905161, + "grad_norm": 3.249464511871338, + "learning_rate": 9.9762659976266e-06, + "loss": 0.0383, + "step": 23745 + }, + { + "epoch": 11.039516503951651, + "grad_norm": 3.5084896087646484, + "learning_rate": 9.9762649976265e-06, + "loss": 0.0467, + "step": 23746 + }, + { + "epoch": 11.03998140399814, + "grad_norm": 3.904383420944214, + "learning_rate": 9.976263997626401e-06, + "loss": 0.0494, + "step": 23747 + }, + { + "epoch": 11.04044630404463, + "grad_norm": 3.2501273155212402, + "learning_rate": 9.9762629976263e-06, + "loss": 0.0363, + "step": 23748 + }, + { + "epoch": 11.04091120409112, + "grad_norm": 5.493012428283691, + "learning_rate": 9.976261997626201e-06, + "loss": 0.0633, + "step": 23749 + }, + { + "epoch": 11.04137610413761, + "grad_norm": 7.523619651794434, + "learning_rate": 9.9762609976261e-06, + "loss": 0.2109, + "step": 23750 + }, + { + "epoch": 11.0418410041841, + "grad_norm": 6.197702884674072, + "learning_rate": 9.976259997626e-06, + "loss": 0.0546, + "step": 23751 + }, + { + "epoch": 11.04230590423059, + "grad_norm": 5.987481594085693, + "learning_rate": 9.976258997625901e-06, + "loss": 0.1157, + "step": 23752 + }, + { + "epoch": 11.04277080427708, + "grad_norm": 8.186620712280273, + "learning_rate": 9.9762579976258e-06, + "loss": 0.0643, + "step": 23753 + }, + { + "epoch": 11.043235704323571, + "grad_norm": 1.4558444023132324, + "learning_rate": 9.976256997625701e-06, + "loss": 0.0089, + "step": 23754 + }, + { + "epoch": 11.043700604370061, + "grad_norm": 3.348477363586426, + "learning_rate": 9.9762559976256e-06, + "loss": 0.0278, + "step": 23755 + }, + { + "epoch": 11.04416550441655, + "grad_norm": 4.388802528381348, + "learning_rate": 9.9762549976255e-06, + "loss": 0.0689, + "step": 23756 + }, + { + "epoch": 11.04463040446304, + "grad_norm": 3.546983003616333, + "learning_rate": 9.9762539976254e-06, + "loss": 0.0312, + "step": 23757 + }, + { + "epoch": 11.04509530450953, + "grad_norm": 11.113842010498047, + "learning_rate": 9.9762529976253e-06, + "loss": 0.1736, + "step": 23758 + }, + { + "epoch": 11.04556020455602, + "grad_norm": 4.964999675750732, + "learning_rate": 9.976251997625201e-06, + "loss": 0.0464, + "step": 23759 + }, + { + "epoch": 11.04602510460251, + "grad_norm": 2.829275131225586, + "learning_rate": 9.976250997625102e-06, + "loss": 0.0271, + "step": 23760 + }, + { + "epoch": 11.046490004649, + "grad_norm": 5.13523530960083, + "learning_rate": 9.976249997625e-06, + "loss": 0.0706, + "step": 23761 + }, + { + "epoch": 11.046954904695491, + "grad_norm": 6.139721870422363, + "learning_rate": 9.976248997624901e-06, + "loss": 0.1225, + "step": 23762 + }, + { + "epoch": 11.047419804741981, + "grad_norm": 5.010382175445557, + "learning_rate": 9.9762479976248e-06, + "loss": 0.1098, + "step": 23763 + }, + { + "epoch": 11.047884704788471, + "grad_norm": 3.693803548812866, + "learning_rate": 9.9762469976247e-06, + "loss": 0.0381, + "step": 23764 + }, + { + "epoch": 11.04834960483496, + "grad_norm": 8.21963882446289, + "learning_rate": 9.976245997624601e-06, + "loss": 0.1637, + "step": 23765 + }, + { + "epoch": 11.04881450488145, + "grad_norm": 3.317018508911133, + "learning_rate": 9.9762449976245e-06, + "loss": 0.0351, + "step": 23766 + }, + { + "epoch": 11.04927940492794, + "grad_norm": 1.6582062244415283, + "learning_rate": 9.976243997624401e-06, + "loss": 0.0121, + "step": 23767 + }, + { + "epoch": 11.04974430497443, + "grad_norm": 5.083511829376221, + "learning_rate": 9.9762429976243e-06, + "loss": 0.0746, + "step": 23768 + }, + { + "epoch": 11.05020920502092, + "grad_norm": 5.6077165603637695, + "learning_rate": 9.9762419976242e-06, + "loss": 0.1116, + "step": 23769 + }, + { + "epoch": 11.05067410506741, + "grad_norm": 2.151237964630127, + "learning_rate": 9.9762409976241e-06, + "loss": 0.0173, + "step": 23770 + }, + { + "epoch": 11.051139005113901, + "grad_norm": 2.955934524536133, + "learning_rate": 9.976239997624e-06, + "loss": 0.0171, + "step": 23771 + }, + { + "epoch": 11.051603905160391, + "grad_norm": 4.5791335105896, + "learning_rate": 9.976238997623901e-06, + "loss": 0.0844, + "step": 23772 + }, + { + "epoch": 11.052068805206881, + "grad_norm": 0.1578976809978485, + "learning_rate": 9.976237997623802e-06, + "loss": 0.0012, + "step": 23773 + }, + { + "epoch": 11.05253370525337, + "grad_norm": 3.9724555015563965, + "learning_rate": 9.9762369976237e-06, + "loss": 0.0651, + "step": 23774 + }, + { + "epoch": 11.05299860529986, + "grad_norm": 4.1832146644592285, + "learning_rate": 9.9762359976236e-06, + "loss": 0.0586, + "step": 23775 + }, + { + "epoch": 11.05346350534635, + "grad_norm": 2.5139544010162354, + "learning_rate": 9.9762349976235e-06, + "loss": 0.0167, + "step": 23776 + }, + { + "epoch": 11.05392840539284, + "grad_norm": 6.841562747955322, + "learning_rate": 9.976233997623401e-06, + "loss": 0.1988, + "step": 23777 + }, + { + "epoch": 11.05439330543933, + "grad_norm": 5.371411323547363, + "learning_rate": 9.976232997623302e-06, + "loss": 0.0461, + "step": 23778 + }, + { + "epoch": 11.05485820548582, + "grad_norm": 5.367513179779053, + "learning_rate": 9.9762319976232e-06, + "loss": 0.0601, + "step": 23779 + }, + { + "epoch": 11.055323105532311, + "grad_norm": 4.429006099700928, + "learning_rate": 9.976230997623101e-06, + "loss": 0.1101, + "step": 23780 + }, + { + "epoch": 11.055788005578801, + "grad_norm": 8.061264991760254, + "learning_rate": 9.976229997623e-06, + "loss": 0.1647, + "step": 23781 + }, + { + "epoch": 11.056252905625291, + "grad_norm": 0.5375755429267883, + "learning_rate": 9.9762289976229e-06, + "loss": 0.0047, + "step": 23782 + }, + { + "epoch": 11.05671780567178, + "grad_norm": 5.685748100280762, + "learning_rate": 9.9762279976228e-06, + "loss": 0.0686, + "step": 23783 + }, + { + "epoch": 11.05718270571827, + "grad_norm": 7.161962509155273, + "learning_rate": 9.9762269976227e-06, + "loss": 0.2049, + "step": 23784 + }, + { + "epoch": 11.05764760576476, + "grad_norm": 8.30440902709961, + "learning_rate": 9.976225997622601e-06, + "loss": 0.1208, + "step": 23785 + }, + { + "epoch": 11.05811250581125, + "grad_norm": 3.104114055633545, + "learning_rate": 9.9762249976225e-06, + "loss": 0.0436, + "step": 23786 + }, + { + "epoch": 11.05857740585774, + "grad_norm": 6.96392297744751, + "learning_rate": 9.9762239976224e-06, + "loss": 0.1104, + "step": 23787 + }, + { + "epoch": 11.05904230590423, + "grad_norm": 2.382476568222046, + "learning_rate": 9.9762229976223e-06, + "loss": 0.0178, + "step": 23788 + }, + { + "epoch": 11.059507205950721, + "grad_norm": 1.7889825105667114, + "learning_rate": 9.9762219976222e-06, + "loss": 0.0109, + "step": 23789 + }, + { + "epoch": 11.059972105997211, + "grad_norm": 6.083587646484375, + "learning_rate": 9.976220997622101e-06, + "loss": 0.0748, + "step": 23790 + }, + { + "epoch": 11.060437006043701, + "grad_norm": 1.1625462770462036, + "learning_rate": 9.976219997622002e-06, + "loss": 0.01, + "step": 23791 + }, + { + "epoch": 11.06090190609019, + "grad_norm": 5.624014377593994, + "learning_rate": 9.9762189976219e-06, + "loss": 0.0639, + "step": 23792 + }, + { + "epoch": 11.06136680613668, + "grad_norm": 2.5262484550476074, + "learning_rate": 9.976217997621801e-06, + "loss": 0.0263, + "step": 23793 + }, + { + "epoch": 11.06183170618317, + "grad_norm": 2.78737211227417, + "learning_rate": 9.9762169976217e-06, + "loss": 0.023, + "step": 23794 + }, + { + "epoch": 11.06229660622966, + "grad_norm": 11.638162612915039, + "learning_rate": 9.9762159976216e-06, + "loss": 0.1794, + "step": 23795 + }, + { + "epoch": 11.06276150627615, + "grad_norm": 8.323854446411133, + "learning_rate": 9.9762149976215e-06, + "loss": 0.1622, + "step": 23796 + }, + { + "epoch": 11.06322640632264, + "grad_norm": 5.686244010925293, + "learning_rate": 9.9762139976214e-06, + "loss": 0.0975, + "step": 23797 + }, + { + "epoch": 11.063691306369131, + "grad_norm": 3.018249750137329, + "learning_rate": 9.976212997621301e-06, + "loss": 0.0233, + "step": 23798 + }, + { + "epoch": 11.064156206415621, + "grad_norm": 4.817468166351318, + "learning_rate": 9.9762119976212e-06, + "loss": 0.0542, + "step": 23799 + }, + { + "epoch": 11.064621106462111, + "grad_norm": 3.1042823791503906, + "learning_rate": 9.9762109976211e-06, + "loss": 0.0266, + "step": 23800 + }, + { + "epoch": 11.0650860065086, + "grad_norm": 2.7347140312194824, + "learning_rate": 9.976209997621e-06, + "loss": 0.0214, + "step": 23801 + }, + { + "epoch": 11.06555090655509, + "grad_norm": 5.167636394500732, + "learning_rate": 9.9762089976209e-06, + "loss": 0.0714, + "step": 23802 + }, + { + "epoch": 11.06601580660158, + "grad_norm": 2.8174991607666016, + "learning_rate": 9.976207997620801e-06, + "loss": 0.0191, + "step": 23803 + }, + { + "epoch": 11.06648070664807, + "grad_norm": 6.095938205718994, + "learning_rate": 9.976206997620702e-06, + "loss": 0.0374, + "step": 23804 + }, + { + "epoch": 11.06694560669456, + "grad_norm": 3.52712345123291, + "learning_rate": 9.9762059976206e-06, + "loss": 0.0466, + "step": 23805 + }, + { + "epoch": 11.067410506741052, + "grad_norm": 3.2894954681396484, + "learning_rate": 9.976204997620501e-06, + "loss": 0.0496, + "step": 23806 + }, + { + "epoch": 11.067875406787541, + "grad_norm": 2.781508684158325, + "learning_rate": 9.9762039976204e-06, + "loss": 0.0357, + "step": 23807 + }, + { + "epoch": 11.068340306834031, + "grad_norm": 9.912799835205078, + "learning_rate": 9.976202997620301e-06, + "loss": 0.1125, + "step": 23808 + }, + { + "epoch": 11.068805206880521, + "grad_norm": 8.40280818939209, + "learning_rate": 9.9762019976202e-06, + "loss": 0.0582, + "step": 23809 + }, + { + "epoch": 11.06927010692701, + "grad_norm": 2.750089406967163, + "learning_rate": 9.9762009976201e-06, + "loss": 0.0403, + "step": 23810 + }, + { + "epoch": 11.0697350069735, + "grad_norm": 4.520620822906494, + "learning_rate": 9.976199997620001e-06, + "loss": 0.0704, + "step": 23811 + }, + { + "epoch": 11.07019990701999, + "grad_norm": 6.07767391204834, + "learning_rate": 9.9761989976199e-06, + "loss": 0.1512, + "step": 23812 + }, + { + "epoch": 11.07066480706648, + "grad_norm": 3.697477102279663, + "learning_rate": 9.9761979976198e-06, + "loss": 0.0553, + "step": 23813 + }, + { + "epoch": 11.07112970711297, + "grad_norm": 2.797329902648926, + "learning_rate": 9.9761969976197e-06, + "loss": 0.0246, + "step": 23814 + }, + { + "epoch": 11.071594607159462, + "grad_norm": 3.359267234802246, + "learning_rate": 9.9761959976196e-06, + "loss": 0.0376, + "step": 23815 + }, + { + "epoch": 11.072059507205951, + "grad_norm": 2.063469648361206, + "learning_rate": 9.976194997619501e-06, + "loss": 0.018, + "step": 23816 + }, + { + "epoch": 11.072524407252441, + "grad_norm": 2.6591906547546387, + "learning_rate": 9.976193997619402e-06, + "loss": 0.0257, + "step": 23817 + }, + { + "epoch": 11.072989307298931, + "grad_norm": 4.349453926086426, + "learning_rate": 9.9761929976193e-06, + "loss": 0.0412, + "step": 23818 + }, + { + "epoch": 11.07345420734542, + "grad_norm": 1.270755648612976, + "learning_rate": 9.9761919976192e-06, + "loss": 0.0149, + "step": 23819 + }, + { + "epoch": 11.07391910739191, + "grad_norm": 2.1462764739990234, + "learning_rate": 9.9761909976191e-06, + "loss": 0.017, + "step": 23820 + }, + { + "epoch": 11.0743840074384, + "grad_norm": 3.658836603164673, + "learning_rate": 9.976189997619001e-06, + "loss": 0.0386, + "step": 23821 + }, + { + "epoch": 11.07484890748489, + "grad_norm": 3.4439938068389893, + "learning_rate": 9.9761889976189e-06, + "loss": 0.0252, + "step": 23822 + }, + { + "epoch": 11.07531380753138, + "grad_norm": 3.905489683151245, + "learning_rate": 9.9761879976188e-06, + "loss": 0.0635, + "step": 23823 + }, + { + "epoch": 11.075778707577872, + "grad_norm": 1.2396752834320068, + "learning_rate": 9.976186997618701e-06, + "loss": 0.008, + "step": 23824 + }, + { + "epoch": 11.076243607624361, + "grad_norm": 4.749523639678955, + "learning_rate": 9.9761859976186e-06, + "loss": 0.0495, + "step": 23825 + }, + { + "epoch": 11.076708507670851, + "grad_norm": 6.110729217529297, + "learning_rate": 9.976184997618501e-06, + "loss": 0.042, + "step": 23826 + }, + { + "epoch": 11.077173407717341, + "grad_norm": 7.748171329498291, + "learning_rate": 9.9761839976184e-06, + "loss": 0.0945, + "step": 23827 + }, + { + "epoch": 11.07763830776383, + "grad_norm": 6.0517578125, + "learning_rate": 9.9761829976183e-06, + "loss": 0.0727, + "step": 23828 + }, + { + "epoch": 11.07810320781032, + "grad_norm": 7.427292346954346, + "learning_rate": 9.976181997618201e-06, + "loss": 0.0181, + "step": 23829 + }, + { + "epoch": 11.07856810785681, + "grad_norm": 4.006938934326172, + "learning_rate": 9.9761809976181e-06, + "loss": 0.0451, + "step": 23830 + }, + { + "epoch": 11.0790330079033, + "grad_norm": 5.177946090698242, + "learning_rate": 9.976179997618e-06, + "loss": 0.0841, + "step": 23831 + }, + { + "epoch": 11.07949790794979, + "grad_norm": 5.7106122970581055, + "learning_rate": 9.9761789976179e-06, + "loss": 0.1248, + "step": 23832 + }, + { + "epoch": 11.079962807996282, + "grad_norm": 1.69294273853302, + "learning_rate": 9.9761779976178e-06, + "loss": 0.0117, + "step": 23833 + }, + { + "epoch": 11.080427708042771, + "grad_norm": 3.68680477142334, + "learning_rate": 9.976176997617701e-06, + "loss": 0.0661, + "step": 23834 + }, + { + "epoch": 11.080892608089261, + "grad_norm": 11.888402938842773, + "learning_rate": 9.976175997617602e-06, + "loss": 0.1864, + "step": 23835 + }, + { + "epoch": 11.081357508135751, + "grad_norm": 1.8629062175750732, + "learning_rate": 9.9761749976175e-06, + "loss": 0.0273, + "step": 23836 + }, + { + "epoch": 11.08182240818224, + "grad_norm": 2.766143321990967, + "learning_rate": 9.976173997617401e-06, + "loss": 0.0238, + "step": 23837 + }, + { + "epoch": 11.08228730822873, + "grad_norm": 1.969123125076294, + "learning_rate": 9.9761729976173e-06, + "loss": 0.0204, + "step": 23838 + }, + { + "epoch": 11.08275220827522, + "grad_norm": 3.991938829421997, + "learning_rate": 9.976171997617201e-06, + "loss": 0.0553, + "step": 23839 + }, + { + "epoch": 11.08321710832171, + "grad_norm": 2.1521666049957275, + "learning_rate": 9.9761709976171e-06, + "loss": 0.0189, + "step": 23840 + }, + { + "epoch": 11.0836820083682, + "grad_norm": 8.491938591003418, + "learning_rate": 9.976169997617e-06, + "loss": 0.2012, + "step": 23841 + }, + { + "epoch": 11.084146908414692, + "grad_norm": 7.738186836242676, + "learning_rate": 9.976168997616901e-06, + "loss": 0.0736, + "step": 23842 + }, + { + "epoch": 11.084611808461181, + "grad_norm": 3.3475983142852783, + "learning_rate": 9.9761679976168e-06, + "loss": 0.0405, + "step": 23843 + }, + { + "epoch": 11.085076708507671, + "grad_norm": 3.9409689903259277, + "learning_rate": 9.976166997616701e-06, + "loss": 0.0453, + "step": 23844 + }, + { + "epoch": 11.085541608554161, + "grad_norm": 9.25457763671875, + "learning_rate": 9.9761659976166e-06, + "loss": 0.1033, + "step": 23845 + }, + { + "epoch": 11.08600650860065, + "grad_norm": 10.003228187561035, + "learning_rate": 9.9761649976165e-06, + "loss": 0.0463, + "step": 23846 + }, + { + "epoch": 11.08647140864714, + "grad_norm": 1.724173665046692, + "learning_rate": 9.976163997616401e-06, + "loss": 0.0147, + "step": 23847 + }, + { + "epoch": 11.08693630869363, + "grad_norm": 6.370265007019043, + "learning_rate": 9.976162997616302e-06, + "loss": 0.1534, + "step": 23848 + }, + { + "epoch": 11.08740120874012, + "grad_norm": 5.191715717315674, + "learning_rate": 9.9761619976162e-06, + "loss": 0.0909, + "step": 23849 + }, + { + "epoch": 11.087866108786612, + "grad_norm": 3.2018980979919434, + "learning_rate": 9.976160997616101e-06, + "loss": 0.0273, + "step": 23850 + }, + { + "epoch": 11.088331008833102, + "grad_norm": 4.637686729431152, + "learning_rate": 9.976159997616e-06, + "loss": 0.0445, + "step": 23851 + }, + { + "epoch": 11.088795908879591, + "grad_norm": 4.195445537567139, + "learning_rate": 9.9761589976159e-06, + "loss": 0.049, + "step": 23852 + }, + { + "epoch": 11.089260808926081, + "grad_norm": 8.081093788146973, + "learning_rate": 9.9761579976158e-06, + "loss": 0.1134, + "step": 23853 + }, + { + "epoch": 11.089725708972571, + "grad_norm": 2.1065971851348877, + "learning_rate": 9.9761569976157e-06, + "loss": 0.0133, + "step": 23854 + }, + { + "epoch": 11.09019060901906, + "grad_norm": 5.781065464019775, + "learning_rate": 9.976155997615601e-06, + "loss": 0.0819, + "step": 23855 + }, + { + "epoch": 11.09065550906555, + "grad_norm": 2.7845587730407715, + "learning_rate": 9.9761549976155e-06, + "loss": 0.0418, + "step": 23856 + }, + { + "epoch": 11.09112040911204, + "grad_norm": 2.1204276084899902, + "learning_rate": 9.976153997615401e-06, + "loss": 0.0171, + "step": 23857 + }, + { + "epoch": 11.09158530915853, + "grad_norm": 14.585049629211426, + "learning_rate": 9.9761529976153e-06, + "loss": 0.2524, + "step": 23858 + }, + { + "epoch": 11.092050209205022, + "grad_norm": 8.614191055297852, + "learning_rate": 9.9761519976152e-06, + "loss": 0.1093, + "step": 23859 + }, + { + "epoch": 11.092515109251512, + "grad_norm": 6.324626922607422, + "learning_rate": 9.976150997615101e-06, + "loss": 0.093, + "step": 23860 + }, + { + "epoch": 11.092980009298001, + "grad_norm": 5.419686317443848, + "learning_rate": 9.976149997615e-06, + "loss": 0.0288, + "step": 23861 + }, + { + "epoch": 11.093444909344491, + "grad_norm": 5.254974365234375, + "learning_rate": 9.976148997614901e-06, + "loss": 0.0916, + "step": 23862 + }, + { + "epoch": 11.093909809390981, + "grad_norm": 1.9745508432388306, + "learning_rate": 9.9761479976148e-06, + "loss": 0.0188, + "step": 23863 + }, + { + "epoch": 11.09437470943747, + "grad_norm": 5.9833598136901855, + "learning_rate": 9.9761469976147e-06, + "loss": 0.0545, + "step": 23864 + }, + { + "epoch": 11.09483960948396, + "grad_norm": 1.5456912517547607, + "learning_rate": 9.9761459976146e-06, + "loss": 0.0074, + "step": 23865 + }, + { + "epoch": 11.09530450953045, + "grad_norm": 2.0148894786834717, + "learning_rate": 9.9761449976145e-06, + "loss": 0.0176, + "step": 23866 + }, + { + "epoch": 11.09576940957694, + "grad_norm": 5.463177680969238, + "learning_rate": 9.9761439976144e-06, + "loss": 0.0308, + "step": 23867 + }, + { + "epoch": 11.096234309623432, + "grad_norm": 2.3951029777526855, + "learning_rate": 9.976142997614301e-06, + "loss": 0.0231, + "step": 23868 + }, + { + "epoch": 11.096699209669922, + "grad_norm": 3.161870002746582, + "learning_rate": 9.9761419976142e-06, + "loss": 0.0452, + "step": 23869 + }, + { + "epoch": 11.097164109716411, + "grad_norm": 6.273830413818359, + "learning_rate": 9.976140997614101e-06, + "loss": 0.0851, + "step": 23870 + }, + { + "epoch": 11.097629009762901, + "grad_norm": 7.642747402191162, + "learning_rate": 9.976139997614e-06, + "loss": 0.0393, + "step": 23871 + }, + { + "epoch": 11.098093909809391, + "grad_norm": 5.1948561668396, + "learning_rate": 9.9761389976139e-06, + "loss": 0.0702, + "step": 23872 + }, + { + "epoch": 11.09855880985588, + "grad_norm": 5.518936634063721, + "learning_rate": 9.976137997613801e-06, + "loss": 0.0592, + "step": 23873 + }, + { + "epoch": 11.09902370990237, + "grad_norm": 5.41964054107666, + "learning_rate": 9.9761369976137e-06, + "loss": 0.0676, + "step": 23874 + }, + { + "epoch": 11.09948860994886, + "grad_norm": 2.7132809162139893, + "learning_rate": 9.976135997613601e-06, + "loss": 0.0225, + "step": 23875 + }, + { + "epoch": 11.09995350999535, + "grad_norm": 6.242812633514404, + "learning_rate": 9.9761349976135e-06, + "loss": 0.0911, + "step": 23876 + }, + { + "epoch": 11.100418410041842, + "grad_norm": 3.259521007537842, + "learning_rate": 9.9761339976134e-06, + "loss": 0.0415, + "step": 23877 + }, + { + "epoch": 11.100883310088332, + "grad_norm": 8.048534393310547, + "learning_rate": 9.9761329976133e-06, + "loss": 0.1588, + "step": 23878 + }, + { + "epoch": 11.101348210134821, + "grad_norm": 3.6518805027008057, + "learning_rate": 9.9761319976132e-06, + "loss": 0.0512, + "step": 23879 + }, + { + "epoch": 11.101813110181311, + "grad_norm": 4.091001510620117, + "learning_rate": 9.9761309976131e-06, + "loss": 0.0443, + "step": 23880 + }, + { + "epoch": 11.102278010227801, + "grad_norm": 5.467336177825928, + "learning_rate": 9.976129997613002e-06, + "loss": 0.0633, + "step": 23881 + }, + { + "epoch": 11.10274291027429, + "grad_norm": 4.745391845703125, + "learning_rate": 9.9761289976129e-06, + "loss": 0.0389, + "step": 23882 + }, + { + "epoch": 11.10320781032078, + "grad_norm": 3.926907539367676, + "learning_rate": 9.9761279976128e-06, + "loss": 0.0444, + "step": 23883 + }, + { + "epoch": 11.10367271036727, + "grad_norm": 3.100032091140747, + "learning_rate": 9.9761269976127e-06, + "loss": 0.0183, + "step": 23884 + }, + { + "epoch": 11.104137610413762, + "grad_norm": 5.997738838195801, + "learning_rate": 9.9761259976126e-06, + "loss": 0.1417, + "step": 23885 + }, + { + "epoch": 11.104602510460252, + "grad_norm": 7.0621819496154785, + "learning_rate": 9.976124997612501e-06, + "loss": 0.1008, + "step": 23886 + }, + { + "epoch": 11.105067410506742, + "grad_norm": 5.482235431671143, + "learning_rate": 9.9761239976124e-06, + "loss": 0.0881, + "step": 23887 + }, + { + "epoch": 11.105532310553231, + "grad_norm": 2.8384814262390137, + "learning_rate": 9.976122997612301e-06, + "loss": 0.0132, + "step": 23888 + }, + { + "epoch": 11.105997210599721, + "grad_norm": 2.4171688556671143, + "learning_rate": 9.9761219976122e-06, + "loss": 0.0206, + "step": 23889 + }, + { + "epoch": 11.106462110646211, + "grad_norm": 1.8208510875701904, + "learning_rate": 9.9761209976121e-06, + "loss": 0.012, + "step": 23890 + }, + { + "epoch": 11.1069270106927, + "grad_norm": 1.6149871349334717, + "learning_rate": 9.976119997612e-06, + "loss": 0.0117, + "step": 23891 + }, + { + "epoch": 11.10739191073919, + "grad_norm": 2.807873010635376, + "learning_rate": 9.9761189976119e-06, + "loss": 0.0151, + "step": 23892 + }, + { + "epoch": 11.10785681078568, + "grad_norm": 0.8679021596908569, + "learning_rate": 9.976117997611801e-06, + "loss": 0.0102, + "step": 23893 + }, + { + "epoch": 11.108321710832172, + "grad_norm": 4.5622663497924805, + "learning_rate": 9.9761169976117e-06, + "loss": 0.0468, + "step": 23894 + }, + { + "epoch": 11.108786610878662, + "grad_norm": 1.276811957359314, + "learning_rate": 9.9761159976116e-06, + "loss": 0.0109, + "step": 23895 + }, + { + "epoch": 11.109251510925152, + "grad_norm": 4.391310214996338, + "learning_rate": 9.9761149976115e-06, + "loss": 0.0812, + "step": 23896 + }, + { + "epoch": 11.109716410971641, + "grad_norm": 1.8774745464324951, + "learning_rate": 9.9761139976114e-06, + "loss": 0.0169, + "step": 23897 + }, + { + "epoch": 11.110181311018131, + "grad_norm": 3.6744272708892822, + "learning_rate": 9.9761129976113e-06, + "loss": 0.036, + "step": 23898 + }, + { + "epoch": 11.110646211064621, + "grad_norm": 5.781975269317627, + "learning_rate": 9.976111997611201e-06, + "loss": 0.0693, + "step": 23899 + }, + { + "epoch": 11.11111111111111, + "grad_norm": 2.9416511058807373, + "learning_rate": 9.9761109976111e-06, + "loss": 0.0248, + "step": 23900 + }, + { + "epoch": 11.1115760111576, + "grad_norm": 0.7926613092422485, + "learning_rate": 9.976109997611001e-06, + "loss": 0.0072, + "step": 23901 + }, + { + "epoch": 11.11204091120409, + "grad_norm": 2.376007080078125, + "learning_rate": 9.9761089976109e-06, + "loss": 0.0118, + "step": 23902 + }, + { + "epoch": 11.112505811250582, + "grad_norm": 3.721083164215088, + "learning_rate": 9.9761079976108e-06, + "loss": 0.037, + "step": 23903 + }, + { + "epoch": 11.112970711297072, + "grad_norm": 5.458309650421143, + "learning_rate": 9.976106997610701e-06, + "loss": 0.0925, + "step": 23904 + }, + { + "epoch": 11.113435611343562, + "grad_norm": 2.9936447143554688, + "learning_rate": 9.9761059976106e-06, + "loss": 0.0361, + "step": 23905 + }, + { + "epoch": 11.113900511390051, + "grad_norm": 1.2984484434127808, + "learning_rate": 9.976104997610501e-06, + "loss": 0.0122, + "step": 23906 + }, + { + "epoch": 11.114365411436541, + "grad_norm": 3.364105701446533, + "learning_rate": 9.9761039976104e-06, + "loss": 0.016, + "step": 23907 + }, + { + "epoch": 11.114830311483031, + "grad_norm": 9.771282196044922, + "learning_rate": 9.9761029976103e-06, + "loss": 0.2293, + "step": 23908 + }, + { + "epoch": 11.11529521152952, + "grad_norm": 3.5826382637023926, + "learning_rate": 9.9761019976102e-06, + "loss": 0.0323, + "step": 23909 + }, + { + "epoch": 11.11576011157601, + "grad_norm": 5.124014377593994, + "learning_rate": 9.9761009976101e-06, + "loss": 0.0634, + "step": 23910 + }, + { + "epoch": 11.1162250116225, + "grad_norm": 2.1759042739868164, + "learning_rate": 9.976099997610001e-06, + "loss": 0.0208, + "step": 23911 + }, + { + "epoch": 11.116689911668992, + "grad_norm": 2.839071273803711, + "learning_rate": 9.976098997609902e-06, + "loss": 0.0275, + "step": 23912 + }, + { + "epoch": 11.117154811715482, + "grad_norm": 5.951653003692627, + "learning_rate": 9.9760979976098e-06, + "loss": 0.129, + "step": 23913 + }, + { + "epoch": 11.117619711761972, + "grad_norm": 4.921796798706055, + "learning_rate": 9.976096997609701e-06, + "loss": 0.0938, + "step": 23914 + }, + { + "epoch": 11.118084611808461, + "grad_norm": 2.5655438899993896, + "learning_rate": 9.9760959976096e-06, + "loss": 0.02, + "step": 23915 + }, + { + "epoch": 11.118549511854951, + "grad_norm": 7.834298610687256, + "learning_rate": 9.9760949976095e-06, + "loss": 0.1007, + "step": 23916 + }, + { + "epoch": 11.119014411901441, + "grad_norm": 7.211017608642578, + "learning_rate": 9.976093997609401e-06, + "loss": 0.1261, + "step": 23917 + }, + { + "epoch": 11.11947931194793, + "grad_norm": 7.283877372741699, + "learning_rate": 9.9760929976093e-06, + "loss": 0.1997, + "step": 23918 + }, + { + "epoch": 11.11994421199442, + "grad_norm": 1.0720723867416382, + "learning_rate": 9.976091997609201e-06, + "loss": 0.0084, + "step": 23919 + }, + { + "epoch": 11.12040911204091, + "grad_norm": 4.915334224700928, + "learning_rate": 9.9760909976091e-06, + "loss": 0.0684, + "step": 23920 + }, + { + "epoch": 11.120874012087402, + "grad_norm": 4.774285793304443, + "learning_rate": 9.976089997609e-06, + "loss": 0.0639, + "step": 23921 + }, + { + "epoch": 11.121338912133892, + "grad_norm": 7.596084117889404, + "learning_rate": 9.9760889976089e-06, + "loss": 0.1006, + "step": 23922 + }, + { + "epoch": 11.121803812180381, + "grad_norm": 7.302844047546387, + "learning_rate": 9.9760879976088e-06, + "loss": 0.1062, + "step": 23923 + }, + { + "epoch": 11.122268712226871, + "grad_norm": 9.131895065307617, + "learning_rate": 9.976086997608701e-06, + "loss": 0.13, + "step": 23924 + }, + { + "epoch": 11.122733612273361, + "grad_norm": 5.645965099334717, + "learning_rate": 9.976085997608602e-06, + "loss": 0.0589, + "step": 23925 + }, + { + "epoch": 11.123198512319851, + "grad_norm": 4.136622428894043, + "learning_rate": 9.9760849976085e-06, + "loss": 0.069, + "step": 23926 + }, + { + "epoch": 11.12366341236634, + "grad_norm": 3.8937957286834717, + "learning_rate": 9.9760839976084e-06, + "loss": 0.0293, + "step": 23927 + }, + { + "epoch": 11.12412831241283, + "grad_norm": 2.6502339839935303, + "learning_rate": 9.9760829976083e-06, + "loss": 0.0231, + "step": 23928 + }, + { + "epoch": 11.124593212459322, + "grad_norm": 3.2944867610931396, + "learning_rate": 9.976081997608201e-06, + "loss": 0.0285, + "step": 23929 + }, + { + "epoch": 11.125058112505812, + "grad_norm": 4.220605850219727, + "learning_rate": 9.976080997608102e-06, + "loss": 0.0755, + "step": 23930 + }, + { + "epoch": 11.125523012552302, + "grad_norm": 5.401728630065918, + "learning_rate": 9.976079997608e-06, + "loss": 0.059, + "step": 23931 + }, + { + "epoch": 11.125987912598791, + "grad_norm": 4.952655792236328, + "learning_rate": 9.976078997607901e-06, + "loss": 0.0283, + "step": 23932 + }, + { + "epoch": 11.126452812645281, + "grad_norm": 5.313394069671631, + "learning_rate": 9.9760779976078e-06, + "loss": 0.0785, + "step": 23933 + }, + { + "epoch": 11.126917712691771, + "grad_norm": 5.7002105712890625, + "learning_rate": 9.9760769976077e-06, + "loss": 0.1074, + "step": 23934 + }, + { + "epoch": 11.127382612738261, + "grad_norm": 2.6365373134613037, + "learning_rate": 9.9760759976076e-06, + "loss": 0.0201, + "step": 23935 + }, + { + "epoch": 11.12784751278475, + "grad_norm": 3.8344852924346924, + "learning_rate": 9.9760749976075e-06, + "loss": 0.0463, + "step": 23936 + }, + { + "epoch": 11.12831241283124, + "grad_norm": 4.907172203063965, + "learning_rate": 9.976073997607401e-06, + "loss": 0.0672, + "step": 23937 + }, + { + "epoch": 11.128777312877732, + "grad_norm": 4.283483505249023, + "learning_rate": 9.9760729976073e-06, + "loss": 0.0349, + "step": 23938 + }, + { + "epoch": 11.129242212924222, + "grad_norm": 6.355648040771484, + "learning_rate": 9.9760719976072e-06, + "loss": 0.0763, + "step": 23939 + }, + { + "epoch": 11.129707112970712, + "grad_norm": 2.029513359069824, + "learning_rate": 9.9760709976071e-06, + "loss": 0.015, + "step": 23940 + }, + { + "epoch": 11.130172013017201, + "grad_norm": 4.895135879516602, + "learning_rate": 9.976069997607e-06, + "loss": 0.0981, + "step": 23941 + }, + { + "epoch": 11.130636913063691, + "grad_norm": 8.893449783325195, + "learning_rate": 9.976068997606901e-06, + "loss": 0.1635, + "step": 23942 + }, + { + "epoch": 11.131101813110181, + "grad_norm": 5.383766174316406, + "learning_rate": 9.976067997606802e-06, + "loss": 0.0587, + "step": 23943 + }, + { + "epoch": 11.131566713156671, + "grad_norm": 5.167491436004639, + "learning_rate": 9.9760669976067e-06, + "loss": 0.0401, + "step": 23944 + }, + { + "epoch": 11.13203161320316, + "grad_norm": 4.201665878295898, + "learning_rate": 9.976065997606601e-06, + "loss": 0.0786, + "step": 23945 + }, + { + "epoch": 11.13249651324965, + "grad_norm": 3.8198492527008057, + "learning_rate": 9.9760649976065e-06, + "loss": 0.0449, + "step": 23946 + }, + { + "epoch": 11.132961413296142, + "grad_norm": 2.123469352722168, + "learning_rate": 9.976063997606401e-06, + "loss": 0.017, + "step": 23947 + }, + { + "epoch": 11.133426313342632, + "grad_norm": 3.776547908782959, + "learning_rate": 9.9760629976063e-06, + "loss": 0.053, + "step": 23948 + }, + { + "epoch": 11.133891213389122, + "grad_norm": 8.823944091796875, + "learning_rate": 9.9760619976062e-06, + "loss": 0.142, + "step": 23949 + }, + { + "epoch": 11.134356113435611, + "grad_norm": 3.170288324356079, + "learning_rate": 9.976060997606101e-06, + "loss": 0.0176, + "step": 23950 + }, + { + "epoch": 11.134821013482101, + "grad_norm": 1.0400620698928833, + "learning_rate": 9.976059997606e-06, + "loss": 0.007, + "step": 23951 + }, + { + "epoch": 11.135285913528591, + "grad_norm": 2.8194854259490967, + "learning_rate": 9.9760589976059e-06, + "loss": 0.0295, + "step": 23952 + }, + { + "epoch": 11.135750813575081, + "grad_norm": 8.745111465454102, + "learning_rate": 9.9760579976058e-06, + "loss": 0.0517, + "step": 23953 + }, + { + "epoch": 11.13621571362157, + "grad_norm": 3.2740285396575928, + "learning_rate": 9.9760569976057e-06, + "loss": 0.0347, + "step": 23954 + }, + { + "epoch": 11.13668061366806, + "grad_norm": 2.463623046875, + "learning_rate": 9.976055997605601e-06, + "loss": 0.0248, + "step": 23955 + }, + { + "epoch": 11.137145513714552, + "grad_norm": 0.9738011956214905, + "learning_rate": 9.976054997605502e-06, + "loss": 0.009, + "step": 23956 + }, + { + "epoch": 11.137610413761042, + "grad_norm": 7.389590263366699, + "learning_rate": 9.9760539976054e-06, + "loss": 0.0926, + "step": 23957 + }, + { + "epoch": 11.138075313807532, + "grad_norm": 2.6917366981506348, + "learning_rate": 9.976052997605301e-06, + "loss": 0.0272, + "step": 23958 + }, + { + "epoch": 11.138540213854021, + "grad_norm": 2.4370779991149902, + "learning_rate": 9.9760519976052e-06, + "loss": 0.0322, + "step": 23959 + }, + { + "epoch": 11.139005113900511, + "grad_norm": 6.8740386962890625, + "learning_rate": 9.9760509976051e-06, + "loss": 0.1266, + "step": 23960 + }, + { + "epoch": 11.139470013947001, + "grad_norm": 3.470750570297241, + "learning_rate": 9.976049997605e-06, + "loss": 0.0585, + "step": 23961 + }, + { + "epoch": 11.139934913993491, + "grad_norm": 2.6602280139923096, + "learning_rate": 9.9760489976049e-06, + "loss": 0.0332, + "step": 23962 + }, + { + "epoch": 11.14039981403998, + "grad_norm": 1.9147720336914062, + "learning_rate": 9.976047997604801e-06, + "loss": 0.0124, + "step": 23963 + }, + { + "epoch": 11.140864714086472, + "grad_norm": 5.872253894805908, + "learning_rate": 9.9760469976047e-06, + "loss": 0.0551, + "step": 23964 + }, + { + "epoch": 11.141329614132962, + "grad_norm": 4.679347038269043, + "learning_rate": 9.976045997604601e-06, + "loss": 0.0742, + "step": 23965 + }, + { + "epoch": 11.141794514179452, + "grad_norm": 2.0522994995117188, + "learning_rate": 9.9760449976045e-06, + "loss": 0.0198, + "step": 23966 + }, + { + "epoch": 11.142259414225942, + "grad_norm": 4.558143138885498, + "learning_rate": 9.9760439976044e-06, + "loss": 0.0528, + "step": 23967 + }, + { + "epoch": 11.142724314272431, + "grad_norm": 5.734438896179199, + "learning_rate": 9.976042997604301e-06, + "loss": 0.0722, + "step": 23968 + }, + { + "epoch": 11.143189214318921, + "grad_norm": 3.393939733505249, + "learning_rate": 9.976041997604202e-06, + "loss": 0.0177, + "step": 23969 + }, + { + "epoch": 11.143654114365411, + "grad_norm": 5.087950706481934, + "learning_rate": 9.9760409976041e-06, + "loss": 0.0785, + "step": 23970 + }, + { + "epoch": 11.144119014411901, + "grad_norm": 6.808443546295166, + "learning_rate": 9.976039997604e-06, + "loss": 0.057, + "step": 23971 + }, + { + "epoch": 11.14458391445839, + "grad_norm": 4.630056381225586, + "learning_rate": 9.9760389976039e-06, + "loss": 0.0254, + "step": 23972 + }, + { + "epoch": 11.145048814504882, + "grad_norm": 3.4732604026794434, + "learning_rate": 9.976037997603801e-06, + "loss": 0.0296, + "step": 23973 + }, + { + "epoch": 11.145513714551372, + "grad_norm": 6.013140678405762, + "learning_rate": 9.9760369976037e-06, + "loss": 0.0478, + "step": 23974 + }, + { + "epoch": 11.145978614597862, + "grad_norm": 6.965941429138184, + "learning_rate": 9.9760359976036e-06, + "loss": 0.1005, + "step": 23975 + }, + { + "epoch": 11.146443514644352, + "grad_norm": 0.5966562628746033, + "learning_rate": 9.976034997603501e-06, + "loss": 0.0039, + "step": 23976 + }, + { + "epoch": 11.146908414690841, + "grad_norm": 3.9451520442962646, + "learning_rate": 9.9760339976034e-06, + "loss": 0.0794, + "step": 23977 + }, + { + "epoch": 11.147373314737331, + "grad_norm": 4.369496822357178, + "learning_rate": 9.976032997603301e-06, + "loss": 0.0608, + "step": 23978 + }, + { + "epoch": 11.147838214783821, + "grad_norm": 2.4322099685668945, + "learning_rate": 9.9760319976032e-06, + "loss": 0.0885, + "step": 23979 + }, + { + "epoch": 11.148303114830311, + "grad_norm": 6.23990535736084, + "learning_rate": 9.9760309976031e-06, + "loss": 0.079, + "step": 23980 + }, + { + "epoch": 11.1487680148768, + "grad_norm": 4.329721927642822, + "learning_rate": 9.976029997603001e-06, + "loss": 0.0767, + "step": 23981 + }, + { + "epoch": 11.149232914923292, + "grad_norm": 3.761274814605713, + "learning_rate": 9.9760289976029e-06, + "loss": 0.0227, + "step": 23982 + }, + { + "epoch": 11.149697814969782, + "grad_norm": 2.226552963256836, + "learning_rate": 9.9760279976028e-06, + "loss": 0.0178, + "step": 23983 + }, + { + "epoch": 11.150162715016272, + "grad_norm": 2.9291977882385254, + "learning_rate": 9.9760269976027e-06, + "loss": 0.0258, + "step": 23984 + }, + { + "epoch": 11.150627615062762, + "grad_norm": 6.9393229484558105, + "learning_rate": 9.9760259976026e-06, + "loss": 0.0788, + "step": 23985 + }, + { + "epoch": 11.151092515109251, + "grad_norm": 0.8548402786254883, + "learning_rate": 9.976024997602501e-06, + "loss": 0.0082, + "step": 23986 + }, + { + "epoch": 11.151557415155741, + "grad_norm": 1.9427701234817505, + "learning_rate": 9.9760239976024e-06, + "loss": 0.0283, + "step": 23987 + }, + { + "epoch": 11.152022315202231, + "grad_norm": 7.559001445770264, + "learning_rate": 9.9760229976023e-06, + "loss": 0.1108, + "step": 23988 + }, + { + "epoch": 11.15248721524872, + "grad_norm": 4.692410945892334, + "learning_rate": 9.976021997602201e-06, + "loss": 0.065, + "step": 23989 + }, + { + "epoch": 11.15295211529521, + "grad_norm": 1.5667692422866821, + "learning_rate": 9.9760209976021e-06, + "loss": 0.0076, + "step": 23990 + }, + { + "epoch": 11.153417015341702, + "grad_norm": 2.285874366760254, + "learning_rate": 9.976019997602001e-06, + "loss": 0.0225, + "step": 23991 + }, + { + "epoch": 11.153881915388192, + "grad_norm": 5.210140705108643, + "learning_rate": 9.9760189976019e-06, + "loss": 0.084, + "step": 23992 + }, + { + "epoch": 11.154346815434682, + "grad_norm": 4.057683944702148, + "learning_rate": 9.9760179976018e-06, + "loss": 0.0463, + "step": 23993 + }, + { + "epoch": 11.154811715481172, + "grad_norm": 4.89300537109375, + "learning_rate": 9.976016997601701e-06, + "loss": 0.0502, + "step": 23994 + }, + { + "epoch": 11.155276615527661, + "grad_norm": 1.9533385038375854, + "learning_rate": 9.9760159976016e-06, + "loss": 0.0279, + "step": 23995 + }, + { + "epoch": 11.155741515574151, + "grad_norm": 4.7919921875, + "learning_rate": 9.976014997601501e-06, + "loss": 0.0392, + "step": 23996 + }, + { + "epoch": 11.156206415620641, + "grad_norm": 6.858912467956543, + "learning_rate": 9.9760139976014e-06, + "loss": 0.1045, + "step": 23997 + }, + { + "epoch": 11.15667131566713, + "grad_norm": 2.92756986618042, + "learning_rate": 9.9760129976013e-06, + "loss": 0.0289, + "step": 23998 + }, + { + "epoch": 11.15713621571362, + "grad_norm": 3.252042531967163, + "learning_rate": 9.976011997601201e-06, + "loss": 0.0358, + "step": 23999 + }, + { + "epoch": 11.157601115760112, + "grad_norm": 1.9673963785171509, + "learning_rate": 9.976010997601102e-06, + "loss": 0.0135, + "step": 24000 + }, + { + "epoch": 11.158066015806602, + "grad_norm": 3.9653284549713135, + "learning_rate": 9.976009997601e-06, + "loss": 0.0712, + "step": 24001 + }, + { + "epoch": 11.158530915853092, + "grad_norm": 3.2725813388824463, + "learning_rate": 9.9760089976009e-06, + "loss": 0.0409, + "step": 24002 + }, + { + "epoch": 11.158995815899582, + "grad_norm": 3.875805139541626, + "learning_rate": 9.9760079976008e-06, + "loss": 0.0678, + "step": 24003 + }, + { + "epoch": 11.159460715946071, + "grad_norm": 6.059844017028809, + "learning_rate": 9.9760069976007e-06, + "loss": 0.0654, + "step": 24004 + }, + { + "epoch": 11.159925615992561, + "grad_norm": 2.18942928314209, + "learning_rate": 9.9760059976006e-06, + "loss": 0.0179, + "step": 24005 + }, + { + "epoch": 11.160390516039051, + "grad_norm": 3.682887315750122, + "learning_rate": 9.9760049976005e-06, + "loss": 0.0252, + "step": 24006 + }, + { + "epoch": 11.16085541608554, + "grad_norm": 8.112459182739258, + "learning_rate": 9.976003997600401e-06, + "loss": 0.0633, + "step": 24007 + }, + { + "epoch": 11.161320316132032, + "grad_norm": 2.8492331504821777, + "learning_rate": 9.9760029976003e-06, + "loss": 0.0348, + "step": 24008 + }, + { + "epoch": 11.161785216178522, + "grad_norm": 5.021193504333496, + "learning_rate": 9.976001997600201e-06, + "loss": 0.0581, + "step": 24009 + }, + { + "epoch": 11.162250116225012, + "grad_norm": 4.760079383850098, + "learning_rate": 9.9760009976001e-06, + "loss": 0.0236, + "step": 24010 + }, + { + "epoch": 11.162715016271502, + "grad_norm": 4.640927314758301, + "learning_rate": 9.9759999976e-06, + "loss": 0.0789, + "step": 24011 + }, + { + "epoch": 11.163179916317992, + "grad_norm": 5.345895290374756, + "learning_rate": 9.975998997599901e-06, + "loss": 0.0776, + "step": 24012 + }, + { + "epoch": 11.163644816364481, + "grad_norm": 1.0547709465026855, + "learning_rate": 9.9759979975998e-06, + "loss": 0.0076, + "step": 24013 + }, + { + "epoch": 11.164109716410971, + "grad_norm": 2.6150825023651123, + "learning_rate": 9.975996997599701e-06, + "loss": 0.0283, + "step": 24014 + }, + { + "epoch": 11.164574616457461, + "grad_norm": 4.3454084396362305, + "learning_rate": 9.9759959975996e-06, + "loss": 0.0673, + "step": 24015 + }, + { + "epoch": 11.16503951650395, + "grad_norm": 6.222078323364258, + "learning_rate": 9.9759949975995e-06, + "loss": 0.0926, + "step": 24016 + }, + { + "epoch": 11.165504416550442, + "grad_norm": 4.804037094116211, + "learning_rate": 9.9759939975994e-06, + "loss": 0.0654, + "step": 24017 + }, + { + "epoch": 11.165969316596932, + "grad_norm": 1.5132527351379395, + "learning_rate": 9.9759929975993e-06, + "loss": 0.0093, + "step": 24018 + }, + { + "epoch": 11.166434216643422, + "grad_norm": 4.625174522399902, + "learning_rate": 9.9759919975992e-06, + "loss": 0.0313, + "step": 24019 + }, + { + "epoch": 11.166899116689912, + "grad_norm": 6.63824987411499, + "learning_rate": 9.975990997599101e-06, + "loss": 0.0663, + "step": 24020 + }, + { + "epoch": 11.167364016736402, + "grad_norm": 5.138673782348633, + "learning_rate": 9.975989997599e-06, + "loss": 0.0311, + "step": 24021 + }, + { + "epoch": 11.167828916782891, + "grad_norm": 6.735617637634277, + "learning_rate": 9.975988997598901e-06, + "loss": 0.0994, + "step": 24022 + }, + { + "epoch": 11.168293816829381, + "grad_norm": 3.02079176902771, + "learning_rate": 9.9759879975988e-06, + "loss": 0.0272, + "step": 24023 + }, + { + "epoch": 11.168758716875871, + "grad_norm": 3.5593838691711426, + "learning_rate": 9.9759869975987e-06, + "loss": 0.0312, + "step": 24024 + }, + { + "epoch": 11.16922361692236, + "grad_norm": 4.035160064697266, + "learning_rate": 9.975985997598601e-06, + "loss": 0.0945, + "step": 24025 + }, + { + "epoch": 11.169688516968852, + "grad_norm": 7.457798004150391, + "learning_rate": 9.9759849975985e-06, + "loss": 0.1842, + "step": 24026 + }, + { + "epoch": 11.170153417015342, + "grad_norm": 7.324087619781494, + "learning_rate": 9.975983997598401e-06, + "loss": 0.157, + "step": 24027 + }, + { + "epoch": 11.170618317061832, + "grad_norm": 5.6963090896606445, + "learning_rate": 9.9759829975983e-06, + "loss": 0.0748, + "step": 24028 + }, + { + "epoch": 11.171083217108322, + "grad_norm": 10.709351539611816, + "learning_rate": 9.9759819975982e-06, + "loss": 0.1869, + "step": 24029 + }, + { + "epoch": 11.171548117154812, + "grad_norm": 2.0523767471313477, + "learning_rate": 9.9759809975981e-06, + "loss": 0.0183, + "step": 24030 + }, + { + "epoch": 11.172013017201301, + "grad_norm": 7.33468770980835, + "learning_rate": 9.975979997598e-06, + "loss": 0.1575, + "step": 24031 + }, + { + "epoch": 11.172477917247791, + "grad_norm": 0.5389769673347473, + "learning_rate": 9.975978997597901e-06, + "loss": 0.0034, + "step": 24032 + }, + { + "epoch": 11.172942817294281, + "grad_norm": 4.385228633880615, + "learning_rate": 9.975977997597802e-06, + "loss": 0.0482, + "step": 24033 + }, + { + "epoch": 11.17340771734077, + "grad_norm": 4.3979668617248535, + "learning_rate": 9.9759769975977e-06, + "loss": 0.0491, + "step": 24034 + }, + { + "epoch": 11.173872617387262, + "grad_norm": 4.749819755554199, + "learning_rate": 9.9759759975976e-06, + "loss": 0.0432, + "step": 24035 + }, + { + "epoch": 11.174337517433752, + "grad_norm": 7.416220188140869, + "learning_rate": 9.9759749975975e-06, + "loss": 0.0789, + "step": 24036 + }, + { + "epoch": 11.174802417480242, + "grad_norm": 4.7598724365234375, + "learning_rate": 9.9759739975974e-06, + "loss": 0.0942, + "step": 24037 + }, + { + "epoch": 11.175267317526732, + "grad_norm": 6.785737037658691, + "learning_rate": 9.975972997597301e-06, + "loss": 0.1946, + "step": 24038 + }, + { + "epoch": 11.175732217573222, + "grad_norm": 4.400453090667725, + "learning_rate": 9.9759719975972e-06, + "loss": 0.0773, + "step": 24039 + }, + { + "epoch": 11.176197117619711, + "grad_norm": 5.525336265563965, + "learning_rate": 9.975970997597101e-06, + "loss": 0.0624, + "step": 24040 + }, + { + "epoch": 11.176662017666201, + "grad_norm": 6.91353178024292, + "learning_rate": 9.975969997597e-06, + "loss": 0.0913, + "step": 24041 + }, + { + "epoch": 11.177126917712691, + "grad_norm": 8.374959945678711, + "learning_rate": 9.9759689975969e-06, + "loss": 0.1323, + "step": 24042 + }, + { + "epoch": 11.177591817759183, + "grad_norm": 2.94085955619812, + "learning_rate": 9.9759679975968e-06, + "loss": 0.0246, + "step": 24043 + }, + { + "epoch": 11.178056717805672, + "grad_norm": 5.7787861824035645, + "learning_rate": 9.9759669975967e-06, + "loss": 0.0217, + "step": 24044 + }, + { + "epoch": 11.178521617852162, + "grad_norm": 9.091426849365234, + "learning_rate": 9.975965997596601e-06, + "loss": 0.1162, + "step": 24045 + }, + { + "epoch": 11.178986517898652, + "grad_norm": 7.790602207183838, + "learning_rate": 9.9759649975965e-06, + "loss": 0.099, + "step": 24046 + }, + { + "epoch": 11.179451417945142, + "grad_norm": 1.6120058298110962, + "learning_rate": 9.9759639975964e-06, + "loss": 0.0119, + "step": 24047 + }, + { + "epoch": 11.179916317991632, + "grad_norm": 7.81130838394165, + "learning_rate": 9.9759629975963e-06, + "loss": 0.0993, + "step": 24048 + }, + { + "epoch": 11.180381218038121, + "grad_norm": 5.583681106567383, + "learning_rate": 9.9759619975962e-06, + "loss": 0.0847, + "step": 24049 + }, + { + "epoch": 11.180846118084611, + "grad_norm": 0.40632712841033936, + "learning_rate": 9.975960997596101e-06, + "loss": 0.0036, + "step": 24050 + }, + { + "epoch": 11.181311018131101, + "grad_norm": 2.999689817428589, + "learning_rate": 9.975959997596002e-06, + "loss": 0.0363, + "step": 24051 + }, + { + "epoch": 11.181775918177593, + "grad_norm": 4.1243109703063965, + "learning_rate": 9.9759589975959e-06, + "loss": 0.0523, + "step": 24052 + }, + { + "epoch": 11.182240818224082, + "grad_norm": 1.8796803951263428, + "learning_rate": 9.975957997595801e-06, + "loss": 0.0203, + "step": 24053 + }, + { + "epoch": 11.182705718270572, + "grad_norm": 2.6577303409576416, + "learning_rate": 9.9759569975957e-06, + "loss": 0.0202, + "step": 24054 + }, + { + "epoch": 11.183170618317062, + "grad_norm": 4.676249027252197, + "learning_rate": 9.9759559975956e-06, + "loss": 0.0319, + "step": 24055 + }, + { + "epoch": 11.183635518363552, + "grad_norm": 4.64652156829834, + "learning_rate": 9.9759549975955e-06, + "loss": 0.0692, + "step": 24056 + }, + { + "epoch": 11.184100418410042, + "grad_norm": 4.882387638092041, + "learning_rate": 9.9759539975954e-06, + "loss": 0.0525, + "step": 24057 + }, + { + "epoch": 11.184565318456531, + "grad_norm": 2.8562095165252686, + "learning_rate": 9.975952997595301e-06, + "loss": 0.0372, + "step": 24058 + }, + { + "epoch": 11.185030218503021, + "grad_norm": 4.672372817993164, + "learning_rate": 9.9759519975952e-06, + "loss": 0.0419, + "step": 24059 + }, + { + "epoch": 11.185495118549511, + "grad_norm": 4.944542407989502, + "learning_rate": 9.9759509975951e-06, + "loss": 0.056, + "step": 24060 + }, + { + "epoch": 11.185960018596003, + "grad_norm": 2.7861387729644775, + "learning_rate": 9.975949997595e-06, + "loss": 0.0239, + "step": 24061 + }, + { + "epoch": 11.186424918642492, + "grad_norm": 2.457876205444336, + "learning_rate": 9.9759489975949e-06, + "loss": 0.0264, + "step": 24062 + }, + { + "epoch": 11.186889818688982, + "grad_norm": 5.065514087677002, + "learning_rate": 9.975947997594801e-06, + "loss": 0.106, + "step": 24063 + }, + { + "epoch": 11.187354718735472, + "grad_norm": 2.1609623432159424, + "learning_rate": 9.975946997594702e-06, + "loss": 0.0188, + "step": 24064 + }, + { + "epoch": 11.187819618781962, + "grad_norm": 5.667187690734863, + "learning_rate": 9.9759459975946e-06, + "loss": 0.0572, + "step": 24065 + }, + { + "epoch": 11.188284518828452, + "grad_norm": 2.535327196121216, + "learning_rate": 9.975944997594501e-06, + "loss": 0.0245, + "step": 24066 + }, + { + "epoch": 11.188749418874941, + "grad_norm": 8.571826934814453, + "learning_rate": 9.9759439975944e-06, + "loss": 0.083, + "step": 24067 + }, + { + "epoch": 11.189214318921431, + "grad_norm": 1.5079340934753418, + "learning_rate": 9.9759429975943e-06, + "loss": 0.0196, + "step": 24068 + }, + { + "epoch": 11.189679218967921, + "grad_norm": 6.1886186599731445, + "learning_rate": 9.975941997594201e-06, + "loss": 0.102, + "step": 24069 + }, + { + "epoch": 11.190144119014413, + "grad_norm": 2.8033273220062256, + "learning_rate": 9.9759409975941e-06, + "loss": 0.0293, + "step": 24070 + }, + { + "epoch": 11.190609019060902, + "grad_norm": 1.800491452217102, + "learning_rate": 9.975939997594001e-06, + "loss": 0.0137, + "step": 24071 + }, + { + "epoch": 11.191073919107392, + "grad_norm": 5.313792705535889, + "learning_rate": 9.9759389975939e-06, + "loss": 0.1071, + "step": 24072 + }, + { + "epoch": 11.191538819153882, + "grad_norm": 2.3175547122955322, + "learning_rate": 9.9759379975938e-06, + "loss": 0.0265, + "step": 24073 + }, + { + "epoch": 11.192003719200372, + "grad_norm": 3.503047466278076, + "learning_rate": 9.9759369975937e-06, + "loss": 0.0577, + "step": 24074 + }, + { + "epoch": 11.192468619246862, + "grad_norm": 3.2273998260498047, + "learning_rate": 9.9759359975936e-06, + "loss": 0.0476, + "step": 24075 + }, + { + "epoch": 11.192933519293351, + "grad_norm": 2.94830060005188, + "learning_rate": 9.975934997593501e-06, + "loss": 0.0387, + "step": 24076 + }, + { + "epoch": 11.193398419339841, + "grad_norm": 5.038949012756348, + "learning_rate": 9.975933997593402e-06, + "loss": 0.0571, + "step": 24077 + }, + { + "epoch": 11.193863319386331, + "grad_norm": 7.800476551055908, + "learning_rate": 9.9759329975933e-06, + "loss": 0.062, + "step": 24078 + }, + { + "epoch": 11.194328219432823, + "grad_norm": 2.140871286392212, + "learning_rate": 9.9759319975932e-06, + "loss": 0.0197, + "step": 24079 + }, + { + "epoch": 11.194793119479312, + "grad_norm": 2.7339985370635986, + "learning_rate": 9.9759309975931e-06, + "loss": 0.0409, + "step": 24080 + }, + { + "epoch": 11.195258019525802, + "grad_norm": 0.3897510766983032, + "learning_rate": 9.975929997593001e-06, + "loss": 0.003, + "step": 24081 + }, + { + "epoch": 11.195722919572292, + "grad_norm": 5.82058572769165, + "learning_rate": 9.975928997592902e-06, + "loss": 0.0543, + "step": 24082 + }, + { + "epoch": 11.196187819618782, + "grad_norm": 5.437811851501465, + "learning_rate": 9.9759279975928e-06, + "loss": 0.0654, + "step": 24083 + }, + { + "epoch": 11.196652719665272, + "grad_norm": 5.533193588256836, + "learning_rate": 9.975926997592701e-06, + "loss": 0.0661, + "step": 24084 + }, + { + "epoch": 11.197117619711761, + "grad_norm": 6.952012062072754, + "learning_rate": 9.9759259975926e-06, + "loss": 0.0637, + "step": 24085 + }, + { + "epoch": 11.197582519758251, + "grad_norm": 5.791411399841309, + "learning_rate": 9.9759249975925e-06, + "loss": 0.05, + "step": 24086 + }, + { + "epoch": 11.198047419804743, + "grad_norm": 9.457024574279785, + "learning_rate": 9.9759239975924e-06, + "loss": 0.2101, + "step": 24087 + }, + { + "epoch": 11.198512319851233, + "grad_norm": 4.6623992919921875, + "learning_rate": 9.9759229975923e-06, + "loss": 0.0309, + "step": 24088 + }, + { + "epoch": 11.198977219897722, + "grad_norm": 7.217007160186768, + "learning_rate": 9.975921997592201e-06, + "loss": 0.178, + "step": 24089 + }, + { + "epoch": 11.199442119944212, + "grad_norm": 2.0806987285614014, + "learning_rate": 9.9759209975921e-06, + "loss": 0.0214, + "step": 24090 + }, + { + "epoch": 11.199907019990702, + "grad_norm": 4.663397789001465, + "learning_rate": 9.975919997592e-06, + "loss": 0.0563, + "step": 24091 + }, + { + "epoch": 11.200371920037192, + "grad_norm": 7.807209014892578, + "learning_rate": 9.9759189975919e-06, + "loss": 0.0537, + "step": 24092 + }, + { + "epoch": 11.200836820083682, + "grad_norm": 4.788042068481445, + "learning_rate": 9.9759179975918e-06, + "loss": 0.0825, + "step": 24093 + }, + { + "epoch": 11.201301720130171, + "grad_norm": 2.8883581161499023, + "learning_rate": 9.975916997591701e-06, + "loss": 0.014, + "step": 24094 + }, + { + "epoch": 11.201766620176661, + "grad_norm": 2.567640781402588, + "learning_rate": 9.975915997591602e-06, + "loss": 0.0366, + "step": 24095 + }, + { + "epoch": 11.202231520223153, + "grad_norm": 4.880044937133789, + "learning_rate": 9.9759149975915e-06, + "loss": 0.1066, + "step": 24096 + }, + { + "epoch": 11.202696420269643, + "grad_norm": 5.82090950012207, + "learning_rate": 9.975913997591401e-06, + "loss": 0.0978, + "step": 24097 + }, + { + "epoch": 11.203161320316132, + "grad_norm": 6.154175758361816, + "learning_rate": 9.9759129975913e-06, + "loss": 0.0818, + "step": 24098 + }, + { + "epoch": 11.203626220362622, + "grad_norm": 5.03303861618042, + "learning_rate": 9.975911997591201e-06, + "loss": 0.0833, + "step": 24099 + }, + { + "epoch": 11.204091120409112, + "grad_norm": 2.5556833744049072, + "learning_rate": 9.9759109975911e-06, + "loss": 0.0287, + "step": 24100 + }, + { + "epoch": 11.204556020455602, + "grad_norm": 6.3657755851745605, + "learning_rate": 9.975909997591e-06, + "loss": 0.0895, + "step": 24101 + }, + { + "epoch": 11.205020920502092, + "grad_norm": 2.7316293716430664, + "learning_rate": 9.975908997590901e-06, + "loss": 0.0318, + "step": 24102 + }, + { + "epoch": 11.205485820548581, + "grad_norm": 0.6500588655471802, + "learning_rate": 9.9759079975908e-06, + "loss": 0.0026, + "step": 24103 + }, + { + "epoch": 11.205950720595071, + "grad_norm": 4.070196628570557, + "learning_rate": 9.9759069975907e-06, + "loss": 0.0507, + "step": 24104 + }, + { + "epoch": 11.206415620641563, + "grad_norm": 5.54578971862793, + "learning_rate": 9.9759059975906e-06, + "loss": 0.0744, + "step": 24105 + }, + { + "epoch": 11.206880520688053, + "grad_norm": 4.0503339767456055, + "learning_rate": 9.9759049975905e-06, + "loss": 0.0418, + "step": 24106 + }, + { + "epoch": 11.207345420734542, + "grad_norm": 2.2569704055786133, + "learning_rate": 9.975903997590401e-06, + "loss": 0.0112, + "step": 24107 + }, + { + "epoch": 11.207810320781032, + "grad_norm": 5.280997276306152, + "learning_rate": 9.975902997590302e-06, + "loss": 0.1128, + "step": 24108 + }, + { + "epoch": 11.208275220827522, + "grad_norm": 6.700904369354248, + "learning_rate": 9.9759019975902e-06, + "loss": 0.0527, + "step": 24109 + }, + { + "epoch": 11.208740120874012, + "grad_norm": 3.6846320629119873, + "learning_rate": 9.975900997590101e-06, + "loss": 0.0281, + "step": 24110 + }, + { + "epoch": 11.209205020920502, + "grad_norm": 2.796502113342285, + "learning_rate": 9.97589999759e-06, + "loss": 0.0242, + "step": 24111 + }, + { + "epoch": 11.209669920966991, + "grad_norm": 4.637854099273682, + "learning_rate": 9.9758989975899e-06, + "loss": 0.0573, + "step": 24112 + }, + { + "epoch": 11.210134821013481, + "grad_norm": 6.249614715576172, + "learning_rate": 9.9758979975898e-06, + "loss": 0.091, + "step": 24113 + }, + { + "epoch": 11.210599721059973, + "grad_norm": 4.1761088371276855, + "learning_rate": 9.9758969975897e-06, + "loss": 0.0367, + "step": 24114 + }, + { + "epoch": 11.211064621106463, + "grad_norm": 5.95216178894043, + "learning_rate": 9.975895997589601e-06, + "loss": 0.0956, + "step": 24115 + }, + { + "epoch": 11.211529521152952, + "grad_norm": 8.891772270202637, + "learning_rate": 9.9758949975895e-06, + "loss": 0.1439, + "step": 24116 + }, + { + "epoch": 11.211994421199442, + "grad_norm": 3.458272695541382, + "learning_rate": 9.975893997589401e-06, + "loss": 0.036, + "step": 24117 + }, + { + "epoch": 11.212459321245932, + "grad_norm": 4.704968452453613, + "learning_rate": 9.9758929975893e-06, + "loss": 0.0597, + "step": 24118 + }, + { + "epoch": 11.212924221292422, + "grad_norm": 5.822587490081787, + "learning_rate": 9.9758919975892e-06, + "loss": 0.1114, + "step": 24119 + }, + { + "epoch": 11.213389121338912, + "grad_norm": 4.052740097045898, + "learning_rate": 9.975890997589101e-06, + "loss": 0.0449, + "step": 24120 + }, + { + "epoch": 11.213854021385401, + "grad_norm": 14.577268600463867, + "learning_rate": 9.975889997589002e-06, + "loss": 0.059, + "step": 24121 + }, + { + "epoch": 11.214318921431893, + "grad_norm": 8.345784187316895, + "learning_rate": 9.9758889975889e-06, + "loss": 0.0625, + "step": 24122 + }, + { + "epoch": 11.214783821478383, + "grad_norm": 10.209118843078613, + "learning_rate": 9.9758879975888e-06, + "loss": 0.0658, + "step": 24123 + }, + { + "epoch": 11.215248721524873, + "grad_norm": 5.530219078063965, + "learning_rate": 9.9758869975887e-06, + "loss": 0.0911, + "step": 24124 + }, + { + "epoch": 11.215713621571362, + "grad_norm": 1.1880078315734863, + "learning_rate": 9.9758859975886e-06, + "loss": 0.0096, + "step": 24125 + }, + { + "epoch": 11.216178521617852, + "grad_norm": 0.2777613401412964, + "learning_rate": 9.9758849975885e-06, + "loss": 0.0027, + "step": 24126 + }, + { + "epoch": 11.216643421664342, + "grad_norm": 3.766094446182251, + "learning_rate": 9.9758839975884e-06, + "loss": 0.0426, + "step": 24127 + }, + { + "epoch": 11.217108321710832, + "grad_norm": 5.490427494049072, + "learning_rate": 9.975882997588301e-06, + "loss": 0.0577, + "step": 24128 + }, + { + "epoch": 11.217573221757322, + "grad_norm": 1.578629493713379, + "learning_rate": 9.9758819975882e-06, + "loss": 0.0136, + "step": 24129 + }, + { + "epoch": 11.218038121803811, + "grad_norm": 3.7229743003845215, + "learning_rate": 9.975880997588101e-06, + "loss": 0.0731, + "step": 24130 + }, + { + "epoch": 11.218503021850303, + "grad_norm": 6.446311950683594, + "learning_rate": 9.975879997588e-06, + "loss": 0.1007, + "step": 24131 + }, + { + "epoch": 11.218967921896793, + "grad_norm": 9.273463249206543, + "learning_rate": 9.9758789975879e-06, + "loss": 0.1005, + "step": 24132 + }, + { + "epoch": 11.219432821943283, + "grad_norm": 3.7018351554870605, + "learning_rate": 9.975877997587801e-06, + "loss": 0.0586, + "step": 24133 + }, + { + "epoch": 11.219897721989772, + "grad_norm": 10.465205192565918, + "learning_rate": 9.9758769975877e-06, + "loss": 0.1663, + "step": 24134 + }, + { + "epoch": 11.220362622036262, + "grad_norm": 2.015561819076538, + "learning_rate": 9.975875997587601e-06, + "loss": 0.0121, + "step": 24135 + }, + { + "epoch": 11.220827522082752, + "grad_norm": 2.381479263305664, + "learning_rate": 9.9758749975875e-06, + "loss": 0.0154, + "step": 24136 + }, + { + "epoch": 11.221292422129242, + "grad_norm": 9.084456443786621, + "learning_rate": 9.9758739975874e-06, + "loss": 0.1474, + "step": 24137 + }, + { + "epoch": 11.221757322175732, + "grad_norm": 7.869433879852295, + "learning_rate": 9.975872997587301e-06, + "loss": 0.1004, + "step": 24138 + }, + { + "epoch": 11.222222222222221, + "grad_norm": 1.6615960597991943, + "learning_rate": 9.9758719975872e-06, + "loss": 0.0182, + "step": 24139 + }, + { + "epoch": 11.222687122268713, + "grad_norm": 1.374893069267273, + "learning_rate": 9.9758709975871e-06, + "loss": 0.0055, + "step": 24140 + }, + { + "epoch": 11.223152022315203, + "grad_norm": 3.3132741451263428, + "learning_rate": 9.975869997587001e-06, + "loss": 0.0427, + "step": 24141 + }, + { + "epoch": 11.223616922361693, + "grad_norm": 8.835244178771973, + "learning_rate": 9.9758689975869e-06, + "loss": 0.1429, + "step": 24142 + }, + { + "epoch": 11.224081822408182, + "grad_norm": 1.7227284908294678, + "learning_rate": 9.9758679975868e-06, + "loss": 0.0109, + "step": 24143 + }, + { + "epoch": 11.224546722454672, + "grad_norm": 5.75783634185791, + "learning_rate": 9.9758669975867e-06, + "loss": 0.069, + "step": 24144 + }, + { + "epoch": 11.225011622501162, + "grad_norm": 5.099205493927002, + "learning_rate": 9.9758659975866e-06, + "loss": 0.081, + "step": 24145 + }, + { + "epoch": 11.225476522547652, + "grad_norm": 4.768218040466309, + "learning_rate": 9.975864997586501e-06, + "loss": 0.0481, + "step": 24146 + }, + { + "epoch": 11.225941422594142, + "grad_norm": 3.8261241912841797, + "learning_rate": 9.9758639975864e-06, + "loss": 0.026, + "step": 24147 + }, + { + "epoch": 11.226406322640631, + "grad_norm": 8.780701637268066, + "learning_rate": 9.975862997586301e-06, + "loss": 0.0473, + "step": 24148 + }, + { + "epoch": 11.226871222687123, + "grad_norm": 2.490805149078369, + "learning_rate": 9.9758619975862e-06, + "loss": 0.0237, + "step": 24149 + }, + { + "epoch": 11.227336122733613, + "grad_norm": 0.5799334049224854, + "learning_rate": 9.9758609975861e-06, + "loss": 0.0046, + "step": 24150 + }, + { + "epoch": 11.227801022780103, + "grad_norm": 5.624022483825684, + "learning_rate": 9.975859997586001e-06, + "loss": 0.0519, + "step": 24151 + }, + { + "epoch": 11.228265922826592, + "grad_norm": 1.7294762134552002, + "learning_rate": 9.975858997585902e-06, + "loss": 0.0218, + "step": 24152 + }, + { + "epoch": 11.228730822873082, + "grad_norm": 0.7369518876075745, + "learning_rate": 9.975857997585801e-06, + "loss": 0.0057, + "step": 24153 + }, + { + "epoch": 11.229195722919572, + "grad_norm": 6.60860538482666, + "learning_rate": 9.9758569975857e-06, + "loss": 0.0883, + "step": 24154 + }, + { + "epoch": 11.229660622966062, + "grad_norm": 5.716125965118408, + "learning_rate": 9.9758559975856e-06, + "loss": 0.1278, + "step": 24155 + }, + { + "epoch": 11.230125523012552, + "grad_norm": 6.607271671295166, + "learning_rate": 9.9758549975855e-06, + "loss": 0.0722, + "step": 24156 + }, + { + "epoch": 11.230590423059041, + "grad_norm": 6.324148654937744, + "learning_rate": 9.9758539975854e-06, + "loss": 0.0991, + "step": 24157 + }, + { + "epoch": 11.231055323105533, + "grad_norm": 5.029421329498291, + "learning_rate": 9.9758529975853e-06, + "loss": 0.0717, + "step": 24158 + }, + { + "epoch": 11.231520223152023, + "grad_norm": 2.0837459564208984, + "learning_rate": 9.975851997585201e-06, + "loss": 0.0148, + "step": 24159 + }, + { + "epoch": 11.231985123198513, + "grad_norm": 3.4656155109405518, + "learning_rate": 9.9758509975851e-06, + "loss": 0.0563, + "step": 24160 + }, + { + "epoch": 11.232450023245002, + "grad_norm": 7.088159084320068, + "learning_rate": 9.975849997585001e-06, + "loss": 0.0628, + "step": 24161 + }, + { + "epoch": 11.232914923291492, + "grad_norm": 9.569660186767578, + "learning_rate": 9.9758489975849e-06, + "loss": 0.1207, + "step": 24162 + }, + { + "epoch": 11.233379823337982, + "grad_norm": 5.87129020690918, + "learning_rate": 9.9758479975848e-06, + "loss": 0.071, + "step": 24163 + }, + { + "epoch": 11.233844723384472, + "grad_norm": 2.055914878845215, + "learning_rate": 9.975846997584701e-06, + "loss": 0.0173, + "step": 24164 + }, + { + "epoch": 11.234309623430962, + "grad_norm": 12.651568412780762, + "learning_rate": 9.9758459975846e-06, + "loss": 0.2721, + "step": 24165 + }, + { + "epoch": 11.234774523477453, + "grad_norm": 5.8847198486328125, + "learning_rate": 9.975844997584501e-06, + "loss": 0.1215, + "step": 24166 + }, + { + "epoch": 11.235239423523943, + "grad_norm": 0.6810613870620728, + "learning_rate": 9.9758439975844e-06, + "loss": 0.0057, + "step": 24167 + }, + { + "epoch": 11.235704323570433, + "grad_norm": 3.054699420928955, + "learning_rate": 9.9758429975843e-06, + "loss": 0.0401, + "step": 24168 + }, + { + "epoch": 11.236169223616923, + "grad_norm": 7.238572120666504, + "learning_rate": 9.9758419975842e-06, + "loss": 0.0959, + "step": 24169 + }, + { + "epoch": 11.236634123663412, + "grad_norm": 7.572151184082031, + "learning_rate": 9.9758409975841e-06, + "loss": 0.0405, + "step": 24170 + }, + { + "epoch": 11.237099023709902, + "grad_norm": 3.7995402812957764, + "learning_rate": 9.975839997584e-06, + "loss": 0.0574, + "step": 24171 + }, + { + "epoch": 11.237563923756392, + "grad_norm": 7.398442268371582, + "learning_rate": 9.975838997583901e-06, + "loss": 0.1263, + "step": 24172 + }, + { + "epoch": 11.238028823802882, + "grad_norm": 7.316294193267822, + "learning_rate": 9.9758379975838e-06, + "loss": 0.0843, + "step": 24173 + }, + { + "epoch": 11.238493723849372, + "grad_norm": 3.27687668800354, + "learning_rate": 9.975836997583701e-06, + "loss": 0.0418, + "step": 24174 + }, + { + "epoch": 11.238958623895863, + "grad_norm": 5.58623743057251, + "learning_rate": 9.9758359975836e-06, + "loss": 0.0829, + "step": 24175 + }, + { + "epoch": 11.239423523942353, + "grad_norm": 4.441181182861328, + "learning_rate": 9.9758349975835e-06, + "loss": 0.0546, + "step": 24176 + }, + { + "epoch": 11.239888423988843, + "grad_norm": 4.321049213409424, + "learning_rate": 9.975833997583401e-06, + "loss": 0.0452, + "step": 24177 + }, + { + "epoch": 11.240353324035333, + "grad_norm": 2.4855170249938965, + "learning_rate": 9.9758329975833e-06, + "loss": 0.0298, + "step": 24178 + }, + { + "epoch": 11.240818224081822, + "grad_norm": 2.038181781768799, + "learning_rate": 9.975831997583201e-06, + "loss": 0.0138, + "step": 24179 + }, + { + "epoch": 11.241283124128312, + "grad_norm": 7.735072612762451, + "learning_rate": 9.9758309975831e-06, + "loss": 0.1173, + "step": 24180 + }, + { + "epoch": 11.241748024174802, + "grad_norm": 6.041223049163818, + "learning_rate": 9.975829997583e-06, + "loss": 0.0869, + "step": 24181 + }, + { + "epoch": 11.242212924221292, + "grad_norm": 3.1885104179382324, + "learning_rate": 9.9758289975829e-06, + "loss": 0.0313, + "step": 24182 + }, + { + "epoch": 11.242677824267782, + "grad_norm": 2.80533504486084, + "learning_rate": 9.9758279975828e-06, + "loss": 0.0148, + "step": 24183 + }, + { + "epoch": 11.243142724314273, + "grad_norm": 6.507983684539795, + "learning_rate": 9.975826997582701e-06, + "loss": 0.0667, + "step": 24184 + }, + { + "epoch": 11.243607624360763, + "grad_norm": 5.855647087097168, + "learning_rate": 9.975825997582602e-06, + "loss": 0.076, + "step": 24185 + }, + { + "epoch": 11.244072524407253, + "grad_norm": 7.243044853210449, + "learning_rate": 9.9758249975825e-06, + "loss": 0.1246, + "step": 24186 + }, + { + "epoch": 11.244537424453743, + "grad_norm": 4.7543745040893555, + "learning_rate": 9.9758239975824e-06, + "loss": 0.1031, + "step": 24187 + }, + { + "epoch": 11.245002324500232, + "grad_norm": 7.455020427703857, + "learning_rate": 9.9758229975823e-06, + "loss": 0.1108, + "step": 24188 + }, + { + "epoch": 11.245467224546722, + "grad_norm": 6.552161693572998, + "learning_rate": 9.9758219975822e-06, + "loss": 0.0959, + "step": 24189 + }, + { + "epoch": 11.245932124593212, + "grad_norm": 5.186428546905518, + "learning_rate": 9.975820997582101e-06, + "loss": 0.0713, + "step": 24190 + }, + { + "epoch": 11.246397024639702, + "grad_norm": 3.326951265335083, + "learning_rate": 9.975819997582e-06, + "loss": 0.0365, + "step": 24191 + }, + { + "epoch": 11.246861924686192, + "grad_norm": 2.7325353622436523, + "learning_rate": 9.975818997581901e-06, + "loss": 0.0223, + "step": 24192 + }, + { + "epoch": 11.247326824732683, + "grad_norm": 4.9112348556518555, + "learning_rate": 9.9758179975818e-06, + "loss": 0.0508, + "step": 24193 + }, + { + "epoch": 11.247791724779173, + "grad_norm": 5.19210147857666, + "learning_rate": 9.9758169975817e-06, + "loss": 0.0549, + "step": 24194 + }, + { + "epoch": 11.248256624825663, + "grad_norm": 1.7581589221954346, + "learning_rate": 9.9758159975816e-06, + "loss": 0.0145, + "step": 24195 + }, + { + "epoch": 11.248721524872153, + "grad_norm": 3.3961522579193115, + "learning_rate": 9.9758149975815e-06, + "loss": 0.0447, + "step": 24196 + }, + { + "epoch": 11.249186424918642, + "grad_norm": 6.199566841125488, + "learning_rate": 9.975813997581401e-06, + "loss": 0.1058, + "step": 24197 + }, + { + "epoch": 11.249651324965132, + "grad_norm": 4.0213165283203125, + "learning_rate": 9.9758129975813e-06, + "loss": 0.0421, + "step": 24198 + }, + { + "epoch": 11.250116225011622, + "grad_norm": 4.854126930236816, + "learning_rate": 9.9758119975812e-06, + "loss": 0.0411, + "step": 24199 + }, + { + "epoch": 11.250581125058112, + "grad_norm": 10.039356231689453, + "learning_rate": 9.9758109975811e-06, + "loss": 0.1145, + "step": 24200 + }, + { + "epoch": 11.251046025104603, + "grad_norm": 9.079814910888672, + "learning_rate": 9.975809997581e-06, + "loss": 0.2243, + "step": 24201 + }, + { + "epoch": 11.251510925151093, + "grad_norm": 4.318491458892822, + "learning_rate": 9.975808997580901e-06, + "loss": 0.0635, + "step": 24202 + }, + { + "epoch": 11.251975825197583, + "grad_norm": 1.4761006832122803, + "learning_rate": 9.975807997580802e-06, + "loss": 0.0104, + "step": 24203 + }, + { + "epoch": 11.252440725244073, + "grad_norm": 3.239910125732422, + "learning_rate": 9.9758069975807e-06, + "loss": 0.0413, + "step": 24204 + }, + { + "epoch": 11.252905625290563, + "grad_norm": 5.970784664154053, + "learning_rate": 9.975805997580601e-06, + "loss": 0.1093, + "step": 24205 + }, + { + "epoch": 11.253370525337052, + "grad_norm": 3.08036470413208, + "learning_rate": 9.9758049975805e-06, + "loss": 0.0169, + "step": 24206 + }, + { + "epoch": 11.253835425383542, + "grad_norm": 3.2362921237945557, + "learning_rate": 9.9758039975804e-06, + "loss": 0.0325, + "step": 24207 + }, + { + "epoch": 11.254300325430032, + "grad_norm": 2.6645069122314453, + "learning_rate": 9.9758029975803e-06, + "loss": 0.0234, + "step": 24208 + }, + { + "epoch": 11.254765225476522, + "grad_norm": 6.378664493560791, + "learning_rate": 9.9758019975802e-06, + "loss": 0.0967, + "step": 24209 + }, + { + "epoch": 11.255230125523013, + "grad_norm": 4.888156890869141, + "learning_rate": 9.975800997580101e-06, + "loss": 0.0824, + "step": 24210 + }, + { + "epoch": 11.255695025569503, + "grad_norm": 4.128586769104004, + "learning_rate": 9.97579999758e-06, + "loss": 0.0458, + "step": 24211 + }, + { + "epoch": 11.256159925615993, + "grad_norm": 4.935535430908203, + "learning_rate": 9.9757989975799e-06, + "loss": 0.0519, + "step": 24212 + }, + { + "epoch": 11.256624825662483, + "grad_norm": 2.403560161590576, + "learning_rate": 9.9757979975798e-06, + "loss": 0.0227, + "step": 24213 + }, + { + "epoch": 11.257089725708973, + "grad_norm": 6.20933198928833, + "learning_rate": 9.9757969975797e-06, + "loss": 0.0291, + "step": 24214 + }, + { + "epoch": 11.257554625755462, + "grad_norm": 2.0453715324401855, + "learning_rate": 9.975795997579601e-06, + "loss": 0.0217, + "step": 24215 + }, + { + "epoch": 11.258019525801952, + "grad_norm": 6.510076999664307, + "learning_rate": 9.975794997579502e-06, + "loss": 0.095, + "step": 24216 + }, + { + "epoch": 11.258484425848442, + "grad_norm": 1.9795689582824707, + "learning_rate": 9.9757939975794e-06, + "loss": 0.0143, + "step": 24217 + }, + { + "epoch": 11.258949325894932, + "grad_norm": 2.571068048477173, + "learning_rate": 9.975792997579301e-06, + "loss": 0.0287, + "step": 24218 + }, + { + "epoch": 11.259414225941423, + "grad_norm": 4.719318866729736, + "learning_rate": 9.9757919975792e-06, + "loss": 0.0609, + "step": 24219 + }, + { + "epoch": 11.259879125987913, + "grad_norm": 2.9732558727264404, + "learning_rate": 9.975790997579101e-06, + "loss": 0.0298, + "step": 24220 + }, + { + "epoch": 11.260344026034403, + "grad_norm": 5.9011311531066895, + "learning_rate": 9.975789997579002e-06, + "loss": 0.1867, + "step": 24221 + }, + { + "epoch": 11.260808926080893, + "grad_norm": 7.358633518218994, + "learning_rate": 9.9757889975789e-06, + "loss": 0.1409, + "step": 24222 + }, + { + "epoch": 11.261273826127383, + "grad_norm": 1.9252532720565796, + "learning_rate": 9.975787997578801e-06, + "loss": 0.0133, + "step": 24223 + }, + { + "epoch": 11.261738726173872, + "grad_norm": 0.9188494682312012, + "learning_rate": 9.9757869975787e-06, + "loss": 0.0073, + "step": 24224 + }, + { + "epoch": 11.262203626220362, + "grad_norm": 4.701203346252441, + "learning_rate": 9.9757859975786e-06, + "loss": 0.0842, + "step": 24225 + }, + { + "epoch": 11.262668526266852, + "grad_norm": 7.841366767883301, + "learning_rate": 9.9757849975785e-06, + "loss": 0.1544, + "step": 24226 + }, + { + "epoch": 11.263133426313342, + "grad_norm": 6.559917449951172, + "learning_rate": 9.9757839975784e-06, + "loss": 0.0735, + "step": 24227 + }, + { + "epoch": 11.263598326359833, + "grad_norm": 3.8244714736938477, + "learning_rate": 9.975782997578301e-06, + "loss": 0.0476, + "step": 24228 + }, + { + "epoch": 11.264063226406323, + "grad_norm": 7.0800700187683105, + "learning_rate": 9.975781997578202e-06, + "loss": 0.0625, + "step": 24229 + }, + { + "epoch": 11.264528126452813, + "grad_norm": 4.288783550262451, + "learning_rate": 9.9757809975781e-06, + "loss": 0.0652, + "step": 24230 + }, + { + "epoch": 11.264993026499303, + "grad_norm": 4.95961856842041, + "learning_rate": 9.975779997578e-06, + "loss": 0.1139, + "step": 24231 + }, + { + "epoch": 11.265457926545793, + "grad_norm": 2.892932415008545, + "learning_rate": 9.9757789975779e-06, + "loss": 0.0257, + "step": 24232 + }, + { + "epoch": 11.265922826592282, + "grad_norm": 3.6806042194366455, + "learning_rate": 9.975777997577801e-06, + "loss": 0.0699, + "step": 24233 + }, + { + "epoch": 11.266387726638772, + "grad_norm": 2.0941944122314453, + "learning_rate": 9.975776997577702e-06, + "loss": 0.0161, + "step": 24234 + }, + { + "epoch": 11.266852626685262, + "grad_norm": 7.097652435302734, + "learning_rate": 9.9757759975776e-06, + "loss": 0.1531, + "step": 24235 + }, + { + "epoch": 11.267317526731752, + "grad_norm": 6.174306869506836, + "learning_rate": 9.975774997577501e-06, + "loss": 0.1147, + "step": 24236 + }, + { + "epoch": 11.267782426778243, + "grad_norm": 8.484996795654297, + "learning_rate": 9.9757739975774e-06, + "loss": 0.121, + "step": 24237 + }, + { + "epoch": 11.268247326824733, + "grad_norm": 4.391073226928711, + "learning_rate": 9.975772997577301e-06, + "loss": 0.0596, + "step": 24238 + }, + { + "epoch": 11.268712226871223, + "grad_norm": 4.870782852172852, + "learning_rate": 9.9757719975772e-06, + "loss": 0.0427, + "step": 24239 + }, + { + "epoch": 11.269177126917713, + "grad_norm": 4.449288368225098, + "learning_rate": 9.9757709975771e-06, + "loss": 0.0955, + "step": 24240 + }, + { + "epoch": 11.269642026964203, + "grad_norm": 5.8658270835876465, + "learning_rate": 9.975769997577001e-06, + "loss": 0.0557, + "step": 24241 + }, + { + "epoch": 11.270106927010692, + "grad_norm": 6.864102840423584, + "learning_rate": 9.9757689975769e-06, + "loss": 0.1237, + "step": 24242 + }, + { + "epoch": 11.270571827057182, + "grad_norm": 6.792637348175049, + "learning_rate": 9.9757679975768e-06, + "loss": 0.0694, + "step": 24243 + }, + { + "epoch": 11.271036727103672, + "grad_norm": 5.91644287109375, + "learning_rate": 9.9757669975767e-06, + "loss": 0.0797, + "step": 24244 + }, + { + "epoch": 11.271501627150164, + "grad_norm": 4.105605125427246, + "learning_rate": 9.9757659975766e-06, + "loss": 0.0369, + "step": 24245 + }, + { + "epoch": 11.271966527196653, + "grad_norm": 4.672788143157959, + "learning_rate": 9.975764997576501e-06, + "loss": 0.0596, + "step": 24246 + }, + { + "epoch": 11.272431427243143, + "grad_norm": 2.5427637100219727, + "learning_rate": 9.975763997576402e-06, + "loss": 0.0152, + "step": 24247 + }, + { + "epoch": 11.272896327289633, + "grad_norm": 1.7613006830215454, + "learning_rate": 9.9757629975763e-06, + "loss": 0.015, + "step": 24248 + }, + { + "epoch": 11.273361227336123, + "grad_norm": 6.343465805053711, + "learning_rate": 9.975761997576201e-06, + "loss": 0.1292, + "step": 24249 + }, + { + "epoch": 11.273826127382613, + "grad_norm": 6.020542144775391, + "learning_rate": 9.9757609975761e-06, + "loss": 0.0752, + "step": 24250 + }, + { + "epoch": 11.274291027429102, + "grad_norm": 2.511319160461426, + "learning_rate": 9.975759997576001e-06, + "loss": 0.0242, + "step": 24251 + }, + { + "epoch": 11.274755927475592, + "grad_norm": 4.940680980682373, + "learning_rate": 9.9757589975759e-06, + "loss": 0.0447, + "step": 24252 + }, + { + "epoch": 11.275220827522082, + "grad_norm": 6.141909122467041, + "learning_rate": 9.9757579975758e-06, + "loss": 0.0888, + "step": 24253 + }, + { + "epoch": 11.275685727568574, + "grad_norm": 2.927715301513672, + "learning_rate": 9.975756997575701e-06, + "loss": 0.0376, + "step": 24254 + }, + { + "epoch": 11.276150627615063, + "grad_norm": 2.8073744773864746, + "learning_rate": 9.9757559975756e-06, + "loss": 0.0227, + "step": 24255 + }, + { + "epoch": 11.276615527661553, + "grad_norm": 3.7073378562927246, + "learning_rate": 9.9757549975755e-06, + "loss": 0.0275, + "step": 24256 + }, + { + "epoch": 11.277080427708043, + "grad_norm": 3.058783531188965, + "learning_rate": 9.9757539975754e-06, + "loss": 0.0351, + "step": 24257 + }, + { + "epoch": 11.277545327754533, + "grad_norm": 6.421762466430664, + "learning_rate": 9.9757529975753e-06, + "loss": 0.0909, + "step": 24258 + }, + { + "epoch": 11.278010227801023, + "grad_norm": 1.6308746337890625, + "learning_rate": 9.975751997575201e-06, + "loss": 0.0112, + "step": 24259 + }, + { + "epoch": 11.278475127847512, + "grad_norm": 5.228917121887207, + "learning_rate": 9.975750997575102e-06, + "loss": 0.1104, + "step": 24260 + }, + { + "epoch": 11.278940027894002, + "grad_norm": 3.0786643028259277, + "learning_rate": 9.975749997575e-06, + "loss": 0.0187, + "step": 24261 + }, + { + "epoch": 11.279404927940492, + "grad_norm": 9.143634796142578, + "learning_rate": 9.975748997574901e-06, + "loss": 0.1907, + "step": 24262 + }, + { + "epoch": 11.279869827986984, + "grad_norm": 7.847959518432617, + "learning_rate": 9.9757479975748e-06, + "loss": 0.0687, + "step": 24263 + }, + { + "epoch": 11.280334728033473, + "grad_norm": 4.484776020050049, + "learning_rate": 9.9757469975747e-06, + "loss": 0.0394, + "step": 24264 + }, + { + "epoch": 11.280799628079963, + "grad_norm": 5.758775234222412, + "learning_rate": 9.9757459975746e-06, + "loss": 0.0865, + "step": 24265 + }, + { + "epoch": 11.281264528126453, + "grad_norm": 5.510590076446533, + "learning_rate": 9.9757449975745e-06, + "loss": 0.1183, + "step": 24266 + }, + { + "epoch": 11.281729428172943, + "grad_norm": 2.369723081588745, + "learning_rate": 9.975743997574401e-06, + "loss": 0.0256, + "step": 24267 + }, + { + "epoch": 11.282194328219433, + "grad_norm": 3.614145517349243, + "learning_rate": 9.9757429975743e-06, + "loss": 0.0321, + "step": 24268 + }, + { + "epoch": 11.282659228265922, + "grad_norm": 4.596938133239746, + "learning_rate": 9.975741997574201e-06, + "loss": 0.0369, + "step": 24269 + }, + { + "epoch": 11.283124128312412, + "grad_norm": 5.992918491363525, + "learning_rate": 9.9757409975741e-06, + "loss": 0.0906, + "step": 24270 + }, + { + "epoch": 11.283589028358902, + "grad_norm": 3.2458386421203613, + "learning_rate": 9.975739997574e-06, + "loss": 0.0899, + "step": 24271 + }, + { + "epoch": 11.284053928405394, + "grad_norm": 3.8598294258117676, + "learning_rate": 9.975738997573901e-06, + "loss": 0.0526, + "step": 24272 + }, + { + "epoch": 11.284518828451883, + "grad_norm": 4.1241068840026855, + "learning_rate": 9.975737997573802e-06, + "loss": 0.0591, + "step": 24273 + }, + { + "epoch": 11.284983728498373, + "grad_norm": 6.229006290435791, + "learning_rate": 9.9757369975737e-06, + "loss": 0.1053, + "step": 24274 + }, + { + "epoch": 11.285448628544863, + "grad_norm": 2.553605079650879, + "learning_rate": 9.9757359975736e-06, + "loss": 0.0385, + "step": 24275 + }, + { + "epoch": 11.285913528591353, + "grad_norm": 5.144227504730225, + "learning_rate": 9.9757349975735e-06, + "loss": 0.0476, + "step": 24276 + }, + { + "epoch": 11.286378428637843, + "grad_norm": 3.261910915374756, + "learning_rate": 9.9757339975734e-06, + "loss": 0.025, + "step": 24277 + }, + { + "epoch": 11.286843328684332, + "grad_norm": 2.31709623336792, + "learning_rate": 9.9757329975733e-06, + "loss": 0.0223, + "step": 24278 + }, + { + "epoch": 11.287308228730822, + "grad_norm": 6.982172012329102, + "learning_rate": 9.9757319975732e-06, + "loss": 0.1027, + "step": 24279 + }, + { + "epoch": 11.287773128777314, + "grad_norm": 5.700996398925781, + "learning_rate": 9.975730997573101e-06, + "loss": 0.0643, + "step": 24280 + }, + { + "epoch": 11.288238028823804, + "grad_norm": 0.9428958296775818, + "learning_rate": 9.975729997573e-06, + "loss": 0.0066, + "step": 24281 + }, + { + "epoch": 11.288702928870293, + "grad_norm": 3.5957112312316895, + "learning_rate": 9.975728997572901e-06, + "loss": 0.041, + "step": 24282 + }, + { + "epoch": 11.289167828916783, + "grad_norm": 2.901172161102295, + "learning_rate": 9.9757279975728e-06, + "loss": 0.0265, + "step": 24283 + }, + { + "epoch": 11.289632728963273, + "grad_norm": 4.198513507843018, + "learning_rate": 9.9757269975727e-06, + "loss": 0.0497, + "step": 24284 + }, + { + "epoch": 11.290097629009763, + "grad_norm": 2.1372344493865967, + "learning_rate": 9.975725997572601e-06, + "loss": 0.0074, + "step": 24285 + }, + { + "epoch": 11.290562529056253, + "grad_norm": 5.137613296508789, + "learning_rate": 9.9757249975725e-06, + "loss": 0.0731, + "step": 24286 + }, + { + "epoch": 11.291027429102742, + "grad_norm": 1.8429830074310303, + "learning_rate": 9.975723997572401e-06, + "loss": 0.0175, + "step": 24287 + }, + { + "epoch": 11.291492329149232, + "grad_norm": 1.7768751382827759, + "learning_rate": 9.9757229975723e-06, + "loss": 0.0126, + "step": 24288 + }, + { + "epoch": 11.291957229195724, + "grad_norm": 4.973364353179932, + "learning_rate": 9.9757219975722e-06, + "loss": 0.0553, + "step": 24289 + }, + { + "epoch": 11.292422129242214, + "grad_norm": 6.740957736968994, + "learning_rate": 9.9757209975721e-06, + "loss": 0.1149, + "step": 24290 + }, + { + "epoch": 11.292887029288703, + "grad_norm": 5.531297206878662, + "learning_rate": 9.975719997572e-06, + "loss": 0.112, + "step": 24291 + }, + { + "epoch": 11.293351929335193, + "grad_norm": 5.616223335266113, + "learning_rate": 9.9757189975719e-06, + "loss": 0.087, + "step": 24292 + }, + { + "epoch": 11.293816829381683, + "grad_norm": 2.666907548904419, + "learning_rate": 9.975717997571801e-06, + "loss": 0.0283, + "step": 24293 + }, + { + "epoch": 11.294281729428173, + "grad_norm": 3.5477662086486816, + "learning_rate": 9.9757169975717e-06, + "loss": 0.0379, + "step": 24294 + }, + { + "epoch": 11.294746629474663, + "grad_norm": 8.61817455291748, + "learning_rate": 9.9757159975716e-06, + "loss": 0.089, + "step": 24295 + }, + { + "epoch": 11.295211529521152, + "grad_norm": 3.712175130844116, + "learning_rate": 9.9757149975715e-06, + "loss": 0.0361, + "step": 24296 + }, + { + "epoch": 11.295676429567642, + "grad_norm": 5.179362773895264, + "learning_rate": 9.9757139975714e-06, + "loss": 0.0841, + "step": 24297 + }, + { + "epoch": 11.296141329614134, + "grad_norm": 3.9865949153900146, + "learning_rate": 9.975712997571301e-06, + "loss": 0.0265, + "step": 24298 + }, + { + "epoch": 11.296606229660624, + "grad_norm": 5.11397647857666, + "learning_rate": 9.9757119975712e-06, + "loss": 0.0523, + "step": 24299 + }, + { + "epoch": 11.297071129707113, + "grad_norm": 5.1583428382873535, + "learning_rate": 9.975710997571101e-06, + "loss": 0.0459, + "step": 24300 + }, + { + "epoch": 11.297536029753603, + "grad_norm": 6.359983444213867, + "learning_rate": 9.975709997571e-06, + "loss": 0.0802, + "step": 24301 + }, + { + "epoch": 11.298000929800093, + "grad_norm": 5.98215389251709, + "learning_rate": 9.9757089975709e-06, + "loss": 0.0812, + "step": 24302 + }, + { + "epoch": 11.298465829846583, + "grad_norm": 2.2554943561553955, + "learning_rate": 9.975707997570801e-06, + "loss": 0.0156, + "step": 24303 + }, + { + "epoch": 11.298930729893073, + "grad_norm": 5.6393022537231445, + "learning_rate": 9.9757069975707e-06, + "loss": 0.0989, + "step": 24304 + }, + { + "epoch": 11.299395629939562, + "grad_norm": 5.728376865386963, + "learning_rate": 9.975705997570601e-06, + "loss": 0.0977, + "step": 24305 + }, + { + "epoch": 11.299860529986052, + "grad_norm": 7.6416401863098145, + "learning_rate": 9.9757049975705e-06, + "loss": 0.1851, + "step": 24306 + }, + { + "epoch": 11.300325430032544, + "grad_norm": 3.449918031692505, + "learning_rate": 9.9757039975704e-06, + "loss": 0.0352, + "step": 24307 + }, + { + "epoch": 11.300790330079034, + "grad_norm": 0.7596992254257202, + "learning_rate": 9.9757029975703e-06, + "loss": 0.0075, + "step": 24308 + }, + { + "epoch": 11.301255230125523, + "grad_norm": 3.25008225440979, + "learning_rate": 9.9757019975702e-06, + "loss": 0.0311, + "step": 24309 + }, + { + "epoch": 11.301720130172013, + "grad_norm": 2.8308568000793457, + "learning_rate": 9.9757009975701e-06, + "loss": 0.0317, + "step": 24310 + }, + { + "epoch": 11.302185030218503, + "grad_norm": 6.690362930297852, + "learning_rate": 9.975699997570001e-06, + "loss": 0.1048, + "step": 24311 + }, + { + "epoch": 11.302649930264993, + "grad_norm": 0.6887123584747314, + "learning_rate": 9.9756989975699e-06, + "loss": 0.0052, + "step": 24312 + }, + { + "epoch": 11.303114830311483, + "grad_norm": 11.238775253295898, + "learning_rate": 9.975697997569801e-06, + "loss": 0.2753, + "step": 24313 + }, + { + "epoch": 11.303579730357972, + "grad_norm": 2.684252977371216, + "learning_rate": 9.9756969975697e-06, + "loss": 0.024, + "step": 24314 + }, + { + "epoch": 11.304044630404462, + "grad_norm": 5.723259449005127, + "learning_rate": 9.9756959975696e-06, + "loss": 0.0964, + "step": 24315 + }, + { + "epoch": 11.304509530450954, + "grad_norm": 4.412761211395264, + "learning_rate": 9.975694997569501e-06, + "loss": 0.052, + "step": 24316 + }, + { + "epoch": 11.304974430497444, + "grad_norm": 2.7361114025115967, + "learning_rate": 9.9756939975694e-06, + "loss": 0.016, + "step": 24317 + }, + { + "epoch": 11.305439330543933, + "grad_norm": 4.170988082885742, + "learning_rate": 9.975692997569301e-06, + "loss": 0.0649, + "step": 24318 + }, + { + "epoch": 11.305904230590423, + "grad_norm": 6.659365177154541, + "learning_rate": 9.9756919975692e-06, + "loss": 0.0836, + "step": 24319 + }, + { + "epoch": 11.306369130636913, + "grad_norm": 2.1358349323272705, + "learning_rate": 9.9756909975691e-06, + "loss": 0.0167, + "step": 24320 + }, + { + "epoch": 11.306834030683403, + "grad_norm": 6.224632263183594, + "learning_rate": 9.975689997569e-06, + "loss": 0.0783, + "step": 24321 + }, + { + "epoch": 11.307298930729893, + "grad_norm": 2.851109266281128, + "learning_rate": 9.9756889975689e-06, + "loss": 0.0243, + "step": 24322 + }, + { + "epoch": 11.307763830776382, + "grad_norm": 4.549407482147217, + "learning_rate": 9.975687997568801e-06, + "loss": 0.1141, + "step": 24323 + }, + { + "epoch": 11.308228730822874, + "grad_norm": 4.511045932769775, + "learning_rate": 9.975686997568702e-06, + "loss": 0.0623, + "step": 24324 + }, + { + "epoch": 11.308693630869364, + "grad_norm": 4.772302627563477, + "learning_rate": 9.9756859975686e-06, + "loss": 0.0282, + "step": 24325 + }, + { + "epoch": 11.309158530915854, + "grad_norm": 4.053487300872803, + "learning_rate": 9.975684997568501e-06, + "loss": 0.0558, + "step": 24326 + }, + { + "epoch": 11.309623430962343, + "grad_norm": 7.473382472991943, + "learning_rate": 9.9756839975684e-06, + "loss": 0.0958, + "step": 24327 + }, + { + "epoch": 11.310088331008833, + "grad_norm": 3.6670448780059814, + "learning_rate": 9.9756829975683e-06, + "loss": 0.0291, + "step": 24328 + }, + { + "epoch": 11.310553231055323, + "grad_norm": 5.412452220916748, + "learning_rate": 9.975681997568201e-06, + "loss": 0.1187, + "step": 24329 + }, + { + "epoch": 11.311018131101813, + "grad_norm": 3.5456793308258057, + "learning_rate": 9.9756809975681e-06, + "loss": 0.0551, + "step": 24330 + }, + { + "epoch": 11.311483031148303, + "grad_norm": 4.224348545074463, + "learning_rate": 9.975679997568001e-06, + "loss": 0.0378, + "step": 24331 + }, + { + "epoch": 11.311947931194792, + "grad_norm": 6.709638595581055, + "learning_rate": 9.9756789975679e-06, + "loss": 0.091, + "step": 24332 + }, + { + "epoch": 11.312412831241284, + "grad_norm": 4.441938877105713, + "learning_rate": 9.9756779975678e-06, + "loss": 0.0673, + "step": 24333 + }, + { + "epoch": 11.312877731287774, + "grad_norm": 1.2468470335006714, + "learning_rate": 9.9756769975677e-06, + "loss": 0.0091, + "step": 24334 + }, + { + "epoch": 11.313342631334264, + "grad_norm": 3.560237169265747, + "learning_rate": 9.9756759975676e-06, + "loss": 0.0552, + "step": 24335 + }, + { + "epoch": 11.313807531380753, + "grad_norm": 3.4513704776763916, + "learning_rate": 9.975674997567501e-06, + "loss": 0.0432, + "step": 24336 + }, + { + "epoch": 11.314272431427243, + "grad_norm": 6.106902599334717, + "learning_rate": 9.975673997567402e-06, + "loss": 0.137, + "step": 24337 + }, + { + "epoch": 11.314737331473733, + "grad_norm": 2.512312650680542, + "learning_rate": 9.9756729975673e-06, + "loss": 0.0269, + "step": 24338 + }, + { + "epoch": 11.315202231520223, + "grad_norm": 3.1615076065063477, + "learning_rate": 9.9756719975672e-06, + "loss": 0.0246, + "step": 24339 + }, + { + "epoch": 11.315667131566713, + "grad_norm": 8.902026176452637, + "learning_rate": 9.9756709975671e-06, + "loss": 0.2458, + "step": 24340 + }, + { + "epoch": 11.316132031613202, + "grad_norm": 6.131204605102539, + "learning_rate": 9.975669997567e-06, + "loss": 0.116, + "step": 24341 + }, + { + "epoch": 11.316596931659694, + "grad_norm": 2.6719141006469727, + "learning_rate": 9.975668997566902e-06, + "loss": 0.0354, + "step": 24342 + }, + { + "epoch": 11.317061831706184, + "grad_norm": 1.914969801902771, + "learning_rate": 9.9756679975668e-06, + "loss": 0.0201, + "step": 24343 + }, + { + "epoch": 11.317526731752674, + "grad_norm": 7.2703094482421875, + "learning_rate": 9.975666997566701e-06, + "loss": 0.1105, + "step": 24344 + }, + { + "epoch": 11.317991631799163, + "grad_norm": 3.794888734817505, + "learning_rate": 9.9756659975666e-06, + "loss": 0.0514, + "step": 24345 + }, + { + "epoch": 11.318456531845653, + "grad_norm": 3.886930227279663, + "learning_rate": 9.9756649975665e-06, + "loss": 0.059, + "step": 24346 + }, + { + "epoch": 11.318921431892143, + "grad_norm": 9.05755615234375, + "learning_rate": 9.9756639975664e-06, + "loss": 0.1331, + "step": 24347 + }, + { + "epoch": 11.319386331938633, + "grad_norm": 2.6394617557525635, + "learning_rate": 9.9756629975663e-06, + "loss": 0.032, + "step": 24348 + }, + { + "epoch": 11.319851231985123, + "grad_norm": 0.857479989528656, + "learning_rate": 9.975661997566201e-06, + "loss": 0.0074, + "step": 24349 + }, + { + "epoch": 11.320316132031612, + "grad_norm": 3.2513558864593506, + "learning_rate": 9.9756609975661e-06, + "loss": 0.0668, + "step": 24350 + }, + { + "epoch": 11.320781032078104, + "grad_norm": 2.849933385848999, + "learning_rate": 9.975659997566e-06, + "loss": 0.0077, + "step": 24351 + }, + { + "epoch": 11.321245932124594, + "grad_norm": 7.137528896331787, + "learning_rate": 9.9756589975659e-06, + "loss": 0.0847, + "step": 24352 + }, + { + "epoch": 11.321710832171084, + "grad_norm": 3.638583183288574, + "learning_rate": 9.9756579975658e-06, + "loss": 0.0326, + "step": 24353 + }, + { + "epoch": 11.322175732217573, + "grad_norm": 5.501625061035156, + "learning_rate": 9.975656997565701e-06, + "loss": 0.0326, + "step": 24354 + }, + { + "epoch": 11.322640632264063, + "grad_norm": 5.815169334411621, + "learning_rate": 9.975655997565602e-06, + "loss": 0.0909, + "step": 24355 + }, + { + "epoch": 11.323105532310553, + "grad_norm": 3.3039700984954834, + "learning_rate": 9.9756549975655e-06, + "loss": 0.0203, + "step": 24356 + }, + { + "epoch": 11.323570432357043, + "grad_norm": 15.894665718078613, + "learning_rate": 9.975653997565401e-06, + "loss": 0.1748, + "step": 24357 + }, + { + "epoch": 11.324035332403533, + "grad_norm": 1.340576410293579, + "learning_rate": 9.9756529975653e-06, + "loss": 0.0134, + "step": 24358 + }, + { + "epoch": 11.324500232450024, + "grad_norm": 4.804885387420654, + "learning_rate": 9.9756519975652e-06, + "loss": 0.0363, + "step": 24359 + }, + { + "epoch": 11.324965132496514, + "grad_norm": 6.835543632507324, + "learning_rate": 9.9756509975651e-06, + "loss": 0.1144, + "step": 24360 + }, + { + "epoch": 11.325430032543004, + "grad_norm": 4.375940799713135, + "learning_rate": 9.975649997565e-06, + "loss": 0.0467, + "step": 24361 + }, + { + "epoch": 11.325894932589494, + "grad_norm": 4.766610145568848, + "learning_rate": 9.975648997564901e-06, + "loss": 0.0675, + "step": 24362 + }, + { + "epoch": 11.326359832635983, + "grad_norm": 4.134673595428467, + "learning_rate": 9.9756479975648e-06, + "loss": 0.0301, + "step": 24363 + }, + { + "epoch": 11.326824732682473, + "grad_norm": 3.218768358230591, + "learning_rate": 9.9756469975647e-06, + "loss": 0.0362, + "step": 24364 + }, + { + "epoch": 11.327289632728963, + "grad_norm": 3.0230977535247803, + "learning_rate": 9.9756459975646e-06, + "loss": 0.0147, + "step": 24365 + }, + { + "epoch": 11.327754532775453, + "grad_norm": 1.750834584236145, + "learning_rate": 9.9756449975645e-06, + "loss": 0.0121, + "step": 24366 + }, + { + "epoch": 11.328219432821943, + "grad_norm": 5.512294769287109, + "learning_rate": 9.975643997564401e-06, + "loss": 0.0971, + "step": 24367 + }, + { + "epoch": 11.328684332868434, + "grad_norm": 2.626619338989258, + "learning_rate": 9.975642997564302e-06, + "loss": 0.0246, + "step": 24368 + }, + { + "epoch": 11.329149232914924, + "grad_norm": 3.674393653869629, + "learning_rate": 9.9756419975642e-06, + "loss": 0.0449, + "step": 24369 + }, + { + "epoch": 11.329614132961414, + "grad_norm": 2.2971835136413574, + "learning_rate": 9.975640997564101e-06, + "loss": 0.0225, + "step": 24370 + }, + { + "epoch": 11.330079033007904, + "grad_norm": 6.592864036560059, + "learning_rate": 9.975639997564e-06, + "loss": 0.1069, + "step": 24371 + }, + { + "epoch": 11.330543933054393, + "grad_norm": 1.6906089782714844, + "learning_rate": 9.975638997563901e-06, + "loss": 0.0122, + "step": 24372 + }, + { + "epoch": 11.331008833100883, + "grad_norm": 8.288280487060547, + "learning_rate": 9.9756379975638e-06, + "loss": 0.1361, + "step": 24373 + }, + { + "epoch": 11.331473733147373, + "grad_norm": 5.688379764556885, + "learning_rate": 9.9756369975637e-06, + "loss": 0.0642, + "step": 24374 + }, + { + "epoch": 11.331938633193863, + "grad_norm": 3.2777798175811768, + "learning_rate": 9.975635997563601e-06, + "loss": 0.0341, + "step": 24375 + }, + { + "epoch": 11.332403533240353, + "grad_norm": 3.349053382873535, + "learning_rate": 9.9756349975635e-06, + "loss": 0.032, + "step": 24376 + }, + { + "epoch": 11.332868433286844, + "grad_norm": 4.112459182739258, + "learning_rate": 9.9756339975634e-06, + "loss": 0.0484, + "step": 24377 + }, + { + "epoch": 11.333333333333334, + "grad_norm": 6.229797840118408, + "learning_rate": 9.9756329975633e-06, + "loss": 0.0821, + "step": 24378 + }, + { + "epoch": 11.333798233379824, + "grad_norm": 6.230852127075195, + "learning_rate": 9.9756319975632e-06, + "loss": 0.0485, + "step": 24379 + }, + { + "epoch": 11.334263133426314, + "grad_norm": 3.703402280807495, + "learning_rate": 9.975630997563101e-06, + "loss": 0.0218, + "step": 24380 + }, + { + "epoch": 11.334728033472803, + "grad_norm": 5.876269817352295, + "learning_rate": 9.975629997563002e-06, + "loss": 0.0835, + "step": 24381 + }, + { + "epoch": 11.335192933519293, + "grad_norm": 2.3785266876220703, + "learning_rate": 9.9756289975629e-06, + "loss": 0.0146, + "step": 24382 + }, + { + "epoch": 11.335657833565783, + "grad_norm": 7.215556621551514, + "learning_rate": 9.9756279975628e-06, + "loss": 0.1852, + "step": 24383 + }, + { + "epoch": 11.336122733612273, + "grad_norm": 4.636837959289551, + "learning_rate": 9.9756269975627e-06, + "loss": 0.0812, + "step": 24384 + }, + { + "epoch": 11.336587633658763, + "grad_norm": 4.905919075012207, + "learning_rate": 9.975625997562601e-06, + "loss": 0.0812, + "step": 24385 + }, + { + "epoch": 11.337052533705254, + "grad_norm": 2.9767463207244873, + "learning_rate": 9.975624997562502e-06, + "loss": 0.0216, + "step": 24386 + }, + { + "epoch": 11.337517433751744, + "grad_norm": 6.333722114562988, + "learning_rate": 9.9756239975624e-06, + "loss": 0.0719, + "step": 24387 + }, + { + "epoch": 11.337982333798234, + "grad_norm": 0.35086309909820557, + "learning_rate": 9.975622997562301e-06, + "loss": 0.0035, + "step": 24388 + }, + { + "epoch": 11.338447233844724, + "grad_norm": 0.8024616241455078, + "learning_rate": 9.9756219975622e-06, + "loss": 0.0048, + "step": 24389 + }, + { + "epoch": 11.338912133891213, + "grad_norm": 4.175436973571777, + "learning_rate": 9.975620997562101e-06, + "loss": 0.0349, + "step": 24390 + }, + { + "epoch": 11.339377033937703, + "grad_norm": 3.900347948074341, + "learning_rate": 9.975619997562e-06, + "loss": 0.029, + "step": 24391 + }, + { + "epoch": 11.339841933984193, + "grad_norm": 2.5388035774230957, + "learning_rate": 9.9756189975619e-06, + "loss": 0.0452, + "step": 24392 + }, + { + "epoch": 11.340306834030683, + "grad_norm": 10.025997161865234, + "learning_rate": 9.975617997561801e-06, + "loss": 0.2091, + "step": 24393 + }, + { + "epoch": 11.340771734077173, + "grad_norm": 5.031567096710205, + "learning_rate": 9.9756169975617e-06, + "loss": 0.0535, + "step": 24394 + }, + { + "epoch": 11.341236634123664, + "grad_norm": 5.9435133934021, + "learning_rate": 9.9756159975616e-06, + "loss": 0.0843, + "step": 24395 + }, + { + "epoch": 11.341701534170154, + "grad_norm": 4.1344428062438965, + "learning_rate": 9.9756149975615e-06, + "loss": 0.0519, + "step": 24396 + }, + { + "epoch": 11.342166434216644, + "grad_norm": 4.420461177825928, + "learning_rate": 9.9756139975614e-06, + "loss": 0.0893, + "step": 24397 + }, + { + "epoch": 11.342631334263134, + "grad_norm": 4.10740852355957, + "learning_rate": 9.975612997561301e-06, + "loss": 0.0504, + "step": 24398 + }, + { + "epoch": 11.343096234309623, + "grad_norm": 1.8917886018753052, + "learning_rate": 9.975611997561202e-06, + "loss": 0.0196, + "step": 24399 + }, + { + "epoch": 11.343561134356113, + "grad_norm": 9.124737739562988, + "learning_rate": 9.9756109975611e-06, + "loss": 0.1752, + "step": 24400 + }, + { + "epoch": 11.344026034402603, + "grad_norm": 8.556554794311523, + "learning_rate": 9.975609997561001e-06, + "loss": 0.2105, + "step": 24401 + }, + { + "epoch": 11.344490934449093, + "grad_norm": 4.02034330368042, + "learning_rate": 9.9756089975609e-06, + "loss": 0.0407, + "step": 24402 + }, + { + "epoch": 11.344955834495583, + "grad_norm": 5.307644367218018, + "learning_rate": 9.975607997560801e-06, + "loss": 0.0669, + "step": 24403 + }, + { + "epoch": 11.345420734542074, + "grad_norm": 7.224976062774658, + "learning_rate": 9.9756069975607e-06, + "loss": 0.0958, + "step": 24404 + }, + { + "epoch": 11.345885634588564, + "grad_norm": 4.437844753265381, + "learning_rate": 9.9756059975606e-06, + "loss": 0.0465, + "step": 24405 + }, + { + "epoch": 11.346350534635054, + "grad_norm": 8.458097457885742, + "learning_rate": 9.975604997560501e-06, + "loss": 0.1898, + "step": 24406 + }, + { + "epoch": 11.346815434681544, + "grad_norm": 2.9909749031066895, + "learning_rate": 9.9756039975604e-06, + "loss": 0.0207, + "step": 24407 + }, + { + "epoch": 11.347280334728033, + "grad_norm": 3.613539218902588, + "learning_rate": 9.975602997560301e-06, + "loss": 0.0458, + "step": 24408 + }, + { + "epoch": 11.347745234774523, + "grad_norm": 5.771446228027344, + "learning_rate": 9.9756019975602e-06, + "loss": 0.1079, + "step": 24409 + }, + { + "epoch": 11.348210134821013, + "grad_norm": 2.4397811889648438, + "learning_rate": 9.9756009975601e-06, + "loss": 0.009, + "step": 24410 + }, + { + "epoch": 11.348675034867503, + "grad_norm": 1.7140101194381714, + "learning_rate": 9.975599997560001e-06, + "loss": 0.0122, + "step": 24411 + }, + { + "epoch": 11.349139934913994, + "grad_norm": 6.0157060623168945, + "learning_rate": 9.975598997559902e-06, + "loss": 0.0583, + "step": 24412 + }, + { + "epoch": 11.349604834960484, + "grad_norm": 2.9334704875946045, + "learning_rate": 9.9755979975598e-06, + "loss": 0.0367, + "step": 24413 + }, + { + "epoch": 11.350069735006974, + "grad_norm": 1.9810556173324585, + "learning_rate": 9.975596997559701e-06, + "loss": 0.0128, + "step": 24414 + }, + { + "epoch": 11.350534635053464, + "grad_norm": 3.011934518814087, + "learning_rate": 9.9755959975596e-06, + "loss": 0.0422, + "step": 24415 + }, + { + "epoch": 11.350999535099954, + "grad_norm": 1.1395756006240845, + "learning_rate": 9.9755949975595e-06, + "loss": 0.0089, + "step": 24416 + }, + { + "epoch": 11.351464435146443, + "grad_norm": 2.92417573928833, + "learning_rate": 9.9755939975594e-06, + "loss": 0.0303, + "step": 24417 + }, + { + "epoch": 11.351929335192933, + "grad_norm": 4.813420295715332, + "learning_rate": 9.9755929975593e-06, + "loss": 0.0397, + "step": 24418 + }, + { + "epoch": 11.352394235239423, + "grad_norm": 5.833065032958984, + "learning_rate": 9.975591997559201e-06, + "loss": 0.113, + "step": 24419 + }, + { + "epoch": 11.352859135285913, + "grad_norm": 10.25600528717041, + "learning_rate": 9.9755909975591e-06, + "loss": 0.1486, + "step": 24420 + }, + { + "epoch": 11.353324035332404, + "grad_norm": 4.5576558113098145, + "learning_rate": 9.975589997559001e-06, + "loss": 0.0526, + "step": 24421 + }, + { + "epoch": 11.353788935378894, + "grad_norm": 4.5781145095825195, + "learning_rate": 9.9755889975589e-06, + "loss": 0.0394, + "step": 24422 + }, + { + "epoch": 11.354253835425384, + "grad_norm": 3.1916112899780273, + "learning_rate": 9.9755879975588e-06, + "loss": 0.0367, + "step": 24423 + }, + { + "epoch": 11.354718735471874, + "grad_norm": 6.184028148651123, + "learning_rate": 9.975586997558701e-06, + "loss": 0.0361, + "step": 24424 + }, + { + "epoch": 11.355183635518364, + "grad_norm": 5.819921493530273, + "learning_rate": 9.9755859975586e-06, + "loss": 0.1288, + "step": 24425 + }, + { + "epoch": 11.355648535564853, + "grad_norm": 1.831249713897705, + "learning_rate": 9.975584997558501e-06, + "loss": 0.0111, + "step": 24426 + }, + { + "epoch": 11.356113435611343, + "grad_norm": 5.897552490234375, + "learning_rate": 9.9755839975584e-06, + "loss": 0.1612, + "step": 24427 + }, + { + "epoch": 11.356578335657833, + "grad_norm": 3.701242685317993, + "learning_rate": 9.9755829975583e-06, + "loss": 0.0575, + "step": 24428 + }, + { + "epoch": 11.357043235704323, + "grad_norm": 5.07051944732666, + "learning_rate": 9.9755819975582e-06, + "loss": 0.0911, + "step": 24429 + }, + { + "epoch": 11.357508135750814, + "grad_norm": 7.178565502166748, + "learning_rate": 9.9755809975581e-06, + "loss": 0.06, + "step": 24430 + }, + { + "epoch": 11.357973035797304, + "grad_norm": 5.742220878601074, + "learning_rate": 9.975579997558e-06, + "loss": 0.0602, + "step": 24431 + }, + { + "epoch": 11.358437935843794, + "grad_norm": 5.4894561767578125, + "learning_rate": 9.975578997557901e-06, + "loss": 0.0856, + "step": 24432 + }, + { + "epoch": 11.358902835890284, + "grad_norm": 5.288408279418945, + "learning_rate": 9.9755779975578e-06, + "loss": 0.0858, + "step": 24433 + }, + { + "epoch": 11.359367735936774, + "grad_norm": 6.412145614624023, + "learning_rate": 9.975576997557701e-06, + "loss": 0.0734, + "step": 24434 + }, + { + "epoch": 11.359832635983263, + "grad_norm": 5.061426639556885, + "learning_rate": 9.9755759975576e-06, + "loss": 0.0787, + "step": 24435 + }, + { + "epoch": 11.360297536029753, + "grad_norm": 6.359889507293701, + "learning_rate": 9.9755749975575e-06, + "loss": 0.0915, + "step": 24436 + }, + { + "epoch": 11.360762436076243, + "grad_norm": 3.807974100112915, + "learning_rate": 9.975573997557401e-06, + "loss": 0.0392, + "step": 24437 + }, + { + "epoch": 11.361227336122734, + "grad_norm": 4.212883472442627, + "learning_rate": 9.9755729975573e-06, + "loss": 0.0669, + "step": 24438 + }, + { + "epoch": 11.361692236169224, + "grad_norm": 1.611816644668579, + "learning_rate": 9.975571997557201e-06, + "loss": 0.0077, + "step": 24439 + }, + { + "epoch": 11.362157136215714, + "grad_norm": 3.868983745574951, + "learning_rate": 9.9755709975571e-06, + "loss": 0.0446, + "step": 24440 + }, + { + "epoch": 11.362622036262204, + "grad_norm": 8.226221084594727, + "learning_rate": 9.975569997557e-06, + "loss": 0.037, + "step": 24441 + }, + { + "epoch": 11.363086936308694, + "grad_norm": 8.203590393066406, + "learning_rate": 9.9755689975569e-06, + "loss": 0.1507, + "step": 24442 + }, + { + "epoch": 11.363551836355184, + "grad_norm": 3.427748203277588, + "learning_rate": 9.9755679975568e-06, + "loss": 0.0504, + "step": 24443 + }, + { + "epoch": 11.364016736401673, + "grad_norm": 3.4798085689544678, + "learning_rate": 9.9755669975567e-06, + "loss": 0.041, + "step": 24444 + }, + { + "epoch": 11.364481636448163, + "grad_norm": 4.60413932800293, + "learning_rate": 9.975565997556601e-06, + "loss": 0.0818, + "step": 24445 + }, + { + "epoch": 11.364946536494653, + "grad_norm": 3.7227590084075928, + "learning_rate": 9.9755649975565e-06, + "loss": 0.0379, + "step": 24446 + }, + { + "epoch": 11.365411436541144, + "grad_norm": 7.9103498458862305, + "learning_rate": 9.9755639975564e-06, + "loss": 0.0402, + "step": 24447 + }, + { + "epoch": 11.365876336587634, + "grad_norm": 3.1587321758270264, + "learning_rate": 9.9755629975563e-06, + "loss": 0.0348, + "step": 24448 + }, + { + "epoch": 11.366341236634124, + "grad_norm": 5.788496494293213, + "learning_rate": 9.9755619975562e-06, + "loss": 0.0612, + "step": 24449 + }, + { + "epoch": 11.366806136680614, + "grad_norm": 2.591609239578247, + "learning_rate": 9.975560997556101e-06, + "loss": 0.0135, + "step": 24450 + }, + { + "epoch": 11.367271036727104, + "grad_norm": 3.7005434036254883, + "learning_rate": 9.975559997556e-06, + "loss": 0.0317, + "step": 24451 + }, + { + "epoch": 11.367735936773594, + "grad_norm": 6.0561394691467285, + "learning_rate": 9.975558997555901e-06, + "loss": 0.081, + "step": 24452 + }, + { + "epoch": 11.368200836820083, + "grad_norm": 3.4424030780792236, + "learning_rate": 9.9755579975558e-06, + "loss": 0.0631, + "step": 24453 + }, + { + "epoch": 11.368665736866573, + "grad_norm": 4.907836437225342, + "learning_rate": 9.9755569975557e-06, + "loss": 0.0523, + "step": 24454 + }, + { + "epoch": 11.369130636913063, + "grad_norm": 8.736602783203125, + "learning_rate": 9.975555997555601e-06, + "loss": 0.1517, + "step": 24455 + }, + { + "epoch": 11.369595536959554, + "grad_norm": 4.721034526824951, + "learning_rate": 9.9755549975555e-06, + "loss": 0.067, + "step": 24456 + }, + { + "epoch": 11.370060437006044, + "grad_norm": 5.512521266937256, + "learning_rate": 9.975553997555401e-06, + "loss": 0.0765, + "step": 24457 + }, + { + "epoch": 11.370525337052534, + "grad_norm": 2.6805243492126465, + "learning_rate": 9.9755529975553e-06, + "loss": 0.0243, + "step": 24458 + }, + { + "epoch": 11.370990237099024, + "grad_norm": 2.3570351600646973, + "learning_rate": 9.9755519975552e-06, + "loss": 0.0323, + "step": 24459 + }, + { + "epoch": 11.371455137145514, + "grad_norm": 7.578428268432617, + "learning_rate": 9.9755509975551e-06, + "loss": 0.03, + "step": 24460 + }, + { + "epoch": 11.371920037192004, + "grad_norm": 7.0638427734375, + "learning_rate": 9.975549997555e-06, + "loss": 0.119, + "step": 24461 + }, + { + "epoch": 11.372384937238493, + "grad_norm": 4.357515335083008, + "learning_rate": 9.9755489975549e-06, + "loss": 0.0551, + "step": 24462 + }, + { + "epoch": 11.372849837284983, + "grad_norm": 13.144501686096191, + "learning_rate": 9.975547997554801e-06, + "loss": 0.1138, + "step": 24463 + }, + { + "epoch": 11.373314737331473, + "grad_norm": 3.5235636234283447, + "learning_rate": 9.9755469975547e-06, + "loss": 0.0365, + "step": 24464 + }, + { + "epoch": 11.373779637377964, + "grad_norm": 4.055490016937256, + "learning_rate": 9.975545997554601e-06, + "loss": 0.072, + "step": 24465 + }, + { + "epoch": 11.374244537424454, + "grad_norm": 4.522370338439941, + "learning_rate": 9.9755449975545e-06, + "loss": 0.0743, + "step": 24466 + }, + { + "epoch": 11.374709437470944, + "grad_norm": 1.3621547222137451, + "learning_rate": 9.9755439975544e-06, + "loss": 0.01, + "step": 24467 + }, + { + "epoch": 11.375174337517434, + "grad_norm": 4.531620979309082, + "learning_rate": 9.975542997554301e-06, + "loss": 0.0824, + "step": 24468 + }, + { + "epoch": 11.375639237563924, + "grad_norm": 2.461326837539673, + "learning_rate": 9.9755419975542e-06, + "loss": 0.0276, + "step": 24469 + }, + { + "epoch": 11.376104137610414, + "grad_norm": 4.29535436630249, + "learning_rate": 9.975540997554101e-06, + "loss": 0.0451, + "step": 24470 + }, + { + "epoch": 11.376569037656903, + "grad_norm": 0.935495913028717, + "learning_rate": 9.975539997554e-06, + "loss": 0.0088, + "step": 24471 + }, + { + "epoch": 11.377033937703393, + "grad_norm": 2.629563808441162, + "learning_rate": 9.9755389975539e-06, + "loss": 0.0214, + "step": 24472 + }, + { + "epoch": 11.377498837749883, + "grad_norm": 5.0838398933410645, + "learning_rate": 9.9755379975538e-06, + "loss": 0.0699, + "step": 24473 + }, + { + "epoch": 11.377963737796374, + "grad_norm": 4.713069438934326, + "learning_rate": 9.9755369975537e-06, + "loss": 0.071, + "step": 24474 + }, + { + "epoch": 11.378428637842864, + "grad_norm": 5.883362770080566, + "learning_rate": 9.975535997553601e-06, + "loss": 0.1206, + "step": 24475 + }, + { + "epoch": 11.378893537889354, + "grad_norm": 1.0751444101333618, + "learning_rate": 9.975534997553502e-06, + "loss": 0.0097, + "step": 24476 + }, + { + "epoch": 11.379358437935844, + "grad_norm": 3.2478909492492676, + "learning_rate": 9.9755339975534e-06, + "loss": 0.0244, + "step": 24477 + }, + { + "epoch": 11.379823337982334, + "grad_norm": 1.5554466247558594, + "learning_rate": 9.975532997553301e-06, + "loss": 0.012, + "step": 24478 + }, + { + "epoch": 11.380288238028823, + "grad_norm": 4.403149604797363, + "learning_rate": 9.9755319975532e-06, + "loss": 0.0468, + "step": 24479 + }, + { + "epoch": 11.380753138075313, + "grad_norm": 7.302948951721191, + "learning_rate": 9.9755309975531e-06, + "loss": 0.0524, + "step": 24480 + }, + { + "epoch": 11.381218038121803, + "grad_norm": 2.9728643894195557, + "learning_rate": 9.975529997553001e-06, + "loss": 0.03, + "step": 24481 + }, + { + "epoch": 11.381682938168293, + "grad_norm": 5.375383377075195, + "learning_rate": 9.9755289975529e-06, + "loss": 0.0491, + "step": 24482 + }, + { + "epoch": 11.382147838214784, + "grad_norm": 2.5144307613372803, + "learning_rate": 9.975527997552801e-06, + "loss": 0.0194, + "step": 24483 + }, + { + "epoch": 11.382612738261274, + "grad_norm": 6.986944675445557, + "learning_rate": 9.9755269975527e-06, + "loss": 0.0807, + "step": 24484 + }, + { + "epoch": 11.383077638307764, + "grad_norm": 3.062453269958496, + "learning_rate": 9.9755259975526e-06, + "loss": 0.0326, + "step": 24485 + }, + { + "epoch": 11.383542538354254, + "grad_norm": 7.213592052459717, + "learning_rate": 9.9755249975525e-06, + "loss": 0.1124, + "step": 24486 + }, + { + "epoch": 11.384007438400744, + "grad_norm": 9.519584655761719, + "learning_rate": 9.9755239975524e-06, + "loss": 0.0303, + "step": 24487 + }, + { + "epoch": 11.384472338447233, + "grad_norm": 2.9019691944122314, + "learning_rate": 9.975522997552301e-06, + "loss": 0.0282, + "step": 24488 + }, + { + "epoch": 11.384937238493723, + "grad_norm": 3.8956401348114014, + "learning_rate": 9.975521997552202e-06, + "loss": 0.0257, + "step": 24489 + }, + { + "epoch": 11.385402138540213, + "grad_norm": 16.521318435668945, + "learning_rate": 9.9755209975521e-06, + "loss": 0.0975, + "step": 24490 + }, + { + "epoch": 11.385867038586705, + "grad_norm": 5.489243984222412, + "learning_rate": 9.975519997552e-06, + "loss": 0.0787, + "step": 24491 + }, + { + "epoch": 11.386331938633194, + "grad_norm": 2.7379720211029053, + "learning_rate": 9.9755189975519e-06, + "loss": 0.0223, + "step": 24492 + }, + { + "epoch": 11.386796838679684, + "grad_norm": 7.41345739364624, + "learning_rate": 9.975517997551801e-06, + "loss": 0.0892, + "step": 24493 + }, + { + "epoch": 11.387261738726174, + "grad_norm": 7.777094841003418, + "learning_rate": 9.975516997551702e-06, + "loss": 0.1209, + "step": 24494 + }, + { + "epoch": 11.387726638772664, + "grad_norm": 2.472867965698242, + "learning_rate": 9.9755159975516e-06, + "loss": 0.0184, + "step": 24495 + }, + { + "epoch": 11.388191538819154, + "grad_norm": 4.5797600746154785, + "learning_rate": 9.975514997551501e-06, + "loss": 0.0816, + "step": 24496 + }, + { + "epoch": 11.388656438865643, + "grad_norm": 3.035377025604248, + "learning_rate": 9.9755139975514e-06, + "loss": 0.0223, + "step": 24497 + }, + { + "epoch": 11.389121338912133, + "grad_norm": 3.696458101272583, + "learning_rate": 9.9755129975513e-06, + "loss": 0.0268, + "step": 24498 + }, + { + "epoch": 11.389586238958623, + "grad_norm": 1.3039727210998535, + "learning_rate": 9.9755119975512e-06, + "loss": 0.0081, + "step": 24499 + }, + { + "epoch": 11.390051139005115, + "grad_norm": 4.027641296386719, + "learning_rate": 9.9755109975511e-06, + "loss": 0.0478, + "step": 24500 + }, + { + "epoch": 11.390516039051604, + "grad_norm": 2.7371129989624023, + "learning_rate": 9.975509997551001e-06, + "loss": 0.0372, + "step": 24501 + }, + { + "epoch": 11.390980939098094, + "grad_norm": 3.1916184425354004, + "learning_rate": 9.9755089975509e-06, + "loss": 0.0235, + "step": 24502 + }, + { + "epoch": 11.391445839144584, + "grad_norm": 4.458575248718262, + "learning_rate": 9.9755079975508e-06, + "loss": 0.0544, + "step": 24503 + }, + { + "epoch": 11.391910739191074, + "grad_norm": 4.969709396362305, + "learning_rate": 9.9755069975507e-06, + "loss": 0.0473, + "step": 24504 + }, + { + "epoch": 11.392375639237564, + "grad_norm": 4.934238433837891, + "learning_rate": 9.9755059975506e-06, + "loss": 0.0174, + "step": 24505 + }, + { + "epoch": 11.392840539284053, + "grad_norm": 6.336254119873047, + "learning_rate": 9.975504997550501e-06, + "loss": 0.0558, + "step": 24506 + }, + { + "epoch": 11.393305439330543, + "grad_norm": 3.3744122982025146, + "learning_rate": 9.975503997550402e-06, + "loss": 0.0467, + "step": 24507 + }, + { + "epoch": 11.393770339377033, + "grad_norm": 1.6997548341751099, + "learning_rate": 9.9755029975503e-06, + "loss": 0.0131, + "step": 24508 + }, + { + "epoch": 11.394235239423525, + "grad_norm": 3.4875667095184326, + "learning_rate": 9.975501997550201e-06, + "loss": 0.038, + "step": 24509 + }, + { + "epoch": 11.394700139470014, + "grad_norm": 4.673761367797852, + "learning_rate": 9.9755009975501e-06, + "loss": 0.0233, + "step": 24510 + }, + { + "epoch": 11.395165039516504, + "grad_norm": 3.026916265487671, + "learning_rate": 9.975499997550001e-06, + "loss": 0.0188, + "step": 24511 + }, + { + "epoch": 11.395629939562994, + "grad_norm": 4.894824504852295, + "learning_rate": 9.9754989975499e-06, + "loss": 0.1068, + "step": 24512 + }, + { + "epoch": 11.396094839609484, + "grad_norm": 2.686277151107788, + "learning_rate": 9.9754979975498e-06, + "loss": 0.026, + "step": 24513 + }, + { + "epoch": 11.396559739655974, + "grad_norm": 1.0381985902786255, + "learning_rate": 9.975496997549701e-06, + "loss": 0.0108, + "step": 24514 + }, + { + "epoch": 11.397024639702463, + "grad_norm": 4.6031951904296875, + "learning_rate": 9.9754959975496e-06, + "loss": 0.0209, + "step": 24515 + }, + { + "epoch": 11.397489539748953, + "grad_norm": 4.896564960479736, + "learning_rate": 9.9754949975495e-06, + "loss": 0.0592, + "step": 24516 + }, + { + "epoch": 11.397954439795445, + "grad_norm": 6.2090559005737305, + "learning_rate": 9.9754939975494e-06, + "loss": 0.1006, + "step": 24517 + }, + { + "epoch": 11.398419339841935, + "grad_norm": 4.665956020355225, + "learning_rate": 9.9754929975493e-06, + "loss": 0.0573, + "step": 24518 + }, + { + "epoch": 11.398884239888424, + "grad_norm": 5.382876873016357, + "learning_rate": 9.975491997549201e-06, + "loss": 0.0417, + "step": 24519 + }, + { + "epoch": 11.399349139934914, + "grad_norm": 6.568290710449219, + "learning_rate": 9.975490997549102e-06, + "loss": 0.1135, + "step": 24520 + }, + { + "epoch": 11.399814039981404, + "grad_norm": 3.4371604919433594, + "learning_rate": 9.975489997549e-06, + "loss": 0.0359, + "step": 24521 + }, + { + "epoch": 11.400278940027894, + "grad_norm": 4.98538875579834, + "learning_rate": 9.975488997548901e-06, + "loss": 0.0582, + "step": 24522 + }, + { + "epoch": 11.400743840074384, + "grad_norm": 7.096787452697754, + "learning_rate": 9.9754879975488e-06, + "loss": 0.0716, + "step": 24523 + }, + { + "epoch": 11.401208740120873, + "grad_norm": 4.132361888885498, + "learning_rate": 9.9754869975487e-06, + "loss": 0.0405, + "step": 24524 + }, + { + "epoch": 11.401673640167363, + "grad_norm": 2.2390341758728027, + "learning_rate": 9.9754859975486e-06, + "loss": 0.0299, + "step": 24525 + }, + { + "epoch": 11.402138540213855, + "grad_norm": 3.122739553451538, + "learning_rate": 9.9754849975485e-06, + "loss": 0.0411, + "step": 24526 + }, + { + "epoch": 11.402603440260345, + "grad_norm": 1.7858669757843018, + "learning_rate": 9.975483997548401e-06, + "loss": 0.0059, + "step": 24527 + }, + { + "epoch": 11.403068340306834, + "grad_norm": 3.20501971244812, + "learning_rate": 9.9754829975483e-06, + "loss": 0.0494, + "step": 24528 + }, + { + "epoch": 11.403533240353324, + "grad_norm": 1.5639127492904663, + "learning_rate": 9.9754819975482e-06, + "loss": 0.0157, + "step": 24529 + }, + { + "epoch": 11.403998140399814, + "grad_norm": 2.6121773719787598, + "learning_rate": 9.9754809975481e-06, + "loss": 0.0204, + "step": 24530 + }, + { + "epoch": 11.404463040446304, + "grad_norm": 7.0881733894348145, + "learning_rate": 9.975479997548e-06, + "loss": 0.0759, + "step": 24531 + }, + { + "epoch": 11.404927940492794, + "grad_norm": 4.982814311981201, + "learning_rate": 9.975478997547901e-06, + "loss": 0.0814, + "step": 24532 + }, + { + "epoch": 11.405392840539283, + "grad_norm": 6.185226917266846, + "learning_rate": 9.975477997547802e-06, + "loss": 0.0725, + "step": 24533 + }, + { + "epoch": 11.405857740585773, + "grad_norm": 1.5283880233764648, + "learning_rate": 9.9754769975477e-06, + "loss": 0.0104, + "step": 24534 + }, + { + "epoch": 11.406322640632265, + "grad_norm": 3.152151107788086, + "learning_rate": 9.9754759975476e-06, + "loss": 0.0304, + "step": 24535 + }, + { + "epoch": 11.406787540678755, + "grad_norm": 4.7166547775268555, + "learning_rate": 9.9754749975475e-06, + "loss": 0.0811, + "step": 24536 + }, + { + "epoch": 11.407252440725244, + "grad_norm": 5.481424808502197, + "learning_rate": 9.975473997547401e-06, + "loss": 0.0374, + "step": 24537 + }, + { + "epoch": 11.407717340771734, + "grad_norm": 5.076603412628174, + "learning_rate": 9.9754729975473e-06, + "loss": 0.0588, + "step": 24538 + }, + { + "epoch": 11.408182240818224, + "grad_norm": 6.332670211791992, + "learning_rate": 9.9754719975472e-06, + "loss": 0.0825, + "step": 24539 + }, + { + "epoch": 11.408647140864714, + "grad_norm": 4.167099475860596, + "learning_rate": 9.975470997547101e-06, + "loss": 0.1119, + "step": 24540 + }, + { + "epoch": 11.409112040911204, + "grad_norm": 2.104279041290283, + "learning_rate": 9.975469997547e-06, + "loss": 0.0167, + "step": 24541 + }, + { + "epoch": 11.409576940957693, + "grad_norm": 2.623370885848999, + "learning_rate": 9.975468997546901e-06, + "loss": 0.018, + "step": 24542 + }, + { + "epoch": 11.410041841004183, + "grad_norm": 4.758053779602051, + "learning_rate": 9.9754679975468e-06, + "loss": 0.0265, + "step": 24543 + }, + { + "epoch": 11.410506741050675, + "grad_norm": 2.0517594814300537, + "learning_rate": 9.9754669975467e-06, + "loss": 0.0167, + "step": 24544 + }, + { + "epoch": 11.410971641097165, + "grad_norm": 5.52709436416626, + "learning_rate": 9.975465997546601e-06, + "loss": 0.0557, + "step": 24545 + }, + { + "epoch": 11.411436541143654, + "grad_norm": 2.4259397983551025, + "learning_rate": 9.9754649975465e-06, + "loss": 0.0184, + "step": 24546 + }, + { + "epoch": 11.411901441190144, + "grad_norm": 5.74703311920166, + "learning_rate": 9.9754639975464e-06, + "loss": 0.0346, + "step": 24547 + }, + { + "epoch": 11.412366341236634, + "grad_norm": 6.431718826293945, + "learning_rate": 9.9754629975463e-06, + "loss": 0.1125, + "step": 24548 + }, + { + "epoch": 11.412831241283124, + "grad_norm": 3.4115560054779053, + "learning_rate": 9.9754619975462e-06, + "loss": 0.0175, + "step": 24549 + }, + { + "epoch": 11.413296141329614, + "grad_norm": 5.192112445831299, + "learning_rate": 9.975460997546101e-06, + "loss": 0.0356, + "step": 24550 + }, + { + "epoch": 11.413761041376103, + "grad_norm": 5.939082622528076, + "learning_rate": 9.975459997546002e-06, + "loss": 0.0911, + "step": 24551 + }, + { + "epoch": 11.414225941422593, + "grad_norm": 5.572096347808838, + "learning_rate": 9.9754589975459e-06, + "loss": 0.0509, + "step": 24552 + }, + { + "epoch": 11.414690841469085, + "grad_norm": 3.5884571075439453, + "learning_rate": 9.975457997545801e-06, + "loss": 0.061, + "step": 24553 + }, + { + "epoch": 11.415155741515575, + "grad_norm": 5.4170684814453125, + "learning_rate": 9.9754569975457e-06, + "loss": 0.0636, + "step": 24554 + }, + { + "epoch": 11.415620641562064, + "grad_norm": 6.960007667541504, + "learning_rate": 9.975455997545601e-06, + "loss": 0.0944, + "step": 24555 + }, + { + "epoch": 11.416085541608554, + "grad_norm": 2.9830896854400635, + "learning_rate": 9.9754549975455e-06, + "loss": 0.0318, + "step": 24556 + }, + { + "epoch": 11.416550441655044, + "grad_norm": 2.9528486728668213, + "learning_rate": 9.9754539975454e-06, + "loss": 0.0391, + "step": 24557 + }, + { + "epoch": 11.417015341701534, + "grad_norm": 0.9119235873222351, + "learning_rate": 9.975452997545301e-06, + "loss": 0.0061, + "step": 24558 + }, + { + "epoch": 11.417480241748024, + "grad_norm": 2.6900696754455566, + "learning_rate": 9.9754519975452e-06, + "loss": 0.0419, + "step": 24559 + }, + { + "epoch": 11.417945141794513, + "grad_norm": 4.228166103363037, + "learning_rate": 9.975450997545101e-06, + "loss": 0.0336, + "step": 24560 + }, + { + "epoch": 11.418410041841003, + "grad_norm": 2.61549973487854, + "learning_rate": 9.975449997545e-06, + "loss": 0.0275, + "step": 24561 + }, + { + "epoch": 11.418874941887495, + "grad_norm": 6.534754753112793, + "learning_rate": 9.9754489975449e-06, + "loss": 0.1589, + "step": 24562 + }, + { + "epoch": 11.419339841933985, + "grad_norm": 4.967537879943848, + "learning_rate": 9.975447997544801e-06, + "loss": 0.0955, + "step": 24563 + }, + { + "epoch": 11.419804741980474, + "grad_norm": 2.5673933029174805, + "learning_rate": 9.975446997544702e-06, + "loss": 0.0279, + "step": 24564 + }, + { + "epoch": 11.420269642026964, + "grad_norm": 4.613594055175781, + "learning_rate": 9.9754459975446e-06, + "loss": 0.0626, + "step": 24565 + }, + { + "epoch": 11.420734542073454, + "grad_norm": 3.455312728881836, + "learning_rate": 9.9754449975445e-06, + "loss": 0.0166, + "step": 24566 + }, + { + "epoch": 11.421199442119944, + "grad_norm": 2.5452327728271484, + "learning_rate": 9.9754439975444e-06, + "loss": 0.0237, + "step": 24567 + }, + { + "epoch": 11.421664342166434, + "grad_norm": 4.286082744598389, + "learning_rate": 9.9754429975443e-06, + "loss": 0.0381, + "step": 24568 + }, + { + "epoch": 11.422129242212923, + "grad_norm": 10.327566146850586, + "learning_rate": 9.9754419975442e-06, + "loss": 0.093, + "step": 24569 + }, + { + "epoch": 11.422594142259415, + "grad_norm": 4.331243991851807, + "learning_rate": 9.9754409975441e-06, + "loss": 0.0296, + "step": 24570 + }, + { + "epoch": 11.423059042305905, + "grad_norm": 4.575425148010254, + "learning_rate": 9.975439997544001e-06, + "loss": 0.0368, + "step": 24571 + }, + { + "epoch": 11.423523942352395, + "grad_norm": 6.112094402313232, + "learning_rate": 9.9754389975439e-06, + "loss": 0.0952, + "step": 24572 + }, + { + "epoch": 11.423988842398884, + "grad_norm": 1.2729462385177612, + "learning_rate": 9.975437997543801e-06, + "loss": 0.0127, + "step": 24573 + }, + { + "epoch": 11.424453742445374, + "grad_norm": 5.862591743469238, + "learning_rate": 9.9754369975437e-06, + "loss": 0.1089, + "step": 24574 + }, + { + "epoch": 11.424918642491864, + "grad_norm": 2.335484743118286, + "learning_rate": 9.9754359975436e-06, + "loss": 0.0167, + "step": 24575 + }, + { + "epoch": 11.425383542538354, + "grad_norm": 2.3151144981384277, + "learning_rate": 9.975434997543501e-06, + "loss": 0.014, + "step": 24576 + }, + { + "epoch": 11.425848442584844, + "grad_norm": 1.9387073516845703, + "learning_rate": 9.9754339975434e-06, + "loss": 0.0136, + "step": 24577 + }, + { + "epoch": 11.426313342631333, + "grad_norm": 4.256075859069824, + "learning_rate": 9.975432997543301e-06, + "loss": 0.0496, + "step": 24578 + }, + { + "epoch": 11.426778242677825, + "grad_norm": 8.140572547912598, + "learning_rate": 9.9754319975432e-06, + "loss": 0.2061, + "step": 24579 + }, + { + "epoch": 11.427243142724315, + "grad_norm": 6.691686153411865, + "learning_rate": 9.9754309975431e-06, + "loss": 0.1148, + "step": 24580 + }, + { + "epoch": 11.427708042770805, + "grad_norm": 0.5873891711235046, + "learning_rate": 9.975429997543e-06, + "loss": 0.0044, + "step": 24581 + }, + { + "epoch": 11.428172942817294, + "grad_norm": 4.9658732414245605, + "learning_rate": 9.9754289975429e-06, + "loss": 0.1136, + "step": 24582 + }, + { + "epoch": 11.428637842863784, + "grad_norm": 3.431973934173584, + "learning_rate": 9.9754279975428e-06, + "loss": 0.0248, + "step": 24583 + }, + { + "epoch": 11.429102742910274, + "grad_norm": 3.8365726470947266, + "learning_rate": 9.975426997542701e-06, + "loss": 0.0372, + "step": 24584 + }, + { + "epoch": 11.429567642956764, + "grad_norm": 6.875255107879639, + "learning_rate": 9.9754259975426e-06, + "loss": 0.0813, + "step": 24585 + }, + { + "epoch": 11.430032543003254, + "grad_norm": 5.353085041046143, + "learning_rate": 9.975424997542501e-06, + "loss": 0.0737, + "step": 24586 + }, + { + "epoch": 11.430497443049743, + "grad_norm": 9.588621139526367, + "learning_rate": 9.9754239975424e-06, + "loss": 0.1028, + "step": 24587 + }, + { + "epoch": 11.430962343096235, + "grad_norm": 3.8511252403259277, + "learning_rate": 9.9754229975423e-06, + "loss": 0.0522, + "step": 24588 + }, + { + "epoch": 11.431427243142725, + "grad_norm": 2.1874632835388184, + "learning_rate": 9.975421997542201e-06, + "loss": 0.0097, + "step": 24589 + }, + { + "epoch": 11.431892143189215, + "grad_norm": 5.087542533874512, + "learning_rate": 9.9754209975421e-06, + "loss": 0.1062, + "step": 24590 + }, + { + "epoch": 11.432357043235704, + "grad_norm": 2.8871593475341797, + "learning_rate": 9.975419997542001e-06, + "loss": 0.027, + "step": 24591 + }, + { + "epoch": 11.432821943282194, + "grad_norm": 6.641626358032227, + "learning_rate": 9.9754189975419e-06, + "loss": 0.1528, + "step": 24592 + }, + { + "epoch": 11.433286843328684, + "grad_norm": 3.283829927444458, + "learning_rate": 9.9754179975418e-06, + "loss": 0.0502, + "step": 24593 + }, + { + "epoch": 11.433751743375174, + "grad_norm": 5.223791599273682, + "learning_rate": 9.9754169975417e-06, + "loss": 0.1135, + "step": 24594 + }, + { + "epoch": 11.434216643421664, + "grad_norm": 7.9978790283203125, + "learning_rate": 9.9754159975416e-06, + "loss": 0.0633, + "step": 24595 + }, + { + "epoch": 11.434681543468155, + "grad_norm": 3.4687881469726562, + "learning_rate": 9.975414997541501e-06, + "loss": 0.0338, + "step": 24596 + }, + { + "epoch": 11.435146443514645, + "grad_norm": 2.6654815673828125, + "learning_rate": 9.975413997541402e-06, + "loss": 0.0365, + "step": 24597 + }, + { + "epoch": 11.435611343561135, + "grad_norm": 4.331491470336914, + "learning_rate": 9.9754129975413e-06, + "loss": 0.0371, + "step": 24598 + }, + { + "epoch": 11.436076243607625, + "grad_norm": 2.148699998855591, + "learning_rate": 9.9754119975412e-06, + "loss": 0.0205, + "step": 24599 + }, + { + "epoch": 11.436541143654114, + "grad_norm": 4.52946662902832, + "learning_rate": 9.9754109975411e-06, + "loss": 0.0384, + "step": 24600 + }, + { + "epoch": 11.437006043700604, + "grad_norm": 2.7353014945983887, + "learning_rate": 9.975409997541e-06, + "loss": 0.018, + "step": 24601 + }, + { + "epoch": 11.437470943747094, + "grad_norm": 3.183131694793701, + "learning_rate": 9.975408997540901e-06, + "loss": 0.0288, + "step": 24602 + }, + { + "epoch": 11.437935843793584, + "grad_norm": 2.894655704498291, + "learning_rate": 9.9754079975408e-06, + "loss": 0.0374, + "step": 24603 + }, + { + "epoch": 11.438400743840074, + "grad_norm": 2.854320526123047, + "learning_rate": 9.975406997540701e-06, + "loss": 0.0229, + "step": 24604 + }, + { + "epoch": 11.438865643886565, + "grad_norm": 4.38140344619751, + "learning_rate": 9.9754059975406e-06, + "loss": 0.0905, + "step": 24605 + }, + { + "epoch": 11.439330543933055, + "grad_norm": 8.457961082458496, + "learning_rate": 9.9754049975405e-06, + "loss": 0.1521, + "step": 24606 + }, + { + "epoch": 11.439795443979545, + "grad_norm": 5.681241989135742, + "learning_rate": 9.9754039975404e-06, + "loss": 0.0551, + "step": 24607 + }, + { + "epoch": 11.440260344026035, + "grad_norm": 4.640463829040527, + "learning_rate": 9.9754029975403e-06, + "loss": 0.0633, + "step": 24608 + }, + { + "epoch": 11.440725244072524, + "grad_norm": 4.041658401489258, + "learning_rate": 9.975401997540201e-06, + "loss": 0.074, + "step": 24609 + }, + { + "epoch": 11.441190144119014, + "grad_norm": 2.7841174602508545, + "learning_rate": 9.9754009975401e-06, + "loss": 0.0229, + "step": 24610 + }, + { + "epoch": 11.441655044165504, + "grad_norm": 6.753603458404541, + "learning_rate": 9.97539999754e-06, + "loss": 0.1075, + "step": 24611 + }, + { + "epoch": 11.442119944211994, + "grad_norm": 9.450061798095703, + "learning_rate": 9.9753989975399e-06, + "loss": 0.1243, + "step": 24612 + }, + { + "epoch": 11.442584844258484, + "grad_norm": 3.8373851776123047, + "learning_rate": 9.9753979975398e-06, + "loss": 0.0392, + "step": 24613 + }, + { + "epoch": 11.443049744304975, + "grad_norm": 8.194404602050781, + "learning_rate": 9.975396997539701e-06, + "loss": 0.1351, + "step": 24614 + }, + { + "epoch": 11.443514644351465, + "grad_norm": 6.078341960906982, + "learning_rate": 9.975395997539602e-06, + "loss": 0.0913, + "step": 24615 + }, + { + "epoch": 11.443979544397955, + "grad_norm": 4.079042434692383, + "learning_rate": 9.9753949975395e-06, + "loss": 0.0846, + "step": 24616 + }, + { + "epoch": 11.444444444444445, + "grad_norm": 2.4731271266937256, + "learning_rate": 9.975393997539401e-06, + "loss": 0.0266, + "step": 24617 + }, + { + "epoch": 11.444909344490934, + "grad_norm": 5.016382694244385, + "learning_rate": 9.9753929975393e-06, + "loss": 0.1045, + "step": 24618 + }, + { + "epoch": 11.445374244537424, + "grad_norm": 5.78524112701416, + "learning_rate": 9.9753919975392e-06, + "loss": 0.1172, + "step": 24619 + }, + { + "epoch": 11.445839144583914, + "grad_norm": 5.641773223876953, + "learning_rate": 9.975390997539101e-06, + "loss": 0.0943, + "step": 24620 + }, + { + "epoch": 11.446304044630404, + "grad_norm": 2.0175118446350098, + "learning_rate": 9.975389997539e-06, + "loss": 0.0192, + "step": 24621 + }, + { + "epoch": 11.446768944676894, + "grad_norm": 0.7414832711219788, + "learning_rate": 9.975388997538901e-06, + "loss": 0.0059, + "step": 24622 + }, + { + "epoch": 11.447233844723385, + "grad_norm": 3.4584763050079346, + "learning_rate": 9.9753879975388e-06, + "loss": 0.0404, + "step": 24623 + }, + { + "epoch": 11.447698744769875, + "grad_norm": 5.8642988204956055, + "learning_rate": 9.9753869975387e-06, + "loss": 0.0561, + "step": 24624 + }, + { + "epoch": 11.448163644816365, + "grad_norm": 4.678701877593994, + "learning_rate": 9.9753859975386e-06, + "loss": 0.0876, + "step": 24625 + }, + { + "epoch": 11.448628544862855, + "grad_norm": 6.050344467163086, + "learning_rate": 9.9753849975385e-06, + "loss": 0.0807, + "step": 24626 + }, + { + "epoch": 11.449093444909344, + "grad_norm": 5.720910549163818, + "learning_rate": 9.975383997538401e-06, + "loss": 0.0707, + "step": 24627 + }, + { + "epoch": 11.449558344955834, + "grad_norm": 3.231091022491455, + "learning_rate": 9.975382997538302e-06, + "loss": 0.0195, + "step": 24628 + }, + { + "epoch": 11.450023245002324, + "grad_norm": 3.2952983379364014, + "learning_rate": 9.9753819975382e-06, + "loss": 0.0181, + "step": 24629 + }, + { + "epoch": 11.450488145048814, + "grad_norm": 12.389906883239746, + "learning_rate": 9.975380997538101e-06, + "loss": 0.2895, + "step": 24630 + }, + { + "epoch": 11.450953045095304, + "grad_norm": 4.7528862953186035, + "learning_rate": 9.975379997538e-06, + "loss": 0.0833, + "step": 24631 + }, + { + "epoch": 11.451417945141795, + "grad_norm": 4.945478916168213, + "learning_rate": 9.9753789975379e-06, + "loss": 0.0587, + "step": 24632 + }, + { + "epoch": 11.451882845188285, + "grad_norm": 2.6229641437530518, + "learning_rate": 9.975377997537801e-06, + "loss": 0.0213, + "step": 24633 + }, + { + "epoch": 11.452347745234775, + "grad_norm": 3.778277635574341, + "learning_rate": 9.9753769975377e-06, + "loss": 0.0814, + "step": 24634 + }, + { + "epoch": 11.452812645281265, + "grad_norm": 5.164357662200928, + "learning_rate": 9.975375997537601e-06, + "loss": 0.0948, + "step": 24635 + }, + { + "epoch": 11.453277545327754, + "grad_norm": 4.929845333099365, + "learning_rate": 9.9753749975375e-06, + "loss": 0.109, + "step": 24636 + }, + { + "epoch": 11.453742445374244, + "grad_norm": 0.9006219506263733, + "learning_rate": 9.9753739975374e-06, + "loss": 0.0081, + "step": 24637 + }, + { + "epoch": 11.454207345420734, + "grad_norm": 3.4852702617645264, + "learning_rate": 9.9753729975373e-06, + "loss": 0.0431, + "step": 24638 + }, + { + "epoch": 11.454672245467224, + "grad_norm": 6.481417179107666, + "learning_rate": 9.9753719975372e-06, + "loss": 0.0738, + "step": 24639 + }, + { + "epoch": 11.455137145513714, + "grad_norm": 5.846408367156982, + "learning_rate": 9.975370997537101e-06, + "loss": 0.1011, + "step": 24640 + }, + { + "epoch": 11.455602045560205, + "grad_norm": 3.6829473972320557, + "learning_rate": 9.975369997537002e-06, + "loss": 0.0327, + "step": 24641 + }, + { + "epoch": 11.456066945606695, + "grad_norm": 5.206666469573975, + "learning_rate": 9.9753689975369e-06, + "loss": 0.0518, + "step": 24642 + }, + { + "epoch": 11.456531845653185, + "grad_norm": 2.8471314907073975, + "learning_rate": 9.9753679975368e-06, + "loss": 0.028, + "step": 24643 + }, + { + "epoch": 11.456996745699675, + "grad_norm": 1.6788690090179443, + "learning_rate": 9.9753669975367e-06, + "loss": 0.0109, + "step": 24644 + }, + { + "epoch": 11.457461645746164, + "grad_norm": 3.5055654048919678, + "learning_rate": 9.975365997536601e-06, + "loss": 0.0547, + "step": 24645 + }, + { + "epoch": 11.457926545792654, + "grad_norm": 4.5845770835876465, + "learning_rate": 9.975364997536502e-06, + "loss": 0.0325, + "step": 24646 + }, + { + "epoch": 11.458391445839144, + "grad_norm": 3.781236410140991, + "learning_rate": 9.9753639975364e-06, + "loss": 0.0462, + "step": 24647 + }, + { + "epoch": 11.458856345885634, + "grad_norm": 8.76888656616211, + "learning_rate": 9.975362997536301e-06, + "loss": 0.0863, + "step": 24648 + }, + { + "epoch": 11.459321245932125, + "grad_norm": 9.360934257507324, + "learning_rate": 9.9753619975362e-06, + "loss": 0.1435, + "step": 24649 + }, + { + "epoch": 11.459786145978615, + "grad_norm": 5.104965686798096, + "learning_rate": 9.9753609975361e-06, + "loss": 0.0572, + "step": 24650 + }, + { + "epoch": 11.460251046025105, + "grad_norm": 3.5997674465179443, + "learning_rate": 9.975359997536e-06, + "loss": 0.0577, + "step": 24651 + }, + { + "epoch": 11.460715946071595, + "grad_norm": 6.585890769958496, + "learning_rate": 9.9753589975359e-06, + "loss": 0.0751, + "step": 24652 + }, + { + "epoch": 11.461180846118085, + "grad_norm": 2.304356098175049, + "learning_rate": 9.975357997535801e-06, + "loss": 0.0239, + "step": 24653 + }, + { + "epoch": 11.461645746164574, + "grad_norm": 1.7915740013122559, + "learning_rate": 9.9753569975357e-06, + "loss": 0.0118, + "step": 24654 + }, + { + "epoch": 11.462110646211064, + "grad_norm": 5.0803093910217285, + "learning_rate": 9.9753559975356e-06, + "loss": 0.0494, + "step": 24655 + }, + { + "epoch": 11.462575546257554, + "grad_norm": 5.886469841003418, + "learning_rate": 9.9753549975355e-06, + "loss": 0.1144, + "step": 24656 + }, + { + "epoch": 11.463040446304044, + "grad_norm": 7.111903190612793, + "learning_rate": 9.9753539975354e-06, + "loss": 0.0919, + "step": 24657 + }, + { + "epoch": 11.463505346350535, + "grad_norm": 3.4621992111206055, + "learning_rate": 9.975352997535301e-06, + "loss": 0.0339, + "step": 24658 + }, + { + "epoch": 11.463970246397025, + "grad_norm": 2.7908380031585693, + "learning_rate": 9.975351997535202e-06, + "loss": 0.0182, + "step": 24659 + }, + { + "epoch": 11.464435146443515, + "grad_norm": 6.3723907470703125, + "learning_rate": 9.9753509975351e-06, + "loss": 0.0934, + "step": 24660 + }, + { + "epoch": 11.464900046490005, + "grad_norm": 2.7003321647644043, + "learning_rate": 9.975349997535001e-06, + "loss": 0.0188, + "step": 24661 + }, + { + "epoch": 11.465364946536495, + "grad_norm": 4.392265796661377, + "learning_rate": 9.9753489975349e-06, + "loss": 0.0534, + "step": 24662 + }, + { + "epoch": 11.465829846582984, + "grad_norm": 4.166280269622803, + "learning_rate": 9.975347997534801e-06, + "loss": 0.0393, + "step": 24663 + }, + { + "epoch": 11.466294746629474, + "grad_norm": 2.337029218673706, + "learning_rate": 9.9753469975347e-06, + "loss": 0.0189, + "step": 24664 + }, + { + "epoch": 11.466759646675964, + "grad_norm": 6.311256408691406, + "learning_rate": 9.9753459975346e-06, + "loss": 0.0665, + "step": 24665 + }, + { + "epoch": 11.467224546722454, + "grad_norm": 4.7357916831970215, + "learning_rate": 9.975344997534501e-06, + "loss": 0.0701, + "step": 24666 + }, + { + "epoch": 11.467689446768945, + "grad_norm": 2.865506649017334, + "learning_rate": 9.9753439975344e-06, + "loss": 0.0213, + "step": 24667 + }, + { + "epoch": 11.468154346815435, + "grad_norm": 4.966041088104248, + "learning_rate": 9.9753429975343e-06, + "loss": 0.0804, + "step": 24668 + }, + { + "epoch": 11.468619246861925, + "grad_norm": 3.543506622314453, + "learning_rate": 9.9753419975342e-06, + "loss": 0.0267, + "step": 24669 + }, + { + "epoch": 11.469084146908415, + "grad_norm": 4.052424430847168, + "learning_rate": 9.9753409975341e-06, + "loss": 0.0256, + "step": 24670 + }, + { + "epoch": 11.469549046954905, + "grad_norm": 6.668074131011963, + "learning_rate": 9.975339997534001e-06, + "loss": 0.0915, + "step": 24671 + }, + { + "epoch": 11.470013947001394, + "grad_norm": 4.824066162109375, + "learning_rate": 9.975338997533902e-06, + "loss": 0.0595, + "step": 24672 + }, + { + "epoch": 11.470478847047884, + "grad_norm": 1.9759312868118286, + "learning_rate": 9.9753379975338e-06, + "loss": 0.0133, + "step": 24673 + }, + { + "epoch": 11.470943747094374, + "grad_norm": 3.181312084197998, + "learning_rate": 9.975336997533701e-06, + "loss": 0.0241, + "step": 24674 + }, + { + "epoch": 11.471408647140866, + "grad_norm": 2.16818904876709, + "learning_rate": 9.9753359975336e-06, + "loss": 0.0105, + "step": 24675 + }, + { + "epoch": 11.471873547187355, + "grad_norm": 5.952610015869141, + "learning_rate": 9.9753349975335e-06, + "loss": 0.0887, + "step": 24676 + }, + { + "epoch": 11.472338447233845, + "grad_norm": 3.0765581130981445, + "learning_rate": 9.9753339975334e-06, + "loss": 0.0354, + "step": 24677 + }, + { + "epoch": 11.472803347280335, + "grad_norm": 7.147959232330322, + "learning_rate": 9.9753329975333e-06, + "loss": 0.1193, + "step": 24678 + }, + { + "epoch": 11.473268247326825, + "grad_norm": 6.7989277839660645, + "learning_rate": 9.975331997533201e-06, + "loss": 0.1037, + "step": 24679 + }, + { + "epoch": 11.473733147373315, + "grad_norm": 4.770583629608154, + "learning_rate": 9.9753309975331e-06, + "loss": 0.0941, + "step": 24680 + }, + { + "epoch": 11.474198047419804, + "grad_norm": 2.196842908859253, + "learning_rate": 9.975329997533001e-06, + "loss": 0.0186, + "step": 24681 + }, + { + "epoch": 11.474662947466294, + "grad_norm": 2.7104580402374268, + "learning_rate": 9.9753289975329e-06, + "loss": 0.0164, + "step": 24682 + }, + { + "epoch": 11.475127847512784, + "grad_norm": 1.3234996795654297, + "learning_rate": 9.9753279975328e-06, + "loss": 0.0126, + "step": 24683 + }, + { + "epoch": 11.475592747559276, + "grad_norm": 2.188023090362549, + "learning_rate": 9.975326997532701e-06, + "loss": 0.016, + "step": 24684 + }, + { + "epoch": 11.476057647605765, + "grad_norm": 4.031405448913574, + "learning_rate": 9.975325997532602e-06, + "loss": 0.0316, + "step": 24685 + }, + { + "epoch": 11.476522547652255, + "grad_norm": 4.59927225112915, + "learning_rate": 9.9753249975325e-06, + "loss": 0.0858, + "step": 24686 + }, + { + "epoch": 11.476987447698745, + "grad_norm": 3.9871318340301514, + "learning_rate": 9.9753239975324e-06, + "loss": 0.0492, + "step": 24687 + }, + { + "epoch": 11.477452347745235, + "grad_norm": 5.496943473815918, + "learning_rate": 9.9753229975323e-06, + "loss": 0.0927, + "step": 24688 + }, + { + "epoch": 11.477917247791725, + "grad_norm": 7.281867027282715, + "learning_rate": 9.975321997532201e-06, + "loss": 0.094, + "step": 24689 + }, + { + "epoch": 11.478382147838214, + "grad_norm": 8.434067726135254, + "learning_rate": 9.9753209975321e-06, + "loss": 0.0765, + "step": 24690 + }, + { + "epoch": 11.478847047884704, + "grad_norm": 4.16391658782959, + "learning_rate": 9.975319997532e-06, + "loss": 0.0311, + "step": 24691 + }, + { + "epoch": 11.479311947931194, + "grad_norm": 4.237071990966797, + "learning_rate": 9.975318997531901e-06, + "loss": 0.0262, + "step": 24692 + }, + { + "epoch": 11.479776847977686, + "grad_norm": 7.198727130889893, + "learning_rate": 9.9753179975318e-06, + "loss": 0.1359, + "step": 24693 + }, + { + "epoch": 11.480241748024175, + "grad_norm": 6.389607906341553, + "learning_rate": 9.975316997531701e-06, + "loss": 0.1577, + "step": 24694 + }, + { + "epoch": 11.480706648070665, + "grad_norm": 3.1548259258270264, + "learning_rate": 9.9753159975316e-06, + "loss": 0.044, + "step": 24695 + }, + { + "epoch": 11.481171548117155, + "grad_norm": 5.673299312591553, + "learning_rate": 9.9753149975315e-06, + "loss": 0.0905, + "step": 24696 + }, + { + "epoch": 11.481636448163645, + "grad_norm": 5.971745491027832, + "learning_rate": 9.975313997531401e-06, + "loss": 0.0895, + "step": 24697 + }, + { + "epoch": 11.482101348210135, + "grad_norm": 5.618170261383057, + "learning_rate": 9.9753129975313e-06, + "loss": 0.0836, + "step": 24698 + }, + { + "epoch": 11.482566248256624, + "grad_norm": 7.54660701751709, + "learning_rate": 9.975311997531201e-06, + "loss": 0.071, + "step": 24699 + }, + { + "epoch": 11.483031148303114, + "grad_norm": 3.2266476154327393, + "learning_rate": 9.9753109975311e-06, + "loss": 0.0409, + "step": 24700 + }, + { + "epoch": 11.483496048349604, + "grad_norm": 5.117408275604248, + "learning_rate": 9.975309997531e-06, + "loss": 0.0418, + "step": 24701 + }, + { + "epoch": 11.483960948396096, + "grad_norm": 3.267585039138794, + "learning_rate": 9.975308997530901e-06, + "loss": 0.0252, + "step": 24702 + }, + { + "epoch": 11.484425848442585, + "grad_norm": 2.073153257369995, + "learning_rate": 9.9753079975308e-06, + "loss": 0.0129, + "step": 24703 + }, + { + "epoch": 11.484890748489075, + "grad_norm": 3.329282760620117, + "learning_rate": 9.9753069975307e-06, + "loss": 0.0189, + "step": 24704 + }, + { + "epoch": 11.485355648535565, + "grad_norm": 12.791723251342773, + "learning_rate": 9.975305997530601e-06, + "loss": 0.1086, + "step": 24705 + }, + { + "epoch": 11.485820548582055, + "grad_norm": 5.17341947555542, + "learning_rate": 9.9753049975305e-06, + "loss": 0.0637, + "step": 24706 + }, + { + "epoch": 11.486285448628545, + "grad_norm": 6.698720455169678, + "learning_rate": 9.9753039975304e-06, + "loss": 0.0865, + "step": 24707 + }, + { + "epoch": 11.486750348675034, + "grad_norm": 3.035001754760742, + "learning_rate": 9.9753029975303e-06, + "loss": 0.0444, + "step": 24708 + }, + { + "epoch": 11.487215248721524, + "grad_norm": 3.615945339202881, + "learning_rate": 9.9753019975302e-06, + "loss": 0.0551, + "step": 24709 + }, + { + "epoch": 11.487680148768014, + "grad_norm": 3.211228847503662, + "learning_rate": 9.975300997530101e-06, + "loss": 0.0876, + "step": 24710 + }, + { + "epoch": 11.488145048814506, + "grad_norm": 5.680140018463135, + "learning_rate": 9.97529999753e-06, + "loss": 0.0467, + "step": 24711 + }, + { + "epoch": 11.488609948860995, + "grad_norm": 1.8863812685012817, + "learning_rate": 9.975298997529901e-06, + "loss": 0.0179, + "step": 24712 + }, + { + "epoch": 11.489074848907485, + "grad_norm": 3.819211721420288, + "learning_rate": 9.9752979975298e-06, + "loss": 0.0434, + "step": 24713 + }, + { + "epoch": 11.489539748953975, + "grad_norm": 2.5619540214538574, + "learning_rate": 9.9752969975297e-06, + "loss": 0.0174, + "step": 24714 + }, + { + "epoch": 11.490004649000465, + "grad_norm": 8.27382755279541, + "learning_rate": 9.975295997529601e-06, + "loss": 0.1815, + "step": 24715 + }, + { + "epoch": 11.490469549046955, + "grad_norm": 3.9349629878997803, + "learning_rate": 9.975294997529502e-06, + "loss": 0.0353, + "step": 24716 + }, + { + "epoch": 11.490934449093444, + "grad_norm": 5.361056327819824, + "learning_rate": 9.9752939975294e-06, + "loss": 0.0893, + "step": 24717 + }, + { + "epoch": 11.491399349139934, + "grad_norm": 6.191969394683838, + "learning_rate": 9.9752929975293e-06, + "loss": 0.0458, + "step": 24718 + }, + { + "epoch": 11.491864249186424, + "grad_norm": 3.076101303100586, + "learning_rate": 9.9752919975292e-06, + "loss": 0.0303, + "step": 24719 + }, + { + "epoch": 11.492329149232916, + "grad_norm": 2.1521923542022705, + "learning_rate": 9.9752909975291e-06, + "loss": 0.0213, + "step": 24720 + }, + { + "epoch": 11.492794049279405, + "grad_norm": 7.652292251586914, + "learning_rate": 9.975289997529e-06, + "loss": 0.1339, + "step": 24721 + }, + { + "epoch": 11.493258949325895, + "grad_norm": 8.777121543884277, + "learning_rate": 9.9752889975289e-06, + "loss": 0.0756, + "step": 24722 + }, + { + "epoch": 11.493723849372385, + "grad_norm": 6.710319995880127, + "learning_rate": 9.975287997528801e-06, + "loss": 0.073, + "step": 24723 + }, + { + "epoch": 11.494188749418875, + "grad_norm": 2.7246005535125732, + "learning_rate": 9.9752869975287e-06, + "loss": 0.0301, + "step": 24724 + }, + { + "epoch": 11.494653649465365, + "grad_norm": 5.524467945098877, + "learning_rate": 9.975285997528601e-06, + "loss": 0.076, + "step": 24725 + }, + { + "epoch": 11.495118549511854, + "grad_norm": 3.866633892059326, + "learning_rate": 9.9752849975285e-06, + "loss": 0.0406, + "step": 24726 + }, + { + "epoch": 11.495583449558344, + "grad_norm": 1.691170573234558, + "learning_rate": 9.9752839975284e-06, + "loss": 0.0214, + "step": 24727 + }, + { + "epoch": 11.496048349604836, + "grad_norm": 5.857227325439453, + "learning_rate": 9.975282997528301e-06, + "loss": 0.081, + "step": 24728 + }, + { + "epoch": 11.496513249651326, + "grad_norm": 4.111086845397949, + "learning_rate": 9.9752819975282e-06, + "loss": 0.037, + "step": 24729 + }, + { + "epoch": 11.496978149697815, + "grad_norm": 7.576366424560547, + "learning_rate": 9.975280997528101e-06, + "loss": 0.1061, + "step": 24730 + }, + { + "epoch": 11.497443049744305, + "grad_norm": 4.175252914428711, + "learning_rate": 9.975279997528e-06, + "loss": 0.0512, + "step": 24731 + }, + { + "epoch": 11.497907949790795, + "grad_norm": 0.7123712301254272, + "learning_rate": 9.9752789975279e-06, + "loss": 0.0079, + "step": 24732 + }, + { + "epoch": 11.498372849837285, + "grad_norm": 6.226868152618408, + "learning_rate": 9.9752779975278e-06, + "loss": 0.1016, + "step": 24733 + }, + { + "epoch": 11.498837749883775, + "grad_norm": 2.5581705570220947, + "learning_rate": 9.9752769975277e-06, + "loss": 0.0245, + "step": 24734 + }, + { + "epoch": 11.499302649930264, + "grad_norm": 4.128758430480957, + "learning_rate": 9.9752759975276e-06, + "loss": 0.0371, + "step": 24735 + }, + { + "epoch": 11.499767549976754, + "grad_norm": 6.097399711608887, + "learning_rate": 9.975274997527501e-06, + "loss": 0.1289, + "step": 24736 + }, + { + "epoch": 11.500232450023246, + "grad_norm": 1.6859291791915894, + "learning_rate": 9.9752739975274e-06, + "loss": 0.0126, + "step": 24737 + }, + { + "epoch": 11.500697350069736, + "grad_norm": 1.2608815431594849, + "learning_rate": 9.975272997527301e-06, + "loss": 0.0048, + "step": 24738 + }, + { + "epoch": 11.501162250116225, + "grad_norm": 8.24564266204834, + "learning_rate": 9.9752719975272e-06, + "loss": 0.1396, + "step": 24739 + }, + { + "epoch": 11.501627150162715, + "grad_norm": 2.5094218254089355, + "learning_rate": 9.9752709975271e-06, + "loss": 0.0183, + "step": 24740 + }, + { + "epoch": 11.502092050209205, + "grad_norm": 5.115262031555176, + "learning_rate": 9.975269997527001e-06, + "loss": 0.0638, + "step": 24741 + }, + { + "epoch": 11.502556950255695, + "grad_norm": 6.437445640563965, + "learning_rate": 9.9752689975269e-06, + "loss": 0.1041, + "step": 24742 + }, + { + "epoch": 11.503021850302185, + "grad_norm": 3.8367204666137695, + "learning_rate": 9.975267997526801e-06, + "loss": 0.049, + "step": 24743 + }, + { + "epoch": 11.503486750348674, + "grad_norm": 1.0523451566696167, + "learning_rate": 9.9752669975267e-06, + "loss": 0.0072, + "step": 24744 + }, + { + "epoch": 11.503951650395164, + "grad_norm": 7.135903835296631, + "learning_rate": 9.9752659975266e-06, + "loss": 0.1558, + "step": 24745 + }, + { + "epoch": 11.504416550441656, + "grad_norm": 3.42683744430542, + "learning_rate": 9.9752649975265e-06, + "loss": 0.0481, + "step": 24746 + }, + { + "epoch": 11.504881450488146, + "grad_norm": 5.1753644943237305, + "learning_rate": 9.9752639975264e-06, + "loss": 0.0533, + "step": 24747 + }, + { + "epoch": 11.505346350534635, + "grad_norm": 8.18799877166748, + "learning_rate": 9.975262997526301e-06, + "loss": 0.0772, + "step": 24748 + }, + { + "epoch": 11.505811250581125, + "grad_norm": 2.4988160133361816, + "learning_rate": 9.975261997526202e-06, + "loss": 0.0199, + "step": 24749 + }, + { + "epoch": 11.506276150627615, + "grad_norm": 8.661211967468262, + "learning_rate": 9.9752609975261e-06, + "loss": 0.1217, + "step": 24750 + }, + { + "epoch": 11.506741050674105, + "grad_norm": 2.6667838096618652, + "learning_rate": 9.975259997526e-06, + "loss": 0.0252, + "step": 24751 + }, + { + "epoch": 11.507205950720595, + "grad_norm": 1.0273562669754028, + "learning_rate": 9.9752589975259e-06, + "loss": 0.0087, + "step": 24752 + }, + { + "epoch": 11.507670850767084, + "grad_norm": 4.548028469085693, + "learning_rate": 9.9752579975258e-06, + "loss": 0.0466, + "step": 24753 + }, + { + "epoch": 11.508135750813576, + "grad_norm": 10.285873413085938, + "learning_rate": 9.975256997525701e-06, + "loss": 0.1524, + "step": 24754 + }, + { + "epoch": 11.508600650860066, + "grad_norm": 4.775996685028076, + "learning_rate": 9.9752559975256e-06, + "loss": 0.0359, + "step": 24755 + }, + { + "epoch": 11.509065550906556, + "grad_norm": 3.971421241760254, + "learning_rate": 9.975254997525501e-06, + "loss": 0.0283, + "step": 24756 + }, + { + "epoch": 11.509530450953045, + "grad_norm": 4.585459232330322, + "learning_rate": 9.9752539975254e-06, + "loss": 0.038, + "step": 24757 + }, + { + "epoch": 11.509995350999535, + "grad_norm": 5.507528781890869, + "learning_rate": 9.9752529975253e-06, + "loss": 0.0869, + "step": 24758 + }, + { + "epoch": 11.510460251046025, + "grad_norm": 9.242963790893555, + "learning_rate": 9.9752519975252e-06, + "loss": 0.43, + "step": 24759 + }, + { + "epoch": 11.510925151092515, + "grad_norm": 4.0745086669921875, + "learning_rate": 9.9752509975251e-06, + "loss": 0.0713, + "step": 24760 + }, + { + "epoch": 11.511390051139005, + "grad_norm": 3.4879324436187744, + "learning_rate": 9.975249997525001e-06, + "loss": 0.024, + "step": 24761 + }, + { + "epoch": 11.511854951185494, + "grad_norm": 5.3771491050720215, + "learning_rate": 9.9752489975249e-06, + "loss": 0.0749, + "step": 24762 + }, + { + "epoch": 11.512319851231986, + "grad_norm": 6.540536880493164, + "learning_rate": 9.9752479975248e-06, + "loss": 0.0898, + "step": 24763 + }, + { + "epoch": 11.512784751278476, + "grad_norm": 5.7682929039001465, + "learning_rate": 9.9752469975247e-06, + "loss": 0.0753, + "step": 24764 + }, + { + "epoch": 11.513249651324966, + "grad_norm": 2.0381813049316406, + "learning_rate": 9.9752459975246e-06, + "loss": 0.0139, + "step": 24765 + }, + { + "epoch": 11.513714551371455, + "grad_norm": 3.3857200145721436, + "learning_rate": 9.975244997524501e-06, + "loss": 0.0346, + "step": 24766 + }, + { + "epoch": 11.514179451417945, + "grad_norm": 6.30335807800293, + "learning_rate": 9.975243997524402e-06, + "loss": 0.0742, + "step": 24767 + }, + { + "epoch": 11.514644351464435, + "grad_norm": 3.8097357749938965, + "learning_rate": 9.9752429975243e-06, + "loss": 0.0227, + "step": 24768 + }, + { + "epoch": 11.515109251510925, + "grad_norm": 4.041451454162598, + "learning_rate": 9.975241997524201e-06, + "loss": 0.0524, + "step": 24769 + }, + { + "epoch": 11.515574151557415, + "grad_norm": 0.29501914978027344, + "learning_rate": 9.9752409975241e-06, + "loss": 0.0027, + "step": 24770 + }, + { + "epoch": 11.516039051603904, + "grad_norm": 6.085726737976074, + "learning_rate": 9.975239997524e-06, + "loss": 0.0605, + "step": 24771 + }, + { + "epoch": 11.516503951650396, + "grad_norm": 1.7292718887329102, + "learning_rate": 9.9752389975239e-06, + "loss": 0.0157, + "step": 24772 + }, + { + "epoch": 11.516968851696886, + "grad_norm": 0.21618585288524628, + "learning_rate": 9.9752379975238e-06, + "loss": 0.0021, + "step": 24773 + }, + { + "epoch": 11.517433751743376, + "grad_norm": 7.894989967346191, + "learning_rate": 9.975236997523701e-06, + "loss": 0.0907, + "step": 24774 + }, + { + "epoch": 11.517898651789865, + "grad_norm": 2.871276617050171, + "learning_rate": 9.9752359975236e-06, + "loss": 0.0372, + "step": 24775 + }, + { + "epoch": 11.518363551836355, + "grad_norm": 8.156793594360352, + "learning_rate": 9.9752349975235e-06, + "loss": 0.1605, + "step": 24776 + }, + { + "epoch": 11.518828451882845, + "grad_norm": 3.8228673934936523, + "learning_rate": 9.9752339975234e-06, + "loss": 0.0424, + "step": 24777 + }, + { + "epoch": 11.519293351929335, + "grad_norm": 2.9189422130584717, + "learning_rate": 9.9752329975233e-06, + "loss": 0.0271, + "step": 24778 + }, + { + "epoch": 11.519758251975825, + "grad_norm": 3.703706741333008, + "learning_rate": 9.975231997523201e-06, + "loss": 0.0235, + "step": 24779 + }, + { + "epoch": 11.520223152022314, + "grad_norm": 6.348263263702393, + "learning_rate": 9.975230997523102e-06, + "loss": 0.0766, + "step": 24780 + }, + { + "epoch": 11.520688052068806, + "grad_norm": 3.933375358581543, + "learning_rate": 9.975229997523e-06, + "loss": 0.0451, + "step": 24781 + }, + { + "epoch": 11.521152952115296, + "grad_norm": 0.17828713357448578, + "learning_rate": 9.975228997522901e-06, + "loss": 0.002, + "step": 24782 + }, + { + "epoch": 11.521617852161786, + "grad_norm": 2.2274630069732666, + "learning_rate": 9.9752279975228e-06, + "loss": 0.0178, + "step": 24783 + }, + { + "epoch": 11.522082752208275, + "grad_norm": 1.7081634998321533, + "learning_rate": 9.975226997522701e-06, + "loss": 0.009, + "step": 24784 + }, + { + "epoch": 11.522547652254765, + "grad_norm": 6.8611836433410645, + "learning_rate": 9.975225997522602e-06, + "loss": 0.1341, + "step": 24785 + }, + { + "epoch": 11.523012552301255, + "grad_norm": 4.3082451820373535, + "learning_rate": 9.9752249975225e-06, + "loss": 0.0636, + "step": 24786 + }, + { + "epoch": 11.523477452347745, + "grad_norm": 7.223718166351318, + "learning_rate": 9.975223997522401e-06, + "loss": 0.1049, + "step": 24787 + }, + { + "epoch": 11.523942352394235, + "grad_norm": 9.012804985046387, + "learning_rate": 9.9752229975223e-06, + "loss": 0.0476, + "step": 24788 + }, + { + "epoch": 11.524407252440724, + "grad_norm": 2.791999101638794, + "learning_rate": 9.9752219975222e-06, + "loss": 0.0275, + "step": 24789 + }, + { + "epoch": 11.524872152487216, + "grad_norm": 2.277616262435913, + "learning_rate": 9.9752209975221e-06, + "loss": 0.0335, + "step": 24790 + }, + { + "epoch": 11.525337052533706, + "grad_norm": 3.470705986022949, + "learning_rate": 9.975219997522e-06, + "loss": 0.041, + "step": 24791 + }, + { + "epoch": 11.525801952580196, + "grad_norm": 2.4443845748901367, + "learning_rate": 9.975218997521901e-06, + "loss": 0.0238, + "step": 24792 + }, + { + "epoch": 11.526266852626685, + "grad_norm": 5.7991719245910645, + "learning_rate": 9.975217997521802e-06, + "loss": 0.0621, + "step": 24793 + }, + { + "epoch": 11.526731752673175, + "grad_norm": 3.586974620819092, + "learning_rate": 9.9752169975217e-06, + "loss": 0.027, + "step": 24794 + }, + { + "epoch": 11.527196652719665, + "grad_norm": 7.383157730102539, + "learning_rate": 9.9752159975216e-06, + "loss": 0.1134, + "step": 24795 + }, + { + "epoch": 11.527661552766155, + "grad_norm": 8.162758827209473, + "learning_rate": 9.9752149975215e-06, + "loss": 0.1334, + "step": 24796 + }, + { + "epoch": 11.528126452812645, + "grad_norm": 1.889672040939331, + "learning_rate": 9.975213997521401e-06, + "loss": 0.017, + "step": 24797 + }, + { + "epoch": 11.528591352859134, + "grad_norm": 2.7834599018096924, + "learning_rate": 9.975212997521302e-06, + "loss": 0.0297, + "step": 24798 + }, + { + "epoch": 11.529056252905626, + "grad_norm": 3.5914595127105713, + "learning_rate": 9.9752119975212e-06, + "loss": 0.0363, + "step": 24799 + }, + { + "epoch": 11.529521152952116, + "grad_norm": 2.9810967445373535, + "learning_rate": 9.975210997521101e-06, + "loss": 0.0236, + "step": 24800 + }, + { + "epoch": 11.529986052998606, + "grad_norm": 6.445918560028076, + "learning_rate": 9.975209997521e-06, + "loss": 0.0728, + "step": 24801 + }, + { + "epoch": 11.530450953045095, + "grad_norm": 2.490661859512329, + "learning_rate": 9.9752089975209e-06, + "loss": 0.0281, + "step": 24802 + }, + { + "epoch": 11.530915853091585, + "grad_norm": 6.748380184173584, + "learning_rate": 9.9752079975208e-06, + "loss": 0.0905, + "step": 24803 + }, + { + "epoch": 11.531380753138075, + "grad_norm": 3.376664161682129, + "learning_rate": 9.9752069975207e-06, + "loss": 0.0462, + "step": 24804 + }, + { + "epoch": 11.531845653184565, + "grad_norm": 6.277265548706055, + "learning_rate": 9.975205997520601e-06, + "loss": 0.1241, + "step": 24805 + }, + { + "epoch": 11.532310553231055, + "grad_norm": 1.9259819984436035, + "learning_rate": 9.9752049975205e-06, + "loss": 0.0143, + "step": 24806 + }, + { + "epoch": 11.532775453277544, + "grad_norm": 1.7649301290512085, + "learning_rate": 9.9752039975204e-06, + "loss": 0.0167, + "step": 24807 + }, + { + "epoch": 11.533240353324036, + "grad_norm": 2.8483195304870605, + "learning_rate": 9.9752029975203e-06, + "loss": 0.0223, + "step": 24808 + }, + { + "epoch": 11.533705253370526, + "grad_norm": 1.9406174421310425, + "learning_rate": 9.9752019975202e-06, + "loss": 0.0191, + "step": 24809 + }, + { + "epoch": 11.534170153417016, + "grad_norm": 3.4873902797698975, + "learning_rate": 9.975200997520101e-06, + "loss": 0.0279, + "step": 24810 + }, + { + "epoch": 11.534635053463505, + "grad_norm": 5.502872943878174, + "learning_rate": 9.975199997520002e-06, + "loss": 0.1164, + "step": 24811 + }, + { + "epoch": 11.535099953509995, + "grad_norm": 2.720820188522339, + "learning_rate": 9.9751989975199e-06, + "loss": 0.0208, + "step": 24812 + }, + { + "epoch": 11.535564853556485, + "grad_norm": 8.562231063842773, + "learning_rate": 9.975197997519801e-06, + "loss": 0.0816, + "step": 24813 + }, + { + "epoch": 11.536029753602975, + "grad_norm": 3.2916810512542725, + "learning_rate": 9.9751969975197e-06, + "loss": 0.0256, + "step": 24814 + }, + { + "epoch": 11.536494653649465, + "grad_norm": 1.7893733978271484, + "learning_rate": 9.975195997519601e-06, + "loss": 0.0138, + "step": 24815 + }, + { + "epoch": 11.536959553695956, + "grad_norm": 2.9152603149414062, + "learning_rate": 9.9751949975195e-06, + "loss": 0.0299, + "step": 24816 + }, + { + "epoch": 11.537424453742446, + "grad_norm": 2.8576056957244873, + "learning_rate": 9.9751939975194e-06, + "loss": 0.0334, + "step": 24817 + }, + { + "epoch": 11.537889353788936, + "grad_norm": 4.554337978363037, + "learning_rate": 9.975192997519301e-06, + "loss": 0.0441, + "step": 24818 + }, + { + "epoch": 11.538354253835426, + "grad_norm": 1.2671561241149902, + "learning_rate": 9.9751919975192e-06, + "loss": 0.009, + "step": 24819 + }, + { + "epoch": 11.538819153881915, + "grad_norm": 6.558575630187988, + "learning_rate": 9.9751909975191e-06, + "loss": 0.0594, + "step": 24820 + }, + { + "epoch": 11.539284053928405, + "grad_norm": 10.936347007751465, + "learning_rate": 9.975189997519e-06, + "loss": 0.2335, + "step": 24821 + }, + { + "epoch": 11.539748953974895, + "grad_norm": 3.2791378498077393, + "learning_rate": 9.9751889975189e-06, + "loss": 0.0401, + "step": 24822 + }, + { + "epoch": 11.540213854021385, + "grad_norm": 5.20433235168457, + "learning_rate": 9.975187997518801e-06, + "loss": 0.0471, + "step": 24823 + }, + { + "epoch": 11.540678754067875, + "grad_norm": 3.3154077529907227, + "learning_rate": 9.975186997518702e-06, + "loss": 0.0364, + "step": 24824 + }, + { + "epoch": 11.541143654114366, + "grad_norm": 6.356187343597412, + "learning_rate": 9.9751859975186e-06, + "loss": 0.0704, + "step": 24825 + }, + { + "epoch": 11.541608554160856, + "grad_norm": 7.486331462860107, + "learning_rate": 9.975184997518501e-06, + "loss": 0.1033, + "step": 24826 + }, + { + "epoch": 11.542073454207346, + "grad_norm": 5.550586223602295, + "learning_rate": 9.9751839975184e-06, + "loss": 0.0848, + "step": 24827 + }, + { + "epoch": 11.542538354253836, + "grad_norm": 5.322463512420654, + "learning_rate": 9.9751829975183e-06, + "loss": 0.0418, + "step": 24828 + }, + { + "epoch": 11.543003254300325, + "grad_norm": 5.128171920776367, + "learning_rate": 9.9751819975182e-06, + "loss": 0.0289, + "step": 24829 + }, + { + "epoch": 11.543468154346815, + "grad_norm": 11.367385864257812, + "learning_rate": 9.9751809975181e-06, + "loss": 0.1284, + "step": 24830 + }, + { + "epoch": 11.543933054393305, + "grad_norm": 4.396881103515625, + "learning_rate": 9.975179997518001e-06, + "loss": 0.0351, + "step": 24831 + }, + { + "epoch": 11.544397954439795, + "grad_norm": 5.399127006530762, + "learning_rate": 9.9751789975179e-06, + "loss": 0.0764, + "step": 24832 + }, + { + "epoch": 11.544862854486286, + "grad_norm": 6.405257701873779, + "learning_rate": 9.975177997517801e-06, + "loss": 0.0944, + "step": 24833 + }, + { + "epoch": 11.545327754532776, + "grad_norm": 4.0537238121032715, + "learning_rate": 9.9751769975177e-06, + "loss": 0.0352, + "step": 24834 + }, + { + "epoch": 11.545792654579266, + "grad_norm": 2.687363624572754, + "learning_rate": 9.9751759975176e-06, + "loss": 0.0182, + "step": 24835 + }, + { + "epoch": 11.546257554625756, + "grad_norm": 4.258168697357178, + "learning_rate": 9.975174997517501e-06, + "loss": 0.0357, + "step": 24836 + }, + { + "epoch": 11.546722454672246, + "grad_norm": 2.810685396194458, + "learning_rate": 9.975173997517402e-06, + "loss": 0.0215, + "step": 24837 + }, + { + "epoch": 11.547187354718735, + "grad_norm": 5.525300025939941, + "learning_rate": 9.9751729975173e-06, + "loss": 0.065, + "step": 24838 + }, + { + "epoch": 11.547652254765225, + "grad_norm": 0.9249053597450256, + "learning_rate": 9.9751719975172e-06, + "loss": 0.0064, + "step": 24839 + }, + { + "epoch": 11.548117154811715, + "grad_norm": 3.8906502723693848, + "learning_rate": 9.9751709975171e-06, + "loss": 0.0473, + "step": 24840 + }, + { + "epoch": 11.548582054858205, + "grad_norm": 5.045830726623535, + "learning_rate": 9.975169997517e-06, + "loss": 0.0754, + "step": 24841 + }, + { + "epoch": 11.549046954904696, + "grad_norm": 6.8939104080200195, + "learning_rate": 9.9751689975169e-06, + "loss": 0.0851, + "step": 24842 + }, + { + "epoch": 11.549511854951186, + "grad_norm": 3.7087903022766113, + "learning_rate": 9.9751679975168e-06, + "loss": 0.0482, + "step": 24843 + }, + { + "epoch": 11.549976754997676, + "grad_norm": 7.1325249671936035, + "learning_rate": 9.975166997516701e-06, + "loss": 0.1508, + "step": 24844 + }, + { + "epoch": 11.550441655044166, + "grad_norm": 5.399710655212402, + "learning_rate": 9.9751659975166e-06, + "loss": 0.0793, + "step": 24845 + }, + { + "epoch": 11.550906555090656, + "grad_norm": 1.9793721437454224, + "learning_rate": 9.975164997516501e-06, + "loss": 0.0178, + "step": 24846 + }, + { + "epoch": 11.551371455137145, + "grad_norm": 1.977480173110962, + "learning_rate": 9.9751639975164e-06, + "loss": 0.0106, + "step": 24847 + }, + { + "epoch": 11.551836355183635, + "grad_norm": 3.854572057723999, + "learning_rate": 9.9751629975163e-06, + "loss": 0.0321, + "step": 24848 + }, + { + "epoch": 11.552301255230125, + "grad_norm": 3.489041566848755, + "learning_rate": 9.975161997516201e-06, + "loss": 0.0464, + "step": 24849 + }, + { + "epoch": 11.552766155276615, + "grad_norm": 0.7576117515563965, + "learning_rate": 9.9751609975161e-06, + "loss": 0.0054, + "step": 24850 + }, + { + "epoch": 11.553231055323106, + "grad_norm": 1.8478132486343384, + "learning_rate": 9.975159997516001e-06, + "loss": 0.0097, + "step": 24851 + }, + { + "epoch": 11.553695955369596, + "grad_norm": 3.075230836868286, + "learning_rate": 9.9751589975159e-06, + "loss": 0.0161, + "step": 24852 + }, + { + "epoch": 11.554160855416086, + "grad_norm": 4.387439250946045, + "learning_rate": 9.9751579975158e-06, + "loss": 0.0724, + "step": 24853 + }, + { + "epoch": 11.554625755462576, + "grad_norm": 6.609029293060303, + "learning_rate": 9.975156997515701e-06, + "loss": 0.0477, + "step": 24854 + }, + { + "epoch": 11.555090655509066, + "grad_norm": 7.703186511993408, + "learning_rate": 9.9751559975156e-06, + "loss": 0.1093, + "step": 24855 + }, + { + "epoch": 11.555555555555555, + "grad_norm": 3.887345790863037, + "learning_rate": 9.9751549975155e-06, + "loss": 0.0558, + "step": 24856 + }, + { + "epoch": 11.556020455602045, + "grad_norm": 3.364300012588501, + "learning_rate": 9.975153997515401e-06, + "loss": 0.0222, + "step": 24857 + }, + { + "epoch": 11.556485355648535, + "grad_norm": 5.865367412567139, + "learning_rate": 9.9751529975153e-06, + "loss": 0.0695, + "step": 24858 + }, + { + "epoch": 11.556950255695025, + "grad_norm": 5.066498756408691, + "learning_rate": 9.9751519975152e-06, + "loss": 0.0914, + "step": 24859 + }, + { + "epoch": 11.557415155741516, + "grad_norm": 2.459066152572632, + "learning_rate": 9.9751509975151e-06, + "loss": 0.0157, + "step": 24860 + }, + { + "epoch": 11.557880055788006, + "grad_norm": 1.1510100364685059, + "learning_rate": 9.975149997515e-06, + "loss": 0.0074, + "step": 24861 + }, + { + "epoch": 11.558344955834496, + "grad_norm": 4.170164585113525, + "learning_rate": 9.975148997514901e-06, + "loss": 0.0462, + "step": 24862 + }, + { + "epoch": 11.558809855880986, + "grad_norm": 3.5865516662597656, + "learning_rate": 9.9751479975148e-06, + "loss": 0.0468, + "step": 24863 + }, + { + "epoch": 11.559274755927476, + "grad_norm": 1.5339255332946777, + "learning_rate": 9.975146997514701e-06, + "loss": 0.0145, + "step": 24864 + }, + { + "epoch": 11.559739655973965, + "grad_norm": 5.737358093261719, + "learning_rate": 9.9751459975146e-06, + "loss": 0.1248, + "step": 24865 + }, + { + "epoch": 11.560204556020455, + "grad_norm": 4.501669883728027, + "learning_rate": 9.9751449975145e-06, + "loss": 0.06, + "step": 24866 + }, + { + "epoch": 11.560669456066945, + "grad_norm": 3.943881034851074, + "learning_rate": 9.975143997514401e-06, + "loss": 0.0481, + "step": 24867 + }, + { + "epoch": 11.561134356113435, + "grad_norm": 3.5932769775390625, + "learning_rate": 9.9751429975143e-06, + "loss": 0.0678, + "step": 24868 + }, + { + "epoch": 11.561599256159926, + "grad_norm": 3.300713062286377, + "learning_rate": 9.975141997514201e-06, + "loss": 0.0379, + "step": 24869 + }, + { + "epoch": 11.562064156206416, + "grad_norm": 4.743597030639648, + "learning_rate": 9.9751409975141e-06, + "loss": 0.0858, + "step": 24870 + }, + { + "epoch": 11.562529056252906, + "grad_norm": 4.8062896728515625, + "learning_rate": 9.975139997514e-06, + "loss": 0.0447, + "step": 24871 + }, + { + "epoch": 11.562993956299396, + "grad_norm": 6.26339864730835, + "learning_rate": 9.9751389975139e-06, + "loss": 0.1455, + "step": 24872 + }, + { + "epoch": 11.563458856345886, + "grad_norm": 3.1669225692749023, + "learning_rate": 9.9751379975138e-06, + "loss": 0.035, + "step": 24873 + }, + { + "epoch": 11.563923756392375, + "grad_norm": 3.3737001419067383, + "learning_rate": 9.9751369975137e-06, + "loss": 0.0445, + "step": 24874 + }, + { + "epoch": 11.564388656438865, + "grad_norm": 5.547180652618408, + "learning_rate": 9.975135997513601e-06, + "loss": 0.0665, + "step": 24875 + }, + { + "epoch": 11.564853556485355, + "grad_norm": 2.640037775039673, + "learning_rate": 9.9751349975135e-06, + "loss": 0.0236, + "step": 24876 + }, + { + "epoch": 11.565318456531845, + "grad_norm": 4.498903274536133, + "learning_rate": 9.975133997513401e-06, + "loss": 0.074, + "step": 24877 + }, + { + "epoch": 11.565783356578336, + "grad_norm": 2.7590150833129883, + "learning_rate": 9.9751329975133e-06, + "loss": 0.0305, + "step": 24878 + }, + { + "epoch": 11.566248256624826, + "grad_norm": 4.804864883422852, + "learning_rate": 9.9751319975132e-06, + "loss": 0.1101, + "step": 24879 + }, + { + "epoch": 11.566713156671316, + "grad_norm": 1.5030255317687988, + "learning_rate": 9.975130997513101e-06, + "loss": 0.013, + "step": 24880 + }, + { + "epoch": 11.567178056717806, + "grad_norm": 8.144789695739746, + "learning_rate": 9.975129997513e-06, + "loss": 0.1027, + "step": 24881 + }, + { + "epoch": 11.567642956764296, + "grad_norm": 3.8593130111694336, + "learning_rate": 9.975128997512901e-06, + "loss": 0.0556, + "step": 24882 + }, + { + "epoch": 11.568107856810785, + "grad_norm": 2.982093572616577, + "learning_rate": 9.9751279975128e-06, + "loss": 0.0353, + "step": 24883 + }, + { + "epoch": 11.568572756857275, + "grad_norm": 5.142355918884277, + "learning_rate": 9.9751269975127e-06, + "loss": 0.0867, + "step": 24884 + }, + { + "epoch": 11.569037656903765, + "grad_norm": 5.50998592376709, + "learning_rate": 9.9751259975126e-06, + "loss": 0.0441, + "step": 24885 + }, + { + "epoch": 11.569502556950255, + "grad_norm": 5.056594371795654, + "learning_rate": 9.9751249975125e-06, + "loss": 0.0718, + "step": 24886 + }, + { + "epoch": 11.569967456996746, + "grad_norm": 4.191969871520996, + "learning_rate": 9.975123997512401e-06, + "loss": 0.0765, + "step": 24887 + }, + { + "epoch": 11.570432357043236, + "grad_norm": 1.787951946258545, + "learning_rate": 9.975122997512302e-06, + "loss": 0.0111, + "step": 24888 + }, + { + "epoch": 11.570897257089726, + "grad_norm": 5.591122627258301, + "learning_rate": 9.9751219975122e-06, + "loss": 0.1093, + "step": 24889 + }, + { + "epoch": 11.571362157136216, + "grad_norm": 3.3334813117980957, + "learning_rate": 9.975120997512101e-06, + "loss": 0.0752, + "step": 24890 + }, + { + "epoch": 11.571827057182706, + "grad_norm": 4.166484355926514, + "learning_rate": 9.975119997512e-06, + "loss": 0.0518, + "step": 24891 + }, + { + "epoch": 11.572291957229195, + "grad_norm": 4.3199567794799805, + "learning_rate": 9.9751189975119e-06, + "loss": 0.0468, + "step": 24892 + }, + { + "epoch": 11.572756857275685, + "grad_norm": 2.861881732940674, + "learning_rate": 9.975117997511801e-06, + "loss": 0.0426, + "step": 24893 + }, + { + "epoch": 11.573221757322175, + "grad_norm": 3.7569339275360107, + "learning_rate": 9.9751169975117e-06, + "loss": 0.0586, + "step": 24894 + }, + { + "epoch": 11.573686657368667, + "grad_norm": 4.38253116607666, + "learning_rate": 9.975115997511601e-06, + "loss": 0.0521, + "step": 24895 + }, + { + "epoch": 11.574151557415156, + "grad_norm": 3.884611129760742, + "learning_rate": 9.9751149975115e-06, + "loss": 0.0442, + "step": 24896 + }, + { + "epoch": 11.574616457461646, + "grad_norm": 10.193238258361816, + "learning_rate": 9.9751139975114e-06, + "loss": 0.1209, + "step": 24897 + }, + { + "epoch": 11.575081357508136, + "grad_norm": 9.091756820678711, + "learning_rate": 9.9751129975113e-06, + "loss": 0.086, + "step": 24898 + }, + { + "epoch": 11.575546257554626, + "grad_norm": 4.407584190368652, + "learning_rate": 9.9751119975112e-06, + "loss": 0.0598, + "step": 24899 + }, + { + "epoch": 11.576011157601116, + "grad_norm": 6.9948296546936035, + "learning_rate": 9.975110997511101e-06, + "loss": 0.1461, + "step": 24900 + }, + { + "epoch": 11.576476057647605, + "grad_norm": 3.955476999282837, + "learning_rate": 9.975109997511002e-06, + "loss": 0.0372, + "step": 24901 + }, + { + "epoch": 11.576940957694095, + "grad_norm": 4.775691986083984, + "learning_rate": 9.9751089975109e-06, + "loss": 0.0595, + "step": 24902 + }, + { + "epoch": 11.577405857740585, + "grad_norm": 10.12955093383789, + "learning_rate": 9.9751079975108e-06, + "loss": 0.1024, + "step": 24903 + }, + { + "epoch": 11.577870757787077, + "grad_norm": 4.041571617126465, + "learning_rate": 9.9751069975107e-06, + "loss": 0.046, + "step": 24904 + }, + { + "epoch": 11.578335657833566, + "grad_norm": 4.951695442199707, + "learning_rate": 9.9751059975106e-06, + "loss": 0.0577, + "step": 24905 + }, + { + "epoch": 11.578800557880056, + "grad_norm": 5.904666900634766, + "learning_rate": 9.975104997510501e-06, + "loss": 0.0782, + "step": 24906 + }, + { + "epoch": 11.579265457926546, + "grad_norm": 4.300762176513672, + "learning_rate": 9.9751039975104e-06, + "loss": 0.0638, + "step": 24907 + }, + { + "epoch": 11.579730357973036, + "grad_norm": 2.533299684524536, + "learning_rate": 9.975102997510301e-06, + "loss": 0.014, + "step": 24908 + }, + { + "epoch": 11.580195258019526, + "grad_norm": 3.3968729972839355, + "learning_rate": 9.9751019975102e-06, + "loss": 0.0293, + "step": 24909 + }, + { + "epoch": 11.580660158066015, + "grad_norm": 6.299307823181152, + "learning_rate": 9.9751009975101e-06, + "loss": 0.1074, + "step": 24910 + }, + { + "epoch": 11.581125058112505, + "grad_norm": 2.0837228298187256, + "learning_rate": 9.97509999751e-06, + "loss": 0.012, + "step": 24911 + }, + { + "epoch": 11.581589958158997, + "grad_norm": 4.110179424285889, + "learning_rate": 9.9750989975099e-06, + "loss": 0.0371, + "step": 24912 + }, + { + "epoch": 11.582054858205487, + "grad_norm": 5.451714038848877, + "learning_rate": 9.975097997509801e-06, + "loss": 0.0761, + "step": 24913 + }, + { + "epoch": 11.582519758251976, + "grad_norm": 3.8488805294036865, + "learning_rate": 9.9750969975097e-06, + "loss": 0.0527, + "step": 24914 + }, + { + "epoch": 11.582984658298466, + "grad_norm": 5.8770246505737305, + "learning_rate": 9.9750959975096e-06, + "loss": 0.0583, + "step": 24915 + }, + { + "epoch": 11.583449558344956, + "grad_norm": 4.828329563140869, + "learning_rate": 9.9750949975095e-06, + "loss": 0.0799, + "step": 24916 + }, + { + "epoch": 11.583914458391446, + "grad_norm": 4.267880916595459, + "learning_rate": 9.9750939975094e-06, + "loss": 0.042, + "step": 24917 + }, + { + "epoch": 11.584379358437936, + "grad_norm": 3.4781606197357178, + "learning_rate": 9.975092997509301e-06, + "loss": 0.0261, + "step": 24918 + }, + { + "epoch": 11.584844258484425, + "grad_norm": 5.0832977294921875, + "learning_rate": 9.975091997509202e-06, + "loss": 0.047, + "step": 24919 + }, + { + "epoch": 11.585309158530915, + "grad_norm": 6.700254917144775, + "learning_rate": 9.9750909975091e-06, + "loss": 0.0463, + "step": 24920 + }, + { + "epoch": 11.585774058577407, + "grad_norm": 4.126446723937988, + "learning_rate": 9.975089997509001e-06, + "loss": 0.0859, + "step": 24921 + }, + { + "epoch": 11.586238958623897, + "grad_norm": 5.372165203094482, + "learning_rate": 9.9750889975089e-06, + "loss": 0.0409, + "step": 24922 + }, + { + "epoch": 11.586703858670386, + "grad_norm": 1.562933087348938, + "learning_rate": 9.9750879975088e-06, + "loss": 0.0088, + "step": 24923 + }, + { + "epoch": 11.587168758716876, + "grad_norm": 4.286819934844971, + "learning_rate": 9.9750869975087e-06, + "loss": 0.0599, + "step": 24924 + }, + { + "epoch": 11.587633658763366, + "grad_norm": 5.956228256225586, + "learning_rate": 9.9750859975086e-06, + "loss": 0.0674, + "step": 24925 + }, + { + "epoch": 11.588098558809856, + "grad_norm": 6.465571880340576, + "learning_rate": 9.975084997508501e-06, + "loss": 0.0731, + "step": 24926 + }, + { + "epoch": 11.588563458856346, + "grad_norm": 6.6207075119018555, + "learning_rate": 9.9750839975084e-06, + "loss": 0.1293, + "step": 24927 + }, + { + "epoch": 11.589028358902835, + "grad_norm": 6.023738861083984, + "learning_rate": 9.9750829975083e-06, + "loss": 0.112, + "step": 24928 + }, + { + "epoch": 11.589493258949325, + "grad_norm": 10.346025466918945, + "learning_rate": 9.9750819975082e-06, + "loss": 0.1739, + "step": 24929 + }, + { + "epoch": 11.589958158995817, + "grad_norm": 4.986145496368408, + "learning_rate": 9.9750809975081e-06, + "loss": 0.0403, + "step": 24930 + }, + { + "epoch": 11.590423059042307, + "grad_norm": 5.58300256729126, + "learning_rate": 9.975079997508001e-06, + "loss": 0.0963, + "step": 24931 + }, + { + "epoch": 11.590887959088796, + "grad_norm": 5.701540470123291, + "learning_rate": 9.975078997507902e-06, + "loss": 0.0599, + "step": 24932 + }, + { + "epoch": 11.591352859135286, + "grad_norm": 4.16494083404541, + "learning_rate": 9.9750779975078e-06, + "loss": 0.0363, + "step": 24933 + }, + { + "epoch": 11.591817759181776, + "grad_norm": 7.851160526275635, + "learning_rate": 9.975076997507701e-06, + "loss": 0.0998, + "step": 24934 + }, + { + "epoch": 11.592282659228266, + "grad_norm": 6.270783424377441, + "learning_rate": 9.9750759975076e-06, + "loss": 0.0518, + "step": 24935 + }, + { + "epoch": 11.592747559274756, + "grad_norm": 5.692477703094482, + "learning_rate": 9.975074997507501e-06, + "loss": 0.0909, + "step": 24936 + }, + { + "epoch": 11.593212459321245, + "grad_norm": 5.032869338989258, + "learning_rate": 9.9750739975074e-06, + "loss": 0.0582, + "step": 24937 + }, + { + "epoch": 11.593677359367735, + "grad_norm": 2.932969570159912, + "learning_rate": 9.9750729975073e-06, + "loss": 0.0372, + "step": 24938 + }, + { + "epoch": 11.594142259414227, + "grad_norm": 1.5136094093322754, + "learning_rate": 9.975071997507201e-06, + "loss": 0.0138, + "step": 24939 + }, + { + "epoch": 11.594607159460717, + "grad_norm": 9.573186874389648, + "learning_rate": 9.9750709975071e-06, + "loss": 0.2292, + "step": 24940 + }, + { + "epoch": 11.595072059507206, + "grad_norm": 3.7993812561035156, + "learning_rate": 9.975069997507e-06, + "loss": 0.0343, + "step": 24941 + }, + { + "epoch": 11.595536959553696, + "grad_norm": 3.312760591506958, + "learning_rate": 9.9750689975069e-06, + "loss": 0.0287, + "step": 24942 + }, + { + "epoch": 11.596001859600186, + "grad_norm": 4.976744651794434, + "learning_rate": 9.9750679975068e-06, + "loss": 0.0819, + "step": 24943 + }, + { + "epoch": 11.596466759646676, + "grad_norm": 4.829837322235107, + "learning_rate": 9.975066997506701e-06, + "loss": 0.1061, + "step": 24944 + }, + { + "epoch": 11.596931659693166, + "grad_norm": 6.563526153564453, + "learning_rate": 9.975065997506602e-06, + "loss": 0.0761, + "step": 24945 + }, + { + "epoch": 11.597396559739655, + "grad_norm": 1.482909083366394, + "learning_rate": 9.9750649975065e-06, + "loss": 0.0082, + "step": 24946 + }, + { + "epoch": 11.597861459786145, + "grad_norm": 2.3722240924835205, + "learning_rate": 9.9750639975064e-06, + "loss": 0.0129, + "step": 24947 + }, + { + "epoch": 11.598326359832637, + "grad_norm": 5.594181060791016, + "learning_rate": 9.9750629975063e-06, + "loss": 0.0727, + "step": 24948 + }, + { + "epoch": 11.598791259879127, + "grad_norm": 5.61313533782959, + "learning_rate": 9.975061997506201e-06, + "loss": 0.0698, + "step": 24949 + }, + { + "epoch": 11.599256159925616, + "grad_norm": 5.950479984283447, + "learning_rate": 9.975060997506102e-06, + "loss": 0.15, + "step": 24950 + }, + { + "epoch": 11.599721059972106, + "grad_norm": 2.5519754886627197, + "learning_rate": 9.975059997506e-06, + "loss": 0.0192, + "step": 24951 + }, + { + "epoch": 11.600185960018596, + "grad_norm": 4.212334632873535, + "learning_rate": 9.975058997505901e-06, + "loss": 0.0474, + "step": 24952 + }, + { + "epoch": 11.600650860065086, + "grad_norm": 6.078497409820557, + "learning_rate": 9.9750579975058e-06, + "loss": 0.0795, + "step": 24953 + }, + { + "epoch": 11.601115760111576, + "grad_norm": 6.165933132171631, + "learning_rate": 9.975056997505701e-06, + "loss": 0.1129, + "step": 24954 + }, + { + "epoch": 11.601580660158065, + "grad_norm": 6.309293270111084, + "learning_rate": 9.9750559975056e-06, + "loss": 0.1468, + "step": 24955 + }, + { + "epoch": 11.602045560204555, + "grad_norm": 6.429317474365234, + "learning_rate": 9.9750549975055e-06, + "loss": 0.0805, + "step": 24956 + }, + { + "epoch": 11.602510460251047, + "grad_norm": 3.7762820720672607, + "learning_rate": 9.975053997505401e-06, + "loss": 0.0376, + "step": 24957 + }, + { + "epoch": 11.602975360297537, + "grad_norm": 5.074337482452393, + "learning_rate": 9.9750529975053e-06, + "loss": 0.0776, + "step": 24958 + }, + { + "epoch": 11.603440260344026, + "grad_norm": 6.1812238693237305, + "learning_rate": 9.9750519975052e-06, + "loss": 0.0897, + "step": 24959 + }, + { + "epoch": 11.603905160390516, + "grad_norm": 3.5482258796691895, + "learning_rate": 9.9750509975051e-06, + "loss": 0.0168, + "step": 24960 + }, + { + "epoch": 11.604370060437006, + "grad_norm": 5.6657395362854, + "learning_rate": 9.975049997505e-06, + "loss": 0.0714, + "step": 24961 + }, + { + "epoch": 11.604834960483496, + "grad_norm": 4.351200580596924, + "learning_rate": 9.975048997504901e-06, + "loss": 0.0308, + "step": 24962 + }, + { + "epoch": 11.605299860529986, + "grad_norm": 6.007862567901611, + "learning_rate": 9.975047997504802e-06, + "loss": 0.0725, + "step": 24963 + }, + { + "epoch": 11.605764760576475, + "grad_norm": 3.93196964263916, + "learning_rate": 9.9750469975047e-06, + "loss": 0.0347, + "step": 24964 + }, + { + "epoch": 11.606229660622965, + "grad_norm": 5.516890525817871, + "learning_rate": 9.975045997504601e-06, + "loss": 0.1358, + "step": 24965 + }, + { + "epoch": 11.606694560669457, + "grad_norm": 3.6164872646331787, + "learning_rate": 9.9750449975045e-06, + "loss": 0.1058, + "step": 24966 + }, + { + "epoch": 11.607159460715947, + "grad_norm": 3.312559127807617, + "learning_rate": 9.975043997504401e-06, + "loss": 0.0256, + "step": 24967 + }, + { + "epoch": 11.607624360762436, + "grad_norm": 7.621626853942871, + "learning_rate": 9.9750429975043e-06, + "loss": 0.1177, + "step": 24968 + }, + { + "epoch": 11.608089260808926, + "grad_norm": 6.649054050445557, + "learning_rate": 9.9750419975042e-06, + "loss": 0.1164, + "step": 24969 + }, + { + "epoch": 11.608554160855416, + "grad_norm": 6.148209095001221, + "learning_rate": 9.975040997504101e-06, + "loss": 0.1008, + "step": 24970 + }, + { + "epoch": 11.609019060901906, + "grad_norm": 3.9242799282073975, + "learning_rate": 9.975039997504e-06, + "loss": 0.0232, + "step": 24971 + }, + { + "epoch": 11.609483960948396, + "grad_norm": 7.528539657592773, + "learning_rate": 9.975038997503901e-06, + "loss": 0.1344, + "step": 24972 + }, + { + "epoch": 11.609948860994885, + "grad_norm": 6.615054130554199, + "learning_rate": 9.9750379975038e-06, + "loss": 0.1888, + "step": 24973 + }, + { + "epoch": 11.610413761041377, + "grad_norm": 7.218775749206543, + "learning_rate": 9.9750369975037e-06, + "loss": 0.1102, + "step": 24974 + }, + { + "epoch": 11.610878661087867, + "grad_norm": 2.9439778327941895, + "learning_rate": 9.975035997503601e-06, + "loss": 0.014, + "step": 24975 + }, + { + "epoch": 11.611343561134357, + "grad_norm": 5.131994724273682, + "learning_rate": 9.975034997503502e-06, + "loss": 0.0648, + "step": 24976 + }, + { + "epoch": 11.611808461180846, + "grad_norm": 4.014848232269287, + "learning_rate": 9.9750339975034e-06, + "loss": 0.0391, + "step": 24977 + }, + { + "epoch": 11.612273361227336, + "grad_norm": 1.655579686164856, + "learning_rate": 9.975032997503301e-06, + "loss": 0.0123, + "step": 24978 + }, + { + "epoch": 11.612738261273826, + "grad_norm": 4.330716609954834, + "learning_rate": 9.9750319975032e-06, + "loss": 0.0582, + "step": 24979 + }, + { + "epoch": 11.613203161320316, + "grad_norm": 4.347203254699707, + "learning_rate": 9.9750309975031e-06, + "loss": 0.0298, + "step": 24980 + }, + { + "epoch": 11.613668061366806, + "grad_norm": 5.431432723999023, + "learning_rate": 9.975029997503e-06, + "loss": 0.1098, + "step": 24981 + }, + { + "epoch": 11.614132961413295, + "grad_norm": 7.130980968475342, + "learning_rate": 9.9750289975029e-06, + "loss": 0.1097, + "step": 24982 + }, + { + "epoch": 11.614597861459787, + "grad_norm": 3.2530596256256104, + "learning_rate": 9.975027997502801e-06, + "loss": 0.0312, + "step": 24983 + }, + { + "epoch": 11.615062761506277, + "grad_norm": 2.260870933532715, + "learning_rate": 9.9750269975027e-06, + "loss": 0.0172, + "step": 24984 + }, + { + "epoch": 11.615527661552767, + "grad_norm": 3.799083948135376, + "learning_rate": 9.975025997502601e-06, + "loss": 0.042, + "step": 24985 + }, + { + "epoch": 11.615992561599256, + "grad_norm": 6.180632591247559, + "learning_rate": 9.9750249975025e-06, + "loss": 0.1068, + "step": 24986 + }, + { + "epoch": 11.616457461645746, + "grad_norm": 3.2701833248138428, + "learning_rate": 9.9750239975024e-06, + "loss": 0.0404, + "step": 24987 + }, + { + "epoch": 11.616922361692236, + "grad_norm": 5.014498233795166, + "learning_rate": 9.975022997502301e-06, + "loss": 0.0568, + "step": 24988 + }, + { + "epoch": 11.617387261738726, + "grad_norm": 2.148442506790161, + "learning_rate": 9.9750219975022e-06, + "loss": 0.0174, + "step": 24989 + }, + { + "epoch": 11.617852161785216, + "grad_norm": 0.2400539070367813, + "learning_rate": 9.9750209975021e-06, + "loss": 0.0018, + "step": 24990 + }, + { + "epoch": 11.618317061831707, + "grad_norm": 0.9374943375587463, + "learning_rate": 9.975019997502e-06, + "loss": 0.0062, + "step": 24991 + }, + { + "epoch": 11.618781961878197, + "grad_norm": 1.518109679222107, + "learning_rate": 9.9750189975019e-06, + "loss": 0.0148, + "step": 24992 + }, + { + "epoch": 11.619246861924687, + "grad_norm": 11.938755989074707, + "learning_rate": 9.9750179975018e-06, + "loss": 0.2119, + "step": 24993 + }, + { + "epoch": 11.619711761971177, + "grad_norm": 4.036336421966553, + "learning_rate": 9.9750169975017e-06, + "loss": 0.0511, + "step": 24994 + }, + { + "epoch": 11.620176662017666, + "grad_norm": 1.5670658349990845, + "learning_rate": 9.9750159975016e-06, + "loss": 0.0105, + "step": 24995 + }, + { + "epoch": 11.620641562064156, + "grad_norm": 3.0487539768218994, + "learning_rate": 9.975014997501501e-06, + "loss": 0.0244, + "step": 24996 + }, + { + "epoch": 11.621106462110646, + "grad_norm": 8.1649751663208, + "learning_rate": 9.9750139975014e-06, + "loss": 0.1465, + "step": 24997 + }, + { + "epoch": 11.621571362157136, + "grad_norm": 10.139805793762207, + "learning_rate": 9.975012997501301e-06, + "loss": 0.091, + "step": 24998 + }, + { + "epoch": 11.622036262203626, + "grad_norm": 8.511774063110352, + "learning_rate": 9.9750119975012e-06, + "loss": 0.0589, + "step": 24999 + }, + { + "epoch": 11.622501162250117, + "grad_norm": 3.1564619541168213, + "learning_rate": 9.9750109975011e-06, + "loss": 0.0471, + "step": 25000 + }, + { + "epoch": 11.622966062296607, + "grad_norm": 3.6023850440979004, + "learning_rate": 9.975009997501001e-06, + "loss": 0.0202, + "step": 25001 + }, + { + "epoch": 11.623430962343097, + "grad_norm": 4.7390522956848145, + "learning_rate": 9.9750089975009e-06, + "loss": 0.0293, + "step": 25002 + }, + { + "epoch": 11.623895862389586, + "grad_norm": 4.692493438720703, + "learning_rate": 9.975007997500801e-06, + "loss": 0.0369, + "step": 25003 + }, + { + "epoch": 11.624360762436076, + "grad_norm": 6.118403434753418, + "learning_rate": 9.9750069975007e-06, + "loss": 0.0731, + "step": 25004 + }, + { + "epoch": 11.624825662482566, + "grad_norm": 5.497860908508301, + "learning_rate": 9.9750059975006e-06, + "loss": 0.0473, + "step": 25005 + }, + { + "epoch": 11.625290562529056, + "grad_norm": 3.7006289958953857, + "learning_rate": 9.9750049975005e-06, + "loss": 0.0213, + "step": 25006 + }, + { + "epoch": 11.625755462575546, + "grad_norm": 5.560757160186768, + "learning_rate": 9.9750039975004e-06, + "loss": 0.0587, + "step": 25007 + }, + { + "epoch": 11.626220362622036, + "grad_norm": 3.663790464401245, + "learning_rate": 9.9750029975003e-06, + "loss": 0.0558, + "step": 25008 + }, + { + "epoch": 11.626685262668527, + "grad_norm": 3.6055686473846436, + "learning_rate": 9.975001997500201e-06, + "loss": 0.0504, + "step": 25009 + }, + { + "epoch": 11.627150162715017, + "grad_norm": 3.2127161026000977, + "learning_rate": 9.9750009975001e-06, + "loss": 0.0239, + "step": 25010 + }, + { + "epoch": 11.627615062761507, + "grad_norm": 5.535305500030518, + "learning_rate": 9.9749999975e-06, + "loss": 0.0779, + "step": 25011 + }, + { + "epoch": 11.628079962807996, + "grad_norm": 3.1218395233154297, + "learning_rate": 9.9749989974999e-06, + "loss": 0.0337, + "step": 25012 + }, + { + "epoch": 11.628544862854486, + "grad_norm": 7.0492048263549805, + "learning_rate": 9.9749979974998e-06, + "loss": 0.0497, + "step": 25013 + }, + { + "epoch": 11.629009762900976, + "grad_norm": 0.37373635172843933, + "learning_rate": 9.974996997499701e-06, + "loss": 0.0041, + "step": 25014 + }, + { + "epoch": 11.629474662947466, + "grad_norm": 1.3443834781646729, + "learning_rate": 9.9749959974996e-06, + "loss": 0.0102, + "step": 25015 + }, + { + "epoch": 11.629939562993956, + "grad_norm": 4.5926337242126465, + "learning_rate": 9.974994997499501e-06, + "loss": 0.044, + "step": 25016 + }, + { + "epoch": 11.630404463040446, + "grad_norm": 2.6803550720214844, + "learning_rate": 9.9749939974994e-06, + "loss": 0.0283, + "step": 25017 + }, + { + "epoch": 11.630869363086937, + "grad_norm": 8.306425094604492, + "learning_rate": 9.9749929974993e-06, + "loss": 0.1495, + "step": 25018 + }, + { + "epoch": 11.631334263133427, + "grad_norm": 4.586706161499023, + "learning_rate": 9.974991997499201e-06, + "loss": 0.0747, + "step": 25019 + }, + { + "epoch": 11.631799163179917, + "grad_norm": 3.403925657272339, + "learning_rate": 9.9749909974991e-06, + "loss": 0.0237, + "step": 25020 + }, + { + "epoch": 11.632264063226406, + "grad_norm": 1.4114834070205688, + "learning_rate": 9.974989997499001e-06, + "loss": 0.0134, + "step": 25021 + }, + { + "epoch": 11.632728963272896, + "grad_norm": 5.741433143615723, + "learning_rate": 9.9749889974989e-06, + "loss": 0.0767, + "step": 25022 + }, + { + "epoch": 11.633193863319386, + "grad_norm": 5.465295314788818, + "learning_rate": 9.9749879974988e-06, + "loss": 0.1118, + "step": 25023 + }, + { + "epoch": 11.633658763365876, + "grad_norm": 2.8890931606292725, + "learning_rate": 9.9749869974987e-06, + "loss": 0.093, + "step": 25024 + }, + { + "epoch": 11.634123663412366, + "grad_norm": 3.5464394092559814, + "learning_rate": 9.9749859974986e-06, + "loss": 0.0225, + "step": 25025 + }, + { + "epoch": 11.634588563458856, + "grad_norm": 2.898759126663208, + "learning_rate": 9.9749849974985e-06, + "loss": 0.0464, + "step": 25026 + }, + { + "epoch": 11.635053463505347, + "grad_norm": 5.382779121398926, + "learning_rate": 9.974983997498401e-06, + "loss": 0.019, + "step": 25027 + }, + { + "epoch": 11.635518363551837, + "grad_norm": 8.426685333251953, + "learning_rate": 9.9749829974983e-06, + "loss": 0.1046, + "step": 25028 + }, + { + "epoch": 11.635983263598327, + "grad_norm": 6.385907173156738, + "learning_rate": 9.974981997498201e-06, + "loss": 0.1563, + "step": 25029 + }, + { + "epoch": 11.636448163644816, + "grad_norm": 4.539150714874268, + "learning_rate": 9.9749809974981e-06, + "loss": 0.0717, + "step": 25030 + }, + { + "epoch": 11.636913063691306, + "grad_norm": 4.014517784118652, + "learning_rate": 9.974979997498e-06, + "loss": 0.0502, + "step": 25031 + }, + { + "epoch": 11.637377963737796, + "grad_norm": 1.238329291343689, + "learning_rate": 9.974978997497901e-06, + "loss": 0.0057, + "step": 25032 + }, + { + "epoch": 11.637842863784286, + "grad_norm": 4.7049641609191895, + "learning_rate": 9.9749779974978e-06, + "loss": 0.0923, + "step": 25033 + }, + { + "epoch": 11.638307763830776, + "grad_norm": 3.621314287185669, + "learning_rate": 9.974976997497701e-06, + "loss": 0.0523, + "step": 25034 + }, + { + "epoch": 11.638772663877266, + "grad_norm": 4.5493669509887695, + "learning_rate": 9.9749759974976e-06, + "loss": 0.0587, + "step": 25035 + }, + { + "epoch": 11.639237563923757, + "grad_norm": 4.436552047729492, + "learning_rate": 9.9749749974975e-06, + "loss": 0.0676, + "step": 25036 + }, + { + "epoch": 11.639702463970247, + "grad_norm": 1.6966663599014282, + "learning_rate": 9.9749739974974e-06, + "loss": 0.0103, + "step": 25037 + }, + { + "epoch": 11.640167364016737, + "grad_norm": 8.51181411743164, + "learning_rate": 9.9749729974973e-06, + "loss": 0.2289, + "step": 25038 + }, + { + "epoch": 11.640632264063226, + "grad_norm": 4.921104907989502, + "learning_rate": 9.974971997497201e-06, + "loss": 0.0725, + "step": 25039 + }, + { + "epoch": 11.641097164109716, + "grad_norm": 8.902563095092773, + "learning_rate": 9.974970997497102e-06, + "loss": 0.0772, + "step": 25040 + }, + { + "epoch": 11.641562064156206, + "grad_norm": 2.562830924987793, + "learning_rate": 9.974969997497e-06, + "loss": 0.0187, + "step": 25041 + }, + { + "epoch": 11.642026964202696, + "grad_norm": 3.275892734527588, + "learning_rate": 9.974968997496901e-06, + "loss": 0.0574, + "step": 25042 + }, + { + "epoch": 11.642491864249186, + "grad_norm": 2.873439311981201, + "learning_rate": 9.9749679974968e-06, + "loss": 0.0385, + "step": 25043 + }, + { + "epoch": 11.642956764295675, + "grad_norm": 5.058699131011963, + "learning_rate": 9.9749669974967e-06, + "loss": 0.0664, + "step": 25044 + }, + { + "epoch": 11.643421664342167, + "grad_norm": 2.974339008331299, + "learning_rate": 9.974965997496601e-06, + "loss": 0.0514, + "step": 25045 + }, + { + "epoch": 11.643886564388657, + "grad_norm": 6.62168550491333, + "learning_rate": 9.9749649974965e-06, + "loss": 0.0965, + "step": 25046 + }, + { + "epoch": 11.644351464435147, + "grad_norm": 1.996811032295227, + "learning_rate": 9.974963997496401e-06, + "loss": 0.0142, + "step": 25047 + }, + { + "epoch": 11.644816364481636, + "grad_norm": 0.8436651229858398, + "learning_rate": 9.9749629974963e-06, + "loss": 0.0081, + "step": 25048 + }, + { + "epoch": 11.645281264528126, + "grad_norm": 5.510472297668457, + "learning_rate": 9.9749619974962e-06, + "loss": 0.08, + "step": 25049 + }, + { + "epoch": 11.645746164574616, + "grad_norm": 3.2485084533691406, + "learning_rate": 9.9749609974961e-06, + "loss": 0.0362, + "step": 25050 + }, + { + "epoch": 11.646211064621106, + "grad_norm": 3.2114570140838623, + "learning_rate": 9.974959997496e-06, + "loss": 0.0279, + "step": 25051 + }, + { + "epoch": 11.646675964667596, + "grad_norm": 3.269329786300659, + "learning_rate": 9.974958997495901e-06, + "loss": 0.022, + "step": 25052 + }, + { + "epoch": 11.647140864714087, + "grad_norm": 2.8498640060424805, + "learning_rate": 9.974957997495802e-06, + "loss": 0.0491, + "step": 25053 + }, + { + "epoch": 11.647605764760577, + "grad_norm": 1.2375493049621582, + "learning_rate": 9.9749569974957e-06, + "loss": 0.0063, + "step": 25054 + }, + { + "epoch": 11.648070664807067, + "grad_norm": 1.3984317779541016, + "learning_rate": 9.9749559974956e-06, + "loss": 0.0083, + "step": 25055 + }, + { + "epoch": 11.648535564853557, + "grad_norm": 4.852440357208252, + "learning_rate": 9.9749549974955e-06, + "loss": 0.0559, + "step": 25056 + }, + { + "epoch": 11.649000464900046, + "grad_norm": 4.957413673400879, + "learning_rate": 9.974953997495401e-06, + "loss": 0.083, + "step": 25057 + }, + { + "epoch": 11.649465364946536, + "grad_norm": 2.851670980453491, + "learning_rate": 9.974952997495302e-06, + "loss": 0.0504, + "step": 25058 + }, + { + "epoch": 11.649930264993026, + "grad_norm": 5.119807720184326, + "learning_rate": 9.9749519974952e-06, + "loss": 0.0653, + "step": 25059 + }, + { + "epoch": 11.650395165039516, + "grad_norm": 2.2199840545654297, + "learning_rate": 9.974950997495101e-06, + "loss": 0.0224, + "step": 25060 + }, + { + "epoch": 11.650860065086006, + "grad_norm": 3.214571237564087, + "learning_rate": 9.974949997495e-06, + "loss": 0.0407, + "step": 25061 + }, + { + "epoch": 11.651324965132497, + "grad_norm": 2.8484153747558594, + "learning_rate": 9.9749489974949e-06, + "loss": 0.0284, + "step": 25062 + }, + { + "epoch": 11.651789865178987, + "grad_norm": 7.142725467681885, + "learning_rate": 9.9749479974948e-06, + "loss": 0.1334, + "step": 25063 + }, + { + "epoch": 11.652254765225477, + "grad_norm": 6.871109485626221, + "learning_rate": 9.9749469974947e-06, + "loss": 0.1108, + "step": 25064 + }, + { + "epoch": 11.652719665271967, + "grad_norm": 1.6105295419692993, + "learning_rate": 9.974945997494601e-06, + "loss": 0.0156, + "step": 25065 + }, + { + "epoch": 11.653184565318456, + "grad_norm": 5.534423828125, + "learning_rate": 9.9749449974945e-06, + "loss": 0.0293, + "step": 25066 + }, + { + "epoch": 11.653649465364946, + "grad_norm": 8.032090187072754, + "learning_rate": 9.9749439974944e-06, + "loss": 0.0957, + "step": 25067 + }, + { + "epoch": 11.654114365411436, + "grad_norm": 2.4957094192504883, + "learning_rate": 9.9749429974943e-06, + "loss": 0.0244, + "step": 25068 + }, + { + "epoch": 11.654579265457926, + "grad_norm": 4.19886589050293, + "learning_rate": 9.9749419974942e-06, + "loss": 0.0302, + "step": 25069 + }, + { + "epoch": 11.655044165504417, + "grad_norm": 4.118383884429932, + "learning_rate": 9.974940997494101e-06, + "loss": 0.0658, + "step": 25070 + }, + { + "epoch": 11.655509065550907, + "grad_norm": 3.813436269760132, + "learning_rate": 9.974939997494002e-06, + "loss": 0.0481, + "step": 25071 + }, + { + "epoch": 11.655973965597397, + "grad_norm": 1.046667218208313, + "learning_rate": 9.9749389974939e-06, + "loss": 0.0074, + "step": 25072 + }, + { + "epoch": 11.656438865643887, + "grad_norm": 4.099620819091797, + "learning_rate": 9.974937997493801e-06, + "loss": 0.0616, + "step": 25073 + }, + { + "epoch": 11.656903765690377, + "grad_norm": 6.527751445770264, + "learning_rate": 9.9749369974937e-06, + "loss": 0.0641, + "step": 25074 + }, + { + "epoch": 11.657368665736866, + "grad_norm": 5.9705119132995605, + "learning_rate": 9.974935997493601e-06, + "loss": 0.1055, + "step": 25075 + }, + { + "epoch": 11.657833565783356, + "grad_norm": 1.0671169757843018, + "learning_rate": 9.9749349974935e-06, + "loss": 0.0075, + "step": 25076 + }, + { + "epoch": 11.658298465829846, + "grad_norm": 2.271469831466675, + "learning_rate": 9.9749339974934e-06, + "loss": 0.0198, + "step": 25077 + }, + { + "epoch": 11.658763365876336, + "grad_norm": 9.219953536987305, + "learning_rate": 9.974932997493301e-06, + "loss": 0.2431, + "step": 25078 + }, + { + "epoch": 11.659228265922827, + "grad_norm": 5.390213966369629, + "learning_rate": 9.9749319974932e-06, + "loss": 0.0713, + "step": 25079 + }, + { + "epoch": 11.659693165969317, + "grad_norm": 3.0500636100769043, + "learning_rate": 9.9749309974931e-06, + "loss": 0.0182, + "step": 25080 + }, + { + "epoch": 11.660158066015807, + "grad_norm": 3.4819629192352295, + "learning_rate": 9.974929997493e-06, + "loss": 0.0491, + "step": 25081 + }, + { + "epoch": 11.660622966062297, + "grad_norm": 3.5833473205566406, + "learning_rate": 9.9749289974929e-06, + "loss": 0.0207, + "step": 25082 + }, + { + "epoch": 11.661087866108787, + "grad_norm": 6.376991271972656, + "learning_rate": 9.974927997492801e-06, + "loss": 0.0672, + "step": 25083 + }, + { + "epoch": 11.661552766155276, + "grad_norm": 2.5363636016845703, + "learning_rate": 9.974926997492702e-06, + "loss": 0.0145, + "step": 25084 + }, + { + "epoch": 11.662017666201766, + "grad_norm": 4.776394367218018, + "learning_rate": 9.9749259974926e-06, + "loss": 0.0552, + "step": 25085 + }, + { + "epoch": 11.662482566248256, + "grad_norm": 2.7707531452178955, + "learning_rate": 9.974924997492501e-06, + "loss": 0.0372, + "step": 25086 + }, + { + "epoch": 11.662947466294746, + "grad_norm": 4.065465450286865, + "learning_rate": 9.9749239974924e-06, + "loss": 0.023, + "step": 25087 + }, + { + "epoch": 11.663412366341237, + "grad_norm": 6.547696590423584, + "learning_rate": 9.974922997492301e-06, + "loss": 0.1195, + "step": 25088 + }, + { + "epoch": 11.663877266387727, + "grad_norm": 5.033039093017578, + "learning_rate": 9.9749219974922e-06, + "loss": 0.0758, + "step": 25089 + }, + { + "epoch": 11.664342166434217, + "grad_norm": 6.658321380615234, + "learning_rate": 9.9749209974921e-06, + "loss": 0.0791, + "step": 25090 + }, + { + "epoch": 11.664807066480707, + "grad_norm": 5.315128326416016, + "learning_rate": 9.974919997492001e-06, + "loss": 0.0364, + "step": 25091 + }, + { + "epoch": 11.665271966527197, + "grad_norm": 4.98252010345459, + "learning_rate": 9.9749189974919e-06, + "loss": 0.0804, + "step": 25092 + }, + { + "epoch": 11.665736866573686, + "grad_norm": 2.9715967178344727, + "learning_rate": 9.9749179974918e-06, + "loss": 0.0156, + "step": 25093 + }, + { + "epoch": 11.666201766620176, + "grad_norm": 3.843377113342285, + "learning_rate": 9.9749169974917e-06, + "loss": 0.0377, + "step": 25094 + }, + { + "epoch": 11.666666666666666, + "grad_norm": 1.2696269750595093, + "learning_rate": 9.9749159974916e-06, + "loss": 0.0114, + "step": 25095 + }, + { + "epoch": 11.667131566713156, + "grad_norm": 6.887218952178955, + "learning_rate": 9.974914997491501e-06, + "loss": 0.0906, + "step": 25096 + }, + { + "epoch": 11.667596466759647, + "grad_norm": 3.0483932495117188, + "learning_rate": 9.974913997491402e-06, + "loss": 0.0374, + "step": 25097 + }, + { + "epoch": 11.668061366806137, + "grad_norm": 4.481689453125, + "learning_rate": 9.9749129974913e-06, + "loss": 0.0628, + "step": 25098 + }, + { + "epoch": 11.668526266852627, + "grad_norm": 2.7158970832824707, + "learning_rate": 9.9749119974912e-06, + "loss": 0.0225, + "step": 25099 + }, + { + "epoch": 11.668991166899117, + "grad_norm": 1.3573923110961914, + "learning_rate": 9.9749109974911e-06, + "loss": 0.0128, + "step": 25100 + }, + { + "epoch": 11.669456066945607, + "grad_norm": 2.27182674407959, + "learning_rate": 9.974909997491001e-06, + "loss": 0.0145, + "step": 25101 + }, + { + "epoch": 11.669920966992096, + "grad_norm": 12.024988174438477, + "learning_rate": 9.9749089974909e-06, + "loss": 0.1661, + "step": 25102 + }, + { + "epoch": 11.670385867038586, + "grad_norm": 6.816710948944092, + "learning_rate": 9.9749079974908e-06, + "loss": 0.0764, + "step": 25103 + }, + { + "epoch": 11.670850767085076, + "grad_norm": 9.970630645751953, + "learning_rate": 9.974906997490701e-06, + "loss": 0.1803, + "step": 25104 + }, + { + "epoch": 11.671315667131566, + "grad_norm": 8.985674858093262, + "learning_rate": 9.9749059974906e-06, + "loss": 0.0694, + "step": 25105 + }, + { + "epoch": 11.671780567178057, + "grad_norm": 2.093766927719116, + "learning_rate": 9.974904997490501e-06, + "loss": 0.0094, + "step": 25106 + }, + { + "epoch": 11.672245467224547, + "grad_norm": 3.9489328861236572, + "learning_rate": 9.9749039974904e-06, + "loss": 0.0455, + "step": 25107 + }, + { + "epoch": 11.672710367271037, + "grad_norm": 4.648226737976074, + "learning_rate": 9.9749029974903e-06, + "loss": 0.0868, + "step": 25108 + }, + { + "epoch": 11.673175267317527, + "grad_norm": 5.325974464416504, + "learning_rate": 9.974901997490201e-06, + "loss": 0.06, + "step": 25109 + }, + { + "epoch": 11.673640167364017, + "grad_norm": 8.413858413696289, + "learning_rate": 9.9749009974901e-06, + "loss": 0.2076, + "step": 25110 + }, + { + "epoch": 11.674105067410506, + "grad_norm": 1.9489877223968506, + "learning_rate": 9.97489999749e-06, + "loss": 0.0119, + "step": 25111 + }, + { + "epoch": 11.674569967456996, + "grad_norm": 4.419486045837402, + "learning_rate": 9.9748989974899e-06, + "loss": 0.0647, + "step": 25112 + }, + { + "epoch": 11.675034867503486, + "grad_norm": 5.09013032913208, + "learning_rate": 9.9748979974898e-06, + "loss": 0.0566, + "step": 25113 + }, + { + "epoch": 11.675499767549976, + "grad_norm": 5.107567310333252, + "learning_rate": 9.974896997489701e-06, + "loss": 0.0699, + "step": 25114 + }, + { + "epoch": 11.675964667596467, + "grad_norm": 6.004267692565918, + "learning_rate": 9.974895997489602e-06, + "loss": 0.0813, + "step": 25115 + }, + { + "epoch": 11.676429567642957, + "grad_norm": 2.9784998893737793, + "learning_rate": 9.9748949974895e-06, + "loss": 0.0259, + "step": 25116 + }, + { + "epoch": 11.676894467689447, + "grad_norm": 9.844196319580078, + "learning_rate": 9.974893997489401e-06, + "loss": 0.1654, + "step": 25117 + }, + { + "epoch": 11.677359367735937, + "grad_norm": 4.512913703918457, + "learning_rate": 9.9748929974893e-06, + "loss": 0.057, + "step": 25118 + }, + { + "epoch": 11.677824267782427, + "grad_norm": 5.210183143615723, + "learning_rate": 9.974891997489201e-06, + "loss": 0.0913, + "step": 25119 + }, + { + "epoch": 11.678289167828916, + "grad_norm": 3.2358360290527344, + "learning_rate": 9.9748909974891e-06, + "loss": 0.0292, + "step": 25120 + }, + { + "epoch": 11.678754067875406, + "grad_norm": 2.198463201522827, + "learning_rate": 9.974889997489e-06, + "loss": 0.0135, + "step": 25121 + }, + { + "epoch": 11.679218967921896, + "grad_norm": 3.1225290298461914, + "learning_rate": 9.974888997488901e-06, + "loss": 0.0276, + "step": 25122 + }, + { + "epoch": 11.679683867968386, + "grad_norm": 3.3947622776031494, + "learning_rate": 9.9748879974888e-06, + "loss": 0.0223, + "step": 25123 + }, + { + "epoch": 11.680148768014877, + "grad_norm": 4.19690465927124, + "learning_rate": 9.974886997488701e-06, + "loss": 0.0476, + "step": 25124 + }, + { + "epoch": 11.680613668061367, + "grad_norm": 3.387826681137085, + "learning_rate": 9.9748859974886e-06, + "loss": 0.032, + "step": 25125 + }, + { + "epoch": 11.681078568107857, + "grad_norm": 5.388323783874512, + "learning_rate": 9.9748849974885e-06, + "loss": 0.0877, + "step": 25126 + }, + { + "epoch": 11.681543468154347, + "grad_norm": 3.9545533657073975, + "learning_rate": 9.974883997488401e-06, + "loss": 0.049, + "step": 25127 + }, + { + "epoch": 11.682008368200837, + "grad_norm": 3.659921884536743, + "learning_rate": 9.974882997488302e-06, + "loss": 0.0398, + "step": 25128 + }, + { + "epoch": 11.682473268247326, + "grad_norm": 9.247089385986328, + "learning_rate": 9.9748819974882e-06, + "loss": 0.077, + "step": 25129 + }, + { + "epoch": 11.682938168293816, + "grad_norm": 5.199463367462158, + "learning_rate": 9.9748809974881e-06, + "loss": 0.0618, + "step": 25130 + }, + { + "epoch": 11.683403068340306, + "grad_norm": 6.102962970733643, + "learning_rate": 9.974879997488e-06, + "loss": 0.096, + "step": 25131 + }, + { + "epoch": 11.683867968386798, + "grad_norm": 5.772517681121826, + "learning_rate": 9.9748789974879e-06, + "loss": 0.0992, + "step": 25132 + }, + { + "epoch": 11.684332868433287, + "grad_norm": 3.9802498817443848, + "learning_rate": 9.9748779974878e-06, + "loss": 0.0354, + "step": 25133 + }, + { + "epoch": 11.684797768479777, + "grad_norm": 6.888803958892822, + "learning_rate": 9.9748769974877e-06, + "loss": 0.0712, + "step": 25134 + }, + { + "epoch": 11.685262668526267, + "grad_norm": 3.853496551513672, + "learning_rate": 9.974875997487601e-06, + "loss": 0.0393, + "step": 25135 + }, + { + "epoch": 11.685727568572757, + "grad_norm": 8.215755462646484, + "learning_rate": 9.9748749974875e-06, + "loss": 0.1241, + "step": 25136 + }, + { + "epoch": 11.686192468619247, + "grad_norm": 6.332451820373535, + "learning_rate": 9.974873997487401e-06, + "loss": 0.1095, + "step": 25137 + }, + { + "epoch": 11.686657368665736, + "grad_norm": 2.846403121948242, + "learning_rate": 9.9748729974873e-06, + "loss": 0.0264, + "step": 25138 + }, + { + "epoch": 11.687122268712226, + "grad_norm": 4.5125837326049805, + "learning_rate": 9.9748719974872e-06, + "loss": 0.0302, + "step": 25139 + }, + { + "epoch": 11.687587168758716, + "grad_norm": 4.00011682510376, + "learning_rate": 9.974870997487101e-06, + "loss": 0.0486, + "step": 25140 + }, + { + "epoch": 11.688052068805208, + "grad_norm": 4.927282333374023, + "learning_rate": 9.974869997487e-06, + "loss": 0.0417, + "step": 25141 + }, + { + "epoch": 11.688516968851697, + "grad_norm": 5.526995658874512, + "learning_rate": 9.974868997486901e-06, + "loss": 0.0563, + "step": 25142 + }, + { + "epoch": 11.688981868898187, + "grad_norm": 4.126929759979248, + "learning_rate": 9.9748679974868e-06, + "loss": 0.0674, + "step": 25143 + }, + { + "epoch": 11.689446768944677, + "grad_norm": 5.046952247619629, + "learning_rate": 9.9748669974867e-06, + "loss": 0.0845, + "step": 25144 + }, + { + "epoch": 11.689911668991167, + "grad_norm": 2.318016290664673, + "learning_rate": 9.9748659974866e-06, + "loss": 0.0204, + "step": 25145 + }, + { + "epoch": 11.690376569037657, + "grad_norm": 4.011483669281006, + "learning_rate": 9.9748649974865e-06, + "loss": 0.0454, + "step": 25146 + }, + { + "epoch": 11.690841469084146, + "grad_norm": 5.863938808441162, + "learning_rate": 9.9748639974864e-06, + "loss": 0.0718, + "step": 25147 + }, + { + "epoch": 11.691306369130636, + "grad_norm": 6.41650915145874, + "learning_rate": 9.974862997486301e-06, + "loss": 0.0571, + "step": 25148 + }, + { + "epoch": 11.691771269177128, + "grad_norm": 8.992515563964844, + "learning_rate": 9.9748619974862e-06, + "loss": 0.1713, + "step": 25149 + }, + { + "epoch": 11.692236169223618, + "grad_norm": 3.3124566078186035, + "learning_rate": 9.974860997486101e-06, + "loss": 0.0343, + "step": 25150 + }, + { + "epoch": 11.692701069270107, + "grad_norm": 3.4536778926849365, + "learning_rate": 9.974859997486e-06, + "loss": 0.0358, + "step": 25151 + }, + { + "epoch": 11.693165969316597, + "grad_norm": 4.386656284332275, + "learning_rate": 9.9748589974859e-06, + "loss": 0.0432, + "step": 25152 + }, + { + "epoch": 11.693630869363087, + "grad_norm": 5.1525187492370605, + "learning_rate": 9.974857997485801e-06, + "loss": 0.0189, + "step": 25153 + }, + { + "epoch": 11.694095769409577, + "grad_norm": 4.279004096984863, + "learning_rate": 9.9748569974857e-06, + "loss": 0.0525, + "step": 25154 + }, + { + "epoch": 11.694560669456067, + "grad_norm": 2.7504100799560547, + "learning_rate": 9.974855997485601e-06, + "loss": 0.0304, + "step": 25155 + }, + { + "epoch": 11.695025569502556, + "grad_norm": 4.325756549835205, + "learning_rate": 9.9748549974855e-06, + "loss": 0.0834, + "step": 25156 + }, + { + "epoch": 11.695490469549046, + "grad_norm": 1.088666558265686, + "learning_rate": 9.9748539974854e-06, + "loss": 0.0072, + "step": 25157 + }, + { + "epoch": 11.695955369595538, + "grad_norm": 3.23195743560791, + "learning_rate": 9.9748529974853e-06, + "loss": 0.0244, + "step": 25158 + }, + { + "epoch": 11.696420269642028, + "grad_norm": 3.6333794593811035, + "learning_rate": 9.9748519974852e-06, + "loss": 0.0271, + "step": 25159 + }, + { + "epoch": 11.696885169688517, + "grad_norm": 5.752721309661865, + "learning_rate": 9.974850997485101e-06, + "loss": 0.0543, + "step": 25160 + }, + { + "epoch": 11.697350069735007, + "grad_norm": 2.6938703060150146, + "learning_rate": 9.974849997485002e-06, + "loss": 0.0215, + "step": 25161 + }, + { + "epoch": 11.697814969781497, + "grad_norm": 5.119892597198486, + "learning_rate": 9.9748489974849e-06, + "loss": 0.0509, + "step": 25162 + }, + { + "epoch": 11.698279869827987, + "grad_norm": 4.113815784454346, + "learning_rate": 9.9748479974848e-06, + "loss": 0.0784, + "step": 25163 + }, + { + "epoch": 11.698744769874477, + "grad_norm": 2.727135181427002, + "learning_rate": 9.9748469974847e-06, + "loss": 0.0224, + "step": 25164 + }, + { + "epoch": 11.699209669920966, + "grad_norm": 1.7918877601623535, + "learning_rate": 9.9748459974846e-06, + "loss": 0.0109, + "step": 25165 + }, + { + "epoch": 11.699674569967456, + "grad_norm": 5.987631320953369, + "learning_rate": 9.974844997484501e-06, + "loss": 0.0505, + "step": 25166 + }, + { + "epoch": 11.700139470013948, + "grad_norm": 3.703524351119995, + "learning_rate": 9.9748439974844e-06, + "loss": 0.0547, + "step": 25167 + }, + { + "epoch": 11.700604370060438, + "grad_norm": 5.161300182342529, + "learning_rate": 9.974842997484301e-06, + "loss": 0.027, + "step": 25168 + }, + { + "epoch": 11.701069270106927, + "grad_norm": 4.5668182373046875, + "learning_rate": 9.9748419974842e-06, + "loss": 0.0486, + "step": 25169 + }, + { + "epoch": 11.701534170153417, + "grad_norm": 10.578994750976562, + "learning_rate": 9.9748409974841e-06, + "loss": 0.1095, + "step": 25170 + }, + { + "epoch": 11.701999070199907, + "grad_norm": 5.279341697692871, + "learning_rate": 9.974839997484e-06, + "loss": 0.0678, + "step": 25171 + }, + { + "epoch": 11.702463970246397, + "grad_norm": 3.3459572792053223, + "learning_rate": 9.9748389974839e-06, + "loss": 0.0367, + "step": 25172 + }, + { + "epoch": 11.702928870292887, + "grad_norm": 3.6157073974609375, + "learning_rate": 9.974837997483801e-06, + "loss": 0.0282, + "step": 25173 + }, + { + "epoch": 11.703393770339376, + "grad_norm": 2.901323080062866, + "learning_rate": 9.9748369974837e-06, + "loss": 0.0227, + "step": 25174 + }, + { + "epoch": 11.703858670385866, + "grad_norm": 4.535750865936279, + "learning_rate": 9.9748359974836e-06, + "loss": 0.0579, + "step": 25175 + }, + { + "epoch": 11.704323570432358, + "grad_norm": 9.690367698669434, + "learning_rate": 9.9748349974835e-06, + "loss": 0.1718, + "step": 25176 + }, + { + "epoch": 11.704788470478848, + "grad_norm": 2.4881484508514404, + "learning_rate": 9.9748339974834e-06, + "loss": 0.0162, + "step": 25177 + }, + { + "epoch": 11.705253370525337, + "grad_norm": 2.2608628273010254, + "learning_rate": 9.9748329974833e-06, + "loss": 0.0179, + "step": 25178 + }, + { + "epoch": 11.705718270571827, + "grad_norm": 3.9391493797302246, + "learning_rate": 9.974831997483201e-06, + "loss": 0.0418, + "step": 25179 + }, + { + "epoch": 11.706183170618317, + "grad_norm": 5.731545925140381, + "learning_rate": 9.9748309974831e-06, + "loss": 0.0243, + "step": 25180 + }, + { + "epoch": 11.706648070664807, + "grad_norm": 6.859394550323486, + "learning_rate": 9.974829997483001e-06, + "loss": 0.1109, + "step": 25181 + }, + { + "epoch": 11.707112970711297, + "grad_norm": 5.420582294464111, + "learning_rate": 9.9748289974829e-06, + "loss": 0.0512, + "step": 25182 + }, + { + "epoch": 11.707577870757786, + "grad_norm": 3.1907122135162354, + "learning_rate": 9.9748279974828e-06, + "loss": 0.0282, + "step": 25183 + }, + { + "epoch": 11.708042770804276, + "grad_norm": 5.14958381652832, + "learning_rate": 9.974826997482701e-06, + "loss": 0.0737, + "step": 25184 + }, + { + "epoch": 11.708507670850768, + "grad_norm": 5.486836910247803, + "learning_rate": 9.9748259974826e-06, + "loss": 0.0968, + "step": 25185 + }, + { + "epoch": 11.708972570897258, + "grad_norm": 8.653276443481445, + "learning_rate": 9.974824997482501e-06, + "loss": 0.0964, + "step": 25186 + }, + { + "epoch": 11.709437470943747, + "grad_norm": 4.045632362365723, + "learning_rate": 9.9748239974824e-06, + "loss": 0.0527, + "step": 25187 + }, + { + "epoch": 11.709902370990237, + "grad_norm": 7.148333549499512, + "learning_rate": 9.9748229974823e-06, + "loss": 0.0683, + "step": 25188 + }, + { + "epoch": 11.710367271036727, + "grad_norm": 7.05297327041626, + "learning_rate": 9.9748219974822e-06, + "loss": 0.0897, + "step": 25189 + }, + { + "epoch": 11.710832171083217, + "grad_norm": 2.561802387237549, + "learning_rate": 9.9748209974821e-06, + "loss": 0.0175, + "step": 25190 + }, + { + "epoch": 11.711297071129707, + "grad_norm": 9.449872016906738, + "learning_rate": 9.974819997482001e-06, + "loss": 0.0706, + "step": 25191 + }, + { + "epoch": 11.711761971176196, + "grad_norm": 1.0245907306671143, + "learning_rate": 9.974818997481902e-06, + "loss": 0.0077, + "step": 25192 + }, + { + "epoch": 11.712226871222686, + "grad_norm": 7.578227996826172, + "learning_rate": 9.9748179974818e-06, + "loss": 0.1444, + "step": 25193 + }, + { + "epoch": 11.712691771269178, + "grad_norm": 5.804296016693115, + "learning_rate": 9.974816997481701e-06, + "loss": 0.0916, + "step": 25194 + }, + { + "epoch": 11.713156671315668, + "grad_norm": 2.707858085632324, + "learning_rate": 9.9748159974816e-06, + "loss": 0.0197, + "step": 25195 + }, + { + "epoch": 11.713621571362157, + "grad_norm": 5.042569160461426, + "learning_rate": 9.9748149974815e-06, + "loss": 0.0571, + "step": 25196 + }, + { + "epoch": 11.714086471408647, + "grad_norm": 7.328393459320068, + "learning_rate": 9.974813997481401e-06, + "loss": 0.028, + "step": 25197 + }, + { + "epoch": 11.714551371455137, + "grad_norm": 4.384845733642578, + "learning_rate": 9.9748129974813e-06, + "loss": 0.0328, + "step": 25198 + }, + { + "epoch": 11.715016271501627, + "grad_norm": 5.804147243499756, + "learning_rate": 9.974811997481201e-06, + "loss": 0.044, + "step": 25199 + }, + { + "epoch": 11.715481171548117, + "grad_norm": 2.616743564605713, + "learning_rate": 9.9748109974811e-06, + "loss": 0.0304, + "step": 25200 + }, + { + "epoch": 11.715946071594606, + "grad_norm": 2.402991771697998, + "learning_rate": 9.974809997481e-06, + "loss": 0.0222, + "step": 25201 + }, + { + "epoch": 11.716410971641096, + "grad_norm": 2.525550365447998, + "learning_rate": 9.9748089974809e-06, + "loss": 0.023, + "step": 25202 + }, + { + "epoch": 11.716875871687588, + "grad_norm": 0.9436042904853821, + "learning_rate": 9.9748079974808e-06, + "loss": 0.0065, + "step": 25203 + }, + { + "epoch": 11.717340771734078, + "grad_norm": 7.025494575500488, + "learning_rate": 9.974806997480701e-06, + "loss": 0.1502, + "step": 25204 + }, + { + "epoch": 11.717805671780567, + "grad_norm": 3.6330349445343018, + "learning_rate": 9.974805997480602e-06, + "loss": 0.0491, + "step": 25205 + }, + { + "epoch": 11.718270571827057, + "grad_norm": 3.8323488235473633, + "learning_rate": 9.9748049974805e-06, + "loss": 0.0498, + "step": 25206 + }, + { + "epoch": 11.718735471873547, + "grad_norm": 2.8405544757843018, + "learning_rate": 9.9748039974804e-06, + "loss": 0.031, + "step": 25207 + }, + { + "epoch": 11.719200371920037, + "grad_norm": 4.590168476104736, + "learning_rate": 9.9748029974803e-06, + "loss": 0.0215, + "step": 25208 + }, + { + "epoch": 11.719665271966527, + "grad_norm": 5.812345027923584, + "learning_rate": 9.974801997480201e-06, + "loss": 0.1105, + "step": 25209 + }, + { + "epoch": 11.720130172013016, + "grad_norm": 4.0655317306518555, + "learning_rate": 9.974800997480102e-06, + "loss": 0.0511, + "step": 25210 + }, + { + "epoch": 11.720595072059508, + "grad_norm": 3.376596689224243, + "learning_rate": 9.97479999748e-06, + "loss": 0.0155, + "step": 25211 + }, + { + "epoch": 11.721059972105998, + "grad_norm": 3.3717148303985596, + "learning_rate": 9.974798997479901e-06, + "loss": 0.0479, + "step": 25212 + }, + { + "epoch": 11.721524872152488, + "grad_norm": 8.946443557739258, + "learning_rate": 9.9747979974798e-06, + "loss": 0.1622, + "step": 25213 + }, + { + "epoch": 11.721989772198977, + "grad_norm": 3.3810009956359863, + "learning_rate": 9.9747969974797e-06, + "loss": 0.0777, + "step": 25214 + }, + { + "epoch": 11.722454672245467, + "grad_norm": 7.406413555145264, + "learning_rate": 9.9747959974796e-06, + "loss": 0.1341, + "step": 25215 + }, + { + "epoch": 11.722919572291957, + "grad_norm": 5.5532965660095215, + "learning_rate": 9.9747949974795e-06, + "loss": 0.0853, + "step": 25216 + }, + { + "epoch": 11.723384472338447, + "grad_norm": 4.548465251922607, + "learning_rate": 9.974793997479401e-06, + "loss": 0.0556, + "step": 25217 + }, + { + "epoch": 11.723849372384937, + "grad_norm": 3.621608018875122, + "learning_rate": 9.9747929974793e-06, + "loss": 0.0258, + "step": 25218 + }, + { + "epoch": 11.724314272431426, + "grad_norm": 3.699960470199585, + "learning_rate": 9.9747919974792e-06, + "loss": 0.0446, + "step": 25219 + }, + { + "epoch": 11.724779172477918, + "grad_norm": 4.011912822723389, + "learning_rate": 9.9747909974791e-06, + "loss": 0.0522, + "step": 25220 + }, + { + "epoch": 11.725244072524408, + "grad_norm": 3.2774298191070557, + "learning_rate": 9.974789997479e-06, + "loss": 0.0184, + "step": 25221 + }, + { + "epoch": 11.725708972570898, + "grad_norm": 4.73477840423584, + "learning_rate": 9.974788997478901e-06, + "loss": 0.0729, + "step": 25222 + }, + { + "epoch": 11.726173872617387, + "grad_norm": 7.234653949737549, + "learning_rate": 9.974787997478802e-06, + "loss": 0.156, + "step": 25223 + }, + { + "epoch": 11.726638772663877, + "grad_norm": 5.030210971832275, + "learning_rate": 9.9747869974787e-06, + "loss": 0.0886, + "step": 25224 + }, + { + "epoch": 11.727103672710367, + "grad_norm": 6.7883381843566895, + "learning_rate": 9.974785997478601e-06, + "loss": 0.0664, + "step": 25225 + }, + { + "epoch": 11.727568572756857, + "grad_norm": 3.965592384338379, + "learning_rate": 9.9747849974785e-06, + "loss": 0.0584, + "step": 25226 + }, + { + "epoch": 11.728033472803347, + "grad_norm": 3.6733086109161377, + "learning_rate": 9.974783997478401e-06, + "loss": 0.0325, + "step": 25227 + }, + { + "epoch": 11.728498372849838, + "grad_norm": 4.699937343597412, + "learning_rate": 9.9747829974783e-06, + "loss": 0.0649, + "step": 25228 + }, + { + "epoch": 11.728963272896328, + "grad_norm": 7.632876873016357, + "learning_rate": 9.9747819974782e-06, + "loss": 0.0289, + "step": 25229 + }, + { + "epoch": 11.729428172942818, + "grad_norm": 6.3128862380981445, + "learning_rate": 9.974780997478101e-06, + "loss": 0.0923, + "step": 25230 + }, + { + "epoch": 11.729893072989308, + "grad_norm": 12.44379711151123, + "learning_rate": 9.974779997478e-06, + "loss": 0.1752, + "step": 25231 + }, + { + "epoch": 11.730357973035797, + "grad_norm": 4.6088128089904785, + "learning_rate": 9.9747789974779e-06, + "loss": 0.0623, + "step": 25232 + }, + { + "epoch": 11.730822873082287, + "grad_norm": 3.2371366024017334, + "learning_rate": 9.9747779974778e-06, + "loss": 0.0193, + "step": 25233 + }, + { + "epoch": 11.731287773128777, + "grad_norm": 2.935347080230713, + "learning_rate": 9.9747769974777e-06, + "loss": 0.0203, + "step": 25234 + }, + { + "epoch": 11.731752673175267, + "grad_norm": 5.162364959716797, + "learning_rate": 9.974775997477601e-06, + "loss": 0.0937, + "step": 25235 + }, + { + "epoch": 11.732217573221757, + "grad_norm": 6.8079986572265625, + "learning_rate": 9.974774997477502e-06, + "loss": 0.053, + "step": 25236 + }, + { + "epoch": 11.732682473268248, + "grad_norm": 14.65825366973877, + "learning_rate": 9.9747739974774e-06, + "loss": 0.0746, + "step": 25237 + }, + { + "epoch": 11.733147373314738, + "grad_norm": 1.462830901145935, + "learning_rate": 9.974772997477301e-06, + "loss": 0.0068, + "step": 25238 + }, + { + "epoch": 11.733612273361228, + "grad_norm": 1.8871166706085205, + "learning_rate": 9.9747719974772e-06, + "loss": 0.0141, + "step": 25239 + }, + { + "epoch": 11.734077173407718, + "grad_norm": 6.063197135925293, + "learning_rate": 9.9747709974771e-06, + "loss": 0.0411, + "step": 25240 + }, + { + "epoch": 11.734542073454207, + "grad_norm": 3.0546891689300537, + "learning_rate": 9.974769997477e-06, + "loss": 0.0314, + "step": 25241 + }, + { + "epoch": 11.735006973500697, + "grad_norm": 4.8128886222839355, + "learning_rate": 9.9747689974769e-06, + "loss": 0.0831, + "step": 25242 + }, + { + "epoch": 11.735471873547187, + "grad_norm": 3.1904311180114746, + "learning_rate": 9.974767997476801e-06, + "loss": 0.0404, + "step": 25243 + }, + { + "epoch": 11.735936773593677, + "grad_norm": 3.985558271408081, + "learning_rate": 9.9747669974767e-06, + "loss": 0.0315, + "step": 25244 + }, + { + "epoch": 11.736401673640167, + "grad_norm": 6.17880392074585, + "learning_rate": 9.974765997476601e-06, + "loss": 0.0799, + "step": 25245 + }, + { + "epoch": 11.736866573686658, + "grad_norm": 5.158433437347412, + "learning_rate": 9.9747649974765e-06, + "loss": 0.0563, + "step": 25246 + }, + { + "epoch": 11.737331473733148, + "grad_norm": 2.760206699371338, + "learning_rate": 9.9747639974764e-06, + "loss": 0.0257, + "step": 25247 + }, + { + "epoch": 11.737796373779638, + "grad_norm": 4.397202968597412, + "learning_rate": 9.974762997476301e-06, + "loss": 0.052, + "step": 25248 + }, + { + "epoch": 11.738261273826128, + "grad_norm": 4.799167633056641, + "learning_rate": 9.974761997476202e-06, + "loss": 0.0558, + "step": 25249 + }, + { + "epoch": 11.738726173872617, + "grad_norm": 3.247189998626709, + "learning_rate": 9.9747609974761e-06, + "loss": 0.0257, + "step": 25250 + }, + { + "epoch": 11.739191073919107, + "grad_norm": 5.995456218719482, + "learning_rate": 9.974759997476e-06, + "loss": 0.085, + "step": 25251 + }, + { + "epoch": 11.739655973965597, + "grad_norm": 6.088917255401611, + "learning_rate": 9.9747589974759e-06, + "loss": 0.1257, + "step": 25252 + }, + { + "epoch": 11.740120874012087, + "grad_norm": 7.884708881378174, + "learning_rate": 9.974757997475801e-06, + "loss": 0.1083, + "step": 25253 + }, + { + "epoch": 11.740585774058577, + "grad_norm": 10.90457820892334, + "learning_rate": 9.9747569974757e-06, + "loss": 0.0673, + "step": 25254 + }, + { + "epoch": 11.741050674105068, + "grad_norm": 5.354421138763428, + "learning_rate": 9.9747559974756e-06, + "loss": 0.0932, + "step": 25255 + }, + { + "epoch": 11.741515574151558, + "grad_norm": 10.575468063354492, + "learning_rate": 9.974754997475501e-06, + "loss": 0.1729, + "step": 25256 + }, + { + "epoch": 11.741980474198048, + "grad_norm": 4.884174346923828, + "learning_rate": 9.9747539974754e-06, + "loss": 0.0576, + "step": 25257 + }, + { + "epoch": 11.742445374244538, + "grad_norm": 4.892212867736816, + "learning_rate": 9.974752997475301e-06, + "loss": 0.0812, + "step": 25258 + }, + { + "epoch": 11.742910274291027, + "grad_norm": 2.915123224258423, + "learning_rate": 9.9747519974752e-06, + "loss": 0.0224, + "step": 25259 + }, + { + "epoch": 11.743375174337517, + "grad_norm": 2.7366511821746826, + "learning_rate": 9.9747509974751e-06, + "loss": 0.0175, + "step": 25260 + }, + { + "epoch": 11.743840074384007, + "grad_norm": 4.776670455932617, + "learning_rate": 9.974749997475001e-06, + "loss": 0.0798, + "step": 25261 + }, + { + "epoch": 11.744304974430497, + "grad_norm": 2.794219732284546, + "learning_rate": 9.9747489974749e-06, + "loss": 0.0254, + "step": 25262 + }, + { + "epoch": 11.744769874476987, + "grad_norm": 2.2161407470703125, + "learning_rate": 9.9747479974748e-06, + "loss": 0.0183, + "step": 25263 + }, + { + "epoch": 11.745234774523478, + "grad_norm": 6.792083263397217, + "learning_rate": 9.9747469974747e-06, + "loss": 0.1001, + "step": 25264 + }, + { + "epoch": 11.745699674569968, + "grad_norm": 2.745567798614502, + "learning_rate": 9.9747459974746e-06, + "loss": 0.0287, + "step": 25265 + }, + { + "epoch": 11.746164574616458, + "grad_norm": 4.38714075088501, + "learning_rate": 9.974744997474501e-06, + "loss": 0.0403, + "step": 25266 + }, + { + "epoch": 11.746629474662948, + "grad_norm": 4.535373687744141, + "learning_rate": 9.9747439974744e-06, + "loss": 0.0441, + "step": 25267 + }, + { + "epoch": 11.747094374709437, + "grad_norm": 3.5928945541381836, + "learning_rate": 9.9747429974743e-06, + "loss": 0.0591, + "step": 25268 + }, + { + "epoch": 11.747559274755927, + "grad_norm": 3.716111660003662, + "learning_rate": 9.974741997474201e-06, + "loss": 0.0392, + "step": 25269 + }, + { + "epoch": 11.748024174802417, + "grad_norm": 3.341982126235962, + "learning_rate": 9.9747409974741e-06, + "loss": 0.0279, + "step": 25270 + }, + { + "epoch": 11.748489074848907, + "grad_norm": 7.159616947174072, + "learning_rate": 9.974739997474e-06, + "loss": 0.1243, + "step": 25271 + }, + { + "epoch": 11.748953974895397, + "grad_norm": 1.817133903503418, + "learning_rate": 9.9747389974739e-06, + "loss": 0.0134, + "step": 25272 + }, + { + "epoch": 11.749418874941888, + "grad_norm": 4.927132606506348, + "learning_rate": 9.9747379974738e-06, + "loss": 0.0756, + "step": 25273 + }, + { + "epoch": 11.749883774988378, + "grad_norm": 5.347421169281006, + "learning_rate": 9.974736997473701e-06, + "loss": 0.1011, + "step": 25274 + }, + { + "epoch": 11.750348675034868, + "grad_norm": 3.6951544284820557, + "learning_rate": 9.9747359974736e-06, + "loss": 0.0379, + "step": 25275 + }, + { + "epoch": 11.750813575081358, + "grad_norm": 4.603103160858154, + "learning_rate": 9.974734997473501e-06, + "loss": 0.0538, + "step": 25276 + }, + { + "epoch": 11.751278475127847, + "grad_norm": 3.526141405105591, + "learning_rate": 9.9747339974734e-06, + "loss": 0.0472, + "step": 25277 + }, + { + "epoch": 11.751743375174337, + "grad_norm": 8.912376403808594, + "learning_rate": 9.9747329974733e-06, + "loss": 0.105, + "step": 25278 + }, + { + "epoch": 11.752208275220827, + "grad_norm": 3.7665064334869385, + "learning_rate": 9.974731997473201e-06, + "loss": 0.0377, + "step": 25279 + }, + { + "epoch": 11.752673175267317, + "grad_norm": 3.345585584640503, + "learning_rate": 9.974730997473102e-06, + "loss": 0.0343, + "step": 25280 + }, + { + "epoch": 11.753138075313807, + "grad_norm": 4.266292572021484, + "learning_rate": 9.974729997473e-06, + "loss": 0.0527, + "step": 25281 + }, + { + "epoch": 11.753602975360298, + "grad_norm": 3.854146718978882, + "learning_rate": 9.9747289974729e-06, + "loss": 0.0304, + "step": 25282 + }, + { + "epoch": 11.754067875406788, + "grad_norm": 6.746329307556152, + "learning_rate": 9.9747279974728e-06, + "loss": 0.0809, + "step": 25283 + }, + { + "epoch": 11.754532775453278, + "grad_norm": 7.204043865203857, + "learning_rate": 9.9747269974727e-06, + "loss": 0.1174, + "step": 25284 + }, + { + "epoch": 11.754997675499768, + "grad_norm": 3.9863107204437256, + "learning_rate": 9.9747259974726e-06, + "loss": 0.0434, + "step": 25285 + }, + { + "epoch": 11.755462575546257, + "grad_norm": 1.992120385169983, + "learning_rate": 9.9747249974725e-06, + "loss": 0.0138, + "step": 25286 + }, + { + "epoch": 11.755927475592747, + "grad_norm": 5.638650894165039, + "learning_rate": 9.974723997472401e-06, + "loss": 0.0763, + "step": 25287 + }, + { + "epoch": 11.756392375639237, + "grad_norm": 5.914666652679443, + "learning_rate": 9.9747229974723e-06, + "loss": 0.0872, + "step": 25288 + }, + { + "epoch": 11.756857275685727, + "grad_norm": 4.5474114418029785, + "learning_rate": 9.974721997472201e-06, + "loss": 0.0903, + "step": 25289 + }, + { + "epoch": 11.757322175732218, + "grad_norm": 5.985111236572266, + "learning_rate": 9.9747209974721e-06, + "loss": 0.1488, + "step": 25290 + }, + { + "epoch": 11.757787075778708, + "grad_norm": 3.2936437129974365, + "learning_rate": 9.974719997472e-06, + "loss": 0.059, + "step": 25291 + }, + { + "epoch": 11.758251975825198, + "grad_norm": 2.274390459060669, + "learning_rate": 9.974718997471901e-06, + "loss": 0.0187, + "step": 25292 + }, + { + "epoch": 11.758716875871688, + "grad_norm": 6.632340431213379, + "learning_rate": 9.9747179974718e-06, + "loss": 0.0863, + "step": 25293 + }, + { + "epoch": 11.759181775918178, + "grad_norm": 3.592966318130493, + "learning_rate": 9.974716997471701e-06, + "loss": 0.0509, + "step": 25294 + }, + { + "epoch": 11.759646675964667, + "grad_norm": 2.329129934310913, + "learning_rate": 9.9747159974716e-06, + "loss": 0.0207, + "step": 25295 + }, + { + "epoch": 11.760111576011157, + "grad_norm": 6.06663179397583, + "learning_rate": 9.9747149974715e-06, + "loss": 0.0688, + "step": 25296 + }, + { + "epoch": 11.760576476057647, + "grad_norm": 6.816576957702637, + "learning_rate": 9.9747139974714e-06, + "loss": 0.0496, + "step": 25297 + }, + { + "epoch": 11.761041376104137, + "grad_norm": 4.191389560699463, + "learning_rate": 9.9747129974713e-06, + "loss": 0.0475, + "step": 25298 + }, + { + "epoch": 11.761506276150628, + "grad_norm": 2.0766355991363525, + "learning_rate": 9.9747119974712e-06, + "loss": 0.0143, + "step": 25299 + }, + { + "epoch": 11.761971176197118, + "grad_norm": 5.3152642250061035, + "learning_rate": 9.974710997471101e-06, + "loss": 0.0652, + "step": 25300 + }, + { + "epoch": 11.762436076243608, + "grad_norm": 2.9727776050567627, + "learning_rate": 9.974709997471e-06, + "loss": 0.0203, + "step": 25301 + }, + { + "epoch": 11.762900976290098, + "grad_norm": 4.445004463195801, + "learning_rate": 9.974708997470901e-06, + "loss": 0.0619, + "step": 25302 + }, + { + "epoch": 11.763365876336588, + "grad_norm": 5.409650802612305, + "learning_rate": 9.9747079974708e-06, + "loss": 0.0872, + "step": 25303 + }, + { + "epoch": 11.763830776383077, + "grad_norm": 3.400301933288574, + "learning_rate": 9.9747069974707e-06, + "loss": 0.0233, + "step": 25304 + }, + { + "epoch": 11.764295676429567, + "grad_norm": 4.942079544067383, + "learning_rate": 9.974705997470601e-06, + "loss": 0.0498, + "step": 25305 + }, + { + "epoch": 11.764760576476057, + "grad_norm": 4.4735107421875, + "learning_rate": 9.9747049974705e-06, + "loss": 0.0632, + "step": 25306 + }, + { + "epoch": 11.765225476522549, + "grad_norm": 2.4905781745910645, + "learning_rate": 9.974703997470401e-06, + "loss": 0.0245, + "step": 25307 + }, + { + "epoch": 11.765690376569038, + "grad_norm": 5.1561598777771, + "learning_rate": 9.9747029974703e-06, + "loss": 0.0643, + "step": 25308 + }, + { + "epoch": 11.766155276615528, + "grad_norm": 5.169783115386963, + "learning_rate": 9.9747019974702e-06, + "loss": 0.083, + "step": 25309 + }, + { + "epoch": 11.766620176662018, + "grad_norm": 4.0047760009765625, + "learning_rate": 9.9747009974701e-06, + "loss": 0.049, + "step": 25310 + }, + { + "epoch": 11.767085076708508, + "grad_norm": 3.336979389190674, + "learning_rate": 9.97469999747e-06, + "loss": 0.0369, + "step": 25311 + }, + { + "epoch": 11.767549976754998, + "grad_norm": 2.6011688709259033, + "learning_rate": 9.974698997469901e-06, + "loss": 0.036, + "step": 25312 + }, + { + "epoch": 11.768014876801487, + "grad_norm": 5.60992431640625, + "learning_rate": 9.974697997469802e-06, + "loss": 0.0924, + "step": 25313 + }, + { + "epoch": 11.768479776847977, + "grad_norm": 12.595205307006836, + "learning_rate": 9.9746969974697e-06, + "loss": 0.0403, + "step": 25314 + }, + { + "epoch": 11.768944676894467, + "grad_norm": 6.409214019775391, + "learning_rate": 9.9746959974696e-06, + "loss": 0.063, + "step": 25315 + }, + { + "epoch": 11.769409576940959, + "grad_norm": 4.835926055908203, + "learning_rate": 9.9746949974695e-06, + "loss": 0.0698, + "step": 25316 + }, + { + "epoch": 11.769874476987448, + "grad_norm": 0.8933964967727661, + "learning_rate": 9.9746939974694e-06, + "loss": 0.0115, + "step": 25317 + }, + { + "epoch": 11.770339377033938, + "grad_norm": 3.493011474609375, + "learning_rate": 9.974692997469301e-06, + "loss": 0.0218, + "step": 25318 + }, + { + "epoch": 11.770804277080428, + "grad_norm": 3.4253838062286377, + "learning_rate": 9.9746919974692e-06, + "loss": 0.0868, + "step": 25319 + }, + { + "epoch": 11.771269177126918, + "grad_norm": 6.791016578674316, + "learning_rate": 9.974690997469101e-06, + "loss": 0.0663, + "step": 25320 + }, + { + "epoch": 11.771734077173408, + "grad_norm": 5.2408766746521, + "learning_rate": 9.974689997469e-06, + "loss": 0.0557, + "step": 25321 + }, + { + "epoch": 11.772198977219897, + "grad_norm": 6.117639064788818, + "learning_rate": 9.9746889974689e-06, + "loss": 0.0597, + "step": 25322 + }, + { + "epoch": 11.772663877266387, + "grad_norm": 6.702134609222412, + "learning_rate": 9.9746879974688e-06, + "loss": 0.1008, + "step": 25323 + }, + { + "epoch": 11.773128777312877, + "grad_norm": 4.922016620635986, + "learning_rate": 9.9746869974687e-06, + "loss": 0.0521, + "step": 25324 + }, + { + "epoch": 11.773593677359369, + "grad_norm": 4.848677635192871, + "learning_rate": 9.974685997468601e-06, + "loss": 0.0646, + "step": 25325 + }, + { + "epoch": 11.774058577405858, + "grad_norm": 3.3290910720825195, + "learning_rate": 9.9746849974685e-06, + "loss": 0.0301, + "step": 25326 + }, + { + "epoch": 11.774523477452348, + "grad_norm": 5.776988983154297, + "learning_rate": 9.9746839974684e-06, + "loss": 0.1264, + "step": 25327 + }, + { + "epoch": 11.774988377498838, + "grad_norm": 7.2600908279418945, + "learning_rate": 9.9746829974683e-06, + "loss": 0.0736, + "step": 25328 + }, + { + "epoch": 11.775453277545328, + "grad_norm": 1.3689948320388794, + "learning_rate": 9.9746819974682e-06, + "loss": 0.0078, + "step": 25329 + }, + { + "epoch": 11.775918177591818, + "grad_norm": 3.5961949825286865, + "learning_rate": 9.974680997468101e-06, + "loss": 0.0369, + "step": 25330 + }, + { + "epoch": 11.776383077638307, + "grad_norm": 5.856704235076904, + "learning_rate": 9.974679997468002e-06, + "loss": 0.0645, + "step": 25331 + }, + { + "epoch": 11.776847977684797, + "grad_norm": 6.081303119659424, + "learning_rate": 9.9746789974679e-06, + "loss": 0.0993, + "step": 25332 + }, + { + "epoch": 11.777312877731287, + "grad_norm": 3.587404727935791, + "learning_rate": 9.974677997467801e-06, + "loss": 0.0534, + "step": 25333 + }, + { + "epoch": 11.777777777777779, + "grad_norm": 6.382213115692139, + "learning_rate": 9.9746769974677e-06, + "loss": 0.0737, + "step": 25334 + }, + { + "epoch": 11.778242677824268, + "grad_norm": 4.03794002532959, + "learning_rate": 9.9746759974676e-06, + "loss": 0.0475, + "step": 25335 + }, + { + "epoch": 11.778707577870758, + "grad_norm": 5.800041198730469, + "learning_rate": 9.9746749974675e-06, + "loss": 0.086, + "step": 25336 + }, + { + "epoch": 11.779172477917248, + "grad_norm": 2.1158623695373535, + "learning_rate": 9.9746739974674e-06, + "loss": 0.0124, + "step": 25337 + }, + { + "epoch": 11.779637377963738, + "grad_norm": 7.429626941680908, + "learning_rate": 9.974672997467301e-06, + "loss": 0.1784, + "step": 25338 + }, + { + "epoch": 11.780102278010228, + "grad_norm": 5.773734092712402, + "learning_rate": 9.9746719974672e-06, + "loss": 0.0582, + "step": 25339 + }, + { + "epoch": 11.780567178056717, + "grad_norm": 6.834418296813965, + "learning_rate": 9.9746709974671e-06, + "loss": 0.0846, + "step": 25340 + }, + { + "epoch": 11.781032078103207, + "grad_norm": 3.213503360748291, + "learning_rate": 9.974669997467e-06, + "loss": 0.0327, + "step": 25341 + }, + { + "epoch": 11.781496978149697, + "grad_norm": 5.146398544311523, + "learning_rate": 9.9746689974669e-06, + "loss": 0.0455, + "step": 25342 + }, + { + "epoch": 11.781961878196189, + "grad_norm": 7.847296714782715, + "learning_rate": 9.974667997466801e-06, + "loss": 0.086, + "step": 25343 + }, + { + "epoch": 11.782426778242678, + "grad_norm": 6.51495885848999, + "learning_rate": 9.974666997466702e-06, + "loss": 0.1117, + "step": 25344 + }, + { + "epoch": 11.782891678289168, + "grad_norm": 7.318968772888184, + "learning_rate": 9.9746659974666e-06, + "loss": 0.1017, + "step": 25345 + }, + { + "epoch": 11.783356578335658, + "grad_norm": 3.3630175590515137, + "learning_rate": 9.974664997466501e-06, + "loss": 0.0313, + "step": 25346 + }, + { + "epoch": 11.783821478382148, + "grad_norm": 1.950803279876709, + "learning_rate": 9.9746639974664e-06, + "loss": 0.0052, + "step": 25347 + }, + { + "epoch": 11.784286378428638, + "grad_norm": 5.669799327850342, + "learning_rate": 9.974662997466301e-06, + "loss": 0.0717, + "step": 25348 + }, + { + "epoch": 11.784751278475127, + "grad_norm": 4.833171844482422, + "learning_rate": 9.974661997466202e-06, + "loss": 0.0654, + "step": 25349 + }, + { + "epoch": 11.785216178521617, + "grad_norm": 5.670680522918701, + "learning_rate": 9.9746609974661e-06, + "loss": 0.0715, + "step": 25350 + }, + { + "epoch": 11.785681078568107, + "grad_norm": 1.9462230205535889, + "learning_rate": 9.974659997466001e-06, + "loss": 0.0129, + "step": 25351 + }, + { + "epoch": 11.786145978614599, + "grad_norm": 2.670381546020508, + "learning_rate": 9.9746589974659e-06, + "loss": 0.0237, + "step": 25352 + }, + { + "epoch": 11.786610878661088, + "grad_norm": 4.680973529815674, + "learning_rate": 9.9746579974658e-06, + "loss": 0.0753, + "step": 25353 + }, + { + "epoch": 11.787075778707578, + "grad_norm": 3.4808154106140137, + "learning_rate": 9.9746569974657e-06, + "loss": 0.0323, + "step": 25354 + }, + { + "epoch": 11.787540678754068, + "grad_norm": 4.968514919281006, + "learning_rate": 9.9746559974656e-06, + "loss": 0.0525, + "step": 25355 + }, + { + "epoch": 11.788005578800558, + "grad_norm": 4.496647357940674, + "learning_rate": 9.974654997465501e-06, + "loss": 0.0557, + "step": 25356 + }, + { + "epoch": 11.788470478847048, + "grad_norm": 5.443962574005127, + "learning_rate": 9.974653997465402e-06, + "loss": 0.0588, + "step": 25357 + }, + { + "epoch": 11.788935378893537, + "grad_norm": 3.779749631881714, + "learning_rate": 9.9746529974653e-06, + "loss": 0.023, + "step": 25358 + }, + { + "epoch": 11.789400278940027, + "grad_norm": 2.7324957847595215, + "learning_rate": 9.9746519974652e-06, + "loss": 0.0205, + "step": 25359 + }, + { + "epoch": 11.789865178986517, + "grad_norm": 1.6608569622039795, + "learning_rate": 9.9746509974651e-06, + "loss": 0.0084, + "step": 25360 + }, + { + "epoch": 11.790330079033009, + "grad_norm": 5.181735515594482, + "learning_rate": 9.974649997465001e-06, + "loss": 0.0659, + "step": 25361 + }, + { + "epoch": 11.790794979079498, + "grad_norm": 4.825450897216797, + "learning_rate": 9.974648997464902e-06, + "loss": 0.0673, + "step": 25362 + }, + { + "epoch": 11.791259879125988, + "grad_norm": 4.627866744995117, + "learning_rate": 9.9746479974648e-06, + "loss": 0.0416, + "step": 25363 + }, + { + "epoch": 11.791724779172478, + "grad_norm": 6.481324195861816, + "learning_rate": 9.974646997464701e-06, + "loss": 0.1126, + "step": 25364 + }, + { + "epoch": 11.792189679218968, + "grad_norm": 4.940171718597412, + "learning_rate": 9.9746459974646e-06, + "loss": 0.0289, + "step": 25365 + }, + { + "epoch": 11.792654579265458, + "grad_norm": 6.375105857849121, + "learning_rate": 9.9746449974645e-06, + "loss": 0.1375, + "step": 25366 + }, + { + "epoch": 11.793119479311947, + "grad_norm": 2.356511354446411, + "learning_rate": 9.9746439974644e-06, + "loss": 0.0175, + "step": 25367 + }, + { + "epoch": 11.793584379358437, + "grad_norm": 5.952409744262695, + "learning_rate": 9.9746429974643e-06, + "loss": 0.1458, + "step": 25368 + }, + { + "epoch": 11.794049279404929, + "grad_norm": 4.06074857711792, + "learning_rate": 9.974641997464201e-06, + "loss": 0.084, + "step": 25369 + }, + { + "epoch": 11.794514179451419, + "grad_norm": 5.149719715118408, + "learning_rate": 9.9746409974641e-06, + "loss": 0.0515, + "step": 25370 + }, + { + "epoch": 11.794979079497908, + "grad_norm": 5.937187194824219, + "learning_rate": 9.974639997464e-06, + "loss": 0.0681, + "step": 25371 + }, + { + "epoch": 11.795443979544398, + "grad_norm": 6.355993747711182, + "learning_rate": 9.9746389974639e-06, + "loss": 0.0932, + "step": 25372 + }, + { + "epoch": 11.795908879590888, + "grad_norm": 4.319134712219238, + "learning_rate": 9.9746379974638e-06, + "loss": 0.0735, + "step": 25373 + }, + { + "epoch": 11.796373779637378, + "grad_norm": 2.3609302043914795, + "learning_rate": 9.974636997463701e-06, + "loss": 0.0148, + "step": 25374 + }, + { + "epoch": 11.796838679683868, + "grad_norm": 0.68253493309021, + "learning_rate": 9.974635997463602e-06, + "loss": 0.0057, + "step": 25375 + }, + { + "epoch": 11.797303579730357, + "grad_norm": 2.745936393737793, + "learning_rate": 9.9746349974635e-06, + "loss": 0.0282, + "step": 25376 + }, + { + "epoch": 11.797768479776847, + "grad_norm": 5.191048622131348, + "learning_rate": 9.974633997463401e-06, + "loss": 0.0955, + "step": 25377 + }, + { + "epoch": 11.798233379823339, + "grad_norm": 4.005604267120361, + "learning_rate": 9.9746329974633e-06, + "loss": 0.0511, + "step": 25378 + }, + { + "epoch": 11.798698279869829, + "grad_norm": 0.7064276933670044, + "learning_rate": 9.974631997463201e-06, + "loss": 0.005, + "step": 25379 + }, + { + "epoch": 11.799163179916318, + "grad_norm": 2.8913750648498535, + "learning_rate": 9.9746309974631e-06, + "loss": 0.0389, + "step": 25380 + }, + { + "epoch": 11.799628079962808, + "grad_norm": 5.658105373382568, + "learning_rate": 9.974629997463e-06, + "loss": 0.0701, + "step": 25381 + }, + { + "epoch": 11.800092980009298, + "grad_norm": 7.00608491897583, + "learning_rate": 9.974628997462901e-06, + "loss": 0.1379, + "step": 25382 + }, + { + "epoch": 11.800557880055788, + "grad_norm": 6.454098701477051, + "learning_rate": 9.9746279974628e-06, + "loss": 0.1127, + "step": 25383 + }, + { + "epoch": 11.801022780102278, + "grad_norm": 2.7032277584075928, + "learning_rate": 9.9746269974627e-06, + "loss": 0.0241, + "step": 25384 + }, + { + "epoch": 11.801487680148767, + "grad_norm": 4.069949150085449, + "learning_rate": 9.9746259974626e-06, + "loss": 0.0383, + "step": 25385 + }, + { + "epoch": 11.801952580195259, + "grad_norm": 0.4582821726799011, + "learning_rate": 9.9746249974625e-06, + "loss": 0.0023, + "step": 25386 + }, + { + "epoch": 11.802417480241749, + "grad_norm": 6.47890567779541, + "learning_rate": 9.974623997462401e-06, + "loss": 0.1505, + "step": 25387 + }, + { + "epoch": 11.802882380288239, + "grad_norm": 3.2536840438842773, + "learning_rate": 9.974622997462302e-06, + "loss": 0.0385, + "step": 25388 + }, + { + "epoch": 11.803347280334728, + "grad_norm": 6.52921199798584, + "learning_rate": 9.9746219974622e-06, + "loss": 0.1152, + "step": 25389 + }, + { + "epoch": 11.803812180381218, + "grad_norm": 5.5571489334106445, + "learning_rate": 9.974620997462101e-06, + "loss": 0.072, + "step": 25390 + }, + { + "epoch": 11.804277080427708, + "grad_norm": 6.158161163330078, + "learning_rate": 9.974619997462e-06, + "loss": 0.0952, + "step": 25391 + }, + { + "epoch": 11.804741980474198, + "grad_norm": 1.1274601221084595, + "learning_rate": 9.9746189974619e-06, + "loss": 0.0109, + "step": 25392 + }, + { + "epoch": 11.805206880520688, + "grad_norm": 4.978218078613281, + "learning_rate": 9.9746179974618e-06, + "loss": 0.0851, + "step": 25393 + }, + { + "epoch": 11.805671780567177, + "grad_norm": 5.228461265563965, + "learning_rate": 9.9746169974617e-06, + "loss": 0.0806, + "step": 25394 + }, + { + "epoch": 11.806136680613669, + "grad_norm": 4.881372451782227, + "learning_rate": 9.974615997461601e-06, + "loss": 0.1047, + "step": 25395 + }, + { + "epoch": 11.806601580660159, + "grad_norm": 9.924484252929688, + "learning_rate": 9.9746149974615e-06, + "loss": 0.0904, + "step": 25396 + }, + { + "epoch": 11.807066480706649, + "grad_norm": 0.7486565709114075, + "learning_rate": 9.974613997461401e-06, + "loss": 0.0078, + "step": 25397 + }, + { + "epoch": 11.807531380753138, + "grad_norm": 0.6356340050697327, + "learning_rate": 9.9746129974613e-06, + "loss": 0.0062, + "step": 25398 + }, + { + "epoch": 11.807996280799628, + "grad_norm": 6.561659336090088, + "learning_rate": 9.9746119974612e-06, + "loss": 0.071, + "step": 25399 + }, + { + "epoch": 11.808461180846118, + "grad_norm": 6.451469898223877, + "learning_rate": 9.974610997461101e-06, + "loss": 0.0744, + "step": 25400 + }, + { + "epoch": 11.808926080892608, + "grad_norm": 2.6553404331207275, + "learning_rate": 9.974609997461002e-06, + "loss": 0.0388, + "step": 25401 + }, + { + "epoch": 11.809390980939098, + "grad_norm": 3.114595651626587, + "learning_rate": 9.9746089974609e-06, + "loss": 0.0139, + "step": 25402 + }, + { + "epoch": 11.809855880985587, + "grad_norm": 2.5337624549865723, + "learning_rate": 9.9746079974608e-06, + "loss": 0.0312, + "step": 25403 + }, + { + "epoch": 11.810320781032079, + "grad_norm": 2.788984537124634, + "learning_rate": 9.9746069974607e-06, + "loss": 0.0226, + "step": 25404 + }, + { + "epoch": 11.810785681078569, + "grad_norm": 2.782437562942505, + "learning_rate": 9.9746059974606e-06, + "loss": 0.0241, + "step": 25405 + }, + { + "epoch": 11.811250581125059, + "grad_norm": 4.601869106292725, + "learning_rate": 9.9746049974605e-06, + "loss": 0.0823, + "step": 25406 + }, + { + "epoch": 11.811715481171548, + "grad_norm": Infinity, + "learning_rate": 9.9746049974605e-06, + "loss": 0.3893, + "step": 25407 + }, + { + "epoch": 11.812180381218038, + "grad_norm": 10.472460746765137, + "learning_rate": 9.9746039974604e-06, + "loss": 0.1081, + "step": 25408 + }, + { + "epoch": 11.812645281264528, + "grad_norm": 7.712444305419922, + "learning_rate": 9.974602997460301e-06, + "loss": 0.1905, + "step": 25409 + }, + { + "epoch": 11.813110181311018, + "grad_norm": 3.612189769744873, + "learning_rate": 9.9746019974602e-06, + "loss": 0.0283, + "step": 25410 + }, + { + "epoch": 11.813575081357508, + "grad_norm": 4.2982916831970215, + "learning_rate": 9.974600997460101e-06, + "loss": 0.0642, + "step": 25411 + }, + { + "epoch": 11.814039981403997, + "grad_norm": 3.994715929031372, + "learning_rate": 9.97459999746e-06, + "loss": 0.0361, + "step": 25412 + }, + { + "epoch": 11.814504881450489, + "grad_norm": 7.059380054473877, + "learning_rate": 9.9745989974599e-06, + "loss": 0.1376, + "step": 25413 + }, + { + "epoch": 11.814969781496979, + "grad_norm": 3.385176658630371, + "learning_rate": 9.974597997459801e-06, + "loss": 0.0241, + "step": 25414 + }, + { + "epoch": 11.815434681543469, + "grad_norm": 5.3353400230407715, + "learning_rate": 9.9745969974597e-06, + "loss": 0.1013, + "step": 25415 + }, + { + "epoch": 11.815899581589958, + "grad_norm": 8.294177055358887, + "learning_rate": 9.974595997459601e-06, + "loss": 0.2919, + "step": 25416 + }, + { + "epoch": 11.816364481636448, + "grad_norm": 2.223284959793091, + "learning_rate": 9.9745949974595e-06, + "loss": 0.0345, + "step": 25417 + }, + { + "epoch": 11.816829381682938, + "grad_norm": 4.367681980133057, + "learning_rate": 9.9745939974594e-06, + "loss": 0.0424, + "step": 25418 + }, + { + "epoch": 11.817294281729428, + "grad_norm": 3.9666495323181152, + "learning_rate": 9.974592997459301e-06, + "loss": 0.0354, + "step": 25419 + }, + { + "epoch": 11.817759181775918, + "grad_norm": 1.9880014657974243, + "learning_rate": 9.9745919974592e-06, + "loss": 0.0171, + "step": 25420 + }, + { + "epoch": 11.818224081822407, + "grad_norm": 7.000672340393066, + "learning_rate": 9.9745909974591e-06, + "loss": 0.2192, + "step": 25421 + }, + { + "epoch": 11.818688981868899, + "grad_norm": 8.018488883972168, + "learning_rate": 9.974589997459001e-06, + "loss": 0.1296, + "step": 25422 + }, + { + "epoch": 11.819153881915389, + "grad_norm": 5.047811985015869, + "learning_rate": 9.9745889974589e-06, + "loss": 0.0796, + "step": 25423 + }, + { + "epoch": 11.819618781961879, + "grad_norm": 5.654405117034912, + "learning_rate": 9.9745879974588e-06, + "loss": 0.0803, + "step": 25424 + }, + { + "epoch": 11.820083682008368, + "grad_norm": 5.385044574737549, + "learning_rate": 9.9745869974587e-06, + "loss": 0.0976, + "step": 25425 + }, + { + "epoch": 11.820548582054858, + "grad_norm": 1.328678846359253, + "learning_rate": 9.9745859974586e-06, + "loss": 0.0119, + "step": 25426 + }, + { + "epoch": 11.821013482101348, + "grad_norm": 3.988204002380371, + "learning_rate": 9.974584997458501e-06, + "loss": 0.0657, + "step": 25427 + }, + { + "epoch": 11.821478382147838, + "grad_norm": 0.25076988339424133, + "learning_rate": 9.9745839974584e-06, + "loss": 0.0019, + "step": 25428 + }, + { + "epoch": 11.821943282194328, + "grad_norm": 5.3868513107299805, + "learning_rate": 9.974582997458301e-06, + "loss": 0.0558, + "step": 25429 + }, + { + "epoch": 11.822408182240817, + "grad_norm": 5.6023173332214355, + "learning_rate": 9.9745819974582e-06, + "loss": 0.0435, + "step": 25430 + }, + { + "epoch": 11.822873082287309, + "grad_norm": 2.961822986602783, + "learning_rate": 9.9745809974581e-06, + "loss": 0.0355, + "step": 25431 + }, + { + "epoch": 11.823337982333799, + "grad_norm": 3.80180287361145, + "learning_rate": 9.974579997458001e-06, + "loss": 0.0313, + "step": 25432 + }, + { + "epoch": 11.823802882380289, + "grad_norm": 2.739652156829834, + "learning_rate": 9.9745789974579e-06, + "loss": 0.0278, + "step": 25433 + }, + { + "epoch": 11.824267782426778, + "grad_norm": 2.2507247924804688, + "learning_rate": 9.974577997457801e-06, + "loss": 0.0176, + "step": 25434 + }, + { + "epoch": 11.824732682473268, + "grad_norm": 5.836026668548584, + "learning_rate": 9.9745769974577e-06, + "loss": 0.0528, + "step": 25435 + }, + { + "epoch": 11.825197582519758, + "grad_norm": 2.2086453437805176, + "learning_rate": 9.9745759974576e-06, + "loss": 0.0266, + "step": 25436 + }, + { + "epoch": 11.825662482566248, + "grad_norm": 7.842169761657715, + "learning_rate": 9.9745749974575e-06, + "loss": 0.07, + "step": 25437 + }, + { + "epoch": 11.826127382612738, + "grad_norm": 1.9742001295089722, + "learning_rate": 9.9745739974574e-06, + "loss": 0.0185, + "step": 25438 + }, + { + "epoch": 11.826592282659227, + "grad_norm": 3.6069250106811523, + "learning_rate": 9.9745729974573e-06, + "loss": 0.0323, + "step": 25439 + }, + { + "epoch": 11.827057182705719, + "grad_norm": 11.307353973388672, + "learning_rate": 9.974571997457201e-06, + "loss": 0.0907, + "step": 25440 + }, + { + "epoch": 11.827522082752209, + "grad_norm": 5.8299384117126465, + "learning_rate": 9.9745709974571e-06, + "loss": 0.1246, + "step": 25441 + }, + { + "epoch": 11.827986982798699, + "grad_norm": 4.819369316101074, + "learning_rate": 9.974569997457001e-06, + "loss": 0.074, + "step": 25442 + }, + { + "epoch": 11.828451882845188, + "grad_norm": 8.85300350189209, + "learning_rate": 9.9745689974569e-06, + "loss": 0.0835, + "step": 25443 + }, + { + "epoch": 11.828916782891678, + "grad_norm": 5.72933292388916, + "learning_rate": 9.9745679974568e-06, + "loss": 0.0675, + "step": 25444 + }, + { + "epoch": 11.829381682938168, + "grad_norm": 1.8163477182388306, + "learning_rate": 9.974566997456701e-06, + "loss": 0.0102, + "step": 25445 + }, + { + "epoch": 11.829846582984658, + "grad_norm": 2.3662145137786865, + "learning_rate": 9.9745659974566e-06, + "loss": 0.0212, + "step": 25446 + }, + { + "epoch": 11.830311483031148, + "grad_norm": 3.9471659660339355, + "learning_rate": 9.974564997456501e-06, + "loss": 0.0651, + "step": 25447 + }, + { + "epoch": 11.83077638307764, + "grad_norm": 0.7026003003120422, + "learning_rate": 9.9745639974564e-06, + "loss": 0.0039, + "step": 25448 + }, + { + "epoch": 11.831241283124129, + "grad_norm": 0.6946069002151489, + "learning_rate": 9.9745629974563e-06, + "loss": 0.0068, + "step": 25449 + }, + { + "epoch": 11.831706183170619, + "grad_norm": 5.720332622528076, + "learning_rate": 9.9745619974562e-06, + "loss": 0.0877, + "step": 25450 + }, + { + "epoch": 11.832171083217109, + "grad_norm": 5.271731853485107, + "learning_rate": 9.9745609974561e-06, + "loss": 0.0732, + "step": 25451 + }, + { + "epoch": 11.832635983263598, + "grad_norm": 6.724781513214111, + "learning_rate": 9.974559997456e-06, + "loss": 0.0661, + "step": 25452 + }, + { + "epoch": 11.833100883310088, + "grad_norm": 5.199872970581055, + "learning_rate": 9.974558997455901e-06, + "loss": 0.0886, + "step": 25453 + }, + { + "epoch": 11.833565783356578, + "grad_norm": 5.43261194229126, + "learning_rate": 9.9745579974558e-06, + "loss": 0.0742, + "step": 25454 + }, + { + "epoch": 11.834030683403068, + "grad_norm": 4.016566753387451, + "learning_rate": 9.974556997455701e-06, + "loss": 0.0671, + "step": 25455 + }, + { + "epoch": 11.834495583449558, + "grad_norm": 3.3792340755462646, + "learning_rate": 9.9745559974556e-06, + "loss": 0.0307, + "step": 25456 + }, + { + "epoch": 11.83496048349605, + "grad_norm": 6.521294116973877, + "learning_rate": 9.9745549974555e-06, + "loss": 0.0726, + "step": 25457 + }, + { + "epoch": 11.835425383542539, + "grad_norm": 3.681889533996582, + "learning_rate": 9.974553997455401e-06, + "loss": 0.0354, + "step": 25458 + }, + { + "epoch": 11.835890283589029, + "grad_norm": 13.119203567504883, + "learning_rate": 9.9745529974553e-06, + "loss": 0.0444, + "step": 25459 + }, + { + "epoch": 11.836355183635519, + "grad_norm": 3.741769313812256, + "learning_rate": 9.974551997455201e-06, + "loss": 0.0489, + "step": 25460 + }, + { + "epoch": 11.836820083682008, + "grad_norm": 4.6294846534729, + "learning_rate": 9.9745509974551e-06, + "loss": 0.0716, + "step": 25461 + }, + { + "epoch": 11.837284983728498, + "grad_norm": 6.156795978546143, + "learning_rate": 9.974549997455e-06, + "loss": 0.0861, + "step": 25462 + }, + { + "epoch": 11.837749883774988, + "grad_norm": 8.019855499267578, + "learning_rate": 9.9745489974549e-06, + "loss": 0.145, + "step": 25463 + }, + { + "epoch": 11.838214783821478, + "grad_norm": 4.964631080627441, + "learning_rate": 9.9745479974548e-06, + "loss": 0.0769, + "step": 25464 + }, + { + "epoch": 11.83867968386797, + "grad_norm": 5.107486248016357, + "learning_rate": 9.974546997454701e-06, + "loss": 0.0731, + "step": 25465 + }, + { + "epoch": 11.83914458391446, + "grad_norm": 4.84255313873291, + "learning_rate": 9.974545997454602e-06, + "loss": 0.0712, + "step": 25466 + }, + { + "epoch": 11.839609483960949, + "grad_norm": 5.371371746063232, + "learning_rate": 9.9745449974545e-06, + "loss": 0.0907, + "step": 25467 + }, + { + "epoch": 11.840074384007439, + "grad_norm": 3.5644655227661133, + "learning_rate": 9.9745439974544e-06, + "loss": 0.048, + "step": 25468 + }, + { + "epoch": 11.840539284053929, + "grad_norm": 3.194622278213501, + "learning_rate": 9.9745429974543e-06, + "loss": 0.0367, + "step": 25469 + }, + { + "epoch": 11.841004184100418, + "grad_norm": 3.77946138381958, + "learning_rate": 9.9745419974542e-06, + "loss": 0.0639, + "step": 25470 + }, + { + "epoch": 11.841469084146908, + "grad_norm": 2.920572280883789, + "learning_rate": 9.974540997454101e-06, + "loss": 0.0408, + "step": 25471 + }, + { + "epoch": 11.841933984193398, + "grad_norm": 5.796776294708252, + "learning_rate": 9.974539997454e-06, + "loss": 0.0609, + "step": 25472 + }, + { + "epoch": 11.842398884239888, + "grad_norm": 4.841434478759766, + "learning_rate": 9.974538997453901e-06, + "loss": 0.0689, + "step": 25473 + }, + { + "epoch": 11.84286378428638, + "grad_norm": 6.229780673980713, + "learning_rate": 9.9745379974538e-06, + "loss": 0.0955, + "step": 25474 + }, + { + "epoch": 11.84332868433287, + "grad_norm": 4.672370433807373, + "learning_rate": 9.9745369974537e-06, + "loss": 0.08, + "step": 25475 + }, + { + "epoch": 11.843793584379359, + "grad_norm": 2.2176928520202637, + "learning_rate": 9.9745359974536e-06, + "loss": 0.0142, + "step": 25476 + }, + { + "epoch": 11.844258484425849, + "grad_norm": 3.815833568572998, + "learning_rate": 9.9745349974535e-06, + "loss": 0.0464, + "step": 25477 + }, + { + "epoch": 11.844723384472339, + "grad_norm": 5.008854866027832, + "learning_rate": 9.974533997453401e-06, + "loss": 0.0687, + "step": 25478 + }, + { + "epoch": 11.845188284518828, + "grad_norm": 6.752854824066162, + "learning_rate": 9.9745329974533e-06, + "loss": 0.129, + "step": 25479 + }, + { + "epoch": 11.845653184565318, + "grad_norm": 3.894209861755371, + "learning_rate": 9.9745319974532e-06, + "loss": 0.0464, + "step": 25480 + }, + { + "epoch": 11.846118084611808, + "grad_norm": 4.316963195800781, + "learning_rate": 9.9745309974531e-06, + "loss": 0.0399, + "step": 25481 + }, + { + "epoch": 11.846582984658298, + "grad_norm": 2.4213802814483643, + "learning_rate": 9.974529997453e-06, + "loss": 0.0359, + "step": 25482 + }, + { + "epoch": 11.84704788470479, + "grad_norm": 2.661633253097534, + "learning_rate": 9.974528997452901e-06, + "loss": 0.0259, + "step": 25483 + }, + { + "epoch": 11.84751278475128, + "grad_norm": 2.3659229278564453, + "learning_rate": 9.974527997452802e-06, + "loss": 0.0151, + "step": 25484 + }, + { + "epoch": 11.847977684797769, + "grad_norm": 3.866089105606079, + "learning_rate": 9.9745269974527e-06, + "loss": 0.0249, + "step": 25485 + }, + { + "epoch": 11.848442584844259, + "grad_norm": 6.809478759765625, + "learning_rate": 9.974525997452601e-06, + "loss": 0.1265, + "step": 25486 + }, + { + "epoch": 11.848907484890749, + "grad_norm": 6.513626575469971, + "learning_rate": 9.9745249974525e-06, + "loss": 0.0233, + "step": 25487 + }, + { + "epoch": 11.849372384937238, + "grad_norm": 5.609416484832764, + "learning_rate": 9.9745239974524e-06, + "loss": 0.0826, + "step": 25488 + }, + { + "epoch": 11.849837284983728, + "grad_norm": 3.191978693008423, + "learning_rate": 9.9745229974523e-06, + "loss": 0.0221, + "step": 25489 + }, + { + "epoch": 11.850302185030218, + "grad_norm": 6.913154125213623, + "learning_rate": 9.9745219974522e-06, + "loss": 0.0942, + "step": 25490 + }, + { + "epoch": 11.850767085076708, + "grad_norm": 8.259578704833984, + "learning_rate": 9.974520997452101e-06, + "loss": 0.1382, + "step": 25491 + }, + { + "epoch": 11.8512319851232, + "grad_norm": 6.800602436065674, + "learning_rate": 9.974519997452e-06, + "loss": 0.0868, + "step": 25492 + }, + { + "epoch": 11.851696885169689, + "grad_norm": 5.806674003601074, + "learning_rate": 9.9745189974519e-06, + "loss": 0.0876, + "step": 25493 + }, + { + "epoch": 11.852161785216179, + "grad_norm": 0.7077040672302246, + "learning_rate": 9.9745179974518e-06, + "loss": 0.0047, + "step": 25494 + }, + { + "epoch": 11.852626685262669, + "grad_norm": 2.2016448974609375, + "learning_rate": 9.9745169974517e-06, + "loss": 0.0104, + "step": 25495 + }, + { + "epoch": 11.853091585309159, + "grad_norm": 6.219363212585449, + "learning_rate": 9.974515997451601e-06, + "loss": 0.0844, + "step": 25496 + }, + { + "epoch": 11.853556485355648, + "grad_norm": 4.932008266448975, + "learning_rate": 9.974514997451502e-06, + "loss": 0.077, + "step": 25497 + }, + { + "epoch": 11.854021385402138, + "grad_norm": 7.14618444442749, + "learning_rate": 9.9745139974514e-06, + "loss": 0.0754, + "step": 25498 + }, + { + "epoch": 11.854486285448628, + "grad_norm": 6.761521339416504, + "learning_rate": 9.974512997451301e-06, + "loss": 0.0791, + "step": 25499 + }, + { + "epoch": 11.854951185495118, + "grad_norm": 2.5017595291137695, + "learning_rate": 9.9745119974512e-06, + "loss": 0.0205, + "step": 25500 + }, + { + "epoch": 11.85541608554161, + "grad_norm": 4.212823390960693, + "learning_rate": 9.974510997451101e-06, + "loss": 0.0333, + "step": 25501 + }, + { + "epoch": 11.855880985588099, + "grad_norm": 4.551053047180176, + "learning_rate": 9.974509997451e-06, + "loss": 0.0438, + "step": 25502 + }, + { + "epoch": 11.856345885634589, + "grad_norm": 3.325047492980957, + "learning_rate": 9.9745089974509e-06, + "loss": 0.0441, + "step": 25503 + }, + { + "epoch": 11.856810785681079, + "grad_norm": 5.672537326812744, + "learning_rate": 9.974507997450801e-06, + "loss": 0.0495, + "step": 25504 + }, + { + "epoch": 11.857275685727569, + "grad_norm": 7.866996765136719, + "learning_rate": 9.9745069974507e-06, + "loss": 0.1957, + "step": 25505 + }, + { + "epoch": 11.857740585774058, + "grad_norm": 8.508785247802734, + "learning_rate": 9.9745059974506e-06, + "loss": 0.1044, + "step": 25506 + }, + { + "epoch": 11.858205485820548, + "grad_norm": 4.722927570343018, + "learning_rate": 9.9745049974505e-06, + "loss": 0.0366, + "step": 25507 + }, + { + "epoch": 11.858670385867038, + "grad_norm": 6.260897636413574, + "learning_rate": 9.9745039974504e-06, + "loss": 0.0531, + "step": 25508 + }, + { + "epoch": 11.859135285913528, + "grad_norm": 3.176657199859619, + "learning_rate": 9.974502997450301e-06, + "loss": 0.0374, + "step": 25509 + }, + { + "epoch": 11.85960018596002, + "grad_norm": 4.684889793395996, + "learning_rate": 9.974501997450202e-06, + "loss": 0.0639, + "step": 25510 + }, + { + "epoch": 11.860065086006509, + "grad_norm": 4.490438461303711, + "learning_rate": 9.9745009974501e-06, + "loss": 0.0515, + "step": 25511 + }, + { + "epoch": 11.860529986052999, + "grad_norm": 6.449231147766113, + "learning_rate": 9.97449999745e-06, + "loss": 0.1004, + "step": 25512 + }, + { + "epoch": 11.860994886099489, + "grad_norm": 5.076138973236084, + "learning_rate": 9.9744989974499e-06, + "loss": 0.0597, + "step": 25513 + }, + { + "epoch": 11.861459786145979, + "grad_norm": 5.957280158996582, + "learning_rate": 9.974497997449801e-06, + "loss": 0.1564, + "step": 25514 + }, + { + "epoch": 11.861924686192468, + "grad_norm": 2.0136938095092773, + "learning_rate": 9.974496997449702e-06, + "loss": 0.0109, + "step": 25515 + }, + { + "epoch": 11.862389586238958, + "grad_norm": 2.3528599739074707, + "learning_rate": 9.9744959974496e-06, + "loss": 0.0243, + "step": 25516 + }, + { + "epoch": 11.862854486285448, + "grad_norm": 2.1483073234558105, + "learning_rate": 9.974494997449501e-06, + "loss": 0.0269, + "step": 25517 + }, + { + "epoch": 11.863319386331938, + "grad_norm": 2.0656354427337646, + "learning_rate": 9.9744939974494e-06, + "loss": 0.0193, + "step": 25518 + }, + { + "epoch": 11.86378428637843, + "grad_norm": 2.216942310333252, + "learning_rate": 9.974492997449301e-06, + "loss": 0.0235, + "step": 25519 + }, + { + "epoch": 11.864249186424919, + "grad_norm": 6.7314887046813965, + "learning_rate": 9.9744919974492e-06, + "loss": 0.1545, + "step": 25520 + }, + { + "epoch": 11.864714086471409, + "grad_norm": 6.30556583404541, + "learning_rate": 9.9744909974491e-06, + "loss": 0.1648, + "step": 25521 + }, + { + "epoch": 11.865178986517899, + "grad_norm": 6.571623802185059, + "learning_rate": 9.974489997449001e-06, + "loss": 0.0766, + "step": 25522 + }, + { + "epoch": 11.865643886564389, + "grad_norm": 4.821353435516357, + "learning_rate": 9.9744889974489e-06, + "loss": 0.0553, + "step": 25523 + }, + { + "epoch": 11.866108786610878, + "grad_norm": 5.344632148742676, + "learning_rate": 9.9744879974488e-06, + "loss": 0.0804, + "step": 25524 + }, + { + "epoch": 11.866573686657368, + "grad_norm": 3.603372573852539, + "learning_rate": 9.9744869974487e-06, + "loss": 0.0447, + "step": 25525 + }, + { + "epoch": 11.867038586703858, + "grad_norm": 4.6044840812683105, + "learning_rate": 9.9744859974486e-06, + "loss": 0.0576, + "step": 25526 + }, + { + "epoch": 11.86750348675035, + "grad_norm": 0.8986131548881531, + "learning_rate": 9.974484997448501e-06, + "loss": 0.0069, + "step": 25527 + }, + { + "epoch": 11.86796838679684, + "grad_norm": 5.713068008422852, + "learning_rate": 9.974483997448402e-06, + "loss": 0.1188, + "step": 25528 + }, + { + "epoch": 11.868433286843329, + "grad_norm": 2.4377548694610596, + "learning_rate": 9.9744829974483e-06, + "loss": 0.0205, + "step": 25529 + }, + { + "epoch": 11.868898186889819, + "grad_norm": 2.6752090454101562, + "learning_rate": 9.974481997448201e-06, + "loss": 0.0229, + "step": 25530 + }, + { + "epoch": 11.869363086936309, + "grad_norm": 2.972158193588257, + "learning_rate": 9.9744809974481e-06, + "loss": 0.0256, + "step": 25531 + }, + { + "epoch": 11.869827986982799, + "grad_norm": 3.2981820106506348, + "learning_rate": 9.974479997448001e-06, + "loss": 0.0422, + "step": 25532 + }, + { + "epoch": 11.870292887029288, + "grad_norm": 5.224760055541992, + "learning_rate": 9.9744789974479e-06, + "loss": 0.0737, + "step": 25533 + }, + { + "epoch": 11.870757787075778, + "grad_norm": 1.86558198928833, + "learning_rate": 9.9744779974478e-06, + "loss": 0.0161, + "step": 25534 + }, + { + "epoch": 11.871222687122268, + "grad_norm": 7.199378967285156, + "learning_rate": 9.974476997447701e-06, + "loss": 0.1445, + "step": 25535 + }, + { + "epoch": 11.87168758716876, + "grad_norm": 4.846868515014648, + "learning_rate": 9.9744759974476e-06, + "loss": 0.0631, + "step": 25536 + }, + { + "epoch": 11.87215248721525, + "grad_norm": 5.29465389251709, + "learning_rate": 9.974474997447501e-06, + "loss": 0.0823, + "step": 25537 + }, + { + "epoch": 11.872617387261739, + "grad_norm": 3.705455780029297, + "learning_rate": 9.9744739974474e-06, + "loss": 0.0442, + "step": 25538 + }, + { + "epoch": 11.873082287308229, + "grad_norm": 2.989377021789551, + "learning_rate": 9.9744729974473e-06, + "loss": 0.0438, + "step": 25539 + }, + { + "epoch": 11.873547187354719, + "grad_norm": 0.6568268537521362, + "learning_rate": 9.974471997447201e-06, + "loss": 0.0047, + "step": 25540 + }, + { + "epoch": 11.874012087401209, + "grad_norm": 6.383808135986328, + "learning_rate": 9.974470997447102e-06, + "loss": 0.1213, + "step": 25541 + }, + { + "epoch": 11.874476987447698, + "grad_norm": 5.371675968170166, + "learning_rate": 9.974469997447e-06, + "loss": 0.0974, + "step": 25542 + }, + { + "epoch": 11.874941887494188, + "grad_norm": 4.659996509552002, + "learning_rate": 9.974468997446901e-06, + "loss": 0.047, + "step": 25543 + }, + { + "epoch": 11.87540678754068, + "grad_norm": 5.960539817810059, + "learning_rate": 9.9744679974468e-06, + "loss": 0.1274, + "step": 25544 + }, + { + "epoch": 11.87587168758717, + "grad_norm": 3.6810100078582764, + "learning_rate": 9.9744669974467e-06, + "loss": 0.0354, + "step": 25545 + }, + { + "epoch": 11.87633658763366, + "grad_norm": 4.687997341156006, + "learning_rate": 9.9744659974466e-06, + "loss": 0.0351, + "step": 25546 + }, + { + "epoch": 11.876801487680149, + "grad_norm": 5.360324859619141, + "learning_rate": 9.9744649974465e-06, + "loss": 0.0821, + "step": 25547 + }, + { + "epoch": 11.877266387726639, + "grad_norm": 6.119870185852051, + "learning_rate": 9.974463997446401e-06, + "loss": 0.1198, + "step": 25548 + }, + { + "epoch": 11.877731287773129, + "grad_norm": 1.7201112508773804, + "learning_rate": 9.9744629974463e-06, + "loss": 0.007, + "step": 25549 + }, + { + "epoch": 11.878196187819619, + "grad_norm": 2.7756378650665283, + "learning_rate": 9.974461997446201e-06, + "loss": 0.018, + "step": 25550 + }, + { + "epoch": 11.878661087866108, + "grad_norm": 5.173830032348633, + "learning_rate": 9.9744609974461e-06, + "loss": 0.0736, + "step": 25551 + }, + { + "epoch": 11.879125987912598, + "grad_norm": 5.365025520324707, + "learning_rate": 9.974459997446e-06, + "loss": 0.0394, + "step": 25552 + }, + { + "epoch": 11.87959088795909, + "grad_norm": 3.446821451187134, + "learning_rate": 9.974458997445901e-06, + "loss": 0.0337, + "step": 25553 + }, + { + "epoch": 11.88005578800558, + "grad_norm": 3.278836727142334, + "learning_rate": 9.9744579974458e-06, + "loss": 0.0297, + "step": 25554 + }, + { + "epoch": 11.88052068805207, + "grad_norm": 0.5575905442237854, + "learning_rate": 9.9744569974457e-06, + "loss": 0.0025, + "step": 25555 + }, + { + "epoch": 11.880985588098559, + "grad_norm": 4.734559535980225, + "learning_rate": 9.9744559974456e-06, + "loss": 0.0663, + "step": 25556 + }, + { + "epoch": 11.881450488145049, + "grad_norm": 5.199069499969482, + "learning_rate": 9.9744549974455e-06, + "loss": 0.0628, + "step": 25557 + }, + { + "epoch": 11.881915388191539, + "grad_norm": 8.506166458129883, + "learning_rate": 9.9744539974454e-06, + "loss": 0.1685, + "step": 25558 + }, + { + "epoch": 11.882380288238028, + "grad_norm": 3.5967118740081787, + "learning_rate": 9.9744529974453e-06, + "loss": 0.0388, + "step": 25559 + }, + { + "epoch": 11.882845188284518, + "grad_norm": 4.261318206787109, + "learning_rate": 9.9744519974452e-06, + "loss": 0.0395, + "step": 25560 + }, + { + "epoch": 11.883310088331008, + "grad_norm": 4.235098838806152, + "learning_rate": 9.974450997445101e-06, + "loss": 0.0628, + "step": 25561 + }, + { + "epoch": 11.8837749883775, + "grad_norm": 7.1299638748168945, + "learning_rate": 9.974449997445e-06, + "loss": 0.1144, + "step": 25562 + }, + { + "epoch": 11.88423988842399, + "grad_norm": 8.480879783630371, + "learning_rate": 9.974448997444901e-06, + "loss": 0.195, + "step": 25563 + }, + { + "epoch": 11.88470478847048, + "grad_norm": 6.125485897064209, + "learning_rate": 9.9744479974448e-06, + "loss": 0.0994, + "step": 25564 + }, + { + "epoch": 11.885169688516969, + "grad_norm": 3.7764060497283936, + "learning_rate": 9.9744469974447e-06, + "loss": 0.0387, + "step": 25565 + }, + { + "epoch": 11.885634588563459, + "grad_norm": 3.879319429397583, + "learning_rate": 9.974445997444601e-06, + "loss": 0.0374, + "step": 25566 + }, + { + "epoch": 11.886099488609949, + "grad_norm": 6.176530361175537, + "learning_rate": 9.9744449974445e-06, + "loss": 0.0974, + "step": 25567 + }, + { + "epoch": 11.886564388656438, + "grad_norm": 13.035266876220703, + "learning_rate": 9.974443997444401e-06, + "loss": 0.082, + "step": 25568 + }, + { + "epoch": 11.887029288702928, + "grad_norm": 8.616996765136719, + "learning_rate": 9.9744429974443e-06, + "loss": 0.1676, + "step": 25569 + }, + { + "epoch": 11.887494188749418, + "grad_norm": 4.351804256439209, + "learning_rate": 9.9744419974442e-06, + "loss": 0.0415, + "step": 25570 + }, + { + "epoch": 11.88795908879591, + "grad_norm": 9.243152618408203, + "learning_rate": 9.9744409974441e-06, + "loss": 0.0906, + "step": 25571 + }, + { + "epoch": 11.8884239888424, + "grad_norm": 5.260819911956787, + "learning_rate": 9.974439997444e-06, + "loss": 0.04, + "step": 25572 + }, + { + "epoch": 11.88888888888889, + "grad_norm": 5.4855756759643555, + "learning_rate": 9.9744389974439e-06, + "loss": 0.0435, + "step": 25573 + }, + { + "epoch": 11.889353788935379, + "grad_norm": 5.9162211418151855, + "learning_rate": 9.974437997443801e-06, + "loss": 0.1015, + "step": 25574 + }, + { + "epoch": 11.889818688981869, + "grad_norm": 6.696743965148926, + "learning_rate": 9.9744369974437e-06, + "loss": 0.1183, + "step": 25575 + }, + { + "epoch": 11.890283589028359, + "grad_norm": 5.437331676483154, + "learning_rate": 9.9744359974436e-06, + "loss": 0.0441, + "step": 25576 + }, + { + "epoch": 11.890748489074848, + "grad_norm": 4.905808925628662, + "learning_rate": 9.9744349974435e-06, + "loss": 0.0671, + "step": 25577 + }, + { + "epoch": 11.891213389121338, + "grad_norm": 7.121293544769287, + "learning_rate": 9.9744339974434e-06, + "loss": 0.1431, + "step": 25578 + }, + { + "epoch": 11.891678289167828, + "grad_norm": 5.442352771759033, + "learning_rate": 9.974432997443301e-06, + "loss": 0.0476, + "step": 25579 + }, + { + "epoch": 11.89214318921432, + "grad_norm": 6.072299957275391, + "learning_rate": 9.9744319974432e-06, + "loss": 0.1269, + "step": 25580 + }, + { + "epoch": 11.89260808926081, + "grad_norm": 4.524803638458252, + "learning_rate": 9.974430997443101e-06, + "loss": 0.0812, + "step": 25581 + }, + { + "epoch": 11.8930729893073, + "grad_norm": 4.862587928771973, + "learning_rate": 9.974429997443e-06, + "loss": 0.0424, + "step": 25582 + }, + { + "epoch": 11.893537889353789, + "grad_norm": 1.2318545579910278, + "learning_rate": 9.9744289974429e-06, + "loss": 0.0123, + "step": 25583 + }, + { + "epoch": 11.894002789400279, + "grad_norm": 8.443967819213867, + "learning_rate": 9.974427997442801e-06, + "loss": 0.048, + "step": 25584 + }, + { + "epoch": 11.894467689446769, + "grad_norm": 2.3667097091674805, + "learning_rate": 9.9744269974427e-06, + "loss": 0.0256, + "step": 25585 + }, + { + "epoch": 11.894932589493258, + "grad_norm": 5.350236415863037, + "learning_rate": 9.974425997442601e-06, + "loss": 0.0945, + "step": 25586 + }, + { + "epoch": 11.895397489539748, + "grad_norm": 1.9700285196304321, + "learning_rate": 9.9744249974425e-06, + "loss": 0.0191, + "step": 25587 + }, + { + "epoch": 11.895862389586238, + "grad_norm": 1.7952066659927368, + "learning_rate": 9.9744239974424e-06, + "loss": 0.0211, + "step": 25588 + }, + { + "epoch": 11.89632728963273, + "grad_norm": 2.0278360843658447, + "learning_rate": 9.9744229974423e-06, + "loss": 0.01, + "step": 25589 + }, + { + "epoch": 11.89679218967922, + "grad_norm": 5.221533298492432, + "learning_rate": 9.9744219974422e-06, + "loss": 0.0925, + "step": 25590 + }, + { + "epoch": 11.89725708972571, + "grad_norm": 4.022729873657227, + "learning_rate": 9.9744209974421e-06, + "loss": 0.0362, + "step": 25591 + }, + { + "epoch": 11.897721989772199, + "grad_norm": 3.847848415374756, + "learning_rate": 9.974419997442001e-06, + "loss": 0.0563, + "step": 25592 + }, + { + "epoch": 11.898186889818689, + "grad_norm": 7.132491111755371, + "learning_rate": 9.9744189974419e-06, + "loss": 0.0782, + "step": 25593 + }, + { + "epoch": 11.898651789865179, + "grad_norm": 1.859285831451416, + "learning_rate": 9.974417997441801e-06, + "loss": 0.0158, + "step": 25594 + }, + { + "epoch": 11.899116689911668, + "grad_norm": 2.5728535652160645, + "learning_rate": 9.9744169974417e-06, + "loss": 0.0187, + "step": 25595 + }, + { + "epoch": 11.899581589958158, + "grad_norm": 6.585933208465576, + "learning_rate": 9.9744159974416e-06, + "loss": 0.0434, + "step": 25596 + }, + { + "epoch": 11.900046490004648, + "grad_norm": 4.474940776824951, + "learning_rate": 9.974414997441501e-06, + "loss": 0.0859, + "step": 25597 + }, + { + "epoch": 11.90051139005114, + "grad_norm": 1.7821364402770996, + "learning_rate": 9.9744139974414e-06, + "loss": 0.0146, + "step": 25598 + }, + { + "epoch": 11.90097629009763, + "grad_norm": 4.989114284515381, + "learning_rate": 9.974412997441301e-06, + "loss": 0.0441, + "step": 25599 + }, + { + "epoch": 11.90144119014412, + "grad_norm": 0.9931098222732544, + "learning_rate": 9.9744119974412e-06, + "loss": 0.0075, + "step": 25600 + }, + { + "epoch": 11.901906090190609, + "grad_norm": 0.927808940410614, + "learning_rate": 9.9744109974411e-06, + "loss": 0.0099, + "step": 25601 + }, + { + "epoch": 11.902370990237099, + "grad_norm": 3.2173221111297607, + "learning_rate": 9.974409997441e-06, + "loss": 0.0349, + "step": 25602 + }, + { + "epoch": 11.902835890283589, + "grad_norm": 3.979079246520996, + "learning_rate": 9.9744089974409e-06, + "loss": 0.044, + "step": 25603 + }, + { + "epoch": 11.903300790330078, + "grad_norm": 5.014030933380127, + "learning_rate": 9.974407997440801e-06, + "loss": 0.0919, + "step": 25604 + }, + { + "epoch": 11.903765690376568, + "grad_norm": 3.4452285766601562, + "learning_rate": 9.974406997440702e-06, + "loss": 0.0269, + "step": 25605 + }, + { + "epoch": 11.90423059042306, + "grad_norm": 2.1244595050811768, + "learning_rate": 9.9744059974406e-06, + "loss": 0.0173, + "step": 25606 + }, + { + "epoch": 11.90469549046955, + "grad_norm": 10.458073616027832, + "learning_rate": 9.974404997440501e-06, + "loss": 0.0775, + "step": 25607 + }, + { + "epoch": 11.90516039051604, + "grad_norm": 5.6428375244140625, + "learning_rate": 9.9744039974404e-06, + "loss": 0.0708, + "step": 25608 + }, + { + "epoch": 11.90562529056253, + "grad_norm": 4.505288600921631, + "learning_rate": 9.9744029974403e-06, + "loss": 0.0544, + "step": 25609 + }, + { + "epoch": 11.906090190609019, + "grad_norm": 4.299890041351318, + "learning_rate": 9.974401997440201e-06, + "loss": 0.0342, + "step": 25610 + }, + { + "epoch": 11.906555090655509, + "grad_norm": 3.0707974433898926, + "learning_rate": 9.9744009974401e-06, + "loss": 0.0321, + "step": 25611 + }, + { + "epoch": 11.907019990701999, + "grad_norm": 5.151762008666992, + "learning_rate": 9.974399997440001e-06, + "loss": 0.0787, + "step": 25612 + }, + { + "epoch": 11.907484890748488, + "grad_norm": 3.892995834350586, + "learning_rate": 9.9743989974399e-06, + "loss": 0.0542, + "step": 25613 + }, + { + "epoch": 11.907949790794978, + "grad_norm": 10.404370307922363, + "learning_rate": 9.9743979974398e-06, + "loss": 0.1674, + "step": 25614 + }, + { + "epoch": 11.90841469084147, + "grad_norm": 1.6324743032455444, + "learning_rate": 9.9743969974397e-06, + "loss": 0.0102, + "step": 25615 + }, + { + "epoch": 11.90887959088796, + "grad_norm": 2.629981279373169, + "learning_rate": 9.9743959974396e-06, + "loss": 0.0243, + "step": 25616 + }, + { + "epoch": 11.90934449093445, + "grad_norm": 6.0376811027526855, + "learning_rate": 9.974394997439501e-06, + "loss": 0.1167, + "step": 25617 + }, + { + "epoch": 11.90980939098094, + "grad_norm": 4.743083953857422, + "learning_rate": 9.974393997439402e-06, + "loss": 0.0359, + "step": 25618 + }, + { + "epoch": 11.910274291027429, + "grad_norm": 5.966331958770752, + "learning_rate": 9.9743929974393e-06, + "loss": 0.0583, + "step": 25619 + }, + { + "epoch": 11.910739191073919, + "grad_norm": 4.711549282073975, + "learning_rate": 9.9743919974392e-06, + "loss": 0.0803, + "step": 25620 + }, + { + "epoch": 11.911204091120409, + "grad_norm": 8.148451805114746, + "learning_rate": 9.9743909974391e-06, + "loss": 0.2155, + "step": 25621 + }, + { + "epoch": 11.911668991166898, + "grad_norm": 4.472845077514648, + "learning_rate": 9.974389997439001e-06, + "loss": 0.0515, + "step": 25622 + }, + { + "epoch": 11.91213389121339, + "grad_norm": 3.063106060028076, + "learning_rate": 9.974388997438902e-06, + "loss": 0.0351, + "step": 25623 + }, + { + "epoch": 11.91259879125988, + "grad_norm": 4.5018134117126465, + "learning_rate": 9.9743879974388e-06, + "loss": 0.0727, + "step": 25624 + }, + { + "epoch": 11.91306369130637, + "grad_norm": 4.94244909286499, + "learning_rate": 9.974386997438701e-06, + "loss": 0.0473, + "step": 25625 + }, + { + "epoch": 11.91352859135286, + "grad_norm": 6.257122993469238, + "learning_rate": 9.9743859974386e-06, + "loss": 0.0501, + "step": 25626 + }, + { + "epoch": 11.91399349139935, + "grad_norm": 9.372636795043945, + "learning_rate": 9.9743849974385e-06, + "loss": 0.1578, + "step": 25627 + }, + { + "epoch": 11.914458391445839, + "grad_norm": 3.568061351776123, + "learning_rate": 9.9743839974384e-06, + "loss": 0.0587, + "step": 25628 + }, + { + "epoch": 11.914923291492329, + "grad_norm": 2.6582891941070557, + "learning_rate": 9.9743829974383e-06, + "loss": 0.0237, + "step": 25629 + }, + { + "epoch": 11.915388191538819, + "grad_norm": 3.5294270515441895, + "learning_rate": 9.974381997438201e-06, + "loss": 0.0324, + "step": 25630 + }, + { + "epoch": 11.915853091585308, + "grad_norm": 1.6578171253204346, + "learning_rate": 9.9743809974381e-06, + "loss": 0.0166, + "step": 25631 + }, + { + "epoch": 11.9163179916318, + "grad_norm": 2.819711446762085, + "learning_rate": 9.974379997438e-06, + "loss": 0.0278, + "step": 25632 + }, + { + "epoch": 11.91678289167829, + "grad_norm": 4.733428955078125, + "learning_rate": 9.9743789974379e-06, + "loss": 0.0561, + "step": 25633 + }, + { + "epoch": 11.91724779172478, + "grad_norm": 4.060757160186768, + "learning_rate": 9.9743779974378e-06, + "loss": 0.0438, + "step": 25634 + }, + { + "epoch": 11.91771269177127, + "grad_norm": 2.2604775428771973, + "learning_rate": 9.974376997437701e-06, + "loss": 0.0162, + "step": 25635 + }, + { + "epoch": 11.91817759181776, + "grad_norm": 1.9453126192092896, + "learning_rate": 9.974375997437602e-06, + "loss": 0.0236, + "step": 25636 + }, + { + "epoch": 11.918642491864249, + "grad_norm": 3.6259288787841797, + "learning_rate": 9.9743749974375e-06, + "loss": 0.0389, + "step": 25637 + }, + { + "epoch": 11.919107391910739, + "grad_norm": 2.213535785675049, + "learning_rate": 9.974373997437401e-06, + "loss": 0.0185, + "step": 25638 + }, + { + "epoch": 11.919572291957229, + "grad_norm": 5.727424621582031, + "learning_rate": 9.9743729974373e-06, + "loss": 0.1285, + "step": 25639 + }, + { + "epoch": 11.920037192003718, + "grad_norm": 5.319018363952637, + "learning_rate": 9.9743719974372e-06, + "loss": 0.067, + "step": 25640 + }, + { + "epoch": 11.92050209205021, + "grad_norm": 1.94673752784729, + "learning_rate": 9.9743709974371e-06, + "loss": 0.0202, + "step": 25641 + }, + { + "epoch": 11.9209669920967, + "grad_norm": 4.7547783851623535, + "learning_rate": 9.974369997437e-06, + "loss": 0.047, + "step": 25642 + }, + { + "epoch": 11.92143189214319, + "grad_norm": 3.04708194732666, + "learning_rate": 9.974368997436901e-06, + "loss": 0.0303, + "step": 25643 + }, + { + "epoch": 11.92189679218968, + "grad_norm": 6.448512077331543, + "learning_rate": 9.9743679974368e-06, + "loss": 0.1204, + "step": 25644 + }, + { + "epoch": 11.92236169223617, + "grad_norm": 13.815359115600586, + "learning_rate": 9.9743669974367e-06, + "loss": 0.0682, + "step": 25645 + }, + { + "epoch": 11.922826592282659, + "grad_norm": 5.7688798904418945, + "learning_rate": 9.9743659974366e-06, + "loss": 0.0767, + "step": 25646 + }, + { + "epoch": 11.923291492329149, + "grad_norm": 3.5237693786621094, + "learning_rate": 9.9743649974365e-06, + "loss": 0.0556, + "step": 25647 + }, + { + "epoch": 11.923756392375639, + "grad_norm": 7.782510280609131, + "learning_rate": 9.974363997436401e-06, + "loss": 0.1334, + "step": 25648 + }, + { + "epoch": 11.924221292422128, + "grad_norm": 4.115792274475098, + "learning_rate": 9.974362997436302e-06, + "loss": 0.0342, + "step": 25649 + }, + { + "epoch": 11.92468619246862, + "grad_norm": 2.783372163772583, + "learning_rate": 9.9743619974362e-06, + "loss": 0.0321, + "step": 25650 + }, + { + "epoch": 11.92515109251511, + "grad_norm": 4.37965202331543, + "learning_rate": 9.974360997436101e-06, + "loss": 0.1002, + "step": 25651 + }, + { + "epoch": 11.9256159925616, + "grad_norm": 1.8080509901046753, + "learning_rate": 9.974359997436e-06, + "loss": 0.0162, + "step": 25652 + }, + { + "epoch": 11.92608089260809, + "grad_norm": 1.7002543210983276, + "learning_rate": 9.974358997435901e-06, + "loss": 0.0136, + "step": 25653 + }, + { + "epoch": 11.92654579265458, + "grad_norm": 5.742440223693848, + "learning_rate": 9.9743579974358e-06, + "loss": 0.0813, + "step": 25654 + }, + { + "epoch": 11.927010692701069, + "grad_norm": 6.624496936798096, + "learning_rate": 9.9743569974357e-06, + "loss": 0.0767, + "step": 25655 + }, + { + "epoch": 11.927475592747559, + "grad_norm": 7.8048505783081055, + "learning_rate": 9.974355997435601e-06, + "loss": 0.1664, + "step": 25656 + }, + { + "epoch": 11.927940492794049, + "grad_norm": 6.567562580108643, + "learning_rate": 9.9743549974355e-06, + "loss": 0.0644, + "step": 25657 + }, + { + "epoch": 11.928405392840538, + "grad_norm": 4.703021049499512, + "learning_rate": 9.9743539974354e-06, + "loss": 0.0647, + "step": 25658 + }, + { + "epoch": 11.92887029288703, + "grad_norm": 3.919708251953125, + "learning_rate": 9.9743529974353e-06, + "loss": 0.0506, + "step": 25659 + }, + { + "epoch": 11.92933519293352, + "grad_norm": 3.5002543926239014, + "learning_rate": 9.9743519974352e-06, + "loss": 0.0308, + "step": 25660 + }, + { + "epoch": 11.92980009298001, + "grad_norm": 5.180441379547119, + "learning_rate": 9.974350997435101e-06, + "loss": 0.083, + "step": 25661 + }, + { + "epoch": 11.9302649930265, + "grad_norm": 5.791745662689209, + "learning_rate": 9.974349997435002e-06, + "loss": 0.0716, + "step": 25662 + }, + { + "epoch": 11.93072989307299, + "grad_norm": 7.186489105224609, + "learning_rate": 9.9743489974349e-06, + "loss": 0.115, + "step": 25663 + }, + { + "epoch": 11.931194793119479, + "grad_norm": 1.6264183521270752, + "learning_rate": 9.9743479974348e-06, + "loss": 0.0123, + "step": 25664 + }, + { + "epoch": 11.931659693165969, + "grad_norm": 6.447575569152832, + "learning_rate": 9.9743469974347e-06, + "loss": 0.0712, + "step": 25665 + }, + { + "epoch": 11.932124593212459, + "grad_norm": 4.857420921325684, + "learning_rate": 9.974345997434601e-06, + "loss": 0.0428, + "step": 25666 + }, + { + "epoch": 11.932589493258948, + "grad_norm": 7.7082743644714355, + "learning_rate": 9.9743449974345e-06, + "loss": 0.0587, + "step": 25667 + }, + { + "epoch": 11.93305439330544, + "grad_norm": 5.394460678100586, + "learning_rate": 9.9743439974344e-06, + "loss": 0.0556, + "step": 25668 + }, + { + "epoch": 11.93351929335193, + "grad_norm": 4.926642417907715, + "learning_rate": 9.974342997434301e-06, + "loss": 0.0497, + "step": 25669 + }, + { + "epoch": 11.93398419339842, + "grad_norm": 3.902202844619751, + "learning_rate": 9.9743419974342e-06, + "loss": 0.0434, + "step": 25670 + }, + { + "epoch": 11.93444909344491, + "grad_norm": 7.007826328277588, + "learning_rate": 9.974340997434101e-06, + "loss": 0.0794, + "step": 25671 + }, + { + "epoch": 11.9349139934914, + "grad_norm": 4.859917640686035, + "learning_rate": 9.974339997434e-06, + "loss": 0.0413, + "step": 25672 + }, + { + "epoch": 11.935378893537889, + "grad_norm": 1.391388177871704, + "learning_rate": 9.9743389974339e-06, + "loss": 0.014, + "step": 25673 + }, + { + "epoch": 11.935843793584379, + "grad_norm": 3.7301883697509766, + "learning_rate": 9.974337997433801e-06, + "loss": 0.056, + "step": 25674 + }, + { + "epoch": 11.936308693630869, + "grad_norm": 8.727446556091309, + "learning_rate": 9.9743369974337e-06, + "loss": 0.1247, + "step": 25675 + }, + { + "epoch": 11.936773593677358, + "grad_norm": 5.03809118270874, + "learning_rate": 9.9743359974336e-06, + "loss": 0.0706, + "step": 25676 + }, + { + "epoch": 11.93723849372385, + "grad_norm": 4.937686920166016, + "learning_rate": 9.9743349974335e-06, + "loss": 0.1005, + "step": 25677 + }, + { + "epoch": 11.93770339377034, + "grad_norm": 5.788722515106201, + "learning_rate": 9.9743339974334e-06, + "loss": 0.1146, + "step": 25678 + }, + { + "epoch": 11.93816829381683, + "grad_norm": 2.5575103759765625, + "learning_rate": 9.974332997433301e-06, + "loss": 0.0212, + "step": 25679 + }, + { + "epoch": 11.93863319386332, + "grad_norm": 3.693593740463257, + "learning_rate": 9.974331997433202e-06, + "loss": 0.0364, + "step": 25680 + }, + { + "epoch": 11.93909809390981, + "grad_norm": 4.9722490310668945, + "learning_rate": 9.9743309974331e-06, + "loss": 0.0653, + "step": 25681 + }, + { + "epoch": 11.939562993956299, + "grad_norm": 7.284401893615723, + "learning_rate": 9.974329997433001e-06, + "loss": 0.122, + "step": 25682 + }, + { + "epoch": 11.940027894002789, + "grad_norm": 4.713525295257568, + "learning_rate": 9.9743289974329e-06, + "loss": 0.0472, + "step": 25683 + }, + { + "epoch": 11.940492794049279, + "grad_norm": 4.9925947189331055, + "learning_rate": 9.974327997432801e-06, + "loss": 0.0386, + "step": 25684 + }, + { + "epoch": 11.94095769409577, + "grad_norm": 3.9970571994781494, + "learning_rate": 9.9743269974327e-06, + "loss": 0.0293, + "step": 25685 + }, + { + "epoch": 11.94142259414226, + "grad_norm": 2.604320526123047, + "learning_rate": 9.9743259974326e-06, + "loss": 0.0187, + "step": 25686 + }, + { + "epoch": 11.94188749418875, + "grad_norm": 2.892357587814331, + "learning_rate": 9.974324997432501e-06, + "loss": 0.0217, + "step": 25687 + }, + { + "epoch": 11.94235239423524, + "grad_norm": 4.676212310791016, + "learning_rate": 9.9743239974324e-06, + "loss": 0.0696, + "step": 25688 + }, + { + "epoch": 11.94281729428173, + "grad_norm": 5.299905776977539, + "learning_rate": 9.974322997432301e-06, + "loss": 0.0796, + "step": 25689 + }, + { + "epoch": 11.94328219432822, + "grad_norm": 2.9035682678222656, + "learning_rate": 9.9743219974322e-06, + "loss": 0.0226, + "step": 25690 + }, + { + "epoch": 11.943747094374709, + "grad_norm": 3.0454139709472656, + "learning_rate": 9.9743209974321e-06, + "loss": 0.0255, + "step": 25691 + }, + { + "epoch": 11.944211994421199, + "grad_norm": 6.084761142730713, + "learning_rate": 9.974319997432001e-06, + "loss": 0.0642, + "step": 25692 + }, + { + "epoch": 11.944676894467689, + "grad_norm": 5.878023147583008, + "learning_rate": 9.974318997431902e-06, + "loss": 0.0945, + "step": 25693 + }, + { + "epoch": 11.94514179451418, + "grad_norm": 6.9016432762146, + "learning_rate": 9.9743179974318e-06, + "loss": 0.1464, + "step": 25694 + }, + { + "epoch": 11.94560669456067, + "grad_norm": 1.8446604013442993, + "learning_rate": 9.9743169974317e-06, + "loss": 0.0128, + "step": 25695 + }, + { + "epoch": 11.94607159460716, + "grad_norm": 3.0348143577575684, + "learning_rate": 9.9743159974316e-06, + "loss": 0.0373, + "step": 25696 + }, + { + "epoch": 11.94653649465365, + "grad_norm": 3.703610897064209, + "learning_rate": 9.9743149974315e-06, + "loss": 0.0281, + "step": 25697 + }, + { + "epoch": 11.94700139470014, + "grad_norm": 4.835699081420898, + "learning_rate": 9.9743139974314e-06, + "loss": 0.0702, + "step": 25698 + }, + { + "epoch": 11.94746629474663, + "grad_norm": 1.135240912437439, + "learning_rate": 9.9743129974313e-06, + "loss": 0.0115, + "step": 25699 + }, + { + "epoch": 11.947931194793119, + "grad_norm": 6.596284866333008, + "learning_rate": 9.974311997431201e-06, + "loss": 0.0756, + "step": 25700 + }, + { + "epoch": 11.948396094839609, + "grad_norm": 5.301661968231201, + "learning_rate": 9.9743109974311e-06, + "loss": 0.0383, + "step": 25701 + }, + { + "epoch": 11.9488609948861, + "grad_norm": 3.1263396739959717, + "learning_rate": 9.974309997431001e-06, + "loss": 0.0286, + "step": 25702 + }, + { + "epoch": 11.94932589493259, + "grad_norm": 3.8358376026153564, + "learning_rate": 9.9743089974309e-06, + "loss": 0.0324, + "step": 25703 + }, + { + "epoch": 11.94979079497908, + "grad_norm": 3.257388114929199, + "learning_rate": 9.9743079974308e-06, + "loss": 0.0343, + "step": 25704 + }, + { + "epoch": 11.95025569502557, + "grad_norm": 3.7285256385803223, + "learning_rate": 9.974306997430701e-06, + "loss": 0.0392, + "step": 25705 + }, + { + "epoch": 11.95072059507206, + "grad_norm": 5.487644195556641, + "learning_rate": 9.9743059974306e-06, + "loss": 0.0298, + "step": 25706 + }, + { + "epoch": 11.95118549511855, + "grad_norm": 7.414536476135254, + "learning_rate": 9.974304997430501e-06, + "loss": 0.1224, + "step": 25707 + }, + { + "epoch": 11.95165039516504, + "grad_norm": 6.199008941650391, + "learning_rate": 9.9743039974304e-06, + "loss": 0.0712, + "step": 25708 + }, + { + "epoch": 11.952115295211529, + "grad_norm": 3.7752983570098877, + "learning_rate": 9.9743029974303e-06, + "loss": 0.0547, + "step": 25709 + }, + { + "epoch": 11.952580195258019, + "grad_norm": 8.359708786010742, + "learning_rate": 9.9743019974302e-06, + "loss": 0.0735, + "step": 25710 + }, + { + "epoch": 11.95304509530451, + "grad_norm": 1.4887619018554688, + "learning_rate": 9.9743009974301e-06, + "loss": 0.014, + "step": 25711 + }, + { + "epoch": 11.953509995351, + "grad_norm": 2.727921962738037, + "learning_rate": 9.97429999743e-06, + "loss": 0.0155, + "step": 25712 + }, + { + "epoch": 11.95397489539749, + "grad_norm": 4.444553852081299, + "learning_rate": 9.974298997429901e-06, + "loss": 0.0607, + "step": 25713 + }, + { + "epoch": 11.95443979544398, + "grad_norm": 4.043735980987549, + "learning_rate": 9.9742979974298e-06, + "loss": 0.0482, + "step": 25714 + }, + { + "epoch": 11.95490469549047, + "grad_norm": 7.545122146606445, + "learning_rate": 9.974296997429701e-06, + "loss": 0.1469, + "step": 25715 + }, + { + "epoch": 11.95536959553696, + "grad_norm": 4.041855335235596, + "learning_rate": 9.9742959974296e-06, + "loss": 0.0547, + "step": 25716 + }, + { + "epoch": 11.95583449558345, + "grad_norm": 5.744326114654541, + "learning_rate": 9.9742949974295e-06, + "loss": 0.029, + "step": 25717 + }, + { + "epoch": 11.956299395629939, + "grad_norm": 4.489951133728027, + "learning_rate": 9.974293997429401e-06, + "loss": 0.0425, + "step": 25718 + }, + { + "epoch": 11.956764295676429, + "grad_norm": 4.947884559631348, + "learning_rate": 9.9742929974293e-06, + "loss": 0.0541, + "step": 25719 + }, + { + "epoch": 11.95722919572292, + "grad_norm": 5.379580020904541, + "learning_rate": 9.974291997429201e-06, + "loss": 0.0476, + "step": 25720 + }, + { + "epoch": 11.95769409576941, + "grad_norm": 5.673581123352051, + "learning_rate": 9.9742909974291e-06, + "loss": 0.0741, + "step": 25721 + }, + { + "epoch": 11.9581589958159, + "grad_norm": 5.362267017364502, + "learning_rate": 9.974289997429e-06, + "loss": 0.0395, + "step": 25722 + }, + { + "epoch": 11.95862389586239, + "grad_norm": 6.009006500244141, + "learning_rate": 9.9742889974289e-06, + "loss": 0.0824, + "step": 25723 + }, + { + "epoch": 11.95908879590888, + "grad_norm": 2.049381732940674, + "learning_rate": 9.9742879974288e-06, + "loss": 0.0187, + "step": 25724 + }, + { + "epoch": 11.95955369595537, + "grad_norm": 2.71186900138855, + "learning_rate": 9.9742869974287e-06, + "loss": 0.0209, + "step": 25725 + }, + { + "epoch": 11.96001859600186, + "grad_norm": 3.281730890274048, + "learning_rate": 9.974285997428602e-06, + "loss": 0.0329, + "step": 25726 + }, + { + "epoch": 11.960483496048349, + "grad_norm": 6.2815093994140625, + "learning_rate": 9.9742849974285e-06, + "loss": 0.0862, + "step": 25727 + }, + { + "epoch": 11.960948396094839, + "grad_norm": 3.439180850982666, + "learning_rate": 9.9742839974284e-06, + "loss": 0.0313, + "step": 25728 + }, + { + "epoch": 11.96141329614133, + "grad_norm": 3.2584755420684814, + "learning_rate": 9.9742829974283e-06, + "loss": 0.034, + "step": 25729 + }, + { + "epoch": 11.96187819618782, + "grad_norm": 5.991876125335693, + "learning_rate": 9.9742819974282e-06, + "loss": 0.0562, + "step": 25730 + }, + { + "epoch": 11.96234309623431, + "grad_norm": 5.386407375335693, + "learning_rate": 9.974280997428101e-06, + "loss": 0.0869, + "step": 25731 + }, + { + "epoch": 11.9628079962808, + "grad_norm": 4.550567150115967, + "learning_rate": 9.974279997428e-06, + "loss": 0.0839, + "step": 25732 + }, + { + "epoch": 11.96327289632729, + "grad_norm": 18.5585880279541, + "learning_rate": 9.974278997427901e-06, + "loss": 0.2338, + "step": 25733 + }, + { + "epoch": 11.96373779637378, + "grad_norm": 6.703153610229492, + "learning_rate": 9.9742779974278e-06, + "loss": 0.0558, + "step": 25734 + }, + { + "epoch": 11.96420269642027, + "grad_norm": 4.3503618240356445, + "learning_rate": 9.9742769974277e-06, + "loss": 0.0493, + "step": 25735 + }, + { + "epoch": 11.964667596466759, + "grad_norm": 1.5547312498092651, + "learning_rate": 9.9742759974276e-06, + "loss": 0.0126, + "step": 25736 + }, + { + "epoch": 11.965132496513249, + "grad_norm": 6.133702754974365, + "learning_rate": 9.9742749974275e-06, + "loss": 0.0812, + "step": 25737 + }, + { + "epoch": 11.96559739655974, + "grad_norm": 2.7165775299072266, + "learning_rate": 9.974273997427401e-06, + "loss": 0.0325, + "step": 25738 + }, + { + "epoch": 11.96606229660623, + "grad_norm": 9.627129554748535, + "learning_rate": 9.9742729974273e-06, + "loss": 0.1068, + "step": 25739 + }, + { + "epoch": 11.96652719665272, + "grad_norm": 3.6901936531066895, + "learning_rate": 9.9742719974272e-06, + "loss": 0.0694, + "step": 25740 + }, + { + "epoch": 11.96699209669921, + "grad_norm": 5.322312831878662, + "learning_rate": 9.9742709974271e-06, + "loss": 0.0569, + "step": 25741 + }, + { + "epoch": 11.9674569967457, + "grad_norm": 3.6099493503570557, + "learning_rate": 9.974269997427e-06, + "loss": 0.067, + "step": 25742 + }, + { + "epoch": 11.96792189679219, + "grad_norm": 1.8380335569381714, + "learning_rate": 9.9742689974269e-06, + "loss": 0.0138, + "step": 25743 + }, + { + "epoch": 11.96838679683868, + "grad_norm": 5.740697383880615, + "learning_rate": 9.974267997426801e-06, + "loss": 0.0832, + "step": 25744 + }, + { + "epoch": 11.968851696885169, + "grad_norm": 2.336998462677002, + "learning_rate": 9.9742669974267e-06, + "loss": 0.0164, + "step": 25745 + }, + { + "epoch": 11.969316596931659, + "grad_norm": 2.1204161643981934, + "learning_rate": 9.974265997426601e-06, + "loss": 0.0193, + "step": 25746 + }, + { + "epoch": 11.96978149697815, + "grad_norm": 5.995691776275635, + "learning_rate": 9.9742649974265e-06, + "loss": 0.114, + "step": 25747 + }, + { + "epoch": 11.97024639702464, + "grad_norm": 2.4748330116271973, + "learning_rate": 9.9742639974264e-06, + "loss": 0.0222, + "step": 25748 + }, + { + "epoch": 11.97071129707113, + "grad_norm": 3.1316609382629395, + "learning_rate": 9.974262997426301e-06, + "loss": 0.0273, + "step": 25749 + }, + { + "epoch": 11.97117619711762, + "grad_norm": 5.130016803741455, + "learning_rate": 9.9742619974262e-06, + "loss": 0.0775, + "step": 25750 + }, + { + "epoch": 11.97164109716411, + "grad_norm": 3.278088092803955, + "learning_rate": 9.974260997426101e-06, + "loss": 0.0211, + "step": 25751 + }, + { + "epoch": 11.9721059972106, + "grad_norm": 3.46280574798584, + "learning_rate": 9.974259997426e-06, + "loss": 0.0208, + "step": 25752 + }, + { + "epoch": 11.97257089725709, + "grad_norm": 37.126625061035156, + "learning_rate": 9.9742589974259e-06, + "loss": 0.0794, + "step": 25753 + }, + { + "epoch": 11.973035797303579, + "grad_norm": 3.9460906982421875, + "learning_rate": 9.9742579974258e-06, + "loss": 0.0247, + "step": 25754 + }, + { + "epoch": 11.973500697350069, + "grad_norm": 5.4069108963012695, + "learning_rate": 9.9742569974257e-06, + "loss": 0.0961, + "step": 25755 + }, + { + "epoch": 11.97396559739656, + "grad_norm": 3.7054636478424072, + "learning_rate": 9.974255997425601e-06, + "loss": 0.0304, + "step": 25756 + }, + { + "epoch": 11.97443049744305, + "grad_norm": 4.891651630401611, + "learning_rate": 9.974254997425502e-06, + "loss": 0.0585, + "step": 25757 + }, + { + "epoch": 11.97489539748954, + "grad_norm": 4.432921409606934, + "learning_rate": 9.9742539974254e-06, + "loss": 0.0519, + "step": 25758 + }, + { + "epoch": 11.97536029753603, + "grad_norm": 5.161764621734619, + "learning_rate": 9.974252997425301e-06, + "loss": 0.0634, + "step": 25759 + }, + { + "epoch": 11.97582519758252, + "grad_norm": 5.5467143058776855, + "learning_rate": 9.9742519974252e-06, + "loss": 0.0549, + "step": 25760 + }, + { + "epoch": 11.97629009762901, + "grad_norm": 5.508586883544922, + "learning_rate": 9.9742509974251e-06, + "loss": 0.0882, + "step": 25761 + }, + { + "epoch": 11.9767549976755, + "grad_norm": 7.64013671875, + "learning_rate": 9.974249997425001e-06, + "loss": 0.1299, + "step": 25762 + }, + { + "epoch": 11.977219897721989, + "grad_norm": 7.353931903839111, + "learning_rate": 9.9742489974249e-06, + "loss": 0.2114, + "step": 25763 + }, + { + "epoch": 11.97768479776848, + "grad_norm": 4.636723041534424, + "learning_rate": 9.974247997424801e-06, + "loss": 0.0572, + "step": 25764 + }, + { + "epoch": 11.97814969781497, + "grad_norm": 4.513576984405518, + "learning_rate": 9.9742469974247e-06, + "loss": 0.0471, + "step": 25765 + }, + { + "epoch": 11.97861459786146, + "grad_norm": 3.378002882003784, + "learning_rate": 9.9742459974246e-06, + "loss": 0.0258, + "step": 25766 + }, + { + "epoch": 11.97907949790795, + "grad_norm": 3.6130878925323486, + "learning_rate": 9.9742449974245e-06, + "loss": 0.0402, + "step": 25767 + }, + { + "epoch": 11.97954439795444, + "grad_norm": 2.0745608806610107, + "learning_rate": 9.9742439974244e-06, + "loss": 0.0146, + "step": 25768 + }, + { + "epoch": 11.98000929800093, + "grad_norm": 4.186197757720947, + "learning_rate": 9.974242997424301e-06, + "loss": 0.0541, + "step": 25769 + }, + { + "epoch": 11.98047419804742, + "grad_norm": 3.8609931468963623, + "learning_rate": 9.974241997424202e-06, + "loss": 0.0206, + "step": 25770 + }, + { + "epoch": 11.98093909809391, + "grad_norm": 4.806670188903809, + "learning_rate": 9.9742409974241e-06, + "loss": 0.0705, + "step": 25771 + }, + { + "epoch": 11.981403998140399, + "grad_norm": 6.186886310577393, + "learning_rate": 9.974239997424e-06, + "loss": 0.0497, + "step": 25772 + }, + { + "epoch": 11.98186889818689, + "grad_norm": 4.9579291343688965, + "learning_rate": 9.9742389974239e-06, + "loss": 0.056, + "step": 25773 + }, + { + "epoch": 11.98233379823338, + "grad_norm": 6.194747447967529, + "learning_rate": 9.974237997423801e-06, + "loss": 0.0733, + "step": 25774 + }, + { + "epoch": 11.98279869827987, + "grad_norm": 7.850194931030273, + "learning_rate": 9.974236997423702e-06, + "loss": 0.1692, + "step": 25775 + }, + { + "epoch": 11.98326359832636, + "grad_norm": 7.774453639984131, + "learning_rate": 9.9742359974236e-06, + "loss": 0.144, + "step": 25776 + }, + { + "epoch": 11.98372849837285, + "grad_norm": 0.7322095036506653, + "learning_rate": 9.974234997423501e-06, + "loss": 0.0051, + "step": 25777 + }, + { + "epoch": 11.98419339841934, + "grad_norm": 9.403812408447266, + "learning_rate": 9.9742339974234e-06, + "loss": 0.1397, + "step": 25778 + }, + { + "epoch": 11.98465829846583, + "grad_norm": 5.459172248840332, + "learning_rate": 9.9742329974233e-06, + "loss": 0.0908, + "step": 25779 + }, + { + "epoch": 11.98512319851232, + "grad_norm": 8.102396011352539, + "learning_rate": 9.9742319974232e-06, + "loss": 0.094, + "step": 25780 + }, + { + "epoch": 11.98558809855881, + "grad_norm": 2.927595853805542, + "learning_rate": 9.9742309974231e-06, + "loss": 0.0259, + "step": 25781 + }, + { + "epoch": 11.9860529986053, + "grad_norm": 2.3584847450256348, + "learning_rate": 9.974229997423001e-06, + "loss": 0.0296, + "step": 25782 + }, + { + "epoch": 11.98651789865179, + "grad_norm": 2.802917242050171, + "learning_rate": 9.9742289974229e-06, + "loss": 0.0264, + "step": 25783 + }, + { + "epoch": 11.98698279869828, + "grad_norm": 8.066703796386719, + "learning_rate": 9.9742279974228e-06, + "loss": 0.1215, + "step": 25784 + }, + { + "epoch": 11.98744769874477, + "grad_norm": 1.0275993347167969, + "learning_rate": 9.9742269974227e-06, + "loss": 0.0052, + "step": 25785 + }, + { + "epoch": 11.98791259879126, + "grad_norm": 3.530146360397339, + "learning_rate": 9.9742259974226e-06, + "loss": 0.0318, + "step": 25786 + }, + { + "epoch": 11.98837749883775, + "grad_norm": 5.244620323181152, + "learning_rate": 9.974224997422501e-06, + "loss": 0.0718, + "step": 25787 + }, + { + "epoch": 11.98884239888424, + "grad_norm": 4.829963684082031, + "learning_rate": 9.974223997422402e-06, + "loss": 0.0922, + "step": 25788 + }, + { + "epoch": 11.98930729893073, + "grad_norm": 4.738605976104736, + "learning_rate": 9.9742229974223e-06, + "loss": 0.0517, + "step": 25789 + }, + { + "epoch": 11.98977219897722, + "grad_norm": 3.4546585083007812, + "learning_rate": 9.974221997422201e-06, + "loss": 0.0296, + "step": 25790 + }, + { + "epoch": 11.99023709902371, + "grad_norm": 2.8389885425567627, + "learning_rate": 9.9742209974221e-06, + "loss": 0.0308, + "step": 25791 + }, + { + "epoch": 11.9907019990702, + "grad_norm": 5.3599700927734375, + "learning_rate": 9.974219997422001e-06, + "loss": 0.0873, + "step": 25792 + }, + { + "epoch": 11.99116689911669, + "grad_norm": 3.604306936264038, + "learning_rate": 9.9742189974219e-06, + "loss": 0.0384, + "step": 25793 + }, + { + "epoch": 11.99163179916318, + "grad_norm": 4.644660472869873, + "learning_rate": 9.9742179974218e-06, + "loss": 0.0567, + "step": 25794 + }, + { + "epoch": 11.99209669920967, + "grad_norm": 3.88372540473938, + "learning_rate": 9.974216997421701e-06, + "loss": 0.0424, + "step": 25795 + }, + { + "epoch": 11.99256159925616, + "grad_norm": 6.212472915649414, + "learning_rate": 9.9742159974216e-06, + "loss": 0.1733, + "step": 25796 + }, + { + "epoch": 11.99302649930265, + "grad_norm": 3.0141968727111816, + "learning_rate": 9.9742149974215e-06, + "loss": 0.0316, + "step": 25797 + }, + { + "epoch": 11.99349139934914, + "grad_norm": 2.685567855834961, + "learning_rate": 9.9742139974214e-06, + "loss": 0.0105, + "step": 25798 + }, + { + "epoch": 11.99395629939563, + "grad_norm": 5.954508304595947, + "learning_rate": 9.9742129974213e-06, + "loss": 0.0821, + "step": 25799 + }, + { + "epoch": 11.99442119944212, + "grad_norm": 2.39445161819458, + "learning_rate": 9.974211997421201e-06, + "loss": 0.019, + "step": 25800 + }, + { + "epoch": 11.99488609948861, + "grad_norm": 5.860004901885986, + "learning_rate": 9.974210997421102e-06, + "loss": 0.1014, + "step": 25801 + }, + { + "epoch": 11.9953509995351, + "grad_norm": 7.662127494812012, + "learning_rate": 9.974209997421e-06, + "loss": 0.0378, + "step": 25802 + }, + { + "epoch": 11.99581589958159, + "grad_norm": 5.406190872192383, + "learning_rate": 9.974208997420901e-06, + "loss": 0.0997, + "step": 25803 + }, + { + "epoch": 11.99628079962808, + "grad_norm": 2.7160098552703857, + "learning_rate": 9.9742079974208e-06, + "loss": 0.0335, + "step": 25804 + }, + { + "epoch": 11.99674569967457, + "grad_norm": 10.289796829223633, + "learning_rate": 9.9742069974207e-06, + "loss": 0.0479, + "step": 25805 + }, + { + "epoch": 11.99721059972106, + "grad_norm": 3.92704176902771, + "learning_rate": 9.9742059974206e-06, + "loss": 0.0289, + "step": 25806 + }, + { + "epoch": 11.99767549976755, + "grad_norm": 2.1003453731536865, + "learning_rate": 9.9742049974205e-06, + "loss": 0.0179, + "step": 25807 + }, + { + "epoch": 11.99814039981404, + "grad_norm": 4.22955846786499, + "learning_rate": 9.974203997420401e-06, + "loss": 0.037, + "step": 25808 + }, + { + "epoch": 11.99860529986053, + "grad_norm": 4.402738571166992, + "learning_rate": 9.9742029974203e-06, + "loss": 0.042, + "step": 25809 + }, + { + "epoch": 11.99907019990702, + "grad_norm": 2.4384233951568604, + "learning_rate": 9.974201997420201e-06, + "loss": 0.0212, + "step": 25810 + }, + { + "epoch": 11.99953509995351, + "grad_norm": 4.112040042877197, + "learning_rate": 9.9742009974201e-06, + "loss": 0.0622, + "step": 25811 + }, + { + "epoch": 12.0, + "grad_norm": 11.957462310791016, + "learning_rate": 9.97419999742e-06, + "loss": 0.1633, + "step": 25812 + }, + { + "epoch": 12.00046490004649, + "grad_norm": 8.391984939575195, + "learning_rate": 9.974198997419901e-06, + "loss": 0.1822, + "step": 25813 + }, + { + "epoch": 12.00092980009298, + "grad_norm": 3.182223320007324, + "learning_rate": 9.974197997419802e-06, + "loss": 0.0207, + "step": 25814 + }, + { + "epoch": 12.00139470013947, + "grad_norm": 5.784937858581543, + "learning_rate": 9.9741969974197e-06, + "loss": 0.0516, + "step": 25815 + }, + { + "epoch": 12.00185960018596, + "grad_norm": 1.7033131122589111, + "learning_rate": 9.9741959974196e-06, + "loss": 0.0144, + "step": 25816 + }, + { + "epoch": 12.00232450023245, + "grad_norm": 3.550417423248291, + "learning_rate": 9.9741949974195e-06, + "loss": 0.0401, + "step": 25817 + }, + { + "epoch": 12.00278940027894, + "grad_norm": 5.485034942626953, + "learning_rate": 9.974193997419401e-06, + "loss": 0.0964, + "step": 25818 + }, + { + "epoch": 12.00325430032543, + "grad_norm": 1.9215335845947266, + "learning_rate": 9.9741929974193e-06, + "loss": 0.0094, + "step": 25819 + }, + { + "epoch": 12.00371920037192, + "grad_norm": 0.5946618914604187, + "learning_rate": 9.9741919974192e-06, + "loss": 0.004, + "step": 25820 + }, + { + "epoch": 12.00418410041841, + "grad_norm": 3.364168167114258, + "learning_rate": 9.974190997419101e-06, + "loss": 0.0597, + "step": 25821 + }, + { + "epoch": 12.0046490004649, + "grad_norm": 4.140407562255859, + "learning_rate": 9.974189997419e-06, + "loss": 0.0354, + "step": 25822 + }, + { + "epoch": 12.00511390051139, + "grad_norm": 3.958726406097412, + "learning_rate": 9.974188997418901e-06, + "loss": 0.0515, + "step": 25823 + }, + { + "epoch": 12.00557880055788, + "grad_norm": 4.034339904785156, + "learning_rate": 9.9741879974188e-06, + "loss": 0.0184, + "step": 25824 + }, + { + "epoch": 12.00604370060437, + "grad_norm": 1.322789192199707, + "learning_rate": 9.9741869974187e-06, + "loss": 0.0099, + "step": 25825 + }, + { + "epoch": 12.00650860065086, + "grad_norm": 4.202818870544434, + "learning_rate": 9.974185997418601e-06, + "loss": 0.0446, + "step": 25826 + }, + { + "epoch": 12.00697350069735, + "grad_norm": 3.3898210525512695, + "learning_rate": 9.9741849974185e-06, + "loss": 0.0566, + "step": 25827 + }, + { + "epoch": 12.00743840074384, + "grad_norm": 4.523471355438232, + "learning_rate": 9.9741839974184e-06, + "loss": 0.043, + "step": 25828 + }, + { + "epoch": 12.00790330079033, + "grad_norm": 6.535004615783691, + "learning_rate": 9.9741829974183e-06, + "loss": 0.0523, + "step": 25829 + }, + { + "epoch": 12.00836820083682, + "grad_norm": 5.02192497253418, + "learning_rate": 9.9741819974182e-06, + "loss": 0.0484, + "step": 25830 + }, + { + "epoch": 12.00883310088331, + "grad_norm": 2.9834978580474854, + "learning_rate": 9.974180997418101e-06, + "loss": 0.0219, + "step": 25831 + }, + { + "epoch": 12.0092980009298, + "grad_norm": 2.0212454795837402, + "learning_rate": 9.974179997418e-06, + "loss": 0.0191, + "step": 25832 + }, + { + "epoch": 12.00976290097629, + "grad_norm": 1.868187427520752, + "learning_rate": 9.9741789974179e-06, + "loss": 0.0105, + "step": 25833 + }, + { + "epoch": 12.010227801022781, + "grad_norm": 3.1524977684020996, + "learning_rate": 9.974177997417801e-06, + "loss": 0.0253, + "step": 25834 + }, + { + "epoch": 12.01069270106927, + "grad_norm": 5.656409740447998, + "learning_rate": 9.9741769974177e-06, + "loss": 0.0705, + "step": 25835 + }, + { + "epoch": 12.01115760111576, + "grad_norm": 5.456111907958984, + "learning_rate": 9.9741759974176e-06, + "loss": 0.0674, + "step": 25836 + }, + { + "epoch": 12.01162250116225, + "grad_norm": 4.131933212280273, + "learning_rate": 9.9741749974175e-06, + "loss": 0.1109, + "step": 25837 + }, + { + "epoch": 12.01208740120874, + "grad_norm": 2.474480628967285, + "learning_rate": 9.9741739974174e-06, + "loss": 0.0192, + "step": 25838 + }, + { + "epoch": 12.01255230125523, + "grad_norm": 4.675340175628662, + "learning_rate": 9.974172997417301e-06, + "loss": 0.0667, + "step": 25839 + }, + { + "epoch": 12.01301720130172, + "grad_norm": 3.0393872261047363, + "learning_rate": 9.9741719974172e-06, + "loss": 0.0218, + "step": 25840 + }, + { + "epoch": 12.01348210134821, + "grad_norm": 3.5040853023529053, + "learning_rate": 9.974170997417101e-06, + "loss": 0.0383, + "step": 25841 + }, + { + "epoch": 12.0139470013947, + "grad_norm": 5.198654651641846, + "learning_rate": 9.974169997417e-06, + "loss": 0.0478, + "step": 25842 + }, + { + "epoch": 12.014411901441191, + "grad_norm": 2.209479570388794, + "learning_rate": 9.9741689974169e-06, + "loss": 0.0224, + "step": 25843 + }, + { + "epoch": 12.01487680148768, + "grad_norm": 2.163343906402588, + "learning_rate": 9.974167997416801e-06, + "loss": 0.0189, + "step": 25844 + }, + { + "epoch": 12.01534170153417, + "grad_norm": 2.8071253299713135, + "learning_rate": 9.974166997416702e-06, + "loss": 0.0156, + "step": 25845 + }, + { + "epoch": 12.01580660158066, + "grad_norm": 9.430892944335938, + "learning_rate": 9.9741659974166e-06, + "loss": 0.1624, + "step": 25846 + }, + { + "epoch": 12.01627150162715, + "grad_norm": 8.512602806091309, + "learning_rate": 9.9741649974165e-06, + "loss": 0.062, + "step": 25847 + }, + { + "epoch": 12.01673640167364, + "grad_norm": 3.8796892166137695, + "learning_rate": 9.9741639974164e-06, + "loss": 0.0394, + "step": 25848 + }, + { + "epoch": 12.01720130172013, + "grad_norm": 5.030316352844238, + "learning_rate": 9.9741629974163e-06, + "loss": 0.0392, + "step": 25849 + }, + { + "epoch": 12.01766620176662, + "grad_norm": 3.5276362895965576, + "learning_rate": 9.9741619974162e-06, + "loss": 0.0312, + "step": 25850 + }, + { + "epoch": 12.01813110181311, + "grad_norm": 1.1385648250579834, + "learning_rate": 9.9741609974161e-06, + "loss": 0.0081, + "step": 25851 + }, + { + "epoch": 12.018596001859601, + "grad_norm": 2.2261388301849365, + "learning_rate": 9.974159997416001e-06, + "loss": 0.0181, + "step": 25852 + }, + { + "epoch": 12.01906090190609, + "grad_norm": 5.233630657196045, + "learning_rate": 9.9741589974159e-06, + "loss": 0.0477, + "step": 25853 + }, + { + "epoch": 12.01952580195258, + "grad_norm": 3.65187406539917, + "learning_rate": 9.974157997415801e-06, + "loss": 0.026, + "step": 25854 + }, + { + "epoch": 12.01999070199907, + "grad_norm": 1.0808579921722412, + "learning_rate": 9.9741569974157e-06, + "loss": 0.0082, + "step": 25855 + }, + { + "epoch": 12.02045560204556, + "grad_norm": 2.7163634300231934, + "learning_rate": 9.9741559974156e-06, + "loss": 0.0252, + "step": 25856 + }, + { + "epoch": 12.02092050209205, + "grad_norm": 6.787975311279297, + "learning_rate": 9.974154997415501e-06, + "loss": 0.1052, + "step": 25857 + }, + { + "epoch": 12.02138540213854, + "grad_norm": 2.0119211673736572, + "learning_rate": 9.9741539974154e-06, + "loss": 0.0087, + "step": 25858 + }, + { + "epoch": 12.02185030218503, + "grad_norm": 1.962134599685669, + "learning_rate": 9.974152997415301e-06, + "loss": 0.0162, + "step": 25859 + }, + { + "epoch": 12.02231520223152, + "grad_norm": 1.5940818786621094, + "learning_rate": 9.9741519974152e-06, + "loss": 0.0084, + "step": 25860 + }, + { + "epoch": 12.022780102278011, + "grad_norm": 5.321466445922852, + "learning_rate": 9.9741509974151e-06, + "loss": 0.0829, + "step": 25861 + }, + { + "epoch": 12.0232450023245, + "grad_norm": 3.7280609607696533, + "learning_rate": 9.974149997415e-06, + "loss": 0.0493, + "step": 25862 + }, + { + "epoch": 12.02370990237099, + "grad_norm": 3.0948359966278076, + "learning_rate": 9.9741489974149e-06, + "loss": 0.0168, + "step": 25863 + }, + { + "epoch": 12.02417480241748, + "grad_norm": 2.0735960006713867, + "learning_rate": 9.9741479974148e-06, + "loss": 0.0165, + "step": 25864 + }, + { + "epoch": 12.02463970246397, + "grad_norm": 2.2775120735168457, + "learning_rate": 9.974146997414701e-06, + "loss": 0.0195, + "step": 25865 + }, + { + "epoch": 12.02510460251046, + "grad_norm": 2.090914726257324, + "learning_rate": 9.9741459974146e-06, + "loss": 0.0144, + "step": 25866 + }, + { + "epoch": 12.02556950255695, + "grad_norm": 3.8368642330169678, + "learning_rate": 9.974144997414501e-06, + "loss": 0.0423, + "step": 25867 + }, + { + "epoch": 12.02603440260344, + "grad_norm": 11.492815971374512, + "learning_rate": 9.9741439974144e-06, + "loss": 0.0675, + "step": 25868 + }, + { + "epoch": 12.02649930264993, + "grad_norm": 9.219521522521973, + "learning_rate": 9.9741429974143e-06, + "loss": 0.1753, + "step": 25869 + }, + { + "epoch": 12.026964202696421, + "grad_norm": 6.042776107788086, + "learning_rate": 9.974141997414201e-06, + "loss": 0.0884, + "step": 25870 + }, + { + "epoch": 12.02742910274291, + "grad_norm": 4.403399467468262, + "learning_rate": 9.9741409974141e-06, + "loss": 0.0504, + "step": 25871 + }, + { + "epoch": 12.0278940027894, + "grad_norm": 4.074985980987549, + "learning_rate": 9.974139997414001e-06, + "loss": 0.0696, + "step": 25872 + }, + { + "epoch": 12.02835890283589, + "grad_norm": 1.2138147354125977, + "learning_rate": 9.9741389974139e-06, + "loss": 0.0155, + "step": 25873 + }, + { + "epoch": 12.02882380288238, + "grad_norm": 3.8672842979431152, + "learning_rate": 9.9741379974138e-06, + "loss": 0.0408, + "step": 25874 + }, + { + "epoch": 12.02928870292887, + "grad_norm": 3.154878616333008, + "learning_rate": 9.9741369974137e-06, + "loss": 0.0453, + "step": 25875 + }, + { + "epoch": 12.02975360297536, + "grad_norm": 2.925893783569336, + "learning_rate": 9.9741359974136e-06, + "loss": 0.0304, + "step": 25876 + }, + { + "epoch": 12.03021850302185, + "grad_norm": 2.055968761444092, + "learning_rate": 9.974134997413501e-06, + "loss": 0.0153, + "step": 25877 + }, + { + "epoch": 12.030683403068341, + "grad_norm": 9.18277645111084, + "learning_rate": 9.974133997413402e-06, + "loss": 0.1887, + "step": 25878 + }, + { + "epoch": 12.031148303114831, + "grad_norm": 4.4675612449646, + "learning_rate": 9.9741329974133e-06, + "loss": 0.0263, + "step": 25879 + }, + { + "epoch": 12.03161320316132, + "grad_norm": 7.615654945373535, + "learning_rate": 9.9741319974132e-06, + "loss": 0.0802, + "step": 25880 + }, + { + "epoch": 12.03207810320781, + "grad_norm": 3.385267734527588, + "learning_rate": 9.9741309974131e-06, + "loss": 0.0268, + "step": 25881 + }, + { + "epoch": 12.0325430032543, + "grad_norm": 1.6847038269042969, + "learning_rate": 9.974129997413e-06, + "loss": 0.0137, + "step": 25882 + }, + { + "epoch": 12.03300790330079, + "grad_norm": 4.135281085968018, + "learning_rate": 9.974128997412901e-06, + "loss": 0.0563, + "step": 25883 + }, + { + "epoch": 12.03347280334728, + "grad_norm": 6.108997821807861, + "learning_rate": 9.9741279974128e-06, + "loss": 0.0692, + "step": 25884 + }, + { + "epoch": 12.03393770339377, + "grad_norm": 1.961689829826355, + "learning_rate": 9.974126997412701e-06, + "loss": 0.0167, + "step": 25885 + }, + { + "epoch": 12.03440260344026, + "grad_norm": 7.335472106933594, + "learning_rate": 9.9741259974126e-06, + "loss": 0.0513, + "step": 25886 + }, + { + "epoch": 12.034867503486751, + "grad_norm": 3.7885069847106934, + "learning_rate": 9.9741249974125e-06, + "loss": 0.0281, + "step": 25887 + }, + { + "epoch": 12.035332403533241, + "grad_norm": 8.651548385620117, + "learning_rate": 9.9741239974124e-06, + "loss": 0.0883, + "step": 25888 + }, + { + "epoch": 12.03579730357973, + "grad_norm": 1.4207050800323486, + "learning_rate": 9.9741229974123e-06, + "loss": 0.0121, + "step": 25889 + }, + { + "epoch": 12.03626220362622, + "grad_norm": 4.877945423126221, + "learning_rate": 9.974121997412201e-06, + "loss": 0.0676, + "step": 25890 + }, + { + "epoch": 12.03672710367271, + "grad_norm": 6.183601379394531, + "learning_rate": 9.9741209974121e-06, + "loss": 0.0803, + "step": 25891 + }, + { + "epoch": 12.0371920037192, + "grad_norm": 6.437891006469727, + "learning_rate": 9.974119997412e-06, + "loss": 0.0595, + "step": 25892 + }, + { + "epoch": 12.03765690376569, + "grad_norm": 2.5207111835479736, + "learning_rate": 9.9741189974119e-06, + "loss": 0.0154, + "step": 25893 + }, + { + "epoch": 12.03812180381218, + "grad_norm": 4.926115989685059, + "learning_rate": 9.9741179974118e-06, + "loss": 0.0968, + "step": 25894 + }, + { + "epoch": 12.03858670385867, + "grad_norm": 5.083611488342285, + "learning_rate": 9.974116997411701e-06, + "loss": 0.0973, + "step": 25895 + }, + { + "epoch": 12.039051603905161, + "grad_norm": 2.668548822402954, + "learning_rate": 9.974115997411602e-06, + "loss": 0.0122, + "step": 25896 + }, + { + "epoch": 12.039516503951651, + "grad_norm": 3.530442953109741, + "learning_rate": 9.9741149974115e-06, + "loss": 0.0485, + "step": 25897 + }, + { + "epoch": 12.03998140399814, + "grad_norm": 4.049719333648682, + "learning_rate": 9.974113997411401e-06, + "loss": 0.061, + "step": 25898 + }, + { + "epoch": 12.04044630404463, + "grad_norm": 5.164884567260742, + "learning_rate": 9.9741129974113e-06, + "loss": 0.0615, + "step": 25899 + }, + { + "epoch": 12.04091120409112, + "grad_norm": 6.190024375915527, + "learning_rate": 9.9741119974112e-06, + "loss": 0.0587, + "step": 25900 + }, + { + "epoch": 12.04137610413761, + "grad_norm": 3.435425281524658, + "learning_rate": 9.9741109974111e-06, + "loss": 0.0473, + "step": 25901 + }, + { + "epoch": 12.0418410041841, + "grad_norm": 2.8575406074523926, + "learning_rate": 9.974109997411e-06, + "loss": 0.0254, + "step": 25902 + }, + { + "epoch": 12.04230590423059, + "grad_norm": 2.996857166290283, + "learning_rate": 9.974108997410901e-06, + "loss": 0.0318, + "step": 25903 + }, + { + "epoch": 12.04277080427708, + "grad_norm": 1.9498496055603027, + "learning_rate": 9.9741079974108e-06, + "loss": 0.0169, + "step": 25904 + }, + { + "epoch": 12.043235704323571, + "grad_norm": 3.052424669265747, + "learning_rate": 9.9741069974107e-06, + "loss": 0.0228, + "step": 25905 + }, + { + "epoch": 12.043700604370061, + "grad_norm": 14.881278991699219, + "learning_rate": 9.9741059974106e-06, + "loss": 0.1371, + "step": 25906 + }, + { + "epoch": 12.04416550441655, + "grad_norm": 7.244996547698975, + "learning_rate": 9.9741049974105e-06, + "loss": 0.088, + "step": 25907 + }, + { + "epoch": 12.04463040446304, + "grad_norm": 5.333559513092041, + "learning_rate": 9.974103997410401e-06, + "loss": 0.0387, + "step": 25908 + }, + { + "epoch": 12.04509530450953, + "grad_norm": 3.024920701980591, + "learning_rate": 9.974102997410302e-06, + "loss": 0.0321, + "step": 25909 + }, + { + "epoch": 12.04556020455602, + "grad_norm": 8.155486106872559, + "learning_rate": 9.9741019974102e-06, + "loss": 0.0694, + "step": 25910 + }, + { + "epoch": 12.04602510460251, + "grad_norm": 6.357056617736816, + "learning_rate": 9.974100997410101e-06, + "loss": 0.0257, + "step": 25911 + }, + { + "epoch": 12.046490004649, + "grad_norm": 2.7295010089874268, + "learning_rate": 9.97409999741e-06, + "loss": 0.022, + "step": 25912 + }, + { + "epoch": 12.046954904695491, + "grad_norm": 5.8679962158203125, + "learning_rate": 9.9740989974099e-06, + "loss": 0.0528, + "step": 25913 + }, + { + "epoch": 12.047419804741981, + "grad_norm": 6.944729328155518, + "learning_rate": 9.974097997409801e-06, + "loss": 0.1005, + "step": 25914 + }, + { + "epoch": 12.047884704788471, + "grad_norm": 6.339533805847168, + "learning_rate": 9.9740969974097e-06, + "loss": 0.0721, + "step": 25915 + }, + { + "epoch": 12.04834960483496, + "grad_norm": 2.401493787765503, + "learning_rate": 9.974095997409601e-06, + "loss": 0.0175, + "step": 25916 + }, + { + "epoch": 12.04881450488145, + "grad_norm": 4.327508926391602, + "learning_rate": 9.9740949974095e-06, + "loss": 0.0474, + "step": 25917 + }, + { + "epoch": 12.04927940492794, + "grad_norm": 1.339160442352295, + "learning_rate": 9.9740939974094e-06, + "loss": 0.0088, + "step": 25918 + }, + { + "epoch": 12.04974430497443, + "grad_norm": 5.406423091888428, + "learning_rate": 9.9740929974093e-06, + "loss": 0.0478, + "step": 25919 + }, + { + "epoch": 12.05020920502092, + "grad_norm": 8.182120323181152, + "learning_rate": 9.9740919974092e-06, + "loss": 0.0775, + "step": 25920 + }, + { + "epoch": 12.05067410506741, + "grad_norm": 8.5064115524292, + "learning_rate": 9.974090997409101e-06, + "loss": 0.1506, + "step": 25921 + }, + { + "epoch": 12.051139005113901, + "grad_norm": 2.133854627609253, + "learning_rate": 9.974089997409002e-06, + "loss": 0.018, + "step": 25922 + }, + { + "epoch": 12.051603905160391, + "grad_norm": 6.136945724487305, + "learning_rate": 9.9740889974089e-06, + "loss": 0.1113, + "step": 25923 + }, + { + "epoch": 12.052068805206881, + "grad_norm": 7.520925521850586, + "learning_rate": 9.9740879974088e-06, + "loss": 0.1657, + "step": 25924 + }, + { + "epoch": 12.05253370525337, + "grad_norm": 3.954803228378296, + "learning_rate": 9.9740869974087e-06, + "loss": 0.0462, + "step": 25925 + }, + { + "epoch": 12.05299860529986, + "grad_norm": 1.1373388767242432, + "learning_rate": 9.974085997408601e-06, + "loss": 0.0089, + "step": 25926 + }, + { + "epoch": 12.05346350534635, + "grad_norm": 1.963047742843628, + "learning_rate": 9.974084997408502e-06, + "loss": 0.0177, + "step": 25927 + }, + { + "epoch": 12.05392840539284, + "grad_norm": 3.4715864658355713, + "learning_rate": 9.9740839974084e-06, + "loss": 0.0179, + "step": 25928 + }, + { + "epoch": 12.05439330543933, + "grad_norm": 2.038778781890869, + "learning_rate": 9.974082997408301e-06, + "loss": 0.0105, + "step": 25929 + }, + { + "epoch": 12.05485820548582, + "grad_norm": 3.234936237335205, + "learning_rate": 9.9740819974082e-06, + "loss": 0.0261, + "step": 25930 + }, + { + "epoch": 12.055323105532311, + "grad_norm": 2.6703457832336426, + "learning_rate": 9.9740809974081e-06, + "loss": 0.0515, + "step": 25931 + }, + { + "epoch": 12.055788005578801, + "grad_norm": 7.198862075805664, + "learning_rate": 9.974079997408e-06, + "loss": 0.0991, + "step": 25932 + }, + { + "epoch": 12.056252905625291, + "grad_norm": 3.897106647491455, + "learning_rate": 9.9740789974079e-06, + "loss": 0.0565, + "step": 25933 + }, + { + "epoch": 12.05671780567178, + "grad_norm": 4.049031734466553, + "learning_rate": 9.974077997407801e-06, + "loss": 0.0209, + "step": 25934 + }, + { + "epoch": 12.05718270571827, + "grad_norm": 0.9485253691673279, + "learning_rate": 9.9740769974077e-06, + "loss": 0.0082, + "step": 25935 + }, + { + "epoch": 12.05764760576476, + "grad_norm": 5.363549709320068, + "learning_rate": 9.9740759974076e-06, + "loss": 0.0805, + "step": 25936 + }, + { + "epoch": 12.05811250581125, + "grad_norm": 3.442352294921875, + "learning_rate": 9.9740749974075e-06, + "loss": 0.0487, + "step": 25937 + }, + { + "epoch": 12.05857740585774, + "grad_norm": 9.15495777130127, + "learning_rate": 9.9740739974074e-06, + "loss": 0.1458, + "step": 25938 + }, + { + "epoch": 12.05904230590423, + "grad_norm": 1.3766711950302124, + "learning_rate": 9.974072997407301e-06, + "loss": 0.0085, + "step": 25939 + }, + { + "epoch": 12.059507205950721, + "grad_norm": 3.964329481124878, + "learning_rate": 9.974071997407202e-06, + "loss": 0.0322, + "step": 25940 + }, + { + "epoch": 12.059972105997211, + "grad_norm": 1.5619292259216309, + "learning_rate": 9.9740709974071e-06, + "loss": 0.0137, + "step": 25941 + }, + { + "epoch": 12.060437006043701, + "grad_norm": 1.9505155086517334, + "learning_rate": 9.974069997407001e-06, + "loss": 0.0073, + "step": 25942 + }, + { + "epoch": 12.06090190609019, + "grad_norm": 1.8867018222808838, + "learning_rate": 9.9740689974069e-06, + "loss": 0.0115, + "step": 25943 + }, + { + "epoch": 12.06136680613668, + "grad_norm": 6.306935787200928, + "learning_rate": 9.974067997406801e-06, + "loss": 0.1535, + "step": 25944 + }, + { + "epoch": 12.06183170618317, + "grad_norm": 6.634018898010254, + "learning_rate": 9.9740669974067e-06, + "loss": 0.1171, + "step": 25945 + }, + { + "epoch": 12.06229660622966, + "grad_norm": 4.448983192443848, + "learning_rate": 9.9740659974066e-06, + "loss": 0.0603, + "step": 25946 + }, + { + "epoch": 12.06276150627615, + "grad_norm": 3.1954379081726074, + "learning_rate": 9.974064997406501e-06, + "loss": 0.0429, + "step": 25947 + }, + { + "epoch": 12.06322640632264, + "grad_norm": 1.4437768459320068, + "learning_rate": 9.9740639974064e-06, + "loss": 0.0089, + "step": 25948 + }, + { + "epoch": 12.063691306369131, + "grad_norm": 2.102411985397339, + "learning_rate": 9.9740629974063e-06, + "loss": 0.0294, + "step": 25949 + }, + { + "epoch": 12.064156206415621, + "grad_norm": 7.713778018951416, + "learning_rate": 9.9740619974062e-06, + "loss": 0.0951, + "step": 25950 + }, + { + "epoch": 12.064621106462111, + "grad_norm": 4.9362874031066895, + "learning_rate": 9.9740609974061e-06, + "loss": 0.0399, + "step": 25951 + }, + { + "epoch": 12.0650860065086, + "grad_norm": 7.469210147857666, + "learning_rate": 9.974059997406001e-06, + "loss": 0.1039, + "step": 25952 + }, + { + "epoch": 12.06555090655509, + "grad_norm": 4.469797134399414, + "learning_rate": 9.974058997405902e-06, + "loss": 0.0479, + "step": 25953 + }, + { + "epoch": 12.06601580660158, + "grad_norm": 4.874453544616699, + "learning_rate": 9.9740579974058e-06, + "loss": 0.0745, + "step": 25954 + }, + { + "epoch": 12.06648070664807, + "grad_norm": 5.5504536628723145, + "learning_rate": 9.974056997405701e-06, + "loss": 0.0659, + "step": 25955 + }, + { + "epoch": 12.06694560669456, + "grad_norm": 6.166810035705566, + "learning_rate": 9.9740559974056e-06, + "loss": 0.1801, + "step": 25956 + }, + { + "epoch": 12.067410506741052, + "grad_norm": 2.2289061546325684, + "learning_rate": 9.9740549974055e-06, + "loss": 0.0141, + "step": 25957 + }, + { + "epoch": 12.067875406787541, + "grad_norm": 1.4799470901489258, + "learning_rate": 9.9740539974054e-06, + "loss": 0.0128, + "step": 25958 + }, + { + "epoch": 12.068340306834031, + "grad_norm": 3.467756748199463, + "learning_rate": 9.9740529974053e-06, + "loss": 0.0202, + "step": 25959 + }, + { + "epoch": 12.068805206880521, + "grad_norm": 3.478628635406494, + "learning_rate": 9.974051997405201e-06, + "loss": 0.0445, + "step": 25960 + }, + { + "epoch": 12.06927010692701, + "grad_norm": 3.687000036239624, + "learning_rate": 9.9740509974051e-06, + "loss": 0.0189, + "step": 25961 + }, + { + "epoch": 12.0697350069735, + "grad_norm": 3.5217268466949463, + "learning_rate": 9.974049997405001e-06, + "loss": 0.0329, + "step": 25962 + }, + { + "epoch": 12.07019990701999, + "grad_norm": 1.3763344287872314, + "learning_rate": 9.9740489974049e-06, + "loss": 0.013, + "step": 25963 + }, + { + "epoch": 12.07066480706648, + "grad_norm": 3.254986524581909, + "learning_rate": 9.9740479974048e-06, + "loss": 0.0337, + "step": 25964 + }, + { + "epoch": 12.07112970711297, + "grad_norm": 6.846736431121826, + "learning_rate": 9.974046997404701e-06, + "loss": 0.0847, + "step": 25965 + }, + { + "epoch": 12.071594607159462, + "grad_norm": 1.6930731534957886, + "learning_rate": 9.974045997404602e-06, + "loss": 0.011, + "step": 25966 + }, + { + "epoch": 12.072059507205951, + "grad_norm": 3.8441414833068848, + "learning_rate": 9.9740449974045e-06, + "loss": 0.0641, + "step": 25967 + }, + { + "epoch": 12.072524407252441, + "grad_norm": 3.8309202194213867, + "learning_rate": 9.9740439974044e-06, + "loss": 0.0331, + "step": 25968 + }, + { + "epoch": 12.072989307298931, + "grad_norm": 7.290744304656982, + "learning_rate": 9.9740429974043e-06, + "loss": 0.1183, + "step": 25969 + }, + { + "epoch": 12.07345420734542, + "grad_norm": 6.385804176330566, + "learning_rate": 9.9740419974042e-06, + "loss": 0.1176, + "step": 25970 + }, + { + "epoch": 12.07391910739191, + "grad_norm": 2.5035128593444824, + "learning_rate": 9.9740409974041e-06, + "loss": 0.009, + "step": 25971 + }, + { + "epoch": 12.0743840074384, + "grad_norm": 1.655387043952942, + "learning_rate": 9.974039997404e-06, + "loss": 0.0126, + "step": 25972 + }, + { + "epoch": 12.07484890748489, + "grad_norm": 1.1729378700256348, + "learning_rate": 9.974038997403901e-06, + "loss": 0.0108, + "step": 25973 + }, + { + "epoch": 12.07531380753138, + "grad_norm": 1.8944793939590454, + "learning_rate": 9.9740379974038e-06, + "loss": 0.0159, + "step": 25974 + }, + { + "epoch": 12.075778707577872, + "grad_norm": 1.3138190507888794, + "learning_rate": 9.974036997403701e-06, + "loss": 0.0098, + "step": 25975 + }, + { + "epoch": 12.076243607624361, + "grad_norm": 3.6691722869873047, + "learning_rate": 9.9740359974036e-06, + "loss": 0.0526, + "step": 25976 + }, + { + "epoch": 12.076708507670851, + "grad_norm": 3.4780941009521484, + "learning_rate": 9.9740349974035e-06, + "loss": 0.0309, + "step": 25977 + }, + { + "epoch": 12.077173407717341, + "grad_norm": 4.0670013427734375, + "learning_rate": 9.974033997403401e-06, + "loss": 0.0981, + "step": 25978 + }, + { + "epoch": 12.07763830776383, + "grad_norm": 0.6804754734039307, + "learning_rate": 9.9740329974033e-06, + "loss": 0.0042, + "step": 25979 + }, + { + "epoch": 12.07810320781032, + "grad_norm": 2.558264970779419, + "learning_rate": 9.974031997403201e-06, + "loss": 0.0166, + "step": 25980 + }, + { + "epoch": 12.07856810785681, + "grad_norm": 4.968965530395508, + "learning_rate": 9.9740309974031e-06, + "loss": 0.0303, + "step": 25981 + }, + { + "epoch": 12.0790330079033, + "grad_norm": 3.8178439140319824, + "learning_rate": 9.974029997403e-06, + "loss": 0.0491, + "step": 25982 + }, + { + "epoch": 12.07949790794979, + "grad_norm": 3.865666389465332, + "learning_rate": 9.974028997402901e-06, + "loss": 0.0603, + "step": 25983 + }, + { + "epoch": 12.079962807996282, + "grad_norm": 3.2285056114196777, + "learning_rate": 9.9740279974028e-06, + "loss": 0.0275, + "step": 25984 + }, + { + "epoch": 12.080427708042771, + "grad_norm": 2.0737571716308594, + "learning_rate": 9.9740269974027e-06, + "loss": 0.0158, + "step": 25985 + }, + { + "epoch": 12.080892608089261, + "grad_norm": 5.077083587646484, + "learning_rate": 9.974025997402601e-06, + "loss": 0.0497, + "step": 25986 + }, + { + "epoch": 12.081357508135751, + "grad_norm": 6.471678256988525, + "learning_rate": 9.9740249974025e-06, + "loss": 0.0815, + "step": 25987 + }, + { + "epoch": 12.08182240818224, + "grad_norm": 9.742145538330078, + "learning_rate": 9.9740239974024e-06, + "loss": 0.1366, + "step": 25988 + }, + { + "epoch": 12.08228730822873, + "grad_norm": 2.5552570819854736, + "learning_rate": 9.9740229974023e-06, + "loss": 0.0283, + "step": 25989 + }, + { + "epoch": 12.08275220827522, + "grad_norm": 5.72283935546875, + "learning_rate": 9.9740219974022e-06, + "loss": 0.0946, + "step": 25990 + }, + { + "epoch": 12.08321710832171, + "grad_norm": 2.4799063205718994, + "learning_rate": 9.974020997402101e-06, + "loss": 0.0255, + "step": 25991 + }, + { + "epoch": 12.0836820083682, + "grad_norm": 4.846177577972412, + "learning_rate": 9.974019997402e-06, + "loss": 0.0588, + "step": 25992 + }, + { + "epoch": 12.084146908414692, + "grad_norm": 3.7874932289123535, + "learning_rate": 9.974018997401901e-06, + "loss": 0.0445, + "step": 25993 + }, + { + "epoch": 12.084611808461181, + "grad_norm": 2.5949747562408447, + "learning_rate": 9.9740179974018e-06, + "loss": 0.0212, + "step": 25994 + }, + { + "epoch": 12.085076708507671, + "grad_norm": 9.096817016601562, + "learning_rate": 9.9740169974017e-06, + "loss": 0.1358, + "step": 25995 + }, + { + "epoch": 12.085541608554161, + "grad_norm": 2.92968487739563, + "learning_rate": 9.974015997401601e-06, + "loss": 0.0296, + "step": 25996 + }, + { + "epoch": 12.08600650860065, + "grad_norm": 6.82484769821167, + "learning_rate": 9.9740149974015e-06, + "loss": 0.0652, + "step": 25997 + }, + { + "epoch": 12.08647140864714, + "grad_norm": 4.597789764404297, + "learning_rate": 9.974013997401401e-06, + "loss": 0.02, + "step": 25998 + }, + { + "epoch": 12.08693630869363, + "grad_norm": 2.0782470703125, + "learning_rate": 9.9740129974013e-06, + "loss": 0.0174, + "step": 25999 + }, + { + "epoch": 12.08740120874012, + "grad_norm": 3.624840021133423, + "learning_rate": 9.9740119974012e-06, + "loss": 0.051, + "step": 26000 + }, + { + "epoch": 12.087866108786612, + "grad_norm": 1.2109743356704712, + "learning_rate": 9.9740109974011e-06, + "loss": 0.0052, + "step": 26001 + }, + { + "epoch": 12.088331008833102, + "grad_norm": 1.545770287513733, + "learning_rate": 9.974009997401e-06, + "loss": 0.0091, + "step": 26002 + }, + { + "epoch": 12.088795908879591, + "grad_norm": 0.08294123411178589, + "learning_rate": 9.9740089974009e-06, + "loss": 0.001, + "step": 26003 + }, + { + "epoch": 12.089260808926081, + "grad_norm": 3.885781764984131, + "learning_rate": 9.974007997400801e-06, + "loss": 0.0477, + "step": 26004 + }, + { + "epoch": 12.089725708972571, + "grad_norm": 4.237422943115234, + "learning_rate": 9.9740069974007e-06, + "loss": 0.0871, + "step": 26005 + }, + { + "epoch": 12.09019060901906, + "grad_norm": 5.476109027862549, + "learning_rate": 9.974005997400601e-06, + "loss": 0.0564, + "step": 26006 + }, + { + "epoch": 12.09065550906555, + "grad_norm": 6.289051055908203, + "learning_rate": 9.9740049974005e-06, + "loss": 0.0645, + "step": 26007 + }, + { + "epoch": 12.09112040911204, + "grad_norm": 3.661281108856201, + "learning_rate": 9.9740039974004e-06, + "loss": 0.0289, + "step": 26008 + }, + { + "epoch": 12.09158530915853, + "grad_norm": 4.265267848968506, + "learning_rate": 9.974002997400301e-06, + "loss": 0.0368, + "step": 26009 + }, + { + "epoch": 12.092050209205022, + "grad_norm": 6.134062767028809, + "learning_rate": 9.9740019974002e-06, + "loss": 0.0767, + "step": 26010 + }, + { + "epoch": 12.092515109251512, + "grad_norm": 3.5949463844299316, + "learning_rate": 9.974000997400101e-06, + "loss": 0.023, + "step": 26011 + }, + { + "epoch": 12.092980009298001, + "grad_norm": 7.858131408691406, + "learning_rate": 9.9739999974e-06, + "loss": 0.062, + "step": 26012 + }, + { + "epoch": 12.093444909344491, + "grad_norm": 1.9761667251586914, + "learning_rate": 9.9739989973999e-06, + "loss": 0.0135, + "step": 26013 + }, + { + "epoch": 12.093909809390981, + "grad_norm": 5.1892805099487305, + "learning_rate": 9.9739979973998e-06, + "loss": 0.1121, + "step": 26014 + }, + { + "epoch": 12.09437470943747, + "grad_norm": 2.7322356700897217, + "learning_rate": 9.9739969973997e-06, + "loss": 0.0259, + "step": 26015 + }, + { + "epoch": 12.09483960948396, + "grad_norm": 5.504369258880615, + "learning_rate": 9.9739959973996e-06, + "loss": 0.0535, + "step": 26016 + }, + { + "epoch": 12.09530450953045, + "grad_norm": 2.695408582687378, + "learning_rate": 9.973994997399501e-06, + "loss": 0.0262, + "step": 26017 + }, + { + "epoch": 12.09576940957694, + "grad_norm": 2.5623738765716553, + "learning_rate": 9.9739939973994e-06, + "loss": 0.0325, + "step": 26018 + }, + { + "epoch": 12.096234309623432, + "grad_norm": 7.845884323120117, + "learning_rate": 9.973992997399301e-06, + "loss": 0.1272, + "step": 26019 + }, + { + "epoch": 12.096699209669922, + "grad_norm": 4.620838642120361, + "learning_rate": 9.9739919973992e-06, + "loss": 0.0425, + "step": 26020 + }, + { + "epoch": 12.097164109716411, + "grad_norm": 5.823617935180664, + "learning_rate": 9.9739909973991e-06, + "loss": 0.0405, + "step": 26021 + }, + { + "epoch": 12.097629009762901, + "grad_norm": 2.6081554889678955, + "learning_rate": 9.973989997399001e-06, + "loss": 0.0151, + "step": 26022 + }, + { + "epoch": 12.098093909809391, + "grad_norm": 3.6028120517730713, + "learning_rate": 9.9739889973989e-06, + "loss": 0.0596, + "step": 26023 + }, + { + "epoch": 12.09855880985588, + "grad_norm": 3.856186866760254, + "learning_rate": 9.973987997398801e-06, + "loss": 0.0268, + "step": 26024 + }, + { + "epoch": 12.09902370990237, + "grad_norm": 2.4819839000701904, + "learning_rate": 9.9739869973987e-06, + "loss": 0.0177, + "step": 26025 + }, + { + "epoch": 12.09948860994886, + "grad_norm": 4.846835136413574, + "learning_rate": 9.9739859973986e-06, + "loss": 0.0733, + "step": 26026 + }, + { + "epoch": 12.09995350999535, + "grad_norm": 3.518371343612671, + "learning_rate": 9.9739849973985e-06, + "loss": 0.0246, + "step": 26027 + }, + { + "epoch": 12.100418410041842, + "grad_norm": 3.7458901405334473, + "learning_rate": 9.9739839973984e-06, + "loss": 0.0712, + "step": 26028 + }, + { + "epoch": 12.100883310088332, + "grad_norm": 2.2859890460968018, + "learning_rate": 9.973982997398301e-06, + "loss": 0.0173, + "step": 26029 + }, + { + "epoch": 12.101348210134821, + "grad_norm": 12.682252883911133, + "learning_rate": 9.973981997398202e-06, + "loss": 0.2048, + "step": 26030 + }, + { + "epoch": 12.101813110181311, + "grad_norm": 5.83231258392334, + "learning_rate": 9.9739809973981e-06, + "loss": 0.0408, + "step": 26031 + }, + { + "epoch": 12.102278010227801, + "grad_norm": 6.274165153503418, + "learning_rate": 9.973979997398e-06, + "loss": 0.1005, + "step": 26032 + }, + { + "epoch": 12.10274291027429, + "grad_norm": 2.9633169174194336, + "learning_rate": 9.9739789973979e-06, + "loss": 0.0306, + "step": 26033 + }, + { + "epoch": 12.10320781032078, + "grad_norm": 1.581075668334961, + "learning_rate": 9.9739779973978e-06, + "loss": 0.0129, + "step": 26034 + }, + { + "epoch": 12.10367271036727, + "grad_norm": 2.516401767730713, + "learning_rate": 9.973976997397701e-06, + "loss": 0.0191, + "step": 26035 + }, + { + "epoch": 12.104137610413762, + "grad_norm": 1.1527972221374512, + "learning_rate": 9.9739759973976e-06, + "loss": 0.0063, + "step": 26036 + }, + { + "epoch": 12.104602510460252, + "grad_norm": 6.756228923797607, + "learning_rate": 9.973974997397501e-06, + "loss": 0.1291, + "step": 26037 + }, + { + "epoch": 12.105067410506742, + "grad_norm": 3.405019760131836, + "learning_rate": 9.9739739973974e-06, + "loss": 0.0348, + "step": 26038 + }, + { + "epoch": 12.105532310553231, + "grad_norm": 3.663595676422119, + "learning_rate": 9.9739729973973e-06, + "loss": 0.0346, + "step": 26039 + }, + { + "epoch": 12.105997210599721, + "grad_norm": 1.1246370077133179, + "learning_rate": 9.9739719973972e-06, + "loss": 0.0075, + "step": 26040 + }, + { + "epoch": 12.106462110646211, + "grad_norm": 2.8478517532348633, + "learning_rate": 9.9739709973971e-06, + "loss": 0.0193, + "step": 26041 + }, + { + "epoch": 12.1069270106927, + "grad_norm": 3.3812620639801025, + "learning_rate": 9.973969997397001e-06, + "loss": 0.0618, + "step": 26042 + }, + { + "epoch": 12.10739191073919, + "grad_norm": 5.726168155670166, + "learning_rate": 9.9739689973969e-06, + "loss": 0.1449, + "step": 26043 + }, + { + "epoch": 12.10785681078568, + "grad_norm": 5.139943599700928, + "learning_rate": 9.9739679973968e-06, + "loss": 0.0961, + "step": 26044 + }, + { + "epoch": 12.108321710832172, + "grad_norm": 1.6288610696792603, + "learning_rate": 9.9739669973967e-06, + "loss": 0.0166, + "step": 26045 + }, + { + "epoch": 12.108786610878662, + "grad_norm": 8.34463882446289, + "learning_rate": 9.9739659973966e-06, + "loss": 0.0519, + "step": 26046 + }, + { + "epoch": 12.109251510925152, + "grad_norm": 11.973578453063965, + "learning_rate": 9.973964997396501e-06, + "loss": 0.0882, + "step": 26047 + }, + { + "epoch": 12.109716410971641, + "grad_norm": 1.8192018270492554, + "learning_rate": 9.973963997396402e-06, + "loss": 0.0141, + "step": 26048 + }, + { + "epoch": 12.110181311018131, + "grad_norm": 4.664975643157959, + "learning_rate": 9.9739629973963e-06, + "loss": 0.0358, + "step": 26049 + }, + { + "epoch": 12.110646211064621, + "grad_norm": 4.664182662963867, + "learning_rate": 9.973961997396201e-06, + "loss": 0.0481, + "step": 26050 + }, + { + "epoch": 12.11111111111111, + "grad_norm": 3.7949509620666504, + "learning_rate": 9.9739609973961e-06, + "loss": 0.0343, + "step": 26051 + }, + { + "epoch": 12.1115760111576, + "grad_norm": 4.216423511505127, + "learning_rate": 9.973959997396e-06, + "loss": 0.078, + "step": 26052 + }, + { + "epoch": 12.11204091120409, + "grad_norm": 5.432210922241211, + "learning_rate": 9.9739589973959e-06, + "loss": 0.0828, + "step": 26053 + }, + { + "epoch": 12.112505811250582, + "grad_norm": 5.515682220458984, + "learning_rate": 9.9739579973958e-06, + "loss": 0.0593, + "step": 26054 + }, + { + "epoch": 12.112970711297072, + "grad_norm": 2.0604825019836426, + "learning_rate": 9.973956997395701e-06, + "loss": 0.0166, + "step": 26055 + }, + { + "epoch": 12.113435611343562, + "grad_norm": 1.8339742422103882, + "learning_rate": 9.9739559973956e-06, + "loss": 0.012, + "step": 26056 + }, + { + "epoch": 12.113900511390051, + "grad_norm": 5.229381084442139, + "learning_rate": 9.9739549973955e-06, + "loss": 0.0257, + "step": 26057 + }, + { + "epoch": 12.114365411436541, + "grad_norm": 2.5553596019744873, + "learning_rate": 9.9739539973954e-06, + "loss": 0.0154, + "step": 26058 + }, + { + "epoch": 12.114830311483031, + "grad_norm": 5.146662712097168, + "learning_rate": 9.9739529973953e-06, + "loss": 0.1034, + "step": 26059 + }, + { + "epoch": 12.11529521152952, + "grad_norm": 2.662144660949707, + "learning_rate": 9.973951997395201e-06, + "loss": 0.0308, + "step": 26060 + }, + { + "epoch": 12.11576011157601, + "grad_norm": 7.1583428382873535, + "learning_rate": 9.973950997395102e-06, + "loss": 0.1478, + "step": 26061 + }, + { + "epoch": 12.1162250116225, + "grad_norm": 2.040924549102783, + "learning_rate": 9.973949997395e-06, + "loss": 0.0129, + "step": 26062 + }, + { + "epoch": 12.116689911668992, + "grad_norm": 6.728091716766357, + "learning_rate": 9.973948997394901e-06, + "loss": 0.0912, + "step": 26063 + }, + { + "epoch": 12.117154811715482, + "grad_norm": 3.1994197368621826, + "learning_rate": 9.9739479973948e-06, + "loss": 0.1256, + "step": 26064 + }, + { + "epoch": 12.117619711761972, + "grad_norm": 1.6405171155929565, + "learning_rate": 9.973946997394701e-06, + "loss": 0.0108, + "step": 26065 + }, + { + "epoch": 12.118084611808461, + "grad_norm": 2.4417121410369873, + "learning_rate": 9.9739459973946e-06, + "loss": 0.022, + "step": 26066 + }, + { + "epoch": 12.118549511854951, + "grad_norm": 8.102197647094727, + "learning_rate": 9.9739449973945e-06, + "loss": 0.0839, + "step": 26067 + }, + { + "epoch": 12.119014411901441, + "grad_norm": 4.826562404632568, + "learning_rate": 9.973943997394401e-06, + "loss": 0.0734, + "step": 26068 + }, + { + "epoch": 12.11947931194793, + "grad_norm": 3.8488287925720215, + "learning_rate": 9.9739429973943e-06, + "loss": 0.0485, + "step": 26069 + }, + { + "epoch": 12.11994421199442, + "grad_norm": 2.3812930583953857, + "learning_rate": 9.9739419973942e-06, + "loss": 0.0168, + "step": 26070 + }, + { + "epoch": 12.12040911204091, + "grad_norm": 5.119785785675049, + "learning_rate": 9.9739409973941e-06, + "loss": 0.038, + "step": 26071 + }, + { + "epoch": 12.120874012087402, + "grad_norm": 7.535277366638184, + "learning_rate": 9.973939997394e-06, + "loss": 0.1091, + "step": 26072 + }, + { + "epoch": 12.121338912133892, + "grad_norm": 9.3848295211792, + "learning_rate": 9.973938997393901e-06, + "loss": 0.1008, + "step": 26073 + }, + { + "epoch": 12.121803812180381, + "grad_norm": 7.571488857269287, + "learning_rate": 9.973937997393802e-06, + "loss": 0.1316, + "step": 26074 + }, + { + "epoch": 12.122268712226871, + "grad_norm": 4.734256744384766, + "learning_rate": 9.9739369973937e-06, + "loss": 0.0607, + "step": 26075 + }, + { + "epoch": 12.122733612273361, + "grad_norm": 1.5876003503799438, + "learning_rate": 9.9739359973936e-06, + "loss": 0.0108, + "step": 26076 + }, + { + "epoch": 12.123198512319851, + "grad_norm": 3.0926008224487305, + "learning_rate": 9.9739349973935e-06, + "loss": 0.0225, + "step": 26077 + }, + { + "epoch": 12.12366341236634, + "grad_norm": 5.237044334411621, + "learning_rate": 9.973933997393401e-06, + "loss": 0.0987, + "step": 26078 + }, + { + "epoch": 12.12412831241283, + "grad_norm": 6.122990131378174, + "learning_rate": 9.973932997393302e-06, + "loss": 0.0932, + "step": 26079 + }, + { + "epoch": 12.124593212459322, + "grad_norm": 0.9616968035697937, + "learning_rate": 9.9739319973932e-06, + "loss": 0.0074, + "step": 26080 + }, + { + "epoch": 12.125058112505812, + "grad_norm": 5.67318868637085, + "learning_rate": 9.973930997393101e-06, + "loss": 0.1011, + "step": 26081 + }, + { + "epoch": 12.125523012552302, + "grad_norm": 9.134212493896484, + "learning_rate": 9.973929997393e-06, + "loss": 0.0618, + "step": 26082 + }, + { + "epoch": 12.125987912598791, + "grad_norm": 5.056088924407959, + "learning_rate": 9.973928997392901e-06, + "loss": 0.0794, + "step": 26083 + }, + { + "epoch": 12.126452812645281, + "grad_norm": 2.0724024772644043, + "learning_rate": 9.9739279973928e-06, + "loss": 0.0106, + "step": 26084 + }, + { + "epoch": 12.126917712691771, + "grad_norm": 13.877533912658691, + "learning_rate": 9.9739269973927e-06, + "loss": 0.3046, + "step": 26085 + }, + { + "epoch": 12.127382612738261, + "grad_norm": 4.136419296264648, + "learning_rate": 9.973925997392601e-06, + "loss": 0.042, + "step": 26086 + }, + { + "epoch": 12.12784751278475, + "grad_norm": 5.704065799713135, + "learning_rate": 9.9739249973925e-06, + "loss": 0.0594, + "step": 26087 + }, + { + "epoch": 12.12831241283124, + "grad_norm": 3.9731802940368652, + "learning_rate": 9.9739239973924e-06, + "loss": 0.0378, + "step": 26088 + }, + { + "epoch": 12.128777312877732, + "grad_norm": 4.456221103668213, + "learning_rate": 9.9739229973923e-06, + "loss": 0.1055, + "step": 26089 + }, + { + "epoch": 12.129242212924222, + "grad_norm": 1.9762780666351318, + "learning_rate": 9.9739219973922e-06, + "loss": 0.0106, + "step": 26090 + }, + { + "epoch": 12.129707112970712, + "grad_norm": 3.768686294555664, + "learning_rate": 9.973920997392101e-06, + "loss": 0.0423, + "step": 26091 + }, + { + "epoch": 12.130172013017201, + "grad_norm": 8.484378814697266, + "learning_rate": 9.973919997392002e-06, + "loss": 0.0971, + "step": 26092 + }, + { + "epoch": 12.130636913063691, + "grad_norm": 0.49606209993362427, + "learning_rate": 9.9739189973919e-06, + "loss": 0.0039, + "step": 26093 + }, + { + "epoch": 12.131101813110181, + "grad_norm": 3.544496774673462, + "learning_rate": 9.973917997391801e-06, + "loss": 0.03, + "step": 26094 + }, + { + "epoch": 12.131566713156671, + "grad_norm": 3.6015124320983887, + "learning_rate": 9.9739169973917e-06, + "loss": 0.0388, + "step": 26095 + }, + { + "epoch": 12.13203161320316, + "grad_norm": 4.391792297363281, + "learning_rate": 9.973915997391601e-06, + "loss": 0.0366, + "step": 26096 + }, + { + "epoch": 12.13249651324965, + "grad_norm": 3.080564498901367, + "learning_rate": 9.9739149973915e-06, + "loss": 0.0237, + "step": 26097 + }, + { + "epoch": 12.132961413296142, + "grad_norm": 4.482898235321045, + "learning_rate": 9.9739139973914e-06, + "loss": 0.0772, + "step": 26098 + }, + { + "epoch": 12.133426313342632, + "grad_norm": 4.924830913543701, + "learning_rate": 9.973912997391301e-06, + "loss": 0.0283, + "step": 26099 + }, + { + "epoch": 12.133891213389122, + "grad_norm": 3.024125337600708, + "learning_rate": 9.9739119973912e-06, + "loss": 0.03, + "step": 26100 + }, + { + "epoch": 12.134356113435611, + "grad_norm": 7.732160568237305, + "learning_rate": 9.9739109973911e-06, + "loss": 0.1252, + "step": 26101 + }, + { + "epoch": 12.134821013482101, + "grad_norm": 10.874968528747559, + "learning_rate": 9.973909997391e-06, + "loss": 0.0245, + "step": 26102 + }, + { + "epoch": 12.135285913528591, + "grad_norm": 2.3951425552368164, + "learning_rate": 9.9739089973909e-06, + "loss": 0.0204, + "step": 26103 + }, + { + "epoch": 12.135750813575081, + "grad_norm": 10.874160766601562, + "learning_rate": 9.973907997390801e-06, + "loss": 0.0928, + "step": 26104 + }, + { + "epoch": 12.13621571362157, + "grad_norm": 4.651556968688965, + "learning_rate": 9.973906997390702e-06, + "loss": 0.0442, + "step": 26105 + }, + { + "epoch": 12.13668061366806, + "grad_norm": 2.509634256362915, + "learning_rate": 9.9739059973906e-06, + "loss": 0.0159, + "step": 26106 + }, + { + "epoch": 12.137145513714552, + "grad_norm": 5.804408550262451, + "learning_rate": 9.973904997390501e-06, + "loss": 0.082, + "step": 26107 + }, + { + "epoch": 12.137610413761042, + "grad_norm": 3.9707350730895996, + "learning_rate": 9.9739039973904e-06, + "loss": 0.0492, + "step": 26108 + }, + { + "epoch": 12.138075313807532, + "grad_norm": 7.0920209884643555, + "learning_rate": 9.9739029973903e-06, + "loss": 0.0281, + "step": 26109 + }, + { + "epoch": 12.138540213854021, + "grad_norm": 5.595985412597656, + "learning_rate": 9.9739019973902e-06, + "loss": 0.0578, + "step": 26110 + }, + { + "epoch": 12.139005113900511, + "grad_norm": 4.964511871337891, + "learning_rate": 9.9739009973901e-06, + "loss": 0.0578, + "step": 26111 + }, + { + "epoch": 12.139470013947001, + "grad_norm": 3.4847631454467773, + "learning_rate": 9.973899997390001e-06, + "loss": 0.0315, + "step": 26112 + }, + { + "epoch": 12.139934913993491, + "grad_norm": 2.0521929264068604, + "learning_rate": 9.9738989973899e-06, + "loss": 0.0149, + "step": 26113 + }, + { + "epoch": 12.14039981403998, + "grad_norm": 7.105027198791504, + "learning_rate": 9.973897997389801e-06, + "loss": 0.1707, + "step": 26114 + }, + { + "epoch": 12.140864714086472, + "grad_norm": 4.13774299621582, + "learning_rate": 9.9738969973897e-06, + "loss": 0.0511, + "step": 26115 + }, + { + "epoch": 12.141329614132962, + "grad_norm": 8.349865913391113, + "learning_rate": 9.9738959973896e-06, + "loss": 0.0909, + "step": 26116 + }, + { + "epoch": 12.141794514179452, + "grad_norm": 5.881717681884766, + "learning_rate": 9.973894997389501e-06, + "loss": 0.0498, + "step": 26117 + }, + { + "epoch": 12.142259414225942, + "grad_norm": 3.4185752868652344, + "learning_rate": 9.9738939973894e-06, + "loss": 0.0263, + "step": 26118 + }, + { + "epoch": 12.142724314272431, + "grad_norm": 3.9556257724761963, + "learning_rate": 9.9738929973893e-06, + "loss": 0.024, + "step": 26119 + }, + { + "epoch": 12.143189214318921, + "grad_norm": 3.552870273590088, + "learning_rate": 9.9738919973892e-06, + "loss": 0.0276, + "step": 26120 + }, + { + "epoch": 12.143654114365411, + "grad_norm": 3.8196566104888916, + "learning_rate": 9.9738909973891e-06, + "loss": 0.0302, + "step": 26121 + }, + { + "epoch": 12.144119014411901, + "grad_norm": 0.7681084871292114, + "learning_rate": 9.973889997389e-06, + "loss": 0.0068, + "step": 26122 + }, + { + "epoch": 12.14458391445839, + "grad_norm": 2.128685474395752, + "learning_rate": 9.9738889973889e-06, + "loss": 0.0135, + "step": 26123 + }, + { + "epoch": 12.145048814504882, + "grad_norm": 3.929251194000244, + "learning_rate": 9.9738879973888e-06, + "loss": 0.0459, + "step": 26124 + }, + { + "epoch": 12.145513714551372, + "grad_norm": 2.9906885623931885, + "learning_rate": 9.973886997388701e-06, + "loss": 0.0279, + "step": 26125 + }, + { + "epoch": 12.145978614597862, + "grad_norm": 3.484440803527832, + "learning_rate": 9.9738859973886e-06, + "loss": 0.0291, + "step": 26126 + }, + { + "epoch": 12.146443514644352, + "grad_norm": 9.858248710632324, + "learning_rate": 9.973884997388501e-06, + "loss": 0.1691, + "step": 26127 + }, + { + "epoch": 12.146908414690841, + "grad_norm": 5.719177722930908, + "learning_rate": 9.9738839973884e-06, + "loss": 0.0651, + "step": 26128 + }, + { + "epoch": 12.147373314737331, + "grad_norm": 4.902647018432617, + "learning_rate": 9.9738829973883e-06, + "loss": 0.0585, + "step": 26129 + }, + { + "epoch": 12.147838214783821, + "grad_norm": 0.4810870885848999, + "learning_rate": 9.973881997388201e-06, + "loss": 0.0046, + "step": 26130 + }, + { + "epoch": 12.148303114830311, + "grad_norm": 2.9649884700775146, + "learning_rate": 9.9738809973881e-06, + "loss": 0.0374, + "step": 26131 + }, + { + "epoch": 12.1487680148768, + "grad_norm": 4.1345319747924805, + "learning_rate": 9.973879997388001e-06, + "loss": 0.0586, + "step": 26132 + }, + { + "epoch": 12.149232914923292, + "grad_norm": 7.327733039855957, + "learning_rate": 9.9738789973879e-06, + "loss": 0.1056, + "step": 26133 + }, + { + "epoch": 12.149697814969782, + "grad_norm": 3.884708881378174, + "learning_rate": 9.9738779973878e-06, + "loss": 0.0603, + "step": 26134 + }, + { + "epoch": 12.150162715016272, + "grad_norm": 3.8059346675872803, + "learning_rate": 9.9738769973877e-06, + "loss": 0.0456, + "step": 26135 + }, + { + "epoch": 12.150627615062762, + "grad_norm": 5.5123162269592285, + "learning_rate": 9.9738759973876e-06, + "loss": 0.0876, + "step": 26136 + }, + { + "epoch": 12.151092515109251, + "grad_norm": 3.162999391555786, + "learning_rate": 9.9738749973875e-06, + "loss": 0.028, + "step": 26137 + }, + { + "epoch": 12.151557415155741, + "grad_norm": 4.103489398956299, + "learning_rate": 9.973873997387401e-06, + "loss": 0.0409, + "step": 26138 + }, + { + "epoch": 12.152022315202231, + "grad_norm": 2.9754819869995117, + "learning_rate": 9.9738729973873e-06, + "loss": 0.0371, + "step": 26139 + }, + { + "epoch": 12.15248721524872, + "grad_norm": 3.1863625049591064, + "learning_rate": 9.9738719973872e-06, + "loss": 0.0274, + "step": 26140 + }, + { + "epoch": 12.15295211529521, + "grad_norm": 5.08774471282959, + "learning_rate": 9.9738709973871e-06, + "loss": 0.0754, + "step": 26141 + }, + { + "epoch": 12.153417015341702, + "grad_norm": 5.119771957397461, + "learning_rate": 9.973869997387e-06, + "loss": 0.0569, + "step": 26142 + }, + { + "epoch": 12.153881915388192, + "grad_norm": 0.8891907334327698, + "learning_rate": 9.973868997386901e-06, + "loss": 0.0067, + "step": 26143 + }, + { + "epoch": 12.154346815434682, + "grad_norm": 4.656991004943848, + "learning_rate": 9.9738679973868e-06, + "loss": 0.0498, + "step": 26144 + }, + { + "epoch": 12.154811715481172, + "grad_norm": 2.710901975631714, + "learning_rate": 9.973866997386701e-06, + "loss": 0.0267, + "step": 26145 + }, + { + "epoch": 12.155276615527661, + "grad_norm": 5.9262824058532715, + "learning_rate": 9.9738659973866e-06, + "loss": 0.0626, + "step": 26146 + }, + { + "epoch": 12.155741515574151, + "grad_norm": 1.1399133205413818, + "learning_rate": 9.9738649973865e-06, + "loss": 0.0085, + "step": 26147 + }, + { + "epoch": 12.156206415620641, + "grad_norm": 3.211240530014038, + "learning_rate": 9.973863997386401e-06, + "loss": 0.027, + "step": 26148 + }, + { + "epoch": 12.15667131566713, + "grad_norm": 6.769516468048096, + "learning_rate": 9.9738629973863e-06, + "loss": 0.0797, + "step": 26149 + }, + { + "epoch": 12.15713621571362, + "grad_norm": 8.485909461975098, + "learning_rate": 9.973861997386201e-06, + "loss": 0.1408, + "step": 26150 + }, + { + "epoch": 12.157601115760112, + "grad_norm": 0.3798184394836426, + "learning_rate": 9.9738609973861e-06, + "loss": 0.0031, + "step": 26151 + }, + { + "epoch": 12.158066015806602, + "grad_norm": 2.776519536972046, + "learning_rate": 9.973859997386e-06, + "loss": 0.0249, + "step": 26152 + }, + { + "epoch": 12.158530915853092, + "grad_norm": 3.3411142826080322, + "learning_rate": 9.9738589973859e-06, + "loss": 0.0427, + "step": 26153 + }, + { + "epoch": 12.158995815899582, + "grad_norm": 2.390380859375, + "learning_rate": 9.9738579973858e-06, + "loss": 0.0151, + "step": 26154 + }, + { + "epoch": 12.159460715946071, + "grad_norm": 2.601381778717041, + "learning_rate": 9.9738569973857e-06, + "loss": 0.0205, + "step": 26155 + }, + { + "epoch": 12.159925615992561, + "grad_norm": 1.6517020463943481, + "learning_rate": 9.973855997385601e-06, + "loss": 0.0148, + "step": 26156 + }, + { + "epoch": 12.160390516039051, + "grad_norm": 4.426846981048584, + "learning_rate": 9.9738549973855e-06, + "loss": 0.074, + "step": 26157 + }, + { + "epoch": 12.16085541608554, + "grad_norm": 7.589346885681152, + "learning_rate": 9.973853997385401e-06, + "loss": 0.1305, + "step": 26158 + }, + { + "epoch": 12.161320316132032, + "grad_norm": 1.146072268486023, + "learning_rate": 9.9738529973853e-06, + "loss": 0.0066, + "step": 26159 + }, + { + "epoch": 12.161785216178522, + "grad_norm": 4.075537204742432, + "learning_rate": 9.9738519973852e-06, + "loss": 0.0526, + "step": 26160 + }, + { + "epoch": 12.162250116225012, + "grad_norm": 7.773307800292969, + "learning_rate": 9.973850997385101e-06, + "loss": 0.0385, + "step": 26161 + }, + { + "epoch": 12.162715016271502, + "grad_norm": 3.214400053024292, + "learning_rate": 9.973849997385e-06, + "loss": 0.0384, + "step": 26162 + }, + { + "epoch": 12.163179916317992, + "grad_norm": 1.5006189346313477, + "learning_rate": 9.973848997384901e-06, + "loss": 0.008, + "step": 26163 + }, + { + "epoch": 12.163644816364481, + "grad_norm": 3.6049020290374756, + "learning_rate": 9.9738479973848e-06, + "loss": 0.0275, + "step": 26164 + }, + { + "epoch": 12.164109716410971, + "grad_norm": 4.681026935577393, + "learning_rate": 9.9738469973847e-06, + "loss": 0.0194, + "step": 26165 + }, + { + "epoch": 12.164574616457461, + "grad_norm": 5.911472320556641, + "learning_rate": 9.9738459973846e-06, + "loss": 0.0879, + "step": 26166 + }, + { + "epoch": 12.16503951650395, + "grad_norm": 4.940888404846191, + "learning_rate": 9.9738449973845e-06, + "loss": 0.0601, + "step": 26167 + }, + { + "epoch": 12.165504416550442, + "grad_norm": 3.3877196311950684, + "learning_rate": 9.973843997384401e-06, + "loss": 0.0332, + "step": 26168 + }, + { + "epoch": 12.165969316596932, + "grad_norm": 3.7383530139923096, + "learning_rate": 9.973842997384302e-06, + "loss": 0.0405, + "step": 26169 + }, + { + "epoch": 12.166434216643422, + "grad_norm": 4.030243396759033, + "learning_rate": 9.9738419973842e-06, + "loss": 0.0514, + "step": 26170 + }, + { + "epoch": 12.166899116689912, + "grad_norm": 2.2620315551757812, + "learning_rate": 9.973840997384101e-06, + "loss": 0.0255, + "step": 26171 + }, + { + "epoch": 12.167364016736402, + "grad_norm": 8.165765762329102, + "learning_rate": 9.973839997384e-06, + "loss": 0.1085, + "step": 26172 + }, + { + "epoch": 12.167828916782891, + "grad_norm": 4.721707344055176, + "learning_rate": 9.9738389973839e-06, + "loss": 0.0405, + "step": 26173 + }, + { + "epoch": 12.168293816829381, + "grad_norm": 3.0845062732696533, + "learning_rate": 9.973837997383801e-06, + "loss": 0.0194, + "step": 26174 + }, + { + "epoch": 12.168758716875871, + "grad_norm": 5.088038921356201, + "learning_rate": 9.9738369973837e-06, + "loss": 0.0759, + "step": 26175 + }, + { + "epoch": 12.16922361692236, + "grad_norm": 4.581505298614502, + "learning_rate": 9.973835997383601e-06, + "loss": 0.065, + "step": 26176 + }, + { + "epoch": 12.169688516968852, + "grad_norm": 4.248290538787842, + "learning_rate": 9.9738349973835e-06, + "loss": 0.0578, + "step": 26177 + }, + { + "epoch": 12.170153417015342, + "grad_norm": 6.993419647216797, + "learning_rate": 9.9738339973834e-06, + "loss": 0.0773, + "step": 26178 + }, + { + "epoch": 12.170618317061832, + "grad_norm": 0.9427377581596375, + "learning_rate": 9.9738329973833e-06, + "loss": 0.0098, + "step": 26179 + }, + { + "epoch": 12.171083217108322, + "grad_norm": 3.057316541671753, + "learning_rate": 9.9738319973832e-06, + "loss": 0.0224, + "step": 26180 + }, + { + "epoch": 12.171548117154812, + "grad_norm": 3.3859481811523438, + "learning_rate": 9.973830997383101e-06, + "loss": 0.039, + "step": 26181 + }, + { + "epoch": 12.172013017201301, + "grad_norm": 5.004383087158203, + "learning_rate": 9.973829997383002e-06, + "loss": 0.0511, + "step": 26182 + }, + { + "epoch": 12.172477917247791, + "grad_norm": 0.5295270085334778, + "learning_rate": 9.9738289973829e-06, + "loss": 0.0042, + "step": 26183 + }, + { + "epoch": 12.172942817294281, + "grad_norm": 5.209824085235596, + "learning_rate": 9.9738279973828e-06, + "loss": 0.0583, + "step": 26184 + }, + { + "epoch": 12.17340771734077, + "grad_norm": 0.8598915934562683, + "learning_rate": 9.9738269973827e-06, + "loss": 0.0049, + "step": 26185 + }, + { + "epoch": 12.173872617387262, + "grad_norm": 5.102747917175293, + "learning_rate": 9.9738259973826e-06, + "loss": 0.0684, + "step": 26186 + }, + { + "epoch": 12.174337517433752, + "grad_norm": 5.29545259475708, + "learning_rate": 9.973824997382502e-06, + "loss": 0.0666, + "step": 26187 + }, + { + "epoch": 12.174802417480242, + "grad_norm": 2.549659013748169, + "learning_rate": 9.9738239973824e-06, + "loss": 0.0175, + "step": 26188 + }, + { + "epoch": 12.175267317526732, + "grad_norm": 7.741885662078857, + "learning_rate": 9.973822997382301e-06, + "loss": 0.1521, + "step": 26189 + }, + { + "epoch": 12.175732217573222, + "grad_norm": 1.9527658224105835, + "learning_rate": 9.9738219973822e-06, + "loss": 0.0196, + "step": 26190 + }, + { + "epoch": 12.176197117619711, + "grad_norm": 4.494850158691406, + "learning_rate": 9.9738209973821e-06, + "loss": 0.029, + "step": 26191 + }, + { + "epoch": 12.176662017666201, + "grad_norm": 4.511179447174072, + "learning_rate": 9.973819997382e-06, + "loss": 0.0321, + "step": 26192 + }, + { + "epoch": 12.177126917712691, + "grad_norm": 6.916108131408691, + "learning_rate": 9.9738189973819e-06, + "loss": 0.1192, + "step": 26193 + }, + { + "epoch": 12.177591817759183, + "grad_norm": 3.337829828262329, + "learning_rate": 9.973817997381801e-06, + "loss": 0.0262, + "step": 26194 + }, + { + "epoch": 12.178056717805672, + "grad_norm": 7.698736190795898, + "learning_rate": 9.9738169973817e-06, + "loss": 0.0627, + "step": 26195 + }, + { + "epoch": 12.178521617852162, + "grad_norm": 2.4810311794281006, + "learning_rate": 9.9738159973816e-06, + "loss": 0.0288, + "step": 26196 + }, + { + "epoch": 12.178986517898652, + "grad_norm": 4.235721111297607, + "learning_rate": 9.9738149973815e-06, + "loss": 0.0333, + "step": 26197 + }, + { + "epoch": 12.179451417945142, + "grad_norm": 3.693782091140747, + "learning_rate": 9.9738139973814e-06, + "loss": 0.0592, + "step": 26198 + }, + { + "epoch": 12.179916317991632, + "grad_norm": 6.0898027420043945, + "learning_rate": 9.973812997381301e-06, + "loss": 0.0777, + "step": 26199 + }, + { + "epoch": 12.180381218038121, + "grad_norm": 2.9377620220184326, + "learning_rate": 9.973811997381202e-06, + "loss": 0.0458, + "step": 26200 + }, + { + "epoch": 12.180846118084611, + "grad_norm": 6.485260486602783, + "learning_rate": 9.9738109973811e-06, + "loss": 0.1391, + "step": 26201 + }, + { + "epoch": 12.181311018131101, + "grad_norm": 2.405585765838623, + "learning_rate": 9.973809997381001e-06, + "loss": 0.0115, + "step": 26202 + }, + { + "epoch": 12.181775918177593, + "grad_norm": 1.4373779296875, + "learning_rate": 9.9738089973809e-06, + "loss": 0.0099, + "step": 26203 + }, + { + "epoch": 12.182240818224082, + "grad_norm": 4.446032524108887, + "learning_rate": 9.9738079973808e-06, + "loss": 0.0324, + "step": 26204 + }, + { + "epoch": 12.182705718270572, + "grad_norm": 2.5758557319641113, + "learning_rate": 9.9738069973807e-06, + "loss": 0.0097, + "step": 26205 + }, + { + "epoch": 12.183170618317062, + "grad_norm": 3.3860111236572266, + "learning_rate": 9.9738059973806e-06, + "loss": 0.0342, + "step": 26206 + }, + { + "epoch": 12.183635518363552, + "grad_norm": 3.3035991191864014, + "learning_rate": 9.973804997380501e-06, + "loss": 0.0349, + "step": 26207 + }, + { + "epoch": 12.184100418410042, + "grad_norm": 3.498471260070801, + "learning_rate": 9.9738039973804e-06, + "loss": 0.0266, + "step": 26208 + }, + { + "epoch": 12.184565318456531, + "grad_norm": 8.063207626342773, + "learning_rate": 9.9738029973803e-06, + "loss": 0.1207, + "step": 26209 + }, + { + "epoch": 12.185030218503021, + "grad_norm": 9.014734268188477, + "learning_rate": 9.9738019973802e-06, + "loss": 0.0908, + "step": 26210 + }, + { + "epoch": 12.185495118549511, + "grad_norm": 6.188704013824463, + "learning_rate": 9.9738009973801e-06, + "loss": 0.0756, + "step": 26211 + }, + { + "epoch": 12.185960018596003, + "grad_norm": 1.4209755659103394, + "learning_rate": 9.973799997380001e-06, + "loss": 0.0074, + "step": 26212 + }, + { + "epoch": 12.186424918642492, + "grad_norm": 2.8036768436431885, + "learning_rate": 9.973798997379902e-06, + "loss": 0.0154, + "step": 26213 + }, + { + "epoch": 12.186889818688982, + "grad_norm": 7.207315921783447, + "learning_rate": 9.9737979973798e-06, + "loss": 0.0968, + "step": 26214 + }, + { + "epoch": 12.187354718735472, + "grad_norm": 3.5034074783325195, + "learning_rate": 9.973796997379701e-06, + "loss": 0.06, + "step": 26215 + }, + { + "epoch": 12.187819618781962, + "grad_norm": 2.044978380203247, + "learning_rate": 9.9737959973796e-06, + "loss": 0.0185, + "step": 26216 + }, + { + "epoch": 12.188284518828452, + "grad_norm": 5.87162971496582, + "learning_rate": 9.973794997379501e-06, + "loss": 0.034, + "step": 26217 + }, + { + "epoch": 12.188749418874941, + "grad_norm": 3.6619324684143066, + "learning_rate": 9.9737939973794e-06, + "loss": 0.0384, + "step": 26218 + }, + { + "epoch": 12.189214318921431, + "grad_norm": 5.9576849937438965, + "learning_rate": 9.9737929973793e-06, + "loss": 0.0499, + "step": 26219 + }, + { + "epoch": 12.189679218967921, + "grad_norm": 4.0590620040893555, + "learning_rate": 9.973791997379201e-06, + "loss": 0.0592, + "step": 26220 + }, + { + "epoch": 12.190144119014413, + "grad_norm": 2.693695306777954, + "learning_rate": 9.9737909973791e-06, + "loss": 0.0231, + "step": 26221 + }, + { + "epoch": 12.190609019060902, + "grad_norm": 4.968390464782715, + "learning_rate": 9.973789997379e-06, + "loss": 0.063, + "step": 26222 + }, + { + "epoch": 12.191073919107392, + "grad_norm": 4.356352806091309, + "learning_rate": 9.9737889973789e-06, + "loss": 0.021, + "step": 26223 + }, + { + "epoch": 12.191538819153882, + "grad_norm": 3.5445525646209717, + "learning_rate": 9.9737879973788e-06, + "loss": 0.0389, + "step": 26224 + }, + { + "epoch": 12.192003719200372, + "grad_norm": 8.219742774963379, + "learning_rate": 9.973786997378701e-06, + "loss": 0.12, + "step": 26225 + }, + { + "epoch": 12.192468619246862, + "grad_norm": 7.700135231018066, + "learning_rate": 9.973785997378602e-06, + "loss": 0.0904, + "step": 26226 + }, + { + "epoch": 12.192933519293351, + "grad_norm": 6.086438179016113, + "learning_rate": 9.9737849973785e-06, + "loss": 0.1029, + "step": 26227 + }, + { + "epoch": 12.193398419339841, + "grad_norm": 2.99617075920105, + "learning_rate": 9.9737839973784e-06, + "loss": 0.0283, + "step": 26228 + }, + { + "epoch": 12.193863319386331, + "grad_norm": 7.048488140106201, + "learning_rate": 9.9737829973783e-06, + "loss": 0.0961, + "step": 26229 + }, + { + "epoch": 12.194328219432823, + "grad_norm": 0.6054759621620178, + "learning_rate": 9.973781997378201e-06, + "loss": 0.0049, + "step": 26230 + }, + { + "epoch": 12.194793119479312, + "grad_norm": 4.755608558654785, + "learning_rate": 9.9737809973781e-06, + "loss": 0.0843, + "step": 26231 + }, + { + "epoch": 12.195258019525802, + "grad_norm": 5.340970039367676, + "learning_rate": 9.973779997378e-06, + "loss": 0.1219, + "step": 26232 + }, + { + "epoch": 12.195722919572292, + "grad_norm": 2.3664464950561523, + "learning_rate": 9.973778997377901e-06, + "loss": 0.0242, + "step": 26233 + }, + { + "epoch": 12.196187819618782, + "grad_norm": 0.9687680602073669, + "learning_rate": 9.9737779973778e-06, + "loss": 0.0079, + "step": 26234 + }, + { + "epoch": 12.196652719665272, + "grad_norm": 2.6762478351593018, + "learning_rate": 9.973776997377701e-06, + "loss": 0.0236, + "step": 26235 + }, + { + "epoch": 12.197117619711761, + "grad_norm": 1.708276391029358, + "learning_rate": 9.9737759973776e-06, + "loss": 0.0188, + "step": 26236 + }, + { + "epoch": 12.197582519758251, + "grad_norm": 3.163727283477783, + "learning_rate": 9.9737749973775e-06, + "loss": 0.0331, + "step": 26237 + }, + { + "epoch": 12.198047419804743, + "grad_norm": 3.342069149017334, + "learning_rate": 9.973773997377401e-06, + "loss": 0.0222, + "step": 26238 + }, + { + "epoch": 12.198512319851233, + "grad_norm": 7.709033012390137, + "learning_rate": 9.9737729973773e-06, + "loss": 0.0705, + "step": 26239 + }, + { + "epoch": 12.198977219897722, + "grad_norm": 3.2846925258636475, + "learning_rate": 9.9737719973772e-06, + "loss": 0.0252, + "step": 26240 + }, + { + "epoch": 12.199442119944212, + "grad_norm": 4.219098091125488, + "learning_rate": 9.9737709973771e-06, + "loss": 0.0478, + "step": 26241 + }, + { + "epoch": 12.199907019990702, + "grad_norm": 5.43328332901001, + "learning_rate": 9.973769997377e-06, + "loss": 0.0473, + "step": 26242 + }, + { + "epoch": 12.200371920037192, + "grad_norm": 7.774013042449951, + "learning_rate": 9.973768997376901e-06, + "loss": 0.0627, + "step": 26243 + }, + { + "epoch": 12.200836820083682, + "grad_norm": 3.6273319721221924, + "learning_rate": 9.973767997376802e-06, + "loss": 0.0425, + "step": 26244 + }, + { + "epoch": 12.201301720130171, + "grad_norm": 7.237170696258545, + "learning_rate": 9.9737669973767e-06, + "loss": 0.097, + "step": 26245 + }, + { + "epoch": 12.201766620176661, + "grad_norm": 2.8500683307647705, + "learning_rate": 9.973765997376601e-06, + "loss": 0.024, + "step": 26246 + }, + { + "epoch": 12.202231520223153, + "grad_norm": 1.520923137664795, + "learning_rate": 9.9737649973765e-06, + "loss": 0.0095, + "step": 26247 + }, + { + "epoch": 12.202696420269643, + "grad_norm": 4.283890247344971, + "learning_rate": 9.973763997376401e-06, + "loss": 0.0537, + "step": 26248 + }, + { + "epoch": 12.203161320316132, + "grad_norm": 2.7154295444488525, + "learning_rate": 9.9737629973763e-06, + "loss": 0.0154, + "step": 26249 + }, + { + "epoch": 12.203626220362622, + "grad_norm": 4.981987953186035, + "learning_rate": 9.9737619973762e-06, + "loss": 0.1045, + "step": 26250 + }, + { + "epoch": 12.204091120409112, + "grad_norm": 7.092301845550537, + "learning_rate": 9.973760997376101e-06, + "loss": 0.1977, + "step": 26251 + }, + { + "epoch": 12.204556020455602, + "grad_norm": 6.35330867767334, + "learning_rate": 9.973759997376e-06, + "loss": 0.0666, + "step": 26252 + }, + { + "epoch": 12.205020920502092, + "grad_norm": 1.080971121788025, + "learning_rate": 9.973758997375901e-06, + "loss": 0.0097, + "step": 26253 + }, + { + "epoch": 12.205485820548581, + "grad_norm": 2.6537437438964844, + "learning_rate": 9.9737579973758e-06, + "loss": 0.0288, + "step": 26254 + }, + { + "epoch": 12.205950720595071, + "grad_norm": 5.706347942352295, + "learning_rate": 9.9737569973757e-06, + "loss": 0.048, + "step": 26255 + }, + { + "epoch": 12.206415620641563, + "grad_norm": 5.923040390014648, + "learning_rate": 9.973755997375601e-06, + "loss": 0.1115, + "step": 26256 + }, + { + "epoch": 12.206880520688053, + "grad_norm": 4.611391067504883, + "learning_rate": 9.973754997375502e-06, + "loss": 0.055, + "step": 26257 + }, + { + "epoch": 12.207345420734542, + "grad_norm": 6.925513744354248, + "learning_rate": 9.9737539973754e-06, + "loss": 0.1638, + "step": 26258 + }, + { + "epoch": 12.207810320781032, + "grad_norm": 10.814409255981445, + "learning_rate": 9.9737529973753e-06, + "loss": 0.1846, + "step": 26259 + }, + { + "epoch": 12.208275220827522, + "grad_norm": 5.324868202209473, + "learning_rate": 9.9737519973752e-06, + "loss": 0.0731, + "step": 26260 + }, + { + "epoch": 12.208740120874012, + "grad_norm": 2.939591884613037, + "learning_rate": 9.9737509973751e-06, + "loss": 0.0393, + "step": 26261 + }, + { + "epoch": 12.209205020920502, + "grad_norm": 4.246665954589844, + "learning_rate": 9.973749997375e-06, + "loss": 0.0626, + "step": 26262 + }, + { + "epoch": 12.209669920966991, + "grad_norm": 1.8182661533355713, + "learning_rate": 9.9737489973749e-06, + "loss": 0.0209, + "step": 26263 + }, + { + "epoch": 12.210134821013481, + "grad_norm": 6.625327110290527, + "learning_rate": 9.973747997374801e-06, + "loss": 0.0616, + "step": 26264 + }, + { + "epoch": 12.210599721059973, + "grad_norm": 1.3441423177719116, + "learning_rate": 9.9737469973747e-06, + "loss": 0.0107, + "step": 26265 + }, + { + "epoch": 12.211064621106463, + "grad_norm": 3.266880750656128, + "learning_rate": 9.973745997374601e-06, + "loss": 0.0318, + "step": 26266 + }, + { + "epoch": 12.211529521152952, + "grad_norm": 12.333740234375, + "learning_rate": 9.9737449973745e-06, + "loss": 0.0341, + "step": 26267 + }, + { + "epoch": 12.211994421199442, + "grad_norm": 3.058340311050415, + "learning_rate": 9.9737439973744e-06, + "loss": 0.0198, + "step": 26268 + }, + { + "epoch": 12.212459321245932, + "grad_norm": 4.090423583984375, + "learning_rate": 9.973742997374301e-06, + "loss": 0.0831, + "step": 26269 + }, + { + "epoch": 12.212924221292422, + "grad_norm": 3.6261110305786133, + "learning_rate": 9.9737419973742e-06, + "loss": 0.0434, + "step": 26270 + }, + { + "epoch": 12.213389121338912, + "grad_norm": 1.333211064338684, + "learning_rate": 9.973740997374101e-06, + "loss": 0.0097, + "step": 26271 + }, + { + "epoch": 12.213854021385401, + "grad_norm": 3.990516185760498, + "learning_rate": 9.973739997374e-06, + "loss": 0.0564, + "step": 26272 + }, + { + "epoch": 12.214318921431893, + "grad_norm": 7.1461100578308105, + "learning_rate": 9.9737389973739e-06, + "loss": 0.0293, + "step": 26273 + }, + { + "epoch": 12.214783821478383, + "grad_norm": 2.4053571224212646, + "learning_rate": 9.9737379973738e-06, + "loss": 0.0218, + "step": 26274 + }, + { + "epoch": 12.215248721524873, + "grad_norm": 2.827373743057251, + "learning_rate": 9.9737369973737e-06, + "loss": 0.0447, + "step": 26275 + }, + { + "epoch": 12.215713621571362, + "grad_norm": 1.2943193912506104, + "learning_rate": 9.9737359973736e-06, + "loss": 0.0097, + "step": 26276 + }, + { + "epoch": 12.216178521617852, + "grad_norm": 3.0947790145874023, + "learning_rate": 9.973734997373501e-06, + "loss": 0.0197, + "step": 26277 + }, + { + "epoch": 12.216643421664342, + "grad_norm": 0.271811306476593, + "learning_rate": 9.9737339973734e-06, + "loss": 0.0015, + "step": 26278 + }, + { + "epoch": 12.217108321710832, + "grad_norm": 4.513382434844971, + "learning_rate": 9.973732997373301e-06, + "loss": 0.0651, + "step": 26279 + }, + { + "epoch": 12.217573221757322, + "grad_norm": 3.3916072845458984, + "learning_rate": 9.9737319973732e-06, + "loss": 0.0294, + "step": 26280 + }, + { + "epoch": 12.218038121803811, + "grad_norm": 5.461623191833496, + "learning_rate": 9.9737309973731e-06, + "loss": 0.0687, + "step": 26281 + }, + { + "epoch": 12.218503021850303, + "grad_norm": 2.5166776180267334, + "learning_rate": 9.973729997373001e-06, + "loss": 0.0154, + "step": 26282 + }, + { + "epoch": 12.218967921896793, + "grad_norm": 5.726863384246826, + "learning_rate": 9.9737289973729e-06, + "loss": 0.0911, + "step": 26283 + }, + { + "epoch": 12.219432821943283, + "grad_norm": 8.670515060424805, + "learning_rate": 9.973727997372801e-06, + "loss": 0.1135, + "step": 26284 + }, + { + "epoch": 12.219897721989772, + "grad_norm": 3.2991855144500732, + "learning_rate": 9.9737269973727e-06, + "loss": 0.0378, + "step": 26285 + }, + { + "epoch": 12.220362622036262, + "grad_norm": 1.9205212593078613, + "learning_rate": 9.9737259973726e-06, + "loss": 0.0106, + "step": 26286 + }, + { + "epoch": 12.220827522082752, + "grad_norm": 1.3103742599487305, + "learning_rate": 9.9737249973725e-06, + "loss": 0.0078, + "step": 26287 + }, + { + "epoch": 12.221292422129242, + "grad_norm": 5.098587989807129, + "learning_rate": 9.9737239973724e-06, + "loss": 0.0973, + "step": 26288 + }, + { + "epoch": 12.221757322175732, + "grad_norm": 4.964829444885254, + "learning_rate": 9.9737229973723e-06, + "loss": 0.0901, + "step": 26289 + }, + { + "epoch": 12.222222222222221, + "grad_norm": 3.8615894317626953, + "learning_rate": 9.973721997372201e-06, + "loss": 0.0207, + "step": 26290 + }, + { + "epoch": 12.222687122268713, + "grad_norm": 4.413858890533447, + "learning_rate": 9.9737209973721e-06, + "loss": 0.0407, + "step": 26291 + }, + { + "epoch": 12.223152022315203, + "grad_norm": 4.1488728523254395, + "learning_rate": 9.973719997372e-06, + "loss": 0.0287, + "step": 26292 + }, + { + "epoch": 12.223616922361693, + "grad_norm": 3.77931809425354, + "learning_rate": 9.9737189973719e-06, + "loss": 0.0244, + "step": 26293 + }, + { + "epoch": 12.224081822408182, + "grad_norm": 4.194464683532715, + "learning_rate": 9.9737179973718e-06, + "loss": 0.0923, + "step": 26294 + }, + { + "epoch": 12.224546722454672, + "grad_norm": 4.9175896644592285, + "learning_rate": 9.973716997371701e-06, + "loss": 0.1049, + "step": 26295 + }, + { + "epoch": 12.225011622501162, + "grad_norm": 6.380578517913818, + "learning_rate": 9.9737159973716e-06, + "loss": 0.0834, + "step": 26296 + }, + { + "epoch": 12.225476522547652, + "grad_norm": 6.826848983764648, + "learning_rate": 9.973714997371501e-06, + "loss": 0.0391, + "step": 26297 + }, + { + "epoch": 12.225941422594142, + "grad_norm": 5.379398822784424, + "learning_rate": 9.9737139973714e-06, + "loss": 0.0448, + "step": 26298 + }, + { + "epoch": 12.226406322640631, + "grad_norm": 2.8816399574279785, + "learning_rate": 9.9737129973713e-06, + "loss": 0.0205, + "step": 26299 + }, + { + "epoch": 12.226871222687123, + "grad_norm": 3.0170602798461914, + "learning_rate": 9.9737119973712e-06, + "loss": 0.0355, + "step": 26300 + }, + { + "epoch": 12.227336122733613, + "grad_norm": 2.7631173133850098, + "learning_rate": 9.9737109973711e-06, + "loss": 0.021, + "step": 26301 + }, + { + "epoch": 12.227801022780103, + "grad_norm": 2.5004918575286865, + "learning_rate": 9.973709997371001e-06, + "loss": 0.0107, + "step": 26302 + }, + { + "epoch": 12.228265922826592, + "grad_norm": 2.848238706588745, + "learning_rate": 9.9737089973709e-06, + "loss": 0.0172, + "step": 26303 + }, + { + "epoch": 12.228730822873082, + "grad_norm": 4.041922092437744, + "learning_rate": 9.9737079973708e-06, + "loss": 0.0395, + "step": 26304 + }, + { + "epoch": 12.229195722919572, + "grad_norm": 2.2466511726379395, + "learning_rate": 9.9737069973707e-06, + "loss": 0.02, + "step": 26305 + }, + { + "epoch": 12.229660622966062, + "grad_norm": 4.005514144897461, + "learning_rate": 9.9737059973706e-06, + "loss": 0.0378, + "step": 26306 + }, + { + "epoch": 12.230125523012552, + "grad_norm": 2.1179065704345703, + "learning_rate": 9.9737049973705e-06, + "loss": 0.0175, + "step": 26307 + }, + { + "epoch": 12.230590423059041, + "grad_norm": 2.208688974380493, + "learning_rate": 9.973703997370401e-06, + "loss": 0.0175, + "step": 26308 + }, + { + "epoch": 12.231055323105533, + "grad_norm": 2.6108314990997314, + "learning_rate": 9.9737029973703e-06, + "loss": 0.017, + "step": 26309 + }, + { + "epoch": 12.231520223152023, + "grad_norm": 8.3438720703125, + "learning_rate": 9.973701997370201e-06, + "loss": 0.1111, + "step": 26310 + }, + { + "epoch": 12.231985123198513, + "grad_norm": 2.913902759552002, + "learning_rate": 9.9737009973701e-06, + "loss": 0.0247, + "step": 26311 + }, + { + "epoch": 12.232450023245002, + "grad_norm": 1.2490953207015991, + "learning_rate": 9.97369999737e-06, + "loss": 0.0101, + "step": 26312 + }, + { + "epoch": 12.232914923291492, + "grad_norm": 6.263229846954346, + "learning_rate": 9.973698997369901e-06, + "loss": 0.1171, + "step": 26313 + }, + { + "epoch": 12.233379823337982, + "grad_norm": 4.856832027435303, + "learning_rate": 9.9736979973698e-06, + "loss": 0.1098, + "step": 26314 + }, + { + "epoch": 12.233844723384472, + "grad_norm": 7.041647434234619, + "learning_rate": 9.973696997369701e-06, + "loss": 0.109, + "step": 26315 + }, + { + "epoch": 12.234309623430962, + "grad_norm": 3.653599977493286, + "learning_rate": 9.9736959973696e-06, + "loss": 0.0408, + "step": 26316 + }, + { + "epoch": 12.234774523477453, + "grad_norm": 8.534360885620117, + "learning_rate": 9.9736949973695e-06, + "loss": 0.1655, + "step": 26317 + }, + { + "epoch": 12.235239423523943, + "grad_norm": 3.6748316287994385, + "learning_rate": 9.9736939973694e-06, + "loss": 0.0288, + "step": 26318 + }, + { + "epoch": 12.235704323570433, + "grad_norm": 3.0253498554229736, + "learning_rate": 9.9736929973693e-06, + "loss": 0.0287, + "step": 26319 + }, + { + "epoch": 12.236169223616923, + "grad_norm": 5.037949085235596, + "learning_rate": 9.973691997369201e-06, + "loss": 0.0736, + "step": 26320 + }, + { + "epoch": 12.236634123663412, + "grad_norm": 4.144909381866455, + "learning_rate": 9.973690997369102e-06, + "loss": 0.0336, + "step": 26321 + }, + { + "epoch": 12.237099023709902, + "grad_norm": 8.646553993225098, + "learning_rate": 9.973689997369e-06, + "loss": 0.101, + "step": 26322 + }, + { + "epoch": 12.237563923756392, + "grad_norm": 5.371391773223877, + "learning_rate": 9.973688997368901e-06, + "loss": 0.0676, + "step": 26323 + }, + { + "epoch": 12.238028823802882, + "grad_norm": 2.2414023876190186, + "learning_rate": 9.9736879973688e-06, + "loss": 0.0157, + "step": 26324 + }, + { + "epoch": 12.238493723849372, + "grad_norm": 4.217485427856445, + "learning_rate": 9.9736869973687e-06, + "loss": 0.0207, + "step": 26325 + }, + { + "epoch": 12.238958623895863, + "grad_norm": 5.715732574462891, + "learning_rate": 9.973685997368601e-06, + "loss": 0.1026, + "step": 26326 + }, + { + "epoch": 12.239423523942353, + "grad_norm": 8.076650619506836, + "learning_rate": 9.9736849973685e-06, + "loss": 0.0947, + "step": 26327 + }, + { + "epoch": 12.239888423988843, + "grad_norm": 4.813732147216797, + "learning_rate": 9.973683997368401e-06, + "loss": 0.0861, + "step": 26328 + }, + { + "epoch": 12.240353324035333, + "grad_norm": 1.6706622838974, + "learning_rate": 9.9736829973683e-06, + "loss": 0.0123, + "step": 26329 + }, + { + "epoch": 12.240818224081822, + "grad_norm": 5.739108562469482, + "learning_rate": 9.9736819973682e-06, + "loss": 0.0713, + "step": 26330 + }, + { + "epoch": 12.241283124128312, + "grad_norm": 1.9552019834518433, + "learning_rate": 9.9736809973681e-06, + "loss": 0.0168, + "step": 26331 + }, + { + "epoch": 12.241748024174802, + "grad_norm": 2.7824878692626953, + "learning_rate": 9.973679997368e-06, + "loss": 0.0317, + "step": 26332 + }, + { + "epoch": 12.242212924221292, + "grad_norm": 2.5393178462982178, + "learning_rate": 9.973678997367901e-06, + "loss": 0.0205, + "step": 26333 + }, + { + "epoch": 12.242677824267782, + "grad_norm": 4.786830902099609, + "learning_rate": 9.973677997367802e-06, + "loss": 0.0533, + "step": 26334 + }, + { + "epoch": 12.243142724314273, + "grad_norm": 2.1276891231536865, + "learning_rate": 9.9736769973677e-06, + "loss": 0.0185, + "step": 26335 + }, + { + "epoch": 12.243607624360763, + "grad_norm": 4.082505702972412, + "learning_rate": 9.9736759973676e-06, + "loss": 0.0271, + "step": 26336 + }, + { + "epoch": 12.244072524407253, + "grad_norm": 2.7884562015533447, + "learning_rate": 9.9736749973675e-06, + "loss": 0.0405, + "step": 26337 + }, + { + "epoch": 12.244537424453743, + "grad_norm": 5.203202247619629, + "learning_rate": 9.973673997367401e-06, + "loss": 0.0545, + "step": 26338 + }, + { + "epoch": 12.245002324500232, + "grad_norm": 5.028334617614746, + "learning_rate": 9.973672997367302e-06, + "loss": 0.0675, + "step": 26339 + }, + { + "epoch": 12.245467224546722, + "grad_norm": 3.3607382774353027, + "learning_rate": 9.9736719973672e-06, + "loss": 0.0225, + "step": 26340 + }, + { + "epoch": 12.245932124593212, + "grad_norm": 5.042446136474609, + "learning_rate": 9.973670997367101e-06, + "loss": 0.048, + "step": 26341 + }, + { + "epoch": 12.246397024639702, + "grad_norm": 5.9494733810424805, + "learning_rate": 9.973669997367e-06, + "loss": 0.0435, + "step": 26342 + }, + { + "epoch": 12.246861924686192, + "grad_norm": 2.937466859817505, + "learning_rate": 9.9736689973669e-06, + "loss": 0.0153, + "step": 26343 + }, + { + "epoch": 12.247326824732683, + "grad_norm": 3.4735352993011475, + "learning_rate": 9.9736679973668e-06, + "loss": 0.0454, + "step": 26344 + }, + { + "epoch": 12.247791724779173, + "grad_norm": 7.9757280349731445, + "learning_rate": 9.9736669973667e-06, + "loss": 0.1192, + "step": 26345 + }, + { + "epoch": 12.248256624825663, + "grad_norm": 12.859465599060059, + "learning_rate": 9.973665997366601e-06, + "loss": 0.2057, + "step": 26346 + }, + { + "epoch": 12.248721524872153, + "grad_norm": 4.8401055335998535, + "learning_rate": 9.9736649973665e-06, + "loss": 0.0553, + "step": 26347 + }, + { + "epoch": 12.249186424918642, + "grad_norm": 1.899580478668213, + "learning_rate": 9.9736639973664e-06, + "loss": 0.015, + "step": 26348 + }, + { + "epoch": 12.249651324965132, + "grad_norm": 2.691619396209717, + "learning_rate": 9.9736629973663e-06, + "loss": 0.0297, + "step": 26349 + }, + { + "epoch": 12.250116225011622, + "grad_norm": 3.4680631160736084, + "learning_rate": 9.9736619973662e-06, + "loss": 0.0265, + "step": 26350 + }, + { + "epoch": 12.250581125058112, + "grad_norm": 5.3589091300964355, + "learning_rate": 9.973660997366101e-06, + "loss": 0.0778, + "step": 26351 + }, + { + "epoch": 12.251046025104603, + "grad_norm": 1.2360146045684814, + "learning_rate": 9.973659997366002e-06, + "loss": 0.0077, + "step": 26352 + }, + { + "epoch": 12.251510925151093, + "grad_norm": 8.107276916503906, + "learning_rate": 9.9736589973659e-06, + "loss": 0.1105, + "step": 26353 + }, + { + "epoch": 12.251975825197583, + "grad_norm": 3.8918795585632324, + "learning_rate": 9.973657997365801e-06, + "loss": 0.0457, + "step": 26354 + }, + { + "epoch": 12.252440725244073, + "grad_norm": 5.728420734405518, + "learning_rate": 9.9736569973657e-06, + "loss": 0.0448, + "step": 26355 + }, + { + "epoch": 12.252905625290563, + "grad_norm": 1.2522488832473755, + "learning_rate": 9.973655997365601e-06, + "loss": 0.0067, + "step": 26356 + }, + { + "epoch": 12.253370525337052, + "grad_norm": 2.863504648208618, + "learning_rate": 9.9736549973655e-06, + "loss": 0.0246, + "step": 26357 + }, + { + "epoch": 12.253835425383542, + "grad_norm": 2.9108338356018066, + "learning_rate": 9.9736539973654e-06, + "loss": 0.0422, + "step": 26358 + }, + { + "epoch": 12.254300325430032, + "grad_norm": 5.772616386413574, + "learning_rate": 9.973652997365301e-06, + "loss": 0.0657, + "step": 26359 + }, + { + "epoch": 12.254765225476522, + "grad_norm": 2.4401164054870605, + "learning_rate": 9.9736519973652e-06, + "loss": 0.0139, + "step": 26360 + }, + { + "epoch": 12.255230125523013, + "grad_norm": 3.176246404647827, + "learning_rate": 9.9736509973651e-06, + "loss": 0.0362, + "step": 26361 + }, + { + "epoch": 12.255695025569503, + "grad_norm": 3.9337403774261475, + "learning_rate": 9.973649997365e-06, + "loss": 0.0388, + "step": 26362 + }, + { + "epoch": 12.256159925615993, + "grad_norm": 3.493433713912964, + "learning_rate": 9.9736489973649e-06, + "loss": 0.0268, + "step": 26363 + }, + { + "epoch": 12.256624825662483, + "grad_norm": 3.269124984741211, + "learning_rate": 9.973647997364801e-06, + "loss": 0.0316, + "step": 26364 + }, + { + "epoch": 12.257089725708973, + "grad_norm": 0.49558910727500916, + "learning_rate": 9.973646997364702e-06, + "loss": 0.004, + "step": 26365 + }, + { + "epoch": 12.257554625755462, + "grad_norm": 5.267884254455566, + "learning_rate": 9.9736459973646e-06, + "loss": 0.0479, + "step": 26366 + }, + { + "epoch": 12.258019525801952, + "grad_norm": 1.0786666870117188, + "learning_rate": 9.973644997364501e-06, + "loss": 0.0097, + "step": 26367 + }, + { + "epoch": 12.258484425848442, + "grad_norm": 3.2131197452545166, + "learning_rate": 9.9736439973644e-06, + "loss": 0.0289, + "step": 26368 + }, + { + "epoch": 12.258949325894932, + "grad_norm": 6.206722736358643, + "learning_rate": 9.9736429973643e-06, + "loss": 0.0836, + "step": 26369 + }, + { + "epoch": 12.259414225941423, + "grad_norm": 2.41216778755188, + "learning_rate": 9.9736419973642e-06, + "loss": 0.0189, + "step": 26370 + }, + { + "epoch": 12.259879125987913, + "grad_norm": 4.5833821296691895, + "learning_rate": 9.9736409973641e-06, + "loss": 0.0252, + "step": 26371 + }, + { + "epoch": 12.260344026034403, + "grad_norm": 7.172004699707031, + "learning_rate": 9.973639997364001e-06, + "loss": 0.175, + "step": 26372 + }, + { + "epoch": 12.260808926080893, + "grad_norm": 3.672125816345215, + "learning_rate": 9.9736389973639e-06, + "loss": 0.0544, + "step": 26373 + }, + { + "epoch": 12.261273826127383, + "grad_norm": 6.195338249206543, + "learning_rate": 9.9736379973638e-06, + "loss": 0.0632, + "step": 26374 + }, + { + "epoch": 12.261738726173872, + "grad_norm": 8.991962432861328, + "learning_rate": 9.9736369973637e-06, + "loss": 0.0962, + "step": 26375 + }, + { + "epoch": 12.262203626220362, + "grad_norm": 4.9048333168029785, + "learning_rate": 9.9736359973636e-06, + "loss": 0.032, + "step": 26376 + }, + { + "epoch": 12.262668526266852, + "grad_norm": 4.154116153717041, + "learning_rate": 9.973634997363501e-06, + "loss": 0.0524, + "step": 26377 + }, + { + "epoch": 12.263133426313342, + "grad_norm": 3.130650043487549, + "learning_rate": 9.973633997363402e-06, + "loss": 0.029, + "step": 26378 + }, + { + "epoch": 12.263598326359833, + "grad_norm": 2.2422797679901123, + "learning_rate": 9.9736329973633e-06, + "loss": 0.0179, + "step": 26379 + }, + { + "epoch": 12.264063226406323, + "grad_norm": 6.4436750411987305, + "learning_rate": 9.9736319973632e-06, + "loss": 0.0291, + "step": 26380 + }, + { + "epoch": 12.264528126452813, + "grad_norm": 13.401341438293457, + "learning_rate": 9.9736309973631e-06, + "loss": 0.1049, + "step": 26381 + }, + { + "epoch": 12.264993026499303, + "grad_norm": 4.387688636779785, + "learning_rate": 9.973629997363001e-06, + "loss": 0.045, + "step": 26382 + }, + { + "epoch": 12.265457926545793, + "grad_norm": 5.942111015319824, + "learning_rate": 9.9736289973629e-06, + "loss": 0.049, + "step": 26383 + }, + { + "epoch": 12.265922826592282, + "grad_norm": 8.099390983581543, + "learning_rate": 9.9736279973628e-06, + "loss": 0.0741, + "step": 26384 + }, + { + "epoch": 12.266387726638772, + "grad_norm": 13.714563369750977, + "learning_rate": 9.973626997362701e-06, + "loss": 0.0971, + "step": 26385 + }, + { + "epoch": 12.266852626685262, + "grad_norm": 3.3323307037353516, + "learning_rate": 9.9736259973626e-06, + "loss": 0.033, + "step": 26386 + }, + { + "epoch": 12.267317526731752, + "grad_norm": 2.2427854537963867, + "learning_rate": 9.973624997362501e-06, + "loss": 0.0079, + "step": 26387 + }, + { + "epoch": 12.267782426778243, + "grad_norm": 1.7095917463302612, + "learning_rate": 9.9736239973624e-06, + "loss": 0.012, + "step": 26388 + }, + { + "epoch": 12.268247326824733, + "grad_norm": 1.7413389682769775, + "learning_rate": 9.9736229973623e-06, + "loss": 0.0079, + "step": 26389 + }, + { + "epoch": 12.268712226871223, + "grad_norm": 2.8733203411102295, + "learning_rate": 9.973621997362201e-06, + "loss": 0.0291, + "step": 26390 + }, + { + "epoch": 12.269177126917713, + "grad_norm": 2.5420119762420654, + "learning_rate": 9.9736209973621e-06, + "loss": 0.0181, + "step": 26391 + }, + { + "epoch": 12.269642026964203, + "grad_norm": 2.5704329013824463, + "learning_rate": 9.973619997362e-06, + "loss": 0.0235, + "step": 26392 + }, + { + "epoch": 12.270106927010692, + "grad_norm": 4.809597015380859, + "learning_rate": 9.9736189973619e-06, + "loss": 0.0612, + "step": 26393 + }, + { + "epoch": 12.270571827057182, + "grad_norm": 4.52910041809082, + "learning_rate": 9.9736179973618e-06, + "loss": 0.0675, + "step": 26394 + }, + { + "epoch": 12.271036727103672, + "grad_norm": 2.374190330505371, + "learning_rate": 9.973616997361701e-06, + "loss": 0.0118, + "step": 26395 + }, + { + "epoch": 12.271501627150164, + "grad_norm": 5.944537162780762, + "learning_rate": 9.9736159973616e-06, + "loss": 0.049, + "step": 26396 + }, + { + "epoch": 12.271966527196653, + "grad_norm": 7.8740315437316895, + "learning_rate": 9.9736149973615e-06, + "loss": 0.1349, + "step": 26397 + }, + { + "epoch": 12.272431427243143, + "grad_norm": 4.617404937744141, + "learning_rate": 9.973613997361401e-06, + "loss": 0.0656, + "step": 26398 + }, + { + "epoch": 12.272896327289633, + "grad_norm": 4.308221340179443, + "learning_rate": 9.9736129973613e-06, + "loss": 0.0631, + "step": 26399 + }, + { + "epoch": 12.273361227336123, + "grad_norm": 2.4540202617645264, + "learning_rate": 9.9736119973612e-06, + "loss": 0.0131, + "step": 26400 + }, + { + "epoch": 12.273826127382613, + "grad_norm": 4.685204982757568, + "learning_rate": 9.9736109973611e-06, + "loss": 0.0629, + "step": 26401 + }, + { + "epoch": 12.274291027429102, + "grad_norm": 2.81326961517334, + "learning_rate": 9.973609997361e-06, + "loss": 0.0272, + "step": 26402 + }, + { + "epoch": 12.274755927475592, + "grad_norm": 3.3955488204956055, + "learning_rate": 9.973608997360901e-06, + "loss": 0.0303, + "step": 26403 + }, + { + "epoch": 12.275220827522082, + "grad_norm": 1.9564075469970703, + "learning_rate": 9.9736079973608e-06, + "loss": 0.0177, + "step": 26404 + }, + { + "epoch": 12.275685727568574, + "grad_norm": 4.1270318031311035, + "learning_rate": 9.973606997360701e-06, + "loss": 0.0233, + "step": 26405 + }, + { + "epoch": 12.276150627615063, + "grad_norm": 4.0492753982543945, + "learning_rate": 9.9736059973606e-06, + "loss": 0.0407, + "step": 26406 + }, + { + "epoch": 12.276615527661553, + "grad_norm": 4.174473285675049, + "learning_rate": 9.9736049973605e-06, + "loss": 0.055, + "step": 26407 + }, + { + "epoch": 12.277080427708043, + "grad_norm": 4.821389198303223, + "learning_rate": 9.973603997360401e-06, + "loss": 0.092, + "step": 26408 + }, + { + "epoch": 12.277545327754533, + "grad_norm": 5.98284912109375, + "learning_rate": 9.973602997360302e-06, + "loss": 0.099, + "step": 26409 + }, + { + "epoch": 12.278010227801023, + "grad_norm": 5.1382622718811035, + "learning_rate": 9.9736019973602e-06, + "loss": 0.0721, + "step": 26410 + }, + { + "epoch": 12.278475127847512, + "grad_norm": 4.4983038902282715, + "learning_rate": 9.9736009973601e-06, + "loss": 0.0432, + "step": 26411 + }, + { + "epoch": 12.278940027894002, + "grad_norm": 4.887149810791016, + "learning_rate": 9.97359999736e-06, + "loss": 0.0457, + "step": 26412 + }, + { + "epoch": 12.279404927940492, + "grad_norm": 8.528799057006836, + "learning_rate": 9.9735989973599e-06, + "loss": 0.0473, + "step": 26413 + }, + { + "epoch": 12.279869827986984, + "grad_norm": 4.144416332244873, + "learning_rate": 9.9735979973598e-06, + "loss": 0.043, + "step": 26414 + }, + { + "epoch": 12.280334728033473, + "grad_norm": 5.142176628112793, + "learning_rate": 9.9735969973597e-06, + "loss": 0.0776, + "step": 26415 + }, + { + "epoch": 12.280799628079963, + "grad_norm": 2.7143425941467285, + "learning_rate": 9.973595997359601e-06, + "loss": 0.0171, + "step": 26416 + }, + { + "epoch": 12.281264528126453, + "grad_norm": 1.768701434135437, + "learning_rate": 9.9735949973595e-06, + "loss": 0.014, + "step": 26417 + }, + { + "epoch": 12.281729428172943, + "grad_norm": 5.787397861480713, + "learning_rate": 9.973593997359401e-06, + "loss": 0.0794, + "step": 26418 + }, + { + "epoch": 12.282194328219433, + "grad_norm": 4.027562618255615, + "learning_rate": 9.9735929973593e-06, + "loss": 0.0566, + "step": 26419 + }, + { + "epoch": 12.282659228265922, + "grad_norm": 2.482100009918213, + "learning_rate": 9.9735919973592e-06, + "loss": 0.0249, + "step": 26420 + }, + { + "epoch": 12.283124128312412, + "grad_norm": 3.3056893348693848, + "learning_rate": 9.973590997359101e-06, + "loss": 0.0392, + "step": 26421 + }, + { + "epoch": 12.283589028358902, + "grad_norm": 5.538564205169678, + "learning_rate": 9.973589997359e-06, + "loss": 0.0748, + "step": 26422 + }, + { + "epoch": 12.284053928405394, + "grad_norm": 5.136740684509277, + "learning_rate": 9.973588997358901e-06, + "loss": 0.111, + "step": 26423 + }, + { + "epoch": 12.284518828451883, + "grad_norm": 3.5912153720855713, + "learning_rate": 9.9735879973588e-06, + "loss": 0.0377, + "step": 26424 + }, + { + "epoch": 12.284983728498373, + "grad_norm": 4.020504474639893, + "learning_rate": 9.9735869973587e-06, + "loss": 0.0305, + "step": 26425 + }, + { + "epoch": 12.285448628544863, + "grad_norm": 3.3020293712615967, + "learning_rate": 9.9735859973586e-06, + "loss": 0.0357, + "step": 26426 + }, + { + "epoch": 12.285913528591353, + "grad_norm": 1.76781165599823, + "learning_rate": 9.9735849973585e-06, + "loss": 0.0167, + "step": 26427 + }, + { + "epoch": 12.286378428637843, + "grad_norm": 0.8807920217514038, + "learning_rate": 9.9735839973584e-06, + "loss": 0.0057, + "step": 26428 + }, + { + "epoch": 12.286843328684332, + "grad_norm": 4.87891149520874, + "learning_rate": 9.973582997358301e-06, + "loss": 0.0718, + "step": 26429 + }, + { + "epoch": 12.287308228730822, + "grad_norm": 3.857936382293701, + "learning_rate": 9.9735819973582e-06, + "loss": 0.0328, + "step": 26430 + }, + { + "epoch": 12.287773128777314, + "grad_norm": 2.2582051753997803, + "learning_rate": 9.973580997358101e-06, + "loss": 0.0116, + "step": 26431 + }, + { + "epoch": 12.288238028823804, + "grad_norm": 4.55913782119751, + "learning_rate": 9.973579997358e-06, + "loss": 0.034, + "step": 26432 + }, + { + "epoch": 12.288702928870293, + "grad_norm": 2.9110519886016846, + "learning_rate": 9.9735789973579e-06, + "loss": 0.0255, + "step": 26433 + }, + { + "epoch": 12.289167828916783, + "grad_norm": 1.3095864057540894, + "learning_rate": 9.973577997357801e-06, + "loss": 0.0092, + "step": 26434 + }, + { + "epoch": 12.289632728963273, + "grad_norm": 6.81386661529541, + "learning_rate": 9.9735769973577e-06, + "loss": 0.1351, + "step": 26435 + }, + { + "epoch": 12.290097629009763, + "grad_norm": 4.710999965667725, + "learning_rate": 9.973575997357601e-06, + "loss": 0.0766, + "step": 26436 + }, + { + "epoch": 12.290562529056253, + "grad_norm": 4.083093166351318, + "learning_rate": 9.9735749973575e-06, + "loss": 0.0321, + "step": 26437 + }, + { + "epoch": 12.291027429102742, + "grad_norm": 7.025228500366211, + "learning_rate": 9.9735739973574e-06, + "loss": 0.1174, + "step": 26438 + }, + { + "epoch": 12.291492329149232, + "grad_norm": 1.9383976459503174, + "learning_rate": 9.9735729973573e-06, + "loss": 0.0143, + "step": 26439 + }, + { + "epoch": 12.291957229195724, + "grad_norm": 3.436126470565796, + "learning_rate": 9.9735719973572e-06, + "loss": 0.0398, + "step": 26440 + }, + { + "epoch": 12.292422129242214, + "grad_norm": 1.8886123895645142, + "learning_rate": 9.973570997357101e-06, + "loss": 0.0104, + "step": 26441 + }, + { + "epoch": 12.292887029288703, + "grad_norm": 2.7886016368865967, + "learning_rate": 9.973569997357002e-06, + "loss": 0.0289, + "step": 26442 + }, + { + "epoch": 12.293351929335193, + "grad_norm": 2.4880895614624023, + "learning_rate": 9.9735689973569e-06, + "loss": 0.016, + "step": 26443 + }, + { + "epoch": 12.293816829381683, + "grad_norm": 4.832181930541992, + "learning_rate": 9.9735679973568e-06, + "loss": 0.0405, + "step": 26444 + }, + { + "epoch": 12.294281729428173, + "grad_norm": 3.365216016769409, + "learning_rate": 9.9735669973567e-06, + "loss": 0.0119, + "step": 26445 + }, + { + "epoch": 12.294746629474663, + "grad_norm": 4.4033894538879395, + "learning_rate": 9.9735659973566e-06, + "loss": 0.0592, + "step": 26446 + }, + { + "epoch": 12.295211529521152, + "grad_norm": 7.415870666503906, + "learning_rate": 9.973564997356501e-06, + "loss": 0.2015, + "step": 26447 + }, + { + "epoch": 12.295676429567642, + "grad_norm": 2.7359611988067627, + "learning_rate": 9.9735639973564e-06, + "loss": 0.0274, + "step": 26448 + }, + { + "epoch": 12.296141329614134, + "grad_norm": 3.12933349609375, + "learning_rate": 9.973562997356301e-06, + "loss": 0.0385, + "step": 26449 + }, + { + "epoch": 12.296606229660624, + "grad_norm": 5.2124247550964355, + "learning_rate": 9.9735619973562e-06, + "loss": 0.0591, + "step": 26450 + }, + { + "epoch": 12.297071129707113, + "grad_norm": 3.9019384384155273, + "learning_rate": 9.9735609973561e-06, + "loss": 0.0408, + "step": 26451 + }, + { + "epoch": 12.297536029753603, + "grad_norm": 6.949101448059082, + "learning_rate": 9.973559997356e-06, + "loss": 0.1116, + "step": 26452 + }, + { + "epoch": 12.298000929800093, + "grad_norm": 6.000412464141846, + "learning_rate": 9.9735589973559e-06, + "loss": 0.0651, + "step": 26453 + }, + { + "epoch": 12.298465829846583, + "grad_norm": 4.958848476409912, + "learning_rate": 9.973557997355801e-06, + "loss": 0.0924, + "step": 26454 + }, + { + "epoch": 12.298930729893073, + "grad_norm": 2.474278450012207, + "learning_rate": 9.9735569973557e-06, + "loss": 0.0219, + "step": 26455 + }, + { + "epoch": 12.299395629939562, + "grad_norm": 7.550038814544678, + "learning_rate": 9.9735559973556e-06, + "loss": 0.0648, + "step": 26456 + }, + { + "epoch": 12.299860529986052, + "grad_norm": 5.003562927246094, + "learning_rate": 9.9735549973555e-06, + "loss": 0.0509, + "step": 26457 + }, + { + "epoch": 12.300325430032544, + "grad_norm": 3.088212013244629, + "learning_rate": 9.9735539973554e-06, + "loss": 0.023, + "step": 26458 + }, + { + "epoch": 12.300790330079034, + "grad_norm": 2.860360622406006, + "learning_rate": 9.973552997355301e-06, + "loss": 0.0365, + "step": 26459 + }, + { + "epoch": 12.301255230125523, + "grad_norm": 2.1905601024627686, + "learning_rate": 9.973551997355202e-06, + "loss": 0.0143, + "step": 26460 + }, + { + "epoch": 12.301720130172013, + "grad_norm": 5.101428985595703, + "learning_rate": 9.9735509973551e-06, + "loss": 0.0708, + "step": 26461 + }, + { + "epoch": 12.302185030218503, + "grad_norm": 4.646651268005371, + "learning_rate": 9.973549997355001e-06, + "loss": 0.0734, + "step": 26462 + }, + { + "epoch": 12.302649930264993, + "grad_norm": 7.086675643920898, + "learning_rate": 9.9735489973549e-06, + "loss": 0.0375, + "step": 26463 + }, + { + "epoch": 12.303114830311483, + "grad_norm": 4.421750068664551, + "learning_rate": 9.9735479973548e-06, + "loss": 0.056, + "step": 26464 + }, + { + "epoch": 12.303579730357972, + "grad_norm": 2.2123258113861084, + "learning_rate": 9.9735469973547e-06, + "loss": 0.0186, + "step": 26465 + }, + { + "epoch": 12.304044630404462, + "grad_norm": 4.849829196929932, + "learning_rate": 9.9735459973546e-06, + "loss": 0.0531, + "step": 26466 + }, + { + "epoch": 12.304509530450954, + "grad_norm": 6.040436744689941, + "learning_rate": 9.973544997354501e-06, + "loss": 0.115, + "step": 26467 + }, + { + "epoch": 12.304974430497444, + "grad_norm": 6.652692794799805, + "learning_rate": 9.9735439973544e-06, + "loss": 0.0751, + "step": 26468 + }, + { + "epoch": 12.305439330543933, + "grad_norm": 2.7957987785339355, + "learning_rate": 9.9735429973543e-06, + "loss": 0.0182, + "step": 26469 + }, + { + "epoch": 12.305904230590423, + "grad_norm": 2.9395647048950195, + "learning_rate": 9.9735419973542e-06, + "loss": 0.0221, + "step": 26470 + }, + { + "epoch": 12.306369130636913, + "grad_norm": 2.5286693572998047, + "learning_rate": 9.9735409973541e-06, + "loss": 0.018, + "step": 26471 + }, + { + "epoch": 12.306834030683403, + "grad_norm": 7.843503952026367, + "learning_rate": 9.973539997354001e-06, + "loss": 0.0772, + "step": 26472 + }, + { + "epoch": 12.307298930729893, + "grad_norm": 3.6504392623901367, + "learning_rate": 9.973538997353902e-06, + "loss": 0.0584, + "step": 26473 + }, + { + "epoch": 12.307763830776382, + "grad_norm": 3.918109178543091, + "learning_rate": 9.9735379973538e-06, + "loss": 0.0531, + "step": 26474 + }, + { + "epoch": 12.308228730822874, + "grad_norm": 2.978410243988037, + "learning_rate": 9.973536997353701e-06, + "loss": 0.0307, + "step": 26475 + }, + { + "epoch": 12.308693630869364, + "grad_norm": 8.475515365600586, + "learning_rate": 9.9735359973536e-06, + "loss": 0.0367, + "step": 26476 + }, + { + "epoch": 12.309158530915854, + "grad_norm": 6.031950950622559, + "learning_rate": 9.9735349973535e-06, + "loss": 0.1055, + "step": 26477 + }, + { + "epoch": 12.309623430962343, + "grad_norm": 4.082828998565674, + "learning_rate": 9.973533997353401e-06, + "loss": 0.0402, + "step": 26478 + }, + { + "epoch": 12.310088331008833, + "grad_norm": 2.4455854892730713, + "learning_rate": 9.9735329973533e-06, + "loss": 0.0216, + "step": 26479 + }, + { + "epoch": 12.310553231055323, + "grad_norm": 5.2360920906066895, + "learning_rate": 9.973531997353201e-06, + "loss": 0.0666, + "step": 26480 + }, + { + "epoch": 12.311018131101813, + "grad_norm": 2.557036876678467, + "learning_rate": 9.9735309973531e-06, + "loss": 0.0315, + "step": 26481 + }, + { + "epoch": 12.311483031148303, + "grad_norm": 1.4474101066589355, + "learning_rate": 9.973529997353e-06, + "loss": 0.0108, + "step": 26482 + }, + { + "epoch": 12.311947931194792, + "grad_norm": 0.3819139003753662, + "learning_rate": 9.9735289973529e-06, + "loss": 0.0017, + "step": 26483 + }, + { + "epoch": 12.312412831241284, + "grad_norm": 4.2759528160095215, + "learning_rate": 9.9735279973528e-06, + "loss": 0.0526, + "step": 26484 + }, + { + "epoch": 12.312877731287774, + "grad_norm": 3.5423591136932373, + "learning_rate": 9.973526997352701e-06, + "loss": 0.0604, + "step": 26485 + }, + { + "epoch": 12.313342631334264, + "grad_norm": 4.316055774688721, + "learning_rate": 9.973525997352602e-06, + "loss": 0.0542, + "step": 26486 + }, + { + "epoch": 12.313807531380753, + "grad_norm": 5.765012741088867, + "learning_rate": 9.9735249973525e-06, + "loss": 0.0557, + "step": 26487 + }, + { + "epoch": 12.314272431427243, + "grad_norm": 5.086000442504883, + "learning_rate": 9.9735239973524e-06, + "loss": 0.0736, + "step": 26488 + }, + { + "epoch": 12.314737331473733, + "grad_norm": 4.022983551025391, + "learning_rate": 9.9735229973523e-06, + "loss": 0.0418, + "step": 26489 + }, + { + "epoch": 12.315202231520223, + "grad_norm": 6.826007843017578, + "learning_rate": 9.973521997352201e-06, + "loss": 0.0862, + "step": 26490 + }, + { + "epoch": 12.315667131566713, + "grad_norm": 2.6089305877685547, + "learning_rate": 9.973520997352102e-06, + "loss": 0.0215, + "step": 26491 + }, + { + "epoch": 12.316132031613202, + "grad_norm": 3.3901712894439697, + "learning_rate": 9.973519997352e-06, + "loss": 0.0245, + "step": 26492 + }, + { + "epoch": 12.316596931659694, + "grad_norm": 3.05891489982605, + "learning_rate": 9.973518997351901e-06, + "loss": 0.0206, + "step": 26493 + }, + { + "epoch": 12.317061831706184, + "grad_norm": 7.100647449493408, + "learning_rate": 9.9735179973518e-06, + "loss": 0.0369, + "step": 26494 + }, + { + "epoch": 12.317526731752674, + "grad_norm": 5.150270462036133, + "learning_rate": 9.9735169973517e-06, + "loss": 0.0782, + "step": 26495 + }, + { + "epoch": 12.317991631799163, + "grad_norm": 3.4675939083099365, + "learning_rate": 9.9735159973516e-06, + "loss": 0.0329, + "step": 26496 + }, + { + "epoch": 12.318456531845653, + "grad_norm": 3.1345551013946533, + "learning_rate": 9.9735149973515e-06, + "loss": 0.0304, + "step": 26497 + }, + { + "epoch": 12.318921431892143, + "grad_norm": 11.923628807067871, + "learning_rate": 9.973513997351401e-06, + "loss": 0.179, + "step": 26498 + }, + { + "epoch": 12.319386331938633, + "grad_norm": 8.354647636413574, + "learning_rate": 9.9735129973513e-06, + "loss": 0.034, + "step": 26499 + }, + { + "epoch": 12.319851231985123, + "grad_norm": 3.6336829662323, + "learning_rate": 9.9735119973512e-06, + "loss": 0.0375, + "step": 26500 + }, + { + "epoch": 12.320316132031612, + "grad_norm": 12.72875690460205, + "learning_rate": 9.9735109973511e-06, + "loss": 0.0537, + "step": 26501 + }, + { + "epoch": 12.320781032078104, + "grad_norm": 2.516263246536255, + "learning_rate": 9.973509997351e-06, + "loss": 0.0376, + "step": 26502 + }, + { + "epoch": 12.321245932124594, + "grad_norm": 4.693097114562988, + "learning_rate": 9.973508997350901e-06, + "loss": 0.0665, + "step": 26503 + }, + { + "epoch": 12.321710832171084, + "grad_norm": 3.7189176082611084, + "learning_rate": 9.973507997350802e-06, + "loss": 0.0264, + "step": 26504 + }, + { + "epoch": 12.322175732217573, + "grad_norm": 3.478605270385742, + "learning_rate": 9.9735069973507e-06, + "loss": 0.025, + "step": 26505 + }, + { + "epoch": 12.322640632264063, + "grad_norm": 4.183009624481201, + "learning_rate": 9.973505997350601e-06, + "loss": 0.0292, + "step": 26506 + }, + { + "epoch": 12.323105532310553, + "grad_norm": 0.7561227679252625, + "learning_rate": 9.9735049973505e-06, + "loss": 0.006, + "step": 26507 + }, + { + "epoch": 12.323570432357043, + "grad_norm": 1.3001327514648438, + "learning_rate": 9.973503997350401e-06, + "loss": 0.0108, + "step": 26508 + }, + { + "epoch": 12.324035332403533, + "grad_norm": 7.365129470825195, + "learning_rate": 9.9735029973503e-06, + "loss": 0.0921, + "step": 26509 + }, + { + "epoch": 12.324500232450024, + "grad_norm": 4.678394317626953, + "learning_rate": 9.9735019973502e-06, + "loss": 0.0624, + "step": 26510 + }, + { + "epoch": 12.324965132496514, + "grad_norm": 3.482764720916748, + "learning_rate": 9.973500997350101e-06, + "loss": 0.0213, + "step": 26511 + }, + { + "epoch": 12.325430032543004, + "grad_norm": 4.093615531921387, + "learning_rate": 9.97349999735e-06, + "loss": 0.0295, + "step": 26512 + }, + { + "epoch": 12.325894932589494, + "grad_norm": 7.3275322914123535, + "learning_rate": 9.9734989973499e-06, + "loss": 0.1026, + "step": 26513 + }, + { + "epoch": 12.326359832635983, + "grad_norm": 3.0292532444000244, + "learning_rate": 9.9734979973498e-06, + "loss": 0.0213, + "step": 26514 + }, + { + "epoch": 12.326824732682473, + "grad_norm": 1.5403294563293457, + "learning_rate": 9.9734969973497e-06, + "loss": 0.0107, + "step": 26515 + }, + { + "epoch": 12.327289632728963, + "grad_norm": 5.058963775634766, + "learning_rate": 9.973495997349601e-06, + "loss": 0.0587, + "step": 26516 + }, + { + "epoch": 12.327754532775453, + "grad_norm": 4.488790035247803, + "learning_rate": 9.973494997349502e-06, + "loss": 0.059, + "step": 26517 + }, + { + "epoch": 12.328219432821943, + "grad_norm": 0.5944567322731018, + "learning_rate": 9.9734939973494e-06, + "loss": 0.0059, + "step": 26518 + }, + { + "epoch": 12.328684332868434, + "grad_norm": 1.0474085807800293, + "learning_rate": 9.973492997349301e-06, + "loss": 0.0059, + "step": 26519 + }, + { + "epoch": 12.329149232914924, + "grad_norm": 4.727181911468506, + "learning_rate": 9.9734919973492e-06, + "loss": 0.0476, + "step": 26520 + }, + { + "epoch": 12.329614132961414, + "grad_norm": 1.4843250513076782, + "learning_rate": 9.9734909973491e-06, + "loss": 0.0089, + "step": 26521 + }, + { + "epoch": 12.330079033007904, + "grad_norm": 5.626049041748047, + "learning_rate": 9.973489997349e-06, + "loss": 0.051, + "step": 26522 + }, + { + "epoch": 12.330543933054393, + "grad_norm": 4.985930442810059, + "learning_rate": 9.9734889973489e-06, + "loss": 0.0354, + "step": 26523 + }, + { + "epoch": 12.331008833100883, + "grad_norm": 6.473629474639893, + "learning_rate": 9.973487997348801e-06, + "loss": 0.0945, + "step": 26524 + }, + { + "epoch": 12.331473733147373, + "grad_norm": 0.9382702708244324, + "learning_rate": 9.9734869973487e-06, + "loss": 0.0056, + "step": 26525 + }, + { + "epoch": 12.331938633193863, + "grad_norm": 4.819591522216797, + "learning_rate": 9.973485997348601e-06, + "loss": 0.078, + "step": 26526 + }, + { + "epoch": 12.332403533240353, + "grad_norm": 5.9409918785095215, + "learning_rate": 9.9734849973485e-06, + "loss": 0.0406, + "step": 26527 + }, + { + "epoch": 12.332868433286844, + "grad_norm": 0.9362910985946655, + "learning_rate": 9.9734839973484e-06, + "loss": 0.0059, + "step": 26528 + }, + { + "epoch": 12.333333333333334, + "grad_norm": 4.4044342041015625, + "learning_rate": 9.973482997348301e-06, + "loss": 0.0475, + "step": 26529 + }, + { + "epoch": 12.333798233379824, + "grad_norm": 1.7928558588027954, + "learning_rate": 9.973481997348202e-06, + "loss": 0.0151, + "step": 26530 + }, + { + "epoch": 12.334263133426314, + "grad_norm": 6.2968645095825195, + "learning_rate": 9.9734809973481e-06, + "loss": 0.0525, + "step": 26531 + }, + { + "epoch": 12.334728033472803, + "grad_norm": 0.8816251754760742, + "learning_rate": 9.973479997348e-06, + "loss": 0.0067, + "step": 26532 + }, + { + "epoch": 12.335192933519293, + "grad_norm": 4.403395175933838, + "learning_rate": 9.9734789973479e-06, + "loss": 0.0455, + "step": 26533 + }, + { + "epoch": 12.335657833565783, + "grad_norm": 8.056838035583496, + "learning_rate": 9.9734779973478e-06, + "loss": 0.0571, + "step": 26534 + }, + { + "epoch": 12.336122733612273, + "grad_norm": 3.615495443344116, + "learning_rate": 9.9734769973477e-06, + "loss": 0.0473, + "step": 26535 + }, + { + "epoch": 12.336587633658763, + "grad_norm": 4.634615421295166, + "learning_rate": 9.9734759973476e-06, + "loss": 0.0656, + "step": 26536 + }, + { + "epoch": 12.337052533705254, + "grad_norm": 5.398848533630371, + "learning_rate": 9.973474997347501e-06, + "loss": 0.0642, + "step": 26537 + }, + { + "epoch": 12.337517433751744, + "grad_norm": 6.0743560791015625, + "learning_rate": 9.9734739973474e-06, + "loss": 0.0588, + "step": 26538 + }, + { + "epoch": 12.337982333798234, + "grad_norm": 6.736973285675049, + "learning_rate": 9.973472997347301e-06, + "loss": 0.0957, + "step": 26539 + }, + { + "epoch": 12.338447233844724, + "grad_norm": 3.0885143280029297, + "learning_rate": 9.9734719973472e-06, + "loss": 0.0295, + "step": 26540 + }, + { + "epoch": 12.338912133891213, + "grad_norm": 7.4806952476501465, + "learning_rate": 9.9734709973471e-06, + "loss": 0.0802, + "step": 26541 + }, + { + "epoch": 12.339377033937703, + "grad_norm": 4.762510776519775, + "learning_rate": 9.973469997347001e-06, + "loss": 0.0994, + "step": 26542 + }, + { + "epoch": 12.339841933984193, + "grad_norm": 4.150334358215332, + "learning_rate": 9.9734689973469e-06, + "loss": 0.0456, + "step": 26543 + }, + { + "epoch": 12.340306834030683, + "grad_norm": 8.1846284866333, + "learning_rate": 9.973467997346801e-06, + "loss": 0.0986, + "step": 26544 + }, + { + "epoch": 12.340771734077173, + "grad_norm": 3.8115108013153076, + "learning_rate": 9.9734669973467e-06, + "loss": 0.0668, + "step": 26545 + }, + { + "epoch": 12.341236634123664, + "grad_norm": 2.9442272186279297, + "learning_rate": 9.9734659973466e-06, + "loss": 0.0135, + "step": 26546 + }, + { + "epoch": 12.341701534170154, + "grad_norm": 5.357098579406738, + "learning_rate": 9.973464997346501e-06, + "loss": 0.0671, + "step": 26547 + }, + { + "epoch": 12.342166434216644, + "grad_norm": 5.567703723907471, + "learning_rate": 9.9734639973464e-06, + "loss": 0.0837, + "step": 26548 + }, + { + "epoch": 12.342631334263134, + "grad_norm": 2.462007761001587, + "learning_rate": 9.9734629973463e-06, + "loss": 0.0156, + "step": 26549 + }, + { + "epoch": 12.343096234309623, + "grad_norm": 5.750202178955078, + "learning_rate": 9.973461997346201e-06, + "loss": 0.1117, + "step": 26550 + }, + { + "epoch": 12.343561134356113, + "grad_norm": 4.533182144165039, + "learning_rate": 9.9734609973461e-06, + "loss": 0.0511, + "step": 26551 + }, + { + "epoch": 12.344026034402603, + "grad_norm": 6.9390339851379395, + "learning_rate": 9.973459997346e-06, + "loss": 0.1244, + "step": 26552 + }, + { + "epoch": 12.344490934449093, + "grad_norm": 7.624881267547607, + "learning_rate": 9.9734589973459e-06, + "loss": 0.095, + "step": 26553 + }, + { + "epoch": 12.344955834495583, + "grad_norm": 3.5646469593048096, + "learning_rate": 9.9734579973458e-06, + "loss": 0.0244, + "step": 26554 + }, + { + "epoch": 12.345420734542074, + "grad_norm": 4.26849365234375, + "learning_rate": 9.973456997345701e-06, + "loss": 0.066, + "step": 26555 + }, + { + "epoch": 12.345885634588564, + "grad_norm": 3.609064817428589, + "learning_rate": 9.9734559973456e-06, + "loss": 0.0222, + "step": 26556 + }, + { + "epoch": 12.346350534635054, + "grad_norm": 2.6699130535125732, + "learning_rate": 9.973454997345501e-06, + "loss": 0.0203, + "step": 26557 + }, + { + "epoch": 12.346815434681544, + "grad_norm": 3.8443329334259033, + "learning_rate": 9.9734539973454e-06, + "loss": 0.0386, + "step": 26558 + }, + { + "epoch": 12.347280334728033, + "grad_norm": 4.434258460998535, + "learning_rate": 9.9734529973453e-06, + "loss": 0.0517, + "step": 26559 + }, + { + "epoch": 12.347745234774523, + "grad_norm": 3.470048427581787, + "learning_rate": 9.973451997345201e-06, + "loss": 0.0242, + "step": 26560 + }, + { + "epoch": 12.348210134821013, + "grad_norm": 2.3879401683807373, + "learning_rate": 9.9734509973451e-06, + "loss": 0.03, + "step": 26561 + }, + { + "epoch": 12.348675034867503, + "grad_norm": 3.7948737144470215, + "learning_rate": 9.973449997345e-06, + "loss": 0.0478, + "step": 26562 + }, + { + "epoch": 12.349139934913994, + "grad_norm": 3.3528740406036377, + "learning_rate": 9.9734489973449e-06, + "loss": 0.0179, + "step": 26563 + }, + { + "epoch": 12.349604834960484, + "grad_norm": 6.670353412628174, + "learning_rate": 9.9734479973448e-06, + "loss": 0.1136, + "step": 26564 + }, + { + "epoch": 12.350069735006974, + "grad_norm": 5.0709757804870605, + "learning_rate": 9.9734469973447e-06, + "loss": 0.0545, + "step": 26565 + }, + { + "epoch": 12.350534635053464, + "grad_norm": 8.474944114685059, + "learning_rate": 9.9734459973446e-06, + "loss": 0.0805, + "step": 26566 + }, + { + "epoch": 12.350999535099954, + "grad_norm": 4.828309059143066, + "learning_rate": 9.9734449973445e-06, + "loss": 0.0737, + "step": 26567 + }, + { + "epoch": 12.351464435146443, + "grad_norm": 6.414262294769287, + "learning_rate": 9.973443997344401e-06, + "loss": 0.0379, + "step": 26568 + }, + { + "epoch": 12.351929335192933, + "grad_norm": 1.721523642539978, + "learning_rate": 9.9734429973443e-06, + "loss": 0.0218, + "step": 26569 + }, + { + "epoch": 12.352394235239423, + "grad_norm": 2.5902669429779053, + "learning_rate": 9.973441997344201e-06, + "loss": 0.0176, + "step": 26570 + }, + { + "epoch": 12.352859135285913, + "grad_norm": 2.626447916030884, + "learning_rate": 9.9734409973441e-06, + "loss": 0.0202, + "step": 26571 + }, + { + "epoch": 12.353324035332404, + "grad_norm": 4.817467212677002, + "learning_rate": 9.973439997344e-06, + "loss": 0.031, + "step": 26572 + }, + { + "epoch": 12.353788935378894, + "grad_norm": 3.419511556625366, + "learning_rate": 9.973438997343901e-06, + "loss": 0.0286, + "step": 26573 + }, + { + "epoch": 12.354253835425384, + "grad_norm": 8.003761291503906, + "learning_rate": 9.9734379973438e-06, + "loss": 0.1329, + "step": 26574 + }, + { + "epoch": 12.354718735471874, + "grad_norm": 3.9311277866363525, + "learning_rate": 9.973436997343701e-06, + "loss": 0.0378, + "step": 26575 + }, + { + "epoch": 12.355183635518364, + "grad_norm": 3.2976765632629395, + "learning_rate": 9.9734359973436e-06, + "loss": 0.0237, + "step": 26576 + }, + { + "epoch": 12.355648535564853, + "grad_norm": 2.942056655883789, + "learning_rate": 9.9734349973435e-06, + "loss": 0.0378, + "step": 26577 + }, + { + "epoch": 12.356113435611343, + "grad_norm": 1.595593810081482, + "learning_rate": 9.9734339973434e-06, + "loss": 0.0129, + "step": 26578 + }, + { + "epoch": 12.356578335657833, + "grad_norm": 1.7885842323303223, + "learning_rate": 9.9734329973433e-06, + "loss": 0.0112, + "step": 26579 + }, + { + "epoch": 12.357043235704323, + "grad_norm": 4.18768835067749, + "learning_rate": 9.9734319973432e-06, + "loss": 0.0506, + "step": 26580 + }, + { + "epoch": 12.357508135750814, + "grad_norm": 3.0884976387023926, + "learning_rate": 9.973430997343101e-06, + "loss": 0.029, + "step": 26581 + }, + { + "epoch": 12.357973035797304, + "grad_norm": 7.9459967613220215, + "learning_rate": 9.973429997343e-06, + "loss": 0.1126, + "step": 26582 + }, + { + "epoch": 12.358437935843794, + "grad_norm": 2.9306857585906982, + "learning_rate": 9.973428997342901e-06, + "loss": 0.0245, + "step": 26583 + }, + { + "epoch": 12.358902835890284, + "grad_norm": 4.058760643005371, + "learning_rate": 9.9734279973428e-06, + "loss": 0.0383, + "step": 26584 + }, + { + "epoch": 12.359367735936774, + "grad_norm": 12.368721961975098, + "learning_rate": 9.9734269973427e-06, + "loss": 0.0921, + "step": 26585 + }, + { + "epoch": 12.359832635983263, + "grad_norm": 10.946958541870117, + "learning_rate": 9.973425997342601e-06, + "loss": 0.0657, + "step": 26586 + }, + { + "epoch": 12.360297536029753, + "grad_norm": 3.939789056777954, + "learning_rate": 9.9734249973425e-06, + "loss": 0.0349, + "step": 26587 + }, + { + "epoch": 12.360762436076243, + "grad_norm": 4.958483695983887, + "learning_rate": 9.973423997342401e-06, + "loss": 0.0447, + "step": 26588 + }, + { + "epoch": 12.361227336122734, + "grad_norm": 4.969325542449951, + "learning_rate": 9.9734229973423e-06, + "loss": 0.0401, + "step": 26589 + }, + { + "epoch": 12.361692236169224, + "grad_norm": 2.0926434993743896, + "learning_rate": 9.9734219973422e-06, + "loss": 0.012, + "step": 26590 + }, + { + "epoch": 12.362157136215714, + "grad_norm": 3.5504658222198486, + "learning_rate": 9.9734209973421e-06, + "loss": 0.0398, + "step": 26591 + }, + { + "epoch": 12.362622036262204, + "grad_norm": 5.06921911239624, + "learning_rate": 9.973419997342e-06, + "loss": 0.0476, + "step": 26592 + }, + { + "epoch": 12.363086936308694, + "grad_norm": 3.3776557445526123, + "learning_rate": 9.973418997341901e-06, + "loss": 0.026, + "step": 26593 + }, + { + "epoch": 12.363551836355184, + "grad_norm": 2.9820616245269775, + "learning_rate": 9.973417997341802e-06, + "loss": 0.019, + "step": 26594 + }, + { + "epoch": 12.364016736401673, + "grad_norm": 2.5895206928253174, + "learning_rate": 9.9734169973417e-06, + "loss": 0.0244, + "step": 26595 + }, + { + "epoch": 12.364481636448163, + "grad_norm": 4.9488019943237305, + "learning_rate": 9.9734159973416e-06, + "loss": 0.1061, + "step": 26596 + }, + { + "epoch": 12.364946536494653, + "grad_norm": 2.5622873306274414, + "learning_rate": 9.9734149973415e-06, + "loss": 0.0246, + "step": 26597 + }, + { + "epoch": 12.365411436541144, + "grad_norm": 2.8840525150299072, + "learning_rate": 9.9734139973414e-06, + "loss": 0.0174, + "step": 26598 + }, + { + "epoch": 12.365876336587634, + "grad_norm": 3.4999656677246094, + "learning_rate": 9.973412997341301e-06, + "loss": 0.0414, + "step": 26599 + }, + { + "epoch": 12.366341236634124, + "grad_norm": 2.2970588207244873, + "learning_rate": 9.9734119973412e-06, + "loss": 0.027, + "step": 26600 + }, + { + "epoch": 12.366806136680614, + "grad_norm": 2.452639102935791, + "learning_rate": 9.973410997341101e-06, + "loss": 0.0148, + "step": 26601 + }, + { + "epoch": 12.367271036727104, + "grad_norm": 18.848419189453125, + "learning_rate": 9.973409997341e-06, + "loss": 0.2288, + "step": 26602 + }, + { + "epoch": 12.367735936773594, + "grad_norm": 3.56526517868042, + "learning_rate": 9.9734089973409e-06, + "loss": 0.0216, + "step": 26603 + }, + { + "epoch": 12.368200836820083, + "grad_norm": 2.0590438842773438, + "learning_rate": 9.9734079973408e-06, + "loss": 0.0156, + "step": 26604 + }, + { + "epoch": 12.368665736866573, + "grad_norm": 10.195674896240234, + "learning_rate": 9.9734069973407e-06, + "loss": 0.1064, + "step": 26605 + }, + { + "epoch": 12.369130636913063, + "grad_norm": 5.067331314086914, + "learning_rate": 9.973405997340601e-06, + "loss": 0.0421, + "step": 26606 + }, + { + "epoch": 12.369595536959554, + "grad_norm": 2.7943999767303467, + "learning_rate": 9.9734049973405e-06, + "loss": 0.0237, + "step": 26607 + }, + { + "epoch": 12.370060437006044, + "grad_norm": 4.56234884262085, + "learning_rate": 9.9734039973404e-06, + "loss": 0.0605, + "step": 26608 + }, + { + "epoch": 12.370525337052534, + "grad_norm": 3.5953898429870605, + "learning_rate": 9.9734029973403e-06, + "loss": 0.039, + "step": 26609 + }, + { + "epoch": 12.370990237099024, + "grad_norm": 2.6534202098846436, + "learning_rate": 9.9734019973402e-06, + "loss": 0.0292, + "step": 26610 + }, + { + "epoch": 12.371455137145514, + "grad_norm": 6.563655376434326, + "learning_rate": 9.973400997340101e-06, + "loss": 0.0355, + "step": 26611 + }, + { + "epoch": 12.371920037192004, + "grad_norm": 3.393991708755493, + "learning_rate": 9.973399997340002e-06, + "loss": 0.0454, + "step": 26612 + }, + { + "epoch": 12.372384937238493, + "grad_norm": 5.387018203735352, + "learning_rate": 9.9733989973399e-06, + "loss": 0.093, + "step": 26613 + }, + { + "epoch": 12.372849837284983, + "grad_norm": 4.134897708892822, + "learning_rate": 9.973397997339801e-06, + "loss": 0.0473, + "step": 26614 + }, + { + "epoch": 12.373314737331473, + "grad_norm": 4.234012126922607, + "learning_rate": 9.9733969973397e-06, + "loss": 0.0534, + "step": 26615 + }, + { + "epoch": 12.373779637377964, + "grad_norm": 6.30555534362793, + "learning_rate": 9.9733959973396e-06, + "loss": 0.0459, + "step": 26616 + }, + { + "epoch": 12.374244537424454, + "grad_norm": 1.527736783027649, + "learning_rate": 9.9733949973395e-06, + "loss": 0.009, + "step": 26617 + }, + { + "epoch": 12.374709437470944, + "grad_norm": 1.7328630685806274, + "learning_rate": 9.9733939973394e-06, + "loss": 0.0132, + "step": 26618 + }, + { + "epoch": 12.375174337517434, + "grad_norm": 6.639041423797607, + "learning_rate": 9.973392997339301e-06, + "loss": 0.0825, + "step": 26619 + }, + { + "epoch": 12.375639237563924, + "grad_norm": 2.0010156631469727, + "learning_rate": 9.9733919973392e-06, + "loss": 0.014, + "step": 26620 + }, + { + "epoch": 12.376104137610414, + "grad_norm": 5.376948833465576, + "learning_rate": 9.9733909973391e-06, + "loss": 0.0606, + "step": 26621 + }, + { + "epoch": 12.376569037656903, + "grad_norm": 4.417965888977051, + "learning_rate": 9.973389997339e-06, + "loss": 0.0694, + "step": 26622 + }, + { + "epoch": 12.377033937703393, + "grad_norm": 3.1213674545288086, + "learning_rate": 9.9733889973389e-06, + "loss": 0.0317, + "step": 26623 + }, + { + "epoch": 12.377498837749883, + "grad_norm": 3.998910665512085, + "learning_rate": 9.973387997338801e-06, + "loss": 0.0329, + "step": 26624 + }, + { + "epoch": 12.377963737796374, + "grad_norm": 2.7651243209838867, + "learning_rate": 9.973386997338702e-06, + "loss": 0.0295, + "step": 26625 + }, + { + "epoch": 12.378428637842864, + "grad_norm": 4.083562850952148, + "learning_rate": 9.9733859973386e-06, + "loss": 0.0386, + "step": 26626 + }, + { + "epoch": 12.378893537889354, + "grad_norm": 6.019838809967041, + "learning_rate": 9.973384997338501e-06, + "loss": 0.0451, + "step": 26627 + }, + { + "epoch": 12.379358437935844, + "grad_norm": 2.179278612136841, + "learning_rate": 9.9733839973384e-06, + "loss": 0.0257, + "step": 26628 + }, + { + "epoch": 12.379823337982334, + "grad_norm": 5.756269454956055, + "learning_rate": 9.973382997338301e-06, + "loss": 0.0693, + "step": 26629 + }, + { + "epoch": 12.380288238028823, + "grad_norm": 7.478024482727051, + "learning_rate": 9.9733819973382e-06, + "loss": 0.1307, + "step": 26630 + }, + { + "epoch": 12.380753138075313, + "grad_norm": 5.0490336418151855, + "learning_rate": 9.9733809973381e-06, + "loss": 0.0555, + "step": 26631 + }, + { + "epoch": 12.381218038121803, + "grad_norm": 1.8932559490203857, + "learning_rate": 9.973379997338001e-06, + "loss": 0.0261, + "step": 26632 + }, + { + "epoch": 12.381682938168293, + "grad_norm": 7.743404865264893, + "learning_rate": 9.9733789973379e-06, + "loss": 0.0538, + "step": 26633 + }, + { + "epoch": 12.382147838214784, + "grad_norm": 3.294062376022339, + "learning_rate": 9.9733779973378e-06, + "loss": 0.0282, + "step": 26634 + }, + { + "epoch": 12.382612738261274, + "grad_norm": 4.427945613861084, + "learning_rate": 9.9733769973377e-06, + "loss": 0.081, + "step": 26635 + }, + { + "epoch": 12.383077638307764, + "grad_norm": 5.467460632324219, + "learning_rate": 9.9733759973376e-06, + "loss": 0.0675, + "step": 26636 + }, + { + "epoch": 12.383542538354254, + "grad_norm": 1.7451759576797485, + "learning_rate": 9.973374997337501e-06, + "loss": 0.0131, + "step": 26637 + }, + { + "epoch": 12.384007438400744, + "grad_norm": 3.384429693222046, + "learning_rate": 9.973373997337402e-06, + "loss": 0.0391, + "step": 26638 + }, + { + "epoch": 12.384472338447233, + "grad_norm": 5.795722484588623, + "learning_rate": 9.9733729973373e-06, + "loss": 0.0577, + "step": 26639 + }, + { + "epoch": 12.384937238493723, + "grad_norm": 4.407012939453125, + "learning_rate": 9.9733719973372e-06, + "loss": 0.0494, + "step": 26640 + }, + { + "epoch": 12.385402138540213, + "grad_norm": 5.371340751647949, + "learning_rate": 9.9733709973371e-06, + "loss": 0.0461, + "step": 26641 + }, + { + "epoch": 12.385867038586705, + "grad_norm": 3.2467944622039795, + "learning_rate": 9.973369997337001e-06, + "loss": 0.0249, + "step": 26642 + }, + { + "epoch": 12.386331938633194, + "grad_norm": 5.557841777801514, + "learning_rate": 9.973368997336902e-06, + "loss": 0.0699, + "step": 26643 + }, + { + "epoch": 12.386796838679684, + "grad_norm": 5.1380438804626465, + "learning_rate": 9.9733679973368e-06, + "loss": 0.0804, + "step": 26644 + }, + { + "epoch": 12.387261738726174, + "grad_norm": 1.8780003786087036, + "learning_rate": 9.973366997336701e-06, + "loss": 0.0108, + "step": 26645 + }, + { + "epoch": 12.387726638772664, + "grad_norm": 4.677558898925781, + "learning_rate": 9.9733659973366e-06, + "loss": 0.0739, + "step": 26646 + }, + { + "epoch": 12.388191538819154, + "grad_norm": 2.8988876342773438, + "learning_rate": 9.9733649973365e-06, + "loss": 0.0242, + "step": 26647 + }, + { + "epoch": 12.388656438865643, + "grad_norm": 0.9995172023773193, + "learning_rate": 9.9733639973364e-06, + "loss": 0.0103, + "step": 26648 + }, + { + "epoch": 12.389121338912133, + "grad_norm": 8.589607238769531, + "learning_rate": 9.9733629973363e-06, + "loss": 0.0872, + "step": 26649 + }, + { + "epoch": 12.389586238958623, + "grad_norm": 3.9927306175231934, + "learning_rate": 9.973361997336201e-06, + "loss": 0.0295, + "step": 26650 + }, + { + "epoch": 12.390051139005115, + "grad_norm": 4.4391069412231445, + "learning_rate": 9.9733609973361e-06, + "loss": 0.0295, + "step": 26651 + }, + { + "epoch": 12.390516039051604, + "grad_norm": 3.450064182281494, + "learning_rate": 9.973359997336e-06, + "loss": 0.0265, + "step": 26652 + }, + { + "epoch": 12.390980939098094, + "grad_norm": 1.9076184034347534, + "learning_rate": 9.9733589973359e-06, + "loss": 0.0195, + "step": 26653 + }, + { + "epoch": 12.391445839144584, + "grad_norm": 4.797369003295898, + "learning_rate": 9.9733579973358e-06, + "loss": 0.0599, + "step": 26654 + }, + { + "epoch": 12.391910739191074, + "grad_norm": 7.221858501434326, + "learning_rate": 9.973356997335701e-06, + "loss": 0.0803, + "step": 26655 + }, + { + "epoch": 12.392375639237564, + "grad_norm": 2.830190420150757, + "learning_rate": 9.973355997335602e-06, + "loss": 0.0264, + "step": 26656 + }, + { + "epoch": 12.392840539284053, + "grad_norm": 3.9937567710876465, + "learning_rate": 9.9733549973355e-06, + "loss": 0.0451, + "step": 26657 + }, + { + "epoch": 12.393305439330543, + "grad_norm": 3.5180282592773438, + "learning_rate": 9.973353997335401e-06, + "loss": 0.0564, + "step": 26658 + }, + { + "epoch": 12.393770339377033, + "grad_norm": 7.667828559875488, + "learning_rate": 9.9733529973353e-06, + "loss": 0.1328, + "step": 26659 + }, + { + "epoch": 12.394235239423525, + "grad_norm": 2.361241102218628, + "learning_rate": 9.973351997335201e-06, + "loss": 0.018, + "step": 26660 + }, + { + "epoch": 12.394700139470014, + "grad_norm": 4.572478294372559, + "learning_rate": 9.9733509973351e-06, + "loss": 0.0421, + "step": 26661 + }, + { + "epoch": 12.395165039516504, + "grad_norm": 4.272923469543457, + "learning_rate": 9.973349997335e-06, + "loss": 0.0281, + "step": 26662 + }, + { + "epoch": 12.395629939562994, + "grad_norm": 3.013446807861328, + "learning_rate": 9.973348997334901e-06, + "loss": 0.0274, + "step": 26663 + }, + { + "epoch": 12.396094839609484, + "grad_norm": 0.38906773924827576, + "learning_rate": 9.9733479973348e-06, + "loss": 0.0033, + "step": 26664 + }, + { + "epoch": 12.396559739655974, + "grad_norm": 3.9471778869628906, + "learning_rate": 9.9733469973347e-06, + "loss": 0.0443, + "step": 26665 + }, + { + "epoch": 12.397024639702463, + "grad_norm": 5.667669773101807, + "learning_rate": 9.9733459973346e-06, + "loss": 0.0862, + "step": 26666 + }, + { + "epoch": 12.397489539748953, + "grad_norm": 0.7777973413467407, + "learning_rate": 9.9733449973345e-06, + "loss": 0.0069, + "step": 26667 + }, + { + "epoch": 12.397954439795445, + "grad_norm": 1.631905436515808, + "learning_rate": 9.973343997334401e-06, + "loss": 0.0134, + "step": 26668 + }, + { + "epoch": 12.398419339841935, + "grad_norm": 3.81465220451355, + "learning_rate": 9.973342997334302e-06, + "loss": 0.0401, + "step": 26669 + }, + { + "epoch": 12.398884239888424, + "grad_norm": 2.9308011531829834, + "learning_rate": 9.9733419973342e-06, + "loss": 0.0266, + "step": 26670 + }, + { + "epoch": 12.399349139934914, + "grad_norm": 1.6335885524749756, + "learning_rate": 9.9733409973341e-06, + "loss": 0.0151, + "step": 26671 + }, + { + "epoch": 12.399814039981404, + "grad_norm": 8.184629440307617, + "learning_rate": 9.973339997334e-06, + "loss": 0.1244, + "step": 26672 + }, + { + "epoch": 12.400278940027894, + "grad_norm": 3.6108040809631348, + "learning_rate": 9.9733389973339e-06, + "loss": 0.0402, + "step": 26673 + }, + { + "epoch": 12.400743840074384, + "grad_norm": 2.6750571727752686, + "learning_rate": 9.9733379973338e-06, + "loss": 0.0175, + "step": 26674 + }, + { + "epoch": 12.401208740120873, + "grad_norm": 3.32997727394104, + "learning_rate": 9.9733369973337e-06, + "loss": 0.0353, + "step": 26675 + }, + { + "epoch": 12.401673640167363, + "grad_norm": 7.758159160614014, + "learning_rate": 9.973335997333601e-06, + "loss": 0.1328, + "step": 26676 + }, + { + "epoch": 12.402138540213855, + "grad_norm": 2.198493242263794, + "learning_rate": 9.9733349973335e-06, + "loss": 0.0372, + "step": 26677 + }, + { + "epoch": 12.402603440260345, + "grad_norm": 2.0659406185150146, + "learning_rate": 9.973333997333401e-06, + "loss": 0.017, + "step": 26678 + }, + { + "epoch": 12.403068340306834, + "grad_norm": 4.564395904541016, + "learning_rate": 9.9733329973333e-06, + "loss": 0.0471, + "step": 26679 + }, + { + "epoch": 12.403533240353324, + "grad_norm": 2.484602212905884, + "learning_rate": 9.9733319973332e-06, + "loss": 0.0186, + "step": 26680 + }, + { + "epoch": 12.403998140399814, + "grad_norm": 6.969846725463867, + "learning_rate": 9.973330997333101e-06, + "loss": 0.086, + "step": 26681 + }, + { + "epoch": 12.404463040446304, + "grad_norm": 5.993853569030762, + "learning_rate": 9.973329997333e-06, + "loss": 0.0584, + "step": 26682 + }, + { + "epoch": 12.404927940492794, + "grad_norm": 3.676887273788452, + "learning_rate": 9.9733289973329e-06, + "loss": 0.0241, + "step": 26683 + }, + { + "epoch": 12.405392840539283, + "grad_norm": 2.4011318683624268, + "learning_rate": 9.9733279973328e-06, + "loss": 0.0448, + "step": 26684 + }, + { + "epoch": 12.405857740585773, + "grad_norm": 4.369837284088135, + "learning_rate": 9.9733269973327e-06, + "loss": 0.0936, + "step": 26685 + }, + { + "epoch": 12.406322640632265, + "grad_norm": 5.065753936767578, + "learning_rate": 9.9733259973326e-06, + "loss": 0.0546, + "step": 26686 + }, + { + "epoch": 12.406787540678755, + "grad_norm": 9.615601539611816, + "learning_rate": 9.9733249973325e-06, + "loss": 0.04, + "step": 26687 + }, + { + "epoch": 12.407252440725244, + "grad_norm": 6.7778730392456055, + "learning_rate": 9.9733239973324e-06, + "loss": 0.11, + "step": 26688 + }, + { + "epoch": 12.407717340771734, + "grad_norm": 4.353189468383789, + "learning_rate": 9.973322997332301e-06, + "loss": 0.047, + "step": 26689 + }, + { + "epoch": 12.408182240818224, + "grad_norm": 4.180752754211426, + "learning_rate": 9.9733219973322e-06, + "loss": 0.0442, + "step": 26690 + }, + { + "epoch": 12.408647140864714, + "grad_norm": 2.181565046310425, + "learning_rate": 9.973320997332101e-06, + "loss": 0.0139, + "step": 26691 + }, + { + "epoch": 12.409112040911204, + "grad_norm": 14.25403881072998, + "learning_rate": 9.973319997332e-06, + "loss": 0.0369, + "step": 26692 + }, + { + "epoch": 12.409576940957693, + "grad_norm": 6.612433910369873, + "learning_rate": 9.9733189973319e-06, + "loss": 0.0776, + "step": 26693 + }, + { + "epoch": 12.410041841004183, + "grad_norm": 2.4989657402038574, + "learning_rate": 9.973317997331801e-06, + "loss": 0.0142, + "step": 26694 + }, + { + "epoch": 12.410506741050675, + "grad_norm": 1.5459569692611694, + "learning_rate": 9.9733169973317e-06, + "loss": 0.0106, + "step": 26695 + }, + { + "epoch": 12.410971641097165, + "grad_norm": 2.496539831161499, + "learning_rate": 9.973315997331601e-06, + "loss": 0.0234, + "step": 26696 + }, + { + "epoch": 12.411436541143654, + "grad_norm": 0.29519250988960266, + "learning_rate": 9.9733149973315e-06, + "loss": 0.0027, + "step": 26697 + }, + { + "epoch": 12.411901441190144, + "grad_norm": 2.7651748657226562, + "learning_rate": 9.9733139973314e-06, + "loss": 0.0172, + "step": 26698 + }, + { + "epoch": 12.412366341236634, + "grad_norm": 3.9733071327209473, + "learning_rate": 9.9733129973313e-06, + "loss": 0.0304, + "step": 26699 + }, + { + "epoch": 12.412831241283124, + "grad_norm": 6.415345668792725, + "learning_rate": 9.9733119973312e-06, + "loss": 0.0746, + "step": 26700 + }, + { + "epoch": 12.413296141329614, + "grad_norm": 7.824082374572754, + "learning_rate": 9.9733109973311e-06, + "loss": 0.1771, + "step": 26701 + }, + { + "epoch": 12.413761041376103, + "grad_norm": 4.135961532592773, + "learning_rate": 9.973309997331001e-06, + "loss": 0.0515, + "step": 26702 + }, + { + "epoch": 12.414225941422593, + "grad_norm": 4.4043803215026855, + "learning_rate": 9.9733089973309e-06, + "loss": 0.0482, + "step": 26703 + }, + { + "epoch": 12.414690841469085, + "grad_norm": 0.4879937767982483, + "learning_rate": 9.9733079973308e-06, + "loss": 0.0041, + "step": 26704 + }, + { + "epoch": 12.415155741515575, + "grad_norm": 3.8927712440490723, + "learning_rate": 9.9733069973307e-06, + "loss": 0.0217, + "step": 26705 + }, + { + "epoch": 12.415620641562064, + "grad_norm": 4.0526814460754395, + "learning_rate": 9.9733059973306e-06, + "loss": 0.0244, + "step": 26706 + }, + { + "epoch": 12.416085541608554, + "grad_norm": 5.817947864532471, + "learning_rate": 9.973304997330501e-06, + "loss": 0.121, + "step": 26707 + }, + { + "epoch": 12.416550441655044, + "grad_norm": 10.009790420532227, + "learning_rate": 9.9733039973304e-06, + "loss": 0.0882, + "step": 26708 + }, + { + "epoch": 12.417015341701534, + "grad_norm": 2.871764898300171, + "learning_rate": 9.973302997330301e-06, + "loss": 0.0289, + "step": 26709 + }, + { + "epoch": 12.417480241748024, + "grad_norm": 4.724532127380371, + "learning_rate": 9.9733019973302e-06, + "loss": 0.0461, + "step": 26710 + }, + { + "epoch": 12.417945141794513, + "grad_norm": 6.632753372192383, + "learning_rate": 9.9733009973301e-06, + "loss": 0.0945, + "step": 26711 + }, + { + "epoch": 12.418410041841003, + "grad_norm": 5.724524021148682, + "learning_rate": 9.973299997330001e-06, + "loss": 0.0407, + "step": 26712 + }, + { + "epoch": 12.418874941887495, + "grad_norm": 5.405686855316162, + "learning_rate": 9.9732989973299e-06, + "loss": 0.0793, + "step": 26713 + }, + { + "epoch": 12.419339841933985, + "grad_norm": 4.735612869262695, + "learning_rate": 9.973297997329801e-06, + "loss": 0.0303, + "step": 26714 + }, + { + "epoch": 12.419804741980474, + "grad_norm": 7.437492370605469, + "learning_rate": 9.9732969973297e-06, + "loss": 0.1183, + "step": 26715 + }, + { + "epoch": 12.420269642026964, + "grad_norm": 6.027035236358643, + "learning_rate": 9.9732959973296e-06, + "loss": 0.0595, + "step": 26716 + }, + { + "epoch": 12.420734542073454, + "grad_norm": 4.908890247344971, + "learning_rate": 9.9732949973295e-06, + "loss": 0.0375, + "step": 26717 + }, + { + "epoch": 12.421199442119944, + "grad_norm": 7.927025318145752, + "learning_rate": 9.9732939973294e-06, + "loss": 0.1274, + "step": 26718 + }, + { + "epoch": 12.421664342166434, + "grad_norm": 5.64941930770874, + "learning_rate": 9.9732929973293e-06, + "loss": 0.0825, + "step": 26719 + }, + { + "epoch": 12.422129242212923, + "grad_norm": 7.475364685058594, + "learning_rate": 9.973291997329201e-06, + "loss": 0.1163, + "step": 26720 + }, + { + "epoch": 12.422594142259415, + "grad_norm": 6.870969772338867, + "learning_rate": 9.9732909973291e-06, + "loss": 0.1763, + "step": 26721 + }, + { + "epoch": 12.423059042305905, + "grad_norm": 2.075533151626587, + "learning_rate": 9.973289997329001e-06, + "loss": 0.0166, + "step": 26722 + }, + { + "epoch": 12.423523942352395, + "grad_norm": 3.861224412918091, + "learning_rate": 9.9732889973289e-06, + "loss": 0.0322, + "step": 26723 + }, + { + "epoch": 12.423988842398884, + "grad_norm": 0.9827479720115662, + "learning_rate": 9.9732879973288e-06, + "loss": 0.0074, + "step": 26724 + }, + { + "epoch": 12.424453742445374, + "grad_norm": 6.772619247436523, + "learning_rate": 9.973286997328701e-06, + "loss": 0.1173, + "step": 26725 + }, + { + "epoch": 12.424918642491864, + "grad_norm": 4.852466106414795, + "learning_rate": 9.9732859973286e-06, + "loss": 0.0408, + "step": 26726 + }, + { + "epoch": 12.425383542538354, + "grad_norm": 1.0051970481872559, + "learning_rate": 9.973284997328501e-06, + "loss": 0.0092, + "step": 26727 + }, + { + "epoch": 12.425848442584844, + "grad_norm": 3.9799113273620605, + "learning_rate": 9.9732839973284e-06, + "loss": 0.0391, + "step": 26728 + }, + { + "epoch": 12.426313342631333, + "grad_norm": 2.064411163330078, + "learning_rate": 9.9732829973283e-06, + "loss": 0.0173, + "step": 26729 + }, + { + "epoch": 12.426778242677825, + "grad_norm": 3.3763043880462646, + "learning_rate": 9.9732819973282e-06, + "loss": 0.0411, + "step": 26730 + }, + { + "epoch": 12.427243142724315, + "grad_norm": 4.041507244110107, + "learning_rate": 9.9732809973281e-06, + "loss": 0.048, + "step": 26731 + }, + { + "epoch": 12.427708042770805, + "grad_norm": 0.6483913064002991, + "learning_rate": 9.973279997328001e-06, + "loss": 0.0034, + "step": 26732 + }, + { + "epoch": 12.428172942817294, + "grad_norm": 4.340888977050781, + "learning_rate": 9.973278997327902e-06, + "loss": 0.0518, + "step": 26733 + }, + { + "epoch": 12.428637842863784, + "grad_norm": 7.186402320861816, + "learning_rate": 9.9732779973278e-06, + "loss": 0.0888, + "step": 26734 + }, + { + "epoch": 12.429102742910274, + "grad_norm": 3.89924955368042, + "learning_rate": 9.973276997327701e-06, + "loss": 0.0278, + "step": 26735 + }, + { + "epoch": 12.429567642956764, + "grad_norm": 7.0409722328186035, + "learning_rate": 9.9732759973276e-06, + "loss": 0.0807, + "step": 26736 + }, + { + "epoch": 12.430032543003254, + "grad_norm": 1.9734375476837158, + "learning_rate": 9.9732749973275e-06, + "loss": 0.014, + "step": 26737 + }, + { + "epoch": 12.430497443049743, + "grad_norm": 1.7536895275115967, + "learning_rate": 9.973273997327401e-06, + "loss": 0.0117, + "step": 26738 + }, + { + "epoch": 12.430962343096235, + "grad_norm": 6.157166481018066, + "learning_rate": 9.9732729973273e-06, + "loss": 0.0714, + "step": 26739 + }, + { + "epoch": 12.431427243142725, + "grad_norm": 5.236515045166016, + "learning_rate": 9.973271997327201e-06, + "loss": 0.1394, + "step": 26740 + }, + { + "epoch": 12.431892143189215, + "grad_norm": 2.4508726596832275, + "learning_rate": 9.9732709973271e-06, + "loss": 0.0156, + "step": 26741 + }, + { + "epoch": 12.432357043235704, + "grad_norm": 7.657028675079346, + "learning_rate": 9.973269997327e-06, + "loss": 0.204, + "step": 26742 + }, + { + "epoch": 12.432821943282194, + "grad_norm": 3.5199472904205322, + "learning_rate": 9.9732689973269e-06, + "loss": 0.0299, + "step": 26743 + }, + { + "epoch": 12.433286843328684, + "grad_norm": 5.746664524078369, + "learning_rate": 9.9732679973268e-06, + "loss": 0.06, + "step": 26744 + }, + { + "epoch": 12.433751743375174, + "grad_norm": 8.194189071655273, + "learning_rate": 9.973266997326701e-06, + "loss": 0.1765, + "step": 26745 + }, + { + "epoch": 12.434216643421664, + "grad_norm": 5.584313869476318, + "learning_rate": 9.973265997326602e-06, + "loss": 0.0374, + "step": 26746 + }, + { + "epoch": 12.434681543468155, + "grad_norm": 4.793488502502441, + "learning_rate": 9.9732649973265e-06, + "loss": 0.0854, + "step": 26747 + }, + { + "epoch": 12.435146443514645, + "grad_norm": 5.62701940536499, + "learning_rate": 9.9732639973264e-06, + "loss": 0.0376, + "step": 26748 + }, + { + "epoch": 12.435611343561135, + "grad_norm": 3.715487480163574, + "learning_rate": 9.9732629973263e-06, + "loss": 0.0354, + "step": 26749 + }, + { + "epoch": 12.436076243607625, + "grad_norm": 3.011986494064331, + "learning_rate": 9.9732619973262e-06, + "loss": 0.0178, + "step": 26750 + }, + { + "epoch": 12.436541143654114, + "grad_norm": 6.238739490509033, + "learning_rate": 9.973260997326101e-06, + "loss": 0.1223, + "step": 26751 + }, + { + "epoch": 12.437006043700604, + "grad_norm": 6.824532985687256, + "learning_rate": 9.973259997326e-06, + "loss": 0.059, + "step": 26752 + }, + { + "epoch": 12.437470943747094, + "grad_norm": 4.633730411529541, + "learning_rate": 9.973258997325901e-06, + "loss": 0.0675, + "step": 26753 + }, + { + "epoch": 12.437935843793584, + "grad_norm": 6.979551315307617, + "learning_rate": 9.9732579973258e-06, + "loss": 0.0993, + "step": 26754 + }, + { + "epoch": 12.438400743840074, + "grad_norm": 3.813445568084717, + "learning_rate": 9.9732569973257e-06, + "loss": 0.0268, + "step": 26755 + }, + { + "epoch": 12.438865643886565, + "grad_norm": 6.436915874481201, + "learning_rate": 9.9732559973256e-06, + "loss": 0.0669, + "step": 26756 + }, + { + "epoch": 12.439330543933055, + "grad_norm": 4.248586177825928, + "learning_rate": 9.9732549973255e-06, + "loss": 0.0578, + "step": 26757 + }, + { + "epoch": 12.439795443979545, + "grad_norm": 11.669404983520508, + "learning_rate": 9.973253997325401e-06, + "loss": 0.2119, + "step": 26758 + }, + { + "epoch": 12.440260344026035, + "grad_norm": 4.039413928985596, + "learning_rate": 9.9732529973253e-06, + "loss": 0.0278, + "step": 26759 + }, + { + "epoch": 12.440725244072524, + "grad_norm": 5.448730945587158, + "learning_rate": 9.9732519973252e-06, + "loss": 0.0569, + "step": 26760 + }, + { + "epoch": 12.441190144119014, + "grad_norm": 6.062410354614258, + "learning_rate": 9.9732509973251e-06, + "loss": 0.054, + "step": 26761 + }, + { + "epoch": 12.441655044165504, + "grad_norm": 4.672779083251953, + "learning_rate": 9.973249997325e-06, + "loss": 0.0611, + "step": 26762 + }, + { + "epoch": 12.442119944211994, + "grad_norm": 4.669919013977051, + "learning_rate": 9.973248997324901e-06, + "loss": 0.09, + "step": 26763 + }, + { + "epoch": 12.442584844258484, + "grad_norm": 6.333220481872559, + "learning_rate": 9.973247997324802e-06, + "loss": 0.083, + "step": 26764 + }, + { + "epoch": 12.443049744304975, + "grad_norm": 4.240508079528809, + "learning_rate": 9.9732469973247e-06, + "loss": 0.0597, + "step": 26765 + }, + { + "epoch": 12.443514644351465, + "grad_norm": 4.229869842529297, + "learning_rate": 9.973245997324601e-06, + "loss": 0.0455, + "step": 26766 + }, + { + "epoch": 12.443979544397955, + "grad_norm": 3.7237586975097656, + "learning_rate": 9.9732449973245e-06, + "loss": 0.0712, + "step": 26767 + }, + { + "epoch": 12.444444444444445, + "grad_norm": 2.3770954608917236, + "learning_rate": 9.9732439973244e-06, + "loss": 0.021, + "step": 26768 + }, + { + "epoch": 12.444909344490934, + "grad_norm": 6.7813191413879395, + "learning_rate": 9.9732429973243e-06, + "loss": 0.0901, + "step": 26769 + }, + { + "epoch": 12.445374244537424, + "grad_norm": 10.201274871826172, + "learning_rate": 9.9732419973242e-06, + "loss": 0.0965, + "step": 26770 + }, + { + "epoch": 12.445839144583914, + "grad_norm": 5.347912311553955, + "learning_rate": 9.973240997324101e-06, + "loss": 0.0667, + "step": 26771 + }, + { + "epoch": 12.446304044630404, + "grad_norm": 5.217689037322998, + "learning_rate": 9.973239997324e-06, + "loss": 0.082, + "step": 26772 + }, + { + "epoch": 12.446768944676894, + "grad_norm": 4.480463027954102, + "learning_rate": 9.9732389973239e-06, + "loss": 0.062, + "step": 26773 + }, + { + "epoch": 12.447233844723385, + "grad_norm": 5.901220798492432, + "learning_rate": 9.9732379973238e-06, + "loss": 0.0816, + "step": 26774 + }, + { + "epoch": 12.447698744769875, + "grad_norm": 2.921037435531616, + "learning_rate": 9.9732369973237e-06, + "loss": 0.0244, + "step": 26775 + }, + { + "epoch": 12.448163644816365, + "grad_norm": 3.320878505706787, + "learning_rate": 9.973235997323601e-06, + "loss": 0.0293, + "step": 26776 + }, + { + "epoch": 12.448628544862855, + "grad_norm": 5.247039318084717, + "learning_rate": 9.973234997323502e-06, + "loss": 0.0682, + "step": 26777 + }, + { + "epoch": 12.449093444909344, + "grad_norm": 4.608414649963379, + "learning_rate": 9.9732339973234e-06, + "loss": 0.0332, + "step": 26778 + }, + { + "epoch": 12.449558344955834, + "grad_norm": 2.8944783210754395, + "learning_rate": 9.973232997323301e-06, + "loss": 0.0141, + "step": 26779 + }, + { + "epoch": 12.450023245002324, + "grad_norm": 5.685665130615234, + "learning_rate": 9.9732319973232e-06, + "loss": 0.081, + "step": 26780 + }, + { + "epoch": 12.450488145048814, + "grad_norm": 6.508766174316406, + "learning_rate": 9.973230997323101e-06, + "loss": 0.0945, + "step": 26781 + }, + { + "epoch": 12.450953045095304, + "grad_norm": 6.66121768951416, + "learning_rate": 9.973229997323e-06, + "loss": 0.132, + "step": 26782 + }, + { + "epoch": 12.451417945141795, + "grad_norm": 3.7930209636688232, + "learning_rate": 9.9732289973229e-06, + "loss": 0.0319, + "step": 26783 + }, + { + "epoch": 12.451882845188285, + "grad_norm": 2.9572336673736572, + "learning_rate": 9.973227997322801e-06, + "loss": 0.0272, + "step": 26784 + }, + { + "epoch": 12.452347745234775, + "grad_norm": 4.408201694488525, + "learning_rate": 9.9732269973227e-06, + "loss": 0.0347, + "step": 26785 + }, + { + "epoch": 12.452812645281265, + "grad_norm": 2.5232045650482178, + "learning_rate": 9.9732259973226e-06, + "loss": 0.0194, + "step": 26786 + }, + { + "epoch": 12.453277545327754, + "grad_norm": 4.561267375946045, + "learning_rate": 9.9732249973225e-06, + "loss": 0.0563, + "step": 26787 + }, + { + "epoch": 12.453742445374244, + "grad_norm": 4.4063591957092285, + "learning_rate": 9.9732239973224e-06, + "loss": 0.0441, + "step": 26788 + }, + { + "epoch": 12.454207345420734, + "grad_norm": 1.3494093418121338, + "learning_rate": 9.973222997322301e-06, + "loss": 0.0096, + "step": 26789 + }, + { + "epoch": 12.454672245467224, + "grad_norm": 5.530512809753418, + "learning_rate": 9.973221997322202e-06, + "loss": 0.0903, + "step": 26790 + }, + { + "epoch": 12.455137145513714, + "grad_norm": 4.366832733154297, + "learning_rate": 9.9732209973221e-06, + "loss": 0.0556, + "step": 26791 + }, + { + "epoch": 12.455602045560205, + "grad_norm": 3.5049941539764404, + "learning_rate": 9.973219997322e-06, + "loss": 0.0596, + "step": 26792 + }, + { + "epoch": 12.456066945606695, + "grad_norm": 7.075345039367676, + "learning_rate": 9.9732189973219e-06, + "loss": 0.1035, + "step": 26793 + }, + { + "epoch": 12.456531845653185, + "grad_norm": 2.203648328781128, + "learning_rate": 9.973217997321801e-06, + "loss": 0.0224, + "step": 26794 + }, + { + "epoch": 12.456996745699675, + "grad_norm": 6.340472221374512, + "learning_rate": 9.973216997321702e-06, + "loss": 0.0951, + "step": 26795 + }, + { + "epoch": 12.457461645746164, + "grad_norm": 6.837998867034912, + "learning_rate": 9.9732159973216e-06, + "loss": 0.1512, + "step": 26796 + }, + { + "epoch": 12.457926545792654, + "grad_norm": 5.735366344451904, + "learning_rate": 9.973214997321501e-06, + "loss": 0.0511, + "step": 26797 + }, + { + "epoch": 12.458391445839144, + "grad_norm": 3.8833932876586914, + "learning_rate": 9.9732139973214e-06, + "loss": 0.05, + "step": 26798 + }, + { + "epoch": 12.458856345885634, + "grad_norm": 4.0330424308776855, + "learning_rate": 9.973212997321301e-06, + "loss": 0.0348, + "step": 26799 + }, + { + "epoch": 12.459321245932125, + "grad_norm": 7.852912425994873, + "learning_rate": 9.9732119973212e-06, + "loss": 0.1088, + "step": 26800 + }, + { + "epoch": 12.459786145978615, + "grad_norm": 5.46663761138916, + "learning_rate": 9.9732109973211e-06, + "loss": 0.0717, + "step": 26801 + }, + { + "epoch": 12.460251046025105, + "grad_norm": 4.2560343742370605, + "learning_rate": 9.973209997321001e-06, + "loss": 0.0501, + "step": 26802 + }, + { + "epoch": 12.460715946071595, + "grad_norm": 5.826141357421875, + "learning_rate": 9.9732089973209e-06, + "loss": 0.0828, + "step": 26803 + }, + { + "epoch": 12.461180846118085, + "grad_norm": 1.442775845527649, + "learning_rate": 9.9732079973208e-06, + "loss": 0.0102, + "step": 26804 + }, + { + "epoch": 12.461645746164574, + "grad_norm": 2.084610939025879, + "learning_rate": 9.9732069973207e-06, + "loss": 0.0175, + "step": 26805 + }, + { + "epoch": 12.462110646211064, + "grad_norm": 2.8353431224823, + "learning_rate": 9.9732059973206e-06, + "loss": 0.0114, + "step": 26806 + }, + { + "epoch": 12.462575546257554, + "grad_norm": 5.609559535980225, + "learning_rate": 9.973204997320501e-06, + "loss": 0.1309, + "step": 26807 + }, + { + "epoch": 12.463040446304044, + "grad_norm": 5.94877290725708, + "learning_rate": 9.973203997320402e-06, + "loss": 0.0715, + "step": 26808 + }, + { + "epoch": 12.463505346350535, + "grad_norm": 0.4708392024040222, + "learning_rate": 9.9732029973203e-06, + "loss": 0.0041, + "step": 26809 + }, + { + "epoch": 12.463970246397025, + "grad_norm": 1.9777652025222778, + "learning_rate": 9.973201997320201e-06, + "loss": 0.0185, + "step": 26810 + }, + { + "epoch": 12.464435146443515, + "grad_norm": 6.874537944793701, + "learning_rate": 9.9732009973201e-06, + "loss": 0.1313, + "step": 26811 + }, + { + "epoch": 12.464900046490005, + "grad_norm": 2.926297426223755, + "learning_rate": 9.97319999732e-06, + "loss": 0.0171, + "step": 26812 + }, + { + "epoch": 12.465364946536495, + "grad_norm": 4.1223578453063965, + "learning_rate": 9.9731989973199e-06, + "loss": 0.03, + "step": 26813 + }, + { + "epoch": 12.465829846582984, + "grad_norm": 5.090960502624512, + "learning_rate": 9.9731979973198e-06, + "loss": 0.0707, + "step": 26814 + }, + { + "epoch": 12.466294746629474, + "grad_norm": 3.7390620708465576, + "learning_rate": 9.973196997319701e-06, + "loss": 0.0395, + "step": 26815 + }, + { + "epoch": 12.466759646675964, + "grad_norm": 3.879152297973633, + "learning_rate": 9.9731959973196e-06, + "loss": 0.0367, + "step": 26816 + }, + { + "epoch": 12.467224546722454, + "grad_norm": 4.762990951538086, + "learning_rate": 9.973194997319501e-06, + "loss": 0.0687, + "step": 26817 + }, + { + "epoch": 12.467689446768945, + "grad_norm": 5.734547138214111, + "learning_rate": 9.9731939973194e-06, + "loss": 0.1274, + "step": 26818 + }, + { + "epoch": 12.468154346815435, + "grad_norm": 3.0007619857788086, + "learning_rate": 9.9731929973193e-06, + "loss": 0.0328, + "step": 26819 + }, + { + "epoch": 12.468619246861925, + "grad_norm": 9.918216705322266, + "learning_rate": 9.973191997319201e-06, + "loss": 0.0647, + "step": 26820 + }, + { + "epoch": 12.469084146908415, + "grad_norm": 5.892980575561523, + "learning_rate": 9.973190997319102e-06, + "loss": 0.0738, + "step": 26821 + }, + { + "epoch": 12.469549046954905, + "grad_norm": 5.544731140136719, + "learning_rate": 9.973189997319e-06, + "loss": 0.0342, + "step": 26822 + }, + { + "epoch": 12.470013947001394, + "grad_norm": 0.749457061290741, + "learning_rate": 9.9731889973189e-06, + "loss": 0.0027, + "step": 26823 + }, + { + "epoch": 12.470478847047884, + "grad_norm": 8.458308219909668, + "learning_rate": 9.9731879973188e-06, + "loss": 0.0992, + "step": 26824 + }, + { + "epoch": 12.470943747094374, + "grad_norm": 6.505987644195557, + "learning_rate": 9.9731869973187e-06, + "loss": 0.1242, + "step": 26825 + }, + { + "epoch": 12.471408647140866, + "grad_norm": 1.9420628547668457, + "learning_rate": 9.9731859973186e-06, + "loss": 0.0133, + "step": 26826 + }, + { + "epoch": 12.471873547187355, + "grad_norm": 3.9950876235961914, + "learning_rate": 9.9731849973185e-06, + "loss": 0.0294, + "step": 26827 + }, + { + "epoch": 12.472338447233845, + "grad_norm": 3.2351067066192627, + "learning_rate": 9.973183997318401e-06, + "loss": 0.0143, + "step": 26828 + }, + { + "epoch": 12.472803347280335, + "grad_norm": 11.145539283752441, + "learning_rate": 9.9731829973183e-06, + "loss": 0.1948, + "step": 26829 + }, + { + "epoch": 12.473268247326825, + "grad_norm": 1.6172549724578857, + "learning_rate": 9.973181997318201e-06, + "loss": 0.0212, + "step": 26830 + }, + { + "epoch": 12.473733147373315, + "grad_norm": 2.125373363494873, + "learning_rate": 9.9731809973181e-06, + "loss": 0.0164, + "step": 26831 + }, + { + "epoch": 12.474198047419804, + "grad_norm": 4.245364189147949, + "learning_rate": 9.973179997318e-06, + "loss": 0.0441, + "step": 26832 + }, + { + "epoch": 12.474662947466294, + "grad_norm": 3.409729480743408, + "learning_rate": 9.973178997317901e-06, + "loss": 0.0371, + "step": 26833 + }, + { + "epoch": 12.475127847512784, + "grad_norm": 14.149453163146973, + "learning_rate": 9.9731779973178e-06, + "loss": 0.0971, + "step": 26834 + }, + { + "epoch": 12.475592747559276, + "grad_norm": 4.6406049728393555, + "learning_rate": 9.9731769973177e-06, + "loss": 0.0475, + "step": 26835 + }, + { + "epoch": 12.476057647605765, + "grad_norm": 1.740416407585144, + "learning_rate": 9.9731759973176e-06, + "loss": 0.0169, + "step": 26836 + }, + { + "epoch": 12.476522547652255, + "grad_norm": 1.5372027158737183, + "learning_rate": 9.9731749973175e-06, + "loss": 0.0102, + "step": 26837 + }, + { + "epoch": 12.476987447698745, + "grad_norm": 4.069766044616699, + "learning_rate": 9.9731739973174e-06, + "loss": 0.0322, + "step": 26838 + }, + { + "epoch": 12.477452347745235, + "grad_norm": 4.702234268188477, + "learning_rate": 9.9731729973173e-06, + "loss": 0.0823, + "step": 26839 + }, + { + "epoch": 12.477917247791725, + "grad_norm": 5.072729110717773, + "learning_rate": 9.9731719973172e-06, + "loss": 0.0739, + "step": 26840 + }, + { + "epoch": 12.478382147838214, + "grad_norm": 10.18645191192627, + "learning_rate": 9.973170997317101e-06, + "loss": 0.0804, + "step": 26841 + }, + { + "epoch": 12.478847047884704, + "grad_norm": 0.8509603142738342, + "learning_rate": 9.973169997317e-06, + "loss": 0.0039, + "step": 26842 + }, + { + "epoch": 12.479311947931194, + "grad_norm": 3.8885061740875244, + "learning_rate": 9.973168997316901e-06, + "loss": 0.0229, + "step": 26843 + }, + { + "epoch": 12.479776847977686, + "grad_norm": 5.036145210266113, + "learning_rate": 9.9731679973168e-06, + "loss": 0.0621, + "step": 26844 + }, + { + "epoch": 12.480241748024175, + "grad_norm": 9.1204252243042, + "learning_rate": 9.9731669973167e-06, + "loss": 0.0873, + "step": 26845 + }, + { + "epoch": 12.480706648070665, + "grad_norm": 4.815985679626465, + "learning_rate": 9.973165997316601e-06, + "loss": 0.0997, + "step": 26846 + }, + { + "epoch": 12.481171548117155, + "grad_norm": 3.7062957286834717, + "learning_rate": 9.9731649973165e-06, + "loss": 0.0398, + "step": 26847 + }, + { + "epoch": 12.481636448163645, + "grad_norm": 0.5863674879074097, + "learning_rate": 9.973163997316401e-06, + "loss": 0.0043, + "step": 26848 + }, + { + "epoch": 12.482101348210135, + "grad_norm": 5.25370454788208, + "learning_rate": 9.9731629973163e-06, + "loss": 0.101, + "step": 26849 + }, + { + "epoch": 12.482566248256624, + "grad_norm": 4.201927661895752, + "learning_rate": 9.9731619973162e-06, + "loss": 0.0453, + "step": 26850 + }, + { + "epoch": 12.483031148303114, + "grad_norm": 8.250824928283691, + "learning_rate": 9.9731609973161e-06, + "loss": 0.1646, + "step": 26851 + }, + { + "epoch": 12.483496048349604, + "grad_norm": 3.6031932830810547, + "learning_rate": 9.973159997316e-06, + "loss": 0.037, + "step": 26852 + }, + { + "epoch": 12.483960948396096, + "grad_norm": 6.603768348693848, + "learning_rate": 9.9731589973159e-06, + "loss": 0.1033, + "step": 26853 + }, + { + "epoch": 12.484425848442585, + "grad_norm": 4.196619510650635, + "learning_rate": 9.973157997315801e-06, + "loss": 0.0503, + "step": 26854 + }, + { + "epoch": 12.484890748489075, + "grad_norm": 5.5646071434021, + "learning_rate": 9.9731569973157e-06, + "loss": 0.033, + "step": 26855 + }, + { + "epoch": 12.485355648535565, + "grad_norm": 3.6773831844329834, + "learning_rate": 9.9731559973156e-06, + "loss": 0.0459, + "step": 26856 + }, + { + "epoch": 12.485820548582055, + "grad_norm": 1.5518900156021118, + "learning_rate": 9.9731549973155e-06, + "loss": 0.0052, + "step": 26857 + }, + { + "epoch": 12.486285448628545, + "grad_norm": 5.613329887390137, + "learning_rate": 9.9731539973154e-06, + "loss": 0.0562, + "step": 26858 + }, + { + "epoch": 12.486750348675034, + "grad_norm": 8.031490325927734, + "learning_rate": 9.973152997315301e-06, + "loss": 0.1625, + "step": 26859 + }, + { + "epoch": 12.487215248721524, + "grad_norm": 5.972825050354004, + "learning_rate": 9.9731519973152e-06, + "loss": 0.0881, + "step": 26860 + }, + { + "epoch": 12.487680148768014, + "grad_norm": 2.17991042137146, + "learning_rate": 9.973150997315101e-06, + "loss": 0.0182, + "step": 26861 + }, + { + "epoch": 12.488145048814506, + "grad_norm": 2.081123113632202, + "learning_rate": 9.973149997315e-06, + "loss": 0.0175, + "step": 26862 + }, + { + "epoch": 12.488609948860995, + "grad_norm": 3.060446262359619, + "learning_rate": 9.9731489973149e-06, + "loss": 0.0103, + "step": 26863 + }, + { + "epoch": 12.489074848907485, + "grad_norm": 4.221108436584473, + "learning_rate": 9.9731479973148e-06, + "loss": 0.047, + "step": 26864 + }, + { + "epoch": 12.489539748953975, + "grad_norm": 5.903130054473877, + "learning_rate": 9.9731469973147e-06, + "loss": 0.087, + "step": 26865 + }, + { + "epoch": 12.490004649000465, + "grad_norm": 3.9580674171447754, + "learning_rate": 9.973145997314601e-06, + "loss": 0.0759, + "step": 26866 + }, + { + "epoch": 12.490469549046955, + "grad_norm": 4.8727569580078125, + "learning_rate": 9.9731449973145e-06, + "loss": 0.0339, + "step": 26867 + }, + { + "epoch": 12.490934449093444, + "grad_norm": 6.116912364959717, + "learning_rate": 9.9731439973144e-06, + "loss": 0.1152, + "step": 26868 + }, + { + "epoch": 12.491399349139934, + "grad_norm": 3.1386046409606934, + "learning_rate": 9.9731429973143e-06, + "loss": 0.0336, + "step": 26869 + }, + { + "epoch": 12.491864249186424, + "grad_norm": 5.651451110839844, + "learning_rate": 9.9731419973142e-06, + "loss": 0.0768, + "step": 26870 + }, + { + "epoch": 12.492329149232916, + "grad_norm": 7.996427536010742, + "learning_rate": 9.9731409973141e-06, + "loss": 0.115, + "step": 26871 + }, + { + "epoch": 12.492794049279405, + "grad_norm": 1.7528109550476074, + "learning_rate": 9.973139997314001e-06, + "loss": 0.0113, + "step": 26872 + }, + { + "epoch": 12.493258949325895, + "grad_norm": 3.8767027854919434, + "learning_rate": 9.9731389973139e-06, + "loss": 0.0145, + "step": 26873 + }, + { + "epoch": 12.493723849372385, + "grad_norm": 0.392727792263031, + "learning_rate": 9.973137997313801e-06, + "loss": 0.0026, + "step": 26874 + }, + { + "epoch": 12.494188749418875, + "grad_norm": 3.2065751552581787, + "learning_rate": 9.9731369973137e-06, + "loss": 0.028, + "step": 26875 + }, + { + "epoch": 12.494653649465365, + "grad_norm": 5.564851760864258, + "learning_rate": 9.9731359973136e-06, + "loss": 0.068, + "step": 26876 + }, + { + "epoch": 12.495118549511854, + "grad_norm": 2.0254807472229004, + "learning_rate": 9.973134997313501e-06, + "loss": 0.0423, + "step": 26877 + }, + { + "epoch": 12.495583449558344, + "grad_norm": 2.509476900100708, + "learning_rate": 9.9731339973134e-06, + "loss": 0.0286, + "step": 26878 + }, + { + "epoch": 12.496048349604836, + "grad_norm": 3.2195966243743896, + "learning_rate": 9.973132997313301e-06, + "loss": 0.0476, + "step": 26879 + }, + { + "epoch": 12.496513249651326, + "grad_norm": 2.405956983566284, + "learning_rate": 9.9731319973132e-06, + "loss": 0.008, + "step": 26880 + }, + { + "epoch": 12.496978149697815, + "grad_norm": 5.500753402709961, + "learning_rate": 9.9731309973131e-06, + "loss": 0.0703, + "step": 26881 + }, + { + "epoch": 12.497443049744305, + "grad_norm": 4.53148889541626, + "learning_rate": 9.973129997313e-06, + "loss": 0.0544, + "step": 26882 + }, + { + "epoch": 12.497907949790795, + "grad_norm": 3.935791254043579, + "learning_rate": 9.9731289973129e-06, + "loss": 0.0456, + "step": 26883 + }, + { + "epoch": 12.498372849837285, + "grad_norm": 5.683193683624268, + "learning_rate": 9.973127997312801e-06, + "loss": 0.04, + "step": 26884 + }, + { + "epoch": 12.498837749883775, + "grad_norm": 8.078156471252441, + "learning_rate": 9.973126997312702e-06, + "loss": 0.1299, + "step": 26885 + }, + { + "epoch": 12.499302649930264, + "grad_norm": 2.71683931350708, + "learning_rate": 9.9731259973126e-06, + "loss": 0.0169, + "step": 26886 + }, + { + "epoch": 12.499767549976754, + "grad_norm": 3.030426502227783, + "learning_rate": 9.973124997312501e-06, + "loss": 0.0361, + "step": 26887 + }, + { + "epoch": 12.500232450023246, + "grad_norm": 5.943614959716797, + "learning_rate": 9.9731239973124e-06, + "loss": 0.1577, + "step": 26888 + }, + { + "epoch": 12.500697350069736, + "grad_norm": 5.496613025665283, + "learning_rate": 9.9731229973123e-06, + "loss": 0.06, + "step": 26889 + }, + { + "epoch": 12.501162250116225, + "grad_norm": 4.550415992736816, + "learning_rate": 9.973121997312201e-06, + "loss": 0.0659, + "step": 26890 + }, + { + "epoch": 12.501627150162715, + "grad_norm": 4.614595890045166, + "learning_rate": 9.9731209973121e-06, + "loss": 0.0517, + "step": 26891 + }, + { + "epoch": 12.502092050209205, + "grad_norm": 5.537369251251221, + "learning_rate": 9.973119997312001e-06, + "loss": 0.0498, + "step": 26892 + }, + { + "epoch": 12.502556950255695, + "grad_norm": 2.229180097579956, + "learning_rate": 9.9731189973119e-06, + "loss": 0.016, + "step": 26893 + }, + { + "epoch": 12.503021850302185, + "grad_norm": 4.216641426086426, + "learning_rate": 9.9731179973118e-06, + "loss": 0.0534, + "step": 26894 + }, + { + "epoch": 12.503486750348674, + "grad_norm": 3.7838380336761475, + "learning_rate": 9.9731169973117e-06, + "loss": 0.0387, + "step": 26895 + }, + { + "epoch": 12.503951650395164, + "grad_norm": 5.625321388244629, + "learning_rate": 9.9731159973116e-06, + "loss": 0.0564, + "step": 26896 + }, + { + "epoch": 12.504416550441656, + "grad_norm": 3.341182231903076, + "learning_rate": 9.973114997311501e-06, + "loss": 0.0291, + "step": 26897 + }, + { + "epoch": 12.504881450488146, + "grad_norm": 2.824125051498413, + "learning_rate": 9.973113997311402e-06, + "loss": 0.0392, + "step": 26898 + }, + { + "epoch": 12.505346350534635, + "grad_norm": 8.164416313171387, + "learning_rate": 9.9731129973113e-06, + "loss": 0.1073, + "step": 26899 + }, + { + "epoch": 12.505811250581125, + "grad_norm": 3.7506322860717773, + "learning_rate": 9.9731119973112e-06, + "loss": 0.0352, + "step": 26900 + } + ], + "logging_steps": 1, + "max_steps": 9999999, + "num_input_tokens_seen": 0, + "num_train_epochs": 4649, + "save_steps": 538, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.7040629286567936e+20, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +}