diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,41988 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 500, + "global_step": 5990, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.000333889816360601, + "grad_norm": 15.91879524315792, + "learning_rate": 1.669449081803005e-08, + "loss": 1.3044, + "step": 1 + }, + { + "epoch": 0.000667779632721202, + "grad_norm": 16.526229811615835, + "learning_rate": 3.33889816360601e-08, + "loss": 1.3681, + "step": 2 + }, + { + "epoch": 0.001001669449081803, + "grad_norm": 16.18161924767993, + "learning_rate": 5.008347245409015e-08, + "loss": 1.3617, + "step": 3 + }, + { + "epoch": 0.001335559265442404, + "grad_norm": 17.086185404051168, + "learning_rate": 6.67779632721202e-08, + "loss": 1.3855, + "step": 4 + }, + { + "epoch": 0.001669449081803005, + "grad_norm": 16.50157256558087, + "learning_rate": 8.347245409015026e-08, + "loss": 1.3555, + "step": 5 + }, + { + "epoch": 0.002003338898163606, + "grad_norm": 16.04224922318888, + "learning_rate": 1.001669449081803e-07, + "loss": 1.3631, + "step": 6 + }, + { + "epoch": 0.002337228714524207, + "grad_norm": 16.324470916826154, + "learning_rate": 1.1686143572621036e-07, + "loss": 1.3747, + "step": 7 + }, + { + "epoch": 0.002671118530884808, + "grad_norm": 15.814665566387113, + "learning_rate": 1.335559265442404e-07, + "loss": 1.3196, + "step": 8 + }, + { + "epoch": 0.003005008347245409, + "grad_norm": 16.26203101195512, + "learning_rate": 1.5025041736227048e-07, + "loss": 1.368, + "step": 9 + }, + { + "epoch": 0.00333889816360601, + "grad_norm": 15.695292966765413, + "learning_rate": 1.6694490818030052e-07, + "loss": 1.3211, + "step": 10 + }, + { + "epoch": 0.003672787979966611, + "grad_norm": 16.130321198615004, + "learning_rate": 1.8363939899833056e-07, + "loss": 1.3439, + "step": 11 + }, + { + "epoch": 0.004006677796327212, + "grad_norm": 15.745363098304072, + "learning_rate": 2.003338898163606e-07, + "loss": 1.3176, + "step": 12 + }, + { + "epoch": 0.004340567612687813, + "grad_norm": 14.922250252346696, + "learning_rate": 2.1702838063439068e-07, + "loss": 1.3247, + "step": 13 + }, + { + "epoch": 0.004674457429048414, + "grad_norm": 15.6970596051512, + "learning_rate": 2.3372287145242072e-07, + "loss": 1.3292, + "step": 14 + }, + { + "epoch": 0.005008347245409015, + "grad_norm": 14.962379666436268, + "learning_rate": 2.504173622704508e-07, + "loss": 1.2991, + "step": 15 + }, + { + "epoch": 0.005342237061769616, + "grad_norm": 16.512160014313793, + "learning_rate": 2.671118530884808e-07, + "loss": 1.3786, + "step": 16 + }, + { + "epoch": 0.005676126878130217, + "grad_norm": 13.495624071389473, + "learning_rate": 2.838063439065109e-07, + "loss": 1.2887, + "step": 17 + }, + { + "epoch": 0.006010016694490818, + "grad_norm": 13.137547825114584, + "learning_rate": 3.0050083472454095e-07, + "loss": 1.269, + "step": 18 + }, + { + "epoch": 0.006343906510851419, + "grad_norm": 12.678350666913708, + "learning_rate": 3.1719532554257097e-07, + "loss": 1.2535, + "step": 19 + }, + { + "epoch": 0.00667779632721202, + "grad_norm": 12.615038777372357, + "learning_rate": 3.3388981636060104e-07, + "loss": 1.2292, + "step": 20 + }, + { + "epoch": 0.007011686143572621, + "grad_norm": 12.792815875833629, + "learning_rate": 3.5058430717863106e-07, + "loss": 1.2602, + "step": 21 + }, + { + "epoch": 0.007345575959933222, + "grad_norm": 12.079181605177427, + "learning_rate": 3.672787979966611e-07, + "loss": 1.2104, + "step": 22 + }, + { + "epoch": 0.007679465776293823, + "grad_norm": 9.22231165358068, + "learning_rate": 3.839732888146912e-07, + "loss": 1.1153, + "step": 23 + }, + { + "epoch": 0.008013355592654424, + "grad_norm": 8.19810084069538, + "learning_rate": 4.006677796327212e-07, + "loss": 1.1071, + "step": 24 + }, + { + "epoch": 0.008347245409015025, + "grad_norm": 8.010727655757375, + "learning_rate": 4.173622704507513e-07, + "loss": 1.0901, + "step": 25 + }, + { + "epoch": 0.008681135225375626, + "grad_norm": 8.06515338245744, + "learning_rate": 4.3405676126878136e-07, + "loss": 1.0794, + "step": 26 + }, + { + "epoch": 0.009015025041736227, + "grad_norm": 7.81962294942953, + "learning_rate": 4.5075125208681137e-07, + "loss": 1.0618, + "step": 27 + }, + { + "epoch": 0.009348914858096828, + "grad_norm": 7.5146854789050295, + "learning_rate": 4.6744574290484144e-07, + "loss": 1.0636, + "step": 28 + }, + { + "epoch": 0.009682804674457429, + "grad_norm": 7.604411968291465, + "learning_rate": 4.841402337228715e-07, + "loss": 1.0515, + "step": 29 + }, + { + "epoch": 0.01001669449081803, + "grad_norm": 7.673671131493553, + "learning_rate": 5.008347245409016e-07, + "loss": 1.0709, + "step": 30 + }, + { + "epoch": 0.010350584307178631, + "grad_norm": 6.66951529487444, + "learning_rate": 5.175292153589316e-07, + "loss": 0.9283, + "step": 31 + }, + { + "epoch": 0.010684474123539232, + "grad_norm": 6.679125462199873, + "learning_rate": 5.342237061769616e-07, + "loss": 0.8452, + "step": 32 + }, + { + "epoch": 0.011018363939899833, + "grad_norm": 7.064091543648469, + "learning_rate": 5.509181969949917e-07, + "loss": 0.8021, + "step": 33 + }, + { + "epoch": 0.011352253756260434, + "grad_norm": 6.970167557242235, + "learning_rate": 5.676126878130218e-07, + "loss": 0.7813, + "step": 34 + }, + { + "epoch": 0.011686143572621035, + "grad_norm": 6.795508224673737, + "learning_rate": 5.843071786310518e-07, + "loss": 0.7656, + "step": 35 + }, + { + "epoch": 0.012020033388981636, + "grad_norm": 6.279177682848721, + "learning_rate": 6.010016694490819e-07, + "loss": 0.7786, + "step": 36 + }, + { + "epoch": 0.012353923205342237, + "grad_norm": 5.808443415937863, + "learning_rate": 6.176961602671119e-07, + "loss": 0.7536, + "step": 37 + }, + { + "epoch": 0.012687813021702838, + "grad_norm": 5.634561385409915, + "learning_rate": 6.343906510851419e-07, + "loss": 0.7498, + "step": 38 + }, + { + "epoch": 0.01302170283806344, + "grad_norm": 5.552126690014672, + "learning_rate": 6.510851419031721e-07, + "loss": 0.7194, + "step": 39 + }, + { + "epoch": 0.01335559265442404, + "grad_norm": 5.285919769022672, + "learning_rate": 6.677796327212021e-07, + "loss": 0.6824, + "step": 40 + }, + { + "epoch": 0.013689482470784642, + "grad_norm": 4.502553315795097, + "learning_rate": 6.844741235392322e-07, + "loss": 0.6498, + "step": 41 + }, + { + "epoch": 0.014023372287145243, + "grad_norm": 5.681258330636049, + "learning_rate": 7.011686143572621e-07, + "loss": 0.5684, + "step": 42 + }, + { + "epoch": 0.014357262103505844, + "grad_norm": 3.3564397479188965, + "learning_rate": 7.178631051752922e-07, + "loss": 0.5776, + "step": 43 + }, + { + "epoch": 0.014691151919866445, + "grad_norm": 2.536125280813532, + "learning_rate": 7.345575959933223e-07, + "loss": 0.5572, + "step": 44 + }, + { + "epoch": 0.015025041736227046, + "grad_norm": 2.5738791531395373, + "learning_rate": 7.512520868113524e-07, + "loss": 0.5244, + "step": 45 + }, + { + "epoch": 0.015358931552587647, + "grad_norm": 2.6444932958812477, + "learning_rate": 7.679465776293824e-07, + "loss": 0.5555, + "step": 46 + }, + { + "epoch": 0.015692821368948246, + "grad_norm": 2.040539154434188, + "learning_rate": 7.846410684474125e-07, + "loss": 0.5301, + "step": 47 + }, + { + "epoch": 0.016026711185308847, + "grad_norm": 1.6551819605945082, + "learning_rate": 8.013355592654424e-07, + "loss": 0.5379, + "step": 48 + }, + { + "epoch": 0.016360601001669448, + "grad_norm": 1.6158987347055733, + "learning_rate": 8.180300500834726e-07, + "loss": 0.4973, + "step": 49 + }, + { + "epoch": 0.01669449081803005, + "grad_norm": 1.616447590189762, + "learning_rate": 8.347245409015026e-07, + "loss": 0.5172, + "step": 50 + }, + { + "epoch": 0.01702838063439065, + "grad_norm": 1.4072233288245004, + "learning_rate": 8.514190317195327e-07, + "loss": 0.4979, + "step": 51 + }, + { + "epoch": 0.01736227045075125, + "grad_norm": 1.5732277558455279, + "learning_rate": 8.681135225375627e-07, + "loss": 0.4942, + "step": 52 + }, + { + "epoch": 0.017696160267111852, + "grad_norm": 1.512316358101188, + "learning_rate": 8.848080133555926e-07, + "loss": 0.4752, + "step": 53 + }, + { + "epoch": 0.018030050083472453, + "grad_norm": 1.4756194351546434, + "learning_rate": 9.015025041736227e-07, + "loss": 0.4953, + "step": 54 + }, + { + "epoch": 0.018363939899833055, + "grad_norm": 1.2321836395596752, + "learning_rate": 9.181969949916528e-07, + "loss": 0.4817, + "step": 55 + }, + { + "epoch": 0.018697829716193656, + "grad_norm": 1.2416134459402868, + "learning_rate": 9.348914858096829e-07, + "loss": 0.4658, + "step": 56 + }, + { + "epoch": 0.019031719532554257, + "grad_norm": 1.4562511556534186, + "learning_rate": 9.515859766277129e-07, + "loss": 0.4793, + "step": 57 + }, + { + "epoch": 0.019365609348914858, + "grad_norm": 1.2702742764485657, + "learning_rate": 9.68280467445743e-07, + "loss": 0.4561, + "step": 58 + }, + { + "epoch": 0.01969949916527546, + "grad_norm": 1.2641269356815843, + "learning_rate": 9.84974958263773e-07, + "loss": 0.4512, + "step": 59 + }, + { + "epoch": 0.02003338898163606, + "grad_norm": 1.2729636983245716, + "learning_rate": 1.0016694490818032e-06, + "loss": 0.4365, + "step": 60 + }, + { + "epoch": 0.02036727879799666, + "grad_norm": 1.253719303453252, + "learning_rate": 1.0183639398998332e-06, + "loss": 0.4651, + "step": 61 + }, + { + "epoch": 0.020701168614357262, + "grad_norm": 1.1651293564702079, + "learning_rate": 1.0350584307178632e-06, + "loss": 0.44, + "step": 62 + }, + { + "epoch": 0.021035058430717863, + "grad_norm": 1.1841726195118423, + "learning_rate": 1.0517529215358932e-06, + "loss": 0.4578, + "step": 63 + }, + { + "epoch": 0.021368948247078464, + "grad_norm": 1.157671154828316, + "learning_rate": 1.0684474123539232e-06, + "loss": 0.45, + "step": 64 + }, + { + "epoch": 0.021702838063439065, + "grad_norm": 1.170815667537297, + "learning_rate": 1.0851419031719533e-06, + "loss": 0.4332, + "step": 65 + }, + { + "epoch": 0.022036727879799666, + "grad_norm": 1.0875695731668278, + "learning_rate": 1.1018363939899835e-06, + "loss": 0.4398, + "step": 66 + }, + { + "epoch": 0.022370617696160267, + "grad_norm": 1.1363189575092736, + "learning_rate": 1.1185308848080135e-06, + "loss": 0.4332, + "step": 67 + }, + { + "epoch": 0.02270450751252087, + "grad_norm": 1.183353721961383, + "learning_rate": 1.1352253756260435e-06, + "loss": 0.4245, + "step": 68 + }, + { + "epoch": 0.02303839732888147, + "grad_norm": 1.1706871726421442, + "learning_rate": 1.1519198664440735e-06, + "loss": 0.424, + "step": 69 + }, + { + "epoch": 0.02337228714524207, + "grad_norm": 1.0940913914795585, + "learning_rate": 1.1686143572621036e-06, + "loss": 0.4214, + "step": 70 + }, + { + "epoch": 0.02370617696160267, + "grad_norm": 1.1145731098785805, + "learning_rate": 1.1853088480801336e-06, + "loss": 0.4301, + "step": 71 + }, + { + "epoch": 0.024040066777963272, + "grad_norm": 1.1093310041589923, + "learning_rate": 1.2020033388981638e-06, + "loss": 0.4175, + "step": 72 + }, + { + "epoch": 0.024373956594323874, + "grad_norm": 1.1078428193895558, + "learning_rate": 1.2186978297161938e-06, + "loss": 0.408, + "step": 73 + }, + { + "epoch": 0.024707846410684475, + "grad_norm": 1.2019889811930684, + "learning_rate": 1.2353923205342238e-06, + "loss": 0.4148, + "step": 74 + }, + { + "epoch": 0.025041736227045076, + "grad_norm": 1.0233129863902992, + "learning_rate": 1.2520868113522539e-06, + "loss": 0.4192, + "step": 75 + }, + { + "epoch": 0.025375626043405677, + "grad_norm": 1.0847249597295265, + "learning_rate": 1.2687813021702839e-06, + "loss": 0.4008, + "step": 76 + }, + { + "epoch": 0.025709515859766278, + "grad_norm": 1.018232776762982, + "learning_rate": 1.285475792988314e-06, + "loss": 0.4298, + "step": 77 + }, + { + "epoch": 0.02604340567612688, + "grad_norm": 1.1252823899172153, + "learning_rate": 1.3021702838063441e-06, + "loss": 0.4079, + "step": 78 + }, + { + "epoch": 0.02637729549248748, + "grad_norm": 0.9598394017236226, + "learning_rate": 1.3188647746243741e-06, + "loss": 0.4023, + "step": 79 + }, + { + "epoch": 0.02671118530884808, + "grad_norm": 1.0231978723269446, + "learning_rate": 1.3355592654424042e-06, + "loss": 0.4093, + "step": 80 + }, + { + "epoch": 0.027045075125208682, + "grad_norm": 0.9790934420398532, + "learning_rate": 1.352253756260434e-06, + "loss": 0.41, + "step": 81 + }, + { + "epoch": 0.027378964941569283, + "grad_norm": 1.0075592888848743, + "learning_rate": 1.3689482470784644e-06, + "loss": 0.4, + "step": 82 + }, + { + "epoch": 0.027712854757929884, + "grad_norm": 0.9884380352206732, + "learning_rate": 1.3856427378964942e-06, + "loss": 0.4119, + "step": 83 + }, + { + "epoch": 0.028046744574290485, + "grad_norm": 1.0652179932072015, + "learning_rate": 1.4023372287145242e-06, + "loss": 0.3778, + "step": 84 + }, + { + "epoch": 0.028380634390651086, + "grad_norm": 1.026023606594425, + "learning_rate": 1.4190317195325542e-06, + "loss": 0.4048, + "step": 85 + }, + { + "epoch": 0.028714524207011687, + "grad_norm": 1.0215734657545492, + "learning_rate": 1.4357262103505845e-06, + "loss": 0.4068, + "step": 86 + }, + { + "epoch": 0.02904841402337229, + "grad_norm": 0.9677393921671519, + "learning_rate": 1.4524207011686145e-06, + "loss": 0.3995, + "step": 87 + }, + { + "epoch": 0.02938230383973289, + "grad_norm": 1.0058941765911058, + "learning_rate": 1.4691151919866445e-06, + "loss": 0.4094, + "step": 88 + }, + { + "epoch": 0.02971619365609349, + "grad_norm": 1.0545623520350786, + "learning_rate": 1.4858096828046745e-06, + "loss": 0.4141, + "step": 89 + }, + { + "epoch": 0.03005008347245409, + "grad_norm": 0.9929478577765223, + "learning_rate": 1.5025041736227048e-06, + "loss": 0.4108, + "step": 90 + }, + { + "epoch": 0.030383973288814693, + "grad_norm": 1.023769426247681, + "learning_rate": 1.5191986644407348e-06, + "loss": 0.3891, + "step": 91 + }, + { + "epoch": 0.030717863105175294, + "grad_norm": 1.0845609320889147, + "learning_rate": 1.5358931552587648e-06, + "loss": 0.3997, + "step": 92 + }, + { + "epoch": 0.031051752921535895, + "grad_norm": 1.0299054226435702, + "learning_rate": 1.5525876460767946e-06, + "loss": 0.4017, + "step": 93 + }, + { + "epoch": 0.03138564273789649, + "grad_norm": 1.041219320613279, + "learning_rate": 1.569282136894825e-06, + "loss": 0.4053, + "step": 94 + }, + { + "epoch": 0.03171953255425709, + "grad_norm": 1.095310031572485, + "learning_rate": 1.5859766277128548e-06, + "loss": 0.3974, + "step": 95 + }, + { + "epoch": 0.032053422370617694, + "grad_norm": 1.1249961131050714, + "learning_rate": 1.6026711185308849e-06, + "loss": 0.398, + "step": 96 + }, + { + "epoch": 0.032387312186978295, + "grad_norm": 1.0040616337279382, + "learning_rate": 1.6193656093489149e-06, + "loss": 0.3946, + "step": 97 + }, + { + "epoch": 0.032721202003338896, + "grad_norm": 1.0062091560216209, + "learning_rate": 1.636060100166945e-06, + "loss": 0.3794, + "step": 98 + }, + { + "epoch": 0.0330550918196995, + "grad_norm": 1.167020250808046, + "learning_rate": 1.6527545909849751e-06, + "loss": 0.3936, + "step": 99 + }, + { + "epoch": 0.0333889816360601, + "grad_norm": 1.086603171260281, + "learning_rate": 1.6694490818030051e-06, + "loss": 0.3892, + "step": 100 + }, + { + "epoch": 0.0337228714524207, + "grad_norm": 1.0227672813314324, + "learning_rate": 1.6861435726210352e-06, + "loss": 0.4012, + "step": 101 + }, + { + "epoch": 0.0340567612687813, + "grad_norm": 1.0312758333750274, + "learning_rate": 1.7028380634390654e-06, + "loss": 0.3927, + "step": 102 + }, + { + "epoch": 0.0343906510851419, + "grad_norm": 1.0266663647519454, + "learning_rate": 1.7195325542570954e-06, + "loss": 0.3778, + "step": 103 + }, + { + "epoch": 0.0347245409015025, + "grad_norm": 1.0299694001464088, + "learning_rate": 1.7362270450751254e-06, + "loss": 0.3948, + "step": 104 + }, + { + "epoch": 0.035058430717863104, + "grad_norm": 1.014847859894219, + "learning_rate": 1.7529215358931554e-06, + "loss": 0.3839, + "step": 105 + }, + { + "epoch": 0.035392320534223705, + "grad_norm": 1.0032622951069858, + "learning_rate": 1.7696160267111852e-06, + "loss": 0.3766, + "step": 106 + }, + { + "epoch": 0.035726210350584306, + "grad_norm": 1.044079901742671, + "learning_rate": 1.7863105175292157e-06, + "loss": 0.4043, + "step": 107 + }, + { + "epoch": 0.03606010016694491, + "grad_norm": 1.0296782173884778, + "learning_rate": 1.8030050083472455e-06, + "loss": 0.3786, + "step": 108 + }, + { + "epoch": 0.03639398998330551, + "grad_norm": 1.0218842676307789, + "learning_rate": 1.8196994991652755e-06, + "loss": 0.3808, + "step": 109 + }, + { + "epoch": 0.03672787979966611, + "grad_norm": 1.014187041376044, + "learning_rate": 1.8363939899833055e-06, + "loss": 0.3844, + "step": 110 + }, + { + "epoch": 0.03706176961602671, + "grad_norm": 1.0773900857308034, + "learning_rate": 1.8530884808013358e-06, + "loss": 0.3898, + "step": 111 + }, + { + "epoch": 0.03739565943238731, + "grad_norm": 0.9948175690453898, + "learning_rate": 1.8697829716193658e-06, + "loss": 0.3698, + "step": 112 + }, + { + "epoch": 0.03772954924874791, + "grad_norm": 0.9800034650272993, + "learning_rate": 1.8864774624373958e-06, + "loss": 0.3773, + "step": 113 + }, + { + "epoch": 0.03806343906510851, + "grad_norm": 1.0439642056331344, + "learning_rate": 1.9031719532554258e-06, + "loss": 0.3814, + "step": 114 + }, + { + "epoch": 0.038397328881469114, + "grad_norm": 1.0189435688327744, + "learning_rate": 1.919866444073456e-06, + "loss": 0.3835, + "step": 115 + }, + { + "epoch": 0.038731218697829715, + "grad_norm": 0.9986785527564727, + "learning_rate": 1.936560934891486e-06, + "loss": 0.3692, + "step": 116 + }, + { + "epoch": 0.039065108514190316, + "grad_norm": 0.9255161547900401, + "learning_rate": 1.953255425709516e-06, + "loss": 0.3861, + "step": 117 + }, + { + "epoch": 0.03939899833055092, + "grad_norm": 0.9764661091350226, + "learning_rate": 1.969949916527546e-06, + "loss": 0.3681, + "step": 118 + }, + { + "epoch": 0.03973288814691152, + "grad_norm": 0.9615290689857015, + "learning_rate": 1.9866444073455763e-06, + "loss": 0.3759, + "step": 119 + }, + { + "epoch": 0.04006677796327212, + "grad_norm": 0.9440356735825131, + "learning_rate": 2.0033388981636063e-06, + "loss": 0.3774, + "step": 120 + }, + { + "epoch": 0.04040066777963272, + "grad_norm": 1.0430916278395903, + "learning_rate": 2.0200333889816364e-06, + "loss": 0.3737, + "step": 121 + }, + { + "epoch": 0.04073455759599332, + "grad_norm": 1.0470863173329976, + "learning_rate": 2.0367278797996664e-06, + "loss": 0.3784, + "step": 122 + }, + { + "epoch": 0.04106844741235392, + "grad_norm": 1.041758178541942, + "learning_rate": 2.0534223706176964e-06, + "loss": 0.3867, + "step": 123 + }, + { + "epoch": 0.041402337228714524, + "grad_norm": 1.0254945248784368, + "learning_rate": 2.0701168614357264e-06, + "loss": 0.3753, + "step": 124 + }, + { + "epoch": 0.041736227045075125, + "grad_norm": 0.9963775789012155, + "learning_rate": 2.0868113522537564e-06, + "loss": 0.3764, + "step": 125 + }, + { + "epoch": 0.042070116861435726, + "grad_norm": 1.0383254070293222, + "learning_rate": 2.1035058430717864e-06, + "loss": 0.3662, + "step": 126 + }, + { + "epoch": 0.04240400667779633, + "grad_norm": 1.0210391939048344, + "learning_rate": 2.1202003338898165e-06, + "loss": 0.3654, + "step": 127 + }, + { + "epoch": 0.04273789649415693, + "grad_norm": 1.038542647749649, + "learning_rate": 2.1368948247078465e-06, + "loss": 0.3748, + "step": 128 + }, + { + "epoch": 0.04307178631051753, + "grad_norm": 0.9857032065880207, + "learning_rate": 2.1535893155258765e-06, + "loss": 0.373, + "step": 129 + }, + { + "epoch": 0.04340567612687813, + "grad_norm": 0.9882965011297845, + "learning_rate": 2.1702838063439065e-06, + "loss": 0.378, + "step": 130 + }, + { + "epoch": 0.04373956594323873, + "grad_norm": 1.0400792784731479, + "learning_rate": 2.1869782971619365e-06, + "loss": 0.3666, + "step": 131 + }, + { + "epoch": 0.04407345575959933, + "grad_norm": 0.9417138144811431, + "learning_rate": 2.203672787979967e-06, + "loss": 0.3603, + "step": 132 + }, + { + "epoch": 0.04440734557595993, + "grad_norm": 1.0344909515215428, + "learning_rate": 2.220367278797997e-06, + "loss": 0.3667, + "step": 133 + }, + { + "epoch": 0.044741235392320534, + "grad_norm": 1.0795237694855764, + "learning_rate": 2.237061769616027e-06, + "loss": 0.3839, + "step": 134 + }, + { + "epoch": 0.045075125208681135, + "grad_norm": 1.034245798691318, + "learning_rate": 2.253756260434057e-06, + "loss": 0.3844, + "step": 135 + }, + { + "epoch": 0.04540901502504174, + "grad_norm": 1.0386837691179716, + "learning_rate": 2.270450751252087e-06, + "loss": 0.3712, + "step": 136 + }, + { + "epoch": 0.04574290484140234, + "grad_norm": 1.0034004833503796, + "learning_rate": 2.287145242070117e-06, + "loss": 0.3614, + "step": 137 + }, + { + "epoch": 0.04607679465776294, + "grad_norm": 0.9867162365874181, + "learning_rate": 2.303839732888147e-06, + "loss": 0.3527, + "step": 138 + }, + { + "epoch": 0.04641068447412354, + "grad_norm": 1.018361456745601, + "learning_rate": 2.320534223706177e-06, + "loss": 0.3631, + "step": 139 + }, + { + "epoch": 0.04674457429048414, + "grad_norm": 1.2114930287255294, + "learning_rate": 2.337228714524207e-06, + "loss": 0.373, + "step": 140 + }, + { + "epoch": 0.04707846410684474, + "grad_norm": 1.0777964732860152, + "learning_rate": 2.353923205342237e-06, + "loss": 0.3662, + "step": 141 + }, + { + "epoch": 0.04741235392320534, + "grad_norm": 0.978344089906831, + "learning_rate": 2.370617696160267e-06, + "loss": 0.3656, + "step": 142 + }, + { + "epoch": 0.047746243739565944, + "grad_norm": 1.0693442615699256, + "learning_rate": 2.387312186978297e-06, + "loss": 0.3689, + "step": 143 + }, + { + "epoch": 0.048080133555926545, + "grad_norm": 1.0629178390578504, + "learning_rate": 2.4040066777963276e-06, + "loss": 0.3579, + "step": 144 + }, + { + "epoch": 0.048414023372287146, + "grad_norm": 1.0555273848515148, + "learning_rate": 2.4207011686143576e-06, + "loss": 0.3535, + "step": 145 + }, + { + "epoch": 0.04874791318864775, + "grad_norm": 1.0837288035495471, + "learning_rate": 2.4373956594323876e-06, + "loss": 0.3601, + "step": 146 + }, + { + "epoch": 0.04908180300500835, + "grad_norm": 1.1181390231095911, + "learning_rate": 2.4540901502504177e-06, + "loss": 0.3643, + "step": 147 + }, + { + "epoch": 0.04941569282136895, + "grad_norm": 1.004628841118291, + "learning_rate": 2.4707846410684477e-06, + "loss": 0.3631, + "step": 148 + }, + { + "epoch": 0.04974958263772955, + "grad_norm": 1.0198650632333248, + "learning_rate": 2.4874791318864777e-06, + "loss": 0.3465, + "step": 149 + }, + { + "epoch": 0.05008347245409015, + "grad_norm": 1.085268247958911, + "learning_rate": 2.5041736227045077e-06, + "loss": 0.3609, + "step": 150 + }, + { + "epoch": 0.05041736227045075, + "grad_norm": 1.0139737878313253, + "learning_rate": 2.5208681135225377e-06, + "loss": 0.3757, + "step": 151 + }, + { + "epoch": 0.05075125208681135, + "grad_norm": 0.9698135018722278, + "learning_rate": 2.5375626043405677e-06, + "loss": 0.3594, + "step": 152 + }, + { + "epoch": 0.051085141903171954, + "grad_norm": 1.1154189184900825, + "learning_rate": 2.5542570951585978e-06, + "loss": 0.3455, + "step": 153 + }, + { + "epoch": 0.051419031719532556, + "grad_norm": 1.0084088899380255, + "learning_rate": 2.570951585976628e-06, + "loss": 0.3591, + "step": 154 + }, + { + "epoch": 0.05175292153589316, + "grad_norm": 0.9678864236777467, + "learning_rate": 2.5876460767946582e-06, + "loss": 0.3548, + "step": 155 + }, + { + "epoch": 0.05208681135225376, + "grad_norm": 1.0605568977526543, + "learning_rate": 2.6043405676126882e-06, + "loss": 0.347, + "step": 156 + }, + { + "epoch": 0.05242070116861436, + "grad_norm": 1.0693467660649567, + "learning_rate": 2.6210350584307183e-06, + "loss": 0.366, + "step": 157 + }, + { + "epoch": 0.05275459098497496, + "grad_norm": 0.9409259751936911, + "learning_rate": 2.6377295492487483e-06, + "loss": 0.3589, + "step": 158 + }, + { + "epoch": 0.05308848080133556, + "grad_norm": 1.0524876520995168, + "learning_rate": 2.6544240400667783e-06, + "loss": 0.3459, + "step": 159 + }, + { + "epoch": 0.05342237061769616, + "grad_norm": 0.9716385401590814, + "learning_rate": 2.6711185308848083e-06, + "loss": 0.3559, + "step": 160 + }, + { + "epoch": 0.05375626043405676, + "grad_norm": 0.9448460083370125, + "learning_rate": 2.687813021702838e-06, + "loss": 0.3385, + "step": 161 + }, + { + "epoch": 0.054090150250417364, + "grad_norm": 1.008644722137227, + "learning_rate": 2.704507512520868e-06, + "loss": 0.3676, + "step": 162 + }, + { + "epoch": 0.054424040066777965, + "grad_norm": 1.0300168560075817, + "learning_rate": 2.7212020033388988e-06, + "loss": 0.3679, + "step": 163 + }, + { + "epoch": 0.054757929883138566, + "grad_norm": 0.9697429152855427, + "learning_rate": 2.737896494156929e-06, + "loss": 0.3543, + "step": 164 + }, + { + "epoch": 0.05509181969949917, + "grad_norm": 0.957100595466383, + "learning_rate": 2.7545909849749584e-06, + "loss": 0.3632, + "step": 165 + }, + { + "epoch": 0.05542570951585977, + "grad_norm": 0.9557746525245039, + "learning_rate": 2.7712854757929884e-06, + "loss": 0.3507, + "step": 166 + }, + { + "epoch": 0.05575959933222037, + "grad_norm": 1.0212694444102262, + "learning_rate": 2.7879799666110184e-06, + "loss": 0.3663, + "step": 167 + }, + { + "epoch": 0.05609348914858097, + "grad_norm": 0.9839408448211211, + "learning_rate": 2.8046744574290484e-06, + "loss": 0.3623, + "step": 168 + }, + { + "epoch": 0.05642737896494157, + "grad_norm": 0.979208334260945, + "learning_rate": 2.8213689482470785e-06, + "loss": 0.3545, + "step": 169 + }, + { + "epoch": 0.05676126878130217, + "grad_norm": 0.9603473394829852, + "learning_rate": 2.8380634390651085e-06, + "loss": 0.3408, + "step": 170 + }, + { + "epoch": 0.05709515859766277, + "grad_norm": 1.0159601040652502, + "learning_rate": 2.854757929883139e-06, + "loss": 0.3382, + "step": 171 + }, + { + "epoch": 0.057429048414023375, + "grad_norm": 1.0203982436264816, + "learning_rate": 2.871452420701169e-06, + "loss": 0.3589, + "step": 172 + }, + { + "epoch": 0.057762938230383976, + "grad_norm": 1.1006085991223393, + "learning_rate": 2.888146911519199e-06, + "loss": 0.3598, + "step": 173 + }, + { + "epoch": 0.05809682804674458, + "grad_norm": 1.1268281962854176, + "learning_rate": 2.904841402337229e-06, + "loss": 0.3507, + "step": 174 + }, + { + "epoch": 0.05843071786310518, + "grad_norm": 1.015876208801789, + "learning_rate": 2.921535893155259e-06, + "loss": 0.3647, + "step": 175 + }, + { + "epoch": 0.05876460767946578, + "grad_norm": 0.8922011595870982, + "learning_rate": 2.938230383973289e-06, + "loss": 0.3489, + "step": 176 + }, + { + "epoch": 0.05909849749582638, + "grad_norm": 1.074326265836239, + "learning_rate": 2.954924874791319e-06, + "loss": 0.3548, + "step": 177 + }, + { + "epoch": 0.05943238731218698, + "grad_norm": 0.9649613367774901, + "learning_rate": 2.971619365609349e-06, + "loss": 0.3667, + "step": 178 + }, + { + "epoch": 0.05976627712854758, + "grad_norm": 1.0143201998506803, + "learning_rate": 2.9883138564273795e-06, + "loss": 0.3647, + "step": 179 + }, + { + "epoch": 0.06010016694490818, + "grad_norm": 1.0175420497878227, + "learning_rate": 3.0050083472454095e-06, + "loss": 0.3429, + "step": 180 + }, + { + "epoch": 0.060434056761268784, + "grad_norm": 0.9629378174355745, + "learning_rate": 3.0217028380634395e-06, + "loss": 0.3515, + "step": 181 + }, + { + "epoch": 0.060767946577629385, + "grad_norm": 0.979397481615563, + "learning_rate": 3.0383973288814695e-06, + "loss": 0.3454, + "step": 182 + }, + { + "epoch": 0.061101836393989986, + "grad_norm": 0.9335637275788823, + "learning_rate": 3.0550918196994996e-06, + "loss": 0.3405, + "step": 183 + }, + { + "epoch": 0.06143572621035059, + "grad_norm": 0.9494244461216702, + "learning_rate": 3.0717863105175296e-06, + "loss": 0.3557, + "step": 184 + }, + { + "epoch": 0.06176961602671119, + "grad_norm": 0.9396210054691426, + "learning_rate": 3.0884808013355596e-06, + "loss": 0.3467, + "step": 185 + }, + { + "epoch": 0.06210350584307179, + "grad_norm": 1.0294745104595333, + "learning_rate": 3.105175292153589e-06, + "loss": 0.3645, + "step": 186 + }, + { + "epoch": 0.06243739565943239, + "grad_norm": 0.9435828435802452, + "learning_rate": 3.121869782971619e-06, + "loss": 0.3532, + "step": 187 + }, + { + "epoch": 0.06277128547579298, + "grad_norm": 0.9373505336617212, + "learning_rate": 3.13856427378965e-06, + "loss": 0.3607, + "step": 188 + }, + { + "epoch": 0.06310517529215359, + "grad_norm": 0.9566026101972563, + "learning_rate": 3.15525876460768e-06, + "loss": 0.3457, + "step": 189 + }, + { + "epoch": 0.06343906510851419, + "grad_norm": 0.9988583470732278, + "learning_rate": 3.1719532554257097e-06, + "loss": 0.347, + "step": 190 + }, + { + "epoch": 0.0637729549248748, + "grad_norm": 0.9227835867490085, + "learning_rate": 3.1886477462437397e-06, + "loss": 0.3378, + "step": 191 + }, + { + "epoch": 0.06410684474123539, + "grad_norm": 0.9703522089471485, + "learning_rate": 3.2053422370617697e-06, + "loss": 0.3525, + "step": 192 + }, + { + "epoch": 0.064440734557596, + "grad_norm": 0.9179058344713982, + "learning_rate": 3.2220367278797997e-06, + "loss": 0.3454, + "step": 193 + }, + { + "epoch": 0.06477462437395659, + "grad_norm": 1.0081277095215906, + "learning_rate": 3.2387312186978297e-06, + "loss": 0.3537, + "step": 194 + }, + { + "epoch": 0.0651085141903172, + "grad_norm": 0.918338081426016, + "learning_rate": 3.2554257095158598e-06, + "loss": 0.3373, + "step": 195 + }, + { + "epoch": 0.06544240400667779, + "grad_norm": 0.9500427844171352, + "learning_rate": 3.27212020033389e-06, + "loss": 0.3477, + "step": 196 + }, + { + "epoch": 0.0657762938230384, + "grad_norm": 0.9742119092033056, + "learning_rate": 3.2888146911519202e-06, + "loss": 0.3569, + "step": 197 + }, + { + "epoch": 0.066110183639399, + "grad_norm": 0.9942627774219589, + "learning_rate": 3.3055091819699502e-06, + "loss": 0.3535, + "step": 198 + }, + { + "epoch": 0.0664440734557596, + "grad_norm": 0.9361183630535979, + "learning_rate": 3.3222036727879803e-06, + "loss": 0.343, + "step": 199 + }, + { + "epoch": 0.0667779632721202, + "grad_norm": 0.9641003661932582, + "learning_rate": 3.3388981636060103e-06, + "loss": 0.3427, + "step": 200 + }, + { + "epoch": 0.0671118530884808, + "grad_norm": 1.0066307167600785, + "learning_rate": 3.3555926544240403e-06, + "loss": 0.3293, + "step": 201 + }, + { + "epoch": 0.0674457429048414, + "grad_norm": 0.9072097301876015, + "learning_rate": 3.3722871452420703e-06, + "loss": 0.3356, + "step": 202 + }, + { + "epoch": 0.06777963272120201, + "grad_norm": 0.920396040755389, + "learning_rate": 3.3889816360601003e-06, + "loss": 0.3426, + "step": 203 + }, + { + "epoch": 0.0681135225375626, + "grad_norm": 1.0069592861331373, + "learning_rate": 3.4056761268781308e-06, + "loss": 0.3539, + "step": 204 + }, + { + "epoch": 0.06844741235392321, + "grad_norm": 0.9662945616029687, + "learning_rate": 3.422370617696161e-06, + "loss": 0.3504, + "step": 205 + }, + { + "epoch": 0.0687813021702838, + "grad_norm": 0.9479449879788858, + "learning_rate": 3.439065108514191e-06, + "loss": 0.3432, + "step": 206 + }, + { + "epoch": 0.06911519198664441, + "grad_norm": 1.0857970032655686, + "learning_rate": 3.455759599332221e-06, + "loss": 0.3431, + "step": 207 + }, + { + "epoch": 0.069449081803005, + "grad_norm": 0.960283892041666, + "learning_rate": 3.472454090150251e-06, + "loss": 0.334, + "step": 208 + }, + { + "epoch": 0.06978297161936561, + "grad_norm": 0.9413821303117863, + "learning_rate": 3.489148580968281e-06, + "loss": 0.3527, + "step": 209 + }, + { + "epoch": 0.07011686143572621, + "grad_norm": 0.9023985925875552, + "learning_rate": 3.505843071786311e-06, + "loss": 0.3519, + "step": 210 + }, + { + "epoch": 0.07045075125208682, + "grad_norm": 0.9743839886033435, + "learning_rate": 3.5225375626043405e-06, + "loss": 0.3465, + "step": 211 + }, + { + "epoch": 0.07078464106844741, + "grad_norm": 1.0025064359735205, + "learning_rate": 3.5392320534223705e-06, + "loss": 0.3627, + "step": 212 + }, + { + "epoch": 0.07111853088480802, + "grad_norm": 1.0675106211377017, + "learning_rate": 3.5559265442404014e-06, + "loss": 0.3405, + "step": 213 + }, + { + "epoch": 0.07145242070116861, + "grad_norm": 0.9569012653267086, + "learning_rate": 3.5726210350584314e-06, + "loss": 0.3572, + "step": 214 + }, + { + "epoch": 0.07178631051752922, + "grad_norm": 0.9971563244484953, + "learning_rate": 3.589315525876461e-06, + "loss": 0.3387, + "step": 215 + }, + { + "epoch": 0.07212020033388981, + "grad_norm": 1.0537754578632943, + "learning_rate": 3.606010016694491e-06, + "loss": 0.3492, + "step": 216 + }, + { + "epoch": 0.07245409015025042, + "grad_norm": 1.007920749633551, + "learning_rate": 3.622704507512521e-06, + "loss": 0.3466, + "step": 217 + }, + { + "epoch": 0.07278797996661102, + "grad_norm": 0.9571604168320659, + "learning_rate": 3.639398998330551e-06, + "loss": 0.3527, + "step": 218 + }, + { + "epoch": 0.07312186978297162, + "grad_norm": 1.0334093553662662, + "learning_rate": 3.656093489148581e-06, + "loss": 0.3484, + "step": 219 + }, + { + "epoch": 0.07345575959933222, + "grad_norm": 0.9583540278652767, + "learning_rate": 3.672787979966611e-06, + "loss": 0.3437, + "step": 220 + }, + { + "epoch": 0.07378964941569283, + "grad_norm": 0.9129699396694247, + "learning_rate": 3.6894824707846415e-06, + "loss": 0.3501, + "step": 221 + }, + { + "epoch": 0.07412353923205342, + "grad_norm": 0.8901175589465687, + "learning_rate": 3.7061769616026715e-06, + "loss": 0.3356, + "step": 222 + }, + { + "epoch": 0.07445742904841403, + "grad_norm": 0.9496143505152863, + "learning_rate": 3.7228714524207015e-06, + "loss": 0.3392, + "step": 223 + }, + { + "epoch": 0.07479131886477462, + "grad_norm": 0.9176789180189876, + "learning_rate": 3.7395659432387315e-06, + "loss": 0.3312, + "step": 224 + }, + { + "epoch": 0.07512520868113523, + "grad_norm": 0.9669660083192729, + "learning_rate": 3.7562604340567616e-06, + "loss": 0.3467, + "step": 225 + }, + { + "epoch": 0.07545909849749582, + "grad_norm": 0.9456447545741663, + "learning_rate": 3.7729549248747916e-06, + "loss": 0.3483, + "step": 226 + }, + { + "epoch": 0.07579298831385643, + "grad_norm": 0.9013027753421927, + "learning_rate": 3.7896494156928216e-06, + "loss": 0.35, + "step": 227 + }, + { + "epoch": 0.07612687813021703, + "grad_norm": 0.8764441982803429, + "learning_rate": 3.8063439065108516e-06, + "loss": 0.3509, + "step": 228 + }, + { + "epoch": 0.07646076794657763, + "grad_norm": 0.8908777170740607, + "learning_rate": 3.823038397328882e-06, + "loss": 0.3324, + "step": 229 + }, + { + "epoch": 0.07679465776293823, + "grad_norm": 0.9250360908276563, + "learning_rate": 3.839732888146912e-06, + "loss": 0.3412, + "step": 230 + }, + { + "epoch": 0.07712854757929884, + "grad_norm": 0.903556541275547, + "learning_rate": 3.856427378964942e-06, + "loss": 0.3506, + "step": 231 + }, + { + "epoch": 0.07746243739565943, + "grad_norm": 0.9746457927101904, + "learning_rate": 3.873121869782972e-06, + "loss": 0.3396, + "step": 232 + }, + { + "epoch": 0.07779632721202004, + "grad_norm": 0.9313572314056592, + "learning_rate": 3.889816360601002e-06, + "loss": 0.3358, + "step": 233 + }, + { + "epoch": 0.07813021702838063, + "grad_norm": 0.8891419515265979, + "learning_rate": 3.906510851419032e-06, + "loss": 0.3418, + "step": 234 + }, + { + "epoch": 0.07846410684474124, + "grad_norm": 0.8856147183569384, + "learning_rate": 3.923205342237062e-06, + "loss": 0.3319, + "step": 235 + }, + { + "epoch": 0.07879799666110184, + "grad_norm": 0.8813812997468566, + "learning_rate": 3.939899833055092e-06, + "loss": 0.3346, + "step": 236 + }, + { + "epoch": 0.07913188647746244, + "grad_norm": 0.8803411152808657, + "learning_rate": 3.956594323873122e-06, + "loss": 0.3474, + "step": 237 + }, + { + "epoch": 0.07946577629382304, + "grad_norm": 0.909985687989464, + "learning_rate": 3.973288814691153e-06, + "loss": 0.3403, + "step": 238 + }, + { + "epoch": 0.07979966611018365, + "grad_norm": 0.9347120585893356, + "learning_rate": 3.989983305509183e-06, + "loss": 0.346, + "step": 239 + }, + { + "epoch": 0.08013355592654424, + "grad_norm": 0.9109542092733524, + "learning_rate": 4.006677796327213e-06, + "loss": 0.3373, + "step": 240 + }, + { + "epoch": 0.08046744574290485, + "grad_norm": 0.9072400694310836, + "learning_rate": 4.023372287145243e-06, + "loss": 0.3519, + "step": 241 + }, + { + "epoch": 0.08080133555926544, + "grad_norm": 0.8886053547801572, + "learning_rate": 4.040066777963273e-06, + "loss": 0.3447, + "step": 242 + }, + { + "epoch": 0.08113522537562605, + "grad_norm": 0.9187687393136285, + "learning_rate": 4.056761268781303e-06, + "loss": 0.3268, + "step": 243 + }, + { + "epoch": 0.08146911519198664, + "grad_norm": 0.8820044441679105, + "learning_rate": 4.073455759599333e-06, + "loss": 0.3462, + "step": 244 + }, + { + "epoch": 0.08180300500834725, + "grad_norm": 0.962962346707197, + "learning_rate": 4.090150250417363e-06, + "loss": 0.3541, + "step": 245 + }, + { + "epoch": 0.08213689482470785, + "grad_norm": 0.9165122331832503, + "learning_rate": 4.106844741235393e-06, + "loss": 0.3461, + "step": 246 + }, + { + "epoch": 0.08247078464106845, + "grad_norm": 0.8650603799014295, + "learning_rate": 4.123539232053423e-06, + "loss": 0.3332, + "step": 247 + }, + { + "epoch": 0.08280467445742905, + "grad_norm": 0.9250463644991029, + "learning_rate": 4.140233722871453e-06, + "loss": 0.3353, + "step": 248 + }, + { + "epoch": 0.08313856427378966, + "grad_norm": 0.9077852351519153, + "learning_rate": 4.156928213689483e-06, + "loss": 0.3395, + "step": 249 + }, + { + "epoch": 0.08347245409015025, + "grad_norm": 0.8954623324538902, + "learning_rate": 4.173622704507513e-06, + "loss": 0.3349, + "step": 250 + }, + { + "epoch": 0.08380634390651086, + "grad_norm": 0.9994334709329052, + "learning_rate": 4.190317195325543e-06, + "loss": 0.3566, + "step": 251 + }, + { + "epoch": 0.08414023372287145, + "grad_norm": 0.9456392341751303, + "learning_rate": 4.207011686143573e-06, + "loss": 0.3262, + "step": 252 + }, + { + "epoch": 0.08447412353923206, + "grad_norm": 0.9182428255214761, + "learning_rate": 4.223706176961603e-06, + "loss": 0.3526, + "step": 253 + }, + { + "epoch": 0.08480801335559265, + "grad_norm": 0.9558985823758236, + "learning_rate": 4.240400667779633e-06, + "loss": 0.3389, + "step": 254 + }, + { + "epoch": 0.08514190317195326, + "grad_norm": 0.8831125617642548, + "learning_rate": 4.257095158597663e-06, + "loss": 0.337, + "step": 255 + }, + { + "epoch": 0.08547579298831386, + "grad_norm": 0.9217955759516457, + "learning_rate": 4.273789649415693e-06, + "loss": 0.3473, + "step": 256 + }, + { + "epoch": 0.08580968280467446, + "grad_norm": 0.9806721322408467, + "learning_rate": 4.290484140233723e-06, + "loss": 0.3442, + "step": 257 + }, + { + "epoch": 0.08614357262103506, + "grad_norm": 0.8661399916322271, + "learning_rate": 4.307178631051753e-06, + "loss": 0.3273, + "step": 258 + }, + { + "epoch": 0.08647746243739567, + "grad_norm": 0.9298300900084301, + "learning_rate": 4.323873121869783e-06, + "loss": 0.3348, + "step": 259 + }, + { + "epoch": 0.08681135225375626, + "grad_norm": 0.9327204102006542, + "learning_rate": 4.340567612687813e-06, + "loss": 0.3311, + "step": 260 + }, + { + "epoch": 0.08714524207011685, + "grad_norm": 0.9496255215224036, + "learning_rate": 4.357262103505843e-06, + "loss": 0.3469, + "step": 261 + }, + { + "epoch": 0.08747913188647746, + "grad_norm": 0.9132406490507898, + "learning_rate": 4.373956594323873e-06, + "loss": 0.3424, + "step": 262 + }, + { + "epoch": 0.08781302170283806, + "grad_norm": 0.93366322743684, + "learning_rate": 4.390651085141904e-06, + "loss": 0.3403, + "step": 263 + }, + { + "epoch": 0.08814691151919866, + "grad_norm": 0.9323312854602842, + "learning_rate": 4.407345575959934e-06, + "loss": 0.344, + "step": 264 + }, + { + "epoch": 0.08848080133555926, + "grad_norm": 0.8841395485853564, + "learning_rate": 4.424040066777964e-06, + "loss": 0.3321, + "step": 265 + }, + { + "epoch": 0.08881469115191987, + "grad_norm": 0.9149191117011519, + "learning_rate": 4.440734557595994e-06, + "loss": 0.3293, + "step": 266 + }, + { + "epoch": 0.08914858096828046, + "grad_norm": 0.880165634767173, + "learning_rate": 4.457429048414024e-06, + "loss": 0.3402, + "step": 267 + }, + { + "epoch": 0.08948247078464107, + "grad_norm": 0.9253105007941015, + "learning_rate": 4.474123539232054e-06, + "loss": 0.3408, + "step": 268 + }, + { + "epoch": 0.08981636060100166, + "grad_norm": 0.9836217359062059, + "learning_rate": 4.490818030050084e-06, + "loss": 0.3461, + "step": 269 + }, + { + "epoch": 0.09015025041736227, + "grad_norm": 0.9580615444066954, + "learning_rate": 4.507512520868114e-06, + "loss": 0.3465, + "step": 270 + }, + { + "epoch": 0.09048414023372287, + "grad_norm": 0.8637234574505968, + "learning_rate": 4.524207011686144e-06, + "loss": 0.3263, + "step": 271 + }, + { + "epoch": 0.09081803005008347, + "grad_norm": 0.9764955382983849, + "learning_rate": 4.540901502504174e-06, + "loss": 0.3369, + "step": 272 + }, + { + "epoch": 0.09115191986644407, + "grad_norm": 0.9025112218567437, + "learning_rate": 4.557595993322204e-06, + "loss": 0.3372, + "step": 273 + }, + { + "epoch": 0.09148580968280468, + "grad_norm": 0.8591678889921395, + "learning_rate": 4.574290484140234e-06, + "loss": 0.3254, + "step": 274 + }, + { + "epoch": 0.09181969949916527, + "grad_norm": 0.9580925657629958, + "learning_rate": 4.590984974958264e-06, + "loss": 0.3459, + "step": 275 + }, + { + "epoch": 0.09215358931552588, + "grad_norm": 0.9323479799932529, + "learning_rate": 4.607679465776294e-06, + "loss": 0.3408, + "step": 276 + }, + { + "epoch": 0.09248747913188647, + "grad_norm": 0.8790698805653868, + "learning_rate": 4.624373956594324e-06, + "loss": 0.3321, + "step": 277 + }, + { + "epoch": 0.09282136894824708, + "grad_norm": 0.9792224197682113, + "learning_rate": 4.641068447412354e-06, + "loss": 0.3349, + "step": 278 + }, + { + "epoch": 0.09315525876460767, + "grad_norm": 0.9580156321261818, + "learning_rate": 4.657762938230384e-06, + "loss": 0.3397, + "step": 279 + }, + { + "epoch": 0.09348914858096828, + "grad_norm": 0.8815985216315623, + "learning_rate": 4.674457429048414e-06, + "loss": 0.3352, + "step": 280 + }, + { + "epoch": 0.09382303839732888, + "grad_norm": 0.9157857887699313, + "learning_rate": 4.691151919866444e-06, + "loss": 0.3371, + "step": 281 + }, + { + "epoch": 0.09415692821368948, + "grad_norm": 1.0792540848763006, + "learning_rate": 4.707846410684474e-06, + "loss": 0.3501, + "step": 282 + }, + { + "epoch": 0.09449081803005008, + "grad_norm": 0.8520908396452329, + "learning_rate": 4.724540901502504e-06, + "loss": 0.3432, + "step": 283 + }, + { + "epoch": 0.09482470784641069, + "grad_norm": 0.9411767728831182, + "learning_rate": 4.741235392320534e-06, + "loss": 0.3379, + "step": 284 + }, + { + "epoch": 0.09515859766277128, + "grad_norm": 0.8589381325959193, + "learning_rate": 4.757929883138564e-06, + "loss": 0.3372, + "step": 285 + }, + { + "epoch": 0.09549248747913189, + "grad_norm": 0.8691240826499841, + "learning_rate": 4.774624373956594e-06, + "loss": 0.3412, + "step": 286 + }, + { + "epoch": 0.09582637729549248, + "grad_norm": 0.869431361034001, + "learning_rate": 4.791318864774624e-06, + "loss": 0.3309, + "step": 287 + }, + { + "epoch": 0.09616026711185309, + "grad_norm": 0.9193403679342965, + "learning_rate": 4.808013355592655e-06, + "loss": 0.3315, + "step": 288 + }, + { + "epoch": 0.09649415692821368, + "grad_norm": 0.8886520890931207, + "learning_rate": 4.824707846410685e-06, + "loss": 0.3347, + "step": 289 + }, + { + "epoch": 0.09682804674457429, + "grad_norm": 0.8722426956956223, + "learning_rate": 4.841402337228715e-06, + "loss": 0.3377, + "step": 290 + }, + { + "epoch": 0.09716193656093489, + "grad_norm": 0.9286404034977652, + "learning_rate": 4.858096828046745e-06, + "loss": 0.3432, + "step": 291 + }, + { + "epoch": 0.0974958263772955, + "grad_norm": 0.9033051422806828, + "learning_rate": 4.874791318864775e-06, + "loss": 0.3372, + "step": 292 + }, + { + "epoch": 0.09782971619365609, + "grad_norm": 0.9948042306557272, + "learning_rate": 4.891485809682805e-06, + "loss": 0.3334, + "step": 293 + }, + { + "epoch": 0.0981636060100167, + "grad_norm": 0.9215669038659123, + "learning_rate": 4.908180300500835e-06, + "loss": 0.3356, + "step": 294 + }, + { + "epoch": 0.09849749582637729, + "grad_norm": 0.8713128278845043, + "learning_rate": 4.924874791318865e-06, + "loss": 0.3307, + "step": 295 + }, + { + "epoch": 0.0988313856427379, + "grad_norm": 0.9302673788962498, + "learning_rate": 4.941569282136895e-06, + "loss": 0.3323, + "step": 296 + }, + { + "epoch": 0.09916527545909849, + "grad_norm": 0.9110339147339949, + "learning_rate": 4.958263772954925e-06, + "loss": 0.3295, + "step": 297 + }, + { + "epoch": 0.0994991652754591, + "grad_norm": 0.8709003024465509, + "learning_rate": 4.974958263772955e-06, + "loss": 0.3278, + "step": 298 + }, + { + "epoch": 0.0998330550918197, + "grad_norm": 0.9213845120517851, + "learning_rate": 4.991652754590985e-06, + "loss": 0.3349, + "step": 299 + }, + { + "epoch": 0.1001669449081803, + "grad_norm": 0.8782810929039643, + "learning_rate": 5.008347245409015e-06, + "loss": 0.325, + "step": 300 + }, + { + "epoch": 0.1005008347245409, + "grad_norm": 0.8613816390322981, + "learning_rate": 5.0250417362270454e-06, + "loss": 0.3377, + "step": 301 + }, + { + "epoch": 0.1008347245409015, + "grad_norm": 0.8458940237482376, + "learning_rate": 5.0417362270450755e-06, + "loss": 0.3249, + "step": 302 + }, + { + "epoch": 0.1011686143572621, + "grad_norm": 0.8446981650666892, + "learning_rate": 5.0584307178631055e-06, + "loss": 0.3272, + "step": 303 + }, + { + "epoch": 0.1015025041736227, + "grad_norm": 0.9430057526301033, + "learning_rate": 5.0751252086811355e-06, + "loss": 0.3489, + "step": 304 + }, + { + "epoch": 0.1018363939899833, + "grad_norm": 0.8797093895197422, + "learning_rate": 5.0918196994991655e-06, + "loss": 0.3396, + "step": 305 + }, + { + "epoch": 0.10217028380634391, + "grad_norm": 0.8606372393248164, + "learning_rate": 5.1085141903171955e-06, + "loss": 0.3361, + "step": 306 + }, + { + "epoch": 0.1025041736227045, + "grad_norm": 0.8276112941999388, + "learning_rate": 5.1252086811352255e-06, + "loss": 0.3403, + "step": 307 + }, + { + "epoch": 0.10283806343906511, + "grad_norm": 0.8434790553921936, + "learning_rate": 5.141903171953256e-06, + "loss": 0.3419, + "step": 308 + }, + { + "epoch": 0.1031719532554257, + "grad_norm": 0.8595236905626144, + "learning_rate": 5.158597662771286e-06, + "loss": 0.342, + "step": 309 + }, + { + "epoch": 0.10350584307178631, + "grad_norm": 0.8743695509242151, + "learning_rate": 5.1752921535893164e-06, + "loss": 0.3382, + "step": 310 + }, + { + "epoch": 0.10383973288814691, + "grad_norm": 0.8683386025203176, + "learning_rate": 5.1919866444073465e-06, + "loss": 0.3209, + "step": 311 + }, + { + "epoch": 0.10417362270450752, + "grad_norm": 0.847654823560234, + "learning_rate": 5.2086811352253765e-06, + "loss": 0.3332, + "step": 312 + }, + { + "epoch": 0.10450751252086811, + "grad_norm": 0.95009715198075, + "learning_rate": 5.2253756260434065e-06, + "loss": 0.3305, + "step": 313 + }, + { + "epoch": 0.10484140233722872, + "grad_norm": 0.9342443179319497, + "learning_rate": 5.2420701168614365e-06, + "loss": 0.3521, + "step": 314 + }, + { + "epoch": 0.10517529215358931, + "grad_norm": 0.8507851854307606, + "learning_rate": 5.2587646076794665e-06, + "loss": 0.334, + "step": 315 + }, + { + "epoch": 0.10550918196994992, + "grad_norm": 0.8778098458778606, + "learning_rate": 5.2754590984974965e-06, + "loss": 0.3432, + "step": 316 + }, + { + "epoch": 0.10584307178631051, + "grad_norm": 0.965834479261762, + "learning_rate": 5.2921535893155266e-06, + "loss": 0.336, + "step": 317 + }, + { + "epoch": 0.10617696160267112, + "grad_norm": 0.9213863032609703, + "learning_rate": 5.308848080133557e-06, + "loss": 0.3243, + "step": 318 + }, + { + "epoch": 0.10651085141903172, + "grad_norm": 0.8522019306713455, + "learning_rate": 5.325542570951587e-06, + "loss": 0.3327, + "step": 319 + }, + { + "epoch": 0.10684474123539232, + "grad_norm": 0.9238072173248589, + "learning_rate": 5.342237061769617e-06, + "loss": 0.3307, + "step": 320 + }, + { + "epoch": 0.10717863105175292, + "grad_norm": 0.8312866401829898, + "learning_rate": 5.358931552587646e-06, + "loss": 0.3291, + "step": 321 + }, + { + "epoch": 0.10751252086811353, + "grad_norm": 0.8077897082369263, + "learning_rate": 5.375626043405676e-06, + "loss": 0.3253, + "step": 322 + }, + { + "epoch": 0.10784641068447412, + "grad_norm": 0.8888966729923714, + "learning_rate": 5.392320534223706e-06, + "loss": 0.3353, + "step": 323 + }, + { + "epoch": 0.10818030050083473, + "grad_norm": 0.9068370537836071, + "learning_rate": 5.409015025041736e-06, + "loss": 0.3404, + "step": 324 + }, + { + "epoch": 0.10851419031719532, + "grad_norm": 0.8519360602267598, + "learning_rate": 5.4257095158597675e-06, + "loss": 0.3161, + "step": 325 + }, + { + "epoch": 0.10884808013355593, + "grad_norm": 0.9212585312137787, + "learning_rate": 5.4424040066777976e-06, + "loss": 0.3326, + "step": 326 + }, + { + "epoch": 0.10918196994991652, + "grad_norm": 0.8411551775156361, + "learning_rate": 5.459098497495828e-06, + "loss": 0.332, + "step": 327 + }, + { + "epoch": 0.10951585976627713, + "grad_norm": 0.9382390511956942, + "learning_rate": 5.475792988313858e-06, + "loss": 0.3398, + "step": 328 + }, + { + "epoch": 0.10984974958263773, + "grad_norm": 0.9752062887714557, + "learning_rate": 5.492487479131887e-06, + "loss": 0.3228, + "step": 329 + }, + { + "epoch": 0.11018363939899833, + "grad_norm": 0.8454053610274145, + "learning_rate": 5.509181969949917e-06, + "loss": 0.3242, + "step": 330 + }, + { + "epoch": 0.11051752921535893, + "grad_norm": 0.8660943375026406, + "learning_rate": 5.525876460767947e-06, + "loss": 0.3372, + "step": 331 + }, + { + "epoch": 0.11085141903171954, + "grad_norm": 0.8563502508539904, + "learning_rate": 5.542570951585977e-06, + "loss": 0.3341, + "step": 332 + }, + { + "epoch": 0.11118530884808013, + "grad_norm": 0.8533695962441461, + "learning_rate": 5.559265442404007e-06, + "loss": 0.3216, + "step": 333 + }, + { + "epoch": 0.11151919866444074, + "grad_norm": 0.7872028730658, + "learning_rate": 5.575959933222037e-06, + "loss": 0.3327, + "step": 334 + }, + { + "epoch": 0.11185308848080133, + "grad_norm": 0.8631343828557739, + "learning_rate": 5.592654424040067e-06, + "loss": 0.3321, + "step": 335 + }, + { + "epoch": 0.11218697829716194, + "grad_norm": 0.8744185420897909, + "learning_rate": 5.609348914858097e-06, + "loss": 0.3437, + "step": 336 + }, + { + "epoch": 0.11252086811352253, + "grad_norm": 0.8500993786921544, + "learning_rate": 5.626043405676127e-06, + "loss": 0.3338, + "step": 337 + }, + { + "epoch": 0.11285475792988314, + "grad_norm": 0.8641135275195633, + "learning_rate": 5.642737896494157e-06, + "loss": 0.3329, + "step": 338 + }, + { + "epoch": 0.11318864774624374, + "grad_norm": 0.8377859739095267, + "learning_rate": 5.659432387312187e-06, + "loss": 0.338, + "step": 339 + }, + { + "epoch": 0.11352253756260434, + "grad_norm": 0.848360523128495, + "learning_rate": 5.676126878130217e-06, + "loss": 0.3285, + "step": 340 + }, + { + "epoch": 0.11385642737896494, + "grad_norm": 0.8806558535982527, + "learning_rate": 5.692821368948247e-06, + "loss": 0.321, + "step": 341 + }, + { + "epoch": 0.11419031719532555, + "grad_norm": 0.8911932698662547, + "learning_rate": 5.709515859766278e-06, + "loss": 0.3437, + "step": 342 + }, + { + "epoch": 0.11452420701168614, + "grad_norm": 0.8789816109714064, + "learning_rate": 5.726210350584308e-06, + "loss": 0.3277, + "step": 343 + }, + { + "epoch": 0.11485809682804675, + "grad_norm": 0.8895965514124407, + "learning_rate": 5.742904841402338e-06, + "loss": 0.3167, + "step": 344 + }, + { + "epoch": 0.11519198664440734, + "grad_norm": 0.8005225949113343, + "learning_rate": 5.759599332220368e-06, + "loss": 0.3281, + "step": 345 + }, + { + "epoch": 0.11552587646076795, + "grad_norm": 0.865345057594085, + "learning_rate": 5.776293823038398e-06, + "loss": 0.3265, + "step": 346 + }, + { + "epoch": 0.11585976627712855, + "grad_norm": 0.8813619998295729, + "learning_rate": 5.792988313856428e-06, + "loss": 0.3259, + "step": 347 + }, + { + "epoch": 0.11619365609348915, + "grad_norm": 0.9543303480238511, + "learning_rate": 5.809682804674458e-06, + "loss": 0.3492, + "step": 348 + }, + { + "epoch": 0.11652754590984975, + "grad_norm": 0.883948557192829, + "learning_rate": 5.826377295492488e-06, + "loss": 0.3315, + "step": 349 + }, + { + "epoch": 0.11686143572621036, + "grad_norm": 0.9173200207228226, + "learning_rate": 5.843071786310518e-06, + "loss": 0.324, + "step": 350 + }, + { + "epoch": 0.11719532554257095, + "grad_norm": 0.9022956605984715, + "learning_rate": 5.859766277128548e-06, + "loss": 0.3318, + "step": 351 + }, + { + "epoch": 0.11752921535893156, + "grad_norm": 0.8857687971842205, + "learning_rate": 5.876460767946578e-06, + "loss": 0.3286, + "step": 352 + }, + { + "epoch": 0.11786310517529215, + "grad_norm": 0.905520621896653, + "learning_rate": 5.893155258764608e-06, + "loss": 0.3328, + "step": 353 + }, + { + "epoch": 0.11819699499165276, + "grad_norm": 0.8384445646032247, + "learning_rate": 5.909849749582638e-06, + "loss": 0.3337, + "step": 354 + }, + { + "epoch": 0.11853088480801335, + "grad_norm": 0.837717217196992, + "learning_rate": 5.926544240400668e-06, + "loss": 0.3294, + "step": 355 + }, + { + "epoch": 0.11886477462437396, + "grad_norm": 0.86668603253307, + "learning_rate": 5.943238731218698e-06, + "loss": 0.3372, + "step": 356 + }, + { + "epoch": 0.11919866444073456, + "grad_norm": 0.810297122099109, + "learning_rate": 5.959933222036728e-06, + "loss": 0.3255, + "step": 357 + }, + { + "epoch": 0.11953255425709516, + "grad_norm": 0.8428332630366204, + "learning_rate": 5.976627712854759e-06, + "loss": 0.3151, + "step": 358 + }, + { + "epoch": 0.11986644407345576, + "grad_norm": 0.7979082948510249, + "learning_rate": 5.993322203672789e-06, + "loss": 0.328, + "step": 359 + }, + { + "epoch": 0.12020033388981637, + "grad_norm": 0.8516069558311554, + "learning_rate": 6.010016694490819e-06, + "loss": 0.3512, + "step": 360 + }, + { + "epoch": 0.12053422370617696, + "grad_norm": 0.8198620013602222, + "learning_rate": 6.026711185308849e-06, + "loss": 0.3373, + "step": 361 + }, + { + "epoch": 0.12086811352253757, + "grad_norm": 0.8868738706978458, + "learning_rate": 6.043405676126879e-06, + "loss": 0.3172, + "step": 362 + }, + { + "epoch": 0.12120200333889816, + "grad_norm": 0.8135957714657285, + "learning_rate": 6.060100166944909e-06, + "loss": 0.3333, + "step": 363 + }, + { + "epoch": 0.12153589315525877, + "grad_norm": 0.8408112653840903, + "learning_rate": 6.076794657762939e-06, + "loss": 0.3353, + "step": 364 + }, + { + "epoch": 0.12186978297161936, + "grad_norm": 0.8648243707216866, + "learning_rate": 6.093489148580969e-06, + "loss": 0.3282, + "step": 365 + }, + { + "epoch": 0.12220367278797997, + "grad_norm": 0.7764350347339721, + "learning_rate": 6.110183639398999e-06, + "loss": 0.3247, + "step": 366 + }, + { + "epoch": 0.12253756260434057, + "grad_norm": 0.8276983279313975, + "learning_rate": 6.126878130217029e-06, + "loss": 0.3326, + "step": 367 + }, + { + "epoch": 0.12287145242070117, + "grad_norm": 0.8343677996585256, + "learning_rate": 6.143572621035059e-06, + "loss": 0.3341, + "step": 368 + }, + { + "epoch": 0.12320534223706177, + "grad_norm": 0.8038712611859956, + "learning_rate": 6.160267111853089e-06, + "loss": 0.3123, + "step": 369 + }, + { + "epoch": 0.12353923205342238, + "grad_norm": 0.8108769999995006, + "learning_rate": 6.176961602671119e-06, + "loss": 0.323, + "step": 370 + }, + { + "epoch": 0.12387312186978297, + "grad_norm": 0.8508992600388378, + "learning_rate": 6.193656093489148e-06, + "loss": 0.3347, + "step": 371 + }, + { + "epoch": 0.12420701168614358, + "grad_norm": 0.8654023360777294, + "learning_rate": 6.210350584307178e-06, + "loss": 0.3256, + "step": 372 + }, + { + "epoch": 0.12454090150250417, + "grad_norm": 0.8461183275376666, + "learning_rate": 6.227045075125208e-06, + "loss": 0.3309, + "step": 373 + }, + { + "epoch": 0.12487479131886478, + "grad_norm": 0.8313843956377666, + "learning_rate": 6.243739565943238e-06, + "loss": 0.3135, + "step": 374 + }, + { + "epoch": 0.12520868113522537, + "grad_norm": 0.8265471634495156, + "learning_rate": 6.26043405676127e-06, + "loss": 0.3272, + "step": 375 + }, + { + "epoch": 0.12554257095158597, + "grad_norm": 0.7637671960183157, + "learning_rate": 6.2771285475793e-06, + "loss": 0.327, + "step": 376 + }, + { + "epoch": 0.1258764607679466, + "grad_norm": 0.8195707514069247, + "learning_rate": 6.29382303839733e-06, + "loss": 0.3233, + "step": 377 + }, + { + "epoch": 0.12621035058430718, + "grad_norm": 0.8739069950360352, + "learning_rate": 6.31051752921536e-06, + "loss": 0.3267, + "step": 378 + }, + { + "epoch": 0.12654424040066778, + "grad_norm": 0.8298175916421785, + "learning_rate": 6.327212020033389e-06, + "loss": 0.3324, + "step": 379 + }, + { + "epoch": 0.12687813021702837, + "grad_norm": 0.8282178596385226, + "learning_rate": 6.343906510851419e-06, + "loss": 0.3272, + "step": 380 + }, + { + "epoch": 0.127212020033389, + "grad_norm": 0.8585919019773953, + "learning_rate": 6.360601001669449e-06, + "loss": 0.3247, + "step": 381 + }, + { + "epoch": 0.1275459098497496, + "grad_norm": 0.818490275166481, + "learning_rate": 6.377295492487479e-06, + "loss": 0.331, + "step": 382 + }, + { + "epoch": 0.12787979966611018, + "grad_norm": 0.7608322690113255, + "learning_rate": 6.393989983305509e-06, + "loss": 0.3201, + "step": 383 + }, + { + "epoch": 0.12821368948247078, + "grad_norm": 0.814691056369039, + "learning_rate": 6.4106844741235394e-06, + "loss": 0.3377, + "step": 384 + }, + { + "epoch": 0.1285475792988314, + "grad_norm": 0.8528866866369844, + "learning_rate": 6.4273789649415694e-06, + "loss": 0.3278, + "step": 385 + }, + { + "epoch": 0.128881469115192, + "grad_norm": 0.8606733996138757, + "learning_rate": 6.4440734557595995e-06, + "loss": 0.3406, + "step": 386 + }, + { + "epoch": 0.1292153589315526, + "grad_norm": 0.8141897560185359, + "learning_rate": 6.4607679465776295e-06, + "loss": 0.3251, + "step": 387 + }, + { + "epoch": 0.12954924874791318, + "grad_norm": 0.7557603930605392, + "learning_rate": 6.4774624373956595e-06, + "loss": 0.325, + "step": 388 + }, + { + "epoch": 0.1298831385642738, + "grad_norm": 0.8439497793698495, + "learning_rate": 6.4941569282136895e-06, + "loss": 0.3299, + "step": 389 + }, + { + "epoch": 0.1302170283806344, + "grad_norm": 0.7852963444394249, + "learning_rate": 6.5108514190317195e-06, + "loss": 0.316, + "step": 390 + }, + { + "epoch": 0.130550918196995, + "grad_norm": 0.7845846751243021, + "learning_rate": 6.5275459098497496e-06, + "loss": 0.3352, + "step": 391 + }, + { + "epoch": 0.13088480801335559, + "grad_norm": 0.8203040503068401, + "learning_rate": 6.54424040066778e-06, + "loss": 0.3159, + "step": 392 + }, + { + "epoch": 0.13121869782971618, + "grad_norm": 0.8126763764582194, + "learning_rate": 6.5609348914858104e-06, + "loss": 0.3349, + "step": 393 + }, + { + "epoch": 0.1315525876460768, + "grad_norm": 0.8429312332818646, + "learning_rate": 6.5776293823038405e-06, + "loss": 0.334, + "step": 394 + }, + { + "epoch": 0.1318864774624374, + "grad_norm": 0.8486192779298156, + "learning_rate": 6.5943238731218705e-06, + "loss": 0.3325, + "step": 395 + }, + { + "epoch": 0.132220367278798, + "grad_norm": 0.8683177004741326, + "learning_rate": 6.6110183639399005e-06, + "loss": 0.3235, + "step": 396 + }, + { + "epoch": 0.13255425709515858, + "grad_norm": 0.810216483701132, + "learning_rate": 6.6277128547579305e-06, + "loss": 0.3363, + "step": 397 + }, + { + "epoch": 0.1328881469115192, + "grad_norm": 0.7947307457131201, + "learning_rate": 6.6444073455759605e-06, + "loss": 0.3334, + "step": 398 + }, + { + "epoch": 0.1332220367278798, + "grad_norm": 0.8374572819775145, + "learning_rate": 6.6611018363939905e-06, + "loss": 0.3212, + "step": 399 + }, + { + "epoch": 0.1335559265442404, + "grad_norm": 0.822519162208116, + "learning_rate": 6.6777963272120206e-06, + "loss": 0.3227, + "step": 400 + }, + { + "epoch": 0.133889816360601, + "grad_norm": 0.770840773736278, + "learning_rate": 6.6944908180300506e-06, + "loss": 0.3393, + "step": 401 + }, + { + "epoch": 0.1342237061769616, + "grad_norm": 0.8527808953819698, + "learning_rate": 6.711185308848081e-06, + "loss": 0.3201, + "step": 402 + }, + { + "epoch": 0.1345575959933222, + "grad_norm": 0.8086954345389686, + "learning_rate": 6.727879799666111e-06, + "loss": 0.333, + "step": 403 + }, + { + "epoch": 0.1348914858096828, + "grad_norm": 0.7913703220538592, + "learning_rate": 6.744574290484141e-06, + "loss": 0.3284, + "step": 404 + }, + { + "epoch": 0.1352253756260434, + "grad_norm": 0.7838272959714487, + "learning_rate": 6.761268781302171e-06, + "loss": 0.3216, + "step": 405 + }, + { + "epoch": 0.13555926544240401, + "grad_norm": 0.7837160696591648, + "learning_rate": 6.777963272120201e-06, + "loss": 0.322, + "step": 406 + }, + { + "epoch": 0.1358931552587646, + "grad_norm": 0.8180461226446389, + "learning_rate": 6.794657762938231e-06, + "loss": 0.317, + "step": 407 + }, + { + "epoch": 0.1362270450751252, + "grad_norm": 0.7940214715360665, + "learning_rate": 6.8113522537562615e-06, + "loss": 0.3291, + "step": 408 + }, + { + "epoch": 0.1365609348914858, + "grad_norm": 0.8161593979607046, + "learning_rate": 6.8280467445742916e-06, + "loss": 0.3372, + "step": 409 + }, + { + "epoch": 0.13689482470784642, + "grad_norm": 0.7381881339280564, + "learning_rate": 6.844741235392322e-06, + "loss": 0.3254, + "step": 410 + }, + { + "epoch": 0.137228714524207, + "grad_norm": 0.8620790703753725, + "learning_rate": 6.861435726210352e-06, + "loss": 0.326, + "step": 411 + }, + { + "epoch": 0.1375626043405676, + "grad_norm": 0.75569445574404, + "learning_rate": 6.878130217028382e-06, + "loss": 0.3204, + "step": 412 + }, + { + "epoch": 0.1378964941569282, + "grad_norm": 0.830696908087793, + "learning_rate": 6.894824707846412e-06, + "loss": 0.3257, + "step": 413 + }, + { + "epoch": 0.13823038397328882, + "grad_norm": 0.8228195228523577, + "learning_rate": 6.911519198664442e-06, + "loss": 0.3184, + "step": 414 + }, + { + "epoch": 0.13856427378964942, + "grad_norm": 0.750075507723078, + "learning_rate": 6.928213689482472e-06, + "loss": 0.3276, + "step": 415 + }, + { + "epoch": 0.13889816360601, + "grad_norm": 0.7586641213172387, + "learning_rate": 6.944908180300502e-06, + "loss": 0.3155, + "step": 416 + }, + { + "epoch": 0.1392320534223706, + "grad_norm": 0.8133320162828276, + "learning_rate": 6.961602671118532e-06, + "loss": 0.3298, + "step": 417 + }, + { + "epoch": 0.13956594323873123, + "grad_norm": 0.8096909492539445, + "learning_rate": 6.978297161936562e-06, + "loss": 0.3232, + "step": 418 + }, + { + "epoch": 0.13989983305509182, + "grad_norm": 0.7784378671770176, + "learning_rate": 6.994991652754592e-06, + "loss": 0.3276, + "step": 419 + }, + { + "epoch": 0.14023372287145242, + "grad_norm": 0.8233390338875066, + "learning_rate": 7.011686143572622e-06, + "loss": 0.3204, + "step": 420 + }, + { + "epoch": 0.140567612687813, + "grad_norm": 0.7910329644800447, + "learning_rate": 7.028380634390651e-06, + "loss": 0.3199, + "step": 421 + }, + { + "epoch": 0.14090150250417363, + "grad_norm": 0.7935100023699913, + "learning_rate": 7.045075125208681e-06, + "loss": 0.3276, + "step": 422 + }, + { + "epoch": 0.14123539232053423, + "grad_norm": 0.7936093356923852, + "learning_rate": 7.061769616026711e-06, + "loss": 0.3335, + "step": 423 + }, + { + "epoch": 0.14156928213689482, + "grad_norm": 0.752004131269019, + "learning_rate": 7.078464106844741e-06, + "loss": 0.3257, + "step": 424 + }, + { + "epoch": 0.1419031719532554, + "grad_norm": 0.844914202830448, + "learning_rate": 7.095158597662773e-06, + "loss": 0.3112, + "step": 425 + }, + { + "epoch": 0.14223706176961604, + "grad_norm": 0.8222150974005993, + "learning_rate": 7.111853088480803e-06, + "loss": 0.3418, + "step": 426 + }, + { + "epoch": 0.14257095158597663, + "grad_norm": 0.8200860989004035, + "learning_rate": 7.128547579298833e-06, + "loss": 0.3366, + "step": 427 + }, + { + "epoch": 0.14290484140233722, + "grad_norm": 0.8073706397363849, + "learning_rate": 7.145242070116863e-06, + "loss": 0.3332, + "step": 428 + }, + { + "epoch": 0.14323873121869782, + "grad_norm": 0.8128276065396839, + "learning_rate": 7.161936560934892e-06, + "loss": 0.3193, + "step": 429 + }, + { + "epoch": 0.14357262103505844, + "grad_norm": 0.8560999218086056, + "learning_rate": 7.178631051752922e-06, + "loss": 0.3302, + "step": 430 + }, + { + "epoch": 0.14390651085141903, + "grad_norm": 0.7884802446569247, + "learning_rate": 7.195325542570952e-06, + "loss": 0.3205, + "step": 431 + }, + { + "epoch": 0.14424040066777963, + "grad_norm": 0.787729998121904, + "learning_rate": 7.212020033388982e-06, + "loss": 0.3144, + "step": 432 + }, + { + "epoch": 0.14457429048414022, + "grad_norm": 0.794053707141777, + "learning_rate": 7.228714524207012e-06, + "loss": 0.3221, + "step": 433 + }, + { + "epoch": 0.14490818030050084, + "grad_norm": 0.8380603044204966, + "learning_rate": 7.245409015025042e-06, + "loss": 0.3318, + "step": 434 + }, + { + "epoch": 0.14524207011686144, + "grad_norm": 0.8506750095076979, + "learning_rate": 7.262103505843072e-06, + "loss": 0.3185, + "step": 435 + }, + { + "epoch": 0.14557595993322203, + "grad_norm": 0.7975335360066015, + "learning_rate": 7.278797996661102e-06, + "loss": 0.327, + "step": 436 + }, + { + "epoch": 0.14590984974958263, + "grad_norm": 0.8298342889752465, + "learning_rate": 7.295492487479132e-06, + "loss": 0.3406, + "step": 437 + }, + { + "epoch": 0.14624373956594325, + "grad_norm": 0.8252696211789099, + "learning_rate": 7.312186978297162e-06, + "loss": 0.333, + "step": 438 + }, + { + "epoch": 0.14657762938230384, + "grad_norm": 0.7989865699215479, + "learning_rate": 7.328881469115192e-06, + "loss": 0.3266, + "step": 439 + }, + { + "epoch": 0.14691151919866444, + "grad_norm": 0.8388964678660478, + "learning_rate": 7.345575959933222e-06, + "loss": 0.3272, + "step": 440 + }, + { + "epoch": 0.14724540901502503, + "grad_norm": 0.7749527194739172, + "learning_rate": 7.362270450751253e-06, + "loss": 0.329, + "step": 441 + }, + { + "epoch": 0.14757929883138565, + "grad_norm": 0.8105450729693944, + "learning_rate": 7.378964941569283e-06, + "loss": 0.3266, + "step": 442 + }, + { + "epoch": 0.14791318864774625, + "grad_norm": 0.8143023003332843, + "learning_rate": 7.395659432387313e-06, + "loss": 0.3205, + "step": 443 + }, + { + "epoch": 0.14824707846410684, + "grad_norm": 0.7840657204520638, + "learning_rate": 7.412353923205343e-06, + "loss": 0.3296, + "step": 444 + }, + { + "epoch": 0.14858096828046743, + "grad_norm": 0.7793997869627715, + "learning_rate": 7.429048414023373e-06, + "loss": 0.3248, + "step": 445 + }, + { + "epoch": 0.14891485809682806, + "grad_norm": 0.8613197225439834, + "learning_rate": 7.445742904841403e-06, + "loss": 0.3185, + "step": 446 + }, + { + "epoch": 0.14924874791318865, + "grad_norm": 0.7484265293563093, + "learning_rate": 7.462437395659433e-06, + "loss": 0.3334, + "step": 447 + }, + { + "epoch": 0.14958263772954924, + "grad_norm": 0.8441286433591095, + "learning_rate": 7.479131886477463e-06, + "loss": 0.3365, + "step": 448 + }, + { + "epoch": 0.14991652754590984, + "grad_norm": 0.8963193258941903, + "learning_rate": 7.495826377295493e-06, + "loss": 0.3295, + "step": 449 + }, + { + "epoch": 0.15025041736227046, + "grad_norm": 0.7982739108147751, + "learning_rate": 7.512520868113523e-06, + "loss": 0.3334, + "step": 450 + }, + { + "epoch": 0.15058430717863105, + "grad_norm": 0.837469409626879, + "learning_rate": 7.529215358931553e-06, + "loss": 0.3288, + "step": 451 + }, + { + "epoch": 0.15091819699499165, + "grad_norm": 0.8178656765451151, + "learning_rate": 7.545909849749583e-06, + "loss": 0.3168, + "step": 452 + }, + { + "epoch": 0.15125208681135224, + "grad_norm": 0.7437610083713088, + "learning_rate": 7.562604340567613e-06, + "loss": 0.3125, + "step": 453 + }, + { + "epoch": 0.15158597662771287, + "grad_norm": 0.7844830651212958, + "learning_rate": 7.579298831385643e-06, + "loss": 0.3163, + "step": 454 + }, + { + "epoch": 0.15191986644407346, + "grad_norm": 0.7775461860244373, + "learning_rate": 7.595993322203673e-06, + "loss": 0.3149, + "step": 455 + }, + { + "epoch": 0.15225375626043405, + "grad_norm": 0.8428807609544684, + "learning_rate": 7.612687813021703e-06, + "loss": 0.3229, + "step": 456 + }, + { + "epoch": 0.15258764607679465, + "grad_norm": 0.8197169737014166, + "learning_rate": 7.629382303839733e-06, + "loss": 0.3307, + "step": 457 + }, + { + "epoch": 0.15292153589315527, + "grad_norm": 0.7488283411066718, + "learning_rate": 7.646076794657763e-06, + "loss": 0.3309, + "step": 458 + }, + { + "epoch": 0.15325542570951586, + "grad_norm": 0.7999229600734714, + "learning_rate": 7.662771285475793e-06, + "loss": 0.3263, + "step": 459 + }, + { + "epoch": 0.15358931552587646, + "grad_norm": 0.7952537108413845, + "learning_rate": 7.679465776293823e-06, + "loss": 0.3381, + "step": 460 + }, + { + "epoch": 0.15392320534223705, + "grad_norm": 0.8026902493054152, + "learning_rate": 7.696160267111853e-06, + "loss": 0.302, + "step": 461 + }, + { + "epoch": 0.15425709515859767, + "grad_norm": 0.7222194415873212, + "learning_rate": 7.712854757929883e-06, + "loss": 0.3142, + "step": 462 + }, + { + "epoch": 0.15459098497495827, + "grad_norm": 0.7561238717833233, + "learning_rate": 7.729549248747913e-06, + "loss": 0.313, + "step": 463 + }, + { + "epoch": 0.15492487479131886, + "grad_norm": 0.7218297920979148, + "learning_rate": 7.746243739565943e-06, + "loss": 0.3268, + "step": 464 + }, + { + "epoch": 0.15525876460767946, + "grad_norm": 0.7359629533957726, + "learning_rate": 7.762938230383973e-06, + "loss": 0.3228, + "step": 465 + }, + { + "epoch": 0.15559265442404008, + "grad_norm": 0.8237333016205883, + "learning_rate": 7.779632721202003e-06, + "loss": 0.3162, + "step": 466 + }, + { + "epoch": 0.15592654424040067, + "grad_norm": 0.7261624044815037, + "learning_rate": 7.796327212020033e-06, + "loss": 0.3178, + "step": 467 + }, + { + "epoch": 0.15626043405676127, + "grad_norm": 0.7219795971307426, + "learning_rate": 7.813021702838063e-06, + "loss": 0.3116, + "step": 468 + }, + { + "epoch": 0.15659432387312186, + "grad_norm": 0.7608376641874803, + "learning_rate": 7.829716193656093e-06, + "loss": 0.3291, + "step": 469 + }, + { + "epoch": 0.15692821368948248, + "grad_norm": 0.7698450495201847, + "learning_rate": 7.846410684474123e-06, + "loss": 0.3267, + "step": 470 + }, + { + "epoch": 0.15726210350584308, + "grad_norm": 0.6979211939754796, + "learning_rate": 7.863105175292153e-06, + "loss": 0.3249, + "step": 471 + }, + { + "epoch": 0.15759599332220367, + "grad_norm": 0.7824298189873251, + "learning_rate": 7.879799666110184e-06, + "loss": 0.3287, + "step": 472 + }, + { + "epoch": 0.15792988313856426, + "grad_norm": 0.7742505247823801, + "learning_rate": 7.896494156928214e-06, + "loss": 0.324, + "step": 473 + }, + { + "epoch": 0.1582637729549249, + "grad_norm": 0.7663683298440186, + "learning_rate": 7.913188647746244e-06, + "loss": 0.3276, + "step": 474 + }, + { + "epoch": 0.15859766277128548, + "grad_norm": 0.7732806470238048, + "learning_rate": 7.929883138564275e-06, + "loss": 0.3289, + "step": 475 + }, + { + "epoch": 0.15893155258764607, + "grad_norm": 0.7935581431107047, + "learning_rate": 7.946577629382305e-06, + "loss": 0.3168, + "step": 476 + }, + { + "epoch": 0.15926544240400667, + "grad_norm": 0.7582011842364209, + "learning_rate": 7.963272120200335e-06, + "loss": 0.3216, + "step": 477 + }, + { + "epoch": 0.1595993322203673, + "grad_norm": 0.7835950376206476, + "learning_rate": 7.979966611018365e-06, + "loss": 0.3275, + "step": 478 + }, + { + "epoch": 0.15993322203672788, + "grad_norm": 0.8191113916842354, + "learning_rate": 7.996661101836395e-06, + "loss": 0.3255, + "step": 479 + }, + { + "epoch": 0.16026711185308848, + "grad_norm": 0.8532390477742875, + "learning_rate": 8.013355592654425e-06, + "loss": 0.327, + "step": 480 + }, + { + "epoch": 0.16060100166944907, + "grad_norm": 0.7377104308576151, + "learning_rate": 8.030050083472455e-06, + "loss": 0.3236, + "step": 481 + }, + { + "epoch": 0.1609348914858097, + "grad_norm": 0.7461194005400179, + "learning_rate": 8.046744574290485e-06, + "loss": 0.3174, + "step": 482 + }, + { + "epoch": 0.1612687813021703, + "grad_norm": 0.7383037891324972, + "learning_rate": 8.063439065108515e-06, + "loss": 0.3347, + "step": 483 + }, + { + "epoch": 0.16160267111853088, + "grad_norm": 0.7869998510266845, + "learning_rate": 8.080133555926545e-06, + "loss": 0.3116, + "step": 484 + }, + { + "epoch": 0.16193656093489148, + "grad_norm": 0.8230004780699617, + "learning_rate": 8.096828046744575e-06, + "loss": 0.3217, + "step": 485 + }, + { + "epoch": 0.1622704507512521, + "grad_norm": 0.7966028210062216, + "learning_rate": 8.113522537562605e-06, + "loss": 0.325, + "step": 486 + }, + { + "epoch": 0.1626043405676127, + "grad_norm": 0.8116848339609104, + "learning_rate": 8.130217028380635e-06, + "loss": 0.3347, + "step": 487 + }, + { + "epoch": 0.1629382303839733, + "grad_norm": 0.7620085553313396, + "learning_rate": 8.146911519198665e-06, + "loss": 0.3256, + "step": 488 + }, + { + "epoch": 0.16327212020033388, + "grad_norm": 0.7487709440881457, + "learning_rate": 8.163606010016696e-06, + "loss": 0.3338, + "step": 489 + }, + { + "epoch": 0.1636060100166945, + "grad_norm": 0.7459436381574598, + "learning_rate": 8.180300500834726e-06, + "loss": 0.3295, + "step": 490 + }, + { + "epoch": 0.1639398998330551, + "grad_norm": 0.8758962495058791, + "learning_rate": 8.196994991652756e-06, + "loss": 0.3285, + "step": 491 + }, + { + "epoch": 0.1642737896494157, + "grad_norm": 0.7612790096285995, + "learning_rate": 8.213689482470786e-06, + "loss": 0.321, + "step": 492 + }, + { + "epoch": 0.16460767946577629, + "grad_norm": 0.7170522817743062, + "learning_rate": 8.230383973288816e-06, + "loss": 0.3075, + "step": 493 + }, + { + "epoch": 0.1649415692821369, + "grad_norm": 0.7456360107844466, + "learning_rate": 8.247078464106846e-06, + "loss": 0.3158, + "step": 494 + }, + { + "epoch": 0.1652754590984975, + "grad_norm": 0.8337088317659194, + "learning_rate": 8.263772954924876e-06, + "loss": 0.3381, + "step": 495 + }, + { + "epoch": 0.1656093489148581, + "grad_norm": 0.7299368993202087, + "learning_rate": 8.280467445742906e-06, + "loss": 0.317, + "step": 496 + }, + { + "epoch": 0.1659432387312187, + "grad_norm": 0.7528020255965956, + "learning_rate": 8.297161936560936e-06, + "loss": 0.3254, + "step": 497 + }, + { + "epoch": 0.1662771285475793, + "grad_norm": 0.804542907489064, + "learning_rate": 8.313856427378966e-06, + "loss": 0.3366, + "step": 498 + }, + { + "epoch": 0.1666110183639399, + "grad_norm": 0.7790426197148583, + "learning_rate": 8.330550918196996e-06, + "loss": 0.3299, + "step": 499 + }, + { + "epoch": 0.1669449081803005, + "grad_norm": 0.7422911191627818, + "learning_rate": 8.347245409015026e-06, + "loss": 0.329, + "step": 500 + }, + { + "epoch": 0.1672787979966611, + "grad_norm": 0.7390176354224128, + "learning_rate": 8.363939899833056e-06, + "loss": 0.3263, + "step": 501 + }, + { + "epoch": 0.16761268781302172, + "grad_norm": 0.7906202733680489, + "learning_rate": 8.380634390651086e-06, + "loss": 0.3264, + "step": 502 + }, + { + "epoch": 0.1679465776293823, + "grad_norm": 0.7324257544247772, + "learning_rate": 8.397328881469116e-06, + "loss": 0.3226, + "step": 503 + }, + { + "epoch": 0.1682804674457429, + "grad_norm": 0.7623698964691746, + "learning_rate": 8.414023372287146e-06, + "loss": 0.3223, + "step": 504 + }, + { + "epoch": 0.1686143572621035, + "grad_norm": 0.7681011807938551, + "learning_rate": 8.430717863105176e-06, + "loss": 0.3205, + "step": 505 + }, + { + "epoch": 0.16894824707846412, + "grad_norm": 0.7683422881391737, + "learning_rate": 8.447412353923206e-06, + "loss": 0.3331, + "step": 506 + }, + { + "epoch": 0.16928213689482471, + "grad_norm": 0.7438173286243479, + "learning_rate": 8.464106844741236e-06, + "loss": 0.33, + "step": 507 + }, + { + "epoch": 0.1696160267111853, + "grad_norm": 0.6862796224431545, + "learning_rate": 8.480801335559266e-06, + "loss": 0.3248, + "step": 508 + }, + { + "epoch": 0.1699499165275459, + "grad_norm": 0.7881676709650292, + "learning_rate": 8.497495826377296e-06, + "loss": 0.328, + "step": 509 + }, + { + "epoch": 0.17028380634390652, + "grad_norm": 0.7556663417852358, + "learning_rate": 8.514190317195326e-06, + "loss": 0.3249, + "step": 510 + }, + { + "epoch": 0.17061769616026712, + "grad_norm": 0.7403876947827457, + "learning_rate": 8.530884808013356e-06, + "loss": 0.329, + "step": 511 + }, + { + "epoch": 0.1709515859766277, + "grad_norm": 0.7411797106076798, + "learning_rate": 8.547579298831386e-06, + "loss": 0.3237, + "step": 512 + }, + { + "epoch": 0.1712854757929883, + "grad_norm": 0.8426910019423023, + "learning_rate": 8.564273789649416e-06, + "loss": 0.3331, + "step": 513 + }, + { + "epoch": 0.17161936560934893, + "grad_norm": 0.7718590621759173, + "learning_rate": 8.580968280467446e-06, + "loss": 0.3258, + "step": 514 + }, + { + "epoch": 0.17195325542570952, + "grad_norm": 0.8082831855823825, + "learning_rate": 8.597662771285476e-06, + "loss": 0.3243, + "step": 515 + }, + { + "epoch": 0.17228714524207012, + "grad_norm": 0.7593356727846491, + "learning_rate": 8.614357262103506e-06, + "loss": 0.3158, + "step": 516 + }, + { + "epoch": 0.1726210350584307, + "grad_norm": 0.7644500579649327, + "learning_rate": 8.631051752921536e-06, + "loss": 0.3252, + "step": 517 + }, + { + "epoch": 0.17295492487479133, + "grad_norm": 0.7580338086453896, + "learning_rate": 8.647746243739566e-06, + "loss": 0.3299, + "step": 518 + }, + { + "epoch": 0.17328881469115193, + "grad_norm": 0.8313062842945559, + "learning_rate": 8.664440734557596e-06, + "loss": 0.3323, + "step": 519 + }, + { + "epoch": 0.17362270450751252, + "grad_norm": 0.7077128447623422, + "learning_rate": 8.681135225375626e-06, + "loss": 0.335, + "step": 520 + }, + { + "epoch": 0.17395659432387311, + "grad_norm": 0.801035000564132, + "learning_rate": 8.697829716193656e-06, + "loss": 0.3138, + "step": 521 + }, + { + "epoch": 0.1742904841402337, + "grad_norm": 0.70640759580946, + "learning_rate": 8.714524207011686e-06, + "loss": 0.3152, + "step": 522 + }, + { + "epoch": 0.17462437395659433, + "grad_norm": 0.7237106201813894, + "learning_rate": 8.731218697829716e-06, + "loss": 0.3087, + "step": 523 + }, + { + "epoch": 0.17495826377295493, + "grad_norm": 0.8270015338066695, + "learning_rate": 8.747913188647746e-06, + "loss": 0.3267, + "step": 524 + }, + { + "epoch": 0.17529215358931552, + "grad_norm": 0.7725097239241071, + "learning_rate": 8.764607679465778e-06, + "loss": 0.3266, + "step": 525 + }, + { + "epoch": 0.1756260434056761, + "grad_norm": 0.7408863166439572, + "learning_rate": 8.781302170283808e-06, + "loss": 0.3371, + "step": 526 + }, + { + "epoch": 0.17595993322203674, + "grad_norm": 0.8208734300609117, + "learning_rate": 8.797996661101838e-06, + "loss": 0.3154, + "step": 527 + }, + { + "epoch": 0.17629382303839733, + "grad_norm": 0.7678526377966658, + "learning_rate": 8.814691151919868e-06, + "loss": 0.3134, + "step": 528 + }, + { + "epoch": 0.17662771285475792, + "grad_norm": 0.7507224538744631, + "learning_rate": 8.831385642737898e-06, + "loss": 0.3243, + "step": 529 + }, + { + "epoch": 0.17696160267111852, + "grad_norm": 0.7072902503931116, + "learning_rate": 8.848080133555928e-06, + "loss": 0.3142, + "step": 530 + }, + { + "epoch": 0.17729549248747914, + "grad_norm": 0.7655800830392933, + "learning_rate": 8.864774624373958e-06, + "loss": 0.3094, + "step": 531 + }, + { + "epoch": 0.17762938230383973, + "grad_norm": 0.7130994908083609, + "learning_rate": 8.881469115191988e-06, + "loss": 0.3088, + "step": 532 + }, + { + "epoch": 0.17796327212020033, + "grad_norm": 0.7165342511079031, + "learning_rate": 8.898163606010018e-06, + "loss": 0.3284, + "step": 533 + }, + { + "epoch": 0.17829716193656092, + "grad_norm": 0.7497726159425944, + "learning_rate": 8.914858096828048e-06, + "loss": 0.3199, + "step": 534 + }, + { + "epoch": 0.17863105175292154, + "grad_norm": 0.7639013240707101, + "learning_rate": 8.931552587646078e-06, + "loss": 0.3241, + "step": 535 + }, + { + "epoch": 0.17896494156928214, + "grad_norm": 0.7190663003130916, + "learning_rate": 8.948247078464108e-06, + "loss": 0.3237, + "step": 536 + }, + { + "epoch": 0.17929883138564273, + "grad_norm": 0.7102976701316389, + "learning_rate": 8.964941569282138e-06, + "loss": 0.3163, + "step": 537 + }, + { + "epoch": 0.17963272120200333, + "grad_norm": 0.6870634419776727, + "learning_rate": 8.981636060100168e-06, + "loss": 0.332, + "step": 538 + }, + { + "epoch": 0.17996661101836395, + "grad_norm": 0.7340047674343365, + "learning_rate": 8.998330550918198e-06, + "loss": 0.324, + "step": 539 + }, + { + "epoch": 0.18030050083472454, + "grad_norm": 0.740848190587793, + "learning_rate": 9.015025041736228e-06, + "loss": 0.312, + "step": 540 + }, + { + "epoch": 0.18063439065108514, + "grad_norm": 0.7185308360475198, + "learning_rate": 9.031719532554258e-06, + "loss": 0.3093, + "step": 541 + }, + { + "epoch": 0.18096828046744573, + "grad_norm": 0.7381886017497423, + "learning_rate": 9.048414023372288e-06, + "loss": 0.3228, + "step": 542 + }, + { + "epoch": 0.18130217028380635, + "grad_norm": 0.7468127417967011, + "learning_rate": 9.065108514190318e-06, + "loss": 0.3053, + "step": 543 + }, + { + "epoch": 0.18163606010016695, + "grad_norm": 0.7267088568995381, + "learning_rate": 9.081803005008348e-06, + "loss": 0.3172, + "step": 544 + }, + { + "epoch": 0.18196994991652754, + "grad_norm": 0.7358538458677804, + "learning_rate": 9.098497495826378e-06, + "loss": 0.3123, + "step": 545 + }, + { + "epoch": 0.18230383973288813, + "grad_norm": 0.7955340069928836, + "learning_rate": 9.115191986644408e-06, + "loss": 0.3164, + "step": 546 + }, + { + "epoch": 0.18263772954924876, + "grad_norm": 0.7247194315298651, + "learning_rate": 9.131886477462438e-06, + "loss": 0.3299, + "step": 547 + }, + { + "epoch": 0.18297161936560935, + "grad_norm": 0.8063912664449436, + "learning_rate": 9.148580968280468e-06, + "loss": 0.3309, + "step": 548 + }, + { + "epoch": 0.18330550918196994, + "grad_norm": 0.7630892435711264, + "learning_rate": 9.165275459098498e-06, + "loss": 0.3066, + "step": 549 + }, + { + "epoch": 0.18363939899833054, + "grad_norm": 0.7585013930897931, + "learning_rate": 9.181969949916528e-06, + "loss": 0.3227, + "step": 550 + }, + { + "epoch": 0.18397328881469116, + "grad_norm": 0.8280224936512043, + "learning_rate": 9.198664440734558e-06, + "loss": 0.3148, + "step": 551 + }, + { + "epoch": 0.18430717863105175, + "grad_norm": 0.7086949447367906, + "learning_rate": 9.215358931552588e-06, + "loss": 0.3172, + "step": 552 + }, + { + "epoch": 0.18464106844741235, + "grad_norm": 0.7194221484277117, + "learning_rate": 9.232053422370618e-06, + "loss": 0.3197, + "step": 553 + }, + { + "epoch": 0.18497495826377294, + "grad_norm": 0.7403500658764484, + "learning_rate": 9.248747913188648e-06, + "loss": 0.3118, + "step": 554 + }, + { + "epoch": 0.18530884808013356, + "grad_norm": 0.7779107440915423, + "learning_rate": 9.265442404006678e-06, + "loss": 0.3414, + "step": 555 + }, + { + "epoch": 0.18564273789649416, + "grad_norm": 0.6703535986503766, + "learning_rate": 9.282136894824708e-06, + "loss": 0.313, + "step": 556 + }, + { + "epoch": 0.18597662771285475, + "grad_norm": 0.7078900711628437, + "learning_rate": 9.298831385642738e-06, + "loss": 0.3208, + "step": 557 + }, + { + "epoch": 0.18631051752921535, + "grad_norm": 0.8267261385248086, + "learning_rate": 9.315525876460768e-06, + "loss": 0.3193, + "step": 558 + }, + { + "epoch": 0.18664440734557597, + "grad_norm": 0.7579189998201027, + "learning_rate": 9.332220367278798e-06, + "loss": 0.3339, + "step": 559 + }, + { + "epoch": 0.18697829716193656, + "grad_norm": 0.786191922043695, + "learning_rate": 9.348914858096828e-06, + "loss": 0.324, + "step": 560 + }, + { + "epoch": 0.18731218697829716, + "grad_norm": 0.7991805807112369, + "learning_rate": 9.365609348914858e-06, + "loss": 0.3245, + "step": 561 + }, + { + "epoch": 0.18764607679465775, + "grad_norm": 0.7423316925027894, + "learning_rate": 9.382303839732888e-06, + "loss": 0.3175, + "step": 562 + }, + { + "epoch": 0.18797996661101837, + "grad_norm": 0.8035610292681677, + "learning_rate": 9.398998330550918e-06, + "loss": 0.3257, + "step": 563 + }, + { + "epoch": 0.18831385642737897, + "grad_norm": 0.831279527860217, + "learning_rate": 9.415692821368949e-06, + "loss": 0.3146, + "step": 564 + }, + { + "epoch": 0.18864774624373956, + "grad_norm": 0.7457948137337536, + "learning_rate": 9.432387312186979e-06, + "loss": 0.318, + "step": 565 + }, + { + "epoch": 0.18898163606010016, + "grad_norm": 0.7671789660863324, + "learning_rate": 9.449081803005009e-06, + "loss": 0.3138, + "step": 566 + }, + { + "epoch": 0.18931552587646078, + "grad_norm": 0.8688023842692277, + "learning_rate": 9.465776293823039e-06, + "loss": 0.3223, + "step": 567 + }, + { + "epoch": 0.18964941569282137, + "grad_norm": 0.7171048768500374, + "learning_rate": 9.482470784641069e-06, + "loss": 0.3243, + "step": 568 + }, + { + "epoch": 0.18998330550918197, + "grad_norm": 0.8462537328172866, + "learning_rate": 9.499165275459099e-06, + "loss": 0.326, + "step": 569 + }, + { + "epoch": 0.19031719532554256, + "grad_norm": 0.7991013102528451, + "learning_rate": 9.515859766277129e-06, + "loss": 0.3233, + "step": 570 + }, + { + "epoch": 0.19065108514190318, + "grad_norm": 0.7376614267798475, + "learning_rate": 9.532554257095159e-06, + "loss": 0.3065, + "step": 571 + }, + { + "epoch": 0.19098497495826378, + "grad_norm": 0.7868040351914186, + "learning_rate": 9.549248747913189e-06, + "loss": 0.2985, + "step": 572 + }, + { + "epoch": 0.19131886477462437, + "grad_norm": 0.7691605972092043, + "learning_rate": 9.565943238731219e-06, + "loss": 0.3272, + "step": 573 + }, + { + "epoch": 0.19165275459098496, + "grad_norm": 0.7773401992026916, + "learning_rate": 9.582637729549249e-06, + "loss": 0.3141, + "step": 574 + }, + { + "epoch": 0.19198664440734559, + "grad_norm": 0.7834157331676491, + "learning_rate": 9.59933222036728e-06, + "loss": 0.3252, + "step": 575 + }, + { + "epoch": 0.19232053422370618, + "grad_norm": 0.7198788874579221, + "learning_rate": 9.61602671118531e-06, + "loss": 0.3087, + "step": 576 + }, + { + "epoch": 0.19265442404006677, + "grad_norm": 0.721654730193635, + "learning_rate": 9.63272120200334e-06, + "loss": 0.3028, + "step": 577 + }, + { + "epoch": 0.19298831385642737, + "grad_norm": 0.7411476830955602, + "learning_rate": 9.64941569282137e-06, + "loss": 0.3318, + "step": 578 + }, + { + "epoch": 0.193322203672788, + "grad_norm": 0.7502640964656166, + "learning_rate": 9.6661101836394e-06, + "loss": 0.3201, + "step": 579 + }, + { + "epoch": 0.19365609348914858, + "grad_norm": 0.6976177939046787, + "learning_rate": 9.68280467445743e-06, + "loss": 0.3243, + "step": 580 + }, + { + "epoch": 0.19398998330550918, + "grad_norm": 0.7050176806810524, + "learning_rate": 9.69949916527546e-06, + "loss": 0.3026, + "step": 581 + }, + { + "epoch": 0.19432387312186977, + "grad_norm": 0.7283130332705378, + "learning_rate": 9.71619365609349e-06, + "loss": 0.3171, + "step": 582 + }, + { + "epoch": 0.1946577629382304, + "grad_norm": 0.6903358621221012, + "learning_rate": 9.73288814691152e-06, + "loss": 0.3192, + "step": 583 + }, + { + "epoch": 0.194991652754591, + "grad_norm": 0.7775600186257118, + "learning_rate": 9.74958263772955e-06, + "loss": 0.3107, + "step": 584 + }, + { + "epoch": 0.19532554257095158, + "grad_norm": 0.7114664511303405, + "learning_rate": 9.76627712854758e-06, + "loss": 0.3207, + "step": 585 + }, + { + "epoch": 0.19565943238731218, + "grad_norm": 0.7413168130522473, + "learning_rate": 9.78297161936561e-06, + "loss": 0.3252, + "step": 586 + }, + { + "epoch": 0.1959933222036728, + "grad_norm": 0.763607282926004, + "learning_rate": 9.79966611018364e-06, + "loss": 0.3148, + "step": 587 + }, + { + "epoch": 0.1963272120200334, + "grad_norm": 0.6985928231477169, + "learning_rate": 9.81636060100167e-06, + "loss": 0.3088, + "step": 588 + }, + { + "epoch": 0.196661101836394, + "grad_norm": 0.7328207500828827, + "learning_rate": 9.8330550918197e-06, + "loss": 0.3197, + "step": 589 + }, + { + "epoch": 0.19699499165275458, + "grad_norm": 0.762407634478981, + "learning_rate": 9.84974958263773e-06, + "loss": 0.3265, + "step": 590 + }, + { + "epoch": 0.1973288814691152, + "grad_norm": 0.6940607229447971, + "learning_rate": 9.86644407345576e-06, + "loss": 0.3168, + "step": 591 + }, + { + "epoch": 0.1976627712854758, + "grad_norm": 0.785712216630924, + "learning_rate": 9.88313856427379e-06, + "loss": 0.3305, + "step": 592 + }, + { + "epoch": 0.1979966611018364, + "grad_norm": 0.6722822710794846, + "learning_rate": 9.89983305509182e-06, + "loss": 0.3212, + "step": 593 + }, + { + "epoch": 0.19833055091819699, + "grad_norm": 0.8232515008170899, + "learning_rate": 9.91652754590985e-06, + "loss": 0.3254, + "step": 594 + }, + { + "epoch": 0.1986644407345576, + "grad_norm": 0.7716441885493642, + "learning_rate": 9.93322203672788e-06, + "loss": 0.3247, + "step": 595 + }, + { + "epoch": 0.1989983305509182, + "grad_norm": 0.6974882203548125, + "learning_rate": 9.94991652754591e-06, + "loss": 0.3009, + "step": 596 + }, + { + "epoch": 0.1993322203672788, + "grad_norm": 0.733854194044891, + "learning_rate": 9.96661101836394e-06, + "loss": 0.3218, + "step": 597 + }, + { + "epoch": 0.1996661101836394, + "grad_norm": 0.7980874175431234, + "learning_rate": 9.98330550918197e-06, + "loss": 0.3281, + "step": 598 + }, + { + "epoch": 0.2, + "grad_norm": 0.7043941458533823, + "learning_rate": 1e-05, + "loss": 0.3086, + "step": 599 + }, + { + "epoch": 0.2003338898163606, + "grad_norm": 0.8566520002555572, + "learning_rate": 9.999999151012927e-06, + "loss": 0.3288, + "step": 600 + }, + { + "epoch": 0.2006677796327212, + "grad_norm": 0.6549504853641911, + "learning_rate": 9.999996604051993e-06, + "loss": 0.3191, + "step": 601 + }, + { + "epoch": 0.2010016694490818, + "grad_norm": 0.696848696347495, + "learning_rate": 9.999992359118065e-06, + "loss": 0.3222, + "step": 602 + }, + { + "epoch": 0.20133555926544242, + "grad_norm": 0.7859875498774094, + "learning_rate": 9.999986416212582e-06, + "loss": 0.3165, + "step": 603 + }, + { + "epoch": 0.201669449081803, + "grad_norm": 0.7187193594328152, + "learning_rate": 9.999978775337564e-06, + "loss": 0.3223, + "step": 604 + }, + { + "epoch": 0.2020033388981636, + "grad_norm": 0.7045752547463391, + "learning_rate": 9.999969436495606e-06, + "loss": 0.3166, + "step": 605 + }, + { + "epoch": 0.2023372287145242, + "grad_norm": 0.8217046464113188, + "learning_rate": 9.99995839968988e-06, + "loss": 0.3221, + "step": 606 + }, + { + "epoch": 0.20267111853088482, + "grad_norm": 0.6949124441786446, + "learning_rate": 9.999945664924132e-06, + "loss": 0.3175, + "step": 607 + }, + { + "epoch": 0.2030050083472454, + "grad_norm": 0.7312484291025569, + "learning_rate": 9.999931232202689e-06, + "loss": 0.3282, + "step": 608 + }, + { + "epoch": 0.203338898163606, + "grad_norm": 0.7373779871547909, + "learning_rate": 9.99991510153045e-06, + "loss": 0.3226, + "step": 609 + }, + { + "epoch": 0.2036727879799666, + "grad_norm": 0.7011515106348708, + "learning_rate": 9.999897272912895e-06, + "loss": 0.3156, + "step": 610 + }, + { + "epoch": 0.20400667779632722, + "grad_norm": 0.7239664802656571, + "learning_rate": 9.999877746356074e-06, + "loss": 0.3345, + "step": 611 + }, + { + "epoch": 0.20434056761268782, + "grad_norm": 0.7408338513712528, + "learning_rate": 9.999856521866625e-06, + "loss": 0.3125, + "step": 612 + }, + { + "epoch": 0.2046744574290484, + "grad_norm": 0.715921130880272, + "learning_rate": 9.999833599451751e-06, + "loss": 0.3149, + "step": 613 + }, + { + "epoch": 0.205008347245409, + "grad_norm": 0.7113906095284361, + "learning_rate": 9.999808979119239e-06, + "loss": 0.3199, + "step": 614 + }, + { + "epoch": 0.20534223706176963, + "grad_norm": 0.7548609216393225, + "learning_rate": 9.999782660877448e-06, + "loss": 0.315, + "step": 615 + }, + { + "epoch": 0.20567612687813022, + "grad_norm": 0.6838507430621388, + "learning_rate": 9.999754644735316e-06, + "loss": 0.3178, + "step": 616 + }, + { + "epoch": 0.20601001669449082, + "grad_norm": 0.7061892687066811, + "learning_rate": 9.999724930702358e-06, + "loss": 0.3138, + "step": 617 + }, + { + "epoch": 0.2063439065108514, + "grad_norm": 0.7098014307269261, + "learning_rate": 9.999693518788662e-06, + "loss": 0.3279, + "step": 618 + }, + { + "epoch": 0.20667779632721203, + "grad_norm": 0.7042688009543399, + "learning_rate": 9.999660409004896e-06, + "loss": 0.3151, + "step": 619 + }, + { + "epoch": 0.20701168614357263, + "grad_norm": 0.7001782566785403, + "learning_rate": 9.999625601362308e-06, + "loss": 0.3242, + "step": 620 + }, + { + "epoch": 0.20734557595993322, + "grad_norm": 0.681982137706268, + "learning_rate": 9.999589095872715e-06, + "loss": 0.311, + "step": 621 + }, + { + "epoch": 0.20767946577629381, + "grad_norm": 0.6640110233762736, + "learning_rate": 9.999550892548515e-06, + "loss": 0.3224, + "step": 622 + }, + { + "epoch": 0.20801335559265444, + "grad_norm": 0.7220821271155453, + "learning_rate": 9.999510991402681e-06, + "loss": 0.3228, + "step": 623 + }, + { + "epoch": 0.20834724540901503, + "grad_norm": 0.7100077787957213, + "learning_rate": 9.999469392448764e-06, + "loss": 0.3253, + "step": 624 + }, + { + "epoch": 0.20868113522537562, + "grad_norm": 0.743147072666147, + "learning_rate": 9.99942609570089e-06, + "loss": 0.327, + "step": 625 + }, + { + "epoch": 0.20901502504173622, + "grad_norm": 0.6712602151429252, + "learning_rate": 9.999381101173765e-06, + "loss": 0.3155, + "step": 626 + }, + { + "epoch": 0.20934891485809684, + "grad_norm": 0.7239763297534358, + "learning_rate": 9.999334408882662e-06, + "loss": 0.3128, + "step": 627 + }, + { + "epoch": 0.20968280467445743, + "grad_norm": 0.7905946176461562, + "learning_rate": 9.999286018843447e-06, + "loss": 0.3081, + "step": 628 + }, + { + "epoch": 0.21001669449081803, + "grad_norm": 0.7020979277377108, + "learning_rate": 9.999235931072544e-06, + "loss": 0.314, + "step": 629 + }, + { + "epoch": 0.21035058430717862, + "grad_norm": 0.7333591366879596, + "learning_rate": 9.999184145586968e-06, + "loss": 0.313, + "step": 630 + }, + { + "epoch": 0.21068447412353924, + "grad_norm": 0.7369589376002899, + "learning_rate": 9.999130662404306e-06, + "loss": 0.3072, + "step": 631 + }, + { + "epoch": 0.21101836393989984, + "grad_norm": 0.7548484006421701, + "learning_rate": 9.999075481542717e-06, + "loss": 0.3265, + "step": 632 + }, + { + "epoch": 0.21135225375626043, + "grad_norm": 0.7003334090642283, + "learning_rate": 9.99901860302094e-06, + "loss": 0.3221, + "step": 633 + }, + { + "epoch": 0.21168614357262103, + "grad_norm": 0.7385223410698561, + "learning_rate": 9.998960026858296e-06, + "loss": 0.3248, + "step": 634 + }, + { + "epoch": 0.21202003338898165, + "grad_norm": 0.7089176434664283, + "learning_rate": 9.99889975307467e-06, + "loss": 0.3241, + "step": 635 + }, + { + "epoch": 0.21235392320534224, + "grad_norm": 0.6676316673228386, + "learning_rate": 9.998837781690535e-06, + "loss": 0.303, + "step": 636 + }, + { + "epoch": 0.21268781302170284, + "grad_norm": 0.7494344764536006, + "learning_rate": 9.998774112726935e-06, + "loss": 0.307, + "step": 637 + }, + { + "epoch": 0.21302170283806343, + "grad_norm": 0.6760983461146339, + "learning_rate": 9.998708746205493e-06, + "loss": 0.3117, + "step": 638 + }, + { + "epoch": 0.21335559265442405, + "grad_norm": 0.6490852494052426, + "learning_rate": 9.998641682148404e-06, + "loss": 0.3118, + "step": 639 + }, + { + "epoch": 0.21368948247078465, + "grad_norm": 0.7500778276241074, + "learning_rate": 9.998572920578444e-06, + "loss": 0.3121, + "step": 640 + }, + { + "epoch": 0.21402337228714524, + "grad_norm": 0.7373509072386051, + "learning_rate": 9.998502461518966e-06, + "loss": 0.3247, + "step": 641 + }, + { + "epoch": 0.21435726210350584, + "grad_norm": 0.6719035133919844, + "learning_rate": 9.998430304993899e-06, + "loss": 0.3083, + "step": 642 + }, + { + "epoch": 0.21469115191986646, + "grad_norm": 0.6740638066296981, + "learning_rate": 9.99835645102774e-06, + "loss": 0.3135, + "step": 643 + }, + { + "epoch": 0.21502504173622705, + "grad_norm": 0.6878968472246421, + "learning_rate": 9.998280899645575e-06, + "loss": 0.3079, + "step": 644 + }, + { + "epoch": 0.21535893155258765, + "grad_norm": 0.710194399655766, + "learning_rate": 9.99820365087306e-06, + "loss": 0.3243, + "step": 645 + }, + { + "epoch": 0.21569282136894824, + "grad_norm": 0.7041017078946005, + "learning_rate": 9.998124704736426e-06, + "loss": 0.3138, + "step": 646 + }, + { + "epoch": 0.21602671118530886, + "grad_norm": 0.6756827800948253, + "learning_rate": 9.998044061262486e-06, + "loss": 0.3097, + "step": 647 + }, + { + "epoch": 0.21636060100166946, + "grad_norm": 0.6997497722368279, + "learning_rate": 9.997961720478623e-06, + "loss": 0.3146, + "step": 648 + }, + { + "epoch": 0.21669449081803005, + "grad_norm": 0.6676015844917575, + "learning_rate": 9.997877682412804e-06, + "loss": 0.3186, + "step": 649 + }, + { + "epoch": 0.21702838063439064, + "grad_norm": 0.6514765310700045, + "learning_rate": 9.997791947093562e-06, + "loss": 0.3167, + "step": 650 + }, + { + "epoch": 0.21736227045075127, + "grad_norm": 0.6730498459139304, + "learning_rate": 9.997704514550017e-06, + "loss": 0.3104, + "step": 651 + }, + { + "epoch": 0.21769616026711186, + "grad_norm": 0.7013020302409152, + "learning_rate": 9.997615384811857e-06, + "loss": 0.3131, + "step": 652 + }, + { + "epoch": 0.21803005008347245, + "grad_norm": 0.7061119682035909, + "learning_rate": 9.997524557909353e-06, + "loss": 0.3215, + "step": 653 + }, + { + "epoch": 0.21836393989983305, + "grad_norm": 0.6274055151177157, + "learning_rate": 9.997432033873348e-06, + "loss": 0.3112, + "step": 654 + }, + { + "epoch": 0.21869782971619364, + "grad_norm": 0.7367993580620047, + "learning_rate": 9.997337812735263e-06, + "loss": 0.3157, + "step": 655 + }, + { + "epoch": 0.21903171953255426, + "grad_norm": 0.6816534896035203, + "learning_rate": 9.997241894527096e-06, + "loss": 0.3309, + "step": 656 + }, + { + "epoch": 0.21936560934891486, + "grad_norm": 0.689512097755835, + "learning_rate": 9.997144279281416e-06, + "loss": 0.3155, + "step": 657 + }, + { + "epoch": 0.21969949916527545, + "grad_norm": 0.7151953715076534, + "learning_rate": 9.99704496703138e-06, + "loss": 0.3127, + "step": 658 + }, + { + "epoch": 0.22003338898163605, + "grad_norm": 0.6777423266862721, + "learning_rate": 9.996943957810708e-06, + "loss": 0.3107, + "step": 659 + }, + { + "epoch": 0.22036727879799667, + "grad_norm": 0.6938602813630486, + "learning_rate": 9.996841251653701e-06, + "loss": 0.3057, + "step": 660 + }, + { + "epoch": 0.22070116861435726, + "grad_norm": 0.7134775745150277, + "learning_rate": 9.996736848595244e-06, + "loss": 0.3252, + "step": 661 + }, + { + "epoch": 0.22103505843071786, + "grad_norm": 0.7140810806644553, + "learning_rate": 9.996630748670788e-06, + "loss": 0.324, + "step": 662 + }, + { + "epoch": 0.22136894824707845, + "grad_norm": 0.6930323813471048, + "learning_rate": 9.996522951916363e-06, + "loss": 0.3175, + "step": 663 + }, + { + "epoch": 0.22170283806343907, + "grad_norm": 0.6489368882137267, + "learning_rate": 9.996413458368578e-06, + "loss": 0.3148, + "step": 664 + }, + { + "epoch": 0.22203672787979967, + "grad_norm": 0.6704245414166704, + "learning_rate": 9.996302268064615e-06, + "loss": 0.3136, + "step": 665 + }, + { + "epoch": 0.22237061769616026, + "grad_norm": 0.6675247501853886, + "learning_rate": 9.996189381042236e-06, + "loss": 0.3081, + "step": 666 + }, + { + "epoch": 0.22270450751252086, + "grad_norm": 0.6637069468988419, + "learning_rate": 9.996074797339774e-06, + "loss": 0.3157, + "step": 667 + }, + { + "epoch": 0.22303839732888148, + "grad_norm": 0.649834904016236, + "learning_rate": 9.995958516996143e-06, + "loss": 0.3114, + "step": 668 + }, + { + "epoch": 0.22337228714524207, + "grad_norm": 0.6794453662434838, + "learning_rate": 9.99584054005083e-06, + "loss": 0.319, + "step": 669 + }, + { + "epoch": 0.22370617696160267, + "grad_norm": 0.6821690899476152, + "learning_rate": 9.9957208665439e-06, + "loss": 0.3253, + "step": 670 + }, + { + "epoch": 0.22404006677796326, + "grad_norm": 0.6968487649536192, + "learning_rate": 9.995599496515996e-06, + "loss": 0.3082, + "step": 671 + }, + { + "epoch": 0.22437395659432388, + "grad_norm": 0.6942819598756247, + "learning_rate": 9.995476430008328e-06, + "loss": 0.3213, + "step": 672 + }, + { + "epoch": 0.22470784641068448, + "grad_norm": 0.6676574077742479, + "learning_rate": 9.995351667062697e-06, + "loss": 0.3269, + "step": 673 + }, + { + "epoch": 0.22504173622704507, + "grad_norm": 0.6422313726383432, + "learning_rate": 9.995225207721465e-06, + "loss": 0.3195, + "step": 674 + }, + { + "epoch": 0.22537562604340566, + "grad_norm": 0.739089667400378, + "learning_rate": 9.995097052027578e-06, + "loss": 0.3076, + "step": 675 + }, + { + "epoch": 0.22570951585976629, + "grad_norm": 0.6657673642361472, + "learning_rate": 9.994967200024562e-06, + "loss": 0.31, + "step": 676 + }, + { + "epoch": 0.22604340567612688, + "grad_norm": 0.6531981810152021, + "learning_rate": 9.99483565175651e-06, + "loss": 0.3151, + "step": 677 + }, + { + "epoch": 0.22637729549248747, + "grad_norm": 0.6292105662195325, + "learning_rate": 9.994702407268093e-06, + "loss": 0.3144, + "step": 678 + }, + { + "epoch": 0.22671118530884807, + "grad_norm": 0.6793498707800965, + "learning_rate": 9.994567466604566e-06, + "loss": 0.3197, + "step": 679 + }, + { + "epoch": 0.2270450751252087, + "grad_norm": 0.6702697674456445, + "learning_rate": 9.99443082981175e-06, + "loss": 0.3153, + "step": 680 + }, + { + "epoch": 0.22737896494156928, + "grad_norm": 0.6950803319161349, + "learning_rate": 9.994292496936048e-06, + "loss": 0.3216, + "step": 681 + }, + { + "epoch": 0.22771285475792988, + "grad_norm": 0.6155384938859741, + "learning_rate": 9.994152468024435e-06, + "loss": 0.3119, + "step": 682 + }, + { + "epoch": 0.22804674457429047, + "grad_norm": 0.6997721567877282, + "learning_rate": 9.994010743124466e-06, + "loss": 0.3087, + "step": 683 + }, + { + "epoch": 0.2283806343906511, + "grad_norm": 0.673304007752359, + "learning_rate": 9.99386732228427e-06, + "loss": 0.3135, + "step": 684 + }, + { + "epoch": 0.2287145242070117, + "grad_norm": 0.655845424765876, + "learning_rate": 9.993722205552552e-06, + "loss": 0.3137, + "step": 685 + }, + { + "epoch": 0.22904841402337228, + "grad_norm": 0.6332802663031555, + "learning_rate": 9.993575392978592e-06, + "loss": 0.3065, + "step": 686 + }, + { + "epoch": 0.22938230383973288, + "grad_norm": 0.6669194798584023, + "learning_rate": 9.993426884612246e-06, + "loss": 0.3193, + "step": 687 + }, + { + "epoch": 0.2297161936560935, + "grad_norm": 0.6754541501944406, + "learning_rate": 9.99327668050395e-06, + "loss": 0.3243, + "step": 688 + }, + { + "epoch": 0.2300500834724541, + "grad_norm": 0.6330173690472923, + "learning_rate": 9.993124780704707e-06, + "loss": 0.3113, + "step": 689 + }, + { + "epoch": 0.2303839732888147, + "grad_norm": 0.617507112093279, + "learning_rate": 9.992971185266107e-06, + "loss": 0.3119, + "step": 690 + }, + { + "epoch": 0.23071786310517528, + "grad_norm": 0.7081635896196787, + "learning_rate": 9.992815894240309e-06, + "loss": 0.3185, + "step": 691 + }, + { + "epoch": 0.2310517529215359, + "grad_norm": 0.6492912628501856, + "learning_rate": 9.992658907680046e-06, + "loss": 0.3087, + "step": 692 + }, + { + "epoch": 0.2313856427378965, + "grad_norm": 0.6300914240785725, + "learning_rate": 9.992500225638633e-06, + "loss": 0.3215, + "step": 693 + }, + { + "epoch": 0.2317195325542571, + "grad_norm": 0.7000910671590395, + "learning_rate": 9.992339848169956e-06, + "loss": 0.3099, + "step": 694 + }, + { + "epoch": 0.23205342237061768, + "grad_norm": 0.7129853499539469, + "learning_rate": 9.992177775328479e-06, + "loss": 0.3038, + "step": 695 + }, + { + "epoch": 0.2323873121869783, + "grad_norm": 0.7185842769867339, + "learning_rate": 9.992014007169241e-06, + "loss": 0.3163, + "step": 696 + }, + { + "epoch": 0.2327212020033389, + "grad_norm": 0.6906021291785269, + "learning_rate": 9.991848543747857e-06, + "loss": 0.3138, + "step": 697 + }, + { + "epoch": 0.2330550918196995, + "grad_norm": 0.7417109459605487, + "learning_rate": 9.991681385120515e-06, + "loss": 0.3121, + "step": 698 + }, + { + "epoch": 0.2333889816360601, + "grad_norm": 0.6858080307617006, + "learning_rate": 9.991512531343987e-06, + "loss": 0.3219, + "step": 699 + }, + { + "epoch": 0.2337228714524207, + "grad_norm": 0.7028720079517522, + "learning_rate": 9.991341982475609e-06, + "loss": 0.3193, + "step": 700 + }, + { + "epoch": 0.2340567612687813, + "grad_norm": 0.6792095163230769, + "learning_rate": 9.991169738573301e-06, + "loss": 0.3216, + "step": 701 + }, + { + "epoch": 0.2343906510851419, + "grad_norm": 0.6407541669286567, + "learning_rate": 9.990995799695557e-06, + "loss": 0.3007, + "step": 702 + }, + { + "epoch": 0.2347245409015025, + "grad_norm": 0.6644490393729044, + "learning_rate": 9.990820165901444e-06, + "loss": 0.3028, + "step": 703 + }, + { + "epoch": 0.23505843071786311, + "grad_norm": 0.6587611650702211, + "learning_rate": 9.990642837250607e-06, + "loss": 0.3033, + "step": 704 + }, + { + "epoch": 0.2353923205342237, + "grad_norm": 0.6513428732911056, + "learning_rate": 9.990463813803266e-06, + "loss": 0.3084, + "step": 705 + }, + { + "epoch": 0.2357262103505843, + "grad_norm": 0.7349681287692489, + "learning_rate": 9.990283095620216e-06, + "loss": 0.3215, + "step": 706 + }, + { + "epoch": 0.2360601001669449, + "grad_norm": 0.6512825190837345, + "learning_rate": 9.99010068276283e-06, + "loss": 0.3123, + "step": 707 + }, + { + "epoch": 0.23639398998330552, + "grad_norm": 0.613325839160056, + "learning_rate": 9.989916575293051e-06, + "loss": 0.3215, + "step": 708 + }, + { + "epoch": 0.2367278797996661, + "grad_norm": 0.6772892470644201, + "learning_rate": 9.989730773273404e-06, + "loss": 0.3193, + "step": 709 + }, + { + "epoch": 0.2370617696160267, + "grad_norm": 0.6877502994373447, + "learning_rate": 9.989543276766983e-06, + "loss": 0.31, + "step": 710 + }, + { + "epoch": 0.2373956594323873, + "grad_norm": 0.6277755783349489, + "learning_rate": 9.989354085837464e-06, + "loss": 0.3089, + "step": 711 + }, + { + "epoch": 0.23772954924874792, + "grad_norm": 0.6399727451873912, + "learning_rate": 9.989163200549095e-06, + "loss": 0.314, + "step": 712 + }, + { + "epoch": 0.23806343906510852, + "grad_norm": 0.6596146572219671, + "learning_rate": 9.988970620966698e-06, + "loss": 0.3066, + "step": 713 + }, + { + "epoch": 0.2383973288814691, + "grad_norm": 0.674518105255004, + "learning_rate": 9.988776347155673e-06, + "loss": 0.3162, + "step": 714 + }, + { + "epoch": 0.2387312186978297, + "grad_norm": 0.6532326307999126, + "learning_rate": 9.988580379181997e-06, + "loss": 0.3065, + "step": 715 + }, + { + "epoch": 0.23906510851419033, + "grad_norm": 0.6744074017659445, + "learning_rate": 9.988382717112213e-06, + "loss": 0.3217, + "step": 716 + }, + { + "epoch": 0.23939899833055092, + "grad_norm": 0.6598376440763581, + "learning_rate": 9.988183361013452e-06, + "loss": 0.3163, + "step": 717 + }, + { + "epoch": 0.23973288814691152, + "grad_norm": 0.6674898112049668, + "learning_rate": 9.987982310953412e-06, + "loss": 0.3174, + "step": 718 + }, + { + "epoch": 0.2400667779632721, + "grad_norm": 0.636869408850705, + "learning_rate": 9.987779567000369e-06, + "loss": 0.3083, + "step": 719 + }, + { + "epoch": 0.24040066777963273, + "grad_norm": 0.6487769825101677, + "learning_rate": 9.987575129223174e-06, + "loss": 0.3146, + "step": 720 + }, + { + "epoch": 0.24073455759599333, + "grad_norm": 0.5945286371869306, + "learning_rate": 9.987368997691253e-06, + "loss": 0.2994, + "step": 721 + }, + { + "epoch": 0.24106844741235392, + "grad_norm": 0.62935352996601, + "learning_rate": 9.987161172474607e-06, + "loss": 0.3213, + "step": 722 + }, + { + "epoch": 0.24140233722871451, + "grad_norm": 0.6352457311886005, + "learning_rate": 9.986951653643811e-06, + "loss": 0.3021, + "step": 723 + }, + { + "epoch": 0.24173622704507514, + "grad_norm": 0.6164153524135821, + "learning_rate": 9.98674044127002e-06, + "loss": 0.3071, + "step": 724 + }, + { + "epoch": 0.24207011686143573, + "grad_norm": 0.6281362542845271, + "learning_rate": 9.986527535424956e-06, + "loss": 0.3249, + "step": 725 + }, + { + "epoch": 0.24240400667779632, + "grad_norm": 0.6293486946764626, + "learning_rate": 9.986312936180924e-06, + "loss": 0.3114, + "step": 726 + }, + { + "epoch": 0.24273789649415692, + "grad_norm": 0.6631557157756965, + "learning_rate": 9.9860966436108e-06, + "loss": 0.3174, + "step": 727 + }, + { + "epoch": 0.24307178631051754, + "grad_norm": 0.6340649747639411, + "learning_rate": 9.985878657788037e-06, + "loss": 0.32, + "step": 728 + }, + { + "epoch": 0.24340567612687813, + "grad_norm": 0.5873245960456505, + "learning_rate": 9.985658978786659e-06, + "loss": 0.3118, + "step": 729 + }, + { + "epoch": 0.24373956594323873, + "grad_norm": 0.6561509827370153, + "learning_rate": 9.985437606681269e-06, + "loss": 0.3176, + "step": 730 + }, + { + "epoch": 0.24407345575959932, + "grad_norm": 0.649633394985232, + "learning_rate": 9.985214541547046e-06, + "loss": 0.3098, + "step": 731 + }, + { + "epoch": 0.24440734557595994, + "grad_norm": 0.6215244635153744, + "learning_rate": 9.98498978345974e-06, + "loss": 0.3226, + "step": 732 + }, + { + "epoch": 0.24474123539232054, + "grad_norm": 0.622067647920881, + "learning_rate": 9.984763332495677e-06, + "loss": 0.3233, + "step": 733 + }, + { + "epoch": 0.24507512520868113, + "grad_norm": 0.6357396995753083, + "learning_rate": 9.98453518873176e-06, + "loss": 0.3077, + "step": 734 + }, + { + "epoch": 0.24540901502504173, + "grad_norm": 0.6407492332582393, + "learning_rate": 9.984305352245465e-06, + "loss": 0.3283, + "step": 735 + }, + { + "epoch": 0.24574290484140235, + "grad_norm": 0.6568086741042303, + "learning_rate": 9.984073823114842e-06, + "loss": 0.3142, + "step": 736 + }, + { + "epoch": 0.24607679465776294, + "grad_norm": 0.6275340335519496, + "learning_rate": 9.983840601418518e-06, + "loss": 0.3049, + "step": 737 + }, + { + "epoch": 0.24641068447412354, + "grad_norm": 0.6463023555674137, + "learning_rate": 9.983605687235695e-06, + "loss": 0.3086, + "step": 738 + }, + { + "epoch": 0.24674457429048413, + "grad_norm": 0.6397985010827647, + "learning_rate": 9.983369080646148e-06, + "loss": 0.3115, + "step": 739 + }, + { + "epoch": 0.24707846410684475, + "grad_norm": 0.645749279799714, + "learning_rate": 9.983130781730225e-06, + "loss": 0.3305, + "step": 740 + }, + { + "epoch": 0.24741235392320535, + "grad_norm": 0.6175934132973404, + "learning_rate": 9.982890790568855e-06, + "loss": 0.3112, + "step": 741 + }, + { + "epoch": 0.24774624373956594, + "grad_norm": 0.654112066136407, + "learning_rate": 9.982649107243537e-06, + "loss": 0.321, + "step": 742 + }, + { + "epoch": 0.24808013355592654, + "grad_norm": 0.6746394103518467, + "learning_rate": 9.982405731836343e-06, + "loss": 0.3093, + "step": 743 + }, + { + "epoch": 0.24841402337228716, + "grad_norm": 0.6370308472030909, + "learning_rate": 9.982160664429921e-06, + "loss": 0.3166, + "step": 744 + }, + { + "epoch": 0.24874791318864775, + "grad_norm": 0.6659386399377829, + "learning_rate": 9.9819139051075e-06, + "loss": 0.3207, + "step": 745 + }, + { + "epoch": 0.24908180300500835, + "grad_norm": 0.6139691141648983, + "learning_rate": 9.981665453952874e-06, + "loss": 0.3052, + "step": 746 + }, + { + "epoch": 0.24941569282136894, + "grad_norm": 0.6666767354518693, + "learning_rate": 9.981415311050417e-06, + "loss": 0.3094, + "step": 747 + }, + { + "epoch": 0.24974958263772956, + "grad_norm": 0.6954025237312208, + "learning_rate": 9.981163476485077e-06, + "loss": 0.3279, + "step": 748 + }, + { + "epoch": 0.2500834724540901, + "grad_norm": 0.6493768950456869, + "learning_rate": 9.980909950342374e-06, + "loss": 0.3044, + "step": 749 + }, + { + "epoch": 0.25041736227045075, + "grad_norm": 0.7062045469172029, + "learning_rate": 9.980654732708404e-06, + "loss": 0.3031, + "step": 750 + }, + { + "epoch": 0.25075125208681137, + "grad_norm": 0.6319424816517013, + "learning_rate": 9.980397823669841e-06, + "loss": 0.3108, + "step": 751 + }, + { + "epoch": 0.25108514190317194, + "grad_norm": 0.6085947420392623, + "learning_rate": 9.980139223313926e-06, + "loss": 0.3002, + "step": 752 + }, + { + "epoch": 0.25141903171953256, + "grad_norm": 0.6777482648226619, + "learning_rate": 9.979878931728479e-06, + "loss": 0.3154, + "step": 753 + }, + { + "epoch": 0.2517529215358932, + "grad_norm": 0.6635467843446102, + "learning_rate": 9.979616949001896e-06, + "loss": 0.3131, + "step": 754 + }, + { + "epoch": 0.25208681135225375, + "grad_norm": 0.6430022723790265, + "learning_rate": 9.979353275223145e-06, + "loss": 0.3142, + "step": 755 + }, + { + "epoch": 0.25242070116861437, + "grad_norm": 0.649219566836954, + "learning_rate": 9.979087910481765e-06, + "loss": 0.3149, + "step": 756 + }, + { + "epoch": 0.25275459098497494, + "grad_norm": 0.6175932866065273, + "learning_rate": 9.978820854867875e-06, + "loss": 0.325, + "step": 757 + }, + { + "epoch": 0.25308848080133556, + "grad_norm": 0.648710607088272, + "learning_rate": 9.978552108472164e-06, + "loss": 0.3216, + "step": 758 + }, + { + "epoch": 0.2534223706176962, + "grad_norm": 0.6296825401342997, + "learning_rate": 9.978281671385899e-06, + "loss": 0.31, + "step": 759 + }, + { + "epoch": 0.25375626043405675, + "grad_norm": 0.6589505148775748, + "learning_rate": 9.978009543700918e-06, + "loss": 0.3133, + "step": 760 + }, + { + "epoch": 0.25409015025041737, + "grad_norm": 0.5972495169454712, + "learning_rate": 9.977735725509632e-06, + "loss": 0.3184, + "step": 761 + }, + { + "epoch": 0.254424040066778, + "grad_norm": 0.6527937848923318, + "learning_rate": 9.977460216905032e-06, + "loss": 0.3118, + "step": 762 + }, + { + "epoch": 0.25475792988313856, + "grad_norm": 0.6606823569793125, + "learning_rate": 9.977183017980677e-06, + "loss": 0.3184, + "step": 763 + }, + { + "epoch": 0.2550918196994992, + "grad_norm": 0.6380265286594664, + "learning_rate": 9.976904128830705e-06, + "loss": 0.3201, + "step": 764 + }, + { + "epoch": 0.25542570951585974, + "grad_norm": 0.593403542000823, + "learning_rate": 9.976623549549821e-06, + "loss": 0.3123, + "step": 765 + }, + { + "epoch": 0.25575959933222037, + "grad_norm": 0.6414387729044886, + "learning_rate": 9.97634128023331e-06, + "loss": 0.3037, + "step": 766 + }, + { + "epoch": 0.256093489148581, + "grad_norm": 0.6552073957370249, + "learning_rate": 9.976057320977033e-06, + "loss": 0.3115, + "step": 767 + }, + { + "epoch": 0.25642737896494155, + "grad_norm": 0.5947098775665278, + "learning_rate": 9.975771671877414e-06, + "loss": 0.2975, + "step": 768 + }, + { + "epoch": 0.2567612687813022, + "grad_norm": 0.6684257304677536, + "learning_rate": 9.975484333031465e-06, + "loss": 0.3258, + "step": 769 + }, + { + "epoch": 0.2570951585976628, + "grad_norm": 0.6252469609537954, + "learning_rate": 9.97519530453676e-06, + "loss": 0.3153, + "step": 770 + }, + { + "epoch": 0.25742904841402336, + "grad_norm": 0.6446100117261462, + "learning_rate": 9.974904586491455e-06, + "loss": 0.3242, + "step": 771 + }, + { + "epoch": 0.257762938230384, + "grad_norm": 0.5885286198221695, + "learning_rate": 9.974612178994271e-06, + "loss": 0.3102, + "step": 772 + }, + { + "epoch": 0.25809682804674455, + "grad_norm": 0.6142941293567612, + "learning_rate": 9.974318082144515e-06, + "loss": 0.3113, + "step": 773 + }, + { + "epoch": 0.2584307178631052, + "grad_norm": 0.6050100134080159, + "learning_rate": 9.974022296042055e-06, + "loss": 0.3045, + "step": 774 + }, + { + "epoch": 0.2587646076794658, + "grad_norm": 0.619261228852053, + "learning_rate": 9.973724820787343e-06, + "loss": 0.3137, + "step": 775 + }, + { + "epoch": 0.25909849749582636, + "grad_norm": 0.5876762285672034, + "learning_rate": 9.973425656481395e-06, + "loss": 0.2973, + "step": 776 + }, + { + "epoch": 0.259432387312187, + "grad_norm": 0.6059762310691418, + "learning_rate": 9.973124803225811e-06, + "loss": 0.3121, + "step": 777 + }, + { + "epoch": 0.2597662771285476, + "grad_norm": 0.6205353342018597, + "learning_rate": 9.972822261122757e-06, + "loss": 0.3075, + "step": 778 + }, + { + "epoch": 0.2601001669449082, + "grad_norm": 0.6119035328080568, + "learning_rate": 9.97251803027497e-06, + "loss": 0.3112, + "step": 779 + }, + { + "epoch": 0.2604340567612688, + "grad_norm": 0.5868391573659103, + "learning_rate": 9.972212110785776e-06, + "loss": 0.3186, + "step": 780 + }, + { + "epoch": 0.26076794657762936, + "grad_norm": 0.6162528166794374, + "learning_rate": 9.971904502759053e-06, + "loss": 0.2982, + "step": 781 + }, + { + "epoch": 0.26110183639399, + "grad_norm": 0.6406567336283016, + "learning_rate": 9.97159520629927e-06, + "loss": 0.2994, + "step": 782 + }, + { + "epoch": 0.2614357262103506, + "grad_norm": 0.5978567040711715, + "learning_rate": 9.971284221511459e-06, + "loss": 0.31, + "step": 783 + }, + { + "epoch": 0.26176961602671117, + "grad_norm": 0.6379382965047393, + "learning_rate": 9.970971548501229e-06, + "loss": 0.3122, + "step": 784 + }, + { + "epoch": 0.2621035058430718, + "grad_norm": 0.6589036243685998, + "learning_rate": 9.970657187374762e-06, + "loss": 0.3132, + "step": 785 + }, + { + "epoch": 0.26243739565943236, + "grad_norm": 0.6274802396042567, + "learning_rate": 9.970341138238817e-06, + "loss": 0.3108, + "step": 786 + }, + { + "epoch": 0.262771285475793, + "grad_norm": 0.5780001244013581, + "learning_rate": 9.970023401200719e-06, + "loss": 0.3057, + "step": 787 + }, + { + "epoch": 0.2631051752921536, + "grad_norm": 0.673173569250029, + "learning_rate": 9.969703976368368e-06, + "loss": 0.3228, + "step": 788 + }, + { + "epoch": 0.26343906510851417, + "grad_norm": 0.6565293937159798, + "learning_rate": 9.969382863850245e-06, + "loss": 0.3065, + "step": 789 + }, + { + "epoch": 0.2637729549248748, + "grad_norm": 0.6409399011729932, + "learning_rate": 9.969060063755395e-06, + "loss": 0.3108, + "step": 790 + }, + { + "epoch": 0.2641068447412354, + "grad_norm": 0.8177205780834107, + "learning_rate": 9.968735576193438e-06, + "loss": 0.3079, + "step": 791 + }, + { + "epoch": 0.264440734557596, + "grad_norm": 0.625131535488349, + "learning_rate": 9.96840940127457e-06, + "loss": 0.3079, + "step": 792 + }, + { + "epoch": 0.2647746243739566, + "grad_norm": 0.6628614020219976, + "learning_rate": 9.968081539109557e-06, + "loss": 0.3098, + "step": 793 + }, + { + "epoch": 0.26510851419031717, + "grad_norm": 0.646207616905766, + "learning_rate": 9.967751989809741e-06, + "loss": 0.3205, + "step": 794 + }, + { + "epoch": 0.2654424040066778, + "grad_norm": 0.6001489997881362, + "learning_rate": 9.967420753487035e-06, + "loss": 0.304, + "step": 795 + }, + { + "epoch": 0.2657762938230384, + "grad_norm": 0.5836823359225947, + "learning_rate": 9.967087830253923e-06, + "loss": 0.2975, + "step": 796 + }, + { + "epoch": 0.266110183639399, + "grad_norm": 0.6732142764197346, + "learning_rate": 9.966753220223466e-06, + "loss": 0.3145, + "step": 797 + }, + { + "epoch": 0.2664440734557596, + "grad_norm": 0.6377990101324601, + "learning_rate": 9.966416923509296e-06, + "loss": 0.3119, + "step": 798 + }, + { + "epoch": 0.2667779632721202, + "grad_norm": 0.645652722369334, + "learning_rate": 9.966078940225617e-06, + "loss": 0.3086, + "step": 799 + }, + { + "epoch": 0.2671118530884808, + "grad_norm": 0.6717049183906165, + "learning_rate": 9.965739270487207e-06, + "loss": 0.3063, + "step": 800 + }, + { + "epoch": 0.2674457429048414, + "grad_norm": 0.6158282076094603, + "learning_rate": 9.965397914409414e-06, + "loss": 0.2976, + "step": 801 + }, + { + "epoch": 0.267779632721202, + "grad_norm": 0.6001733467038547, + "learning_rate": 9.965054872108161e-06, + "loss": 0.3088, + "step": 802 + }, + { + "epoch": 0.2681135225375626, + "grad_norm": 0.6282539065846117, + "learning_rate": 9.964710143699948e-06, + "loss": 0.3048, + "step": 803 + }, + { + "epoch": 0.2684474123539232, + "grad_norm": 0.6701301013931206, + "learning_rate": 9.964363729301838e-06, + "loss": 0.3045, + "step": 804 + }, + { + "epoch": 0.2687813021702838, + "grad_norm": 0.681220178422193, + "learning_rate": 9.964015629031472e-06, + "loss": 0.3158, + "step": 805 + }, + { + "epoch": 0.2691151919866444, + "grad_norm": 0.6350987182465024, + "learning_rate": 9.963665843007066e-06, + "loss": 0.3129, + "step": 806 + }, + { + "epoch": 0.26944908180300503, + "grad_norm": 0.6276708598655996, + "learning_rate": 9.963314371347403e-06, + "loss": 0.307, + "step": 807 + }, + { + "epoch": 0.2697829716193656, + "grad_norm": 0.6311569423244825, + "learning_rate": 9.96296121417184e-06, + "loss": 0.3261, + "step": 808 + }, + { + "epoch": 0.2701168614357262, + "grad_norm": 0.580271587038022, + "learning_rate": 9.962606371600311e-06, + "loss": 0.3019, + "step": 809 + }, + { + "epoch": 0.2704507512520868, + "grad_norm": 0.6559436191915577, + "learning_rate": 9.962249843753317e-06, + "loss": 0.3072, + "step": 810 + }, + { + "epoch": 0.2707846410684474, + "grad_norm": 0.600062062263448, + "learning_rate": 9.961891630751931e-06, + "loss": 0.3216, + "step": 811 + }, + { + "epoch": 0.27111853088480803, + "grad_norm": 0.6273016362187429, + "learning_rate": 9.961531732717803e-06, + "loss": 0.3293, + "step": 812 + }, + { + "epoch": 0.2714524207011686, + "grad_norm": 0.5759138448802362, + "learning_rate": 9.961170149773152e-06, + "loss": 0.2996, + "step": 813 + }, + { + "epoch": 0.2717863105175292, + "grad_norm": 0.6384149258474171, + "learning_rate": 9.960806882040767e-06, + "loss": 0.3156, + "step": 814 + }, + { + "epoch": 0.27212020033388984, + "grad_norm": 0.5913620181404787, + "learning_rate": 9.960441929644017e-06, + "loss": 0.3122, + "step": 815 + }, + { + "epoch": 0.2724540901502504, + "grad_norm": 0.584316980706626, + "learning_rate": 9.960075292706834e-06, + "loss": 0.3033, + "step": 816 + }, + { + "epoch": 0.272787979966611, + "grad_norm": 0.6279215889504319, + "learning_rate": 9.959706971353727e-06, + "loss": 0.2968, + "step": 817 + }, + { + "epoch": 0.2731218697829716, + "grad_norm": 0.6378668514547856, + "learning_rate": 9.959336965709776e-06, + "loss": 0.3125, + "step": 818 + }, + { + "epoch": 0.2734557595993322, + "grad_norm": 0.6083284073674683, + "learning_rate": 9.958965275900632e-06, + "loss": 0.3148, + "step": 819 + }, + { + "epoch": 0.27378964941569284, + "grad_norm": 0.6149421651109955, + "learning_rate": 9.958591902052521e-06, + "loss": 0.3036, + "step": 820 + }, + { + "epoch": 0.2741235392320534, + "grad_norm": 0.6279228412457, + "learning_rate": 9.958216844292238e-06, + "loss": 0.3044, + "step": 821 + }, + { + "epoch": 0.274457429048414, + "grad_norm": 0.6144754730683342, + "learning_rate": 9.957840102747149e-06, + "loss": 0.3077, + "step": 822 + }, + { + "epoch": 0.27479131886477465, + "grad_norm": 0.6555367147344426, + "learning_rate": 9.957461677545198e-06, + "loss": 0.309, + "step": 823 + }, + { + "epoch": 0.2751252086811352, + "grad_norm": 0.6358275469095108, + "learning_rate": 9.95708156881489e-06, + "loss": 0.3169, + "step": 824 + }, + { + "epoch": 0.27545909849749584, + "grad_norm": 0.5472475213285918, + "learning_rate": 9.956699776685312e-06, + "loss": 0.306, + "step": 825 + }, + { + "epoch": 0.2757929883138564, + "grad_norm": 0.6730166870878292, + "learning_rate": 9.95631630128612e-06, + "loss": 0.306, + "step": 826 + }, + { + "epoch": 0.276126878130217, + "grad_norm": 0.6384414369767303, + "learning_rate": 9.955931142747534e-06, + "loss": 0.3062, + "step": 827 + }, + { + "epoch": 0.27646076794657765, + "grad_norm": 0.6081049338004099, + "learning_rate": 9.95554430120036e-06, + "loss": 0.3052, + "step": 828 + }, + { + "epoch": 0.2767946577629382, + "grad_norm": 0.6118840611801132, + "learning_rate": 9.955155776775961e-06, + "loss": 0.3121, + "step": 829 + }, + { + "epoch": 0.27712854757929883, + "grad_norm": 0.6045958092288646, + "learning_rate": 9.95476556960628e-06, + "loss": 0.3144, + "step": 830 + }, + { + "epoch": 0.27746243739565946, + "grad_norm": 0.6546322527742332, + "learning_rate": 9.954373679823831e-06, + "loss": 0.3228, + "step": 831 + }, + { + "epoch": 0.27779632721202, + "grad_norm": 0.6703196495250673, + "learning_rate": 9.953980107561695e-06, + "loss": 0.3144, + "step": 832 + }, + { + "epoch": 0.27813021702838064, + "grad_norm": 0.5728891650899754, + "learning_rate": 9.95358485295353e-06, + "loss": 0.2989, + "step": 833 + }, + { + "epoch": 0.2784641068447412, + "grad_norm": 0.600667356911053, + "learning_rate": 9.95318791613356e-06, + "loss": 0.3092, + "step": 834 + }, + { + "epoch": 0.27879799666110183, + "grad_norm": 0.6252146905608394, + "learning_rate": 9.952789297236583e-06, + "loss": 0.3249, + "step": 835 + }, + { + "epoch": 0.27913188647746245, + "grad_norm": 0.561479696676962, + "learning_rate": 9.952388996397968e-06, + "loss": 0.2882, + "step": 836 + }, + { + "epoch": 0.279465776293823, + "grad_norm": 0.6031264684709128, + "learning_rate": 9.951987013753657e-06, + "loss": 0.3142, + "step": 837 + }, + { + "epoch": 0.27979966611018364, + "grad_norm": 0.5958619592081246, + "learning_rate": 9.951583349440157e-06, + "loss": 0.3132, + "step": 838 + }, + { + "epoch": 0.28013355592654426, + "grad_norm": 0.6542396307096576, + "learning_rate": 9.951178003594556e-06, + "loss": 0.3268, + "step": 839 + }, + { + "epoch": 0.28046744574290483, + "grad_norm": 0.6135038168458617, + "learning_rate": 9.950770976354503e-06, + "loss": 0.312, + "step": 840 + }, + { + "epoch": 0.28080133555926545, + "grad_norm": 0.6157053571661313, + "learning_rate": 9.950362267858225e-06, + "loss": 0.308, + "step": 841 + }, + { + "epoch": 0.281135225375626, + "grad_norm": 0.6330173355934376, + "learning_rate": 9.949951878244514e-06, + "loss": 0.313, + "step": 842 + }, + { + "epoch": 0.28146911519198664, + "grad_norm": 0.6123887797042682, + "learning_rate": 9.949539807652739e-06, + "loss": 0.3077, + "step": 843 + }, + { + "epoch": 0.28180300500834726, + "grad_norm": 0.630502207334047, + "learning_rate": 9.949126056222837e-06, + "loss": 0.3136, + "step": 844 + }, + { + "epoch": 0.28213689482470783, + "grad_norm": 0.638922108869299, + "learning_rate": 9.948710624095315e-06, + "loss": 0.2998, + "step": 845 + }, + { + "epoch": 0.28247078464106845, + "grad_norm": 0.650500842814557, + "learning_rate": 9.948293511411251e-06, + "loss": 0.3034, + "step": 846 + }, + { + "epoch": 0.2828046744574291, + "grad_norm": 0.5814351959854284, + "learning_rate": 9.947874718312295e-06, + "loss": 0.3087, + "step": 847 + }, + { + "epoch": 0.28313856427378964, + "grad_norm": 0.6287239204801921, + "learning_rate": 9.947454244940668e-06, + "loss": 0.3163, + "step": 848 + }, + { + "epoch": 0.28347245409015026, + "grad_norm": 0.6019418257870377, + "learning_rate": 9.947032091439159e-06, + "loss": 0.3139, + "step": 849 + }, + { + "epoch": 0.2838063439065108, + "grad_norm": 0.5979390239562166, + "learning_rate": 9.94660825795113e-06, + "loss": 0.3047, + "step": 850 + }, + { + "epoch": 0.28414023372287145, + "grad_norm": 0.6302662571322311, + "learning_rate": 9.946182744620512e-06, + "loss": 0.3089, + "step": 851 + }, + { + "epoch": 0.28447412353923207, + "grad_norm": 0.5961305187813384, + "learning_rate": 9.945755551591808e-06, + "loss": 0.3038, + "step": 852 + }, + { + "epoch": 0.28480801335559264, + "grad_norm": 0.6297686619540529, + "learning_rate": 9.94532667901009e-06, + "loss": 0.3175, + "step": 853 + }, + { + "epoch": 0.28514190317195326, + "grad_norm": 0.6571487678194401, + "learning_rate": 9.944896127020999e-06, + "loss": 0.3058, + "step": 854 + }, + { + "epoch": 0.2854757929883139, + "grad_norm": 0.6006638467764017, + "learning_rate": 9.944463895770754e-06, + "loss": 0.3135, + "step": 855 + }, + { + "epoch": 0.28580968280467445, + "grad_norm": 0.5971973789235672, + "learning_rate": 9.944029985406131e-06, + "loss": 0.304, + "step": 856 + }, + { + "epoch": 0.28614357262103507, + "grad_norm": 0.6234322478998101, + "learning_rate": 9.94359439607449e-06, + "loss": 0.3077, + "step": 857 + }, + { + "epoch": 0.28647746243739564, + "grad_norm": 0.6068124603154574, + "learning_rate": 9.943157127923751e-06, + "loss": 0.3129, + "step": 858 + }, + { + "epoch": 0.28681135225375626, + "grad_norm": 0.5895365093104074, + "learning_rate": 9.942718181102408e-06, + "loss": 0.323, + "step": 859 + }, + { + "epoch": 0.2871452420701169, + "grad_norm": 0.6455484559534321, + "learning_rate": 9.94227755575953e-06, + "loss": 0.315, + "step": 860 + }, + { + "epoch": 0.28747913188647745, + "grad_norm": 0.616420947904619, + "learning_rate": 9.941835252044744e-06, + "loss": 0.2995, + "step": 861 + }, + { + "epoch": 0.28781302170283807, + "grad_norm": 0.634640993915691, + "learning_rate": 9.941391270108261e-06, + "loss": 0.2973, + "step": 862 + }, + { + "epoch": 0.2881469115191987, + "grad_norm": 0.6259253158626111, + "learning_rate": 9.940945610100849e-06, + "loss": 0.3058, + "step": 863 + }, + { + "epoch": 0.28848080133555926, + "grad_norm": 0.631191316803027, + "learning_rate": 9.940498272173855e-06, + "loss": 0.306, + "step": 864 + }, + { + "epoch": 0.2888146911519199, + "grad_norm": 0.5998640078762677, + "learning_rate": 9.94004925647919e-06, + "loss": 0.3104, + "step": 865 + }, + { + "epoch": 0.28914858096828044, + "grad_norm": 0.6156671378039831, + "learning_rate": 9.939598563169342e-06, + "loss": 0.3171, + "step": 866 + }, + { + "epoch": 0.28948247078464107, + "grad_norm": 0.6819628803818827, + "learning_rate": 9.93914619239736e-06, + "loss": 0.3297, + "step": 867 + }, + { + "epoch": 0.2898163606010017, + "grad_norm": 0.661258225940603, + "learning_rate": 9.938692144316869e-06, + "loss": 0.3213, + "step": 868 + }, + { + "epoch": 0.29015025041736225, + "grad_norm": 0.5900726763916676, + "learning_rate": 9.93823641908206e-06, + "loss": 0.3038, + "step": 869 + }, + { + "epoch": 0.2904841402337229, + "grad_norm": 0.6212445636277458, + "learning_rate": 9.937779016847698e-06, + "loss": 0.3027, + "step": 870 + }, + { + "epoch": 0.2908180300500835, + "grad_norm": 0.6539321538885854, + "learning_rate": 9.937319937769108e-06, + "loss": 0.3099, + "step": 871 + }, + { + "epoch": 0.29115191986644406, + "grad_norm": 0.6428136114041385, + "learning_rate": 9.936859182002199e-06, + "loss": 0.3177, + "step": 872 + }, + { + "epoch": 0.2914858096828047, + "grad_norm": 0.6129270552971511, + "learning_rate": 9.936396749703435e-06, + "loss": 0.301, + "step": 873 + }, + { + "epoch": 0.29181969949916525, + "grad_norm": 0.6265365291294976, + "learning_rate": 9.935932641029857e-06, + "loss": 0.3184, + "step": 874 + }, + { + "epoch": 0.2921535893155259, + "grad_norm": 0.6189004042144778, + "learning_rate": 9.935466856139075e-06, + "loss": 0.3031, + "step": 875 + }, + { + "epoch": 0.2924874791318865, + "grad_norm": 0.5998489396255959, + "learning_rate": 9.934999395189269e-06, + "loss": 0.3019, + "step": 876 + }, + { + "epoch": 0.29282136894824706, + "grad_norm": 0.6248384280710946, + "learning_rate": 9.93453025833918e-06, + "loss": 0.3198, + "step": 877 + }, + { + "epoch": 0.2931552587646077, + "grad_norm": 0.6036881338884938, + "learning_rate": 9.934059445748134e-06, + "loss": 0.3003, + "step": 878 + }, + { + "epoch": 0.2934891485809683, + "grad_norm": 0.6197791294956829, + "learning_rate": 9.933586957576006e-06, + "loss": 0.3011, + "step": 879 + }, + { + "epoch": 0.2938230383973289, + "grad_norm": 0.5968907195054848, + "learning_rate": 9.93311279398326e-06, + "loss": 0.3039, + "step": 880 + }, + { + "epoch": 0.2941569282136895, + "grad_norm": 0.5826857878604631, + "learning_rate": 9.932636955130913e-06, + "loss": 0.3142, + "step": 881 + }, + { + "epoch": 0.29449081803005006, + "grad_norm": 0.611459702189933, + "learning_rate": 9.93215944118056e-06, + "loss": 0.3099, + "step": 882 + }, + { + "epoch": 0.2948247078464107, + "grad_norm": 0.6198352916800216, + "learning_rate": 9.931680252294363e-06, + "loss": 0.3056, + "step": 883 + }, + { + "epoch": 0.2951585976627713, + "grad_norm": 0.5806707618585026, + "learning_rate": 9.931199388635048e-06, + "loss": 0.3043, + "step": 884 + }, + { + "epoch": 0.29549248747913187, + "grad_norm": 0.6073718073926797, + "learning_rate": 9.930716850365919e-06, + "loss": 0.3067, + "step": 885 + }, + { + "epoch": 0.2958263772954925, + "grad_norm": 0.5652545339563153, + "learning_rate": 9.93023263765084e-06, + "loss": 0.3107, + "step": 886 + }, + { + "epoch": 0.2961602671118531, + "grad_norm": 0.6151798723288843, + "learning_rate": 9.92974675065425e-06, + "loss": 0.3034, + "step": 887 + }, + { + "epoch": 0.2964941569282137, + "grad_norm": 0.5966499336220145, + "learning_rate": 9.92925918954115e-06, + "loss": 0.3148, + "step": 888 + }, + { + "epoch": 0.2968280467445743, + "grad_norm": 0.6286876192494795, + "learning_rate": 9.928769954477117e-06, + "loss": 0.3003, + "step": 889 + }, + { + "epoch": 0.29716193656093487, + "grad_norm": 0.5894046699170157, + "learning_rate": 9.928279045628291e-06, + "loss": 0.3043, + "step": 890 + }, + { + "epoch": 0.2974958263772955, + "grad_norm": 0.612360918976436, + "learning_rate": 9.927786463161382e-06, + "loss": 0.2976, + "step": 891 + }, + { + "epoch": 0.2978297161936561, + "grad_norm": 0.6301047605113264, + "learning_rate": 9.927292207243667e-06, + "loss": 0.3052, + "step": 892 + }, + { + "epoch": 0.2981636060100167, + "grad_norm": 0.6092746864864261, + "learning_rate": 9.926796278042997e-06, + "loss": 0.3194, + "step": 893 + }, + { + "epoch": 0.2984974958263773, + "grad_norm": 0.6156641290646087, + "learning_rate": 9.92629867572778e-06, + "loss": 0.3031, + "step": 894 + }, + { + "epoch": 0.2988313856427379, + "grad_norm": 0.6322363950690639, + "learning_rate": 9.925799400467008e-06, + "loss": 0.3018, + "step": 895 + }, + { + "epoch": 0.2991652754590985, + "grad_norm": 0.6528206115577453, + "learning_rate": 9.925298452430225e-06, + "loss": 0.3063, + "step": 896 + }, + { + "epoch": 0.2994991652754591, + "grad_norm": 0.6718744478615776, + "learning_rate": 9.924795831787555e-06, + "loss": 0.304, + "step": 897 + }, + { + "epoch": 0.2998330550918197, + "grad_norm": 0.6102327634060157, + "learning_rate": 9.924291538709683e-06, + "loss": 0.3085, + "step": 898 + }, + { + "epoch": 0.3001669449081803, + "grad_norm": 0.5892192124082466, + "learning_rate": 9.923785573367865e-06, + "loss": 0.3018, + "step": 899 + }, + { + "epoch": 0.3005008347245409, + "grad_norm": 0.6393837053129542, + "learning_rate": 9.923277935933925e-06, + "loss": 0.3146, + "step": 900 + }, + { + "epoch": 0.3008347245409015, + "grad_norm": 0.6399687824980628, + "learning_rate": 9.922768626580251e-06, + "loss": 0.3123, + "step": 901 + }, + { + "epoch": 0.3011686143572621, + "grad_norm": 0.6305766133789136, + "learning_rate": 9.922257645479803e-06, + "loss": 0.3011, + "step": 902 + }, + { + "epoch": 0.30150250417362273, + "grad_norm": 0.6198973397404037, + "learning_rate": 9.92174499280611e-06, + "loss": 0.308, + "step": 903 + }, + { + "epoch": 0.3018363939899833, + "grad_norm": 0.6283096314652501, + "learning_rate": 9.921230668733265e-06, + "loss": 0.3165, + "step": 904 + }, + { + "epoch": 0.3021702838063439, + "grad_norm": 0.6425794965454609, + "learning_rate": 9.920714673435931e-06, + "loss": 0.32, + "step": 905 + }, + { + "epoch": 0.3025041736227045, + "grad_norm": 0.610463604147464, + "learning_rate": 9.920197007089334e-06, + "loss": 0.3063, + "step": 906 + }, + { + "epoch": 0.3028380634390651, + "grad_norm": 0.5806825161814771, + "learning_rate": 9.919677669869272e-06, + "loss": 0.3128, + "step": 907 + }, + { + "epoch": 0.30317195325542573, + "grad_norm": 0.6162549772294891, + "learning_rate": 9.919156661952113e-06, + "loss": 0.3173, + "step": 908 + }, + { + "epoch": 0.3035058430717863, + "grad_norm": 0.615303368956951, + "learning_rate": 9.918633983514782e-06, + "loss": 0.3153, + "step": 909 + }, + { + "epoch": 0.3038397328881469, + "grad_norm": 0.580419554740561, + "learning_rate": 9.918109634734783e-06, + "loss": 0.316, + "step": 910 + }, + { + "epoch": 0.30417362270450754, + "grad_norm": 0.5855338492866154, + "learning_rate": 9.917583615790178e-06, + "loss": 0.3047, + "step": 911 + }, + { + "epoch": 0.3045075125208681, + "grad_norm": 0.5713157341293394, + "learning_rate": 9.917055926859606e-06, + "loss": 0.3013, + "step": 912 + }, + { + "epoch": 0.30484140233722873, + "grad_norm": 0.6390439175167857, + "learning_rate": 9.916526568122262e-06, + "loss": 0.3118, + "step": 913 + }, + { + "epoch": 0.3051752921535893, + "grad_norm": 0.6396809779271058, + "learning_rate": 9.915995539757918e-06, + "loss": 0.3109, + "step": 914 + }, + { + "epoch": 0.3055091819699499, + "grad_norm": 0.5898667073427104, + "learning_rate": 9.915462841946905e-06, + "loss": 0.2977, + "step": 915 + }, + { + "epoch": 0.30584307178631054, + "grad_norm": 0.5991208827032682, + "learning_rate": 9.914928474870125e-06, + "loss": 0.3136, + "step": 916 + }, + { + "epoch": 0.3061769616026711, + "grad_norm": 0.5747940856902647, + "learning_rate": 9.914392438709049e-06, + "loss": 0.3131, + "step": 917 + }, + { + "epoch": 0.3065108514190317, + "grad_norm": 0.5797945195579927, + "learning_rate": 9.913854733645708e-06, + "loss": 0.3198, + "step": 918 + }, + { + "epoch": 0.3068447412353923, + "grad_norm": 0.5756691860699671, + "learning_rate": 9.913315359862705e-06, + "loss": 0.2973, + "step": 919 + }, + { + "epoch": 0.3071786310517529, + "grad_norm": 0.5888275992106462, + "learning_rate": 9.912774317543212e-06, + "loss": 0.2877, + "step": 920 + }, + { + "epoch": 0.30751252086811354, + "grad_norm": 0.6096492910396711, + "learning_rate": 9.912231606870961e-06, + "loss": 0.3116, + "step": 921 + }, + { + "epoch": 0.3078464106844741, + "grad_norm": 0.6075529630126482, + "learning_rate": 9.911687228030253e-06, + "loss": 0.3006, + "step": 922 + }, + { + "epoch": 0.3081803005008347, + "grad_norm": 0.5939477640034415, + "learning_rate": 9.91114118120596e-06, + "loss": 0.2942, + "step": 923 + }, + { + "epoch": 0.30851419031719535, + "grad_norm": 0.6137987887456721, + "learning_rate": 9.910593466583512e-06, + "loss": 0.3032, + "step": 924 + }, + { + "epoch": 0.3088480801335559, + "grad_norm": 0.6183827825714345, + "learning_rate": 9.910044084348913e-06, + "loss": 0.3002, + "step": 925 + }, + { + "epoch": 0.30918196994991654, + "grad_norm": 0.5939518104266519, + "learning_rate": 9.90949303468873e-06, + "loss": 0.3002, + "step": 926 + }, + { + "epoch": 0.3095158597662771, + "grad_norm": 0.5878305945902575, + "learning_rate": 9.908940317790096e-06, + "loss": 0.2941, + "step": 927 + }, + { + "epoch": 0.3098497495826377, + "grad_norm": 0.6162066212665044, + "learning_rate": 9.908385933840711e-06, + "loss": 0.3076, + "step": 928 + }, + { + "epoch": 0.31018363939899835, + "grad_norm": 0.5794159626420173, + "learning_rate": 9.907829883028842e-06, + "loss": 0.3091, + "step": 929 + }, + { + "epoch": 0.3105175292153589, + "grad_norm": 0.6025058756667004, + "learning_rate": 9.907272165543317e-06, + "loss": 0.3082, + "step": 930 + }, + { + "epoch": 0.31085141903171953, + "grad_norm": 0.6285631554193102, + "learning_rate": 9.90671278157354e-06, + "loss": 0.3147, + "step": 931 + }, + { + "epoch": 0.31118530884808016, + "grad_norm": 0.6031757119907064, + "learning_rate": 9.906151731309472e-06, + "loss": 0.3082, + "step": 932 + }, + { + "epoch": 0.3115191986644407, + "grad_norm": 0.5741482015967575, + "learning_rate": 9.905589014941641e-06, + "loss": 0.3069, + "step": 933 + }, + { + "epoch": 0.31185308848080134, + "grad_norm": 0.606115608559058, + "learning_rate": 9.905024632661146e-06, + "loss": 0.3086, + "step": 934 + }, + { + "epoch": 0.3121869782971619, + "grad_norm": 0.5903380094158948, + "learning_rate": 9.904458584659645e-06, + "loss": 0.3115, + "step": 935 + }, + { + "epoch": 0.31252086811352253, + "grad_norm": 0.6088421013459345, + "learning_rate": 9.903890871129367e-06, + "loss": 0.3169, + "step": 936 + }, + { + "epoch": 0.31285475792988315, + "grad_norm": 0.6083470577556701, + "learning_rate": 9.903321492263106e-06, + "loss": 0.2959, + "step": 937 + }, + { + "epoch": 0.3131886477462437, + "grad_norm": 0.587984830223083, + "learning_rate": 9.902750448254214e-06, + "loss": 0.3053, + "step": 938 + }, + { + "epoch": 0.31352253756260434, + "grad_norm": 0.6160429401184823, + "learning_rate": 9.902177739296621e-06, + "loss": 0.3155, + "step": 939 + }, + { + "epoch": 0.31385642737896496, + "grad_norm": 0.5779814539377545, + "learning_rate": 9.901603365584816e-06, + "loss": 0.2984, + "step": 940 + }, + { + "epoch": 0.31419031719532553, + "grad_norm": 0.623950988120686, + "learning_rate": 9.901027327313847e-06, + "loss": 0.3212, + "step": 941 + }, + { + "epoch": 0.31452420701168615, + "grad_norm": 0.6489183720666848, + "learning_rate": 9.90044962467934e-06, + "loss": 0.3018, + "step": 942 + }, + { + "epoch": 0.3148580968280467, + "grad_norm": 0.5996934093973265, + "learning_rate": 9.89987025787748e-06, + "loss": 0.3033, + "step": 943 + }, + { + "epoch": 0.31519198664440734, + "grad_norm": 0.6183589240973589, + "learning_rate": 9.899289227105012e-06, + "loss": 0.3145, + "step": 944 + }, + { + "epoch": 0.31552587646076796, + "grad_norm": 0.6657754519961233, + "learning_rate": 9.898706532559252e-06, + "loss": 0.2932, + "step": 945 + }, + { + "epoch": 0.31585976627712853, + "grad_norm": 0.6390446615691743, + "learning_rate": 9.898122174438083e-06, + "loss": 0.2985, + "step": 946 + }, + { + "epoch": 0.31619365609348915, + "grad_norm": 0.6037443942866899, + "learning_rate": 9.897536152939949e-06, + "loss": 0.3115, + "step": 947 + }, + { + "epoch": 0.3165275459098498, + "grad_norm": 0.6073307634860344, + "learning_rate": 9.89694846826386e-06, + "loss": 0.2978, + "step": 948 + }, + { + "epoch": 0.31686143572621034, + "grad_norm": 0.6146007705282659, + "learning_rate": 9.896359120609389e-06, + "loss": 0.2917, + "step": 949 + }, + { + "epoch": 0.31719532554257096, + "grad_norm": 0.648693822695636, + "learning_rate": 9.895768110176677e-06, + "loss": 0.3142, + "step": 950 + }, + { + "epoch": 0.3175292153589315, + "grad_norm": 0.5717551881140968, + "learning_rate": 9.89517543716643e-06, + "loss": 0.2938, + "step": 951 + }, + { + "epoch": 0.31786310517529215, + "grad_norm": 0.6448068620681516, + "learning_rate": 9.894581101779912e-06, + "loss": 0.313, + "step": 952 + }, + { + "epoch": 0.31819699499165277, + "grad_norm": 0.6047593205071065, + "learning_rate": 9.893985104218959e-06, + "loss": 0.3048, + "step": 953 + }, + { + "epoch": 0.31853088480801334, + "grad_norm": 0.5719658064999276, + "learning_rate": 9.893387444685969e-06, + "loss": 0.3018, + "step": 954 + }, + { + "epoch": 0.31886477462437396, + "grad_norm": 0.6099649552932656, + "learning_rate": 9.892788123383901e-06, + "loss": 0.308, + "step": 955 + }, + { + "epoch": 0.3191986644407346, + "grad_norm": 0.5653706030914476, + "learning_rate": 9.892187140516285e-06, + "loss": 0.3056, + "step": 956 + }, + { + "epoch": 0.31953255425709515, + "grad_norm": 0.5736228273815476, + "learning_rate": 9.891584496287212e-06, + "loss": 0.3087, + "step": 957 + }, + { + "epoch": 0.31986644407345577, + "grad_norm": 0.6269217658930452, + "learning_rate": 9.890980190901336e-06, + "loss": 0.3286, + "step": 958 + }, + { + "epoch": 0.32020033388981634, + "grad_norm": 0.6344691770807388, + "learning_rate": 9.890374224563872e-06, + "loss": 0.2923, + "step": 959 + }, + { + "epoch": 0.32053422370617696, + "grad_norm": 0.5804374329830767, + "learning_rate": 9.889766597480609e-06, + "loss": 0.3101, + "step": 960 + }, + { + "epoch": 0.3208681135225376, + "grad_norm": 0.6168739282716879, + "learning_rate": 9.88915730985789e-06, + "loss": 0.3029, + "step": 961 + }, + { + "epoch": 0.32120200333889815, + "grad_norm": 0.6208506214144326, + "learning_rate": 9.888546361902629e-06, + "loss": 0.3098, + "step": 962 + }, + { + "epoch": 0.32153589315525877, + "grad_norm": 0.5859414011396221, + "learning_rate": 9.887933753822297e-06, + "loss": 0.3063, + "step": 963 + }, + { + "epoch": 0.3218697829716194, + "grad_norm": 0.6407388860443601, + "learning_rate": 9.887319485824935e-06, + "loss": 0.312, + "step": 964 + }, + { + "epoch": 0.32220367278797996, + "grad_norm": 0.6217988103195133, + "learning_rate": 9.886703558119146e-06, + "loss": 0.3013, + "step": 965 + }, + { + "epoch": 0.3225375626043406, + "grad_norm": 0.5670886196149892, + "learning_rate": 9.886085970914095e-06, + "loss": 0.2996, + "step": 966 + }, + { + "epoch": 0.32287145242070114, + "grad_norm": 0.5771448794414745, + "learning_rate": 9.88546672441951e-06, + "loss": 0.3048, + "step": 967 + }, + { + "epoch": 0.32320534223706177, + "grad_norm": 0.5860927919772361, + "learning_rate": 9.884845818845685e-06, + "loss": 0.3075, + "step": 968 + }, + { + "epoch": 0.3235392320534224, + "grad_norm": 0.617737356683965, + "learning_rate": 9.884223254403478e-06, + "loss": 0.2992, + "step": 969 + }, + { + "epoch": 0.32387312186978295, + "grad_norm": 0.5847235349905233, + "learning_rate": 9.883599031304305e-06, + "loss": 0.3061, + "step": 970 + }, + { + "epoch": 0.3242070116861436, + "grad_norm": 0.5777239833237784, + "learning_rate": 9.882973149760151e-06, + "loss": 0.3058, + "step": 971 + }, + { + "epoch": 0.3245409015025042, + "grad_norm": 0.6101749888771139, + "learning_rate": 9.882345609983564e-06, + "loss": 0.3124, + "step": 972 + }, + { + "epoch": 0.32487479131886476, + "grad_norm": 0.5924854741548193, + "learning_rate": 9.88171641218765e-06, + "loss": 0.3146, + "step": 973 + }, + { + "epoch": 0.3252086811352254, + "grad_norm": 0.5669357927709503, + "learning_rate": 9.881085556586083e-06, + "loss": 0.3051, + "step": 974 + }, + { + "epoch": 0.32554257095158595, + "grad_norm": 0.6454809193991066, + "learning_rate": 9.880453043393098e-06, + "loss": 0.3032, + "step": 975 + }, + { + "epoch": 0.3258764607679466, + "grad_norm": 0.5556890289646113, + "learning_rate": 9.879818872823492e-06, + "loss": 0.2874, + "step": 976 + }, + { + "epoch": 0.3262103505843072, + "grad_norm": 0.5835874096260433, + "learning_rate": 9.879183045092628e-06, + "loss": 0.3004, + "step": 977 + }, + { + "epoch": 0.32654424040066776, + "grad_norm": 0.5924377005508791, + "learning_rate": 9.87854556041643e-06, + "loss": 0.3119, + "step": 978 + }, + { + "epoch": 0.3268781302170284, + "grad_norm": 0.6035687688481293, + "learning_rate": 9.877906419011384e-06, + "loss": 0.3033, + "step": 979 + }, + { + "epoch": 0.327212020033389, + "grad_norm": 0.5717077769181176, + "learning_rate": 9.877265621094537e-06, + "loss": 0.3027, + "step": 980 + }, + { + "epoch": 0.3275459098497496, + "grad_norm": 0.584536556395434, + "learning_rate": 9.876623166883503e-06, + "loss": 0.3053, + "step": 981 + }, + { + "epoch": 0.3278797996661102, + "grad_norm": 0.5956701718831322, + "learning_rate": 9.875979056596457e-06, + "loss": 0.3044, + "step": 982 + }, + { + "epoch": 0.32821368948247076, + "grad_norm": 0.6558968226737262, + "learning_rate": 9.875333290452131e-06, + "loss": 0.3082, + "step": 983 + }, + { + "epoch": 0.3285475792988314, + "grad_norm": 0.5666958584315439, + "learning_rate": 9.874685868669829e-06, + "loss": 0.3132, + "step": 984 + }, + { + "epoch": 0.328881469115192, + "grad_norm": 0.5997837723679186, + "learning_rate": 9.87403679146941e-06, + "loss": 0.303, + "step": 985 + }, + { + "epoch": 0.32921535893155257, + "grad_norm": 0.5994134366693209, + "learning_rate": 9.873386059071294e-06, + "loss": 0.3188, + "step": 986 + }, + { + "epoch": 0.3295492487479132, + "grad_norm": 0.5833652121972631, + "learning_rate": 9.872733671696473e-06, + "loss": 0.3102, + "step": 987 + }, + { + "epoch": 0.3298831385642738, + "grad_norm": 0.6028069176748835, + "learning_rate": 9.872079629566492e-06, + "loss": 0.3246, + "step": 988 + }, + { + "epoch": 0.3302170283806344, + "grad_norm": 0.5990571020173145, + "learning_rate": 9.871423932903456e-06, + "loss": 0.2909, + "step": 989 + }, + { + "epoch": 0.330550918196995, + "grad_norm": 0.5683423727489445, + "learning_rate": 9.87076658193004e-06, + "loss": 0.3016, + "step": 990 + }, + { + "epoch": 0.33088480801335557, + "grad_norm": 0.6133546204079519, + "learning_rate": 9.870107576869478e-06, + "loss": 0.3014, + "step": 991 + }, + { + "epoch": 0.3312186978297162, + "grad_norm": 0.5927343606254837, + "learning_rate": 9.869446917945561e-06, + "loss": 0.3119, + "step": 992 + }, + { + "epoch": 0.3315525876460768, + "grad_norm": 0.6082425987149992, + "learning_rate": 9.868784605382649e-06, + "loss": 0.3203, + "step": 993 + }, + { + "epoch": 0.3318864774624374, + "grad_norm": 0.5779024414231667, + "learning_rate": 9.868120639405659e-06, + "loss": 0.3056, + "step": 994 + }, + { + "epoch": 0.332220367278798, + "grad_norm": 0.6079272910599491, + "learning_rate": 9.86745502024007e-06, + "loss": 0.3121, + "step": 995 + }, + { + "epoch": 0.3325542570951586, + "grad_norm": 0.6109787362072164, + "learning_rate": 9.86678774811192e-06, + "loss": 0.3129, + "step": 996 + }, + { + "epoch": 0.3328881469115192, + "grad_norm": 0.6116973190995274, + "learning_rate": 9.866118823247816e-06, + "loss": 0.306, + "step": 997 + }, + { + "epoch": 0.3332220367278798, + "grad_norm": 0.5970167347990984, + "learning_rate": 9.865448245874918e-06, + "loss": 0.3042, + "step": 998 + }, + { + "epoch": 0.3335559265442404, + "grad_norm": 0.5714368756846021, + "learning_rate": 9.864776016220952e-06, + "loss": 0.3107, + "step": 999 + }, + { + "epoch": 0.333889816360601, + "grad_norm": 0.5830980796375943, + "learning_rate": 9.864102134514205e-06, + "loss": 0.3248, + "step": 1000 + }, + { + "epoch": 0.3342237061769616, + "grad_norm": 0.6453019684681336, + "learning_rate": 9.86342660098352e-06, + "loss": 0.3112, + "step": 1001 + }, + { + "epoch": 0.3345575959933222, + "grad_norm": 0.5633889785626045, + "learning_rate": 9.86274941585831e-06, + "loss": 0.3012, + "step": 1002 + }, + { + "epoch": 0.3348914858096828, + "grad_norm": 0.6309583532474389, + "learning_rate": 9.862070579368538e-06, + "loss": 0.3075, + "step": 1003 + }, + { + "epoch": 0.33522537562604343, + "grad_norm": 0.5893685219336443, + "learning_rate": 9.861390091744738e-06, + "loss": 0.2985, + "step": 1004 + }, + { + "epoch": 0.335559265442404, + "grad_norm": 0.5762289972352435, + "learning_rate": 9.860707953217996e-06, + "loss": 0.3048, + "step": 1005 + }, + { + "epoch": 0.3358931552587646, + "grad_norm": 0.6323147992017026, + "learning_rate": 9.860024164019968e-06, + "loss": 0.317, + "step": 1006 + }, + { + "epoch": 0.3362270450751252, + "grad_norm": 0.571837738939002, + "learning_rate": 9.859338724382859e-06, + "loss": 0.3089, + "step": 1007 + }, + { + "epoch": 0.3365609348914858, + "grad_norm": 0.6044878335727768, + "learning_rate": 9.858651634539444e-06, + "loss": 0.3037, + "step": 1008 + }, + { + "epoch": 0.33689482470784643, + "grad_norm": 0.5893949601007495, + "learning_rate": 9.857962894723056e-06, + "loss": 0.3164, + "step": 1009 + }, + { + "epoch": 0.337228714524207, + "grad_norm": 0.5561608716513259, + "learning_rate": 9.857272505167585e-06, + "loss": 0.3045, + "step": 1010 + }, + { + "epoch": 0.3375626043405676, + "grad_norm": 0.6203875802627289, + "learning_rate": 9.856580466107487e-06, + "loss": 0.3251, + "step": 1011 + }, + { + "epoch": 0.33789649415692824, + "grad_norm": 0.5847216747546243, + "learning_rate": 9.855886777777772e-06, + "loss": 0.3115, + "step": 1012 + }, + { + "epoch": 0.3382303839732888, + "grad_norm": 0.583598975848231, + "learning_rate": 9.855191440414014e-06, + "loss": 0.2978, + "step": 1013 + }, + { + "epoch": 0.33856427378964943, + "grad_norm": 0.5622426190479687, + "learning_rate": 9.854494454252345e-06, + "loss": 0.3052, + "step": 1014 + }, + { + "epoch": 0.33889816360601, + "grad_norm": 0.6002838807577378, + "learning_rate": 9.85379581952946e-06, + "loss": 0.3065, + "step": 1015 + }, + { + "epoch": 0.3392320534223706, + "grad_norm": 0.6517270597284957, + "learning_rate": 9.85309553648261e-06, + "loss": 0.3195, + "step": 1016 + }, + { + "epoch": 0.33956594323873124, + "grad_norm": 0.5690615062253946, + "learning_rate": 9.85239360534961e-06, + "loss": 0.2971, + "step": 1017 + }, + { + "epoch": 0.3398998330550918, + "grad_norm": 0.580097930001129, + "learning_rate": 9.851690026368829e-06, + "loss": 0.3118, + "step": 1018 + }, + { + "epoch": 0.3402337228714524, + "grad_norm": 0.621428728969123, + "learning_rate": 9.850984799779199e-06, + "loss": 0.3165, + "step": 1019 + }, + { + "epoch": 0.34056761268781305, + "grad_norm": 0.605768898479493, + "learning_rate": 9.850277925820215e-06, + "loss": 0.3127, + "step": 1020 + }, + { + "epoch": 0.3409015025041736, + "grad_norm": 0.6418956600168322, + "learning_rate": 9.849569404731922e-06, + "loss": 0.3186, + "step": 1021 + }, + { + "epoch": 0.34123539232053424, + "grad_norm": 0.618942551733838, + "learning_rate": 9.848859236754936e-06, + "loss": 0.3, + "step": 1022 + }, + { + "epoch": 0.3415692821368948, + "grad_norm": 0.5915948373962187, + "learning_rate": 9.848147422130421e-06, + "loss": 0.3032, + "step": 1023 + }, + { + "epoch": 0.3419031719532554, + "grad_norm": 0.5792441904832164, + "learning_rate": 9.84743396110011e-06, + "loss": 0.3011, + "step": 1024 + }, + { + "epoch": 0.34223706176961605, + "grad_norm": 0.628650169431253, + "learning_rate": 9.846718853906289e-06, + "loss": 0.3124, + "step": 1025 + }, + { + "epoch": 0.3425709515859766, + "grad_norm": 0.5369309745154097, + "learning_rate": 9.846002100791802e-06, + "loss": 0.3023, + "step": 1026 + }, + { + "epoch": 0.34290484140233723, + "grad_norm": 0.5791541929304539, + "learning_rate": 9.84528370200006e-06, + "loss": 0.3073, + "step": 1027 + }, + { + "epoch": 0.34323873121869786, + "grad_norm": 0.5763727282434769, + "learning_rate": 9.844563657775024e-06, + "loss": 0.3114, + "step": 1028 + }, + { + "epoch": 0.3435726210350584, + "grad_norm": 0.5809561965018103, + "learning_rate": 9.843841968361217e-06, + "loss": 0.3145, + "step": 1029 + }, + { + "epoch": 0.34390651085141904, + "grad_norm": 0.6033962562460989, + "learning_rate": 9.843118634003724e-06, + "loss": 0.3177, + "step": 1030 + }, + { + "epoch": 0.3442404006677796, + "grad_norm": 0.5833412198729083, + "learning_rate": 9.84239365494818e-06, + "loss": 0.3051, + "step": 1031 + }, + { + "epoch": 0.34457429048414023, + "grad_norm": 0.5747189868749905, + "learning_rate": 9.841667031440791e-06, + "loss": 0.3, + "step": 1032 + }, + { + "epoch": 0.34490818030050086, + "grad_norm": 0.6306969313392421, + "learning_rate": 9.840938763728312e-06, + "loss": 0.3125, + "step": 1033 + }, + { + "epoch": 0.3452420701168614, + "grad_norm": 0.5570068110687038, + "learning_rate": 9.840208852058055e-06, + "loss": 0.3029, + "step": 1034 + }, + { + "epoch": 0.34557595993322204, + "grad_norm": 0.5554764540241462, + "learning_rate": 9.8394772966779e-06, + "loss": 0.2937, + "step": 1035 + }, + { + "epoch": 0.34590984974958267, + "grad_norm": 0.5733775018441675, + "learning_rate": 9.838744097836276e-06, + "loss": 0.3117, + "step": 1036 + }, + { + "epoch": 0.34624373956594323, + "grad_norm": 0.565204043118247, + "learning_rate": 9.838009255782176e-06, + "loss": 0.2972, + "step": 1037 + }, + { + "epoch": 0.34657762938230385, + "grad_norm": 0.5536758392750166, + "learning_rate": 9.837272770765143e-06, + "loss": 0.2947, + "step": 1038 + }, + { + "epoch": 0.3469115191986644, + "grad_norm": 0.5936273820285527, + "learning_rate": 9.836534643035292e-06, + "loss": 0.2933, + "step": 1039 + }, + { + "epoch": 0.34724540901502504, + "grad_norm": 0.5967603462176918, + "learning_rate": 9.835794872843281e-06, + "loss": 0.2977, + "step": 1040 + }, + { + "epoch": 0.34757929883138566, + "grad_norm": 0.583394338220006, + "learning_rate": 9.835053460440334e-06, + "loss": 0.3072, + "step": 1041 + }, + { + "epoch": 0.34791318864774623, + "grad_norm": 0.5553432875180079, + "learning_rate": 9.83431040607823e-06, + "loss": 0.2979, + "step": 1042 + }, + { + "epoch": 0.34824707846410685, + "grad_norm": 0.5819072813823893, + "learning_rate": 9.833565710009308e-06, + "loss": 0.2896, + "step": 1043 + }, + { + "epoch": 0.3485809682804674, + "grad_norm": 0.6244468227476485, + "learning_rate": 9.832819372486463e-06, + "loss": 0.3072, + "step": 1044 + }, + { + "epoch": 0.34891485809682804, + "grad_norm": 0.5971719109417338, + "learning_rate": 9.832071393763145e-06, + "loss": 0.3179, + "step": 1045 + }, + { + "epoch": 0.34924874791318866, + "grad_norm": 0.6028669531219605, + "learning_rate": 9.831321774093366e-06, + "loss": 0.2997, + "step": 1046 + }, + { + "epoch": 0.34958263772954923, + "grad_norm": 0.570411791693014, + "learning_rate": 9.830570513731691e-06, + "loss": 0.3054, + "step": 1047 + }, + { + "epoch": 0.34991652754590985, + "grad_norm": 0.5742931441584476, + "learning_rate": 9.829817612933248e-06, + "loss": 0.2998, + "step": 1048 + }, + { + "epoch": 0.35025041736227047, + "grad_norm": 0.5913544997748079, + "learning_rate": 9.829063071953715e-06, + "loss": 0.3102, + "step": 1049 + }, + { + "epoch": 0.35058430717863104, + "grad_norm": 0.5643036860613097, + "learning_rate": 9.82830689104933e-06, + "loss": 0.2931, + "step": 1050 + }, + { + "epoch": 0.35091819699499166, + "grad_norm": 0.5797909309189535, + "learning_rate": 9.82754907047689e-06, + "loss": 0.3113, + "step": 1051 + }, + { + "epoch": 0.3512520868113522, + "grad_norm": 0.5817920129376497, + "learning_rate": 9.826789610493744e-06, + "loss": 0.3131, + "step": 1052 + }, + { + "epoch": 0.35158597662771285, + "grad_norm": 0.5801970903387936, + "learning_rate": 9.826028511357804e-06, + "loss": 0.2892, + "step": 1053 + }, + { + "epoch": 0.35191986644407347, + "grad_norm": 0.6178107571245536, + "learning_rate": 9.825265773327533e-06, + "loss": 0.3044, + "step": 1054 + }, + { + "epoch": 0.35225375626043404, + "grad_norm": 0.5706158771108824, + "learning_rate": 9.824501396661955e-06, + "loss": 0.3033, + "step": 1055 + }, + { + "epoch": 0.35258764607679466, + "grad_norm": 0.5805570447506774, + "learning_rate": 9.823735381620645e-06, + "loss": 0.3047, + "step": 1056 + }, + { + "epoch": 0.3529215358931553, + "grad_norm": 0.5916719754199041, + "learning_rate": 9.822967728463741e-06, + "loss": 0.3014, + "step": 1057 + }, + { + "epoch": 0.35325542570951585, + "grad_norm": 0.617485132950589, + "learning_rate": 9.822198437451933e-06, + "loss": 0.312, + "step": 1058 + }, + { + "epoch": 0.35358931552587647, + "grad_norm": 0.5976691079138409, + "learning_rate": 9.821427508846469e-06, + "loss": 0.3225, + "step": 1059 + }, + { + "epoch": 0.35392320534223703, + "grad_norm": 0.6120135660244067, + "learning_rate": 9.820654942909152e-06, + "loss": 0.3137, + "step": 1060 + }, + { + "epoch": 0.35425709515859766, + "grad_norm": 0.5548545950382316, + "learning_rate": 9.819880739902339e-06, + "loss": 0.3093, + "step": 1061 + }, + { + "epoch": 0.3545909849749583, + "grad_norm": 0.5862194776443842, + "learning_rate": 9.819104900088948e-06, + "loss": 0.2975, + "step": 1062 + }, + { + "epoch": 0.35492487479131885, + "grad_norm": 0.5633408107835643, + "learning_rate": 9.81832742373245e-06, + "loss": 0.3036, + "step": 1063 + }, + { + "epoch": 0.35525876460767947, + "grad_norm": 0.5685058183159014, + "learning_rate": 9.81754831109687e-06, + "loss": 0.3038, + "step": 1064 + }, + { + "epoch": 0.3555926544240401, + "grad_norm": 0.5321866122630453, + "learning_rate": 9.816767562446794e-06, + "loss": 0.2892, + "step": 1065 + }, + { + "epoch": 0.35592654424040066, + "grad_norm": 0.5925110072780826, + "learning_rate": 9.815985178047357e-06, + "loss": 0.3026, + "step": 1066 + }, + { + "epoch": 0.3562604340567613, + "grad_norm": 0.5659185202468268, + "learning_rate": 9.815201158164253e-06, + "loss": 0.2932, + "step": 1067 + }, + { + "epoch": 0.35659432387312184, + "grad_norm": 0.5636675798163887, + "learning_rate": 9.814415503063735e-06, + "loss": 0.3063, + "step": 1068 + }, + { + "epoch": 0.35692821368948247, + "grad_norm": 0.5555853422836861, + "learning_rate": 9.813628213012603e-06, + "loss": 0.293, + "step": 1069 + }, + { + "epoch": 0.3572621035058431, + "grad_norm": 0.5462157458769987, + "learning_rate": 9.812839288278218e-06, + "loss": 0.2958, + "step": 1070 + }, + { + "epoch": 0.35759599332220365, + "grad_norm": 0.5533275166028977, + "learning_rate": 9.812048729128494e-06, + "loss": 0.3057, + "step": 1071 + }, + { + "epoch": 0.3579298831385643, + "grad_norm": 0.5505172323152103, + "learning_rate": 9.811256535831904e-06, + "loss": 0.3029, + "step": 1072 + }, + { + "epoch": 0.3582637729549249, + "grad_norm": 0.5694428067177824, + "learning_rate": 9.810462708657467e-06, + "loss": 0.3171, + "step": 1073 + }, + { + "epoch": 0.35859766277128546, + "grad_norm": 0.5607570633063697, + "learning_rate": 9.809667247874768e-06, + "loss": 0.3012, + "step": 1074 + }, + { + "epoch": 0.3589315525876461, + "grad_norm": 0.5483989213386087, + "learning_rate": 9.80887015375394e-06, + "loss": 0.2943, + "step": 1075 + }, + { + "epoch": 0.35926544240400665, + "grad_norm": 0.5364806450522815, + "learning_rate": 9.808071426565671e-06, + "loss": 0.2906, + "step": 1076 + }, + { + "epoch": 0.3595993322203673, + "grad_norm": 0.5796905684412624, + "learning_rate": 9.807271066581203e-06, + "loss": 0.308, + "step": 1077 + }, + { + "epoch": 0.3599332220367279, + "grad_norm": 0.5533070940529261, + "learning_rate": 9.806469074072339e-06, + "loss": 0.2998, + "step": 1078 + }, + { + "epoch": 0.36026711185308846, + "grad_norm": 0.5381317391085549, + "learning_rate": 9.805665449311428e-06, + "loss": 0.305, + "step": 1079 + }, + { + "epoch": 0.3606010016694491, + "grad_norm": 0.5355733683084221, + "learning_rate": 9.804860192571376e-06, + "loss": 0.2998, + "step": 1080 + }, + { + "epoch": 0.3609348914858097, + "grad_norm": 0.6152079491346703, + "learning_rate": 9.804053304125646e-06, + "loss": 0.3118, + "step": 1081 + }, + { + "epoch": 0.36126878130217027, + "grad_norm": 0.5733018220128471, + "learning_rate": 9.803244784248253e-06, + "loss": 0.3106, + "step": 1082 + }, + { + "epoch": 0.3616026711185309, + "grad_norm": 0.5349879376241709, + "learning_rate": 9.802434633213767e-06, + "loss": 0.2993, + "step": 1083 + }, + { + "epoch": 0.36193656093489146, + "grad_norm": 0.5847075736142808, + "learning_rate": 9.801622851297309e-06, + "loss": 0.2927, + "step": 1084 + }, + { + "epoch": 0.3622704507512521, + "grad_norm": 0.549238826178887, + "learning_rate": 9.800809438774557e-06, + "loss": 0.3029, + "step": 1085 + }, + { + "epoch": 0.3626043405676127, + "grad_norm": 0.560095035325085, + "learning_rate": 9.79999439592174e-06, + "loss": 0.2995, + "step": 1086 + }, + { + "epoch": 0.36293823038397327, + "grad_norm": 0.5639002944157065, + "learning_rate": 9.799177723015646e-06, + "loss": 0.2958, + "step": 1087 + }, + { + "epoch": 0.3632721202003339, + "grad_norm": 0.5333904867383243, + "learning_rate": 9.798359420333609e-06, + "loss": 0.2975, + "step": 1088 + }, + { + "epoch": 0.3636060100166945, + "grad_norm": 0.5609413464711082, + "learning_rate": 9.797539488153525e-06, + "loss": 0.3049, + "step": 1089 + }, + { + "epoch": 0.3639398998330551, + "grad_norm": 0.5802330792947921, + "learning_rate": 9.796717926753832e-06, + "loss": 0.2971, + "step": 1090 + }, + { + "epoch": 0.3642737896494157, + "grad_norm": 0.5375610505085358, + "learning_rate": 9.795894736413535e-06, + "loss": 0.3028, + "step": 1091 + }, + { + "epoch": 0.36460767946577627, + "grad_norm": 0.5800116797751771, + "learning_rate": 9.79506991741218e-06, + "loss": 0.3074, + "step": 1092 + }, + { + "epoch": 0.3649415692821369, + "grad_norm": 0.5522679135364318, + "learning_rate": 9.794243470029875e-06, + "loss": 0.3013, + "step": 1093 + }, + { + "epoch": 0.3652754590984975, + "grad_norm": 0.56280404595334, + "learning_rate": 9.793415394547274e-06, + "loss": 0.3054, + "step": 1094 + }, + { + "epoch": 0.3656093489148581, + "grad_norm": 0.5273551084983782, + "learning_rate": 9.792585691245591e-06, + "loss": 0.3054, + "step": 1095 + }, + { + "epoch": 0.3659432387312187, + "grad_norm": 0.5589603881493072, + "learning_rate": 9.791754360406584e-06, + "loss": 0.3057, + "step": 1096 + }, + { + "epoch": 0.3662771285475793, + "grad_norm": 0.566661223176785, + "learning_rate": 9.790921402312573e-06, + "loss": 0.3003, + "step": 1097 + }, + { + "epoch": 0.3666110183639399, + "grad_norm": 0.5637473478542079, + "learning_rate": 9.790086817246425e-06, + "loss": 0.2962, + "step": 1098 + }, + { + "epoch": 0.3669449081803005, + "grad_norm": 0.5522886000953054, + "learning_rate": 9.789250605491558e-06, + "loss": 0.3014, + "step": 1099 + }, + { + "epoch": 0.3672787979966611, + "grad_norm": 0.5654982998182645, + "learning_rate": 9.78841276733195e-06, + "loss": 0.2954, + "step": 1100 + }, + { + "epoch": 0.3676126878130217, + "grad_norm": 0.5425147047295557, + "learning_rate": 9.787573303052123e-06, + "loss": 0.3077, + "step": 1101 + }, + { + "epoch": 0.3679465776293823, + "grad_norm": 0.5694226890400224, + "learning_rate": 9.786732212937156e-06, + "loss": 0.3128, + "step": 1102 + }, + { + "epoch": 0.3682804674457429, + "grad_norm": 0.5242138396035159, + "learning_rate": 9.785889497272678e-06, + "loss": 0.3064, + "step": 1103 + }, + { + "epoch": 0.3686143572621035, + "grad_norm": 0.566185638036358, + "learning_rate": 9.785045156344872e-06, + "loss": 0.2997, + "step": 1104 + }, + { + "epoch": 0.36894824707846413, + "grad_norm": 0.5511264072047883, + "learning_rate": 9.784199190440473e-06, + "loss": 0.3076, + "step": 1105 + }, + { + "epoch": 0.3692821368948247, + "grad_norm": 0.5478451471043818, + "learning_rate": 9.783351599846762e-06, + "loss": 0.2993, + "step": 1106 + }, + { + "epoch": 0.3696160267111853, + "grad_norm": 0.5755901067987882, + "learning_rate": 9.782502384851582e-06, + "loss": 0.2997, + "step": 1107 + }, + { + "epoch": 0.3699499165275459, + "grad_norm": 0.5648686426440034, + "learning_rate": 9.781651545743319e-06, + "loss": 0.3055, + "step": 1108 + }, + { + "epoch": 0.3702838063439065, + "grad_norm": 0.5366645195289919, + "learning_rate": 9.780799082810912e-06, + "loss": 0.2956, + "step": 1109 + }, + { + "epoch": 0.37061769616026713, + "grad_norm": 0.5568957012919856, + "learning_rate": 9.779944996343858e-06, + "loss": 0.3109, + "step": 1110 + }, + { + "epoch": 0.3709515859766277, + "grad_norm": 0.5623039028298573, + "learning_rate": 9.779089286632194e-06, + "loss": 0.3031, + "step": 1111 + }, + { + "epoch": 0.3712854757929883, + "grad_norm": 0.5439821823871229, + "learning_rate": 9.778231953966519e-06, + "loss": 0.2963, + "step": 1112 + }, + { + "epoch": 0.37161936560934894, + "grad_norm": 0.5284277978718807, + "learning_rate": 9.77737299863798e-06, + "loss": 0.3051, + "step": 1113 + }, + { + "epoch": 0.3719532554257095, + "grad_norm": 0.5764544260859259, + "learning_rate": 9.77651242093827e-06, + "loss": 0.2985, + "step": 1114 + }, + { + "epoch": 0.3722871452420701, + "grad_norm": 0.5462409166699114, + "learning_rate": 9.775650221159635e-06, + "loss": 0.2932, + "step": 1115 + }, + { + "epoch": 0.3726210350584307, + "grad_norm": 0.5343835850205473, + "learning_rate": 9.774786399594881e-06, + "loss": 0.2861, + "step": 1116 + }, + { + "epoch": 0.3729549248747913, + "grad_norm": 0.5884833894610786, + "learning_rate": 9.773920956537351e-06, + "loss": 0.316, + "step": 1117 + }, + { + "epoch": 0.37328881469115194, + "grad_norm": 0.547504442820607, + "learning_rate": 9.773053892280949e-06, + "loss": 0.2968, + "step": 1118 + }, + { + "epoch": 0.3736227045075125, + "grad_norm": 0.5601394241365709, + "learning_rate": 9.772185207120122e-06, + "loss": 0.3052, + "step": 1119 + }, + { + "epoch": 0.3739565943238731, + "grad_norm": 0.5739119450001532, + "learning_rate": 9.771314901349873e-06, + "loss": 0.3053, + "step": 1120 + }, + { + "epoch": 0.37429048414023375, + "grad_norm": 0.5413550460105998, + "learning_rate": 9.770442975265753e-06, + "loss": 0.2931, + "step": 1121 + }, + { + "epoch": 0.3746243739565943, + "grad_norm": 0.5823746869013713, + "learning_rate": 9.769569429163863e-06, + "loss": 0.2882, + "step": 1122 + }, + { + "epoch": 0.37495826377295494, + "grad_norm": 0.594316216550019, + "learning_rate": 9.768694263340856e-06, + "loss": 0.3136, + "step": 1123 + }, + { + "epoch": 0.3752921535893155, + "grad_norm": 0.5359870864700037, + "learning_rate": 9.767817478093933e-06, + "loss": 0.2978, + "step": 1124 + }, + { + "epoch": 0.3756260434056761, + "grad_norm": 0.5939303577626737, + "learning_rate": 9.766939073720846e-06, + "loss": 0.3072, + "step": 1125 + }, + { + "epoch": 0.37595993322203675, + "grad_norm": 0.5636707991404845, + "learning_rate": 9.766059050519897e-06, + "loss": 0.2999, + "step": 1126 + }, + { + "epoch": 0.3762938230383973, + "grad_norm": 0.5347667886833426, + "learning_rate": 9.765177408789935e-06, + "loss": 0.3024, + "step": 1127 + }, + { + "epoch": 0.37662771285475793, + "grad_norm": 0.5439002037358984, + "learning_rate": 9.764294148830363e-06, + "loss": 0.2967, + "step": 1128 + }, + { + "epoch": 0.37696160267111856, + "grad_norm": 0.5615831382280493, + "learning_rate": 9.763409270941132e-06, + "loss": 0.3105, + "step": 1129 + }, + { + "epoch": 0.3772954924874791, + "grad_norm": 0.5950595714252847, + "learning_rate": 9.762522775422741e-06, + "loss": 0.301, + "step": 1130 + }, + { + "epoch": 0.37762938230383974, + "grad_norm": 0.5380571741680282, + "learning_rate": 9.76163466257624e-06, + "loss": 0.2909, + "step": 1131 + }, + { + "epoch": 0.3779632721202003, + "grad_norm": 0.5335958985138369, + "learning_rate": 9.760744932703225e-06, + "loss": 0.3035, + "step": 1132 + }, + { + "epoch": 0.37829716193656093, + "grad_norm": 0.5715892806797341, + "learning_rate": 9.759853586105848e-06, + "loss": 0.3083, + "step": 1133 + }, + { + "epoch": 0.37863105175292155, + "grad_norm": 0.5733957390442936, + "learning_rate": 9.758960623086802e-06, + "loss": 0.3027, + "step": 1134 + }, + { + "epoch": 0.3789649415692821, + "grad_norm": 0.5654415053449496, + "learning_rate": 9.758066043949335e-06, + "loss": 0.2966, + "step": 1135 + }, + { + "epoch": 0.37929883138564274, + "grad_norm": 0.5475996804862423, + "learning_rate": 9.75716984899724e-06, + "loss": 0.3001, + "step": 1136 + }, + { + "epoch": 0.37963272120200336, + "grad_norm": 0.5390078911862362, + "learning_rate": 9.75627203853486e-06, + "loss": 0.3082, + "step": 1137 + }, + { + "epoch": 0.37996661101836393, + "grad_norm": 0.563645572314964, + "learning_rate": 9.755372612867087e-06, + "loss": 0.3, + "step": 1138 + }, + { + "epoch": 0.38030050083472455, + "grad_norm": 0.5665229211828521, + "learning_rate": 9.754471572299363e-06, + "loss": 0.3092, + "step": 1139 + }, + { + "epoch": 0.3806343906510851, + "grad_norm": 0.5301112426251547, + "learning_rate": 9.753568917137675e-06, + "loss": 0.2967, + "step": 1140 + }, + { + "epoch": 0.38096828046744574, + "grad_norm": 0.5763955736877978, + "learning_rate": 9.752664647688562e-06, + "loss": 0.3047, + "step": 1141 + }, + { + "epoch": 0.38130217028380636, + "grad_norm": 0.5704264377657354, + "learning_rate": 9.751758764259104e-06, + "loss": 0.292, + "step": 1142 + }, + { + "epoch": 0.38163606010016693, + "grad_norm": 0.5401170217081893, + "learning_rate": 9.75085126715694e-06, + "loss": 0.2939, + "step": 1143 + }, + { + "epoch": 0.38196994991652755, + "grad_norm": 0.5549752199814894, + "learning_rate": 9.749942156690251e-06, + "loss": 0.2907, + "step": 1144 + }, + { + "epoch": 0.3823038397328882, + "grad_norm": 0.5713762564874129, + "learning_rate": 9.749031433167761e-06, + "loss": 0.3014, + "step": 1145 + }, + { + "epoch": 0.38263772954924874, + "grad_norm": 0.5735125929152683, + "learning_rate": 9.748119096898754e-06, + "loss": 0.3018, + "step": 1146 + }, + { + "epoch": 0.38297161936560936, + "grad_norm": 0.5672897610227269, + "learning_rate": 9.747205148193048e-06, + "loss": 0.3016, + "step": 1147 + }, + { + "epoch": 0.38330550918196993, + "grad_norm": 0.5649006112228457, + "learning_rate": 9.746289587361021e-06, + "loss": 0.3066, + "step": 1148 + }, + { + "epoch": 0.38363939899833055, + "grad_norm": 0.5516513784284546, + "learning_rate": 9.74537241471359e-06, + "loss": 0.2975, + "step": 1149 + }, + { + "epoch": 0.38397328881469117, + "grad_norm": 0.5869259842453701, + "learning_rate": 9.74445363056222e-06, + "loss": 0.3162, + "step": 1150 + }, + { + "epoch": 0.38430717863105174, + "grad_norm": 0.6086630180051924, + "learning_rate": 9.743533235218931e-06, + "loss": 0.284, + "step": 1151 + }, + { + "epoch": 0.38464106844741236, + "grad_norm": 0.6527621189753554, + "learning_rate": 9.74261122899628e-06, + "loss": 0.3023, + "step": 1152 + }, + { + "epoch": 0.384974958263773, + "grad_norm": 0.5213844616086313, + "learning_rate": 9.741687612207375e-06, + "loss": 0.2976, + "step": 1153 + }, + { + "epoch": 0.38530884808013355, + "grad_norm": 0.5756565644063859, + "learning_rate": 9.740762385165874e-06, + "loss": 0.3065, + "step": 1154 + }, + { + "epoch": 0.38564273789649417, + "grad_norm": 0.579537938162433, + "learning_rate": 9.739835548185978e-06, + "loss": 0.3221, + "step": 1155 + }, + { + "epoch": 0.38597662771285474, + "grad_norm": 0.5533703537681108, + "learning_rate": 9.738907101582439e-06, + "loss": 0.2971, + "step": 1156 + }, + { + "epoch": 0.38631051752921536, + "grad_norm": 0.5502077408004404, + "learning_rate": 9.737977045670549e-06, + "loss": 0.3002, + "step": 1157 + }, + { + "epoch": 0.386644407345576, + "grad_norm": 0.5902437766757179, + "learning_rate": 9.73704538076615e-06, + "loss": 0.3168, + "step": 1158 + }, + { + "epoch": 0.38697829716193655, + "grad_norm": 0.5792304434491469, + "learning_rate": 9.736112107185634e-06, + "loss": 0.2944, + "step": 1159 + }, + { + "epoch": 0.38731218697829717, + "grad_norm": 0.5752368565391309, + "learning_rate": 9.73517722524593e-06, + "loss": 0.3052, + "step": 1160 + }, + { + "epoch": 0.3876460767946578, + "grad_norm": 0.5545505751326958, + "learning_rate": 9.734240735264526e-06, + "loss": 0.2986, + "step": 1161 + }, + { + "epoch": 0.38797996661101836, + "grad_norm": 0.5438353699564492, + "learning_rate": 9.733302637559447e-06, + "loss": 0.2897, + "step": 1162 + }, + { + "epoch": 0.388313856427379, + "grad_norm": 0.5716629535712481, + "learning_rate": 9.732362932449264e-06, + "loss": 0.302, + "step": 1163 + }, + { + "epoch": 0.38864774624373954, + "grad_norm": 0.5569960272846087, + "learning_rate": 9.731421620253097e-06, + "loss": 0.3032, + "step": 1164 + }, + { + "epoch": 0.38898163606010017, + "grad_norm": 0.6246474054452963, + "learning_rate": 9.730478701290611e-06, + "loss": 0.3023, + "step": 1165 + }, + { + "epoch": 0.3893155258764608, + "grad_norm": 0.5826367471042718, + "learning_rate": 9.729534175882016e-06, + "loss": 0.2909, + "step": 1166 + }, + { + "epoch": 0.38964941569282135, + "grad_norm": 0.5220632428887642, + "learning_rate": 9.72858804434807e-06, + "loss": 0.295, + "step": 1167 + }, + { + "epoch": 0.389983305509182, + "grad_norm": 0.5949051900257158, + "learning_rate": 9.72764030701007e-06, + "loss": 0.3051, + "step": 1168 + }, + { + "epoch": 0.3903171953255426, + "grad_norm": 0.6020125535806303, + "learning_rate": 9.726690964189866e-06, + "loss": 0.3051, + "step": 1169 + }, + { + "epoch": 0.39065108514190316, + "grad_norm": 0.5500008999622386, + "learning_rate": 9.725740016209851e-06, + "loss": 0.3033, + "step": 1170 + }, + { + "epoch": 0.3909849749582638, + "grad_norm": 0.5990727033663708, + "learning_rate": 9.72478746339296e-06, + "loss": 0.3027, + "step": 1171 + }, + { + "epoch": 0.39131886477462435, + "grad_norm": 0.6003793339155197, + "learning_rate": 9.723833306062674e-06, + "loss": 0.3069, + "step": 1172 + }, + { + "epoch": 0.391652754590985, + "grad_norm": 0.5612692493248552, + "learning_rate": 9.722877544543023e-06, + "loss": 0.2874, + "step": 1173 + }, + { + "epoch": 0.3919866444073456, + "grad_norm": 0.64259219192056, + "learning_rate": 9.721920179158575e-06, + "loss": 0.2999, + "step": 1174 + }, + { + "epoch": 0.39232053422370616, + "grad_norm": 0.5665868500638949, + "learning_rate": 9.72096121023445e-06, + "loss": 0.2894, + "step": 1175 + }, + { + "epoch": 0.3926544240400668, + "grad_norm": 0.5550360077835788, + "learning_rate": 9.720000638096306e-06, + "loss": 0.295, + "step": 1176 + }, + { + "epoch": 0.39298831385642735, + "grad_norm": 0.5388826655840278, + "learning_rate": 9.71903846307035e-06, + "loss": 0.307, + "step": 1177 + }, + { + "epoch": 0.393322203672788, + "grad_norm": 0.5254865153189809, + "learning_rate": 9.71807468548333e-06, + "loss": 0.2986, + "step": 1178 + }, + { + "epoch": 0.3936560934891486, + "grad_norm": 0.5751151930968073, + "learning_rate": 9.717109305662542e-06, + "loss": 0.3114, + "step": 1179 + }, + { + "epoch": 0.39398998330550916, + "grad_norm": 0.5318662473048824, + "learning_rate": 9.716142323935823e-06, + "loss": 0.3003, + "step": 1180 + }, + { + "epoch": 0.3943238731218698, + "grad_norm": 0.5764077530555939, + "learning_rate": 9.715173740631554e-06, + "loss": 0.3073, + "step": 1181 + }, + { + "epoch": 0.3946577629382304, + "grad_norm": 0.5627005702121924, + "learning_rate": 9.714203556078662e-06, + "loss": 0.2988, + "step": 1182 + }, + { + "epoch": 0.39499165275459097, + "grad_norm": 0.5603314949671782, + "learning_rate": 9.713231770606616e-06, + "loss": 0.2938, + "step": 1183 + }, + { + "epoch": 0.3953255425709516, + "grad_norm": 0.5526056833050033, + "learning_rate": 9.712258384545432e-06, + "loss": 0.2938, + "step": 1184 + }, + { + "epoch": 0.39565943238731216, + "grad_norm": 0.5194022324906876, + "learning_rate": 9.711283398225664e-06, + "loss": 0.3117, + "step": 1185 + }, + { + "epoch": 0.3959933222036728, + "grad_norm": 0.5629310163032546, + "learning_rate": 9.710306811978413e-06, + "loss": 0.2985, + "step": 1186 + }, + { + "epoch": 0.3963272120200334, + "grad_norm": 0.5607486970818731, + "learning_rate": 9.709328626135322e-06, + "loss": 0.2963, + "step": 1187 + }, + { + "epoch": 0.39666110183639397, + "grad_norm": 0.5539826134029737, + "learning_rate": 9.70834884102858e-06, + "loss": 0.3028, + "step": 1188 + }, + { + "epoch": 0.3969949916527546, + "grad_norm": 0.6261311586983064, + "learning_rate": 9.707367456990913e-06, + "loss": 0.3147, + "step": 1189 + }, + { + "epoch": 0.3973288814691152, + "grad_norm": 0.5723062465637354, + "learning_rate": 9.7063844743556e-06, + "loss": 0.298, + "step": 1190 + }, + { + "epoch": 0.3976627712854758, + "grad_norm": 0.5870148777224925, + "learning_rate": 9.705399893456449e-06, + "loss": 0.3146, + "step": 1191 + }, + { + "epoch": 0.3979966611018364, + "grad_norm": 0.5865330170548101, + "learning_rate": 9.704413714627825e-06, + "loss": 0.3041, + "step": 1192 + }, + { + "epoch": 0.39833055091819697, + "grad_norm": 0.5843146187985424, + "learning_rate": 9.703425938204627e-06, + "loss": 0.296, + "step": 1193 + }, + { + "epoch": 0.3986644407345576, + "grad_norm": 0.5464089388607848, + "learning_rate": 9.7024365645223e-06, + "loss": 0.302, + "step": 1194 + }, + { + "epoch": 0.3989983305509182, + "grad_norm": 0.6319476236520657, + "learning_rate": 9.701445593916827e-06, + "loss": 0.2909, + "step": 1195 + }, + { + "epoch": 0.3993322203672788, + "grad_norm": 0.5345331989733932, + "learning_rate": 9.70045302672474e-06, + "loss": 0.2952, + "step": 1196 + }, + { + "epoch": 0.3996661101836394, + "grad_norm": 0.5519692534416208, + "learning_rate": 9.699458863283106e-06, + "loss": 0.3107, + "step": 1197 + }, + { + "epoch": 0.4, + "grad_norm": 0.5838957507648197, + "learning_rate": 9.698463103929542e-06, + "loss": 0.304, + "step": 1198 + }, + { + "epoch": 0.4003338898163606, + "grad_norm": 0.5522635908043041, + "learning_rate": 9.697465749002202e-06, + "loss": 0.2982, + "step": 1199 + }, + { + "epoch": 0.4006677796327212, + "grad_norm": 0.5728355133435783, + "learning_rate": 9.69646679883978e-06, + "loss": 0.2964, + "step": 1200 + }, + { + "epoch": 0.4010016694490818, + "grad_norm": 0.516286538413823, + "learning_rate": 9.695466253781514e-06, + "loss": 0.3064, + "step": 1201 + }, + { + "epoch": 0.4013355592654424, + "grad_norm": 0.5521859804491117, + "learning_rate": 9.694464114167185e-06, + "loss": 0.2897, + "step": 1202 + }, + { + "epoch": 0.401669449081803, + "grad_norm": 0.5650950277755699, + "learning_rate": 9.693460380337118e-06, + "loss": 0.3083, + "step": 1203 + }, + { + "epoch": 0.4020033388981636, + "grad_norm": 0.5565357210766206, + "learning_rate": 9.69245505263217e-06, + "loss": 0.3122, + "step": 1204 + }, + { + "epoch": 0.4023372287145242, + "grad_norm": 0.5560348552069969, + "learning_rate": 9.69144813139375e-06, + "loss": 0.291, + "step": 1205 + }, + { + "epoch": 0.40267111853088483, + "grad_norm": 0.560692089626692, + "learning_rate": 9.6904396169638e-06, + "loss": 0.295, + "step": 1206 + }, + { + "epoch": 0.4030050083472454, + "grad_norm": 0.5139116187672647, + "learning_rate": 9.689429509684808e-06, + "loss": 0.293, + "step": 1207 + }, + { + "epoch": 0.403338898163606, + "grad_norm": 0.5417014042411912, + "learning_rate": 9.688417809899799e-06, + "loss": 0.2912, + "step": 1208 + }, + { + "epoch": 0.4036727879799666, + "grad_norm": 0.5525996580065495, + "learning_rate": 9.687404517952345e-06, + "loss": 0.3051, + "step": 1209 + }, + { + "epoch": 0.4040066777963272, + "grad_norm": 0.5213181115015675, + "learning_rate": 9.68638963418655e-06, + "loss": 0.2965, + "step": 1210 + }, + { + "epoch": 0.40434056761268783, + "grad_norm": 0.5318917311807938, + "learning_rate": 9.685373158947067e-06, + "loss": 0.2874, + "step": 1211 + }, + { + "epoch": 0.4046744574290484, + "grad_norm": 0.5711050265128256, + "learning_rate": 9.684355092579085e-06, + "loss": 0.3036, + "step": 1212 + }, + { + "epoch": 0.405008347245409, + "grad_norm": 0.5610710480357389, + "learning_rate": 9.683335435428332e-06, + "loss": 0.3056, + "step": 1213 + }, + { + "epoch": 0.40534223706176964, + "grad_norm": 0.5436409682723612, + "learning_rate": 9.68231418784108e-06, + "loss": 0.2914, + "step": 1214 + }, + { + "epoch": 0.4056761268781302, + "grad_norm": 0.5906697972062736, + "learning_rate": 9.681291350164139e-06, + "loss": 0.2919, + "step": 1215 + }, + { + "epoch": 0.4060100166944908, + "grad_norm": 0.5859071316899785, + "learning_rate": 9.680266922744862e-06, + "loss": 0.3139, + "step": 1216 + }, + { + "epoch": 0.4063439065108514, + "grad_norm": 0.6074984980148432, + "learning_rate": 9.679240905931132e-06, + "loss": 0.3062, + "step": 1217 + }, + { + "epoch": 0.406677796327212, + "grad_norm": 0.5865179034668015, + "learning_rate": 9.678213300071387e-06, + "loss": 0.3086, + "step": 1218 + }, + { + "epoch": 0.40701168614357264, + "grad_norm": 0.6070760250741978, + "learning_rate": 9.677184105514592e-06, + "loss": 0.3167, + "step": 1219 + }, + { + "epoch": 0.4073455759599332, + "grad_norm": 0.590732884649166, + "learning_rate": 9.676153322610259e-06, + "loss": 0.3027, + "step": 1220 + }, + { + "epoch": 0.4076794657762938, + "grad_norm": 0.6151634532398677, + "learning_rate": 9.675120951708435e-06, + "loss": 0.3016, + "step": 1221 + }, + { + "epoch": 0.40801335559265445, + "grad_norm": 0.6363558602713849, + "learning_rate": 9.674086993159706e-06, + "loss": 0.3127, + "step": 1222 + }, + { + "epoch": 0.408347245409015, + "grad_norm": 0.5168442543052959, + "learning_rate": 9.673051447315202e-06, + "loss": 0.3075, + "step": 1223 + }, + { + "epoch": 0.40868113522537564, + "grad_norm": 0.5330952239512846, + "learning_rate": 9.672014314526588e-06, + "loss": 0.2896, + "step": 1224 + }, + { + "epoch": 0.4090150250417362, + "grad_norm": 0.56589887161752, + "learning_rate": 9.670975595146068e-06, + "loss": 0.2974, + "step": 1225 + }, + { + "epoch": 0.4093489148580968, + "grad_norm": 0.5111551537199233, + "learning_rate": 9.66993528952639e-06, + "loss": 0.2909, + "step": 1226 + }, + { + "epoch": 0.40968280467445745, + "grad_norm": 0.5351831782831177, + "learning_rate": 9.668893398020828e-06, + "loss": 0.2935, + "step": 1227 + }, + { + "epoch": 0.410016694490818, + "grad_norm": 0.5259432910511279, + "learning_rate": 9.66784992098321e-06, + "loss": 0.3096, + "step": 1228 + }, + { + "epoch": 0.41035058430717863, + "grad_norm": 0.5908438033944042, + "learning_rate": 9.666804858767894e-06, + "loss": 0.3037, + "step": 1229 + }, + { + "epoch": 0.41068447412353926, + "grad_norm": 0.5712044385230338, + "learning_rate": 9.665758211729776e-06, + "loss": 0.3004, + "step": 1230 + }, + { + "epoch": 0.4110183639398998, + "grad_norm": 0.5512384438906807, + "learning_rate": 9.664709980224295e-06, + "loss": 0.2898, + "step": 1231 + }, + { + "epoch": 0.41135225375626044, + "grad_norm": 0.5468730434988037, + "learning_rate": 9.66366016460742e-06, + "loss": 0.3001, + "step": 1232 + }, + { + "epoch": 0.411686143572621, + "grad_norm": 0.5650366046843425, + "learning_rate": 9.662608765235668e-06, + "loss": 0.3049, + "step": 1233 + }, + { + "epoch": 0.41202003338898163, + "grad_norm": 0.5521671648832258, + "learning_rate": 9.661555782466087e-06, + "loss": 0.2957, + "step": 1234 + }, + { + "epoch": 0.41235392320534225, + "grad_norm": 0.5572010081847173, + "learning_rate": 9.660501216656264e-06, + "loss": 0.3068, + "step": 1235 + }, + { + "epoch": 0.4126878130217028, + "grad_norm": 0.5742610063808997, + "learning_rate": 9.659445068164323e-06, + "loss": 0.3144, + "step": 1236 + }, + { + "epoch": 0.41302170283806344, + "grad_norm": 0.5519180103006606, + "learning_rate": 9.658387337348928e-06, + "loss": 0.3034, + "step": 1237 + }, + { + "epoch": 0.41335559265442406, + "grad_norm": 0.5815932719133985, + "learning_rate": 9.65732802456928e-06, + "loss": 0.3042, + "step": 1238 + }, + { + "epoch": 0.41368948247078463, + "grad_norm": 0.5284341657620202, + "learning_rate": 9.656267130185116e-06, + "loss": 0.2929, + "step": 1239 + }, + { + "epoch": 0.41402337228714525, + "grad_norm": 0.5845627083538882, + "learning_rate": 9.655204654556707e-06, + "loss": 0.3018, + "step": 1240 + }, + { + "epoch": 0.4143572621035058, + "grad_norm": 0.5592042633964295, + "learning_rate": 9.654140598044865e-06, + "loss": 0.299, + "step": 1241 + }, + { + "epoch": 0.41469115191986644, + "grad_norm": 0.5842006119789035, + "learning_rate": 9.653074961010944e-06, + "loss": 0.3024, + "step": 1242 + }, + { + "epoch": 0.41502504173622706, + "grad_norm": 0.5572016313161785, + "learning_rate": 9.65200774381682e-06, + "loss": 0.3104, + "step": 1243 + }, + { + "epoch": 0.41535893155258763, + "grad_norm": 0.5463513709776471, + "learning_rate": 9.650938946824921e-06, + "loss": 0.3202, + "step": 1244 + }, + { + "epoch": 0.41569282136894825, + "grad_norm": 0.5245642832747773, + "learning_rate": 9.649868570398204e-06, + "loss": 0.2954, + "step": 1245 + }, + { + "epoch": 0.4160267111853089, + "grad_norm": 0.5163540347800899, + "learning_rate": 9.64879661490016e-06, + "loss": 0.2898, + "step": 1246 + }, + { + "epoch": 0.41636060100166944, + "grad_norm": 0.5534838621722216, + "learning_rate": 9.647723080694822e-06, + "loss": 0.2955, + "step": 1247 + }, + { + "epoch": 0.41669449081803006, + "grad_norm": 0.5152066205911658, + "learning_rate": 9.646647968146757e-06, + "loss": 0.2977, + "step": 1248 + }, + { + "epoch": 0.4170283806343906, + "grad_norm": 0.5501246181817072, + "learning_rate": 9.645571277621066e-06, + "loss": 0.2895, + "step": 1249 + }, + { + "epoch": 0.41736227045075125, + "grad_norm": 0.5301671630413882, + "learning_rate": 9.644493009483389e-06, + "loss": 0.2947, + "step": 1250 + }, + { + "epoch": 0.41769616026711187, + "grad_norm": 0.5244215601463759, + "learning_rate": 9.6434131640999e-06, + "loss": 0.2981, + "step": 1251 + }, + { + "epoch": 0.41803005008347244, + "grad_norm": 0.5615774738408453, + "learning_rate": 9.642331741837307e-06, + "loss": 0.3058, + "step": 1252 + }, + { + "epoch": 0.41836393989983306, + "grad_norm": 0.5471985857205439, + "learning_rate": 9.641248743062858e-06, + "loss": 0.3087, + "step": 1253 + }, + { + "epoch": 0.4186978297161937, + "grad_norm": 0.5523449812980197, + "learning_rate": 9.640164168144332e-06, + "loss": 0.2881, + "step": 1254 + }, + { + "epoch": 0.41903171953255425, + "grad_norm": 0.56947894265333, + "learning_rate": 9.639078017450048e-06, + "loss": 0.3053, + "step": 1255 + }, + { + "epoch": 0.41936560934891487, + "grad_norm": 0.5266551079912, + "learning_rate": 9.637990291348853e-06, + "loss": 0.2998, + "step": 1256 + }, + { + "epoch": 0.41969949916527544, + "grad_norm": 0.5364026665958708, + "learning_rate": 9.636900990210136e-06, + "loss": 0.2931, + "step": 1257 + }, + { + "epoch": 0.42003338898163606, + "grad_norm": 0.5361579829994001, + "learning_rate": 9.635810114403818e-06, + "loss": 0.2933, + "step": 1258 + }, + { + "epoch": 0.4203672787979967, + "grad_norm": 0.5249718024570145, + "learning_rate": 9.634717664300352e-06, + "loss": 0.3078, + "step": 1259 + }, + { + "epoch": 0.42070116861435725, + "grad_norm": 0.5408474400256322, + "learning_rate": 9.633623640270733e-06, + "loss": 0.2948, + "step": 1260 + }, + { + "epoch": 0.42103505843071787, + "grad_norm": 0.5311288366614254, + "learning_rate": 9.632528042686482e-06, + "loss": 0.3068, + "step": 1261 + }, + { + "epoch": 0.4213689482470785, + "grad_norm": 0.5540851165525306, + "learning_rate": 9.631430871919659e-06, + "loss": 0.3125, + "step": 1262 + }, + { + "epoch": 0.42170283806343906, + "grad_norm": 0.5471114219357124, + "learning_rate": 9.630332128342858e-06, + "loss": 0.2972, + "step": 1263 + }, + { + "epoch": 0.4220367278797997, + "grad_norm": 0.5762114505810717, + "learning_rate": 9.629231812329207e-06, + "loss": 0.3117, + "step": 1264 + }, + { + "epoch": 0.42237061769616024, + "grad_norm": 0.49537883013774037, + "learning_rate": 9.628129924252368e-06, + "loss": 0.2868, + "step": 1265 + }, + { + "epoch": 0.42270450751252087, + "grad_norm": 0.5571681385249568, + "learning_rate": 9.627026464486535e-06, + "loss": 0.2864, + "step": 1266 + }, + { + "epoch": 0.4230383973288815, + "grad_norm": 0.5553613781084981, + "learning_rate": 9.62592143340644e-06, + "loss": 0.2955, + "step": 1267 + }, + { + "epoch": 0.42337228714524205, + "grad_norm": 0.5585769058336895, + "learning_rate": 9.624814831387342e-06, + "loss": 0.3072, + "step": 1268 + }, + { + "epoch": 0.4237061769616027, + "grad_norm": 0.5377093487602751, + "learning_rate": 9.62370665880504e-06, + "loss": 0.2882, + "step": 1269 + }, + { + "epoch": 0.4240400667779633, + "grad_norm": 0.546543632666723, + "learning_rate": 9.622596916035862e-06, + "loss": 0.3009, + "step": 1270 + }, + { + "epoch": 0.42437395659432386, + "grad_norm": 0.5475282060716415, + "learning_rate": 9.621485603456673e-06, + "loss": 0.3037, + "step": 1271 + }, + { + "epoch": 0.4247078464106845, + "grad_norm": 0.5687255993925933, + "learning_rate": 9.620372721444868e-06, + "loss": 0.3116, + "step": 1272 + }, + { + "epoch": 0.42504173622704505, + "grad_norm": 0.5581094852805137, + "learning_rate": 9.619258270378375e-06, + "loss": 0.303, + "step": 1273 + }, + { + "epoch": 0.4253756260434057, + "grad_norm": 0.6196835765753521, + "learning_rate": 9.618142250635658e-06, + "loss": 0.3109, + "step": 1274 + }, + { + "epoch": 0.4257095158597663, + "grad_norm": 0.5271374821802719, + "learning_rate": 9.61702466259571e-06, + "loss": 0.2908, + "step": 1275 + }, + { + "epoch": 0.42604340567612686, + "grad_norm": 0.5541278409309868, + "learning_rate": 9.615905506638056e-06, + "loss": 0.2997, + "step": 1276 + }, + { + "epoch": 0.4263772954924875, + "grad_norm": 0.5656872256523298, + "learning_rate": 9.614784783142759e-06, + "loss": 0.2926, + "step": 1277 + }, + { + "epoch": 0.4267111853088481, + "grad_norm": 0.5355707640270149, + "learning_rate": 9.613662492490409e-06, + "loss": 0.2955, + "step": 1278 + }, + { + "epoch": 0.4270450751252087, + "grad_norm": 0.5942210115051261, + "learning_rate": 9.612538635062134e-06, + "loss": 0.2915, + "step": 1279 + }, + { + "epoch": 0.4273789649415693, + "grad_norm": 0.5764678944893045, + "learning_rate": 9.611413211239585e-06, + "loss": 0.2915, + "step": 1280 + }, + { + "epoch": 0.42771285475792986, + "grad_norm": 0.530091965453539, + "learning_rate": 9.610286221404951e-06, + "loss": 0.305, + "step": 1281 + }, + { + "epoch": 0.4280467445742905, + "grad_norm": 0.5494839841400664, + "learning_rate": 9.609157665940956e-06, + "loss": 0.3094, + "step": 1282 + }, + { + "epoch": 0.4283806343906511, + "grad_norm": 0.5252597376389204, + "learning_rate": 9.608027545230847e-06, + "loss": 0.2984, + "step": 1283 + }, + { + "epoch": 0.42871452420701167, + "grad_norm": 0.5585599836261066, + "learning_rate": 9.606895859658408e-06, + "loss": 0.2964, + "step": 1284 + }, + { + "epoch": 0.4290484140233723, + "grad_norm": 0.5540080298072316, + "learning_rate": 9.605762609607955e-06, + "loss": 0.3161, + "step": 1285 + }, + { + "epoch": 0.4293823038397329, + "grad_norm": 0.5551516465719891, + "learning_rate": 9.604627795464335e-06, + "loss": 0.3042, + "step": 1286 + }, + { + "epoch": 0.4297161936560935, + "grad_norm": 0.6336663629097926, + "learning_rate": 9.60349141761292e-06, + "loss": 0.3043, + "step": 1287 + }, + { + "epoch": 0.4300500834724541, + "grad_norm": 0.5334894094431695, + "learning_rate": 9.602353476439624e-06, + "loss": 0.306, + "step": 1288 + }, + { + "epoch": 0.43038397328881467, + "grad_norm": 0.527209258392072, + "learning_rate": 9.601213972330883e-06, + "loss": 0.3137, + "step": 1289 + }, + { + "epoch": 0.4307178631051753, + "grad_norm": 0.5952846356350251, + "learning_rate": 9.600072905673665e-06, + "loss": 0.3042, + "step": 1290 + }, + { + "epoch": 0.4310517529215359, + "grad_norm": 0.5602989730892238, + "learning_rate": 9.598930276855474e-06, + "loss": 0.3086, + "step": 1291 + }, + { + "epoch": 0.4313856427378965, + "grad_norm": 0.5320832129268293, + "learning_rate": 9.597786086264339e-06, + "loss": 0.3028, + "step": 1292 + }, + { + "epoch": 0.4317195325542571, + "grad_norm": 0.5670939025122391, + "learning_rate": 9.596640334288822e-06, + "loss": 0.3154, + "step": 1293 + }, + { + "epoch": 0.4320534223706177, + "grad_norm": 0.5859769878978323, + "learning_rate": 9.595493021318012e-06, + "loss": 0.3001, + "step": 1294 + }, + { + "epoch": 0.4323873121869783, + "grad_norm": 0.5344698555651072, + "learning_rate": 9.594344147741534e-06, + "loss": 0.2858, + "step": 1295 + }, + { + "epoch": 0.4327212020033389, + "grad_norm": 0.5350063831062745, + "learning_rate": 9.593193713949537e-06, + "loss": 0.2912, + "step": 1296 + }, + { + "epoch": 0.4330550918196995, + "grad_norm": 0.5418231181346506, + "learning_rate": 9.592041720332703e-06, + "loss": 0.308, + "step": 1297 + }, + { + "epoch": 0.4333889816360601, + "grad_norm": 0.5767610177247011, + "learning_rate": 9.590888167282245e-06, + "loss": 0.3005, + "step": 1298 + }, + { + "epoch": 0.4337228714524207, + "grad_norm": 0.5284612330462783, + "learning_rate": 9.5897330551899e-06, + "loss": 0.3004, + "step": 1299 + }, + { + "epoch": 0.4340567612687813, + "grad_norm": 0.5595979475668135, + "learning_rate": 9.58857638444794e-06, + "loss": 0.3053, + "step": 1300 + }, + { + "epoch": 0.4343906510851419, + "grad_norm": 0.641971687130429, + "learning_rate": 9.587418155449167e-06, + "loss": 0.2947, + "step": 1301 + }, + { + "epoch": 0.43472454090150253, + "grad_norm": 0.5588665476946711, + "learning_rate": 9.586258368586906e-06, + "loss": 0.3108, + "step": 1302 + }, + { + "epoch": 0.4350584307178631, + "grad_norm": 0.5293125361128987, + "learning_rate": 9.585097024255013e-06, + "loss": 0.2932, + "step": 1303 + }, + { + "epoch": 0.4353923205342237, + "grad_norm": 0.5419949614024012, + "learning_rate": 9.58393412284788e-06, + "loss": 0.2995, + "step": 1304 + }, + { + "epoch": 0.4357262103505843, + "grad_norm": 0.6029592716866153, + "learning_rate": 9.582769664760418e-06, + "loss": 0.3126, + "step": 1305 + }, + { + "epoch": 0.4360601001669449, + "grad_norm": 0.5288675727847671, + "learning_rate": 9.581603650388073e-06, + "loss": 0.2942, + "step": 1306 + }, + { + "epoch": 0.43639398998330553, + "grad_norm": 0.5473515370138343, + "learning_rate": 9.580436080126819e-06, + "loss": 0.3046, + "step": 1307 + }, + { + "epoch": 0.4367278797996661, + "grad_norm": 0.5401523800799727, + "learning_rate": 9.579266954373152e-06, + "loss": 0.3007, + "step": 1308 + }, + { + "epoch": 0.4370617696160267, + "grad_norm": 0.5283115315624536, + "learning_rate": 9.578096273524104e-06, + "loss": 0.31, + "step": 1309 + }, + { + "epoch": 0.4373956594323873, + "grad_norm": 0.5255765399122286, + "learning_rate": 9.576924037977233e-06, + "loss": 0.2983, + "step": 1310 + }, + { + "epoch": 0.4377295492487479, + "grad_norm": 0.5758580308207852, + "learning_rate": 9.575750248130623e-06, + "loss": 0.2971, + "step": 1311 + }, + { + "epoch": 0.43806343906510853, + "grad_norm": 0.5688979236305017, + "learning_rate": 9.574574904382887e-06, + "loss": 0.3074, + "step": 1312 + }, + { + "epoch": 0.4383973288814691, + "grad_norm": 0.5534655263495765, + "learning_rate": 9.573398007133166e-06, + "loss": 0.2961, + "step": 1313 + }, + { + "epoch": 0.4387312186978297, + "grad_norm": 0.5351312447075476, + "learning_rate": 9.572219556781126e-06, + "loss": 0.3102, + "step": 1314 + }, + { + "epoch": 0.43906510851419034, + "grad_norm": 0.5519599682845804, + "learning_rate": 9.571039553726966e-06, + "loss": 0.2957, + "step": 1315 + }, + { + "epoch": 0.4393989983305509, + "grad_norm": 0.5857085492033481, + "learning_rate": 9.569857998371408e-06, + "loss": 0.3096, + "step": 1316 + }, + { + "epoch": 0.4397328881469115, + "grad_norm": 0.5086322284643087, + "learning_rate": 9.5686748911157e-06, + "loss": 0.2911, + "step": 1317 + }, + { + "epoch": 0.4400667779632721, + "grad_norm": 0.5539388745461533, + "learning_rate": 9.567490232361623e-06, + "loss": 0.3081, + "step": 1318 + }, + { + "epoch": 0.4404006677796327, + "grad_norm": 0.5199732188250088, + "learning_rate": 9.566304022511477e-06, + "loss": 0.2846, + "step": 1319 + }, + { + "epoch": 0.44073455759599334, + "grad_norm": 0.5541766933610613, + "learning_rate": 9.565116261968095e-06, + "loss": 0.2984, + "step": 1320 + }, + { + "epoch": 0.4410684474123539, + "grad_norm": 0.5147298635976479, + "learning_rate": 9.563926951134834e-06, + "loss": 0.3001, + "step": 1321 + }, + { + "epoch": 0.4414023372287145, + "grad_norm": 0.5069931069072989, + "learning_rate": 9.562736090415577e-06, + "loss": 0.2903, + "step": 1322 + }, + { + "epoch": 0.44173622704507515, + "grad_norm": 0.5194849623989878, + "learning_rate": 9.561543680214735e-06, + "loss": 0.2917, + "step": 1323 + }, + { + "epoch": 0.4420701168614357, + "grad_norm": 0.5574203139795856, + "learning_rate": 9.560349720937246e-06, + "loss": 0.3025, + "step": 1324 + }, + { + "epoch": 0.44240400667779634, + "grad_norm": 0.5629801166291726, + "learning_rate": 9.559154212988566e-06, + "loss": 0.2935, + "step": 1325 + }, + { + "epoch": 0.4427378964941569, + "grad_norm": 0.5289970761915245, + "learning_rate": 9.55795715677469e-06, + "loss": 0.2815, + "step": 1326 + }, + { + "epoch": 0.4430717863105175, + "grad_norm": 0.5088579715144295, + "learning_rate": 9.556758552702132e-06, + "loss": 0.3103, + "step": 1327 + }, + { + "epoch": 0.44340567612687815, + "grad_norm": 0.5451705734463403, + "learning_rate": 9.555558401177927e-06, + "loss": 0.3145, + "step": 1328 + }, + { + "epoch": 0.4437395659432387, + "grad_norm": 0.536532035655797, + "learning_rate": 9.554356702609645e-06, + "loss": 0.2916, + "step": 1329 + }, + { + "epoch": 0.44407345575959933, + "grad_norm": 0.5536592744549156, + "learning_rate": 9.553153457405373e-06, + "loss": 0.286, + "step": 1330 + }, + { + "epoch": 0.44440734557595996, + "grad_norm": 0.5251201621820176, + "learning_rate": 9.551948665973728e-06, + "loss": 0.3061, + "step": 1331 + }, + { + "epoch": 0.4447412353923205, + "grad_norm": 0.5233246352076496, + "learning_rate": 9.550742328723852e-06, + "loss": 0.2902, + "step": 1332 + }, + { + "epoch": 0.44507512520868114, + "grad_norm": 0.5403890557149773, + "learning_rate": 9.54953444606541e-06, + "loss": 0.2858, + "step": 1333 + }, + { + "epoch": 0.4454090150250417, + "grad_norm": 0.5277525452543673, + "learning_rate": 9.548325018408592e-06, + "loss": 0.2929, + "step": 1334 + }, + { + "epoch": 0.44574290484140233, + "grad_norm": 0.5204604845590859, + "learning_rate": 9.547114046164116e-06, + "loss": 0.296, + "step": 1335 + }, + { + "epoch": 0.44607679465776295, + "grad_norm": 0.5160849657956021, + "learning_rate": 9.54590152974322e-06, + "loss": 0.2871, + "step": 1336 + }, + { + "epoch": 0.4464106844741235, + "grad_norm": 0.5473436548332656, + "learning_rate": 9.544687469557667e-06, + "loss": 0.2959, + "step": 1337 + }, + { + "epoch": 0.44674457429048414, + "grad_norm": 0.5182203802196708, + "learning_rate": 9.543471866019748e-06, + "loss": 0.3068, + "step": 1338 + }, + { + "epoch": 0.44707846410684476, + "grad_norm": 0.5157608820284589, + "learning_rate": 9.542254719542274e-06, + "loss": 0.2888, + "step": 1339 + }, + { + "epoch": 0.44741235392320533, + "grad_norm": 0.5421160723835827, + "learning_rate": 9.541036030538584e-06, + "loss": 0.3047, + "step": 1340 + }, + { + "epoch": 0.44774624373956595, + "grad_norm": 0.5210602370605848, + "learning_rate": 9.539815799422536e-06, + "loss": 0.3059, + "step": 1341 + }, + { + "epoch": 0.4480801335559265, + "grad_norm": 0.5169521578998862, + "learning_rate": 9.538594026608514e-06, + "loss": 0.3019, + "step": 1342 + }, + { + "epoch": 0.44841402337228714, + "grad_norm": 0.5328147586982214, + "learning_rate": 9.537370712511427e-06, + "loss": 0.3028, + "step": 1343 + }, + { + "epoch": 0.44874791318864776, + "grad_norm": 0.5388042400726079, + "learning_rate": 9.536145857546708e-06, + "loss": 0.2937, + "step": 1344 + }, + { + "epoch": 0.44908180300500833, + "grad_norm": 0.5127827771362384, + "learning_rate": 9.534919462130308e-06, + "loss": 0.2823, + "step": 1345 + }, + { + "epoch": 0.44941569282136895, + "grad_norm": 0.5298121554492677, + "learning_rate": 9.533691526678705e-06, + "loss": 0.2868, + "step": 1346 + }, + { + "epoch": 0.4497495826377296, + "grad_norm": 0.49035926725963647, + "learning_rate": 9.532462051608901e-06, + "loss": 0.3014, + "step": 1347 + }, + { + "epoch": 0.45008347245409014, + "grad_norm": 0.5114865019388333, + "learning_rate": 9.53123103733842e-06, + "loss": 0.3046, + "step": 1348 + }, + { + "epoch": 0.45041736227045076, + "grad_norm": 0.5618633438957532, + "learning_rate": 9.529998484285306e-06, + "loss": 0.2888, + "step": 1349 + }, + { + "epoch": 0.4507512520868113, + "grad_norm": 0.5650173727249157, + "learning_rate": 9.528764392868129e-06, + "loss": 0.301, + "step": 1350 + }, + { + "epoch": 0.45108514190317195, + "grad_norm": 0.5268869489595138, + "learning_rate": 9.527528763505978e-06, + "loss": 0.3038, + "step": 1351 + }, + { + "epoch": 0.45141903171953257, + "grad_norm": 0.5569471476250318, + "learning_rate": 9.526291596618469e-06, + "loss": 0.3233, + "step": 1352 + }, + { + "epoch": 0.45175292153589314, + "grad_norm": 0.5314512031159738, + "learning_rate": 9.525052892625734e-06, + "loss": 0.2991, + "step": 1353 + }, + { + "epoch": 0.45208681135225376, + "grad_norm": 0.5347885069331869, + "learning_rate": 9.523812651948437e-06, + "loss": 0.292, + "step": 1354 + }, + { + "epoch": 0.4524207011686144, + "grad_norm": 0.5371337503963696, + "learning_rate": 9.52257087500775e-06, + "loss": 0.3032, + "step": 1355 + }, + { + "epoch": 0.45275459098497495, + "grad_norm": 0.5612730842946464, + "learning_rate": 9.521327562225378e-06, + "loss": 0.3041, + "step": 1356 + }, + { + "epoch": 0.45308848080133557, + "grad_norm": 0.5013811030870067, + "learning_rate": 9.520082714023544e-06, + "loss": 0.284, + "step": 1357 + }, + { + "epoch": 0.45342237061769614, + "grad_norm": 0.508046347723654, + "learning_rate": 9.51883633082499e-06, + "loss": 0.3066, + "step": 1358 + }, + { + "epoch": 0.45375626043405676, + "grad_norm": 0.5116271869855235, + "learning_rate": 9.517588413052983e-06, + "loss": 0.2965, + "step": 1359 + }, + { + "epoch": 0.4540901502504174, + "grad_norm": 0.5386515677082221, + "learning_rate": 9.516338961131305e-06, + "loss": 0.2899, + "step": 1360 + }, + { + "epoch": 0.45442404006677795, + "grad_norm": 0.5447373690787114, + "learning_rate": 9.515087975484271e-06, + "loss": 0.2895, + "step": 1361 + }, + { + "epoch": 0.45475792988313857, + "grad_norm": 0.508366388180615, + "learning_rate": 9.513835456536704e-06, + "loss": 0.2947, + "step": 1362 + }, + { + "epoch": 0.4550918196994992, + "grad_norm": 0.563943487791508, + "learning_rate": 9.512581404713954e-06, + "loss": 0.303, + "step": 1363 + }, + { + "epoch": 0.45542570951585976, + "grad_norm": 0.5479482515066728, + "learning_rate": 9.51132582044189e-06, + "loss": 0.298, + "step": 1364 + }, + { + "epoch": 0.4557595993322204, + "grad_norm": 0.5260017880492549, + "learning_rate": 9.510068704146905e-06, + "loss": 0.2954, + "step": 1365 + }, + { + "epoch": 0.45609348914858094, + "grad_norm": 0.5169454932558921, + "learning_rate": 9.508810056255903e-06, + "loss": 0.2893, + "step": 1366 + }, + { + "epoch": 0.45642737896494157, + "grad_norm": 0.5360744753729512, + "learning_rate": 9.50754987719632e-06, + "loss": 0.315, + "step": 1367 + }, + { + "epoch": 0.4567612687813022, + "grad_norm": 0.5224151119896789, + "learning_rate": 9.506288167396106e-06, + "loss": 0.3059, + "step": 1368 + }, + { + "epoch": 0.45709515859766275, + "grad_norm": 0.5166269968688816, + "learning_rate": 9.505024927283729e-06, + "loss": 0.2933, + "step": 1369 + }, + { + "epoch": 0.4574290484140234, + "grad_norm": 0.5120953078088812, + "learning_rate": 9.503760157288178e-06, + "loss": 0.2883, + "step": 1370 + }, + { + "epoch": 0.457762938230384, + "grad_norm": 0.5391588591206656, + "learning_rate": 9.502493857838963e-06, + "loss": 0.2993, + "step": 1371 + }, + { + "epoch": 0.45809682804674456, + "grad_norm": 0.525997405236799, + "learning_rate": 9.501226029366115e-06, + "loss": 0.2984, + "step": 1372 + }, + { + "epoch": 0.4584307178631052, + "grad_norm": 0.5349980972727626, + "learning_rate": 9.49995667230018e-06, + "loss": 0.2993, + "step": 1373 + }, + { + "epoch": 0.45876460767946575, + "grad_norm": 0.5337180105996955, + "learning_rate": 9.498685787072225e-06, + "loss": 0.3144, + "step": 1374 + }, + { + "epoch": 0.4590984974958264, + "grad_norm": 0.5325557358735868, + "learning_rate": 9.497413374113838e-06, + "loss": 0.305, + "step": 1375 + }, + { + "epoch": 0.459432387312187, + "grad_norm": 0.5654856210951742, + "learning_rate": 9.49613943385712e-06, + "loss": 0.2978, + "step": 1376 + }, + { + "epoch": 0.45976627712854756, + "grad_norm": 0.528861157248461, + "learning_rate": 9.494863966734698e-06, + "loss": 0.2927, + "step": 1377 + }, + { + "epoch": 0.4601001669449082, + "grad_norm": 0.5676591915965933, + "learning_rate": 9.493586973179714e-06, + "loss": 0.2986, + "step": 1378 + }, + { + "epoch": 0.4604340567612688, + "grad_norm": 0.5228376473966729, + "learning_rate": 9.492308453625828e-06, + "loss": 0.2971, + "step": 1379 + }, + { + "epoch": 0.4607679465776294, + "grad_norm": 0.5386374313346948, + "learning_rate": 9.491028408507215e-06, + "loss": 0.3018, + "step": 1380 + }, + { + "epoch": 0.46110183639399, + "grad_norm": 0.598948572070358, + "learning_rate": 9.489746838258578e-06, + "loss": 0.3056, + "step": 1381 + }, + { + "epoch": 0.46143572621035056, + "grad_norm": 0.5369176185004423, + "learning_rate": 9.488463743315126e-06, + "loss": 0.2928, + "step": 1382 + }, + { + "epoch": 0.4617696160267112, + "grad_norm": 0.5298069943833881, + "learning_rate": 9.487179124112596e-06, + "loss": 0.3039, + "step": 1383 + }, + { + "epoch": 0.4621035058430718, + "grad_norm": 0.5666498816292825, + "learning_rate": 9.485892981087234e-06, + "loss": 0.3045, + "step": 1384 + }, + { + "epoch": 0.46243739565943237, + "grad_norm": 0.5509388939957, + "learning_rate": 9.48460531467581e-06, + "loss": 0.2906, + "step": 1385 + }, + { + "epoch": 0.462771285475793, + "grad_norm": 0.5255599607029671, + "learning_rate": 9.483316125315607e-06, + "loss": 0.2944, + "step": 1386 + }, + { + "epoch": 0.4631051752921536, + "grad_norm": 0.5229939853896968, + "learning_rate": 9.48202541344443e-06, + "loss": 0.2875, + "step": 1387 + }, + { + "epoch": 0.4634390651085142, + "grad_norm": 0.5840283539932745, + "learning_rate": 9.480733179500596e-06, + "loss": 0.3016, + "step": 1388 + }, + { + "epoch": 0.4637729549248748, + "grad_norm": 0.6502792216581386, + "learning_rate": 9.47943942392294e-06, + "loss": 0.3088, + "step": 1389 + }, + { + "epoch": 0.46410684474123537, + "grad_norm": 0.532773816540503, + "learning_rate": 9.478144147150818e-06, + "loss": 0.3113, + "step": 1390 + }, + { + "epoch": 0.464440734557596, + "grad_norm": 0.5523472472251545, + "learning_rate": 9.476847349624097e-06, + "loss": 0.3099, + "step": 1391 + }, + { + "epoch": 0.4647746243739566, + "grad_norm": 0.6282109751948276, + "learning_rate": 9.475549031783161e-06, + "loss": 0.2975, + "step": 1392 + }, + { + "epoch": 0.4651085141903172, + "grad_norm": 0.5596679298027152, + "learning_rate": 9.474249194068915e-06, + "loss": 0.3013, + "step": 1393 + }, + { + "epoch": 0.4654424040066778, + "grad_norm": 0.5412908518717302, + "learning_rate": 9.472947836922776e-06, + "loss": 0.2963, + "step": 1394 + }, + { + "epoch": 0.4657762938230384, + "grad_norm": 0.5580863318279337, + "learning_rate": 9.47164496078668e-06, + "loss": 0.3009, + "step": 1395 + }, + { + "epoch": 0.466110183639399, + "grad_norm": 0.5515053169154321, + "learning_rate": 9.470340566103074e-06, + "loss": 0.2932, + "step": 1396 + }, + { + "epoch": 0.4664440734557596, + "grad_norm": 0.5512982187051885, + "learning_rate": 9.469034653314927e-06, + "loss": 0.2929, + "step": 1397 + }, + { + "epoch": 0.4667779632721202, + "grad_norm": 0.5813956285889033, + "learning_rate": 9.467727222865715e-06, + "loss": 0.3065, + "step": 1398 + }, + { + "epoch": 0.4671118530884808, + "grad_norm": 0.5336198594455567, + "learning_rate": 9.46641827519944e-06, + "loss": 0.2975, + "step": 1399 + }, + { + "epoch": 0.4674457429048414, + "grad_norm": 0.5289011306590928, + "learning_rate": 9.46510781076061e-06, + "loss": 0.2957, + "step": 1400 + }, + { + "epoch": 0.467779632721202, + "grad_norm": 0.5601901926538205, + "learning_rate": 9.463795829994255e-06, + "loss": 0.2984, + "step": 1401 + }, + { + "epoch": 0.4681135225375626, + "grad_norm": 0.559465202404947, + "learning_rate": 9.462482333345917e-06, + "loss": 0.2963, + "step": 1402 + }, + { + "epoch": 0.46844741235392323, + "grad_norm": 0.531521786796366, + "learning_rate": 9.46116732126165e-06, + "loss": 0.293, + "step": 1403 + }, + { + "epoch": 0.4687813021702838, + "grad_norm": 0.5483608878898986, + "learning_rate": 9.459850794188026e-06, + "loss": 0.3094, + "step": 1404 + }, + { + "epoch": 0.4691151919866444, + "grad_norm": 0.5327871632496454, + "learning_rate": 9.458532752572131e-06, + "loss": 0.31, + "step": 1405 + }, + { + "epoch": 0.469449081803005, + "grad_norm": 0.5422669025509353, + "learning_rate": 9.457213196861567e-06, + "loss": 0.3011, + "step": 1406 + }, + { + "epoch": 0.4697829716193656, + "grad_norm": 0.604834913326236, + "learning_rate": 9.455892127504447e-06, + "loss": 0.2912, + "step": 1407 + }, + { + "epoch": 0.47011686143572623, + "grad_norm": 0.5090166304034505, + "learning_rate": 9.454569544949397e-06, + "loss": 0.2906, + "step": 1408 + }, + { + "epoch": 0.4704507512520868, + "grad_norm": 0.5386516225647907, + "learning_rate": 9.453245449645562e-06, + "loss": 0.2874, + "step": 1409 + }, + { + "epoch": 0.4707846410684474, + "grad_norm": 0.5289619070635977, + "learning_rate": 9.451919842042599e-06, + "loss": 0.3072, + "step": 1410 + }, + { + "epoch": 0.47111853088480804, + "grad_norm": 0.5549075726915047, + "learning_rate": 9.450592722590675e-06, + "loss": 0.2977, + "step": 1411 + }, + { + "epoch": 0.4714524207011686, + "grad_norm": 0.5458874938771535, + "learning_rate": 9.449264091740473e-06, + "loss": 0.3005, + "step": 1412 + }, + { + "epoch": 0.47178631051752923, + "grad_norm": 0.5176103324234522, + "learning_rate": 9.447933949943188e-06, + "loss": 0.2911, + "step": 1413 + }, + { + "epoch": 0.4721202003338898, + "grad_norm": 0.5548560025418969, + "learning_rate": 9.446602297650533e-06, + "loss": 0.2964, + "step": 1414 + }, + { + "epoch": 0.4724540901502504, + "grad_norm": 0.5234196984551414, + "learning_rate": 9.445269135314729e-06, + "loss": 0.2868, + "step": 1415 + }, + { + "epoch": 0.47278797996661104, + "grad_norm": 0.5541770426664847, + "learning_rate": 9.443934463388509e-06, + "loss": 0.2978, + "step": 1416 + }, + { + "epoch": 0.4731218697829716, + "grad_norm": 0.5421518784389697, + "learning_rate": 9.442598282325124e-06, + "loss": 0.2866, + "step": 1417 + }, + { + "epoch": 0.4734557595993322, + "grad_norm": 0.5393371606474158, + "learning_rate": 9.44126059257833e-06, + "loss": 0.3018, + "step": 1418 + }, + { + "epoch": 0.47378964941569285, + "grad_norm": 0.5123603437120414, + "learning_rate": 9.439921394602403e-06, + "loss": 0.3041, + "step": 1419 + }, + { + "epoch": 0.4741235392320534, + "grad_norm": 0.5159057782096642, + "learning_rate": 9.438580688852126e-06, + "loss": 0.2964, + "step": 1420 + }, + { + "epoch": 0.47445742904841404, + "grad_norm": 0.5680453048254843, + "learning_rate": 9.437238475782795e-06, + "loss": 0.2849, + "step": 1421 + }, + { + "epoch": 0.4747913188647746, + "grad_norm": 0.5678997720118169, + "learning_rate": 9.43589475585022e-06, + "loss": 0.3138, + "step": 1422 + }, + { + "epoch": 0.4751252086811352, + "grad_norm": 0.5099324389746028, + "learning_rate": 9.434549529510724e-06, + "loss": 0.2884, + "step": 1423 + }, + { + "epoch": 0.47545909849749585, + "grad_norm": 0.5596741951721991, + "learning_rate": 9.433202797221132e-06, + "loss": 0.2996, + "step": 1424 + }, + { + "epoch": 0.4757929883138564, + "grad_norm": 0.5368014359113862, + "learning_rate": 9.431854559438795e-06, + "loss": 0.3101, + "step": 1425 + }, + { + "epoch": 0.47612687813021703, + "grad_norm": 0.5690073354713124, + "learning_rate": 9.430504816621562e-06, + "loss": 0.3131, + "step": 1426 + }, + { + "epoch": 0.47646076794657766, + "grad_norm": 0.518928365164336, + "learning_rate": 9.4291535692278e-06, + "loss": 0.2899, + "step": 1427 + }, + { + "epoch": 0.4767946577629382, + "grad_norm": 0.5010790799906752, + "learning_rate": 9.427800817716388e-06, + "loss": 0.2848, + "step": 1428 + }, + { + "epoch": 0.47712854757929885, + "grad_norm": 0.5110806956881834, + "learning_rate": 9.426446562546711e-06, + "loss": 0.2826, + "step": 1429 + }, + { + "epoch": 0.4774624373956594, + "grad_norm": 0.5392888773565416, + "learning_rate": 9.425090804178667e-06, + "loss": 0.2985, + "step": 1430 + }, + { + "epoch": 0.47779632721202003, + "grad_norm": 0.5358209940931588, + "learning_rate": 9.423733543072666e-06, + "loss": 0.2951, + "step": 1431 + }, + { + "epoch": 0.47813021702838066, + "grad_norm": 0.5006563636397046, + "learning_rate": 9.422374779689625e-06, + "loss": 0.2943, + "step": 1432 + }, + { + "epoch": 0.4784641068447412, + "grad_norm": 0.5201171085186808, + "learning_rate": 9.421014514490977e-06, + "loss": 0.295, + "step": 1433 + }, + { + "epoch": 0.47879799666110184, + "grad_norm": 0.5339236085344619, + "learning_rate": 9.419652747938655e-06, + "loss": 0.3049, + "step": 1434 + }, + { + "epoch": 0.4791318864774624, + "grad_norm": 0.49973961019267693, + "learning_rate": 9.418289480495114e-06, + "loss": 0.279, + "step": 1435 + }, + { + "epoch": 0.47946577629382303, + "grad_norm": 0.5565671753366227, + "learning_rate": 9.416924712623305e-06, + "loss": 0.3016, + "step": 1436 + }, + { + "epoch": 0.47979966611018365, + "grad_norm": 0.528593474591798, + "learning_rate": 9.415558444786706e-06, + "loss": 0.2907, + "step": 1437 + }, + { + "epoch": 0.4801335559265442, + "grad_norm": 0.5351159334880345, + "learning_rate": 9.414190677449285e-06, + "loss": 0.3027, + "step": 1438 + }, + { + "epoch": 0.48046744574290484, + "grad_norm": 0.5473225197861933, + "learning_rate": 9.412821411075535e-06, + "loss": 0.2998, + "step": 1439 + }, + { + "epoch": 0.48080133555926546, + "grad_norm": 0.49704032252032326, + "learning_rate": 9.411450646130448e-06, + "loss": 0.2996, + "step": 1440 + }, + { + "epoch": 0.48113522537562603, + "grad_norm": 0.49878457433805434, + "learning_rate": 9.410078383079532e-06, + "loss": 0.295, + "step": 1441 + }, + { + "epoch": 0.48146911519198665, + "grad_norm": 0.5110800066394249, + "learning_rate": 9.4087046223888e-06, + "loss": 0.2917, + "step": 1442 + }, + { + "epoch": 0.4818030050083472, + "grad_norm": 0.48538237976961035, + "learning_rate": 9.40732936452477e-06, + "loss": 0.2903, + "step": 1443 + }, + { + "epoch": 0.48213689482470784, + "grad_norm": 0.5168800619542065, + "learning_rate": 9.405952609954478e-06, + "loss": 0.2816, + "step": 1444 + }, + { + "epoch": 0.48247078464106846, + "grad_norm": 0.5416468534828895, + "learning_rate": 9.40457435914546e-06, + "loss": 0.3012, + "step": 1445 + }, + { + "epoch": 0.48280467445742903, + "grad_norm": 0.5338609939340626, + "learning_rate": 9.40319461256576e-06, + "loss": 0.3054, + "step": 1446 + }, + { + "epoch": 0.48313856427378965, + "grad_norm": 0.5335812234450379, + "learning_rate": 9.401813370683938e-06, + "loss": 0.2957, + "step": 1447 + }, + { + "epoch": 0.48347245409015027, + "grad_norm": 0.49346589849626954, + "learning_rate": 9.400430633969054e-06, + "loss": 0.2869, + "step": 1448 + }, + { + "epoch": 0.48380634390651084, + "grad_norm": 0.5285995416526085, + "learning_rate": 9.39904640289068e-06, + "loss": 0.291, + "step": 1449 + }, + { + "epoch": 0.48414023372287146, + "grad_norm": 0.4944627975585281, + "learning_rate": 9.39766067791889e-06, + "loss": 0.2873, + "step": 1450 + }, + { + "epoch": 0.484474123539232, + "grad_norm": 0.5196538712421545, + "learning_rate": 9.396273459524272e-06, + "loss": 0.3006, + "step": 1451 + }, + { + "epoch": 0.48480801335559265, + "grad_norm": 0.5344641159334709, + "learning_rate": 9.394884748177917e-06, + "loss": 0.2905, + "step": 1452 + }, + { + "epoch": 0.48514190317195327, + "grad_norm": 0.4999831867671009, + "learning_rate": 9.393494544351425e-06, + "loss": 0.2964, + "step": 1453 + }, + { + "epoch": 0.48547579298831384, + "grad_norm": 0.5054673672154285, + "learning_rate": 9.392102848516901e-06, + "loss": 0.2803, + "step": 1454 + }, + { + "epoch": 0.48580968280467446, + "grad_norm": 0.49329425468244137, + "learning_rate": 9.39070966114696e-06, + "loss": 0.2931, + "step": 1455 + }, + { + "epoch": 0.4861435726210351, + "grad_norm": 0.515091593170369, + "learning_rate": 9.389314982714718e-06, + "loss": 0.2993, + "step": 1456 + }, + { + "epoch": 0.48647746243739565, + "grad_norm": 0.5085262662418445, + "learning_rate": 9.387918813693805e-06, + "loss": 0.303, + "step": 1457 + }, + { + "epoch": 0.48681135225375627, + "grad_norm": 0.5273045474850394, + "learning_rate": 9.386521154558347e-06, + "loss": 0.3014, + "step": 1458 + }, + { + "epoch": 0.48714524207011684, + "grad_norm": 0.5070425174179797, + "learning_rate": 9.385122005782986e-06, + "loss": 0.2913, + "step": 1459 + }, + { + "epoch": 0.48747913188647746, + "grad_norm": 0.5082781952196401, + "learning_rate": 9.383721367842867e-06, + "loss": 0.2956, + "step": 1460 + }, + { + "epoch": 0.4878130217028381, + "grad_norm": 0.5297329535598102, + "learning_rate": 9.382319241213633e-06, + "loss": 0.3048, + "step": 1461 + }, + { + "epoch": 0.48814691151919865, + "grad_norm": 0.5296936016363766, + "learning_rate": 9.380915626371445e-06, + "loss": 0.3128, + "step": 1462 + }, + { + "epoch": 0.48848080133555927, + "grad_norm": 0.4953997988950679, + "learning_rate": 9.37951052379296e-06, + "loss": 0.287, + "step": 1463 + }, + { + "epoch": 0.4888146911519199, + "grad_norm": 0.5258513951922617, + "learning_rate": 9.378103933955345e-06, + "loss": 0.3078, + "step": 1464 + }, + { + "epoch": 0.48914858096828046, + "grad_norm": 0.5148051892537676, + "learning_rate": 9.37669585733627e-06, + "loss": 0.292, + "step": 1465 + }, + { + "epoch": 0.4894824707846411, + "grad_norm": 0.5301948778160731, + "learning_rate": 9.375286294413914e-06, + "loss": 0.307, + "step": 1466 + }, + { + "epoch": 0.48981636060100164, + "grad_norm": 0.519513204745318, + "learning_rate": 9.373875245666953e-06, + "loss": 0.2958, + "step": 1467 + }, + { + "epoch": 0.49015025041736227, + "grad_norm": 0.5082189516798558, + "learning_rate": 9.37246271157457e-06, + "loss": 0.3089, + "step": 1468 + }, + { + "epoch": 0.4904841402337229, + "grad_norm": 0.5332747047898574, + "learning_rate": 9.371048692616461e-06, + "loss": 0.3, + "step": 1469 + }, + { + "epoch": 0.49081803005008345, + "grad_norm": 0.5259089006554704, + "learning_rate": 9.369633189272815e-06, + "loss": 0.2958, + "step": 1470 + }, + { + "epoch": 0.4911519198664441, + "grad_norm": 0.5125665753364036, + "learning_rate": 9.368216202024329e-06, + "loss": 0.2987, + "step": 1471 + }, + { + "epoch": 0.4914858096828047, + "grad_norm": 0.5303637085327176, + "learning_rate": 9.36679773135221e-06, + "loss": 0.3032, + "step": 1472 + }, + { + "epoch": 0.49181969949916526, + "grad_norm": 0.5390834917745546, + "learning_rate": 9.365377777738155e-06, + "loss": 0.3114, + "step": 1473 + }, + { + "epoch": 0.4921535893155259, + "grad_norm": 0.5267175805067364, + "learning_rate": 9.36395634166438e-06, + "loss": 0.3023, + "step": 1474 + }, + { + "epoch": 0.49248747913188645, + "grad_norm": 0.5546149519974832, + "learning_rate": 9.362533423613592e-06, + "loss": 0.3056, + "step": 1475 + }, + { + "epoch": 0.4928213689482471, + "grad_norm": 0.49972859691659005, + "learning_rate": 9.361109024069011e-06, + "loss": 0.3011, + "step": 1476 + }, + { + "epoch": 0.4931552587646077, + "grad_norm": 0.5512056103145353, + "learning_rate": 9.359683143514354e-06, + "loss": 0.2981, + "step": 1477 + }, + { + "epoch": 0.49348914858096826, + "grad_norm": 0.5742776253634453, + "learning_rate": 9.358255782433842e-06, + "loss": 0.2936, + "step": 1478 + }, + { + "epoch": 0.4938230383973289, + "grad_norm": 0.518097152416003, + "learning_rate": 9.3568269413122e-06, + "loss": 0.2891, + "step": 1479 + }, + { + "epoch": 0.4941569282136895, + "grad_norm": 0.5355107846117838, + "learning_rate": 9.355396620634657e-06, + "loss": 0.2787, + "step": 1480 + }, + { + "epoch": 0.49449081803005007, + "grad_norm": 0.5638301777566492, + "learning_rate": 9.353964820886938e-06, + "loss": 0.2902, + "step": 1481 + }, + { + "epoch": 0.4948247078464107, + "grad_norm": 0.5122064250410363, + "learning_rate": 9.35253154255528e-06, + "loss": 0.2921, + "step": 1482 + }, + { + "epoch": 0.49515859766277126, + "grad_norm": 0.5661462651625967, + "learning_rate": 9.351096786126413e-06, + "loss": 0.2888, + "step": 1483 + }, + { + "epoch": 0.4954924874791319, + "grad_norm": 0.5145053572616372, + "learning_rate": 9.349660552087574e-06, + "loss": 0.2872, + "step": 1484 + }, + { + "epoch": 0.4958263772954925, + "grad_norm": 0.5491811165175631, + "learning_rate": 9.3482228409265e-06, + "loss": 0.2995, + "step": 1485 + }, + { + "epoch": 0.49616026711185307, + "grad_norm": 0.5700928741642592, + "learning_rate": 9.346783653131432e-06, + "loss": 0.3068, + "step": 1486 + }, + { + "epoch": 0.4964941569282137, + "grad_norm": 0.5468816267360477, + "learning_rate": 9.34534298919111e-06, + "loss": 0.2911, + "step": 1487 + }, + { + "epoch": 0.4968280467445743, + "grad_norm": 0.49701729947603057, + "learning_rate": 9.343900849594777e-06, + "loss": 0.2874, + "step": 1488 + }, + { + "epoch": 0.4971619365609349, + "grad_norm": 0.5311755218067368, + "learning_rate": 9.342457234832175e-06, + "loss": 0.3018, + "step": 1489 + }, + { + "epoch": 0.4974958263772955, + "grad_norm": 0.5344349099123405, + "learning_rate": 9.341012145393546e-06, + "loss": 0.2973, + "step": 1490 + }, + { + "epoch": 0.49782971619365607, + "grad_norm": 0.5157392661184167, + "learning_rate": 9.339565581769639e-06, + "loss": 0.297, + "step": 1491 + }, + { + "epoch": 0.4981636060100167, + "grad_norm": 0.49937066135290764, + "learning_rate": 9.338117544451697e-06, + "loss": 0.2899, + "step": 1492 + }, + { + "epoch": 0.4984974958263773, + "grad_norm": 0.5433330118133204, + "learning_rate": 9.336668033931466e-06, + "loss": 0.3065, + "step": 1493 + }, + { + "epoch": 0.4988313856427379, + "grad_norm": 0.5305482092069712, + "learning_rate": 9.335217050701193e-06, + "loss": 0.3151, + "step": 1494 + }, + { + "epoch": 0.4991652754590985, + "grad_norm": 0.5079256667100854, + "learning_rate": 9.333764595253625e-06, + "loss": 0.2894, + "step": 1495 + }, + { + "epoch": 0.4994991652754591, + "grad_norm": 0.5425225900113028, + "learning_rate": 9.332310668082007e-06, + "loss": 0.3066, + "step": 1496 + }, + { + "epoch": 0.4998330550918197, + "grad_norm": 0.5326458349831882, + "learning_rate": 9.330855269680085e-06, + "loss": 0.2996, + "step": 1497 + }, + { + "epoch": 0.5001669449081803, + "grad_norm": 0.5041061735722463, + "learning_rate": 9.329398400542106e-06, + "loss": 0.2997, + "step": 1498 + }, + { + "epoch": 0.5005008347245409, + "grad_norm": 0.5369230054814815, + "learning_rate": 9.327940061162816e-06, + "loss": 0.3069, + "step": 1499 + }, + { + "epoch": 0.5008347245409015, + "grad_norm": 0.5209531530825317, + "learning_rate": 9.326480252037457e-06, + "loss": 0.2793, + "step": 1500 + }, + { + "epoch": 0.5011686143572621, + "grad_norm": 0.526791899451195, + "learning_rate": 9.325018973661776e-06, + "loss": 0.2982, + "step": 1501 + }, + { + "epoch": 0.5015025041736227, + "grad_norm": 0.5171980005382196, + "learning_rate": 9.32355622653201e-06, + "loss": 0.2948, + "step": 1502 + }, + { + "epoch": 0.5018363939899833, + "grad_norm": 0.545129864246647, + "learning_rate": 9.322092011144907e-06, + "loss": 0.2964, + "step": 1503 + }, + { + "epoch": 0.5021702838063439, + "grad_norm": 0.5212971177026743, + "learning_rate": 9.320626327997703e-06, + "loss": 0.291, + "step": 1504 + }, + { + "epoch": 0.5025041736227045, + "grad_norm": 0.5185390358383788, + "learning_rate": 9.319159177588137e-06, + "loss": 0.3007, + "step": 1505 + }, + { + "epoch": 0.5028380634390651, + "grad_norm": 0.5034549131542907, + "learning_rate": 9.317690560414445e-06, + "loss": 0.2951, + "step": 1506 + }, + { + "epoch": 0.5031719532554257, + "grad_norm": 0.5194081495232304, + "learning_rate": 9.316220476975364e-06, + "loss": 0.2982, + "step": 1507 + }, + { + "epoch": 0.5035058430717864, + "grad_norm": 0.49633939932646864, + "learning_rate": 9.314748927770126e-06, + "loss": 0.2925, + "step": 1508 + }, + { + "epoch": 0.5038397328881469, + "grad_norm": 0.5135850009880143, + "learning_rate": 9.313275913298458e-06, + "loss": 0.2941, + "step": 1509 + }, + { + "epoch": 0.5041736227045075, + "grad_norm": 0.5562925344140361, + "learning_rate": 9.311801434060594e-06, + "loss": 0.289, + "step": 1510 + }, + { + "epoch": 0.5045075125208681, + "grad_norm": 0.516528722936568, + "learning_rate": 9.310325490557254e-06, + "loss": 0.3047, + "step": 1511 + }, + { + "epoch": 0.5048414023372287, + "grad_norm": 0.5137511158451734, + "learning_rate": 9.308848083289666e-06, + "loss": 0.3009, + "step": 1512 + }, + { + "epoch": 0.5051752921535894, + "grad_norm": 0.5315506812409, + "learning_rate": 9.307369212759547e-06, + "loss": 0.3093, + "step": 1513 + }, + { + "epoch": 0.5055091819699499, + "grad_norm": 0.5013477132955758, + "learning_rate": 9.305888879469112e-06, + "loss": 0.3007, + "step": 1514 + }, + { + "epoch": 0.5058430717863105, + "grad_norm": 0.5093206569902777, + "learning_rate": 9.30440708392108e-06, + "loss": 0.3012, + "step": 1515 + }, + { + "epoch": 0.5061769616026711, + "grad_norm": 0.49317573237002466, + "learning_rate": 9.302923826618654e-06, + "loss": 0.2984, + "step": 1516 + }, + { + "epoch": 0.5065108514190317, + "grad_norm": 0.5237801764233629, + "learning_rate": 9.301439108065546e-06, + "loss": 0.2876, + "step": 1517 + }, + { + "epoch": 0.5068447412353924, + "grad_norm": 0.5106877728691126, + "learning_rate": 9.299952928765956e-06, + "loss": 0.298, + "step": 1518 + }, + { + "epoch": 0.5071786310517529, + "grad_norm": 0.49900847078089466, + "learning_rate": 9.298465289224585e-06, + "loss": 0.2887, + "step": 1519 + }, + { + "epoch": 0.5075125208681135, + "grad_norm": 0.5017167855594853, + "learning_rate": 9.296976189946625e-06, + "loss": 0.2965, + "step": 1520 + }, + { + "epoch": 0.5078464106844741, + "grad_norm": 0.5472847088722247, + "learning_rate": 9.295485631437768e-06, + "loss": 0.2847, + "step": 1521 + }, + { + "epoch": 0.5081803005008347, + "grad_norm": 0.5439384601225326, + "learning_rate": 9.293993614204202e-06, + "loss": 0.3024, + "step": 1522 + }, + { + "epoch": 0.5085141903171954, + "grad_norm": 0.5311422438374932, + "learning_rate": 9.292500138752606e-06, + "loss": 0.2875, + "step": 1523 + }, + { + "epoch": 0.508848080133556, + "grad_norm": 0.5112745054790727, + "learning_rate": 9.291005205590153e-06, + "loss": 0.2899, + "step": 1524 + }, + { + "epoch": 0.5091819699499165, + "grad_norm": 0.5121052288261757, + "learning_rate": 9.289508815224522e-06, + "loss": 0.2811, + "step": 1525 + }, + { + "epoch": 0.5095158597662771, + "grad_norm": 0.5809762572637023, + "learning_rate": 9.288010968163874e-06, + "loss": 0.3036, + "step": 1526 + }, + { + "epoch": 0.5098497495826377, + "grad_norm": 0.5613999026292603, + "learning_rate": 9.28651166491687e-06, + "loss": 0.2964, + "step": 1527 + }, + { + "epoch": 0.5101836393989984, + "grad_norm": 0.5460499001045575, + "learning_rate": 9.28501090599267e-06, + "loss": 0.297, + "step": 1528 + }, + { + "epoch": 0.510517529215359, + "grad_norm": 0.4977453570962863, + "learning_rate": 9.283508691900922e-06, + "loss": 0.2913, + "step": 1529 + }, + { + "epoch": 0.5108514190317195, + "grad_norm": 0.5442823714534827, + "learning_rate": 9.282005023151767e-06, + "loss": 0.2904, + "step": 1530 + }, + { + "epoch": 0.5111853088480801, + "grad_norm": 0.5500284851717324, + "learning_rate": 9.280499900255848e-06, + "loss": 0.3137, + "step": 1531 + }, + { + "epoch": 0.5115191986644407, + "grad_norm": 0.5248739118741362, + "learning_rate": 9.278993323724293e-06, + "loss": 0.3046, + "step": 1532 + }, + { + "epoch": 0.5118530884808014, + "grad_norm": 0.5247658099690419, + "learning_rate": 9.27748529406873e-06, + "loss": 0.3111, + "step": 1533 + }, + { + "epoch": 0.512186978297162, + "grad_norm": 0.532969823087454, + "learning_rate": 9.275975811801278e-06, + "loss": 0.2833, + "step": 1534 + }, + { + "epoch": 0.5125208681135225, + "grad_norm": 0.5396327042933182, + "learning_rate": 9.274464877434548e-06, + "loss": 0.2992, + "step": 1535 + }, + { + "epoch": 0.5128547579298831, + "grad_norm": 0.5650583785211244, + "learning_rate": 9.272952491481646e-06, + "loss": 0.3067, + "step": 1536 + }, + { + "epoch": 0.5131886477462437, + "grad_norm": 0.5234343193987386, + "learning_rate": 9.271438654456171e-06, + "loss": 0.3025, + "step": 1537 + }, + { + "epoch": 0.5135225375626044, + "grad_norm": 0.5364797856942086, + "learning_rate": 9.269923366872213e-06, + "loss": 0.2876, + "step": 1538 + }, + { + "epoch": 0.513856427378965, + "grad_norm": 0.48243146988117996, + "learning_rate": 9.268406629244358e-06, + "loss": 0.281, + "step": 1539 + }, + { + "epoch": 0.5141903171953256, + "grad_norm": 0.5222687880324574, + "learning_rate": 9.266888442087681e-06, + "loss": 0.2934, + "step": 1540 + }, + { + "epoch": 0.5145242070116861, + "grad_norm": 0.5643642193285364, + "learning_rate": 9.26536880591775e-06, + "loss": 0.2976, + "step": 1541 + }, + { + "epoch": 0.5148580968280467, + "grad_norm": 0.4961450154285068, + "learning_rate": 9.263847721250624e-06, + "loss": 0.2725, + "step": 1542 + }, + { + "epoch": 0.5151919866444074, + "grad_norm": 0.5117454630037116, + "learning_rate": 9.262325188602861e-06, + "loss": 0.2903, + "step": 1543 + }, + { + "epoch": 0.515525876460768, + "grad_norm": 0.5097787397644629, + "learning_rate": 9.260801208491499e-06, + "loss": 0.2995, + "step": 1544 + }, + { + "epoch": 0.5158597662771286, + "grad_norm": 0.5527035929584001, + "learning_rate": 9.259275781434079e-06, + "loss": 0.2917, + "step": 1545 + }, + { + "epoch": 0.5161936560934891, + "grad_norm": 0.5202232221418576, + "learning_rate": 9.257748907948623e-06, + "loss": 0.3007, + "step": 1546 + }, + { + "epoch": 0.5165275459098497, + "grad_norm": 0.5178626419352864, + "learning_rate": 9.256220588553653e-06, + "loss": 0.2966, + "step": 1547 + }, + { + "epoch": 0.5168614357262103, + "grad_norm": 0.5267813100385837, + "learning_rate": 9.254690823768178e-06, + "loss": 0.3026, + "step": 1548 + }, + { + "epoch": 0.517195325542571, + "grad_norm": 0.5104941483092635, + "learning_rate": 9.253159614111698e-06, + "loss": 0.2836, + "step": 1549 + }, + { + "epoch": 0.5175292153589316, + "grad_norm": 0.4716115508874349, + "learning_rate": 9.251626960104201e-06, + "loss": 0.2783, + "step": 1550 + }, + { + "epoch": 0.5178631051752921, + "grad_norm": 0.4865670202993944, + "learning_rate": 9.250092862266174e-06, + "loss": 0.2931, + "step": 1551 + }, + { + "epoch": 0.5181969949916527, + "grad_norm": 0.49812069360060257, + "learning_rate": 9.248557321118583e-06, + "loss": 0.2857, + "step": 1552 + }, + { + "epoch": 0.5185308848080133, + "grad_norm": 0.5378012107224576, + "learning_rate": 9.247020337182892e-06, + "loss": 0.3036, + "step": 1553 + }, + { + "epoch": 0.518864774624374, + "grad_norm": 0.5032862411331327, + "learning_rate": 9.245481910981056e-06, + "loss": 0.2935, + "step": 1554 + }, + { + "epoch": 0.5191986644407346, + "grad_norm": 0.5170305936276663, + "learning_rate": 9.243942043035513e-06, + "loss": 0.296, + "step": 1555 + }, + { + "epoch": 0.5195325542570952, + "grad_norm": 0.5439467851821183, + "learning_rate": 9.242400733869193e-06, + "loss": 0.2911, + "step": 1556 + }, + { + "epoch": 0.5198664440734557, + "grad_norm": 0.5473079901462166, + "learning_rate": 9.240857984005521e-06, + "loss": 0.298, + "step": 1557 + }, + { + "epoch": 0.5202003338898163, + "grad_norm": 0.514340518505882, + "learning_rate": 9.239313793968402e-06, + "loss": 0.288, + "step": 1558 + }, + { + "epoch": 0.520534223706177, + "grad_norm": 0.5092230049979944, + "learning_rate": 9.237768164282239e-06, + "loss": 0.3028, + "step": 1559 + }, + { + "epoch": 0.5208681135225376, + "grad_norm": 0.5165354234871864, + "learning_rate": 9.236221095471918e-06, + "loss": 0.2998, + "step": 1560 + }, + { + "epoch": 0.5212020033388982, + "grad_norm": 0.5165557543158519, + "learning_rate": 9.234672588062813e-06, + "loss": 0.2966, + "step": 1561 + }, + { + "epoch": 0.5215358931552587, + "grad_norm": 0.5036155171615572, + "learning_rate": 9.233122642580796e-06, + "loss": 0.2986, + "step": 1562 + }, + { + "epoch": 0.5218697829716193, + "grad_norm": 0.5102935860464628, + "learning_rate": 9.231571259552214e-06, + "loss": 0.3069, + "step": 1563 + }, + { + "epoch": 0.52220367278798, + "grad_norm": 0.5294891410070776, + "learning_rate": 9.230018439503913e-06, + "loss": 0.3006, + "step": 1564 + }, + { + "epoch": 0.5225375626043406, + "grad_norm": 0.501589076069789, + "learning_rate": 9.22846418296322e-06, + "loss": 0.2982, + "step": 1565 + }, + { + "epoch": 0.5228714524207012, + "grad_norm": 0.494030824061187, + "learning_rate": 9.226908490457952e-06, + "loss": 0.2815, + "step": 1566 + }, + { + "epoch": 0.5232053422370617, + "grad_norm": 0.5008241794250871, + "learning_rate": 9.225351362516418e-06, + "loss": 0.3042, + "step": 1567 + }, + { + "epoch": 0.5235392320534223, + "grad_norm": 0.544285912239712, + "learning_rate": 9.223792799667403e-06, + "loss": 0.3019, + "step": 1568 + }, + { + "epoch": 0.523873121869783, + "grad_norm": 0.4857456785848327, + "learning_rate": 9.222232802440197e-06, + "loss": 0.2917, + "step": 1569 + }, + { + "epoch": 0.5242070116861436, + "grad_norm": 0.49454635089939314, + "learning_rate": 9.220671371364559e-06, + "loss": 0.2854, + "step": 1570 + }, + { + "epoch": 0.5245409015025042, + "grad_norm": 0.5187917765662672, + "learning_rate": 9.219108506970747e-06, + "loss": 0.2948, + "step": 1571 + }, + { + "epoch": 0.5248747913188647, + "grad_norm": 0.5028500116641059, + "learning_rate": 9.2175442097895e-06, + "loss": 0.2991, + "step": 1572 + }, + { + "epoch": 0.5252086811352253, + "grad_norm": 0.46942382084305856, + "learning_rate": 9.215978480352043e-06, + "loss": 0.287, + "step": 1573 + }, + { + "epoch": 0.525542570951586, + "grad_norm": 0.49169695378575135, + "learning_rate": 9.214411319190095e-06, + "loss": 0.2899, + "step": 1574 + }, + { + "epoch": 0.5258764607679466, + "grad_norm": 0.5070517886010653, + "learning_rate": 9.212842726835854e-06, + "loss": 0.2958, + "step": 1575 + }, + { + "epoch": 0.5262103505843072, + "grad_norm": 0.5351311257183571, + "learning_rate": 9.211272703822003e-06, + "loss": 0.3031, + "step": 1576 + }, + { + "epoch": 0.5265442404006678, + "grad_norm": 0.47916032526412244, + "learning_rate": 9.209701250681715e-06, + "loss": 0.2931, + "step": 1577 + }, + { + "epoch": 0.5268781302170283, + "grad_norm": 0.5057359368938524, + "learning_rate": 9.20812836794865e-06, + "loss": 0.2969, + "step": 1578 + }, + { + "epoch": 0.527212020033389, + "grad_norm": 0.571687032586194, + "learning_rate": 9.206554056156948e-06, + "loss": 0.3015, + "step": 1579 + }, + { + "epoch": 0.5275459098497496, + "grad_norm": 0.49161568285393414, + "learning_rate": 9.204978315841238e-06, + "loss": 0.2861, + "step": 1580 + }, + { + "epoch": 0.5278797996661102, + "grad_norm": 0.4832938493021482, + "learning_rate": 9.203401147536633e-06, + "loss": 0.2994, + "step": 1581 + }, + { + "epoch": 0.5282136894824708, + "grad_norm": 0.5174666834099655, + "learning_rate": 9.201822551778731e-06, + "loss": 0.2872, + "step": 1582 + }, + { + "epoch": 0.5285475792988313, + "grad_norm": 0.5389644203906737, + "learning_rate": 9.200242529103615e-06, + "loss": 0.3, + "step": 1583 + }, + { + "epoch": 0.528881469115192, + "grad_norm": 0.526390168726205, + "learning_rate": 9.198661080047855e-06, + "loss": 0.3017, + "step": 1584 + }, + { + "epoch": 0.5292153589315526, + "grad_norm": 0.5087001925834114, + "learning_rate": 9.1970782051485e-06, + "loss": 0.2964, + "step": 1585 + }, + { + "epoch": 0.5295492487479132, + "grad_norm": 0.5542909618831897, + "learning_rate": 9.195493904943087e-06, + "loss": 0.2989, + "step": 1586 + }, + { + "epoch": 0.5298831385642738, + "grad_norm": 0.5176078274072029, + "learning_rate": 9.193908179969635e-06, + "loss": 0.2938, + "step": 1587 + }, + { + "epoch": 0.5302170283806343, + "grad_norm": 0.5115953563409725, + "learning_rate": 9.192321030766652e-06, + "loss": 0.3033, + "step": 1588 + }, + { + "epoch": 0.530550918196995, + "grad_norm": 0.5097321848875267, + "learning_rate": 9.19073245787312e-06, + "loss": 0.2965, + "step": 1589 + }, + { + "epoch": 0.5308848080133556, + "grad_norm": 0.5397160375808504, + "learning_rate": 9.189142461828512e-06, + "loss": 0.3031, + "step": 1590 + }, + { + "epoch": 0.5312186978297162, + "grad_norm": 0.5219821222500399, + "learning_rate": 9.187551043172787e-06, + "loss": 0.3001, + "step": 1591 + }, + { + "epoch": 0.5315525876460768, + "grad_norm": 0.542304148722276, + "learning_rate": 9.185958202446375e-06, + "loss": 0.3083, + "step": 1592 + }, + { + "epoch": 0.5318864774624374, + "grad_norm": 0.5356634402029818, + "learning_rate": 9.184363940190204e-06, + "loss": 0.3018, + "step": 1593 + }, + { + "epoch": 0.532220367278798, + "grad_norm": 0.4906492051348479, + "learning_rate": 9.182768256945671e-06, + "loss": 0.2908, + "step": 1594 + }, + { + "epoch": 0.5325542570951586, + "grad_norm": 0.5148652072278146, + "learning_rate": 9.181171153254666e-06, + "loss": 0.2993, + "step": 1595 + }, + { + "epoch": 0.5328881469115192, + "grad_norm": 0.5159906092108616, + "learning_rate": 9.179572629659554e-06, + "loss": 0.2988, + "step": 1596 + }, + { + "epoch": 0.5332220367278798, + "grad_norm": 0.5231362028988326, + "learning_rate": 9.17797268670319e-06, + "loss": 0.2941, + "step": 1597 + }, + { + "epoch": 0.5335559265442404, + "grad_norm": 0.5019209871386232, + "learning_rate": 9.1763713249289e-06, + "loss": 0.2832, + "step": 1598 + }, + { + "epoch": 0.533889816360601, + "grad_norm": 0.4935216971875969, + "learning_rate": 9.174768544880503e-06, + "loss": 0.2983, + "step": 1599 + }, + { + "epoch": 0.5342237061769616, + "grad_norm": 0.5185072897270906, + "learning_rate": 9.173164347102293e-06, + "loss": 0.2935, + "step": 1600 + }, + { + "epoch": 0.5345575959933222, + "grad_norm": 0.5674961767887532, + "learning_rate": 9.171558732139046e-06, + "loss": 0.2863, + "step": 1601 + }, + { + "epoch": 0.5348914858096828, + "grad_norm": 0.5164417556617377, + "learning_rate": 9.169951700536024e-06, + "loss": 0.2905, + "step": 1602 + }, + { + "epoch": 0.5352253756260434, + "grad_norm": 0.5099748024529122, + "learning_rate": 9.168343252838963e-06, + "loss": 0.2966, + "step": 1603 + }, + { + "epoch": 0.535559265442404, + "grad_norm": 0.5228858345165949, + "learning_rate": 9.166733389594087e-06, + "loss": 0.2862, + "step": 1604 + }, + { + "epoch": 0.5358931552587646, + "grad_norm": 0.5453545717100521, + "learning_rate": 9.165122111348095e-06, + "loss": 0.3038, + "step": 1605 + }, + { + "epoch": 0.5362270450751252, + "grad_norm": 0.4907046064487329, + "learning_rate": 9.16350941864817e-06, + "loss": 0.2856, + "step": 1606 + }, + { + "epoch": 0.5365609348914858, + "grad_norm": 0.5053280712559297, + "learning_rate": 9.16189531204197e-06, + "loss": 0.2939, + "step": 1607 + }, + { + "epoch": 0.5368948247078464, + "grad_norm": 0.4942846389252826, + "learning_rate": 9.160279792077645e-06, + "loss": 0.2858, + "step": 1608 + }, + { + "epoch": 0.5372287145242071, + "grad_norm": 0.4874924450817432, + "learning_rate": 9.158662859303809e-06, + "loss": 0.2909, + "step": 1609 + }, + { + "epoch": 0.5375626043405676, + "grad_norm": 0.4821447118023549, + "learning_rate": 9.15704451426957e-06, + "loss": 0.302, + "step": 1610 + }, + { + "epoch": 0.5378964941569282, + "grad_norm": 0.482491009213483, + "learning_rate": 9.155424757524505e-06, + "loss": 0.301, + "step": 1611 + }, + { + "epoch": 0.5382303839732888, + "grad_norm": 0.5501751085277582, + "learning_rate": 9.153803589618681e-06, + "loss": 0.292, + "step": 1612 + }, + { + "epoch": 0.5385642737896494, + "grad_norm": 0.5441757416457015, + "learning_rate": 9.152181011102632e-06, + "loss": 0.3012, + "step": 1613 + }, + { + "epoch": 0.5388981636060101, + "grad_norm": 0.4847259559908357, + "learning_rate": 9.150557022527382e-06, + "loss": 0.3034, + "step": 1614 + }, + { + "epoch": 0.5392320534223706, + "grad_norm": 0.48740672188123935, + "learning_rate": 9.148931624444425e-06, + "loss": 0.3039, + "step": 1615 + }, + { + "epoch": 0.5395659432387312, + "grad_norm": 0.5241143239265353, + "learning_rate": 9.147304817405741e-06, + "loss": 0.2825, + "step": 1616 + }, + { + "epoch": 0.5398998330550918, + "grad_norm": 0.5312526880709922, + "learning_rate": 9.145676601963783e-06, + "loss": 0.3021, + "step": 1617 + }, + { + "epoch": 0.5402337228714524, + "grad_norm": 0.5083425699187554, + "learning_rate": 9.144046978671486e-06, + "loss": 0.3036, + "step": 1618 + }, + { + "epoch": 0.5405676126878131, + "grad_norm": 0.5065381679566716, + "learning_rate": 9.142415948082264e-06, + "loss": 0.2932, + "step": 1619 + }, + { + "epoch": 0.5409015025041736, + "grad_norm": 0.5376921571073677, + "learning_rate": 9.14078351075e-06, + "loss": 0.3014, + "step": 1620 + }, + { + "epoch": 0.5412353923205342, + "grad_norm": 0.5198146854113334, + "learning_rate": 9.139149667229067e-06, + "loss": 0.2921, + "step": 1621 + }, + { + "epoch": 0.5415692821368948, + "grad_norm": 0.49313693884507814, + "learning_rate": 9.137514418074307e-06, + "loss": 0.2842, + "step": 1622 + }, + { + "epoch": 0.5419031719532554, + "grad_norm": 0.5540669033960939, + "learning_rate": 9.135877763841046e-06, + "loss": 0.2965, + "step": 1623 + }, + { + "epoch": 0.5422370617696161, + "grad_norm": 0.5151727388331636, + "learning_rate": 9.134239705085077e-06, + "loss": 0.2913, + "step": 1624 + }, + { + "epoch": 0.5425709515859767, + "grad_norm": 0.4910393336074433, + "learning_rate": 9.132600242362682e-06, + "loss": 0.2874, + "step": 1625 + }, + { + "epoch": 0.5429048414023372, + "grad_norm": 0.5012844872925795, + "learning_rate": 9.13095937623061e-06, + "loss": 0.2911, + "step": 1626 + }, + { + "epoch": 0.5432387312186978, + "grad_norm": 0.556686286334535, + "learning_rate": 9.129317107246091e-06, + "loss": 0.3093, + "step": 1627 + }, + { + "epoch": 0.5435726210350584, + "grad_norm": 0.5380898245781904, + "learning_rate": 9.127673435966837e-06, + "loss": 0.2816, + "step": 1628 + }, + { + "epoch": 0.5439065108514191, + "grad_norm": 0.4890224324448751, + "learning_rate": 9.126028362951023e-06, + "loss": 0.2945, + "step": 1629 + }, + { + "epoch": 0.5442404006677797, + "grad_norm": 0.5350374035845251, + "learning_rate": 9.124381888757309e-06, + "loss": 0.3102, + "step": 1630 + }, + { + "epoch": 0.5445742904841402, + "grad_norm": 0.5393746667630365, + "learning_rate": 9.122734013944832e-06, + "loss": 0.2924, + "step": 1631 + }, + { + "epoch": 0.5449081803005008, + "grad_norm": 0.5101080762108411, + "learning_rate": 9.1210847390732e-06, + "loss": 0.2948, + "step": 1632 + }, + { + "epoch": 0.5452420701168614, + "grad_norm": 0.5176000177545594, + "learning_rate": 9.119434064702496e-06, + "loss": 0.2898, + "step": 1633 + }, + { + "epoch": 0.545575959933222, + "grad_norm": 0.5204812551416635, + "learning_rate": 9.117781991393285e-06, + "loss": 0.278, + "step": 1634 + }, + { + "epoch": 0.5459098497495827, + "grad_norm": 0.5047256683283994, + "learning_rate": 9.1161285197066e-06, + "loss": 0.2882, + "step": 1635 + }, + { + "epoch": 0.5462437395659432, + "grad_norm": 0.5527611071660795, + "learning_rate": 9.114473650203949e-06, + "loss": 0.303, + "step": 1636 + }, + { + "epoch": 0.5465776293823038, + "grad_norm": 0.5013641464678863, + "learning_rate": 9.112817383447323e-06, + "loss": 0.3011, + "step": 1637 + }, + { + "epoch": 0.5469115191986644, + "grad_norm": 0.5025119775445912, + "learning_rate": 9.111159719999176e-06, + "loss": 0.2882, + "step": 1638 + }, + { + "epoch": 0.547245409015025, + "grad_norm": 0.45923239293648566, + "learning_rate": 9.109500660422448e-06, + "loss": 0.2763, + "step": 1639 + }, + { + "epoch": 0.5475792988313857, + "grad_norm": 0.5449030988064698, + "learning_rate": 9.10784020528054e-06, + "loss": 0.3015, + "step": 1640 + }, + { + "epoch": 0.5479131886477463, + "grad_norm": 0.5229742328294368, + "learning_rate": 9.106178355137338e-06, + "loss": 0.2944, + "step": 1641 + }, + { + "epoch": 0.5482470784641068, + "grad_norm": 0.531095995512509, + "learning_rate": 9.104515110557197e-06, + "loss": 0.2925, + "step": 1642 + }, + { + "epoch": 0.5485809682804674, + "grad_norm": 0.5005615217076512, + "learning_rate": 9.102850472104945e-06, + "loss": 0.2933, + "step": 1643 + }, + { + "epoch": 0.548914858096828, + "grad_norm": 0.5212781375994004, + "learning_rate": 9.101184440345886e-06, + "loss": 0.2931, + "step": 1644 + }, + { + "epoch": 0.5492487479131887, + "grad_norm": 0.5228209654488285, + "learning_rate": 9.099517015845799e-06, + "loss": 0.295, + "step": 1645 + }, + { + "epoch": 0.5495826377295493, + "grad_norm": 0.5151969813081891, + "learning_rate": 9.097848199170928e-06, + "loss": 0.2985, + "step": 1646 + }, + { + "epoch": 0.5499165275459098, + "grad_norm": 0.4973286632746772, + "learning_rate": 9.096177990887995e-06, + "loss": 0.3169, + "step": 1647 + }, + { + "epoch": 0.5502504173622704, + "grad_norm": 0.5305997109612315, + "learning_rate": 9.094506391564196e-06, + "loss": 0.3076, + "step": 1648 + }, + { + "epoch": 0.550584307178631, + "grad_norm": 0.5063307276209618, + "learning_rate": 9.092833401767194e-06, + "loss": 0.2957, + "step": 1649 + }, + { + "epoch": 0.5509181969949917, + "grad_norm": 0.5269992415763571, + "learning_rate": 9.091159022065134e-06, + "loss": 0.2962, + "step": 1650 + }, + { + "epoch": 0.5512520868113523, + "grad_norm": 0.5054813454591253, + "learning_rate": 9.08948325302662e-06, + "loss": 0.2803, + "step": 1651 + }, + { + "epoch": 0.5515859766277128, + "grad_norm": 0.509538129904705, + "learning_rate": 9.08780609522074e-06, + "loss": 0.2951, + "step": 1652 + }, + { + "epoch": 0.5519198664440734, + "grad_norm": 0.5273871521175263, + "learning_rate": 9.086127549217044e-06, + "loss": 0.2723, + "step": 1653 + }, + { + "epoch": 0.552253756260434, + "grad_norm": 0.5157790525709752, + "learning_rate": 9.08444761558556e-06, + "loss": 0.2918, + "step": 1654 + }, + { + "epoch": 0.5525876460767947, + "grad_norm": 0.5263722593387334, + "learning_rate": 9.082766294896781e-06, + "loss": 0.2979, + "step": 1655 + }, + { + "epoch": 0.5529215358931553, + "grad_norm": 0.5011566969695254, + "learning_rate": 9.08108358772168e-06, + "loss": 0.298, + "step": 1656 + }, + { + "epoch": 0.5532554257095159, + "grad_norm": 0.49360381027770306, + "learning_rate": 9.079399494631693e-06, + "loss": 0.2934, + "step": 1657 + }, + { + "epoch": 0.5535893155258764, + "grad_norm": 0.46846659868195345, + "learning_rate": 9.077714016198729e-06, + "loss": 0.293, + "step": 1658 + }, + { + "epoch": 0.553923205342237, + "grad_norm": 0.5030986032217886, + "learning_rate": 9.07602715299517e-06, + "loss": 0.3159, + "step": 1659 + }, + { + "epoch": 0.5542570951585977, + "grad_norm": 0.49576432887805166, + "learning_rate": 9.074338905593862e-06, + "loss": 0.3088, + "step": 1660 + }, + { + "epoch": 0.5545909849749583, + "grad_norm": 0.5023074770984823, + "learning_rate": 9.07264927456813e-06, + "loss": 0.2909, + "step": 1661 + }, + { + "epoch": 0.5549248747913189, + "grad_norm": 0.4723988539695311, + "learning_rate": 9.070958260491761e-06, + "loss": 0.2939, + "step": 1662 + }, + { + "epoch": 0.5552587646076794, + "grad_norm": 0.47793781383107337, + "learning_rate": 9.069265863939012e-06, + "loss": 0.2829, + "step": 1663 + }, + { + "epoch": 0.55559265442404, + "grad_norm": 0.49000188806551054, + "learning_rate": 9.067572085484618e-06, + "loss": 0.2911, + "step": 1664 + }, + { + "epoch": 0.5559265442404007, + "grad_norm": 0.4634598439436136, + "learning_rate": 9.065876925703775e-06, + "loss": 0.2854, + "step": 1665 + }, + { + "epoch": 0.5562604340567613, + "grad_norm": 0.4906370579835716, + "learning_rate": 9.064180385172148e-06, + "loss": 0.2895, + "step": 1666 + }, + { + "epoch": 0.5565943238731219, + "grad_norm": 0.46974825656308483, + "learning_rate": 9.062482464465877e-06, + "loss": 0.2844, + "step": 1667 + }, + { + "epoch": 0.5569282136894824, + "grad_norm": 0.4987815734598347, + "learning_rate": 9.060783164161567e-06, + "loss": 0.2977, + "step": 1668 + }, + { + "epoch": 0.557262103505843, + "grad_norm": 0.47711050128275756, + "learning_rate": 9.059082484836287e-06, + "loss": 0.2891, + "step": 1669 + }, + { + "epoch": 0.5575959933222037, + "grad_norm": 0.49223013674097926, + "learning_rate": 9.057380427067584e-06, + "loss": 0.2975, + "step": 1670 + }, + { + "epoch": 0.5579298831385643, + "grad_norm": 0.48953467478861123, + "learning_rate": 9.055676991433467e-06, + "loss": 0.2955, + "step": 1671 + }, + { + "epoch": 0.5582637729549249, + "grad_norm": 0.5194458725729548, + "learning_rate": 9.05397217851241e-06, + "loss": 0.2975, + "step": 1672 + }, + { + "epoch": 0.5585976627712855, + "grad_norm": 0.5028865338683661, + "learning_rate": 9.052265988883365e-06, + "loss": 0.2981, + "step": 1673 + }, + { + "epoch": 0.558931552587646, + "grad_norm": 0.4740409925480656, + "learning_rate": 9.050558423125742e-06, + "loss": 0.2938, + "step": 1674 + }, + { + "epoch": 0.5592654424040067, + "grad_norm": 0.4882817952473548, + "learning_rate": 9.048849481819418e-06, + "loss": 0.2982, + "step": 1675 + }, + { + "epoch": 0.5595993322203673, + "grad_norm": 0.4920651574001583, + "learning_rate": 9.047139165544748e-06, + "loss": 0.2946, + "step": 1676 + }, + { + "epoch": 0.5599332220367279, + "grad_norm": 0.4977932486985738, + "learning_rate": 9.04542747488254e-06, + "loss": 0.3169, + "step": 1677 + }, + { + "epoch": 0.5602671118530885, + "grad_norm": 0.49149574187272943, + "learning_rate": 9.043714410414078e-06, + "loss": 0.2892, + "step": 1678 + }, + { + "epoch": 0.560601001669449, + "grad_norm": 0.4816029229228286, + "learning_rate": 9.04199997272111e-06, + "loss": 0.2858, + "step": 1679 + }, + { + "epoch": 0.5609348914858097, + "grad_norm": 0.49086675351266057, + "learning_rate": 9.04028416238585e-06, + "loss": 0.3009, + "step": 1680 + }, + { + "epoch": 0.5612687813021703, + "grad_norm": 0.5144960138583805, + "learning_rate": 9.038566979990978e-06, + "loss": 0.2878, + "step": 1681 + }, + { + "epoch": 0.5616026711185309, + "grad_norm": 0.5207718556185451, + "learning_rate": 9.03684842611964e-06, + "loss": 0.2883, + "step": 1682 + }, + { + "epoch": 0.5619365609348915, + "grad_norm": 0.5043711274460629, + "learning_rate": 9.03512850135545e-06, + "loss": 0.2949, + "step": 1683 + }, + { + "epoch": 0.562270450751252, + "grad_norm": 0.4873859355828317, + "learning_rate": 9.033407206282484e-06, + "loss": 0.3025, + "step": 1684 + }, + { + "epoch": 0.5626043405676127, + "grad_norm": 0.47773097643083456, + "learning_rate": 9.031684541485285e-06, + "loss": 0.2972, + "step": 1685 + }, + { + "epoch": 0.5629382303839733, + "grad_norm": 0.5424490689616456, + "learning_rate": 9.029960507548858e-06, + "loss": 0.2896, + "step": 1686 + }, + { + "epoch": 0.5632721202003339, + "grad_norm": 0.4811625548463843, + "learning_rate": 9.02823510505868e-06, + "loss": 0.2913, + "step": 1687 + }, + { + "epoch": 0.5636060100166945, + "grad_norm": 0.5184501223084643, + "learning_rate": 9.02650833460069e-06, + "loss": 0.2846, + "step": 1688 + }, + { + "epoch": 0.5639398998330551, + "grad_norm": 0.505562340609353, + "learning_rate": 9.024780196761287e-06, + "loss": 0.306, + "step": 1689 + }, + { + "epoch": 0.5642737896494157, + "grad_norm": 0.5060803539547568, + "learning_rate": 9.02305069212734e-06, + "loss": 0.2945, + "step": 1690 + }, + { + "epoch": 0.5646076794657763, + "grad_norm": 0.49588386821526875, + "learning_rate": 9.021319821286176e-06, + "loss": 0.2991, + "step": 1691 + }, + { + "epoch": 0.5649415692821369, + "grad_norm": 0.5261226662257844, + "learning_rate": 9.019587584825596e-06, + "loss": 0.3057, + "step": 1692 + }, + { + "epoch": 0.5652754590984975, + "grad_norm": 0.5162741477965035, + "learning_rate": 9.01785398333385e-06, + "loss": 0.2903, + "step": 1693 + }, + { + "epoch": 0.5656093489148581, + "grad_norm": 0.5076912126001749, + "learning_rate": 9.01611901739967e-06, + "loss": 0.2955, + "step": 1694 + }, + { + "epoch": 0.5659432387312187, + "grad_norm": 0.4843048815041193, + "learning_rate": 9.014382687612235e-06, + "loss": 0.2981, + "step": 1695 + }, + { + "epoch": 0.5662771285475793, + "grad_norm": 0.5003123805752508, + "learning_rate": 9.012644994561193e-06, + "loss": 0.3007, + "step": 1696 + }, + { + "epoch": 0.5666110183639399, + "grad_norm": 0.4998685948736363, + "learning_rate": 9.01090593883666e-06, + "loss": 0.2897, + "step": 1697 + }, + { + "epoch": 0.5669449081803005, + "grad_norm": 0.5176304039705191, + "learning_rate": 9.009165521029206e-06, + "loss": 0.2982, + "step": 1698 + }, + { + "epoch": 0.5672787979966611, + "grad_norm": 0.507709286476404, + "learning_rate": 9.007423741729873e-06, + "loss": 0.2879, + "step": 1699 + }, + { + "epoch": 0.5676126878130217, + "grad_norm": 0.5358247208976142, + "learning_rate": 9.005680601530154e-06, + "loss": 0.3003, + "step": 1700 + }, + { + "epoch": 0.5679465776293823, + "grad_norm": 0.5151594309735712, + "learning_rate": 9.003936101022015e-06, + "loss": 0.2976, + "step": 1701 + }, + { + "epoch": 0.5682804674457429, + "grad_norm": 0.5233176408230965, + "learning_rate": 9.002190240797878e-06, + "loss": 0.3076, + "step": 1702 + }, + { + "epoch": 0.5686143572621035, + "grad_norm": 0.5097384610571203, + "learning_rate": 9.000443021450627e-06, + "loss": 0.2904, + "step": 1703 + }, + { + "epoch": 0.5689482470784641, + "grad_norm": 0.49345794341066673, + "learning_rate": 8.998694443573612e-06, + "loss": 0.2892, + "step": 1704 + }, + { + "epoch": 0.5692821368948247, + "grad_norm": 0.5105509728584734, + "learning_rate": 8.996944507760636e-06, + "loss": 0.2839, + "step": 1705 + }, + { + "epoch": 0.5696160267111853, + "grad_norm": 0.5089067914256306, + "learning_rate": 8.995193214605972e-06, + "loss": 0.3006, + "step": 1706 + }, + { + "epoch": 0.5699499165275459, + "grad_norm": 0.5270342251541181, + "learning_rate": 8.99344056470435e-06, + "loss": 0.2974, + "step": 1707 + }, + { + "epoch": 0.5702838063439065, + "grad_norm": 0.5040413992011619, + "learning_rate": 8.991686558650959e-06, + "loss": 0.2844, + "step": 1708 + }, + { + "epoch": 0.5706176961602671, + "grad_norm": 0.5224548766546886, + "learning_rate": 8.989931197041451e-06, + "loss": 0.3026, + "step": 1709 + }, + { + "epoch": 0.5709515859766278, + "grad_norm": 0.5126850013877743, + "learning_rate": 8.988174480471937e-06, + "loss": 0.2965, + "step": 1710 + }, + { + "epoch": 0.5712854757929883, + "grad_norm": 0.5143228397743289, + "learning_rate": 8.98641640953899e-06, + "loss": 0.2906, + "step": 1711 + }, + { + "epoch": 0.5716193656093489, + "grad_norm": 0.508933211315629, + "learning_rate": 8.984656984839643e-06, + "loss": 0.2888, + "step": 1712 + }, + { + "epoch": 0.5719532554257095, + "grad_norm": 0.5422847979117157, + "learning_rate": 8.982896206971385e-06, + "loss": 0.3023, + "step": 1713 + }, + { + "epoch": 0.5722871452420701, + "grad_norm": 0.5266269027515925, + "learning_rate": 8.981134076532169e-06, + "loss": 0.3022, + "step": 1714 + }, + { + "epoch": 0.5726210350584308, + "grad_norm": 0.5218972378735852, + "learning_rate": 8.979370594120404e-06, + "loss": 0.2946, + "step": 1715 + }, + { + "epoch": 0.5729549248747913, + "grad_norm": 0.5182224907725962, + "learning_rate": 8.97760576033496e-06, + "loss": 0.2965, + "step": 1716 + }, + { + "epoch": 0.5732888146911519, + "grad_norm": 0.4956215428913995, + "learning_rate": 8.975839575775165e-06, + "loss": 0.2986, + "step": 1717 + }, + { + "epoch": 0.5736227045075125, + "grad_norm": 0.5025725112755058, + "learning_rate": 8.974072041040807e-06, + "loss": 0.2947, + "step": 1718 + }, + { + "epoch": 0.5739565943238731, + "grad_norm": 0.48260922894778546, + "learning_rate": 8.972303156732131e-06, + "loss": 0.2822, + "step": 1719 + }, + { + "epoch": 0.5742904841402338, + "grad_norm": 0.5216483253312642, + "learning_rate": 8.970532923449841e-06, + "loss": 0.2998, + "step": 1720 + }, + { + "epoch": 0.5746243739565943, + "grad_norm": 0.5089310751493624, + "learning_rate": 8.968761341795102e-06, + "loss": 0.299, + "step": 1721 + }, + { + "epoch": 0.5749582637729549, + "grad_norm": 0.5169843669078298, + "learning_rate": 8.966988412369528e-06, + "loss": 0.3093, + "step": 1722 + }, + { + "epoch": 0.5752921535893155, + "grad_norm": 0.534328609720625, + "learning_rate": 8.965214135775202e-06, + "loss": 0.2979, + "step": 1723 + }, + { + "epoch": 0.5756260434056761, + "grad_norm": 0.4688647812085704, + "learning_rate": 8.963438512614657e-06, + "loss": 0.2751, + "step": 1724 + }, + { + "epoch": 0.5759599332220368, + "grad_norm": 0.4965536554458997, + "learning_rate": 8.961661543490884e-06, + "loss": 0.2972, + "step": 1725 + }, + { + "epoch": 0.5762938230383974, + "grad_norm": 0.5258571663202141, + "learning_rate": 8.959883229007335e-06, + "loss": 0.2967, + "step": 1726 + }, + { + "epoch": 0.5766277128547579, + "grad_norm": 0.5294663816506548, + "learning_rate": 8.958103569767915e-06, + "loss": 0.2953, + "step": 1727 + }, + { + "epoch": 0.5769616026711185, + "grad_norm": 0.504240559707574, + "learning_rate": 8.956322566376989e-06, + "loss": 0.3, + "step": 1728 + }, + { + "epoch": 0.5772954924874791, + "grad_norm": 0.4896440428218204, + "learning_rate": 8.954540219439374e-06, + "loss": 0.2744, + "step": 1729 + }, + { + "epoch": 0.5776293823038398, + "grad_norm": 0.520779493157124, + "learning_rate": 8.952756529560346e-06, + "loss": 0.2943, + "step": 1730 + }, + { + "epoch": 0.5779632721202004, + "grad_norm": 0.5175437329317898, + "learning_rate": 8.95097149734564e-06, + "loss": 0.2951, + "step": 1731 + }, + { + "epoch": 0.5782971619365609, + "grad_norm": 0.47799087870926305, + "learning_rate": 8.94918512340144e-06, + "loss": 0.2804, + "step": 1732 + }, + { + "epoch": 0.5786310517529215, + "grad_norm": 0.49733896336025785, + "learning_rate": 8.94739740833439e-06, + "loss": 0.3054, + "step": 1733 + }, + { + "epoch": 0.5789649415692821, + "grad_norm": 0.5186345922467819, + "learning_rate": 8.945608352751592e-06, + "loss": 0.2992, + "step": 1734 + }, + { + "epoch": 0.5792988313856428, + "grad_norm": 0.5259303868512638, + "learning_rate": 8.943817957260596e-06, + "loss": 0.3106, + "step": 1735 + }, + { + "epoch": 0.5796327212020034, + "grad_norm": 0.5529497421632273, + "learning_rate": 8.942026222469414e-06, + "loss": 0.3034, + "step": 1736 + }, + { + "epoch": 0.5799666110183639, + "grad_norm": 0.47685596368544697, + "learning_rate": 8.940233148986507e-06, + "loss": 0.2866, + "step": 1737 + }, + { + "epoch": 0.5803005008347245, + "grad_norm": 0.4981677105852916, + "learning_rate": 8.938438737420797e-06, + "loss": 0.2843, + "step": 1738 + }, + { + "epoch": 0.5806343906510851, + "grad_norm": 0.5371502491013, + "learning_rate": 8.936642988381652e-06, + "loss": 0.2983, + "step": 1739 + }, + { + "epoch": 0.5809682804674458, + "grad_norm": 0.51051452345514, + "learning_rate": 8.934845902478903e-06, + "loss": 0.2973, + "step": 1740 + }, + { + "epoch": 0.5813021702838064, + "grad_norm": 0.5362943865389812, + "learning_rate": 8.933047480322832e-06, + "loss": 0.2933, + "step": 1741 + }, + { + "epoch": 0.581636060100167, + "grad_norm": 0.530441832975243, + "learning_rate": 8.93124772252417e-06, + "loss": 0.2801, + "step": 1742 + }, + { + "epoch": 0.5819699499165275, + "grad_norm": 0.4984929175047108, + "learning_rate": 8.929446629694109e-06, + "loss": 0.2853, + "step": 1743 + }, + { + "epoch": 0.5823038397328881, + "grad_norm": 0.5058770254027217, + "learning_rate": 8.927644202444287e-06, + "loss": 0.2878, + "step": 1744 + }, + { + "epoch": 0.5826377295492488, + "grad_norm": 0.49810898838663964, + "learning_rate": 8.925840441386803e-06, + "loss": 0.2919, + "step": 1745 + }, + { + "epoch": 0.5829716193656094, + "grad_norm": 0.5175747200445235, + "learning_rate": 8.924035347134204e-06, + "loss": 0.2826, + "step": 1746 + }, + { + "epoch": 0.58330550918197, + "grad_norm": 0.5044299186015049, + "learning_rate": 8.922228920299489e-06, + "loss": 0.288, + "step": 1747 + }, + { + "epoch": 0.5836393989983305, + "grad_norm": 0.4727399647813488, + "learning_rate": 8.92042116149611e-06, + "loss": 0.287, + "step": 1748 + }, + { + "epoch": 0.5839732888146911, + "grad_norm": 0.4734178522224389, + "learning_rate": 8.918612071337978e-06, + "loss": 0.2814, + "step": 1749 + }, + { + "epoch": 0.5843071786310517, + "grad_norm": 0.49690049939616493, + "learning_rate": 8.916801650439446e-06, + "loss": 0.2872, + "step": 1750 + }, + { + "epoch": 0.5846410684474124, + "grad_norm": 0.4996088271677074, + "learning_rate": 8.914989899415323e-06, + "loss": 0.2803, + "step": 1751 + }, + { + "epoch": 0.584974958263773, + "grad_norm": 0.4936286455441914, + "learning_rate": 8.913176818880874e-06, + "loss": 0.2924, + "step": 1752 + }, + { + "epoch": 0.5853088480801335, + "grad_norm": 0.4718753967671852, + "learning_rate": 8.911362409451808e-06, + "loss": 0.2816, + "step": 1753 + }, + { + "epoch": 0.5856427378964941, + "grad_norm": 0.5102692636536437, + "learning_rate": 8.909546671744294e-06, + "loss": 0.2921, + "step": 1754 + }, + { + "epoch": 0.5859766277128547, + "grad_norm": 0.5107111874737701, + "learning_rate": 8.907729606374941e-06, + "loss": 0.2829, + "step": 1755 + }, + { + "epoch": 0.5863105175292154, + "grad_norm": 0.46886818732626023, + "learning_rate": 8.905911213960821e-06, + "loss": 0.286, + "step": 1756 + }, + { + "epoch": 0.586644407345576, + "grad_norm": 0.4907840465811522, + "learning_rate": 8.904091495119447e-06, + "loss": 0.2875, + "step": 1757 + }, + { + "epoch": 0.5869782971619366, + "grad_norm": 0.512428598145084, + "learning_rate": 8.902270450468785e-06, + "loss": 0.2838, + "step": 1758 + }, + { + "epoch": 0.5873121869782971, + "grad_norm": 0.48529269480798926, + "learning_rate": 8.900448080627255e-06, + "loss": 0.2878, + "step": 1759 + }, + { + "epoch": 0.5876460767946577, + "grad_norm": 0.503996411156769, + "learning_rate": 8.898624386213724e-06, + "loss": 0.2941, + "step": 1760 + }, + { + "epoch": 0.5879799666110184, + "grad_norm": 0.5282919690908265, + "learning_rate": 8.896799367847508e-06, + "loss": 0.2841, + "step": 1761 + }, + { + "epoch": 0.588313856427379, + "grad_norm": 0.49021668255746564, + "learning_rate": 8.894973026148376e-06, + "loss": 0.2934, + "step": 1762 + }, + { + "epoch": 0.5886477462437396, + "grad_norm": 0.4771552238257979, + "learning_rate": 8.893145361736541e-06, + "loss": 0.291, + "step": 1763 + }, + { + "epoch": 0.5889816360601001, + "grad_norm": 0.4786630916249514, + "learning_rate": 8.891316375232673e-06, + "loss": 0.2807, + "step": 1764 + }, + { + "epoch": 0.5893155258764607, + "grad_norm": 0.5064014010917812, + "learning_rate": 8.88948606725788e-06, + "loss": 0.2853, + "step": 1765 + }, + { + "epoch": 0.5896494156928214, + "grad_norm": 0.5020168668691624, + "learning_rate": 8.88765443843373e-06, + "loss": 0.3058, + "step": 1766 + }, + { + "epoch": 0.589983305509182, + "grad_norm": 0.5054743077427825, + "learning_rate": 8.885821489382235e-06, + "loss": 0.287, + "step": 1767 + }, + { + "epoch": 0.5903171953255426, + "grad_norm": 0.5156615244181213, + "learning_rate": 8.883987220725852e-06, + "loss": 0.3126, + "step": 1768 + }, + { + "epoch": 0.5906510851419031, + "grad_norm": 0.509857313807726, + "learning_rate": 8.88215163308749e-06, + "loss": 0.2908, + "step": 1769 + }, + { + "epoch": 0.5909849749582637, + "grad_norm": 0.5007544469640427, + "learning_rate": 8.880314727090506e-06, + "loss": 0.3011, + "step": 1770 + }, + { + "epoch": 0.5913188647746244, + "grad_norm": 0.47945867789879265, + "learning_rate": 8.878476503358703e-06, + "loss": 0.2912, + "step": 1771 + }, + { + "epoch": 0.591652754590985, + "grad_norm": 0.5032925162258762, + "learning_rate": 8.876636962516333e-06, + "loss": 0.298, + "step": 1772 + }, + { + "epoch": 0.5919866444073456, + "grad_norm": 0.4809189365280513, + "learning_rate": 8.874796105188095e-06, + "loss": 0.2979, + "step": 1773 + }, + { + "epoch": 0.5923205342237062, + "grad_norm": 0.5105152098680715, + "learning_rate": 8.872953931999132e-06, + "loss": 0.3011, + "step": 1774 + }, + { + "epoch": 0.5926544240400667, + "grad_norm": 0.49133940917012126, + "learning_rate": 8.87111044357504e-06, + "loss": 0.2922, + "step": 1775 + }, + { + "epoch": 0.5929883138564274, + "grad_norm": 0.5037164918438566, + "learning_rate": 8.869265640541854e-06, + "loss": 0.2985, + "step": 1776 + }, + { + "epoch": 0.593322203672788, + "grad_norm": 0.5025327350665917, + "learning_rate": 8.867419523526065e-06, + "loss": 0.2847, + "step": 1777 + }, + { + "epoch": 0.5936560934891486, + "grad_norm": 0.48472109785508133, + "learning_rate": 8.8655720931546e-06, + "loss": 0.2985, + "step": 1778 + }, + { + "epoch": 0.5939899833055092, + "grad_norm": 0.4996771041968084, + "learning_rate": 8.863723350054838e-06, + "loss": 0.3007, + "step": 1779 + }, + { + "epoch": 0.5943238731218697, + "grad_norm": 0.5148210071896765, + "learning_rate": 8.861873294854605e-06, + "loss": 0.2901, + "step": 1780 + }, + { + "epoch": 0.5946577629382304, + "grad_norm": 0.4983564318272678, + "learning_rate": 8.860021928182167e-06, + "loss": 0.2897, + "step": 1781 + }, + { + "epoch": 0.594991652754591, + "grad_norm": 0.4955991702201647, + "learning_rate": 8.85816925066624e-06, + "loss": 0.2819, + "step": 1782 + }, + { + "epoch": 0.5953255425709516, + "grad_norm": 0.5059047904319821, + "learning_rate": 8.856315262935986e-06, + "loss": 0.295, + "step": 1783 + }, + { + "epoch": 0.5956594323873122, + "grad_norm": 0.5150720902570767, + "learning_rate": 8.854459965621005e-06, + "loss": 0.2928, + "step": 1784 + }, + { + "epoch": 0.5959933222036727, + "grad_norm": 0.5066870378572839, + "learning_rate": 8.852603359351348e-06, + "loss": 0.282, + "step": 1785 + }, + { + "epoch": 0.5963272120200334, + "grad_norm": 0.4948316754652034, + "learning_rate": 8.850745444757512e-06, + "loss": 0.291, + "step": 1786 + }, + { + "epoch": 0.596661101836394, + "grad_norm": 0.49116015396620094, + "learning_rate": 8.84888622247043e-06, + "loss": 0.2845, + "step": 1787 + }, + { + "epoch": 0.5969949916527546, + "grad_norm": 0.4796974615089585, + "learning_rate": 8.84702569312149e-06, + "loss": 0.2876, + "step": 1788 + }, + { + "epoch": 0.5973288814691152, + "grad_norm": 0.5170137729239467, + "learning_rate": 8.84516385734251e-06, + "loss": 0.2894, + "step": 1789 + }, + { + "epoch": 0.5976627712854758, + "grad_norm": 0.512036309801671, + "learning_rate": 8.84330071576577e-06, + "loss": 0.2802, + "step": 1790 + }, + { + "epoch": 0.5979966611018364, + "grad_norm": 0.5067766042434214, + "learning_rate": 8.841436269023978e-06, + "loss": 0.272, + "step": 1791 + }, + { + "epoch": 0.598330550918197, + "grad_norm": 0.5030442824956934, + "learning_rate": 8.839570517750289e-06, + "loss": 0.2929, + "step": 1792 + }, + { + "epoch": 0.5986644407345576, + "grad_norm": 0.5062103606107402, + "learning_rate": 8.837703462578304e-06, + "loss": 0.2939, + "step": 1793 + }, + { + "epoch": 0.5989983305509182, + "grad_norm": 0.5073129111222326, + "learning_rate": 8.835835104142066e-06, + "loss": 0.3087, + "step": 1794 + }, + { + "epoch": 0.5993322203672788, + "grad_norm": 0.49137190497073163, + "learning_rate": 8.83396544307606e-06, + "loss": 0.2979, + "step": 1795 + }, + { + "epoch": 0.5996661101836394, + "grad_norm": 0.4920918959295422, + "learning_rate": 8.832094480015211e-06, + "loss": 0.2889, + "step": 1796 + }, + { + "epoch": 0.6, + "grad_norm": 0.5280718159383158, + "learning_rate": 8.83022221559489e-06, + "loss": 0.2836, + "step": 1797 + }, + { + "epoch": 0.6003338898163606, + "grad_norm": 0.4953416386521143, + "learning_rate": 8.82834865045091e-06, + "loss": 0.2949, + "step": 1798 + }, + { + "epoch": 0.6006677796327212, + "grad_norm": 0.4819573166736904, + "learning_rate": 8.82647378521952e-06, + "loss": 0.298, + "step": 1799 + }, + { + "epoch": 0.6010016694490818, + "grad_norm": 0.5043905513725476, + "learning_rate": 8.82459762053742e-06, + "loss": 0.289, + "step": 1800 + }, + { + "epoch": 0.6013355592654424, + "grad_norm": 0.5243090450284964, + "learning_rate": 8.82272015704174e-06, + "loss": 0.2876, + "step": 1801 + }, + { + "epoch": 0.601669449081803, + "grad_norm": 0.5146220695707918, + "learning_rate": 8.820841395370061e-06, + "loss": 0.291, + "step": 1802 + }, + { + "epoch": 0.6020033388981636, + "grad_norm": 0.4769252235780995, + "learning_rate": 8.818961336160399e-06, + "loss": 0.299, + "step": 1803 + }, + { + "epoch": 0.6023372287145242, + "grad_norm": 0.49331062695651207, + "learning_rate": 8.817079980051211e-06, + "loss": 0.2903, + "step": 1804 + }, + { + "epoch": 0.6026711185308848, + "grad_norm": 0.541338030778918, + "learning_rate": 8.815197327681399e-06, + "loss": 0.3093, + "step": 1805 + }, + { + "epoch": 0.6030050083472455, + "grad_norm": 0.4925116442045908, + "learning_rate": 8.8133133796903e-06, + "loss": 0.2922, + "step": 1806 + }, + { + "epoch": 0.603338898163606, + "grad_norm": 0.5111894223300436, + "learning_rate": 8.811428136717694e-06, + "loss": 0.2857, + "step": 1807 + }, + { + "epoch": 0.6036727879799666, + "grad_norm": 0.5224146694682967, + "learning_rate": 8.8095415994038e-06, + "loss": 0.2985, + "step": 1808 + }, + { + "epoch": 0.6040066777963272, + "grad_norm": 0.5389757703223177, + "learning_rate": 8.807653768389274e-06, + "loss": 0.2958, + "step": 1809 + }, + { + "epoch": 0.6043405676126878, + "grad_norm": 0.5368601432726852, + "learning_rate": 8.805764644315216e-06, + "loss": 0.2905, + "step": 1810 + }, + { + "epoch": 0.6046744574290485, + "grad_norm": 0.4967402853508359, + "learning_rate": 8.803874227823161e-06, + "loss": 0.2914, + "step": 1811 + }, + { + "epoch": 0.605008347245409, + "grad_norm": 0.5262087893774048, + "learning_rate": 8.801982519555087e-06, + "loss": 0.3015, + "step": 1812 + }, + { + "epoch": 0.6053422370617696, + "grad_norm": 0.5295349116637149, + "learning_rate": 8.800089520153406e-06, + "loss": 0.2903, + "step": 1813 + }, + { + "epoch": 0.6056761268781302, + "grad_norm": 0.5105936199873661, + "learning_rate": 8.798195230260973e-06, + "loss": 0.2938, + "step": 1814 + }, + { + "epoch": 0.6060100166944908, + "grad_norm": 0.49983362036745843, + "learning_rate": 8.796299650521076e-06, + "loss": 0.2975, + "step": 1815 + }, + { + "epoch": 0.6063439065108515, + "grad_norm": 0.5024080973571774, + "learning_rate": 8.794402781577448e-06, + "loss": 0.2953, + "step": 1816 + }, + { + "epoch": 0.606677796327212, + "grad_norm": 0.5307868272245722, + "learning_rate": 8.792504624074253e-06, + "loss": 0.3064, + "step": 1817 + }, + { + "epoch": 0.6070116861435726, + "grad_norm": 0.5080614299273786, + "learning_rate": 8.790605178656097e-06, + "loss": 0.289, + "step": 1818 + }, + { + "epoch": 0.6073455759599332, + "grad_norm": 0.5007914225207722, + "learning_rate": 8.788704445968019e-06, + "loss": 0.2879, + "step": 1819 + }, + { + "epoch": 0.6076794657762938, + "grad_norm": 0.5237186522997884, + "learning_rate": 8.786802426655502e-06, + "loss": 0.2875, + "step": 1820 + }, + { + "epoch": 0.6080133555926545, + "grad_norm": 0.507827580316375, + "learning_rate": 8.78489912136446e-06, + "loss": 0.2583, + "step": 1821 + }, + { + "epoch": 0.6083472454090151, + "grad_norm": 0.47619513943390646, + "learning_rate": 8.782994530741245e-06, + "loss": 0.2927, + "step": 1822 + }, + { + "epoch": 0.6086811352253756, + "grad_norm": 0.4923484535244514, + "learning_rate": 8.781088655432648e-06, + "loss": 0.2905, + "step": 1823 + }, + { + "epoch": 0.6090150250417362, + "grad_norm": 0.5197787276669431, + "learning_rate": 8.779181496085893e-06, + "loss": 0.3074, + "step": 1824 + }, + { + "epoch": 0.6093489148580968, + "grad_norm": 0.5173308384583759, + "learning_rate": 8.77727305334864e-06, + "loss": 0.2947, + "step": 1825 + }, + { + "epoch": 0.6096828046744575, + "grad_norm": 0.46827071849036467, + "learning_rate": 8.77536332786899e-06, + "loss": 0.29, + "step": 1826 + }, + { + "epoch": 0.6100166944908181, + "grad_norm": 0.5059238757498284, + "learning_rate": 8.773452320295472e-06, + "loss": 0.2881, + "step": 1827 + }, + { + "epoch": 0.6103505843071786, + "grad_norm": 0.5687376319883146, + "learning_rate": 8.771540031277058e-06, + "loss": 0.2837, + "step": 1828 + }, + { + "epoch": 0.6106844741235392, + "grad_norm": 0.4854492784486606, + "learning_rate": 8.769626461463147e-06, + "loss": 0.3024, + "step": 1829 + }, + { + "epoch": 0.6110183639398998, + "grad_norm": 0.48945805359010003, + "learning_rate": 8.767711611503584e-06, + "loss": 0.2955, + "step": 1830 + }, + { + "epoch": 0.6113522537562605, + "grad_norm": 0.49551321399113757, + "learning_rate": 8.765795482048635e-06, + "loss": 0.297, + "step": 1831 + }, + { + "epoch": 0.6116861435726211, + "grad_norm": 0.49809978367131785, + "learning_rate": 8.763878073749013e-06, + "loss": 0.2803, + "step": 1832 + }, + { + "epoch": 0.6120200333889816, + "grad_norm": 0.49037193564292064, + "learning_rate": 8.761959387255857e-06, + "loss": 0.2917, + "step": 1833 + }, + { + "epoch": 0.6123539232053422, + "grad_norm": 0.5251808050427448, + "learning_rate": 8.760039423220741e-06, + "loss": 0.3037, + "step": 1834 + }, + { + "epoch": 0.6126878130217028, + "grad_norm": 0.4760259551275175, + "learning_rate": 8.758118182295682e-06, + "loss": 0.2835, + "step": 1835 + }, + { + "epoch": 0.6130217028380635, + "grad_norm": 0.4862256382967078, + "learning_rate": 8.756195665133115e-06, + "loss": 0.2885, + "step": 1836 + }, + { + "epoch": 0.6133555926544241, + "grad_norm": 0.5142966979139144, + "learning_rate": 8.754271872385925e-06, + "loss": 0.2941, + "step": 1837 + }, + { + "epoch": 0.6136894824707846, + "grad_norm": 0.5440081325519155, + "learning_rate": 8.752346804707417e-06, + "loss": 0.2965, + "step": 1838 + }, + { + "epoch": 0.6140233722871452, + "grad_norm": 0.4918340566749256, + "learning_rate": 8.750420462751334e-06, + "loss": 0.2944, + "step": 1839 + }, + { + "epoch": 0.6143572621035058, + "grad_norm": 0.4847306043801533, + "learning_rate": 8.748492847171853e-06, + "loss": 0.2946, + "step": 1840 + }, + { + "epoch": 0.6146911519198665, + "grad_norm": 0.5212153065103182, + "learning_rate": 8.746563958623584e-06, + "loss": 0.299, + "step": 1841 + }, + { + "epoch": 0.6150250417362271, + "grad_norm": 0.48204148744026537, + "learning_rate": 8.744633797761564e-06, + "loss": 0.2877, + "step": 1842 + }, + { + "epoch": 0.6153589315525877, + "grad_norm": 0.5059823821234395, + "learning_rate": 8.742702365241266e-06, + "loss": 0.2942, + "step": 1843 + }, + { + "epoch": 0.6156928213689482, + "grad_norm": 0.5061532142302817, + "learning_rate": 8.7407696617186e-06, + "loss": 0.2909, + "step": 1844 + }, + { + "epoch": 0.6160267111853088, + "grad_norm": 0.5117450051045148, + "learning_rate": 8.738835687849896e-06, + "loss": 0.297, + "step": 1845 + }, + { + "epoch": 0.6163606010016695, + "grad_norm": 0.48648739629129667, + "learning_rate": 8.736900444291923e-06, + "loss": 0.2933, + "step": 1846 + }, + { + "epoch": 0.6166944908180301, + "grad_norm": 0.5211955145824826, + "learning_rate": 8.73496393170188e-06, + "loss": 0.2851, + "step": 1847 + }, + { + "epoch": 0.6170283806343907, + "grad_norm": 0.4989761592730222, + "learning_rate": 8.733026150737398e-06, + "loss": 0.284, + "step": 1848 + }, + { + "epoch": 0.6173622704507512, + "grad_norm": 0.5146931299000661, + "learning_rate": 8.731087102056536e-06, + "loss": 0.2999, + "step": 1849 + }, + { + "epoch": 0.6176961602671118, + "grad_norm": 0.4958133646234376, + "learning_rate": 8.729146786317787e-06, + "loss": 0.2836, + "step": 1850 + }, + { + "epoch": 0.6180300500834724, + "grad_norm": 0.5529901665250311, + "learning_rate": 8.727205204180067e-06, + "loss": 0.299, + "step": 1851 + }, + { + "epoch": 0.6183639398998331, + "grad_norm": 0.4901894806571664, + "learning_rate": 8.725262356302732e-06, + "loss": 0.289, + "step": 1852 + }, + { + "epoch": 0.6186978297161937, + "grad_norm": 0.49729943113412955, + "learning_rate": 8.723318243345562e-06, + "loss": 0.2915, + "step": 1853 + }, + { + "epoch": 0.6190317195325542, + "grad_norm": 0.5094512142424891, + "learning_rate": 8.721372865968766e-06, + "loss": 0.2949, + "step": 1854 + }, + { + "epoch": 0.6193656093489148, + "grad_norm": 0.4832089452575637, + "learning_rate": 8.71942622483299e-06, + "loss": 0.285, + "step": 1855 + }, + { + "epoch": 0.6196994991652754, + "grad_norm": 0.5037412556826626, + "learning_rate": 8.717478320599295e-06, + "loss": 0.2952, + "step": 1856 + }, + { + "epoch": 0.6200333889816361, + "grad_norm": 0.5254575483547358, + "learning_rate": 8.715529153929182e-06, + "loss": 0.3003, + "step": 1857 + }, + { + "epoch": 0.6203672787979967, + "grad_norm": 0.4911595872253772, + "learning_rate": 8.713578725484582e-06, + "loss": 0.2881, + "step": 1858 + }, + { + "epoch": 0.6207011686143573, + "grad_norm": 0.49845692512549905, + "learning_rate": 8.711627035927848e-06, + "loss": 0.2926, + "step": 1859 + }, + { + "epoch": 0.6210350584307178, + "grad_norm": 0.49067438901264004, + "learning_rate": 8.70967408592176e-06, + "loss": 0.2909, + "step": 1860 + }, + { + "epoch": 0.6213689482470784, + "grad_norm": 0.4489135644474314, + "learning_rate": 8.707719876129535e-06, + "loss": 0.2931, + "step": 1861 + }, + { + "epoch": 0.6217028380634391, + "grad_norm": 0.46892907310013715, + "learning_rate": 8.70576440721481e-06, + "loss": 0.2921, + "step": 1862 + }, + { + "epoch": 0.6220367278797997, + "grad_norm": 0.5040939807894319, + "learning_rate": 8.703807679841651e-06, + "loss": 0.287, + "step": 1863 + }, + { + "epoch": 0.6223706176961603, + "grad_norm": 0.4781167537081272, + "learning_rate": 8.701849694674555e-06, + "loss": 0.2959, + "step": 1864 + }, + { + "epoch": 0.6227045075125208, + "grad_norm": 0.486504075469845, + "learning_rate": 8.699890452378442e-06, + "loss": 0.2885, + "step": 1865 + }, + { + "epoch": 0.6230383973288814, + "grad_norm": 0.49597529000074986, + "learning_rate": 8.697929953618663e-06, + "loss": 0.2982, + "step": 1866 + }, + { + "epoch": 0.6233722871452421, + "grad_norm": 0.4940840120105481, + "learning_rate": 8.695968199060991e-06, + "loss": 0.293, + "step": 1867 + }, + { + "epoch": 0.6237061769616027, + "grad_norm": 0.4910800057531951, + "learning_rate": 8.694005189371627e-06, + "loss": 0.2882, + "step": 1868 + }, + { + "epoch": 0.6240400667779633, + "grad_norm": 0.47479449966234893, + "learning_rate": 8.6920409252172e-06, + "loss": 0.2926, + "step": 1869 + }, + { + "epoch": 0.6243739565943238, + "grad_norm": 0.5277599298258983, + "learning_rate": 8.690075407264766e-06, + "loss": 0.2872, + "step": 1870 + }, + { + "epoch": 0.6247078464106844, + "grad_norm": 0.49363500658235865, + "learning_rate": 8.6881086361818e-06, + "loss": 0.296, + "step": 1871 + }, + { + "epoch": 0.6250417362270451, + "grad_norm": 0.49931324060193844, + "learning_rate": 8.68614061263621e-06, + "loss": 0.2942, + "step": 1872 + }, + { + "epoch": 0.6253756260434057, + "grad_norm": 0.4869660644271303, + "learning_rate": 8.68417133729633e-06, + "loss": 0.3007, + "step": 1873 + }, + { + "epoch": 0.6257095158597663, + "grad_norm": 0.5089077230344302, + "learning_rate": 8.68220081083091e-06, + "loss": 0.2902, + "step": 1874 + }, + { + "epoch": 0.6260434056761269, + "grad_norm": 0.46899143538589266, + "learning_rate": 8.680229033909134e-06, + "loss": 0.2728, + "step": 1875 + }, + { + "epoch": 0.6263772954924874, + "grad_norm": 0.49346871136954534, + "learning_rate": 8.678256007200604e-06, + "loss": 0.2868, + "step": 1876 + }, + { + "epoch": 0.6267111853088481, + "grad_norm": 0.4770891370528182, + "learning_rate": 8.676281731375355e-06, + "loss": 0.2896, + "step": 1877 + }, + { + "epoch": 0.6270450751252087, + "grad_norm": 0.5146127223005229, + "learning_rate": 8.674306207103836e-06, + "loss": 0.2872, + "step": 1878 + }, + { + "epoch": 0.6273789649415693, + "grad_norm": 0.5070283085348711, + "learning_rate": 8.672329435056926e-06, + "loss": 0.3016, + "step": 1879 + }, + { + "epoch": 0.6277128547579299, + "grad_norm": 0.4914719198954913, + "learning_rate": 8.670351415905928e-06, + "loss": 0.2946, + "step": 1880 + }, + { + "epoch": 0.6280467445742904, + "grad_norm": 0.4629629333931339, + "learning_rate": 8.668372150322566e-06, + "loss": 0.2948, + "step": 1881 + }, + { + "epoch": 0.6283806343906511, + "grad_norm": 0.5071697117559244, + "learning_rate": 8.66639163897899e-06, + "loss": 0.274, + "step": 1882 + }, + { + "epoch": 0.6287145242070117, + "grad_norm": 0.4883595194988578, + "learning_rate": 8.664409882547769e-06, + "loss": 0.2917, + "step": 1883 + }, + { + "epoch": 0.6290484140233723, + "grad_norm": 0.4775882962022977, + "learning_rate": 8.662426881701897e-06, + "loss": 0.2956, + "step": 1884 + }, + { + "epoch": 0.6293823038397329, + "grad_norm": 0.5016040203066215, + "learning_rate": 8.660442637114794e-06, + "loss": 0.2908, + "step": 1885 + }, + { + "epoch": 0.6297161936560934, + "grad_norm": 0.4775379722731503, + "learning_rate": 8.658457149460296e-06, + "loss": 0.2858, + "step": 1886 + }, + { + "epoch": 0.6300500834724541, + "grad_norm": 0.48739563311102246, + "learning_rate": 8.656470419412666e-06, + "loss": 0.2979, + "step": 1887 + }, + { + "epoch": 0.6303839732888147, + "grad_norm": 0.48473166507443227, + "learning_rate": 8.654482447646586e-06, + "loss": 0.2867, + "step": 1888 + }, + { + "epoch": 0.6307178631051753, + "grad_norm": 0.4788211943312986, + "learning_rate": 8.652493234837162e-06, + "loss": 0.293, + "step": 1889 + }, + { + "epoch": 0.6310517529215359, + "grad_norm": 0.46489477919738814, + "learning_rate": 8.650502781659922e-06, + "loss": 0.2826, + "step": 1890 + }, + { + "epoch": 0.6313856427378965, + "grad_norm": 0.49658481495867157, + "learning_rate": 8.64851108879081e-06, + "loss": 0.2973, + "step": 1891 + }, + { + "epoch": 0.6317195325542571, + "grad_norm": 0.4668130000219823, + "learning_rate": 8.646518156906198e-06, + "loss": 0.2878, + "step": 1892 + }, + { + "epoch": 0.6320534223706177, + "grad_norm": 0.5015940675113589, + "learning_rate": 8.64452398668287e-06, + "loss": 0.2799, + "step": 1893 + }, + { + "epoch": 0.6323873121869783, + "grad_norm": 0.503929226214965, + "learning_rate": 8.642528578798044e-06, + "loss": 0.2915, + "step": 1894 + }, + { + "epoch": 0.6327212020033389, + "grad_norm": 0.4719824043970263, + "learning_rate": 8.640531933929344e-06, + "loss": 0.297, + "step": 1895 + }, + { + "epoch": 0.6330550918196995, + "grad_norm": 0.5043210985753614, + "learning_rate": 8.638534052754823e-06, + "loss": 0.2918, + "step": 1896 + }, + { + "epoch": 0.6333889816360601, + "grad_norm": 0.4759283351508078, + "learning_rate": 8.636534935952949e-06, + "loss": 0.2809, + "step": 1897 + }, + { + "epoch": 0.6337228714524207, + "grad_norm": 0.4810300434357238, + "learning_rate": 8.634534584202615e-06, + "loss": 0.2821, + "step": 1898 + }, + { + "epoch": 0.6340567612687813, + "grad_norm": 0.495547222710604, + "learning_rate": 8.632532998183126e-06, + "loss": 0.302, + "step": 1899 + }, + { + "epoch": 0.6343906510851419, + "grad_norm": 0.4953540862339173, + "learning_rate": 8.630530178574212e-06, + "loss": 0.2972, + "step": 1900 + }, + { + "epoch": 0.6347245409015025, + "grad_norm": 0.49347790644770456, + "learning_rate": 8.628526126056023e-06, + "loss": 0.2897, + "step": 1901 + }, + { + "epoch": 0.635058430717863, + "grad_norm": 0.4935460532004309, + "learning_rate": 8.626520841309121e-06, + "loss": 0.3005, + "step": 1902 + }, + { + "epoch": 0.6353923205342237, + "grad_norm": 0.49331715107199153, + "learning_rate": 8.62451432501449e-06, + "loss": 0.311, + "step": 1903 + }, + { + "epoch": 0.6357262103505843, + "grad_norm": 0.46541945842187493, + "learning_rate": 8.622506577853538e-06, + "loss": 0.2849, + "step": 1904 + }, + { + "epoch": 0.6360601001669449, + "grad_norm": 0.4883706544126179, + "learning_rate": 8.62049760050808e-06, + "loss": 0.3007, + "step": 1905 + }, + { + "epoch": 0.6363939899833055, + "grad_norm": 0.46407114714439934, + "learning_rate": 8.618487393660355e-06, + "loss": 0.2832, + "step": 1906 + }, + { + "epoch": 0.6367278797996662, + "grad_norm": 0.46552952057691116, + "learning_rate": 8.616475957993021e-06, + "loss": 0.2921, + "step": 1907 + }, + { + "epoch": 0.6370617696160267, + "grad_norm": 0.47821831143344795, + "learning_rate": 8.61446329418915e-06, + "loss": 0.2806, + "step": 1908 + }, + { + "epoch": 0.6373956594323873, + "grad_norm": 0.4624415294106391, + "learning_rate": 8.612449402932233e-06, + "loss": 0.2931, + "step": 1909 + }, + { + "epoch": 0.6377295492487479, + "grad_norm": 0.4675228546734753, + "learning_rate": 8.610434284906175e-06, + "loss": 0.2867, + "step": 1910 + }, + { + "epoch": 0.6380634390651085, + "grad_norm": 0.4987265099018842, + "learning_rate": 8.608417940795304e-06, + "loss": 0.297, + "step": 1911 + }, + { + "epoch": 0.6383973288814692, + "grad_norm": 0.4834473075557041, + "learning_rate": 8.606400371284356e-06, + "loss": 0.294, + "step": 1912 + }, + { + "epoch": 0.6387312186978297, + "grad_norm": 0.5268127231737136, + "learning_rate": 8.604381577058486e-06, + "loss": 0.2933, + "step": 1913 + }, + { + "epoch": 0.6390651085141903, + "grad_norm": 0.48183615816690345, + "learning_rate": 8.602361558803272e-06, + "loss": 0.291, + "step": 1914 + }, + { + "epoch": 0.6393989983305509, + "grad_norm": 0.4910289874829549, + "learning_rate": 8.600340317204697e-06, + "loss": 0.2875, + "step": 1915 + }, + { + "epoch": 0.6397328881469115, + "grad_norm": 0.49827937392000043, + "learning_rate": 8.598317852949163e-06, + "loss": 0.2797, + "step": 1916 + }, + { + "epoch": 0.6400667779632722, + "grad_norm": 0.4642132493514586, + "learning_rate": 8.596294166723492e-06, + "loss": 0.2953, + "step": 1917 + }, + { + "epoch": 0.6404006677796327, + "grad_norm": 0.47700466589622065, + "learning_rate": 8.594269259214918e-06, + "loss": 0.2928, + "step": 1918 + }, + { + "epoch": 0.6407345575959933, + "grad_norm": 0.514982139731658, + "learning_rate": 8.592243131111085e-06, + "loss": 0.312, + "step": 1919 + }, + { + "epoch": 0.6410684474123539, + "grad_norm": 0.4835626417873231, + "learning_rate": 8.590215783100058e-06, + "loss": 0.2993, + "step": 1920 + }, + { + "epoch": 0.6414023372287145, + "grad_norm": 0.4703239851091885, + "learning_rate": 8.588187215870315e-06, + "loss": 0.2794, + "step": 1921 + }, + { + "epoch": 0.6417362270450752, + "grad_norm": 0.48213975928613617, + "learning_rate": 8.586157430110747e-06, + "loss": 0.2889, + "step": 1922 + }, + { + "epoch": 0.6420701168614358, + "grad_norm": 0.4729620695852385, + "learning_rate": 8.584126426510655e-06, + "loss": 0.3031, + "step": 1923 + }, + { + "epoch": 0.6424040066777963, + "grad_norm": 0.46537953127438997, + "learning_rate": 8.582094205759762e-06, + "loss": 0.3001, + "step": 1924 + }, + { + "epoch": 0.6427378964941569, + "grad_norm": 0.4976942147141687, + "learning_rate": 8.580060768548194e-06, + "loss": 0.2818, + "step": 1925 + }, + { + "epoch": 0.6430717863105175, + "grad_norm": 0.49925306075939413, + "learning_rate": 8.578026115566503e-06, + "loss": 0.288, + "step": 1926 + }, + { + "epoch": 0.6434056761268782, + "grad_norm": 0.486278515966021, + "learning_rate": 8.57599024750564e-06, + "loss": 0.2897, + "step": 1927 + }, + { + "epoch": 0.6437395659432388, + "grad_norm": 0.49422464460083587, + "learning_rate": 8.573953165056978e-06, + "loss": 0.3052, + "step": 1928 + }, + { + "epoch": 0.6440734557595993, + "grad_norm": 0.48390926308533294, + "learning_rate": 8.571914868912303e-06, + "loss": 0.2779, + "step": 1929 + }, + { + "epoch": 0.6444073455759599, + "grad_norm": 0.5112891073011894, + "learning_rate": 8.569875359763804e-06, + "loss": 0.2894, + "step": 1930 + }, + { + "epoch": 0.6447412353923205, + "grad_norm": 0.5045446078879864, + "learning_rate": 8.56783463830409e-06, + "loss": 0.298, + "step": 1931 + }, + { + "epoch": 0.6450751252086812, + "grad_norm": 0.4658005984731866, + "learning_rate": 8.565792705226183e-06, + "loss": 0.2953, + "step": 1932 + }, + { + "epoch": 0.6454090150250418, + "grad_norm": 0.48661255471272896, + "learning_rate": 8.563749561223507e-06, + "loss": 0.2844, + "step": 1933 + }, + { + "epoch": 0.6457429048414023, + "grad_norm": 0.48553179227087806, + "learning_rate": 8.561705206989907e-06, + "loss": 0.2929, + "step": 1934 + }, + { + "epoch": 0.6460767946577629, + "grad_norm": 0.5169083356732077, + "learning_rate": 8.559659643219634e-06, + "loss": 0.3048, + "step": 1935 + }, + { + "epoch": 0.6464106844741235, + "grad_norm": 0.471814386165013, + "learning_rate": 8.55761287060735e-06, + "loss": 0.2906, + "step": 1936 + }, + { + "epoch": 0.6467445742904842, + "grad_norm": 0.48579015598634795, + "learning_rate": 8.55556488984813e-06, + "loss": 0.2883, + "step": 1937 + }, + { + "epoch": 0.6470784641068448, + "grad_norm": 0.49995409616333664, + "learning_rate": 8.553515701637458e-06, + "loss": 0.295, + "step": 1938 + }, + { + "epoch": 0.6474123539232054, + "grad_norm": 0.4884538490217293, + "learning_rate": 8.551465306671225e-06, + "loss": 0.2906, + "step": 1939 + }, + { + "epoch": 0.6477462437395659, + "grad_norm": 0.4594075005829511, + "learning_rate": 8.549413705645737e-06, + "loss": 0.2839, + "step": 1940 + }, + { + "epoch": 0.6480801335559265, + "grad_norm": 0.46303268685039406, + "learning_rate": 8.547360899257707e-06, + "loss": 0.2881, + "step": 1941 + }, + { + "epoch": 0.6484140233722872, + "grad_norm": 0.47419809190203344, + "learning_rate": 8.545306888204255e-06, + "loss": 0.2879, + "step": 1942 + }, + { + "epoch": 0.6487479131886478, + "grad_norm": 0.4800146608496853, + "learning_rate": 8.543251673182915e-06, + "loss": 0.3027, + "step": 1943 + }, + { + "epoch": 0.6490818030050084, + "grad_norm": 0.48868955109403006, + "learning_rate": 8.541195254891628e-06, + "loss": 0.2979, + "step": 1944 + }, + { + "epoch": 0.6494156928213689, + "grad_norm": 0.5039122168151536, + "learning_rate": 8.53913763402874e-06, + "loss": 0.2886, + "step": 1945 + }, + { + "epoch": 0.6497495826377295, + "grad_norm": 0.477409645297758, + "learning_rate": 8.537078811293012e-06, + "loss": 0.2976, + "step": 1946 + }, + { + "epoch": 0.6500834724540901, + "grad_norm": 0.4611988379710201, + "learning_rate": 8.535018787383606e-06, + "loss": 0.2931, + "step": 1947 + }, + { + "epoch": 0.6504173622704508, + "grad_norm": 0.486070527771172, + "learning_rate": 8.532957563000096e-06, + "loss": 0.2966, + "step": 1948 + }, + { + "epoch": 0.6507512520868114, + "grad_norm": 0.5001111504780358, + "learning_rate": 8.530895138842467e-06, + "loss": 0.2947, + "step": 1949 + }, + { + "epoch": 0.6510851419031719, + "grad_norm": 0.4702442545395896, + "learning_rate": 8.528831515611105e-06, + "loss": 0.2868, + "step": 1950 + }, + { + "epoch": 0.6514190317195325, + "grad_norm": 0.5064896911992782, + "learning_rate": 8.526766694006804e-06, + "loss": 0.2915, + "step": 1951 + }, + { + "epoch": 0.6517529215358931, + "grad_norm": 0.5013736512860885, + "learning_rate": 8.524700674730769e-06, + "loss": 0.2942, + "step": 1952 + }, + { + "epoch": 0.6520868113522538, + "grad_norm": 0.4752137694291365, + "learning_rate": 8.52263345848461e-06, + "loss": 0.2948, + "step": 1953 + }, + { + "epoch": 0.6524207011686144, + "grad_norm": 0.47691032600485195, + "learning_rate": 8.52056504597034e-06, + "loss": 0.2933, + "step": 1954 + }, + { + "epoch": 0.6527545909849749, + "grad_norm": 0.47718183125943625, + "learning_rate": 8.518495437890383e-06, + "loss": 0.2957, + "step": 1955 + }, + { + "epoch": 0.6530884808013355, + "grad_norm": 0.47514537677859153, + "learning_rate": 8.516424634947569e-06, + "loss": 0.2753, + "step": 1956 + }, + { + "epoch": 0.6534223706176961, + "grad_norm": 0.45869367032171304, + "learning_rate": 8.514352637845128e-06, + "loss": 0.294, + "step": 1957 + }, + { + "epoch": 0.6537562604340568, + "grad_norm": 0.4603302155565696, + "learning_rate": 8.512279447286704e-06, + "loss": 0.2916, + "step": 1958 + }, + { + "epoch": 0.6540901502504174, + "grad_norm": 0.4710495151599902, + "learning_rate": 8.510205063976338e-06, + "loss": 0.2917, + "step": 1959 + }, + { + "epoch": 0.654424040066778, + "grad_norm": 0.4955318001952851, + "learning_rate": 8.50812948861848e-06, + "loss": 0.2972, + "step": 1960 + }, + { + "epoch": 0.6547579298831385, + "grad_norm": 0.4948974785216569, + "learning_rate": 8.506052721917989e-06, + "loss": 0.2998, + "step": 1961 + }, + { + "epoch": 0.6550918196994991, + "grad_norm": 0.4853417200048731, + "learning_rate": 8.50397476458012e-06, + "loss": 0.3025, + "step": 1962 + }, + { + "epoch": 0.6554257095158598, + "grad_norm": 0.48636647622326595, + "learning_rate": 8.501895617310538e-06, + "loss": 0.2934, + "step": 1963 + }, + { + "epoch": 0.6557595993322204, + "grad_norm": 0.4830859544549205, + "learning_rate": 8.499815280815309e-06, + "loss": 0.2973, + "step": 1964 + }, + { + "epoch": 0.656093489148581, + "grad_norm": 0.5104649998060025, + "learning_rate": 8.497733755800906e-06, + "loss": 0.2989, + "step": 1965 + }, + { + "epoch": 0.6564273789649415, + "grad_norm": 0.4827152874144299, + "learning_rate": 8.495651042974206e-06, + "loss": 0.3045, + "step": 1966 + }, + { + "epoch": 0.6567612687813021, + "grad_norm": 0.498597578221063, + "learning_rate": 8.493567143042485e-06, + "loss": 0.2967, + "step": 1967 + }, + { + "epoch": 0.6570951585976628, + "grad_norm": 0.5001608299632879, + "learning_rate": 8.491482056713425e-06, + "loss": 0.2852, + "step": 1968 + }, + { + "epoch": 0.6574290484140234, + "grad_norm": 0.47601642499812497, + "learning_rate": 8.48939578469511e-06, + "loss": 0.2834, + "step": 1969 + }, + { + "epoch": 0.657762938230384, + "grad_norm": 0.4782664332632602, + "learning_rate": 8.487308327696028e-06, + "loss": 0.2929, + "step": 1970 + }, + { + "epoch": 0.6580968280467445, + "grad_norm": 0.503675723073325, + "learning_rate": 8.48521968642507e-06, + "loss": 0.2982, + "step": 1971 + }, + { + "epoch": 0.6584307178631051, + "grad_norm": 0.4783788771006154, + "learning_rate": 8.483129861591525e-06, + "loss": 0.2883, + "step": 1972 + }, + { + "epoch": 0.6587646076794658, + "grad_norm": 0.4810778489338461, + "learning_rate": 8.481038853905091e-06, + "loss": 0.2964, + "step": 1973 + }, + { + "epoch": 0.6590984974958264, + "grad_norm": 0.4990411116440297, + "learning_rate": 8.478946664075857e-06, + "loss": 0.2791, + "step": 1974 + }, + { + "epoch": 0.659432387312187, + "grad_norm": 0.482275151831079, + "learning_rate": 8.476853292814324e-06, + "loss": 0.2834, + "step": 1975 + }, + { + "epoch": 0.6597662771285476, + "grad_norm": 0.4868567946537464, + "learning_rate": 8.47475874083139e-06, + "loss": 0.2811, + "step": 1976 + }, + { + "epoch": 0.6601001669449081, + "grad_norm": 0.4647604073776631, + "learning_rate": 8.472663008838353e-06, + "loss": 0.2868, + "step": 1977 + }, + { + "epoch": 0.6604340567612688, + "grad_norm": 0.4664539773664884, + "learning_rate": 8.470566097546915e-06, + "loss": 0.2806, + "step": 1978 + }, + { + "epoch": 0.6607679465776294, + "grad_norm": 0.4922648815930715, + "learning_rate": 8.468468007669173e-06, + "loss": 0.2959, + "step": 1979 + }, + { + "epoch": 0.66110183639399, + "grad_norm": 0.4934346869993678, + "learning_rate": 8.466368739917628e-06, + "loss": 0.2786, + "step": 1980 + }, + { + "epoch": 0.6614357262103506, + "grad_norm": 0.45703297314016145, + "learning_rate": 8.46426829500518e-06, + "loss": 0.2903, + "step": 1981 + }, + { + "epoch": 0.6617696160267111, + "grad_norm": 0.5100152681579417, + "learning_rate": 8.462166673645134e-06, + "loss": 0.2838, + "step": 1982 + }, + { + "epoch": 0.6621035058430718, + "grad_norm": 0.4958417545747194, + "learning_rate": 8.460063876551186e-06, + "loss": 0.2964, + "step": 1983 + }, + { + "epoch": 0.6624373956594324, + "grad_norm": 0.4639312105881056, + "learning_rate": 8.457959904437434e-06, + "loss": 0.2858, + "step": 1984 + }, + { + "epoch": 0.662771285475793, + "grad_norm": 0.4537448893198869, + "learning_rate": 8.455854758018377e-06, + "loss": 0.2862, + "step": 1985 + }, + { + "epoch": 0.6631051752921536, + "grad_norm": 0.4685886517648523, + "learning_rate": 8.453748438008912e-06, + "loss": 0.2903, + "step": 1986 + }, + { + "epoch": 0.6634390651085141, + "grad_norm": 0.4907383464406096, + "learning_rate": 8.451640945124336e-06, + "loss": 0.2919, + "step": 1987 + }, + { + "epoch": 0.6637729549248748, + "grad_norm": 0.5063063732225663, + "learning_rate": 8.449532280080341e-06, + "loss": 0.297, + "step": 1988 + }, + { + "epoch": 0.6641068447412354, + "grad_norm": 0.4708009088188142, + "learning_rate": 8.447422443593018e-06, + "loss": 0.2832, + "step": 1989 + }, + { + "epoch": 0.664440734557596, + "grad_norm": 0.47592975502978296, + "learning_rate": 8.445311436378859e-06, + "loss": 0.2971, + "step": 1990 + }, + { + "epoch": 0.6647746243739566, + "grad_norm": 0.49657524658362606, + "learning_rate": 8.443199259154748e-06, + "loss": 0.2997, + "step": 1991 + }, + { + "epoch": 0.6651085141903172, + "grad_norm": 0.49326969365131534, + "learning_rate": 8.441085912637973e-06, + "loss": 0.2777, + "step": 1992 + }, + { + "epoch": 0.6654424040066778, + "grad_norm": 0.45999156277588926, + "learning_rate": 8.438971397546214e-06, + "loss": 0.2892, + "step": 1993 + }, + { + "epoch": 0.6657762938230384, + "grad_norm": 0.4823027991513219, + "learning_rate": 8.436855714597548e-06, + "loss": 0.2965, + "step": 1994 + }, + { + "epoch": 0.666110183639399, + "grad_norm": 0.48763173609570803, + "learning_rate": 8.43473886451045e-06, + "loss": 0.2977, + "step": 1995 + }, + { + "epoch": 0.6664440734557596, + "grad_norm": 0.4694083011651597, + "learning_rate": 8.432620848003795e-06, + "loss": 0.2832, + "step": 1996 + }, + { + "epoch": 0.6667779632721202, + "grad_norm": 0.4752011813247931, + "learning_rate": 8.430501665796846e-06, + "loss": 0.2862, + "step": 1997 + }, + { + "epoch": 0.6671118530884808, + "grad_norm": 0.4631610424111215, + "learning_rate": 8.428381318609268e-06, + "loss": 0.285, + "step": 1998 + }, + { + "epoch": 0.6674457429048414, + "grad_norm": 0.47483942572242605, + "learning_rate": 8.426259807161123e-06, + "loss": 0.2952, + "step": 1999 + }, + { + "epoch": 0.667779632721202, + "grad_norm": 0.4744748733926771, + "learning_rate": 8.42413713217286e-06, + "loss": 0.2826, + "step": 2000 + }, + { + "epoch": 0.6681135225375626, + "grad_norm": 0.4670053824008421, + "learning_rate": 8.422013294365332e-06, + "loss": 0.2943, + "step": 2001 + }, + { + "epoch": 0.6684474123539232, + "grad_norm": 0.46376614691480944, + "learning_rate": 8.419888294459782e-06, + "loss": 0.2883, + "step": 2002 + }, + { + "epoch": 0.6687813021702838, + "grad_norm": 0.4816114351725496, + "learning_rate": 8.417762133177849e-06, + "loss": 0.2908, + "step": 2003 + }, + { + "epoch": 0.6691151919866444, + "grad_norm": 0.4687177438287116, + "learning_rate": 8.415634811241568e-06, + "loss": 0.2874, + "step": 2004 + }, + { + "epoch": 0.669449081803005, + "grad_norm": 0.4966233842101538, + "learning_rate": 8.413506329373364e-06, + "loss": 0.2895, + "step": 2005 + }, + { + "epoch": 0.6697829716193656, + "grad_norm": 0.47811626994861645, + "learning_rate": 8.41137668829606e-06, + "loss": 0.2889, + "step": 2006 + }, + { + "epoch": 0.6701168614357262, + "grad_norm": 0.5032323522097676, + "learning_rate": 8.40924588873287e-06, + "loss": 0.2996, + "step": 2007 + }, + { + "epoch": 0.6704507512520869, + "grad_norm": 0.490574482224379, + "learning_rate": 8.407113931407404e-06, + "loss": 0.2788, + "step": 2008 + }, + { + "epoch": 0.6707846410684474, + "grad_norm": 0.48642093837918565, + "learning_rate": 8.404980817043661e-06, + "loss": 0.3004, + "step": 2009 + }, + { + "epoch": 0.671118530884808, + "grad_norm": 0.4950133612519524, + "learning_rate": 8.40284654636604e-06, + "loss": 0.2871, + "step": 2010 + }, + { + "epoch": 0.6714524207011686, + "grad_norm": 0.47409155238120754, + "learning_rate": 8.400711120099325e-06, + "loss": 0.2902, + "step": 2011 + }, + { + "epoch": 0.6717863105175292, + "grad_norm": 0.48783085506795176, + "learning_rate": 8.398574538968697e-06, + "loss": 0.2895, + "step": 2012 + }, + { + "epoch": 0.6721202003338899, + "grad_norm": 0.4690633095131049, + "learning_rate": 8.396436803699728e-06, + "loss": 0.2847, + "step": 2013 + }, + { + "epoch": 0.6724540901502504, + "grad_norm": 0.4691931253200212, + "learning_rate": 8.394297915018378e-06, + "loss": 0.277, + "step": 2014 + }, + { + "epoch": 0.672787979966611, + "grad_norm": 0.48283320362629745, + "learning_rate": 8.39215787365101e-06, + "loss": 0.2864, + "step": 2015 + }, + { + "epoch": 0.6731218697829716, + "grad_norm": 0.47959278045700304, + "learning_rate": 8.390016680324364e-06, + "loss": 0.2804, + "step": 2016 + }, + { + "epoch": 0.6734557595993322, + "grad_norm": 0.49582636847562006, + "learning_rate": 8.387874335765582e-06, + "loss": 0.3062, + "step": 2017 + }, + { + "epoch": 0.6737896494156929, + "grad_norm": 0.49259264556894006, + "learning_rate": 8.385730840702193e-06, + "loss": 0.2965, + "step": 2018 + }, + { + "epoch": 0.6741235392320534, + "grad_norm": 0.47860770073710424, + "learning_rate": 8.383586195862114e-06, + "loss": 0.2726, + "step": 2019 + }, + { + "epoch": 0.674457429048414, + "grad_norm": 0.494598069204726, + "learning_rate": 8.38144040197366e-06, + "loss": 0.3052, + "step": 2020 + }, + { + "epoch": 0.6747913188647746, + "grad_norm": 0.4742789173337979, + "learning_rate": 8.379293459765527e-06, + "loss": 0.2874, + "step": 2021 + }, + { + "epoch": 0.6751252086811352, + "grad_norm": 0.4816621044801787, + "learning_rate": 8.377145369966806e-06, + "loss": 0.2895, + "step": 2022 + }, + { + "epoch": 0.6754590984974959, + "grad_norm": 0.47692832673316876, + "learning_rate": 8.374996133306981e-06, + "loss": 0.2925, + "step": 2023 + }, + { + "epoch": 0.6757929883138565, + "grad_norm": 0.4790359556203248, + "learning_rate": 8.372845750515917e-06, + "loss": 0.2928, + "step": 2024 + }, + { + "epoch": 0.676126878130217, + "grad_norm": 0.467473487806262, + "learning_rate": 8.370694222323876e-06, + "loss": 0.2986, + "step": 2025 + }, + { + "epoch": 0.6764607679465776, + "grad_norm": 0.4717659900672362, + "learning_rate": 8.368541549461504e-06, + "loss": 0.2974, + "step": 2026 + }, + { + "epoch": 0.6767946577629382, + "grad_norm": 0.4652918240331178, + "learning_rate": 8.36638773265984e-06, + "loss": 0.2889, + "step": 2027 + }, + { + "epoch": 0.6771285475792989, + "grad_norm": 0.4881275365647966, + "learning_rate": 8.364232772650306e-06, + "loss": 0.2851, + "step": 2028 + }, + { + "epoch": 0.6774624373956595, + "grad_norm": 0.5073581090486833, + "learning_rate": 8.362076670164718e-06, + "loss": 0.3012, + "step": 2029 + }, + { + "epoch": 0.67779632721202, + "grad_norm": 0.4652676485316276, + "learning_rate": 8.359919425935276e-06, + "loss": 0.2905, + "step": 2030 + }, + { + "epoch": 0.6781302170283806, + "grad_norm": 0.5024085244543889, + "learning_rate": 8.357761040694568e-06, + "loss": 0.3098, + "step": 2031 + }, + { + "epoch": 0.6784641068447412, + "grad_norm": 0.5026023746330929, + "learning_rate": 8.35560151517557e-06, + "loss": 0.2941, + "step": 2032 + }, + { + "epoch": 0.6787979966611019, + "grad_norm": 0.49261322039026406, + "learning_rate": 8.353440850111651e-06, + "loss": 0.296, + "step": 2033 + }, + { + "epoch": 0.6791318864774625, + "grad_norm": 0.5033975986039905, + "learning_rate": 8.351279046236556e-06, + "loss": 0.2891, + "step": 2034 + }, + { + "epoch": 0.679465776293823, + "grad_norm": 0.46262512738238243, + "learning_rate": 8.349116104284423e-06, + "loss": 0.2819, + "step": 2035 + }, + { + "epoch": 0.6797996661101836, + "grad_norm": 0.4728978723374501, + "learning_rate": 8.346952024989778e-06, + "loss": 0.2741, + "step": 2036 + }, + { + "epoch": 0.6801335559265442, + "grad_norm": 0.4838239120770336, + "learning_rate": 8.34478680908753e-06, + "loss": 0.2938, + "step": 2037 + }, + { + "epoch": 0.6804674457429049, + "grad_norm": 0.4699349332718422, + "learning_rate": 8.342620457312977e-06, + "loss": 0.292, + "step": 2038 + }, + { + "epoch": 0.6808013355592655, + "grad_norm": 0.45476074989197124, + "learning_rate": 8.340452970401798e-06, + "loss": 0.2816, + "step": 2039 + }, + { + "epoch": 0.6811352253756261, + "grad_norm": 0.5024731355967874, + "learning_rate": 8.338284349090063e-06, + "loss": 0.2934, + "step": 2040 + }, + { + "epoch": 0.6814691151919866, + "grad_norm": 0.46831982591051324, + "learning_rate": 8.33611459411422e-06, + "loss": 0.2827, + "step": 2041 + }, + { + "epoch": 0.6818030050083472, + "grad_norm": 0.48296667173694247, + "learning_rate": 8.333943706211112e-06, + "loss": 0.2884, + "step": 2042 + }, + { + "epoch": 0.6821368948247079, + "grad_norm": 0.47104095386125594, + "learning_rate": 8.331771686117959e-06, + "loss": 0.284, + "step": 2043 + }, + { + "epoch": 0.6824707846410685, + "grad_norm": 0.4709184205296841, + "learning_rate": 8.329598534572368e-06, + "loss": 0.2858, + "step": 2044 + }, + { + "epoch": 0.6828046744574291, + "grad_norm": 0.46167855097333493, + "learning_rate": 8.327424252312328e-06, + "loss": 0.2841, + "step": 2045 + }, + { + "epoch": 0.6831385642737896, + "grad_norm": 0.46539074302145456, + "learning_rate": 8.325248840076219e-06, + "loss": 0.2884, + "step": 2046 + }, + { + "epoch": 0.6834724540901502, + "grad_norm": 0.5007054790597592, + "learning_rate": 8.323072298602795e-06, + "loss": 0.2884, + "step": 2047 + }, + { + "epoch": 0.6838063439065108, + "grad_norm": 0.4751012658331925, + "learning_rate": 8.3208946286312e-06, + "loss": 0.2885, + "step": 2048 + }, + { + "epoch": 0.6841402337228715, + "grad_norm": 0.46496526938914307, + "learning_rate": 8.318715830900959e-06, + "loss": 0.2816, + "step": 2049 + }, + { + "epoch": 0.6844741235392321, + "grad_norm": 0.49157498747997513, + "learning_rate": 8.316535906151981e-06, + "loss": 0.2806, + "step": 2050 + }, + { + "epoch": 0.6848080133555926, + "grad_norm": 0.4741233049223069, + "learning_rate": 8.314354855124559e-06, + "loss": 0.2912, + "step": 2051 + }, + { + "epoch": 0.6851419031719532, + "grad_norm": 0.4932571734245302, + "learning_rate": 8.312172678559363e-06, + "loss": 0.2813, + "step": 2052 + }, + { + "epoch": 0.6854757929883138, + "grad_norm": 0.48006331930865204, + "learning_rate": 8.30998937719745e-06, + "loss": 0.2864, + "step": 2053 + }, + { + "epoch": 0.6858096828046745, + "grad_norm": 0.4877894695889098, + "learning_rate": 8.307804951780258e-06, + "loss": 0.2814, + "step": 2054 + }, + { + "epoch": 0.6861435726210351, + "grad_norm": 0.4513893719433002, + "learning_rate": 8.305619403049607e-06, + "loss": 0.2903, + "step": 2055 + }, + { + "epoch": 0.6864774624373957, + "grad_norm": 0.46900157426508526, + "learning_rate": 8.3034327317477e-06, + "loss": 0.2886, + "step": 2056 + }, + { + "epoch": 0.6868113522537562, + "grad_norm": 0.4811169236135389, + "learning_rate": 8.301244938617117e-06, + "loss": 0.2885, + "step": 2057 + }, + { + "epoch": 0.6871452420701168, + "grad_norm": 0.4624988066848961, + "learning_rate": 8.29905602440082e-06, + "loss": 0.2912, + "step": 2058 + }, + { + "epoch": 0.6874791318864775, + "grad_norm": 0.4498700289420117, + "learning_rate": 8.296865989842154e-06, + "loss": 0.2758, + "step": 2059 + }, + { + "epoch": 0.6878130217028381, + "grad_norm": 0.47331109135010047, + "learning_rate": 8.294674835684846e-06, + "loss": 0.2966, + "step": 2060 + }, + { + "epoch": 0.6881469115191987, + "grad_norm": 0.48080370477080325, + "learning_rate": 8.292482562672997e-06, + "loss": 0.2856, + "step": 2061 + }, + { + "epoch": 0.6884808013355592, + "grad_norm": 0.498422525497153, + "learning_rate": 8.290289171551093e-06, + "loss": 0.293, + "step": 2062 + }, + { + "epoch": 0.6888146911519198, + "grad_norm": 0.49001722296900935, + "learning_rate": 8.288094663064e-06, + "loss": 0.2936, + "step": 2063 + }, + { + "epoch": 0.6891485809682805, + "grad_norm": 0.4967165214241459, + "learning_rate": 8.285899037956958e-06, + "loss": 0.2886, + "step": 2064 + }, + { + "epoch": 0.6894824707846411, + "grad_norm": 0.45024905420484024, + "learning_rate": 8.283702296975593e-06, + "loss": 0.292, + "step": 2065 + }, + { + "epoch": 0.6898163606010017, + "grad_norm": 0.4876387460120092, + "learning_rate": 8.281504440865905e-06, + "loss": 0.2834, + "step": 2066 + }, + { + "epoch": 0.6901502504173622, + "grad_norm": 0.46807946875149586, + "learning_rate": 8.279305470374278e-06, + "loss": 0.2865, + "step": 2067 + }, + { + "epoch": 0.6904841402337228, + "grad_norm": 0.47711627292626746, + "learning_rate": 8.277105386247466e-06, + "loss": 0.2861, + "step": 2068 + }, + { + "epoch": 0.6908180300500835, + "grad_norm": 0.46566624294586, + "learning_rate": 8.274904189232609e-06, + "loss": 0.2972, + "step": 2069 + }, + { + "epoch": 0.6911519198664441, + "grad_norm": 0.46950288077669533, + "learning_rate": 8.272701880077221e-06, + "loss": 0.2844, + "step": 2070 + }, + { + "epoch": 0.6914858096828047, + "grad_norm": 0.4712753904674148, + "learning_rate": 8.270498459529197e-06, + "loss": 0.2914, + "step": 2071 + }, + { + "epoch": 0.6918196994991653, + "grad_norm": 0.45922625571887027, + "learning_rate": 8.268293928336805e-06, + "loss": 0.2833, + "step": 2072 + }, + { + "epoch": 0.6921535893155258, + "grad_norm": 0.4773445592672361, + "learning_rate": 8.266088287248696e-06, + "loss": 0.289, + "step": 2073 + }, + { + "epoch": 0.6924874791318865, + "grad_norm": 0.4744245621471659, + "learning_rate": 8.263881537013888e-06, + "loss": 0.2954, + "step": 2074 + }, + { + "epoch": 0.6928213689482471, + "grad_norm": 0.4697444460553732, + "learning_rate": 8.261673678381786e-06, + "loss": 0.2958, + "step": 2075 + }, + { + "epoch": 0.6931552587646077, + "grad_norm": 0.45386955624072084, + "learning_rate": 8.259464712102168e-06, + "loss": 0.287, + "step": 2076 + }, + { + "epoch": 0.6934891485809683, + "grad_norm": 0.48441596408475557, + "learning_rate": 8.257254638925188e-06, + "loss": 0.302, + "step": 2077 + }, + { + "epoch": 0.6938230383973288, + "grad_norm": 0.4658813868392801, + "learning_rate": 8.25504345960137e-06, + "loss": 0.2847, + "step": 2078 + }, + { + "epoch": 0.6941569282136895, + "grad_norm": 0.481119493613914, + "learning_rate": 8.252831174881626e-06, + "loss": 0.2965, + "step": 2079 + }, + { + "epoch": 0.6944908180300501, + "grad_norm": 0.482383191010165, + "learning_rate": 8.250617785517232e-06, + "loss": 0.286, + "step": 2080 + }, + { + "epoch": 0.6948247078464107, + "grad_norm": 0.4993473742735979, + "learning_rate": 8.248403292259845e-06, + "loss": 0.2891, + "step": 2081 + }, + { + "epoch": 0.6951585976627713, + "grad_norm": 0.4491647634149955, + "learning_rate": 8.246187695861495e-06, + "loss": 0.2804, + "step": 2082 + }, + { + "epoch": 0.6954924874791318, + "grad_norm": 0.47759949400981716, + "learning_rate": 8.243970997074589e-06, + "loss": 0.2945, + "step": 2083 + }, + { + "epoch": 0.6958263772954925, + "grad_norm": 0.47611528009348636, + "learning_rate": 8.241753196651903e-06, + "loss": 0.3025, + "step": 2084 + }, + { + "epoch": 0.6961602671118531, + "grad_norm": 0.44991512253929655, + "learning_rate": 8.239534295346593e-06, + "loss": 0.291, + "step": 2085 + }, + { + "epoch": 0.6964941569282137, + "grad_norm": 0.4694770356815053, + "learning_rate": 8.237314293912186e-06, + "loss": 0.3001, + "step": 2086 + }, + { + "epoch": 0.6968280467445743, + "grad_norm": 0.47563455609820576, + "learning_rate": 8.235093193102583e-06, + "loss": 0.2983, + "step": 2087 + }, + { + "epoch": 0.6971619365609348, + "grad_norm": 0.47852272754468006, + "learning_rate": 8.23287099367206e-06, + "loss": 0.2875, + "step": 2088 + }, + { + "epoch": 0.6974958263772955, + "grad_norm": 0.4548416673996801, + "learning_rate": 8.23064769637526e-06, + "loss": 0.2775, + "step": 2089 + }, + { + "epoch": 0.6978297161936561, + "grad_norm": 0.47483811279638194, + "learning_rate": 8.228423301967207e-06, + "loss": 0.2879, + "step": 2090 + }, + { + "epoch": 0.6981636060100167, + "grad_norm": 0.4893907231638347, + "learning_rate": 8.226197811203294e-06, + "loss": 0.2943, + "step": 2091 + }, + { + "epoch": 0.6984974958263773, + "grad_norm": 0.4598597457188059, + "learning_rate": 8.223971224839282e-06, + "loss": 0.2753, + "step": 2092 + }, + { + "epoch": 0.698831385642738, + "grad_norm": 0.48903645070770746, + "learning_rate": 8.221743543631314e-06, + "loss": 0.2987, + "step": 2093 + }, + { + "epoch": 0.6991652754590985, + "grad_norm": 0.4722399542300538, + "learning_rate": 8.219514768335895e-06, + "loss": 0.2842, + "step": 2094 + }, + { + "epoch": 0.6994991652754591, + "grad_norm": 0.47010241643266215, + "learning_rate": 8.217284899709907e-06, + "loss": 0.2981, + "step": 2095 + }, + { + "epoch": 0.6998330550918197, + "grad_norm": 0.4839423781033225, + "learning_rate": 8.215053938510602e-06, + "loss": 0.2868, + "step": 2096 + }, + { + "epoch": 0.7001669449081803, + "grad_norm": 0.4964822824781643, + "learning_rate": 8.212821885495602e-06, + "loss": 0.2797, + "step": 2097 + }, + { + "epoch": 0.7005008347245409, + "grad_norm": 0.5045012091921524, + "learning_rate": 8.2105887414229e-06, + "loss": 0.2921, + "step": 2098 + }, + { + "epoch": 0.7008347245409015, + "grad_norm": 0.479862442107164, + "learning_rate": 8.208354507050865e-06, + "loss": 0.2755, + "step": 2099 + }, + { + "epoch": 0.7011686143572621, + "grad_norm": 0.47110880144039563, + "learning_rate": 8.206119183138223e-06, + "loss": 0.2811, + "step": 2100 + }, + { + "epoch": 0.7015025041736227, + "grad_norm": 0.458818406218651, + "learning_rate": 8.203882770444086e-06, + "loss": 0.2822, + "step": 2101 + }, + { + "epoch": 0.7018363939899833, + "grad_norm": 0.49471100158076664, + "learning_rate": 8.201645269727924e-06, + "loss": 0.2821, + "step": 2102 + }, + { + "epoch": 0.7021702838063439, + "grad_norm": 0.4878038946130286, + "learning_rate": 8.199406681749585e-06, + "loss": 0.2927, + "step": 2103 + }, + { + "epoch": 0.7025041736227045, + "grad_norm": 0.5112676463924098, + "learning_rate": 8.197167007269275e-06, + "loss": 0.293, + "step": 2104 + }, + { + "epoch": 0.7028380634390651, + "grad_norm": 0.4913265650900159, + "learning_rate": 8.194926247047584e-06, + "loss": 0.2966, + "step": 2105 + }, + { + "epoch": 0.7031719532554257, + "grad_norm": 0.46726700381480546, + "learning_rate": 8.192684401845454e-06, + "loss": 0.2848, + "step": 2106 + }, + { + "epoch": 0.7035058430717863, + "grad_norm": 0.4848234227280876, + "learning_rate": 8.190441472424213e-06, + "loss": 0.285, + "step": 2107 + }, + { + "epoch": 0.7038397328881469, + "grad_norm": 0.48022650342423756, + "learning_rate": 8.188197459545541e-06, + "loss": 0.2831, + "step": 2108 + }, + { + "epoch": 0.7041736227045076, + "grad_norm": 0.4934332689446164, + "learning_rate": 8.185952363971495e-06, + "loss": 0.2933, + "step": 2109 + }, + { + "epoch": 0.7045075125208681, + "grad_norm": 0.5119004812326595, + "learning_rate": 8.183706186464502e-06, + "loss": 0.304, + "step": 2110 + }, + { + "epoch": 0.7048414023372287, + "grad_norm": 0.5182207897367869, + "learning_rate": 8.181458927787347e-06, + "loss": 0.286, + "step": 2111 + }, + { + "epoch": 0.7051752921535893, + "grad_norm": 0.5217422205898063, + "learning_rate": 8.17921058870319e-06, + "loss": 0.2916, + "step": 2112 + }, + { + "epoch": 0.7055091819699499, + "grad_norm": 0.43664844711580875, + "learning_rate": 8.176961169975553e-06, + "loss": 0.2792, + "step": 2113 + }, + { + "epoch": 0.7058430717863106, + "grad_norm": 0.4686633578758849, + "learning_rate": 8.17471067236833e-06, + "loss": 0.2905, + "step": 2114 + }, + { + "epoch": 0.7061769616026711, + "grad_norm": 0.4836585379058877, + "learning_rate": 8.172459096645778e-06, + "loss": 0.2949, + "step": 2115 + }, + { + "epoch": 0.7065108514190317, + "grad_norm": 0.4840213257560288, + "learning_rate": 8.170206443572518e-06, + "loss": 0.2897, + "step": 2116 + }, + { + "epoch": 0.7068447412353923, + "grad_norm": 0.47774363389391405, + "learning_rate": 8.167952713913541e-06, + "loss": 0.2772, + "step": 2117 + }, + { + "epoch": 0.7071786310517529, + "grad_norm": 0.4648774146622713, + "learning_rate": 8.165697908434204e-06, + "loss": 0.2751, + "step": 2118 + }, + { + "epoch": 0.7075125208681136, + "grad_norm": 0.46818457113191125, + "learning_rate": 8.163442027900224e-06, + "loss": 0.2809, + "step": 2119 + }, + { + "epoch": 0.7078464106844741, + "grad_norm": 0.4947594051560071, + "learning_rate": 8.161185073077686e-06, + "loss": 0.2879, + "step": 2120 + }, + { + "epoch": 0.7081803005008347, + "grad_norm": 0.47655358955993055, + "learning_rate": 8.158927044733043e-06, + "loss": 0.2908, + "step": 2121 + }, + { + "epoch": 0.7085141903171953, + "grad_norm": 0.474505713709861, + "learning_rate": 8.156667943633107e-06, + "loss": 0.2821, + "step": 2122 + }, + { + "epoch": 0.7088480801335559, + "grad_norm": 0.47621756639848406, + "learning_rate": 8.154407770545061e-06, + "loss": 0.2896, + "step": 2123 + }, + { + "epoch": 0.7091819699499166, + "grad_norm": 0.4850787856228742, + "learning_rate": 8.152146526236444e-06, + "loss": 0.2872, + "step": 2124 + }, + { + "epoch": 0.7095158597662772, + "grad_norm": 0.4823939329486205, + "learning_rate": 8.149884211475166e-06, + "loss": 0.287, + "step": 2125 + }, + { + "epoch": 0.7098497495826377, + "grad_norm": 0.47495613255422253, + "learning_rate": 8.147620827029493e-06, + "loss": 0.2837, + "step": 2126 + }, + { + "epoch": 0.7101836393989983, + "grad_norm": 0.4838294773099701, + "learning_rate": 8.145356373668063e-06, + "loss": 0.2842, + "step": 2127 + }, + { + "epoch": 0.7105175292153589, + "grad_norm": 0.4764366154999356, + "learning_rate": 8.143090852159871e-06, + "loss": 0.2799, + "step": 2128 + }, + { + "epoch": 0.7108514190317196, + "grad_norm": 0.5022361491997055, + "learning_rate": 8.140824263274278e-06, + "loss": 0.2971, + "step": 2129 + }, + { + "epoch": 0.7111853088480802, + "grad_norm": 0.496925462860447, + "learning_rate": 8.138556607781003e-06, + "loss": 0.2941, + "step": 2130 + }, + { + "epoch": 0.7115191986644407, + "grad_norm": 0.4618340435146242, + "learning_rate": 8.136287886450132e-06, + "loss": 0.2901, + "step": 2131 + }, + { + "epoch": 0.7118530884808013, + "grad_norm": 0.4794492041378135, + "learning_rate": 8.13401810005211e-06, + "loss": 0.283, + "step": 2132 + }, + { + "epoch": 0.7121869782971619, + "grad_norm": 0.4892111658172144, + "learning_rate": 8.131747249357745e-06, + "loss": 0.2884, + "step": 2133 + }, + { + "epoch": 0.7125208681135226, + "grad_norm": 0.4602451861124919, + "learning_rate": 8.129475335138207e-06, + "loss": 0.2856, + "step": 2134 + }, + { + "epoch": 0.7128547579298832, + "grad_norm": 0.49174959312455835, + "learning_rate": 8.127202358165025e-06, + "loss": 0.2857, + "step": 2135 + }, + { + "epoch": 0.7131886477462437, + "grad_norm": 0.458956944946768, + "learning_rate": 8.124928319210093e-06, + "loss": 0.2883, + "step": 2136 + }, + { + "epoch": 0.7135225375626043, + "grad_norm": 0.4609727592853188, + "learning_rate": 8.12265321904566e-06, + "loss": 0.2803, + "step": 2137 + }, + { + "epoch": 0.7138564273789649, + "grad_norm": 0.48382982662022966, + "learning_rate": 8.120377058444336e-06, + "loss": 0.2966, + "step": 2138 + }, + { + "epoch": 0.7141903171953256, + "grad_norm": 0.49256105420710466, + "learning_rate": 8.1180998381791e-06, + "loss": 0.2942, + "step": 2139 + }, + { + "epoch": 0.7145242070116862, + "grad_norm": 0.48785115846313853, + "learning_rate": 8.11582155902328e-06, + "loss": 0.2904, + "step": 2140 + }, + { + "epoch": 0.7148580968280468, + "grad_norm": 0.48767616263486724, + "learning_rate": 8.113542221750567e-06, + "loss": 0.2882, + "step": 2141 + }, + { + "epoch": 0.7151919866444073, + "grad_norm": 0.44782402188976217, + "learning_rate": 8.111261827135016e-06, + "loss": 0.2705, + "step": 2142 + }, + { + "epoch": 0.7155258764607679, + "grad_norm": 0.5106897195007291, + "learning_rate": 8.108980375951034e-06, + "loss": 0.2901, + "step": 2143 + }, + { + "epoch": 0.7158597662771286, + "grad_norm": 0.4782478792333198, + "learning_rate": 8.10669786897339e-06, + "loss": 0.2913, + "step": 2144 + }, + { + "epoch": 0.7161936560934892, + "grad_norm": 0.48501675758503826, + "learning_rate": 8.104414306977213e-06, + "loss": 0.2911, + "step": 2145 + }, + { + "epoch": 0.7165275459098498, + "grad_norm": 0.47843734647797925, + "learning_rate": 8.102129690737988e-06, + "loss": 0.2915, + "step": 2146 + }, + { + "epoch": 0.7168614357262103, + "grad_norm": 0.48130751600535854, + "learning_rate": 8.099844021031559e-06, + "loss": 0.3013, + "step": 2147 + }, + { + "epoch": 0.7171953255425709, + "grad_norm": 0.5105289562342964, + "learning_rate": 8.09755729863413e-06, + "loss": 0.2962, + "step": 2148 + }, + { + "epoch": 0.7175292153589315, + "grad_norm": 0.5145254164444284, + "learning_rate": 8.095269524322256e-06, + "loss": 0.2855, + "step": 2149 + }, + { + "epoch": 0.7178631051752922, + "grad_norm": 0.47323685267089877, + "learning_rate": 8.092980698872855e-06, + "loss": 0.2864, + "step": 2150 + }, + { + "epoch": 0.7181969949916528, + "grad_norm": 0.4641197893555527, + "learning_rate": 8.0906908230632e-06, + "loss": 0.2831, + "step": 2151 + }, + { + "epoch": 0.7185308848080133, + "grad_norm": 0.46408420862374666, + "learning_rate": 8.088399897670922e-06, + "loss": 0.2864, + "step": 2152 + }, + { + "epoch": 0.7188647746243739, + "grad_norm": 0.4703258291785907, + "learning_rate": 8.086107923474008e-06, + "loss": 0.2811, + "step": 2153 + }, + { + "epoch": 0.7191986644407345, + "grad_norm": 0.48270936870209213, + "learning_rate": 8.0838149012508e-06, + "loss": 0.2728, + "step": 2154 + }, + { + "epoch": 0.7195325542570952, + "grad_norm": 0.47772056249058537, + "learning_rate": 8.081520831779995e-06, + "loss": 0.2772, + "step": 2155 + }, + { + "epoch": 0.7198664440734558, + "grad_norm": 0.47667364080774866, + "learning_rate": 8.079225715840646e-06, + "loss": 0.2836, + "step": 2156 + }, + { + "epoch": 0.7202003338898164, + "grad_norm": 0.5002991459458154, + "learning_rate": 8.076929554212166e-06, + "loss": 0.2878, + "step": 2157 + }, + { + "epoch": 0.7205342237061769, + "grad_norm": 0.4742677758244001, + "learning_rate": 8.07463234767432e-06, + "loss": 0.2818, + "step": 2158 + }, + { + "epoch": 0.7208681135225375, + "grad_norm": 0.4604194891293071, + "learning_rate": 8.072334097007224e-06, + "loss": 0.2755, + "step": 2159 + }, + { + "epoch": 0.7212020033388982, + "grad_norm": 0.4464109835419069, + "learning_rate": 8.070034802991356e-06, + "loss": 0.2786, + "step": 2160 + }, + { + "epoch": 0.7215358931552588, + "grad_norm": 0.45560952020570394, + "learning_rate": 8.067734466407539e-06, + "loss": 0.2926, + "step": 2161 + }, + { + "epoch": 0.7218697829716194, + "grad_norm": 0.46106495981043844, + "learning_rate": 8.06543308803696e-06, + "loss": 0.2924, + "step": 2162 + }, + { + "epoch": 0.7222036727879799, + "grad_norm": 0.4725005474501425, + "learning_rate": 8.063130668661154e-06, + "loss": 0.2949, + "step": 2163 + }, + { + "epoch": 0.7225375626043405, + "grad_norm": 0.4484225240398611, + "learning_rate": 8.06082720906201e-06, + "loss": 0.285, + "step": 2164 + }, + { + "epoch": 0.7228714524207012, + "grad_norm": 0.477261977743606, + "learning_rate": 8.058522710021773e-06, + "loss": 0.2897, + "step": 2165 + }, + { + "epoch": 0.7232053422370618, + "grad_norm": 0.47661974320136474, + "learning_rate": 8.056217172323035e-06, + "loss": 0.2874, + "step": 2166 + }, + { + "epoch": 0.7235392320534224, + "grad_norm": 0.4678415201609025, + "learning_rate": 8.053910596748748e-06, + "loss": 0.28, + "step": 2167 + }, + { + "epoch": 0.7238731218697829, + "grad_norm": 0.5095242997313502, + "learning_rate": 8.051602984082212e-06, + "loss": 0.2886, + "step": 2168 + }, + { + "epoch": 0.7242070116861435, + "grad_norm": 0.4582126020697294, + "learning_rate": 8.04929433510708e-06, + "loss": 0.2813, + "step": 2169 + }, + { + "epoch": 0.7245409015025042, + "grad_norm": 0.4642782335445266, + "learning_rate": 8.04698465060736e-06, + "loss": 0.2803, + "step": 2170 + }, + { + "epoch": 0.7248747913188648, + "grad_norm": 0.5165710174995551, + "learning_rate": 8.044673931367404e-06, + "loss": 0.2916, + "step": 2171 + }, + { + "epoch": 0.7252086811352254, + "grad_norm": 0.479702745414395, + "learning_rate": 8.042362178171926e-06, + "loss": 0.2871, + "step": 2172 + }, + { + "epoch": 0.725542570951586, + "grad_norm": 0.4746390021143665, + "learning_rate": 8.04004939180598e-06, + "loss": 0.2764, + "step": 2173 + }, + { + "epoch": 0.7258764607679465, + "grad_norm": 0.4654338736328087, + "learning_rate": 8.037735573054979e-06, + "loss": 0.2836, + "step": 2174 + }, + { + "epoch": 0.7262103505843072, + "grad_norm": 0.4828011938984545, + "learning_rate": 8.035420722704683e-06, + "loss": 0.2882, + "step": 2175 + }, + { + "epoch": 0.7265442404006678, + "grad_norm": 0.4753249898926171, + "learning_rate": 8.033104841541204e-06, + "loss": 0.302, + "step": 2176 + }, + { + "epoch": 0.7268781302170284, + "grad_norm": 0.4670183943557341, + "learning_rate": 8.030787930351005e-06, + "loss": 0.2816, + "step": 2177 + }, + { + "epoch": 0.727212020033389, + "grad_norm": 0.4750393870469586, + "learning_rate": 8.028469989920893e-06, + "loss": 0.2825, + "step": 2178 + }, + { + "epoch": 0.7275459098497495, + "grad_norm": 0.49769988844001595, + "learning_rate": 8.026151021038032e-06, + "loss": 0.2777, + "step": 2179 + }, + { + "epoch": 0.7278797996661102, + "grad_norm": 0.4889299669899723, + "learning_rate": 8.02383102448993e-06, + "loss": 0.2967, + "step": 2180 + }, + { + "epoch": 0.7282136894824708, + "grad_norm": 0.4873448114407203, + "learning_rate": 8.021510001064448e-06, + "loss": 0.2822, + "step": 2181 + }, + { + "epoch": 0.7285475792988314, + "grad_norm": 0.46372353302800057, + "learning_rate": 8.01918795154979e-06, + "loss": 0.2973, + "step": 2182 + }, + { + "epoch": 0.728881469115192, + "grad_norm": 0.45982312139139614, + "learning_rate": 8.016864876734514e-06, + "loss": 0.2815, + "step": 2183 + }, + { + "epoch": 0.7292153589315525, + "grad_norm": 0.5031261173423953, + "learning_rate": 8.014540777407525e-06, + "loss": 0.2917, + "step": 2184 + }, + { + "epoch": 0.7295492487479132, + "grad_norm": 0.4964757696693711, + "learning_rate": 8.012215654358073e-06, + "loss": 0.2956, + "step": 2185 + }, + { + "epoch": 0.7298831385642738, + "grad_norm": 0.4517334921123558, + "learning_rate": 8.00988950837576e-06, + "loss": 0.2791, + "step": 2186 + }, + { + "epoch": 0.7302170283806344, + "grad_norm": 0.4757848401665604, + "learning_rate": 8.007562340250533e-06, + "loss": 0.2894, + "step": 2187 + }, + { + "epoch": 0.730550918196995, + "grad_norm": 0.4892651343159623, + "learning_rate": 8.005234150772685e-06, + "loss": 0.2796, + "step": 2188 + }, + { + "epoch": 0.7308848080133556, + "grad_norm": 0.45560321365809997, + "learning_rate": 8.002904940732857e-06, + "loss": 0.2873, + "step": 2189 + }, + { + "epoch": 0.7312186978297162, + "grad_norm": 0.4875669568288951, + "learning_rate": 8.000574710922037e-06, + "loss": 0.2981, + "step": 2190 + }, + { + "epoch": 0.7315525876460768, + "grad_norm": 0.4608324960996222, + "learning_rate": 7.99824346213156e-06, + "loss": 0.2896, + "step": 2191 + }, + { + "epoch": 0.7318864774624374, + "grad_norm": 0.4674326539903456, + "learning_rate": 7.995911195153105e-06, + "loss": 0.2895, + "step": 2192 + }, + { + "epoch": 0.732220367278798, + "grad_norm": 0.45848347588317206, + "learning_rate": 7.993577910778698e-06, + "loss": 0.2859, + "step": 2193 + }, + { + "epoch": 0.7325542570951586, + "grad_norm": 0.4559688769983784, + "learning_rate": 7.991243609800711e-06, + "loss": 0.2846, + "step": 2194 + }, + { + "epoch": 0.7328881469115192, + "grad_norm": 0.44156251193872714, + "learning_rate": 7.98890829301186e-06, + "loss": 0.2788, + "step": 2195 + }, + { + "epoch": 0.7332220367278798, + "grad_norm": 0.46503795267053677, + "learning_rate": 7.986571961205206e-06, + "loss": 0.2991, + "step": 2196 + }, + { + "epoch": 0.7335559265442404, + "grad_norm": 0.4703739344065135, + "learning_rate": 7.984234615174156e-06, + "loss": 0.2858, + "step": 2197 + }, + { + "epoch": 0.733889816360601, + "grad_norm": 0.45880750053454183, + "learning_rate": 7.981896255712458e-06, + "loss": 0.2766, + "step": 2198 + }, + { + "epoch": 0.7342237061769616, + "grad_norm": 0.45647081745672285, + "learning_rate": 7.979556883614212e-06, + "loss": 0.2979, + "step": 2199 + }, + { + "epoch": 0.7345575959933222, + "grad_norm": 0.48415946528453485, + "learning_rate": 7.977216499673854e-06, + "loss": 0.2953, + "step": 2200 + }, + { + "epoch": 0.7348914858096828, + "grad_norm": 0.46162422710420303, + "learning_rate": 7.974875104686164e-06, + "loss": 0.2703, + "step": 2201 + }, + { + "epoch": 0.7352253756260434, + "grad_norm": 0.47227699931703004, + "learning_rate": 7.97253269944627e-06, + "loss": 0.302, + "step": 2202 + }, + { + "epoch": 0.735559265442404, + "grad_norm": 0.46814226576060475, + "learning_rate": 7.970189284749641e-06, + "loss": 0.3039, + "step": 2203 + }, + { + "epoch": 0.7358931552587646, + "grad_norm": 0.4636341624364872, + "learning_rate": 7.967844861392087e-06, + "loss": 0.2824, + "step": 2204 + }, + { + "epoch": 0.7362270450751253, + "grad_norm": 0.4404885902188228, + "learning_rate": 7.965499430169764e-06, + "loss": 0.29, + "step": 2205 + }, + { + "epoch": 0.7365609348914858, + "grad_norm": 0.45462877169212895, + "learning_rate": 7.963152991879167e-06, + "loss": 0.2893, + "step": 2206 + }, + { + "epoch": 0.7368948247078464, + "grad_norm": 0.47884173684725617, + "learning_rate": 7.960805547317134e-06, + "loss": 0.287, + "step": 2207 + }, + { + "epoch": 0.737228714524207, + "grad_norm": 0.47097659631192224, + "learning_rate": 7.958457097280845e-06, + "loss": 0.2885, + "step": 2208 + }, + { + "epoch": 0.7375626043405676, + "grad_norm": 0.46208294614085393, + "learning_rate": 7.956107642567824e-06, + "loss": 0.2729, + "step": 2209 + }, + { + "epoch": 0.7378964941569283, + "grad_norm": 0.4864542532377652, + "learning_rate": 7.95375718397593e-06, + "loss": 0.2962, + "step": 2210 + }, + { + "epoch": 0.7382303839732888, + "grad_norm": 0.47707724827697406, + "learning_rate": 7.95140572230337e-06, + "loss": 0.2937, + "step": 2211 + }, + { + "epoch": 0.7385642737896494, + "grad_norm": 0.4777669506765303, + "learning_rate": 7.949053258348684e-06, + "loss": 0.2855, + "step": 2212 + }, + { + "epoch": 0.73889816360601, + "grad_norm": 0.4588429051622455, + "learning_rate": 7.946699792910761e-06, + "loss": 0.2771, + "step": 2213 + }, + { + "epoch": 0.7392320534223706, + "grad_norm": 0.4763041681066591, + "learning_rate": 7.944345326788825e-06, + "loss": 0.2927, + "step": 2214 + }, + { + "epoch": 0.7395659432387313, + "grad_norm": 0.46276160639354585, + "learning_rate": 7.941989860782438e-06, + "loss": 0.2847, + "step": 2215 + }, + { + "epoch": 0.7398998330550918, + "grad_norm": 0.4724558661828109, + "learning_rate": 7.939633395691505e-06, + "loss": 0.2771, + "step": 2216 + }, + { + "epoch": 0.7402337228714524, + "grad_norm": 0.45708491802662565, + "learning_rate": 7.93727593231627e-06, + "loss": 0.2978, + "step": 2217 + }, + { + "epoch": 0.740567612687813, + "grad_norm": 0.5028792999712678, + "learning_rate": 7.934917471457316e-06, + "loss": 0.2898, + "step": 2218 + }, + { + "epoch": 0.7409015025041736, + "grad_norm": 0.4721455927571573, + "learning_rate": 7.932558013915561e-06, + "loss": 0.2876, + "step": 2219 + }, + { + "epoch": 0.7412353923205343, + "grad_norm": 0.4739033384298291, + "learning_rate": 7.93019756049227e-06, + "loss": 0.2807, + "step": 2220 + }, + { + "epoch": 0.7415692821368948, + "grad_norm": 0.4546208194052608, + "learning_rate": 7.927836111989035e-06, + "loss": 0.2919, + "step": 2221 + }, + { + "epoch": 0.7419031719532554, + "grad_norm": 0.48949778393788834, + "learning_rate": 7.925473669207797e-06, + "loss": 0.2889, + "step": 2222 + }, + { + "epoch": 0.742237061769616, + "grad_norm": 0.48279910513271707, + "learning_rate": 7.923110232950826e-06, + "loss": 0.2944, + "step": 2223 + }, + { + "epoch": 0.7425709515859766, + "grad_norm": 0.45741895799030446, + "learning_rate": 7.920745804020733e-06, + "loss": 0.2801, + "step": 2224 + }, + { + "epoch": 0.7429048414023373, + "grad_norm": 0.4451896979136234, + "learning_rate": 7.918380383220466e-06, + "loss": 0.2825, + "step": 2225 + }, + { + "epoch": 0.7432387312186979, + "grad_norm": 0.4843294234098612, + "learning_rate": 7.916013971353309e-06, + "loss": 0.292, + "step": 2226 + }, + { + "epoch": 0.7435726210350584, + "grad_norm": 0.4705858585386276, + "learning_rate": 7.913646569222885e-06, + "loss": 0.2965, + "step": 2227 + }, + { + "epoch": 0.743906510851419, + "grad_norm": 0.46004426606383036, + "learning_rate": 7.911278177633151e-06, + "loss": 0.2827, + "step": 2228 + }, + { + "epoch": 0.7442404006677796, + "grad_norm": 0.4582089710232431, + "learning_rate": 7.9089087973884e-06, + "loss": 0.2905, + "step": 2229 + }, + { + "epoch": 0.7445742904841403, + "grad_norm": 0.4683618004595708, + "learning_rate": 7.906538429293262e-06, + "loss": 0.2806, + "step": 2230 + }, + { + "epoch": 0.7449081803005009, + "grad_norm": 0.4892056124384463, + "learning_rate": 7.904167074152703e-06, + "loss": 0.2839, + "step": 2231 + }, + { + "epoch": 0.7452420701168614, + "grad_norm": 0.4657999962102446, + "learning_rate": 7.901794732772017e-06, + "loss": 0.2861, + "step": 2232 + }, + { + "epoch": 0.745575959933222, + "grad_norm": 0.4725659032324827, + "learning_rate": 7.899421405956847e-06, + "loss": 0.3021, + "step": 2233 + }, + { + "epoch": 0.7459098497495826, + "grad_norm": 0.4834477339278701, + "learning_rate": 7.897047094513157e-06, + "loss": 0.2848, + "step": 2234 + }, + { + "epoch": 0.7462437395659433, + "grad_norm": 0.4574465369298466, + "learning_rate": 7.894671799247251e-06, + "loss": 0.2897, + "step": 2235 + }, + { + "epoch": 0.7465776293823039, + "grad_norm": 0.5164342469542552, + "learning_rate": 7.89229552096577e-06, + "loss": 0.2764, + "step": 2236 + }, + { + "epoch": 0.7469115191986644, + "grad_norm": 0.4952207149837578, + "learning_rate": 7.889918260475685e-06, + "loss": 0.2739, + "step": 2237 + }, + { + "epoch": 0.747245409015025, + "grad_norm": 0.5004403971827338, + "learning_rate": 7.8875400185843e-06, + "loss": 0.286, + "step": 2238 + }, + { + "epoch": 0.7475792988313856, + "grad_norm": 0.48925600934800834, + "learning_rate": 7.885160796099254e-06, + "loss": 0.2974, + "step": 2239 + }, + { + "epoch": 0.7479131886477463, + "grad_norm": 0.5017618193359147, + "learning_rate": 7.88278059382852e-06, + "loss": 0.2922, + "step": 2240 + }, + { + "epoch": 0.7482470784641069, + "grad_norm": 0.483577467587952, + "learning_rate": 7.880399412580401e-06, + "loss": 0.2871, + "step": 2241 + }, + { + "epoch": 0.7485809682804675, + "grad_norm": 0.4494647836252735, + "learning_rate": 7.878017253163533e-06, + "loss": 0.2812, + "step": 2242 + }, + { + "epoch": 0.748914858096828, + "grad_norm": 0.4661261136170103, + "learning_rate": 7.875634116386888e-06, + "loss": 0.2797, + "step": 2243 + }, + { + "epoch": 0.7492487479131886, + "grad_norm": 0.46959546375557354, + "learning_rate": 7.873250003059765e-06, + "loss": 0.287, + "step": 2244 + }, + { + "epoch": 0.7495826377295493, + "grad_norm": 0.4643398842638502, + "learning_rate": 7.870864913991796e-06, + "loss": 0.2876, + "step": 2245 + }, + { + "epoch": 0.7499165275459099, + "grad_norm": 0.4808491211620409, + "learning_rate": 7.868478849992944e-06, + "loss": 0.2931, + "step": 2246 + }, + { + "epoch": 0.7502504173622705, + "grad_norm": 0.46357050878777384, + "learning_rate": 7.866091811873509e-06, + "loss": 0.2923, + "step": 2247 + }, + { + "epoch": 0.750584307178631, + "grad_norm": 0.48696669909591617, + "learning_rate": 7.863703800444111e-06, + "loss": 0.2933, + "step": 2248 + }, + { + "epoch": 0.7509181969949916, + "grad_norm": 0.4937283917184074, + "learning_rate": 7.86131481651571e-06, + "loss": 0.2905, + "step": 2249 + }, + { + "epoch": 0.7512520868113522, + "grad_norm": 0.47795243168537294, + "learning_rate": 7.85892486089959e-06, + "loss": 0.2984, + "step": 2250 + }, + { + "epoch": 0.7515859766277129, + "grad_norm": 0.45735824702309036, + "learning_rate": 7.856533934407371e-06, + "loss": 0.2878, + "step": 2251 + }, + { + "epoch": 0.7519198664440735, + "grad_norm": 0.5288530481571914, + "learning_rate": 7.854142037850995e-06, + "loss": 0.3045, + "step": 2252 + }, + { + "epoch": 0.752253756260434, + "grad_norm": 0.48024188246381655, + "learning_rate": 7.851749172042741e-06, + "loss": 0.2739, + "step": 2253 + }, + { + "epoch": 0.7525876460767946, + "grad_norm": 0.470670269279635, + "learning_rate": 7.849355337795213e-06, + "loss": 0.2849, + "step": 2254 + }, + { + "epoch": 0.7529215358931552, + "grad_norm": 0.478826041734232, + "learning_rate": 7.846960535921344e-06, + "loss": 0.2792, + "step": 2255 + }, + { + "epoch": 0.7532554257095159, + "grad_norm": 0.474983148888579, + "learning_rate": 7.844564767234398e-06, + "loss": 0.2886, + "step": 2256 + }, + { + "epoch": 0.7535893155258765, + "grad_norm": 0.4938106751879456, + "learning_rate": 7.842168032547963e-06, + "loss": 0.2898, + "step": 2257 + }, + { + "epoch": 0.7539232053422371, + "grad_norm": 0.47599761873005886, + "learning_rate": 7.83977033267596e-06, + "loss": 0.2794, + "step": 2258 + }, + { + "epoch": 0.7542570951585976, + "grad_norm": 0.45583565930090353, + "learning_rate": 7.837371668432635e-06, + "loss": 0.2879, + "step": 2259 + }, + { + "epoch": 0.7545909849749582, + "grad_norm": 0.4741870103301954, + "learning_rate": 7.83497204063256e-06, + "loss": 0.2759, + "step": 2260 + }, + { + "epoch": 0.7549248747913189, + "grad_norm": 0.4824821381590354, + "learning_rate": 7.832571450090639e-06, + "loss": 0.2792, + "step": 2261 + }, + { + "epoch": 0.7552587646076795, + "grad_norm": 0.47758705368655696, + "learning_rate": 7.8301698976221e-06, + "loss": 0.2953, + "step": 2262 + }, + { + "epoch": 0.7555926544240401, + "grad_norm": 0.447347102161142, + "learning_rate": 7.827767384042495e-06, + "loss": 0.2764, + "step": 2263 + }, + { + "epoch": 0.7559265442404006, + "grad_norm": 0.46462826654480216, + "learning_rate": 7.825363910167709e-06, + "loss": 0.2925, + "step": 2264 + }, + { + "epoch": 0.7562604340567612, + "grad_norm": 0.4764919644433483, + "learning_rate": 7.822959476813946e-06, + "loss": 0.2956, + "step": 2265 + }, + { + "epoch": 0.7565943238731219, + "grad_norm": 0.48578842567483477, + "learning_rate": 7.820554084797739e-06, + "loss": 0.2805, + "step": 2266 + }, + { + "epoch": 0.7569282136894825, + "grad_norm": 0.49968818548442007, + "learning_rate": 7.81814773493595e-06, + "loss": 0.2873, + "step": 2267 + }, + { + "epoch": 0.7572621035058431, + "grad_norm": 0.465610840016353, + "learning_rate": 7.81574042804576e-06, + "loss": 0.2947, + "step": 2268 + }, + { + "epoch": 0.7575959933222036, + "grad_norm": 0.44724117674211905, + "learning_rate": 7.81333216494468e-06, + "loss": 0.2761, + "step": 2269 + }, + { + "epoch": 0.7579298831385642, + "grad_norm": 0.4888497657207646, + "learning_rate": 7.810922946450543e-06, + "loss": 0.2939, + "step": 2270 + }, + { + "epoch": 0.7582637729549249, + "grad_norm": 0.47034566692725166, + "learning_rate": 7.808512773381506e-06, + "loss": 0.2866, + "step": 2271 + }, + { + "epoch": 0.7585976627712855, + "grad_norm": 0.4641542064187547, + "learning_rate": 7.806101646556052e-06, + "loss": 0.2841, + "step": 2272 + }, + { + "epoch": 0.7589315525876461, + "grad_norm": 0.509747448243513, + "learning_rate": 7.803689566792989e-06, + "loss": 0.2909, + "step": 2273 + }, + { + "epoch": 0.7592654424040067, + "grad_norm": 0.4788521692935791, + "learning_rate": 7.801276534911442e-06, + "loss": 0.2992, + "step": 2274 + }, + { + "epoch": 0.7595993322203672, + "grad_norm": 0.4378353469832804, + "learning_rate": 7.79886255173087e-06, + "loss": 0.2768, + "step": 2275 + }, + { + "epoch": 0.7599332220367279, + "grad_norm": 0.4704035488423027, + "learning_rate": 7.796447618071047e-06, + "loss": 0.2752, + "step": 2276 + }, + { + "epoch": 0.7602671118530885, + "grad_norm": 0.4404151447847845, + "learning_rate": 7.79403173475207e-06, + "loss": 0.2715, + "step": 2277 + }, + { + "epoch": 0.7606010016694491, + "grad_norm": 0.4630715177371395, + "learning_rate": 7.791614902594363e-06, + "loss": 0.2801, + "step": 2278 + }, + { + "epoch": 0.7609348914858097, + "grad_norm": 0.45462350518063255, + "learning_rate": 7.789197122418668e-06, + "loss": 0.2826, + "step": 2279 + }, + { + "epoch": 0.7612687813021702, + "grad_norm": 0.4589678483985069, + "learning_rate": 7.786778395046053e-06, + "loss": 0.2737, + "step": 2280 + }, + { + "epoch": 0.7616026711185309, + "grad_norm": 0.4625337786290638, + "learning_rate": 7.784358721297902e-06, + "loss": 0.2817, + "step": 2281 + }, + { + "epoch": 0.7619365609348915, + "grad_norm": 0.47904502204801114, + "learning_rate": 7.781938101995928e-06, + "loss": 0.2904, + "step": 2282 + }, + { + "epoch": 0.7622704507512521, + "grad_norm": 0.47828811131518933, + "learning_rate": 7.779516537962154e-06, + "loss": 0.2906, + "step": 2283 + }, + { + "epoch": 0.7626043405676127, + "grad_norm": 0.459877472757626, + "learning_rate": 7.777094030018936e-06, + "loss": 0.3034, + "step": 2284 + }, + { + "epoch": 0.7629382303839732, + "grad_norm": 0.4650503016931924, + "learning_rate": 7.774670578988946e-06, + "loss": 0.2688, + "step": 2285 + }, + { + "epoch": 0.7632721202003339, + "grad_norm": 0.4571466404061035, + "learning_rate": 7.772246185695174e-06, + "loss": 0.2806, + "step": 2286 + }, + { + "epoch": 0.7636060100166945, + "grad_norm": 0.44305936561607734, + "learning_rate": 7.769820850960928e-06, + "loss": 0.2935, + "step": 2287 + }, + { + "epoch": 0.7639398998330551, + "grad_norm": 0.45883171631436614, + "learning_rate": 7.767394575609843e-06, + "loss": 0.296, + "step": 2288 + }, + { + "epoch": 0.7642737896494157, + "grad_norm": 0.45762087006984653, + "learning_rate": 7.76496736046587e-06, + "loss": 0.2894, + "step": 2289 + }, + { + "epoch": 0.7646076794657763, + "grad_norm": 0.46710430203377074, + "learning_rate": 7.762539206353276e-06, + "loss": 0.2867, + "step": 2290 + }, + { + "epoch": 0.7649415692821369, + "grad_norm": 0.46143272647790645, + "learning_rate": 7.76011011409665e-06, + "loss": 0.2874, + "step": 2291 + }, + { + "epoch": 0.7652754590984975, + "grad_norm": 0.4600905803181871, + "learning_rate": 7.757680084520902e-06, + "loss": 0.2914, + "step": 2292 + }, + { + "epoch": 0.7656093489148581, + "grad_norm": 0.44551973684043206, + "learning_rate": 7.755249118451256e-06, + "loss": 0.2835, + "step": 2293 + }, + { + "epoch": 0.7659432387312187, + "grad_norm": 0.4430878302622179, + "learning_rate": 7.752817216713254e-06, + "loss": 0.2838, + "step": 2294 + }, + { + "epoch": 0.7662771285475793, + "grad_norm": 0.48002025053690284, + "learning_rate": 7.750384380132756e-06, + "loss": 0.2876, + "step": 2295 + }, + { + "epoch": 0.7666110183639399, + "grad_norm": 0.45214766929362354, + "learning_rate": 7.747950609535947e-06, + "loss": 0.2776, + "step": 2296 + }, + { + "epoch": 0.7669449081803005, + "grad_norm": 0.4722539462941386, + "learning_rate": 7.745515905749315e-06, + "loss": 0.2801, + "step": 2297 + }, + { + "epoch": 0.7672787979966611, + "grad_norm": 0.46970286977664993, + "learning_rate": 7.74308026959968e-06, + "loss": 0.2853, + "step": 2298 + }, + { + "epoch": 0.7676126878130217, + "grad_norm": 0.4869176189906157, + "learning_rate": 7.740643701914166e-06, + "loss": 0.2798, + "step": 2299 + }, + { + "epoch": 0.7679465776293823, + "grad_norm": 0.454102404261234, + "learning_rate": 7.738206203520223e-06, + "loss": 0.2804, + "step": 2300 + }, + { + "epoch": 0.7682804674457429, + "grad_norm": 0.4457089048257066, + "learning_rate": 7.735767775245608e-06, + "loss": 0.281, + "step": 2301 + }, + { + "epoch": 0.7686143572621035, + "grad_norm": 0.450755835938465, + "learning_rate": 7.733328417918405e-06, + "loss": 0.2798, + "step": 2302 + }, + { + "epoch": 0.7689482470784641, + "grad_norm": 0.4812672681261803, + "learning_rate": 7.730888132367e-06, + "loss": 0.2856, + "step": 2303 + }, + { + "epoch": 0.7692821368948247, + "grad_norm": 0.46022587602304543, + "learning_rate": 7.728446919420108e-06, + "loss": 0.2923, + "step": 2304 + }, + { + "epoch": 0.7696160267111853, + "grad_norm": 0.47390216313521316, + "learning_rate": 7.726004779906748e-06, + "loss": 0.2949, + "step": 2305 + }, + { + "epoch": 0.769949916527546, + "grad_norm": 0.47080042063384103, + "learning_rate": 7.723561714656258e-06, + "loss": 0.281, + "step": 2306 + }, + { + "epoch": 0.7702838063439065, + "grad_norm": 0.4524995027988834, + "learning_rate": 7.721117724498293e-06, + "loss": 0.2881, + "step": 2307 + }, + { + "epoch": 0.7706176961602671, + "grad_norm": 0.46523188050608644, + "learning_rate": 7.718672810262818e-06, + "loss": 0.2817, + "step": 2308 + }, + { + "epoch": 0.7709515859766277, + "grad_norm": 0.4498202806790611, + "learning_rate": 7.716226972780111e-06, + "loss": 0.2876, + "step": 2309 + }, + { + "epoch": 0.7712854757929883, + "grad_norm": 0.4751394328349239, + "learning_rate": 7.71378021288077e-06, + "loss": 0.2933, + "step": 2310 + }, + { + "epoch": 0.771619365609349, + "grad_norm": 0.5009226084279902, + "learning_rate": 7.7113325313957e-06, + "loss": 0.3052, + "step": 2311 + }, + { + "epoch": 0.7719532554257095, + "grad_norm": 0.46285349003125714, + "learning_rate": 7.70888392915612e-06, + "loss": 0.2878, + "step": 2312 + }, + { + "epoch": 0.7722871452420701, + "grad_norm": 0.45395938358338805, + "learning_rate": 7.706434406993566e-06, + "loss": 0.2831, + "step": 2313 + }, + { + "epoch": 0.7726210350584307, + "grad_norm": 0.4629279164874938, + "learning_rate": 7.703983965739878e-06, + "loss": 0.2778, + "step": 2314 + }, + { + "epoch": 0.7729549248747913, + "grad_norm": 0.48306700124547897, + "learning_rate": 7.701532606227216e-06, + "loss": 0.2994, + "step": 2315 + }, + { + "epoch": 0.773288814691152, + "grad_norm": 0.4420751387889217, + "learning_rate": 7.69908032928805e-06, + "loss": 0.278, + "step": 2316 + }, + { + "epoch": 0.7736227045075125, + "grad_norm": 0.4573150975333957, + "learning_rate": 7.696627135755159e-06, + "loss": 0.2768, + "step": 2317 + }, + { + "epoch": 0.7739565943238731, + "grad_norm": 0.47034477148797277, + "learning_rate": 7.694173026461634e-06, + "loss": 0.2875, + "step": 2318 + }, + { + "epoch": 0.7742904841402337, + "grad_norm": 0.475693073738245, + "learning_rate": 7.69171800224088e-06, + "loss": 0.2872, + "step": 2319 + }, + { + "epoch": 0.7746243739565943, + "grad_norm": 0.4754874822316868, + "learning_rate": 7.68926206392661e-06, + "loss": 0.2875, + "step": 2320 + }, + { + "epoch": 0.774958263772955, + "grad_norm": 0.47243518085269226, + "learning_rate": 7.686805212352848e-06, + "loss": 0.2825, + "step": 2321 + }, + { + "epoch": 0.7752921535893156, + "grad_norm": 0.46059621230878933, + "learning_rate": 7.684347448353926e-06, + "loss": 0.2866, + "step": 2322 + }, + { + "epoch": 0.7756260434056761, + "grad_norm": 0.4706092401660333, + "learning_rate": 7.681888772764493e-06, + "loss": 0.2951, + "step": 2323 + }, + { + "epoch": 0.7759599332220367, + "grad_norm": 0.49008280936226817, + "learning_rate": 7.679429186419496e-06, + "loss": 0.2894, + "step": 2324 + }, + { + "epoch": 0.7762938230383973, + "grad_norm": 0.4654657225429712, + "learning_rate": 7.6769686901542e-06, + "loss": 0.278, + "step": 2325 + }, + { + "epoch": 0.776627712854758, + "grad_norm": 0.4750391465837793, + "learning_rate": 7.67450728480418e-06, + "loss": 0.2845, + "step": 2326 + }, + { + "epoch": 0.7769616026711186, + "grad_norm": 0.46123967214047257, + "learning_rate": 7.672044971205315e-06, + "loss": 0.2869, + "step": 2327 + }, + { + "epoch": 0.7772954924874791, + "grad_norm": 0.46864494047920724, + "learning_rate": 7.669581750193791e-06, + "loss": 0.289, + "step": 2328 + }, + { + "epoch": 0.7776293823038397, + "grad_norm": 0.4768845998525581, + "learning_rate": 7.667117622606108e-06, + "loss": 0.2847, + "step": 2329 + }, + { + "epoch": 0.7779632721202003, + "grad_norm": 0.4747098983301631, + "learning_rate": 7.664652589279071e-06, + "loss": 0.2913, + "step": 2330 + }, + { + "epoch": 0.778297161936561, + "grad_norm": 0.4503108010468766, + "learning_rate": 7.66218665104979e-06, + "loss": 0.2844, + "step": 2331 + }, + { + "epoch": 0.7786310517529216, + "grad_norm": 0.4771482695823306, + "learning_rate": 7.65971980875569e-06, + "loss": 0.2868, + "step": 2332 + }, + { + "epoch": 0.7789649415692821, + "grad_norm": 0.48396120569583667, + "learning_rate": 7.657252063234493e-06, + "loss": 0.2918, + "step": 2333 + }, + { + "epoch": 0.7792988313856427, + "grad_norm": 0.4504325266288573, + "learning_rate": 7.654783415324234e-06, + "loss": 0.279, + "step": 2334 + }, + { + "epoch": 0.7796327212020033, + "grad_norm": 0.4607556918389292, + "learning_rate": 7.652313865863254e-06, + "loss": 0.2799, + "step": 2335 + }, + { + "epoch": 0.779966611018364, + "grad_norm": 0.4897860050196247, + "learning_rate": 7.649843415690198e-06, + "loss": 0.2831, + "step": 2336 + }, + { + "epoch": 0.7803005008347246, + "grad_norm": 0.4657281197888195, + "learning_rate": 7.647372065644019e-06, + "loss": 0.2875, + "step": 2337 + }, + { + "epoch": 0.7806343906510852, + "grad_norm": 0.4945477707424227, + "learning_rate": 7.644899816563975e-06, + "loss": 0.2923, + "step": 2338 + }, + { + "epoch": 0.7809682804674457, + "grad_norm": 0.43625429736837423, + "learning_rate": 7.642426669289627e-06, + "loss": 0.2779, + "step": 2339 + }, + { + "epoch": 0.7813021702838063, + "grad_norm": 0.4516424366329817, + "learning_rate": 7.639952624660847e-06, + "loss": 0.2997, + "step": 2340 + }, + { + "epoch": 0.781636060100167, + "grad_norm": 0.44931623306382446, + "learning_rate": 7.637477683517802e-06, + "loss": 0.2827, + "step": 2341 + }, + { + "epoch": 0.7819699499165276, + "grad_norm": 0.4749718888668058, + "learning_rate": 7.635001846700973e-06, + "loss": 0.286, + "step": 2342 + }, + { + "epoch": 0.7823038397328882, + "grad_norm": 0.4811848187541129, + "learning_rate": 7.632525115051142e-06, + "loss": 0.2876, + "step": 2343 + }, + { + "epoch": 0.7826377295492487, + "grad_norm": 0.45002724196044913, + "learning_rate": 7.630047489409393e-06, + "loss": 0.2811, + "step": 2344 + }, + { + "epoch": 0.7829716193656093, + "grad_norm": 0.44949391521088616, + "learning_rate": 7.627568970617114e-06, + "loss": 0.2863, + "step": 2345 + }, + { + "epoch": 0.78330550918197, + "grad_norm": 0.4404687745971873, + "learning_rate": 7.625089559515999e-06, + "loss": 0.2815, + "step": 2346 + }, + { + "epoch": 0.7836393989983306, + "grad_norm": 0.4645847628636068, + "learning_rate": 7.622609256948043e-06, + "loss": 0.291, + "step": 2347 + }, + { + "epoch": 0.7839732888146912, + "grad_norm": 0.4885880731768247, + "learning_rate": 7.620128063755541e-06, + "loss": 0.294, + "step": 2348 + }, + { + "epoch": 0.7843071786310517, + "grad_norm": 0.47563889162145145, + "learning_rate": 7.617645980781097e-06, + "loss": 0.2927, + "step": 2349 + }, + { + "epoch": 0.7846410684474123, + "grad_norm": 0.4947211410340266, + "learning_rate": 7.6151630088676125e-06, + "loss": 0.2959, + "step": 2350 + }, + { + "epoch": 0.784974958263773, + "grad_norm": 0.4414392619587441, + "learning_rate": 7.612679148858291e-06, + "loss": 0.287, + "step": 2351 + }, + { + "epoch": 0.7853088480801336, + "grad_norm": 0.4496243451528003, + "learning_rate": 7.610194401596638e-06, + "loss": 0.2933, + "step": 2352 + }, + { + "epoch": 0.7856427378964942, + "grad_norm": 0.4962094709245855, + "learning_rate": 7.607708767926463e-06, + "loss": 0.2941, + "step": 2353 + }, + { + "epoch": 0.7859766277128547, + "grad_norm": 0.47289010078700255, + "learning_rate": 7.6052222486918725e-06, + "loss": 0.2908, + "step": 2354 + }, + { + "epoch": 0.7863105175292153, + "grad_norm": 0.4607679514657715, + "learning_rate": 7.602734844737278e-06, + "loss": 0.2779, + "step": 2355 + }, + { + "epoch": 0.786644407345576, + "grad_norm": 0.493726800024575, + "learning_rate": 7.600246556907386e-06, + "loss": 0.2836, + "step": 2356 + }, + { + "epoch": 0.7869782971619366, + "grad_norm": 0.5036697553542113, + "learning_rate": 7.5977573860472065e-06, + "loss": 0.2697, + "step": 2357 + }, + { + "epoch": 0.7873121869782972, + "grad_norm": 0.49562661690283377, + "learning_rate": 7.595267333002052e-06, + "loss": 0.2804, + "step": 2358 + }, + { + "epoch": 0.7876460767946578, + "grad_norm": 0.4714049710237643, + "learning_rate": 7.592776398617528e-06, + "loss": 0.2847, + "step": 2359 + }, + { + "epoch": 0.7879799666110183, + "grad_norm": 0.4578953486883747, + "learning_rate": 7.590284583739547e-06, + "loss": 0.2772, + "step": 2360 + }, + { + "epoch": 0.788313856427379, + "grad_norm": 0.4889019599222521, + "learning_rate": 7.587791889214312e-06, + "loss": 0.2814, + "step": 2361 + }, + { + "epoch": 0.7886477462437396, + "grad_norm": 0.49322532629441684, + "learning_rate": 7.585298315888332e-06, + "loss": 0.2816, + "step": 2362 + }, + { + "epoch": 0.7889816360601002, + "grad_norm": 0.5002392529010802, + "learning_rate": 7.582803864608411e-06, + "loss": 0.2933, + "step": 2363 + }, + { + "epoch": 0.7893155258764608, + "grad_norm": 0.4791209343802673, + "learning_rate": 7.580308536221651e-06, + "loss": 0.2699, + "step": 2364 + }, + { + "epoch": 0.7896494156928213, + "grad_norm": 0.47731941692419355, + "learning_rate": 7.5778123315754535e-06, + "loss": 0.2949, + "step": 2365 + }, + { + "epoch": 0.7899833055091819, + "grad_norm": 0.513558395541488, + "learning_rate": 7.5753152515175165e-06, + "loss": 0.2944, + "step": 2366 + }, + { + "epoch": 0.7903171953255426, + "grad_norm": 0.445638866386001, + "learning_rate": 7.572817296895835e-06, + "loss": 0.2729, + "step": 2367 + }, + { + "epoch": 0.7906510851419032, + "grad_norm": 0.45368669950970364, + "learning_rate": 7.570318468558704e-06, + "loss": 0.2919, + "step": 2368 + }, + { + "epoch": 0.7909849749582638, + "grad_norm": 0.4406942516084294, + "learning_rate": 7.567818767354708e-06, + "loss": 0.2678, + "step": 2369 + }, + { + "epoch": 0.7913188647746243, + "grad_norm": 0.449785990394564, + "learning_rate": 7.565318194132736e-06, + "loss": 0.2837, + "step": 2370 + }, + { + "epoch": 0.7916527545909849, + "grad_norm": 0.4474613583721199, + "learning_rate": 7.56281674974197e-06, + "loss": 0.2878, + "step": 2371 + }, + { + "epoch": 0.7919866444073456, + "grad_norm": 0.46606632217260063, + "learning_rate": 7.560314435031886e-06, + "loss": 0.2934, + "step": 2372 + }, + { + "epoch": 0.7923205342237062, + "grad_norm": 0.4461975689841608, + "learning_rate": 7.557811250852256e-06, + "loss": 0.2925, + "step": 2373 + }, + { + "epoch": 0.7926544240400668, + "grad_norm": 0.4488774077456846, + "learning_rate": 7.5553071980531526e-06, + "loss": 0.2741, + "step": 2374 + }, + { + "epoch": 0.7929883138564274, + "grad_norm": 0.44992915107448106, + "learning_rate": 7.5528022774849344e-06, + "loss": 0.2845, + "step": 2375 + }, + { + "epoch": 0.7933222036727879, + "grad_norm": 0.4586470369954414, + "learning_rate": 7.550296489998264e-06, + "loss": 0.2837, + "step": 2376 + }, + { + "epoch": 0.7936560934891486, + "grad_norm": 0.45838943947312705, + "learning_rate": 7.547789836444088e-06, + "loss": 0.2707, + "step": 2377 + }, + { + "epoch": 0.7939899833055092, + "grad_norm": 0.45272432780195454, + "learning_rate": 7.54528231767366e-06, + "loss": 0.2728, + "step": 2378 + }, + { + "epoch": 0.7943238731218698, + "grad_norm": 0.47416592203594327, + "learning_rate": 7.542773934538515e-06, + "loss": 0.2852, + "step": 2379 + }, + { + "epoch": 0.7946577629382304, + "grad_norm": 0.4577876011080357, + "learning_rate": 7.540264687890489e-06, + "loss": 0.2779, + "step": 2380 + }, + { + "epoch": 0.7949916527545909, + "grad_norm": 0.46673106073653386, + "learning_rate": 7.537754578581711e-06, + "loss": 0.2867, + "step": 2381 + }, + { + "epoch": 0.7953255425709516, + "grad_norm": 0.4570130832078113, + "learning_rate": 7.535243607464597e-06, + "loss": 0.281, + "step": 2382 + }, + { + "epoch": 0.7956594323873122, + "grad_norm": 0.4836369913794118, + "learning_rate": 7.532731775391864e-06, + "loss": 0.2902, + "step": 2383 + }, + { + "epoch": 0.7959933222036728, + "grad_norm": 0.44791406832810193, + "learning_rate": 7.530219083216512e-06, + "loss": 0.2873, + "step": 2384 + }, + { + "epoch": 0.7963272120200334, + "grad_norm": 0.43421706212423383, + "learning_rate": 7.527705531791844e-06, + "loss": 0.2793, + "step": 2385 + }, + { + "epoch": 0.7966611018363939, + "grad_norm": 0.4645657979402405, + "learning_rate": 7.525191121971447e-06, + "loss": 0.2917, + "step": 2386 + }, + { + "epoch": 0.7969949916527546, + "grad_norm": 0.48756368972548564, + "learning_rate": 7.5226758546092e-06, + "loss": 0.2831, + "step": 2387 + }, + { + "epoch": 0.7973288814691152, + "grad_norm": 0.4712840713885653, + "learning_rate": 7.520159730559278e-06, + "loss": 0.2802, + "step": 2388 + }, + { + "epoch": 0.7976627712854758, + "grad_norm": 0.46577080761627726, + "learning_rate": 7.51764275067614e-06, + "loss": 0.2932, + "step": 2389 + }, + { + "epoch": 0.7979966611018364, + "grad_norm": 0.448353706861509, + "learning_rate": 7.51512491581454e-06, + "loss": 0.2849, + "step": 2390 + }, + { + "epoch": 0.798330550918197, + "grad_norm": 0.48017314508233405, + "learning_rate": 7.512606226829524e-06, + "loss": 0.2946, + "step": 2391 + }, + { + "epoch": 0.7986644407345576, + "grad_norm": 0.47823364200254714, + "learning_rate": 7.510086684576425e-06, + "loss": 0.2889, + "step": 2392 + }, + { + "epoch": 0.7989983305509182, + "grad_norm": 0.47662757362594593, + "learning_rate": 7.5075662899108636e-06, + "loss": 0.2802, + "step": 2393 + }, + { + "epoch": 0.7993322203672788, + "grad_norm": 0.46391185609195396, + "learning_rate": 7.505045043688758e-06, + "loss": 0.2868, + "step": 2394 + }, + { + "epoch": 0.7996661101836394, + "grad_norm": 0.49982352737696834, + "learning_rate": 7.502522946766306e-06, + "loss": 0.3009, + "step": 2395 + }, + { + "epoch": 0.8, + "grad_norm": 0.46699647775147934, + "learning_rate": 7.500000000000001e-06, + "loss": 0.2813, + "step": 2396 + }, + { + "epoch": 0.8003338898163606, + "grad_norm": 0.5097749312256774, + "learning_rate": 7.497476204246621e-06, + "loss": 0.3023, + "step": 2397 + }, + { + "epoch": 0.8006677796327212, + "grad_norm": 0.4640214017100237, + "learning_rate": 7.494951560363235e-06, + "loss": 0.2858, + "step": 2398 + }, + { + "epoch": 0.8010016694490818, + "grad_norm": 0.5011585521381812, + "learning_rate": 7.4924260692072e-06, + "loss": 0.2886, + "step": 2399 + }, + { + "epoch": 0.8013355592654424, + "grad_norm": 0.4532095030437002, + "learning_rate": 7.489899731636159e-06, + "loss": 0.2829, + "step": 2400 + }, + { + "epoch": 0.801669449081803, + "grad_norm": 0.45907576134891404, + "learning_rate": 7.487372548508041e-06, + "loss": 0.2866, + "step": 2401 + }, + { + "epoch": 0.8020033388981636, + "grad_norm": 0.49431326851727597, + "learning_rate": 7.484844520681066e-06, + "loss": 0.2833, + "step": 2402 + }, + { + "epoch": 0.8023372287145242, + "grad_norm": 0.5029577274692207, + "learning_rate": 7.482315649013741e-06, + "loss": 0.2893, + "step": 2403 + }, + { + "epoch": 0.8026711185308848, + "grad_norm": 0.4560755504617047, + "learning_rate": 7.479785934364856e-06, + "loss": 0.2883, + "step": 2404 + }, + { + "epoch": 0.8030050083472454, + "grad_norm": 0.4709808779133249, + "learning_rate": 7.477255377593489e-06, + "loss": 0.2795, + "step": 2405 + }, + { + "epoch": 0.803338898163606, + "grad_norm": 0.4940154035386246, + "learning_rate": 7.474723979559003e-06, + "loss": 0.2902, + "step": 2406 + }, + { + "epoch": 0.8036727879799667, + "grad_norm": 0.4659214278768805, + "learning_rate": 7.4721917411210484e-06, + "loss": 0.2911, + "step": 2407 + }, + { + "epoch": 0.8040066777963272, + "grad_norm": 0.4599319455960184, + "learning_rate": 7.4696586631395626e-06, + "loss": 0.2839, + "step": 2408 + }, + { + "epoch": 0.8043405676126878, + "grad_norm": 0.4591012840567587, + "learning_rate": 7.4671247464747635e-06, + "loss": 0.2882, + "step": 2409 + }, + { + "epoch": 0.8046744574290484, + "grad_norm": 0.47736316948432916, + "learning_rate": 7.4645899919871545e-06, + "loss": 0.2738, + "step": 2410 + }, + { + "epoch": 0.805008347245409, + "grad_norm": 0.5026367652032746, + "learning_rate": 7.462054400537529e-06, + "loss": 0.2795, + "step": 2411 + }, + { + "epoch": 0.8053422370617697, + "grad_norm": 0.4479662027511734, + "learning_rate": 7.4595179729869584e-06, + "loss": 0.2866, + "step": 2412 + }, + { + "epoch": 0.8056761268781302, + "grad_norm": 0.4644836022211116, + "learning_rate": 7.456980710196801e-06, + "loss": 0.2954, + "step": 2413 + }, + { + "epoch": 0.8060100166944908, + "grad_norm": 0.4961413723314931, + "learning_rate": 7.454442613028696e-06, + "loss": 0.28, + "step": 2414 + }, + { + "epoch": 0.8063439065108514, + "grad_norm": 0.5000469326528832, + "learning_rate": 7.451903682344572e-06, + "loss": 0.2984, + "step": 2415 + }, + { + "epoch": 0.806677796327212, + "grad_norm": 0.48875603300640363, + "learning_rate": 7.449363919006635e-06, + "loss": 0.278, + "step": 2416 + }, + { + "epoch": 0.8070116861435727, + "grad_norm": 0.47032521568357866, + "learning_rate": 7.446823323877375e-06, + "loss": 0.2955, + "step": 2417 + }, + { + "epoch": 0.8073455759599332, + "grad_norm": 0.4820853356597098, + "learning_rate": 7.444281897819563e-06, + "loss": 0.2957, + "step": 2418 + }, + { + "epoch": 0.8076794657762938, + "grad_norm": 0.44723168716310735, + "learning_rate": 7.44173964169626e-06, + "loss": 0.2922, + "step": 2419 + }, + { + "epoch": 0.8080133555926544, + "grad_norm": 0.46169475243326186, + "learning_rate": 7.439196556370796e-06, + "loss": 0.3096, + "step": 2420 + }, + { + "epoch": 0.808347245409015, + "grad_norm": 0.44420439488083274, + "learning_rate": 7.436652642706795e-06, + "loss": 0.2873, + "step": 2421 + }, + { + "epoch": 0.8086811352253757, + "grad_norm": 0.45244988682846743, + "learning_rate": 7.434107901568155e-06, + "loss": 0.2723, + "step": 2422 + }, + { + "epoch": 0.8090150250417363, + "grad_norm": 0.4955591266505262, + "learning_rate": 7.431562333819054e-06, + "loss": 0.2985, + "step": 2423 + }, + { + "epoch": 0.8093489148580968, + "grad_norm": 0.45276770891504997, + "learning_rate": 7.429015940323961e-06, + "loss": 0.2828, + "step": 2424 + }, + { + "epoch": 0.8096828046744574, + "grad_norm": 0.47539722313380134, + "learning_rate": 7.426468721947611e-06, + "loss": 0.2864, + "step": 2425 + }, + { + "epoch": 0.810016694490818, + "grad_norm": 0.4497227837598779, + "learning_rate": 7.423920679555029e-06, + "loss": 0.2668, + "step": 2426 + }, + { + "epoch": 0.8103505843071787, + "grad_norm": 0.47230640195452905, + "learning_rate": 7.421371814011516e-06, + "loss": 0.2943, + "step": 2427 + }, + { + "epoch": 0.8106844741235393, + "grad_norm": 0.44943444629717544, + "learning_rate": 7.418822126182655e-06, + "loss": 0.278, + "step": 2428 + }, + { + "epoch": 0.8110183639398998, + "grad_norm": 0.4777899827100713, + "learning_rate": 7.416271616934308e-06, + "loss": 0.2921, + "step": 2429 + }, + { + "epoch": 0.8113522537562604, + "grad_norm": 0.4606159204609906, + "learning_rate": 7.41372028713261e-06, + "loss": 0.2843, + "step": 2430 + }, + { + "epoch": 0.811686143572621, + "grad_norm": 0.44767350689135305, + "learning_rate": 7.4111681376439824e-06, + "loss": 0.2756, + "step": 2431 + }, + { + "epoch": 0.8120200333889817, + "grad_norm": 0.45602009856944464, + "learning_rate": 7.4086151693351235e-06, + "loss": 0.2774, + "step": 2432 + }, + { + "epoch": 0.8123539232053423, + "grad_norm": 0.4596264472511512, + "learning_rate": 7.406061383073007e-06, + "loss": 0.2838, + "step": 2433 + }, + { + "epoch": 0.8126878130217028, + "grad_norm": 0.45252992033319434, + "learning_rate": 7.403506779724883e-06, + "loss": 0.2805, + "step": 2434 + }, + { + "epoch": 0.8130217028380634, + "grad_norm": 0.45219562603891034, + "learning_rate": 7.400951360158285e-06, + "loss": 0.2828, + "step": 2435 + }, + { + "epoch": 0.813355592654424, + "grad_norm": 0.4646976436726408, + "learning_rate": 7.398395125241019e-06, + "loss": 0.2946, + "step": 2436 + }, + { + "epoch": 0.8136894824707847, + "grad_norm": 0.4647360850085069, + "learning_rate": 7.395838075841169e-06, + "loss": 0.2846, + "step": 2437 + }, + { + "epoch": 0.8140233722871453, + "grad_norm": 0.45132547138196577, + "learning_rate": 7.393280212827095e-06, + "loss": 0.2759, + "step": 2438 + }, + { + "epoch": 0.8143572621035059, + "grad_norm": 0.4702420511981817, + "learning_rate": 7.390721537067436e-06, + "loss": 0.2912, + "step": 2439 + }, + { + "epoch": 0.8146911519198664, + "grad_norm": 0.45550837758989504, + "learning_rate": 7.388162049431102e-06, + "loss": 0.281, + "step": 2440 + }, + { + "epoch": 0.815025041736227, + "grad_norm": 0.4847015833556582, + "learning_rate": 7.385601750787285e-06, + "loss": 0.3025, + "step": 2441 + }, + { + "epoch": 0.8153589315525877, + "grad_norm": 0.47992164034785983, + "learning_rate": 7.383040642005448e-06, + "loss": 0.2921, + "step": 2442 + }, + { + "epoch": 0.8156928213689483, + "grad_norm": 0.45099540463459165, + "learning_rate": 7.38047872395533e-06, + "loss": 0.2913, + "step": 2443 + }, + { + "epoch": 0.8160267111853089, + "grad_norm": 0.4402318611592542, + "learning_rate": 7.377915997506945e-06, + "loss": 0.2802, + "step": 2444 + }, + { + "epoch": 0.8163606010016694, + "grad_norm": 0.4751662476254857, + "learning_rate": 7.375352463530583e-06, + "loss": 0.2862, + "step": 2445 + }, + { + "epoch": 0.81669449081803, + "grad_norm": 0.48785281880994485, + "learning_rate": 7.3727881228968045e-06, + "loss": 0.283, + "step": 2446 + }, + { + "epoch": 0.8170283806343906, + "grad_norm": 0.4624621575236823, + "learning_rate": 7.370222976476449e-06, + "loss": 0.2944, + "step": 2447 + }, + { + "epoch": 0.8173622704507513, + "grad_norm": 0.44821662510958615, + "learning_rate": 7.367657025140623e-06, + "loss": 0.2887, + "step": 2448 + }, + { + "epoch": 0.8176961602671119, + "grad_norm": 0.47392605551461614, + "learning_rate": 7.365090269760715e-06, + "loss": 0.2993, + "step": 2449 + }, + { + "epoch": 0.8180300500834724, + "grad_norm": 0.4728395842604744, + "learning_rate": 7.3625227112083796e-06, + "loss": 0.2953, + "step": 2450 + }, + { + "epoch": 0.818363939899833, + "grad_norm": 0.45559781450205256, + "learning_rate": 7.359954350355547e-06, + "loss": 0.2873, + "step": 2451 + }, + { + "epoch": 0.8186978297161936, + "grad_norm": 0.4491610617379849, + "learning_rate": 7.357385188074419e-06, + "loss": 0.2996, + "step": 2452 + }, + { + "epoch": 0.8190317195325543, + "grad_norm": 0.4306978617873916, + "learning_rate": 7.354815225237468e-06, + "loss": 0.2819, + "step": 2453 + }, + { + "epoch": 0.8193656093489149, + "grad_norm": 0.4437469946929909, + "learning_rate": 7.352244462717444e-06, + "loss": 0.2866, + "step": 2454 + }, + { + "epoch": 0.8196994991652755, + "grad_norm": 0.4715591794093596, + "learning_rate": 7.349672901387362e-06, + "loss": 0.2806, + "step": 2455 + }, + { + "epoch": 0.820033388981636, + "grad_norm": 0.4642645164489163, + "learning_rate": 7.34710054212051e-06, + "loss": 0.2806, + "step": 2456 + }, + { + "epoch": 0.8203672787979966, + "grad_norm": 0.4470470727850028, + "learning_rate": 7.344527385790449e-06, + "loss": 0.2936, + "step": 2457 + }, + { + "epoch": 0.8207011686143573, + "grad_norm": 0.46303358487382557, + "learning_rate": 7.341953433271011e-06, + "loss": 0.2952, + "step": 2458 + }, + { + "epoch": 0.8210350584307179, + "grad_norm": 0.45532084601549744, + "learning_rate": 7.339378685436296e-06, + "loss": 0.2967, + "step": 2459 + }, + { + "epoch": 0.8213689482470785, + "grad_norm": 0.47750964143400465, + "learning_rate": 7.336803143160674e-06, + "loss": 0.2811, + "step": 2460 + }, + { + "epoch": 0.821702838063439, + "grad_norm": 0.4420435732341081, + "learning_rate": 7.334226807318787e-06, + "loss": 0.2822, + "step": 2461 + }, + { + "epoch": 0.8220367278797996, + "grad_norm": 0.4609017793229579, + "learning_rate": 7.331649678785545e-06, + "loss": 0.3014, + "step": 2462 + }, + { + "epoch": 0.8223706176961603, + "grad_norm": 0.4590845802163479, + "learning_rate": 7.329071758436128e-06, + "loss": 0.2818, + "step": 2463 + }, + { + "epoch": 0.8227045075125209, + "grad_norm": 0.4708472664571321, + "learning_rate": 7.3264930471459835e-06, + "loss": 0.2944, + "step": 2464 + }, + { + "epoch": 0.8230383973288815, + "grad_norm": 0.45827716196903656, + "learning_rate": 7.3239135457908295e-06, + "loss": 0.2894, + "step": 2465 + }, + { + "epoch": 0.823372287145242, + "grad_norm": 0.47432921815762996, + "learning_rate": 7.32133325524665e-06, + "loss": 0.2773, + "step": 2466 + }, + { + "epoch": 0.8237061769616026, + "grad_norm": 0.4776190578708721, + "learning_rate": 7.318752176389701e-06, + "loss": 0.2998, + "step": 2467 + }, + { + "epoch": 0.8240400667779633, + "grad_norm": 0.455418859314778, + "learning_rate": 7.3161703100964995e-06, + "loss": 0.2762, + "step": 2468 + }, + { + "epoch": 0.8243739565943239, + "grad_norm": 0.46531625523740067, + "learning_rate": 7.313587657243837e-06, + "loss": 0.2899, + "step": 2469 + }, + { + "epoch": 0.8247078464106845, + "grad_norm": 0.4861318033399232, + "learning_rate": 7.311004218708769e-06, + "loss": 0.2905, + "step": 2470 + }, + { + "epoch": 0.8250417362270451, + "grad_norm": 0.47900898334920866, + "learning_rate": 7.308419995368616e-06, + "loss": 0.2985, + "step": 2471 + }, + { + "epoch": 0.8253756260434056, + "grad_norm": 0.4839354611707689, + "learning_rate": 7.30583498810097e-06, + "loss": 0.2899, + "step": 2472 + }, + { + "epoch": 0.8257095158597663, + "grad_norm": 0.45644822255594153, + "learning_rate": 7.303249197783682e-06, + "loss": 0.2845, + "step": 2473 + }, + { + "epoch": 0.8260434056761269, + "grad_norm": 0.4779868287437593, + "learning_rate": 7.300662625294874e-06, + "loss": 0.2938, + "step": 2474 + }, + { + "epoch": 0.8263772954924875, + "grad_norm": 0.4561313649035521, + "learning_rate": 7.2980752715129354e-06, + "loss": 0.2855, + "step": 2475 + }, + { + "epoch": 0.8267111853088481, + "grad_norm": 0.5009475038763934, + "learning_rate": 7.295487137316517e-06, + "loss": 0.2867, + "step": 2476 + }, + { + "epoch": 0.8270450751252086, + "grad_norm": 0.4636273555632232, + "learning_rate": 7.292898223584535e-06, + "loss": 0.2936, + "step": 2477 + }, + { + "epoch": 0.8273789649415693, + "grad_norm": 0.4495437493629532, + "learning_rate": 7.290308531196171e-06, + "loss": 0.2877, + "step": 2478 + }, + { + "epoch": 0.8277128547579299, + "grad_norm": 0.44253899599144203, + "learning_rate": 7.287718061030872e-06, + "loss": 0.2823, + "step": 2479 + }, + { + "epoch": 0.8280467445742905, + "grad_norm": 0.4870146965592383, + "learning_rate": 7.285126813968346e-06, + "loss": 0.2887, + "step": 2480 + }, + { + "epoch": 0.8283806343906511, + "grad_norm": 0.46676137182815924, + "learning_rate": 7.282534790888571e-06, + "loss": 0.2867, + "step": 2481 + }, + { + "epoch": 0.8287145242070116, + "grad_norm": 0.4357934785752925, + "learning_rate": 7.27994199267178e-06, + "loss": 0.2795, + "step": 2482 + }, + { + "epoch": 0.8290484140233723, + "grad_norm": 0.4402917516259569, + "learning_rate": 7.27734842019848e-06, + "loss": 0.288, + "step": 2483 + }, + { + "epoch": 0.8293823038397329, + "grad_norm": 0.4578470835158642, + "learning_rate": 7.274754074349429e-06, + "loss": 0.2816, + "step": 2484 + }, + { + "epoch": 0.8297161936560935, + "grad_norm": 0.43906086015115714, + "learning_rate": 7.2721589560056575e-06, + "loss": 0.2841, + "step": 2485 + }, + { + "epoch": 0.8300500834724541, + "grad_norm": 0.45791194154204556, + "learning_rate": 7.269563066048451e-06, + "loss": 0.2708, + "step": 2486 + }, + { + "epoch": 0.8303839732888146, + "grad_norm": 0.48036399937414515, + "learning_rate": 7.266966405359361e-06, + "loss": 0.2866, + "step": 2487 + }, + { + "epoch": 0.8307178631051753, + "grad_norm": 0.45678735893894223, + "learning_rate": 7.264368974820202e-06, + "loss": 0.2868, + "step": 2488 + }, + { + "epoch": 0.8310517529215359, + "grad_norm": 0.447973055428677, + "learning_rate": 7.2617707753130465e-06, + "loss": 0.2835, + "step": 2489 + }, + { + "epoch": 0.8313856427378965, + "grad_norm": 0.47800716169931573, + "learning_rate": 7.259171807720231e-06, + "loss": 0.2928, + "step": 2490 + }, + { + "epoch": 0.8317195325542571, + "grad_norm": 0.4458925630564376, + "learning_rate": 7.256572072924347e-06, + "loss": 0.2868, + "step": 2491 + }, + { + "epoch": 0.8320534223706177, + "grad_norm": 0.451155642294844, + "learning_rate": 7.253971571808258e-06, + "loss": 0.2841, + "step": 2492 + }, + { + "epoch": 0.8323873121869783, + "grad_norm": 0.4404846847411471, + "learning_rate": 7.2513703052550746e-06, + "loss": 0.2846, + "step": 2493 + }, + { + "epoch": 0.8327212020033389, + "grad_norm": 0.4683161061328168, + "learning_rate": 7.248768274148178e-06, + "loss": 0.2875, + "step": 2494 + }, + { + "epoch": 0.8330550918196995, + "grad_norm": 0.4587094399172571, + "learning_rate": 7.246165479371201e-06, + "loss": 0.2844, + "step": 2495 + }, + { + "epoch": 0.8333889816360601, + "grad_norm": 0.4502865514427051, + "learning_rate": 7.243561921808041e-06, + "loss": 0.278, + "step": 2496 + }, + { + "epoch": 0.8337228714524207, + "grad_norm": 0.45494954498701035, + "learning_rate": 7.240957602342853e-06, + "loss": 0.2815, + "step": 2497 + }, + { + "epoch": 0.8340567612687813, + "grad_norm": 0.47052605395966873, + "learning_rate": 7.238352521860049e-06, + "loss": 0.2874, + "step": 2498 + }, + { + "epoch": 0.8343906510851419, + "grad_norm": 0.4499367541604685, + "learning_rate": 7.235746681244302e-06, + "loss": 0.2888, + "step": 2499 + }, + { + "epoch": 0.8347245409015025, + "grad_norm": 0.4389029210888611, + "learning_rate": 7.233140081380541e-06, + "loss": 0.2878, + "step": 2500 + }, + { + "epoch": 0.8350584307178631, + "grad_norm": 0.4792953987888749, + "learning_rate": 7.230532723153956e-06, + "loss": 0.3007, + "step": 2501 + }, + { + "epoch": 0.8353923205342237, + "grad_norm": 0.4616736326740636, + "learning_rate": 7.2279246074499895e-06, + "loss": 0.2928, + "step": 2502 + }, + { + "epoch": 0.8357262103505843, + "grad_norm": 0.46629638374774335, + "learning_rate": 7.225315735154346e-06, + "loss": 0.2729, + "step": 2503 + }, + { + "epoch": 0.8360601001669449, + "grad_norm": 0.4392611366639684, + "learning_rate": 7.222706107152986e-06, + "loss": 0.276, + "step": 2504 + }, + { + "epoch": 0.8363939899833055, + "grad_norm": 0.4433005059381957, + "learning_rate": 7.220095724332122e-06, + "loss": 0.2941, + "step": 2505 + }, + { + "epoch": 0.8367278797996661, + "grad_norm": 0.4479015097377681, + "learning_rate": 7.2174845875782305e-06, + "loss": 0.2931, + "step": 2506 + }, + { + "epoch": 0.8370617696160267, + "grad_norm": 0.47994365060736727, + "learning_rate": 7.214872697778037e-06, + "loss": 0.2909, + "step": 2507 + }, + { + "epoch": 0.8373956594323874, + "grad_norm": 0.4571509625692555, + "learning_rate": 7.212260055818528e-06, + "loss": 0.2866, + "step": 2508 + }, + { + "epoch": 0.8377295492487479, + "grad_norm": 0.4741739596287434, + "learning_rate": 7.209646662586941e-06, + "loss": 0.2889, + "step": 2509 + }, + { + "epoch": 0.8380634390651085, + "grad_norm": 0.45948019995945594, + "learning_rate": 7.207032518970774e-06, + "loss": 0.2869, + "step": 2510 + }, + { + "epoch": 0.8383973288814691, + "grad_norm": 0.4487301832848612, + "learning_rate": 7.204417625857772e-06, + "loss": 0.2877, + "step": 2511 + }, + { + "epoch": 0.8387312186978297, + "grad_norm": 0.43083801154813445, + "learning_rate": 7.2018019841359435e-06, + "loss": 0.2833, + "step": 2512 + }, + { + "epoch": 0.8390651085141904, + "grad_norm": 0.4590393599208242, + "learning_rate": 7.199185594693545e-06, + "loss": 0.2878, + "step": 2513 + }, + { + "epoch": 0.8393989983305509, + "grad_norm": 0.46023687278883546, + "learning_rate": 7.1965684584190894e-06, + "loss": 0.2859, + "step": 2514 + }, + { + "epoch": 0.8397328881469115, + "grad_norm": 0.4600448187265532, + "learning_rate": 7.193950576201343e-06, + "loss": 0.2856, + "step": 2515 + }, + { + "epoch": 0.8400667779632721, + "grad_norm": 0.4247840005425592, + "learning_rate": 7.191331948929323e-06, + "loss": 0.282, + "step": 2516 + }, + { + "epoch": 0.8404006677796327, + "grad_norm": 0.44825970107464796, + "learning_rate": 7.188712577492305e-06, + "loss": 0.287, + "step": 2517 + }, + { + "epoch": 0.8407345575959934, + "grad_norm": 0.4486242346854203, + "learning_rate": 7.18609246277981e-06, + "loss": 0.287, + "step": 2518 + }, + { + "epoch": 0.8410684474123539, + "grad_norm": 0.4545875295933295, + "learning_rate": 7.18347160568162e-06, + "loss": 0.2872, + "step": 2519 + }, + { + "epoch": 0.8414023372287145, + "grad_norm": 0.445773195329347, + "learning_rate": 7.180850007087761e-06, + "loss": 0.2691, + "step": 2520 + }, + { + "epoch": 0.8417362270450751, + "grad_norm": 0.4822050504988853, + "learning_rate": 7.178227667888513e-06, + "loss": 0.303, + "step": 2521 + }, + { + "epoch": 0.8420701168614357, + "grad_norm": 0.46626369398762774, + "learning_rate": 7.1756045889744145e-06, + "loss": 0.2769, + "step": 2522 + }, + { + "epoch": 0.8424040066777964, + "grad_norm": 0.4540393709757939, + "learning_rate": 7.1729807712362455e-06, + "loss": 0.2895, + "step": 2523 + }, + { + "epoch": 0.842737896494157, + "grad_norm": 0.44017500036839885, + "learning_rate": 7.170356215565042e-06, + "loss": 0.2739, + "step": 2524 + }, + { + "epoch": 0.8430717863105175, + "grad_norm": 0.45714145225745784, + "learning_rate": 7.1677309228520865e-06, + "loss": 0.288, + "step": 2525 + }, + { + "epoch": 0.8434056761268781, + "grad_norm": 0.46474192749999754, + "learning_rate": 7.16510489398892e-06, + "loss": 0.2875, + "step": 2526 + }, + { + "epoch": 0.8437395659432387, + "grad_norm": 0.4405028010635881, + "learning_rate": 7.162478129867326e-06, + "loss": 0.2727, + "step": 2527 + }, + { + "epoch": 0.8440734557595994, + "grad_norm": 0.46280694439635456, + "learning_rate": 7.1598506313793395e-06, + "loss": 0.2898, + "step": 2528 + }, + { + "epoch": 0.84440734557596, + "grad_norm": 0.4502053261512242, + "learning_rate": 7.157222399417245e-06, + "loss": 0.2854, + "step": 2529 + }, + { + "epoch": 0.8447412353923205, + "grad_norm": 0.444437174205444, + "learning_rate": 7.154593434873578e-06, + "loss": 0.2745, + "step": 2530 + }, + { + "epoch": 0.8450751252086811, + "grad_norm": 0.43688350266815495, + "learning_rate": 7.151963738641121e-06, + "loss": 0.2678, + "step": 2531 + }, + { + "epoch": 0.8454090150250417, + "grad_norm": 0.457848448908849, + "learning_rate": 7.149333311612905e-06, + "loss": 0.2902, + "step": 2532 + }, + { + "epoch": 0.8457429048414024, + "grad_norm": 0.43724324525853897, + "learning_rate": 7.146702154682208e-06, + "loss": 0.2755, + "step": 2533 + }, + { + "epoch": 0.846076794657763, + "grad_norm": 0.4715485421292701, + "learning_rate": 7.14407026874256e-06, + "loss": 0.2876, + "step": 2534 + }, + { + "epoch": 0.8464106844741235, + "grad_norm": 0.4500645811375087, + "learning_rate": 7.141437654687733e-06, + "loss": 0.2811, + "step": 2535 + }, + { + "epoch": 0.8467445742904841, + "grad_norm": 0.49631127461856994, + "learning_rate": 7.138804313411751e-06, + "loss": 0.296, + "step": 2536 + }, + { + "epoch": 0.8470784641068447, + "grad_norm": 0.47575818406153203, + "learning_rate": 7.136170245808881e-06, + "loss": 0.2962, + "step": 2537 + }, + { + "epoch": 0.8474123539232054, + "grad_norm": 0.4771237244162959, + "learning_rate": 7.133535452773643e-06, + "loss": 0.279, + "step": 2538 + }, + { + "epoch": 0.847746243739566, + "grad_norm": 0.4277081676302324, + "learning_rate": 7.130899935200794e-06, + "loss": 0.2748, + "step": 2539 + }, + { + "epoch": 0.8480801335559266, + "grad_norm": 0.46593065048850424, + "learning_rate": 7.128263693985347e-06, + "loss": 0.2883, + "step": 2540 + }, + { + "epoch": 0.8484140233722871, + "grad_norm": 0.4661433057267899, + "learning_rate": 7.1256267300225525e-06, + "loss": 0.2994, + "step": 2541 + }, + { + "epoch": 0.8487479131886477, + "grad_norm": 0.4480236426139591, + "learning_rate": 7.12298904420791e-06, + "loss": 0.2755, + "step": 2542 + }, + { + "epoch": 0.8490818030050084, + "grad_norm": 0.4406423062550371, + "learning_rate": 7.120350637437166e-06, + "loss": 0.277, + "step": 2543 + }, + { + "epoch": 0.849415692821369, + "grad_norm": 0.47707515127262223, + "learning_rate": 7.117711510606306e-06, + "loss": 0.2915, + "step": 2544 + }, + { + "epoch": 0.8497495826377296, + "grad_norm": 0.46005993971462306, + "learning_rate": 7.11507166461157e-06, + "loss": 0.2803, + "step": 2545 + }, + { + "epoch": 0.8500834724540901, + "grad_norm": 0.4575058172414909, + "learning_rate": 7.112431100349429e-06, + "loss": 0.2858, + "step": 2546 + }, + { + "epoch": 0.8504173622704507, + "grad_norm": 0.45306789765347427, + "learning_rate": 7.10978981871661e-06, + "loss": 0.2763, + "step": 2547 + }, + { + "epoch": 0.8507512520868113, + "grad_norm": 0.4309216273902561, + "learning_rate": 7.107147820610077e-06, + "loss": 0.2673, + "step": 2548 + }, + { + "epoch": 0.851085141903172, + "grad_norm": 0.46743906446747335, + "learning_rate": 7.1045051069270395e-06, + "loss": 0.2809, + "step": 2549 + }, + { + "epoch": 0.8514190317195326, + "grad_norm": 0.46158531774578415, + "learning_rate": 7.101861678564947e-06, + "loss": 0.2848, + "step": 2550 + }, + { + "epoch": 0.8517529215358931, + "grad_norm": 0.4701804537083453, + "learning_rate": 7.099217536421497e-06, + "loss": 0.3029, + "step": 2551 + }, + { + "epoch": 0.8520868113522537, + "grad_norm": 0.43316691684168773, + "learning_rate": 7.096572681394625e-06, + "loss": 0.2777, + "step": 2552 + }, + { + "epoch": 0.8524207011686143, + "grad_norm": 0.4941632513406352, + "learning_rate": 7.093927114382511e-06, + "loss": 0.292, + "step": 2553 + }, + { + "epoch": 0.852754590984975, + "grad_norm": 0.4721560473485308, + "learning_rate": 7.091280836283574e-06, + "loss": 0.2901, + "step": 2554 + }, + { + "epoch": 0.8530884808013356, + "grad_norm": 0.44306486546062457, + "learning_rate": 7.088633847996479e-06, + "loss": 0.2804, + "step": 2555 + }, + { + "epoch": 0.8534223706176962, + "grad_norm": 0.4527496324223066, + "learning_rate": 7.085986150420128e-06, + "loss": 0.2897, + "step": 2556 + }, + { + "epoch": 0.8537562604340567, + "grad_norm": 0.41986798373822454, + "learning_rate": 7.083337744453666e-06, + "loss": 0.2743, + "step": 2557 + }, + { + "epoch": 0.8540901502504173, + "grad_norm": 0.4716997706214041, + "learning_rate": 7.080688630996477e-06, + "loss": 0.2897, + "step": 2558 + }, + { + "epoch": 0.854424040066778, + "grad_norm": 0.4682160092618649, + "learning_rate": 7.078038810948187e-06, + "loss": 0.2808, + "step": 2559 + }, + { + "epoch": 0.8547579298831386, + "grad_norm": 0.4699668012035714, + "learning_rate": 7.07538828520866e-06, + "loss": 0.295, + "step": 2560 + }, + { + "epoch": 0.8550918196994992, + "grad_norm": 0.4668182678770529, + "learning_rate": 7.072737054678004e-06, + "loss": 0.3024, + "step": 2561 + }, + { + "epoch": 0.8554257095158597, + "grad_norm": 0.4705219658701618, + "learning_rate": 7.070085120256559e-06, + "loss": 0.2941, + "step": 2562 + }, + { + "epoch": 0.8557595993322203, + "grad_norm": 0.4678407100773061, + "learning_rate": 7.067432482844911e-06, + "loss": 0.2819, + "step": 2563 + }, + { + "epoch": 0.856093489148581, + "grad_norm": 0.47051705642434954, + "learning_rate": 7.064779143343881e-06, + "loss": 0.2827, + "step": 2564 + }, + { + "epoch": 0.8564273789649416, + "grad_norm": 0.4544695529820105, + "learning_rate": 7.06212510265453e-06, + "loss": 0.2783, + "step": 2565 + }, + { + "epoch": 0.8567612687813022, + "grad_norm": 0.4751280841870434, + "learning_rate": 7.059470361678156e-06, + "loss": 0.2803, + "step": 2566 + }, + { + "epoch": 0.8570951585976627, + "grad_norm": 0.46871685752883785, + "learning_rate": 7.056814921316295e-06, + "loss": 0.2925, + "step": 2567 + }, + { + "epoch": 0.8574290484140233, + "grad_norm": 0.4631173416236228, + "learning_rate": 7.054158782470722e-06, + "loss": 0.2766, + "step": 2568 + }, + { + "epoch": 0.857762938230384, + "grad_norm": 0.4538735602084514, + "learning_rate": 7.051501946043446e-06, + "loss": 0.2881, + "step": 2569 + }, + { + "epoch": 0.8580968280467446, + "grad_norm": 0.47366486458231627, + "learning_rate": 7.048844412936718e-06, + "loss": 0.2835, + "step": 2570 + }, + { + "epoch": 0.8584307178631052, + "grad_norm": 0.4476293239290854, + "learning_rate": 7.046186184053019e-06, + "loss": 0.2742, + "step": 2571 + }, + { + "epoch": 0.8587646076794658, + "grad_norm": 0.4827429728074449, + "learning_rate": 7.043527260295073e-06, + "loss": 0.2867, + "step": 2572 + }, + { + "epoch": 0.8590984974958263, + "grad_norm": 0.45483591392176026, + "learning_rate": 7.040867642565834e-06, + "loss": 0.2794, + "step": 2573 + }, + { + "epoch": 0.859432387312187, + "grad_norm": 0.44005772402635507, + "learning_rate": 7.038207331768496e-06, + "loss": 0.28, + "step": 2574 + }, + { + "epoch": 0.8597662771285476, + "grad_norm": 0.45622500532482885, + "learning_rate": 7.0355463288064865e-06, + "loss": 0.2905, + "step": 2575 + }, + { + "epoch": 0.8601001669449082, + "grad_norm": 0.4703151591171797, + "learning_rate": 7.032884634583468e-06, + "loss": 0.2925, + "step": 2576 + }, + { + "epoch": 0.8604340567612688, + "grad_norm": 0.45052549720921653, + "learning_rate": 7.030222250003338e-06, + "loss": 0.2721, + "step": 2577 + }, + { + "epoch": 0.8607679465776293, + "grad_norm": 0.42895714506848376, + "learning_rate": 7.027559175970229e-06, + "loss": 0.274, + "step": 2578 + }, + { + "epoch": 0.86110183639399, + "grad_norm": 0.4471265751611153, + "learning_rate": 7.024895413388508e-06, + "loss": 0.2911, + "step": 2579 + }, + { + "epoch": 0.8614357262103506, + "grad_norm": 0.47584964266511554, + "learning_rate": 7.022230963162773e-06, + "loss": 0.2806, + "step": 2580 + }, + { + "epoch": 0.8617696160267112, + "grad_norm": 0.44963815687407566, + "learning_rate": 7.019565826197859e-06, + "loss": 0.2891, + "step": 2581 + }, + { + "epoch": 0.8621035058430718, + "grad_norm": 0.4303973162426582, + "learning_rate": 7.016900003398832e-06, + "loss": 0.2717, + "step": 2582 + }, + { + "epoch": 0.8624373956594323, + "grad_norm": 0.46933329065938706, + "learning_rate": 7.014233495670994e-06, + "loss": 0.2864, + "step": 2583 + }, + { + "epoch": 0.862771285475793, + "grad_norm": 0.45519665388831976, + "learning_rate": 7.011566303919872e-06, + "loss": 0.2841, + "step": 2584 + }, + { + "epoch": 0.8631051752921536, + "grad_norm": 0.46791591132451393, + "learning_rate": 7.008898429051235e-06, + "loss": 0.2891, + "step": 2585 + }, + { + "epoch": 0.8634390651085142, + "grad_norm": 0.4462422624965346, + "learning_rate": 7.00622987197108e-06, + "loss": 0.2917, + "step": 2586 + }, + { + "epoch": 0.8637729549248748, + "grad_norm": 0.460405226329125, + "learning_rate": 7.003560633585631e-06, + "loss": 0.2928, + "step": 2587 + }, + { + "epoch": 0.8641068447412354, + "grad_norm": 0.4428283313242832, + "learning_rate": 7.000890714801352e-06, + "loss": 0.2918, + "step": 2588 + }, + { + "epoch": 0.864440734557596, + "grad_norm": 0.4742980714221235, + "learning_rate": 6.99822011652493e-06, + "loss": 0.2931, + "step": 2589 + }, + { + "epoch": 0.8647746243739566, + "grad_norm": 0.4672226274289243, + "learning_rate": 6.995548839663289e-06, + "loss": 0.3012, + "step": 2590 + }, + { + "epoch": 0.8651085141903172, + "grad_norm": 0.43607597699032774, + "learning_rate": 6.992876885123578e-06, + "loss": 0.2856, + "step": 2591 + }, + { + "epoch": 0.8654424040066778, + "grad_norm": 0.4672581961528126, + "learning_rate": 6.990204253813182e-06, + "loss": 0.2882, + "step": 2592 + }, + { + "epoch": 0.8657762938230384, + "grad_norm": 0.4474881087565465, + "learning_rate": 6.987530946639712e-06, + "loss": 0.2819, + "step": 2593 + }, + { + "epoch": 0.866110183639399, + "grad_norm": 0.44403042573587836, + "learning_rate": 6.984856964511007e-06, + "loss": 0.2814, + "step": 2594 + }, + { + "epoch": 0.8664440734557596, + "grad_norm": 0.46744019446424506, + "learning_rate": 6.982182308335142e-06, + "loss": 0.2929, + "step": 2595 + }, + { + "epoch": 0.8667779632721202, + "grad_norm": 0.4445841376005458, + "learning_rate": 6.979506979020412e-06, + "loss": 0.2807, + "step": 2596 + }, + { + "epoch": 0.8671118530884808, + "grad_norm": 0.43092758705267514, + "learning_rate": 6.976830977475346e-06, + "loss": 0.2831, + "step": 2597 + }, + { + "epoch": 0.8674457429048414, + "grad_norm": 0.4424297221885524, + "learning_rate": 6.974154304608703e-06, + "loss": 0.2669, + "step": 2598 + }, + { + "epoch": 0.867779632721202, + "grad_norm": 0.4585271487924436, + "learning_rate": 6.971476961329464e-06, + "loss": 0.2937, + "step": 2599 + }, + { + "epoch": 0.8681135225375626, + "grad_norm": 0.4475696511111327, + "learning_rate": 6.968798948546842e-06, + "loss": 0.2919, + "step": 2600 + }, + { + "epoch": 0.8684474123539232, + "grad_norm": 0.5019209839693296, + "learning_rate": 6.9661202671702765e-06, + "loss": 0.2924, + "step": 2601 + }, + { + "epoch": 0.8687813021702838, + "grad_norm": 0.46745936273178773, + "learning_rate": 6.9634409181094355e-06, + "loss": 0.2964, + "step": 2602 + }, + { + "epoch": 0.8691151919866444, + "grad_norm": 0.4556788223929218, + "learning_rate": 6.9607609022742085e-06, + "loss": 0.2864, + "step": 2603 + }, + { + "epoch": 0.8694490818030051, + "grad_norm": 0.4467304422106497, + "learning_rate": 6.958080220574719e-06, + "loss": 0.2772, + "step": 2604 + }, + { + "epoch": 0.8697829716193656, + "grad_norm": 0.47193609573438877, + "learning_rate": 6.9553988739213075e-06, + "loss": 0.2921, + "step": 2605 + }, + { + "epoch": 0.8701168614357262, + "grad_norm": 0.44567408011640164, + "learning_rate": 6.952716863224551e-06, + "loss": 0.2823, + "step": 2606 + }, + { + "epoch": 0.8704507512520868, + "grad_norm": 0.45802425289833654, + "learning_rate": 6.950034189395243e-06, + "loss": 0.2887, + "step": 2607 + }, + { + "epoch": 0.8707846410684474, + "grad_norm": 0.44569610219002465, + "learning_rate": 6.947350853344408e-06, + "loss": 0.2792, + "step": 2608 + }, + { + "epoch": 0.8711185308848081, + "grad_norm": 0.4317219541009044, + "learning_rate": 6.9446668559832895e-06, + "loss": 0.279, + "step": 2609 + }, + { + "epoch": 0.8714524207011686, + "grad_norm": 0.4442205356763654, + "learning_rate": 6.941982198223362e-06, + "loss": 0.2732, + "step": 2610 + }, + { + "epoch": 0.8717863105175292, + "grad_norm": 0.45618614958814024, + "learning_rate": 6.939296880976322e-06, + "loss": 0.2908, + "step": 2611 + }, + { + "epoch": 0.8721202003338898, + "grad_norm": 0.46919683315595706, + "learning_rate": 6.936610905154088e-06, + "loss": 0.2924, + "step": 2612 + }, + { + "epoch": 0.8724540901502504, + "grad_norm": 0.46135888461187224, + "learning_rate": 6.933924271668802e-06, + "loss": 0.283, + "step": 2613 + }, + { + "epoch": 0.8727879799666111, + "grad_norm": 0.46413435848021717, + "learning_rate": 6.931236981432833e-06, + "loss": 0.281, + "step": 2614 + }, + { + "epoch": 0.8731218697829716, + "grad_norm": 0.4704883188587761, + "learning_rate": 6.928549035358772e-06, + "loss": 0.2823, + "step": 2615 + }, + { + "epoch": 0.8734557595993322, + "grad_norm": 0.44629783682345686, + "learning_rate": 6.925860434359427e-06, + "loss": 0.2719, + "step": 2616 + }, + { + "epoch": 0.8737896494156928, + "grad_norm": 0.46798641410413044, + "learning_rate": 6.923171179347838e-06, + "loss": 0.2959, + "step": 2617 + }, + { + "epoch": 0.8741235392320534, + "grad_norm": 0.4615627537571001, + "learning_rate": 6.92048127123726e-06, + "loss": 0.2894, + "step": 2618 + }, + { + "epoch": 0.8744574290484141, + "grad_norm": 0.4676244553418392, + "learning_rate": 6.917790710941171e-06, + "loss": 0.2872, + "step": 2619 + }, + { + "epoch": 0.8747913188647746, + "grad_norm": 0.4593050809371556, + "learning_rate": 6.915099499373273e-06, + "loss": 0.2737, + "step": 2620 + }, + { + "epoch": 0.8751252086811352, + "grad_norm": 0.45036959043388125, + "learning_rate": 6.912407637447487e-06, + "loss": 0.2748, + "step": 2621 + }, + { + "epoch": 0.8754590984974958, + "grad_norm": 0.45039774454184894, + "learning_rate": 6.909715126077953e-06, + "loss": 0.2811, + "step": 2622 + }, + { + "epoch": 0.8757929883138564, + "grad_norm": 0.43493405497682786, + "learning_rate": 6.907021966179039e-06, + "loss": 0.2699, + "step": 2623 + }, + { + "epoch": 0.8761268781302171, + "grad_norm": 0.46222495705031363, + "learning_rate": 6.904328158665323e-06, + "loss": 0.2804, + "step": 2624 + }, + { + "epoch": 0.8764607679465777, + "grad_norm": 0.4538736966073643, + "learning_rate": 6.901633704451614e-06, + "loss": 0.2783, + "step": 2625 + }, + { + "epoch": 0.8767946577629382, + "grad_norm": 0.45029808702413127, + "learning_rate": 6.898938604452926e-06, + "loss": 0.2716, + "step": 2626 + }, + { + "epoch": 0.8771285475792988, + "grad_norm": 0.4720425663666722, + "learning_rate": 6.89624285958451e-06, + "loss": 0.2766, + "step": 2627 + }, + { + "epoch": 0.8774624373956594, + "grad_norm": 0.44961393853769943, + "learning_rate": 6.89354647076182e-06, + "loss": 0.2803, + "step": 2628 + }, + { + "epoch": 0.8777963272120201, + "grad_norm": 0.4384764144994483, + "learning_rate": 6.890849438900542e-06, + "loss": 0.283, + "step": 2629 + }, + { + "epoch": 0.8781302170283807, + "grad_norm": 0.4361683098213447, + "learning_rate": 6.88815176491657e-06, + "loss": 0.2706, + "step": 2630 + }, + { + "epoch": 0.8784641068447412, + "grad_norm": 0.4803035126771615, + "learning_rate": 6.88545344972602e-06, + "loss": 0.2744, + "step": 2631 + }, + { + "epoch": 0.8787979966611018, + "grad_norm": 0.46806137455364144, + "learning_rate": 6.882754494245229e-06, + "loss": 0.288, + "step": 2632 + }, + { + "epoch": 0.8791318864774624, + "grad_norm": 0.4482808959315587, + "learning_rate": 6.880054899390744e-06, + "loss": 0.2899, + "step": 2633 + }, + { + "epoch": 0.879465776293823, + "grad_norm": 0.430750744534407, + "learning_rate": 6.8773546660793374e-06, + "loss": 0.2852, + "step": 2634 + }, + { + "epoch": 0.8797996661101837, + "grad_norm": 0.47732083306286166, + "learning_rate": 6.87465379522799e-06, + "loss": 0.2926, + "step": 2635 + }, + { + "epoch": 0.8801335559265442, + "grad_norm": 0.4550135866657338, + "learning_rate": 6.87195228775391e-06, + "loss": 0.2797, + "step": 2636 + }, + { + "epoch": 0.8804674457429048, + "grad_norm": 0.43283983796272696, + "learning_rate": 6.8692501445745094e-06, + "loss": 0.271, + "step": 2637 + }, + { + "epoch": 0.8808013355592654, + "grad_norm": 0.4670822842432418, + "learning_rate": 6.866547366607427e-06, + "loss": 0.302, + "step": 2638 + }, + { + "epoch": 0.881135225375626, + "grad_norm": 0.45111086885787305, + "learning_rate": 6.863843954770507e-06, + "loss": 0.2941, + "step": 2639 + }, + { + "epoch": 0.8814691151919867, + "grad_norm": 0.45548531570552964, + "learning_rate": 6.861139909981818e-06, + "loss": 0.2889, + "step": 2640 + }, + { + "epoch": 0.8818030050083473, + "grad_norm": 0.4286044150947517, + "learning_rate": 6.858435233159638e-06, + "loss": 0.2766, + "step": 2641 + }, + { + "epoch": 0.8821368948247078, + "grad_norm": 0.4495911706610504, + "learning_rate": 6.855729925222462e-06, + "loss": 0.2865, + "step": 2642 + }, + { + "epoch": 0.8824707846410684, + "grad_norm": 0.4746431784739942, + "learning_rate": 6.853023987088997e-06, + "loss": 0.2955, + "step": 2643 + }, + { + "epoch": 0.882804674457429, + "grad_norm": 0.42079050679647567, + "learning_rate": 6.850317419678168e-06, + "loss": 0.2743, + "step": 2644 + }, + { + "epoch": 0.8831385642737897, + "grad_norm": 0.46137629516006945, + "learning_rate": 6.847610223909109e-06, + "loss": 0.2948, + "step": 2645 + }, + { + "epoch": 0.8834724540901503, + "grad_norm": 0.4549546990794912, + "learning_rate": 6.844902400701172e-06, + "loss": 0.2857, + "step": 2646 + }, + { + "epoch": 0.8838063439065108, + "grad_norm": 0.49558165872172677, + "learning_rate": 6.842193950973919e-06, + "loss": 0.2807, + "step": 2647 + }, + { + "epoch": 0.8841402337228714, + "grad_norm": 0.45181213261790915, + "learning_rate": 6.839484875647124e-06, + "loss": 0.2725, + "step": 2648 + }, + { + "epoch": 0.884474123539232, + "grad_norm": 0.44798491133758744, + "learning_rate": 6.836775175640775e-06, + "loss": 0.2861, + "step": 2649 + }, + { + "epoch": 0.8848080133555927, + "grad_norm": 0.46106198450631797, + "learning_rate": 6.834064851875075e-06, + "loss": 0.2869, + "step": 2650 + }, + { + "epoch": 0.8851419031719533, + "grad_norm": 0.4716076815878801, + "learning_rate": 6.831353905270433e-06, + "loss": 0.2825, + "step": 2651 + }, + { + "epoch": 0.8854757929883138, + "grad_norm": 0.4694114462759815, + "learning_rate": 6.828642336747474e-06, + "loss": 0.2831, + "step": 2652 + }, + { + "epoch": 0.8858096828046744, + "grad_norm": 0.4690007350033732, + "learning_rate": 6.8259301472270314e-06, + "loss": 0.2904, + "step": 2653 + }, + { + "epoch": 0.886143572621035, + "grad_norm": 0.4366035915320944, + "learning_rate": 6.823217337630153e-06, + "loss": 0.2837, + "step": 2654 + }, + { + "epoch": 0.8864774624373957, + "grad_norm": 0.4798804600547086, + "learning_rate": 6.820503908878093e-06, + "loss": 0.2907, + "step": 2655 + }, + { + "epoch": 0.8868113522537563, + "grad_norm": 0.4433181002921653, + "learning_rate": 6.817789861892318e-06, + "loss": 0.2778, + "step": 2656 + }, + { + "epoch": 0.8871452420701169, + "grad_norm": 0.4728929485240123, + "learning_rate": 6.815075197594505e-06, + "loss": 0.2846, + "step": 2657 + }, + { + "epoch": 0.8874791318864774, + "grad_norm": 0.4495400373351323, + "learning_rate": 6.812359916906539e-06, + "loss": 0.2793, + "step": 2658 + }, + { + "epoch": 0.887813021702838, + "grad_norm": 0.4817818841616904, + "learning_rate": 6.809644020750517e-06, + "loss": 0.2953, + "step": 2659 + }, + { + "epoch": 0.8881469115191987, + "grad_norm": 0.4607049295565626, + "learning_rate": 6.806927510048738e-06, + "loss": 0.2869, + "step": 2660 + }, + { + "epoch": 0.8884808013355593, + "grad_norm": 0.4448564068890994, + "learning_rate": 6.804210385723723e-06, + "loss": 0.2862, + "step": 2661 + }, + { + "epoch": 0.8888146911519199, + "grad_norm": 0.45121499813339955, + "learning_rate": 6.801492648698188e-06, + "loss": 0.2902, + "step": 2662 + }, + { + "epoch": 0.8891485809682804, + "grad_norm": 0.4361192225346147, + "learning_rate": 6.798774299895066e-06, + "loss": 0.2838, + "step": 2663 + }, + { + "epoch": 0.889482470784641, + "grad_norm": 0.4535577278249873, + "learning_rate": 6.79605534023749e-06, + "loss": 0.2767, + "step": 2664 + }, + { + "epoch": 0.8898163606010017, + "grad_norm": 0.4886006703201498, + "learning_rate": 6.793335770648806e-06, + "loss": 0.2993, + "step": 2665 + }, + { + "epoch": 0.8901502504173623, + "grad_norm": 0.44301650775471224, + "learning_rate": 6.790615592052568e-06, + "loss": 0.277, + "step": 2666 + }, + { + "epoch": 0.8904841402337229, + "grad_norm": 0.4457842467135025, + "learning_rate": 6.787894805372532e-06, + "loss": 0.2698, + "step": 2667 + }, + { + "epoch": 0.8908180300500834, + "grad_norm": 0.43591646661542627, + "learning_rate": 6.785173411532666e-06, + "loss": 0.2773, + "step": 2668 + }, + { + "epoch": 0.891151919866444, + "grad_norm": 0.44136164929114735, + "learning_rate": 6.782451411457137e-06, + "loss": 0.2884, + "step": 2669 + }, + { + "epoch": 0.8914858096828047, + "grad_norm": 0.4451127172193005, + "learning_rate": 6.779728806070327e-06, + "loss": 0.2841, + "step": 2670 + }, + { + "epoch": 0.8918196994991653, + "grad_norm": 0.45915790614782825, + "learning_rate": 6.777005596296815e-06, + "loss": 0.2982, + "step": 2671 + }, + { + "epoch": 0.8921535893155259, + "grad_norm": 0.46641518949036087, + "learning_rate": 6.774281783061391e-06, + "loss": 0.2838, + "step": 2672 + }, + { + "epoch": 0.8924874791318865, + "grad_norm": 0.43601664954328917, + "learning_rate": 6.7715573672890455e-06, + "loss": 0.281, + "step": 2673 + }, + { + "epoch": 0.892821368948247, + "grad_norm": 0.4250529536498888, + "learning_rate": 6.768832349904979e-06, + "loss": 0.2726, + "step": 2674 + }, + { + "epoch": 0.8931552587646077, + "grad_norm": 0.451867930483308, + "learning_rate": 6.766106731834592e-06, + "loss": 0.283, + "step": 2675 + }, + { + "epoch": 0.8934891485809683, + "grad_norm": 0.45641463979801034, + "learning_rate": 6.76338051400349e-06, + "loss": 0.2848, + "step": 2676 + }, + { + "epoch": 0.8938230383973289, + "grad_norm": 0.47175747525150313, + "learning_rate": 6.760653697337483e-06, + "loss": 0.2804, + "step": 2677 + }, + { + "epoch": 0.8941569282136895, + "grad_norm": 0.44675960761590755, + "learning_rate": 6.757926282762583e-06, + "loss": 0.2661, + "step": 2678 + }, + { + "epoch": 0.89449081803005, + "grad_norm": 0.4540045165327667, + "learning_rate": 6.755198271205008e-06, + "loss": 0.2876, + "step": 2679 + }, + { + "epoch": 0.8948247078464107, + "grad_norm": 0.44103714830535184, + "learning_rate": 6.752469663591174e-06, + "loss": 0.2808, + "step": 2680 + }, + { + "epoch": 0.8951585976627713, + "grad_norm": 0.4353283224652031, + "learning_rate": 6.7497404608477035e-06, + "loss": 0.2775, + "step": 2681 + }, + { + "epoch": 0.8954924874791319, + "grad_norm": 0.4521458694209575, + "learning_rate": 6.747010663901419e-06, + "loss": 0.2746, + "step": 2682 + }, + { + "epoch": 0.8958263772954925, + "grad_norm": 0.45731412971822527, + "learning_rate": 6.744280273679345e-06, + "loss": 0.2857, + "step": 2683 + }, + { + "epoch": 0.896160267111853, + "grad_norm": 0.44913748557913885, + "learning_rate": 6.74154929110871e-06, + "loss": 0.2908, + "step": 2684 + }, + { + "epoch": 0.8964941569282137, + "grad_norm": 0.4623193036902147, + "learning_rate": 6.738817717116939e-06, + "loss": 0.2809, + "step": 2685 + }, + { + "epoch": 0.8968280467445743, + "grad_norm": 0.4458092640441154, + "learning_rate": 6.736085552631663e-06, + "loss": 0.2778, + "step": 2686 + }, + { + "epoch": 0.8971619365609349, + "grad_norm": 0.4518051159970357, + "learning_rate": 6.733352798580708e-06, + "loss": 0.2947, + "step": 2687 + }, + { + "epoch": 0.8974958263772955, + "grad_norm": 0.413136934099687, + "learning_rate": 6.730619455892106e-06, + "loss": 0.2671, + "step": 2688 + }, + { + "epoch": 0.8978297161936561, + "grad_norm": 0.44107910834598524, + "learning_rate": 6.727885525494083e-06, + "loss": 0.2743, + "step": 2689 + }, + { + "epoch": 0.8981636060100167, + "grad_norm": 0.438446285553196, + "learning_rate": 6.725151008315071e-06, + "loss": 0.2746, + "step": 2690 + }, + { + "epoch": 0.8984974958263773, + "grad_norm": 0.4498910406063331, + "learning_rate": 6.722415905283696e-06, + "loss": 0.2915, + "step": 2691 + }, + { + "epoch": 0.8988313856427379, + "grad_norm": 0.4822421127914717, + "learning_rate": 6.719680217328784e-06, + "loss": 0.2974, + "step": 2692 + }, + { + "epoch": 0.8991652754590985, + "grad_norm": 0.44975178493310974, + "learning_rate": 6.716943945379364e-06, + "loss": 0.2909, + "step": 2693 + }, + { + "epoch": 0.8994991652754591, + "grad_norm": 0.4515526618612398, + "learning_rate": 6.714207090364654e-06, + "loss": 0.2883, + "step": 2694 + }, + { + "epoch": 0.8998330550918197, + "grad_norm": 0.4373662791244928, + "learning_rate": 6.7114696532140825e-06, + "loss": 0.2792, + "step": 2695 + }, + { + "epoch": 0.9001669449081803, + "grad_norm": 0.4584902016334691, + "learning_rate": 6.7087316348572626e-06, + "loss": 0.2841, + "step": 2696 + }, + { + "epoch": 0.9005008347245409, + "grad_norm": 0.44390347392205076, + "learning_rate": 6.705993036224017e-06, + "loss": 0.2768, + "step": 2697 + }, + { + "epoch": 0.9008347245409015, + "grad_norm": 0.43771423486666194, + "learning_rate": 6.703253858244355e-06, + "loss": 0.2806, + "step": 2698 + }, + { + "epoch": 0.9011686143572621, + "grad_norm": 0.4280542138545146, + "learning_rate": 6.70051410184849e-06, + "loss": 0.2813, + "step": 2699 + }, + { + "epoch": 0.9015025041736227, + "grad_norm": 0.45113104357188216, + "learning_rate": 6.697773767966829e-06, + "loss": 0.2763, + "step": 2700 + }, + { + "epoch": 0.9018363939899833, + "grad_norm": 0.4430281434619522, + "learning_rate": 6.695032857529973e-06, + "loss": 0.2855, + "step": 2701 + }, + { + "epoch": 0.9021702838063439, + "grad_norm": 0.4667618617624305, + "learning_rate": 6.692291371468724e-06, + "loss": 0.2819, + "step": 2702 + }, + { + "epoch": 0.9025041736227045, + "grad_norm": 0.44553178966326695, + "learning_rate": 6.689549310714073e-06, + "loss": 0.2893, + "step": 2703 + }, + { + "epoch": 0.9028380634390651, + "grad_norm": 0.43929192831850966, + "learning_rate": 6.686806676197214e-06, + "loss": 0.2789, + "step": 2704 + }, + { + "epoch": 0.9031719532554258, + "grad_norm": 0.450241176318801, + "learning_rate": 6.684063468849528e-06, + "loss": 0.2825, + "step": 2705 + }, + { + "epoch": 0.9035058430717863, + "grad_norm": 0.4515850943852716, + "learning_rate": 6.681319689602596e-06, + "loss": 0.2845, + "step": 2706 + }, + { + "epoch": 0.9038397328881469, + "grad_norm": 0.44183345050004114, + "learning_rate": 6.678575339388189e-06, + "loss": 0.2825, + "step": 2707 + }, + { + "epoch": 0.9041736227045075, + "grad_norm": 0.4518012848919926, + "learning_rate": 6.6758304191382764e-06, + "loss": 0.2818, + "step": 2708 + }, + { + "epoch": 0.9045075125208681, + "grad_norm": 0.44329721239382003, + "learning_rate": 6.673084929785019e-06, + "loss": 0.2725, + "step": 2709 + }, + { + "epoch": 0.9048414023372288, + "grad_norm": 0.45732358299101666, + "learning_rate": 6.670338872260768e-06, + "loss": 0.287, + "step": 2710 + }, + { + "epoch": 0.9051752921535893, + "grad_norm": 0.46179277634819693, + "learning_rate": 6.667592247498074e-06, + "loss": 0.2911, + "step": 2711 + }, + { + "epoch": 0.9055091819699499, + "grad_norm": 0.4696999832767792, + "learning_rate": 6.664845056429673e-06, + "loss": 0.2828, + "step": 2712 + }, + { + "epoch": 0.9058430717863105, + "grad_norm": 0.4572210081919441, + "learning_rate": 6.662097299988502e-06, + "loss": 0.2909, + "step": 2713 + }, + { + "epoch": 0.9061769616026711, + "grad_norm": 0.4347244868227132, + "learning_rate": 6.659348979107679e-06, + "loss": 0.2781, + "step": 2714 + }, + { + "epoch": 0.9065108514190318, + "grad_norm": 0.44680189849145197, + "learning_rate": 6.656600094720522e-06, + "loss": 0.2859, + "step": 2715 + }, + { + "epoch": 0.9068447412353923, + "grad_norm": 0.45996142838887194, + "learning_rate": 6.653850647760538e-06, + "loss": 0.2944, + "step": 2716 + }, + { + "epoch": 0.9071786310517529, + "grad_norm": 0.4456809502823781, + "learning_rate": 6.651100639161425e-06, + "loss": 0.2785, + "step": 2717 + }, + { + "epoch": 0.9075125208681135, + "grad_norm": 0.4555545564227511, + "learning_rate": 6.6483500698570734e-06, + "loss": 0.2812, + "step": 2718 + }, + { + "epoch": 0.9078464106844741, + "grad_norm": 0.4457468116814632, + "learning_rate": 6.645598940781559e-06, + "loss": 0.2804, + "step": 2719 + }, + { + "epoch": 0.9081803005008348, + "grad_norm": 0.43441214577893106, + "learning_rate": 6.642847252869153e-06, + "loss": 0.2702, + "step": 2720 + }, + { + "epoch": 0.9085141903171954, + "grad_norm": 0.44486388767446505, + "learning_rate": 6.6400950070543155e-06, + "loss": 0.2887, + "step": 2721 + }, + { + "epoch": 0.9088480801335559, + "grad_norm": 0.4701730817638189, + "learning_rate": 6.637342204271693e-06, + "loss": 0.2967, + "step": 2722 + }, + { + "epoch": 0.9091819699499165, + "grad_norm": 0.42334569670978295, + "learning_rate": 6.634588845456122e-06, + "loss": 0.2768, + "step": 2723 + }, + { + "epoch": 0.9095158597662771, + "grad_norm": 0.44978378615096754, + "learning_rate": 6.631834931542631e-06, + "loss": 0.2804, + "step": 2724 + }, + { + "epoch": 0.9098497495826378, + "grad_norm": 0.42213318740795586, + "learning_rate": 6.629080463466435e-06, + "loss": 0.2873, + "step": 2725 + }, + { + "epoch": 0.9101836393989984, + "grad_norm": 0.449783060427424, + "learning_rate": 6.626325442162936e-06, + "loss": 0.2782, + "step": 2726 + }, + { + "epoch": 0.9105175292153589, + "grad_norm": 0.42092533919069797, + "learning_rate": 6.623569868567727e-06, + "loss": 0.2776, + "step": 2727 + }, + { + "epoch": 0.9108514190317195, + "grad_norm": 0.4530207343026499, + "learning_rate": 6.620813743616584e-06, + "loss": 0.2898, + "step": 2728 + }, + { + "epoch": 0.9111853088480801, + "grad_norm": 0.4386713189098542, + "learning_rate": 6.6180570682454734e-06, + "loss": 0.2888, + "step": 2729 + }, + { + "epoch": 0.9115191986644408, + "grad_norm": 0.45047646170759376, + "learning_rate": 6.61529984339055e-06, + "loss": 0.2787, + "step": 2730 + }, + { + "epoch": 0.9118530884808014, + "grad_norm": 0.452371158783999, + "learning_rate": 6.61254206998815e-06, + "loss": 0.2886, + "step": 2731 + }, + { + "epoch": 0.9121869782971619, + "grad_norm": 0.44195368718852296, + "learning_rate": 6.609783748974802e-06, + "loss": 0.2681, + "step": 2732 + }, + { + "epoch": 0.9125208681135225, + "grad_norm": 0.456826773212481, + "learning_rate": 6.607024881287213e-06, + "loss": 0.2885, + "step": 2733 + }, + { + "epoch": 0.9128547579298831, + "grad_norm": 0.46489548432715233, + "learning_rate": 6.604265467862286e-06, + "loss": 0.2878, + "step": 2734 + }, + { + "epoch": 0.9131886477462438, + "grad_norm": 0.4626889664889108, + "learning_rate": 6.6015055096371e-06, + "loss": 0.291, + "step": 2735 + }, + { + "epoch": 0.9135225375626044, + "grad_norm": 0.44872780142370494, + "learning_rate": 6.5987450075489234e-06, + "loss": 0.2841, + "step": 2736 + }, + { + "epoch": 0.9138564273789649, + "grad_norm": 0.4505352627561478, + "learning_rate": 6.595983962535209e-06, + "loss": 0.2815, + "step": 2737 + }, + { + "epoch": 0.9141903171953255, + "grad_norm": 0.4630230391819299, + "learning_rate": 6.59322237553359e-06, + "loss": 0.2865, + "step": 2738 + }, + { + "epoch": 0.9145242070116861, + "grad_norm": 0.41528192416341936, + "learning_rate": 6.5904602474818924e-06, + "loss": 0.2676, + "step": 2739 + }, + { + "epoch": 0.9148580968280468, + "grad_norm": 0.44946377906828944, + "learning_rate": 6.587697579318118e-06, + "loss": 0.2839, + "step": 2740 + }, + { + "epoch": 0.9151919866444074, + "grad_norm": 0.44866977959548193, + "learning_rate": 6.584934371980452e-06, + "loss": 0.2797, + "step": 2741 + }, + { + "epoch": 0.915525876460768, + "grad_norm": 0.46478093070960574, + "learning_rate": 6.582170626407271e-06, + "loss": 0.2863, + "step": 2742 + }, + { + "epoch": 0.9158597662771285, + "grad_norm": 0.43424631776161504, + "learning_rate": 6.579406343537125e-06, + "loss": 0.2818, + "step": 2743 + }, + { + "epoch": 0.9161936560934891, + "grad_norm": 0.43822117788986115, + "learning_rate": 6.57664152430875e-06, + "loss": 0.2782, + "step": 2744 + }, + { + "epoch": 0.9165275459098498, + "grad_norm": 0.46958293898741144, + "learning_rate": 6.573876169661065e-06, + "loss": 0.2737, + "step": 2745 + }, + { + "epoch": 0.9168614357262104, + "grad_norm": 0.4419870474241023, + "learning_rate": 6.571110280533171e-06, + "loss": 0.275, + "step": 2746 + }, + { + "epoch": 0.917195325542571, + "grad_norm": 0.45288649657725666, + "learning_rate": 6.56834385786435e-06, + "loss": 0.2773, + "step": 2747 + }, + { + "epoch": 0.9175292153589315, + "grad_norm": 0.4624720325717901, + "learning_rate": 6.565576902594063e-06, + "loss": 0.2902, + "step": 2748 + }, + { + "epoch": 0.9178631051752921, + "grad_norm": 0.4476036866535528, + "learning_rate": 6.562809415661954e-06, + "loss": 0.2843, + "step": 2749 + }, + { + "epoch": 0.9181969949916527, + "grad_norm": 0.44442424171404865, + "learning_rate": 6.560041398007847e-06, + "loss": 0.2852, + "step": 2750 + }, + { + "epoch": 0.9185308848080134, + "grad_norm": 0.45338911716154856, + "learning_rate": 6.5572728505717475e-06, + "loss": 0.2703, + "step": 2751 + }, + { + "epoch": 0.918864774624374, + "grad_norm": 0.456189447808685, + "learning_rate": 6.554503774293841e-06, + "loss": 0.2892, + "step": 2752 + }, + { + "epoch": 0.9191986644407345, + "grad_norm": 0.44547772280479064, + "learning_rate": 6.55173417011449e-06, + "loss": 0.2862, + "step": 2753 + }, + { + "epoch": 0.9195325542570951, + "grad_norm": 0.46690993678896164, + "learning_rate": 6.548964038974236e-06, + "loss": 0.2941, + "step": 2754 + }, + { + "epoch": 0.9198664440734557, + "grad_norm": 0.4468468545172953, + "learning_rate": 6.546193381813806e-06, + "loss": 0.2866, + "step": 2755 + }, + { + "epoch": 0.9202003338898164, + "grad_norm": 0.4341810417017788, + "learning_rate": 6.543422199574096e-06, + "loss": 0.2786, + "step": 2756 + }, + { + "epoch": 0.920534223706177, + "grad_norm": 0.47709623429709336, + "learning_rate": 6.540650493196187e-06, + "loss": 0.2933, + "step": 2757 + }, + { + "epoch": 0.9208681135225376, + "grad_norm": 0.4247700159456189, + "learning_rate": 6.537878263621335e-06, + "loss": 0.2602, + "step": 2758 + }, + { + "epoch": 0.9212020033388981, + "grad_norm": 0.44464510566709897, + "learning_rate": 6.535105511790979e-06, + "loss": 0.2835, + "step": 2759 + }, + { + "epoch": 0.9215358931552587, + "grad_norm": 0.43915110100501903, + "learning_rate": 6.532332238646726e-06, + "loss": 0.2889, + "step": 2760 + }, + { + "epoch": 0.9218697829716194, + "grad_norm": 0.4506062888764894, + "learning_rate": 6.5295584451303694e-06, + "loss": 0.2852, + "step": 2761 + }, + { + "epoch": 0.92220367278798, + "grad_norm": 0.46576811863550616, + "learning_rate": 6.526784132183872e-06, + "loss": 0.2823, + "step": 2762 + }, + { + "epoch": 0.9225375626043406, + "grad_norm": 0.43701462964732435, + "learning_rate": 6.524009300749377e-06, + "loss": 0.2734, + "step": 2763 + }, + { + "epoch": 0.9228714524207011, + "grad_norm": 0.47213913946600067, + "learning_rate": 6.521233951769204e-06, + "loss": 0.2746, + "step": 2764 + }, + { + "epoch": 0.9232053422370617, + "grad_norm": 0.4364222845294654, + "learning_rate": 6.518458086185846e-06, + "loss": 0.2737, + "step": 2765 + }, + { + "epoch": 0.9235392320534224, + "grad_norm": 0.4510424191586167, + "learning_rate": 6.515681704941974e-06, + "loss": 0.2803, + "step": 2766 + }, + { + "epoch": 0.923873121869783, + "grad_norm": 0.4638366661214214, + "learning_rate": 6.512904808980429e-06, + "loss": 0.2868, + "step": 2767 + }, + { + "epoch": 0.9242070116861436, + "grad_norm": 0.45629956077059525, + "learning_rate": 6.510127399244235e-06, + "loss": 0.2784, + "step": 2768 + }, + { + "epoch": 0.9245409015025041, + "grad_norm": 0.4627142800080594, + "learning_rate": 6.5073494766765835e-06, + "loss": 0.2896, + "step": 2769 + }, + { + "epoch": 0.9248747913188647, + "grad_norm": 0.4700431905197205, + "learning_rate": 6.5045710422208445e-06, + "loss": 0.277, + "step": 2770 + }, + { + "epoch": 0.9252086811352254, + "grad_norm": 0.42662732022059496, + "learning_rate": 6.5017920968205575e-06, + "loss": 0.2714, + "step": 2771 + }, + { + "epoch": 0.925542570951586, + "grad_norm": 0.46599962405872236, + "learning_rate": 6.499012641419438e-06, + "loss": 0.2934, + "step": 2772 + }, + { + "epoch": 0.9258764607679466, + "grad_norm": 0.4462877432424602, + "learning_rate": 6.496232676961379e-06, + "loss": 0.2846, + "step": 2773 + }, + { + "epoch": 0.9262103505843072, + "grad_norm": 0.45066625153970835, + "learning_rate": 6.493452204390438e-06, + "loss": 0.292, + "step": 2774 + }, + { + "epoch": 0.9265442404006677, + "grad_norm": 0.46254607078887783, + "learning_rate": 6.4906712246508484e-06, + "loss": 0.2885, + "step": 2775 + }, + { + "epoch": 0.9268781302170284, + "grad_norm": 0.44534130337512035, + "learning_rate": 6.487889738687019e-06, + "loss": 0.2778, + "step": 2776 + }, + { + "epoch": 0.927212020033389, + "grad_norm": 0.44503225231832155, + "learning_rate": 6.485107747443529e-06, + "loss": 0.2871, + "step": 2777 + }, + { + "epoch": 0.9275459098497496, + "grad_norm": 0.46408300247166223, + "learning_rate": 6.482325251865125e-06, + "loss": 0.2842, + "step": 2778 + }, + { + "epoch": 0.9278797996661102, + "grad_norm": 0.4656542581813288, + "learning_rate": 6.479542252896729e-06, + "loss": 0.2887, + "step": 2779 + }, + { + "epoch": 0.9282136894824707, + "grad_norm": 0.4435372837140093, + "learning_rate": 6.476758751483435e-06, + "loss": 0.2941, + "step": 2780 + }, + { + "epoch": 0.9285475792988314, + "grad_norm": 0.4507776527802763, + "learning_rate": 6.473974748570504e-06, + "loss": 0.2765, + "step": 2781 + }, + { + "epoch": 0.928881469115192, + "grad_norm": 0.4627206063065108, + "learning_rate": 6.47119024510337e-06, + "loss": 0.2922, + "step": 2782 + }, + { + "epoch": 0.9292153589315526, + "grad_norm": 0.45361421828538806, + "learning_rate": 6.4684052420276335e-06, + "loss": 0.2767, + "step": 2783 + }, + { + "epoch": 0.9295492487479132, + "grad_norm": 0.4291039655480181, + "learning_rate": 6.465619740289071e-06, + "loss": 0.2728, + "step": 2784 + }, + { + "epoch": 0.9298831385642737, + "grad_norm": 0.420162909624095, + "learning_rate": 6.46283374083362e-06, + "loss": 0.2765, + "step": 2785 + }, + { + "epoch": 0.9302170283806344, + "grad_norm": 0.4464740544877944, + "learning_rate": 6.460047244607397e-06, + "loss": 0.2824, + "step": 2786 + }, + { + "epoch": 0.930550918196995, + "grad_norm": 0.4617451142372585, + "learning_rate": 6.457260252556677e-06, + "loss": 0.2863, + "step": 2787 + }, + { + "epoch": 0.9308848080133556, + "grad_norm": 0.4493554037714004, + "learning_rate": 6.454472765627911e-06, + "loss": 0.2827, + "step": 2788 + }, + { + "epoch": 0.9312186978297162, + "grad_norm": 0.45938862598860225, + "learning_rate": 6.451684784767714e-06, + "loss": 0.2896, + "step": 2789 + }, + { + "epoch": 0.9315525876460768, + "grad_norm": 0.44029399756870263, + "learning_rate": 6.4488963109228675e-06, + "loss": 0.287, + "step": 2790 + }, + { + "epoch": 0.9318864774624374, + "grad_norm": 0.4454271852546989, + "learning_rate": 6.446107345040328e-06, + "loss": 0.283, + "step": 2791 + }, + { + "epoch": 0.932220367278798, + "grad_norm": 0.4896046925819746, + "learning_rate": 6.443317888067207e-06, + "loss": 0.2856, + "step": 2792 + }, + { + "epoch": 0.9325542570951586, + "grad_norm": 0.42331101280250294, + "learning_rate": 6.4405279409507984e-06, + "loss": 0.2753, + "step": 2793 + }, + { + "epoch": 0.9328881469115192, + "grad_norm": 0.4379681338560488, + "learning_rate": 6.437737504638546e-06, + "loss": 0.2828, + "step": 2794 + }, + { + "epoch": 0.9332220367278798, + "grad_norm": 0.4345880558016942, + "learning_rate": 6.434946580078072e-06, + "loss": 0.279, + "step": 2795 + }, + { + "epoch": 0.9335559265442404, + "grad_norm": 0.4450185235828327, + "learning_rate": 6.432155168217158e-06, + "loss": 0.2865, + "step": 2796 + }, + { + "epoch": 0.933889816360601, + "grad_norm": 0.45079692105514974, + "learning_rate": 6.429363270003755e-06, + "loss": 0.2831, + "step": 2797 + }, + { + "epoch": 0.9342237061769616, + "grad_norm": 0.45886572381183016, + "learning_rate": 6.426570886385974e-06, + "loss": 0.283, + "step": 2798 + }, + { + "epoch": 0.9345575959933222, + "grad_norm": 0.45100835946377393, + "learning_rate": 6.423778018312097e-06, + "loss": 0.2796, + "step": 2799 + }, + { + "epoch": 0.9348914858096828, + "grad_norm": 0.4398982479513788, + "learning_rate": 6.420984666730565e-06, + "loss": 0.2763, + "step": 2800 + }, + { + "epoch": 0.9352253756260434, + "grad_norm": 0.4638678223681745, + "learning_rate": 6.4181908325899885e-06, + "loss": 0.3018, + "step": 2801 + }, + { + "epoch": 0.935559265442404, + "grad_norm": 0.4827494479159568, + "learning_rate": 6.415396516839139e-06, + "loss": 0.282, + "step": 2802 + }, + { + "epoch": 0.9358931552587646, + "grad_norm": 0.44617728882591956, + "learning_rate": 6.412601720426948e-06, + "loss": 0.2806, + "step": 2803 + }, + { + "epoch": 0.9362270450751252, + "grad_norm": 0.42484344432570503, + "learning_rate": 6.409806444302519e-06, + "loss": 0.2696, + "step": 2804 + }, + { + "epoch": 0.9365609348914858, + "grad_norm": 0.44325082574267605, + "learning_rate": 6.407010689415109e-06, + "loss": 0.2778, + "step": 2805 + }, + { + "epoch": 0.9368948247078465, + "grad_norm": 0.45452140674377206, + "learning_rate": 6.404214456714143e-06, + "loss": 0.2805, + "step": 2806 + }, + { + "epoch": 0.937228714524207, + "grad_norm": 0.4309226594190592, + "learning_rate": 6.40141774714921e-06, + "loss": 0.2907, + "step": 2807 + }, + { + "epoch": 0.9375626043405676, + "grad_norm": 0.44867036269346516, + "learning_rate": 6.3986205616700534e-06, + "loss": 0.2812, + "step": 2808 + }, + { + "epoch": 0.9378964941569282, + "grad_norm": 0.45234312663404347, + "learning_rate": 6.395822901226587e-06, + "loss": 0.2825, + "step": 2809 + }, + { + "epoch": 0.9382303839732888, + "grad_norm": 0.45208082329925675, + "learning_rate": 6.39302476676888e-06, + "loss": 0.2813, + "step": 2810 + }, + { + "epoch": 0.9385642737896495, + "grad_norm": 0.4241509605357387, + "learning_rate": 6.390226159247165e-06, + "loss": 0.2825, + "step": 2811 + }, + { + "epoch": 0.93889816360601, + "grad_norm": 0.436171197672562, + "learning_rate": 6.3874270796118345e-06, + "loss": 0.2809, + "step": 2812 + }, + { + "epoch": 0.9392320534223706, + "grad_norm": 0.4646052866291057, + "learning_rate": 6.384627528813439e-06, + "loss": 0.2864, + "step": 2813 + }, + { + "epoch": 0.9395659432387312, + "grad_norm": 0.45487353717276857, + "learning_rate": 6.381827507802696e-06, + "loss": 0.2777, + "step": 2814 + }, + { + "epoch": 0.9398998330550918, + "grad_norm": 0.4631820082322171, + "learning_rate": 6.379027017530476e-06, + "loss": 0.2831, + "step": 2815 + }, + { + "epoch": 0.9402337228714525, + "grad_norm": 0.42059680571953434, + "learning_rate": 6.376226058947811e-06, + "loss": 0.278, + "step": 2816 + }, + { + "epoch": 0.940567612687813, + "grad_norm": 0.4460216617507853, + "learning_rate": 6.373424633005889e-06, + "loss": 0.2913, + "step": 2817 + }, + { + "epoch": 0.9409015025041736, + "grad_norm": 0.44929165097819224, + "learning_rate": 6.370622740656066e-06, + "loss": 0.2907, + "step": 2818 + }, + { + "epoch": 0.9412353923205342, + "grad_norm": 0.4560960316959701, + "learning_rate": 6.367820382849847e-06, + "loss": 0.286, + "step": 2819 + }, + { + "epoch": 0.9415692821368948, + "grad_norm": 0.4355061369334617, + "learning_rate": 6.365017560538896e-06, + "loss": 0.2906, + "step": 2820 + }, + { + "epoch": 0.9419031719532555, + "grad_norm": 0.4759292178328558, + "learning_rate": 6.362214274675042e-06, + "loss": 0.279, + "step": 2821 + }, + { + "epoch": 0.9422370617696161, + "grad_norm": 0.4629434120958936, + "learning_rate": 6.359410526210259e-06, + "loss": 0.2993, + "step": 2822 + }, + { + "epoch": 0.9425709515859766, + "grad_norm": 0.44746826762866443, + "learning_rate": 6.356606316096694e-06, + "loss": 0.2807, + "step": 2823 + }, + { + "epoch": 0.9429048414023372, + "grad_norm": 0.46752417389815637, + "learning_rate": 6.353801645286636e-06, + "loss": 0.2858, + "step": 2824 + }, + { + "epoch": 0.9432387312186978, + "grad_norm": 0.46948621096367316, + "learning_rate": 6.35099651473254e-06, + "loss": 0.2853, + "step": 2825 + }, + { + "epoch": 0.9435726210350585, + "grad_norm": 0.43556530732349286, + "learning_rate": 6.348190925387012e-06, + "loss": 0.2814, + "step": 2826 + }, + { + "epoch": 0.9439065108514191, + "grad_norm": 0.45038153611316795, + "learning_rate": 6.345384878202817e-06, + "loss": 0.2942, + "step": 2827 + }, + { + "epoch": 0.9442404006677796, + "grad_norm": 0.4264166653274924, + "learning_rate": 6.342578374132874e-06, + "loss": 0.2774, + "step": 2828 + }, + { + "epoch": 0.9445742904841402, + "grad_norm": 0.48731714643599844, + "learning_rate": 6.339771414130256e-06, + "loss": 0.2842, + "step": 2829 + }, + { + "epoch": 0.9449081803005008, + "grad_norm": 0.46800897042488504, + "learning_rate": 6.336963999148193e-06, + "loss": 0.291, + "step": 2830 + }, + { + "epoch": 0.9452420701168615, + "grad_norm": 0.4340276021758347, + "learning_rate": 6.334156130140068e-06, + "loss": 0.2796, + "step": 2831 + }, + { + "epoch": 0.9455759599332221, + "grad_norm": 0.4533941267648751, + "learning_rate": 6.331347808059419e-06, + "loss": 0.281, + "step": 2832 + }, + { + "epoch": 0.9459098497495826, + "grad_norm": 0.44657288251968114, + "learning_rate": 6.328539033859939e-06, + "loss": 0.2634, + "step": 2833 + }, + { + "epoch": 0.9462437395659432, + "grad_norm": 0.47409550216601803, + "learning_rate": 6.325729808495471e-06, + "loss": 0.2798, + "step": 2834 + }, + { + "epoch": 0.9465776293823038, + "grad_norm": 0.45264915990969834, + "learning_rate": 6.3229201329200165e-06, + "loss": 0.2941, + "step": 2835 + }, + { + "epoch": 0.9469115191986645, + "grad_norm": 0.4680536725852588, + "learning_rate": 6.3201100080877235e-06, + "loss": 0.2879, + "step": 2836 + }, + { + "epoch": 0.9472454090150251, + "grad_norm": 0.4571461411461232, + "learning_rate": 6.3172994349528984e-06, + "loss": 0.2783, + "step": 2837 + }, + { + "epoch": 0.9475792988313857, + "grad_norm": 0.4483353203257524, + "learning_rate": 6.314488414469995e-06, + "loss": 0.2833, + "step": 2838 + }, + { + "epoch": 0.9479131886477462, + "grad_norm": 0.4206236071416488, + "learning_rate": 6.311676947593621e-06, + "loss": 0.2771, + "step": 2839 + }, + { + "epoch": 0.9482470784641068, + "grad_norm": 0.45643546871962043, + "learning_rate": 6.308865035278539e-06, + "loss": 0.2894, + "step": 2840 + }, + { + "epoch": 0.9485809682804675, + "grad_norm": 0.46007529180960954, + "learning_rate": 6.30605267847966e-06, + "loss": 0.2838, + "step": 2841 + }, + { + "epoch": 0.9489148580968281, + "grad_norm": 0.45588157167938104, + "learning_rate": 6.303239878152041e-06, + "loss": 0.2852, + "step": 2842 + }, + { + "epoch": 0.9492487479131887, + "grad_norm": 0.452118921442403, + "learning_rate": 6.300426635250899e-06, + "loss": 0.2835, + "step": 2843 + }, + { + "epoch": 0.9495826377295492, + "grad_norm": 0.4348604184746589, + "learning_rate": 6.297612950731596e-06, + "loss": 0.2794, + "step": 2844 + }, + { + "epoch": 0.9499165275459098, + "grad_norm": 0.45597428613482427, + "learning_rate": 6.294798825549643e-06, + "loss": 0.275, + "step": 2845 + }, + { + "epoch": 0.9502504173622704, + "grad_norm": 0.4416953951288915, + "learning_rate": 6.291984260660705e-06, + "loss": 0.2729, + "step": 2846 + }, + { + "epoch": 0.9505843071786311, + "grad_norm": 0.44326552226815086, + "learning_rate": 6.28916925702059e-06, + "loss": 0.2911, + "step": 2847 + }, + { + "epoch": 0.9509181969949917, + "grad_norm": 0.4365780756722261, + "learning_rate": 6.286353815585262e-06, + "loss": 0.2817, + "step": 2848 + }, + { + "epoch": 0.9512520868113522, + "grad_norm": 0.4358677312972007, + "learning_rate": 6.283537937310829e-06, + "loss": 0.2791, + "step": 2849 + }, + { + "epoch": 0.9515859766277128, + "grad_norm": 0.4645363611359433, + "learning_rate": 6.280721623153549e-06, + "loss": 0.2831, + "step": 2850 + }, + { + "epoch": 0.9519198664440734, + "grad_norm": 0.4359643733148081, + "learning_rate": 6.277904874069827e-06, + "loss": 0.2813, + "step": 2851 + }, + { + "epoch": 0.9522537562604341, + "grad_norm": 0.4599024624711476, + "learning_rate": 6.275087691016217e-06, + "loss": 0.2742, + "step": 2852 + }, + { + "epoch": 0.9525876460767947, + "grad_norm": 0.4226000277953542, + "learning_rate": 6.272270074949421e-06, + "loss": 0.2708, + "step": 2853 + }, + { + "epoch": 0.9529215358931553, + "grad_norm": 0.43414031506918715, + "learning_rate": 6.269452026826285e-06, + "loss": 0.2822, + "step": 2854 + }, + { + "epoch": 0.9532554257095158, + "grad_norm": 0.4512122216721651, + "learning_rate": 6.266633547603804e-06, + "loss": 0.2819, + "step": 2855 + }, + { + "epoch": 0.9535893155258764, + "grad_norm": 0.428668150358751, + "learning_rate": 6.263814638239118e-06, + "loss": 0.2826, + "step": 2856 + }, + { + "epoch": 0.9539232053422371, + "grad_norm": 0.4740616818154354, + "learning_rate": 6.2609952996895185e-06, + "loss": 0.2782, + "step": 2857 + }, + { + "epoch": 0.9542570951585977, + "grad_norm": 0.42856089813796955, + "learning_rate": 6.258175532912432e-06, + "loss": 0.2755, + "step": 2858 + }, + { + "epoch": 0.9545909849749583, + "grad_norm": 0.43735496754191433, + "learning_rate": 6.255355338865442e-06, + "loss": 0.2848, + "step": 2859 + }, + { + "epoch": 0.9549248747913188, + "grad_norm": 0.4512791437274907, + "learning_rate": 6.252534718506268e-06, + "loss": 0.2731, + "step": 2860 + }, + { + "epoch": 0.9552587646076794, + "grad_norm": 0.4516414199725579, + "learning_rate": 6.24971367279278e-06, + "loss": 0.2853, + "step": 2861 + }, + { + "epoch": 0.9555926544240401, + "grad_norm": 0.4356576784038092, + "learning_rate": 6.24689220268299e-06, + "loss": 0.2763, + "step": 2862 + }, + { + "epoch": 0.9559265442404007, + "grad_norm": 0.4161099537943426, + "learning_rate": 6.244070309135055e-06, + "loss": 0.2663, + "step": 2863 + }, + { + "epoch": 0.9562604340567613, + "grad_norm": 0.4484128310103563, + "learning_rate": 6.2412479931072754e-06, + "loss": 0.2989, + "step": 2864 + }, + { + "epoch": 0.9565943238731218, + "grad_norm": 0.44651296827100534, + "learning_rate": 6.238425255558095e-06, + "loss": 0.2672, + "step": 2865 + }, + { + "epoch": 0.9569282136894824, + "grad_norm": 0.4493091435262233, + "learning_rate": 6.235602097446102e-06, + "loss": 0.2782, + "step": 2866 + }, + { + "epoch": 0.9572621035058431, + "grad_norm": 0.42683556605926837, + "learning_rate": 6.232778519730024e-06, + "loss": 0.279, + "step": 2867 + }, + { + "epoch": 0.9575959933222037, + "grad_norm": 0.4230924240166505, + "learning_rate": 6.2299545233687346e-06, + "loss": 0.2719, + "step": 2868 + }, + { + "epoch": 0.9579298831385643, + "grad_norm": 0.42430586290078576, + "learning_rate": 6.227130109321249e-06, + "loss": 0.2784, + "step": 2869 + }, + { + "epoch": 0.9582637729549248, + "grad_norm": 0.4532687710205164, + "learning_rate": 6.224305278546723e-06, + "loss": 0.2913, + "step": 2870 + }, + { + "epoch": 0.9585976627712854, + "grad_norm": 0.4233497699021432, + "learning_rate": 6.221480032004455e-06, + "loss": 0.2809, + "step": 2871 + }, + { + "epoch": 0.9589315525876461, + "grad_norm": 0.42724256866439614, + "learning_rate": 6.218654370653881e-06, + "loss": 0.2799, + "step": 2872 + }, + { + "epoch": 0.9592654424040067, + "grad_norm": 0.4249316600592112, + "learning_rate": 6.215828295454585e-06, + "loss": 0.2811, + "step": 2873 + }, + { + "epoch": 0.9595993322203673, + "grad_norm": 0.45911821962303595, + "learning_rate": 6.213001807366287e-06, + "loss": 0.2877, + "step": 2874 + }, + { + "epoch": 0.9599332220367279, + "grad_norm": 0.45600397165485235, + "learning_rate": 6.210174907348847e-06, + "loss": 0.2968, + "step": 2875 + }, + { + "epoch": 0.9602671118530884, + "grad_norm": 0.43591899197807127, + "learning_rate": 6.207347596362265e-06, + "loss": 0.2772, + "step": 2876 + }, + { + "epoch": 0.9606010016694491, + "grad_norm": 0.4647608869411942, + "learning_rate": 6.204519875366681e-06, + "loss": 0.2905, + "step": 2877 + }, + { + "epoch": 0.9609348914858097, + "grad_norm": 0.46424199943327576, + "learning_rate": 6.201691745322377e-06, + "loss": 0.2853, + "step": 2878 + }, + { + "epoch": 0.9612687813021703, + "grad_norm": 0.47371996555228324, + "learning_rate": 6.198863207189768e-06, + "loss": 0.2944, + "step": 2879 + }, + { + "epoch": 0.9616026711185309, + "grad_norm": 0.4311334885060702, + "learning_rate": 6.196034261929413e-06, + "loss": 0.2757, + "step": 2880 + }, + { + "epoch": 0.9619365609348914, + "grad_norm": 0.4329962095906153, + "learning_rate": 6.193204910502006e-06, + "loss": 0.2846, + "step": 2881 + }, + { + "epoch": 0.9622704507512521, + "grad_norm": 0.47429556762737884, + "learning_rate": 6.190375153868381e-06, + "loss": 0.2953, + "step": 2882 + }, + { + "epoch": 0.9626043405676127, + "grad_norm": 0.471683916936054, + "learning_rate": 6.187544992989508e-06, + "loss": 0.2858, + "step": 2883 + }, + { + "epoch": 0.9629382303839733, + "grad_norm": 0.4606291949452427, + "learning_rate": 6.184714428826496e-06, + "loss": 0.2982, + "step": 2884 + }, + { + "epoch": 0.9632721202003339, + "grad_norm": 0.4408430206193836, + "learning_rate": 6.181883462340589e-06, + "loss": 0.2777, + "step": 2885 + }, + { + "epoch": 0.9636060100166944, + "grad_norm": 0.44057406888246214, + "learning_rate": 6.179052094493169e-06, + "loss": 0.2736, + "step": 2886 + }, + { + "epoch": 0.9639398998330551, + "grad_norm": 0.46212511878995116, + "learning_rate": 6.176220326245754e-06, + "loss": 0.2817, + "step": 2887 + }, + { + "epoch": 0.9642737896494157, + "grad_norm": 0.4406045251970601, + "learning_rate": 6.1733881585599985e-06, + "loss": 0.2782, + "step": 2888 + }, + { + "epoch": 0.9646076794657763, + "grad_norm": 0.45824901716589267, + "learning_rate": 6.170555592397691e-06, + "loss": 0.2902, + "step": 2889 + }, + { + "epoch": 0.9649415692821369, + "grad_norm": 0.46823936723886556, + "learning_rate": 6.1677226287207545e-06, + "loss": 0.2874, + "step": 2890 + }, + { + "epoch": 0.9652754590984975, + "grad_norm": 0.47596421395733496, + "learning_rate": 6.164889268491253e-06, + "loss": 0.2812, + "step": 2891 + }, + { + "epoch": 0.9656093489148581, + "grad_norm": 0.45448249972740057, + "learning_rate": 6.162055512671378e-06, + "loss": 0.2862, + "step": 2892 + }, + { + "epoch": 0.9659432387312187, + "grad_norm": 0.45475439382728483, + "learning_rate": 6.15922136222346e-06, + "loss": 0.2802, + "step": 2893 + }, + { + "epoch": 0.9662771285475793, + "grad_norm": 0.42002283859398, + "learning_rate": 6.1563868181099596e-06, + "loss": 0.2749, + "step": 2894 + }, + { + "epoch": 0.9666110183639399, + "grad_norm": 0.4402890860898216, + "learning_rate": 6.153551881293475e-06, + "loss": 0.2811, + "step": 2895 + }, + { + "epoch": 0.9669449081803005, + "grad_norm": 0.4508090456866796, + "learning_rate": 6.1507165527367376e-06, + "loss": 0.2915, + "step": 2896 + }, + { + "epoch": 0.967278797996661, + "grad_norm": 0.48911629645004057, + "learning_rate": 6.147880833402606e-06, + "loss": 0.2879, + "step": 2897 + }, + { + "epoch": 0.9676126878130217, + "grad_norm": 0.4409751372381057, + "learning_rate": 6.145044724254079e-06, + "loss": 0.2849, + "step": 2898 + }, + { + "epoch": 0.9679465776293823, + "grad_norm": 0.450406206221538, + "learning_rate": 6.142208226254285e-06, + "loss": 0.2897, + "step": 2899 + }, + { + "epoch": 0.9682804674457429, + "grad_norm": 0.4454873353555689, + "learning_rate": 6.139371340366482e-06, + "loss": 0.2917, + "step": 2900 + }, + { + "epoch": 0.9686143572621035, + "grad_norm": 0.46414075746578104, + "learning_rate": 6.136534067554063e-06, + "loss": 0.2869, + "step": 2901 + }, + { + "epoch": 0.968948247078464, + "grad_norm": 0.4416875705312836, + "learning_rate": 6.133696408780551e-06, + "loss": 0.2838, + "step": 2902 + }, + { + "epoch": 0.9692821368948247, + "grad_norm": 0.43830943497559227, + "learning_rate": 6.130858365009601e-06, + "loss": 0.2807, + "step": 2903 + }, + { + "epoch": 0.9696160267111853, + "grad_norm": 0.4439209148989399, + "learning_rate": 6.128019937204997e-06, + "loss": 0.2765, + "step": 2904 + }, + { + "epoch": 0.9699499165275459, + "grad_norm": 0.471642113443311, + "learning_rate": 6.1251811263306535e-06, + "loss": 0.2977, + "step": 2905 + }, + { + "epoch": 0.9702838063439065, + "grad_norm": 0.47212437671251734, + "learning_rate": 6.1223419333506185e-06, + "loss": 0.2914, + "step": 2906 + }, + { + "epoch": 0.9706176961602672, + "grad_norm": 0.47326219603572645, + "learning_rate": 6.119502359229065e-06, + "loss": 0.279, + "step": 2907 + }, + { + "epoch": 0.9709515859766277, + "grad_norm": 0.4098758053774518, + "learning_rate": 6.116662404930298e-06, + "loss": 0.2624, + "step": 2908 + }, + { + "epoch": 0.9712854757929883, + "grad_norm": 0.4518529345523616, + "learning_rate": 6.1138220714187535e-06, + "loss": 0.2636, + "step": 2909 + }, + { + "epoch": 0.9716193656093489, + "grad_norm": 0.477216087186937, + "learning_rate": 6.110981359658992e-06, + "loss": 0.2872, + "step": 2910 + }, + { + "epoch": 0.9719532554257095, + "grad_norm": 0.4549472880481955, + "learning_rate": 6.108140270615702e-06, + "loss": 0.2856, + "step": 2911 + }, + { + "epoch": 0.9722871452420702, + "grad_norm": 0.4431754136285394, + "learning_rate": 6.105298805253709e-06, + "loss": 0.2806, + "step": 2912 + }, + { + "epoch": 0.9726210350584307, + "grad_norm": 0.43277388015102375, + "learning_rate": 6.102456964537955e-06, + "loss": 0.2769, + "step": 2913 + }, + { + "epoch": 0.9729549248747913, + "grad_norm": 0.4363914580196727, + "learning_rate": 6.0996147494335174e-06, + "loss": 0.2805, + "step": 2914 + }, + { + "epoch": 0.9732888146911519, + "grad_norm": 0.44465972336168913, + "learning_rate": 6.0967721609055956e-06, + "loss": 0.2882, + "step": 2915 + }, + { + "epoch": 0.9736227045075125, + "grad_norm": 0.43182882898693364, + "learning_rate": 6.093929199919517e-06, + "loss": 0.283, + "step": 2916 + }, + { + "epoch": 0.9739565943238732, + "grad_norm": 0.46318149884407284, + "learning_rate": 6.0910858674407406e-06, + "loss": 0.2972, + "step": 2917 + }, + { + "epoch": 0.9742904841402337, + "grad_norm": 0.427203303306548, + "learning_rate": 6.0882421644348435e-06, + "loss": 0.2785, + "step": 2918 + }, + { + "epoch": 0.9746243739565943, + "grad_norm": 0.4472550223522689, + "learning_rate": 6.085398091867534e-06, + "loss": 0.2819, + "step": 2919 + }, + { + "epoch": 0.9749582637729549, + "grad_norm": 0.44879249909757324, + "learning_rate": 6.082553650704644e-06, + "loss": 0.2644, + "step": 2920 + }, + { + "epoch": 0.9752921535893155, + "grad_norm": 0.4224246849187929, + "learning_rate": 6.079708841912133e-06, + "loss": 0.2792, + "step": 2921 + }, + { + "epoch": 0.9756260434056762, + "grad_norm": 0.41553766089998584, + "learning_rate": 6.076863666456081e-06, + "loss": 0.2654, + "step": 2922 + }, + { + "epoch": 0.9759599332220368, + "grad_norm": 0.44301757040251905, + "learning_rate": 6.074018125302697e-06, + "loss": 0.2758, + "step": 2923 + }, + { + "epoch": 0.9762938230383973, + "grad_norm": 0.43417290056965196, + "learning_rate": 6.071172219418309e-06, + "loss": 0.285, + "step": 2924 + }, + { + "epoch": 0.9766277128547579, + "grad_norm": 0.43627864656767495, + "learning_rate": 6.0683259497693736e-06, + "loss": 0.274, + "step": 2925 + }, + { + "epoch": 0.9769616026711185, + "grad_norm": 0.4262864710894089, + "learning_rate": 6.0654793173224706e-06, + "loss": 0.2914, + "step": 2926 + }, + { + "epoch": 0.9772954924874792, + "grad_norm": 0.4476570110789471, + "learning_rate": 6.062632323044299e-06, + "loss": 0.2737, + "step": 2927 + }, + { + "epoch": 0.9776293823038398, + "grad_norm": 0.4350706905537517, + "learning_rate": 6.059784967901684e-06, + "loss": 0.2853, + "step": 2928 + }, + { + "epoch": 0.9779632721202003, + "grad_norm": 0.4446066285849519, + "learning_rate": 6.056937252861574e-06, + "loss": 0.2663, + "step": 2929 + }, + { + "epoch": 0.9782971619365609, + "grad_norm": 0.44236882353953616, + "learning_rate": 6.054089178891039e-06, + "loss": 0.2731, + "step": 2930 + }, + { + "epoch": 0.9786310517529215, + "grad_norm": 0.453443264432735, + "learning_rate": 6.0512407469572675e-06, + "loss": 0.2804, + "step": 2931 + }, + { + "epoch": 0.9789649415692822, + "grad_norm": 0.4294055926076345, + "learning_rate": 6.048391958027574e-06, + "loss": 0.266, + "step": 2932 + }, + { + "epoch": 0.9792988313856428, + "grad_norm": 0.42660336162064283, + "learning_rate": 6.045542813069394e-06, + "loss": 0.2815, + "step": 2933 + }, + { + "epoch": 0.9796327212020033, + "grad_norm": 0.4465774873791782, + "learning_rate": 6.0426933130502786e-06, + "loss": 0.2813, + "step": 2934 + }, + { + "epoch": 0.9799666110183639, + "grad_norm": 0.4493960696428338, + "learning_rate": 6.039843458937905e-06, + "loss": 0.2867, + "step": 2935 + }, + { + "epoch": 0.9803005008347245, + "grad_norm": 0.4401804835180483, + "learning_rate": 6.036993251700069e-06, + "loss": 0.2877, + "step": 2936 + }, + { + "epoch": 0.9806343906510852, + "grad_norm": 0.4607977649751314, + "learning_rate": 6.034142692304688e-06, + "loss": 0.3047, + "step": 2937 + }, + { + "epoch": 0.9809682804674458, + "grad_norm": 0.42609400025084665, + "learning_rate": 6.031291781719793e-06, + "loss": 0.2786, + "step": 2938 + }, + { + "epoch": 0.9813021702838064, + "grad_norm": 0.44419853700605294, + "learning_rate": 6.028440520913545e-06, + "loss": 0.2794, + "step": 2939 + }, + { + "epoch": 0.9816360601001669, + "grad_norm": 0.4341157069206071, + "learning_rate": 6.02558891085421e-06, + "loss": 0.2695, + "step": 2940 + }, + { + "epoch": 0.9819699499165275, + "grad_norm": 0.45559591892817297, + "learning_rate": 6.022736952510184e-06, + "loss": 0.2798, + "step": 2941 + }, + { + "epoch": 0.9823038397328882, + "grad_norm": 0.43719467066966994, + "learning_rate": 6.019884646849978e-06, + "loss": 0.2715, + "step": 2942 + }, + { + "epoch": 0.9826377295492488, + "grad_norm": 0.4225692939558709, + "learning_rate": 6.017031994842218e-06, + "loss": 0.2748, + "step": 2943 + }, + { + "epoch": 0.9829716193656094, + "grad_norm": 0.42783970926018405, + "learning_rate": 6.014178997455653e-06, + "loss": 0.2777, + "step": 2944 + }, + { + "epoch": 0.9833055091819699, + "grad_norm": 0.4438913495402183, + "learning_rate": 6.0113256556591415e-06, + "loss": 0.2759, + "step": 2945 + }, + { + "epoch": 0.9836393989983305, + "grad_norm": 0.4421330874482738, + "learning_rate": 6.008471970421667e-06, + "loss": 0.2743, + "step": 2946 + }, + { + "epoch": 0.9839732888146911, + "grad_norm": 0.449047135488092, + "learning_rate": 6.0056179427123266e-06, + "loss": 0.2879, + "step": 2947 + }, + { + "epoch": 0.9843071786310518, + "grad_norm": 0.46116493823587607, + "learning_rate": 6.002763573500332e-06, + "loss": 0.287, + "step": 2948 + }, + { + "epoch": 0.9846410684474124, + "grad_norm": 0.42950697846418323, + "learning_rate": 5.9999088637550126e-06, + "loss": 0.2812, + "step": 2949 + }, + { + "epoch": 0.9849749582637729, + "grad_norm": 0.429243384169998, + "learning_rate": 5.997053814445812e-06, + "loss": 0.2753, + "step": 2950 + }, + { + "epoch": 0.9853088480801335, + "grad_norm": 0.42893737081650535, + "learning_rate": 5.994198426542293e-06, + "loss": 0.2727, + "step": 2951 + }, + { + "epoch": 0.9856427378964941, + "grad_norm": 0.4369600931820618, + "learning_rate": 5.991342701014127e-06, + "loss": 0.2722, + "step": 2952 + }, + { + "epoch": 0.9859766277128548, + "grad_norm": 0.4330961710845029, + "learning_rate": 5.988486638831105e-06, + "loss": 0.275, + "step": 2953 + }, + { + "epoch": 0.9863105175292154, + "grad_norm": 0.4346644950932211, + "learning_rate": 5.985630240963132e-06, + "loss": 0.2757, + "step": 2954 + }, + { + "epoch": 0.986644407345576, + "grad_norm": 0.42910115637722973, + "learning_rate": 5.982773508380226e-06, + "loss": 0.2884, + "step": 2955 + }, + { + "epoch": 0.9869782971619365, + "grad_norm": 0.44573623086773706, + "learning_rate": 5.979916442052517e-06, + "loss": 0.2892, + "step": 2956 + }, + { + "epoch": 0.9873121869782971, + "grad_norm": 0.43768228959212085, + "learning_rate": 5.977059042950252e-06, + "loss": 0.2827, + "step": 2957 + }, + { + "epoch": 0.9876460767946578, + "grad_norm": 0.42963286016203067, + "learning_rate": 5.974201312043786e-06, + "loss": 0.2736, + "step": 2958 + }, + { + "epoch": 0.9879799666110184, + "grad_norm": 0.4412309467421957, + "learning_rate": 5.971343250303592e-06, + "loss": 0.2716, + "step": 2959 + }, + { + "epoch": 0.988313856427379, + "grad_norm": 0.43474037120555264, + "learning_rate": 5.968484858700253e-06, + "loss": 0.279, + "step": 2960 + }, + { + "epoch": 0.9886477462437395, + "grad_norm": 0.43623570080823043, + "learning_rate": 5.965626138204463e-06, + "loss": 0.2833, + "step": 2961 + }, + { + "epoch": 0.9889816360601001, + "grad_norm": 0.4439402790834901, + "learning_rate": 5.9627670897870285e-06, + "loss": 0.2849, + "step": 2962 + }, + { + "epoch": 0.9893155258764608, + "grad_norm": 0.47954642783051965, + "learning_rate": 5.959907714418868e-06, + "loss": 0.2941, + "step": 2963 + }, + { + "epoch": 0.9896494156928214, + "grad_norm": 0.4404325511356687, + "learning_rate": 5.957048013071013e-06, + "loss": 0.2865, + "step": 2964 + }, + { + "epoch": 0.989983305509182, + "grad_norm": 0.440918704532239, + "learning_rate": 5.954187986714599e-06, + "loss": 0.2759, + "step": 2965 + }, + { + "epoch": 0.9903171953255425, + "grad_norm": 0.4714351937324036, + "learning_rate": 5.951327636320878e-06, + "loss": 0.2798, + "step": 2966 + }, + { + "epoch": 0.9906510851419031, + "grad_norm": 0.4596060909779612, + "learning_rate": 5.948466962861212e-06, + "loss": 0.284, + "step": 2967 + }, + { + "epoch": 0.9909849749582638, + "grad_norm": 0.4444320704921734, + "learning_rate": 5.9456059673070685e-06, + "loss": 0.2767, + "step": 2968 + }, + { + "epoch": 0.9913188647746244, + "grad_norm": 0.43544789745291884, + "learning_rate": 5.942744650630028e-06, + "loss": 0.2824, + "step": 2969 + }, + { + "epoch": 0.991652754590985, + "grad_norm": 0.43789595973183293, + "learning_rate": 5.9398830138017774e-06, + "loss": 0.2738, + "step": 2970 + }, + { + "epoch": 0.9919866444073456, + "grad_norm": 0.4281751697560679, + "learning_rate": 5.937021057794116e-06, + "loss": 0.2833, + "step": 2971 + }, + { + "epoch": 0.9923205342237061, + "grad_norm": 0.44805771376572523, + "learning_rate": 5.934158783578947e-06, + "loss": 0.2851, + "step": 2972 + }, + { + "epoch": 0.9926544240400668, + "grad_norm": 0.46991904509498955, + "learning_rate": 5.9312961921282866e-06, + "loss": 0.2815, + "step": 2973 + }, + { + "epoch": 0.9929883138564274, + "grad_norm": 0.44498415267147734, + "learning_rate": 5.928433284414253e-06, + "loss": 0.2745, + "step": 2974 + }, + { + "epoch": 0.993322203672788, + "grad_norm": 0.43231462531898945, + "learning_rate": 5.925570061409077e-06, + "loss": 0.2838, + "step": 2975 + }, + { + "epoch": 0.9936560934891486, + "grad_norm": 0.4342909803920257, + "learning_rate": 5.922706524085094e-06, + "loss": 0.2841, + "step": 2976 + }, + { + "epoch": 0.9939899833055091, + "grad_norm": 0.4632335062374389, + "learning_rate": 5.919842673414746e-06, + "loss": 0.27, + "step": 2977 + }, + { + "epoch": 0.9943238731218698, + "grad_norm": 0.4971475287125927, + "learning_rate": 5.9169785103705824e-06, + "loss": 0.2806, + "step": 2978 + }, + { + "epoch": 0.9946577629382304, + "grad_norm": 0.46737399426148235, + "learning_rate": 5.914114035925256e-06, + "loss": 0.2897, + "step": 2979 + }, + { + "epoch": 0.994991652754591, + "grad_norm": 0.44361345517475026, + "learning_rate": 5.911249251051532e-06, + "loss": 0.2801, + "step": 2980 + }, + { + "epoch": 0.9953255425709516, + "grad_norm": 0.4386193546171152, + "learning_rate": 5.908384156722271e-06, + "loss": 0.2748, + "step": 2981 + }, + { + "epoch": 0.9956594323873121, + "grad_norm": 0.4457823420021135, + "learning_rate": 5.90551875391045e-06, + "loss": 0.2781, + "step": 2982 + }, + { + "epoch": 0.9959933222036728, + "grad_norm": 0.49624653541718283, + "learning_rate": 5.9026530435891405e-06, + "loss": 0.2879, + "step": 2983 + }, + { + "epoch": 0.9963272120200334, + "grad_norm": 0.445027285918299, + "learning_rate": 5.899787026731524e-06, + "loss": 0.2801, + "step": 2984 + }, + { + "epoch": 0.996661101836394, + "grad_norm": 0.4792640860092878, + "learning_rate": 5.8969207043108865e-06, + "loss": 0.2828, + "step": 2985 + }, + { + "epoch": 0.9969949916527546, + "grad_norm": 0.485193817477391, + "learning_rate": 5.894054077300614e-06, + "loss": 0.2938, + "step": 2986 + }, + { + "epoch": 0.9973288814691152, + "grad_norm": 0.45409889074534243, + "learning_rate": 5.8911871466742e-06, + "loss": 0.2772, + "step": 2987 + }, + { + "epoch": 0.9976627712854758, + "grad_norm": 0.42105953632209353, + "learning_rate": 5.888319913405238e-06, + "loss": 0.2749, + "step": 2988 + }, + { + "epoch": 0.9979966611018364, + "grad_norm": 0.4852529534598636, + "learning_rate": 5.885452378467429e-06, + "loss": 0.2779, + "step": 2989 + }, + { + "epoch": 0.998330550918197, + "grad_norm": 0.4612250035108816, + "learning_rate": 5.8825845428345674e-06, + "loss": 0.2864, + "step": 2990 + }, + { + "epoch": 0.9986644407345576, + "grad_norm": 0.4513524833741755, + "learning_rate": 5.879716407480559e-06, + "loss": 0.2882, + "step": 2991 + }, + { + "epoch": 0.9989983305509182, + "grad_norm": 0.4638846696057987, + "learning_rate": 5.876847973379409e-06, + "loss": 0.2744, + "step": 2992 + }, + { + "epoch": 0.9993322203672788, + "grad_norm": 0.43122240305314136, + "learning_rate": 5.873979241505219e-06, + "loss": 0.2776, + "step": 2993 + }, + { + "epoch": 0.9996661101836394, + "grad_norm": 0.46756526505481033, + "learning_rate": 5.871110212832199e-06, + "loss": 0.2924, + "step": 2994 + }, + { + "epoch": 1.0, + "grad_norm": 0.44450699964228596, + "learning_rate": 5.8682408883346535e-06, + "loss": 0.2817, + "step": 2995 + }, + { + "epoch": 1.0, + "eval_loss": 0.2817595899105072, + "eval_runtime": 184.6484, + "eval_samples_per_second": 109.278, + "eval_steps_per_second": 0.856, + "step": 2995 + }, + { + "epoch": 1.0003338898163605, + "grad_norm": 0.44960559750987666, + "learning_rate": 5.86537126898699e-06, + "loss": 0.2545, + "step": 2996 + }, + { + "epoch": 1.0006677796327212, + "grad_norm": 0.4437018059155895, + "learning_rate": 5.862501355763719e-06, + "loss": 0.2606, + "step": 2997 + }, + { + "epoch": 1.0010016694490818, + "grad_norm": 0.4413237077907396, + "learning_rate": 5.859631149639449e-06, + "loss": 0.2713, + "step": 2998 + }, + { + "epoch": 1.0013355592654425, + "grad_norm": 0.4368017638592634, + "learning_rate": 5.856760651588882e-06, + "loss": 0.2614, + "step": 2999 + }, + { + "epoch": 1.001669449081803, + "grad_norm": 0.4281093123380615, + "learning_rate": 5.853889862586829e-06, + "loss": 0.2555, + "step": 3000 + }, + { + "epoch": 1.0020033388981635, + "grad_norm": 0.4422745628935671, + "learning_rate": 5.851018783608195e-06, + "loss": 0.2577, + "step": 3001 + }, + { + "epoch": 1.0023372287145242, + "grad_norm": 0.44795744173308877, + "learning_rate": 5.848147415627981e-06, + "loss": 0.2526, + "step": 3002 + }, + { + "epoch": 1.0026711185308848, + "grad_norm": 0.47550862760142015, + "learning_rate": 5.84527575962129e-06, + "loss": 0.269, + "step": 3003 + }, + { + "epoch": 1.0030050083472455, + "grad_norm": 0.4502890521909984, + "learning_rate": 5.842403816563319e-06, + "loss": 0.247, + "step": 3004 + }, + { + "epoch": 1.003338898163606, + "grad_norm": 0.46180607326177975, + "learning_rate": 5.83953158742937e-06, + "loss": 0.2574, + "step": 3005 + }, + { + "epoch": 1.0036727879799665, + "grad_norm": 0.455488644156617, + "learning_rate": 5.836659073194835e-06, + "loss": 0.2432, + "step": 3006 + }, + { + "epoch": 1.0040066777963272, + "grad_norm": 0.4655480268895469, + "learning_rate": 5.833786274835204e-06, + "loss": 0.2584, + "step": 3007 + }, + { + "epoch": 1.0043405676126878, + "grad_norm": 0.42041366144205244, + "learning_rate": 5.830913193326064e-06, + "loss": 0.2502, + "step": 3008 + }, + { + "epoch": 1.0046744574290485, + "grad_norm": 0.4460649264196421, + "learning_rate": 5.828039829643101e-06, + "loss": 0.2542, + "step": 3009 + }, + { + "epoch": 1.005008347245409, + "grad_norm": 0.448832009458717, + "learning_rate": 5.825166184762092e-06, + "loss": 0.2584, + "step": 3010 + }, + { + "epoch": 1.0053422370617695, + "grad_norm": 0.4570163345857488, + "learning_rate": 5.822292259658914e-06, + "loss": 0.2557, + "step": 3011 + }, + { + "epoch": 1.0056761268781302, + "grad_norm": 0.44066343667971924, + "learning_rate": 5.819418055309536e-06, + "loss": 0.2611, + "step": 3012 + }, + { + "epoch": 1.0060100166944907, + "grad_norm": 0.4261077632743608, + "learning_rate": 5.816543572690023e-06, + "loss": 0.2439, + "step": 3013 + }, + { + "epoch": 1.0063439065108515, + "grad_norm": 0.4554180376662268, + "learning_rate": 5.813668812776534e-06, + "loss": 0.2556, + "step": 3014 + }, + { + "epoch": 1.006677796327212, + "grad_norm": 0.46650470607584127, + "learning_rate": 5.810793776545325e-06, + "loss": 0.2622, + "step": 3015 + }, + { + "epoch": 1.0070116861435727, + "grad_norm": 0.4662632932129285, + "learning_rate": 5.807918464972739e-06, + "loss": 0.2649, + "step": 3016 + }, + { + "epoch": 1.0073455759599332, + "grad_norm": 0.42864003753914937, + "learning_rate": 5.805042879035221e-06, + "loss": 0.257, + "step": 3017 + }, + { + "epoch": 1.0076794657762937, + "grad_norm": 0.43653071580441044, + "learning_rate": 5.802167019709303e-06, + "loss": 0.246, + "step": 3018 + }, + { + "epoch": 1.0080133555926545, + "grad_norm": 0.4678799191115644, + "learning_rate": 5.799290887971614e-06, + "loss": 0.2571, + "step": 3019 + }, + { + "epoch": 1.008347245409015, + "grad_norm": 0.4512010745071946, + "learning_rate": 5.79641448479887e-06, + "loss": 0.2552, + "step": 3020 + }, + { + "epoch": 1.0086811352253757, + "grad_norm": 0.4417518841155561, + "learning_rate": 5.793537811167885e-06, + "loss": 0.2528, + "step": 3021 + }, + { + "epoch": 1.0090150250417362, + "grad_norm": 0.450626865301417, + "learning_rate": 5.790660868055564e-06, + "loss": 0.2551, + "step": 3022 + }, + { + "epoch": 1.0093489148580967, + "grad_norm": 0.4519104402934889, + "learning_rate": 5.787783656438899e-06, + "loss": 0.2508, + "step": 3023 + }, + { + "epoch": 1.0096828046744575, + "grad_norm": 0.438773817411675, + "learning_rate": 5.7849061772949785e-06, + "loss": 0.2455, + "step": 3024 + }, + { + "epoch": 1.010016694490818, + "grad_norm": 0.4679306757579102, + "learning_rate": 5.782028431600977e-06, + "loss": 0.2602, + "step": 3025 + }, + { + "epoch": 1.0103505843071787, + "grad_norm": 0.4545912810576913, + "learning_rate": 5.7791504203341645e-06, + "loss": 0.2598, + "step": 3026 + }, + { + "epoch": 1.0106844741235392, + "grad_norm": 0.44660182138407545, + "learning_rate": 5.776272144471896e-06, + "loss": 0.2558, + "step": 3027 + }, + { + "epoch": 1.0110183639398997, + "grad_norm": 0.47753956115613144, + "learning_rate": 5.773393604991623e-06, + "loss": 0.2494, + "step": 3028 + }, + { + "epoch": 1.0113522537562605, + "grad_norm": 0.43822636989091646, + "learning_rate": 5.770514802870879e-06, + "loss": 0.2559, + "step": 3029 + }, + { + "epoch": 1.011686143572621, + "grad_norm": 0.45825272419163166, + "learning_rate": 5.7676357390872925e-06, + "loss": 0.2585, + "step": 3030 + }, + { + "epoch": 1.0120200333889817, + "grad_norm": 0.46462282915679914, + "learning_rate": 5.764756414618579e-06, + "loss": 0.2546, + "step": 3031 + }, + { + "epoch": 1.0123539232053422, + "grad_norm": 0.5023785167422805, + "learning_rate": 5.76187683044254e-06, + "loss": 0.2732, + "step": 3032 + }, + { + "epoch": 1.0126878130217027, + "grad_norm": 0.4500090676139374, + "learning_rate": 5.7589969875370685e-06, + "loss": 0.2629, + "step": 3033 + }, + { + "epoch": 1.0130217028380635, + "grad_norm": 0.4402752915573057, + "learning_rate": 5.756116886880143e-06, + "loss": 0.2459, + "step": 3034 + }, + { + "epoch": 1.013355592654424, + "grad_norm": 0.4704533023271832, + "learning_rate": 5.753236529449835e-06, + "loss": 0.258, + "step": 3035 + }, + { + "epoch": 1.0136894824707847, + "grad_norm": 0.44372271644046896, + "learning_rate": 5.750355916224294e-06, + "loss": 0.2571, + "step": 3036 + }, + { + "epoch": 1.0140233722871452, + "grad_norm": 0.4962160614966293, + "learning_rate": 5.747475048181765e-06, + "loss": 0.2642, + "step": 3037 + }, + { + "epoch": 1.0143572621035057, + "grad_norm": 0.45225224347285575, + "learning_rate": 5.744593926300573e-06, + "loss": 0.2529, + "step": 3038 + }, + { + "epoch": 1.0146911519198665, + "grad_norm": 0.4650191744890438, + "learning_rate": 5.741712551559133e-06, + "loss": 0.2591, + "step": 3039 + }, + { + "epoch": 1.015025041736227, + "grad_norm": 0.4738201976723723, + "learning_rate": 5.738830924935948e-06, + "loss": 0.2673, + "step": 3040 + }, + { + "epoch": 1.0153589315525877, + "grad_norm": 0.4610517063401495, + "learning_rate": 5.735949047409599e-06, + "loss": 0.2496, + "step": 3041 + }, + { + "epoch": 1.0156928213689482, + "grad_norm": 0.44478170785663085, + "learning_rate": 5.7330669199587585e-06, + "loss": 0.2606, + "step": 3042 + }, + { + "epoch": 1.0160267111853087, + "grad_norm": 0.4330903963151855, + "learning_rate": 5.730184543562183e-06, + "loss": 0.2488, + "step": 3043 + }, + { + "epoch": 1.0163606010016695, + "grad_norm": 0.4431853544619894, + "learning_rate": 5.7273019191987115e-06, + "loss": 0.2545, + "step": 3044 + }, + { + "epoch": 1.01669449081803, + "grad_norm": 0.4582633193634833, + "learning_rate": 5.724419047847268e-06, + "loss": 0.2602, + "step": 3045 + }, + { + "epoch": 1.0170283806343907, + "grad_norm": 0.46319578323287774, + "learning_rate": 5.721535930486863e-06, + "loss": 0.2588, + "step": 3046 + }, + { + "epoch": 1.0173622704507512, + "grad_norm": 0.43507639077842386, + "learning_rate": 5.718652568096585e-06, + "loss": 0.2488, + "step": 3047 + }, + { + "epoch": 1.017696160267112, + "grad_norm": 0.4439408702733472, + "learning_rate": 5.715768961655611e-06, + "loss": 0.2573, + "step": 3048 + }, + { + "epoch": 1.0180300500834725, + "grad_norm": 0.47555332944964196, + "learning_rate": 5.7128851121431996e-06, + "loss": 0.2602, + "step": 3049 + }, + { + "epoch": 1.018363939899833, + "grad_norm": 0.454945889373208, + "learning_rate": 5.710001020538688e-06, + "loss": 0.2605, + "step": 3050 + }, + { + "epoch": 1.0186978297161937, + "grad_norm": 0.45322794485189977, + "learning_rate": 5.707116687821501e-06, + "loss": 0.2595, + "step": 3051 + }, + { + "epoch": 1.0190317195325542, + "grad_norm": 0.43854966777296456, + "learning_rate": 5.704232114971144e-06, + "loss": 0.2585, + "step": 3052 + }, + { + "epoch": 1.019365609348915, + "grad_norm": 0.43492907709300616, + "learning_rate": 5.701347302967203e-06, + "loss": 0.2585, + "step": 3053 + }, + { + "epoch": 1.0196994991652755, + "grad_norm": 0.43522681662080864, + "learning_rate": 5.698462252789342e-06, + "loss": 0.2531, + "step": 3054 + }, + { + "epoch": 1.020033388981636, + "grad_norm": 0.46512959938310555, + "learning_rate": 5.695576965417311e-06, + "loss": 0.2601, + "step": 3055 + }, + { + "epoch": 1.0203672787979967, + "grad_norm": 0.46399572666292266, + "learning_rate": 5.6926914418309405e-06, + "loss": 0.262, + "step": 3056 + }, + { + "epoch": 1.0207011686143572, + "grad_norm": 0.4375756195793228, + "learning_rate": 5.689805683010137e-06, + "loss": 0.2434, + "step": 3057 + }, + { + "epoch": 1.021035058430718, + "grad_norm": 0.4414429046734702, + "learning_rate": 5.686919689934891e-06, + "loss": 0.2574, + "step": 3058 + }, + { + "epoch": 1.0213689482470785, + "grad_norm": 0.45777111388543296, + "learning_rate": 5.684033463585268e-06, + "loss": 0.2619, + "step": 3059 + }, + { + "epoch": 1.021702838063439, + "grad_norm": 0.45484349585543504, + "learning_rate": 5.68114700494142e-06, + "loss": 0.2598, + "step": 3060 + }, + { + "epoch": 1.0220367278797997, + "grad_norm": 0.45810150208096256, + "learning_rate": 5.678260314983569e-06, + "loss": 0.2712, + "step": 3061 + }, + { + "epoch": 1.0223706176961602, + "grad_norm": 0.4449020914385383, + "learning_rate": 5.675373394692023e-06, + "loss": 0.2597, + "step": 3062 + }, + { + "epoch": 1.022704507512521, + "grad_norm": 0.481574018649305, + "learning_rate": 5.672486245047163e-06, + "loss": 0.2615, + "step": 3063 + }, + { + "epoch": 1.0230383973288815, + "grad_norm": 0.4596218089724499, + "learning_rate": 5.6695988670294515e-06, + "loss": 0.2607, + "step": 3064 + }, + { + "epoch": 1.023372287145242, + "grad_norm": 0.4403955073485268, + "learning_rate": 5.666711261619429e-06, + "loss": 0.2477, + "step": 3065 + }, + { + "epoch": 1.0237061769616027, + "grad_norm": 0.44318798905819357, + "learning_rate": 5.663823429797706e-06, + "loss": 0.2579, + "step": 3066 + }, + { + "epoch": 1.0240400667779632, + "grad_norm": 0.45307657847656724, + "learning_rate": 5.6609353725449804e-06, + "loss": 0.2577, + "step": 3067 + }, + { + "epoch": 1.024373956594324, + "grad_norm": 0.46029057501108295, + "learning_rate": 5.658047090842017e-06, + "loss": 0.2589, + "step": 3068 + }, + { + "epoch": 1.0247078464106845, + "grad_norm": 0.44861522255790964, + "learning_rate": 5.6551585856696665e-06, + "loss": 0.2499, + "step": 3069 + }, + { + "epoch": 1.025041736227045, + "grad_norm": 0.46699848878521316, + "learning_rate": 5.652269858008848e-06, + "loss": 0.2614, + "step": 3070 + }, + { + "epoch": 1.0253756260434057, + "grad_norm": 0.441413092851511, + "learning_rate": 5.649380908840557e-06, + "loss": 0.2586, + "step": 3071 + }, + { + "epoch": 1.0257095158597662, + "grad_norm": 0.45540842277655075, + "learning_rate": 5.646491739145867e-06, + "loss": 0.2654, + "step": 3072 + }, + { + "epoch": 1.026043405676127, + "grad_norm": 0.4427897949209429, + "learning_rate": 5.643602349905924e-06, + "loss": 0.2482, + "step": 3073 + }, + { + "epoch": 1.0263772954924875, + "grad_norm": 0.48410624079035364, + "learning_rate": 5.640712742101954e-06, + "loss": 0.2709, + "step": 3074 + }, + { + "epoch": 1.026711185308848, + "grad_norm": 0.46096506785407626, + "learning_rate": 5.637822916715247e-06, + "loss": 0.2696, + "step": 3075 + }, + { + "epoch": 1.0270450751252087, + "grad_norm": 0.43276984663739243, + "learning_rate": 5.634932874727175e-06, + "loss": 0.2407, + "step": 3076 + }, + { + "epoch": 1.0273789649415692, + "grad_norm": 0.45331347551666845, + "learning_rate": 5.632042617119181e-06, + "loss": 0.2645, + "step": 3077 + }, + { + "epoch": 1.02771285475793, + "grad_norm": 0.4415962884503581, + "learning_rate": 5.629152144872784e-06, + "loss": 0.2602, + "step": 3078 + }, + { + "epoch": 1.0280467445742905, + "grad_norm": 0.4573542430983963, + "learning_rate": 5.626261458969571e-06, + "loss": 0.26, + "step": 3079 + }, + { + "epoch": 1.0283806343906512, + "grad_norm": 0.4560409844848656, + "learning_rate": 5.623370560391205e-06, + "loss": 0.2618, + "step": 3080 + }, + { + "epoch": 1.0287145242070117, + "grad_norm": 0.4577666730477791, + "learning_rate": 5.6204794501194185e-06, + "loss": 0.2591, + "step": 3081 + }, + { + "epoch": 1.0290484140233722, + "grad_norm": 0.43151657689880546, + "learning_rate": 5.617588129136019e-06, + "loss": 0.2487, + "step": 3082 + }, + { + "epoch": 1.029382303839733, + "grad_norm": 0.4469093553846591, + "learning_rate": 5.614696598422885e-06, + "loss": 0.2573, + "step": 3083 + }, + { + "epoch": 1.0297161936560935, + "grad_norm": 0.4496361427949099, + "learning_rate": 5.611804858961963e-06, + "loss": 0.262, + "step": 3084 + }, + { + "epoch": 1.0300500834724542, + "grad_norm": 0.4587622549809809, + "learning_rate": 5.608912911735275e-06, + "loss": 0.267, + "step": 3085 + }, + { + "epoch": 1.0303839732888147, + "grad_norm": 0.4725141699507035, + "learning_rate": 5.6060207577249094e-06, + "loss": 0.26, + "step": 3086 + }, + { + "epoch": 1.0307178631051752, + "grad_norm": 0.45256972924010763, + "learning_rate": 5.6031283979130305e-06, + "loss": 0.258, + "step": 3087 + }, + { + "epoch": 1.031051752921536, + "grad_norm": 0.4574673547985797, + "learning_rate": 5.600235833281864e-06, + "loss": 0.2624, + "step": 3088 + }, + { + "epoch": 1.0313856427378965, + "grad_norm": 0.4713202032217356, + "learning_rate": 5.597343064813712e-06, + "loss": 0.2756, + "step": 3089 + }, + { + "epoch": 1.0317195325542572, + "grad_norm": 0.4530044147432069, + "learning_rate": 5.594450093490944e-06, + "loss": 0.2536, + "step": 3090 + }, + { + "epoch": 1.0320534223706177, + "grad_norm": 0.4469710221998817, + "learning_rate": 5.591556920295998e-06, + "loss": 0.254, + "step": 3091 + }, + { + "epoch": 1.0323873121869782, + "grad_norm": 0.4581803310612515, + "learning_rate": 5.5886635462113805e-06, + "loss": 0.2673, + "step": 3092 + }, + { + "epoch": 1.032721202003339, + "grad_norm": 0.4551674079351838, + "learning_rate": 5.585769972219664e-06, + "loss": 0.2555, + "step": 3093 + }, + { + "epoch": 1.0330550918196995, + "grad_norm": 0.4566454014835286, + "learning_rate": 5.582876199303497e-06, + "loss": 0.2634, + "step": 3094 + }, + { + "epoch": 1.0333889816360602, + "grad_norm": 0.45832793841766106, + "learning_rate": 5.579982228445585e-06, + "loss": 0.2625, + "step": 3095 + }, + { + "epoch": 1.0337228714524207, + "grad_norm": 0.4470599439009321, + "learning_rate": 5.577088060628708e-06, + "loss": 0.2618, + "step": 3096 + }, + { + "epoch": 1.0340567612687812, + "grad_norm": 0.4535741763364359, + "learning_rate": 5.574193696835709e-06, + "loss": 0.2555, + "step": 3097 + }, + { + "epoch": 1.034390651085142, + "grad_norm": 0.44914968669727223, + "learning_rate": 5.571299138049498e-06, + "loss": 0.2531, + "step": 3098 + }, + { + "epoch": 1.0347245409015025, + "grad_norm": 0.4307782843750989, + "learning_rate": 5.568404385253054e-06, + "loss": 0.2456, + "step": 3099 + }, + { + "epoch": 1.0350584307178632, + "grad_norm": 0.4538831106411402, + "learning_rate": 5.56550943942942e-06, + "loss": 0.2599, + "step": 3100 + }, + { + "epoch": 1.0353923205342237, + "grad_norm": 0.4563513351135983, + "learning_rate": 5.562614301561704e-06, + "loss": 0.261, + "step": 3101 + }, + { + "epoch": 1.0357262103505842, + "grad_norm": 0.46670366467770685, + "learning_rate": 5.559718972633079e-06, + "loss": 0.2646, + "step": 3102 + }, + { + "epoch": 1.036060100166945, + "grad_norm": 0.45151461933537235, + "learning_rate": 5.556823453626788e-06, + "loss": 0.2604, + "step": 3103 + }, + { + "epoch": 1.0363939899833055, + "grad_norm": 0.46821511986248543, + "learning_rate": 5.553927745526129e-06, + "loss": 0.2703, + "step": 3104 + }, + { + "epoch": 1.0367278797996662, + "grad_norm": 0.474730732392154, + "learning_rate": 5.551031849314471e-06, + "loss": 0.2591, + "step": 3105 + }, + { + "epoch": 1.0370617696160267, + "grad_norm": 0.45117496572292554, + "learning_rate": 5.548135765975247e-06, + "loss": 0.2581, + "step": 3106 + }, + { + "epoch": 1.0373956594323872, + "grad_norm": 0.4384893226679777, + "learning_rate": 5.54523949649195e-06, + "loss": 0.2605, + "step": 3107 + }, + { + "epoch": 1.037729549248748, + "grad_norm": 0.4575518515067703, + "learning_rate": 5.542343041848141e-06, + "loss": 0.2643, + "step": 3108 + }, + { + "epoch": 1.0380634390651085, + "grad_norm": 0.44208316450559404, + "learning_rate": 5.539446403027438e-06, + "loss": 0.2509, + "step": 3109 + }, + { + "epoch": 1.0383973288814692, + "grad_norm": 0.47816612547053433, + "learning_rate": 5.536549581013525e-06, + "loss": 0.2599, + "step": 3110 + }, + { + "epoch": 1.0387312186978297, + "grad_norm": 0.46106839734500826, + "learning_rate": 5.53365257679015e-06, + "loss": 0.2698, + "step": 3111 + }, + { + "epoch": 1.0390651085141904, + "grad_norm": 0.46307961882835014, + "learning_rate": 5.530755391341117e-06, + "loss": 0.2605, + "step": 3112 + }, + { + "epoch": 1.039398998330551, + "grad_norm": 0.48211684308679, + "learning_rate": 5.5278580256503e-06, + "loss": 0.256, + "step": 3113 + }, + { + "epoch": 1.0397328881469114, + "grad_norm": 0.4749371718767197, + "learning_rate": 5.524960480701625e-06, + "loss": 0.2598, + "step": 3114 + }, + { + "epoch": 1.0400667779632722, + "grad_norm": 0.44556930566436104, + "learning_rate": 5.522062757479084e-06, + "loss": 0.2525, + "step": 3115 + }, + { + "epoch": 1.0404006677796327, + "grad_norm": 0.44941983334008967, + "learning_rate": 5.5191648569667314e-06, + "loss": 0.2583, + "step": 3116 + }, + { + "epoch": 1.0407345575959934, + "grad_norm": 0.47340987396103523, + "learning_rate": 5.516266780148678e-06, + "loss": 0.2696, + "step": 3117 + }, + { + "epoch": 1.041068447412354, + "grad_norm": 0.4784925634118586, + "learning_rate": 5.513368528009094e-06, + "loss": 0.2496, + "step": 3118 + }, + { + "epoch": 1.0414023372287144, + "grad_norm": 0.4680049955517439, + "learning_rate": 5.510470101532213e-06, + "loss": 0.2567, + "step": 3119 + }, + { + "epoch": 1.0417362270450752, + "grad_norm": 0.4496356643050343, + "learning_rate": 5.5075715017023244e-06, + "loss": 0.2528, + "step": 3120 + }, + { + "epoch": 1.0420701168614357, + "grad_norm": 0.4615677243420764, + "learning_rate": 5.504672729503778e-06, + "loss": 0.2605, + "step": 3121 + }, + { + "epoch": 1.0424040066777964, + "grad_norm": 0.4666867143729317, + "learning_rate": 5.5017737859209845e-06, + "loss": 0.2566, + "step": 3122 + }, + { + "epoch": 1.042737896494157, + "grad_norm": 0.4640913529468494, + "learning_rate": 5.498874671938403e-06, + "loss": 0.262, + "step": 3123 + }, + { + "epoch": 1.0430717863105174, + "grad_norm": 0.4515309499166379, + "learning_rate": 5.495975388540567e-06, + "loss": 0.2531, + "step": 3124 + }, + { + "epoch": 1.0434056761268782, + "grad_norm": 0.46617358172672113, + "learning_rate": 5.49307593671205e-06, + "loss": 0.2595, + "step": 3125 + }, + { + "epoch": 1.0437395659432387, + "grad_norm": 0.45160848588418995, + "learning_rate": 5.490176317437497e-06, + "loss": 0.2524, + "step": 3126 + }, + { + "epoch": 1.0440734557595994, + "grad_norm": 0.4472816518239624, + "learning_rate": 5.487276531701599e-06, + "loss": 0.257, + "step": 3127 + }, + { + "epoch": 1.04440734557596, + "grad_norm": 0.4513186507585648, + "learning_rate": 5.48437658048911e-06, + "loss": 0.2613, + "step": 3128 + }, + { + "epoch": 1.0447412353923204, + "grad_norm": 0.4585782226614117, + "learning_rate": 5.48147646478484e-06, + "loss": 0.2618, + "step": 3129 + }, + { + "epoch": 1.0450751252086812, + "grad_norm": 0.4755196542131541, + "learning_rate": 5.47857618557365e-06, + "loss": 0.2567, + "step": 3130 + }, + { + "epoch": 1.0454090150250417, + "grad_norm": 0.44955204381903713, + "learning_rate": 5.475675743840465e-06, + "loss": 0.2468, + "step": 3131 + }, + { + "epoch": 1.0457429048414024, + "grad_norm": 0.4335766695693397, + "learning_rate": 5.4727751405702535e-06, + "loss": 0.2533, + "step": 3132 + }, + { + "epoch": 1.046076794657763, + "grad_norm": 0.4647424023541543, + "learning_rate": 5.469874376748051e-06, + "loss": 0.262, + "step": 3133 + }, + { + "epoch": 1.0464106844741234, + "grad_norm": 0.45510498628032126, + "learning_rate": 5.466973453358938e-06, + "loss": 0.2562, + "step": 3134 + }, + { + "epoch": 1.0467445742904842, + "grad_norm": 0.4525870428382878, + "learning_rate": 5.464072371388056e-06, + "loss": 0.2467, + "step": 3135 + }, + { + "epoch": 1.0470784641068447, + "grad_norm": 0.45449824719804455, + "learning_rate": 5.461171131820596e-06, + "loss": 0.2486, + "step": 3136 + }, + { + "epoch": 1.0474123539232054, + "grad_norm": 0.4387665422463441, + "learning_rate": 5.4582697356418036e-06, + "loss": 0.2409, + "step": 3137 + }, + { + "epoch": 1.047746243739566, + "grad_norm": 0.4480349590666483, + "learning_rate": 5.4553681838369784e-06, + "loss": 0.2524, + "step": 3138 + }, + { + "epoch": 1.0480801335559264, + "grad_norm": 0.43799918008788985, + "learning_rate": 5.4524664773914725e-06, + "loss": 0.2527, + "step": 3139 + }, + { + "epoch": 1.0484140233722872, + "grad_norm": 0.4467319534370973, + "learning_rate": 5.44956461729069e-06, + "loss": 0.2536, + "step": 3140 + }, + { + "epoch": 1.0487479131886477, + "grad_norm": 0.4532497475677387, + "learning_rate": 5.446662604520088e-06, + "loss": 0.2442, + "step": 3141 + }, + { + "epoch": 1.0490818030050084, + "grad_norm": 0.4925555391407359, + "learning_rate": 5.443760440065176e-06, + "loss": 0.2674, + "step": 3142 + }, + { + "epoch": 1.049415692821369, + "grad_norm": 0.43778945233785266, + "learning_rate": 5.440858124911511e-06, + "loss": 0.245, + "step": 3143 + }, + { + "epoch": 1.0497495826377294, + "grad_norm": 0.4723390820338071, + "learning_rate": 5.437955660044707e-06, + "loss": 0.2634, + "step": 3144 + }, + { + "epoch": 1.0500834724540902, + "grad_norm": 0.43665459209884, + "learning_rate": 5.435053046450426e-06, + "loss": 0.2508, + "step": 3145 + }, + { + "epoch": 1.0504173622704507, + "grad_norm": 0.4611322049529888, + "learning_rate": 5.4321502851143785e-06, + "loss": 0.2726, + "step": 3146 + }, + { + "epoch": 1.0507512520868114, + "grad_norm": 0.47189696539810116, + "learning_rate": 5.429247377022329e-06, + "loss": 0.2581, + "step": 3147 + }, + { + "epoch": 1.051085141903172, + "grad_norm": 0.4481651888927194, + "learning_rate": 5.426344323160091e-06, + "loss": 0.2556, + "step": 3148 + }, + { + "epoch": 1.0514190317195327, + "grad_norm": 0.4557933113042225, + "learning_rate": 5.423441124513523e-06, + "loss": 0.2486, + "step": 3149 + }, + { + "epoch": 1.0517529215358932, + "grad_norm": 0.49005828135869123, + "learning_rate": 5.420537782068539e-06, + "loss": 0.2657, + "step": 3150 + }, + { + "epoch": 1.0520868113522537, + "grad_norm": 0.4846462262425801, + "learning_rate": 5.4176342968111005e-06, + "loss": 0.2563, + "step": 3151 + }, + { + "epoch": 1.0524207011686144, + "grad_norm": 0.44657704142053317, + "learning_rate": 5.414730669727212e-06, + "loss": 0.2542, + "step": 3152 + }, + { + "epoch": 1.052754590984975, + "grad_norm": 0.4630633180724599, + "learning_rate": 5.411826901802933e-06, + "loss": 0.2572, + "step": 3153 + }, + { + "epoch": 1.0530884808013357, + "grad_norm": 0.48100419371120984, + "learning_rate": 5.4089229940243685e-06, + "loss": 0.2596, + "step": 3154 + }, + { + "epoch": 1.0534223706176962, + "grad_norm": 0.467070240936253, + "learning_rate": 5.406018947377668e-06, + "loss": 0.2527, + "step": 3155 + }, + { + "epoch": 1.0537562604340567, + "grad_norm": 0.4606964247943917, + "learning_rate": 5.403114762849034e-06, + "loss": 0.2565, + "step": 3156 + }, + { + "epoch": 1.0540901502504174, + "grad_norm": 0.4502103131697735, + "learning_rate": 5.400210441424707e-06, + "loss": 0.2406, + "step": 3157 + }, + { + "epoch": 1.054424040066778, + "grad_norm": 0.4696073417025046, + "learning_rate": 5.397305984090987e-06, + "loss": 0.2561, + "step": 3158 + }, + { + "epoch": 1.0547579298831387, + "grad_norm": 0.43345982157413115, + "learning_rate": 5.394401391834207e-06, + "loss": 0.2394, + "step": 3159 + }, + { + "epoch": 1.0550918196994992, + "grad_norm": 0.45829755890973967, + "learning_rate": 5.391496665640757e-06, + "loss": 0.2435, + "step": 3160 + }, + { + "epoch": 1.0554257095158597, + "grad_norm": 0.45050987580781315, + "learning_rate": 5.38859180649706e-06, + "loss": 0.248, + "step": 3161 + }, + { + "epoch": 1.0557595993322204, + "grad_norm": 0.4671930836599733, + "learning_rate": 5.385686815389595e-06, + "loss": 0.2652, + "step": 3162 + }, + { + "epoch": 1.056093489148581, + "grad_norm": 0.44402392303879296, + "learning_rate": 5.382781693304883e-06, + "loss": 0.2468, + "step": 3163 + }, + { + "epoch": 1.0564273789649417, + "grad_norm": 0.4802765240172819, + "learning_rate": 5.379876441229486e-06, + "loss": 0.2629, + "step": 3164 + }, + { + "epoch": 1.0567612687813022, + "grad_norm": 0.4650774866952289, + "learning_rate": 5.376971060150015e-06, + "loss": 0.2656, + "step": 3165 + }, + { + "epoch": 1.0570951585976627, + "grad_norm": 0.4510842458643156, + "learning_rate": 5.374065551053118e-06, + "loss": 0.2543, + "step": 3166 + }, + { + "epoch": 1.0574290484140234, + "grad_norm": 0.45139966853054025, + "learning_rate": 5.3711599149254966e-06, + "loss": 0.2588, + "step": 3167 + }, + { + "epoch": 1.057762938230384, + "grad_norm": 0.45889547926398544, + "learning_rate": 5.368254152753887e-06, + "loss": 0.2502, + "step": 3168 + }, + { + "epoch": 1.0580968280467447, + "grad_norm": 0.45762304896667355, + "learning_rate": 5.365348265525071e-06, + "loss": 0.2682, + "step": 3169 + }, + { + "epoch": 1.0584307178631052, + "grad_norm": 0.4483004481256225, + "learning_rate": 5.362442254225872e-06, + "loss": 0.2653, + "step": 3170 + }, + { + "epoch": 1.0587646076794657, + "grad_norm": 0.45726689732182263, + "learning_rate": 5.359536119843158e-06, + "loss": 0.2494, + "step": 3171 + }, + { + "epoch": 1.0590984974958264, + "grad_norm": 0.4383801169976877, + "learning_rate": 5.356629863363838e-06, + "loss": 0.2563, + "step": 3172 + }, + { + "epoch": 1.059432387312187, + "grad_norm": 0.4425651465653449, + "learning_rate": 5.353723485774859e-06, + "loss": 0.2594, + "step": 3173 + }, + { + "epoch": 1.0597662771285477, + "grad_norm": 0.44946387477491173, + "learning_rate": 5.350816988063212e-06, + "loss": 0.2602, + "step": 3174 + }, + { + "epoch": 1.0601001669449082, + "grad_norm": 0.4411100907563107, + "learning_rate": 5.34791037121593e-06, + "loss": 0.2628, + "step": 3175 + }, + { + "epoch": 1.060434056761269, + "grad_norm": 0.4406374076722808, + "learning_rate": 5.345003636220088e-06, + "loss": 0.2616, + "step": 3176 + }, + { + "epoch": 1.0607679465776294, + "grad_norm": 0.4735209591706196, + "learning_rate": 5.342096784062792e-06, + "loss": 0.2565, + "step": 3177 + }, + { + "epoch": 1.06110183639399, + "grad_norm": 0.4459338774628588, + "learning_rate": 5.339189815731196e-06, + "loss": 0.2537, + "step": 3178 + }, + { + "epoch": 1.0614357262103506, + "grad_norm": 0.4663922119227229, + "learning_rate": 5.336282732212494e-06, + "loss": 0.2563, + "step": 3179 + }, + { + "epoch": 1.0617696160267112, + "grad_norm": 0.4610327638821071, + "learning_rate": 5.333375534493915e-06, + "loss": 0.2452, + "step": 3180 + }, + { + "epoch": 1.062103505843072, + "grad_norm": 0.46732931190113214, + "learning_rate": 5.330468223562728e-06, + "loss": 0.2547, + "step": 3181 + }, + { + "epoch": 1.0624373956594324, + "grad_norm": 0.47876536151653865, + "learning_rate": 5.327560800406241e-06, + "loss": 0.2505, + "step": 3182 + }, + { + "epoch": 1.062771285475793, + "grad_norm": 0.4701372102425428, + "learning_rate": 5.324653266011799e-06, + "loss": 0.2449, + "step": 3183 + }, + { + "epoch": 1.0631051752921536, + "grad_norm": 0.44656195439846075, + "learning_rate": 5.321745621366786e-06, + "loss": 0.2517, + "step": 3184 + }, + { + "epoch": 1.0634390651085142, + "grad_norm": 0.4403804906031, + "learning_rate": 5.318837867458626e-06, + "loss": 0.2491, + "step": 3185 + }, + { + "epoch": 1.063772954924875, + "grad_norm": 0.47379710723500473, + "learning_rate": 5.315930005274773e-06, + "loss": 0.2628, + "step": 3186 + }, + { + "epoch": 1.0641068447412354, + "grad_norm": 0.46914961222653273, + "learning_rate": 5.313022035802723e-06, + "loss": 0.2577, + "step": 3187 + }, + { + "epoch": 1.064440734557596, + "grad_norm": 0.44579935002799975, + "learning_rate": 5.31011396003001e-06, + "loss": 0.2538, + "step": 3188 + }, + { + "epoch": 1.0647746243739566, + "grad_norm": 0.45881283079367696, + "learning_rate": 5.307205778944198e-06, + "loss": 0.2569, + "step": 3189 + }, + { + "epoch": 1.0651085141903172, + "grad_norm": 0.45088757304147953, + "learning_rate": 5.3042974935328925e-06, + "loss": 0.2593, + "step": 3190 + }, + { + "epoch": 1.065442404006678, + "grad_norm": 0.44065174690650916, + "learning_rate": 5.301389104783731e-06, + "loss": 0.2572, + "step": 3191 + }, + { + "epoch": 1.0657762938230384, + "grad_norm": 0.4282167313828538, + "learning_rate": 5.298480613684388e-06, + "loss": 0.2534, + "step": 3192 + }, + { + "epoch": 1.066110183639399, + "grad_norm": 0.43855767758715763, + "learning_rate": 5.295572021222573e-06, + "loss": 0.2427, + "step": 3193 + }, + { + "epoch": 1.0664440734557596, + "grad_norm": 0.4464861676435093, + "learning_rate": 5.292663328386028e-06, + "loss": 0.2532, + "step": 3194 + }, + { + "epoch": 1.0667779632721202, + "grad_norm": 0.4575198738954354, + "learning_rate": 5.289754536162527e-06, + "loss": 0.2512, + "step": 3195 + }, + { + "epoch": 1.0671118530884809, + "grad_norm": 0.4567237084408686, + "learning_rate": 5.2868456455398855e-06, + "loss": 0.2574, + "step": 3196 + }, + { + "epoch": 1.0674457429048414, + "grad_norm": 0.44774926316844754, + "learning_rate": 5.283936657505946e-06, + "loss": 0.2519, + "step": 3197 + }, + { + "epoch": 1.067779632721202, + "grad_norm": 0.43471021206581684, + "learning_rate": 5.281027573048585e-06, + "loss": 0.2506, + "step": 3198 + }, + { + "epoch": 1.0681135225375626, + "grad_norm": 0.4359479585085106, + "learning_rate": 5.278118393155713e-06, + "loss": 0.2533, + "step": 3199 + }, + { + "epoch": 1.0684474123539232, + "grad_norm": 0.4613560019406218, + "learning_rate": 5.275209118815274e-06, + "loss": 0.2649, + "step": 3200 + }, + { + "epoch": 1.0687813021702839, + "grad_norm": 0.45635640776334285, + "learning_rate": 5.272299751015241e-06, + "loss": 0.2484, + "step": 3201 + }, + { + "epoch": 1.0691151919866444, + "grad_norm": 0.4406972597187109, + "learning_rate": 5.2693902907436204e-06, + "loss": 0.249, + "step": 3202 + }, + { + "epoch": 1.069449081803005, + "grad_norm": 0.45807738507630247, + "learning_rate": 5.266480738988449e-06, + "loss": 0.2546, + "step": 3203 + }, + { + "epoch": 1.0697829716193656, + "grad_norm": 0.4698656063655063, + "learning_rate": 5.263571096737796e-06, + "loss": 0.2594, + "step": 3204 + }, + { + "epoch": 1.0701168614357262, + "grad_norm": 0.44065470368159537, + "learning_rate": 5.2606613649797625e-06, + "loss": 0.2552, + "step": 3205 + }, + { + "epoch": 1.0704507512520869, + "grad_norm": 0.4585616570031521, + "learning_rate": 5.257751544702478e-06, + "loss": 0.2617, + "step": 3206 + }, + { + "epoch": 1.0707846410684474, + "grad_norm": 0.45861161851995746, + "learning_rate": 5.254841636894099e-06, + "loss": 0.2616, + "step": 3207 + }, + { + "epoch": 1.071118530884808, + "grad_norm": 0.46454376230843153, + "learning_rate": 5.251931642542819e-06, + "loss": 0.2626, + "step": 3208 + }, + { + "epoch": 1.0714524207011686, + "grad_norm": 0.4591615950936803, + "learning_rate": 5.249021562636857e-06, + "loss": 0.2575, + "step": 3209 + }, + { + "epoch": 1.0717863105175292, + "grad_norm": 0.4482358202439751, + "learning_rate": 5.246111398164459e-06, + "loss": 0.2554, + "step": 3210 + }, + { + "epoch": 1.0721202003338899, + "grad_norm": 0.46238829667638076, + "learning_rate": 5.243201150113902e-06, + "loss": 0.2634, + "step": 3211 + }, + { + "epoch": 1.0724540901502504, + "grad_norm": 0.46353164354659504, + "learning_rate": 5.240290819473492e-06, + "loss": 0.2621, + "step": 3212 + }, + { + "epoch": 1.072787979966611, + "grad_norm": 0.4680634235176216, + "learning_rate": 5.2373804072315625e-06, + "loss": 0.2558, + "step": 3213 + }, + { + "epoch": 1.0731218697829716, + "grad_norm": 0.4566926742868297, + "learning_rate": 5.2344699143764735e-06, + "loss": 0.2602, + "step": 3214 + }, + { + "epoch": 1.0734557595993321, + "grad_norm": 0.4620924229481597, + "learning_rate": 5.231559341896615e-06, + "loss": 0.2504, + "step": 3215 + }, + { + "epoch": 1.0737896494156929, + "grad_norm": 0.4694673246271609, + "learning_rate": 5.2286486907804e-06, + "loss": 0.2472, + "step": 3216 + }, + { + "epoch": 1.0741235392320534, + "grad_norm": 0.44818083361960925, + "learning_rate": 5.225737962016273e-06, + "loss": 0.262, + "step": 3217 + }, + { + "epoch": 1.0744574290484141, + "grad_norm": 0.47346305760212126, + "learning_rate": 5.222827156592701e-06, + "loss": 0.2608, + "step": 3218 + }, + { + "epoch": 1.0747913188647746, + "grad_norm": 0.45352568127682097, + "learning_rate": 5.219916275498178e-06, + "loss": 0.2492, + "step": 3219 + }, + { + "epoch": 1.0751252086811351, + "grad_norm": 0.47846803160389945, + "learning_rate": 5.217005319721225e-06, + "loss": 0.261, + "step": 3220 + }, + { + "epoch": 1.0754590984974959, + "grad_norm": 0.46028830585982505, + "learning_rate": 5.214094290250386e-06, + "loss": 0.2618, + "step": 3221 + }, + { + "epoch": 1.0757929883138564, + "grad_norm": 0.4614435243666677, + "learning_rate": 5.211183188074237e-06, + "loss": 0.2546, + "step": 3222 + }, + { + "epoch": 1.0761268781302171, + "grad_norm": 0.48995995983143936, + "learning_rate": 5.208272014181366e-06, + "loss": 0.2655, + "step": 3223 + }, + { + "epoch": 1.0764607679465776, + "grad_norm": 0.47790718177277924, + "learning_rate": 5.205360769560398e-06, + "loss": 0.2726, + "step": 3224 + }, + { + "epoch": 1.0767946577629381, + "grad_norm": 0.4602336435139318, + "learning_rate": 5.2024494551999725e-06, + "loss": 0.2484, + "step": 3225 + }, + { + "epoch": 1.0771285475792989, + "grad_norm": 0.4598639732603491, + "learning_rate": 5.199538072088759e-06, + "loss": 0.2598, + "step": 3226 + }, + { + "epoch": 1.0774624373956594, + "grad_norm": 0.45620904825276026, + "learning_rate": 5.196626621215449e-06, + "loss": 0.2601, + "step": 3227 + }, + { + "epoch": 1.0777963272120201, + "grad_norm": 0.45985553663482875, + "learning_rate": 5.193715103568754e-06, + "loss": 0.2584, + "step": 3228 + }, + { + "epoch": 1.0781302170283806, + "grad_norm": 0.48780028979595, + "learning_rate": 5.190803520137412e-06, + "loss": 0.2529, + "step": 3229 + }, + { + "epoch": 1.0784641068447411, + "grad_norm": 0.4676488923315673, + "learning_rate": 5.18789187191018e-06, + "loss": 0.2549, + "step": 3230 + }, + { + "epoch": 1.0787979966611019, + "grad_norm": 0.4543694811498327, + "learning_rate": 5.184980159875841e-06, + "loss": 0.2525, + "step": 3231 + }, + { + "epoch": 1.0791318864774624, + "grad_norm": 0.45026437039792805, + "learning_rate": 5.182068385023196e-06, + "loss": 0.2414, + "step": 3232 + }, + { + "epoch": 1.0794657762938231, + "grad_norm": 0.45272284162222753, + "learning_rate": 5.179156548341068e-06, + "loss": 0.2562, + "step": 3233 + }, + { + "epoch": 1.0797996661101836, + "grad_norm": 0.47070106399742434, + "learning_rate": 5.176244650818304e-06, + "loss": 0.2584, + "step": 3234 + }, + { + "epoch": 1.0801335559265441, + "grad_norm": 0.4604572788471965, + "learning_rate": 5.1733326934437665e-06, + "loss": 0.2586, + "step": 3235 + }, + { + "epoch": 1.0804674457429049, + "grad_norm": 0.43002502898564765, + "learning_rate": 5.170420677206344e-06, + "loss": 0.2419, + "step": 3236 + }, + { + "epoch": 1.0808013355592654, + "grad_norm": 0.45680065128450437, + "learning_rate": 5.167508603094938e-06, + "loss": 0.26, + "step": 3237 + }, + { + "epoch": 1.0811352253756261, + "grad_norm": 0.44545764843697444, + "learning_rate": 5.164596472098479e-06, + "loss": 0.2592, + "step": 3238 + }, + { + "epoch": 1.0814691151919866, + "grad_norm": 0.44666491296498123, + "learning_rate": 5.161684285205907e-06, + "loss": 0.252, + "step": 3239 + }, + { + "epoch": 1.0818030050083474, + "grad_norm": 0.46468311585062144, + "learning_rate": 5.1587720434061895e-06, + "loss": 0.2576, + "step": 3240 + }, + { + "epoch": 1.0821368948247079, + "grad_norm": 0.46035954031484677, + "learning_rate": 5.155859747688306e-06, + "loss": 0.2627, + "step": 3241 + }, + { + "epoch": 1.0824707846410684, + "grad_norm": 0.4671239784382539, + "learning_rate": 5.152947399041258e-06, + "loss": 0.2613, + "step": 3242 + }, + { + "epoch": 1.0828046744574291, + "grad_norm": 0.443867444041447, + "learning_rate": 5.1500349984540655e-06, + "loss": 0.2435, + "step": 3243 + }, + { + "epoch": 1.0831385642737896, + "grad_norm": 0.4710698678422595, + "learning_rate": 5.147122546915762e-06, + "loss": 0.2669, + "step": 3244 + }, + { + "epoch": 1.0834724540901504, + "grad_norm": 0.44064503675354516, + "learning_rate": 5.144210045415402e-06, + "loss": 0.2529, + "step": 3245 + }, + { + "epoch": 1.0838063439065109, + "grad_norm": 0.4432984611181617, + "learning_rate": 5.1412974949420555e-06, + "loss": 0.2558, + "step": 3246 + }, + { + "epoch": 1.0841402337228714, + "grad_norm": 0.46272973608897683, + "learning_rate": 5.138384896484813e-06, + "loss": 0.2655, + "step": 3247 + }, + { + "epoch": 1.0844741235392321, + "grad_norm": 0.4823623385897424, + "learning_rate": 5.135472251032772e-06, + "loss": 0.2644, + "step": 3248 + }, + { + "epoch": 1.0848080133555926, + "grad_norm": 0.43936662744280036, + "learning_rate": 5.132559559575058e-06, + "loss": 0.2574, + "step": 3249 + }, + { + "epoch": 1.0851419031719534, + "grad_norm": 0.4777179483733499, + "learning_rate": 5.129646823100801e-06, + "loss": 0.2697, + "step": 3250 + }, + { + "epoch": 1.0854757929883139, + "grad_norm": 0.41948445400620205, + "learning_rate": 5.126734042599153e-06, + "loss": 0.2495, + "step": 3251 + }, + { + "epoch": 1.0858096828046744, + "grad_norm": 0.44800391488128805, + "learning_rate": 5.12382121905928e-06, + "loss": 0.257, + "step": 3252 + }, + { + "epoch": 1.0861435726210351, + "grad_norm": 0.479122136850862, + "learning_rate": 5.120908353470362e-06, + "loss": 0.2604, + "step": 3253 + }, + { + "epoch": 1.0864774624373956, + "grad_norm": 0.4452199701298003, + "learning_rate": 5.117995446821592e-06, + "loss": 0.2583, + "step": 3254 + }, + { + "epoch": 1.0868113522537564, + "grad_norm": 0.45526654931853167, + "learning_rate": 5.115082500102177e-06, + "loss": 0.252, + "step": 3255 + }, + { + "epoch": 1.0871452420701169, + "grad_norm": 0.45133300283429667, + "learning_rate": 5.112169514301341e-06, + "loss": 0.2625, + "step": 3256 + }, + { + "epoch": 1.0874791318864774, + "grad_norm": 0.4457683738903199, + "learning_rate": 5.109256490408317e-06, + "loss": 0.2528, + "step": 3257 + }, + { + "epoch": 1.0878130217028381, + "grad_norm": 0.4528116319757895, + "learning_rate": 5.1063434294123556e-06, + "loss": 0.243, + "step": 3258 + }, + { + "epoch": 1.0881469115191986, + "grad_norm": 0.5025214107176771, + "learning_rate": 5.103430332302715e-06, + "loss": 0.2617, + "step": 3259 + }, + { + "epoch": 1.0884808013355594, + "grad_norm": 0.4561510784485599, + "learning_rate": 5.1005172000686675e-06, + "loss": 0.2521, + "step": 3260 + }, + { + "epoch": 1.0888146911519199, + "grad_norm": 0.4500461834133341, + "learning_rate": 5.097604033699499e-06, + "loss": 0.2653, + "step": 3261 + }, + { + "epoch": 1.0891485809682804, + "grad_norm": 0.4740584187063796, + "learning_rate": 5.094690834184505e-06, + "loss": 0.2516, + "step": 3262 + }, + { + "epoch": 1.089482470784641, + "grad_norm": 0.43904374479813807, + "learning_rate": 5.091777602512993e-06, + "loss": 0.2505, + "step": 3263 + }, + { + "epoch": 1.0898163606010016, + "grad_norm": 0.4479886249415303, + "learning_rate": 5.088864339674282e-06, + "loss": 0.2592, + "step": 3264 + }, + { + "epoch": 1.0901502504173624, + "grad_norm": 0.44569408993756415, + "learning_rate": 5.085951046657701e-06, + "loss": 0.2575, + "step": 3265 + }, + { + "epoch": 1.0904841402337229, + "grad_norm": 0.45812490194254674, + "learning_rate": 5.083037724452589e-06, + "loss": 0.2506, + "step": 3266 + }, + { + "epoch": 1.0908180300500834, + "grad_norm": 0.46747522891073673, + "learning_rate": 5.080124374048294e-06, + "loss": 0.2583, + "step": 3267 + }, + { + "epoch": 1.091151919866444, + "grad_norm": 0.4604302569978002, + "learning_rate": 5.077210996434177e-06, + "loss": 0.2599, + "step": 3268 + }, + { + "epoch": 1.0914858096828046, + "grad_norm": 0.44336741294903936, + "learning_rate": 5.074297592599604e-06, + "loss": 0.2545, + "step": 3269 + }, + { + "epoch": 1.0918196994991654, + "grad_norm": 0.4492018324004123, + "learning_rate": 5.071384163533955e-06, + "loss": 0.2539, + "step": 3270 + }, + { + "epoch": 1.0921535893155259, + "grad_norm": 0.44492655041376894, + "learning_rate": 5.068470710226611e-06, + "loss": 0.2527, + "step": 3271 + }, + { + "epoch": 1.0924874791318864, + "grad_norm": 0.4658482964273919, + "learning_rate": 5.065557233666968e-06, + "loss": 0.263, + "step": 3272 + }, + { + "epoch": 1.092821368948247, + "grad_norm": 0.44935700263561645, + "learning_rate": 5.062643734844427e-06, + "loss": 0.2662, + "step": 3273 + }, + { + "epoch": 1.0931552587646076, + "grad_norm": 0.47011350693225323, + "learning_rate": 5.059730214748398e-06, + "loss": 0.2592, + "step": 3274 + }, + { + "epoch": 1.0934891485809684, + "grad_norm": 0.4866110438773455, + "learning_rate": 5.0568166743682964e-06, + "loss": 0.2636, + "step": 3275 + }, + { + "epoch": 1.0938230383973289, + "grad_norm": 0.4545797367615475, + "learning_rate": 5.053903114693547e-06, + "loss": 0.2651, + "step": 3276 + }, + { + "epoch": 1.0941569282136894, + "grad_norm": 0.4558087112385125, + "learning_rate": 5.050989536713578e-06, + "loss": 0.2589, + "step": 3277 + }, + { + "epoch": 1.09449081803005, + "grad_norm": 0.4479078146614079, + "learning_rate": 5.048075941417827e-06, + "loss": 0.2537, + "step": 3278 + }, + { + "epoch": 1.0948247078464106, + "grad_norm": 0.453888853360913, + "learning_rate": 5.0451623297957345e-06, + "loss": 0.2622, + "step": 3279 + }, + { + "epoch": 1.0951585976627713, + "grad_norm": 0.46771869330228616, + "learning_rate": 5.042248702836746e-06, + "loss": 0.2593, + "step": 3280 + }, + { + "epoch": 1.0954924874791319, + "grad_norm": 0.45150799445250545, + "learning_rate": 5.039335061530319e-06, + "loss": 0.2586, + "step": 3281 + }, + { + "epoch": 1.0958263772954924, + "grad_norm": 0.4692696900284393, + "learning_rate": 5.036421406865908e-06, + "loss": 0.2447, + "step": 3282 + }, + { + "epoch": 1.096160267111853, + "grad_norm": 0.46430049571095866, + "learning_rate": 5.033507739832976e-06, + "loss": 0.2481, + "step": 3283 + }, + { + "epoch": 1.0964941569282136, + "grad_norm": 0.4619569639772674, + "learning_rate": 5.030594061420987e-06, + "loss": 0.2621, + "step": 3284 + }, + { + "epoch": 1.0968280467445743, + "grad_norm": 0.4523413875196469, + "learning_rate": 5.027680372619414e-06, + "loss": 0.2622, + "step": 3285 + }, + { + "epoch": 1.0971619365609349, + "grad_norm": 0.4764970250101189, + "learning_rate": 5.02476667441773e-06, + "loss": 0.2478, + "step": 3286 + }, + { + "epoch": 1.0974958263772956, + "grad_norm": 0.4933082636965013, + "learning_rate": 5.021852967805411e-06, + "loss": 0.2751, + "step": 3287 + }, + { + "epoch": 1.097829716193656, + "grad_norm": 0.46669856151647693, + "learning_rate": 5.018939253771937e-06, + "loss": 0.2546, + "step": 3288 + }, + { + "epoch": 1.0981636060100166, + "grad_norm": 0.4409841999060258, + "learning_rate": 5.0160255333067895e-06, + "loss": 0.2596, + "step": 3289 + }, + { + "epoch": 1.0984974958263773, + "grad_norm": 0.4643314796625189, + "learning_rate": 5.013111807399455e-06, + "loss": 0.2451, + "step": 3290 + }, + { + "epoch": 1.0988313856427379, + "grad_norm": 0.4493327828633093, + "learning_rate": 5.010198077039418e-06, + "loss": 0.2451, + "step": 3291 + }, + { + "epoch": 1.0991652754590986, + "grad_norm": 0.44634518989179756, + "learning_rate": 5.007284343216168e-06, + "loss": 0.2526, + "step": 3292 + }, + { + "epoch": 1.099499165275459, + "grad_norm": 0.48751199413387114, + "learning_rate": 5.004370606919191e-06, + "loss": 0.2668, + "step": 3293 + }, + { + "epoch": 1.0998330550918196, + "grad_norm": 0.457469302870812, + "learning_rate": 5.001456869137979e-06, + "loss": 0.2517, + "step": 3294 + }, + { + "epoch": 1.1001669449081803, + "grad_norm": 0.45300126860377005, + "learning_rate": 4.998543130862021e-06, + "loss": 0.2568, + "step": 3295 + }, + { + "epoch": 1.1005008347245409, + "grad_norm": 0.46098347255941247, + "learning_rate": 4.995629393080809e-06, + "loss": 0.2636, + "step": 3296 + }, + { + "epoch": 1.1008347245409016, + "grad_norm": 0.4659318699443641, + "learning_rate": 4.992715656783834e-06, + "loss": 0.255, + "step": 3297 + }, + { + "epoch": 1.101168614357262, + "grad_norm": 0.45782846702549085, + "learning_rate": 4.989801922960583e-06, + "loss": 0.2551, + "step": 3298 + }, + { + "epoch": 1.1015025041736226, + "grad_norm": 0.46183114787351986, + "learning_rate": 4.986888192600546e-06, + "loss": 0.2563, + "step": 3299 + }, + { + "epoch": 1.1018363939899833, + "grad_norm": 0.47043229718765983, + "learning_rate": 4.983974466693212e-06, + "loss": 0.2667, + "step": 3300 + }, + { + "epoch": 1.1021702838063439, + "grad_norm": 0.4324470954830299, + "learning_rate": 4.9810607462280655e-06, + "loss": 0.2517, + "step": 3301 + }, + { + "epoch": 1.1025041736227046, + "grad_norm": 0.4661654918700883, + "learning_rate": 4.978147032194592e-06, + "loss": 0.2626, + "step": 3302 + }, + { + "epoch": 1.102838063439065, + "grad_norm": 0.45488218008500253, + "learning_rate": 4.975233325582271e-06, + "loss": 0.2586, + "step": 3303 + }, + { + "epoch": 1.1031719532554256, + "grad_norm": 0.458609307084958, + "learning_rate": 4.972319627380587e-06, + "loss": 0.2609, + "step": 3304 + }, + { + "epoch": 1.1035058430717863, + "grad_norm": 0.47243253290203935, + "learning_rate": 4.969405938579014e-06, + "loss": 0.2544, + "step": 3305 + }, + { + "epoch": 1.1038397328881469, + "grad_norm": 0.4643269561354744, + "learning_rate": 4.9664922601670265e-06, + "loss": 0.2615, + "step": 3306 + }, + { + "epoch": 1.1041736227045076, + "grad_norm": 0.4944296974657985, + "learning_rate": 4.963578593134094e-06, + "loss": 0.2745, + "step": 3307 + }, + { + "epoch": 1.104507512520868, + "grad_norm": 0.4612880699377389, + "learning_rate": 4.960664938469683e-06, + "loss": 0.2576, + "step": 3308 + }, + { + "epoch": 1.1048414023372288, + "grad_norm": 0.4789286216978341, + "learning_rate": 4.9577512971632555e-06, + "loss": 0.2547, + "step": 3309 + }, + { + "epoch": 1.1051752921535893, + "grad_norm": 0.4630810084781392, + "learning_rate": 4.954837670204268e-06, + "loss": 0.2588, + "step": 3310 + }, + { + "epoch": 1.1055091819699499, + "grad_norm": 0.43937951166439626, + "learning_rate": 4.951924058582176e-06, + "loss": 0.2531, + "step": 3311 + }, + { + "epoch": 1.1058430717863106, + "grad_norm": 0.44245001446859844, + "learning_rate": 4.949010463286422e-06, + "loss": 0.2583, + "step": 3312 + }, + { + "epoch": 1.106176961602671, + "grad_norm": 0.47301451283747287, + "learning_rate": 4.9460968853064535e-06, + "loss": 0.2567, + "step": 3313 + }, + { + "epoch": 1.1065108514190318, + "grad_norm": 0.4504253775116624, + "learning_rate": 4.943183325631704e-06, + "loss": 0.2514, + "step": 3314 + }, + { + "epoch": 1.1068447412353923, + "grad_norm": 0.4365238619841919, + "learning_rate": 4.940269785251603e-06, + "loss": 0.2396, + "step": 3315 + }, + { + "epoch": 1.1071786310517528, + "grad_norm": 0.4492956811928863, + "learning_rate": 4.937356265155575e-06, + "loss": 0.2527, + "step": 3316 + }, + { + "epoch": 1.1075125208681136, + "grad_norm": 0.44790878116809535, + "learning_rate": 4.934442766333034e-06, + "loss": 0.2574, + "step": 3317 + }, + { + "epoch": 1.107846410684474, + "grad_norm": 0.4802311352292578, + "learning_rate": 4.931529289773392e-06, + "loss": 0.2611, + "step": 3318 + }, + { + "epoch": 1.1081803005008348, + "grad_norm": 0.4825403926660598, + "learning_rate": 4.928615836466048e-06, + "loss": 0.2635, + "step": 3319 + }, + { + "epoch": 1.1085141903171953, + "grad_norm": 0.4661169385561199, + "learning_rate": 4.925702407400396e-06, + "loss": 0.2524, + "step": 3320 + }, + { + "epoch": 1.1088480801335558, + "grad_norm": 0.44655957938439683, + "learning_rate": 4.922789003565824e-06, + "loss": 0.2516, + "step": 3321 + }, + { + "epoch": 1.1091819699499166, + "grad_norm": 0.4502219066794303, + "learning_rate": 4.919875625951707e-06, + "loss": 0.2522, + "step": 3322 + }, + { + "epoch": 1.109515859766277, + "grad_norm": 0.46571585318424413, + "learning_rate": 4.916962275547413e-06, + "loss": 0.252, + "step": 3323 + }, + { + "epoch": 1.1098497495826378, + "grad_norm": 0.4677525541080259, + "learning_rate": 4.9140489533423e-06, + "loss": 0.2712, + "step": 3324 + }, + { + "epoch": 1.1101836393989983, + "grad_norm": 0.4629671411361042, + "learning_rate": 4.91113566032572e-06, + "loss": 0.2542, + "step": 3325 + }, + { + "epoch": 1.1105175292153588, + "grad_norm": 0.47393482743201254, + "learning_rate": 4.908222397487009e-06, + "loss": 0.2554, + "step": 3326 + }, + { + "epoch": 1.1108514190317196, + "grad_norm": 0.4599345607492182, + "learning_rate": 4.9053091658154975e-06, + "loss": 0.2543, + "step": 3327 + }, + { + "epoch": 1.11118530884808, + "grad_norm": 0.4543139616756665, + "learning_rate": 4.902395966300503e-06, + "loss": 0.2572, + "step": 3328 + }, + { + "epoch": 1.1115191986644408, + "grad_norm": 0.4499185432034435, + "learning_rate": 4.899482799931333e-06, + "loss": 0.2597, + "step": 3329 + }, + { + "epoch": 1.1118530884808013, + "grad_norm": 0.4369072190663253, + "learning_rate": 4.896569667697287e-06, + "loss": 0.2515, + "step": 3330 + }, + { + "epoch": 1.1121869782971618, + "grad_norm": 0.4532516255085451, + "learning_rate": 4.893656570587645e-06, + "loss": 0.261, + "step": 3331 + }, + { + "epoch": 1.1125208681135226, + "grad_norm": 0.43873815878753175, + "learning_rate": 4.890743509591684e-06, + "loss": 0.2512, + "step": 3332 + }, + { + "epoch": 1.112854757929883, + "grad_norm": 0.45841330319341816, + "learning_rate": 4.887830485698659e-06, + "loss": 0.2437, + "step": 3333 + }, + { + "epoch": 1.1131886477462438, + "grad_norm": 0.47630229090456966, + "learning_rate": 4.884917499897824e-06, + "loss": 0.2698, + "step": 3334 + }, + { + "epoch": 1.1135225375626043, + "grad_norm": 0.44242629769526426, + "learning_rate": 4.88200455317841e-06, + "loss": 0.2533, + "step": 3335 + }, + { + "epoch": 1.1138564273789648, + "grad_norm": 0.4611541777099399, + "learning_rate": 4.87909164652964e-06, + "loss": 0.2571, + "step": 3336 + }, + { + "epoch": 1.1141903171953256, + "grad_norm": 0.45035585661931343, + "learning_rate": 4.87617878094072e-06, + "loss": 0.2602, + "step": 3337 + }, + { + "epoch": 1.114524207011686, + "grad_norm": 0.45988422222505143, + "learning_rate": 4.873265957400847e-06, + "loss": 0.2478, + "step": 3338 + }, + { + "epoch": 1.1148580968280468, + "grad_norm": 0.4548295918254399, + "learning_rate": 4.8703531768992e-06, + "loss": 0.2419, + "step": 3339 + }, + { + "epoch": 1.1151919866444073, + "grad_norm": 0.426143776532123, + "learning_rate": 4.867440440424944e-06, + "loss": 0.2453, + "step": 3340 + }, + { + "epoch": 1.1155258764607678, + "grad_norm": 0.44913210147526905, + "learning_rate": 4.864527748967229e-06, + "loss": 0.2571, + "step": 3341 + }, + { + "epoch": 1.1158597662771286, + "grad_norm": 0.45572979846875233, + "learning_rate": 4.86161510351519e-06, + "loss": 0.252, + "step": 3342 + }, + { + "epoch": 1.116193656093489, + "grad_norm": 0.4590076435038289, + "learning_rate": 4.858702505057945e-06, + "loss": 0.2586, + "step": 3343 + }, + { + "epoch": 1.1165275459098498, + "grad_norm": 0.46085494261060256, + "learning_rate": 4.8557899545846e-06, + "loss": 0.2554, + "step": 3344 + }, + { + "epoch": 1.1168614357262103, + "grad_norm": 0.45962449937600963, + "learning_rate": 4.852877453084241e-06, + "loss": 0.2636, + "step": 3345 + }, + { + "epoch": 1.1171953255425708, + "grad_norm": 0.44436916460973563, + "learning_rate": 4.849965001545936e-06, + "loss": 0.2533, + "step": 3346 + }, + { + "epoch": 1.1175292153589316, + "grad_norm": 0.4423994227192061, + "learning_rate": 4.847052600958743e-06, + "loss": 0.2596, + "step": 3347 + }, + { + "epoch": 1.117863105175292, + "grad_norm": 0.4592624552044963, + "learning_rate": 4.844140252311695e-06, + "loss": 0.2539, + "step": 3348 + }, + { + "epoch": 1.1181969949916528, + "grad_norm": 0.4430152219365055, + "learning_rate": 4.841227956593812e-06, + "loss": 0.2443, + "step": 3349 + }, + { + "epoch": 1.1185308848080133, + "grad_norm": 0.46090661408085937, + "learning_rate": 4.838315714794094e-06, + "loss": 0.2548, + "step": 3350 + }, + { + "epoch": 1.118864774624374, + "grad_norm": 0.44220167328864074, + "learning_rate": 4.8354035279015235e-06, + "loss": 0.254, + "step": 3351 + }, + { + "epoch": 1.1191986644407346, + "grad_norm": 0.44781120625279114, + "learning_rate": 4.8324913969050645e-06, + "loss": 0.2453, + "step": 3352 + }, + { + "epoch": 1.119532554257095, + "grad_norm": 0.4536258212369617, + "learning_rate": 4.829579322793659e-06, + "loss": 0.2479, + "step": 3353 + }, + { + "epoch": 1.1198664440734558, + "grad_norm": 0.46622540021263104, + "learning_rate": 4.826667306556234e-06, + "loss": 0.2616, + "step": 3354 + }, + { + "epoch": 1.1202003338898163, + "grad_norm": 0.439346385105829, + "learning_rate": 4.823755349181698e-06, + "loss": 0.2433, + "step": 3355 + }, + { + "epoch": 1.120534223706177, + "grad_norm": 0.4438675932190583, + "learning_rate": 4.820843451658932e-06, + "loss": 0.2495, + "step": 3356 + }, + { + "epoch": 1.1208681135225376, + "grad_norm": 0.44773652244320183, + "learning_rate": 4.817931614976806e-06, + "loss": 0.2449, + "step": 3357 + }, + { + "epoch": 1.121202003338898, + "grad_norm": 0.4596777702876801, + "learning_rate": 4.81501984012416e-06, + "loss": 0.2586, + "step": 3358 + }, + { + "epoch": 1.1215358931552588, + "grad_norm": 0.4656878844431221, + "learning_rate": 4.812108128089821e-06, + "loss": 0.2478, + "step": 3359 + }, + { + "epoch": 1.1218697829716193, + "grad_norm": 0.4564045623896528, + "learning_rate": 4.8091964798625905e-06, + "loss": 0.2513, + "step": 3360 + }, + { + "epoch": 1.12220367278798, + "grad_norm": 0.4562805247099522, + "learning_rate": 4.8062848964312485e-06, + "loss": 0.2533, + "step": 3361 + }, + { + "epoch": 1.1225375626043406, + "grad_norm": 0.45395064288457276, + "learning_rate": 4.8033733787845535e-06, + "loss": 0.2502, + "step": 3362 + }, + { + "epoch": 1.122871452420701, + "grad_norm": 0.43057926013335485, + "learning_rate": 4.800461927911241e-06, + "loss": 0.258, + "step": 3363 + }, + { + "epoch": 1.1232053422370618, + "grad_norm": 0.44287063149156186, + "learning_rate": 4.797550544800029e-06, + "loss": 0.2554, + "step": 3364 + }, + { + "epoch": 1.1235392320534223, + "grad_norm": 0.4197247670019723, + "learning_rate": 4.794639230439604e-06, + "loss": 0.2531, + "step": 3365 + }, + { + "epoch": 1.123873121869783, + "grad_norm": 0.43584056712797026, + "learning_rate": 4.791727985818636e-06, + "loss": 0.2543, + "step": 3366 + }, + { + "epoch": 1.1242070116861436, + "grad_norm": 0.4477582089431661, + "learning_rate": 4.788816811925765e-06, + "loss": 0.2583, + "step": 3367 + }, + { + "epoch": 1.124540901502504, + "grad_norm": 0.45798340671892324, + "learning_rate": 4.7859057097496145e-06, + "loss": 0.2716, + "step": 3368 + }, + { + "epoch": 1.1248747913188648, + "grad_norm": 0.435085491998603, + "learning_rate": 4.782994680278776e-06, + "loss": 0.251, + "step": 3369 + }, + { + "epoch": 1.1252086811352253, + "grad_norm": 0.4470967266226729, + "learning_rate": 4.780083724501825e-06, + "loss": 0.2448, + "step": 3370 + }, + { + "epoch": 1.125542570951586, + "grad_norm": 0.45779491879234, + "learning_rate": 4.7771728434073005e-06, + "loss": 0.2603, + "step": 3371 + }, + { + "epoch": 1.1258764607679466, + "grad_norm": 0.47490707635346496, + "learning_rate": 4.7742620379837276e-06, + "loss": 0.2597, + "step": 3372 + }, + { + "epoch": 1.1262103505843073, + "grad_norm": 0.4547828233711212, + "learning_rate": 4.771351309219601e-06, + "loss": 0.2609, + "step": 3373 + }, + { + "epoch": 1.1265442404006678, + "grad_norm": 0.4509066857494809, + "learning_rate": 4.768440658103386e-06, + "loss": 0.2577, + "step": 3374 + }, + { + "epoch": 1.1268781302170283, + "grad_norm": 0.4639963854343385, + "learning_rate": 4.765530085623527e-06, + "loss": 0.2551, + "step": 3375 + }, + { + "epoch": 1.127212020033389, + "grad_norm": 0.4728731218443318, + "learning_rate": 4.762619592768439e-06, + "loss": 0.2611, + "step": 3376 + }, + { + "epoch": 1.1275459098497496, + "grad_norm": 0.45411461236401107, + "learning_rate": 4.75970918052651e-06, + "loss": 0.2492, + "step": 3377 + }, + { + "epoch": 1.1278797996661103, + "grad_norm": 0.45164400525382975, + "learning_rate": 4.7567988498861e-06, + "loss": 0.2537, + "step": 3378 + }, + { + "epoch": 1.1282136894824708, + "grad_norm": 0.46287094444541205, + "learning_rate": 4.753888601835544e-06, + "loss": 0.2474, + "step": 3379 + }, + { + "epoch": 1.1285475792988313, + "grad_norm": 0.46100402127174217, + "learning_rate": 4.7509784373631446e-06, + "loss": 0.2605, + "step": 3380 + }, + { + "epoch": 1.128881469115192, + "grad_norm": 0.46701156086223805, + "learning_rate": 4.748068357457182e-06, + "loss": 0.2594, + "step": 3381 + }, + { + "epoch": 1.1292153589315526, + "grad_norm": 0.45715745094026067, + "learning_rate": 4.7451583631059026e-06, + "loss": 0.2618, + "step": 3382 + }, + { + "epoch": 1.1295492487479133, + "grad_norm": 0.46002602389050967, + "learning_rate": 4.742248455297524e-06, + "loss": 0.265, + "step": 3383 + }, + { + "epoch": 1.1298831385642738, + "grad_norm": 0.45927866469411915, + "learning_rate": 4.739338635020239e-06, + "loss": 0.2568, + "step": 3384 + }, + { + "epoch": 1.1302170283806343, + "grad_norm": 0.4555956503394849, + "learning_rate": 4.736428903262205e-06, + "loss": 0.2566, + "step": 3385 + }, + { + "epoch": 1.130550918196995, + "grad_norm": 0.4377433589146595, + "learning_rate": 4.7335192610115535e-06, + "loss": 0.2323, + "step": 3386 + }, + { + "epoch": 1.1308848080133556, + "grad_norm": 0.4620737672813431, + "learning_rate": 4.730609709256382e-06, + "loss": 0.2548, + "step": 3387 + }, + { + "epoch": 1.1312186978297163, + "grad_norm": 0.44664056289156234, + "learning_rate": 4.727700248984759e-06, + "loss": 0.2532, + "step": 3388 + }, + { + "epoch": 1.1315525876460768, + "grad_norm": 0.47947257677568217, + "learning_rate": 4.724790881184727e-06, + "loss": 0.2611, + "step": 3389 + }, + { + "epoch": 1.1318864774624373, + "grad_norm": 0.45391312967270125, + "learning_rate": 4.721881606844287e-06, + "loss": 0.2593, + "step": 3390 + }, + { + "epoch": 1.132220367278798, + "grad_norm": 0.45350983130234623, + "learning_rate": 4.718972426951417e-06, + "loss": 0.2689, + "step": 3391 + }, + { + "epoch": 1.1325542570951586, + "grad_norm": 0.4418666318680249, + "learning_rate": 4.716063342494055e-06, + "loss": 0.2472, + "step": 3392 + }, + { + "epoch": 1.1328881469115193, + "grad_norm": 0.4493857429797891, + "learning_rate": 4.713154354460116e-06, + "loss": 0.2515, + "step": 3393 + }, + { + "epoch": 1.1332220367278798, + "grad_norm": 0.45885528842333184, + "learning_rate": 4.710245463837475e-06, + "loss": 0.2678, + "step": 3394 + }, + { + "epoch": 1.1335559265442403, + "grad_norm": 0.4381318441894104, + "learning_rate": 4.707336671613976e-06, + "loss": 0.2499, + "step": 3395 + }, + { + "epoch": 1.133889816360601, + "grad_norm": 0.44676257897500543, + "learning_rate": 4.7044279787774296e-06, + "loss": 0.2505, + "step": 3396 + }, + { + "epoch": 1.1342237061769616, + "grad_norm": 0.45961201607719354, + "learning_rate": 4.701519386315612e-06, + "loss": 0.2609, + "step": 3397 + }, + { + "epoch": 1.1345575959933223, + "grad_norm": 0.4771162138244539, + "learning_rate": 4.69861089521627e-06, + "loss": 0.2598, + "step": 3398 + }, + { + "epoch": 1.1348914858096828, + "grad_norm": 0.48050396728926853, + "learning_rate": 4.695702506467108e-06, + "loss": 0.2654, + "step": 3399 + }, + { + "epoch": 1.1352253756260433, + "grad_norm": 0.4601270329627936, + "learning_rate": 4.692794221055804e-06, + "loss": 0.2563, + "step": 3400 + }, + { + "epoch": 1.135559265442404, + "grad_norm": 0.43761393588695285, + "learning_rate": 4.689886039969991e-06, + "loss": 0.2455, + "step": 3401 + }, + { + "epoch": 1.1358931552587646, + "grad_norm": 0.46118918808682463, + "learning_rate": 4.6869779641972785e-06, + "loss": 0.2701, + "step": 3402 + }, + { + "epoch": 1.1362270450751253, + "grad_norm": 0.46647344348501235, + "learning_rate": 4.684069994725229e-06, + "loss": 0.2631, + "step": 3403 + }, + { + "epoch": 1.1365609348914858, + "grad_norm": 0.468456631791398, + "learning_rate": 4.681162132541377e-06, + "loss": 0.2543, + "step": 3404 + }, + { + "epoch": 1.1368948247078463, + "grad_norm": 0.4767856780682363, + "learning_rate": 4.678254378633214e-06, + "loss": 0.2634, + "step": 3405 + }, + { + "epoch": 1.137228714524207, + "grad_norm": 0.4595857113477975, + "learning_rate": 4.675346733988201e-06, + "loss": 0.2522, + "step": 3406 + }, + { + "epoch": 1.1375626043405676, + "grad_norm": 0.45885775090239544, + "learning_rate": 4.672439199593761e-06, + "loss": 0.26, + "step": 3407 + }, + { + "epoch": 1.1378964941569283, + "grad_norm": 0.47087359765292075, + "learning_rate": 4.669531776437273e-06, + "loss": 0.2626, + "step": 3408 + }, + { + "epoch": 1.1382303839732888, + "grad_norm": 0.4666928833742033, + "learning_rate": 4.6666244655060866e-06, + "loss": 0.2618, + "step": 3409 + }, + { + "epoch": 1.1385642737896493, + "grad_norm": 0.4631481040436322, + "learning_rate": 4.663717267787507e-06, + "loss": 0.2467, + "step": 3410 + }, + { + "epoch": 1.13889816360601, + "grad_norm": 0.46940177760703267, + "learning_rate": 4.660810184268805e-06, + "loss": 0.2627, + "step": 3411 + }, + { + "epoch": 1.1392320534223705, + "grad_norm": 0.46815726549101094, + "learning_rate": 4.65790321593721e-06, + "loss": 0.2489, + "step": 3412 + }, + { + "epoch": 1.1395659432387313, + "grad_norm": 0.46047836744932324, + "learning_rate": 4.654996363779916e-06, + "loss": 0.2605, + "step": 3413 + }, + { + "epoch": 1.1398998330550918, + "grad_norm": 0.48942620672289244, + "learning_rate": 4.65208962878407e-06, + "loss": 0.2673, + "step": 3414 + }, + { + "epoch": 1.1402337228714523, + "grad_norm": 0.46296518879337606, + "learning_rate": 4.649183011936788e-06, + "loss": 0.2626, + "step": 3415 + }, + { + "epoch": 1.140567612687813, + "grad_norm": 0.45752710134045127, + "learning_rate": 4.646276514225143e-06, + "loss": 0.2582, + "step": 3416 + }, + { + "epoch": 1.1409015025041735, + "grad_norm": 0.4722277479186175, + "learning_rate": 4.643370136636164e-06, + "loss": 0.2552, + "step": 3417 + }, + { + "epoch": 1.1412353923205343, + "grad_norm": 0.45504819686609377, + "learning_rate": 4.6404638801568435e-06, + "loss": 0.2545, + "step": 3418 + }, + { + "epoch": 1.1415692821368948, + "grad_norm": 0.44765723989473477, + "learning_rate": 4.637557745774129e-06, + "loss": 0.2467, + "step": 3419 + }, + { + "epoch": 1.1419031719532553, + "grad_norm": 0.45701178852690033, + "learning_rate": 4.634651734474932e-06, + "loss": 0.2486, + "step": 3420 + }, + { + "epoch": 1.142237061769616, + "grad_norm": 0.4464980469782843, + "learning_rate": 4.631745847246115e-06, + "loss": 0.2482, + "step": 3421 + }, + { + "epoch": 1.1425709515859765, + "grad_norm": 0.44354815717259166, + "learning_rate": 4.628840085074503e-06, + "loss": 0.2617, + "step": 3422 + }, + { + "epoch": 1.1429048414023373, + "grad_norm": 0.43730980129846975, + "learning_rate": 4.6259344489468825e-06, + "loss": 0.2457, + "step": 3423 + }, + { + "epoch": 1.1432387312186978, + "grad_norm": 0.4672459065648512, + "learning_rate": 4.623028939849987e-06, + "loss": 0.259, + "step": 3424 + }, + { + "epoch": 1.1435726210350585, + "grad_norm": 0.45172467771622987, + "learning_rate": 4.6201235587705155e-06, + "loss": 0.2478, + "step": 3425 + }, + { + "epoch": 1.143906510851419, + "grad_norm": 0.47805573239546956, + "learning_rate": 4.617218306695119e-06, + "loss": 0.2622, + "step": 3426 + }, + { + "epoch": 1.1442404006677795, + "grad_norm": 0.4352050851726695, + "learning_rate": 4.614313184610407e-06, + "loss": 0.2512, + "step": 3427 + }, + { + "epoch": 1.1445742904841403, + "grad_norm": 0.4454580343233765, + "learning_rate": 4.611408193502942e-06, + "loss": 0.2513, + "step": 3428 + }, + { + "epoch": 1.1449081803005008, + "grad_norm": 0.46257545159177166, + "learning_rate": 4.608503334359247e-06, + "loss": 0.2561, + "step": 3429 + }, + { + "epoch": 1.1452420701168615, + "grad_norm": 0.4654075447178096, + "learning_rate": 4.605598608165794e-06, + "loss": 0.2545, + "step": 3430 + }, + { + "epoch": 1.145575959933222, + "grad_norm": 0.45629367563548284, + "learning_rate": 4.602694015909014e-06, + "loss": 0.2518, + "step": 3431 + }, + { + "epoch": 1.1459098497495825, + "grad_norm": 0.4524277856944719, + "learning_rate": 4.599789558575293e-06, + "loss": 0.2548, + "step": 3432 + }, + { + "epoch": 1.1462437395659433, + "grad_norm": 0.44624342985762927, + "learning_rate": 4.596885237150968e-06, + "loss": 0.2537, + "step": 3433 + }, + { + "epoch": 1.1465776293823038, + "grad_norm": 0.454800153635006, + "learning_rate": 4.593981052622334e-06, + "loss": 0.2555, + "step": 3434 + }, + { + "epoch": 1.1469115191986645, + "grad_norm": 0.45055406906341583, + "learning_rate": 4.591077005975633e-06, + "loss": 0.2401, + "step": 3435 + }, + { + "epoch": 1.147245409015025, + "grad_norm": 0.4661086896461625, + "learning_rate": 4.588173098197068e-06, + "loss": 0.2563, + "step": 3436 + }, + { + "epoch": 1.1475792988313858, + "grad_norm": 0.4581184088646506, + "learning_rate": 4.585269330272789e-06, + "loss": 0.2561, + "step": 3437 + }, + { + "epoch": 1.1479131886477463, + "grad_norm": 0.4571179678580447, + "learning_rate": 4.582365703188902e-06, + "loss": 0.2539, + "step": 3438 + }, + { + "epoch": 1.1482470784641068, + "grad_norm": 0.4724170044365395, + "learning_rate": 4.5794622179314605e-06, + "loss": 0.259, + "step": 3439 + }, + { + "epoch": 1.1485809682804675, + "grad_norm": 0.46350254756967113, + "learning_rate": 4.576558875486477e-06, + "loss": 0.2496, + "step": 3440 + }, + { + "epoch": 1.148914858096828, + "grad_norm": 0.4652275031977234, + "learning_rate": 4.573655676839911e-06, + "loss": 0.265, + "step": 3441 + }, + { + "epoch": 1.1492487479131888, + "grad_norm": 0.4588631805479502, + "learning_rate": 4.5707526229776715e-06, + "loss": 0.2526, + "step": 3442 + }, + { + "epoch": 1.1495826377295493, + "grad_norm": 0.4648550213020831, + "learning_rate": 4.567849714885622e-06, + "loss": 0.2624, + "step": 3443 + }, + { + "epoch": 1.1499165275459098, + "grad_norm": 0.48393448433432534, + "learning_rate": 4.564946953549576e-06, + "loss": 0.2589, + "step": 3444 + }, + { + "epoch": 1.1502504173622705, + "grad_norm": 0.45120975408919056, + "learning_rate": 4.562044339955295e-06, + "loss": 0.2641, + "step": 3445 + }, + { + "epoch": 1.150584307178631, + "grad_norm": 0.44730576174107595, + "learning_rate": 4.559141875088491e-06, + "loss": 0.249, + "step": 3446 + }, + { + "epoch": 1.1509181969949918, + "grad_norm": 0.45313184025004377, + "learning_rate": 4.556239559934828e-06, + "loss": 0.2616, + "step": 3447 + }, + { + "epoch": 1.1512520868113523, + "grad_norm": 0.4680838548971961, + "learning_rate": 4.553337395479913e-06, + "loss": 0.2466, + "step": 3448 + }, + { + "epoch": 1.1515859766277128, + "grad_norm": 0.4578830248265678, + "learning_rate": 4.550435382709311e-06, + "loss": 0.2533, + "step": 3449 + }, + { + "epoch": 1.1519198664440735, + "grad_norm": 0.45097972475278153, + "learning_rate": 4.547533522608529e-06, + "loss": 0.2634, + "step": 3450 + }, + { + "epoch": 1.152253756260434, + "grad_norm": 0.47256511507009946, + "learning_rate": 4.544631816163022e-06, + "loss": 0.2617, + "step": 3451 + }, + { + "epoch": 1.1525876460767948, + "grad_norm": 0.4727088173726251, + "learning_rate": 4.541730264358198e-06, + "loss": 0.2518, + "step": 3452 + }, + { + "epoch": 1.1529215358931553, + "grad_norm": 0.4835539326415661, + "learning_rate": 4.5388288681794056e-06, + "loss": 0.2495, + "step": 3453 + }, + { + "epoch": 1.1532554257095158, + "grad_norm": 0.45529115528017466, + "learning_rate": 4.535927628611946e-06, + "loss": 0.2498, + "step": 3454 + }, + { + "epoch": 1.1535893155258765, + "grad_norm": 0.43819877858885875, + "learning_rate": 4.5330265466410625e-06, + "loss": 0.2504, + "step": 3455 + }, + { + "epoch": 1.153923205342237, + "grad_norm": 0.4533695837499993, + "learning_rate": 4.5301256232519495e-06, + "loss": 0.2541, + "step": 3456 + }, + { + "epoch": 1.1542570951585978, + "grad_norm": 0.4408105865718172, + "learning_rate": 4.5272248594297465e-06, + "loss": 0.2487, + "step": 3457 + }, + { + "epoch": 1.1545909849749583, + "grad_norm": 0.4639857571593701, + "learning_rate": 4.524324256159536e-06, + "loss": 0.2661, + "step": 3458 + }, + { + "epoch": 1.1549248747913188, + "grad_norm": 0.435858309344594, + "learning_rate": 4.5214238144263504e-06, + "loss": 0.2448, + "step": 3459 + }, + { + "epoch": 1.1552587646076795, + "grad_norm": 0.4556001999410124, + "learning_rate": 4.518523535215161e-06, + "loss": 0.2604, + "step": 3460 + }, + { + "epoch": 1.15559265442404, + "grad_norm": 0.45333622113080413, + "learning_rate": 4.515623419510891e-06, + "loss": 0.2509, + "step": 3461 + }, + { + "epoch": 1.1559265442404008, + "grad_norm": 0.44969426720419187, + "learning_rate": 4.512723468298403e-06, + "loss": 0.2552, + "step": 3462 + }, + { + "epoch": 1.1562604340567613, + "grad_norm": 0.4738965438549234, + "learning_rate": 4.509823682562507e-06, + "loss": 0.2608, + "step": 3463 + }, + { + "epoch": 1.1565943238731218, + "grad_norm": 0.44387742208348024, + "learning_rate": 4.506924063287951e-06, + "loss": 0.2556, + "step": 3464 + }, + { + "epoch": 1.1569282136894825, + "grad_norm": 0.467425205300095, + "learning_rate": 4.504024611459434e-06, + "loss": 0.2635, + "step": 3465 + }, + { + "epoch": 1.157262103505843, + "grad_norm": 0.46322004722275933, + "learning_rate": 4.501125328061597e-06, + "loss": 0.252, + "step": 3466 + }, + { + "epoch": 1.1575959933222038, + "grad_norm": 0.4824644401886665, + "learning_rate": 4.498226214079018e-06, + "loss": 0.2728, + "step": 3467 + }, + { + "epoch": 1.1579298831385643, + "grad_norm": 0.4693106423743159, + "learning_rate": 4.495327270496222e-06, + "loss": 0.2393, + "step": 3468 + }, + { + "epoch": 1.1582637729549248, + "grad_norm": 0.4729758805797848, + "learning_rate": 4.492428498297676e-06, + "loss": 0.2576, + "step": 3469 + }, + { + "epoch": 1.1585976627712855, + "grad_norm": 0.46996337541359795, + "learning_rate": 4.489529898467789e-06, + "loss": 0.2466, + "step": 3470 + }, + { + "epoch": 1.158931552587646, + "grad_norm": 0.4599180467903547, + "learning_rate": 4.486631471990907e-06, + "loss": 0.2512, + "step": 3471 + }, + { + "epoch": 1.1592654424040068, + "grad_norm": 0.4543742458627394, + "learning_rate": 4.483733219851324e-06, + "loss": 0.2535, + "step": 3472 + }, + { + "epoch": 1.1595993322203673, + "grad_norm": 0.4655362720658223, + "learning_rate": 4.480835143033269e-06, + "loss": 0.2576, + "step": 3473 + }, + { + "epoch": 1.1599332220367278, + "grad_norm": 0.4720841205082019, + "learning_rate": 4.477937242520915e-06, + "loss": 0.2517, + "step": 3474 + }, + { + "epoch": 1.1602671118530885, + "grad_norm": 0.47412764821029224, + "learning_rate": 4.475039519298376e-06, + "loss": 0.2594, + "step": 3475 + }, + { + "epoch": 1.160601001669449, + "grad_norm": 0.47326231364464955, + "learning_rate": 4.472141974349702e-06, + "loss": 0.2552, + "step": 3476 + }, + { + "epoch": 1.1609348914858098, + "grad_norm": 0.43934852470009217, + "learning_rate": 4.469244608658884e-06, + "loss": 0.2507, + "step": 3477 + }, + { + "epoch": 1.1612687813021703, + "grad_norm": 0.4541647993464101, + "learning_rate": 4.466347423209852e-06, + "loss": 0.254, + "step": 3478 + }, + { + "epoch": 1.1616026711185308, + "grad_norm": 0.46350432878813735, + "learning_rate": 4.463450418986477e-06, + "loss": 0.2534, + "step": 3479 + }, + { + "epoch": 1.1619365609348915, + "grad_norm": 0.4487684769751664, + "learning_rate": 4.460553596972565e-06, + "loss": 0.244, + "step": 3480 + }, + { + "epoch": 1.162270450751252, + "grad_norm": 0.45860798579118883, + "learning_rate": 4.457656958151862e-06, + "loss": 0.2584, + "step": 3481 + }, + { + "epoch": 1.1626043405676127, + "grad_norm": 0.43695588534013635, + "learning_rate": 4.45476050350805e-06, + "loss": 0.2499, + "step": 3482 + }, + { + "epoch": 1.1629382303839733, + "grad_norm": 0.4598738238513854, + "learning_rate": 4.451864234024754e-06, + "loss": 0.2556, + "step": 3483 + }, + { + "epoch": 1.1632721202003338, + "grad_norm": 0.4532349556671438, + "learning_rate": 4.44896815068553e-06, + "loss": 0.2592, + "step": 3484 + }, + { + "epoch": 1.1636060100166945, + "grad_norm": 0.46253274458616567, + "learning_rate": 4.446072254473873e-06, + "loss": 0.2666, + "step": 3485 + }, + { + "epoch": 1.163939899833055, + "grad_norm": 0.4496397361186143, + "learning_rate": 4.443176546373215e-06, + "loss": 0.2541, + "step": 3486 + }, + { + "epoch": 1.1642737896494157, + "grad_norm": 0.44767261335480557, + "learning_rate": 4.440281027366922e-06, + "loss": 0.2456, + "step": 3487 + }, + { + "epoch": 1.1646076794657763, + "grad_norm": 0.4499527671582899, + "learning_rate": 4.4373856984382985e-06, + "loss": 0.2466, + "step": 3488 + }, + { + "epoch": 1.164941569282137, + "grad_norm": 0.45332262432228426, + "learning_rate": 4.434490560570583e-06, + "loss": 0.2641, + "step": 3489 + }, + { + "epoch": 1.1652754590984975, + "grad_norm": 0.47145899419117593, + "learning_rate": 4.431595614746946e-06, + "loss": 0.2459, + "step": 3490 + }, + { + "epoch": 1.165609348914858, + "grad_norm": 0.4460938924109197, + "learning_rate": 4.428700861950503e-06, + "loss": 0.2522, + "step": 3491 + }, + { + "epoch": 1.1659432387312187, + "grad_norm": 0.43202543128465276, + "learning_rate": 4.425806303164293e-06, + "loss": 0.2517, + "step": 3492 + }, + { + "epoch": 1.1662771285475793, + "grad_norm": 0.4491010077182344, + "learning_rate": 4.422911939371294e-06, + "loss": 0.257, + "step": 3493 + }, + { + "epoch": 1.16661101836394, + "grad_norm": 0.46680392058382963, + "learning_rate": 4.420017771554417e-06, + "loss": 0.2518, + "step": 3494 + }, + { + "epoch": 1.1669449081803005, + "grad_norm": 0.4789126137494215, + "learning_rate": 4.4171238006965046e-06, + "loss": 0.2671, + "step": 3495 + }, + { + "epoch": 1.167278797996661, + "grad_norm": 0.4424167012223468, + "learning_rate": 4.4142300277803365e-06, + "loss": 0.2545, + "step": 3496 + }, + { + "epoch": 1.1676126878130217, + "grad_norm": 0.4272445918947194, + "learning_rate": 4.411336453788622e-06, + "loss": 0.2473, + "step": 3497 + }, + { + "epoch": 1.1679465776293823, + "grad_norm": 0.4465689066349861, + "learning_rate": 4.408443079704005e-06, + "loss": 0.2524, + "step": 3498 + }, + { + "epoch": 1.168280467445743, + "grad_norm": 0.46666908996866957, + "learning_rate": 4.4055499065090575e-06, + "loss": 0.264, + "step": 3499 + }, + { + "epoch": 1.1686143572621035, + "grad_norm": 0.464442188357313, + "learning_rate": 4.402656935186289e-06, + "loss": 0.253, + "step": 3500 + }, + { + "epoch": 1.1689482470784642, + "grad_norm": 0.44232346640171066, + "learning_rate": 4.399764166718138e-06, + "loss": 0.2477, + "step": 3501 + }, + { + "epoch": 1.1692821368948247, + "grad_norm": 0.49040906246350385, + "learning_rate": 4.396871602086972e-06, + "loss": 0.2606, + "step": 3502 + }, + { + "epoch": 1.1696160267111853, + "grad_norm": 0.4478506095770969, + "learning_rate": 4.393979242275091e-06, + "loss": 0.2559, + "step": 3503 + }, + { + "epoch": 1.169949916527546, + "grad_norm": 0.46091446864806357, + "learning_rate": 4.3910870882647265e-06, + "loss": 0.251, + "step": 3504 + }, + { + "epoch": 1.1702838063439065, + "grad_norm": 0.4691308134837677, + "learning_rate": 4.38819514103804e-06, + "loss": 0.2539, + "step": 3505 + }, + { + "epoch": 1.1706176961602672, + "grad_norm": 0.45446686881099235, + "learning_rate": 4.3853034015771176e-06, + "loss": 0.2425, + "step": 3506 + }, + { + "epoch": 1.1709515859766277, + "grad_norm": 0.45401873639731244, + "learning_rate": 4.382411870863982e-06, + "loss": 0.2676, + "step": 3507 + }, + { + "epoch": 1.1712854757929883, + "grad_norm": 0.45697115158096185, + "learning_rate": 4.379520549880583e-06, + "loss": 0.2467, + "step": 3508 + }, + { + "epoch": 1.171619365609349, + "grad_norm": 0.47075849811094056, + "learning_rate": 4.376629439608797e-06, + "loss": 0.266, + "step": 3509 + }, + { + "epoch": 1.1719532554257095, + "grad_norm": 0.46278059583531855, + "learning_rate": 4.3737385410304314e-06, + "loss": 0.2502, + "step": 3510 + }, + { + "epoch": 1.1722871452420702, + "grad_norm": 0.46343779675674746, + "learning_rate": 4.370847855127218e-06, + "loss": 0.2549, + "step": 3511 + }, + { + "epoch": 1.1726210350584307, + "grad_norm": 0.44334378117139384, + "learning_rate": 4.36795738288082e-06, + "loss": 0.2444, + "step": 3512 + }, + { + "epoch": 1.1729549248747912, + "grad_norm": 0.4805007948606413, + "learning_rate": 4.365067125272827e-06, + "loss": 0.2575, + "step": 3513 + }, + { + "epoch": 1.173288814691152, + "grad_norm": 0.460851808817436, + "learning_rate": 4.362177083284757e-06, + "loss": 0.2635, + "step": 3514 + }, + { + "epoch": 1.1736227045075125, + "grad_norm": 0.4656154480683378, + "learning_rate": 4.3592872578980495e-06, + "loss": 0.2533, + "step": 3515 + }, + { + "epoch": 1.1739565943238732, + "grad_norm": 0.44482892719131106, + "learning_rate": 4.356397650094076e-06, + "loss": 0.2491, + "step": 3516 + }, + { + "epoch": 1.1742904841402337, + "grad_norm": 0.4466411514592481, + "learning_rate": 4.353508260854134e-06, + "loss": 0.257, + "step": 3517 + }, + { + "epoch": 1.1746243739565942, + "grad_norm": 0.4504352280889102, + "learning_rate": 4.350619091159444e-06, + "loss": 0.26, + "step": 3518 + }, + { + "epoch": 1.174958263772955, + "grad_norm": 0.4448911112512274, + "learning_rate": 4.347730141991155e-06, + "loss": 0.2543, + "step": 3519 + }, + { + "epoch": 1.1752921535893155, + "grad_norm": 0.4723590497329978, + "learning_rate": 4.344841414330334e-06, + "loss": 0.2605, + "step": 3520 + }, + { + "epoch": 1.1756260434056762, + "grad_norm": 0.459278063195679, + "learning_rate": 4.341952909157984e-06, + "loss": 0.2629, + "step": 3521 + }, + { + "epoch": 1.1759599332220367, + "grad_norm": 0.4554754675743274, + "learning_rate": 4.339064627455022e-06, + "loss": 0.2487, + "step": 3522 + }, + { + "epoch": 1.1762938230383972, + "grad_norm": 0.4447051394852914, + "learning_rate": 4.3361765702022965e-06, + "loss": 0.2353, + "step": 3523 + }, + { + "epoch": 1.176627712854758, + "grad_norm": 0.46258649497982635, + "learning_rate": 4.333288738380574e-06, + "loss": 0.2673, + "step": 3524 + }, + { + "epoch": 1.1769616026711185, + "grad_norm": 0.44759522310711736, + "learning_rate": 4.3304011329705485e-06, + "loss": 0.2605, + "step": 3525 + }, + { + "epoch": 1.1772954924874792, + "grad_norm": 0.46749688597395817, + "learning_rate": 4.327513754952838e-06, + "loss": 0.2652, + "step": 3526 + }, + { + "epoch": 1.1776293823038397, + "grad_norm": 0.44824374002240225, + "learning_rate": 4.3246266053079785e-06, + "loss": 0.2645, + "step": 3527 + }, + { + "epoch": 1.1779632721202002, + "grad_norm": 0.46931520518265346, + "learning_rate": 4.321739685016433e-06, + "loss": 0.2623, + "step": 3528 + }, + { + "epoch": 1.178297161936561, + "grad_norm": 0.48073872805320217, + "learning_rate": 4.318852995058582e-06, + "loss": 0.2601, + "step": 3529 + }, + { + "epoch": 1.1786310517529215, + "grad_norm": 0.46831158619658664, + "learning_rate": 4.315966536414733e-06, + "loss": 0.2663, + "step": 3530 + }, + { + "epoch": 1.1789649415692822, + "grad_norm": 0.4493562691230495, + "learning_rate": 4.313080310065111e-06, + "loss": 0.2495, + "step": 3531 + }, + { + "epoch": 1.1792988313856427, + "grad_norm": 0.44983785745141486, + "learning_rate": 4.3101943169898655e-06, + "loss": 0.2444, + "step": 3532 + }, + { + "epoch": 1.1796327212020032, + "grad_norm": 0.45350169719722944, + "learning_rate": 4.30730855816906e-06, + "loss": 0.253, + "step": 3533 + }, + { + "epoch": 1.179966611018364, + "grad_norm": 0.4659899811474826, + "learning_rate": 4.3044230345826885e-06, + "loss": 0.254, + "step": 3534 + }, + { + "epoch": 1.1803005008347245, + "grad_norm": 0.45097240626952056, + "learning_rate": 4.30153774721066e-06, + "loss": 0.2563, + "step": 3535 + }, + { + "epoch": 1.1806343906510852, + "grad_norm": 0.45398218922426276, + "learning_rate": 4.298652697032799e-06, + "loss": 0.2445, + "step": 3536 + }, + { + "epoch": 1.1809682804674457, + "grad_norm": 0.468340810476045, + "learning_rate": 4.2957678850288574e-06, + "loss": 0.2696, + "step": 3537 + }, + { + "epoch": 1.1813021702838062, + "grad_norm": 0.4473082455297555, + "learning_rate": 4.2928833121785e-06, + "loss": 0.247, + "step": 3538 + }, + { + "epoch": 1.181636060100167, + "grad_norm": 0.481966921541516, + "learning_rate": 4.289998979461314e-06, + "loss": 0.2683, + "step": 3539 + }, + { + "epoch": 1.1819699499165275, + "grad_norm": 0.4696747976430095, + "learning_rate": 4.287114887856803e-06, + "loss": 0.2568, + "step": 3540 + }, + { + "epoch": 1.1823038397328882, + "grad_norm": 0.45446347257676667, + "learning_rate": 4.284231038344389e-06, + "loss": 0.2582, + "step": 3541 + }, + { + "epoch": 1.1826377295492487, + "grad_norm": 0.4692061787512795, + "learning_rate": 4.281347431903416e-06, + "loss": 0.2657, + "step": 3542 + }, + { + "epoch": 1.1829716193656092, + "grad_norm": 0.44873534372934426, + "learning_rate": 4.278464069513138e-06, + "loss": 0.244, + "step": 3543 + }, + { + "epoch": 1.18330550918197, + "grad_norm": 0.4749334634786587, + "learning_rate": 4.275580952152733e-06, + "loss": 0.2677, + "step": 3544 + }, + { + "epoch": 1.1836393989983305, + "grad_norm": 0.46343197868055375, + "learning_rate": 4.272698080801289e-06, + "loss": 0.2656, + "step": 3545 + }, + { + "epoch": 1.1839732888146912, + "grad_norm": 0.4533795456409976, + "learning_rate": 4.269815456437819e-06, + "loss": 0.2582, + "step": 3546 + }, + { + "epoch": 1.1843071786310517, + "grad_norm": 0.44684422053829226, + "learning_rate": 4.266933080041243e-06, + "loss": 0.2553, + "step": 3547 + }, + { + "epoch": 1.1846410684474122, + "grad_norm": 0.43723128549865353, + "learning_rate": 4.264050952590404e-06, + "loss": 0.2473, + "step": 3548 + }, + { + "epoch": 1.184974958263773, + "grad_norm": 0.4623695185070162, + "learning_rate": 4.261169075064054e-06, + "loss": 0.2592, + "step": 3549 + }, + { + "epoch": 1.1853088480801335, + "grad_norm": 0.4743850722570731, + "learning_rate": 4.258287448440867e-06, + "loss": 0.2639, + "step": 3550 + }, + { + "epoch": 1.1856427378964942, + "grad_norm": 0.48067421530668053, + "learning_rate": 4.255406073699428e-06, + "loss": 0.25, + "step": 3551 + }, + { + "epoch": 1.1859766277128547, + "grad_norm": 0.468829678492854, + "learning_rate": 4.252524951818236e-06, + "loss": 0.2689, + "step": 3552 + }, + { + "epoch": 1.1863105175292152, + "grad_norm": 0.45890129560568976, + "learning_rate": 4.2496440837757076e-06, + "loss": 0.2535, + "step": 3553 + }, + { + "epoch": 1.186644407345576, + "grad_norm": 0.463195409068598, + "learning_rate": 4.246763470550167e-06, + "loss": 0.2564, + "step": 3554 + }, + { + "epoch": 1.1869782971619365, + "grad_norm": 0.46204795118519904, + "learning_rate": 4.243883113119858e-06, + "loss": 0.2608, + "step": 3555 + }, + { + "epoch": 1.1873121869782972, + "grad_norm": 0.45802556587855275, + "learning_rate": 4.241003012462934e-06, + "loss": 0.2529, + "step": 3556 + }, + { + "epoch": 1.1876460767946577, + "grad_norm": 0.44584960365880044, + "learning_rate": 4.2381231695574635e-06, + "loss": 0.2531, + "step": 3557 + }, + { + "epoch": 1.1879799666110185, + "grad_norm": 0.4490679773048986, + "learning_rate": 4.235243585381423e-06, + "loss": 0.2489, + "step": 3558 + }, + { + "epoch": 1.188313856427379, + "grad_norm": 0.46284341443141297, + "learning_rate": 4.2323642609127075e-06, + "loss": 0.2571, + "step": 3559 + }, + { + "epoch": 1.1886477462437395, + "grad_norm": 0.47936767386435203, + "learning_rate": 4.229485197129122e-06, + "loss": 0.2639, + "step": 3560 + }, + { + "epoch": 1.1889816360601002, + "grad_norm": 0.4460036494355798, + "learning_rate": 4.226606395008378e-06, + "loss": 0.2661, + "step": 3561 + }, + { + "epoch": 1.1893155258764607, + "grad_norm": 0.4447814462553888, + "learning_rate": 4.2237278555281045e-06, + "loss": 0.2531, + "step": 3562 + }, + { + "epoch": 1.1896494156928215, + "grad_norm": 0.471249436473038, + "learning_rate": 4.220849579665838e-06, + "loss": 0.262, + "step": 3563 + }, + { + "epoch": 1.189983305509182, + "grad_norm": 0.44233159442257236, + "learning_rate": 4.217971568399025e-06, + "loss": 0.2495, + "step": 3564 + }, + { + "epoch": 1.1903171953255425, + "grad_norm": 0.4640498890018557, + "learning_rate": 4.215093822705024e-06, + "loss": 0.2576, + "step": 3565 + }, + { + "epoch": 1.1906510851419032, + "grad_norm": 0.4545187715276503, + "learning_rate": 4.212216343561103e-06, + "loss": 0.273, + "step": 3566 + }, + { + "epoch": 1.1909849749582637, + "grad_norm": 0.4433178005677119, + "learning_rate": 4.209339131944437e-06, + "loss": 0.2596, + "step": 3567 + }, + { + "epoch": 1.1913188647746245, + "grad_norm": 0.4273335548561304, + "learning_rate": 4.206462188832114e-06, + "loss": 0.2569, + "step": 3568 + }, + { + "epoch": 1.191652754590985, + "grad_norm": 0.44866266156026563, + "learning_rate": 4.203585515201131e-06, + "loss": 0.2584, + "step": 3569 + }, + { + "epoch": 1.1919866444073457, + "grad_norm": 0.45186888543959525, + "learning_rate": 4.200709112028389e-06, + "loss": 0.2561, + "step": 3570 + }, + { + "epoch": 1.1923205342237062, + "grad_norm": 0.44609052074172767, + "learning_rate": 4.197832980290699e-06, + "loss": 0.2459, + "step": 3571 + }, + { + "epoch": 1.1926544240400667, + "grad_norm": 0.4654046084840568, + "learning_rate": 4.194957120964781e-06, + "loss": 0.2665, + "step": 3572 + }, + { + "epoch": 1.1929883138564275, + "grad_norm": 0.4435246209719868, + "learning_rate": 4.192081535027263e-06, + "loss": 0.2413, + "step": 3573 + }, + { + "epoch": 1.193322203672788, + "grad_norm": 0.44025521796322503, + "learning_rate": 4.189206223454678e-06, + "loss": 0.2544, + "step": 3574 + }, + { + "epoch": 1.1936560934891487, + "grad_norm": 0.4607722011314423, + "learning_rate": 4.1863311872234666e-06, + "loss": 0.2522, + "step": 3575 + }, + { + "epoch": 1.1939899833055092, + "grad_norm": 0.4629705528097409, + "learning_rate": 4.183456427309979e-06, + "loss": 0.2575, + "step": 3576 + }, + { + "epoch": 1.1943238731218697, + "grad_norm": 0.4580500800586019, + "learning_rate": 4.180581944690465e-06, + "loss": 0.2495, + "step": 3577 + }, + { + "epoch": 1.1946577629382304, + "grad_norm": 0.45666526637440663, + "learning_rate": 4.177707740341088e-06, + "loss": 0.2585, + "step": 3578 + }, + { + "epoch": 1.194991652754591, + "grad_norm": 0.4472547973843682, + "learning_rate": 4.1748338152379085e-06, + "loss": 0.252, + "step": 3579 + }, + { + "epoch": 1.1953255425709517, + "grad_norm": 0.43012243831613045, + "learning_rate": 4.171960170356901e-06, + "loss": 0.2463, + "step": 3580 + }, + { + "epoch": 1.1956594323873122, + "grad_norm": 0.4638809202405797, + "learning_rate": 4.169086806673938e-06, + "loss": 0.2708, + "step": 3581 + }, + { + "epoch": 1.1959933222036727, + "grad_norm": 0.45661793301717113, + "learning_rate": 4.1662137251647985e-06, + "loss": 0.26, + "step": 3582 + }, + { + "epoch": 1.1963272120200334, + "grad_norm": 0.4459616864358611, + "learning_rate": 4.1633409268051675e-06, + "loss": 0.2669, + "step": 3583 + }, + { + "epoch": 1.196661101836394, + "grad_norm": 0.4535912714167784, + "learning_rate": 4.1604684125706296e-06, + "loss": 0.2555, + "step": 3584 + }, + { + "epoch": 1.1969949916527547, + "grad_norm": 0.4644509885529257, + "learning_rate": 4.157596183436681e-06, + "loss": 0.2614, + "step": 3585 + }, + { + "epoch": 1.1973288814691152, + "grad_norm": 0.4536057299212921, + "learning_rate": 4.154724240378712e-06, + "loss": 0.2566, + "step": 3586 + }, + { + "epoch": 1.1976627712854757, + "grad_norm": 0.463004046343585, + "learning_rate": 4.151852584372021e-06, + "loss": 0.2621, + "step": 3587 + }, + { + "epoch": 1.1979966611018364, + "grad_norm": 0.4573891396622969, + "learning_rate": 4.148981216391807e-06, + "loss": 0.2593, + "step": 3588 + }, + { + "epoch": 1.198330550918197, + "grad_norm": 0.46322547246451873, + "learning_rate": 4.146110137413172e-06, + "loss": 0.2643, + "step": 3589 + }, + { + "epoch": 1.1986644407345577, + "grad_norm": 0.45158143436028303, + "learning_rate": 4.143239348411119e-06, + "loss": 0.247, + "step": 3590 + }, + { + "epoch": 1.1989983305509182, + "grad_norm": 0.4762163411915143, + "learning_rate": 4.1403688503605545e-06, + "loss": 0.2565, + "step": 3591 + }, + { + "epoch": 1.1993322203672787, + "grad_norm": 0.4735649880728535, + "learning_rate": 4.137498644236281e-06, + "loss": 0.2624, + "step": 3592 + }, + { + "epoch": 1.1996661101836394, + "grad_norm": 0.43980386950194333, + "learning_rate": 4.13462873101301e-06, + "loss": 0.2559, + "step": 3593 + }, + { + "epoch": 1.2, + "grad_norm": 0.48667331267817493, + "learning_rate": 4.131759111665349e-06, + "loss": 0.2591, + "step": 3594 + }, + { + "epoch": 1.2003338898163607, + "grad_norm": 0.46526906348632213, + "learning_rate": 4.128889787167803e-06, + "loss": 0.2569, + "step": 3595 + }, + { + "epoch": 1.2006677796327212, + "grad_norm": 0.47671623832476623, + "learning_rate": 4.1260207584947825e-06, + "loss": 0.2543, + "step": 3596 + }, + { + "epoch": 1.2010016694490817, + "grad_norm": 0.42953906431207756, + "learning_rate": 4.123152026620592e-06, + "loss": 0.251, + "step": 3597 + }, + { + "epoch": 1.2013355592654424, + "grad_norm": 0.4230112644923775, + "learning_rate": 4.120283592519442e-06, + "loss": 0.231, + "step": 3598 + }, + { + "epoch": 1.201669449081803, + "grad_norm": 0.4371425175068066, + "learning_rate": 4.117415457165434e-06, + "loss": 0.2555, + "step": 3599 + }, + { + "epoch": 1.2020033388981637, + "grad_norm": 0.4548080631639388, + "learning_rate": 4.1145476215325745e-06, + "loss": 0.2451, + "step": 3600 + }, + { + "epoch": 1.2023372287145242, + "grad_norm": 0.4428989442677222, + "learning_rate": 4.111680086594762e-06, + "loss": 0.2554, + "step": 3601 + }, + { + "epoch": 1.2026711185308847, + "grad_norm": 0.4395122949443739, + "learning_rate": 4.1088128533258005e-06, + "loss": 0.2588, + "step": 3602 + }, + { + "epoch": 1.2030050083472454, + "grad_norm": 0.4592952670372389, + "learning_rate": 4.105945922699387e-06, + "loss": 0.2646, + "step": 3603 + }, + { + "epoch": 1.203338898163606, + "grad_norm": 0.4438467841076115, + "learning_rate": 4.103079295689115e-06, + "loss": 0.252, + "step": 3604 + }, + { + "epoch": 1.2036727879799667, + "grad_norm": 0.45866060624272686, + "learning_rate": 4.100212973268478e-06, + "loss": 0.2614, + "step": 3605 + }, + { + "epoch": 1.2040066777963272, + "grad_norm": 0.4606488858291694, + "learning_rate": 4.097346956410861e-06, + "loss": 0.2381, + "step": 3606 + }, + { + "epoch": 1.2043405676126877, + "grad_norm": 0.4695474198763743, + "learning_rate": 4.094481246089553e-06, + "loss": 0.2534, + "step": 3607 + }, + { + "epoch": 1.2046744574290484, + "grad_norm": 0.44361550845460773, + "learning_rate": 4.09161584327773e-06, + "loss": 0.252, + "step": 3608 + }, + { + "epoch": 1.205008347245409, + "grad_norm": 0.4693375865271115, + "learning_rate": 4.088750748948469e-06, + "loss": 0.2594, + "step": 3609 + }, + { + "epoch": 1.2053422370617697, + "grad_norm": 0.4413883174321409, + "learning_rate": 4.0858859640747435e-06, + "loss": 0.2548, + "step": 3610 + }, + { + "epoch": 1.2056761268781302, + "grad_norm": 0.4396784819587877, + "learning_rate": 4.083021489629418e-06, + "loss": 0.2488, + "step": 3611 + }, + { + "epoch": 1.2060100166944907, + "grad_norm": 0.4565961638590267, + "learning_rate": 4.080157326585256e-06, + "loss": 0.2514, + "step": 3612 + }, + { + "epoch": 1.2063439065108514, + "grad_norm": 0.4708340340980189, + "learning_rate": 4.077293475914907e-06, + "loss": 0.2573, + "step": 3613 + }, + { + "epoch": 1.206677796327212, + "grad_norm": 0.456964961225834, + "learning_rate": 4.074429938590924e-06, + "loss": 0.2576, + "step": 3614 + }, + { + "epoch": 1.2070116861435727, + "grad_norm": 0.4396068430428025, + "learning_rate": 4.071566715585748e-06, + "loss": 0.2501, + "step": 3615 + }, + { + "epoch": 1.2073455759599332, + "grad_norm": 0.45978936443211366, + "learning_rate": 4.068703807871716e-06, + "loss": 0.2626, + "step": 3616 + }, + { + "epoch": 1.2076794657762937, + "grad_norm": 0.44802894844212554, + "learning_rate": 4.065841216421054e-06, + "loss": 0.2607, + "step": 3617 + }, + { + "epoch": 1.2080133555926544, + "grad_norm": 0.44672182371824737, + "learning_rate": 4.062978942205885e-06, + "loss": 0.2563, + "step": 3618 + }, + { + "epoch": 1.208347245409015, + "grad_norm": 0.4692549135248673, + "learning_rate": 4.060116986198223e-06, + "loss": 0.2534, + "step": 3619 + }, + { + "epoch": 1.2086811352253757, + "grad_norm": 0.4243906052871389, + "learning_rate": 4.057255349369974e-06, + "loss": 0.2368, + "step": 3620 + }, + { + "epoch": 1.2090150250417362, + "grad_norm": 0.45886632467351374, + "learning_rate": 4.054394032692933e-06, + "loss": 0.2514, + "step": 3621 + }, + { + "epoch": 1.209348914858097, + "grad_norm": 0.44882401331572025, + "learning_rate": 4.051533037138789e-06, + "loss": 0.2626, + "step": 3622 + }, + { + "epoch": 1.2096828046744574, + "grad_norm": 0.4473346037590162, + "learning_rate": 4.048672363679124e-06, + "loss": 0.2463, + "step": 3623 + }, + { + "epoch": 1.210016694490818, + "grad_norm": 0.4510642913457738, + "learning_rate": 4.045812013285403e-06, + "loss": 0.2451, + "step": 3624 + }, + { + "epoch": 1.2103505843071787, + "grad_norm": 0.4558859020367237, + "learning_rate": 4.04295198692899e-06, + "loss": 0.252, + "step": 3625 + }, + { + "epoch": 1.2106844741235392, + "grad_norm": 0.46902732328794144, + "learning_rate": 4.040092285581132e-06, + "loss": 0.251, + "step": 3626 + }, + { + "epoch": 1.2110183639399, + "grad_norm": 0.4772362207747422, + "learning_rate": 4.037232910212972e-06, + "loss": 0.2695, + "step": 3627 + }, + { + "epoch": 1.2113522537562604, + "grad_norm": 0.47273875051838016, + "learning_rate": 4.034373861795538e-06, + "loss": 0.2678, + "step": 3628 + }, + { + "epoch": 1.211686143572621, + "grad_norm": 0.4507978559764304, + "learning_rate": 4.031515141299748e-06, + "loss": 0.2495, + "step": 3629 + }, + { + "epoch": 1.2120200333889817, + "grad_norm": 0.4537528268762178, + "learning_rate": 4.02865674969641e-06, + "loss": 0.2552, + "step": 3630 + }, + { + "epoch": 1.2123539232053422, + "grad_norm": 0.4543865044860425, + "learning_rate": 4.025798687956215e-06, + "loss": 0.2562, + "step": 3631 + }, + { + "epoch": 1.212687813021703, + "grad_norm": 0.4428784780423466, + "learning_rate": 4.022940957049752e-06, + "loss": 0.2476, + "step": 3632 + }, + { + "epoch": 1.2130217028380634, + "grad_norm": 0.4518090683381747, + "learning_rate": 4.020083557947485e-06, + "loss": 0.2659, + "step": 3633 + }, + { + "epoch": 1.2133555926544242, + "grad_norm": 0.4510620826624123, + "learning_rate": 4.0172264916197765e-06, + "loss": 0.2649, + "step": 3634 + }, + { + "epoch": 1.2136894824707847, + "grad_norm": 0.45113987876357053, + "learning_rate": 4.014369759036869e-06, + "loss": 0.2554, + "step": 3635 + }, + { + "epoch": 1.2140233722871452, + "grad_norm": 0.44933794956938783, + "learning_rate": 4.011513361168895e-06, + "loss": 0.2481, + "step": 3636 + }, + { + "epoch": 1.214357262103506, + "grad_norm": 0.45751929864444785, + "learning_rate": 4.0086572989858745e-06, + "loss": 0.2538, + "step": 3637 + }, + { + "epoch": 1.2146911519198664, + "grad_norm": 0.4351287158029505, + "learning_rate": 4.005801573457709e-06, + "loss": 0.2433, + "step": 3638 + }, + { + "epoch": 1.2150250417362272, + "grad_norm": 0.4475369129104344, + "learning_rate": 4.0029461855541895e-06, + "loss": 0.2434, + "step": 3639 + }, + { + "epoch": 1.2153589315525877, + "grad_norm": 0.43533313595765777, + "learning_rate": 4.000091136244989e-06, + "loss": 0.2506, + "step": 3640 + }, + { + "epoch": 1.2156928213689482, + "grad_norm": 0.4609871270789042, + "learning_rate": 3.99723642649967e-06, + "loss": 0.2627, + "step": 3641 + }, + { + "epoch": 1.216026711185309, + "grad_norm": 0.47927530483979774, + "learning_rate": 3.994382057287674e-06, + "loss": 0.2606, + "step": 3642 + }, + { + "epoch": 1.2163606010016694, + "grad_norm": 0.45371391650315007, + "learning_rate": 3.991528029578332e-06, + "loss": 0.266, + "step": 3643 + }, + { + "epoch": 1.2166944908180302, + "grad_norm": 0.4496337317939919, + "learning_rate": 3.988674344340859e-06, + "loss": 0.2567, + "step": 3644 + }, + { + "epoch": 1.2170283806343907, + "grad_norm": 0.4719865370492984, + "learning_rate": 3.985821002544348e-06, + "loss": 0.2649, + "step": 3645 + }, + { + "epoch": 1.2173622704507512, + "grad_norm": 0.449800496725345, + "learning_rate": 3.982968005157783e-06, + "loss": 0.25, + "step": 3646 + }, + { + "epoch": 1.217696160267112, + "grad_norm": 0.48062547585825227, + "learning_rate": 3.9801153531500234e-06, + "loss": 0.2628, + "step": 3647 + }, + { + "epoch": 1.2180300500834724, + "grad_norm": 0.449943374326205, + "learning_rate": 3.9772630474898175e-06, + "loss": 0.2551, + "step": 3648 + }, + { + "epoch": 1.2183639398998332, + "grad_norm": 0.4548326259774112, + "learning_rate": 3.974411089145792e-06, + "loss": 0.2425, + "step": 3649 + }, + { + "epoch": 1.2186978297161937, + "grad_norm": 0.4401025515912481, + "learning_rate": 3.971559479086459e-06, + "loss": 0.2434, + "step": 3650 + }, + { + "epoch": 1.2190317195325542, + "grad_norm": 0.4365200416014323, + "learning_rate": 3.9687082182802076e-06, + "loss": 0.2445, + "step": 3651 + }, + { + "epoch": 1.219365609348915, + "grad_norm": 0.4625869291544512, + "learning_rate": 3.965857307695313e-06, + "loss": 0.2519, + "step": 3652 + }, + { + "epoch": 1.2196994991652754, + "grad_norm": 0.4614937695283631, + "learning_rate": 3.963006748299932e-06, + "loss": 0.2575, + "step": 3653 + }, + { + "epoch": 1.2200333889816362, + "grad_norm": 0.4340677856243796, + "learning_rate": 3.9601565410620956e-06, + "loss": 0.2492, + "step": 3654 + }, + { + "epoch": 1.2203672787979967, + "grad_norm": 0.44903175714869353, + "learning_rate": 3.957306686949723e-06, + "loss": 0.2495, + "step": 3655 + }, + { + "epoch": 1.2207011686143572, + "grad_norm": 0.46276858808502497, + "learning_rate": 3.954457186930608e-06, + "loss": 0.2507, + "step": 3656 + }, + { + "epoch": 1.221035058430718, + "grad_norm": 0.4496319481287274, + "learning_rate": 3.951608041972427e-06, + "loss": 0.2556, + "step": 3657 + }, + { + "epoch": 1.2213689482470784, + "grad_norm": 0.4501874016032137, + "learning_rate": 3.948759253042733e-06, + "loss": 0.2604, + "step": 3658 + }, + { + "epoch": 1.2217028380634392, + "grad_norm": 0.4330549952731009, + "learning_rate": 3.945910821108963e-06, + "loss": 0.2563, + "step": 3659 + }, + { + "epoch": 1.2220367278797997, + "grad_norm": 0.45210097836619906, + "learning_rate": 3.9430627471384256e-06, + "loss": 0.2665, + "step": 3660 + }, + { + "epoch": 1.2223706176961602, + "grad_norm": 0.436613694352032, + "learning_rate": 3.940215032098316e-06, + "loss": 0.2641, + "step": 3661 + }, + { + "epoch": 1.222704507512521, + "grad_norm": 0.4532034427660456, + "learning_rate": 3.937367676955703e-06, + "loss": 0.2633, + "step": 3662 + }, + { + "epoch": 1.2230383973288814, + "grad_norm": 0.44753961889946875, + "learning_rate": 3.934520682677532e-06, + "loss": 0.2604, + "step": 3663 + }, + { + "epoch": 1.2233722871452422, + "grad_norm": 0.46984772883776754, + "learning_rate": 3.931674050230628e-06, + "loss": 0.2645, + "step": 3664 + }, + { + "epoch": 1.2237061769616027, + "grad_norm": 0.4640155735606919, + "learning_rate": 3.928827780581693e-06, + "loss": 0.2643, + "step": 3665 + }, + { + "epoch": 1.2240400667779632, + "grad_norm": 0.47419217985187473, + "learning_rate": 3.925981874697306e-06, + "loss": 0.2616, + "step": 3666 + }, + { + "epoch": 1.224373956594324, + "grad_norm": 0.48041538192121114, + "learning_rate": 3.92313633354392e-06, + "loss": 0.2688, + "step": 3667 + }, + { + "epoch": 1.2247078464106844, + "grad_norm": 0.4485025615953906, + "learning_rate": 3.9202911580878685e-06, + "loss": 0.2537, + "step": 3668 + }, + { + "epoch": 1.2250417362270452, + "grad_norm": 0.4441639456777698, + "learning_rate": 3.917446349295356e-06, + "loss": 0.2516, + "step": 3669 + }, + { + "epoch": 1.2253756260434057, + "grad_norm": 0.46697243115629194, + "learning_rate": 3.914601908132466e-06, + "loss": 0.2677, + "step": 3670 + }, + { + "epoch": 1.2257095158597662, + "grad_norm": 0.44840009972726935, + "learning_rate": 3.911757835565157e-06, + "loss": 0.2475, + "step": 3671 + }, + { + "epoch": 1.226043405676127, + "grad_norm": 0.4657536908302201, + "learning_rate": 3.90891413255926e-06, + "loss": 0.2551, + "step": 3672 + }, + { + "epoch": 1.2263772954924874, + "grad_norm": 0.4643576695238107, + "learning_rate": 3.906070800080484e-06, + "loss": 0.2544, + "step": 3673 + }, + { + "epoch": 1.2267111853088482, + "grad_norm": 0.44876013384153685, + "learning_rate": 3.903227839094406e-06, + "loss": 0.2552, + "step": 3674 + }, + { + "epoch": 1.2270450751252087, + "grad_norm": 0.45902860028118136, + "learning_rate": 3.900385250566485e-06, + "loss": 0.2576, + "step": 3675 + }, + { + "epoch": 1.2273789649415692, + "grad_norm": 0.46623030934690946, + "learning_rate": 3.897543035462047e-06, + "loss": 0.2613, + "step": 3676 + }, + { + "epoch": 1.22771285475793, + "grad_norm": 0.4668752484612016, + "learning_rate": 3.894701194746291e-06, + "loss": 0.2568, + "step": 3677 + }, + { + "epoch": 1.2280467445742904, + "grad_norm": 0.46938263439740047, + "learning_rate": 3.891859729384298e-06, + "loss": 0.259, + "step": 3678 + }, + { + "epoch": 1.2283806343906511, + "grad_norm": 0.4714635984441706, + "learning_rate": 3.88901864034101e-06, + "loss": 0.256, + "step": 3679 + }, + { + "epoch": 1.2287145242070117, + "grad_norm": 0.4583963201819993, + "learning_rate": 3.886177928581248e-06, + "loss": 0.2574, + "step": 3680 + }, + { + "epoch": 1.2290484140233722, + "grad_norm": 0.4505004710157989, + "learning_rate": 3.883337595069704e-06, + "loss": 0.251, + "step": 3681 + }, + { + "epoch": 1.229382303839733, + "grad_norm": 0.44723352657872684, + "learning_rate": 3.8804976407709375e-06, + "loss": 0.2487, + "step": 3682 + }, + { + "epoch": 1.2297161936560934, + "grad_norm": 0.4767561499103855, + "learning_rate": 3.877658066649385e-06, + "loss": 0.2601, + "step": 3683 + }, + { + "epoch": 1.2300500834724541, + "grad_norm": 0.45563510260346674, + "learning_rate": 3.874818873669348e-06, + "loss": 0.2501, + "step": 3684 + }, + { + "epoch": 1.2303839732888147, + "grad_norm": 0.4457978108844736, + "learning_rate": 3.871980062795006e-06, + "loss": 0.2412, + "step": 3685 + }, + { + "epoch": 1.2307178631051752, + "grad_norm": 0.4551817174426536, + "learning_rate": 3.869141634990399e-06, + "loss": 0.2481, + "step": 3686 + }, + { + "epoch": 1.231051752921536, + "grad_norm": 0.4593226266604026, + "learning_rate": 3.866303591219449e-06, + "loss": 0.2474, + "step": 3687 + }, + { + "epoch": 1.2313856427378964, + "grad_norm": 0.4429493493414806, + "learning_rate": 3.8634659324459375e-06, + "loss": 0.2476, + "step": 3688 + }, + { + "epoch": 1.2317195325542571, + "grad_norm": 0.4811420457595111, + "learning_rate": 3.8606286596335195e-06, + "loss": 0.2743, + "step": 3689 + }, + { + "epoch": 1.2320534223706177, + "grad_norm": 0.4499880316857853, + "learning_rate": 3.857791773745717e-06, + "loss": 0.2523, + "step": 3690 + }, + { + "epoch": 1.2323873121869784, + "grad_norm": 0.4467503467058336, + "learning_rate": 3.854955275745922e-06, + "loss": 0.2609, + "step": 3691 + }, + { + "epoch": 1.232721202003339, + "grad_norm": 0.4601294533093509, + "learning_rate": 3.8521191665973965e-06, + "loss": 0.2512, + "step": 3692 + }, + { + "epoch": 1.2330550918196994, + "grad_norm": 0.45334204332476935, + "learning_rate": 3.849283447263266e-06, + "loss": 0.2686, + "step": 3693 + }, + { + "epoch": 1.2333889816360601, + "grad_norm": 0.44242064880495524, + "learning_rate": 3.846448118706525e-06, + "loss": 0.2513, + "step": 3694 + }, + { + "epoch": 1.2337228714524207, + "grad_norm": 0.4552309841001592, + "learning_rate": 3.843613181890042e-06, + "loss": 0.2622, + "step": 3695 + }, + { + "epoch": 1.2340567612687814, + "grad_norm": 0.43993977274109475, + "learning_rate": 3.840778637776541e-06, + "loss": 0.2505, + "step": 3696 + }, + { + "epoch": 1.234390651085142, + "grad_norm": 0.4660836274947334, + "learning_rate": 3.837944487328624e-06, + "loss": 0.2654, + "step": 3697 + }, + { + "epoch": 1.2347245409015024, + "grad_norm": 0.46166580221989306, + "learning_rate": 3.835110731508749e-06, + "loss": 0.2668, + "step": 3698 + }, + { + "epoch": 1.2350584307178631, + "grad_norm": 0.4517426170110665, + "learning_rate": 3.832277371279247e-06, + "loss": 0.2602, + "step": 3699 + }, + { + "epoch": 1.2353923205342237, + "grad_norm": 0.4614472385513623, + "learning_rate": 3.8294444076023115e-06, + "loss": 0.2649, + "step": 3700 + }, + { + "epoch": 1.2357262103505844, + "grad_norm": 0.46242985873383613, + "learning_rate": 3.826611841440005e-06, + "loss": 0.2711, + "step": 3701 + }, + { + "epoch": 1.236060100166945, + "grad_norm": 0.44130958518116836, + "learning_rate": 3.823779673754247e-06, + "loss": 0.2503, + "step": 3702 + }, + { + "epoch": 1.2363939899833056, + "grad_norm": 0.448781703518994, + "learning_rate": 3.820947905506831e-06, + "loss": 0.2524, + "step": 3703 + }, + { + "epoch": 1.2367278797996661, + "grad_norm": 0.44377459631633376, + "learning_rate": 3.818116537659412e-06, + "loss": 0.2508, + "step": 3704 + }, + { + "epoch": 1.2370617696160267, + "grad_norm": 0.45950486973015825, + "learning_rate": 3.8152855711735045e-06, + "loss": 0.2508, + "step": 3705 + }, + { + "epoch": 1.2373956594323874, + "grad_norm": 0.4427300288378183, + "learning_rate": 3.8124550070104938e-06, + "loss": 0.2456, + "step": 3706 + }, + { + "epoch": 1.237729549248748, + "grad_norm": 0.45118854109162215, + "learning_rate": 3.809624846131621e-06, + "loss": 0.2653, + "step": 3707 + }, + { + "epoch": 1.2380634390651086, + "grad_norm": 0.47294085873284053, + "learning_rate": 3.8067950894979965e-06, + "loss": 0.2574, + "step": 3708 + }, + { + "epoch": 1.2383973288814691, + "grad_norm": 0.45390797548809436, + "learning_rate": 3.8039657380705895e-06, + "loss": 0.2489, + "step": 3709 + }, + { + "epoch": 1.2387312186978297, + "grad_norm": 0.4515474017042673, + "learning_rate": 3.801136792810235e-06, + "loss": 0.2469, + "step": 3710 + }, + { + "epoch": 1.2390651085141904, + "grad_norm": 0.46177287792704846, + "learning_rate": 3.7983082546776244e-06, + "loss": 0.2525, + "step": 3711 + }, + { + "epoch": 1.239398998330551, + "grad_norm": 0.4644175414030979, + "learning_rate": 3.7954801246333196e-06, + "loss": 0.2598, + "step": 3712 + }, + { + "epoch": 1.2397328881469116, + "grad_norm": 0.4442203224401926, + "learning_rate": 3.7926524036377366e-06, + "loss": 0.2461, + "step": 3713 + }, + { + "epoch": 1.2400667779632721, + "grad_norm": 0.4568139961282154, + "learning_rate": 3.789825092651154e-06, + "loss": 0.2509, + "step": 3714 + }, + { + "epoch": 1.2404006677796326, + "grad_norm": 0.44697597528064614, + "learning_rate": 3.7869981926337146e-06, + "loss": 0.2501, + "step": 3715 + }, + { + "epoch": 1.2407345575959934, + "grad_norm": 0.4529861097602015, + "learning_rate": 3.784171704545416e-06, + "loss": 0.2498, + "step": 3716 + }, + { + "epoch": 1.241068447412354, + "grad_norm": 0.44284611781421573, + "learning_rate": 3.781345629346121e-06, + "loss": 0.2532, + "step": 3717 + }, + { + "epoch": 1.2414023372287146, + "grad_norm": 0.4522290754061744, + "learning_rate": 3.778519967995548e-06, + "loss": 0.2634, + "step": 3718 + }, + { + "epoch": 1.2417362270450751, + "grad_norm": 0.43823484329773527, + "learning_rate": 3.775694721453278e-06, + "loss": 0.2549, + "step": 3719 + }, + { + "epoch": 1.2420701168614356, + "grad_norm": 0.4532980195416731, + "learning_rate": 3.7728698906787527e-06, + "loss": 0.2503, + "step": 3720 + }, + { + "epoch": 1.2424040066777964, + "grad_norm": 0.4658910606407133, + "learning_rate": 3.7700454766312662e-06, + "loss": 0.2554, + "step": 3721 + }, + { + "epoch": 1.242737896494157, + "grad_norm": 0.4406907759037928, + "learning_rate": 3.767221480269978e-06, + "loss": 0.2569, + "step": 3722 + }, + { + "epoch": 1.2430717863105176, + "grad_norm": 0.46627540502076054, + "learning_rate": 3.7643979025538995e-06, + "loss": 0.2715, + "step": 3723 + }, + { + "epoch": 1.2434056761268781, + "grad_norm": 0.4482575660714646, + "learning_rate": 3.7615747444419064e-06, + "loss": 0.2528, + "step": 3724 + }, + { + "epoch": 1.2437395659432386, + "grad_norm": 0.43402645895868536, + "learning_rate": 3.758752006892726e-06, + "loss": 0.258, + "step": 3725 + }, + { + "epoch": 1.2440734557595994, + "grad_norm": 0.43156194988952984, + "learning_rate": 3.7559296908649467e-06, + "loss": 0.2445, + "step": 3726 + }, + { + "epoch": 1.2444073455759599, + "grad_norm": 0.4552006125944719, + "learning_rate": 3.7531077973170117e-06, + "loss": 0.2574, + "step": 3727 + }, + { + "epoch": 1.2447412353923206, + "grad_norm": 0.4686604867061722, + "learning_rate": 3.750286327207221e-06, + "loss": 0.2542, + "step": 3728 + }, + { + "epoch": 1.2450751252086811, + "grad_norm": 0.45402899832640625, + "learning_rate": 3.7474652814937332e-06, + "loss": 0.261, + "step": 3729 + }, + { + "epoch": 1.2454090150250416, + "grad_norm": 0.4397064445504983, + "learning_rate": 3.74464466113456e-06, + "loss": 0.2658, + "step": 3730 + }, + { + "epoch": 1.2457429048414024, + "grad_norm": 0.44280666548280356, + "learning_rate": 3.741824467087569e-06, + "loss": 0.253, + "step": 3731 + }, + { + "epoch": 1.2460767946577629, + "grad_norm": 0.4406903774195919, + "learning_rate": 3.7390047003104835e-06, + "loss": 0.2532, + "step": 3732 + }, + { + "epoch": 1.2464106844741236, + "grad_norm": 0.4838016510627983, + "learning_rate": 3.736185361760882e-06, + "loss": 0.2612, + "step": 3733 + }, + { + "epoch": 1.2467445742904841, + "grad_norm": 0.456890282764742, + "learning_rate": 3.7333664523961975e-06, + "loss": 0.2522, + "step": 3734 + }, + { + "epoch": 1.2470784641068446, + "grad_norm": 0.45652240113770776, + "learning_rate": 3.730547973173717e-06, + "loss": 0.2515, + "step": 3735 + }, + { + "epoch": 1.2474123539232054, + "grad_norm": 0.44478449866584374, + "learning_rate": 3.72772992505058e-06, + "loss": 0.2485, + "step": 3736 + }, + { + "epoch": 1.2477462437395659, + "grad_norm": 0.45064553259389273, + "learning_rate": 3.724912308983783e-06, + "loss": 0.2521, + "step": 3737 + }, + { + "epoch": 1.2480801335559266, + "grad_norm": 0.46186348156260004, + "learning_rate": 3.7220951259301742e-06, + "loss": 0.2645, + "step": 3738 + }, + { + "epoch": 1.2484140233722871, + "grad_norm": 0.46016790297273147, + "learning_rate": 3.7192783768464523e-06, + "loss": 0.2575, + "step": 3739 + }, + { + "epoch": 1.2487479131886476, + "grad_norm": 0.47945430908823594, + "learning_rate": 3.7164620626891724e-06, + "loss": 0.2647, + "step": 3740 + }, + { + "epoch": 1.2490818030050084, + "grad_norm": 0.4625868729632643, + "learning_rate": 3.7136461844147392e-06, + "loss": 0.2581, + "step": 3741 + }, + { + "epoch": 1.2494156928213689, + "grad_norm": 0.45541637079794367, + "learning_rate": 3.710830742979412e-06, + "loss": 0.2638, + "step": 3742 + }, + { + "epoch": 1.2497495826377296, + "grad_norm": 0.44116499309499707, + "learning_rate": 3.708015739339298e-06, + "loss": 0.2415, + "step": 3743 + }, + { + "epoch": 1.2500834724540901, + "grad_norm": 0.45586925697960345, + "learning_rate": 3.7052011744503593e-06, + "loss": 0.2541, + "step": 3744 + }, + { + "epoch": 1.2504173622704506, + "grad_norm": 0.4496871346281184, + "learning_rate": 3.7023870492684055e-06, + "loss": 0.2573, + "step": 3745 + }, + { + "epoch": 1.2507512520868114, + "grad_norm": 0.48596468769359114, + "learning_rate": 3.6995733647491016e-06, + "loss": 0.2554, + "step": 3746 + }, + { + "epoch": 1.2510851419031719, + "grad_norm": 0.4528407982051353, + "learning_rate": 3.6967601218479604e-06, + "loss": 0.2488, + "step": 3747 + }, + { + "epoch": 1.2514190317195326, + "grad_norm": 0.4443302674444962, + "learning_rate": 3.6939473215203424e-06, + "loss": 0.2377, + "step": 3748 + }, + { + "epoch": 1.2517529215358931, + "grad_norm": 0.4496872982713112, + "learning_rate": 3.6911349647214623e-06, + "loss": 0.2555, + "step": 3749 + }, + { + "epoch": 1.2520868113522536, + "grad_norm": 0.4430027306163416, + "learning_rate": 3.6883230524063802e-06, + "loss": 0.2492, + "step": 3750 + }, + { + "epoch": 1.2524207011686144, + "grad_norm": 0.4633468031259995, + "learning_rate": 3.685511585530008e-06, + "loss": 0.2617, + "step": 3751 + }, + { + "epoch": 1.2527545909849749, + "grad_norm": 0.4775954578237915, + "learning_rate": 3.682700565047105e-06, + "loss": 0.2669, + "step": 3752 + }, + { + "epoch": 1.2530884808013356, + "grad_norm": 0.44946215233921, + "learning_rate": 3.6798899919122773e-06, + "loss": 0.262, + "step": 3753 + }, + { + "epoch": 1.2534223706176961, + "grad_norm": 0.4551421050269747, + "learning_rate": 3.677079867079985e-06, + "loss": 0.2483, + "step": 3754 + }, + { + "epoch": 1.2537562604340566, + "grad_norm": 0.4569375326753469, + "learning_rate": 3.6742701915045285e-06, + "loss": 0.2565, + "step": 3755 + }, + { + "epoch": 1.2540901502504174, + "grad_norm": 0.4612633658659809, + "learning_rate": 3.6714609661400624e-06, + "loss": 0.2618, + "step": 3756 + }, + { + "epoch": 1.254424040066778, + "grad_norm": 0.4493931304563577, + "learning_rate": 3.6686521919405814e-06, + "loss": 0.2421, + "step": 3757 + }, + { + "epoch": 1.2547579298831386, + "grad_norm": 0.46864068794198466, + "learning_rate": 3.665843869859934e-06, + "loss": 0.2507, + "step": 3758 + }, + { + "epoch": 1.2550918196994991, + "grad_norm": 0.48994281920169597, + "learning_rate": 3.6630360008518094e-06, + "loss": 0.2586, + "step": 3759 + }, + { + "epoch": 1.2554257095158596, + "grad_norm": 0.4851264509713167, + "learning_rate": 3.6602285858697474e-06, + "loss": 0.249, + "step": 3760 + }, + { + "epoch": 1.2557595993322204, + "grad_norm": 0.46904575262045384, + "learning_rate": 3.6574216258671287e-06, + "loss": 0.2537, + "step": 3761 + }, + { + "epoch": 1.256093489148581, + "grad_norm": 0.46419351351574123, + "learning_rate": 3.6546151217971836e-06, + "loss": 0.2588, + "step": 3762 + }, + { + "epoch": 1.2564273789649416, + "grad_norm": 0.4532476295292924, + "learning_rate": 3.6518090746129884e-06, + "loss": 0.2499, + "step": 3763 + }, + { + "epoch": 1.2567612687813021, + "grad_norm": 0.4691564757306817, + "learning_rate": 3.649003485267461e-06, + "loss": 0.2556, + "step": 3764 + }, + { + "epoch": 1.2570951585976629, + "grad_norm": 0.4574486944236121, + "learning_rate": 3.646198354713365e-06, + "loss": 0.2568, + "step": 3765 + }, + { + "epoch": 1.2574290484140234, + "grad_norm": 0.44263375461842835, + "learning_rate": 3.643393683903308e-06, + "loss": 0.2521, + "step": 3766 + }, + { + "epoch": 1.257762938230384, + "grad_norm": 0.4530461450167133, + "learning_rate": 3.640589473789742e-06, + "loss": 0.2496, + "step": 3767 + }, + { + "epoch": 1.2580968280467446, + "grad_norm": 0.4537818333267677, + "learning_rate": 3.6377857253249616e-06, + "loss": 0.2611, + "step": 3768 + }, + { + "epoch": 1.2584307178631051, + "grad_norm": 0.4585227463154636, + "learning_rate": 3.634982439461106e-06, + "loss": 0.2517, + "step": 3769 + }, + { + "epoch": 1.2587646076794659, + "grad_norm": 0.45363842156042183, + "learning_rate": 3.6321796171501546e-06, + "loss": 0.2583, + "step": 3770 + }, + { + "epoch": 1.2590984974958264, + "grad_norm": 0.4430763453390645, + "learning_rate": 3.629377259343934e-06, + "loss": 0.2566, + "step": 3771 + }, + { + "epoch": 1.259432387312187, + "grad_norm": 0.4417470271075849, + "learning_rate": 3.6265753669941104e-06, + "loss": 0.2606, + "step": 3772 + }, + { + "epoch": 1.2597662771285476, + "grad_norm": 0.4452557153602873, + "learning_rate": 3.623773941052191e-06, + "loss": 0.2413, + "step": 3773 + }, + { + "epoch": 1.2601001669449081, + "grad_norm": 0.44243983463084363, + "learning_rate": 3.6209729824695262e-06, + "loss": 0.2455, + "step": 3774 + }, + { + "epoch": 1.2604340567612689, + "grad_norm": 0.46478853935474523, + "learning_rate": 3.618172492197305e-06, + "loss": 0.2646, + "step": 3775 + }, + { + "epoch": 1.2607679465776294, + "grad_norm": 0.4656089511781191, + "learning_rate": 3.6153724711865623e-06, + "loss": 0.255, + "step": 3776 + }, + { + "epoch": 1.26110183639399, + "grad_norm": 0.4340415603022293, + "learning_rate": 3.6125729203881676e-06, + "loss": 0.2541, + "step": 3777 + }, + { + "epoch": 1.2614357262103506, + "grad_norm": 0.44452128140804587, + "learning_rate": 3.609773840752837e-06, + "loss": 0.2491, + "step": 3778 + }, + { + "epoch": 1.2617696160267111, + "grad_norm": 0.43754151958522747, + "learning_rate": 3.6069752332311203e-06, + "loss": 0.249, + "step": 3779 + }, + { + "epoch": 1.2621035058430718, + "grad_norm": 0.4618335762106439, + "learning_rate": 3.604177098773414e-06, + "loss": 0.2591, + "step": 3780 + }, + { + "epoch": 1.2624373956594324, + "grad_norm": 0.46932229464348624, + "learning_rate": 3.6013794383299474e-06, + "loss": 0.249, + "step": 3781 + }, + { + "epoch": 1.262771285475793, + "grad_norm": 0.477061154943901, + "learning_rate": 3.598582252850792e-06, + "loss": 0.2641, + "step": 3782 + }, + { + "epoch": 1.2631051752921536, + "grad_norm": 0.45887501971746647, + "learning_rate": 3.595785543285858e-06, + "loss": 0.2522, + "step": 3783 + }, + { + "epoch": 1.2634390651085141, + "grad_norm": 0.4649830383028245, + "learning_rate": 3.592989310584893e-06, + "loss": 0.2599, + "step": 3784 + }, + { + "epoch": 1.2637729549248748, + "grad_norm": 0.4687332523888459, + "learning_rate": 3.5901935556974837e-06, + "loss": 0.26, + "step": 3785 + }, + { + "epoch": 1.2641068447412354, + "grad_norm": 0.4665713252411812, + "learning_rate": 3.5873982795730533e-06, + "loss": 0.2564, + "step": 3786 + }, + { + "epoch": 1.264440734557596, + "grad_norm": 0.4644417695497607, + "learning_rate": 3.5846034831608617e-06, + "loss": 0.2567, + "step": 3787 + }, + { + "epoch": 1.2647746243739566, + "grad_norm": 0.44584668606543487, + "learning_rate": 3.581809167410012e-06, + "loss": 0.2425, + "step": 3788 + }, + { + "epoch": 1.2651085141903171, + "grad_norm": 0.46717306240394524, + "learning_rate": 3.5790153332694344e-06, + "loss": 0.2496, + "step": 3789 + }, + { + "epoch": 1.2654424040066778, + "grad_norm": 0.4742052656764265, + "learning_rate": 3.576221981687904e-06, + "loss": 0.2583, + "step": 3790 + }, + { + "epoch": 1.2657762938230384, + "grad_norm": 0.4669002984559574, + "learning_rate": 3.5734291136140265e-06, + "loss": 0.2506, + "step": 3791 + }, + { + "epoch": 1.266110183639399, + "grad_norm": 0.4637678151153122, + "learning_rate": 3.5706367299962474e-06, + "loss": 0.2643, + "step": 3792 + }, + { + "epoch": 1.2664440734557596, + "grad_norm": 0.45348298539779985, + "learning_rate": 3.5678448317828427e-06, + "loss": 0.2549, + "step": 3793 + }, + { + "epoch": 1.2667779632721201, + "grad_norm": 0.45022100926704295, + "learning_rate": 3.56505341992193e-06, + "loss": 0.2566, + "step": 3794 + }, + { + "epoch": 1.2671118530884808, + "grad_norm": 0.4405188559419773, + "learning_rate": 3.5622624953614554e-06, + "loss": 0.2488, + "step": 3795 + }, + { + "epoch": 1.2674457429048414, + "grad_norm": 0.4452924833431293, + "learning_rate": 3.559472059049203e-06, + "loss": 0.2563, + "step": 3796 + }, + { + "epoch": 1.267779632721202, + "grad_norm": 0.4375931008436299, + "learning_rate": 3.5566821119327927e-06, + "loss": 0.2492, + "step": 3797 + }, + { + "epoch": 1.2681135225375626, + "grad_norm": 0.439126740627671, + "learning_rate": 3.553892654959674e-06, + "loss": 0.254, + "step": 3798 + }, + { + "epoch": 1.268447412353923, + "grad_norm": 0.44954377778125126, + "learning_rate": 3.5511036890771337e-06, + "loss": 0.2479, + "step": 3799 + }, + { + "epoch": 1.2687813021702838, + "grad_norm": 0.4665759643603475, + "learning_rate": 3.5483152152322884e-06, + "loss": 0.2554, + "step": 3800 + }, + { + "epoch": 1.2691151919866444, + "grad_norm": 0.45383168424685033, + "learning_rate": 3.5455272343720915e-06, + "loss": 0.243, + "step": 3801 + }, + { + "epoch": 1.269449081803005, + "grad_norm": 0.44423326287685055, + "learning_rate": 3.5427397474433238e-06, + "loss": 0.2551, + "step": 3802 + }, + { + "epoch": 1.2697829716193656, + "grad_norm": 0.46732826157596713, + "learning_rate": 3.539952755392605e-06, + "loss": 0.2613, + "step": 3803 + }, + { + "epoch": 1.270116861435726, + "grad_norm": 0.4637387821816155, + "learning_rate": 3.537166259166379e-06, + "loss": 0.264, + "step": 3804 + }, + { + "epoch": 1.2704507512520868, + "grad_norm": 0.47470187188496643, + "learning_rate": 3.53438025971093e-06, + "loss": 0.2601, + "step": 3805 + }, + { + "epoch": 1.2707846410684474, + "grad_norm": 0.455522527094802, + "learning_rate": 3.5315947579723677e-06, + "loss": 0.2499, + "step": 3806 + }, + { + "epoch": 1.271118530884808, + "grad_norm": 0.439807544028952, + "learning_rate": 3.5288097548966314e-06, + "loss": 0.2487, + "step": 3807 + }, + { + "epoch": 1.2714524207011686, + "grad_norm": 0.45303452447574744, + "learning_rate": 3.5260252514294974e-06, + "loss": 0.2534, + "step": 3808 + }, + { + "epoch": 1.271786310517529, + "grad_norm": 0.421115819489531, + "learning_rate": 3.5232412485165657e-06, + "loss": 0.237, + "step": 3809 + }, + { + "epoch": 1.2721202003338898, + "grad_norm": 0.4747696322070975, + "learning_rate": 3.520457747103272e-06, + "loss": 0.2581, + "step": 3810 + }, + { + "epoch": 1.2724540901502503, + "grad_norm": 0.46183245023973357, + "learning_rate": 3.5176747481348774e-06, + "loss": 0.2635, + "step": 3811 + }, + { + "epoch": 1.272787979966611, + "grad_norm": 0.4523858756274051, + "learning_rate": 3.514892252556474e-06, + "loss": 0.2436, + "step": 3812 + }, + { + "epoch": 1.2731218697829716, + "grad_norm": 0.44324465379745415, + "learning_rate": 3.512110261312981e-06, + "loss": 0.2476, + "step": 3813 + }, + { + "epoch": 1.273455759599332, + "grad_norm": 0.4891124518295324, + "learning_rate": 3.5093287753491515e-06, + "loss": 0.2652, + "step": 3814 + }, + { + "epoch": 1.2737896494156928, + "grad_norm": 0.4619333620567034, + "learning_rate": 3.506547795609564e-06, + "loss": 0.2559, + "step": 3815 + }, + { + "epoch": 1.2741235392320533, + "grad_norm": 0.4534032122675086, + "learning_rate": 3.503767323038622e-06, + "loss": 0.2643, + "step": 3816 + }, + { + "epoch": 1.274457429048414, + "grad_norm": 0.45641292786116927, + "learning_rate": 3.500987358580562e-06, + "loss": 0.2474, + "step": 3817 + }, + { + "epoch": 1.2747913188647746, + "grad_norm": 0.45783781936441786, + "learning_rate": 3.498207903179444e-06, + "loss": 0.2572, + "step": 3818 + }, + { + "epoch": 1.275125208681135, + "grad_norm": 0.473135381378946, + "learning_rate": 3.495428957779158e-06, + "loss": 0.2575, + "step": 3819 + }, + { + "epoch": 1.2754590984974958, + "grad_norm": 0.4433032560681486, + "learning_rate": 3.4926505233234177e-06, + "loss": 0.243, + "step": 3820 + }, + { + "epoch": 1.2757929883138563, + "grad_norm": 0.45582621033185855, + "learning_rate": 3.4898726007557655e-06, + "loss": 0.2461, + "step": 3821 + }, + { + "epoch": 1.276126878130217, + "grad_norm": 0.4475570867166873, + "learning_rate": 3.4870951910195716e-06, + "loss": 0.2605, + "step": 3822 + }, + { + "epoch": 1.2764607679465776, + "grad_norm": 0.4539904999527105, + "learning_rate": 3.484318295058028e-06, + "loss": 0.2516, + "step": 3823 + }, + { + "epoch": 1.276794657762938, + "grad_norm": 0.45066616232021584, + "learning_rate": 3.481541913814156e-06, + "loss": 0.242, + "step": 3824 + }, + { + "epoch": 1.2771285475792988, + "grad_norm": 0.45686475194594844, + "learning_rate": 3.478766048230797e-06, + "loss": 0.2596, + "step": 3825 + }, + { + "epoch": 1.2774624373956596, + "grad_norm": 0.4761594773445954, + "learning_rate": 3.4759906992506244e-06, + "loss": 0.2591, + "step": 3826 + }, + { + "epoch": 1.27779632721202, + "grad_norm": 0.43984842412026115, + "learning_rate": 3.47321586781613e-06, + "loss": 0.24, + "step": 3827 + }, + { + "epoch": 1.2781302170283806, + "grad_norm": 0.4594919423533571, + "learning_rate": 3.4704415548696335e-06, + "loss": 0.2518, + "step": 3828 + }, + { + "epoch": 1.278464106844741, + "grad_norm": 0.47135392272898813, + "learning_rate": 3.467667761353275e-06, + "loss": 0.2647, + "step": 3829 + }, + { + "epoch": 1.2787979966611018, + "grad_norm": 0.45762936546807387, + "learning_rate": 3.464894488209022e-06, + "loss": 0.2651, + "step": 3830 + }, + { + "epoch": 1.2791318864774626, + "grad_norm": 0.47427027858061194, + "learning_rate": 3.462121736378665e-06, + "loss": 0.2616, + "step": 3831 + }, + { + "epoch": 1.279465776293823, + "grad_norm": 0.44390171724887345, + "learning_rate": 3.4593495068038142e-06, + "loss": 0.2379, + "step": 3832 + }, + { + "epoch": 1.2797996661101836, + "grad_norm": 0.49479481726867913, + "learning_rate": 3.4565778004259064e-06, + "loss": 0.2597, + "step": 3833 + }, + { + "epoch": 1.2801335559265443, + "grad_norm": 0.4517124951545255, + "learning_rate": 3.4538066181861964e-06, + "loss": 0.259, + "step": 3834 + }, + { + "epoch": 1.2804674457429048, + "grad_norm": 0.4400562889502232, + "learning_rate": 3.451035961025765e-06, + "loss": 0.2571, + "step": 3835 + }, + { + "epoch": 1.2808013355592656, + "grad_norm": 0.4383673021901867, + "learning_rate": 3.448265829885512e-06, + "loss": 0.2503, + "step": 3836 + }, + { + "epoch": 1.281135225375626, + "grad_norm": 0.44680542292611397, + "learning_rate": 3.445496225706161e-06, + "loss": 0.2474, + "step": 3837 + }, + { + "epoch": 1.2814691151919866, + "grad_norm": 0.46104284338220586, + "learning_rate": 3.442727149428252e-06, + "loss": 0.2491, + "step": 3838 + }, + { + "epoch": 1.2818030050083473, + "grad_norm": 0.45671207826333915, + "learning_rate": 3.439958601992153e-06, + "loss": 0.2536, + "step": 3839 + }, + { + "epoch": 1.2821368948247078, + "grad_norm": 0.46619378207447376, + "learning_rate": 3.4371905843380483e-06, + "loss": 0.2504, + "step": 3840 + }, + { + "epoch": 1.2824707846410686, + "grad_norm": 0.453357890954882, + "learning_rate": 3.4344230974059388e-06, + "loss": 0.2508, + "step": 3841 + }, + { + "epoch": 1.282804674457429, + "grad_norm": 0.4532752590070186, + "learning_rate": 3.431656142135652e-06, + "loss": 0.2471, + "step": 3842 + }, + { + "epoch": 1.2831385642737896, + "grad_norm": 0.4541523637353768, + "learning_rate": 3.428889719466829e-06, + "loss": 0.2404, + "step": 3843 + }, + { + "epoch": 1.2834724540901503, + "grad_norm": 0.5003719747504092, + "learning_rate": 3.426123830338936e-06, + "loss": 0.2679, + "step": 3844 + }, + { + "epoch": 1.2838063439065108, + "grad_norm": 0.45364665260380566, + "learning_rate": 3.4233584756912513e-06, + "loss": 0.2532, + "step": 3845 + }, + { + "epoch": 1.2841402337228716, + "grad_norm": 0.4490860774538949, + "learning_rate": 3.420593656462877e-06, + "loss": 0.2495, + "step": 3846 + }, + { + "epoch": 1.284474123539232, + "grad_norm": 0.45185607686512513, + "learning_rate": 3.4178293735927298e-06, + "loss": 0.2535, + "step": 3847 + }, + { + "epoch": 1.2848080133555926, + "grad_norm": 0.47793296496232873, + "learning_rate": 3.415065628019547e-06, + "loss": 0.2624, + "step": 3848 + }, + { + "epoch": 1.2851419031719533, + "grad_norm": 0.4569520639205432, + "learning_rate": 3.4123024206818846e-06, + "loss": 0.251, + "step": 3849 + }, + { + "epoch": 1.2854757929883138, + "grad_norm": 0.43538663841027647, + "learning_rate": 3.4095397525181084e-06, + "loss": 0.2427, + "step": 3850 + }, + { + "epoch": 1.2858096828046746, + "grad_norm": 0.44961296896662595, + "learning_rate": 3.406777624466411e-06, + "loss": 0.2524, + "step": 3851 + }, + { + "epoch": 1.286143572621035, + "grad_norm": 0.45247128553667243, + "learning_rate": 3.4040160374647936e-06, + "loss": 0.255, + "step": 3852 + }, + { + "epoch": 1.2864774624373956, + "grad_norm": 0.4408055578330588, + "learning_rate": 3.401254992451079e-06, + "loss": 0.2513, + "step": 3853 + }, + { + "epoch": 1.2868113522537563, + "grad_norm": 0.45139467231286534, + "learning_rate": 3.398494490362902e-06, + "loss": 0.2539, + "step": 3854 + }, + { + "epoch": 1.2871452420701168, + "grad_norm": 0.4617024846061824, + "learning_rate": 3.3957345321377143e-06, + "loss": 0.2607, + "step": 3855 + }, + { + "epoch": 1.2874791318864776, + "grad_norm": 0.45234977566790213, + "learning_rate": 3.392975118712787e-06, + "loss": 0.2601, + "step": 3856 + }, + { + "epoch": 1.287813021702838, + "grad_norm": 0.44956005437754204, + "learning_rate": 3.3902162510252e-06, + "loss": 0.2425, + "step": 3857 + }, + { + "epoch": 1.2881469115191986, + "grad_norm": 0.44633186193687596, + "learning_rate": 3.3874579300118515e-06, + "loss": 0.2493, + "step": 3858 + }, + { + "epoch": 1.2884808013355593, + "grad_norm": 0.4415992392682934, + "learning_rate": 3.3847001566094518e-06, + "loss": 0.2441, + "step": 3859 + }, + { + "epoch": 1.2888146911519198, + "grad_norm": 0.43747812016074994, + "learning_rate": 3.381942931754528e-06, + "loss": 0.2526, + "step": 3860 + }, + { + "epoch": 1.2891485809682806, + "grad_norm": 0.4617081669899412, + "learning_rate": 3.379186256383418e-06, + "loss": 0.259, + "step": 3861 + }, + { + "epoch": 1.289482470784641, + "grad_norm": 0.4681556898914822, + "learning_rate": 3.376430131432275e-06, + "loss": 0.2566, + "step": 3862 + }, + { + "epoch": 1.2898163606010016, + "grad_norm": 0.4639059064779521, + "learning_rate": 3.3736745578370654e-06, + "loss": 0.2596, + "step": 3863 + }, + { + "epoch": 1.2901502504173623, + "grad_norm": 0.4514113348984263, + "learning_rate": 3.3709195365335646e-06, + "loss": 0.2479, + "step": 3864 + }, + { + "epoch": 1.2904841402337228, + "grad_norm": 0.44644267134550497, + "learning_rate": 3.3681650684573695e-06, + "loss": 0.2487, + "step": 3865 + }, + { + "epoch": 1.2908180300500836, + "grad_norm": 0.4466130236362784, + "learning_rate": 3.365411154543878e-06, + "loss": 0.2514, + "step": 3866 + }, + { + "epoch": 1.291151919866444, + "grad_norm": 0.461302426589445, + "learning_rate": 3.362657795728309e-06, + "loss": 0.2544, + "step": 3867 + }, + { + "epoch": 1.2914858096828046, + "grad_norm": 0.44962610397940234, + "learning_rate": 3.3599049929456857e-06, + "loss": 0.2607, + "step": 3868 + }, + { + "epoch": 1.2918196994991653, + "grad_norm": 0.45713631226778867, + "learning_rate": 3.3571527471308476e-06, + "loss": 0.2448, + "step": 3869 + }, + { + "epoch": 1.2921535893155258, + "grad_norm": 0.47215512323344805, + "learning_rate": 3.3544010592184428e-06, + "loss": 0.2591, + "step": 3870 + }, + { + "epoch": 1.2924874791318866, + "grad_norm": 0.4726473559260972, + "learning_rate": 3.3516499301429282e-06, + "loss": 0.248, + "step": 3871 + }, + { + "epoch": 1.292821368948247, + "grad_norm": 0.4563259491499186, + "learning_rate": 3.3488993608385745e-06, + "loss": 0.2556, + "step": 3872 + }, + { + "epoch": 1.2931552587646076, + "grad_norm": 0.4258781213554337, + "learning_rate": 3.3461493522394616e-06, + "loss": 0.2369, + "step": 3873 + }, + { + "epoch": 1.2934891485809683, + "grad_norm": 0.4400258349316607, + "learning_rate": 3.343399905279479e-06, + "loss": 0.2496, + "step": 3874 + }, + { + "epoch": 1.2938230383973288, + "grad_norm": 0.44743324630544157, + "learning_rate": 3.3406510208923228e-06, + "loss": 0.2575, + "step": 3875 + }, + { + "epoch": 1.2941569282136896, + "grad_norm": 0.46659786884031607, + "learning_rate": 3.3379027000115005e-06, + "loss": 0.2522, + "step": 3876 + }, + { + "epoch": 1.29449081803005, + "grad_norm": 0.44928478082175377, + "learning_rate": 3.335154943570328e-06, + "loss": 0.2545, + "step": 3877 + }, + { + "epoch": 1.2948247078464106, + "grad_norm": 0.4538762258124005, + "learning_rate": 3.332407752501928e-06, + "loss": 0.2543, + "step": 3878 + }, + { + "epoch": 1.2951585976627713, + "grad_norm": 0.46644115678568526, + "learning_rate": 3.329661127739234e-06, + "loss": 0.2474, + "step": 3879 + }, + { + "epoch": 1.2954924874791318, + "grad_norm": 0.4350550696866848, + "learning_rate": 3.3269150702149834e-06, + "loss": 0.2519, + "step": 3880 + }, + { + "epoch": 1.2958263772954925, + "grad_norm": 0.45106977367358325, + "learning_rate": 3.3241695808617235e-06, + "loss": 0.2499, + "step": 3881 + }, + { + "epoch": 1.296160267111853, + "grad_norm": 0.46161000139074815, + "learning_rate": 3.3214246606118116e-06, + "loss": 0.2523, + "step": 3882 + }, + { + "epoch": 1.2964941569282136, + "grad_norm": 0.4815964426339726, + "learning_rate": 3.3186803103974054e-06, + "loss": 0.2677, + "step": 3883 + }, + { + "epoch": 1.2968280467445743, + "grad_norm": 0.46300596084771956, + "learning_rate": 3.3159365311504732e-06, + "loss": 0.2518, + "step": 3884 + }, + { + "epoch": 1.2971619365609348, + "grad_norm": 0.43853400364797956, + "learning_rate": 3.3131933238027877e-06, + "loss": 0.2409, + "step": 3885 + }, + { + "epoch": 1.2974958263772955, + "grad_norm": 0.4516424567787992, + "learning_rate": 3.310450689285928e-06, + "loss": 0.2508, + "step": 3886 + }, + { + "epoch": 1.297829716193656, + "grad_norm": 0.4645652937118097, + "learning_rate": 3.307708628531278e-06, + "loss": 0.2621, + "step": 3887 + }, + { + "epoch": 1.2981636060100166, + "grad_norm": 0.4547554917184767, + "learning_rate": 3.3049671424700293e-06, + "loss": 0.2695, + "step": 3888 + }, + { + "epoch": 1.2984974958263773, + "grad_norm": 0.4724158736755732, + "learning_rate": 3.302226232033172e-06, + "loss": 0.2617, + "step": 3889 + }, + { + "epoch": 1.298831385642738, + "grad_norm": 0.4453206244709202, + "learning_rate": 3.2994858981515103e-06, + "loss": 0.2551, + "step": 3890 + }, + { + "epoch": 1.2991652754590985, + "grad_norm": 0.43781899853878364, + "learning_rate": 3.296746141755646e-06, + "loss": 0.2539, + "step": 3891 + }, + { + "epoch": 1.299499165275459, + "grad_norm": 0.44010027350810327, + "learning_rate": 3.2940069637759843e-06, + "loss": 0.251, + "step": 3892 + }, + { + "epoch": 1.2998330550918196, + "grad_norm": 0.45184007981060237, + "learning_rate": 3.291268365142738e-06, + "loss": 0.2559, + "step": 3893 + }, + { + "epoch": 1.3001669449081803, + "grad_norm": 0.4493010278783117, + "learning_rate": 3.2885303467859196e-06, + "loss": 0.2529, + "step": 3894 + }, + { + "epoch": 1.300500834724541, + "grad_norm": 0.47434329088485533, + "learning_rate": 3.285792909635347e-06, + "loss": 0.2577, + "step": 3895 + }, + { + "epoch": 1.3008347245409015, + "grad_norm": 0.47834971254839914, + "learning_rate": 3.2830560546206392e-06, + "loss": 0.2647, + "step": 3896 + }, + { + "epoch": 1.301168614357262, + "grad_norm": 0.4493678582519355, + "learning_rate": 3.280319782671218e-06, + "loss": 0.2557, + "step": 3897 + }, + { + "epoch": 1.3015025041736228, + "grad_norm": 0.4455775849558845, + "learning_rate": 3.277584094716305e-06, + "loss": 0.2558, + "step": 3898 + }, + { + "epoch": 1.3018363939899833, + "grad_norm": 0.4569636276839816, + "learning_rate": 3.27484899168493e-06, + "loss": 0.2507, + "step": 3899 + }, + { + "epoch": 1.302170283806344, + "grad_norm": 0.46907451489747193, + "learning_rate": 3.2721144745059173e-06, + "loss": 0.2631, + "step": 3900 + }, + { + "epoch": 1.3025041736227045, + "grad_norm": 0.43975623581395706, + "learning_rate": 3.2693805441078956e-06, + "loss": 0.252, + "step": 3901 + }, + { + "epoch": 1.302838063439065, + "grad_norm": 0.4525395772602552, + "learning_rate": 3.2666472014192942e-06, + "loss": 0.2532, + "step": 3902 + }, + { + "epoch": 1.3031719532554258, + "grad_norm": 0.45146541624297637, + "learning_rate": 3.2639144473683393e-06, + "loss": 0.2513, + "step": 3903 + }, + { + "epoch": 1.3035058430717863, + "grad_norm": 0.43351301058849206, + "learning_rate": 3.261182282883063e-06, + "loss": 0.2399, + "step": 3904 + }, + { + "epoch": 1.303839732888147, + "grad_norm": 0.4695196311483116, + "learning_rate": 3.258450708891292e-06, + "loss": 0.2549, + "step": 3905 + }, + { + "epoch": 1.3041736227045075, + "grad_norm": 0.45865501643487866, + "learning_rate": 3.255719726320655e-06, + "loss": 0.2418, + "step": 3906 + }, + { + "epoch": 1.304507512520868, + "grad_norm": 0.43416630349427243, + "learning_rate": 3.252989336098583e-06, + "loss": 0.2401, + "step": 3907 + }, + { + "epoch": 1.3048414023372288, + "grad_norm": 0.463116274664072, + "learning_rate": 3.2502595391522985e-06, + "loss": 0.2503, + "step": 3908 + }, + { + "epoch": 1.3051752921535893, + "grad_norm": 0.44161065361658847, + "learning_rate": 3.2475303364088283e-06, + "loss": 0.242, + "step": 3909 + }, + { + "epoch": 1.30550918196995, + "grad_norm": 0.458466326454986, + "learning_rate": 3.244801728794994e-06, + "loss": 0.2582, + "step": 3910 + }, + { + "epoch": 1.3058430717863105, + "grad_norm": 0.45834230326610714, + "learning_rate": 3.2420737172374184e-06, + "loss": 0.2416, + "step": 3911 + }, + { + "epoch": 1.306176961602671, + "grad_norm": 0.4476277411819857, + "learning_rate": 3.2393463026625183e-06, + "loss": 0.25, + "step": 3912 + }, + { + "epoch": 1.3065108514190318, + "grad_norm": 0.4443607666660848, + "learning_rate": 3.236619485996512e-06, + "loss": 0.2467, + "step": 3913 + }, + { + "epoch": 1.3068447412353923, + "grad_norm": 0.4472077936332228, + "learning_rate": 3.23389326816541e-06, + "loss": 0.2529, + "step": 3914 + }, + { + "epoch": 1.307178631051753, + "grad_norm": 0.45547577880758483, + "learning_rate": 3.2311676500950217e-06, + "loss": 0.2499, + "step": 3915 + }, + { + "epoch": 1.3075125208681135, + "grad_norm": 0.4523073935057141, + "learning_rate": 3.2284426327109553e-06, + "loss": 0.2613, + "step": 3916 + }, + { + "epoch": 1.307846410684474, + "grad_norm": 0.46696947418262363, + "learning_rate": 3.2257182169386102e-06, + "loss": 0.2477, + "step": 3917 + }, + { + "epoch": 1.3081803005008348, + "grad_norm": 0.4795781983958322, + "learning_rate": 3.222994403703187e-06, + "loss": 0.2631, + "step": 3918 + }, + { + "epoch": 1.3085141903171953, + "grad_norm": 0.45364174298378895, + "learning_rate": 3.220271193929675e-06, + "loss": 0.2446, + "step": 3919 + }, + { + "epoch": 1.308848080133556, + "grad_norm": 0.4469264579727855, + "learning_rate": 3.217548588542864e-06, + "loss": 0.2569, + "step": 3920 + }, + { + "epoch": 1.3091819699499165, + "grad_norm": 0.4365208649067791, + "learning_rate": 3.2148265884673365e-06, + "loss": 0.2491, + "step": 3921 + }, + { + "epoch": 1.309515859766277, + "grad_norm": 0.4532187876903356, + "learning_rate": 3.2121051946274694e-06, + "loss": 0.2485, + "step": 3922 + }, + { + "epoch": 1.3098497495826378, + "grad_norm": 0.4488341932349452, + "learning_rate": 3.2093844079474335e-06, + "loss": 0.2466, + "step": 3923 + }, + { + "epoch": 1.3101836393989983, + "grad_norm": 0.44930488325572315, + "learning_rate": 3.206664229351194e-06, + "loss": 0.2538, + "step": 3924 + }, + { + "epoch": 1.310517529215359, + "grad_norm": 0.4555457467931871, + "learning_rate": 3.2039446597625124e-06, + "loss": 0.2486, + "step": 3925 + }, + { + "epoch": 1.3108514190317195, + "grad_norm": 0.4505696713850277, + "learning_rate": 3.2012257001049364e-06, + "loss": 0.2421, + "step": 3926 + }, + { + "epoch": 1.31118530884808, + "grad_norm": 0.440498063763997, + "learning_rate": 3.198507351301813e-06, + "loss": 0.2507, + "step": 3927 + }, + { + "epoch": 1.3115191986644408, + "grad_norm": 0.4482380012574031, + "learning_rate": 3.1957896142762778e-06, + "loss": 0.2489, + "step": 3928 + }, + { + "epoch": 1.3118530884808013, + "grad_norm": 0.4698562234264339, + "learning_rate": 3.1930724899512633e-06, + "loss": 0.2623, + "step": 3929 + }, + { + "epoch": 1.312186978297162, + "grad_norm": 0.46385652545645556, + "learning_rate": 3.1903559792494864e-06, + "loss": 0.2495, + "step": 3930 + }, + { + "epoch": 1.3125208681135225, + "grad_norm": 0.44978395120006925, + "learning_rate": 3.1876400830934637e-06, + "loss": 0.2514, + "step": 3931 + }, + { + "epoch": 1.312854757929883, + "grad_norm": 0.45184287041151905, + "learning_rate": 3.184924802405497e-06, + "loss": 0.2437, + "step": 3932 + }, + { + "epoch": 1.3131886477462438, + "grad_norm": 0.4659464794154328, + "learning_rate": 3.1822101381076832e-06, + "loss": 0.2486, + "step": 3933 + }, + { + "epoch": 1.3135225375626043, + "grad_norm": 0.4591611460071791, + "learning_rate": 3.1794960911219085e-06, + "loss": 0.2469, + "step": 3934 + }, + { + "epoch": 1.313856427378965, + "grad_norm": 0.46788607187666625, + "learning_rate": 3.1767826623698483e-06, + "loss": 0.2639, + "step": 3935 + }, + { + "epoch": 1.3141903171953255, + "grad_norm": 0.4425004350832933, + "learning_rate": 3.17406985277297e-06, + "loss": 0.247, + "step": 3936 + }, + { + "epoch": 1.314524207011686, + "grad_norm": 0.44113295705027034, + "learning_rate": 3.1713576632525273e-06, + "loss": 0.2545, + "step": 3937 + }, + { + "epoch": 1.3148580968280468, + "grad_norm": 0.4567580718893521, + "learning_rate": 3.1686460947295695e-06, + "loss": 0.2604, + "step": 3938 + }, + { + "epoch": 1.3151919866444073, + "grad_norm": 0.4473081524778561, + "learning_rate": 3.165935148124927e-06, + "loss": 0.2516, + "step": 3939 + }, + { + "epoch": 1.315525876460768, + "grad_norm": 0.45521533283723575, + "learning_rate": 3.163224824359225e-06, + "loss": 0.2562, + "step": 3940 + }, + { + "epoch": 1.3158597662771285, + "grad_norm": 0.44525026780939525, + "learning_rate": 3.160515124352878e-06, + "loss": 0.2479, + "step": 3941 + }, + { + "epoch": 1.316193656093489, + "grad_norm": 0.45922648685316164, + "learning_rate": 3.157806049026083e-06, + "loss": 0.2562, + "step": 3942 + }, + { + "epoch": 1.3165275459098498, + "grad_norm": 0.4474782738340382, + "learning_rate": 3.1550975992988298e-06, + "loss": 0.248, + "step": 3943 + }, + { + "epoch": 1.3168614357262103, + "grad_norm": 0.44398666609262627, + "learning_rate": 3.1523897760908916e-06, + "loss": 0.2524, + "step": 3944 + }, + { + "epoch": 1.317195325542571, + "grad_norm": 0.4552975515411492, + "learning_rate": 3.1496825803218344e-06, + "loss": 0.2636, + "step": 3945 + }, + { + "epoch": 1.3175292153589315, + "grad_norm": 0.4482957504642507, + "learning_rate": 3.1469760129110046e-06, + "loss": 0.258, + "step": 3946 + }, + { + "epoch": 1.317863105175292, + "grad_norm": 0.4474644763382885, + "learning_rate": 3.1442700747775413e-06, + "loss": 0.2364, + "step": 3947 + }, + { + "epoch": 1.3181969949916528, + "grad_norm": 0.466789798189974, + "learning_rate": 3.1415647668403643e-06, + "loss": 0.252, + "step": 3948 + }, + { + "epoch": 1.3185308848080133, + "grad_norm": 0.44443603116404984, + "learning_rate": 3.1388600900181832e-06, + "loss": 0.2536, + "step": 3949 + }, + { + "epoch": 1.318864774624374, + "grad_norm": 0.45781230466868883, + "learning_rate": 3.136156045229494e-06, + "loss": 0.2522, + "step": 3950 + }, + { + "epoch": 1.3191986644407345, + "grad_norm": 0.4345726222529469, + "learning_rate": 3.1334526333925754e-06, + "loss": 0.2523, + "step": 3951 + }, + { + "epoch": 1.319532554257095, + "grad_norm": 0.45282247368553663, + "learning_rate": 3.130749855425491e-06, + "loss": 0.2478, + "step": 3952 + }, + { + "epoch": 1.3198664440734558, + "grad_norm": 0.4651572981520585, + "learning_rate": 3.1280477122460917e-06, + "loss": 0.249, + "step": 3953 + }, + { + "epoch": 1.3202003338898163, + "grad_norm": 0.4574796239250155, + "learning_rate": 3.1253462047720107e-06, + "loss": 0.2502, + "step": 3954 + }, + { + "epoch": 1.320534223706177, + "grad_norm": 0.46272988795185355, + "learning_rate": 3.122645333920665e-06, + "loss": 0.2506, + "step": 3955 + }, + { + "epoch": 1.3208681135225375, + "grad_norm": 0.4654342422698836, + "learning_rate": 3.1199451006092586e-06, + "loss": 0.2544, + "step": 3956 + }, + { + "epoch": 1.321202003338898, + "grad_norm": 0.4689930672397519, + "learning_rate": 3.117245505754773e-06, + "loss": 0.2525, + "step": 3957 + }, + { + "epoch": 1.3215358931552588, + "grad_norm": 0.4808339763342675, + "learning_rate": 3.11454655027398e-06, + "loss": 0.2633, + "step": 3958 + }, + { + "epoch": 1.3218697829716195, + "grad_norm": 0.47641839089003146, + "learning_rate": 3.1118482350834316e-06, + "loss": 0.2636, + "step": 3959 + }, + { + "epoch": 1.32220367278798, + "grad_norm": 0.483812216787538, + "learning_rate": 3.109150561099459e-06, + "loss": 0.2578, + "step": 3960 + }, + { + "epoch": 1.3225375626043405, + "grad_norm": 0.4603924131933953, + "learning_rate": 3.1064535292381803e-06, + "loss": 0.2488, + "step": 3961 + }, + { + "epoch": 1.322871452420701, + "grad_norm": 0.4656187743462153, + "learning_rate": 3.103757140415492e-06, + "loss": 0.2483, + "step": 3962 + }, + { + "epoch": 1.3232053422370618, + "grad_norm": 0.44001043131004997, + "learning_rate": 3.1010613955470754e-06, + "loss": 0.2538, + "step": 3963 + }, + { + "epoch": 1.3235392320534225, + "grad_norm": 0.474933012515487, + "learning_rate": 3.0983662955483895e-06, + "loss": 0.2542, + "step": 3964 + }, + { + "epoch": 1.323873121869783, + "grad_norm": 0.46927664433683314, + "learning_rate": 3.0956718413346785e-06, + "loss": 0.2539, + "step": 3965 + }, + { + "epoch": 1.3242070116861435, + "grad_norm": 0.46131994666311343, + "learning_rate": 3.0929780338209627e-06, + "loss": 0.26, + "step": 3966 + }, + { + "epoch": 1.3245409015025043, + "grad_norm": 0.44011453585662264, + "learning_rate": 3.090284873922047e-06, + "loss": 0.237, + "step": 3967 + }, + { + "epoch": 1.3248747913188648, + "grad_norm": 0.46623212565027483, + "learning_rate": 3.087592362552515e-06, + "loss": 0.2607, + "step": 3968 + }, + { + "epoch": 1.3252086811352255, + "grad_norm": 0.4451649987842149, + "learning_rate": 3.084900500626728e-06, + "loss": 0.2489, + "step": 3969 + }, + { + "epoch": 1.325542570951586, + "grad_norm": 0.46945670786556065, + "learning_rate": 3.0822092890588307e-06, + "loss": 0.259, + "step": 3970 + }, + { + "epoch": 1.3258764607679465, + "grad_norm": 0.4400644053042205, + "learning_rate": 3.0795187287627414e-06, + "loss": 0.2427, + "step": 3971 + }, + { + "epoch": 1.3262103505843073, + "grad_norm": 0.4620446039837612, + "learning_rate": 3.0768288206521636e-06, + "loss": 0.2596, + "step": 3972 + }, + { + "epoch": 1.3265442404006678, + "grad_norm": 0.44545243747562174, + "learning_rate": 3.0741395656405735e-06, + "loss": 0.2509, + "step": 3973 + }, + { + "epoch": 1.3268781302170285, + "grad_norm": 0.4470561022736912, + "learning_rate": 3.0714509646412295e-06, + "loss": 0.2607, + "step": 3974 + }, + { + "epoch": 1.327212020033389, + "grad_norm": 0.4546284072232968, + "learning_rate": 3.0687630185671674e-06, + "loss": 0.2638, + "step": 3975 + }, + { + "epoch": 1.3275459098497495, + "grad_norm": 0.4662516792870794, + "learning_rate": 3.0660757283311986e-06, + "loss": 0.2619, + "step": 3976 + }, + { + "epoch": 1.3278797996661102, + "grad_norm": 0.46988261847127755, + "learning_rate": 3.0633890948459142e-06, + "loss": 0.2477, + "step": 3977 + }, + { + "epoch": 1.3282136894824708, + "grad_norm": 0.45532646288140016, + "learning_rate": 3.060703119023679e-06, + "loss": 0.2654, + "step": 3978 + }, + { + "epoch": 1.3285475792988315, + "grad_norm": 0.4267485080415923, + "learning_rate": 3.0580178017766393e-06, + "loss": 0.2482, + "step": 3979 + }, + { + "epoch": 1.328881469115192, + "grad_norm": 0.4568809918000089, + "learning_rate": 3.055333144016712e-06, + "loss": 0.2617, + "step": 3980 + }, + { + "epoch": 1.3292153589315525, + "grad_norm": 0.4387921020337651, + "learning_rate": 3.052649146655595e-06, + "loss": 0.2499, + "step": 3981 + }, + { + "epoch": 1.3295492487479132, + "grad_norm": 0.4442904897230734, + "learning_rate": 3.049965810604759e-06, + "loss": 0.2466, + "step": 3982 + }, + { + "epoch": 1.3298831385642738, + "grad_norm": 0.4496460487250851, + "learning_rate": 3.0472831367754496e-06, + "loss": 0.2501, + "step": 3983 + }, + { + "epoch": 1.3302170283806345, + "grad_norm": 0.45373329883809976, + "learning_rate": 3.0446011260786933e-06, + "loss": 0.2586, + "step": 3984 + }, + { + "epoch": 1.330550918196995, + "grad_norm": 0.4402379087350798, + "learning_rate": 3.041919779425283e-06, + "loss": 0.245, + "step": 3985 + }, + { + "epoch": 1.3308848080133555, + "grad_norm": 0.46553260167448135, + "learning_rate": 3.0392390977257928e-06, + "loss": 0.2559, + "step": 3986 + }, + { + "epoch": 1.3312186978297162, + "grad_norm": 0.4365702525160463, + "learning_rate": 3.036559081890566e-06, + "loss": 0.2377, + "step": 3987 + }, + { + "epoch": 1.3315525876460768, + "grad_norm": 0.4502996671722102, + "learning_rate": 3.0338797328297243e-06, + "loss": 0.2529, + "step": 3988 + }, + { + "epoch": 1.3318864774624375, + "grad_norm": 0.44078774343589444, + "learning_rate": 3.0312010514531592e-06, + "loss": 0.2502, + "step": 3989 + }, + { + "epoch": 1.332220367278798, + "grad_norm": 0.4536124697787074, + "learning_rate": 3.028523038670538e-06, + "loss": 0.2539, + "step": 3990 + }, + { + "epoch": 1.3325542570951585, + "grad_norm": 0.45805216385745634, + "learning_rate": 3.0258456953912985e-06, + "loss": 0.2555, + "step": 3991 + }, + { + "epoch": 1.3328881469115192, + "grad_norm": 0.4372183406418791, + "learning_rate": 3.0231690225246537e-06, + "loss": 0.2386, + "step": 3992 + }, + { + "epoch": 1.3332220367278798, + "grad_norm": 0.44167587163032934, + "learning_rate": 3.020493020979589e-06, + "loss": 0.242, + "step": 3993 + }, + { + "epoch": 1.3335559265442405, + "grad_norm": 0.45595119012804636, + "learning_rate": 3.0178176916648595e-06, + "loss": 0.2563, + "step": 3994 + }, + { + "epoch": 1.333889816360601, + "grad_norm": 0.47536649017528365, + "learning_rate": 3.015143035488993e-06, + "loss": 0.2517, + "step": 3995 + }, + { + "epoch": 1.3342237061769615, + "grad_norm": 0.45460589275270796, + "learning_rate": 3.0124690533602895e-06, + "loss": 0.2472, + "step": 3996 + }, + { + "epoch": 1.3345575959933222, + "grad_norm": 0.4483501472500109, + "learning_rate": 3.00979574618682e-06, + "loss": 0.2463, + "step": 3997 + }, + { + "epoch": 1.3348914858096828, + "grad_norm": 0.48001747739899325, + "learning_rate": 3.0071231148764235e-06, + "loss": 0.2651, + "step": 3998 + }, + { + "epoch": 1.3352253756260435, + "grad_norm": 0.46463427764275445, + "learning_rate": 3.0044511603367143e-06, + "loss": 0.2583, + "step": 3999 + }, + { + "epoch": 1.335559265442404, + "grad_norm": 0.45600800487171067, + "learning_rate": 3.001779883475071e-06, + "loss": 0.2563, + "step": 4000 + }, + { + "epoch": 1.3358931552587645, + "grad_norm": 0.4566947582579546, + "learning_rate": 2.999109285198649e-06, + "loss": 0.2425, + "step": 4001 + }, + { + "epoch": 1.3362270450751252, + "grad_norm": 0.43881270888955093, + "learning_rate": 2.99643936641437e-06, + "loss": 0.2486, + "step": 4002 + }, + { + "epoch": 1.3365609348914858, + "grad_norm": 0.45714354839897653, + "learning_rate": 2.9937701280289216e-06, + "loss": 0.262, + "step": 4003 + }, + { + "epoch": 1.3368948247078465, + "grad_norm": 0.4702264826840724, + "learning_rate": 2.9911015709487654e-06, + "loss": 0.2642, + "step": 4004 + }, + { + "epoch": 1.337228714524207, + "grad_norm": 0.4585783366038096, + "learning_rate": 2.9884336960801284e-06, + "loss": 0.2546, + "step": 4005 + }, + { + "epoch": 1.3375626043405675, + "grad_norm": 0.4487222378648368, + "learning_rate": 2.985766504329009e-06, + "loss": 0.2483, + "step": 4006 + }, + { + "epoch": 1.3378964941569282, + "grad_norm": 0.45946478593544093, + "learning_rate": 2.9830999966011688e-06, + "loss": 0.2543, + "step": 4007 + }, + { + "epoch": 1.3382303839732888, + "grad_norm": 0.47678092849556764, + "learning_rate": 2.9804341738021407e-06, + "loss": 0.2591, + "step": 4008 + }, + { + "epoch": 1.3385642737896495, + "grad_norm": 0.4316541092810723, + "learning_rate": 2.977769036837228e-06, + "loss": 0.2456, + "step": 4009 + }, + { + "epoch": 1.33889816360601, + "grad_norm": 0.44833131388757125, + "learning_rate": 2.9751045866114926e-06, + "loss": 0.2363, + "step": 4010 + }, + { + "epoch": 1.3392320534223705, + "grad_norm": 0.4442421085906493, + "learning_rate": 2.972440824029772e-06, + "loss": 0.248, + "step": 4011 + }, + { + "epoch": 1.3395659432387312, + "grad_norm": 0.46151336607500976, + "learning_rate": 2.969777749996663e-06, + "loss": 0.2489, + "step": 4012 + }, + { + "epoch": 1.3398998330550917, + "grad_norm": 0.4621322031580347, + "learning_rate": 2.967115365416534e-06, + "loss": 0.2482, + "step": 4013 + }, + { + "epoch": 1.3402337228714525, + "grad_norm": 0.455828423753715, + "learning_rate": 2.964453671193515e-06, + "loss": 0.2602, + "step": 4014 + }, + { + "epoch": 1.340567612687813, + "grad_norm": 0.45539544907993706, + "learning_rate": 2.9617926682315062e-06, + "loss": 0.2458, + "step": 4015 + }, + { + "epoch": 1.3409015025041735, + "grad_norm": 0.44682755989914713, + "learning_rate": 2.959132357434168e-06, + "loss": 0.2505, + "step": 4016 + }, + { + "epoch": 1.3412353923205342, + "grad_norm": 0.44772146200208673, + "learning_rate": 2.9564727397049274e-06, + "loss": 0.2565, + "step": 4017 + }, + { + "epoch": 1.3415692821368947, + "grad_norm": 0.44893199828492214, + "learning_rate": 2.9538138159469814e-06, + "loss": 0.257, + "step": 4018 + }, + { + "epoch": 1.3419031719532555, + "grad_norm": 0.4643653897879117, + "learning_rate": 2.9511555870632824e-06, + "loss": 0.2562, + "step": 4019 + }, + { + "epoch": 1.342237061769616, + "grad_norm": 0.4601563334974004, + "learning_rate": 2.9484980539565546e-06, + "loss": 0.2675, + "step": 4020 + }, + { + "epoch": 1.3425709515859765, + "grad_norm": 0.4768078595048585, + "learning_rate": 2.945841217529279e-06, + "loss": 0.2497, + "step": 4021 + }, + { + "epoch": 1.3429048414023372, + "grad_norm": 0.4571675656602561, + "learning_rate": 2.9431850786837063e-06, + "loss": 0.2478, + "step": 4022 + }, + { + "epoch": 1.343238731218698, + "grad_norm": 0.44517224207994355, + "learning_rate": 2.940529638321845e-06, + "loss": 0.2554, + "step": 4023 + }, + { + "epoch": 1.3435726210350585, + "grad_norm": 0.45049491316888063, + "learning_rate": 2.937874897345472e-06, + "loss": 0.25, + "step": 4024 + }, + { + "epoch": 1.343906510851419, + "grad_norm": 0.4362403822929042, + "learning_rate": 2.935220856656119e-06, + "loss": 0.2421, + "step": 4025 + }, + { + "epoch": 1.3442404006677795, + "grad_norm": 0.4608503302085293, + "learning_rate": 2.9325675171550895e-06, + "loss": 0.268, + "step": 4026 + }, + { + "epoch": 1.3445742904841402, + "grad_norm": 0.4507079171475488, + "learning_rate": 2.9299148797434425e-06, + "loss": 0.2488, + "step": 4027 + }, + { + "epoch": 1.344908180300501, + "grad_norm": 0.44177659333732056, + "learning_rate": 2.927262945321998e-06, + "loss": 0.253, + "step": 4028 + }, + { + "epoch": 1.3452420701168615, + "grad_norm": 0.46542961793755727, + "learning_rate": 2.9246117147913412e-06, + "loss": 0.2557, + "step": 4029 + }, + { + "epoch": 1.345575959933222, + "grad_norm": 0.4445010186730351, + "learning_rate": 2.921961189051815e-06, + "loss": 0.2445, + "step": 4030 + }, + { + "epoch": 1.3459098497495827, + "grad_norm": 0.4635737568940452, + "learning_rate": 2.9193113690035256e-06, + "loss": 0.249, + "step": 4031 + }, + { + "epoch": 1.3462437395659432, + "grad_norm": 0.42716082804022376, + "learning_rate": 2.9166622555463366e-06, + "loss": 0.2432, + "step": 4032 + }, + { + "epoch": 1.346577629382304, + "grad_norm": 0.44089508113626474, + "learning_rate": 2.9140138495798742e-06, + "loss": 0.243, + "step": 4033 + }, + { + "epoch": 1.3469115191986645, + "grad_norm": 0.45392725909124915, + "learning_rate": 2.9113661520035218e-06, + "loss": 0.2663, + "step": 4034 + }, + { + "epoch": 1.347245409015025, + "grad_norm": 0.4581975602736771, + "learning_rate": 2.9087191637164268e-06, + "loss": 0.2619, + "step": 4035 + }, + { + "epoch": 1.3475792988313857, + "grad_norm": 0.4656344169413579, + "learning_rate": 2.9060728856174898e-06, + "loss": 0.2634, + "step": 4036 + }, + { + "epoch": 1.3479131886477462, + "grad_norm": 0.45404597092425725, + "learning_rate": 2.903427318605376e-06, + "loss": 0.2525, + "step": 4037 + }, + { + "epoch": 1.348247078464107, + "grad_norm": 0.44768208049585795, + "learning_rate": 2.9007824635785047e-06, + "loss": 0.243, + "step": 4038 + }, + { + "epoch": 1.3485809682804675, + "grad_norm": 0.4640181713396613, + "learning_rate": 2.898138321435055e-06, + "loss": 0.2585, + "step": 4039 + }, + { + "epoch": 1.348914858096828, + "grad_norm": 0.44785149363168786, + "learning_rate": 2.895494893072962e-06, + "loss": 0.2569, + "step": 4040 + }, + { + "epoch": 1.3492487479131887, + "grad_norm": 0.4357587318924515, + "learning_rate": 2.8928521793899246e-06, + "loss": 0.2475, + "step": 4041 + }, + { + "epoch": 1.3495826377295492, + "grad_norm": 0.4409948703768042, + "learning_rate": 2.89021018128339e-06, + "loss": 0.2562, + "step": 4042 + }, + { + "epoch": 1.34991652754591, + "grad_norm": 0.44082278492827676, + "learning_rate": 2.8875688996505713e-06, + "loss": 0.2499, + "step": 4043 + }, + { + "epoch": 1.3502504173622705, + "grad_norm": 0.44648567746383006, + "learning_rate": 2.884928335388433e-06, + "loss": 0.2447, + "step": 4044 + }, + { + "epoch": 1.350584307178631, + "grad_norm": 0.48115768546364907, + "learning_rate": 2.8822884893936933e-06, + "loss": 0.2445, + "step": 4045 + }, + { + "epoch": 1.3509181969949917, + "grad_norm": 0.43153923532354543, + "learning_rate": 2.8796493625628357e-06, + "loss": 0.2434, + "step": 4046 + }, + { + "epoch": 1.3512520868113522, + "grad_norm": 0.4666147715408986, + "learning_rate": 2.8770109557920916e-06, + "loss": 0.2655, + "step": 4047 + }, + { + "epoch": 1.351585976627713, + "grad_norm": 0.4536896675224663, + "learning_rate": 2.87437326997745e-06, + "loss": 0.2575, + "step": 4048 + }, + { + "epoch": 1.3519198664440735, + "grad_norm": 0.45883829298628553, + "learning_rate": 2.8717363060146543e-06, + "loss": 0.2576, + "step": 4049 + }, + { + "epoch": 1.352253756260434, + "grad_norm": 0.44995935744485566, + "learning_rate": 2.869100064799206e-06, + "loss": 0.2595, + "step": 4050 + }, + { + "epoch": 1.3525876460767947, + "grad_norm": 0.4680929523539071, + "learning_rate": 2.8664645472263575e-06, + "loss": 0.2433, + "step": 4051 + }, + { + "epoch": 1.3529215358931552, + "grad_norm": 0.46441301500128523, + "learning_rate": 2.8638297541911187e-06, + "loss": 0.2481, + "step": 4052 + }, + { + "epoch": 1.353255425709516, + "grad_norm": 0.46326252541194507, + "learning_rate": 2.8611956865882513e-06, + "loss": 0.2563, + "step": 4053 + }, + { + "epoch": 1.3535893155258765, + "grad_norm": 0.45792838556052795, + "learning_rate": 2.858562345312268e-06, + "loss": 0.248, + "step": 4054 + }, + { + "epoch": 1.353923205342237, + "grad_norm": 0.48078621208425104, + "learning_rate": 2.8559297312574417e-06, + "loss": 0.2508, + "step": 4055 + }, + { + "epoch": 1.3542570951585977, + "grad_norm": 0.4313036324625314, + "learning_rate": 2.853297845317794e-06, + "loss": 0.2438, + "step": 4056 + }, + { + "epoch": 1.3545909849749582, + "grad_norm": 0.4580423703118545, + "learning_rate": 2.8506666883870982e-06, + "loss": 0.2499, + "step": 4057 + }, + { + "epoch": 1.354924874791319, + "grad_norm": 0.4571685213117817, + "learning_rate": 2.8480362613588817e-06, + "loss": 0.2608, + "step": 4058 + }, + { + "epoch": 1.3552587646076795, + "grad_norm": 0.4401070935062678, + "learning_rate": 2.8454065651264212e-06, + "loss": 0.2509, + "step": 4059 + }, + { + "epoch": 1.35559265442404, + "grad_norm": 0.4677286624469556, + "learning_rate": 2.8427776005827546e-06, + "loss": 0.2509, + "step": 4060 + }, + { + "epoch": 1.3559265442404007, + "grad_norm": 0.46606883414267214, + "learning_rate": 2.8401493686206618e-06, + "loss": 0.2656, + "step": 4061 + }, + { + "epoch": 1.3562604340567612, + "grad_norm": 0.46212593152918646, + "learning_rate": 2.8375218701326757e-06, + "loss": 0.2535, + "step": 4062 + }, + { + "epoch": 1.356594323873122, + "grad_norm": 0.45450429507276985, + "learning_rate": 2.8348951060110796e-06, + "loss": 0.255, + "step": 4063 + }, + { + "epoch": 1.3569282136894825, + "grad_norm": 0.47972186909842895, + "learning_rate": 2.8322690771479135e-06, + "loss": 0.2667, + "step": 4064 + }, + { + "epoch": 1.357262103505843, + "grad_norm": 0.45414552300764044, + "learning_rate": 2.829643784434961e-06, + "loss": 0.2481, + "step": 4065 + }, + { + "epoch": 1.3575959933222037, + "grad_norm": 0.45464389785179543, + "learning_rate": 2.827019228763758e-06, + "loss": 0.2595, + "step": 4066 + }, + { + "epoch": 1.3579298831385642, + "grad_norm": 0.4575953435388372, + "learning_rate": 2.824395411025589e-06, + "loss": 0.2431, + "step": 4067 + }, + { + "epoch": 1.358263772954925, + "grad_norm": 0.4426066864961646, + "learning_rate": 2.8217723321114855e-06, + "loss": 0.2506, + "step": 4068 + }, + { + "epoch": 1.3585976627712855, + "grad_norm": 0.43846091101173257, + "learning_rate": 2.81914999291224e-06, + "loss": 0.2517, + "step": 4069 + }, + { + "epoch": 1.358931552587646, + "grad_norm": 0.44129950583805333, + "learning_rate": 2.816528394318382e-06, + "loss": 0.2528, + "step": 4070 + }, + { + "epoch": 1.3592654424040067, + "grad_norm": 0.4344061047764903, + "learning_rate": 2.813907537220191e-06, + "loss": 0.2548, + "step": 4071 + }, + { + "epoch": 1.3595993322203672, + "grad_norm": 0.44945858617057177, + "learning_rate": 2.811287422507698e-06, + "loss": 0.2493, + "step": 4072 + }, + { + "epoch": 1.359933222036728, + "grad_norm": 0.4439290673056922, + "learning_rate": 2.808668051070678e-06, + "loss": 0.2481, + "step": 4073 + }, + { + "epoch": 1.3602671118530885, + "grad_norm": 0.4729997515894317, + "learning_rate": 2.806049423798659e-06, + "loss": 0.2601, + "step": 4074 + }, + { + "epoch": 1.360601001669449, + "grad_norm": 0.46637265885567913, + "learning_rate": 2.8034315415809126e-06, + "loss": 0.2601, + "step": 4075 + }, + { + "epoch": 1.3609348914858097, + "grad_norm": 0.4468194004902941, + "learning_rate": 2.8008144053064557e-06, + "loss": 0.2449, + "step": 4076 + }, + { + "epoch": 1.3612687813021702, + "grad_norm": 0.43946664240993355, + "learning_rate": 2.798198015864058e-06, + "loss": 0.2516, + "step": 4077 + }, + { + "epoch": 1.361602671118531, + "grad_norm": 0.44228589569771043, + "learning_rate": 2.7955823741422285e-06, + "loss": 0.2509, + "step": 4078 + }, + { + "epoch": 1.3619365609348915, + "grad_norm": 0.4375904789919822, + "learning_rate": 2.7929674810292284e-06, + "loss": 0.2533, + "step": 4079 + }, + { + "epoch": 1.362270450751252, + "grad_norm": 0.44839389211746433, + "learning_rate": 2.7903533374130607e-06, + "loss": 0.2592, + "step": 4080 + }, + { + "epoch": 1.3626043405676127, + "grad_norm": 0.4426312715726812, + "learning_rate": 2.787739944181475e-06, + "loss": 0.2503, + "step": 4081 + }, + { + "epoch": 1.3629382303839732, + "grad_norm": 0.4564066741136666, + "learning_rate": 2.7851273022219645e-06, + "loss": 0.2573, + "step": 4082 + }, + { + "epoch": 1.363272120200334, + "grad_norm": 0.4587177939715424, + "learning_rate": 2.7825154124217716e-06, + "loss": 0.2643, + "step": 4083 + }, + { + "epoch": 1.3636060100166945, + "grad_norm": 0.4562558627044749, + "learning_rate": 2.77990427566788e-06, + "loss": 0.2624, + "step": 4084 + }, + { + "epoch": 1.363939899833055, + "grad_norm": 0.44987959648863696, + "learning_rate": 2.777293892847016e-06, + "loss": 0.2532, + "step": 4085 + }, + { + "epoch": 1.3642737896494157, + "grad_norm": 0.4636427647376755, + "learning_rate": 2.7746842648456552e-06, + "loss": 0.2468, + "step": 4086 + }, + { + "epoch": 1.3646076794657762, + "grad_norm": 0.4734480851190995, + "learning_rate": 2.7720753925500105e-06, + "loss": 0.2638, + "step": 4087 + }, + { + "epoch": 1.364941569282137, + "grad_norm": 0.44359238550209823, + "learning_rate": 2.769467276846045e-06, + "loss": 0.253, + "step": 4088 + }, + { + "epoch": 1.3652754590984975, + "grad_norm": 0.4428506955737794, + "learning_rate": 2.76685991861946e-06, + "loss": 0.2589, + "step": 4089 + }, + { + "epoch": 1.365609348914858, + "grad_norm": 0.438691805279262, + "learning_rate": 2.764253318755701e-06, + "loss": 0.2447, + "step": 4090 + }, + { + "epoch": 1.3659432387312187, + "grad_norm": 0.4367077606291315, + "learning_rate": 2.7616474781399527e-06, + "loss": 0.257, + "step": 4091 + }, + { + "epoch": 1.3662771285475794, + "grad_norm": 0.44195385725592873, + "learning_rate": 2.759042397657149e-06, + "loss": 0.2544, + "step": 4092 + }, + { + "epoch": 1.36661101836394, + "grad_norm": 0.4415769422274932, + "learning_rate": 2.75643807819196e-06, + "loss": 0.2602, + "step": 4093 + }, + { + "epoch": 1.3669449081803005, + "grad_norm": 0.4419170225340206, + "learning_rate": 2.7538345206288005e-06, + "loss": 0.2495, + "step": 4094 + }, + { + "epoch": 1.367278797996661, + "grad_norm": 0.4587385060539545, + "learning_rate": 2.7512317258518244e-06, + "loss": 0.2506, + "step": 4095 + }, + { + "epoch": 1.3676126878130217, + "grad_norm": 0.47474838726713015, + "learning_rate": 2.7486296947449254e-06, + "loss": 0.2658, + "step": 4096 + }, + { + "epoch": 1.3679465776293824, + "grad_norm": 0.44237663341814853, + "learning_rate": 2.746028428191744e-06, + "loss": 0.2481, + "step": 4097 + }, + { + "epoch": 1.368280467445743, + "grad_norm": 0.4757133716420115, + "learning_rate": 2.7434279270756536e-06, + "loss": 0.2626, + "step": 4098 + }, + { + "epoch": 1.3686143572621035, + "grad_norm": 0.44430440891560874, + "learning_rate": 2.740828192279773e-06, + "loss": 0.2432, + "step": 4099 + }, + { + "epoch": 1.3689482470784642, + "grad_norm": 0.4638826669207696, + "learning_rate": 2.7382292246869548e-06, + "loss": 0.2549, + "step": 4100 + }, + { + "epoch": 1.3692821368948247, + "grad_norm": 0.43241302731889697, + "learning_rate": 2.7356310251798e-06, + "loss": 0.2433, + "step": 4101 + }, + { + "epoch": 1.3696160267111854, + "grad_norm": 0.4372335504532919, + "learning_rate": 2.7330335946406394e-06, + "loss": 0.238, + "step": 4102 + }, + { + "epoch": 1.369949916527546, + "grad_norm": 0.46256188496047607, + "learning_rate": 2.7304369339515515e-06, + "loss": 0.2533, + "step": 4103 + }, + { + "epoch": 1.3702838063439065, + "grad_norm": 0.463196023343467, + "learning_rate": 2.7278410439943454e-06, + "loss": 0.2615, + "step": 4104 + }, + { + "epoch": 1.3706176961602672, + "grad_norm": 0.45266526084295644, + "learning_rate": 2.7252459256505715e-06, + "loss": 0.2443, + "step": 4105 + }, + { + "epoch": 1.3709515859766277, + "grad_norm": 0.467394381458765, + "learning_rate": 2.7226515798015214e-06, + "loss": 0.2453, + "step": 4106 + }, + { + "epoch": 1.3712854757929884, + "grad_norm": 0.46489230175061697, + "learning_rate": 2.72005800732822e-06, + "loss": 0.2552, + "step": 4107 + }, + { + "epoch": 1.371619365609349, + "grad_norm": 0.4404299443275045, + "learning_rate": 2.7174652091114316e-06, + "loss": 0.2486, + "step": 4108 + }, + { + "epoch": 1.3719532554257095, + "grad_norm": 0.4723590976282228, + "learning_rate": 2.7148731860316544e-06, + "loss": 0.2535, + "step": 4109 + }, + { + "epoch": 1.3722871452420702, + "grad_norm": 0.4620842554452276, + "learning_rate": 2.712281938969129e-06, + "loss": 0.2543, + "step": 4110 + }, + { + "epoch": 1.3726210350584307, + "grad_norm": 0.466407862947831, + "learning_rate": 2.70969146880383e-06, + "loss": 0.2486, + "step": 4111 + }, + { + "epoch": 1.3729549248747914, + "grad_norm": 0.4580617734516804, + "learning_rate": 2.707101776415467e-06, + "loss": 0.2584, + "step": 4112 + }, + { + "epoch": 1.373288814691152, + "grad_norm": 0.4692037897841373, + "learning_rate": 2.704512862683485e-06, + "loss": 0.2617, + "step": 4113 + }, + { + "epoch": 1.3736227045075124, + "grad_norm": 0.45117739210954794, + "learning_rate": 2.701924728487065e-06, + "loss": 0.2538, + "step": 4114 + }, + { + "epoch": 1.3739565943238732, + "grad_norm": 0.42118588905156457, + "learning_rate": 2.6993373747051265e-06, + "loss": 0.2426, + "step": 4115 + }, + { + "epoch": 1.3742904841402337, + "grad_norm": 0.45388967291167814, + "learning_rate": 2.696750802216321e-06, + "loss": 0.2443, + "step": 4116 + }, + { + "epoch": 1.3746243739565944, + "grad_norm": 0.4792415053285697, + "learning_rate": 2.6941650118990338e-06, + "loss": 0.2562, + "step": 4117 + }, + { + "epoch": 1.374958263772955, + "grad_norm": 0.453985913015437, + "learning_rate": 2.6915800046313852e-06, + "loss": 0.2544, + "step": 4118 + }, + { + "epoch": 1.3752921535893154, + "grad_norm": 0.4549061481859068, + "learning_rate": 2.688995781291231e-06, + "loss": 0.2529, + "step": 4119 + }, + { + "epoch": 1.3756260434056762, + "grad_norm": 0.447885071132297, + "learning_rate": 2.6864123427561627e-06, + "loss": 0.252, + "step": 4120 + }, + { + "epoch": 1.3759599332220367, + "grad_norm": 0.4657576591287802, + "learning_rate": 2.683829689903501e-06, + "loss": 0.2596, + "step": 4121 + }, + { + "epoch": 1.3762938230383974, + "grad_norm": 0.4658608662073216, + "learning_rate": 2.6812478236103014e-06, + "loss": 0.2621, + "step": 4122 + }, + { + "epoch": 1.376627712854758, + "grad_norm": 0.45363199700171963, + "learning_rate": 2.6786667447533498e-06, + "loss": 0.2489, + "step": 4123 + }, + { + "epoch": 1.3769616026711184, + "grad_norm": 0.4521524520506427, + "learning_rate": 2.6760864542091717e-06, + "loss": 0.251, + "step": 4124 + }, + { + "epoch": 1.3772954924874792, + "grad_norm": 0.46034732028966324, + "learning_rate": 2.673506952854018e-06, + "loss": 0.251, + "step": 4125 + }, + { + "epoch": 1.3776293823038397, + "grad_norm": 0.4673142838894917, + "learning_rate": 2.670928241563875e-06, + "loss": 0.251, + "step": 4126 + }, + { + "epoch": 1.3779632721202004, + "grad_norm": 0.46508483358131886, + "learning_rate": 2.6683503212144563e-06, + "loss": 0.261, + "step": 4127 + }, + { + "epoch": 1.378297161936561, + "grad_norm": 0.4647576926587593, + "learning_rate": 2.6657731926812132e-06, + "loss": 0.2634, + "step": 4128 + }, + { + "epoch": 1.3786310517529214, + "grad_norm": 0.4636298927829521, + "learning_rate": 2.6631968568393273e-06, + "loss": 0.241, + "step": 4129 + }, + { + "epoch": 1.3789649415692822, + "grad_norm": 0.44226574144603786, + "learning_rate": 2.660621314563706e-06, + "loss": 0.2401, + "step": 4130 + }, + { + "epoch": 1.3792988313856427, + "grad_norm": 0.4601762151634015, + "learning_rate": 2.6580465667289912e-06, + "loss": 0.2599, + "step": 4131 + }, + { + "epoch": 1.3796327212020034, + "grad_norm": 0.44661248123329106, + "learning_rate": 2.6554726142095524e-06, + "loss": 0.2454, + "step": 4132 + }, + { + "epoch": 1.379966611018364, + "grad_norm": 0.45454050934189993, + "learning_rate": 2.652899457879492e-06, + "loss": 0.2436, + "step": 4133 + }, + { + "epoch": 1.3803005008347244, + "grad_norm": 0.44852140399280593, + "learning_rate": 2.6503270986126417e-06, + "loss": 0.2498, + "step": 4134 + }, + { + "epoch": 1.3806343906510852, + "grad_norm": 0.4449519832676643, + "learning_rate": 2.6477555372825573e-06, + "loss": 0.2458, + "step": 4135 + }, + { + "epoch": 1.3809682804674457, + "grad_norm": 0.44871068397020514, + "learning_rate": 2.645184774762533e-06, + "loss": 0.2514, + "step": 4136 + }, + { + "epoch": 1.3813021702838064, + "grad_norm": 0.4386599684409481, + "learning_rate": 2.6426148119255817e-06, + "loss": 0.2531, + "step": 4137 + }, + { + "epoch": 1.381636060100167, + "grad_norm": 0.4701741876514327, + "learning_rate": 2.6400456496444536e-06, + "loss": 0.2693, + "step": 4138 + }, + { + "epoch": 1.3819699499165274, + "grad_norm": 0.46568194539061464, + "learning_rate": 2.6374772887916213e-06, + "loss": 0.2445, + "step": 4139 + }, + { + "epoch": 1.3823038397328882, + "grad_norm": 0.4540243609335272, + "learning_rate": 2.6349097302392867e-06, + "loss": 0.249, + "step": 4140 + }, + { + "epoch": 1.3826377295492487, + "grad_norm": 0.4625491186684048, + "learning_rate": 2.632342974859377e-06, + "loss": 0.2605, + "step": 4141 + }, + { + "epoch": 1.3829716193656094, + "grad_norm": 0.46339675952686327, + "learning_rate": 2.629777023523553e-06, + "loss": 0.2554, + "step": 4142 + }, + { + "epoch": 1.38330550918197, + "grad_norm": 0.45941774467408164, + "learning_rate": 2.627211877103198e-06, + "loss": 0.2434, + "step": 4143 + }, + { + "epoch": 1.3836393989983304, + "grad_norm": 0.4537449144472196, + "learning_rate": 2.6246475364694183e-06, + "loss": 0.2465, + "step": 4144 + }, + { + "epoch": 1.3839732888146912, + "grad_norm": 0.4461926148178033, + "learning_rate": 2.6220840024930562e-06, + "loss": 0.2465, + "step": 4145 + }, + { + "epoch": 1.3843071786310517, + "grad_norm": 0.4611099212391055, + "learning_rate": 2.61952127604467e-06, + "loss": 0.252, + "step": 4146 + }, + { + "epoch": 1.3846410684474124, + "grad_norm": 0.46583084659975543, + "learning_rate": 2.616959357994553e-06, + "loss": 0.261, + "step": 4147 + }, + { + "epoch": 1.384974958263773, + "grad_norm": 0.4498233808481512, + "learning_rate": 2.614398249212716e-06, + "loss": 0.2531, + "step": 4148 + }, + { + "epoch": 1.3853088480801334, + "grad_norm": 0.4447890757587739, + "learning_rate": 2.6118379505689003e-06, + "loss": 0.2469, + "step": 4149 + }, + { + "epoch": 1.3856427378964942, + "grad_norm": 0.4497799576036312, + "learning_rate": 2.609278462932566e-06, + "loss": 0.2475, + "step": 4150 + }, + { + "epoch": 1.3859766277128547, + "grad_norm": 0.45713263629828493, + "learning_rate": 2.606719787172907e-06, + "loss": 0.2521, + "step": 4151 + }, + { + "epoch": 1.3863105175292154, + "grad_norm": 0.464228376309319, + "learning_rate": 2.604161924158832e-06, + "loss": 0.2486, + "step": 4152 + }, + { + "epoch": 1.386644407345576, + "grad_norm": 0.4561432030617271, + "learning_rate": 2.601604874758982e-06, + "loss": 0.2641, + "step": 4153 + }, + { + "epoch": 1.3869782971619364, + "grad_norm": 0.45733667517444004, + "learning_rate": 2.599048639841717e-06, + "loss": 0.2531, + "step": 4154 + }, + { + "epoch": 1.3873121869782972, + "grad_norm": 0.4721813302441456, + "learning_rate": 2.596493220275117e-06, + "loss": 0.2544, + "step": 4155 + }, + { + "epoch": 1.387646076794658, + "grad_norm": 0.4427954568605357, + "learning_rate": 2.5939386169269947e-06, + "loss": 0.2435, + "step": 4156 + }, + { + "epoch": 1.3879799666110184, + "grad_norm": 0.44663282689639333, + "learning_rate": 2.591384830664878e-06, + "loss": 0.2476, + "step": 4157 + }, + { + "epoch": 1.388313856427379, + "grad_norm": 0.45373994816040164, + "learning_rate": 2.5888318623560192e-06, + "loss": 0.2505, + "step": 4158 + }, + { + "epoch": 1.3886477462437394, + "grad_norm": 0.4456754989712732, + "learning_rate": 2.586279712867391e-06, + "loss": 0.2442, + "step": 4159 + }, + { + "epoch": 1.3889816360601002, + "grad_norm": 0.4387259105174294, + "learning_rate": 2.5837283830656955e-06, + "loss": 0.2605, + "step": 4160 + }, + { + "epoch": 1.389315525876461, + "grad_norm": 0.4538053675202521, + "learning_rate": 2.5811778738173453e-06, + "loss": 0.2507, + "step": 4161 + }, + { + "epoch": 1.3896494156928214, + "grad_norm": 0.45397938850570013, + "learning_rate": 2.5786281859884856e-06, + "loss": 0.2542, + "step": 4162 + }, + { + "epoch": 1.389983305509182, + "grad_norm": 0.44317067163068946, + "learning_rate": 2.5760793204449737e-06, + "loss": 0.2522, + "step": 4163 + }, + { + "epoch": 1.3903171953255427, + "grad_norm": 0.45016413262853516, + "learning_rate": 2.57353127805239e-06, + "loss": 0.2451, + "step": 4164 + }, + { + "epoch": 1.3906510851419032, + "grad_norm": 0.4488511096994649, + "learning_rate": 2.570984059676041e-06, + "loss": 0.2504, + "step": 4165 + }, + { + "epoch": 1.390984974958264, + "grad_norm": 0.4660581793259283, + "learning_rate": 2.5684376661809465e-06, + "loss": 0.254, + "step": 4166 + }, + { + "epoch": 1.3913188647746244, + "grad_norm": 0.4957563922408119, + "learning_rate": 2.565892098431848e-06, + "loss": 0.2489, + "step": 4167 + }, + { + "epoch": 1.391652754590985, + "grad_norm": 0.4777383760232454, + "learning_rate": 2.5633473572932066e-06, + "loss": 0.2552, + "step": 4168 + }, + { + "epoch": 1.3919866444073457, + "grad_norm": 0.460649701185088, + "learning_rate": 2.5608034436292036e-06, + "loss": 0.2476, + "step": 4169 + }, + { + "epoch": 1.3923205342237062, + "grad_norm": 0.45080066930112345, + "learning_rate": 2.5582603583037415e-06, + "loss": 0.2469, + "step": 4170 + }, + { + "epoch": 1.392654424040067, + "grad_norm": 0.46496410303531616, + "learning_rate": 2.555718102180437e-06, + "loss": 0.2516, + "step": 4171 + }, + { + "epoch": 1.3929883138564274, + "grad_norm": 0.4976214708487347, + "learning_rate": 2.553176676122627e-06, + "loss": 0.2663, + "step": 4172 + }, + { + "epoch": 1.393322203672788, + "grad_norm": 0.4816134682960262, + "learning_rate": 2.5506360809933657e-06, + "loss": 0.2589, + "step": 4173 + }, + { + "epoch": 1.3936560934891487, + "grad_norm": 0.45027493601002166, + "learning_rate": 2.5480963176554285e-06, + "loss": 0.2462, + "step": 4174 + }, + { + "epoch": 1.3939899833055092, + "grad_norm": 0.4485009788090508, + "learning_rate": 2.5455573869713046e-06, + "loss": 0.2395, + "step": 4175 + }, + { + "epoch": 1.39432387312187, + "grad_norm": 0.4725630510814808, + "learning_rate": 2.543019289803201e-06, + "loss": 0.2566, + "step": 4176 + }, + { + "epoch": 1.3946577629382304, + "grad_norm": 0.4650079771247164, + "learning_rate": 2.5404820270130424e-06, + "loss": 0.2628, + "step": 4177 + }, + { + "epoch": 1.394991652754591, + "grad_norm": 0.4454096597422792, + "learning_rate": 2.537945599462471e-06, + "loss": 0.2596, + "step": 4178 + }, + { + "epoch": 1.3953255425709516, + "grad_norm": 0.4567106888002051, + "learning_rate": 2.535410008012845e-06, + "loss": 0.253, + "step": 4179 + }, + { + "epoch": 1.3956594323873122, + "grad_norm": 0.48692081141017945, + "learning_rate": 2.532875253525238e-06, + "loss": 0.2552, + "step": 4180 + }, + { + "epoch": 1.395993322203673, + "grad_norm": 0.4578651213974916, + "learning_rate": 2.530341336860439e-06, + "loss": 0.2443, + "step": 4181 + }, + { + "epoch": 1.3963272120200334, + "grad_norm": 0.45955274827840403, + "learning_rate": 2.527808258878951e-06, + "loss": 0.2564, + "step": 4182 + }, + { + "epoch": 1.396661101836394, + "grad_norm": 0.4426189980197254, + "learning_rate": 2.525276020440998e-06, + "loss": 0.2584, + "step": 4183 + }, + { + "epoch": 1.3969949916527546, + "grad_norm": 0.45469260219760493, + "learning_rate": 2.5227446224065134e-06, + "loss": 0.2612, + "step": 4184 + }, + { + "epoch": 1.3973288814691152, + "grad_norm": 0.4384805647665657, + "learning_rate": 2.520214065635147e-06, + "loss": 0.2433, + "step": 4185 + }, + { + "epoch": 1.397662771285476, + "grad_norm": 0.4493013133954873, + "learning_rate": 2.5176843509862604e-06, + "loss": 0.2395, + "step": 4186 + }, + { + "epoch": 1.3979966611018364, + "grad_norm": 0.4424665179659106, + "learning_rate": 2.5151554793189325e-06, + "loss": 0.2539, + "step": 4187 + }, + { + "epoch": 1.398330550918197, + "grad_norm": 0.4726737743154938, + "learning_rate": 2.5126274514919597e-06, + "loss": 0.2542, + "step": 4188 + }, + { + "epoch": 1.3986644407345576, + "grad_norm": 0.44703958509016034, + "learning_rate": 2.510100268363843e-06, + "loss": 0.2501, + "step": 4189 + }, + { + "epoch": 1.3989983305509182, + "grad_norm": 0.445945880880112, + "learning_rate": 2.5075739307928017e-06, + "loss": 0.2504, + "step": 4190 + }, + { + "epoch": 1.399332220367279, + "grad_norm": 0.4329703301896131, + "learning_rate": 2.505048439636765e-06, + "loss": 0.239, + "step": 4191 + }, + { + "epoch": 1.3996661101836394, + "grad_norm": 0.43967417539624176, + "learning_rate": 2.5025237957533797e-06, + "loss": 0.2465, + "step": 4192 + }, + { + "epoch": 1.4, + "grad_norm": 0.453008460153347, + "learning_rate": 2.5000000000000015e-06, + "loss": 0.2518, + "step": 4193 + }, + { + "epoch": 1.4003338898163606, + "grad_norm": 0.47020588937598834, + "learning_rate": 2.4974770532336967e-06, + "loss": 0.2591, + "step": 4194 + }, + { + "epoch": 1.4006677796327212, + "grad_norm": 0.45802802922086533, + "learning_rate": 2.4949549563112434e-06, + "loss": 0.2541, + "step": 4195 + }, + { + "epoch": 1.4010016694490819, + "grad_norm": 0.433118490780594, + "learning_rate": 2.4924337100891355e-06, + "loss": 0.2428, + "step": 4196 + }, + { + "epoch": 1.4013355592654424, + "grad_norm": 0.46299213942065076, + "learning_rate": 2.489913315423576e-06, + "loss": 0.2502, + "step": 4197 + }, + { + "epoch": 1.401669449081803, + "grad_norm": 0.4662752892295688, + "learning_rate": 2.487393773170477e-06, + "loss": 0.2489, + "step": 4198 + }, + { + "epoch": 1.4020033388981636, + "grad_norm": 0.4612319053576625, + "learning_rate": 2.484875084185462e-06, + "loss": 0.2654, + "step": 4199 + }, + { + "epoch": 1.4023372287145242, + "grad_norm": 0.44464245662463553, + "learning_rate": 2.4823572493238613e-06, + "loss": 0.2478, + "step": 4200 + }, + { + "epoch": 1.4026711185308849, + "grad_norm": 0.4613118634793046, + "learning_rate": 2.479840269440724e-06, + "loss": 0.2577, + "step": 4201 + }, + { + "epoch": 1.4030050083472454, + "grad_norm": 0.4691031763179424, + "learning_rate": 2.477324145390801e-06, + "loss": 0.2434, + "step": 4202 + }, + { + "epoch": 1.403338898163606, + "grad_norm": 0.4579615826332775, + "learning_rate": 2.4748088780285533e-06, + "loss": 0.2658, + "step": 4203 + }, + { + "epoch": 1.4036727879799666, + "grad_norm": 0.45119290409047824, + "learning_rate": 2.472294468208156e-06, + "loss": 0.244, + "step": 4204 + }, + { + "epoch": 1.4040066777963272, + "grad_norm": 0.46069735591678224, + "learning_rate": 2.4697809167834863e-06, + "loss": 0.2576, + "step": 4205 + }, + { + "epoch": 1.4043405676126879, + "grad_norm": 0.4494041525066538, + "learning_rate": 2.4672682246081375e-06, + "loss": 0.2506, + "step": 4206 + }, + { + "epoch": 1.4046744574290484, + "grad_norm": 0.4407173476809594, + "learning_rate": 2.464756392535404e-06, + "loss": 0.2418, + "step": 4207 + }, + { + "epoch": 1.405008347245409, + "grad_norm": 0.4520157850918208, + "learning_rate": 2.462245421418292e-06, + "loss": 0.245, + "step": 4208 + }, + { + "epoch": 1.4053422370617696, + "grad_norm": 0.44732110120509483, + "learning_rate": 2.459735312109511e-06, + "loss": 0.2499, + "step": 4209 + }, + { + "epoch": 1.4056761268781301, + "grad_norm": 0.424482599825315, + "learning_rate": 2.457226065461486e-06, + "loss": 0.2351, + "step": 4210 + }, + { + "epoch": 1.4060100166944909, + "grad_norm": 0.4516244386359752, + "learning_rate": 2.4547176823263423e-06, + "loss": 0.2524, + "step": 4211 + }, + { + "epoch": 1.4063439065108514, + "grad_norm": 0.46276595333961223, + "learning_rate": 2.452210163555912e-06, + "loss": 0.2574, + "step": 4212 + }, + { + "epoch": 1.406677796327212, + "grad_norm": 0.45125380912298474, + "learning_rate": 2.4497035100017384e-06, + "loss": 0.253, + "step": 4213 + }, + { + "epoch": 1.4070116861435726, + "grad_norm": 0.4466147365209044, + "learning_rate": 2.4471977225150655e-06, + "loss": 0.2594, + "step": 4214 + }, + { + "epoch": 1.4073455759599331, + "grad_norm": 0.4448904291698702, + "learning_rate": 2.444692801946849e-06, + "loss": 0.256, + "step": 4215 + }, + { + "epoch": 1.4076794657762939, + "grad_norm": 0.4643891818710284, + "learning_rate": 2.442188749147745e-06, + "loss": 0.255, + "step": 4216 + }, + { + "epoch": 1.4080133555926544, + "grad_norm": 0.4357669107064095, + "learning_rate": 2.4396855649681166e-06, + "loss": 0.237, + "step": 4217 + }, + { + "epoch": 1.408347245409015, + "grad_norm": 0.4635764331900659, + "learning_rate": 2.437183250258031e-06, + "loss": 0.2527, + "step": 4218 + }, + { + "epoch": 1.4086811352253756, + "grad_norm": 0.45936843954391227, + "learning_rate": 2.4346818058672645e-06, + "loss": 0.245, + "step": 4219 + }, + { + "epoch": 1.4090150250417361, + "grad_norm": 0.4320171816586275, + "learning_rate": 2.4321812326452916e-06, + "loss": 0.2412, + "step": 4220 + }, + { + "epoch": 1.4093489148580969, + "grad_norm": 0.45967707195372576, + "learning_rate": 2.4296815314412976e-06, + "loss": 0.2597, + "step": 4221 + }, + { + "epoch": 1.4096828046744574, + "grad_norm": 0.44574346661577774, + "learning_rate": 2.4271827031041654e-06, + "loss": 0.2643, + "step": 4222 + }, + { + "epoch": 1.410016694490818, + "grad_norm": 0.4495105275356669, + "learning_rate": 2.424684748482483e-06, + "loss": 0.2523, + "step": 4223 + }, + { + "epoch": 1.4103505843071786, + "grad_norm": 0.4618000517164061, + "learning_rate": 2.422187668424547e-06, + "loss": 0.2564, + "step": 4224 + }, + { + "epoch": 1.4106844741235394, + "grad_norm": 0.48245321809436637, + "learning_rate": 2.4196914637783504e-06, + "loss": 0.2609, + "step": 4225 + }, + { + "epoch": 1.4110183639398999, + "grad_norm": 0.4522822113614776, + "learning_rate": 2.4171961353915914e-06, + "loss": 0.2441, + "step": 4226 + }, + { + "epoch": 1.4113522537562604, + "grad_norm": 0.4629664526817858, + "learning_rate": 2.4147016841116687e-06, + "loss": 0.2584, + "step": 4227 + }, + { + "epoch": 1.411686143572621, + "grad_norm": 0.47204332900797036, + "learning_rate": 2.412208110785689e-06, + "loss": 0.2684, + "step": 4228 + }, + { + "epoch": 1.4120200333889816, + "grad_norm": 0.4515387473125136, + "learning_rate": 2.4097154162604537e-06, + "loss": 0.2444, + "step": 4229 + }, + { + "epoch": 1.4123539232053424, + "grad_norm": 0.4549640810345601, + "learning_rate": 2.4072236013824718e-06, + "loss": 0.2459, + "step": 4230 + }, + { + "epoch": 1.4126878130217029, + "grad_norm": 0.45125668425668924, + "learning_rate": 2.40473266699795e-06, + "loss": 0.2493, + "step": 4231 + }, + { + "epoch": 1.4130217028380634, + "grad_norm": 0.43237344554162926, + "learning_rate": 2.4022426139527934e-06, + "loss": 0.2488, + "step": 4232 + }, + { + "epoch": 1.4133555926544241, + "grad_norm": 0.45063309442002636, + "learning_rate": 2.3997534430926155e-06, + "loss": 0.2471, + "step": 4233 + }, + { + "epoch": 1.4136894824707846, + "grad_norm": 0.4376209303231064, + "learning_rate": 2.397265155262724e-06, + "loss": 0.2531, + "step": 4234 + }, + { + "epoch": 1.4140233722871454, + "grad_norm": 0.4566657561198127, + "learning_rate": 2.394777751308129e-06, + "loss": 0.248, + "step": 4235 + }, + { + "epoch": 1.4143572621035059, + "grad_norm": 0.4651751987023715, + "learning_rate": 2.392291232073538e-06, + "loss": 0.2602, + "step": 4236 + }, + { + "epoch": 1.4146911519198664, + "grad_norm": 0.4519724798379954, + "learning_rate": 2.3898055984033615e-06, + "loss": 0.2475, + "step": 4237 + }, + { + "epoch": 1.4150250417362271, + "grad_norm": 0.453245294295696, + "learning_rate": 2.3873208511417102e-06, + "loss": 0.256, + "step": 4238 + }, + { + "epoch": 1.4153589315525876, + "grad_norm": 0.4544983217501033, + "learning_rate": 2.384836991132389e-06, + "loss": 0.259, + "step": 4239 + }, + { + "epoch": 1.4156928213689484, + "grad_norm": 0.4478100727281565, + "learning_rate": 2.3823540192189047e-06, + "loss": 0.2582, + "step": 4240 + }, + { + "epoch": 1.4160267111853089, + "grad_norm": 0.4664979011177337, + "learning_rate": 2.379871936244459e-06, + "loss": 0.2556, + "step": 4241 + }, + { + "epoch": 1.4163606010016694, + "grad_norm": 0.45930618953282637, + "learning_rate": 2.377390743051959e-06, + "loss": 0.2484, + "step": 4242 + }, + { + "epoch": 1.4166944908180301, + "grad_norm": 0.4391194569215555, + "learning_rate": 2.3749104404840023e-06, + "loss": 0.2529, + "step": 4243 + }, + { + "epoch": 1.4170283806343906, + "grad_norm": 0.44914740215229465, + "learning_rate": 2.372431029382888e-06, + "loss": 0.2468, + "step": 4244 + }, + { + "epoch": 1.4173622704507514, + "grad_norm": 0.45526959760854024, + "learning_rate": 2.3699525105906095e-06, + "loss": 0.2627, + "step": 4245 + }, + { + "epoch": 1.4176961602671119, + "grad_norm": 0.45442590665617644, + "learning_rate": 2.367474884948857e-06, + "loss": 0.2484, + "step": 4246 + }, + { + "epoch": 1.4180300500834724, + "grad_norm": 0.46726872206178016, + "learning_rate": 2.364998153299027e-06, + "loss": 0.2638, + "step": 4247 + }, + { + "epoch": 1.4183639398998331, + "grad_norm": 0.4532547428096288, + "learning_rate": 2.3625223164821992e-06, + "loss": 0.2505, + "step": 4248 + }, + { + "epoch": 1.4186978297161936, + "grad_norm": 0.45502528313998, + "learning_rate": 2.360047375339156e-06, + "loss": 0.2579, + "step": 4249 + }, + { + "epoch": 1.4190317195325544, + "grad_norm": 0.4525101063676052, + "learning_rate": 2.357573330710373e-06, + "loss": 0.2543, + "step": 4250 + }, + { + "epoch": 1.4193656093489149, + "grad_norm": 0.43904288635926736, + "learning_rate": 2.3551001834360263e-06, + "loss": 0.2346, + "step": 4251 + }, + { + "epoch": 1.4196994991652754, + "grad_norm": 0.46582431701067856, + "learning_rate": 2.3526279343559826e-06, + "loss": 0.2561, + "step": 4252 + }, + { + "epoch": 1.4200333889816361, + "grad_norm": 0.4617147155250402, + "learning_rate": 2.350156584309804e-06, + "loss": 0.2541, + "step": 4253 + }, + { + "epoch": 1.4203672787979966, + "grad_norm": 0.44181187064727984, + "learning_rate": 2.3476861341367474e-06, + "loss": 0.2517, + "step": 4254 + }, + { + "epoch": 1.4207011686143574, + "grad_norm": 0.4734050466472813, + "learning_rate": 2.345216584675766e-06, + "loss": 0.2583, + "step": 4255 + }, + { + "epoch": 1.4210350584307179, + "grad_norm": 0.4515575887199224, + "learning_rate": 2.3427479367655076e-06, + "loss": 0.2387, + "step": 4256 + }, + { + "epoch": 1.4213689482470784, + "grad_norm": 0.471481421468873, + "learning_rate": 2.340280191244311e-06, + "loss": 0.2637, + "step": 4257 + }, + { + "epoch": 1.4217028380634391, + "grad_norm": 0.472597058911026, + "learning_rate": 2.3378133489502105e-06, + "loss": 0.2533, + "step": 4258 + }, + { + "epoch": 1.4220367278797996, + "grad_norm": 0.4639821285034645, + "learning_rate": 2.335347410720931e-06, + "loss": 0.2597, + "step": 4259 + }, + { + "epoch": 1.4223706176961604, + "grad_norm": 0.45968082316502396, + "learning_rate": 2.332882377393893e-06, + "loss": 0.2543, + "step": 4260 + }, + { + "epoch": 1.4227045075125209, + "grad_norm": 0.4690550276748994, + "learning_rate": 2.3304182498062104e-06, + "loss": 0.2571, + "step": 4261 + }, + { + "epoch": 1.4230383973288814, + "grad_norm": 0.4389223962867522, + "learning_rate": 2.3279550287946883e-06, + "loss": 0.2396, + "step": 4262 + }, + { + "epoch": 1.423372287145242, + "grad_norm": 0.46533079770268476, + "learning_rate": 2.325492715195821e-06, + "loss": 0.2604, + "step": 4263 + }, + { + "epoch": 1.4237061769616026, + "grad_norm": 0.4712132884845062, + "learning_rate": 2.3230313098457992e-06, + "loss": 0.2527, + "step": 4264 + }, + { + "epoch": 1.4240400667779634, + "grad_norm": 0.44122968467376633, + "learning_rate": 2.320570813580505e-06, + "loss": 0.2485, + "step": 4265 + }, + { + "epoch": 1.4243739565943239, + "grad_norm": 0.4609433929728791, + "learning_rate": 2.3181112272355095e-06, + "loss": 0.2476, + "step": 4266 + }, + { + "epoch": 1.4247078464106844, + "grad_norm": 0.45021358079917273, + "learning_rate": 2.315652551646075e-06, + "loss": 0.2457, + "step": 4267 + }, + { + "epoch": 1.425041736227045, + "grad_norm": 0.4512159671619874, + "learning_rate": 2.3131947876471543e-06, + "loss": 0.2506, + "step": 4268 + }, + { + "epoch": 1.4253756260434056, + "grad_norm": 0.45323358344667025, + "learning_rate": 2.3107379360733905e-06, + "loss": 0.2549, + "step": 4269 + }, + { + "epoch": 1.4257095158597664, + "grad_norm": 0.4491480722797947, + "learning_rate": 2.308281997759121e-06, + "loss": 0.2539, + "step": 4270 + }, + { + "epoch": 1.4260434056761269, + "grad_norm": 0.4595271552279523, + "learning_rate": 2.305826973538366e-06, + "loss": 0.2528, + "step": 4271 + }, + { + "epoch": 1.4263772954924874, + "grad_norm": 0.4514706131662384, + "learning_rate": 2.3033728642448426e-06, + "loss": 0.2543, + "step": 4272 + }, + { + "epoch": 1.426711185308848, + "grad_norm": 0.44710936441810656, + "learning_rate": 2.3009196707119526e-06, + "loss": 0.2358, + "step": 4273 + }, + { + "epoch": 1.4270450751252086, + "grad_norm": 0.4451270297529489, + "learning_rate": 2.298467393772784e-06, + "loss": 0.2466, + "step": 4274 + }, + { + "epoch": 1.4273789649415694, + "grad_norm": 0.4805022082397593, + "learning_rate": 2.2960160342601234e-06, + "loss": 0.2538, + "step": 4275 + }, + { + "epoch": 1.4277128547579299, + "grad_norm": 0.4407386944643034, + "learning_rate": 2.293565593006437e-06, + "loss": 0.2503, + "step": 4276 + }, + { + "epoch": 1.4280467445742904, + "grad_norm": 0.48758531617316325, + "learning_rate": 2.2911160708438816e-06, + "loss": 0.2705, + "step": 4277 + }, + { + "epoch": 1.428380634390651, + "grad_norm": 0.4700080235822836, + "learning_rate": 2.2886674686043008e-06, + "loss": 0.2506, + "step": 4278 + }, + { + "epoch": 1.4287145242070116, + "grad_norm": 0.4750518934358717, + "learning_rate": 2.2862197871192314e-06, + "loss": 0.2528, + "step": 4279 + }, + { + "epoch": 1.4290484140233723, + "grad_norm": 0.4496971016834056, + "learning_rate": 2.2837730272198886e-06, + "loss": 0.2525, + "step": 4280 + }, + { + "epoch": 1.4293823038397329, + "grad_norm": 0.47016367080359334, + "learning_rate": 2.281327189737184e-06, + "loss": 0.2428, + "step": 4281 + }, + { + "epoch": 1.4297161936560934, + "grad_norm": 0.46134383406418233, + "learning_rate": 2.278882275501709e-06, + "loss": 0.2499, + "step": 4282 + }, + { + "epoch": 1.430050083472454, + "grad_norm": 0.43241194493044144, + "learning_rate": 2.2764382853437427e-06, + "loss": 0.2311, + "step": 4283 + }, + { + "epoch": 1.4303839732888146, + "grad_norm": 0.4511435518941323, + "learning_rate": 2.2739952200932543e-06, + "loss": 0.2473, + "step": 4284 + }, + { + "epoch": 1.4307178631051753, + "grad_norm": 0.44528933968405165, + "learning_rate": 2.2715530805798945e-06, + "loss": 0.2479, + "step": 4285 + }, + { + "epoch": 1.4310517529215359, + "grad_norm": 0.45218738338169906, + "learning_rate": 2.269111867633002e-06, + "loss": 0.2439, + "step": 4286 + }, + { + "epoch": 1.4313856427378964, + "grad_norm": 0.4385742815460764, + "learning_rate": 2.2666715820815975e-06, + "loss": 0.2515, + "step": 4287 + }, + { + "epoch": 1.431719532554257, + "grad_norm": 0.4713309077381171, + "learning_rate": 2.264232224754391e-06, + "loss": 0.2582, + "step": 4288 + }, + { + "epoch": 1.4320534223706178, + "grad_norm": 0.46321696340849516, + "learning_rate": 2.2617937964797786e-06, + "loss": 0.2581, + "step": 4289 + }, + { + "epoch": 1.4323873121869783, + "grad_norm": 0.46421693021152505, + "learning_rate": 2.2593562980858353e-06, + "loss": 0.2522, + "step": 4290 + }, + { + "epoch": 1.4327212020033389, + "grad_norm": 0.4443626425490951, + "learning_rate": 2.2569197304003227e-06, + "loss": 0.2499, + "step": 4291 + }, + { + "epoch": 1.4330550918196994, + "grad_norm": 0.4382523532677162, + "learning_rate": 2.254484094250685e-06, + "loss": 0.2399, + "step": 4292 + }, + { + "epoch": 1.43338898163606, + "grad_norm": 0.47295100489856123, + "learning_rate": 2.2520493904640555e-06, + "loss": 0.2599, + "step": 4293 + }, + { + "epoch": 1.4337228714524208, + "grad_norm": 0.43452977078624627, + "learning_rate": 2.2496156198672453e-06, + "loss": 0.2414, + "step": 4294 + }, + { + "epoch": 1.4340567612687813, + "grad_norm": 0.44807752294210396, + "learning_rate": 2.24718278328675e-06, + "loss": 0.2427, + "step": 4295 + }, + { + "epoch": 1.4343906510851419, + "grad_norm": 0.4485355833007276, + "learning_rate": 2.244750881548746e-06, + "loss": 0.2479, + "step": 4296 + }, + { + "epoch": 1.4347245409015026, + "grad_norm": 0.4688514817596998, + "learning_rate": 2.2423199154790975e-06, + "loss": 0.2565, + "step": 4297 + }, + { + "epoch": 1.435058430717863, + "grad_norm": 0.4453160535121835, + "learning_rate": 2.2398898859033496e-06, + "loss": 0.2526, + "step": 4298 + }, + { + "epoch": 1.4353923205342238, + "grad_norm": 0.46666238405490756, + "learning_rate": 2.237460793646725e-06, + "loss": 0.2554, + "step": 4299 + }, + { + "epoch": 1.4357262103505843, + "grad_norm": 0.4599395850752696, + "learning_rate": 2.235032639534132e-06, + "loss": 0.2592, + "step": 4300 + }, + { + "epoch": 1.4360601001669449, + "grad_norm": 0.43814637416156826, + "learning_rate": 2.2326054243901567e-06, + "loss": 0.2526, + "step": 4301 + }, + { + "epoch": 1.4363939899833056, + "grad_norm": 0.44704637480553183, + "learning_rate": 2.2301791490390727e-06, + "loss": 0.2424, + "step": 4302 + }, + { + "epoch": 1.436727879799666, + "grad_norm": 0.4659500769506624, + "learning_rate": 2.227753814304829e-06, + "loss": 0.26, + "step": 4303 + }, + { + "epoch": 1.4370617696160268, + "grad_norm": 0.48098131002183203, + "learning_rate": 2.225329421011056e-06, + "loss": 0.2595, + "step": 4304 + }, + { + "epoch": 1.4373956594323873, + "grad_norm": 0.46915236731483545, + "learning_rate": 2.222905969981064e-06, + "loss": 0.258, + "step": 4305 + }, + { + "epoch": 1.4377295492487479, + "grad_norm": 0.4635584185962312, + "learning_rate": 2.2204834620378455e-06, + "loss": 0.2506, + "step": 4306 + }, + { + "epoch": 1.4380634390651086, + "grad_norm": 0.471473098195624, + "learning_rate": 2.2180618980040747e-06, + "loss": 0.2559, + "step": 4307 + }, + { + "epoch": 1.438397328881469, + "grad_norm": 0.469994076527872, + "learning_rate": 2.2156412787020993e-06, + "loss": 0.25, + "step": 4308 + }, + { + "epoch": 1.4387312186978298, + "grad_norm": 0.46671702942496684, + "learning_rate": 2.2132216049539495e-06, + "loss": 0.2456, + "step": 4309 + }, + { + "epoch": 1.4390651085141903, + "grad_norm": 0.4466296771253124, + "learning_rate": 2.210802877581332e-06, + "loss": 0.2404, + "step": 4310 + }, + { + "epoch": 1.4393989983305508, + "grad_norm": 0.44202907720816764, + "learning_rate": 2.208385097405638e-06, + "loss": 0.2493, + "step": 4311 + }, + { + "epoch": 1.4397328881469116, + "grad_norm": 0.49578114346858426, + "learning_rate": 2.2059682652479313e-06, + "loss": 0.2647, + "step": 4312 + }, + { + "epoch": 1.440066777963272, + "grad_norm": 0.4667699802700531, + "learning_rate": 2.203552381928956e-06, + "loss": 0.2606, + "step": 4313 + }, + { + "epoch": 1.4404006677796328, + "grad_norm": 0.4331490410032939, + "learning_rate": 2.2011374482691306e-06, + "loss": 0.2399, + "step": 4314 + }, + { + "epoch": 1.4407345575959933, + "grad_norm": 0.44611025499766443, + "learning_rate": 2.198723465088557e-06, + "loss": 0.2448, + "step": 4315 + }, + { + "epoch": 1.4410684474123538, + "grad_norm": 0.4390832197691051, + "learning_rate": 2.196310433207013e-06, + "loss": 0.2393, + "step": 4316 + }, + { + "epoch": 1.4414023372287146, + "grad_norm": 0.4582088704090086, + "learning_rate": 2.1938983534439496e-06, + "loss": 0.2578, + "step": 4317 + }, + { + "epoch": 1.441736227045075, + "grad_norm": 0.44529683952351706, + "learning_rate": 2.191487226618497e-06, + "loss": 0.2514, + "step": 4318 + }, + { + "epoch": 1.4420701168614358, + "grad_norm": 0.4675944950292698, + "learning_rate": 2.189077053549459e-06, + "loss": 0.2641, + "step": 4319 + }, + { + "epoch": 1.4424040066777963, + "grad_norm": 0.4524799788880056, + "learning_rate": 2.1866678350553215e-06, + "loss": 0.2559, + "step": 4320 + }, + { + "epoch": 1.4427378964941568, + "grad_norm": 0.44027421545224443, + "learning_rate": 2.184259571954242e-06, + "loss": 0.2571, + "step": 4321 + }, + { + "epoch": 1.4430717863105176, + "grad_norm": 0.44928775668052773, + "learning_rate": 2.181852265064051e-06, + "loss": 0.2475, + "step": 4322 + }, + { + "epoch": 1.443405676126878, + "grad_norm": 0.46618174794863554, + "learning_rate": 2.179445915202262e-06, + "loss": 0.2573, + "step": 4323 + }, + { + "epoch": 1.4437395659432388, + "grad_norm": 0.44930191859598273, + "learning_rate": 2.1770405231860557e-06, + "loss": 0.247, + "step": 4324 + }, + { + "epoch": 1.4440734557595993, + "grad_norm": 0.45838576978251766, + "learning_rate": 2.174636089832293e-06, + "loss": 0.2494, + "step": 4325 + }, + { + "epoch": 1.4444073455759598, + "grad_norm": 0.44860753427747124, + "learning_rate": 2.1722326159575068e-06, + "loss": 0.2569, + "step": 4326 + }, + { + "epoch": 1.4447412353923206, + "grad_norm": 0.46939515552737204, + "learning_rate": 2.169830102377903e-06, + "loss": 0.2607, + "step": 4327 + }, + { + "epoch": 1.445075125208681, + "grad_norm": 0.45874052373859087, + "learning_rate": 2.1674285499093618e-06, + "loss": 0.2541, + "step": 4328 + }, + { + "epoch": 1.4454090150250418, + "grad_norm": 0.4592639100892928, + "learning_rate": 2.165027959367441e-06, + "loss": 0.253, + "step": 4329 + }, + { + "epoch": 1.4457429048414023, + "grad_norm": 0.46670888740331246, + "learning_rate": 2.162628331567368e-06, + "loss": 0.2517, + "step": 4330 + }, + { + "epoch": 1.4460767946577628, + "grad_norm": 0.4601883822171726, + "learning_rate": 2.1602296673240416e-06, + "loss": 0.2457, + "step": 4331 + }, + { + "epoch": 1.4464106844741236, + "grad_norm": 0.44846323659062615, + "learning_rate": 2.1578319674520385e-06, + "loss": 0.2549, + "step": 4332 + }, + { + "epoch": 1.446744574290484, + "grad_norm": 0.4387290479456937, + "learning_rate": 2.155435232765603e-06, + "loss": 0.2453, + "step": 4333 + }, + { + "epoch": 1.4470784641068448, + "grad_norm": 0.4467497097932715, + "learning_rate": 2.1530394640786567e-06, + "loss": 0.2639, + "step": 4334 + }, + { + "epoch": 1.4474123539232053, + "grad_norm": 0.4268354076318614, + "learning_rate": 2.1506446622047887e-06, + "loss": 0.243, + "step": 4335 + }, + { + "epoch": 1.4477462437395658, + "grad_norm": 0.44811223001524414, + "learning_rate": 2.148250827957261e-06, + "loss": 0.2441, + "step": 4336 + }, + { + "epoch": 1.4480801335559266, + "grad_norm": 0.4236006843680659, + "learning_rate": 2.1458579621490056e-06, + "loss": 0.2449, + "step": 4337 + }, + { + "epoch": 1.448414023372287, + "grad_norm": 0.4753902344801476, + "learning_rate": 2.1434660655926316e-06, + "loss": 0.2586, + "step": 4338 + }, + { + "epoch": 1.4487479131886478, + "grad_norm": 0.4485047492165278, + "learning_rate": 2.1410751391004105e-06, + "loss": 0.2507, + "step": 4339 + }, + { + "epoch": 1.4490818030050083, + "grad_norm": 0.4473056267779963, + "learning_rate": 2.1386851834842916e-06, + "loss": 0.2338, + "step": 4340 + }, + { + "epoch": 1.4494156928213688, + "grad_norm": 0.4503315208908993, + "learning_rate": 2.1362961995558904e-06, + "loss": 0.2395, + "step": 4341 + }, + { + "epoch": 1.4497495826377296, + "grad_norm": 0.4693848688233949, + "learning_rate": 2.133908188126492e-06, + "loss": 0.26, + "step": 4342 + }, + { + "epoch": 1.45008347245409, + "grad_norm": 0.463310587144955, + "learning_rate": 2.131521150007056e-06, + "loss": 0.2479, + "step": 4343 + }, + { + "epoch": 1.4504173622704508, + "grad_norm": 0.4663086391779457, + "learning_rate": 2.1291350860082066e-06, + "loss": 0.2536, + "step": 4344 + }, + { + "epoch": 1.4507512520868113, + "grad_norm": 0.4675583739677409, + "learning_rate": 2.1267499969402383e-06, + "loss": 0.255, + "step": 4345 + }, + { + "epoch": 1.4510851419031718, + "grad_norm": 0.4548676146749028, + "learning_rate": 2.1243658836131133e-06, + "loss": 0.2527, + "step": 4346 + }, + { + "epoch": 1.4514190317195326, + "grad_norm": 0.46426377180824013, + "learning_rate": 2.121982746836468e-06, + "loss": 0.2572, + "step": 4347 + }, + { + "epoch": 1.451752921535893, + "grad_norm": 0.4464428141773434, + "learning_rate": 2.1196005874195995e-06, + "loss": 0.2392, + "step": 4348 + }, + { + "epoch": 1.4520868113522538, + "grad_norm": 0.45427564762549794, + "learning_rate": 2.117219406171481e-06, + "loss": 0.2472, + "step": 4349 + }, + { + "epoch": 1.4524207011686143, + "grad_norm": 0.4494718405994756, + "learning_rate": 2.114839203900747e-06, + "loss": 0.2465, + "step": 4350 + }, + { + "epoch": 1.4527545909849748, + "grad_norm": 0.4917005776237037, + "learning_rate": 2.1124599814157006e-06, + "loss": 0.2625, + "step": 4351 + }, + { + "epoch": 1.4530884808013356, + "grad_norm": 0.4340381225707385, + "learning_rate": 2.110081739524316e-06, + "loss": 0.252, + "step": 4352 + }, + { + "epoch": 1.453422370617696, + "grad_norm": 0.461781049256306, + "learning_rate": 2.107704479034231e-06, + "loss": 0.246, + "step": 4353 + }, + { + "epoch": 1.4537562604340568, + "grad_norm": 0.4535241872543963, + "learning_rate": 2.1053282007527503e-06, + "loss": 0.2582, + "step": 4354 + }, + { + "epoch": 1.4540901502504173, + "grad_norm": 0.45062908437404686, + "learning_rate": 2.1029529054868447e-06, + "loss": 0.2567, + "step": 4355 + }, + { + "epoch": 1.4544240400667778, + "grad_norm": 0.4649682913079168, + "learning_rate": 2.1005785940431533e-06, + "loss": 0.2471, + "step": 4356 + }, + { + "epoch": 1.4547579298831386, + "grad_norm": 0.4449367622402766, + "learning_rate": 2.0982052672279825e-06, + "loss": 0.2529, + "step": 4357 + }, + { + "epoch": 1.4550918196994993, + "grad_norm": 0.4487140391771142, + "learning_rate": 2.095832925847299e-06, + "loss": 0.2427, + "step": 4358 + }, + { + "epoch": 1.4554257095158598, + "grad_norm": 0.4555519451200949, + "learning_rate": 2.093461570706739e-06, + "loss": 0.2428, + "step": 4359 + }, + { + "epoch": 1.4557595993322203, + "grad_norm": 0.4676603056385322, + "learning_rate": 2.0910912026116e-06, + "loss": 0.2552, + "step": 4360 + }, + { + "epoch": 1.4560934891485808, + "grad_norm": 0.4659455736412407, + "learning_rate": 2.0887218223668493e-06, + "loss": 0.2481, + "step": 4361 + }, + { + "epoch": 1.4564273789649416, + "grad_norm": 0.4651000643923732, + "learning_rate": 2.086353430777116e-06, + "loss": 0.2633, + "step": 4362 + }, + { + "epoch": 1.4567612687813023, + "grad_norm": 0.45820574978697604, + "learning_rate": 2.0839860286466927e-06, + "loss": 0.2612, + "step": 4363 + }, + { + "epoch": 1.4570951585976628, + "grad_norm": 0.4553923612700206, + "learning_rate": 2.081619616779536e-06, + "loss": 0.2439, + "step": 4364 + }, + { + "epoch": 1.4574290484140233, + "grad_norm": 0.45267933878891575, + "learning_rate": 2.0792541959792676e-06, + "loss": 0.2419, + "step": 4365 + }, + { + "epoch": 1.457762938230384, + "grad_norm": 0.45168961759243875, + "learning_rate": 2.076889767049175e-06, + "loss": 0.2578, + "step": 4366 + }, + { + "epoch": 1.4580968280467446, + "grad_norm": 0.45094558056225986, + "learning_rate": 2.0745263307922043e-06, + "loss": 0.2385, + "step": 4367 + }, + { + "epoch": 1.4584307178631053, + "grad_norm": 0.4489749513440614, + "learning_rate": 2.072163888010966e-06, + "loss": 0.2438, + "step": 4368 + }, + { + "epoch": 1.4587646076794658, + "grad_norm": 0.4513424227603912, + "learning_rate": 2.0698024395077303e-06, + "loss": 0.2491, + "step": 4369 + }, + { + "epoch": 1.4590984974958263, + "grad_norm": 0.4517847644098309, + "learning_rate": 2.0674419860844385e-06, + "loss": 0.2372, + "step": 4370 + }, + { + "epoch": 1.459432387312187, + "grad_norm": 0.4435055354021787, + "learning_rate": 2.065082528542686e-06, + "loss": 0.2461, + "step": 4371 + }, + { + "epoch": 1.4597662771285476, + "grad_norm": 0.4629080050049465, + "learning_rate": 2.0627240676837325e-06, + "loss": 0.2511, + "step": 4372 + }, + { + "epoch": 1.4601001669449083, + "grad_norm": 0.45017092824087607, + "learning_rate": 2.0603666043084964e-06, + "loss": 0.2566, + "step": 4373 + }, + { + "epoch": 1.4604340567612688, + "grad_norm": 0.46951223046710727, + "learning_rate": 2.0580101392175626e-06, + "loss": 0.2548, + "step": 4374 + }, + { + "epoch": 1.4607679465776293, + "grad_norm": 0.4569798921281256, + "learning_rate": 2.0556546732111758e-06, + "loss": 0.2541, + "step": 4375 + }, + { + "epoch": 1.46110183639399, + "grad_norm": 0.4450019885329337, + "learning_rate": 2.0533002070892395e-06, + "loss": 0.2426, + "step": 4376 + }, + { + "epoch": 1.4614357262103506, + "grad_norm": 0.44907969813427584, + "learning_rate": 2.050946741651317e-06, + "loss": 0.2509, + "step": 4377 + }, + { + "epoch": 1.4617696160267113, + "grad_norm": 0.4732590315962745, + "learning_rate": 2.0485942776966313e-06, + "loss": 0.2514, + "step": 4378 + }, + { + "epoch": 1.4621035058430718, + "grad_norm": 0.4493758935453374, + "learning_rate": 2.046242816024071e-06, + "loss": 0.2405, + "step": 4379 + }, + { + "epoch": 1.4624373956594323, + "grad_norm": 0.44093870323477596, + "learning_rate": 2.043892357432178e-06, + "loss": 0.248, + "step": 4380 + }, + { + "epoch": 1.462771285475793, + "grad_norm": 0.45957683486297807, + "learning_rate": 2.041542902719157e-06, + "loss": 0.2478, + "step": 4381 + }, + { + "epoch": 1.4631051752921536, + "grad_norm": 0.4621090110985951, + "learning_rate": 2.039194452682867e-06, + "loss": 0.2618, + "step": 4382 + }, + { + "epoch": 1.4634390651085143, + "grad_norm": 0.45844139005923595, + "learning_rate": 2.0368470081208335e-06, + "loss": 0.2543, + "step": 4383 + }, + { + "epoch": 1.4637729549248748, + "grad_norm": 0.44553263922737113, + "learning_rate": 2.0345005698302367e-06, + "loss": 0.2459, + "step": 4384 + }, + { + "epoch": 1.4641068447412353, + "grad_norm": 0.4494391070643716, + "learning_rate": 2.032155138607914e-06, + "loss": 0.2581, + "step": 4385 + }, + { + "epoch": 1.464440734557596, + "grad_norm": 0.4600442693679695, + "learning_rate": 2.029810715250361e-06, + "loss": 0.2492, + "step": 4386 + }, + { + "epoch": 1.4647746243739566, + "grad_norm": 0.4578507134062572, + "learning_rate": 2.0274673005537305e-06, + "loss": 0.254, + "step": 4387 + }, + { + "epoch": 1.4651085141903173, + "grad_norm": 0.44955704362023563, + "learning_rate": 2.0251248953138377e-06, + "loss": 0.2487, + "step": 4388 + }, + { + "epoch": 1.4654424040066778, + "grad_norm": 0.4455359153691052, + "learning_rate": 2.0227835003261486e-06, + "loss": 0.2396, + "step": 4389 + }, + { + "epoch": 1.4657762938230383, + "grad_norm": 0.47502699220638434, + "learning_rate": 2.0204431163857887e-06, + "loss": 0.2606, + "step": 4390 + }, + { + "epoch": 1.466110183639399, + "grad_norm": 0.4468476954721313, + "learning_rate": 2.0181037442875417e-06, + "loss": 0.2489, + "step": 4391 + }, + { + "epoch": 1.4664440734557596, + "grad_norm": 0.46007974656446987, + "learning_rate": 2.0157653848258445e-06, + "loss": 0.2553, + "step": 4392 + }, + { + "epoch": 1.4667779632721203, + "grad_norm": 0.461570254668703, + "learning_rate": 2.0134280387947947e-06, + "loss": 0.2507, + "step": 4393 + }, + { + "epoch": 1.4671118530884808, + "grad_norm": 0.45728175366593754, + "learning_rate": 2.0110917069881413e-06, + "loss": 0.2519, + "step": 4394 + }, + { + "epoch": 1.4674457429048413, + "grad_norm": 0.46124200029736584, + "learning_rate": 2.008756390199291e-06, + "loss": 0.2486, + "step": 4395 + }, + { + "epoch": 1.467779632721202, + "grad_norm": 0.4532590455437685, + "learning_rate": 2.006422089221302e-06, + "loss": 0.2474, + "step": 4396 + }, + { + "epoch": 1.4681135225375626, + "grad_norm": 0.4290772815900165, + "learning_rate": 2.0040888048468954e-06, + "loss": 0.2527, + "step": 4397 + }, + { + "epoch": 1.4684474123539233, + "grad_norm": 0.4513427298097657, + "learning_rate": 2.001756537868441e-06, + "loss": 0.2574, + "step": 4398 + }, + { + "epoch": 1.4687813021702838, + "grad_norm": 0.4641400116075369, + "learning_rate": 1.999425289077963e-06, + "loss": 0.2619, + "step": 4399 + }, + { + "epoch": 1.4691151919866443, + "grad_norm": 0.45509435952583427, + "learning_rate": 1.997095059267144e-06, + "loss": 0.2562, + "step": 4400 + }, + { + "epoch": 1.469449081803005, + "grad_norm": 0.4563484336844938, + "learning_rate": 1.9947658492273153e-06, + "loss": 0.2547, + "step": 4401 + }, + { + "epoch": 1.4697829716193656, + "grad_norm": 0.4541384972190946, + "learning_rate": 1.9924376597494676e-06, + "loss": 0.2495, + "step": 4402 + }, + { + "epoch": 1.4701168614357263, + "grad_norm": 0.45121093671800394, + "learning_rate": 1.9901104916242403e-06, + "loss": 0.2466, + "step": 4403 + }, + { + "epoch": 1.4704507512520868, + "grad_norm": 0.429163881965237, + "learning_rate": 1.987784345641928e-06, + "loss": 0.2466, + "step": 4404 + }, + { + "epoch": 1.4707846410684473, + "grad_norm": 0.48278808380904087, + "learning_rate": 1.985459222592476e-06, + "loss": 0.2632, + "step": 4405 + }, + { + "epoch": 1.471118530884808, + "grad_norm": 0.44720227029471166, + "learning_rate": 1.9831351232654874e-06, + "loss": 0.2535, + "step": 4406 + }, + { + "epoch": 1.4714524207011686, + "grad_norm": 0.4414300708334768, + "learning_rate": 1.98081204845021e-06, + "loss": 0.2475, + "step": 4407 + }, + { + "epoch": 1.4717863105175293, + "grad_norm": 0.4738938990908965, + "learning_rate": 1.978489998935554e-06, + "loss": 0.2658, + "step": 4408 + }, + { + "epoch": 1.4721202003338898, + "grad_norm": 0.44036635425967713, + "learning_rate": 1.976168975510071e-06, + "loss": 0.2441, + "step": 4409 + }, + { + "epoch": 1.4724540901502503, + "grad_norm": 0.4419343379378365, + "learning_rate": 1.973848978961968e-06, + "loss": 0.2454, + "step": 4410 + }, + { + "epoch": 1.472787979966611, + "grad_norm": 0.4559815964096995, + "learning_rate": 1.9715300100791073e-06, + "loss": 0.2494, + "step": 4411 + }, + { + "epoch": 1.4731218697829715, + "grad_norm": 0.4539136655405124, + "learning_rate": 1.969212069648997e-06, + "loss": 0.2429, + "step": 4412 + }, + { + "epoch": 1.4734557595993323, + "grad_norm": 0.4529910071025664, + "learning_rate": 1.966895158458797e-06, + "loss": 0.2506, + "step": 4413 + }, + { + "epoch": 1.4737896494156928, + "grad_norm": 0.44562777423668715, + "learning_rate": 1.9645792772953175e-06, + "loss": 0.2453, + "step": 4414 + }, + { + "epoch": 1.4741235392320533, + "grad_norm": 0.47387186762408345, + "learning_rate": 1.962264426945023e-06, + "loss": 0.2527, + "step": 4415 + }, + { + "epoch": 1.474457429048414, + "grad_norm": 0.4656225845809016, + "learning_rate": 1.9599506081940212e-06, + "loss": 0.2511, + "step": 4416 + }, + { + "epoch": 1.4747913188647745, + "grad_norm": 0.46415531233565094, + "learning_rate": 1.9576378218280763e-06, + "loss": 0.2569, + "step": 4417 + }, + { + "epoch": 1.4751252086811353, + "grad_norm": 0.45940072007943183, + "learning_rate": 1.955326068632597e-06, + "loss": 0.2584, + "step": 4418 + }, + { + "epoch": 1.4754590984974958, + "grad_norm": 0.46505459819281453, + "learning_rate": 1.953015349392641e-06, + "loss": 0.2548, + "step": 4419 + }, + { + "epoch": 1.4757929883138563, + "grad_norm": 0.45317506670241314, + "learning_rate": 1.95070566489292e-06, + "loss": 0.2474, + "step": 4420 + }, + { + "epoch": 1.476126878130217, + "grad_norm": 0.4565149611887697, + "learning_rate": 1.9483970159177896e-06, + "loss": 0.2431, + "step": 4421 + }, + { + "epoch": 1.4764607679465778, + "grad_norm": 0.45018750180417183, + "learning_rate": 1.9460894032512538e-06, + "loss": 0.246, + "step": 4422 + }, + { + "epoch": 1.4767946577629383, + "grad_norm": 0.4506933605242086, + "learning_rate": 1.943782827676966e-06, + "loss": 0.248, + "step": 4423 + }, + { + "epoch": 1.4771285475792988, + "grad_norm": 0.4494272482851282, + "learning_rate": 1.9414772899782274e-06, + "loss": 0.2544, + "step": 4424 + }, + { + "epoch": 1.4774624373956593, + "grad_norm": 0.4577829204570946, + "learning_rate": 1.9391727909379897e-06, + "loss": 0.261, + "step": 4425 + }, + { + "epoch": 1.47779632721202, + "grad_norm": 0.46381498522772907, + "learning_rate": 1.9368693313388465e-06, + "loss": 0.2483, + "step": 4426 + }, + { + "epoch": 1.4781302170283808, + "grad_norm": 0.44739269641963075, + "learning_rate": 1.9345669119630413e-06, + "loss": 0.249, + "step": 4427 + }, + { + "epoch": 1.4784641068447413, + "grad_norm": 0.45341927785230146, + "learning_rate": 1.9322655335924605e-06, + "loss": 0.2486, + "step": 4428 + }, + { + "epoch": 1.4787979966611018, + "grad_norm": 0.46099940060961847, + "learning_rate": 1.9299651970086457e-06, + "loss": 0.2589, + "step": 4429 + }, + { + "epoch": 1.4791318864774623, + "grad_norm": 0.44766699378387864, + "learning_rate": 1.9276659029927768e-06, + "loss": 0.2513, + "step": 4430 + }, + { + "epoch": 1.479465776293823, + "grad_norm": 0.46506777014273926, + "learning_rate": 1.925367652325682e-06, + "loss": 0.2459, + "step": 4431 + }, + { + "epoch": 1.4797996661101838, + "grad_norm": 0.4327885999452541, + "learning_rate": 1.9230704457878335e-06, + "loss": 0.2424, + "step": 4432 + }, + { + "epoch": 1.4801335559265443, + "grad_norm": 0.4624757370068729, + "learning_rate": 1.920774284159353e-06, + "loss": 0.2506, + "step": 4433 + }, + { + "epoch": 1.4804674457429048, + "grad_norm": 0.4750937815387872, + "learning_rate": 1.9184791682200064e-06, + "loss": 0.2496, + "step": 4434 + }, + { + "epoch": 1.4808013355592655, + "grad_norm": 0.46062163028960573, + "learning_rate": 1.916185098749202e-06, + "loss": 0.2649, + "step": 4435 + }, + { + "epoch": 1.481135225375626, + "grad_norm": 0.46737329335806277, + "learning_rate": 1.913892076525993e-06, + "loss": 0.2593, + "step": 4436 + }, + { + "epoch": 1.4814691151919868, + "grad_norm": 0.45000666792090177, + "learning_rate": 1.9116001023290776e-06, + "loss": 0.2394, + "step": 4437 + }, + { + "epoch": 1.4818030050083473, + "grad_norm": 0.45361914361802547, + "learning_rate": 1.9093091769368006e-06, + "loss": 0.2436, + "step": 4438 + }, + { + "epoch": 1.4821368948247078, + "grad_norm": 0.45147111265192147, + "learning_rate": 1.907019301127147e-06, + "loss": 0.2537, + "step": 4439 + }, + { + "epoch": 1.4824707846410685, + "grad_norm": 0.4563167080800117, + "learning_rate": 1.9047304756777473e-06, + "loss": 0.2548, + "step": 4440 + }, + { + "epoch": 1.482804674457429, + "grad_norm": 0.47059986680261184, + "learning_rate": 1.9024427013658725e-06, + "loss": 0.2526, + "step": 4441 + }, + { + "epoch": 1.4831385642737898, + "grad_norm": 0.46864903513775114, + "learning_rate": 1.9001559789684403e-06, + "loss": 0.2387, + "step": 4442 + }, + { + "epoch": 1.4834724540901503, + "grad_norm": 0.44513222800183333, + "learning_rate": 1.8978703092620128e-06, + "loss": 0.244, + "step": 4443 + }, + { + "epoch": 1.4838063439065108, + "grad_norm": 0.4708603895691993, + "learning_rate": 1.8955856930227884e-06, + "loss": 0.2536, + "step": 4444 + }, + { + "epoch": 1.4841402337228715, + "grad_norm": 0.4462047815177883, + "learning_rate": 1.893302131026612e-06, + "loss": 0.2481, + "step": 4445 + }, + { + "epoch": 1.484474123539232, + "grad_norm": 0.4420037154349952, + "learning_rate": 1.8910196240489676e-06, + "loss": 0.2492, + "step": 4446 + }, + { + "epoch": 1.4848080133555928, + "grad_norm": 0.472567272542776, + "learning_rate": 1.8887381728649856e-06, + "loss": 0.2645, + "step": 4447 + }, + { + "epoch": 1.4851419031719533, + "grad_norm": 0.46720379577785903, + "learning_rate": 1.8864577782494343e-06, + "loss": 0.2524, + "step": 4448 + }, + { + "epoch": 1.4854757929883138, + "grad_norm": 0.497891112884876, + "learning_rate": 1.8841784409767228e-06, + "loss": 0.2656, + "step": 4449 + }, + { + "epoch": 1.4858096828046745, + "grad_norm": 0.459910752443667, + "learning_rate": 1.881900161820901e-06, + "loss": 0.2571, + "step": 4450 + }, + { + "epoch": 1.486143572621035, + "grad_norm": 0.4557623118365, + "learning_rate": 1.8796229415556628e-06, + "loss": 0.2502, + "step": 4451 + }, + { + "epoch": 1.4864774624373958, + "grad_norm": 0.45413943794217504, + "learning_rate": 1.8773467809543422e-06, + "loss": 0.2542, + "step": 4452 + }, + { + "epoch": 1.4868113522537563, + "grad_norm": 0.4451745119541537, + "learning_rate": 1.8750716807899089e-06, + "loss": 0.2559, + "step": 4453 + }, + { + "epoch": 1.4871452420701168, + "grad_norm": 0.4618656415623637, + "learning_rate": 1.8727976418349758e-06, + "loss": 0.2553, + "step": 4454 + }, + { + "epoch": 1.4874791318864775, + "grad_norm": 0.45886226141557984, + "learning_rate": 1.8705246648617953e-06, + "loss": 0.2536, + "step": 4455 + }, + { + "epoch": 1.487813021702838, + "grad_norm": 0.43141494307476297, + "learning_rate": 1.868252750642256e-06, + "loss": 0.2302, + "step": 4456 + }, + { + "epoch": 1.4881469115191988, + "grad_norm": 0.4481994347970021, + "learning_rate": 1.8659818999478918e-06, + "loss": 0.2585, + "step": 4457 + }, + { + "epoch": 1.4884808013355593, + "grad_norm": 0.4986302479522215, + "learning_rate": 1.8637121135498693e-06, + "loss": 0.2704, + "step": 4458 + }, + { + "epoch": 1.4888146911519198, + "grad_norm": 0.44655334489113424, + "learning_rate": 1.8614433922189984e-06, + "loss": 0.2487, + "step": 4459 + }, + { + "epoch": 1.4891485809682805, + "grad_norm": 0.4468344647961806, + "learning_rate": 1.859175736725724e-06, + "loss": 0.2449, + "step": 4460 + }, + { + "epoch": 1.489482470784641, + "grad_norm": 0.44365352207953185, + "learning_rate": 1.8569091478401286e-06, + "loss": 0.2509, + "step": 4461 + }, + { + "epoch": 1.4898163606010018, + "grad_norm": 0.4572275241736437, + "learning_rate": 1.8546436263319379e-06, + "loss": 0.2551, + "step": 4462 + }, + { + "epoch": 1.4901502504173623, + "grad_norm": 0.44081219660770726, + "learning_rate": 1.8523791729705082e-06, + "loss": 0.2532, + "step": 4463 + }, + { + "epoch": 1.4904841402337228, + "grad_norm": 0.4401163505508295, + "learning_rate": 1.8501157885248373e-06, + "loss": 0.2471, + "step": 4464 + }, + { + "epoch": 1.4908180300500835, + "grad_norm": 0.4544784757741447, + "learning_rate": 1.8478534737635568e-06, + "loss": 0.2508, + "step": 4465 + }, + { + "epoch": 1.491151919866444, + "grad_norm": 0.45228497902840553, + "learning_rate": 1.845592229454941e-06, + "loss": 0.249, + "step": 4466 + }, + { + "epoch": 1.4914858096828048, + "grad_norm": 0.4712149469095369, + "learning_rate": 1.8433320563668921e-06, + "loss": 0.2655, + "step": 4467 + }, + { + "epoch": 1.4918196994991653, + "grad_norm": 0.4754024485757981, + "learning_rate": 1.8410729552669582e-06, + "loss": 0.263, + "step": 4468 + }, + { + "epoch": 1.4921535893155258, + "grad_norm": 0.44416460886718684, + "learning_rate": 1.8388149269223153e-06, + "loss": 0.2582, + "step": 4469 + }, + { + "epoch": 1.4924874791318865, + "grad_norm": 0.44269962105750427, + "learning_rate": 1.8365579720997779e-06, + "loss": 0.2502, + "step": 4470 + }, + { + "epoch": 1.492821368948247, + "grad_norm": 0.4601866961956389, + "learning_rate": 1.8343020915657977e-06, + "loss": 0.2564, + "step": 4471 + }, + { + "epoch": 1.4931552587646078, + "grad_norm": 0.45716329920342713, + "learning_rate": 1.8320472860864596e-06, + "loss": 0.2606, + "step": 4472 + }, + { + "epoch": 1.4934891485809683, + "grad_norm": 0.4423349386290148, + "learning_rate": 1.8297935564274844e-06, + "loss": 0.2467, + "step": 4473 + }, + { + "epoch": 1.4938230383973288, + "grad_norm": 0.4530289386662238, + "learning_rate": 1.827540903354224e-06, + "loss": 0.2519, + "step": 4474 + }, + { + "epoch": 1.4941569282136895, + "grad_norm": 0.45282121393976404, + "learning_rate": 1.8252893276316696e-06, + "loss": 0.2545, + "step": 4475 + }, + { + "epoch": 1.49449081803005, + "grad_norm": 0.446551441535247, + "learning_rate": 1.823038830024448e-06, + "loss": 0.2509, + "step": 4476 + }, + { + "epoch": 1.4948247078464107, + "grad_norm": 0.4423184278017057, + "learning_rate": 1.8207894112968126e-06, + "loss": 0.2499, + "step": 4477 + }, + { + "epoch": 1.4951585976627713, + "grad_norm": 0.45893584693354056, + "learning_rate": 1.8185410722126556e-06, + "loss": 0.2582, + "step": 4478 + }, + { + "epoch": 1.4954924874791318, + "grad_norm": 0.45533858623007356, + "learning_rate": 1.8162938135354995e-06, + "loss": 0.2592, + "step": 4479 + }, + { + "epoch": 1.4958263772954925, + "grad_norm": 0.4531259230989938, + "learning_rate": 1.8140476360285053e-06, + "loss": 0.2468, + "step": 4480 + }, + { + "epoch": 1.496160267111853, + "grad_norm": 0.4409150976917041, + "learning_rate": 1.8118025404544614e-06, + "loss": 0.2445, + "step": 4481 + }, + { + "epoch": 1.4964941569282137, + "grad_norm": 0.4434647054059321, + "learning_rate": 1.8095585275757904e-06, + "loss": 0.2413, + "step": 4482 + }, + { + "epoch": 1.4968280467445743, + "grad_norm": 0.47006158902766537, + "learning_rate": 1.8073155981545465e-06, + "loss": 0.2472, + "step": 4483 + }, + { + "epoch": 1.4971619365609348, + "grad_norm": 0.4494213727189019, + "learning_rate": 1.8050737529524176e-06, + "loss": 0.2441, + "step": 4484 + }, + { + "epoch": 1.4974958263772955, + "grad_norm": 0.4488278324907001, + "learning_rate": 1.8028329927307247e-06, + "loss": 0.2536, + "step": 4485 + }, + { + "epoch": 1.497829716193656, + "grad_norm": 0.455554051277272, + "learning_rate": 1.8005933182504175e-06, + "loss": 0.2517, + "step": 4486 + }, + { + "epoch": 1.4981636060100167, + "grad_norm": 0.4559890813209985, + "learning_rate": 1.7983547302720773e-06, + "loss": 0.2511, + "step": 4487 + }, + { + "epoch": 1.4984974958263773, + "grad_norm": 0.4424016720786547, + "learning_rate": 1.796117229555915e-06, + "loss": 0.2456, + "step": 4488 + }, + { + "epoch": 1.4988313856427378, + "grad_norm": 0.4436836463175337, + "learning_rate": 1.7938808168617776e-06, + "loss": 0.2467, + "step": 4489 + }, + { + "epoch": 1.4991652754590985, + "grad_norm": 0.4626056267503057, + "learning_rate": 1.791645492949139e-06, + "loss": 0.2511, + "step": 4490 + }, + { + "epoch": 1.4994991652754592, + "grad_norm": 0.4606241605150028, + "learning_rate": 1.7894112585771017e-06, + "loss": 0.244, + "step": 4491 + }, + { + "epoch": 1.4998330550918197, + "grad_norm": 0.44305415137380055, + "learning_rate": 1.7871781145044e-06, + "loss": 0.2472, + "step": 4492 + }, + { + "epoch": 1.5001669449081803, + "grad_norm": 0.4593487130076081, + "learning_rate": 1.7849460614893987e-06, + "loss": 0.2538, + "step": 4493 + }, + { + "epoch": 1.5005008347245408, + "grad_norm": 0.47519141376564117, + "learning_rate": 1.7827151002900938e-06, + "loss": 0.2536, + "step": 4494 + }, + { + "epoch": 1.5008347245409015, + "grad_norm": 0.45430263083115585, + "learning_rate": 1.7804852316641063e-06, + "loss": 0.2497, + "step": 4495 + }, + { + "epoch": 1.5011686143572622, + "grad_norm": 0.44301919238294696, + "learning_rate": 1.7782564563686882e-06, + "loss": 0.2479, + "step": 4496 + }, + { + "epoch": 1.5015025041736227, + "grad_norm": 0.45502786774449056, + "learning_rate": 1.776028775160718e-06, + "loss": 0.2611, + "step": 4497 + }, + { + "epoch": 1.5018363939899833, + "grad_norm": 0.4583764929797273, + "learning_rate": 1.7738021887967083e-06, + "loss": 0.2526, + "step": 4498 + }, + { + "epoch": 1.5021702838063438, + "grad_norm": 0.4464787452859947, + "learning_rate": 1.7715766980327947e-06, + "loss": 0.2454, + "step": 4499 + }, + { + "epoch": 1.5025041736227045, + "grad_norm": 0.48548656826073516, + "learning_rate": 1.7693523036247423e-06, + "loss": 0.2583, + "step": 4500 + }, + { + "epoch": 1.5028380634390652, + "grad_norm": 0.4471529413493275, + "learning_rate": 1.7671290063279423e-06, + "loss": 0.2443, + "step": 4501 + }, + { + "epoch": 1.5031719532554257, + "grad_norm": 0.4505348419812305, + "learning_rate": 1.7649068068974167e-06, + "loss": 0.2493, + "step": 4502 + }, + { + "epoch": 1.5035058430717863, + "grad_norm": 0.44743290259540447, + "learning_rate": 1.7626857060878143e-06, + "loss": 0.2517, + "step": 4503 + }, + { + "epoch": 1.5038397328881468, + "grad_norm": 0.44106602844394277, + "learning_rate": 1.760465704653408e-06, + "loss": 0.2493, + "step": 4504 + }, + { + "epoch": 1.5041736227045075, + "grad_norm": 0.4524812539472182, + "learning_rate": 1.7582468033480992e-06, + "loss": 0.2523, + "step": 4505 + }, + { + "epoch": 1.5045075125208682, + "grad_norm": 0.4584467474998276, + "learning_rate": 1.7560290029254129e-06, + "loss": 0.2528, + "step": 4506 + }, + { + "epoch": 1.5048414023372287, + "grad_norm": 0.4467866883929527, + "learning_rate": 1.7538123041385064e-06, + "loss": 0.2444, + "step": 4507 + }, + { + "epoch": 1.5051752921535893, + "grad_norm": 0.4803234315604944, + "learning_rate": 1.7515967077401574e-06, + "loss": 0.2574, + "step": 4508 + }, + { + "epoch": 1.5055091819699498, + "grad_norm": 0.4656374856355367, + "learning_rate": 1.7493822144827694e-06, + "loss": 0.2526, + "step": 4509 + }, + { + "epoch": 1.5058430717863105, + "grad_norm": 0.4518413197611426, + "learning_rate": 1.747168825118376e-06, + "loss": 0.2454, + "step": 4510 + }, + { + "epoch": 1.5061769616026712, + "grad_norm": 0.4502426559844138, + "learning_rate": 1.7449565403986301e-06, + "loss": 0.2564, + "step": 4511 + }, + { + "epoch": 1.5065108514190317, + "grad_norm": 0.4418413395189127, + "learning_rate": 1.7427453610748146e-06, + "loss": 0.2448, + "step": 4512 + }, + { + "epoch": 1.5068447412353922, + "grad_norm": 0.437057283881869, + "learning_rate": 1.7405352878978332e-06, + "loss": 0.2425, + "step": 4513 + }, + { + "epoch": 1.5071786310517528, + "grad_norm": 0.45879266153647374, + "learning_rate": 1.7383263216182155e-06, + "loss": 0.2465, + "step": 4514 + }, + { + "epoch": 1.5075125208681135, + "grad_norm": 0.44823398740768744, + "learning_rate": 1.7361184629861133e-06, + "loss": 0.2483, + "step": 4515 + }, + { + "epoch": 1.5078464106844742, + "grad_norm": 0.4408094494447716, + "learning_rate": 1.7339117127513072e-06, + "loss": 0.2424, + "step": 4516 + }, + { + "epoch": 1.5081803005008347, + "grad_norm": 0.4546667335521452, + "learning_rate": 1.7317060716631967e-06, + "loss": 0.2517, + "step": 4517 + }, + { + "epoch": 1.5085141903171952, + "grad_norm": 0.4547856885807853, + "learning_rate": 1.7295015404708038e-06, + "loss": 0.2463, + "step": 4518 + }, + { + "epoch": 1.508848080133556, + "grad_norm": 0.47166953484133745, + "learning_rate": 1.7272981199227802e-06, + "loss": 0.2486, + "step": 4519 + }, + { + "epoch": 1.5091819699499165, + "grad_norm": 0.4403546037048267, + "learning_rate": 1.725095810767392e-06, + "loss": 0.2463, + "step": 4520 + }, + { + "epoch": 1.5095158597662772, + "grad_norm": 0.4587869176605608, + "learning_rate": 1.7228946137525366e-06, + "loss": 0.2489, + "step": 4521 + }, + { + "epoch": 1.5098497495826377, + "grad_norm": 0.4562062027839798, + "learning_rate": 1.7206945296257255e-06, + "loss": 0.2656, + "step": 4522 + }, + { + "epoch": 1.5101836393989982, + "grad_norm": 0.44833646587380044, + "learning_rate": 1.7184955591340974e-06, + "loss": 0.252, + "step": 4523 + }, + { + "epoch": 1.510517529215359, + "grad_norm": 0.4485761974311179, + "learning_rate": 1.7162977030244088e-06, + "loss": 0.2435, + "step": 4524 + }, + { + "epoch": 1.5108514190317195, + "grad_norm": 0.439102459190539, + "learning_rate": 1.7141009620430442e-06, + "loss": 0.2561, + "step": 4525 + }, + { + "epoch": 1.5111853088480802, + "grad_norm": 0.46149847934432897, + "learning_rate": 1.7119053369360023e-06, + "loss": 0.2596, + "step": 4526 + }, + { + "epoch": 1.5115191986644407, + "grad_norm": 0.44694570291245767, + "learning_rate": 1.7097108284489084e-06, + "loss": 0.2507, + "step": 4527 + }, + { + "epoch": 1.5118530884808012, + "grad_norm": 0.43274998889824773, + "learning_rate": 1.7075174373270053e-06, + "loss": 0.2453, + "step": 4528 + }, + { + "epoch": 1.512186978297162, + "grad_norm": 0.45758433846223706, + "learning_rate": 1.7053251643151554e-06, + "loss": 0.254, + "step": 4529 + }, + { + "epoch": 1.5125208681135225, + "grad_norm": 0.4474502359045679, + "learning_rate": 1.7031340101578469e-06, + "loss": 0.24, + "step": 4530 + }, + { + "epoch": 1.5128547579298832, + "grad_norm": 0.4464743274875594, + "learning_rate": 1.7009439755991825e-06, + "loss": 0.2426, + "step": 4531 + }, + { + "epoch": 1.5131886477462437, + "grad_norm": 0.46589868902144305, + "learning_rate": 1.6987550613828863e-06, + "loss": 0.2527, + "step": 4532 + }, + { + "epoch": 1.5135225375626042, + "grad_norm": 0.4536118265843622, + "learning_rate": 1.6965672682523015e-06, + "loss": 0.2413, + "step": 4533 + }, + { + "epoch": 1.513856427378965, + "grad_norm": 0.4609976867653153, + "learning_rate": 1.6943805969503936e-06, + "loss": 0.2513, + "step": 4534 + }, + { + "epoch": 1.5141903171953257, + "grad_norm": 0.4681761559994737, + "learning_rate": 1.6921950482197425e-06, + "loss": 0.2599, + "step": 4535 + }, + { + "epoch": 1.5145242070116862, + "grad_norm": 0.44343573982701046, + "learning_rate": 1.6900106228025516e-06, + "loss": 0.2526, + "step": 4536 + }, + { + "epoch": 1.5148580968280467, + "grad_norm": 0.4621045206107263, + "learning_rate": 1.6878273214406393e-06, + "loss": 0.2542, + "step": 4537 + }, + { + "epoch": 1.5151919866444072, + "grad_norm": 0.4674473560267757, + "learning_rate": 1.6856451448754423e-06, + "loss": 0.2572, + "step": 4538 + }, + { + "epoch": 1.515525876460768, + "grad_norm": 0.4630308431573608, + "learning_rate": 1.683464093848019e-06, + "loss": 0.2452, + "step": 4539 + }, + { + "epoch": 1.5158597662771287, + "grad_norm": 0.447578577483464, + "learning_rate": 1.6812841690990423e-06, + "loss": 0.245, + "step": 4540 + }, + { + "epoch": 1.5161936560934892, + "grad_norm": 0.45771466880599787, + "learning_rate": 1.6791053713688022e-06, + "loss": 0.2498, + "step": 4541 + }, + { + "epoch": 1.5165275459098497, + "grad_norm": 0.45298017523231987, + "learning_rate": 1.6769277013972064e-06, + "loss": 0.2522, + "step": 4542 + }, + { + "epoch": 1.5168614357262102, + "grad_norm": 0.44442422067712206, + "learning_rate": 1.6747511599237816e-06, + "loss": 0.2353, + "step": 4543 + }, + { + "epoch": 1.517195325542571, + "grad_norm": 0.44083372934386833, + "learning_rate": 1.6725757476876714e-06, + "loss": 0.2457, + "step": 4544 + }, + { + "epoch": 1.5175292153589317, + "grad_norm": 0.45291845035298817, + "learning_rate": 1.6704014654276335e-06, + "loss": 0.2574, + "step": 4545 + }, + { + "epoch": 1.5178631051752922, + "grad_norm": 0.44486115224083883, + "learning_rate": 1.6682283138820426e-06, + "loss": 0.2546, + "step": 4546 + }, + { + "epoch": 1.5181969949916527, + "grad_norm": 0.438623898380514, + "learning_rate": 1.6660562937888886e-06, + "loss": 0.2495, + "step": 4547 + }, + { + "epoch": 1.5185308848080132, + "grad_norm": 0.44410838226823884, + "learning_rate": 1.6638854058857806e-06, + "loss": 0.2392, + "step": 4548 + }, + { + "epoch": 1.518864774624374, + "grad_norm": 0.46114083389884897, + "learning_rate": 1.6617156509099403e-06, + "loss": 0.2489, + "step": 4549 + }, + { + "epoch": 1.5191986644407347, + "grad_norm": 0.46094919828087777, + "learning_rate": 1.6595470295982047e-06, + "loss": 0.2491, + "step": 4550 + }, + { + "epoch": 1.5195325542570952, + "grad_norm": 0.45428513730780595, + "learning_rate": 1.6573795426870247e-06, + "loss": 0.2428, + "step": 4551 + }, + { + "epoch": 1.5198664440734557, + "grad_norm": 0.4472519126378265, + "learning_rate": 1.655213190912469e-06, + "loss": 0.2521, + "step": 4552 + }, + { + "epoch": 1.5202003338898162, + "grad_norm": 0.438539418441718, + "learning_rate": 1.6530479750102224e-06, + "loss": 0.2485, + "step": 4553 + }, + { + "epoch": 1.520534223706177, + "grad_norm": 0.46163776122541444, + "learning_rate": 1.6508838957155782e-06, + "loss": 0.2628, + "step": 4554 + }, + { + "epoch": 1.5208681135225377, + "grad_norm": 0.4554285283574826, + "learning_rate": 1.6487209537634463e-06, + "loss": 0.2601, + "step": 4555 + }, + { + "epoch": 1.5212020033388982, + "grad_norm": 0.4519484171715965, + "learning_rate": 1.6465591498883503e-06, + "loss": 0.248, + "step": 4556 + }, + { + "epoch": 1.5215358931552587, + "grad_norm": 0.46023445339508934, + "learning_rate": 1.64439848482443e-06, + "loss": 0.2472, + "step": 4557 + }, + { + "epoch": 1.5218697829716192, + "grad_norm": 0.45637178476939383, + "learning_rate": 1.6422389593054338e-06, + "loss": 0.2499, + "step": 4558 + }, + { + "epoch": 1.52220367278798, + "grad_norm": 0.4678865509734864, + "learning_rate": 1.6400805740647269e-06, + "loss": 0.2575, + "step": 4559 + }, + { + "epoch": 1.5225375626043407, + "grad_norm": 0.4356260506424383, + "learning_rate": 1.6379233298352832e-06, + "loss": 0.2444, + "step": 4560 + }, + { + "epoch": 1.5228714524207012, + "grad_norm": 0.45102301152862856, + "learning_rate": 1.6357672273496932e-06, + "loss": 0.2501, + "step": 4561 + }, + { + "epoch": 1.5232053422370617, + "grad_norm": 0.43787789176337527, + "learning_rate": 1.6336122673401606e-06, + "loss": 0.2419, + "step": 4562 + }, + { + "epoch": 1.5235392320534222, + "grad_norm": 0.44827871610198866, + "learning_rate": 1.6314584505384961e-06, + "loss": 0.2589, + "step": 4563 + }, + { + "epoch": 1.523873121869783, + "grad_norm": 0.45816013720737847, + "learning_rate": 1.6293057776761257e-06, + "loss": 0.2567, + "step": 4564 + }, + { + "epoch": 1.5242070116861437, + "grad_norm": 0.48971538479852517, + "learning_rate": 1.6271542494840832e-06, + "loss": 0.261, + "step": 4565 + }, + { + "epoch": 1.5245409015025042, + "grad_norm": 0.4443882832411184, + "learning_rate": 1.6250038666930207e-06, + "loss": 0.2432, + "step": 4566 + }, + { + "epoch": 1.5248747913188647, + "grad_norm": 0.45530642652644826, + "learning_rate": 1.6228546300331953e-06, + "loss": 0.2518, + "step": 4567 + }, + { + "epoch": 1.5252086811352252, + "grad_norm": 0.45245080990800224, + "learning_rate": 1.6207065402344746e-06, + "loss": 0.2538, + "step": 4568 + }, + { + "epoch": 1.525542570951586, + "grad_norm": 0.44453592377507295, + "learning_rate": 1.6185595980263423e-06, + "loss": 0.2422, + "step": 4569 + }, + { + "epoch": 1.5258764607679467, + "grad_norm": 0.4590751816863805, + "learning_rate": 1.6164138041378857e-06, + "loss": 0.2585, + "step": 4570 + }, + { + "epoch": 1.5262103505843072, + "grad_norm": 0.43687110185549033, + "learning_rate": 1.6142691592978083e-06, + "loss": 0.2465, + "step": 4571 + }, + { + "epoch": 1.5265442404006677, + "grad_norm": 0.43330730326649325, + "learning_rate": 1.6121256642344186e-06, + "loss": 0.2375, + "step": 4572 + }, + { + "epoch": 1.5268781302170282, + "grad_norm": 0.45187764938959296, + "learning_rate": 1.609983319675637e-06, + "loss": 0.2538, + "step": 4573 + }, + { + "epoch": 1.527212020033389, + "grad_norm": 0.43836410226976974, + "learning_rate": 1.6078421263489914e-06, + "loss": 0.25, + "step": 4574 + }, + { + "epoch": 1.5275459098497497, + "grad_norm": 0.4490767016197745, + "learning_rate": 1.605702084981622e-06, + "loss": 0.2463, + "step": 4575 + }, + { + "epoch": 1.5278797996661102, + "grad_norm": 0.466001395251547, + "learning_rate": 1.6035631963002745e-06, + "loss": 0.2484, + "step": 4576 + }, + { + "epoch": 1.5282136894824707, + "grad_norm": 0.466172783683788, + "learning_rate": 1.6014254610313035e-06, + "loss": 0.2542, + "step": 4577 + }, + { + "epoch": 1.5285475792988312, + "grad_norm": 0.4500819760966404, + "learning_rate": 1.5992888799006756e-06, + "loss": 0.2444, + "step": 4578 + }, + { + "epoch": 1.528881469115192, + "grad_norm": 0.4526680319089892, + "learning_rate": 1.59715345363396e-06, + "loss": 0.2547, + "step": 4579 + }, + { + "epoch": 1.5292153589315527, + "grad_norm": 0.4513456771790534, + "learning_rate": 1.5950191829563382e-06, + "loss": 0.2445, + "step": 4580 + }, + { + "epoch": 1.5295492487479132, + "grad_norm": 0.46265510836386126, + "learning_rate": 1.592886068592598e-06, + "loss": 0.2515, + "step": 4581 + }, + { + "epoch": 1.5298831385642737, + "grad_norm": 0.4663203675047444, + "learning_rate": 1.5907541112671321e-06, + "loss": 0.2521, + "step": 4582 + }, + { + "epoch": 1.5302170283806342, + "grad_norm": 0.45921507681256674, + "learning_rate": 1.5886233117039412e-06, + "loss": 0.2497, + "step": 4583 + }, + { + "epoch": 1.530550918196995, + "grad_norm": 0.46286772811359517, + "learning_rate": 1.5864936706266376e-06, + "loss": 0.2521, + "step": 4584 + }, + { + "epoch": 1.5308848080133557, + "grad_norm": 0.4607771628402319, + "learning_rate": 1.5843651887584326e-06, + "loss": 0.2428, + "step": 4585 + }, + { + "epoch": 1.5312186978297162, + "grad_norm": 0.4637718000690064, + "learning_rate": 1.5822378668221511e-06, + "loss": 0.2549, + "step": 4586 + }, + { + "epoch": 1.5315525876460767, + "grad_norm": 0.4401020644507396, + "learning_rate": 1.5801117055402193e-06, + "loss": 0.241, + "step": 4587 + }, + { + "epoch": 1.5318864774624374, + "grad_norm": 0.4745525197393475, + "learning_rate": 1.5779867056346682e-06, + "loss": 0.264, + "step": 4588 + }, + { + "epoch": 1.532220367278798, + "grad_norm": 0.4397734206065471, + "learning_rate": 1.5758628678271403e-06, + "loss": 0.238, + "step": 4589 + }, + { + "epoch": 1.5325542570951587, + "grad_norm": 0.47016452151671323, + "learning_rate": 1.5737401928388785e-06, + "loss": 0.2568, + "step": 4590 + }, + { + "epoch": 1.5328881469115192, + "grad_norm": 0.45597149055723774, + "learning_rate": 1.5716186813907325e-06, + "loss": 0.2556, + "step": 4591 + }, + { + "epoch": 1.5332220367278797, + "grad_norm": 0.4606127668909029, + "learning_rate": 1.569498334203154e-06, + "loss": 0.2491, + "step": 4592 + }, + { + "epoch": 1.5335559265442404, + "grad_norm": 0.44297469782719173, + "learning_rate": 1.5673791519962061e-06, + "loss": 0.2476, + "step": 4593 + }, + { + "epoch": 1.533889816360601, + "grad_norm": 0.4783085058801766, + "learning_rate": 1.565261135489549e-06, + "loss": 0.2531, + "step": 4594 + }, + { + "epoch": 1.5342237061769617, + "grad_norm": 0.46444260655028924, + "learning_rate": 1.563144285402453e-06, + "loss": 0.2632, + "step": 4595 + }, + { + "epoch": 1.5345575959933222, + "grad_norm": 0.45422900274935935, + "learning_rate": 1.561028602453788e-06, + "loss": 0.2543, + "step": 4596 + }, + { + "epoch": 1.5348914858096827, + "grad_norm": 0.4568180952467144, + "learning_rate": 1.5589140873620267e-06, + "loss": 0.2477, + "step": 4597 + }, + { + "epoch": 1.5352253756260434, + "grad_norm": 0.4715251584466027, + "learning_rate": 1.5568007408452519e-06, + "loss": 0.2559, + "step": 4598 + }, + { + "epoch": 1.535559265442404, + "grad_norm": 0.4395625015692386, + "learning_rate": 1.5546885636211422e-06, + "loss": 0.2499, + "step": 4599 + }, + { + "epoch": 1.5358931552587647, + "grad_norm": 0.46060658532976273, + "learning_rate": 1.5525775564069833e-06, + "loss": 0.2554, + "step": 4600 + }, + { + "epoch": 1.5362270450751252, + "grad_norm": 0.44983320370008634, + "learning_rate": 1.55046771991966e-06, + "loss": 0.2439, + "step": 4601 + }, + { + "epoch": 1.5365609348914857, + "grad_norm": 0.43565642735141236, + "learning_rate": 1.5483590548756633e-06, + "loss": 0.2302, + "step": 4602 + }, + { + "epoch": 1.5368948247078464, + "grad_norm": 0.45794655419323294, + "learning_rate": 1.5462515619910873e-06, + "loss": 0.2558, + "step": 4603 + }, + { + "epoch": 1.5372287145242072, + "grad_norm": 0.4400775694723566, + "learning_rate": 1.5441452419816238e-06, + "loss": 0.2377, + "step": 4604 + }, + { + "epoch": 1.5375626043405677, + "grad_norm": 0.4781960842211436, + "learning_rate": 1.542040095562568e-06, + "loss": 0.2537, + "step": 4605 + }, + { + "epoch": 1.5378964941569282, + "grad_norm": 0.45746277940064, + "learning_rate": 1.5399361234488154e-06, + "loss": 0.2434, + "step": 4606 + }, + { + "epoch": 1.5382303839732887, + "grad_norm": 0.47014747346479097, + "learning_rate": 1.5378333263548667e-06, + "loss": 0.2573, + "step": 4607 + }, + { + "epoch": 1.5385642737896494, + "grad_norm": 0.4649180609803252, + "learning_rate": 1.5357317049948194e-06, + "loss": 0.2602, + "step": 4608 + }, + { + "epoch": 1.5388981636060102, + "grad_norm": 0.4355648828915874, + "learning_rate": 1.5336312600823743e-06, + "loss": 0.2498, + "step": 4609 + }, + { + "epoch": 1.5392320534223707, + "grad_norm": 0.4633966622312195, + "learning_rate": 1.5315319923308291e-06, + "loss": 0.2532, + "step": 4610 + }, + { + "epoch": 1.5395659432387312, + "grad_norm": 0.5291998168279793, + "learning_rate": 1.5294339024530858e-06, + "loss": 0.243, + "step": 4611 + }, + { + "epoch": 1.5398998330550917, + "grad_norm": 0.4710007074777216, + "learning_rate": 1.5273369911616465e-06, + "loss": 0.2634, + "step": 4612 + }, + { + "epoch": 1.5402337228714524, + "grad_norm": 0.47391884769287157, + "learning_rate": 1.5252412591686105e-06, + "loss": 0.2491, + "step": 4613 + }, + { + "epoch": 1.5405676126878132, + "grad_norm": 0.4582976121086286, + "learning_rate": 1.523146707185677e-06, + "loss": 0.2641, + "step": 4614 + }, + { + "epoch": 1.5409015025041737, + "grad_norm": 0.43793525520284826, + "learning_rate": 1.5210533359241436e-06, + "loss": 0.2572, + "step": 4615 + }, + { + "epoch": 1.5412353923205342, + "grad_norm": 0.4468823848010383, + "learning_rate": 1.5189611460949112e-06, + "loss": 0.2452, + "step": 4616 + }, + { + "epoch": 1.5415692821368947, + "grad_norm": 0.4844012739945484, + "learning_rate": 1.5168701384084754e-06, + "loss": 0.2555, + "step": 4617 + }, + { + "epoch": 1.5419031719532554, + "grad_norm": 0.4328146160736993, + "learning_rate": 1.5147803135749318e-06, + "loss": 0.2314, + "step": 4618 + }, + { + "epoch": 1.5422370617696162, + "grad_norm": 0.45087801018026374, + "learning_rate": 1.512691672303972e-06, + "loss": 0.2424, + "step": 4619 + }, + { + "epoch": 1.5425709515859767, + "grad_norm": 0.4443721720572599, + "learning_rate": 1.5106042153048894e-06, + "loss": 0.2552, + "step": 4620 + }, + { + "epoch": 1.5429048414023372, + "grad_norm": 0.4617438186622842, + "learning_rate": 1.508517943286576e-06, + "loss": 0.2587, + "step": 4621 + }, + { + "epoch": 1.5432387312186977, + "grad_norm": 0.44881498676573695, + "learning_rate": 1.5064328569575166e-06, + "loss": 0.253, + "step": 4622 + }, + { + "epoch": 1.5435726210350584, + "grad_norm": 0.47039617634143893, + "learning_rate": 1.5043489570257957e-06, + "loss": 0.2533, + "step": 4623 + }, + { + "epoch": 1.5439065108514192, + "grad_norm": 0.44107000402358526, + "learning_rate": 1.5022662441990938e-06, + "loss": 0.2505, + "step": 4624 + }, + { + "epoch": 1.5442404006677797, + "grad_norm": 0.45798632467327044, + "learning_rate": 1.5001847191846924e-06, + "loss": 0.2555, + "step": 4625 + }, + { + "epoch": 1.5445742904841402, + "grad_norm": 0.4525311991666009, + "learning_rate": 1.4981043826894648e-06, + "loss": 0.243, + "step": 4626 + }, + { + "epoch": 1.5449081803005007, + "grad_norm": 0.4651019489869386, + "learning_rate": 1.4960252354198828e-06, + "loss": 0.2526, + "step": 4627 + }, + { + "epoch": 1.5452420701168614, + "grad_norm": 0.4476722869936872, + "learning_rate": 1.4939472780820125e-06, + "loss": 0.2457, + "step": 4628 + }, + { + "epoch": 1.5455759599332222, + "grad_norm": 0.4303668565253337, + "learning_rate": 1.491870511381519e-06, + "loss": 0.2361, + "step": 4629 + }, + { + "epoch": 1.5459098497495827, + "grad_norm": 0.46007536785120096, + "learning_rate": 1.4897949360236635e-06, + "loss": 0.2503, + "step": 4630 + }, + { + "epoch": 1.5462437395659432, + "grad_norm": 0.46121701523351366, + "learning_rate": 1.4877205527132983e-06, + "loss": 0.2502, + "step": 4631 + }, + { + "epoch": 1.5465776293823037, + "grad_norm": 0.46134638002272116, + "learning_rate": 1.4856473621548733e-06, + "loss": 0.2514, + "step": 4632 + }, + { + "epoch": 1.5469115191986644, + "grad_norm": 0.45357405121133376, + "learning_rate": 1.4835753650524326e-06, + "loss": 0.2485, + "step": 4633 + }, + { + "epoch": 1.5472454090150252, + "grad_norm": 0.45814707195788645, + "learning_rate": 1.481504562109618e-06, + "loss": 0.25, + "step": 4634 + }, + { + "epoch": 1.5475792988313857, + "grad_norm": 0.4630857876492156, + "learning_rate": 1.4794349540296622e-06, + "loss": 0.2536, + "step": 4635 + }, + { + "epoch": 1.5479131886477462, + "grad_norm": 0.45767126619754334, + "learning_rate": 1.4773665415153915e-06, + "loss": 0.2486, + "step": 4636 + }, + { + "epoch": 1.5482470784641067, + "grad_norm": 0.46475984471787396, + "learning_rate": 1.475299325269232e-06, + "loss": 0.2488, + "step": 4637 + }, + { + "epoch": 1.5485809682804674, + "grad_norm": 0.4396559267368927, + "learning_rate": 1.4732333059931958e-06, + "loss": 0.2425, + "step": 4638 + }, + { + "epoch": 1.5489148580968282, + "grad_norm": 0.43139072927838706, + "learning_rate": 1.4711684843888963e-06, + "loss": 0.2385, + "step": 4639 + }, + { + "epoch": 1.5492487479131887, + "grad_norm": 0.44777215092860123, + "learning_rate": 1.4691048611575337e-06, + "loss": 0.2484, + "step": 4640 + }, + { + "epoch": 1.5495826377295492, + "grad_norm": 0.4607466776458701, + "learning_rate": 1.4670424369999043e-06, + "loss": 0.2471, + "step": 4641 + }, + { + "epoch": 1.5499165275459097, + "grad_norm": 0.4798509669502947, + "learning_rate": 1.4649812126163964e-06, + "loss": 0.2474, + "step": 4642 + }, + { + "epoch": 1.5502504173622704, + "grad_norm": 0.45286571228968114, + "learning_rate": 1.46292118870699e-06, + "loss": 0.2536, + "step": 4643 + }, + { + "epoch": 1.5505843071786312, + "grad_norm": 0.43971080051632194, + "learning_rate": 1.4608623659712612e-06, + "loss": 0.2465, + "step": 4644 + }, + { + "epoch": 1.5509181969949917, + "grad_norm": 0.4652189026352351, + "learning_rate": 1.458804745108373e-06, + "loss": 0.2487, + "step": 4645 + }, + { + "epoch": 1.5512520868113522, + "grad_norm": 0.4683338661203451, + "learning_rate": 1.4567483268170856e-06, + "loss": 0.2597, + "step": 4646 + }, + { + "epoch": 1.5515859766277127, + "grad_norm": 0.45130873875106153, + "learning_rate": 1.4546931117957452e-06, + "loss": 0.2455, + "step": 4647 + }, + { + "epoch": 1.5519198664440734, + "grad_norm": 0.4809184487655275, + "learning_rate": 1.4526391007422946e-06, + "loss": 0.2582, + "step": 4648 + }, + { + "epoch": 1.5522537562604342, + "grad_norm": 0.4661229936672732, + "learning_rate": 1.4505862943542643e-06, + "loss": 0.2594, + "step": 4649 + }, + { + "epoch": 1.5525876460767947, + "grad_norm": 0.46476866797198646, + "learning_rate": 1.4485346933287768e-06, + "loss": 0.2479, + "step": 4650 + }, + { + "epoch": 1.5529215358931552, + "grad_norm": 0.4714373845926809, + "learning_rate": 1.446484298362545e-06, + "loss": 0.251, + "step": 4651 + }, + { + "epoch": 1.553255425709516, + "grad_norm": 0.4814742324200845, + "learning_rate": 1.444435110151871e-06, + "loss": 0.265, + "step": 4652 + }, + { + "epoch": 1.5535893155258764, + "grad_norm": 0.4533776414834825, + "learning_rate": 1.4423871293926501e-06, + "loss": 0.2541, + "step": 4653 + }, + { + "epoch": 1.5539232053422372, + "grad_norm": 0.48171604582851907, + "learning_rate": 1.440340356780367e-06, + "loss": 0.2594, + "step": 4654 + }, + { + "epoch": 1.5542570951585977, + "grad_norm": 0.43588543266291974, + "learning_rate": 1.4382947930100944e-06, + "loss": 0.2512, + "step": 4655 + }, + { + "epoch": 1.5545909849749582, + "grad_norm": 0.45359442240528486, + "learning_rate": 1.4362504387764947e-06, + "loss": 0.25, + "step": 4656 + }, + { + "epoch": 1.554924874791319, + "grad_norm": 0.4551560606614718, + "learning_rate": 1.4342072947738183e-06, + "loss": 0.2383, + "step": 4657 + }, + { + "epoch": 1.5552587646076794, + "grad_norm": 0.5317385903770661, + "learning_rate": 1.4321653616959096e-06, + "loss": 0.2444, + "step": 4658 + }, + { + "epoch": 1.5555926544240402, + "grad_norm": 0.4486665634838382, + "learning_rate": 1.430124640236198e-06, + "loss": 0.253, + "step": 4659 + }, + { + "epoch": 1.5559265442404007, + "grad_norm": 0.4683437103760997, + "learning_rate": 1.4280851310876998e-06, + "loss": 0.2469, + "step": 4660 + }, + { + "epoch": 1.5562604340567612, + "grad_norm": 0.44250680199858566, + "learning_rate": 1.4260468349430224e-06, + "loss": 0.244, + "step": 4661 + }, + { + "epoch": 1.556594323873122, + "grad_norm": 0.4536050096242991, + "learning_rate": 1.4240097524943602e-06, + "loss": 0.2436, + "step": 4662 + }, + { + "epoch": 1.5569282136894824, + "grad_norm": 0.4578205497324621, + "learning_rate": 1.421973884433499e-06, + "loss": 0.2536, + "step": 4663 + }, + { + "epoch": 1.5572621035058432, + "grad_norm": 0.4509590324193091, + "learning_rate": 1.4199392314518068e-06, + "loss": 0.2432, + "step": 4664 + }, + { + "epoch": 1.5575959933222037, + "grad_norm": 0.45794663321547124, + "learning_rate": 1.417905794240242e-06, + "loss": 0.2401, + "step": 4665 + }, + { + "epoch": 1.5579298831385642, + "grad_norm": 0.43967928807661294, + "learning_rate": 1.4158735734893464e-06, + "loss": 0.2357, + "step": 4666 + }, + { + "epoch": 1.558263772954925, + "grad_norm": 0.45162638609869377, + "learning_rate": 1.4138425698892555e-06, + "loss": 0.2521, + "step": 4667 + }, + { + "epoch": 1.5585976627712856, + "grad_norm": 0.4566642579949405, + "learning_rate": 1.4118127841296864e-06, + "loss": 0.2565, + "step": 4668 + }, + { + "epoch": 1.5589315525876462, + "grad_norm": 0.4784726338702912, + "learning_rate": 1.4097842168999437e-06, + "loss": 0.2481, + "step": 4669 + }, + { + "epoch": 1.5592654424040067, + "grad_norm": 0.46374066512684753, + "learning_rate": 1.4077568688889165e-06, + "loss": 0.2555, + "step": 4670 + }, + { + "epoch": 1.5595993322203672, + "grad_norm": 0.46129074780833856, + "learning_rate": 1.4057307407850834e-06, + "loss": 0.2634, + "step": 4671 + }, + { + "epoch": 1.559933222036728, + "grad_norm": 0.4544590740430966, + "learning_rate": 1.4037058332765085e-06, + "loss": 0.248, + "step": 4672 + }, + { + "epoch": 1.5602671118530886, + "grad_norm": 0.4389000064351568, + "learning_rate": 1.4016821470508385e-06, + "loss": 0.2448, + "step": 4673 + }, + { + "epoch": 1.5606010016694492, + "grad_norm": 0.449521146213102, + "learning_rate": 1.3996596827953063e-06, + "loss": 0.2435, + "step": 4674 + }, + { + "epoch": 1.5609348914858097, + "grad_norm": 0.4668710058142303, + "learning_rate": 1.3976384411967297e-06, + "loss": 0.2586, + "step": 4675 + }, + { + "epoch": 1.5612687813021702, + "grad_norm": 0.4463063495432483, + "learning_rate": 1.3956184229415148e-06, + "loss": 0.2525, + "step": 4676 + }, + { + "epoch": 1.561602671118531, + "grad_norm": 0.47049566778663415, + "learning_rate": 1.3935996287156467e-06, + "loss": 0.2604, + "step": 4677 + }, + { + "epoch": 1.5619365609348916, + "grad_norm": 0.45332908484221596, + "learning_rate": 1.3915820592046986e-06, + "loss": 0.2442, + "step": 4678 + }, + { + "epoch": 1.5622704507512521, + "grad_norm": 0.4497864125000331, + "learning_rate": 1.3895657150938252e-06, + "loss": 0.2452, + "step": 4679 + }, + { + "epoch": 1.5626043405676127, + "grad_norm": 0.45096570305424827, + "learning_rate": 1.3875505970677672e-06, + "loss": 0.2447, + "step": 4680 + }, + { + "epoch": 1.5629382303839732, + "grad_norm": 0.4598990025236315, + "learning_rate": 1.3855367058108504e-06, + "loss": 0.2523, + "step": 4681 + }, + { + "epoch": 1.563272120200334, + "grad_norm": 0.44860747865115147, + "learning_rate": 1.3835240420069801e-06, + "loss": 0.2545, + "step": 4682 + }, + { + "epoch": 1.5636060100166946, + "grad_norm": 0.4434952321274702, + "learning_rate": 1.3815126063396467e-06, + "loss": 0.2491, + "step": 4683 + }, + { + "epoch": 1.5639398998330551, + "grad_norm": 0.4642770017993263, + "learning_rate": 1.379502399491922e-06, + "loss": 0.2543, + "step": 4684 + }, + { + "epoch": 1.5642737896494157, + "grad_norm": 0.45610353682759003, + "learning_rate": 1.3774934221464643e-06, + "loss": 0.2478, + "step": 4685 + }, + { + "epoch": 1.5646076794657762, + "grad_norm": 0.463061315319157, + "learning_rate": 1.3754856749855105e-06, + "loss": 0.2522, + "step": 4686 + }, + { + "epoch": 1.564941569282137, + "grad_norm": 0.44870586473113594, + "learning_rate": 1.3734791586908808e-06, + "loss": 0.2377, + "step": 4687 + }, + { + "epoch": 1.5652754590984976, + "grad_norm": 0.4523132763871824, + "learning_rate": 1.3714738739439792e-06, + "loss": 0.2504, + "step": 4688 + }, + { + "epoch": 1.5656093489148581, + "grad_norm": 0.4489102280393724, + "learning_rate": 1.3694698214257879e-06, + "loss": 0.242, + "step": 4689 + }, + { + "epoch": 1.5659432387312187, + "grad_norm": 0.4494955375093671, + "learning_rate": 1.367467001816875e-06, + "loss": 0.2506, + "step": 4690 + }, + { + "epoch": 1.5662771285475792, + "grad_norm": 0.4594175406296867, + "learning_rate": 1.3654654157973874e-06, + "loss": 0.2532, + "step": 4691 + }, + { + "epoch": 1.56661101836394, + "grad_norm": 0.4450655043415581, + "learning_rate": 1.3634650640470525e-06, + "loss": 0.2438, + "step": 4692 + }, + { + "epoch": 1.5669449081803006, + "grad_norm": 0.48444768849422204, + "learning_rate": 1.361465947245178e-06, + "loss": 0.2497, + "step": 4693 + }, + { + "epoch": 1.5672787979966611, + "grad_norm": 0.4742040438439686, + "learning_rate": 1.3594680660706572e-06, + "loss": 0.2516, + "step": 4694 + }, + { + "epoch": 1.5676126878130217, + "grad_norm": 0.4428574759060811, + "learning_rate": 1.357471421201958e-06, + "loss": 0.246, + "step": 4695 + }, + { + "epoch": 1.5679465776293822, + "grad_norm": 0.4721198041836463, + "learning_rate": 1.355476013317129e-06, + "loss": 0.2628, + "step": 4696 + }, + { + "epoch": 1.568280467445743, + "grad_norm": 0.4445617015792463, + "learning_rate": 1.3534818430938046e-06, + "loss": 0.2401, + "step": 4697 + }, + { + "epoch": 1.5686143572621036, + "grad_norm": 0.4650689664539967, + "learning_rate": 1.3514889112091905e-06, + "loss": 0.2541, + "step": 4698 + }, + { + "epoch": 1.5689482470784641, + "grad_norm": 0.4666800492420589, + "learning_rate": 1.3494972183400795e-06, + "loss": 0.2402, + "step": 4699 + }, + { + "epoch": 1.5692821368948247, + "grad_norm": 0.4581826824342904, + "learning_rate": 1.3475067651628388e-06, + "loss": 0.2619, + "step": 4700 + }, + { + "epoch": 1.5696160267111852, + "grad_norm": 0.45691364924803435, + "learning_rate": 1.3455175523534158e-06, + "loss": 0.2505, + "step": 4701 + }, + { + "epoch": 1.569949916527546, + "grad_norm": 0.47916547330276094, + "learning_rate": 1.3435295805873361e-06, + "loss": 0.2521, + "step": 4702 + }, + { + "epoch": 1.5702838063439066, + "grad_norm": 0.44964637474795277, + "learning_rate": 1.341542850539706e-06, + "loss": 0.2562, + "step": 4703 + }, + { + "epoch": 1.5706176961602671, + "grad_norm": 0.4284830949224782, + "learning_rate": 1.3395573628852076e-06, + "loss": 0.2444, + "step": 4704 + }, + { + "epoch": 1.5709515859766277, + "grad_norm": 0.4648273196412473, + "learning_rate": 1.337573118298104e-06, + "loss": 0.249, + "step": 4705 + }, + { + "epoch": 1.5712854757929882, + "grad_norm": 0.46445909153630544, + "learning_rate": 1.3355901174522334e-06, + "loss": 0.2524, + "step": 4706 + }, + { + "epoch": 1.571619365609349, + "grad_norm": 0.4547400024055269, + "learning_rate": 1.3336083610210115e-06, + "loss": 0.2498, + "step": 4707 + }, + { + "epoch": 1.5719532554257096, + "grad_norm": 0.4574201014327143, + "learning_rate": 1.331627849677435e-06, + "loss": 0.2414, + "step": 4708 + }, + { + "epoch": 1.5722871452420701, + "grad_norm": 0.4525324547587566, + "learning_rate": 1.3296485840940732e-06, + "loss": 0.2449, + "step": 4709 + }, + { + "epoch": 1.5726210350584306, + "grad_norm": 0.44539727481174185, + "learning_rate": 1.3276705649430754e-06, + "loss": 0.2517, + "step": 4710 + }, + { + "epoch": 1.5729549248747912, + "grad_norm": 0.4622232065733278, + "learning_rate": 1.3256937928961655e-06, + "loss": 0.2504, + "step": 4711 + }, + { + "epoch": 1.573288814691152, + "grad_norm": 0.4456006966881543, + "learning_rate": 1.323718268624647e-06, + "loss": 0.2583, + "step": 4712 + }, + { + "epoch": 1.5736227045075126, + "grad_norm": 0.45817754357186957, + "learning_rate": 1.3217439927993957e-06, + "loss": 0.2476, + "step": 4713 + }, + { + "epoch": 1.5739565943238731, + "grad_norm": 0.44744782800480076, + "learning_rate": 1.3197709660908676e-06, + "loss": 0.2494, + "step": 4714 + }, + { + "epoch": 1.5742904841402336, + "grad_norm": 0.4676803588339746, + "learning_rate": 1.3177991891690917e-06, + "loss": 0.2534, + "step": 4715 + }, + { + "epoch": 1.5746243739565942, + "grad_norm": 0.43450646730949316, + "learning_rate": 1.3158286627036709e-06, + "loss": 0.2424, + "step": 4716 + }, + { + "epoch": 1.574958263772955, + "grad_norm": 0.441988503986452, + "learning_rate": 1.3138593873637894e-06, + "loss": 0.2381, + "step": 4717 + }, + { + "epoch": 1.5752921535893156, + "grad_norm": 0.44357592119288897, + "learning_rate": 1.3118913638182012e-06, + "loss": 0.2487, + "step": 4718 + }, + { + "epoch": 1.5756260434056761, + "grad_norm": 0.4254437089922603, + "learning_rate": 1.3099245927352367e-06, + "loss": 0.2281, + "step": 4719 + }, + { + "epoch": 1.5759599332220366, + "grad_norm": 0.4430836852642295, + "learning_rate": 1.3079590747828003e-06, + "loss": 0.246, + "step": 4720 + }, + { + "epoch": 1.5762938230383974, + "grad_norm": 0.4556791442136155, + "learning_rate": 1.3059948106283725e-06, + "loss": 0.2557, + "step": 4721 + }, + { + "epoch": 1.576627712854758, + "grad_norm": 0.4533000153207651, + "learning_rate": 1.3040318009390096e-06, + "loss": 0.2451, + "step": 4722 + }, + { + "epoch": 1.5769616026711186, + "grad_norm": 0.4530242959669288, + "learning_rate": 1.3020700463813374e-06, + "loss": 0.2576, + "step": 4723 + }, + { + "epoch": 1.5772954924874791, + "grad_norm": 0.4576366421086781, + "learning_rate": 1.3001095476215581e-06, + "loss": 0.2489, + "step": 4724 + }, + { + "epoch": 1.5776293823038396, + "grad_norm": 0.45020519717148594, + "learning_rate": 1.2981503053254446e-06, + "loss": 0.2585, + "step": 4725 + }, + { + "epoch": 1.5779632721202004, + "grad_norm": 0.4404391409556741, + "learning_rate": 1.2961923201583493e-06, + "loss": 0.2487, + "step": 4726 + }, + { + "epoch": 1.5782971619365609, + "grad_norm": 0.4343048327374323, + "learning_rate": 1.2942355927851919e-06, + "loss": 0.2458, + "step": 4727 + }, + { + "epoch": 1.5786310517529216, + "grad_norm": 0.4491009136348061, + "learning_rate": 1.2922801238704675e-06, + "loss": 0.2564, + "step": 4728 + }, + { + "epoch": 1.5789649415692821, + "grad_norm": 0.4425497453079414, + "learning_rate": 1.2903259140782404e-06, + "loss": 0.2409, + "step": 4729 + }, + { + "epoch": 1.5792988313856426, + "grad_norm": 0.45214955538660995, + "learning_rate": 1.2883729640721532e-06, + "loss": 0.2484, + "step": 4730 + }, + { + "epoch": 1.5796327212020034, + "grad_norm": 0.43577683264444195, + "learning_rate": 1.286421274515418e-06, + "loss": 0.2437, + "step": 4731 + }, + { + "epoch": 1.5799666110183639, + "grad_norm": 0.4437354743949751, + "learning_rate": 1.2844708460708177e-06, + "loss": 0.2465, + "step": 4732 + }, + { + "epoch": 1.5803005008347246, + "grad_norm": 0.4629879156360492, + "learning_rate": 1.2825216794007073e-06, + "loss": 0.2537, + "step": 4733 + }, + { + "epoch": 1.5806343906510851, + "grad_norm": 0.46496557344261147, + "learning_rate": 1.2805737751670122e-06, + "loss": 0.2595, + "step": 4734 + }, + { + "epoch": 1.5809682804674456, + "grad_norm": 0.4409642444506756, + "learning_rate": 1.278627134031234e-06, + "loss": 0.2505, + "step": 4735 + }, + { + "epoch": 1.5813021702838064, + "grad_norm": 0.4439941172545017, + "learning_rate": 1.2766817566544398e-06, + "loss": 0.2437, + "step": 4736 + }, + { + "epoch": 1.581636060100167, + "grad_norm": 0.4538097259905716, + "learning_rate": 1.2747376436972702e-06, + "loss": 0.2504, + "step": 4737 + }, + { + "epoch": 1.5819699499165276, + "grad_norm": 0.4421467510710132, + "learning_rate": 1.2727947958199349e-06, + "loss": 0.2461, + "step": 4738 + }, + { + "epoch": 1.5823038397328881, + "grad_norm": 0.44909633529089776, + "learning_rate": 1.2708532136822156e-06, + "loss": 0.249, + "step": 4739 + }, + { + "epoch": 1.5826377295492486, + "grad_norm": 0.4523828805982266, + "learning_rate": 1.268912897943465e-06, + "loss": 0.24, + "step": 4740 + }, + { + "epoch": 1.5829716193656094, + "grad_norm": 0.4567978031715271, + "learning_rate": 1.2669738492626033e-06, + "loss": 0.2513, + "step": 4741 + }, + { + "epoch": 1.58330550918197, + "grad_norm": 0.452742379281957, + "learning_rate": 1.2650360682981216e-06, + "loss": 0.2507, + "step": 4742 + }, + { + "epoch": 1.5836393989983306, + "grad_norm": 0.4738419483253126, + "learning_rate": 1.2630995557080783e-06, + "loss": 0.2648, + "step": 4743 + }, + { + "epoch": 1.5839732888146911, + "grad_norm": 0.4682766433803461, + "learning_rate": 1.2611643121501061e-06, + "loss": 0.2525, + "step": 4744 + }, + { + "epoch": 1.5843071786310516, + "grad_norm": 0.4463003602107445, + "learning_rate": 1.2592303382814024e-06, + "loss": 0.2432, + "step": 4745 + }, + { + "epoch": 1.5846410684474124, + "grad_norm": 0.4532841278150243, + "learning_rate": 1.2572976347587345e-06, + "loss": 0.2479, + "step": 4746 + }, + { + "epoch": 1.584974958263773, + "grad_norm": 0.45051052022983257, + "learning_rate": 1.2553662022384377e-06, + "loss": 0.2458, + "step": 4747 + }, + { + "epoch": 1.5853088480801336, + "grad_norm": 0.46414018666363177, + "learning_rate": 1.253436041376417e-06, + "loss": 0.2484, + "step": 4748 + }, + { + "epoch": 1.5856427378964941, + "grad_norm": 0.4607700026810279, + "learning_rate": 1.2515071528281474e-06, + "loss": 0.2522, + "step": 4749 + }, + { + "epoch": 1.5859766277128546, + "grad_norm": 0.4570769591119826, + "learning_rate": 1.2495795372486673e-06, + "loss": 0.2558, + "step": 4750 + }, + { + "epoch": 1.5863105175292154, + "grad_norm": 0.4906020002723191, + "learning_rate": 1.2476531952925852e-06, + "loss": 0.2553, + "step": 4751 + }, + { + "epoch": 1.586644407345576, + "grad_norm": 0.45141253026749967, + "learning_rate": 1.2457281276140759e-06, + "loss": 0.2515, + "step": 4752 + }, + { + "epoch": 1.5869782971619366, + "grad_norm": 0.45071595898911837, + "learning_rate": 1.243804334866885e-06, + "loss": 0.2536, + "step": 4753 + }, + { + "epoch": 1.5873121869782971, + "grad_norm": 0.45436469990066125, + "learning_rate": 1.2418818177043207e-06, + "loss": 0.2447, + "step": 4754 + }, + { + "epoch": 1.5876460767946576, + "grad_norm": 0.4562150665706285, + "learning_rate": 1.2399605767792588e-06, + "loss": 0.2489, + "step": 4755 + }, + { + "epoch": 1.5879799666110184, + "grad_norm": 0.46127543905363644, + "learning_rate": 1.2380406127441457e-06, + "loss": 0.261, + "step": 4756 + }, + { + "epoch": 1.588313856427379, + "grad_norm": 0.4434867116451677, + "learning_rate": 1.2361219262509882e-06, + "loss": 0.2425, + "step": 4757 + }, + { + "epoch": 1.5886477462437396, + "grad_norm": 0.4359318298644504, + "learning_rate": 1.2342045179513655e-06, + "loss": 0.2465, + "step": 4758 + }, + { + "epoch": 1.5889816360601001, + "grad_norm": 0.4380302502399583, + "learning_rate": 1.2322883884964181e-06, + "loss": 0.245, + "step": 4759 + }, + { + "epoch": 1.5893155258764606, + "grad_norm": 0.45213785295307984, + "learning_rate": 1.2303735385368538e-06, + "loss": 0.2468, + "step": 4760 + }, + { + "epoch": 1.5896494156928214, + "grad_norm": 0.45352999597053517, + "learning_rate": 1.228459968722943e-06, + "loss": 0.2569, + "step": 4761 + }, + { + "epoch": 1.589983305509182, + "grad_norm": 0.459805304021665, + "learning_rate": 1.2265476797045288e-06, + "loss": 0.2506, + "step": 4762 + }, + { + "epoch": 1.5903171953255426, + "grad_norm": 0.44999763168960494, + "learning_rate": 1.2246366721310126e-06, + "loss": 0.2395, + "step": 4763 + }, + { + "epoch": 1.5906510851419031, + "grad_norm": 0.4586064444479258, + "learning_rate": 1.2227269466513608e-06, + "loss": 0.241, + "step": 4764 + }, + { + "epoch": 1.5909849749582636, + "grad_norm": 0.4488475744758149, + "learning_rate": 1.2208185039141096e-06, + "loss": 0.2477, + "step": 4765 + }, + { + "epoch": 1.5913188647746244, + "grad_norm": 0.4559143345893874, + "learning_rate": 1.2189113445673528e-06, + "loss": 0.2479, + "step": 4766 + }, + { + "epoch": 1.591652754590985, + "grad_norm": 0.43962346441793526, + "learning_rate": 1.2170054692587556e-06, + "loss": 0.2432, + "step": 4767 + }, + { + "epoch": 1.5919866444073456, + "grad_norm": 0.46608567630331316, + "learning_rate": 1.2151008786355411e-06, + "loss": 0.2536, + "step": 4768 + }, + { + "epoch": 1.5923205342237061, + "grad_norm": 0.4513907551588063, + "learning_rate": 1.2131975733444996e-06, + "loss": 0.2422, + "step": 4769 + }, + { + "epoch": 1.5926544240400666, + "grad_norm": 0.45053702850821914, + "learning_rate": 1.211295554031981e-06, + "loss": 0.2462, + "step": 4770 + }, + { + "epoch": 1.5929883138564274, + "grad_norm": 0.4555752986289785, + "learning_rate": 1.2093948213439049e-06, + "loss": 0.2513, + "step": 4771 + }, + { + "epoch": 1.593322203672788, + "grad_norm": 0.4319728120471387, + "learning_rate": 1.2074953759257474e-06, + "loss": 0.2276, + "step": 4772 + }, + { + "epoch": 1.5936560934891486, + "grad_norm": 0.45855560932820666, + "learning_rate": 1.2055972184225528e-06, + "loss": 0.2481, + "step": 4773 + }, + { + "epoch": 1.5939899833055091, + "grad_norm": 0.47538508897322945, + "learning_rate": 1.2037003494789246e-06, + "loss": 0.2518, + "step": 4774 + }, + { + "epoch": 1.5943238731218696, + "grad_norm": 0.4571243199201818, + "learning_rate": 1.201804769739028e-06, + "loss": 0.251, + "step": 4775 + }, + { + "epoch": 1.5946577629382304, + "grad_norm": 0.4535942339982754, + "learning_rate": 1.1999104798465943e-06, + "loss": 0.234, + "step": 4776 + }, + { + "epoch": 1.594991652754591, + "grad_norm": 0.4575231269700912, + "learning_rate": 1.1980174804449145e-06, + "loss": 0.2597, + "step": 4777 + }, + { + "epoch": 1.5953255425709516, + "grad_norm": 0.4345663507739573, + "learning_rate": 1.1961257721768405e-06, + "loss": 0.2396, + "step": 4778 + }, + { + "epoch": 1.5956594323873121, + "grad_norm": 0.44887474792367293, + "learning_rate": 1.1942353556847852e-06, + "loss": 0.2429, + "step": 4779 + }, + { + "epoch": 1.5959933222036726, + "grad_norm": 0.4619485471624482, + "learning_rate": 1.1923462316107276e-06, + "loss": 0.2479, + "step": 4780 + }, + { + "epoch": 1.5963272120200334, + "grad_norm": 0.4645820933988809, + "learning_rate": 1.190458400596201e-06, + "loss": 0.2535, + "step": 4781 + }, + { + "epoch": 1.596661101836394, + "grad_norm": 0.4405934139784985, + "learning_rate": 1.1885718632823068e-06, + "loss": 0.2442, + "step": 4782 + }, + { + "epoch": 1.5969949916527546, + "grad_norm": 0.45116885934561446, + "learning_rate": 1.186686620309701e-06, + "loss": 0.2385, + "step": 4783 + }, + { + "epoch": 1.5973288814691151, + "grad_norm": 0.44894062552403213, + "learning_rate": 1.1848026723186013e-06, + "loss": 0.2414, + "step": 4784 + }, + { + "epoch": 1.5976627712854758, + "grad_norm": 0.43993917318633013, + "learning_rate": 1.1829200199487894e-06, + "loss": 0.2447, + "step": 4785 + }, + { + "epoch": 1.5979966611018364, + "grad_norm": 0.44889373943139776, + "learning_rate": 1.1810386638396032e-06, + "loss": 0.2469, + "step": 4786 + }, + { + "epoch": 1.598330550918197, + "grad_norm": 0.4523652709425815, + "learning_rate": 1.1791586046299414e-06, + "loss": 0.2513, + "step": 4787 + }, + { + "epoch": 1.5986644407345576, + "grad_norm": 0.4520382987793894, + "learning_rate": 1.177279842958261e-06, + "loss": 0.2525, + "step": 4788 + }, + { + "epoch": 1.5989983305509181, + "grad_norm": 0.44144098147669775, + "learning_rate": 1.1754023794625806e-06, + "loss": 0.2426, + "step": 4789 + }, + { + "epoch": 1.5993322203672788, + "grad_norm": 0.4358415506163421, + "learning_rate": 1.1735262147804793e-06, + "loss": 0.2377, + "step": 4790 + }, + { + "epoch": 1.5996661101836394, + "grad_norm": 0.4722837609114571, + "learning_rate": 1.1716513495490912e-06, + "loss": 0.2538, + "step": 4791 + }, + { + "epoch": 1.6, + "grad_norm": 0.4994505636826688, + "learning_rate": 1.1697777844051105e-06, + "loss": 0.2585, + "step": 4792 + }, + { + "epoch": 1.6003338898163606, + "grad_norm": 0.44886503208461964, + "learning_rate": 1.1679055199847894e-06, + "loss": 0.2383, + "step": 4793 + }, + { + "epoch": 1.600667779632721, + "grad_norm": 0.4338868534378518, + "learning_rate": 1.1660345569239418e-06, + "loss": 0.2484, + "step": 4794 + }, + { + "epoch": 1.6010016694490818, + "grad_norm": 0.4571473989707943, + "learning_rate": 1.1641648958579354e-06, + "loss": 0.2501, + "step": 4795 + }, + { + "epoch": 1.6013355592654424, + "grad_norm": 0.4451545336461222, + "learning_rate": 1.1622965374216976e-06, + "loss": 0.2536, + "step": 4796 + }, + { + "epoch": 1.601669449081803, + "grad_norm": 0.4394711265712638, + "learning_rate": 1.1604294822497125e-06, + "loss": 0.2534, + "step": 4797 + }, + { + "epoch": 1.6020033388981636, + "grad_norm": 0.4662562335502728, + "learning_rate": 1.1585637309760229e-06, + "loss": 0.2594, + "step": 4798 + }, + { + "epoch": 1.602337228714524, + "grad_norm": 0.44070728521991864, + "learning_rate": 1.15669928423423e-06, + "loss": 0.2481, + "step": 4799 + }, + { + "epoch": 1.6026711185308848, + "grad_norm": 0.4511441965163916, + "learning_rate": 1.154836142657489e-06, + "loss": 0.246, + "step": 4800 + }, + { + "epoch": 1.6030050083472456, + "grad_norm": 0.45454165872402746, + "learning_rate": 1.1529743068785131e-06, + "loss": 0.2473, + "step": 4801 + }, + { + "epoch": 1.603338898163606, + "grad_norm": 0.47296890133290037, + "learning_rate": 1.1511137775295705e-06, + "loss": 0.2514, + "step": 4802 + }, + { + "epoch": 1.6036727879799666, + "grad_norm": 0.44400701352770044, + "learning_rate": 1.14925455524249e-06, + "loss": 0.2403, + "step": 4803 + }, + { + "epoch": 1.604006677796327, + "grad_norm": 0.47810458337017414, + "learning_rate": 1.147396640648653e-06, + "loss": 0.2677, + "step": 4804 + }, + { + "epoch": 1.6043405676126878, + "grad_norm": 0.46969565683444947, + "learning_rate": 1.145540034378998e-06, + "loss": 0.2575, + "step": 4805 + }, + { + "epoch": 1.6046744574290486, + "grad_norm": 0.45309902094205345, + "learning_rate": 1.1436847370640164e-06, + "loss": 0.2417, + "step": 4806 + }, + { + "epoch": 1.605008347245409, + "grad_norm": 0.4729675913621866, + "learning_rate": 1.1418307493337594e-06, + "loss": 0.2612, + "step": 4807 + }, + { + "epoch": 1.6053422370617696, + "grad_norm": 0.4456198362491413, + "learning_rate": 1.1399780718178332e-06, + "loss": 0.2597, + "step": 4808 + }, + { + "epoch": 1.60567612687813, + "grad_norm": 0.4511311378935977, + "learning_rate": 1.1381267051453965e-06, + "loss": 0.2498, + "step": 4809 + }, + { + "epoch": 1.6060100166944908, + "grad_norm": 0.42562034341469696, + "learning_rate": 1.1362766499451633e-06, + "loss": 0.2425, + "step": 4810 + }, + { + "epoch": 1.6063439065108516, + "grad_norm": 0.44666736062429196, + "learning_rate": 1.1344279068454012e-06, + "loss": 0.2603, + "step": 4811 + }, + { + "epoch": 1.606677796327212, + "grad_norm": 0.46650093977041696, + "learning_rate": 1.1325804764739372e-06, + "loss": 0.254, + "step": 4812 + }, + { + "epoch": 1.6070116861435726, + "grad_norm": 0.4548223685047317, + "learning_rate": 1.130734359458147e-06, + "loss": 0.2545, + "step": 4813 + }, + { + "epoch": 1.607345575959933, + "grad_norm": 0.4455446812722191, + "learning_rate": 1.128889556424963e-06, + "loss": 0.2542, + "step": 4814 + }, + { + "epoch": 1.6076794657762938, + "grad_norm": 0.4459763767817864, + "learning_rate": 1.1270460680008693e-06, + "loss": 0.2594, + "step": 4815 + }, + { + "epoch": 1.6080133555926546, + "grad_norm": 0.4576505537237573, + "learning_rate": 1.1252038948119065e-06, + "loss": 0.2426, + "step": 4816 + }, + { + "epoch": 1.608347245409015, + "grad_norm": 0.4519679986127231, + "learning_rate": 1.1233630374836679e-06, + "loss": 0.2506, + "step": 4817 + }, + { + "epoch": 1.6086811352253756, + "grad_norm": 0.42416392396395564, + "learning_rate": 1.121523496641298e-06, + "loss": 0.2374, + "step": 4818 + }, + { + "epoch": 1.609015025041736, + "grad_norm": 0.450308590492201, + "learning_rate": 1.119685272909496e-06, + "loss": 0.2404, + "step": 4819 + }, + { + "epoch": 1.6093489148580968, + "grad_norm": 0.4453628125960755, + "learning_rate": 1.1178483669125112e-06, + "loss": 0.2392, + "step": 4820 + }, + { + "epoch": 1.6096828046744576, + "grad_norm": 0.44188650831613224, + "learning_rate": 1.11601277927415e-06, + "loss": 0.2533, + "step": 4821 + }, + { + "epoch": 1.610016694490818, + "grad_norm": 0.4547812654897812, + "learning_rate": 1.1141785106177672e-06, + "loss": 0.2506, + "step": 4822 + }, + { + "epoch": 1.6103505843071786, + "grad_norm": 0.46137922299608264, + "learning_rate": 1.1123455615662698e-06, + "loss": 0.2447, + "step": 4823 + }, + { + "epoch": 1.610684474123539, + "grad_norm": 0.4597646578464569, + "learning_rate": 1.1105139327421204e-06, + "loss": 0.2471, + "step": 4824 + }, + { + "epoch": 1.6110183639398998, + "grad_norm": 0.4707387644366269, + "learning_rate": 1.1086836247673282e-06, + "loss": 0.245, + "step": 4825 + }, + { + "epoch": 1.6113522537562606, + "grad_norm": 0.4564857703626363, + "learning_rate": 1.1068546382634587e-06, + "loss": 0.2454, + "step": 4826 + }, + { + "epoch": 1.611686143572621, + "grad_norm": 0.4556024282195579, + "learning_rate": 1.105026973851625e-06, + "loss": 0.2442, + "step": 4827 + }, + { + "epoch": 1.6120200333889816, + "grad_norm": 0.4389173420576472, + "learning_rate": 1.1032006321524929e-06, + "loss": 0.2439, + "step": 4828 + }, + { + "epoch": 1.612353923205342, + "grad_norm": 0.45655175959430966, + "learning_rate": 1.101375613786278e-06, + "loss": 0.2487, + "step": 4829 + }, + { + "epoch": 1.6126878130217028, + "grad_norm": 0.4603421842926232, + "learning_rate": 1.0995519193727462e-06, + "loss": 0.2469, + "step": 4830 + }, + { + "epoch": 1.6130217028380636, + "grad_norm": 0.45737707770754327, + "learning_rate": 1.0977295495312167e-06, + "loss": 0.2497, + "step": 4831 + }, + { + "epoch": 1.613355592654424, + "grad_norm": 0.44160597502742693, + "learning_rate": 1.0959085048805546e-06, + "loss": 0.2515, + "step": 4832 + }, + { + "epoch": 1.6136894824707846, + "grad_norm": 0.44050964982188456, + "learning_rate": 1.0940887860391802e-06, + "loss": 0.2457, + "step": 4833 + }, + { + "epoch": 1.614023372287145, + "grad_norm": 0.466632473971184, + "learning_rate": 1.0922703936250578e-06, + "loss": 0.2517, + "step": 4834 + }, + { + "epoch": 1.6143572621035058, + "grad_norm": 0.44561703241711315, + "learning_rate": 1.0904533282557066e-06, + "loss": 0.246, + "step": 4835 + }, + { + "epoch": 1.6146911519198666, + "grad_norm": 0.46542434655058074, + "learning_rate": 1.0886375905481917e-06, + "loss": 0.2472, + "step": 4836 + }, + { + "epoch": 1.615025041736227, + "grad_norm": 0.43629587163524247, + "learning_rate": 1.0868231811191276e-06, + "loss": 0.2455, + "step": 4837 + }, + { + "epoch": 1.6153589315525876, + "grad_norm": 0.44210989679871127, + "learning_rate": 1.0850101005846787e-06, + "loss": 0.247, + "step": 4838 + }, + { + "epoch": 1.615692821368948, + "grad_norm": 0.4700352526368341, + "learning_rate": 1.0831983495605557e-06, + "loss": 0.2538, + "step": 4839 + }, + { + "epoch": 1.6160267111853088, + "grad_norm": 0.4405337198380049, + "learning_rate": 1.0813879286620227e-06, + "loss": 0.2434, + "step": 4840 + }, + { + "epoch": 1.6163606010016696, + "grad_norm": 0.44892130222623405, + "learning_rate": 1.0795788385038892e-06, + "loss": 0.2445, + "step": 4841 + }, + { + "epoch": 1.61669449081803, + "grad_norm": 0.45467825384928046, + "learning_rate": 1.0777710797005124e-06, + "loss": 0.2503, + "step": 4842 + }, + { + "epoch": 1.6170283806343906, + "grad_norm": 0.4256878384793147, + "learning_rate": 1.075964652865798e-06, + "loss": 0.237, + "step": 4843 + }, + { + "epoch": 1.617362270450751, + "grad_norm": 0.450980207528074, + "learning_rate": 1.074159558613197e-06, + "loss": 0.2441, + "step": 4844 + }, + { + "epoch": 1.6176961602671118, + "grad_norm": 0.44109222718338625, + "learning_rate": 1.072355797555713e-06, + "loss": 0.2504, + "step": 4845 + }, + { + "epoch": 1.6180300500834726, + "grad_norm": 0.44224237334998534, + "learning_rate": 1.0705533703058928e-06, + "loss": 0.2456, + "step": 4846 + }, + { + "epoch": 1.618363939899833, + "grad_norm": 0.4563699001376963, + "learning_rate": 1.0687522774758318e-06, + "loss": 0.2442, + "step": 4847 + }, + { + "epoch": 1.6186978297161936, + "grad_norm": 0.463281169129271, + "learning_rate": 1.0669525196771695e-06, + "loss": 0.2491, + "step": 4848 + }, + { + "epoch": 1.619031719532554, + "grad_norm": 0.4534260777057201, + "learning_rate": 1.0651540975210966e-06, + "loss": 0.2517, + "step": 4849 + }, + { + "epoch": 1.6193656093489148, + "grad_norm": 0.47486358562827524, + "learning_rate": 1.063357011618349e-06, + "loss": 0.2553, + "step": 4850 + }, + { + "epoch": 1.6196994991652756, + "grad_norm": 0.4668378649645729, + "learning_rate": 1.061561262579206e-06, + "loss": 0.246, + "step": 4851 + }, + { + "epoch": 1.620033388981636, + "grad_norm": 0.4732743049765437, + "learning_rate": 1.059766851013495e-06, + "loss": 0.2503, + "step": 4852 + }, + { + "epoch": 1.6203672787979966, + "grad_norm": 0.45297433611468896, + "learning_rate": 1.0579737775305881e-06, + "loss": 0.2532, + "step": 4853 + }, + { + "epoch": 1.6207011686143573, + "grad_norm": 0.4429360601901967, + "learning_rate": 1.0561820427394054e-06, + "loss": 0.2541, + "step": 4854 + }, + { + "epoch": 1.6210350584307178, + "grad_norm": 0.4573573215762885, + "learning_rate": 1.0543916472484105e-06, + "loss": 0.2461, + "step": 4855 + }, + { + "epoch": 1.6213689482470786, + "grad_norm": 0.43462424985379644, + "learning_rate": 1.052602591665612e-06, + "loss": 0.2427, + "step": 4856 + }, + { + "epoch": 1.621702838063439, + "grad_norm": 0.46030038790232186, + "learning_rate": 1.0508148765985621e-06, + "loss": 0.2536, + "step": 4857 + }, + { + "epoch": 1.6220367278797996, + "grad_norm": 0.45160645742322175, + "learning_rate": 1.0490285026543612e-06, + "loss": 0.2449, + "step": 4858 + }, + { + "epoch": 1.6223706176961603, + "grad_norm": 0.4207463040650711, + "learning_rate": 1.0472434704396544e-06, + "loss": 0.2336, + "step": 4859 + }, + { + "epoch": 1.6227045075125208, + "grad_norm": 0.44882850182490164, + "learning_rate": 1.0454597805606282e-06, + "loss": 0.2407, + "step": 4860 + }, + { + "epoch": 1.6230383973288816, + "grad_norm": 0.4624569500563338, + "learning_rate": 1.0436774336230137e-06, + "loss": 0.2428, + "step": 4861 + }, + { + "epoch": 1.623372287145242, + "grad_norm": 0.44624185050815796, + "learning_rate": 1.0418964302320854e-06, + "loss": 0.2389, + "step": 4862 + }, + { + "epoch": 1.6237061769616026, + "grad_norm": 0.45606909768826887, + "learning_rate": 1.0401167709926663e-06, + "loss": 0.2525, + "step": 4863 + }, + { + "epoch": 1.6240400667779633, + "grad_norm": 0.46431992671022676, + "learning_rate": 1.038338456509118e-06, + "loss": 0.2429, + "step": 4864 + }, + { + "epoch": 1.6243739565943238, + "grad_norm": 0.45497051759066165, + "learning_rate": 1.0365614873853463e-06, + "loss": 0.2427, + "step": 4865 + }, + { + "epoch": 1.6247078464106846, + "grad_norm": 0.45500317103062743, + "learning_rate": 1.0347858642247998e-06, + "loss": 0.25, + "step": 4866 + }, + { + "epoch": 1.625041736227045, + "grad_norm": 0.4493672186025745, + "learning_rate": 1.033011587630472e-06, + "loss": 0.25, + "step": 4867 + }, + { + "epoch": 1.6253756260434056, + "grad_norm": 0.4622376802903284, + "learning_rate": 1.0312386582048994e-06, + "loss": 0.2531, + "step": 4868 + }, + { + "epoch": 1.6257095158597663, + "grad_norm": 0.467148200367176, + "learning_rate": 1.0294670765501592e-06, + "loss": 0.2588, + "step": 4869 + }, + { + "epoch": 1.626043405676127, + "grad_norm": 0.4486066451263577, + "learning_rate": 1.0276968432678702e-06, + "loss": 0.2503, + "step": 4870 + }, + { + "epoch": 1.6263772954924876, + "grad_norm": 0.45699590775799775, + "learning_rate": 1.0259279589591936e-06, + "loss": 0.256, + "step": 4871 + }, + { + "epoch": 1.626711185308848, + "grad_norm": 0.45960192604415945, + "learning_rate": 1.0241604242248365e-06, + "loss": 0.2423, + "step": 4872 + }, + { + "epoch": 1.6270450751252086, + "grad_norm": 0.4494883891506108, + "learning_rate": 1.022394239665042e-06, + "loss": 0.2529, + "step": 4873 + }, + { + "epoch": 1.6273789649415693, + "grad_norm": 0.4491002311726028, + "learning_rate": 1.0206294058795974e-06, + "loss": 0.2461, + "step": 4874 + }, + { + "epoch": 1.62771285475793, + "grad_norm": 0.45949508387997245, + "learning_rate": 1.0188659234678327e-06, + "loss": 0.2456, + "step": 4875 + }, + { + "epoch": 1.6280467445742905, + "grad_norm": 0.45681828013115316, + "learning_rate": 1.0171037930286154e-06, + "loss": 0.257, + "step": 4876 + }, + { + "epoch": 1.628380634390651, + "grad_norm": 0.46015879113696384, + "learning_rate": 1.0153430151603583e-06, + "loss": 0.2446, + "step": 4877 + }, + { + "epoch": 1.6287145242070116, + "grad_norm": 0.45090618366566876, + "learning_rate": 1.013583590461011e-06, + "loss": 0.2524, + "step": 4878 + }, + { + "epoch": 1.6290484140233723, + "grad_norm": 0.4543756851799613, + "learning_rate": 1.0118255195280646e-06, + "loss": 0.2458, + "step": 4879 + }, + { + "epoch": 1.629382303839733, + "grad_norm": 0.44575137953367006, + "learning_rate": 1.010068802958551e-06, + "loss": 0.2523, + "step": 4880 + }, + { + "epoch": 1.6297161936560935, + "grad_norm": 0.4418112583801264, + "learning_rate": 1.0083134413490431e-06, + "loss": 0.2477, + "step": 4881 + }, + { + "epoch": 1.630050083472454, + "grad_norm": 0.45319761823584886, + "learning_rate": 1.0065594352956526e-06, + "loss": 0.251, + "step": 4882 + }, + { + "epoch": 1.6303839732888146, + "grad_norm": 0.5030604970420042, + "learning_rate": 1.0048067853940286e-06, + "loss": 0.2516, + "step": 4883 + }, + { + "epoch": 1.6307178631051753, + "grad_norm": 0.453512042749142, + "learning_rate": 1.0030554922393648e-06, + "loss": 0.2364, + "step": 4884 + }, + { + "epoch": 1.631051752921536, + "grad_norm": 0.44760889669215237, + "learning_rate": 1.001305556426389e-06, + "loss": 0.247, + "step": 4885 + }, + { + "epoch": 1.6313856427378965, + "grad_norm": 0.4891463819112501, + "learning_rate": 9.99556978549373e-07, + "loss": 0.2462, + "step": 4886 + }, + { + "epoch": 1.631719532554257, + "grad_norm": 0.45362443978565875, + "learning_rate": 9.978097592021236e-07, + "loss": 0.2472, + "step": 4887 + }, + { + "epoch": 1.6320534223706176, + "grad_norm": 0.47838422597281766, + "learning_rate": 9.96063898977987e-07, + "loss": 0.2565, + "step": 4888 + }, + { + "epoch": 1.6323873121869783, + "grad_norm": 0.45326862331929757, + "learning_rate": 9.943193984698469e-07, + "loss": 0.2447, + "step": 4889 + }, + { + "epoch": 1.632721202003339, + "grad_norm": 0.4431785714732806, + "learning_rate": 9.9257625827013e-07, + "loss": 0.2412, + "step": 4890 + }, + { + "epoch": 1.6330550918196995, + "grad_norm": 0.48739428048912636, + "learning_rate": 9.90834478970794e-07, + "loss": 0.2486, + "step": 4891 + }, + { + "epoch": 1.63338898163606, + "grad_norm": 0.45696840369514413, + "learning_rate": 9.890940611633416e-07, + "loss": 0.2569, + "step": 4892 + }, + { + "epoch": 1.6337228714524206, + "grad_norm": 0.4518721204237414, + "learning_rate": 9.873550054388082e-07, + "loss": 0.2465, + "step": 4893 + }, + { + "epoch": 1.6340567612687813, + "grad_norm": 0.4432479659038795, + "learning_rate": 9.856173123877667e-07, + "loss": 0.2559, + "step": 4894 + }, + { + "epoch": 1.634390651085142, + "grad_norm": 0.4437214729257971, + "learning_rate": 9.838809826003315e-07, + "loss": 0.2518, + "step": 4895 + }, + { + "epoch": 1.6347245409015025, + "grad_norm": 0.45720541309596996, + "learning_rate": 9.8214601666615e-07, + "loss": 0.2427, + "step": 4896 + }, + { + "epoch": 1.635058430717863, + "grad_norm": 0.4701106549117815, + "learning_rate": 9.80412415174407e-07, + "loss": 0.2503, + "step": 4897 + }, + { + "epoch": 1.6353923205342236, + "grad_norm": 0.4650269018502718, + "learning_rate": 9.786801787138245e-07, + "loss": 0.2648, + "step": 4898 + }, + { + "epoch": 1.6357262103505843, + "grad_norm": 0.43997592107513644, + "learning_rate": 9.769493078726622e-07, + "loss": 0.232, + "step": 4899 + }, + { + "epoch": 1.636060100166945, + "grad_norm": 0.45066894546428077, + "learning_rate": 9.752198032387133e-07, + "loss": 0.2339, + "step": 4900 + }, + { + "epoch": 1.6363939899833055, + "grad_norm": 0.45648081316088374, + "learning_rate": 9.734916653993104e-07, + "loss": 0.2513, + "step": 4901 + }, + { + "epoch": 1.636727879799666, + "grad_norm": 0.4397278923598162, + "learning_rate": 9.7176489494132e-07, + "loss": 0.243, + "step": 4902 + }, + { + "epoch": 1.6370617696160266, + "grad_norm": 0.4561154521620516, + "learning_rate": 9.700394924511424e-07, + "loss": 0.2438, + "step": 4903 + }, + { + "epoch": 1.6373956594323873, + "grad_norm": 0.4500074750490054, + "learning_rate": 9.68315458514718e-07, + "loss": 0.2505, + "step": 4904 + }, + { + "epoch": 1.637729549248748, + "grad_norm": 0.4561350353946136, + "learning_rate": 9.665927937175184e-07, + "loss": 0.254, + "step": 4905 + }, + { + "epoch": 1.6380634390651085, + "grad_norm": 0.4544729188031643, + "learning_rate": 9.648714986445523e-07, + "loss": 0.253, + "step": 4906 + }, + { + "epoch": 1.638397328881469, + "grad_norm": 0.4691131852348111, + "learning_rate": 9.631515738803604e-07, + "loss": 0.2538, + "step": 4907 + }, + { + "epoch": 1.6387312186978296, + "grad_norm": 0.4583099073999997, + "learning_rate": 9.614330200090221e-07, + "loss": 0.2529, + "step": 4908 + }, + { + "epoch": 1.6390651085141903, + "grad_norm": 0.46137964891012934, + "learning_rate": 9.597158376141513e-07, + "loss": 0.2506, + "step": 4909 + }, + { + "epoch": 1.639398998330551, + "grad_norm": 0.455758218412398, + "learning_rate": 9.580000272788915e-07, + "loss": 0.243, + "step": 4910 + }, + { + "epoch": 1.6397328881469115, + "grad_norm": 0.4544772708468556, + "learning_rate": 9.56285589585924e-07, + "loss": 0.2425, + "step": 4911 + }, + { + "epoch": 1.640066777963272, + "grad_norm": 0.44727553038514983, + "learning_rate": 9.545725251174616e-07, + "loss": 0.2415, + "step": 4912 + }, + { + "epoch": 1.6404006677796326, + "grad_norm": 0.4459143713119041, + "learning_rate": 9.528608344552548e-07, + "loss": 0.2498, + "step": 4913 + }, + { + "epoch": 1.6407345575959933, + "grad_norm": 0.4559148733418392, + "learning_rate": 9.51150518180583e-07, + "loss": 0.2492, + "step": 4914 + }, + { + "epoch": 1.641068447412354, + "grad_norm": 0.4488240106866857, + "learning_rate": 9.494415768742616e-07, + "loss": 0.2433, + "step": 4915 + }, + { + "epoch": 1.6414023372287145, + "grad_norm": 0.4509153511856908, + "learning_rate": 9.477340111166361e-07, + "loss": 0.2464, + "step": 4916 + }, + { + "epoch": 1.641736227045075, + "grad_norm": 0.43588483132038053, + "learning_rate": 9.460278214875884e-07, + "loss": 0.246, + "step": 4917 + }, + { + "epoch": 1.6420701168614358, + "grad_norm": 0.4542745702990466, + "learning_rate": 9.443230085665339e-07, + "loss": 0.2497, + "step": 4918 + }, + { + "epoch": 1.6424040066777963, + "grad_norm": 0.4477133547036678, + "learning_rate": 9.426195729324161e-07, + "loss": 0.2435, + "step": 4919 + }, + { + "epoch": 1.642737896494157, + "grad_norm": 0.4483405537079852, + "learning_rate": 9.409175151637134e-07, + "loss": 0.2423, + "step": 4920 + }, + { + "epoch": 1.6430717863105175, + "grad_norm": 0.445507746785088, + "learning_rate": 9.392168358384346e-07, + "loss": 0.2395, + "step": 4921 + }, + { + "epoch": 1.643405676126878, + "grad_norm": 0.44309147442316743, + "learning_rate": 9.375175355341232e-07, + "loss": 0.2478, + "step": 4922 + }, + { + "epoch": 1.6437395659432388, + "grad_norm": 0.45720910281244725, + "learning_rate": 9.358196148278525e-07, + "loss": 0.2515, + "step": 4923 + }, + { + "epoch": 1.6440734557595993, + "grad_norm": 0.4648586021402646, + "learning_rate": 9.34123074296227e-07, + "loss": 0.256, + "step": 4924 + }, + { + "epoch": 1.64440734557596, + "grad_norm": 0.45108781917324575, + "learning_rate": 9.324279145153825e-07, + "loss": 0.2457, + "step": 4925 + }, + { + "epoch": 1.6447412353923205, + "grad_norm": 0.44644137777148796, + "learning_rate": 9.307341360609873e-07, + "loss": 0.2503, + "step": 4926 + }, + { + "epoch": 1.645075125208681, + "grad_norm": 0.44913422535241065, + "learning_rate": 9.290417395082413e-07, + "loss": 0.2425, + "step": 4927 + }, + { + "epoch": 1.6454090150250418, + "grad_norm": 0.4404592611271814, + "learning_rate": 9.27350725431872e-07, + "loss": 0.2379, + "step": 4928 + }, + { + "epoch": 1.6457429048414023, + "grad_norm": 0.45428215281841033, + "learning_rate": 9.25661094406139e-07, + "loss": 0.256, + "step": 4929 + }, + { + "epoch": 1.646076794657763, + "grad_norm": 0.4678478900102358, + "learning_rate": 9.239728470048315e-07, + "loss": 0.2547, + "step": 4930 + }, + { + "epoch": 1.6464106844741235, + "grad_norm": 0.4314479387347268, + "learning_rate": 9.222859838012721e-07, + "loss": 0.2375, + "step": 4931 + }, + { + "epoch": 1.646744574290484, + "grad_norm": 0.4625302315067015, + "learning_rate": 9.206005053683093e-07, + "loss": 0.2442, + "step": 4932 + }, + { + "epoch": 1.6470784641068448, + "grad_norm": 0.47068770271669585, + "learning_rate": 9.189164122783222e-07, + "loss": 0.2498, + "step": 4933 + }, + { + "epoch": 1.6474123539232055, + "grad_norm": 0.46773429815118156, + "learning_rate": 9.172337051032204e-07, + "loss": 0.255, + "step": 4934 + }, + { + "epoch": 1.647746243739566, + "grad_norm": 0.4621349702684146, + "learning_rate": 9.15552384414442e-07, + "loss": 0.248, + "step": 4935 + }, + { + "epoch": 1.6480801335559265, + "grad_norm": 0.4662042704468137, + "learning_rate": 9.138724507829577e-07, + "loss": 0.2563, + "step": 4936 + }, + { + "epoch": 1.648414023372287, + "grad_norm": 0.45596780964546535, + "learning_rate": 9.121939047792622e-07, + "loss": 0.2407, + "step": 4937 + }, + { + "epoch": 1.6487479131886478, + "grad_norm": 0.46385245397667146, + "learning_rate": 9.105167469733811e-07, + "loss": 0.2412, + "step": 4938 + }, + { + "epoch": 1.6490818030050085, + "grad_norm": 0.4480710007018189, + "learning_rate": 9.088409779348673e-07, + "loss": 0.2511, + "step": 4939 + }, + { + "epoch": 1.649415692821369, + "grad_norm": 0.4787438496542397, + "learning_rate": 9.07166598232806e-07, + "loss": 0.2533, + "step": 4940 + }, + { + "epoch": 1.6497495826377295, + "grad_norm": 0.46230086933512715, + "learning_rate": 9.054936084358062e-07, + "loss": 0.2484, + "step": 4941 + }, + { + "epoch": 1.65008347245409, + "grad_norm": 0.4459643961828653, + "learning_rate": 9.03822009112006e-07, + "loss": 0.2452, + "step": 4942 + }, + { + "epoch": 1.6504173622704508, + "grad_norm": 0.4479810575696321, + "learning_rate": 9.021518008290741e-07, + "loss": 0.2403, + "step": 4943 + }, + { + "epoch": 1.6507512520868115, + "grad_norm": 0.4774413423445045, + "learning_rate": 9.004829841542018e-07, + "loss": 0.2541, + "step": 4944 + }, + { + "epoch": 1.651085141903172, + "grad_norm": 0.4557133133997602, + "learning_rate": 8.988155596541131e-07, + "loss": 0.2558, + "step": 4945 + }, + { + "epoch": 1.6514190317195325, + "grad_norm": 0.4556081757600427, + "learning_rate": 8.971495278950559e-07, + "loss": 0.2488, + "step": 4946 + }, + { + "epoch": 1.651752921535893, + "grad_norm": 0.43390467580356995, + "learning_rate": 8.954848894428053e-07, + "loss": 0.241, + "step": 4947 + }, + { + "epoch": 1.6520868113522538, + "grad_norm": 0.45312741417964586, + "learning_rate": 8.938216448626635e-07, + "loss": 0.2559, + "step": 4948 + }, + { + "epoch": 1.6524207011686145, + "grad_norm": 0.45440967868544174, + "learning_rate": 8.921597947194621e-07, + "loss": 0.238, + "step": 4949 + }, + { + "epoch": 1.652754590984975, + "grad_norm": 0.4421167964400811, + "learning_rate": 8.904993395775552e-07, + "loss": 0.2514, + "step": 4950 + }, + { + "epoch": 1.6530884808013355, + "grad_norm": 0.4527455475339987, + "learning_rate": 8.888402800008239e-07, + "loss": 0.2416, + "step": 4951 + }, + { + "epoch": 1.653422370617696, + "grad_norm": 0.4469528428149118, + "learning_rate": 8.871826165526781e-07, + "loss": 0.2567, + "step": 4952 + }, + { + "epoch": 1.6537562604340568, + "grad_norm": 0.46260546691903326, + "learning_rate": 8.855263497960504e-07, + "loss": 0.2501, + "step": 4953 + }, + { + "epoch": 1.6540901502504175, + "grad_norm": 0.4470968718880415, + "learning_rate": 8.838714802934017e-07, + "loss": 0.2429, + "step": 4954 + }, + { + "epoch": 1.654424040066778, + "grad_norm": 0.4369878675230337, + "learning_rate": 8.822180086067161e-07, + "loss": 0.249, + "step": 4955 + }, + { + "epoch": 1.6547579298831385, + "grad_norm": 0.4638955003792206, + "learning_rate": 8.805659352975049e-07, + "loss": 0.2523, + "step": 4956 + }, + { + "epoch": 1.655091819699499, + "grad_norm": 0.44275427092913927, + "learning_rate": 8.789152609268015e-07, + "loss": 0.2432, + "step": 4957 + }, + { + "epoch": 1.6554257095158598, + "grad_norm": 0.46217753628700636, + "learning_rate": 8.772659860551691e-07, + "loss": 0.2592, + "step": 4958 + }, + { + "epoch": 1.6557595993322205, + "grad_norm": 0.4467446395017067, + "learning_rate": 8.756181112426909e-07, + "loss": 0.243, + "step": 4959 + }, + { + "epoch": 1.656093489148581, + "grad_norm": 0.4569568561752167, + "learning_rate": 8.739716370489787e-07, + "loss": 0.2573, + "step": 4960 + }, + { + "epoch": 1.6564273789649415, + "grad_norm": 0.4732945162269539, + "learning_rate": 8.723265640331652e-07, + "loss": 0.2558, + "step": 4961 + }, + { + "epoch": 1.656761268781302, + "grad_norm": 0.43699916354060486, + "learning_rate": 8.706828927539079e-07, + "loss": 0.2385, + "step": 4962 + }, + { + "epoch": 1.6570951585976628, + "grad_norm": 0.4462289327186133, + "learning_rate": 8.690406237693916e-07, + "loss": 0.2451, + "step": 4963 + }, + { + "epoch": 1.6574290484140235, + "grad_norm": 0.44743867320605774, + "learning_rate": 8.673997576373205e-07, + "loss": 0.2516, + "step": 4964 + }, + { + "epoch": 1.657762938230384, + "grad_norm": 0.4566444381750538, + "learning_rate": 8.657602949149252e-07, + "loss": 0.2528, + "step": 4965 + }, + { + "epoch": 1.6580968280467445, + "grad_norm": 0.4670770218096883, + "learning_rate": 8.641222361589563e-07, + "loss": 0.2506, + "step": 4966 + }, + { + "epoch": 1.658430717863105, + "grad_norm": 0.47094121358144647, + "learning_rate": 8.624855819256922e-07, + "loss": 0.2521, + "step": 4967 + }, + { + "epoch": 1.6587646076794658, + "grad_norm": 0.43584329444644027, + "learning_rate": 8.608503327709333e-07, + "loss": 0.2414, + "step": 4968 + }, + { + "epoch": 1.6590984974958265, + "grad_norm": 0.4601988212087873, + "learning_rate": 8.592164892500005e-07, + "loss": 0.2451, + "step": 4969 + }, + { + "epoch": 1.659432387312187, + "grad_norm": 0.4529405199950001, + "learning_rate": 8.575840519177386e-07, + "loss": 0.2488, + "step": 4970 + }, + { + "epoch": 1.6597662771285475, + "grad_norm": 0.44883542575738233, + "learning_rate": 8.559530213285133e-07, + "loss": 0.2354, + "step": 4971 + }, + { + "epoch": 1.660100166944908, + "grad_norm": 0.4512953816203883, + "learning_rate": 8.543233980362175e-07, + "loss": 0.2462, + "step": 4972 + }, + { + "epoch": 1.6604340567612688, + "grad_norm": 0.44692567575881587, + "learning_rate": 8.526951825942608e-07, + "loss": 0.243, + "step": 4973 + }, + { + "epoch": 1.6607679465776295, + "grad_norm": 0.4608743157126187, + "learning_rate": 8.510683755555771e-07, + "loss": 0.2501, + "step": 4974 + }, + { + "epoch": 1.66110183639399, + "grad_norm": 0.4480689106610217, + "learning_rate": 8.494429774726198e-07, + "loss": 0.2495, + "step": 4975 + }, + { + "epoch": 1.6614357262103505, + "grad_norm": 0.4724967255017713, + "learning_rate": 8.478189888973676e-07, + "loss": 0.2574, + "step": 4976 + }, + { + "epoch": 1.661769616026711, + "grad_norm": 0.43508357080406496, + "learning_rate": 8.461964103813203e-07, + "loss": 0.2358, + "step": 4977 + }, + { + "epoch": 1.6621035058430718, + "grad_norm": 0.4579897245145642, + "learning_rate": 8.44575242475495e-07, + "loss": 0.2447, + "step": 4978 + }, + { + "epoch": 1.6624373956594325, + "grad_norm": 0.44745789202067504, + "learning_rate": 8.429554857304318e-07, + "loss": 0.2419, + "step": 4979 + }, + { + "epoch": 1.662771285475793, + "grad_norm": 0.4795170095942476, + "learning_rate": 8.413371406961918e-07, + "loss": 0.2605, + "step": 4980 + }, + { + "epoch": 1.6631051752921535, + "grad_norm": 0.45516646443472825, + "learning_rate": 8.397202079223577e-07, + "loss": 0.2469, + "step": 4981 + }, + { + "epoch": 1.663439065108514, + "grad_norm": 0.4513654347024176, + "learning_rate": 8.381046879580307e-07, + "loss": 0.2455, + "step": 4982 + }, + { + "epoch": 1.6637729549248748, + "grad_norm": 0.45893907227406194, + "learning_rate": 8.364905813518331e-07, + "loss": 0.2475, + "step": 4983 + }, + { + "epoch": 1.6641068447412355, + "grad_norm": 0.46841323002845897, + "learning_rate": 8.348778886519065e-07, + "loss": 0.251, + "step": 4984 + }, + { + "epoch": 1.664440734557596, + "grad_norm": 0.4599741331316511, + "learning_rate": 8.332666104059134e-07, + "loss": 0.2566, + "step": 4985 + }, + { + "epoch": 1.6647746243739565, + "grad_norm": 0.468507681202396, + "learning_rate": 8.316567471610371e-07, + "loss": 0.249, + "step": 4986 + }, + { + "epoch": 1.6651085141903172, + "grad_norm": 0.4661835352111316, + "learning_rate": 8.300482994639775e-07, + "loss": 0.2491, + "step": 4987 + }, + { + "epoch": 1.6654424040066778, + "grad_norm": 0.4593239332679216, + "learning_rate": 8.284412678609555e-07, + "loss": 0.2483, + "step": 4988 + }, + { + "epoch": 1.6657762938230385, + "grad_norm": 0.42899738316499014, + "learning_rate": 8.268356528977089e-07, + "loss": 0.2338, + "step": 4989 + }, + { + "epoch": 1.666110183639399, + "grad_norm": 0.4418493052273266, + "learning_rate": 8.25231455119499e-07, + "loss": 0.2437, + "step": 4990 + }, + { + "epoch": 1.6664440734557595, + "grad_norm": 0.4563959029152161, + "learning_rate": 8.23628675071102e-07, + "loss": 0.2472, + "step": 4991 + }, + { + "epoch": 1.6667779632721202, + "grad_norm": 0.4608917465802201, + "learning_rate": 8.220273132968137e-07, + "loss": 0.2509, + "step": 4992 + }, + { + "epoch": 1.6671118530884808, + "grad_norm": 0.45630872457287847, + "learning_rate": 8.204273703404463e-07, + "loss": 0.2456, + "step": 4993 + }, + { + "epoch": 1.6674457429048415, + "grad_norm": 0.43827556779444277, + "learning_rate": 8.188288467453348e-07, + "loss": 0.2428, + "step": 4994 + }, + { + "epoch": 1.667779632721202, + "grad_norm": 0.44646703692754774, + "learning_rate": 8.172317430543292e-07, + "loss": 0.2406, + "step": 4995 + }, + { + "epoch": 1.6681135225375625, + "grad_norm": 0.46080152275038727, + "learning_rate": 8.156360598097979e-07, + "loss": 0.2435, + "step": 4996 + }, + { + "epoch": 1.6684474123539232, + "grad_norm": 0.45189056532854655, + "learning_rate": 8.140417975536258e-07, + "loss": 0.2494, + "step": 4997 + }, + { + "epoch": 1.6687813021702838, + "grad_norm": 0.4575835894175166, + "learning_rate": 8.124489568272148e-07, + "loss": 0.2402, + "step": 4998 + }, + { + "epoch": 1.6691151919866445, + "grad_norm": 0.4456578093532183, + "learning_rate": 8.108575381714883e-07, + "loss": 0.2526, + "step": 4999 + }, + { + "epoch": 1.669449081803005, + "grad_norm": 0.4661989859569466, + "learning_rate": 8.092675421268825e-07, + "loss": 0.2582, + "step": 5000 + }, + { + "epoch": 1.6697829716193655, + "grad_norm": 0.4663856486622323, + "learning_rate": 8.076789692333503e-07, + "loss": 0.2474, + "step": 5001 + }, + { + "epoch": 1.6701168614357262, + "grad_norm": 0.45650783370945985, + "learning_rate": 8.060918200303658e-07, + "loss": 0.2507, + "step": 5002 + }, + { + "epoch": 1.670450751252087, + "grad_norm": 0.46156175763494656, + "learning_rate": 8.045060950569139e-07, + "loss": 0.2458, + "step": 5003 + }, + { + "epoch": 1.6707846410684475, + "grad_norm": 0.4625348658598544, + "learning_rate": 8.029217948515006e-07, + "loss": 0.2554, + "step": 5004 + }, + { + "epoch": 1.671118530884808, + "grad_norm": 0.4432673905826133, + "learning_rate": 8.013389199521459e-07, + "loss": 0.2475, + "step": 5005 + }, + { + "epoch": 1.6714524207011685, + "grad_norm": 0.4545670663707961, + "learning_rate": 7.997574708963852e-07, + "loss": 0.2494, + "step": 5006 + }, + { + "epoch": 1.6717863105175292, + "grad_norm": 0.44118571469835755, + "learning_rate": 7.981774482212695e-07, + "loss": 0.2349, + "step": 5007 + }, + { + "epoch": 1.67212020033389, + "grad_norm": 0.45825719748694327, + "learning_rate": 7.965988524633683e-07, + "loss": 0.2528, + "step": 5008 + }, + { + "epoch": 1.6724540901502505, + "grad_norm": 0.4412907472954062, + "learning_rate": 7.950216841587638e-07, + "loss": 0.2408, + "step": 5009 + }, + { + "epoch": 1.672787979966611, + "grad_norm": 0.4443874786635421, + "learning_rate": 7.934459438430531e-07, + "loss": 0.2376, + "step": 5010 + }, + { + "epoch": 1.6731218697829715, + "grad_norm": 0.4542674974163823, + "learning_rate": 7.918716320513514e-07, + "loss": 0.2503, + "step": 5011 + }, + { + "epoch": 1.6734557595993322, + "grad_norm": 0.44330422775252776, + "learning_rate": 7.902987493182846e-07, + "loss": 0.2429, + "step": 5012 + }, + { + "epoch": 1.673789649415693, + "grad_norm": 0.4475912873142635, + "learning_rate": 7.887272961779979e-07, + "loss": 0.2435, + "step": 5013 + }, + { + "epoch": 1.6741235392320535, + "grad_norm": 0.451612105599901, + "learning_rate": 7.871572731641481e-07, + "loss": 0.2431, + "step": 5014 + }, + { + "epoch": 1.674457429048414, + "grad_norm": 0.4583353876993538, + "learning_rate": 7.855886808099062e-07, + "loss": 0.2518, + "step": 5015 + }, + { + "epoch": 1.6747913188647745, + "grad_norm": 0.43661989897171793, + "learning_rate": 7.840215196479567e-07, + "loss": 0.2397, + "step": 5016 + }, + { + "epoch": 1.6751252086811352, + "grad_norm": 0.467656341422543, + "learning_rate": 7.824557902105024e-07, + "loss": 0.2607, + "step": 5017 + }, + { + "epoch": 1.675459098497496, + "grad_norm": 0.4516524695089928, + "learning_rate": 7.808914930292544e-07, + "loss": 0.2457, + "step": 5018 + }, + { + "epoch": 1.6757929883138565, + "grad_norm": 0.475825189203761, + "learning_rate": 7.793286286354418e-07, + "loss": 0.2663, + "step": 5019 + }, + { + "epoch": 1.676126878130217, + "grad_norm": 0.4514118612587378, + "learning_rate": 7.77767197559805e-07, + "loss": 0.2637, + "step": 5020 + }, + { + "epoch": 1.6764607679465775, + "grad_norm": 0.4531707892057287, + "learning_rate": 7.76207200332596e-07, + "loss": 0.2526, + "step": 5021 + }, + { + "epoch": 1.6767946577629382, + "grad_norm": 0.4355946884864162, + "learning_rate": 7.746486374835843e-07, + "loss": 0.2463, + "step": 5022 + }, + { + "epoch": 1.677128547579299, + "grad_norm": 0.45600170443569743, + "learning_rate": 7.730915095420489e-07, + "loss": 0.2475, + "step": 5023 + }, + { + "epoch": 1.6774624373956595, + "grad_norm": 0.4461590456911649, + "learning_rate": 7.71535817036782e-07, + "loss": 0.2412, + "step": 5024 + }, + { + "epoch": 1.67779632721202, + "grad_norm": 0.45091895530538456, + "learning_rate": 7.699815604960892e-07, + "loss": 0.242, + "step": 5025 + }, + { + "epoch": 1.6781302170283805, + "grad_norm": 0.44913420322461034, + "learning_rate": 7.684287404477864e-07, + "loss": 0.2455, + "step": 5026 + }, + { + "epoch": 1.6784641068447412, + "grad_norm": 0.4458262027099179, + "learning_rate": 7.66877357419204e-07, + "loss": 0.2465, + "step": 5027 + }, + { + "epoch": 1.678797996661102, + "grad_norm": 0.4822682152678643, + "learning_rate": 7.653274119371862e-07, + "loss": 0.2571, + "step": 5028 + }, + { + "epoch": 1.6791318864774625, + "grad_norm": 0.43064437669723854, + "learning_rate": 7.637789045280841e-07, + "loss": 0.2353, + "step": 5029 + }, + { + "epoch": 1.679465776293823, + "grad_norm": 0.4401367776131267, + "learning_rate": 7.622318357177633e-07, + "loss": 0.2408, + "step": 5030 + }, + { + "epoch": 1.6797996661101835, + "grad_norm": 0.46831190225753083, + "learning_rate": 7.606862060315989e-07, + "loss": 0.254, + "step": 5031 + }, + { + "epoch": 1.6801335559265442, + "grad_norm": 0.45779560410261244, + "learning_rate": 7.591420159944813e-07, + "loss": 0.2516, + "step": 5032 + }, + { + "epoch": 1.680467445742905, + "grad_norm": 0.4581771078979595, + "learning_rate": 7.575992661308085e-07, + "loss": 0.2386, + "step": 5033 + }, + { + "epoch": 1.6808013355592655, + "grad_norm": 0.46649975096329593, + "learning_rate": 7.5605795696449e-07, + "loss": 0.2573, + "step": 5034 + }, + { + "epoch": 1.681135225375626, + "grad_norm": 0.46576102113383205, + "learning_rate": 7.545180890189441e-07, + "loss": 0.2556, + "step": 5035 + }, + { + "epoch": 1.6814691151919865, + "grad_norm": 0.45181432832268675, + "learning_rate": 7.529796628171071e-07, + "loss": 0.2499, + "step": 5036 + }, + { + "epoch": 1.6818030050083472, + "grad_norm": 0.43541453147164444, + "learning_rate": 7.514426788814178e-07, + "loss": 0.2363, + "step": 5037 + }, + { + "epoch": 1.682136894824708, + "grad_norm": 0.455984295905807, + "learning_rate": 7.499071377338279e-07, + "loss": 0.2513, + "step": 5038 + }, + { + "epoch": 1.6824707846410685, + "grad_norm": 0.4592193209739497, + "learning_rate": 7.483730398958001e-07, + "loss": 0.2362, + "step": 5039 + }, + { + "epoch": 1.682804674457429, + "grad_norm": 0.4412060403746792, + "learning_rate": 7.468403858883038e-07, + "loss": 0.2411, + "step": 5040 + }, + { + "epoch": 1.6831385642737895, + "grad_norm": 0.4582278706629776, + "learning_rate": 7.453091762318232e-07, + "loss": 0.2476, + "step": 5041 + }, + { + "epoch": 1.6834724540901502, + "grad_norm": 0.4638298383139332, + "learning_rate": 7.437794114463482e-07, + "loss": 0.2457, + "step": 5042 + }, + { + "epoch": 1.683806343906511, + "grad_norm": 0.46839973253104444, + "learning_rate": 7.422510920513792e-07, + "loss": 0.2476, + "step": 5043 + }, + { + "epoch": 1.6841402337228715, + "grad_norm": 0.47201372817087445, + "learning_rate": 7.407242185659236e-07, + "loss": 0.2335, + "step": 5044 + }, + { + "epoch": 1.684474123539232, + "grad_norm": 0.4584513428052339, + "learning_rate": 7.391987915085014e-07, + "loss": 0.247, + "step": 5045 + }, + { + "epoch": 1.6848080133555925, + "grad_norm": 0.4359755756979785, + "learning_rate": 7.376748113971404e-07, + "loss": 0.2353, + "step": 5046 + }, + { + "epoch": 1.6851419031719532, + "grad_norm": 0.4476636910788323, + "learning_rate": 7.361522787493763e-07, + "loss": 0.2527, + "step": 5047 + }, + { + "epoch": 1.685475792988314, + "grad_norm": 0.43154958948895394, + "learning_rate": 7.346311940822526e-07, + "loss": 0.2401, + "step": 5048 + }, + { + "epoch": 1.6858096828046745, + "grad_norm": 0.4580855266248229, + "learning_rate": 7.331115579123205e-07, + "loss": 0.2517, + "step": 5049 + }, + { + "epoch": 1.686143572621035, + "grad_norm": 0.46102980132522886, + "learning_rate": 7.315933707556433e-07, + "loss": 0.2457, + "step": 5050 + }, + { + "epoch": 1.6864774624373957, + "grad_norm": 0.4611589850276941, + "learning_rate": 7.300766331277881e-07, + "loss": 0.2508, + "step": 5051 + }, + { + "epoch": 1.6868113522537562, + "grad_norm": 0.45296314627976003, + "learning_rate": 7.285613455438301e-07, + "loss": 0.2458, + "step": 5052 + }, + { + "epoch": 1.687145242070117, + "grad_norm": 0.44939770635257137, + "learning_rate": 7.270475085183554e-07, + "loss": 0.2432, + "step": 5053 + }, + { + "epoch": 1.6874791318864775, + "grad_norm": 0.4676955441954793, + "learning_rate": 7.255351225654527e-07, + "loss": 0.2535, + "step": 5054 + }, + { + "epoch": 1.687813021702838, + "grad_norm": 0.46070777805437274, + "learning_rate": 7.240241881987231e-07, + "loss": 0.2524, + "step": 5055 + }, + { + "epoch": 1.6881469115191987, + "grad_norm": 0.4579841791422698, + "learning_rate": 7.225147059312704e-07, + "loss": 0.2293, + "step": 5056 + }, + { + "epoch": 1.6884808013355592, + "grad_norm": 0.47083852190888437, + "learning_rate": 7.210066762757079e-07, + "loss": 0.2572, + "step": 5057 + }, + { + "epoch": 1.68881469115192, + "grad_norm": 0.44206343709494006, + "learning_rate": 7.195000997441532e-07, + "loss": 0.2485, + "step": 5058 + }, + { + "epoch": 1.6891485809682805, + "grad_norm": 0.4438660931330386, + "learning_rate": 7.179949768482337e-07, + "loss": 0.2482, + "step": 5059 + }, + { + "epoch": 1.689482470784641, + "grad_norm": 0.44015447856659345, + "learning_rate": 7.164913080990804e-07, + "loss": 0.2437, + "step": 5060 + }, + { + "epoch": 1.6898163606010017, + "grad_norm": 0.4489517452207698, + "learning_rate": 7.149890940073306e-07, + "loss": 0.2451, + "step": 5061 + }, + { + "epoch": 1.6901502504173622, + "grad_norm": 0.44736059098485464, + "learning_rate": 7.134883350831301e-07, + "loss": 0.2465, + "step": 5062 + }, + { + "epoch": 1.690484140233723, + "grad_norm": 0.44472566992398915, + "learning_rate": 7.119890318361278e-07, + "loss": 0.2488, + "step": 5063 + }, + { + "epoch": 1.6908180300500835, + "grad_norm": 0.4461006356666508, + "learning_rate": 7.104911847754803e-07, + "loss": 0.2399, + "step": 5064 + }, + { + "epoch": 1.691151919866444, + "grad_norm": 0.45104897234117713, + "learning_rate": 7.08994794409848e-07, + "loss": 0.2422, + "step": 5065 + }, + { + "epoch": 1.6914858096828047, + "grad_norm": 0.449045364043578, + "learning_rate": 7.074998612473977e-07, + "loss": 0.2456, + "step": 5066 + }, + { + "epoch": 1.6918196994991654, + "grad_norm": 0.45879976269934675, + "learning_rate": 7.060063857957994e-07, + "loss": 0.2505, + "step": 5067 + }, + { + "epoch": 1.692153589315526, + "grad_norm": 0.4594411904160147, + "learning_rate": 7.045143685622324e-07, + "loss": 0.2484, + "step": 5068 + }, + { + "epoch": 1.6924874791318865, + "grad_norm": 0.4464431925621404, + "learning_rate": 7.030238100533754e-07, + "loss": 0.234, + "step": 5069 + }, + { + "epoch": 1.692821368948247, + "grad_norm": 0.4483311817253807, + "learning_rate": 7.015347107754161e-07, + "loss": 0.2421, + "step": 5070 + }, + { + "epoch": 1.6931552587646077, + "grad_norm": 0.443831098764209, + "learning_rate": 7.000470712340451e-07, + "loss": 0.2537, + "step": 5071 + }, + { + "epoch": 1.6934891485809684, + "grad_norm": 0.4516433876876618, + "learning_rate": 6.98560891934455e-07, + "loss": 0.2431, + "step": 5072 + }, + { + "epoch": 1.693823038397329, + "grad_norm": 0.4571374372407988, + "learning_rate": 6.97076173381347e-07, + "loss": 0.244, + "step": 5073 + }, + { + "epoch": 1.6941569282136895, + "grad_norm": 0.4524675243285974, + "learning_rate": 6.955929160789227e-07, + "loss": 0.246, + "step": 5074 + }, + { + "epoch": 1.69449081803005, + "grad_norm": 0.44631358054407294, + "learning_rate": 6.941111205308893e-07, + "loss": 0.2456, + "step": 5075 + }, + { + "epoch": 1.6948247078464107, + "grad_norm": 0.45765406708183526, + "learning_rate": 6.926307872404548e-07, + "loss": 0.2565, + "step": 5076 + }, + { + "epoch": 1.6951585976627714, + "grad_norm": 0.4438575283734923, + "learning_rate": 6.911519167103353e-07, + "loss": 0.2516, + "step": 5077 + }, + { + "epoch": 1.695492487479132, + "grad_norm": 0.45177121799342257, + "learning_rate": 6.896745094427454e-07, + "loss": 0.247, + "step": 5078 + }, + { + "epoch": 1.6958263772954925, + "grad_norm": 0.4497893024979225, + "learning_rate": 6.881985659394075e-07, + "loss": 0.2411, + "step": 5079 + }, + { + "epoch": 1.696160267111853, + "grad_norm": 0.45215028438876925, + "learning_rate": 6.867240867015429e-07, + "loss": 0.251, + "step": 5080 + }, + { + "epoch": 1.6964941569282137, + "grad_norm": 0.45954686871527184, + "learning_rate": 6.852510722298761e-07, + "loss": 0.2577, + "step": 5081 + }, + { + "epoch": 1.6968280467445744, + "grad_norm": 0.44184369165437387, + "learning_rate": 6.837795230246375e-07, + "loss": 0.2425, + "step": 5082 + }, + { + "epoch": 1.697161936560935, + "grad_norm": 0.44664478060994145, + "learning_rate": 6.823094395855562e-07, + "loss": 0.2517, + "step": 5083 + }, + { + "epoch": 1.6974958263772955, + "grad_norm": 0.46739133160267904, + "learning_rate": 6.808408224118651e-07, + "loss": 0.2477, + "step": 5084 + }, + { + "epoch": 1.697829716193656, + "grad_norm": 0.45679820051655057, + "learning_rate": 6.793736720022981e-07, + "loss": 0.2525, + "step": 5085 + }, + { + "epoch": 1.6981636060100167, + "grad_norm": 0.4264353557047915, + "learning_rate": 6.779079888550933e-07, + "loss": 0.2355, + "step": 5086 + }, + { + "epoch": 1.6984974958263774, + "grad_norm": 0.4499731661470423, + "learning_rate": 6.764437734679891e-07, + "loss": 0.2457, + "step": 5087 + }, + { + "epoch": 1.698831385642738, + "grad_norm": 0.4777655295498718, + "learning_rate": 6.749810263382256e-07, + "loss": 0.2583, + "step": 5088 + }, + { + "epoch": 1.6991652754590985, + "grad_norm": 0.45260313522266155, + "learning_rate": 6.735197479625433e-07, + "loss": 0.2558, + "step": 5089 + }, + { + "epoch": 1.699499165275459, + "grad_norm": 0.4460293139615854, + "learning_rate": 6.72059938837184e-07, + "loss": 0.2491, + "step": 5090 + }, + { + "epoch": 1.6998330550918197, + "grad_norm": 0.4503223100012358, + "learning_rate": 6.706015994578935e-07, + "loss": 0.2466, + "step": 5091 + }, + { + "epoch": 1.7001669449081804, + "grad_norm": 0.44520766630535596, + "learning_rate": 6.691447303199156e-07, + "loss": 0.2484, + "step": 5092 + }, + { + "epoch": 1.700500834724541, + "grad_norm": 0.45961908140105795, + "learning_rate": 6.67689331917995e-07, + "loss": 0.2448, + "step": 5093 + }, + { + "epoch": 1.7008347245409015, + "grad_norm": 0.44876997223822873, + "learning_rate": 6.66235404746376e-07, + "loss": 0.2485, + "step": 5094 + }, + { + "epoch": 1.701168614357262, + "grad_norm": 0.4485018109990208, + "learning_rate": 6.647829492988072e-07, + "loss": 0.2387, + "step": 5095 + }, + { + "epoch": 1.7015025041736227, + "grad_norm": 0.43347116670568947, + "learning_rate": 6.633319660685345e-07, + "loss": 0.2386, + "step": 5096 + }, + { + "epoch": 1.7018363939899834, + "grad_norm": 0.44786954094084447, + "learning_rate": 6.618824555483044e-07, + "loss": 0.2476, + "step": 5097 + }, + { + "epoch": 1.702170283806344, + "grad_norm": 0.45205904623967663, + "learning_rate": 6.604344182303629e-07, + "loss": 0.2506, + "step": 5098 + }, + { + "epoch": 1.7025041736227045, + "grad_norm": 0.4433996876587384, + "learning_rate": 6.589878546064544e-07, + "loss": 0.2408, + "step": 5099 + }, + { + "epoch": 1.702838063439065, + "grad_norm": 0.45568041449947344, + "learning_rate": 6.575427651678267e-07, + "loss": 0.2512, + "step": 5100 + }, + { + "epoch": 1.7031719532554257, + "grad_norm": 0.45565800679624596, + "learning_rate": 6.560991504052244e-07, + "loss": 0.2608, + "step": 5101 + }, + { + "epoch": 1.7035058430717864, + "grad_norm": 0.45464171082474936, + "learning_rate": 6.546570108088907e-07, + "loss": 0.2542, + "step": 5102 + }, + { + "epoch": 1.703839732888147, + "grad_norm": 0.47257450320183253, + "learning_rate": 6.532163468685682e-07, + "loss": 0.2534, + "step": 5103 + }, + { + "epoch": 1.7041736227045075, + "grad_norm": 0.45723475444507994, + "learning_rate": 6.517771590734995e-07, + "loss": 0.2505, + "step": 5104 + }, + { + "epoch": 1.704507512520868, + "grad_norm": 0.44858326133820453, + "learning_rate": 6.50339447912427e-07, + "loss": 0.2462, + "step": 5105 + }, + { + "epoch": 1.7048414023372287, + "grad_norm": 0.4547635178797795, + "learning_rate": 6.489032138735884e-07, + "loss": 0.2481, + "step": 5106 + }, + { + "epoch": 1.7051752921535894, + "grad_norm": 0.4508524854831446, + "learning_rate": 6.474684574447221e-07, + "loss": 0.2465, + "step": 5107 + }, + { + "epoch": 1.70550918196995, + "grad_norm": 0.4499925366467944, + "learning_rate": 6.46035179113062e-07, + "loss": 0.2409, + "step": 5108 + }, + { + "epoch": 1.7058430717863104, + "grad_norm": 0.45312261240944457, + "learning_rate": 6.446033793653445e-07, + "loss": 0.236, + "step": 5109 + }, + { + "epoch": 1.706176961602671, + "grad_norm": 0.45276154529158813, + "learning_rate": 6.431730586878004e-07, + "loss": 0.2396, + "step": 5110 + }, + { + "epoch": 1.7065108514190317, + "grad_norm": 0.4435663560145072, + "learning_rate": 6.417442175661598e-07, + "loss": 0.2407, + "step": 5111 + }, + { + "epoch": 1.7068447412353924, + "grad_norm": 0.44071133369776866, + "learning_rate": 6.403168564856472e-07, + "loss": 0.2413, + "step": 5112 + }, + { + "epoch": 1.707178631051753, + "grad_norm": 0.44651045465134304, + "learning_rate": 6.388909759309892e-07, + "loss": 0.254, + "step": 5113 + }, + { + "epoch": 1.7075125208681134, + "grad_norm": 0.4543869972959644, + "learning_rate": 6.374665763864085e-07, + "loss": 0.2391, + "step": 5114 + }, + { + "epoch": 1.707846410684474, + "grad_norm": 0.44054572590312946, + "learning_rate": 6.360436583356222e-07, + "loss": 0.2495, + "step": 5115 + }, + { + "epoch": 1.7081803005008347, + "grad_norm": 0.45073803549676206, + "learning_rate": 6.34622222261847e-07, + "loss": 0.245, + "step": 5116 + }, + { + "epoch": 1.7085141903171954, + "grad_norm": 0.44240188805117187, + "learning_rate": 6.332022686477929e-07, + "loss": 0.2453, + "step": 5117 + }, + { + "epoch": 1.708848080133556, + "grad_norm": 0.43746724532666637, + "learning_rate": 6.317837979756713e-07, + "loss": 0.2428, + "step": 5118 + }, + { + "epoch": 1.7091819699499164, + "grad_norm": 0.46094574321452697, + "learning_rate": 6.303668107271871e-07, + "loss": 0.2495, + "step": 5119 + }, + { + "epoch": 1.7095158597662772, + "grad_norm": 0.44980001109588397, + "learning_rate": 6.289513073835396e-07, + "loss": 0.2398, + "step": 5120 + }, + { + "epoch": 1.7098497495826377, + "grad_norm": 0.4455515518258393, + "learning_rate": 6.275372884254299e-07, + "loss": 0.2466, + "step": 5121 + }, + { + "epoch": 1.7101836393989984, + "grad_norm": 0.44456938295262305, + "learning_rate": 6.261247543330484e-07, + "loss": 0.2506, + "step": 5122 + }, + { + "epoch": 1.710517529215359, + "grad_norm": 0.4441528306905688, + "learning_rate": 6.247137055860869e-07, + "loss": 0.2402, + "step": 5123 + }, + { + "epoch": 1.7108514190317194, + "grad_norm": 0.4578665364531156, + "learning_rate": 6.233041426637293e-07, + "loss": 0.2551, + "step": 5124 + }, + { + "epoch": 1.7111853088480802, + "grad_norm": 0.4620301346121438, + "learning_rate": 6.218960660446554e-07, + "loss": 0.2479, + "step": 5125 + }, + { + "epoch": 1.7115191986644407, + "grad_norm": 0.4530786221861537, + "learning_rate": 6.204894762070407e-07, + "loss": 0.251, + "step": 5126 + }, + { + "epoch": 1.7118530884808014, + "grad_norm": 0.44818596082983636, + "learning_rate": 6.190843736285568e-07, + "loss": 0.258, + "step": 5127 + }, + { + "epoch": 1.712186978297162, + "grad_norm": 0.44206755123540764, + "learning_rate": 6.176807587863681e-07, + "loss": 0.2292, + "step": 5128 + }, + { + "epoch": 1.7125208681135224, + "grad_norm": 0.43607094375285, + "learning_rate": 6.162786321571351e-07, + "loss": 0.2424, + "step": 5129 + }, + { + "epoch": 1.7128547579298832, + "grad_norm": 0.4441977007525159, + "learning_rate": 6.148779942170141e-07, + "loss": 0.2467, + "step": 5130 + }, + { + "epoch": 1.7131886477462437, + "grad_norm": 0.4512538933503769, + "learning_rate": 6.134788454416529e-07, + "loss": 0.2542, + "step": 5131 + }, + { + "epoch": 1.7135225375626044, + "grad_norm": 0.45281148249916464, + "learning_rate": 6.120811863061965e-07, + "loss": 0.2546, + "step": 5132 + }, + { + "epoch": 1.713856427378965, + "grad_norm": 0.4406824468430129, + "learning_rate": 6.106850172852819e-07, + "loss": 0.2442, + "step": 5133 + }, + { + "epoch": 1.7141903171953254, + "grad_norm": 0.4471205577832461, + "learning_rate": 6.092903388530413e-07, + "loss": 0.2438, + "step": 5134 + }, + { + "epoch": 1.7145242070116862, + "grad_norm": 0.4551475997145213, + "learning_rate": 6.078971514830989e-07, + "loss": 0.2524, + "step": 5135 + }, + { + "epoch": 1.714858096828047, + "grad_norm": 0.43144890328908153, + "learning_rate": 6.065054556485761e-07, + "loss": 0.232, + "step": 5136 + }, + { + "epoch": 1.7151919866444074, + "grad_norm": 0.46003162887880267, + "learning_rate": 6.051152518220838e-07, + "loss": 0.2446, + "step": 5137 + }, + { + "epoch": 1.715525876460768, + "grad_norm": 0.43758656959871606, + "learning_rate": 6.037265404757292e-07, + "loss": 0.2413, + "step": 5138 + }, + { + "epoch": 1.7158597662771284, + "grad_norm": 0.4533517691711473, + "learning_rate": 6.023393220811119e-07, + "loss": 0.2575, + "step": 5139 + }, + { + "epoch": 1.7161936560934892, + "grad_norm": 0.4682554171486951, + "learning_rate": 6.009535971093216e-07, + "loss": 0.2506, + "step": 5140 + }, + { + "epoch": 1.71652754590985, + "grad_norm": 0.45642153677602704, + "learning_rate": 5.995693660309466e-07, + "loss": 0.2548, + "step": 5141 + }, + { + "epoch": 1.7168614357262104, + "grad_norm": 0.45238342808865195, + "learning_rate": 5.981866293160627e-07, + "loss": 0.2454, + "step": 5142 + }, + { + "epoch": 1.717195325542571, + "grad_norm": 0.4686176430359024, + "learning_rate": 5.968053874342405e-07, + "loss": 0.2481, + "step": 5143 + }, + { + "epoch": 1.7175292153589314, + "grad_norm": 0.4579923874843899, + "learning_rate": 5.954256408545417e-07, + "loss": 0.2465, + "step": 5144 + }, + { + "epoch": 1.7178631051752922, + "grad_norm": 0.4512859372174023, + "learning_rate": 5.94047390045523e-07, + "loss": 0.2425, + "step": 5145 + }, + { + "epoch": 1.718196994991653, + "grad_norm": 0.4648009476634029, + "learning_rate": 5.926706354752293e-07, + "loss": 0.2535, + "step": 5146 + }, + { + "epoch": 1.7185308848080134, + "grad_norm": 0.44798178149150847, + "learning_rate": 5.912953776112007e-07, + "loss": 0.2406, + "step": 5147 + }, + { + "epoch": 1.718864774624374, + "grad_norm": 0.45421673303505533, + "learning_rate": 5.89921616920468e-07, + "loss": 0.2454, + "step": 5148 + }, + { + "epoch": 1.7191986644407344, + "grad_norm": 0.450963911470753, + "learning_rate": 5.885493538695508e-07, + "loss": 0.2459, + "step": 5149 + }, + { + "epoch": 1.7195325542570952, + "grad_norm": 0.4322090624997725, + "learning_rate": 5.871785889244657e-07, + "loss": 0.2485, + "step": 5150 + }, + { + "epoch": 1.719866444073456, + "grad_norm": 0.4511616153940949, + "learning_rate": 5.85809322550716e-07, + "loss": 0.2499, + "step": 5151 + }, + { + "epoch": 1.7202003338898164, + "grad_norm": 0.45021625174001567, + "learning_rate": 5.84441555213297e-07, + "loss": 0.251, + "step": 5152 + }, + { + "epoch": 1.720534223706177, + "grad_norm": 0.4517931639153106, + "learning_rate": 5.830752873766948e-07, + "loss": 0.2442, + "step": 5153 + }, + { + "epoch": 1.7208681135225374, + "grad_norm": 0.43905482667022927, + "learning_rate": 5.817105195048878e-07, + "loss": 0.2371, + "step": 5154 + }, + { + "epoch": 1.7212020033388982, + "grad_norm": 0.4516818827916672, + "learning_rate": 5.803472520613457e-07, + "loss": 0.2536, + "step": 5155 + }, + { + "epoch": 1.721535893155259, + "grad_norm": 0.46664640381565065, + "learning_rate": 5.789854855090249e-07, + "loss": 0.2372, + "step": 5156 + }, + { + "epoch": 1.7218697829716194, + "grad_norm": 0.4583828906583453, + "learning_rate": 5.776252203103755e-07, + "loss": 0.2418, + "step": 5157 + }, + { + "epoch": 1.72220367278798, + "grad_norm": 0.4501496234419443, + "learning_rate": 5.762664569273346e-07, + "loss": 0.2473, + "step": 5158 + }, + { + "epoch": 1.7225375626043404, + "grad_norm": 0.4500294996520143, + "learning_rate": 5.749091958213338e-07, + "loss": 0.2383, + "step": 5159 + }, + { + "epoch": 1.7228714524207012, + "grad_norm": 0.4662367349551767, + "learning_rate": 5.735534374532908e-07, + "loss": 0.245, + "step": 5160 + }, + { + "epoch": 1.723205342237062, + "grad_norm": 0.44519549692909044, + "learning_rate": 5.721991822836137e-07, + "loss": 0.2429, + "step": 5161 + }, + { + "epoch": 1.7235392320534224, + "grad_norm": 0.45725842647927223, + "learning_rate": 5.708464307722006e-07, + "loss": 0.2462, + "step": 5162 + }, + { + "epoch": 1.723873121869783, + "grad_norm": 0.4483426556062558, + "learning_rate": 5.694951833784384e-07, + "loss": 0.2373, + "step": 5163 + }, + { + "epoch": 1.7242070116861434, + "grad_norm": 0.45147019214388456, + "learning_rate": 5.681454405612063e-07, + "loss": 0.2388, + "step": 5164 + }, + { + "epoch": 1.7245409015025042, + "grad_norm": 0.43848189650429586, + "learning_rate": 5.667972027788682e-07, + "loss": 0.2437, + "step": 5165 + }, + { + "epoch": 1.724874791318865, + "grad_norm": 0.45345338664057505, + "learning_rate": 5.654504704892782e-07, + "loss": 0.2425, + "step": 5166 + }, + { + "epoch": 1.7252086811352254, + "grad_norm": 0.4652744341149004, + "learning_rate": 5.641052441497797e-07, + "loss": 0.2455, + "step": 5167 + }, + { + "epoch": 1.725542570951586, + "grad_norm": 0.461824794049695, + "learning_rate": 5.627615242172057e-07, + "loss": 0.2494, + "step": 5168 + }, + { + "epoch": 1.7258764607679464, + "grad_norm": 0.4671197708993026, + "learning_rate": 5.614193111478761e-07, + "loss": 0.2552, + "step": 5169 + }, + { + "epoch": 1.7262103505843072, + "grad_norm": 0.45412870658793314, + "learning_rate": 5.600786053975993e-07, + "loss": 0.2433, + "step": 5170 + }, + { + "epoch": 1.726544240400668, + "grad_norm": 0.4496294880592402, + "learning_rate": 5.587394074216712e-07, + "loss": 0.2373, + "step": 5171 + }, + { + "epoch": 1.7268781302170284, + "grad_norm": 0.4644753396080483, + "learning_rate": 5.574017176748769e-07, + "loss": 0.2678, + "step": 5172 + }, + { + "epoch": 1.727212020033389, + "grad_norm": 0.45475833913514957, + "learning_rate": 5.560655366114909e-07, + "loss": 0.2531, + "step": 5173 + }, + { + "epoch": 1.7275459098497494, + "grad_norm": 0.45536764831911836, + "learning_rate": 5.547308646852717e-07, + "loss": 0.2491, + "step": 5174 + }, + { + "epoch": 1.7278797996661102, + "grad_norm": 0.4462483397433719, + "learning_rate": 5.533977023494675e-07, + "loss": 0.249, + "step": 5175 + }, + { + "epoch": 1.728213689482471, + "grad_norm": 0.45617452049089385, + "learning_rate": 5.520660500568115e-07, + "loss": 0.2482, + "step": 5176 + }, + { + "epoch": 1.7285475792988314, + "grad_norm": 0.4650271069487119, + "learning_rate": 5.507359082595287e-07, + "loss": 0.2434, + "step": 5177 + }, + { + "epoch": 1.728881469115192, + "grad_norm": 0.4589942381720827, + "learning_rate": 5.494072774093273e-07, + "loss": 0.2478, + "step": 5178 + }, + { + "epoch": 1.7292153589315524, + "grad_norm": 0.46951673124863796, + "learning_rate": 5.48080157957403e-07, + "loss": 0.2617, + "step": 5179 + }, + { + "epoch": 1.7295492487479132, + "grad_norm": 0.4520058586565226, + "learning_rate": 5.467545503544381e-07, + "loss": 0.24, + "step": 5180 + }, + { + "epoch": 1.729883138564274, + "grad_norm": 0.4584700789046871, + "learning_rate": 5.454304550506029e-07, + "loss": 0.2446, + "step": 5181 + }, + { + "epoch": 1.7302170283806344, + "grad_norm": 0.44709492400106887, + "learning_rate": 5.441078724955545e-07, + "loss": 0.2422, + "step": 5182 + }, + { + "epoch": 1.730550918196995, + "grad_norm": 0.4363370153343696, + "learning_rate": 5.427868031384342e-07, + "loss": 0.2457, + "step": 5183 + }, + { + "epoch": 1.7308848080133556, + "grad_norm": 0.4864831659259961, + "learning_rate": 5.414672474278698e-07, + "loss": 0.2551, + "step": 5184 + }, + { + "epoch": 1.7312186978297162, + "grad_norm": 0.45665151977461976, + "learning_rate": 5.401492058119751e-07, + "loss": 0.245, + "step": 5185 + }, + { + "epoch": 1.731552587646077, + "grad_norm": 0.45511171409575757, + "learning_rate": 5.388326787383518e-07, + "loss": 0.2627, + "step": 5186 + }, + { + "epoch": 1.7318864774624374, + "grad_norm": 0.44061113236866273, + "learning_rate": 5.375176666540848e-07, + "loss": 0.2405, + "step": 5187 + }, + { + "epoch": 1.732220367278798, + "grad_norm": 0.4586612445517302, + "learning_rate": 5.362041700057452e-07, + "loss": 0.2482, + "step": 5188 + }, + { + "epoch": 1.7325542570951586, + "grad_norm": 0.4718340166416909, + "learning_rate": 5.348921892393905e-07, + "loss": 0.2551, + "step": 5189 + }, + { + "epoch": 1.7328881469115192, + "grad_norm": 0.4458814985832397, + "learning_rate": 5.33581724800561e-07, + "loss": 0.2514, + "step": 5190 + }, + { + "epoch": 1.7332220367278799, + "grad_norm": 0.45926010651770194, + "learning_rate": 5.322727771342856e-07, + "loss": 0.2537, + "step": 5191 + }, + { + "epoch": 1.7335559265442404, + "grad_norm": 0.439224589992814, + "learning_rate": 5.309653466850756e-07, + "loss": 0.249, + "step": 5192 + }, + { + "epoch": 1.733889816360601, + "grad_norm": 0.4317070430764539, + "learning_rate": 5.296594338969274e-07, + "loss": 0.2477, + "step": 5193 + }, + { + "epoch": 1.7342237061769616, + "grad_norm": 0.4391288458441448, + "learning_rate": 5.283550392133208e-07, + "loss": 0.2397, + "step": 5194 + }, + { + "epoch": 1.7345575959933222, + "grad_norm": 0.45316205738428733, + "learning_rate": 5.270521630772241e-07, + "loss": 0.2484, + "step": 5195 + }, + { + "epoch": 1.7348914858096829, + "grad_norm": 0.44176478683844406, + "learning_rate": 5.257508059310862e-07, + "loss": 0.2377, + "step": 5196 + }, + { + "epoch": 1.7352253756260434, + "grad_norm": 0.45004275863836735, + "learning_rate": 5.244509682168397e-07, + "loss": 0.2415, + "step": 5197 + }, + { + "epoch": 1.735559265442404, + "grad_norm": 0.4426838947368507, + "learning_rate": 5.231526503759055e-07, + "loss": 0.245, + "step": 5198 + }, + { + "epoch": 1.7358931552587646, + "grad_norm": 0.43978214361425827, + "learning_rate": 5.218558528491829e-07, + "loss": 0.2393, + "step": 5199 + }, + { + "epoch": 1.7362270450751254, + "grad_norm": 0.46483708479666364, + "learning_rate": 5.205605760770605e-07, + "loss": 0.2502, + "step": 5200 + }, + { + "epoch": 1.7365609348914859, + "grad_norm": 0.45387019141049717, + "learning_rate": 5.192668204994056e-07, + "loss": 0.244, + "step": 5201 + }, + { + "epoch": 1.7368948247078464, + "grad_norm": 0.47030824053315257, + "learning_rate": 5.179745865555713e-07, + "loss": 0.2482, + "step": 5202 + }, + { + "epoch": 1.737228714524207, + "grad_norm": 0.466627288325223, + "learning_rate": 5.166838746843933e-07, + "loss": 0.2502, + "step": 5203 + }, + { + "epoch": 1.7375626043405676, + "grad_norm": 0.4426691048954312, + "learning_rate": 5.153946853241914e-07, + "loss": 0.2431, + "step": 5204 + }, + { + "epoch": 1.7378964941569284, + "grad_norm": 0.46966344174805097, + "learning_rate": 5.141070189127668e-07, + "loss": 0.2581, + "step": 5205 + }, + { + "epoch": 1.7382303839732889, + "grad_norm": 0.4620097490554972, + "learning_rate": 5.128208758874059e-07, + "loss": 0.2412, + "step": 5206 + }, + { + "epoch": 1.7385642737896494, + "grad_norm": 0.4583352703558653, + "learning_rate": 5.115362566848747e-07, + "loss": 0.2469, + "step": 5207 + }, + { + "epoch": 1.73889816360601, + "grad_norm": 0.45370732402262925, + "learning_rate": 5.10253161741423e-07, + "loss": 0.2403, + "step": 5208 + }, + { + "epoch": 1.7392320534223706, + "grad_norm": 0.4555434108870677, + "learning_rate": 5.089715914927851e-07, + "loss": 0.2469, + "step": 5209 + }, + { + "epoch": 1.7395659432387314, + "grad_norm": 0.4563856750546455, + "learning_rate": 5.076915463741744e-07, + "loss": 0.2515, + "step": 5210 + }, + { + "epoch": 1.7398998330550919, + "grad_norm": 0.44523813640016274, + "learning_rate": 5.064130268202877e-07, + "loss": 0.2459, + "step": 5211 + }, + { + "epoch": 1.7402337228714524, + "grad_norm": 0.480937576188989, + "learning_rate": 5.051360332653032e-07, + "loss": 0.258, + "step": 5212 + }, + { + "epoch": 1.740567612687813, + "grad_norm": 0.44483119267970567, + "learning_rate": 5.038605661428809e-07, + "loss": 0.2505, + "step": 5213 + }, + { + "epoch": 1.7409015025041736, + "grad_norm": 0.45828297468504103, + "learning_rate": 5.025866258861639e-07, + "loss": 0.2362, + "step": 5214 + }, + { + "epoch": 1.7412353923205344, + "grad_norm": 0.4734235981178256, + "learning_rate": 5.013142129277759e-07, + "loss": 0.2501, + "step": 5215 + }, + { + "epoch": 1.7415692821368949, + "grad_norm": 0.4612709137028769, + "learning_rate": 5.000433276998218e-07, + "loss": 0.2543, + "step": 5216 + }, + { + "epoch": 1.7419031719532554, + "grad_norm": 0.44812240577941703, + "learning_rate": 4.987739706338862e-07, + "loss": 0.2407, + "step": 5217 + }, + { + "epoch": 1.742237061769616, + "grad_norm": 0.4595428150215036, + "learning_rate": 4.975061421610377e-07, + "loss": 0.2532, + "step": 5218 + }, + { + "epoch": 1.7425709515859766, + "grad_norm": 0.46877077848426607, + "learning_rate": 4.962398427118243e-07, + "loss": 0.2538, + "step": 5219 + }, + { + "epoch": 1.7429048414023374, + "grad_norm": 0.46023912402463996, + "learning_rate": 4.949750727162738e-07, + "loss": 0.2488, + "step": 5220 + }, + { + "epoch": 1.7432387312186979, + "grad_norm": 0.4379374344062306, + "learning_rate": 4.937118326038964e-07, + "loss": 0.2479, + "step": 5221 + }, + { + "epoch": 1.7435726210350584, + "grad_norm": 0.4562956067362719, + "learning_rate": 4.924501228036793e-07, + "loss": 0.2496, + "step": 5222 + }, + { + "epoch": 1.743906510851419, + "grad_norm": 0.4313995512676906, + "learning_rate": 4.911899437440964e-07, + "loss": 0.2343, + "step": 5223 + }, + { + "epoch": 1.7442404006677796, + "grad_norm": 0.46058930808464493, + "learning_rate": 4.899312958530966e-07, + "loss": 0.2537, + "step": 5224 + }, + { + "epoch": 1.7445742904841404, + "grad_norm": 0.4452358697005718, + "learning_rate": 4.886741795581101e-07, + "loss": 0.2486, + "step": 5225 + }, + { + "epoch": 1.7449081803005009, + "grad_norm": 0.4475260620927993, + "learning_rate": 4.874185952860471e-07, + "loss": 0.2494, + "step": 5226 + }, + { + "epoch": 1.7452420701168614, + "grad_norm": 0.44266182255161785, + "learning_rate": 4.861645434632967e-07, + "loss": 0.2366, + "step": 5227 + }, + { + "epoch": 1.745575959933222, + "grad_norm": 0.44372114566271653, + "learning_rate": 4.849120245157296e-07, + "loss": 0.2515, + "step": 5228 + }, + { + "epoch": 1.7459098497495826, + "grad_norm": 0.4565062068337323, + "learning_rate": 4.836610388686947e-07, + "loss": 0.2503, + "step": 5229 + }, + { + "epoch": 1.7462437395659434, + "grad_norm": 0.4792705662318239, + "learning_rate": 4.824115869470197e-07, + "loss": 0.2522, + "step": 5230 + }, + { + "epoch": 1.7465776293823039, + "grad_norm": 0.45017662688912047, + "learning_rate": 4.81163669175011e-07, + "loss": 0.2427, + "step": 5231 + }, + { + "epoch": 1.7469115191986644, + "grad_norm": 0.43367828974700057, + "learning_rate": 4.799172859764567e-07, + "loss": 0.24, + "step": 5232 + }, + { + "epoch": 1.747245409015025, + "grad_norm": 0.44563791280971327, + "learning_rate": 4.786724377746221e-07, + "loss": 0.2449, + "step": 5233 + }, + { + "epoch": 1.7475792988313856, + "grad_norm": 0.4495528021657842, + "learning_rate": 4.774291249922508e-07, + "loss": 0.249, + "step": 5234 + }, + { + "epoch": 1.7479131886477464, + "grad_norm": 0.45129376058814646, + "learning_rate": 4.76187348051565e-07, + "loss": 0.2343, + "step": 5235 + }, + { + "epoch": 1.7482470784641069, + "grad_norm": 0.4716836954188935, + "learning_rate": 4.749471073742656e-07, + "loss": 0.2541, + "step": 5236 + }, + { + "epoch": 1.7485809682804674, + "grad_norm": 0.46720374705290807, + "learning_rate": 4.7370840338153313e-07, + "loss": 0.2594, + "step": 5237 + }, + { + "epoch": 1.748914858096828, + "grad_norm": 0.44058287787410083, + "learning_rate": 4.7247123649402427e-07, + "loss": 0.2365, + "step": 5238 + }, + { + "epoch": 1.7492487479131886, + "grad_norm": 0.44017650665016017, + "learning_rate": 4.7123560713187353e-07, + "loss": 0.2425, + "step": 5239 + }, + { + "epoch": 1.7495826377295494, + "grad_norm": 0.4652617256098495, + "learning_rate": 4.700015157146959e-07, + "loss": 0.2606, + "step": 5240 + }, + { + "epoch": 1.7499165275459099, + "grad_norm": 0.4547482025203008, + "learning_rate": 4.68768962661581e-07, + "loss": 0.248, + "step": 5241 + }, + { + "epoch": 1.7502504173622704, + "grad_norm": 0.42685829568012956, + "learning_rate": 4.6753794839109966e-07, + "loss": 0.2404, + "step": 5242 + }, + { + "epoch": 1.750584307178631, + "grad_norm": 0.4587270787016151, + "learning_rate": 4.663084733212958e-07, + "loss": 0.2498, + "step": 5243 + }, + { + "epoch": 1.7509181969949916, + "grad_norm": 0.45534478219437075, + "learning_rate": 4.6508053786969375e-07, + "loss": 0.2443, + "step": 5244 + }, + { + "epoch": 1.7512520868113524, + "grad_norm": 0.44990688967005216, + "learning_rate": 4.638541424532933e-07, + "loss": 0.2429, + "step": 5245 + }, + { + "epoch": 1.7515859766277129, + "grad_norm": 0.43734437216484123, + "learning_rate": 4.6262928748857296e-07, + "loss": 0.2461, + "step": 5246 + }, + { + "epoch": 1.7519198664440734, + "grad_norm": 0.4710581780130505, + "learning_rate": 4.614059733914872e-07, + "loss": 0.2585, + "step": 5247 + }, + { + "epoch": 1.752253756260434, + "grad_norm": 0.4581966631475156, + "learning_rate": 4.6018420057746537e-07, + "loss": 0.2539, + "step": 5248 + }, + { + "epoch": 1.7525876460767946, + "grad_norm": 0.44337637962875964, + "learning_rate": 4.5896396946141784e-07, + "loss": 0.2397, + "step": 5249 + }, + { + "epoch": 1.7529215358931554, + "grad_norm": 0.44661926072139496, + "learning_rate": 4.577452804577259e-07, + "loss": 0.238, + "step": 5250 + }, + { + "epoch": 1.7532554257095159, + "grad_norm": 0.4485954405708441, + "learning_rate": 4.56528133980253e-07, + "loss": 0.2463, + "step": 5251 + }, + { + "epoch": 1.7535893155258764, + "grad_norm": 0.47039469263679523, + "learning_rate": 4.5531253044233393e-07, + "loss": 0.2509, + "step": 5252 + }, + { + "epoch": 1.7539232053422371, + "grad_norm": 0.44222338851529275, + "learning_rate": 4.5409847025678246e-07, + "loss": 0.246, + "step": 5253 + }, + { + "epoch": 1.7542570951585976, + "grad_norm": 0.4545987735286972, + "learning_rate": 4.5288595383588484e-07, + "loss": 0.2475, + "step": 5254 + }, + { + "epoch": 1.7545909849749584, + "grad_norm": 0.4468175692550485, + "learning_rate": 4.5167498159140845e-07, + "loss": 0.25, + "step": 5255 + }, + { + "epoch": 1.7549248747913189, + "grad_norm": 0.45074224642771604, + "learning_rate": 4.504655539345909e-07, + "loss": 0.2464, + "step": 5256 + }, + { + "epoch": 1.7552587646076794, + "grad_norm": 0.44879228608171545, + "learning_rate": 4.4925767127614927e-07, + "loss": 0.2536, + "step": 5257 + }, + { + "epoch": 1.7555926544240401, + "grad_norm": 0.44026952758780385, + "learning_rate": 4.4805133402627323e-07, + "loss": 0.2448, + "step": 5258 + }, + { + "epoch": 1.7559265442404006, + "grad_norm": 0.4540118202785224, + "learning_rate": 4.4684654259462836e-07, + "loss": 0.2401, + "step": 5259 + }, + { + "epoch": 1.7562604340567614, + "grad_norm": 0.4445528999172309, + "learning_rate": 4.4564329739035684e-07, + "loss": 0.2516, + "step": 5260 + }, + { + "epoch": 1.7565943238731219, + "grad_norm": 0.4475672121713866, + "learning_rate": 4.4444159882207406e-07, + "loss": 0.2313, + "step": 5261 + }, + { + "epoch": 1.7569282136894824, + "grad_norm": 0.4540625145194913, + "learning_rate": 4.432414472978702e-07, + "loss": 0.2513, + "step": 5262 + }, + { + "epoch": 1.757262103505843, + "grad_norm": 0.43648965797524486, + "learning_rate": 4.4204284322530933e-07, + "loss": 0.2393, + "step": 5263 + }, + { + "epoch": 1.7575959933222036, + "grad_norm": 0.46050692386228487, + "learning_rate": 4.408457870114341e-07, + "loss": 0.2453, + "step": 5264 + }, + { + "epoch": 1.7579298831385644, + "grad_norm": 0.43837043256690683, + "learning_rate": 4.3965027906275604e-07, + "loss": 0.2464, + "step": 5265 + }, + { + "epoch": 1.7582637729549249, + "grad_norm": 0.42946552666042853, + "learning_rate": 4.3845631978526536e-07, + "loss": 0.2331, + "step": 5266 + }, + { + "epoch": 1.7585976627712854, + "grad_norm": 0.4445892848227378, + "learning_rate": 4.3726390958442386e-07, + "loss": 0.2448, + "step": 5267 + }, + { + "epoch": 1.758931552587646, + "grad_norm": 0.45039309493794794, + "learning_rate": 4.3607304886516655e-07, + "loss": 0.2454, + "step": 5268 + }, + { + "epoch": 1.7592654424040068, + "grad_norm": 0.4433314438795035, + "learning_rate": 4.3488373803190545e-07, + "loss": 0.2507, + "step": 5269 + }, + { + "epoch": 1.7595993322203674, + "grad_norm": 0.4359487597552127, + "learning_rate": 4.336959774885241e-07, + "loss": 0.2434, + "step": 5270 + }, + { + "epoch": 1.7599332220367279, + "grad_norm": 0.4428147682307642, + "learning_rate": 4.325097676383788e-07, + "loss": 0.2415, + "step": 5271 + }, + { + "epoch": 1.7602671118530884, + "grad_norm": 0.4767398613131912, + "learning_rate": 4.3132510888430015e-07, + "loss": 0.2456, + "step": 5272 + }, + { + "epoch": 1.760601001669449, + "grad_norm": 0.45788697700690534, + "learning_rate": 4.3014200162859233e-07, + "loss": 0.2517, + "step": 5273 + }, + { + "epoch": 1.7609348914858098, + "grad_norm": 0.44569986124275035, + "learning_rate": 4.28960446273034e-07, + "loss": 0.2395, + "step": 5274 + }, + { + "epoch": 1.7612687813021703, + "grad_norm": 0.4528069283745822, + "learning_rate": 4.277804432188748e-07, + "loss": 0.2463, + "step": 5275 + }, + { + "epoch": 1.7616026711185309, + "grad_norm": 0.4582867385114381, + "learning_rate": 4.2660199286683636e-07, + "loss": 0.2483, + "step": 5276 + }, + { + "epoch": 1.7619365609348914, + "grad_norm": 0.42829654433225545, + "learning_rate": 4.2542509561711407e-07, + "loss": 0.2414, + "step": 5277 + }, + { + "epoch": 1.762270450751252, + "grad_norm": 0.45669418929991085, + "learning_rate": 4.242497518693783e-07, + "loss": 0.2554, + "step": 5278 + }, + { + "epoch": 1.7626043405676128, + "grad_norm": 0.46601334517339194, + "learning_rate": 4.230759620227681e-07, + "loss": 0.2565, + "step": 5279 + }, + { + "epoch": 1.7629382303839733, + "grad_norm": 0.4602281147651698, + "learning_rate": 4.2190372647589736e-07, + "loss": 0.2502, + "step": 5280 + }, + { + "epoch": 1.7632721202003339, + "grad_norm": 0.44177239109999916, + "learning_rate": 4.2073304562684934e-07, + "loss": 0.2453, + "step": 5281 + }, + { + "epoch": 1.7636060100166944, + "grad_norm": 0.44507546670099485, + "learning_rate": 4.1956391987318266e-07, + "loss": 0.2457, + "step": 5282 + }, + { + "epoch": 1.763939899833055, + "grad_norm": 0.43922562821619904, + "learning_rate": 4.1839634961192697e-07, + "loss": 0.2394, + "step": 5283 + }, + { + "epoch": 1.7642737896494158, + "grad_norm": 0.44659604704182776, + "learning_rate": 4.1723033523958235e-07, + "loss": 0.244, + "step": 5284 + }, + { + "epoch": 1.7646076794657763, + "grad_norm": 0.45273574945178885, + "learning_rate": 4.160658771521214e-07, + "loss": 0.2487, + "step": 5285 + }, + { + "epoch": 1.7649415692821369, + "grad_norm": 0.4479079190752488, + "learning_rate": 4.149029757449874e-07, + "loss": 0.2447, + "step": 5286 + }, + { + "epoch": 1.7652754590984974, + "grad_norm": 0.45069032461693564, + "learning_rate": 4.137416314130965e-07, + "loss": 0.25, + "step": 5287 + }, + { + "epoch": 1.765609348914858, + "grad_norm": 0.43863680465291, + "learning_rate": 4.125818445508351e-07, + "loss": 0.2445, + "step": 5288 + }, + { + "epoch": 1.7659432387312188, + "grad_norm": 0.4643076494810265, + "learning_rate": 4.1142361555206066e-07, + "loss": 0.2463, + "step": 5289 + }, + { + "epoch": 1.7662771285475793, + "grad_norm": 0.45466114921081185, + "learning_rate": 4.10266944810101e-07, + "loss": 0.2521, + "step": 5290 + }, + { + "epoch": 1.7666110183639399, + "grad_norm": 0.4581395067314624, + "learning_rate": 4.0911183271775625e-07, + "loss": 0.2544, + "step": 5291 + }, + { + "epoch": 1.7669449081803004, + "grad_norm": 0.44616587001165514, + "learning_rate": 4.079582796672976e-07, + "loss": 0.2351, + "step": 5292 + }, + { + "epoch": 1.767278797996661, + "grad_norm": 0.4561375852599334, + "learning_rate": 4.06806286050464e-07, + "loss": 0.2593, + "step": 5293 + }, + { + "epoch": 1.7676126878130218, + "grad_norm": 0.4387734383137366, + "learning_rate": 4.0565585225846806e-07, + "loss": 0.2357, + "step": 5294 + }, + { + "epoch": 1.7679465776293823, + "grad_norm": 0.46151869535049117, + "learning_rate": 4.0450697868198885e-07, + "loss": 0.2476, + "step": 5295 + }, + { + "epoch": 1.7682804674457429, + "grad_norm": 0.4445710201822182, + "learning_rate": 4.033596657111799e-07, + "loss": 0.2406, + "step": 5296 + }, + { + "epoch": 1.7686143572621034, + "grad_norm": 0.4330667783905077, + "learning_rate": 4.022139137356623e-07, + "loss": 0.236, + "step": 5297 + }, + { + "epoch": 1.768948247078464, + "grad_norm": 0.45886251763559127, + "learning_rate": 4.010697231445276e-07, + "loss": 0.2516, + "step": 5298 + }, + { + "epoch": 1.7692821368948248, + "grad_norm": 0.4377619128122364, + "learning_rate": 3.999270943263356e-07, + "loss": 0.2351, + "step": 5299 + }, + { + "epoch": 1.7696160267111853, + "grad_norm": 0.45287783025191813, + "learning_rate": 3.987860276691186e-07, + "loss": 0.2414, + "step": 5300 + }, + { + "epoch": 1.7699499165275459, + "grad_norm": 0.4521999634009879, + "learning_rate": 3.9764652356037715e-07, + "loss": 0.2481, + "step": 5301 + }, + { + "epoch": 1.7702838063439064, + "grad_norm": 0.4458130246046085, + "learning_rate": 3.965085823870807e-07, + "loss": 0.2436, + "step": 5302 + }, + { + "epoch": 1.770617696160267, + "grad_norm": 0.47124177644740656, + "learning_rate": 3.9537220453566784e-07, + "loss": 0.2437, + "step": 5303 + }, + { + "epoch": 1.7709515859766278, + "grad_norm": 0.4487843239135025, + "learning_rate": 3.9423739039204546e-07, + "loss": 0.2567, + "step": 5304 + }, + { + "epoch": 1.7712854757929883, + "grad_norm": 0.4350190182123882, + "learning_rate": 3.9310414034159306e-07, + "loss": 0.2398, + "step": 5305 + }, + { + "epoch": 1.7716193656093489, + "grad_norm": 0.4538348315953314, + "learning_rate": 3.9197245476915556e-07, + "loss": 0.2496, + "step": 5306 + }, + { + "epoch": 1.7719532554257094, + "grad_norm": 0.4549401213698778, + "learning_rate": 3.908423340590456e-07, + "loss": 0.2484, + "step": 5307 + }, + { + "epoch": 1.77228714524207, + "grad_norm": 0.4472741284578394, + "learning_rate": 3.8971377859504944e-07, + "loss": 0.2411, + "step": 5308 + }, + { + "epoch": 1.7726210350584308, + "grad_norm": 0.45139472818549936, + "learning_rate": 3.885867887604161e-07, + "loss": 0.2494, + "step": 5309 + }, + { + "epoch": 1.7729549248747913, + "grad_norm": 0.47090489539888053, + "learning_rate": 3.8746136493786725e-07, + "loss": 0.2508, + "step": 5310 + }, + { + "epoch": 1.7732888146911518, + "grad_norm": 0.46447661557580583, + "learning_rate": 3.8633750750959097e-07, + "loss": 0.2493, + "step": 5311 + }, + { + "epoch": 1.7736227045075124, + "grad_norm": 0.46364452892833075, + "learning_rate": 3.8521521685724204e-07, + "loss": 0.2512, + "step": 5312 + }, + { + "epoch": 1.773956594323873, + "grad_norm": 0.46434699671651736, + "learning_rate": 3.8409449336194505e-07, + "loss": 0.2587, + "step": 5313 + }, + { + "epoch": 1.7742904841402338, + "grad_norm": 0.4567393844101892, + "learning_rate": 3.829753374042927e-07, + "loss": 0.2493, + "step": 5314 + }, + { + "epoch": 1.7746243739565943, + "grad_norm": 0.4259107927972579, + "learning_rate": 3.818577493643444e-07, + "loss": 0.2354, + "step": 5315 + }, + { + "epoch": 1.7749582637729548, + "grad_norm": 0.4510461810666371, + "learning_rate": 3.807417296216254e-07, + "loss": 0.2467, + "step": 5316 + }, + { + "epoch": 1.7752921535893156, + "grad_norm": 0.4548329257751433, + "learning_rate": 3.7962727855513305e-07, + "loss": 0.2367, + "step": 5317 + }, + { + "epoch": 1.775626043405676, + "grad_norm": 0.4393975838954521, + "learning_rate": 3.785143965433269e-07, + "loss": 0.2374, + "step": 5318 + }, + { + "epoch": 1.7759599332220368, + "grad_norm": 0.4444140281284795, + "learning_rate": 3.7740308396413794e-07, + "loss": 0.2435, + "step": 5319 + }, + { + "epoch": 1.7762938230383973, + "grad_norm": 0.44765338154988105, + "learning_rate": 3.76293341194961e-07, + "loss": 0.2409, + "step": 5320 + }, + { + "epoch": 1.7766277128547578, + "grad_norm": 0.4575761013944382, + "learning_rate": 3.75185168612659e-07, + "loss": 0.2468, + "step": 5321 + }, + { + "epoch": 1.7769616026711186, + "grad_norm": 0.44379637574836966, + "learning_rate": 3.7407856659356147e-07, + "loss": 0.2421, + "step": 5322 + }, + { + "epoch": 1.777295492487479, + "grad_norm": 0.4641630733322113, + "learning_rate": 3.7297353551346507e-07, + "loss": 0.2571, + "step": 5323 + }, + { + "epoch": 1.7776293823038398, + "grad_norm": 0.45321115059890327, + "learning_rate": 3.718700757476323e-07, + "loss": 0.2488, + "step": 5324 + }, + { + "epoch": 1.7779632721202003, + "grad_norm": 0.4512429998427221, + "learning_rate": 3.7076818767079304e-07, + "loss": 0.2513, + "step": 5325 + }, + { + "epoch": 1.7782971619365608, + "grad_norm": 0.4496238875056177, + "learning_rate": 3.696678716571428e-07, + "loss": 0.2427, + "step": 5326 + }, + { + "epoch": 1.7786310517529216, + "grad_norm": 0.45565778091330156, + "learning_rate": 3.6856912808034164e-07, + "loss": 0.2392, + "step": 5327 + }, + { + "epoch": 1.778964941569282, + "grad_norm": 0.44674329544993635, + "learning_rate": 3.6747195731351935e-07, + "loss": 0.2394, + "step": 5328 + }, + { + "epoch": 1.7792988313856428, + "grad_norm": 0.4592614063860769, + "learning_rate": 3.6637635972926853e-07, + "loss": 0.2463, + "step": 5329 + }, + { + "epoch": 1.7796327212020033, + "grad_norm": 0.44511897734229616, + "learning_rate": 3.652823356996488e-07, + "loss": 0.2486, + "step": 5330 + }, + { + "epoch": 1.7799666110183638, + "grad_norm": 0.4437369593208335, + "learning_rate": 3.641898855961834e-07, + "loss": 0.241, + "step": 5331 + }, + { + "epoch": 1.7803005008347246, + "grad_norm": 0.46286724473198326, + "learning_rate": 3.6309900978986454e-07, + "loss": 0.2464, + "step": 5332 + }, + { + "epoch": 1.7806343906510853, + "grad_norm": 0.43745866441654296, + "learning_rate": 3.6200970865114705e-07, + "loss": 0.2383, + "step": 5333 + }, + { + "epoch": 1.7809682804674458, + "grad_norm": 0.45866815769256314, + "learning_rate": 3.6092198254995326e-07, + "loss": 0.2547, + "step": 5334 + }, + { + "epoch": 1.7813021702838063, + "grad_norm": 0.4366686227026687, + "learning_rate": 3.598358318556683e-07, + "loss": 0.2413, + "step": 5335 + }, + { + "epoch": 1.7816360601001668, + "grad_norm": 0.4551802475423785, + "learning_rate": 3.5875125693714265e-07, + "loss": 0.2433, + "step": 5336 + }, + { + "epoch": 1.7819699499165276, + "grad_norm": 0.4548653449621556, + "learning_rate": 3.576682581626939e-07, + "loss": 0.2329, + "step": 5337 + }, + { + "epoch": 1.7823038397328883, + "grad_norm": 0.4506868489691408, + "learning_rate": 3.5658683590010237e-07, + "loss": 0.2384, + "step": 5338 + }, + { + "epoch": 1.7826377295492488, + "grad_norm": 0.45402543743073825, + "learning_rate": 3.555069905166131e-07, + "loss": 0.2424, + "step": 5339 + }, + { + "epoch": 1.7829716193656093, + "grad_norm": 0.45908767677602275, + "learning_rate": 3.5442872237893554e-07, + "loss": 0.2551, + "step": 5340 + }, + { + "epoch": 1.7833055091819698, + "grad_norm": 0.47319309798196885, + "learning_rate": 3.5335203185324406e-07, + "loss": 0.254, + "step": 5341 + }, + { + "epoch": 1.7836393989983306, + "grad_norm": 0.43830854386573076, + "learning_rate": 3.5227691930517895e-07, + "loss": 0.2489, + "step": 5342 + }, + { + "epoch": 1.7839732888146913, + "grad_norm": 0.4670017184065928, + "learning_rate": 3.5120338509984143e-07, + "loss": 0.2442, + "step": 5343 + }, + { + "epoch": 1.7843071786310518, + "grad_norm": 0.43098767943734856, + "learning_rate": 3.5013142960179824e-07, + "loss": 0.2349, + "step": 5344 + }, + { + "epoch": 1.7846410684474123, + "grad_norm": 0.4688963334663138, + "learning_rate": 3.490610531750788e-07, + "loss": 0.2509, + "step": 5345 + }, + { + "epoch": 1.7849749582637728, + "grad_norm": 0.44713836110589134, + "learning_rate": 3.4799225618318e-07, + "loss": 0.2412, + "step": 5346 + }, + { + "epoch": 1.7853088480801336, + "grad_norm": 0.4666964773651079, + "learning_rate": 3.469250389890583e-07, + "loss": 0.2465, + "step": 5347 + }, + { + "epoch": 1.7856427378964943, + "grad_norm": 0.4504380834851406, + "learning_rate": 3.4585940195513537e-07, + "loss": 0.2484, + "step": 5348 + }, + { + "epoch": 1.7859766277128548, + "grad_norm": 0.4603126987421969, + "learning_rate": 3.447953454432951e-07, + "loss": 0.2425, + "step": 5349 + }, + { + "epoch": 1.7863105175292153, + "grad_norm": 0.4636486706036487, + "learning_rate": 3.437328698148862e-07, + "loss": 0.249, + "step": 5350 + }, + { + "epoch": 1.7866444073455758, + "grad_norm": 0.4473153259445572, + "learning_rate": 3.426719754307206e-07, + "loss": 0.2513, + "step": 5351 + }, + { + "epoch": 1.7869782971619366, + "grad_norm": 0.4537033354540247, + "learning_rate": 3.416126626510724e-07, + "loss": 0.253, + "step": 5352 + }, + { + "epoch": 1.7873121869782973, + "grad_norm": 0.4612826083471079, + "learning_rate": 3.405549318356788e-07, + "loss": 0.2457, + "step": 5353 + }, + { + "epoch": 1.7876460767946578, + "grad_norm": 0.4465812475083103, + "learning_rate": 3.3949878334373796e-07, + "loss": 0.2434, + "step": 5354 + }, + { + "epoch": 1.7879799666110183, + "grad_norm": 0.4577722281858029, + "learning_rate": 3.384442175339148e-07, + "loss": 0.2457, + "step": 5355 + }, + { + "epoch": 1.7883138564273788, + "grad_norm": 0.44548113715298526, + "learning_rate": 3.373912347643332e-07, + "loss": 0.2436, + "step": 5356 + }, + { + "epoch": 1.7886477462437396, + "grad_norm": 0.44442149902048117, + "learning_rate": 3.3633983539258065e-07, + "loss": 0.241, + "step": 5357 + }, + { + "epoch": 1.7889816360601003, + "grad_norm": 0.47166400580448276, + "learning_rate": 3.3529001977570695e-07, + "loss": 0.2471, + "step": 5358 + }, + { + "epoch": 1.7893155258764608, + "grad_norm": 0.45032543503942035, + "learning_rate": 3.342417882702237e-07, + "loss": 0.2509, + "step": 5359 + }, + { + "epoch": 1.7896494156928213, + "grad_norm": 0.44670703148445984, + "learning_rate": 3.331951412321066e-07, + "loss": 0.2384, + "step": 5360 + }, + { + "epoch": 1.7899833055091818, + "grad_norm": 0.4454423986283281, + "learning_rate": 3.321500790167903e-07, + "loss": 0.2344, + "step": 5361 + }, + { + "epoch": 1.7903171953255426, + "grad_norm": 0.42586454650046085, + "learning_rate": 3.311066019791731e-07, + "loss": 0.2415, + "step": 5362 + }, + { + "epoch": 1.7906510851419033, + "grad_norm": 0.4651640364622151, + "learning_rate": 3.3006471047361267e-07, + "loss": 0.2515, + "step": 5363 + }, + { + "epoch": 1.7909849749582638, + "grad_norm": 0.46111030671959613, + "learning_rate": 3.2902440485393204e-07, + "loss": 0.2518, + "step": 5364 + }, + { + "epoch": 1.7913188647746243, + "grad_norm": 0.4573201920732719, + "learning_rate": 3.2798568547341315e-07, + "loss": 0.2452, + "step": 5365 + }, + { + "epoch": 1.7916527545909848, + "grad_norm": 0.44651406664857196, + "learning_rate": 3.269485526847993e-07, + "loss": 0.2358, + "step": 5366 + }, + { + "epoch": 1.7919866444073456, + "grad_norm": 0.4580455529504217, + "learning_rate": 3.2591300684029494e-07, + "loss": 0.2427, + "step": 5367 + }, + { + "epoch": 1.7923205342237063, + "grad_norm": 0.44133246150877253, + "learning_rate": 3.2487904829156644e-07, + "loss": 0.2496, + "step": 5368 + }, + { + "epoch": 1.7926544240400668, + "grad_norm": 0.45518355066668936, + "learning_rate": 3.23846677389742e-07, + "loss": 0.2546, + "step": 5369 + }, + { + "epoch": 1.7929883138564273, + "grad_norm": 0.45854622411502455, + "learning_rate": 3.228158944854082e-07, + "loss": 0.2614, + "step": 5370 + }, + { + "epoch": 1.7933222036727878, + "grad_norm": 0.4568207611017781, + "learning_rate": 3.21786699928614e-07, + "loss": 0.2443, + "step": 5371 + }, + { + "epoch": 1.7936560934891486, + "grad_norm": 0.44383653081756635, + "learning_rate": 3.207590940688682e-07, + "loss": 0.2429, + "step": 5372 + }, + { + "epoch": 1.7939899833055093, + "grad_norm": 0.45800070653773917, + "learning_rate": 3.1973307725514047e-07, + "loss": 0.2403, + "step": 5373 + }, + { + "epoch": 1.7943238731218698, + "grad_norm": 0.44244397469678876, + "learning_rate": 3.187086498358616e-07, + "loss": 0.2449, + "step": 5374 + }, + { + "epoch": 1.7946577629382303, + "grad_norm": 0.4517216381944342, + "learning_rate": 3.1768581215892046e-07, + "loss": 0.2509, + "step": 5375 + }, + { + "epoch": 1.7949916527545908, + "grad_norm": 0.4534457632008208, + "learning_rate": 3.166645645716687e-07, + "loss": 0.2506, + "step": 5376 + }, + { + "epoch": 1.7953255425709516, + "grad_norm": 0.4507114400007608, + "learning_rate": 3.156449074209156e-07, + "loss": 0.2557, + "step": 5377 + }, + { + "epoch": 1.7956594323873123, + "grad_norm": 0.46397613434128454, + "learning_rate": 3.14626841052933e-07, + "loss": 0.2535, + "step": 5378 + }, + { + "epoch": 1.7959933222036728, + "grad_norm": 0.4598519963039894, + "learning_rate": 3.1361036581344985e-07, + "loss": 0.2502, + "step": 5379 + }, + { + "epoch": 1.7963272120200333, + "grad_norm": 0.4495504779793958, + "learning_rate": 3.125954820476562e-07, + "loss": 0.2507, + "step": 5380 + }, + { + "epoch": 1.7966611018363938, + "grad_norm": 0.4313906141112962, + "learning_rate": 3.1158219010020087e-07, + "loss": 0.2362, + "step": 5381 + }, + { + "epoch": 1.7969949916527546, + "grad_norm": 0.44309570556783806, + "learning_rate": 3.105704903151929e-07, + "loss": 0.2414, + "step": 5382 + }, + { + "epoch": 1.7973288814691153, + "grad_norm": 0.4523457003612914, + "learning_rate": 3.095603830362004e-07, + "loss": 0.2512, + "step": 5383 + }, + { + "epoch": 1.7976627712854758, + "grad_norm": 0.4619360566333035, + "learning_rate": 3.0855186860625054e-07, + "loss": 0.2522, + "step": 5384 + }, + { + "epoch": 1.7979966611018363, + "grad_norm": 0.4622627066809767, + "learning_rate": 3.075449473678299e-07, + "loss": 0.2418, + "step": 5385 + }, + { + "epoch": 1.798330550918197, + "grad_norm": 0.46805290642178066, + "learning_rate": 3.0653961966288283e-07, + "loss": 0.2533, + "step": 5386 + }, + { + "epoch": 1.7986644407345576, + "grad_norm": 0.4599956052505175, + "learning_rate": 3.0553588583281446e-07, + "loss": 0.2556, + "step": 5387 + }, + { + "epoch": 1.7989983305509183, + "grad_norm": 0.4594334883713135, + "learning_rate": 3.045337462184877e-07, + "loss": 0.2394, + "step": 5388 + }, + { + "epoch": 1.7993322203672788, + "grad_norm": 0.4539481420238498, + "learning_rate": 3.0353320116022267e-07, + "loss": 0.2511, + "step": 5389 + }, + { + "epoch": 1.7996661101836393, + "grad_norm": 0.442842883729244, + "learning_rate": 3.0253425099779963e-07, + "loss": 0.2376, + "step": 5390 + }, + { + "epoch": 1.8, + "grad_norm": 0.45568707825965193, + "learning_rate": 3.015368960704584e-07, + "loss": 0.2471, + "step": 5391 + }, + { + "epoch": 1.8003338898163606, + "grad_norm": 0.4475286742437988, + "learning_rate": 3.0054113671689365e-07, + "loss": 0.2519, + "step": 5392 + }, + { + "epoch": 1.8006677796327213, + "grad_norm": 0.45814368566987634, + "learning_rate": 2.995469732752615e-07, + "loss": 0.2527, + "step": 5393 + }, + { + "epoch": 1.8010016694490818, + "grad_norm": 0.45682921852804753, + "learning_rate": 2.985544060831741e-07, + "loss": 0.2526, + "step": 5394 + }, + { + "epoch": 1.8013355592654423, + "grad_norm": 0.4813541487524955, + "learning_rate": 2.9756343547770127e-07, + "loss": 0.2581, + "step": 5395 + }, + { + "epoch": 1.801669449081803, + "grad_norm": 0.4575725238790243, + "learning_rate": 2.965740617953733e-07, + "loss": 0.2512, + "step": 5396 + }, + { + "epoch": 1.8020033388981636, + "grad_norm": 0.45603548666323795, + "learning_rate": 2.9558628537217583e-07, + "loss": 0.2408, + "step": 5397 + }, + { + "epoch": 1.8023372287145243, + "grad_norm": 0.4498894062233273, + "learning_rate": 2.9460010654355165e-07, + "loss": 0.2465, + "step": 5398 + }, + { + "epoch": 1.8026711185308848, + "grad_norm": 0.45282140388707637, + "learning_rate": 2.936155256444023e-07, + "loss": 0.247, + "step": 5399 + }, + { + "epoch": 1.8030050083472453, + "grad_norm": 0.4380240191252072, + "learning_rate": 2.926325430090865e-07, + "loss": 0.2431, + "step": 5400 + }, + { + "epoch": 1.803338898163606, + "grad_norm": 0.44602775098973646, + "learning_rate": 2.916511589714216e-07, + "loss": 0.2386, + "step": 5401 + }, + { + "epoch": 1.8036727879799668, + "grad_norm": 0.46059083223349406, + "learning_rate": 2.9067137386467823e-07, + "loss": 0.2448, + "step": 5402 + }, + { + "epoch": 1.8040066777963273, + "grad_norm": 0.45301707168732647, + "learning_rate": 2.896931880215881e-07, + "loss": 0.2484, + "step": 5403 + }, + { + "epoch": 1.8043405676126878, + "grad_norm": 0.442612434753218, + "learning_rate": 2.8871660177433646e-07, + "loss": 0.2437, + "step": 5404 + }, + { + "epoch": 1.8046744574290483, + "grad_norm": 0.46394341927909033, + "learning_rate": 2.877416154545681e-07, + "loss": 0.2444, + "step": 5405 + }, + { + "epoch": 1.805008347245409, + "grad_norm": 0.4521373911555106, + "learning_rate": 2.8676822939338375e-07, + "loss": 0.2436, + "step": 5406 + }, + { + "epoch": 1.8053422370617698, + "grad_norm": 0.4557557624734399, + "learning_rate": 2.8579644392133896e-07, + "loss": 0.2563, + "step": 5407 + }, + { + "epoch": 1.8056761268781303, + "grad_norm": 0.46925550080087325, + "learning_rate": 2.8482625936844755e-07, + "loss": 0.2534, + "step": 5408 + }, + { + "epoch": 1.8060100166944908, + "grad_norm": 0.4569912656103889, + "learning_rate": 2.8385767606417814e-07, + "loss": 0.2433, + "step": 5409 + }, + { + "epoch": 1.8063439065108513, + "grad_norm": 0.42674618826084787, + "learning_rate": 2.828906943374593e-07, + "loss": 0.2407, + "step": 5410 + }, + { + "epoch": 1.806677796327212, + "grad_norm": 0.45330143800476447, + "learning_rate": 2.819253145166712e-07, + "loss": 0.2403, + "step": 5411 + }, + { + "epoch": 1.8070116861435728, + "grad_norm": 0.46675866279042977, + "learning_rate": 2.80961536929652e-07, + "loss": 0.2419, + "step": 5412 + }, + { + "epoch": 1.8073455759599333, + "grad_norm": 0.4428196489206015, + "learning_rate": 2.7999936190369615e-07, + "loss": 0.2426, + "step": 5413 + }, + { + "epoch": 1.8076794657762938, + "grad_norm": 0.4504974735423462, + "learning_rate": 2.7903878976555165e-07, + "loss": 0.253, + "step": 5414 + }, + { + "epoch": 1.8080133555926543, + "grad_norm": 0.45440152643261306, + "learning_rate": 2.7807982084142595e-07, + "loss": 0.2516, + "step": 5415 + }, + { + "epoch": 1.808347245409015, + "grad_norm": 0.44958389548654976, + "learning_rate": 2.7712245545697903e-07, + "loss": 0.2514, + "step": 5416 + }, + { + "epoch": 1.8086811352253758, + "grad_norm": 0.44362363189132903, + "learning_rate": 2.7616669393732643e-07, + "loss": 0.2456, + "step": 5417 + }, + { + "epoch": 1.8090150250417363, + "grad_norm": 0.4525329133490123, + "learning_rate": 2.7521253660704075e-07, + "loss": 0.2444, + "step": 5418 + }, + { + "epoch": 1.8093489148580968, + "grad_norm": 0.4796895899892305, + "learning_rate": 2.742599837901494e-07, + "loss": 0.2486, + "step": 5419 + }, + { + "epoch": 1.8096828046744573, + "grad_norm": 0.4533675819362655, + "learning_rate": 2.733090358101337e-07, + "loss": 0.2569, + "step": 5420 + }, + { + "epoch": 1.810016694490818, + "grad_norm": 0.4323798081528751, + "learning_rate": 2.723596929899308e-07, + "loss": 0.248, + "step": 5421 + }, + { + "epoch": 1.8103505843071788, + "grad_norm": 0.4417245154418586, + "learning_rate": 2.7141195565193237e-07, + "loss": 0.2349, + "step": 5422 + }, + { + "epoch": 1.8106844741235393, + "grad_norm": 0.45120665949156147, + "learning_rate": 2.704658241179847e-07, + "loss": 0.2435, + "step": 5423 + }, + { + "epoch": 1.8110183639398998, + "grad_norm": 0.46414532256036417, + "learning_rate": 2.6952129870939035e-07, + "loss": 0.247, + "step": 5424 + }, + { + "epoch": 1.8113522537562603, + "grad_norm": 0.45498390053642124, + "learning_rate": 2.685783797469049e-07, + "loss": 0.2557, + "step": 5425 + }, + { + "epoch": 1.811686143572621, + "grad_norm": 0.46355450240108553, + "learning_rate": 2.6763706755073726e-07, + "loss": 0.2571, + "step": 5426 + }, + { + "epoch": 1.8120200333889818, + "grad_norm": 0.44039552656086, + "learning_rate": 2.666973624405544e-07, + "loss": 0.2455, + "step": 5427 + }, + { + "epoch": 1.8123539232053423, + "grad_norm": 0.4496784372605509, + "learning_rate": 2.657592647354734e-07, + "loss": 0.2485, + "step": 5428 + }, + { + "epoch": 1.8126878130217028, + "grad_norm": 0.4444566506173863, + "learning_rate": 2.6482277475406935e-07, + "loss": 0.2362, + "step": 5429 + }, + { + "epoch": 1.8130217028380633, + "grad_norm": 0.46643522254196945, + "learning_rate": 2.6388789281436844e-07, + "loss": 0.249, + "step": 5430 + }, + { + "epoch": 1.813355592654424, + "grad_norm": 0.4753744840807337, + "learning_rate": 2.629546192338517e-07, + "loss": 0.2535, + "step": 5431 + }, + { + "epoch": 1.8136894824707848, + "grad_norm": 0.46378879332339296, + "learning_rate": 2.620229543294528e-07, + "loss": 0.2589, + "step": 5432 + }, + { + "epoch": 1.8140233722871453, + "grad_norm": 0.46259594436889334, + "learning_rate": 2.6109289841756257e-07, + "loss": 0.2449, + "step": 5433 + }, + { + "epoch": 1.8143572621035058, + "grad_norm": 0.44511869848320595, + "learning_rate": 2.601644518140217e-07, + "loss": 0.2435, + "step": 5434 + }, + { + "epoch": 1.8146911519198663, + "grad_norm": 0.4422144832106498, + "learning_rate": 2.5923761483412635e-07, + "loss": 0.2564, + "step": 5435 + }, + { + "epoch": 1.815025041736227, + "grad_norm": 0.4703354690702091, + "learning_rate": 2.583123877926258e-07, + "loss": 0.2562, + "step": 5436 + }, + { + "epoch": 1.8153589315525878, + "grad_norm": 0.46047556325965294, + "learning_rate": 2.573887710037215e-07, + "loss": 0.2533, + "step": 5437 + }, + { + "epoch": 1.8156928213689483, + "grad_norm": 0.4445908324400765, + "learning_rate": 2.5646676478107034e-07, + "loss": 0.2472, + "step": 5438 + }, + { + "epoch": 1.8160267111853088, + "grad_norm": 0.4510629464133699, + "learning_rate": 2.5554636943777965e-07, + "loss": 0.258, + "step": 5439 + }, + { + "epoch": 1.8163606010016693, + "grad_norm": 0.46132202488554014, + "learning_rate": 2.546275852864116e-07, + "loss": 0.2465, + "step": 5440 + }, + { + "epoch": 1.81669449081803, + "grad_norm": 0.45762655273256997, + "learning_rate": 2.5371041263897945e-07, + "loss": 0.243, + "step": 5441 + }, + { + "epoch": 1.8170283806343908, + "grad_norm": 0.44405297458678655, + "learning_rate": 2.527948518069517e-07, + "loss": 0.2484, + "step": 5442 + }, + { + "epoch": 1.8173622704507513, + "grad_norm": 0.4523822580250063, + "learning_rate": 2.51880903101247e-07, + "loss": 0.2528, + "step": 5443 + }, + { + "epoch": 1.8176961602671118, + "grad_norm": 0.4308537588650928, + "learning_rate": 2.509685668322387e-07, + "loss": 0.2272, + "step": 5444 + }, + { + "epoch": 1.8180300500834723, + "grad_norm": 0.4609131521945385, + "learning_rate": 2.500578433097511e-07, + "loss": 0.2467, + "step": 5445 + }, + { + "epoch": 1.818363939899833, + "grad_norm": 0.4662781295078944, + "learning_rate": 2.491487328430603e-07, + "loss": 0.2419, + "step": 5446 + }, + { + "epoch": 1.8186978297161938, + "grad_norm": 0.4349020376786924, + "learning_rate": 2.482412357408964e-07, + "loss": 0.2495, + "step": 5447 + }, + { + "epoch": 1.8190317195325543, + "grad_norm": 0.44463625852705096, + "learning_rate": 2.473353523114408e-07, + "loss": 0.2485, + "step": 5448 + }, + { + "epoch": 1.8193656093489148, + "grad_norm": 0.4472485007489611, + "learning_rate": 2.464310828623262e-07, + "loss": 0.2516, + "step": 5449 + }, + { + "epoch": 1.8196994991652755, + "grad_norm": 0.4765467973527427, + "learning_rate": 2.4552842770063756e-07, + "loss": 0.2576, + "step": 5450 + }, + { + "epoch": 1.820033388981636, + "grad_norm": 0.4368925266209682, + "learning_rate": 2.44627387132913e-07, + "loss": 0.2384, + "step": 5451 + }, + { + "epoch": 1.8203672787979968, + "grad_norm": 0.44800418782291535, + "learning_rate": 2.4372796146514056e-07, + "loss": 0.2459, + "step": 5452 + }, + { + "epoch": 1.8207011686143573, + "grad_norm": 0.44966009207110486, + "learning_rate": 2.428301510027614e-07, + "loss": 0.2409, + "step": 5453 + }, + { + "epoch": 1.8210350584307178, + "grad_norm": 0.4429186916871473, + "learning_rate": 2.419339560506667e-07, + "loss": 0.2434, + "step": 5454 + }, + { + "epoch": 1.8213689482470785, + "grad_norm": 0.44020677072900866, + "learning_rate": 2.4103937691319856e-07, + "loss": 0.2441, + "step": 5455 + }, + { + "epoch": 1.821702838063439, + "grad_norm": 0.47770480795570974, + "learning_rate": 2.401464138941528e-07, + "loss": 0.2456, + "step": 5456 + }, + { + "epoch": 1.8220367278797998, + "grad_norm": 0.4517527295217175, + "learning_rate": 2.392550672967753e-07, + "loss": 0.2481, + "step": 5457 + }, + { + "epoch": 1.8223706176961603, + "grad_norm": 0.4618685304551235, + "learning_rate": 2.3836533742376212e-07, + "loss": 0.2523, + "step": 5458 + }, + { + "epoch": 1.8227045075125208, + "grad_norm": 0.4444006751089145, + "learning_rate": 2.3747722457725996e-07, + "loss": 0.2475, + "step": 5459 + }, + { + "epoch": 1.8230383973288815, + "grad_norm": 0.454197561495216, + "learning_rate": 2.3659072905886815e-07, + "loss": 0.2534, + "step": 5460 + }, + { + "epoch": 1.823372287145242, + "grad_norm": 0.4433835333926558, + "learning_rate": 2.357058511696375e-07, + "loss": 0.2459, + "step": 5461 + }, + { + "epoch": 1.8237061769616028, + "grad_norm": 0.46727487651992083, + "learning_rate": 2.34822591210066e-07, + "loss": 0.2431, + "step": 5462 + }, + { + "epoch": 1.8240400667779633, + "grad_norm": 0.4379356202832459, + "learning_rate": 2.3394094948010539e-07, + "loss": 0.2348, + "step": 5463 + }, + { + "epoch": 1.8243739565943238, + "grad_norm": 0.4469526849702228, + "learning_rate": 2.33060926279155e-07, + "loss": 0.2468, + "step": 5464 + }, + { + "epoch": 1.8247078464106845, + "grad_norm": 0.4657246539407155, + "learning_rate": 2.3218252190606804e-07, + "loss": 0.2518, + "step": 5465 + }, + { + "epoch": 1.8250417362270452, + "grad_norm": 0.45467877397215806, + "learning_rate": 2.313057366591448e-07, + "loss": 0.2457, + "step": 5466 + }, + { + "epoch": 1.8253756260434058, + "grad_norm": 0.44386994567730476, + "learning_rate": 2.3043057083613818e-07, + "loss": 0.2503, + "step": 5467 + }, + { + "epoch": 1.8257095158597663, + "grad_norm": 0.46409313414838715, + "learning_rate": 2.2955702473424824e-07, + "loss": 0.2597, + "step": 5468 + }, + { + "epoch": 1.8260434056761268, + "grad_norm": 0.447189308625288, + "learning_rate": 2.2868509865012767e-07, + "loss": 0.2352, + "step": 5469 + }, + { + "epoch": 1.8263772954924875, + "grad_norm": 0.4567040526902787, + "learning_rate": 2.2781479287987908e-07, + "loss": 0.2448, + "step": 5470 + }, + { + "epoch": 1.8267111853088482, + "grad_norm": 0.46911419604745647, + "learning_rate": 2.2694610771905267e-07, + "loss": 0.2564, + "step": 5471 + }, + { + "epoch": 1.8270450751252088, + "grad_norm": 0.4438367286714941, + "learning_rate": 2.260790434626492e-07, + "loss": 0.2453, + "step": 5472 + }, + { + "epoch": 1.8273789649415693, + "grad_norm": 0.46339550906933646, + "learning_rate": 2.2521360040511974e-07, + "loss": 0.2573, + "step": 5473 + }, + { + "epoch": 1.8277128547579298, + "grad_norm": 0.467531998433807, + "learning_rate": 2.2434977884036424e-07, + "loss": 0.2511, + "step": 5474 + }, + { + "epoch": 1.8280467445742905, + "grad_norm": 0.46417258300787967, + "learning_rate": 2.2348757906173245e-07, + "loss": 0.25, + "step": 5475 + }, + { + "epoch": 1.8283806343906512, + "grad_norm": 0.44117518822575175, + "learning_rate": 2.2262700136202241e-07, + "loss": 0.2445, + "step": 5476 + }, + { + "epoch": 1.8287145242070117, + "grad_norm": 0.45335659566726505, + "learning_rate": 2.217680460334809e-07, + "loss": 0.2524, + "step": 5477 + }, + { + "epoch": 1.8290484140233723, + "grad_norm": 0.44023250405981446, + "learning_rate": 2.2091071336780622e-07, + "loss": 0.2431, + "step": 5478 + }, + { + "epoch": 1.8293823038397328, + "grad_norm": 0.4507522332181317, + "learning_rate": 2.2005500365614386e-07, + "loss": 0.2585, + "step": 5479 + }, + { + "epoch": 1.8297161936560935, + "grad_norm": 0.4557670600253893, + "learning_rate": 2.1920091718908853e-07, + "loss": 0.2589, + "step": 5480 + }, + { + "epoch": 1.8300500834724542, + "grad_norm": 0.4463920526080001, + "learning_rate": 2.1834845425668327e-07, + "loss": 0.2407, + "step": 5481 + }, + { + "epoch": 1.8303839732888147, + "grad_norm": 0.45066402004420514, + "learning_rate": 2.1749761514841871e-07, + "loss": 0.2464, + "step": 5482 + }, + { + "epoch": 1.8307178631051753, + "grad_norm": 0.45491529374758544, + "learning_rate": 2.1664840015323818e-07, + "loss": 0.2493, + "step": 5483 + }, + { + "epoch": 1.8310517529215358, + "grad_norm": 0.43604798218222973, + "learning_rate": 2.1580080955952875e-07, + "loss": 0.2422, + "step": 5484 + }, + { + "epoch": 1.8313856427378965, + "grad_norm": 0.4417568823566669, + "learning_rate": 2.1495484365512797e-07, + "loss": 0.2478, + "step": 5485 + }, + { + "epoch": 1.8317195325542572, + "grad_norm": 0.4652962409949888, + "learning_rate": 2.141105027273227e-07, + "loss": 0.2668, + "step": 5486 + }, + { + "epoch": 1.8320534223706177, + "grad_norm": 0.444370819265364, + "learning_rate": 2.1326778706284468e-07, + "loss": 0.2445, + "step": 5487 + }, + { + "epoch": 1.8323873121869783, + "grad_norm": 0.438251734560701, + "learning_rate": 2.1242669694787776e-07, + "loss": 0.2387, + "step": 5488 + }, + { + "epoch": 1.8327212020033388, + "grad_norm": 0.45187541888467103, + "learning_rate": 2.1158723266805125e-07, + "loss": 0.2422, + "step": 5489 + }, + { + "epoch": 1.8330550918196995, + "grad_norm": 0.44200996703431494, + "learning_rate": 2.1074939450844267e-07, + "loss": 0.2425, + "step": 5490 + }, + { + "epoch": 1.8333889816360602, + "grad_norm": 0.4458482619013215, + "learning_rate": 2.0991318275357663e-07, + "loss": 0.2435, + "step": 5491 + }, + { + "epoch": 1.8337228714524207, + "grad_norm": 0.42991565278205124, + "learning_rate": 2.0907859768742823e-07, + "loss": 0.2431, + "step": 5492 + }, + { + "epoch": 1.8340567612687813, + "grad_norm": 0.4490306883990684, + "learning_rate": 2.0824563959341682e-07, + "loss": 0.2491, + "step": 5493 + }, + { + "epoch": 1.8343906510851418, + "grad_norm": 0.46786204296819445, + "learning_rate": 2.074143087544106e-07, + "loss": 0.2556, + "step": 5494 + }, + { + "epoch": 1.8347245409015025, + "grad_norm": 0.4360453360991183, + "learning_rate": 2.0658460545272653e-07, + "loss": 0.2279, + "step": 5495 + }, + { + "epoch": 1.8350584307178632, + "grad_norm": 0.43959609544404027, + "learning_rate": 2.0575652997012586e-07, + "loss": 0.2384, + "step": 5496 + }, + { + "epoch": 1.8353923205342237, + "grad_norm": 0.46230707518500536, + "learning_rate": 2.0493008258782033e-07, + "loss": 0.2466, + "step": 5497 + }, + { + "epoch": 1.8357262103505843, + "grad_norm": 0.4502794114791781, + "learning_rate": 2.0410526358646654e-07, + "loss": 0.2477, + "step": 5498 + }, + { + "epoch": 1.8360601001669448, + "grad_norm": 0.4475884949647476, + "learning_rate": 2.0328207324616877e-07, + "loss": 0.2532, + "step": 5499 + }, + { + "epoch": 1.8363939899833055, + "grad_norm": 0.45686924593115386, + "learning_rate": 2.024605118464773e-07, + "loss": 0.2471, + "step": 5500 + }, + { + "epoch": 1.8367278797996662, + "grad_norm": 0.43830212524665, + "learning_rate": 2.0164057966639116e-07, + "loss": 0.2445, + "step": 5501 + }, + { + "epoch": 1.8370617696160267, + "grad_norm": 0.4560458015516495, + "learning_rate": 2.0082227698435485e-07, + "loss": 0.2485, + "step": 5502 + }, + { + "epoch": 1.8373956594323873, + "grad_norm": 0.45445000777285255, + "learning_rate": 2.0000560407825999e-07, + "loss": 0.2573, + "step": 5503 + }, + { + "epoch": 1.8377295492487478, + "grad_norm": 0.4357395826250604, + "learning_rate": 1.9919056122544467e-07, + "loss": 0.2447, + "step": 5504 + }, + { + "epoch": 1.8380634390651085, + "grad_norm": 0.444935773127674, + "learning_rate": 1.9837714870269198e-07, + "loss": 0.2535, + "step": 5505 + }, + { + "epoch": 1.8383973288814692, + "grad_norm": 0.4626297393015913, + "learning_rate": 1.9756536678623429e-07, + "loss": 0.2478, + "step": 5506 + }, + { + "epoch": 1.8387312186978297, + "grad_norm": 0.4476209866370699, + "learning_rate": 1.9675521575174717e-07, + "loss": 0.252, + "step": 5507 + }, + { + "epoch": 1.8390651085141902, + "grad_norm": 0.44450688952486905, + "learning_rate": 1.9594669587435445e-07, + "loss": 0.2387, + "step": 5508 + }, + { + "epoch": 1.8393989983305508, + "grad_norm": 0.44452495228098965, + "learning_rate": 1.9513980742862482e-07, + "loss": 0.2485, + "step": 5509 + }, + { + "epoch": 1.8397328881469115, + "grad_norm": 0.4577774847070869, + "learning_rate": 1.943345506885741e-07, + "loss": 0.2489, + "step": 5510 + }, + { + "epoch": 1.8400667779632722, + "grad_norm": 0.46432308240297476, + "learning_rate": 1.9353092592766186e-07, + "loss": 0.248, + "step": 5511 + }, + { + "epoch": 1.8404006677796327, + "grad_norm": 0.4851176550625034, + "learning_rate": 1.9272893341879707e-07, + "loss": 0.2564, + "step": 5512 + }, + { + "epoch": 1.8407345575959932, + "grad_norm": 0.45025607442978705, + "learning_rate": 1.9192857343433069e-07, + "loss": 0.2456, + "step": 5513 + }, + { + "epoch": 1.8410684474123538, + "grad_norm": 0.4543249838160914, + "learning_rate": 1.911298462460609e-07, + "loss": 0.2572, + "step": 5514 + }, + { + "epoch": 1.8414023372287145, + "grad_norm": 0.46040535683033895, + "learning_rate": 1.9033275212523238e-07, + "loss": 0.2405, + "step": 5515 + }, + { + "epoch": 1.8417362270450752, + "grad_norm": 0.4542705159154179, + "learning_rate": 1.8953729134253362e-07, + "loss": 0.2483, + "step": 5516 + }, + { + "epoch": 1.8420701168614357, + "grad_norm": 0.46506915033301366, + "learning_rate": 1.8874346416809853e-07, + "loss": 0.254, + "step": 5517 + }, + { + "epoch": 1.8424040066777962, + "grad_norm": 0.46104789973062293, + "learning_rate": 1.8795127087150644e-07, + "loss": 0.2477, + "step": 5518 + }, + { + "epoch": 1.842737896494157, + "grad_norm": 0.4446835717885217, + "learning_rate": 1.871607117217833e-07, + "loss": 0.2379, + "step": 5519 + }, + { + "epoch": 1.8430717863105175, + "grad_norm": 0.4655278023056049, + "learning_rate": 1.863717869873982e-07, + "loss": 0.2524, + "step": 5520 + }, + { + "epoch": 1.8434056761268782, + "grad_norm": 0.4463190523418212, + "learning_rate": 1.8558449693626633e-07, + "loss": 0.242, + "step": 5521 + }, + { + "epoch": 1.8437395659432387, + "grad_norm": 0.44618654269908054, + "learning_rate": 1.847988418357466e-07, + "loss": 0.2398, + "step": 5522 + }, + { + "epoch": 1.8440734557595992, + "grad_norm": 0.4502232902049994, + "learning_rate": 1.8401482195264386e-07, + "loss": 0.251, + "step": 5523 + }, + { + "epoch": 1.84440734557596, + "grad_norm": 0.45552737207237143, + "learning_rate": 1.8323243755320741e-07, + "loss": 0.2336, + "step": 5524 + }, + { + "epoch": 1.8447412353923205, + "grad_norm": 0.43981735079307016, + "learning_rate": 1.824516889031308e-07, + "loss": 0.2444, + "step": 5525 + }, + { + "epoch": 1.8450751252086812, + "grad_norm": 0.46853464379304305, + "learning_rate": 1.8167257626755196e-07, + "loss": 0.2464, + "step": 5526 + }, + { + "epoch": 1.8454090150250417, + "grad_norm": 0.45138634259698296, + "learning_rate": 1.8089509991105315e-07, + "loss": 0.2362, + "step": 5527 + }, + { + "epoch": 1.8457429048414022, + "grad_norm": 0.4484117402702327, + "learning_rate": 1.8011926009766202e-07, + "loss": 0.2402, + "step": 5528 + }, + { + "epoch": 1.846076794657763, + "grad_norm": 0.43541670743800576, + "learning_rate": 1.7934505709085004e-07, + "loss": 0.2426, + "step": 5529 + }, + { + "epoch": 1.8464106844741235, + "grad_norm": 0.4424765067684759, + "learning_rate": 1.7857249115353192e-07, + "loss": 0.2456, + "step": 5530 + }, + { + "epoch": 1.8467445742904842, + "grad_norm": 0.44363764050269094, + "learning_rate": 1.778015625480678e-07, + "loss": 0.2433, + "step": 5531 + }, + { + "epoch": 1.8470784641068447, + "grad_norm": 0.45184464234680777, + "learning_rate": 1.7703227153625934e-07, + "loss": 0.2485, + "step": 5532 + }, + { + "epoch": 1.8474123539232052, + "grad_norm": 0.46235391092970685, + "learning_rate": 1.7626461837935592e-07, + "loss": 0.263, + "step": 5533 + }, + { + "epoch": 1.847746243739566, + "grad_norm": 0.4323116411515287, + "learning_rate": 1.7549860333804735e-07, + "loss": 0.2329, + "step": 5534 + }, + { + "epoch": 1.8480801335559267, + "grad_norm": 0.455169456600778, + "learning_rate": 1.7473422667246886e-07, + "loss": 0.2477, + "step": 5535 + }, + { + "epoch": 1.8484140233722872, + "grad_norm": 0.4430360589573115, + "learning_rate": 1.7397148864219782e-07, + "loss": 0.2426, + "step": 5536 + }, + { + "epoch": 1.8487479131886477, + "grad_norm": 0.4528248006158325, + "learning_rate": 1.7321038950625702e-07, + "loss": 0.2457, + "step": 5537 + }, + { + "epoch": 1.8490818030050082, + "grad_norm": 0.4478879232986762, + "learning_rate": 1.7245092952311193e-07, + "loss": 0.2453, + "step": 5538 + }, + { + "epoch": 1.849415692821369, + "grad_norm": 0.44434383854203746, + "learning_rate": 1.7169310895067126e-07, + "loss": 0.2387, + "step": 5539 + }, + { + "epoch": 1.8497495826377297, + "grad_norm": 0.4475578940022303, + "learning_rate": 1.7093692804628637e-07, + "loss": 0.2463, + "step": 5540 + }, + { + "epoch": 1.8500834724540902, + "grad_norm": 0.45063818554756446, + "learning_rate": 1.7018238706675237e-07, + "loss": 0.2411, + "step": 5541 + }, + { + "epoch": 1.8504173622704507, + "grad_norm": 0.4457019585501006, + "learning_rate": 1.6942948626830824e-07, + "loss": 0.2499, + "step": 5542 + }, + { + "epoch": 1.8507512520868112, + "grad_norm": 0.45180527549384686, + "learning_rate": 1.6867822590663497e-07, + "loss": 0.2449, + "step": 5543 + }, + { + "epoch": 1.851085141903172, + "grad_norm": 0.45171560827577745, + "learning_rate": 1.6792860623685624e-07, + "loss": 0.2426, + "step": 5544 + }, + { + "epoch": 1.8514190317195327, + "grad_norm": 0.46429412610154547, + "learning_rate": 1.6718062751353902e-07, + "loss": 0.253, + "step": 5545 + }, + { + "epoch": 1.8517529215358932, + "grad_norm": 0.44516630065109264, + "learning_rate": 1.664342899906929e-07, + "loss": 0.2354, + "step": 5546 + }, + { + "epoch": 1.8520868113522537, + "grad_norm": 0.4586826111890537, + "learning_rate": 1.656895939217712e-07, + "loss": 0.2597, + "step": 5547 + }, + { + "epoch": 1.8524207011686142, + "grad_norm": 0.4321576213764627, + "learning_rate": 1.6494653955966833e-07, + "loss": 0.2373, + "step": 5548 + }, + { + "epoch": 1.852754590984975, + "grad_norm": 0.4661022269049113, + "learning_rate": 1.642051271567213e-07, + "loss": 0.2566, + "step": 5549 + }, + { + "epoch": 1.8530884808013357, + "grad_norm": 0.4402429930928559, + "learning_rate": 1.634653569647099e-07, + "loss": 0.2435, + "step": 5550 + }, + { + "epoch": 1.8534223706176962, + "grad_norm": 0.46571559094626513, + "learning_rate": 1.627272292348575e-07, + "loss": 0.2533, + "step": 5551 + }, + { + "epoch": 1.8537562604340567, + "grad_norm": 0.46270301107429385, + "learning_rate": 1.6199074421782702e-07, + "loss": 0.2501, + "step": 5552 + }, + { + "epoch": 1.8540901502504172, + "grad_norm": 0.42851468492035055, + "learning_rate": 1.61255902163725e-07, + "loss": 0.2389, + "step": 5553 + }, + { + "epoch": 1.854424040066778, + "grad_norm": 0.44274240358477374, + "learning_rate": 1.605227033221013e-07, + "loss": 0.2404, + "step": 5554 + }, + { + "epoch": 1.8547579298831387, + "grad_norm": 0.43921991568784907, + "learning_rate": 1.5979114794194506e-07, + "loss": 0.2509, + "step": 5555 + }, + { + "epoch": 1.8550918196994992, + "grad_norm": 0.452038204922528, + "learning_rate": 1.590612362716898e-07, + "loss": 0.2401, + "step": 5556 + }, + { + "epoch": 1.8554257095158597, + "grad_norm": 0.45085996096196795, + "learning_rate": 1.5833296855920944e-07, + "loss": 0.2519, + "step": 5557 + }, + { + "epoch": 1.8557595993322202, + "grad_norm": 0.43200541194712877, + "learning_rate": 1.5760634505182004e-07, + "loss": 0.236, + "step": 5558 + }, + { + "epoch": 1.856093489148581, + "grad_norm": 0.45592511159182153, + "learning_rate": 1.5688136599627757e-07, + "loss": 0.2529, + "step": 5559 + }, + { + "epoch": 1.8564273789649417, + "grad_norm": 0.4511144571085775, + "learning_rate": 1.561580316387834e-07, + "loss": 0.2519, + "step": 5560 + }, + { + "epoch": 1.8567612687813022, + "grad_norm": 0.43520246630798964, + "learning_rate": 1.5543634222497718e-07, + "loss": 0.2465, + "step": 5561 + }, + { + "epoch": 1.8570951585976627, + "grad_norm": 0.4523670584259471, + "learning_rate": 1.547162979999406e-07, + "loss": 0.246, + "step": 5562 + }, + { + "epoch": 1.8574290484140232, + "grad_norm": 0.45179183758394936, + "learning_rate": 1.5399789920819808e-07, + "loss": 0.2437, + "step": 5563 + }, + { + "epoch": 1.857762938230384, + "grad_norm": 0.44201491334099613, + "learning_rate": 1.5328114609371224e-07, + "loss": 0.2439, + "step": 5564 + }, + { + "epoch": 1.8580968280467447, + "grad_norm": 0.4468370162668023, + "learning_rate": 1.5256603889989063e-07, + "loss": 0.2479, + "step": 5565 + }, + { + "epoch": 1.8584307178631052, + "grad_norm": 0.4439158426685447, + "learning_rate": 1.518525778695795e-07, + "loss": 0.2448, + "step": 5566 + }, + { + "epoch": 1.8587646076794657, + "grad_norm": 0.4595766663709916, + "learning_rate": 1.5114076324506567e-07, + "loss": 0.2486, + "step": 5567 + }, + { + "epoch": 1.8590984974958262, + "grad_norm": 0.4511201212435555, + "learning_rate": 1.504305952680779e-07, + "loss": 0.2497, + "step": 5568 + }, + { + "epoch": 1.859432387312187, + "grad_norm": 0.45910311486502564, + "learning_rate": 1.4972207417978669e-07, + "loss": 0.2585, + "step": 5569 + }, + { + "epoch": 1.8597662771285477, + "grad_norm": 0.44932397503401555, + "learning_rate": 1.4901520022080063e-07, + "loss": 0.2452, + "step": 5570 + }, + { + "epoch": 1.8601001669449082, + "grad_norm": 0.43617972195612814, + "learning_rate": 1.4830997363117218e-07, + "loss": 0.2382, + "step": 5571 + }, + { + "epoch": 1.8604340567612687, + "grad_norm": 0.4381299459784258, + "learning_rate": 1.4760639465039084e-07, + "loss": 0.2341, + "step": 5572 + }, + { + "epoch": 1.8607679465776292, + "grad_norm": 0.4497206687246771, + "learning_rate": 1.4690446351738942e-07, + "loss": 0.2556, + "step": 5573 + }, + { + "epoch": 1.86110183639399, + "grad_norm": 0.4741557409591402, + "learning_rate": 1.4620418047053998e-07, + "loss": 0.2535, + "step": 5574 + }, + { + "epoch": 1.8614357262103507, + "grad_norm": 0.45469281872248024, + "learning_rate": 1.4550554574765507e-07, + "loss": 0.2523, + "step": 5575 + }, + { + "epoch": 1.8617696160267112, + "grad_norm": 0.46451337138913545, + "learning_rate": 1.4480855958598716e-07, + "loss": 0.2492, + "step": 5576 + }, + { + "epoch": 1.8621035058430717, + "grad_norm": 0.44528976898750616, + "learning_rate": 1.4411322222222913e-07, + "loss": 0.2513, + "step": 5577 + }, + { + "epoch": 1.8624373956594322, + "grad_norm": 0.4778027818342519, + "learning_rate": 1.4341953389251427e-07, + "loss": 0.2518, + "step": 5578 + }, + { + "epoch": 1.862771285475793, + "grad_norm": 0.43187512703886954, + "learning_rate": 1.4272749483241477e-07, + "loss": 0.2438, + "step": 5579 + }, + { + "epoch": 1.8631051752921537, + "grad_norm": 0.4440688559254374, + "learning_rate": 1.420371052769448e-07, + "loss": 0.2503, + "step": 5580 + }, + { + "epoch": 1.8634390651085142, + "grad_norm": 0.45345690462046795, + "learning_rate": 1.4134836546055686e-07, + "loss": 0.2481, + "step": 5581 + }, + { + "epoch": 1.8637729549248747, + "grad_norm": 0.44443476227269324, + "learning_rate": 1.4066127561714215e-07, + "loss": 0.2434, + "step": 5582 + }, + { + "epoch": 1.8641068447412354, + "grad_norm": 0.4616396561328826, + "learning_rate": 1.3997583598003405e-07, + "loss": 0.2418, + "step": 5583 + }, + { + "epoch": 1.864440734557596, + "grad_norm": 0.44310820804494494, + "learning_rate": 1.3929204678200415e-07, + "loss": 0.2328, + "step": 5584 + }, + { + "epoch": 1.8647746243739567, + "grad_norm": 0.43746502867482095, + "learning_rate": 1.3860990825526334e-07, + "loss": 0.2381, + "step": 5585 + }, + { + "epoch": 1.8651085141903172, + "grad_norm": 0.44010983687353245, + "learning_rate": 1.3792942063146188e-07, + "loss": 0.2438, + "step": 5586 + }, + { + "epoch": 1.8654424040066777, + "grad_norm": 0.44282450752577657, + "learning_rate": 1.3725058414169047e-07, + "loss": 0.2487, + "step": 5587 + }, + { + "epoch": 1.8657762938230384, + "grad_norm": 0.4476777842984589, + "learning_rate": 1.3657339901647914e-07, + "loss": 0.2411, + "step": 5588 + }, + { + "epoch": 1.866110183639399, + "grad_norm": 0.43796258761254886, + "learning_rate": 1.3589786548579563e-07, + "loss": 0.2472, + "step": 5589 + }, + { + "epoch": 1.8664440734557597, + "grad_norm": 0.44888427585202584, + "learning_rate": 1.352239837790481e-07, + "loss": 0.2462, + "step": 5590 + }, + { + "epoch": 1.8667779632721202, + "grad_norm": 0.42165171939081453, + "learning_rate": 1.3455175412508237e-07, + "loss": 0.2335, + "step": 5591 + }, + { + "epoch": 1.8671118530884807, + "grad_norm": 0.4532636120075771, + "learning_rate": 1.3388117675218525e-07, + "loss": 0.2475, + "step": 5592 + }, + { + "epoch": 1.8674457429048414, + "grad_norm": 0.4344408751267592, + "learning_rate": 1.3321225188808075e-07, + "loss": 0.2469, + "step": 5593 + }, + { + "epoch": 1.867779632721202, + "grad_norm": 0.4630120495477217, + "learning_rate": 1.3254497975993263e-07, + "loss": 0.249, + "step": 5594 + }, + { + "epoch": 1.8681135225375627, + "grad_norm": 0.4522123564675102, + "learning_rate": 1.3187936059434303e-07, + "loss": 0.2418, + "step": 5595 + }, + { + "epoch": 1.8684474123539232, + "grad_norm": 0.4444171162434614, + "learning_rate": 1.3121539461735112e-07, + "loss": 0.2414, + "step": 5596 + }, + { + "epoch": 1.8687813021702837, + "grad_norm": 0.4445166147152175, + "learning_rate": 1.3055308205443928e-07, + "loss": 0.239, + "step": 5597 + }, + { + "epoch": 1.8691151919866444, + "grad_norm": 0.45963988336124256, + "learning_rate": 1.2989242313052318e-07, + "loss": 0.2478, + "step": 5598 + }, + { + "epoch": 1.8694490818030052, + "grad_norm": 0.4425440516208252, + "learning_rate": 1.292334180699606e-07, + "loss": 0.249, + "step": 5599 + }, + { + "epoch": 1.8697829716193657, + "grad_norm": 0.45941519517914853, + "learning_rate": 1.285760670965447e-07, + "loss": 0.2537, + "step": 5600 + }, + { + "epoch": 1.8701168614357262, + "grad_norm": 0.45609413667446924, + "learning_rate": 1.2792037043351023e-07, + "loss": 0.247, + "step": 5601 + }, + { + "epoch": 1.8704507512520867, + "grad_norm": 0.4526964873997279, + "learning_rate": 1.2726632830352748e-07, + "loss": 0.2474, + "step": 5602 + }, + { + "epoch": 1.8707846410684474, + "grad_norm": 0.4538161108750504, + "learning_rate": 1.266139409287054e-07, + "loss": 0.2503, + "step": 5603 + }, + { + "epoch": 1.8711185308848082, + "grad_norm": 0.4469451371738491, + "learning_rate": 1.2596320853059175e-07, + "loss": 0.2477, + "step": 5604 + }, + { + "epoch": 1.8714524207011687, + "grad_norm": 0.45714113244361937, + "learning_rate": 1.2531413133017146e-07, + "loss": 0.2421, + "step": 5605 + }, + { + "epoch": 1.8717863105175292, + "grad_norm": 0.4437948696661533, + "learning_rate": 1.246667095478693e-07, + "loss": 0.2461, + "step": 5606 + }, + { + "epoch": 1.8721202003338897, + "grad_norm": 0.4423232975995517, + "learning_rate": 1.2402094340354497e-07, + "loss": 0.2459, + "step": 5607 + }, + { + "epoch": 1.8724540901502504, + "grad_norm": 0.44362993036627135, + "learning_rate": 1.233768331164975e-07, + "loss": 0.2429, + "step": 5608 + }, + { + "epoch": 1.8727879799666112, + "grad_norm": 0.4607489462780422, + "learning_rate": 1.2273437890546357e-07, + "loss": 0.2491, + "step": 5609 + }, + { + "epoch": 1.8731218697829717, + "grad_norm": 0.4361482457706481, + "learning_rate": 1.22093580988617e-07, + "loss": 0.2432, + "step": 5610 + }, + { + "epoch": 1.8734557595993322, + "grad_norm": 0.4447049994140928, + "learning_rate": 1.214544395835704e-07, + "loss": 0.2415, + "step": 5611 + }, + { + "epoch": 1.8737896494156927, + "grad_norm": 0.4259467266747376, + "learning_rate": 1.2081695490737177e-07, + "loss": 0.2364, + "step": 5612 + }, + { + "epoch": 1.8741235392320534, + "grad_norm": 0.4502415350305598, + "learning_rate": 1.2018112717650797e-07, + "loss": 0.249, + "step": 5613 + }, + { + "epoch": 1.8744574290484142, + "grad_norm": 0.4488759915795211, + "learning_rate": 1.1954695660690353e-07, + "loss": 0.2379, + "step": 5614 + }, + { + "epoch": 1.8747913188647747, + "grad_norm": 0.4449744100443262, + "learning_rate": 1.1891444341391778e-07, + "loss": 0.2446, + "step": 5615 + }, + { + "epoch": 1.8751252086811352, + "grad_norm": 0.44929816547816914, + "learning_rate": 1.1828358781235117e-07, + "loss": 0.2509, + "step": 5616 + }, + { + "epoch": 1.8754590984974957, + "grad_norm": 0.44659029593280314, + "learning_rate": 1.176543900164373e-07, + "loss": 0.2401, + "step": 5617 + }, + { + "epoch": 1.8757929883138564, + "grad_norm": 0.4624410985961442, + "learning_rate": 1.1702685023984972e-07, + "loss": 0.2523, + "step": 5618 + }, + { + "epoch": 1.8761268781302172, + "grad_norm": 0.4450661511901236, + "learning_rate": 1.1640096869569628e-07, + "loss": 0.2397, + "step": 5619 + }, + { + "epoch": 1.8764607679465777, + "grad_norm": 0.44144347720007165, + "learning_rate": 1.1577674559652419e-07, + "loss": 0.2305, + "step": 5620 + }, + { + "epoch": 1.8767946577629382, + "grad_norm": 0.4557875147840734, + "learning_rate": 1.1515418115431554e-07, + "loss": 0.2548, + "step": 5621 + }, + { + "epoch": 1.8771285475792987, + "grad_norm": 0.4702008214258934, + "learning_rate": 1.1453327558049122e-07, + "loss": 0.2456, + "step": 5622 + }, + { + "epoch": 1.8774624373956594, + "grad_norm": 0.456015409972349, + "learning_rate": 1.1391402908590644e-07, + "loss": 0.2436, + "step": 5623 + }, + { + "epoch": 1.8777963272120202, + "grad_norm": 0.4499286755402619, + "learning_rate": 1.132964418808541e-07, + "loss": 0.2433, + "step": 5624 + }, + { + "epoch": 1.8781302170283807, + "grad_norm": 0.45431774501357425, + "learning_rate": 1.1268051417506476e-07, + "loss": 0.2494, + "step": 5625 + }, + { + "epoch": 1.8784641068447412, + "grad_norm": 0.45271108934746174, + "learning_rate": 1.120662461777039e-07, + "loss": 0.2552, + "step": 5626 + }, + { + "epoch": 1.8787979966611017, + "grad_norm": 0.46675153412766546, + "learning_rate": 1.1145363809737297e-07, + "loss": 0.2429, + "step": 5627 + }, + { + "epoch": 1.8791318864774624, + "grad_norm": 0.43974142047001985, + "learning_rate": 1.1084269014211058e-07, + "loss": 0.2508, + "step": 5628 + }, + { + "epoch": 1.8794657762938232, + "grad_norm": 0.46259895470824397, + "learning_rate": 1.102334025193924e-07, + "loss": 0.248, + "step": 5629 + }, + { + "epoch": 1.8797996661101837, + "grad_norm": 0.4390673022447401, + "learning_rate": 1.0962577543612796e-07, + "loss": 0.2486, + "step": 5630 + }, + { + "epoch": 1.8801335559265442, + "grad_norm": 0.46547107675773625, + "learning_rate": 1.0901980909866605e-07, + "loss": 0.2492, + "step": 5631 + }, + { + "epoch": 1.8804674457429047, + "grad_norm": 0.4556923416014421, + "learning_rate": 1.0841550371278875e-07, + "loss": 0.2315, + "step": 5632 + }, + { + "epoch": 1.8808013355592654, + "grad_norm": 0.43595187451589734, + "learning_rate": 1.0781285948371467e-07, + "loss": 0.2373, + "step": 5633 + }, + { + "epoch": 1.8811352253756262, + "grad_norm": 0.4467402526377735, + "learning_rate": 1.07211876616099e-07, + "loss": 0.2479, + "step": 5634 + }, + { + "epoch": 1.8814691151919867, + "grad_norm": 0.4424111994381275, + "learning_rate": 1.0661255531403292e-07, + "loss": 0.2392, + "step": 5635 + }, + { + "epoch": 1.8818030050083472, + "grad_norm": 0.44313196515598224, + "learning_rate": 1.0601489578104252e-07, + "loss": 0.2455, + "step": 5636 + }, + { + "epoch": 1.8821368948247077, + "grad_norm": 0.4601439044278716, + "learning_rate": 1.0541889822008932e-07, + "loss": 0.2415, + "step": 5637 + }, + { + "epoch": 1.8824707846410684, + "grad_norm": 0.4746028473444561, + "learning_rate": 1.0482456283357145e-07, + "loss": 0.2573, + "step": 5638 + }, + { + "epoch": 1.8828046744574292, + "grad_norm": 0.44613931846806976, + "learning_rate": 1.04231889823323e-07, + "loss": 0.2466, + "step": 5639 + }, + { + "epoch": 1.8831385642737897, + "grad_norm": 0.46023860037414116, + "learning_rate": 1.0364087939061129e-07, + "loss": 0.2488, + "step": 5640 + }, + { + "epoch": 1.8834724540901502, + "grad_norm": 0.43683021689043355, + "learning_rate": 1.0305153173614135e-07, + "loss": 0.2429, + "step": 5641 + }, + { + "epoch": 1.8838063439065107, + "grad_norm": 0.46250198081899563, + "learning_rate": 1.0246384706005142e-07, + "loss": 0.252, + "step": 5642 + }, + { + "epoch": 1.8841402337228714, + "grad_norm": 0.4374224307532263, + "learning_rate": 1.0187782556191739e-07, + "loss": 0.2384, + "step": 5643 + }, + { + "epoch": 1.8844741235392322, + "grad_norm": 0.4596969631963904, + "learning_rate": 1.0129346744074897e-07, + "loss": 0.2389, + "step": 5644 + }, + { + "epoch": 1.8848080133555927, + "grad_norm": 0.4519606502265985, + "learning_rate": 1.0071077289499076e-07, + "loss": 0.2473, + "step": 5645 + }, + { + "epoch": 1.8851419031719532, + "grad_norm": 0.4568588262004344, + "learning_rate": 1.0012974212252224e-07, + "loss": 0.2475, + "step": 5646 + }, + { + "epoch": 1.8854757929883137, + "grad_norm": 0.43724314044548585, + "learning_rate": 9.955037532065947e-08, + "loss": 0.2395, + "step": 5647 + }, + { + "epoch": 1.8858096828046744, + "grad_norm": 0.4465861937759533, + "learning_rate": 9.897267268615285e-08, + "loss": 0.2457, + "step": 5648 + }, + { + "epoch": 1.8861435726210352, + "grad_norm": 0.4418805556992922, + "learning_rate": 9.839663441518598e-08, + "loss": 0.24, + "step": 5649 + }, + { + "epoch": 1.8864774624373957, + "grad_norm": 0.44055179042488657, + "learning_rate": 9.782226070337908e-08, + "loss": 0.246, + "step": 5650 + }, + { + "epoch": 1.8868113522537562, + "grad_norm": 0.44295071040792144, + "learning_rate": 9.724955174578609e-08, + "loss": 0.2359, + "step": 5651 + }, + { + "epoch": 1.887145242070117, + "grad_norm": 0.4556061243340427, + "learning_rate": 9.667850773689647e-08, + "loss": 0.2483, + "step": 5652 + }, + { + "epoch": 1.8874791318864774, + "grad_norm": 0.4496329434884927, + "learning_rate": 9.610912887063395e-08, + "loss": 0.2448, + "step": 5653 + }, + { + "epoch": 1.8878130217028382, + "grad_norm": 0.45419656121227386, + "learning_rate": 9.554141534035611e-08, + "loss": 0.2486, + "step": 5654 + }, + { + "epoch": 1.8881469115191987, + "grad_norm": 0.4487737719314641, + "learning_rate": 9.497536733885537e-08, + "loss": 0.2516, + "step": 5655 + }, + { + "epoch": 1.8884808013355592, + "grad_norm": 0.4535471926328394, + "learning_rate": 9.441098505835911e-08, + "loss": 0.247, + "step": 5656 + }, + { + "epoch": 1.88881469115192, + "grad_norm": 0.44617973680113326, + "learning_rate": 9.384826869052899e-08, + "loss": 0.2451, + "step": 5657 + }, + { + "epoch": 1.8891485809682804, + "grad_norm": 0.4446804146331147, + "learning_rate": 9.328721842646049e-08, + "loss": 0.2502, + "step": 5658 + }, + { + "epoch": 1.8894824707846412, + "grad_norm": 0.4619521368261692, + "learning_rate": 9.272783445668287e-08, + "loss": 0.2612, + "step": 5659 + }, + { + "epoch": 1.8898163606010017, + "grad_norm": 0.45439778213354215, + "learning_rate": 9.217011697115974e-08, + "loss": 0.242, + "step": 5660 + }, + { + "epoch": 1.8901502504173622, + "grad_norm": 0.45276851372117904, + "learning_rate": 9.161406615929013e-08, + "loss": 0.2483, + "step": 5661 + }, + { + "epoch": 1.890484140233723, + "grad_norm": 0.44573332867927495, + "learning_rate": 9.105968220990524e-08, + "loss": 0.2436, + "step": 5662 + }, + { + "epoch": 1.8908180300500834, + "grad_norm": 0.46500489823664354, + "learning_rate": 9.050696531127167e-08, + "loss": 0.2505, + "step": 5663 + }, + { + "epoch": 1.8911519198664442, + "grad_norm": 0.44512200412919467, + "learning_rate": 8.995591565108819e-08, + "loss": 0.252, + "step": 5664 + }, + { + "epoch": 1.8914858096828047, + "grad_norm": 0.45230391512825263, + "learning_rate": 8.940653341648897e-08, + "loss": 0.2358, + "step": 5665 + }, + { + "epoch": 1.8918196994991652, + "grad_norm": 0.43442770600830305, + "learning_rate": 8.885881879404201e-08, + "loss": 0.2367, + "step": 5666 + }, + { + "epoch": 1.892153589315526, + "grad_norm": 0.43047335351640115, + "learning_rate": 8.831277196974741e-08, + "loss": 0.2465, + "step": 5667 + }, + { + "epoch": 1.8924874791318866, + "grad_norm": 0.4702491001730818, + "learning_rate": 8.776839312904074e-08, + "loss": 0.248, + "step": 5668 + }, + { + "epoch": 1.8928213689482472, + "grad_norm": 0.4622585959675401, + "learning_rate": 8.722568245678908e-08, + "loss": 0.2601, + "step": 5669 + }, + { + "epoch": 1.8931552587646077, + "grad_norm": 0.4382041733002877, + "learning_rate": 8.668464013729506e-08, + "loss": 0.2412, + "step": 5670 + }, + { + "epoch": 1.8934891485809682, + "grad_norm": 0.44638078165719114, + "learning_rate": 8.614526635429388e-08, + "loss": 0.244, + "step": 5671 + }, + { + "epoch": 1.893823038397329, + "grad_norm": 0.4349248511987183, + "learning_rate": 8.560756129095294e-08, + "loss": 0.2254, + "step": 5672 + }, + { + "epoch": 1.8941569282136896, + "grad_norm": 0.442186150521781, + "learning_rate": 8.50715251298756e-08, + "loss": 0.2366, + "step": 5673 + }, + { + "epoch": 1.8944908180300501, + "grad_norm": 0.4324922874604291, + "learning_rate": 8.45371580530957e-08, + "loss": 0.2281, + "step": 5674 + }, + { + "epoch": 1.8948247078464107, + "grad_norm": 0.44902647901890413, + "learning_rate": 8.400446024208309e-08, + "loss": 0.2471, + "step": 5675 + }, + { + "epoch": 1.8951585976627712, + "grad_norm": 0.45271967624322673, + "learning_rate": 8.347343187773805e-08, + "loss": 0.2454, + "step": 5676 + }, + { + "epoch": 1.895492487479132, + "grad_norm": 0.4361995210001215, + "learning_rate": 8.294407314039521e-08, + "loss": 0.2423, + "step": 5677 + }, + { + "epoch": 1.8958263772954926, + "grad_norm": 0.43266117961262235, + "learning_rate": 8.24163842098219e-08, + "loss": 0.2362, + "step": 5678 + }, + { + "epoch": 1.8961602671118531, + "grad_norm": 0.4493880520580965, + "learning_rate": 8.189036526521921e-08, + "loss": 0.2402, + "step": 5679 + }, + { + "epoch": 1.8964941569282137, + "grad_norm": 0.4616539701280541, + "learning_rate": 8.13660164852198e-08, + "loss": 0.2466, + "step": 5680 + }, + { + "epoch": 1.8968280467445742, + "grad_norm": 0.44799864949028395, + "learning_rate": 8.084333804788958e-08, + "loss": 0.2374, + "step": 5681 + }, + { + "epoch": 1.897161936560935, + "grad_norm": 0.4549289654285618, + "learning_rate": 8.032233013072821e-08, + "loss": 0.2556, + "step": 5682 + }, + { + "epoch": 1.8974958263772956, + "grad_norm": 0.45725609409989937, + "learning_rate": 7.980299291066696e-08, + "loss": 0.2509, + "step": 5683 + }, + { + "epoch": 1.8978297161936561, + "grad_norm": 0.44352701644852544, + "learning_rate": 7.928532656407029e-08, + "loss": 0.2413, + "step": 5684 + }, + { + "epoch": 1.8981636060100167, + "grad_norm": 0.4411579267233682, + "learning_rate": 7.876933126673481e-08, + "loss": 0.2291, + "step": 5685 + }, + { + "epoch": 1.8984974958263772, + "grad_norm": 0.4550226608765151, + "learning_rate": 7.825500719389035e-08, + "loss": 0.2461, + "step": 5686 + }, + { + "epoch": 1.898831385642738, + "grad_norm": 0.43841041897510513, + "learning_rate": 7.774235452019718e-08, + "loss": 0.2411, + "step": 5687 + }, + { + "epoch": 1.8991652754590986, + "grad_norm": 0.4635783706520524, + "learning_rate": 7.723137341975107e-08, + "loss": 0.2444, + "step": 5688 + }, + { + "epoch": 1.8994991652754591, + "grad_norm": 0.4645108169258164, + "learning_rate": 7.672206406607763e-08, + "loss": 0.2555, + "step": 5689 + }, + { + "epoch": 1.8998330550918197, + "grad_norm": 0.43206658887765215, + "learning_rate": 7.62144266321363e-08, + "loss": 0.2406, + "step": 5690 + }, + { + "epoch": 1.9001669449081802, + "grad_norm": 0.4440335635742115, + "learning_rate": 7.570846129031861e-08, + "loss": 0.2446, + "step": 5691 + }, + { + "epoch": 1.900500834724541, + "grad_norm": 0.44442950062145675, + "learning_rate": 7.520416821244603e-08, + "loss": 0.2401, + "step": 5692 + }, + { + "epoch": 1.9008347245409016, + "grad_norm": 0.4343096538522302, + "learning_rate": 7.470154756977544e-08, + "loss": 0.2454, + "step": 5693 + }, + { + "epoch": 1.9011686143572621, + "grad_norm": 0.44472759839908554, + "learning_rate": 7.420059953299363e-08, + "loss": 0.2538, + "step": 5694 + }, + { + "epoch": 1.9015025041736227, + "grad_norm": 0.45427154937116765, + "learning_rate": 7.37013242722201e-08, + "loss": 0.2526, + "step": 5695 + }, + { + "epoch": 1.9018363939899832, + "grad_norm": 0.4424905005784402, + "learning_rate": 7.320372195700532e-08, + "loss": 0.2512, + "step": 5696 + }, + { + "epoch": 1.902170283806344, + "grad_norm": 0.4488859362543764, + "learning_rate": 7.270779275633411e-08, + "loss": 0.242, + "step": 5697 + }, + { + "epoch": 1.9025041736227046, + "grad_norm": 0.4510596780595988, + "learning_rate": 7.22135368386201e-08, + "loss": 0.2462, + "step": 5698 + }, + { + "epoch": 1.9028380634390651, + "grad_norm": 0.4510736476451831, + "learning_rate": 7.172095437171067e-08, + "loss": 0.2529, + "step": 5699 + }, + { + "epoch": 1.9031719532554257, + "grad_norm": 0.45066017136976994, + "learning_rate": 7.123004552288427e-08, + "loss": 0.2563, + "step": 5700 + }, + { + "epoch": 1.9035058430717862, + "grad_norm": 0.43945685841381715, + "learning_rate": 7.074081045885028e-08, + "loss": 0.2423, + "step": 5701 + }, + { + "epoch": 1.903839732888147, + "grad_norm": 0.4412762591159135, + "learning_rate": 7.02532493457514e-08, + "loss": 0.2362, + "step": 5702 + }, + { + "epoch": 1.9041736227045076, + "grad_norm": 0.4507465375132684, + "learning_rate": 6.97673623491607e-08, + "loss": 0.2385, + "step": 5703 + }, + { + "epoch": 1.9045075125208681, + "grad_norm": 0.4445232219998077, + "learning_rate": 6.928314963408234e-08, + "loss": 0.2432, + "step": 5704 + }, + { + "epoch": 1.9048414023372287, + "grad_norm": 0.46467350324380635, + "learning_rate": 6.880061136495252e-08, + "loss": 0.2527, + "step": 5705 + }, + { + "epoch": 1.9051752921535892, + "grad_norm": 0.45641934936311224, + "learning_rate": 6.831974770563909e-08, + "loss": 0.2474, + "step": 5706 + }, + { + "epoch": 1.90550918196995, + "grad_norm": 0.44073775289389144, + "learning_rate": 6.784055881944085e-08, + "loss": 0.2399, + "step": 5707 + }, + { + "epoch": 1.9058430717863106, + "grad_norm": 0.4421160885477975, + "learning_rate": 6.736304486908763e-08, + "loss": 0.2434, + "step": 5708 + }, + { + "epoch": 1.9061769616026711, + "grad_norm": 0.45883120131870003, + "learning_rate": 6.688720601674137e-08, + "loss": 0.246, + "step": 5709 + }, + { + "epoch": 1.9065108514190316, + "grad_norm": 0.44569849489266805, + "learning_rate": 6.641304242399282e-08, + "loss": 0.2402, + "step": 5710 + }, + { + "epoch": 1.9068447412353922, + "grad_norm": 0.45097022995084074, + "learning_rate": 6.594055425186763e-08, + "loss": 0.2465, + "step": 5711 + }, + { + "epoch": 1.907178631051753, + "grad_norm": 0.4406023883703869, + "learning_rate": 6.54697416608191e-08, + "loss": 0.2505, + "step": 5712 + }, + { + "epoch": 1.9075125208681136, + "grad_norm": 0.4565296210219915, + "learning_rate": 6.500060481073267e-08, + "loss": 0.2468, + "step": 5713 + }, + { + "epoch": 1.9078464106844741, + "grad_norm": 0.45275723763240444, + "learning_rate": 6.453314386092535e-08, + "loss": 0.2475, + "step": 5714 + }, + { + "epoch": 1.9081803005008346, + "grad_norm": 0.4423086956420608, + "learning_rate": 6.406735897014348e-08, + "loss": 0.2461, + "step": 5715 + }, + { + "epoch": 1.9085141903171954, + "grad_norm": 0.4562776611171476, + "learning_rate": 6.360325029656667e-08, + "loss": 0.2484, + "step": 5716 + }, + { + "epoch": 1.908848080133556, + "grad_norm": 0.4587860325774291, + "learning_rate": 6.314081799780269e-08, + "loss": 0.2491, + "step": 5717 + }, + { + "epoch": 1.9091819699499166, + "grad_norm": 0.4565299892394236, + "learning_rate": 6.268006223089207e-08, + "loss": 0.2404, + "step": 5718 + }, + { + "epoch": 1.9095158597662771, + "grad_norm": 0.4524672850611653, + "learning_rate": 6.222098315230407e-08, + "loss": 0.2528, + "step": 5719 + }, + { + "epoch": 1.9098497495826376, + "grad_norm": 0.44801853120248725, + "learning_rate": 6.176358091794011e-08, + "loss": 0.2411, + "step": 5720 + }, + { + "epoch": 1.9101836393989984, + "grad_norm": 0.44552003699971887, + "learning_rate": 6.130785568313202e-08, + "loss": 0.2431, + "step": 5721 + }, + { + "epoch": 1.910517529215359, + "grad_norm": 0.470582502452664, + "learning_rate": 6.085380760264103e-08, + "loss": 0.2613, + "step": 5722 + }, + { + "epoch": 1.9108514190317196, + "grad_norm": 0.44058338633914723, + "learning_rate": 6.040143683065936e-08, + "loss": 0.242, + "step": 5723 + }, + { + "epoch": 1.9111853088480801, + "grad_norm": 0.45764473446483006, + "learning_rate": 5.995074352081021e-08, + "loss": 0.2532, + "step": 5724 + }, + { + "epoch": 1.9115191986644406, + "grad_norm": 0.4427463715209946, + "learning_rate": 5.950172782614672e-08, + "loss": 0.2445, + "step": 5725 + }, + { + "epoch": 1.9118530884808014, + "grad_norm": 0.44615254277866917, + "learning_rate": 5.905438989915246e-08, + "loss": 0.2508, + "step": 5726 + }, + { + "epoch": 1.9121869782971619, + "grad_norm": 0.45521391807968614, + "learning_rate": 5.860872989174149e-08, + "loss": 0.235, + "step": 5727 + }, + { + "epoch": 1.9125208681135226, + "grad_norm": 0.43808248386056126, + "learning_rate": 5.8164747955256064e-08, + "loss": 0.2473, + "step": 5728 + }, + { + "epoch": 1.9128547579298831, + "grad_norm": 0.4409672591534923, + "learning_rate": 5.772244424047169e-08, + "loss": 0.243, + "step": 5729 + }, + { + "epoch": 1.9131886477462436, + "grad_norm": 0.4396703384067556, + "learning_rate": 5.728181889759155e-08, + "loss": 0.2541, + "step": 5730 + }, + { + "epoch": 1.9135225375626044, + "grad_norm": 0.4445314621000635, + "learning_rate": 5.684287207625039e-08, + "loss": 0.2388, + "step": 5731 + }, + { + "epoch": 1.9138564273789649, + "grad_norm": 0.44696044913182503, + "learning_rate": 5.6405603925511196e-08, + "loss": 0.2385, + "step": 5732 + }, + { + "epoch": 1.9141903171953256, + "grad_norm": 0.4609708709505417, + "learning_rate": 5.597001459386908e-08, + "loss": 0.2443, + "step": 5733 + }, + { + "epoch": 1.9145242070116861, + "grad_norm": 0.4596449067239604, + "learning_rate": 5.5536104229247934e-08, + "loss": 0.2568, + "step": 5734 + }, + { + "epoch": 1.9148580968280466, + "grad_norm": 0.4373856035577545, + "learning_rate": 5.510387297900099e-08, + "loss": 0.2397, + "step": 5735 + }, + { + "epoch": 1.9151919866444074, + "grad_norm": 0.43622067974664847, + "learning_rate": 5.46733209899114e-08, + "loss": 0.2362, + "step": 5736 + }, + { + "epoch": 1.915525876460768, + "grad_norm": 0.4459892121581979, + "learning_rate": 5.4244448408192764e-08, + "loss": 0.238, + "step": 5737 + }, + { + "epoch": 1.9158597662771286, + "grad_norm": 0.4500207102253361, + "learning_rate": 5.3817255379488565e-08, + "loss": 0.2402, + "step": 5738 + }, + { + "epoch": 1.9161936560934891, + "grad_norm": 0.44588822961587615, + "learning_rate": 5.33917420488711e-08, + "loss": 0.2362, + "step": 5739 + }, + { + "epoch": 1.9165275459098496, + "grad_norm": 0.4740871907417988, + "learning_rate": 5.2967908560841444e-08, + "loss": 0.2517, + "step": 5740 + }, + { + "epoch": 1.9168614357262104, + "grad_norm": 0.44079544894565303, + "learning_rate": 5.254575505933279e-08, + "loss": 0.2391, + "step": 5741 + }, + { + "epoch": 1.917195325542571, + "grad_norm": 0.47332542767652436, + "learning_rate": 5.2125281687704924e-08, + "loss": 0.257, + "step": 5742 + }, + { + "epoch": 1.9175292153589316, + "grad_norm": 0.43797902477112266, + "learning_rate": 5.170648858874916e-08, + "loss": 0.2385, + "step": 5743 + }, + { + "epoch": 1.9178631051752921, + "grad_norm": 0.4581676603329248, + "learning_rate": 5.1289375904685637e-08, + "loss": 0.2535, + "step": 5744 + }, + { + "epoch": 1.9181969949916526, + "grad_norm": 0.4640479560124321, + "learning_rate": 5.0873943777163814e-08, + "loss": 0.247, + "step": 5745 + }, + { + "epoch": 1.9185308848080134, + "grad_norm": 0.45023986873585764, + "learning_rate": 5.046019234726085e-08, + "loss": 0.2487, + "step": 5746 + }, + { + "epoch": 1.918864774624374, + "grad_norm": 0.44202238039149255, + "learning_rate": 5.004812175548657e-08, + "loss": 0.2457, + "step": 5747 + }, + { + "epoch": 1.9191986644407346, + "grad_norm": 0.4559764400992642, + "learning_rate": 4.963773214177681e-08, + "loss": 0.2433, + "step": 5748 + }, + { + "epoch": 1.9195325542570951, + "grad_norm": 0.46383489027424124, + "learning_rate": 4.922902364549731e-08, + "loss": 0.2574, + "step": 5749 + }, + { + "epoch": 1.9198664440734556, + "grad_norm": 0.4401495141943513, + "learning_rate": 4.8821996405444825e-08, + "loss": 0.2395, + "step": 5750 + }, + { + "epoch": 1.9202003338898164, + "grad_norm": 0.45058034024364085, + "learning_rate": 4.8416650559842685e-08, + "loss": 0.2354, + "step": 5751 + }, + { + "epoch": 1.920534223706177, + "grad_norm": 0.4418920726842569, + "learning_rate": 4.8012986246344653e-08, + "loss": 0.2381, + "step": 5752 + }, + { + "epoch": 1.9208681135225376, + "grad_norm": 0.43635719503467474, + "learning_rate": 4.761100360203274e-08, + "loss": 0.2396, + "step": 5753 + }, + { + "epoch": 1.9212020033388981, + "grad_norm": 0.4667686655891189, + "learning_rate": 4.7210702763418324e-08, + "loss": 0.2421, + "step": 5754 + }, + { + "epoch": 1.9215358931552586, + "grad_norm": 0.4564314903837557, + "learning_rate": 4.6812083866441535e-08, + "loss": 0.2357, + "step": 5755 + }, + { + "epoch": 1.9218697829716194, + "grad_norm": 0.4420348746592532, + "learning_rate": 4.641514704647132e-08, + "loss": 0.243, + "step": 5756 + }, + { + "epoch": 1.92220367278798, + "grad_norm": 0.44861749506327153, + "learning_rate": 4.6019892438304845e-08, + "loss": 0.2456, + "step": 5757 + }, + { + "epoch": 1.9225375626043406, + "grad_norm": 0.45733757949912346, + "learning_rate": 4.562632017616975e-08, + "loss": 0.2554, + "step": 5758 + }, + { + "epoch": 1.9228714524207011, + "grad_norm": 0.4418497042979992, + "learning_rate": 4.523443039372022e-08, + "loss": 0.2356, + "step": 5759 + }, + { + "epoch": 1.9232053422370616, + "grad_norm": 0.4689028183974601, + "learning_rate": 4.484422322403981e-08, + "loss": 0.2548, + "step": 5760 + }, + { + "epoch": 1.9235392320534224, + "grad_norm": 0.43228389491953434, + "learning_rate": 4.44556987996414e-08, + "loss": 0.2382, + "step": 5761 + }, + { + "epoch": 1.923873121869783, + "grad_norm": 0.4411343808071002, + "learning_rate": 4.406885725246557e-08, + "loss": 0.2362, + "step": 5762 + }, + { + "epoch": 1.9242070116861436, + "grad_norm": 0.4650767407896124, + "learning_rate": 4.3683698713882226e-08, + "loss": 0.2517, + "step": 5763 + }, + { + "epoch": 1.9245409015025041, + "grad_norm": 0.4392472990753455, + "learning_rate": 4.3300223314688416e-08, + "loss": 0.244, + "step": 5764 + }, + { + "epoch": 1.9248747913188646, + "grad_norm": 0.45530649540094636, + "learning_rate": 4.2918431185110523e-08, + "loss": 0.2379, + "step": 5765 + }, + { + "epoch": 1.9252086811352254, + "grad_norm": 0.4597932600641774, + "learning_rate": 4.253832245480372e-08, + "loss": 0.2487, + "step": 5766 + }, + { + "epoch": 1.925542570951586, + "grad_norm": 0.45064887395576086, + "learning_rate": 4.2159897252850854e-08, + "loss": 0.2472, + "step": 5767 + }, + { + "epoch": 1.9258764607679466, + "grad_norm": 0.45180438966822445, + "learning_rate": 4.178315570776303e-08, + "loss": 0.243, + "step": 5768 + }, + { + "epoch": 1.9262103505843071, + "grad_norm": 0.4574623710607537, + "learning_rate": 4.140809794747958e-08, + "loss": 0.2507, + "step": 5769 + }, + { + "epoch": 1.9265442404006676, + "grad_norm": 0.4527984209950441, + "learning_rate": 4.103472409936859e-08, + "loss": 0.2503, + "step": 5770 + }, + { + "epoch": 1.9268781302170284, + "grad_norm": 0.44908973879786246, + "learning_rate": 4.0663034290225333e-08, + "loss": 0.2412, + "step": 5771 + }, + { + "epoch": 1.927212020033389, + "grad_norm": 0.43687828990827393, + "learning_rate": 4.029302864627438e-08, + "loss": 0.2493, + "step": 5772 + }, + { + "epoch": 1.9275459098497496, + "grad_norm": 0.45031813969573564, + "learning_rate": 3.992470729316689e-08, + "loss": 0.2408, + "step": 5773 + }, + { + "epoch": 1.9278797996661101, + "grad_norm": 0.46517446566566306, + "learning_rate": 3.9558070355983356e-08, + "loss": 0.2421, + "step": 5774 + }, + { + "epoch": 1.9282136894824706, + "grad_norm": 0.4566299141043284, + "learning_rate": 3.919311795923253e-08, + "loss": 0.2417, + "step": 5775 + }, + { + "epoch": 1.9285475792988314, + "grad_norm": 0.4559519431019001, + "learning_rate": 3.882985022684915e-08, + "loss": 0.2397, + "step": 5776 + }, + { + "epoch": 1.928881469115192, + "grad_norm": 0.461053808349028, + "learning_rate": 3.846826728219788e-08, + "loss": 0.2439, + "step": 5777 + }, + { + "epoch": 1.9292153589315526, + "grad_norm": 0.46122191421384084, + "learning_rate": 3.8108369248069376e-08, + "loss": 0.2573, + "step": 5778 + }, + { + "epoch": 1.9295492487479131, + "grad_norm": 0.4703311137933205, + "learning_rate": 3.775015624668477e-08, + "loss": 0.2528, + "step": 5779 + }, + { + "epoch": 1.9298831385642736, + "grad_norm": 0.46522282950398053, + "learning_rate": 3.739362839968952e-08, + "loss": 0.2387, + "step": 5780 + }, + { + "epoch": 1.9302170283806344, + "grad_norm": 0.4728471147728339, + "learning_rate": 3.703878582816012e-08, + "loss": 0.2498, + "step": 5781 + }, + { + "epoch": 1.930550918196995, + "grad_norm": 0.4684190195768486, + "learning_rate": 3.66856286525985e-08, + "loss": 0.2448, + "step": 5782 + }, + { + "epoch": 1.9308848080133556, + "grad_norm": 0.45733434863577344, + "learning_rate": 3.633415699293541e-08, + "loss": 0.2622, + "step": 5783 + }, + { + "epoch": 1.9312186978297161, + "grad_norm": 0.44082222504516166, + "learning_rate": 3.598437096852869e-08, + "loss": 0.2484, + "step": 5784 + }, + { + "epoch": 1.9315525876460768, + "grad_norm": 0.45150850683423044, + "learning_rate": 3.5636270698163336e-08, + "loss": 0.2473, + "step": 5785 + }, + { + "epoch": 1.9318864774624374, + "grad_norm": 0.44610904134127866, + "learning_rate": 3.528985630005366e-08, + "loss": 0.2477, + "step": 5786 + }, + { + "epoch": 1.932220367278798, + "grad_norm": 0.4431710268769536, + "learning_rate": 3.494512789183835e-08, + "loss": 0.2414, + "step": 5787 + }, + { + "epoch": 1.9325542570951586, + "grad_norm": 0.4603329421633306, + "learning_rate": 3.4602085590587084e-08, + "loss": 0.2517, + "step": 5788 + }, + { + "epoch": 1.9328881469115191, + "grad_norm": 0.4572137928308264, + "learning_rate": 3.426072951279502e-08, + "loss": 0.2553, + "step": 5789 + }, + { + "epoch": 1.9332220367278798, + "grad_norm": 0.44770351263863556, + "learning_rate": 3.392105977438387e-08, + "loss": 0.2484, + "step": 5790 + }, + { + "epoch": 1.9335559265442404, + "grad_norm": 0.4515470006635667, + "learning_rate": 3.3583076490704134e-08, + "loss": 0.2432, + "step": 5791 + }, + { + "epoch": 1.933889816360601, + "grad_norm": 0.4463100811552544, + "learning_rate": 3.324677977653401e-08, + "loss": 0.2408, + "step": 5792 + }, + { + "epoch": 1.9342237061769616, + "grad_norm": 0.45791497931498726, + "learning_rate": 3.2912169746077136e-08, + "loss": 0.2494, + "step": 5793 + }, + { + "epoch": 1.934557595993322, + "grad_norm": 0.47737354092369577, + "learning_rate": 3.2579246512965954e-08, + "loss": 0.2599, + "step": 5794 + }, + { + "epoch": 1.9348914858096828, + "grad_norm": 0.45268827945499734, + "learning_rate": 3.224801019025947e-08, + "loss": 0.2473, + "step": 5795 + }, + { + "epoch": 1.9352253756260434, + "grad_norm": 0.44726909816283333, + "learning_rate": 3.1918460890443256e-08, + "loss": 0.249, + "step": 5796 + }, + { + "epoch": 1.935559265442404, + "grad_norm": 0.4398624791257476, + "learning_rate": 3.159059872543113e-08, + "loss": 0.2367, + "step": 5797 + }, + { + "epoch": 1.9358931552587646, + "grad_norm": 0.46156961876182695, + "learning_rate": 3.126442380656292e-08, + "loss": 0.2559, + "step": 5798 + }, + { + "epoch": 1.936227045075125, + "grad_norm": 0.45534941942899193, + "learning_rate": 3.093993624460612e-08, + "loss": 0.2473, + "step": 5799 + }, + { + "epoch": 1.9365609348914858, + "grad_norm": 0.4539607324674014, + "learning_rate": 3.0617136149755365e-08, + "loss": 0.2469, + "step": 5800 + }, + { + "epoch": 1.9368948247078466, + "grad_norm": 0.45436079073139063, + "learning_rate": 3.0296023631631866e-08, + "loss": 0.2516, + "step": 5801 + }, + { + "epoch": 1.937228714524207, + "grad_norm": 0.44074074354535697, + "learning_rate": 2.997659879928283e-08, + "loss": 0.2275, + "step": 5802 + }, + { + "epoch": 1.9375626043405676, + "grad_norm": 0.4558012280395735, + "learning_rate": 2.965886176118482e-08, + "loss": 0.2535, + "step": 5803 + }, + { + "epoch": 1.937896494156928, + "grad_norm": 0.4675894435180009, + "learning_rate": 2.9342812625238194e-08, + "loss": 0.2429, + "step": 5804 + }, + { + "epoch": 1.9382303839732888, + "grad_norm": 0.4574854982853331, + "learning_rate": 2.9028451498772646e-08, + "loss": 0.2403, + "step": 5805 + }, + { + "epoch": 1.9385642737896496, + "grad_norm": 0.45941059049884597, + "learning_rate": 2.8715778488543344e-08, + "loss": 0.2479, + "step": 5806 + }, + { + "epoch": 1.93889816360601, + "grad_norm": 0.4479108700718104, + "learning_rate": 2.8404793700732012e-08, + "loss": 0.2517, + "step": 5807 + }, + { + "epoch": 1.9392320534223706, + "grad_norm": 0.44853219727819377, + "learning_rate": 2.8095497240947512e-08, + "loss": 0.2516, + "step": 5808 + }, + { + "epoch": 1.939565943238731, + "grad_norm": 0.45471605163304346, + "learning_rate": 2.7787889214225817e-08, + "loss": 0.2406, + "step": 5809 + }, + { + "epoch": 1.9398998330550918, + "grad_norm": 0.46518892613635926, + "learning_rate": 2.7481969725028923e-08, + "loss": 0.2514, + "step": 5810 + }, + { + "epoch": 1.9402337228714526, + "grad_norm": 0.4681319175694748, + "learning_rate": 2.7177738877244842e-08, + "loss": 0.2508, + "step": 5811 + }, + { + "epoch": 1.940567612687813, + "grad_norm": 0.4544112821102379, + "learning_rate": 2.6875196774189815e-08, + "loss": 0.2446, + "step": 5812 + }, + { + "epoch": 1.9409015025041736, + "grad_norm": 0.44769935596581867, + "learning_rate": 2.6574343518604995e-08, + "loss": 0.2487, + "step": 5813 + }, + { + "epoch": 1.941235392320534, + "grad_norm": 0.46102911220286585, + "learning_rate": 2.6275179212658652e-08, + "loss": 0.2518, + "step": 5814 + }, + { + "epoch": 1.9415692821368948, + "grad_norm": 0.45523929958663717, + "learning_rate": 2.5977703957945632e-08, + "loss": 0.2509, + "step": 5815 + }, + { + "epoch": 1.9419031719532556, + "grad_norm": 0.4285403554250903, + "learning_rate": 2.5681917855486238e-08, + "loss": 0.2311, + "step": 5816 + }, + { + "epoch": 1.942237061769616, + "grad_norm": 0.46309950815760265, + "learning_rate": 2.5387821005729008e-08, + "loss": 0.2412, + "step": 5817 + }, + { + "epoch": 1.9425709515859766, + "grad_norm": 0.4573879506438098, + "learning_rate": 2.5095413508547385e-08, + "loss": 0.2524, + "step": 5818 + }, + { + "epoch": 1.942904841402337, + "grad_norm": 0.4294548191015719, + "learning_rate": 2.4804695463240825e-08, + "loss": 0.2344, + "step": 5819 + }, + { + "epoch": 1.9432387312186978, + "grad_norm": 0.4323550176494346, + "learning_rate": 2.451566696853591e-08, + "loss": 0.2402, + "step": 5820 + }, + { + "epoch": 1.9435726210350586, + "grad_norm": 0.4526665586078463, + "learning_rate": 2.4228328122585797e-08, + "loss": 0.2547, + "step": 5821 + }, + { + "epoch": 1.943906510851419, + "grad_norm": 0.4491897242633116, + "learning_rate": 2.394267902296854e-08, + "loss": 0.2481, + "step": 5822 + }, + { + "epoch": 1.9442404006677796, + "grad_norm": 0.45724048376117304, + "learning_rate": 2.3658719766689874e-08, + "loss": 0.2447, + "step": 5823 + }, + { + "epoch": 1.94457429048414, + "grad_norm": 0.4505239559766953, + "learning_rate": 2.3376450450179888e-08, + "loss": 0.2513, + "step": 5824 + }, + { + "epoch": 1.9449081803005008, + "grad_norm": 0.43999220214324325, + "learning_rate": 2.30958711692969e-08, + "loss": 0.2454, + "step": 5825 + }, + { + "epoch": 1.9452420701168616, + "grad_norm": 0.44543763677704673, + "learning_rate": 2.281698201932303e-08, + "loss": 0.2431, + "step": 5826 + }, + { + "epoch": 1.945575959933222, + "grad_norm": 0.4631221978517737, + "learning_rate": 2.253978309496918e-08, + "loss": 0.2417, + "step": 5827 + }, + { + "epoch": 1.9459098497495826, + "grad_norm": 0.4495744456346734, + "learning_rate": 2.226427449036894e-08, + "loss": 0.2457, + "step": 5828 + }, + { + "epoch": 1.946243739565943, + "grad_norm": 0.4343153892701046, + "learning_rate": 2.199045629908414e-08, + "loss": 0.2476, + "step": 5829 + }, + { + "epoch": 1.9465776293823038, + "grad_norm": 0.4532026535251867, + "learning_rate": 2.1718328614102614e-08, + "loss": 0.2461, + "step": 5830 + }, + { + "epoch": 1.9469115191986646, + "grad_norm": 0.45080496799196973, + "learning_rate": 2.1447891527837107e-08, + "loss": 0.2407, + "step": 5831 + }, + { + "epoch": 1.947245409015025, + "grad_norm": 0.4340880950769532, + "learning_rate": 2.1179145132126933e-08, + "loss": 0.2332, + "step": 5832 + }, + { + "epoch": 1.9475792988313856, + "grad_norm": 0.4647211518268712, + "learning_rate": 2.091208951823631e-08, + "loss": 0.2532, + "step": 5833 + }, + { + "epoch": 1.947913188647746, + "grad_norm": 0.44387827074073966, + "learning_rate": 2.0646724776856586e-08, + "loss": 0.2355, + "step": 5834 + }, + { + "epoch": 1.9482470784641068, + "grad_norm": 0.46284819481771183, + "learning_rate": 2.0383050998104004e-08, + "loss": 0.2582, + "step": 5835 + }, + { + "epoch": 1.9485809682804676, + "grad_norm": 0.45679745283034584, + "learning_rate": 2.0121068271520826e-08, + "loss": 0.2579, + "step": 5836 + }, + { + "epoch": 1.948914858096828, + "grad_norm": 0.4452832214194113, + "learning_rate": 1.9860776686075332e-08, + "loss": 0.2469, + "step": 5837 + }, + { + "epoch": 1.9492487479131886, + "grad_norm": 0.45555386673944787, + "learning_rate": 1.960217633016015e-08, + "loss": 0.2525, + "step": 5838 + }, + { + "epoch": 1.949582637729549, + "grad_norm": 0.46333888685258917, + "learning_rate": 1.9345267291596136e-08, + "loss": 0.2434, + "step": 5839 + }, + { + "epoch": 1.9499165275459098, + "grad_norm": 0.45839203517982163, + "learning_rate": 1.9090049657627397e-08, + "loss": 0.2422, + "step": 5840 + }, + { + "epoch": 1.9502504173622706, + "grad_norm": 0.4560735997503801, + "learning_rate": 1.88365235149246e-08, + "loss": 0.24, + "step": 5841 + }, + { + "epoch": 1.950584307178631, + "grad_norm": 0.4517601024546749, + "learning_rate": 1.8584688949583872e-08, + "loss": 0.2471, + "step": 5842 + }, + { + "epoch": 1.9509181969949916, + "grad_norm": 0.4577325654554013, + "learning_rate": 1.833454604712681e-08, + "loss": 0.2497, + "step": 5843 + }, + { + "epoch": 1.951252086811352, + "grad_norm": 0.46835315674636785, + "learning_rate": 1.8086094892501015e-08, + "loss": 0.2486, + "step": 5844 + }, + { + "epoch": 1.9515859766277128, + "grad_norm": 0.43668842790468293, + "learning_rate": 1.7839335570079553e-08, + "loss": 0.239, + "step": 5845 + }, + { + "epoch": 1.9519198664440736, + "grad_norm": 0.44873118906648674, + "learning_rate": 1.7594268163659277e-08, + "loss": 0.2427, + "step": 5846 + }, + { + "epoch": 1.952253756260434, + "grad_norm": 0.44286153090543035, + "learning_rate": 1.7350892756465286e-08, + "loss": 0.2538, + "step": 5847 + }, + { + "epoch": 1.9525876460767946, + "grad_norm": 0.458866108407277, + "learning_rate": 1.7109209431145913e-08, + "loss": 0.2508, + "step": 5848 + }, + { + "epoch": 1.9529215358931553, + "grad_norm": 0.4456207236237469, + "learning_rate": 1.68692182697755e-08, + "loss": 0.2508, + "step": 5849 + }, + { + "epoch": 1.9532554257095158, + "grad_norm": 0.44601025954369183, + "learning_rate": 1.663091935385386e-08, + "loss": 0.2437, + "step": 5850 + }, + { + "epoch": 1.9535893155258766, + "grad_norm": 0.4623837112097835, + "learning_rate": 1.639431276430625e-08, + "loss": 0.2469, + "step": 5851 + }, + { + "epoch": 1.953923205342237, + "grad_norm": 0.45795006095919527, + "learning_rate": 1.6159398581482855e-08, + "loss": 0.2455, + "step": 5852 + }, + { + "epoch": 1.9542570951585976, + "grad_norm": 0.4577881319135945, + "learning_rate": 1.5926176885159296e-08, + "loss": 0.2423, + "step": 5853 + }, + { + "epoch": 1.9545909849749583, + "grad_norm": 0.4619187109838988, + "learning_rate": 1.5694647754537217e-08, + "loss": 0.2529, + "step": 5854 + }, + { + "epoch": 1.9549248747913188, + "grad_norm": 0.44438582831769774, + "learning_rate": 1.546481126824151e-08, + "loss": 0.2417, + "step": 5855 + }, + { + "epoch": 1.9552587646076796, + "grad_norm": 0.4563607247667431, + "learning_rate": 1.523666750432362e-08, + "loss": 0.2516, + "step": 5856 + }, + { + "epoch": 1.95559265442404, + "grad_norm": 0.43894044429145707, + "learning_rate": 1.5010216540261026e-08, + "loss": 0.2325, + "step": 5857 + }, + { + "epoch": 1.9559265442404006, + "grad_norm": 0.45742563580237294, + "learning_rate": 1.4785458452954427e-08, + "loss": 0.2464, + "step": 5858 + }, + { + "epoch": 1.9562604340567613, + "grad_norm": 0.4549712965056846, + "learning_rate": 1.4562393318731106e-08, + "loss": 0.2437, + "step": 5859 + }, + { + "epoch": 1.9565943238731218, + "grad_norm": 0.45833982574702276, + "learning_rate": 1.4341021213342133e-08, + "loss": 0.2433, + "step": 5860 + }, + { + "epoch": 1.9569282136894826, + "grad_norm": 0.43499196195334705, + "learning_rate": 1.4121342211964595e-08, + "loss": 0.2431, + "step": 5861 + }, + { + "epoch": 1.957262103505843, + "grad_norm": 0.46603305582833854, + "learning_rate": 1.3903356389200484e-08, + "loss": 0.2477, + "step": 5862 + }, + { + "epoch": 1.9575959933222036, + "grad_norm": 0.45308112107190773, + "learning_rate": 1.3687063819076697e-08, + "loss": 0.2493, + "step": 5863 + }, + { + "epoch": 1.9579298831385643, + "grad_norm": 0.45470228466138113, + "learning_rate": 1.347246457504503e-08, + "loss": 0.238, + "step": 5864 + }, + { + "epoch": 1.9582637729549248, + "grad_norm": 0.4392821316165472, + "learning_rate": 1.3259558729981637e-08, + "loss": 0.2316, + "step": 5865 + }, + { + "epoch": 1.9585976627712856, + "grad_norm": 0.4512486002436541, + "learning_rate": 1.3048346356189234e-08, + "loss": 0.2556, + "step": 5866 + }, + { + "epoch": 1.958931552587646, + "grad_norm": 0.4434941777443223, + "learning_rate": 1.2838827525394337e-08, + "loss": 0.2413, + "step": 5867 + }, + { + "epoch": 1.9592654424040066, + "grad_norm": 0.4383254468817122, + "learning_rate": 1.2631002308747809e-08, + "loss": 0.2425, + "step": 5868 + }, + { + "epoch": 1.9595993322203673, + "grad_norm": 0.46346404195955604, + "learning_rate": 1.2424870776826526e-08, + "loss": 0.2464, + "step": 5869 + }, + { + "epoch": 1.959933222036728, + "grad_norm": 0.45236341389817364, + "learning_rate": 1.222043299963116e-08, + "loss": 0.2366, + "step": 5870 + }, + { + "epoch": 1.9602671118530886, + "grad_norm": 0.4557548149029034, + "learning_rate": 1.20176890465884e-08, + "loss": 0.2456, + "step": 5871 + }, + { + "epoch": 1.960601001669449, + "grad_norm": 0.46237645808331446, + "learning_rate": 1.1816638986548723e-08, + "loss": 0.2554, + "step": 5872 + }, + { + "epoch": 1.9609348914858096, + "grad_norm": 0.4722951239934201, + "learning_rate": 1.1617282887787517e-08, + "loss": 0.2538, + "step": 5873 + }, + { + "epoch": 1.9612687813021703, + "grad_norm": 0.44474043568830146, + "learning_rate": 1.1419620818005072e-08, + "loss": 0.244, + "step": 5874 + }, + { + "epoch": 1.961602671118531, + "grad_norm": 0.4581728134731523, + "learning_rate": 1.1223652844327137e-08, + "loss": 0.2428, + "step": 5875 + }, + { + "epoch": 1.9619365609348915, + "grad_norm": 0.4502884806662179, + "learning_rate": 1.1029379033302145e-08, + "loss": 0.2411, + "step": 5876 + }, + { + "epoch": 1.962270450751252, + "grad_norm": 0.4477509357609229, + "learning_rate": 1.083679945090621e-08, + "loss": 0.2502, + "step": 5877 + }, + { + "epoch": 1.9626043405676126, + "grad_norm": 0.44310683655552124, + "learning_rate": 1.0645914162536463e-08, + "loss": 0.2532, + "step": 5878 + }, + { + "epoch": 1.9629382303839733, + "grad_norm": 0.45482342135372894, + "learning_rate": 1.0456723233017718e-08, + "loss": 0.2502, + "step": 5879 + }, + { + "epoch": 1.963272120200334, + "grad_norm": 0.46080747831745683, + "learning_rate": 1.026922672659858e-08, + "loss": 0.2552, + "step": 5880 + }, + { + "epoch": 1.9636060100166945, + "grad_norm": 0.4661887169635394, + "learning_rate": 1.0083424706950894e-08, + "loss": 0.2524, + "step": 5881 + }, + { + "epoch": 1.963939899833055, + "grad_norm": 0.4476036657870536, + "learning_rate": 9.899317237172524e-09, + "loss": 0.2472, + "step": 5882 + }, + { + "epoch": 1.9642737896494156, + "grad_norm": 0.4530585237237698, + "learning_rate": 9.71690437978512e-09, + "loss": 0.2439, + "step": 5883 + }, + { + "epoch": 1.9646076794657763, + "grad_norm": 0.44694485875187673, + "learning_rate": 9.536186196735242e-09, + "loss": 0.2537, + "step": 5884 + }, + { + "epoch": 1.964941569282137, + "grad_norm": 0.43382938726091885, + "learning_rate": 9.357162749394355e-09, + "loss": 0.2467, + "step": 5885 + }, + { + "epoch": 1.9652754590984975, + "grad_norm": 0.44835361905161436, + "learning_rate": 9.17983409855716e-09, + "loss": 0.2413, + "step": 5886 + }, + { + "epoch": 1.965609348914858, + "grad_norm": 0.4458084655019593, + "learning_rate": 9.004200304444377e-09, + "loss": 0.2273, + "step": 5887 + }, + { + "epoch": 1.9659432387312186, + "grad_norm": 0.4443755564699868, + "learning_rate": 8.83026142669996e-09, + "loss": 0.2452, + "step": 5888 + }, + { + "epoch": 1.9662771285475793, + "grad_norm": 0.44650962167144154, + "learning_rate": 8.658017524392216e-09, + "loss": 0.2506, + "step": 5889 + }, + { + "epoch": 1.96661101836394, + "grad_norm": 0.4539774098093091, + "learning_rate": 8.487468656014908e-09, + "loss": 0.2394, + "step": 5890 + }, + { + "epoch": 1.9669449081803005, + "grad_norm": 0.4466878652172313, + "learning_rate": 8.318614879485044e-09, + "loss": 0.2502, + "step": 5891 + }, + { + "epoch": 1.967278797996661, + "grad_norm": 0.441357439374316, + "learning_rate": 8.151456252144529e-09, + "loss": 0.253, + "step": 5892 + }, + { + "epoch": 1.9676126878130216, + "grad_norm": 0.43761625343048266, + "learning_rate": 7.98599283076018e-09, + "loss": 0.2397, + "step": 5893 + }, + { + "epoch": 1.9679465776293823, + "grad_norm": 0.44558078242709215, + "learning_rate": 7.82222467152205e-09, + "loss": 0.2366, + "step": 5894 + }, + { + "epoch": 1.968280467445743, + "grad_norm": 0.4640203199629371, + "learning_rate": 7.660151830045093e-09, + "loss": 0.2535, + "step": 5895 + }, + { + "epoch": 1.9686143572621035, + "grad_norm": 0.44879386056793397, + "learning_rate": 7.499774361368062e-09, + "loss": 0.2437, + "step": 5896 + }, + { + "epoch": 1.968948247078464, + "grad_norm": 0.4572093248546252, + "learning_rate": 7.341092319954612e-09, + "loss": 0.2385, + "step": 5897 + }, + { + "epoch": 1.9692821368948246, + "grad_norm": 0.42871377490100737, + "learning_rate": 7.184105759692195e-09, + "loss": 0.2377, + "step": 5898 + }, + { + "epoch": 1.9696160267111853, + "grad_norm": 0.44886808690295643, + "learning_rate": 7.028814733893163e-09, + "loss": 0.2417, + "step": 5899 + }, + { + "epoch": 1.969949916527546, + "grad_norm": 0.45472179562445525, + "learning_rate": 6.8752192952931115e-09, + "loss": 0.2458, + "step": 5900 + }, + { + "epoch": 1.9702838063439065, + "grad_norm": 0.4540952046585316, + "learning_rate": 6.723319496051983e-09, + "loss": 0.2527, + "step": 5901 + }, + { + "epoch": 1.970617696160267, + "grad_norm": 0.46278958720900415, + "learning_rate": 6.573115387754625e-09, + "loss": 0.2531, + "step": 5902 + }, + { + "epoch": 1.9709515859766276, + "grad_norm": 0.43651223982135273, + "learning_rate": 6.424607021409679e-09, + "loss": 0.2341, + "step": 5903 + }, + { + "epoch": 1.9712854757929883, + "grad_norm": 0.45072943248921915, + "learning_rate": 6.277794447449026e-09, + "loss": 0.2472, + "step": 5904 + }, + { + "epoch": 1.971619365609349, + "grad_norm": 0.4552382782498121, + "learning_rate": 6.132677715730562e-09, + "loss": 0.2498, + "step": 5905 + }, + { + "epoch": 1.9719532554257095, + "grad_norm": 0.4551644247386207, + "learning_rate": 5.98925687553431e-09, + "loss": 0.2465, + "step": 5906 + }, + { + "epoch": 1.97228714524207, + "grad_norm": 0.449093361331099, + "learning_rate": 5.8475319755657566e-09, + "loss": 0.2501, + "step": 5907 + }, + { + "epoch": 1.9726210350584306, + "grad_norm": 0.4337324303798181, + "learning_rate": 5.707503063953623e-09, + "loss": 0.2462, + "step": 5908 + }, + { + "epoch": 1.9729549248747913, + "grad_norm": 0.4893717370752528, + "learning_rate": 5.569170188250983e-09, + "loss": 0.2626, + "step": 5909 + }, + { + "epoch": 1.973288814691152, + "grad_norm": 0.45591325865039206, + "learning_rate": 5.4325333954347024e-09, + "loss": 0.2398, + "step": 5910 + }, + { + "epoch": 1.9736227045075125, + "grad_norm": 0.4556742924269473, + "learning_rate": 5.297592731907109e-09, + "loss": 0.2386, + "step": 5911 + }, + { + "epoch": 1.973956594323873, + "grad_norm": 0.43343759353080474, + "learning_rate": 5.164348243492101e-09, + "loss": 0.2492, + "step": 5912 + }, + { + "epoch": 1.9742904841402336, + "grad_norm": 0.44865420260852723, + "learning_rate": 5.032799975439595e-09, + "loss": 0.2548, + "step": 5913 + }, + { + "epoch": 1.9746243739565943, + "grad_norm": 0.45308511218925607, + "learning_rate": 4.90294797242219e-09, + "loss": 0.2476, + "step": 5914 + }, + { + "epoch": 1.974958263772955, + "grad_norm": 0.4604020317798855, + "learning_rate": 4.7747922785373875e-09, + "loss": 0.251, + "step": 5915 + }, + { + "epoch": 1.9752921535893155, + "grad_norm": 0.4376794657616519, + "learning_rate": 4.648332937305377e-09, + "loss": 0.2303, + "step": 5916 + }, + { + "epoch": 1.975626043405676, + "grad_norm": 0.4749567446931136, + "learning_rate": 4.523569991672361e-09, + "loss": 0.2571, + "step": 5917 + }, + { + "epoch": 1.9759599332220368, + "grad_norm": 0.45940604073642355, + "learning_rate": 4.4005034840061135e-09, + "loss": 0.247, + "step": 5918 + }, + { + "epoch": 1.9762938230383973, + "grad_norm": 0.4640167433193488, + "learning_rate": 4.279133456099871e-09, + "loss": 0.2492, + "step": 5919 + }, + { + "epoch": 1.976627712854758, + "grad_norm": 0.45252682533589916, + "learning_rate": 4.159459949170108e-09, + "loss": 0.2478, + "step": 5920 + }, + { + "epoch": 1.9769616026711185, + "grad_norm": 0.4450832579091648, + "learning_rate": 4.041483003857649e-09, + "loss": 0.2386, + "step": 5921 + }, + { + "epoch": 1.977295492487479, + "grad_norm": 0.442413342227629, + "learning_rate": 3.925202660226557e-09, + "loss": 0.2386, + "step": 5922 + }, + { + "epoch": 1.9776293823038398, + "grad_norm": 0.4417926744852581, + "learning_rate": 3.810618957765244e-09, + "loss": 0.2513, + "step": 5923 + }, + { + "epoch": 1.9779632721202003, + "grad_norm": 0.45098980027576313, + "learning_rate": 3.6977319353853626e-09, + "loss": 0.2468, + "step": 5924 + }, + { + "epoch": 1.978297161936561, + "grad_norm": 0.44818612625615045, + "learning_rate": 3.5865416314229127e-09, + "loss": 0.2337, + "step": 5925 + }, + { + "epoch": 1.9786310517529215, + "grad_norm": 0.4278181131979253, + "learning_rate": 3.4770480836376908e-09, + "loss": 0.2347, + "step": 5926 + }, + { + "epoch": 1.978964941569282, + "grad_norm": 0.45849198520420964, + "learning_rate": 3.3692513292132855e-09, + "loss": 0.255, + "step": 5927 + }, + { + "epoch": 1.9792988313856428, + "grad_norm": 0.4635950399865771, + "learning_rate": 3.2631514047565262e-09, + "loss": 0.255, + "step": 5928 + }, + { + "epoch": 1.9796327212020033, + "grad_norm": 0.436196266093366, + "learning_rate": 3.15874834629859e-09, + "loss": 0.2394, + "step": 5929 + }, + { + "epoch": 1.979966611018364, + "grad_norm": 0.4441414247814038, + "learning_rate": 3.0560421892938952e-09, + "loss": 0.2455, + "step": 5930 + }, + { + "epoch": 1.9803005008347245, + "grad_norm": 0.43048950461827196, + "learning_rate": 2.9550329686217625e-09, + "loss": 0.2396, + "step": 5931 + }, + { + "epoch": 1.980634390651085, + "grad_norm": 0.45041259810822093, + "learning_rate": 2.8557207185836435e-09, + "loss": 0.2348, + "step": 5932 + }, + { + "epoch": 1.9809682804674458, + "grad_norm": 0.4698013656945959, + "learning_rate": 2.758105472905337e-09, + "loss": 0.2549, + "step": 5933 + }, + { + "epoch": 1.9813021702838065, + "grad_norm": 0.4609906460537087, + "learning_rate": 2.662187264737548e-09, + "loss": 0.2504, + "step": 5934 + }, + { + "epoch": 1.981636060100167, + "grad_norm": 0.45929465248264023, + "learning_rate": 2.5679661266525544e-09, + "loss": 0.2477, + "step": 5935 + }, + { + "epoch": 1.9819699499165275, + "grad_norm": 0.44325909152858384, + "learning_rate": 2.4754420906475396e-09, + "loss": 0.2495, + "step": 5936 + }, + { + "epoch": 1.982303839732888, + "grad_norm": 0.46830959553411383, + "learning_rate": 2.38461518814348e-09, + "loss": 0.2362, + "step": 5937 + }, + { + "epoch": 1.9826377295492488, + "grad_norm": 0.4474728225314958, + "learning_rate": 2.2954854499845914e-09, + "loss": 0.2428, + "step": 5938 + }, + { + "epoch": 1.9829716193656095, + "grad_norm": 0.4376759362396262, + "learning_rate": 2.2080529064388845e-09, + "loss": 0.2509, + "step": 5939 + }, + { + "epoch": 1.98330550918197, + "grad_norm": 0.4547553736529194, + "learning_rate": 2.122317587197609e-09, + "loss": 0.2539, + "step": 5940 + }, + { + "epoch": 1.9836393989983305, + "grad_norm": 0.440463337991286, + "learning_rate": 2.038279521376918e-09, + "loss": 0.2497, + "step": 5941 + }, + { + "epoch": 1.983973288814691, + "grad_norm": 0.46177067285661455, + "learning_rate": 1.955938737515095e-09, + "loss": 0.2459, + "step": 5942 + }, + { + "epoch": 1.9843071786310518, + "grad_norm": 0.44983902595771874, + "learning_rate": 1.875295263574772e-09, + "loss": 0.2505, + "step": 5943 + }, + { + "epoch": 1.9846410684474125, + "grad_norm": 0.4689442696681314, + "learning_rate": 1.796349126941821e-09, + "loss": 0.248, + "step": 5944 + }, + { + "epoch": 1.984974958263773, + "grad_norm": 0.4627902299700257, + "learning_rate": 1.7191003544259067e-09, + "loss": 0.2556, + "step": 5945 + }, + { + "epoch": 1.9853088480801335, + "grad_norm": 0.44435527736053454, + "learning_rate": 1.6435489722610442e-09, + "loss": 0.2422, + "step": 5946 + }, + { + "epoch": 1.985642737896494, + "grad_norm": 0.4343997729478612, + "learning_rate": 1.5696950061033777e-09, + "loss": 0.2538, + "step": 5947 + }, + { + "epoch": 1.9859766277128548, + "grad_norm": 0.4582408515250657, + "learning_rate": 1.4975384810339555e-09, + "loss": 0.2473, + "step": 5948 + }, + { + "epoch": 1.9863105175292155, + "grad_norm": 0.45534717160025817, + "learning_rate": 1.4270794215559548e-09, + "loss": 0.2438, + "step": 5949 + }, + { + "epoch": 1.986644407345576, + "grad_norm": 0.4521687002459953, + "learning_rate": 1.358317851597457e-09, + "loss": 0.254, + "step": 5950 + }, + { + "epoch": 1.9869782971619365, + "grad_norm": 0.45041457947257346, + "learning_rate": 1.2912537945092285e-09, + "loss": 0.2484, + "step": 5951 + }, + { + "epoch": 1.987312186978297, + "grad_norm": 0.4402153638166802, + "learning_rate": 1.225887273066384e-09, + "loss": 0.2343, + "step": 5952 + }, + { + "epoch": 1.9876460767946578, + "grad_norm": 0.45867856249685385, + "learning_rate": 1.1622183094661677e-09, + "loss": 0.2479, + "step": 5953 + }, + { + "epoch": 1.9879799666110185, + "grad_norm": 0.45812805844283416, + "learning_rate": 1.100246925331283e-09, + "loss": 0.2486, + "step": 5954 + }, + { + "epoch": 1.988313856427379, + "grad_norm": 0.45582576076821124, + "learning_rate": 1.0399731417060077e-09, + "loss": 0.2475, + "step": 5955 + }, + { + "epoch": 1.9886477462437395, + "grad_norm": 0.4546117294834438, + "learning_rate": 9.813969790595235e-10, + "loss": 0.2419, + "step": 5956 + }, + { + "epoch": 1.9889816360601, + "grad_norm": 0.4474076028870332, + "learning_rate": 9.245184572842514e-10, + "loss": 0.2503, + "step": 5957 + }, + { + "epoch": 1.9893155258764608, + "grad_norm": 0.45845223330588464, + "learning_rate": 8.693375956947414e-10, + "loss": 0.2526, + "step": 5958 + }, + { + "epoch": 1.9896494156928215, + "grad_norm": 0.4828373887902602, + "learning_rate": 8.158544130315582e-10, + "loss": 0.2526, + "step": 5959 + }, + { + "epoch": 1.989983305509182, + "grad_norm": 0.45422537480124237, + "learning_rate": 7.640689274562851e-10, + "loss": 0.2438, + "step": 5960 + }, + { + "epoch": 1.9903171953255425, + "grad_norm": 0.4352065052854263, + "learning_rate": 7.139811565548549e-10, + "loss": 0.2488, + "step": 5961 + }, + { + "epoch": 1.990651085141903, + "grad_norm": 0.4413498249811745, + "learning_rate": 6.655911173381047e-10, + "loss": 0.2453, + "step": 5962 + }, + { + "epoch": 1.9909849749582638, + "grad_norm": 0.4561956634651036, + "learning_rate": 6.188988262373353e-10, + "loss": 0.2438, + "step": 5963 + }, + { + "epoch": 1.9913188647746245, + "grad_norm": 0.45248835305199714, + "learning_rate": 5.739042991104171e-10, + "loss": 0.2385, + "step": 5964 + }, + { + "epoch": 1.991652754590985, + "grad_norm": 0.44852597548223155, + "learning_rate": 5.306075512367947e-10, + "loss": 0.2448, + "step": 5965 + }, + { + "epoch": 1.9919866444073455, + "grad_norm": 0.45550315628080257, + "learning_rate": 4.890085973197067e-10, + "loss": 0.2402, + "step": 5966 + }, + { + "epoch": 1.992320534223706, + "grad_norm": 0.4527018486161834, + "learning_rate": 4.4910745148563083e-10, + "loss": 0.2478, + "step": 5967 + }, + { + "epoch": 1.9926544240400668, + "grad_norm": 0.4498848894389342, + "learning_rate": 4.1090412728539417e-10, + "loss": 0.2407, + "step": 5968 + }, + { + "epoch": 1.9929883138564275, + "grad_norm": 0.44515534169199694, + "learning_rate": 3.7439863769250793e-10, + "loss": 0.2456, + "step": 5969 + }, + { + "epoch": 1.993322203672788, + "grad_norm": 0.46326898137975203, + "learning_rate": 3.395909951037224e-10, + "loss": 0.2514, + "step": 5970 + }, + { + "epoch": 1.9936560934891485, + "grad_norm": 0.4481531798594148, + "learning_rate": 3.0648121133958206e-10, + "loss": 0.2396, + "step": 5971 + }, + { + "epoch": 1.993989983305509, + "grad_norm": 0.4490190103995155, + "learning_rate": 2.750692976444258e-10, + "loss": 0.2448, + "step": 5972 + }, + { + "epoch": 1.9943238731218698, + "grad_norm": 0.44750478843417185, + "learning_rate": 2.4535526468527636e-10, + "loss": 0.2561, + "step": 5973 + }, + { + "epoch": 1.9946577629382305, + "grad_norm": 0.4331324455979931, + "learning_rate": 2.1733912255295087e-10, + "loss": 0.2413, + "step": 5974 + }, + { + "epoch": 1.994991652754591, + "grad_norm": 0.4507621219708399, + "learning_rate": 1.910208807615055e-10, + "loss": 0.2551, + "step": 5975 + }, + { + "epoch": 1.9953255425709515, + "grad_norm": 0.4524621215694584, + "learning_rate": 1.664005482487907e-10, + "loss": 0.2506, + "step": 5976 + }, + { + "epoch": 1.995659432387312, + "grad_norm": 0.4488302089476692, + "learning_rate": 1.43478133375341e-10, + "loss": 0.2485, + "step": 5977 + }, + { + "epoch": 1.9959933222036728, + "grad_norm": 0.4525250408138406, + "learning_rate": 1.2225364392548512e-10, + "loss": 0.2386, + "step": 5978 + }, + { + "epoch": 1.9963272120200335, + "grad_norm": 0.4407413041885691, + "learning_rate": 1.0272708710679091e-10, + "loss": 0.2396, + "step": 5979 + }, + { + "epoch": 1.996661101836394, + "grad_norm": 0.4505497133964645, + "learning_rate": 8.489846955117565e-11, + "loss": 0.2443, + "step": 5980 + }, + { + "epoch": 1.9969949916527545, + "grad_norm": 0.4643315464617044, + "learning_rate": 6.876779731213035e-11, + "loss": 0.2503, + "step": 5981 + }, + { + "epoch": 1.9973288814691152, + "grad_norm": 0.44384435064039407, + "learning_rate": 5.4335075868605647e-11, + "loss": 0.2449, + "step": 5982 + }, + { + "epoch": 1.9976627712854758, + "grad_norm": 0.4437795413760093, + "learning_rate": 4.160031012057086e-11, + "loss": 0.246, + "step": 5983 + }, + { + "epoch": 1.9979966611018365, + "grad_norm": 0.44532330972238715, + "learning_rate": 3.056350439401001e-11, + "loss": 0.2505, + "step": 5984 + }, + { + "epoch": 1.998330550918197, + "grad_norm": 0.4580439817149641, + "learning_rate": 2.1224662436480913e-11, + "loss": 0.2478, + "step": 5985 + }, + { + "epoch": 1.9986644407345575, + "grad_norm": 0.4599163582795626, + "learning_rate": 1.3583787418780525e-11, + "loss": 0.2564, + "step": 5986 + }, + { + "epoch": 1.9989983305509182, + "grad_norm": 0.464035594547468, + "learning_rate": 7.640881936610278e-12, + "loss": 0.2457, + "step": 5987 + }, + { + "epoch": 1.9993322203672788, + "grad_norm": 0.45495124599611436, + "learning_rate": 3.3959480078005203e-12, + "loss": 0.2543, + "step": 5988 + }, + { + "epoch": 1.9996661101836395, + "grad_norm": 0.48593710633874654, + "learning_rate": 8.489870739758487e-13, + "loss": 0.2615, + "step": 5989 + }, + { + "epoch": 2.0, + "grad_norm": 0.42755929175093554, + "learning_rate": 0.0, + "loss": 0.2432, + "step": 5990 + }, + { + "epoch": 2.0, + "eval_loss": 0.27092963457107544, + "eval_runtime": 181.0355, + "eval_samples_per_second": 111.459, + "eval_steps_per_second": 0.873, + "step": 5990 + }, + { + "epoch": 2.0, + "step": 5990, + "total_flos": 2071187864551424.0, + "train_loss": 0.28312984947270664, + "train_runtime": 26606.5881, + "train_samples_per_second": 28.815, + "train_steps_per_second": 0.225 + } + ], + "logging_steps": 1, + "max_steps": 5990, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 2071187864551424.0, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}