diff --git "a/checkpoint-1830/trainer_state.json" "b/checkpoint-1830/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-1830/trainer_state.json" @@ -0,0 +1,13811 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.8011811023622046, + "eval_steps": 153, + "global_step": 1830, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.000984251968503937, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 6.0688, + "step": 1 + }, + { + "epoch": 0.001968503937007874, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 7.5576, + "step": 2 + }, + { + "epoch": 0.002952755905511811, + "grad_norm": Infinity, + "learning_rate": 0.0, + "loss": 4.6849, + "step": 3 + }, + { + "epoch": 0.003937007874015748, + "grad_norm": Infinity, + "learning_rate": 0.0, + "loss": 5.4503, + "step": 4 + }, + { + "epoch": 0.004921259842519685, + "grad_norm": 32.947513580322266, + "learning_rate": 9.940357852882705e-10, + "loss": 5.6057, + "step": 5 + }, + { + "epoch": 0.005905511811023622, + "grad_norm": 40.39633560180664, + "learning_rate": 1.988071570576541e-09, + "loss": 6.3049, + "step": 6 + }, + { + "epoch": 0.006889763779527559, + "grad_norm": 44.38965606689453, + "learning_rate": 2.9821073558648116e-09, + "loss": 6.8336, + "step": 7 + }, + { + "epoch": 0.007874015748031496, + "grad_norm": 28.46092987060547, + "learning_rate": 3.976143141153082e-09, + "loss": 5.0777, + "step": 8 + }, + { + "epoch": 0.008858267716535433, + "grad_norm": 25.817373275756836, + "learning_rate": 4.970178926441353e-09, + "loss": 4.8358, + "step": 9 + }, + { + "epoch": 0.00984251968503937, + "grad_norm": 27.06571388244629, + "learning_rate": 5.964214711729623e-09, + "loss": 4.641, + "step": 10 + }, + { + "epoch": 0.010826771653543307, + "grad_norm": 27.857297897338867, + "learning_rate": 6.9582504970178946e-09, + "loss": 4.828, + "step": 11 + }, + { + "epoch": 0.011811023622047244, + "grad_norm": 29.35377311706543, + "learning_rate": 7.952286282306164e-09, + "loss": 5.2269, + "step": 12 + }, + { + "epoch": 0.012795275590551181, + "grad_norm": 32.06682586669922, + "learning_rate": 8.946322067594435e-09, + "loss": 5.6772, + "step": 13 + }, + { + "epoch": 0.013779527559055118, + "grad_norm": 30.774206161499023, + "learning_rate": 9.940357852882705e-09, + "loss": 5.1422, + "step": 14 + }, + { + "epoch": 0.014763779527559055, + "grad_norm": 45.917659759521484, + "learning_rate": 1.0934393638170978e-08, + "loss": 6.2469, + "step": 15 + }, + { + "epoch": 0.015748031496062992, + "grad_norm": 30.189363479614258, + "learning_rate": 1.1928429423459246e-08, + "loss": 4.6802, + "step": 16 + }, + { + "epoch": 0.01673228346456693, + "grad_norm": 25.023223876953125, + "learning_rate": 1.2922465208747517e-08, + "loss": 4.5492, + "step": 17 + }, + { + "epoch": 0.017716535433070866, + "grad_norm": 26.336748123168945, + "learning_rate": 1.3916500994035789e-08, + "loss": 4.8062, + "step": 18 + }, + { + "epoch": 0.018700787401574805, + "grad_norm": 51.350032806396484, + "learning_rate": 1.4910536779324056e-08, + "loss": 7.5141, + "step": 19 + }, + { + "epoch": 0.01968503937007874, + "grad_norm": 33.953224182128906, + "learning_rate": 1.590457256461233e-08, + "loss": 5.5202, + "step": 20 + }, + { + "epoch": 0.02066929133858268, + "grad_norm": 36.828433990478516, + "learning_rate": 1.68986083499006e-08, + "loss": 6.5025, + "step": 21 + }, + { + "epoch": 0.021653543307086614, + "grad_norm": 43.68119812011719, + "learning_rate": 1.789264413518887e-08, + "loss": 7.318, + "step": 22 + }, + { + "epoch": 0.022637795275590553, + "grad_norm": 23.9952392578125, + "learning_rate": 1.888667992047714e-08, + "loss": 4.6458, + "step": 23 + }, + { + "epoch": 0.023622047244094488, + "grad_norm": 19.315523147583008, + "learning_rate": 1.988071570576541e-08, + "loss": 4.6191, + "step": 24 + }, + { + "epoch": 0.024606299212598427, + "grad_norm": 22.747207641601562, + "learning_rate": 2.087475149105368e-08, + "loss": 4.3159, + "step": 25 + }, + { + "epoch": 0.025590551181102362, + "grad_norm": 32.90699768066406, + "learning_rate": 2.1868787276341955e-08, + "loss": 6.3677, + "step": 26 + }, + { + "epoch": 0.0265748031496063, + "grad_norm": 25.224443435668945, + "learning_rate": 2.2862823061630224e-08, + "loss": 5.6052, + "step": 27 + }, + { + "epoch": 0.027559055118110236, + "grad_norm": 16.32752799987793, + "learning_rate": 2.3856858846918493e-08, + "loss": 4.196, + "step": 28 + }, + { + "epoch": 0.028543307086614175, + "grad_norm": 19.50617218017578, + "learning_rate": 2.4850894632206765e-08, + "loss": 4.4802, + "step": 29 + }, + { + "epoch": 0.02952755905511811, + "grad_norm": 21.613880157470703, + "learning_rate": 2.5844930417495034e-08, + "loss": 4.9193, + "step": 30 + }, + { + "epoch": 0.03051181102362205, + "grad_norm": 16.059247970581055, + "learning_rate": 2.6838966202783303e-08, + "loss": 4.0996, + "step": 31 + }, + { + "epoch": 0.031496062992125984, + "grad_norm": 27.45536994934082, + "learning_rate": 2.7833001988071578e-08, + "loss": 5.6307, + "step": 32 + }, + { + "epoch": 0.03248031496062992, + "grad_norm": 19.249103546142578, + "learning_rate": 2.8827037773359847e-08, + "loss": 4.5745, + "step": 33 + }, + { + "epoch": 0.03346456692913386, + "grad_norm": 17.516521453857422, + "learning_rate": 2.982107355864811e-08, + "loss": 4.4514, + "step": 34 + }, + { + "epoch": 0.0344488188976378, + "grad_norm": 11.684857368469238, + "learning_rate": 3.081510934393639e-08, + "loss": 4.0617, + "step": 35 + }, + { + "epoch": 0.03543307086614173, + "grad_norm": 16.813453674316406, + "learning_rate": 3.180914512922466e-08, + "loss": 5.0298, + "step": 36 + }, + { + "epoch": 0.03641732283464567, + "grad_norm": 10.92259407043457, + "learning_rate": 3.280318091451293e-08, + "loss": 3.9815, + "step": 37 + }, + { + "epoch": 0.03740157480314961, + "grad_norm": 11.449631690979004, + "learning_rate": 3.37972166998012e-08, + "loss": 4.0871, + "step": 38 + }, + { + "epoch": 0.038385826771653545, + "grad_norm": 11.223074913024902, + "learning_rate": 3.479125248508947e-08, + "loss": 4.2378, + "step": 39 + }, + { + "epoch": 0.03937007874015748, + "grad_norm": 9.864290237426758, + "learning_rate": 3.578528827037774e-08, + "loss": 3.8226, + "step": 40 + }, + { + "epoch": 0.040354330708661415, + "grad_norm": 11.084328651428223, + "learning_rate": 3.6779324055666005e-08, + "loss": 4.3519, + "step": 41 + }, + { + "epoch": 0.04133858267716536, + "grad_norm": 9.965168952941895, + "learning_rate": 3.777335984095428e-08, + "loss": 3.6345, + "step": 42 + }, + { + "epoch": 0.04232283464566929, + "grad_norm": 18.375577926635742, + "learning_rate": 3.8767395626242556e-08, + "loss": 5.0829, + "step": 43 + }, + { + "epoch": 0.04330708661417323, + "grad_norm": 11.232723236083984, + "learning_rate": 3.976143141153082e-08, + "loss": 4.6701, + "step": 44 + }, + { + "epoch": 0.04429133858267716, + "grad_norm": 10.896126747131348, + "learning_rate": 4.0755467196819094e-08, + "loss": 4.1371, + "step": 45 + }, + { + "epoch": 0.045275590551181105, + "grad_norm": 10.237566947937012, + "learning_rate": 4.174950298210736e-08, + "loss": 4.2418, + "step": 46 + }, + { + "epoch": 0.04625984251968504, + "grad_norm": 13.484391212463379, + "learning_rate": 4.274353876739563e-08, + "loss": 4.4766, + "step": 47 + }, + { + "epoch": 0.047244094488188976, + "grad_norm": 9.769365310668945, + "learning_rate": 4.373757455268391e-08, + "loss": 4.4797, + "step": 48 + }, + { + "epoch": 0.04822834645669291, + "grad_norm": 9.953036308288574, + "learning_rate": 4.4731610337972176e-08, + "loss": 3.8471, + "step": 49 + }, + { + "epoch": 0.04921259842519685, + "grad_norm": 15.378759384155273, + "learning_rate": 4.572564612326045e-08, + "loss": 4.3194, + "step": 50 + }, + { + "epoch": 0.05019685039370079, + "grad_norm": 9.474905967712402, + "learning_rate": 4.6719681908548713e-08, + "loss": 3.9426, + "step": 51 + }, + { + "epoch": 0.051181102362204724, + "grad_norm": 8.882713317871094, + "learning_rate": 4.7713717693836986e-08, + "loss": 3.5333, + "step": 52 + }, + { + "epoch": 0.05216535433070866, + "grad_norm": 14.956167221069336, + "learning_rate": 4.870775347912525e-08, + "loss": 4.2426, + "step": 53 + }, + { + "epoch": 0.0531496062992126, + "grad_norm": 9.333856582641602, + "learning_rate": 4.970178926441353e-08, + "loss": 3.9816, + "step": 54 + }, + { + "epoch": 0.054133858267716536, + "grad_norm": 9.034503936767578, + "learning_rate": 5.06958250497018e-08, + "loss": 3.663, + "step": 55 + }, + { + "epoch": 0.05511811023622047, + "grad_norm": 10.114209175109863, + "learning_rate": 5.168986083499007e-08, + "loss": 3.9057, + "step": 56 + }, + { + "epoch": 0.05610236220472441, + "grad_norm": 9.802759170532227, + "learning_rate": 5.268389662027834e-08, + "loss": 4.0345, + "step": 57 + }, + { + "epoch": 0.05708661417322835, + "grad_norm": 10.378804206848145, + "learning_rate": 5.3677932405566605e-08, + "loss": 3.5233, + "step": 58 + }, + { + "epoch": 0.058070866141732284, + "grad_norm": 10.237386703491211, + "learning_rate": 5.467196819085488e-08, + "loss": 3.7999, + "step": 59 + }, + { + "epoch": 0.05905511811023622, + "grad_norm": 11.103493690490723, + "learning_rate": 5.5666003976143156e-08, + "loss": 3.1885, + "step": 60 + }, + { + "epoch": 0.060039370078740155, + "grad_norm": 11.967391014099121, + "learning_rate": 5.666003976143142e-08, + "loss": 3.6013, + "step": 61 + }, + { + "epoch": 0.0610236220472441, + "grad_norm": 10.79468822479248, + "learning_rate": 5.7654075546719694e-08, + "loss": 3.392, + "step": 62 + }, + { + "epoch": 0.06200787401574803, + "grad_norm": 13.18323802947998, + "learning_rate": 5.864811133200796e-08, + "loss": 3.3814, + "step": 63 + }, + { + "epoch": 0.06299212598425197, + "grad_norm": 15.541842460632324, + "learning_rate": 5.964214711729623e-08, + "loss": 4.0428, + "step": 64 + }, + { + "epoch": 0.0639763779527559, + "grad_norm": 13.166780471801758, + "learning_rate": 6.06361829025845e-08, + "loss": 3.7825, + "step": 65 + }, + { + "epoch": 0.06496062992125984, + "grad_norm": 12.48865795135498, + "learning_rate": 6.163021868787278e-08, + "loss": 3.4181, + "step": 66 + }, + { + "epoch": 0.06594488188976377, + "grad_norm": 14.15877628326416, + "learning_rate": 6.262425447316104e-08, + "loss": 3.7793, + "step": 67 + }, + { + "epoch": 0.06692913385826772, + "grad_norm": 18.16546630859375, + "learning_rate": 6.361829025844931e-08, + "loss": 3.8344, + "step": 68 + }, + { + "epoch": 0.06791338582677166, + "grad_norm": 14.758286476135254, + "learning_rate": 6.461232604373759e-08, + "loss": 3.2165, + "step": 69 + }, + { + "epoch": 0.0688976377952756, + "grad_norm": 15.924538612365723, + "learning_rate": 6.560636182902586e-08, + "loss": 3.3811, + "step": 70 + }, + { + "epoch": 0.06988188976377953, + "grad_norm": 17.181415557861328, + "learning_rate": 6.660039761431412e-08, + "loss": 3.5984, + "step": 71 + }, + { + "epoch": 0.07086614173228346, + "grad_norm": 16.691272735595703, + "learning_rate": 6.75944333996024e-08, + "loss": 3.8583, + "step": 72 + }, + { + "epoch": 0.0718503937007874, + "grad_norm": 17.478809356689453, + "learning_rate": 6.858846918489067e-08, + "loss": 3.296, + "step": 73 + }, + { + "epoch": 0.07283464566929133, + "grad_norm": 15.149248123168945, + "learning_rate": 6.958250497017893e-08, + "loss": 2.7661, + "step": 74 + }, + { + "epoch": 0.07381889763779527, + "grad_norm": 15.139701843261719, + "learning_rate": 7.057654075546721e-08, + "loss": 2.9805, + "step": 75 + }, + { + "epoch": 0.07480314960629922, + "grad_norm": 14.794472694396973, + "learning_rate": 7.157057654075548e-08, + "loss": 2.566, + "step": 76 + }, + { + "epoch": 0.07578740157480315, + "grad_norm": 15.433120727539062, + "learning_rate": 7.256461232604374e-08, + "loss": 3.258, + "step": 77 + }, + { + "epoch": 0.07677165354330709, + "grad_norm": 22.558115005493164, + "learning_rate": 7.355864811133201e-08, + "loss": 3.3804, + "step": 78 + }, + { + "epoch": 0.07775590551181102, + "grad_norm": 19.630340576171875, + "learning_rate": 7.455268389662029e-08, + "loss": 2.8828, + "step": 79 + }, + { + "epoch": 0.07874015748031496, + "grad_norm": 17.1998233795166, + "learning_rate": 7.554671968190855e-08, + "loss": 3.1077, + "step": 80 + }, + { + "epoch": 0.0797244094488189, + "grad_norm": 13.535666465759277, + "learning_rate": 7.654075546719683e-08, + "loss": 2.9441, + "step": 81 + }, + { + "epoch": 0.08070866141732283, + "grad_norm": 14.031355857849121, + "learning_rate": 7.753479125248511e-08, + "loss": 2.9465, + "step": 82 + }, + { + "epoch": 0.08169291338582677, + "grad_norm": 13.701343536376953, + "learning_rate": 7.852882703777338e-08, + "loss": 2.7088, + "step": 83 + }, + { + "epoch": 0.08267716535433071, + "grad_norm": 15.102946281433105, + "learning_rate": 7.952286282306164e-08, + "loss": 2.9215, + "step": 84 + }, + { + "epoch": 0.08366141732283465, + "grad_norm": 20.15928077697754, + "learning_rate": 8.051689860834992e-08, + "loss": 3.4698, + "step": 85 + }, + { + "epoch": 0.08464566929133858, + "grad_norm": 14.807561874389648, + "learning_rate": 8.151093439363819e-08, + "loss": 2.2414, + "step": 86 + }, + { + "epoch": 0.08562992125984252, + "grad_norm": 18.390995025634766, + "learning_rate": 8.250497017892645e-08, + "loss": 3.1601, + "step": 87 + }, + { + "epoch": 0.08661417322834646, + "grad_norm": 17.47426986694336, + "learning_rate": 8.349900596421472e-08, + "loss": 2.7714, + "step": 88 + }, + { + "epoch": 0.08759842519685039, + "grad_norm": 19.336660385131836, + "learning_rate": 8.4493041749503e-08, + "loss": 3.0311, + "step": 89 + }, + { + "epoch": 0.08858267716535433, + "grad_norm": 22.308195114135742, + "learning_rate": 8.548707753479126e-08, + "loss": 3.0336, + "step": 90 + }, + { + "epoch": 0.08956692913385826, + "grad_norm": 16.85112762451172, + "learning_rate": 8.648111332007953e-08, + "loss": 1.9358, + "step": 91 + }, + { + "epoch": 0.09055118110236221, + "grad_norm": 18.357120513916016, + "learning_rate": 8.747514910536782e-08, + "loss": 2.6031, + "step": 92 + }, + { + "epoch": 0.09153543307086615, + "grad_norm": 21.082101821899414, + "learning_rate": 8.846918489065609e-08, + "loss": 2.7515, + "step": 93 + }, + { + "epoch": 0.09251968503937008, + "grad_norm": 21.116485595703125, + "learning_rate": 8.946322067594435e-08, + "loss": 2.8496, + "step": 94 + }, + { + "epoch": 0.09350393700787402, + "grad_norm": 15.967782020568848, + "learning_rate": 9.045725646123262e-08, + "loss": 1.8015, + "step": 95 + }, + { + "epoch": 0.09448818897637795, + "grad_norm": 20.430376052856445, + "learning_rate": 9.14512922465209e-08, + "loss": 2.8138, + "step": 96 + }, + { + "epoch": 0.09547244094488189, + "grad_norm": 18.769359588623047, + "learning_rate": 9.244532803180916e-08, + "loss": 2.0597, + "step": 97 + }, + { + "epoch": 0.09645669291338582, + "grad_norm": 19.360021591186523, + "learning_rate": 9.343936381709743e-08, + "loss": 2.1053, + "step": 98 + }, + { + "epoch": 0.09744094488188976, + "grad_norm": 22.48013687133789, + "learning_rate": 9.44333996023857e-08, + "loss": 2.6785, + "step": 99 + }, + { + "epoch": 0.0984251968503937, + "grad_norm": 21.50940704345703, + "learning_rate": 9.542743538767397e-08, + "loss": 2.588, + "step": 100 + }, + { + "epoch": 0.09940944881889764, + "grad_norm": 17.409393310546875, + "learning_rate": 9.642147117296224e-08, + "loss": 2.0099, + "step": 101 + }, + { + "epoch": 0.10039370078740158, + "grad_norm": 22.641996383666992, + "learning_rate": 9.74155069582505e-08, + "loss": 2.7947, + "step": 102 + }, + { + "epoch": 0.10137795275590551, + "grad_norm": 18.91701889038086, + "learning_rate": 9.840954274353878e-08, + "loss": 2.3274, + "step": 103 + }, + { + "epoch": 0.10236220472440945, + "grad_norm": 21.317251205444336, + "learning_rate": 9.940357852882706e-08, + "loss": 2.2545, + "step": 104 + }, + { + "epoch": 0.10334645669291338, + "grad_norm": 24.298248291015625, + "learning_rate": 1.0039761431411533e-07, + "loss": 2.4575, + "step": 105 + }, + { + "epoch": 0.10433070866141732, + "grad_norm": 19.284252166748047, + "learning_rate": 1.013916500994036e-07, + "loss": 2.4413, + "step": 106 + }, + { + "epoch": 0.10531496062992125, + "grad_norm": 19.619426727294922, + "learning_rate": 1.0238568588469187e-07, + "loss": 2.3185, + "step": 107 + }, + { + "epoch": 0.1062992125984252, + "grad_norm": 20.323917388916016, + "learning_rate": 1.0337972166998014e-07, + "loss": 2.1577, + "step": 108 + }, + { + "epoch": 0.10728346456692914, + "grad_norm": 20.885499954223633, + "learning_rate": 1.043737574552684e-07, + "loss": 2.1278, + "step": 109 + }, + { + "epoch": 0.10826771653543307, + "grad_norm": 21.687036514282227, + "learning_rate": 1.0536779324055668e-07, + "loss": 2.0967, + "step": 110 + }, + { + "epoch": 0.10925196850393701, + "grad_norm": 23.031898498535156, + "learning_rate": 1.0636182902584495e-07, + "loss": 2.6142, + "step": 111 + }, + { + "epoch": 0.11023622047244094, + "grad_norm": 20.64027976989746, + "learning_rate": 1.0735586481113321e-07, + "loss": 1.8553, + "step": 112 + }, + { + "epoch": 0.11122047244094488, + "grad_norm": 18.268877029418945, + "learning_rate": 1.0834990059642149e-07, + "loss": 2.1523, + "step": 113 + }, + { + "epoch": 0.11220472440944881, + "grad_norm": 22.190549850463867, + "learning_rate": 1.0934393638170976e-07, + "loss": 2.1726, + "step": 114 + }, + { + "epoch": 0.11318897637795275, + "grad_norm": 18.314851760864258, + "learning_rate": 1.1033797216699802e-07, + "loss": 1.8564, + "step": 115 + }, + { + "epoch": 0.1141732283464567, + "grad_norm": 19.17725944519043, + "learning_rate": 1.1133200795228631e-07, + "loss": 1.8413, + "step": 116 + }, + { + "epoch": 0.11515748031496063, + "grad_norm": 24.518449783325195, + "learning_rate": 1.1232604373757458e-07, + "loss": 2.0441, + "step": 117 + }, + { + "epoch": 0.11614173228346457, + "grad_norm": 24.78040313720703, + "learning_rate": 1.1332007952286284e-07, + "loss": 2.2159, + "step": 118 + }, + { + "epoch": 0.1171259842519685, + "grad_norm": 28.552289962768555, + "learning_rate": 1.1431411530815111e-07, + "loss": 2.6779, + "step": 119 + }, + { + "epoch": 0.11811023622047244, + "grad_norm": 21.049219131469727, + "learning_rate": 1.1530815109343939e-07, + "loss": 2.2976, + "step": 120 + }, + { + "epoch": 0.11909448818897637, + "grad_norm": 23.962575912475586, + "learning_rate": 1.1630218687872765e-07, + "loss": 1.9407, + "step": 121 + }, + { + "epoch": 0.12007874015748031, + "grad_norm": 20.15193748474121, + "learning_rate": 1.1729622266401592e-07, + "loss": 1.9019, + "step": 122 + }, + { + "epoch": 0.12106299212598425, + "grad_norm": 22.563459396362305, + "learning_rate": 1.182902584493042e-07, + "loss": 2.2149, + "step": 123 + }, + { + "epoch": 0.1220472440944882, + "grad_norm": 16.86638832092285, + "learning_rate": 1.1928429423459245e-07, + "loss": 1.6823, + "step": 124 + }, + { + "epoch": 0.12303149606299213, + "grad_norm": 20.249584197998047, + "learning_rate": 1.2027833001988073e-07, + "loss": 1.8402, + "step": 125 + }, + { + "epoch": 0.12401574803149606, + "grad_norm": 19.5487117767334, + "learning_rate": 1.21272365805169e-07, + "loss": 1.6914, + "step": 126 + }, + { + "epoch": 0.125, + "grad_norm": 19.093505859375, + "learning_rate": 1.222664015904573e-07, + "loss": 2.1626, + "step": 127 + }, + { + "epoch": 0.12598425196850394, + "grad_norm": 17.598743438720703, + "learning_rate": 1.2326043737574557e-07, + "loss": 1.6414, + "step": 128 + }, + { + "epoch": 0.12696850393700787, + "grad_norm": 22.712181091308594, + "learning_rate": 1.2425447316103382e-07, + "loss": 2.2043, + "step": 129 + }, + { + "epoch": 0.1279527559055118, + "grad_norm": 20.516571044921875, + "learning_rate": 1.2524850894632207e-07, + "loss": 1.9987, + "step": 130 + }, + { + "epoch": 0.12893700787401574, + "grad_norm": 22.58023452758789, + "learning_rate": 1.2624254473161035e-07, + "loss": 1.8868, + "step": 131 + }, + { + "epoch": 0.12992125984251968, + "grad_norm": 17.800390243530273, + "learning_rate": 1.2723658051689863e-07, + "loss": 1.8262, + "step": 132 + }, + { + "epoch": 0.1309055118110236, + "grad_norm": 20.7684383392334, + "learning_rate": 1.282306163021869e-07, + "loss": 2.0404, + "step": 133 + }, + { + "epoch": 0.13188976377952755, + "grad_norm": 20.88973617553711, + "learning_rate": 1.2922465208747519e-07, + "loss": 1.9134, + "step": 134 + }, + { + "epoch": 0.1328740157480315, + "grad_norm": 21.78728485107422, + "learning_rate": 1.3021868787276344e-07, + "loss": 2.3725, + "step": 135 + }, + { + "epoch": 0.13385826771653545, + "grad_norm": 16.885374069213867, + "learning_rate": 1.3121272365805172e-07, + "loss": 1.4127, + "step": 136 + }, + { + "epoch": 0.13484251968503938, + "grad_norm": 18.944265365600586, + "learning_rate": 1.3220675944333997e-07, + "loss": 1.6876, + "step": 137 + }, + { + "epoch": 0.13582677165354332, + "grad_norm": 18.579692840576172, + "learning_rate": 1.3320079522862825e-07, + "loss": 1.8376, + "step": 138 + }, + { + "epoch": 0.13681102362204725, + "grad_norm": 19.68845558166504, + "learning_rate": 1.3419483101391653e-07, + "loss": 1.6992, + "step": 139 + }, + { + "epoch": 0.1377952755905512, + "grad_norm": 21.27842903137207, + "learning_rate": 1.351888667992048e-07, + "loss": 1.5032, + "step": 140 + }, + { + "epoch": 0.13877952755905512, + "grad_norm": 28.506824493408203, + "learning_rate": 1.3618290258449306e-07, + "loss": 2.0334, + "step": 141 + }, + { + "epoch": 0.13976377952755906, + "grad_norm": 24.97527313232422, + "learning_rate": 1.3717693836978134e-07, + "loss": 2.3581, + "step": 142 + }, + { + "epoch": 0.140748031496063, + "grad_norm": 22.391145706176758, + "learning_rate": 1.381709741550696e-07, + "loss": 1.4236, + "step": 143 + }, + { + "epoch": 0.14173228346456693, + "grad_norm": 22.977148056030273, + "learning_rate": 1.3916500994035787e-07, + "loss": 2.202, + "step": 144 + }, + { + "epoch": 0.14271653543307086, + "grad_norm": 25.096189498901367, + "learning_rate": 1.4015904572564615e-07, + "loss": 1.7654, + "step": 145 + }, + { + "epoch": 0.1437007874015748, + "grad_norm": 28.528640747070312, + "learning_rate": 1.4115308151093443e-07, + "loss": 1.5748, + "step": 146 + }, + { + "epoch": 0.14468503937007873, + "grad_norm": 30.80350685119629, + "learning_rate": 1.421471172962227e-07, + "loss": 1.7996, + "step": 147 + }, + { + "epoch": 0.14566929133858267, + "grad_norm": 22.726850509643555, + "learning_rate": 1.4314115308151096e-07, + "loss": 1.7517, + "step": 148 + }, + { + "epoch": 0.1466535433070866, + "grad_norm": 24.926170349121094, + "learning_rate": 1.4413518886679924e-07, + "loss": 1.8933, + "step": 149 + }, + { + "epoch": 0.14763779527559054, + "grad_norm": 20.570375442504883, + "learning_rate": 1.451292246520875e-07, + "loss": 1.2836, + "step": 150 + }, + { + "epoch": 0.1486220472440945, + "grad_norm": 21.08650779724121, + "learning_rate": 1.4612326043737577e-07, + "loss": 1.7145, + "step": 151 + }, + { + "epoch": 0.14960629921259844, + "grad_norm": 24.37209701538086, + "learning_rate": 1.4711729622266402e-07, + "loss": 1.6499, + "step": 152 + }, + { + "epoch": 0.15059055118110237, + "grad_norm": 26.801511764526367, + "learning_rate": 1.4811133200795232e-07, + "loss": 1.8273, + "step": 153 + }, + { + "epoch": 0.15059055118110237, + "eval_Qnli-dev_cosine_accuracy": 0.646484375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8057259321212769, + "eval_Qnli-dev_cosine_ap": 0.6720663622193426, + "eval_Qnli-dev_cosine_f1": 0.6688102893890675, + "eval_Qnli-dev_cosine_f1_threshold": 0.7187118530273438, + "eval_Qnli-dev_cosine_precision": 0.538860103626943, + "eval_Qnli-dev_cosine_recall": 0.8813559322033898, + "eval_Qnli-dev_dot_accuracy": 0.646484375, + "eval_Qnli-dev_dot_accuracy_threshold": 618.8643798828125, + "eval_Qnli-dev_dot_ap": 0.672083506527328, + "eval_Qnli-dev_dot_f1": 0.6688102893890675, + "eval_Qnli-dev_dot_f1_threshold": 552.0260009765625, + "eval_Qnli-dev_dot_precision": 0.538860103626943, + "eval_Qnli-dev_dot_recall": 0.8813559322033898, + "eval_Qnli-dev_euclidean_accuracy": 0.646484375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.27533721923828, + "eval_Qnli-dev_euclidean_ap": 0.6720591998758361, + "eval_Qnli-dev_euclidean_f1": 0.6688102893890675, + "eval_Qnli-dev_euclidean_f1_threshold": 20.787063598632812, + "eval_Qnli-dev_euclidean_precision": 0.538860103626943, + "eval_Qnli-dev_euclidean_recall": 0.8813559322033898, + "eval_Qnli-dev_manhattan_accuracy": 0.6484375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 386.58905029296875, + "eval_Qnli-dev_manhattan_ap": 0.6724653688821339, + "eval_Qnli-dev_manhattan_f1": 0.6645569620253164, + "eval_Qnli-dev_manhattan_f1_threshold": 462.609130859375, + "eval_Qnli-dev_manhattan_precision": 0.5303030303030303, + "eval_Qnli-dev_manhattan_recall": 0.8898305084745762, + "eval_Qnli-dev_max_accuracy": 0.6484375, + "eval_Qnli-dev_max_accuracy_threshold": 618.8643798828125, + "eval_Qnli-dev_max_ap": 0.6724653688821339, + "eval_Qnli-dev_max_f1": 0.6688102893890675, + "eval_Qnli-dev_max_f1_threshold": 552.0260009765625, + "eval_Qnli-dev_max_precision": 0.538860103626943, + "eval_Qnli-dev_max_recall": 0.8898305084745762, + "eval_allNLI-dev_cosine_accuracy": 0.67578125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9427558183670044, + "eval_allNLI-dev_cosine_ap": 0.4368751759846574, + "eval_allNLI-dev_cosine_f1": 0.5225225225225225, + "eval_allNLI-dev_cosine_f1_threshold": 0.8046966791152954, + "eval_allNLI-dev_cosine_precision": 0.3795811518324607, + "eval_allNLI-dev_cosine_recall": 0.838150289017341, + "eval_allNLI-dev_dot_accuracy": 0.67578125, + "eval_allNLI-dev_dot_accuracy_threshold": 724.1080322265625, + "eval_allNLI-dev_dot_ap": 0.436842886797982, + "eval_allNLI-dev_dot_f1": 0.5225225225225225, + "eval_allNLI-dev_dot_f1_threshold": 618.074951171875, + "eval_allNLI-dev_dot_precision": 0.3795811518324607, + "eval_allNLI-dev_dot_recall": 0.838150289017341, + "eval_allNLI-dev_euclidean_accuracy": 0.67578125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.377331733703613, + "eval_allNLI-dev_euclidean_ap": 0.4368602200677977, + "eval_allNLI-dev_euclidean_f1": 0.5225225225225225, + "eval_allNLI-dev_euclidean_f1_threshold": 17.321048736572266, + "eval_allNLI-dev_euclidean_precision": 0.3795811518324607, + "eval_allNLI-dev_euclidean_recall": 0.838150289017341, + "eval_allNLI-dev_manhattan_accuracy": 0.677734375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 223.6764373779297, + "eval_allNLI-dev_manhattan_ap": 0.43892484929307635, + "eval_allNLI-dev_manhattan_f1": 0.5239852398523985, + "eval_allNLI-dev_manhattan_f1_threshold": 372.31396484375, + "eval_allNLI-dev_manhattan_precision": 0.38482384823848237, + "eval_allNLI-dev_manhattan_recall": 0.8208092485549133, + "eval_allNLI-dev_max_accuracy": 0.677734375, + "eval_allNLI-dev_max_accuracy_threshold": 724.1080322265625, + "eval_allNLI-dev_max_ap": 0.43892484929307635, + "eval_allNLI-dev_max_f1": 0.5239852398523985, + "eval_allNLI-dev_max_f1_threshold": 618.074951171875, + "eval_allNLI-dev_max_precision": 0.38482384823848237, + "eval_allNLI-dev_max_recall": 0.838150289017341, + "eval_runtime": 1.2344, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.6724653688821339, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.3774946012125992, + "eval_sts-test_pearson_dot": 0.3774648453085433, + "eval_sts-test_pearson_euclidean": 0.38652243004790016, + "eval_sts-test_pearson_manhattan": 0.3861982631744407, + "eval_sts-test_pearson_max": 0.38652243004790016, + "eval_sts-test_spearman_cosine": 0.4056589966976888, + "eval_sts-test_spearman_dot": 0.40563469676275316, + "eval_sts-test_spearman_euclidean": 0.4056589966976888, + "eval_sts-test_spearman_manhattan": 0.4059364545183154, + "eval_sts-test_spearman_max": 0.4059364545183154, + "step": 153 + }, + { + "epoch": 0.1515748031496063, + "grad_norm": 27.144521713256836, + "learning_rate": 1.4910536779324058e-07, + "loss": 2.2859, + "step": 154 + }, + { + "epoch": 0.15255905511811024, + "grad_norm": 16.310321807861328, + "learning_rate": 1.5009940357852886e-07, + "loss": 1.0833, + "step": 155 + }, + { + "epoch": 0.15354330708661418, + "grad_norm": 18.788604736328125, + "learning_rate": 1.510934393638171e-07, + "loss": 1.6829, + "step": 156 + }, + { + "epoch": 0.1545275590551181, + "grad_norm": 23.34528923034668, + "learning_rate": 1.5208747514910539e-07, + "loss": 2.1464, + "step": 157 + }, + { + "epoch": 0.15551181102362205, + "grad_norm": 19.06661033630371, + "learning_rate": 1.5308151093439367e-07, + "loss": 1.745, + "step": 158 + }, + { + "epoch": 0.15649606299212598, + "grad_norm": 23.650297164916992, + "learning_rate": 1.5407554671968192e-07, + "loss": 1.7319, + "step": 159 + }, + { + "epoch": 0.15748031496062992, + "grad_norm": 22.01628303527832, + "learning_rate": 1.5506958250497022e-07, + "loss": 1.6968, + "step": 160 + }, + { + "epoch": 0.15846456692913385, + "grad_norm": 20.905855178833008, + "learning_rate": 1.5606361829025848e-07, + "loss": 1.7401, + "step": 161 + }, + { + "epoch": 0.1594488188976378, + "grad_norm": 22.40980339050293, + "learning_rate": 1.5705765407554675e-07, + "loss": 1.729, + "step": 162 + }, + { + "epoch": 0.16043307086614172, + "grad_norm": 23.574447631835938, + "learning_rate": 1.58051689860835e-07, + "loss": 2.0782, + "step": 163 + }, + { + "epoch": 0.16141732283464566, + "grad_norm": 28.43741798400879, + "learning_rate": 1.5904572564612329e-07, + "loss": 2.6545, + "step": 164 + }, + { + "epoch": 0.1624015748031496, + "grad_norm": 18.536827087402344, + "learning_rate": 1.6003976143141154e-07, + "loss": 1.4045, + "step": 165 + }, + { + "epoch": 0.16338582677165353, + "grad_norm": 17.013965606689453, + "learning_rate": 1.6103379721669984e-07, + "loss": 1.2937, + "step": 166 + }, + { + "epoch": 0.1643700787401575, + "grad_norm": 16.021455764770508, + "learning_rate": 1.620278330019881e-07, + "loss": 1.1171, + "step": 167 + }, + { + "epoch": 0.16535433070866143, + "grad_norm": 20.248655319213867, + "learning_rate": 1.6302186878727637e-07, + "loss": 1.3537, + "step": 168 + }, + { + "epoch": 0.16633858267716536, + "grad_norm": 21.731477737426758, + "learning_rate": 1.6401590457256465e-07, + "loss": 1.7028, + "step": 169 + }, + { + "epoch": 0.1673228346456693, + "grad_norm": 20.372398376464844, + "learning_rate": 1.650099403578529e-07, + "loss": 1.4143, + "step": 170 + }, + { + "epoch": 0.16830708661417323, + "grad_norm": 21.89658546447754, + "learning_rate": 1.6600397614314118e-07, + "loss": 1.8648, + "step": 171 + }, + { + "epoch": 0.16929133858267717, + "grad_norm": 24.402984619140625, + "learning_rate": 1.6699801192842944e-07, + "loss": 1.6768, + "step": 172 + }, + { + "epoch": 0.1702755905511811, + "grad_norm": 28.650039672851562, + "learning_rate": 1.6799204771371774e-07, + "loss": 1.9528, + "step": 173 + }, + { + "epoch": 0.17125984251968504, + "grad_norm": 18.586729049682617, + "learning_rate": 1.68986083499006e-07, + "loss": 1.1718, + "step": 174 + }, + { + "epoch": 0.17224409448818898, + "grad_norm": 22.66750717163086, + "learning_rate": 1.6998011928429427e-07, + "loss": 1.8176, + "step": 175 + }, + { + "epoch": 0.1732283464566929, + "grad_norm": 15.157270431518555, + "learning_rate": 1.7097415506958253e-07, + "loss": 0.8439, + "step": 176 + }, + { + "epoch": 0.17421259842519685, + "grad_norm": 23.08325958251953, + "learning_rate": 1.719681908548708e-07, + "loss": 1.5092, + "step": 177 + }, + { + "epoch": 0.17519685039370078, + "grad_norm": 19.215696334838867, + "learning_rate": 1.7296222664015906e-07, + "loss": 1.1947, + "step": 178 + }, + { + "epoch": 0.17618110236220472, + "grad_norm": 23.268930435180664, + "learning_rate": 1.7395626242544734e-07, + "loss": 1.6395, + "step": 179 + }, + { + "epoch": 0.17716535433070865, + "grad_norm": 24.91169548034668, + "learning_rate": 1.7495029821073564e-07, + "loss": 1.4394, + "step": 180 + }, + { + "epoch": 0.1781496062992126, + "grad_norm": 26.92424201965332, + "learning_rate": 1.759443339960239e-07, + "loss": 1.7548, + "step": 181 + }, + { + "epoch": 0.17913385826771652, + "grad_norm": 22.831315994262695, + "learning_rate": 1.7693836978131217e-07, + "loss": 1.1181, + "step": 182 + }, + { + "epoch": 0.18011811023622049, + "grad_norm": 18.39922332763672, + "learning_rate": 1.7793240556660042e-07, + "loss": 1.0271, + "step": 183 + }, + { + "epoch": 0.18110236220472442, + "grad_norm": 30.0064697265625, + "learning_rate": 1.789264413518887e-07, + "loss": 2.3108, + "step": 184 + }, + { + "epoch": 0.18208661417322836, + "grad_norm": 25.531293869018555, + "learning_rate": 1.7992047713717695e-07, + "loss": 2.1242, + "step": 185 + }, + { + "epoch": 0.1830708661417323, + "grad_norm": 25.3813533782959, + "learning_rate": 1.8091451292246523e-07, + "loss": 1.9822, + "step": 186 + }, + { + "epoch": 0.18405511811023623, + "grad_norm": 22.24681282043457, + "learning_rate": 1.819085487077535e-07, + "loss": 2.3605, + "step": 187 + }, + { + "epoch": 0.18503937007874016, + "grad_norm": 19.291316986083984, + "learning_rate": 1.829025844930418e-07, + "loss": 1.5251, + "step": 188 + }, + { + "epoch": 0.1860236220472441, + "grad_norm": 18.280820846557617, + "learning_rate": 1.8389662027833004e-07, + "loss": 1.2351, + "step": 189 + }, + { + "epoch": 0.18700787401574803, + "grad_norm": 18.341115951538086, + "learning_rate": 1.8489065606361832e-07, + "loss": 1.5859, + "step": 190 + }, + { + "epoch": 0.18799212598425197, + "grad_norm": 21.981475830078125, + "learning_rate": 1.8588469184890657e-07, + "loss": 1.8056, + "step": 191 + }, + { + "epoch": 0.1889763779527559, + "grad_norm": 23.482725143432617, + "learning_rate": 1.8687872763419485e-07, + "loss": 1.349, + "step": 192 + }, + { + "epoch": 0.18996062992125984, + "grad_norm": 15.451715469360352, + "learning_rate": 1.8787276341948313e-07, + "loss": 0.893, + "step": 193 + }, + { + "epoch": 0.19094488188976377, + "grad_norm": 25.24738121032715, + "learning_rate": 1.888667992047714e-07, + "loss": 1.5122, + "step": 194 + }, + { + "epoch": 0.1919291338582677, + "grad_norm": 22.348981857299805, + "learning_rate": 1.898608349900597e-07, + "loss": 1.3875, + "step": 195 + }, + { + "epoch": 0.19291338582677164, + "grad_norm": 22.770132064819336, + "learning_rate": 1.9085487077534794e-07, + "loss": 1.29, + "step": 196 + }, + { + "epoch": 0.19389763779527558, + "grad_norm": 31.219194412231445, + "learning_rate": 1.9184890656063622e-07, + "loss": 2.2931, + "step": 197 + }, + { + "epoch": 0.19488188976377951, + "grad_norm": 24.657859802246094, + "learning_rate": 1.9284294234592447e-07, + "loss": 1.2663, + "step": 198 + }, + { + "epoch": 0.19586614173228348, + "grad_norm": 26.825185775756836, + "learning_rate": 1.9383697813121275e-07, + "loss": 1.9712, + "step": 199 + }, + { + "epoch": 0.1968503937007874, + "grad_norm": 30.97873306274414, + "learning_rate": 1.94831013916501e-07, + "loss": 2.3307, + "step": 200 + }, + { + "epoch": 0.19783464566929135, + "grad_norm": 24.922365188598633, + "learning_rate": 1.958250497017893e-07, + "loss": 1.6544, + "step": 201 + }, + { + "epoch": 0.19881889763779528, + "grad_norm": 22.842063903808594, + "learning_rate": 1.9681908548707756e-07, + "loss": 1.638, + "step": 202 + }, + { + "epoch": 0.19980314960629922, + "grad_norm": 17.376962661743164, + "learning_rate": 1.9781312127236584e-07, + "loss": 1.3412, + "step": 203 + }, + { + "epoch": 0.20078740157480315, + "grad_norm": 17.828535079956055, + "learning_rate": 1.9880715705765412e-07, + "loss": 1.4454, + "step": 204 + }, + { + "epoch": 0.2017716535433071, + "grad_norm": 18.73072624206543, + "learning_rate": 1.9980119284294237e-07, + "loss": 1.5437, + "step": 205 + }, + { + "epoch": 0.20275590551181102, + "grad_norm": 22.05230140686035, + "learning_rate": 2.0079522862823065e-07, + "loss": 1.4921, + "step": 206 + }, + { + "epoch": 0.20374015748031496, + "grad_norm": 18.090892791748047, + "learning_rate": 2.017892644135189e-07, + "loss": 1.4298, + "step": 207 + }, + { + "epoch": 0.2047244094488189, + "grad_norm": 22.94427490234375, + "learning_rate": 2.027833001988072e-07, + "loss": 1.6174, + "step": 208 + }, + { + "epoch": 0.20570866141732283, + "grad_norm": 20.903331756591797, + "learning_rate": 2.0377733598409546e-07, + "loss": 1.4137, + "step": 209 + }, + { + "epoch": 0.20669291338582677, + "grad_norm": 22.67809295654297, + "learning_rate": 2.0477137176938374e-07, + "loss": 1.5652, + "step": 210 + }, + { + "epoch": 0.2076771653543307, + "grad_norm": 16.824064254760742, + "learning_rate": 2.05765407554672e-07, + "loss": 1.1631, + "step": 211 + }, + { + "epoch": 0.20866141732283464, + "grad_norm": 21.46413230895996, + "learning_rate": 2.0675944333996027e-07, + "loss": 1.2351, + "step": 212 + }, + { + "epoch": 0.20964566929133857, + "grad_norm": 22.461610794067383, + "learning_rate": 2.0775347912524852e-07, + "loss": 1.7537, + "step": 213 + }, + { + "epoch": 0.2106299212598425, + "grad_norm": 22.150938034057617, + "learning_rate": 2.087475149105368e-07, + "loss": 1.3186, + "step": 214 + }, + { + "epoch": 0.21161417322834647, + "grad_norm": 22.82622718811035, + "learning_rate": 2.097415506958251e-07, + "loss": 1.2258, + "step": 215 + }, + { + "epoch": 0.2125984251968504, + "grad_norm": 13.514169692993164, + "learning_rate": 2.1073558648111336e-07, + "loss": 0.7695, + "step": 216 + }, + { + "epoch": 0.21358267716535434, + "grad_norm": 22.254596710205078, + "learning_rate": 2.1172962226640164e-07, + "loss": 1.2775, + "step": 217 + }, + { + "epoch": 0.21456692913385828, + "grad_norm": 23.31028938293457, + "learning_rate": 2.127236580516899e-07, + "loss": 1.6795, + "step": 218 + }, + { + "epoch": 0.2155511811023622, + "grad_norm": 24.057025909423828, + "learning_rate": 2.1371769383697817e-07, + "loss": 1.2862, + "step": 219 + }, + { + "epoch": 0.21653543307086615, + "grad_norm": 17.363821029663086, + "learning_rate": 2.1471172962226642e-07, + "loss": 1.1723, + "step": 220 + }, + { + "epoch": 0.21751968503937008, + "grad_norm": 22.69175148010254, + "learning_rate": 2.1570576540755473e-07, + "loss": 1.3322, + "step": 221 + }, + { + "epoch": 0.21850393700787402, + "grad_norm": 24.083465576171875, + "learning_rate": 2.1669980119284298e-07, + "loss": 1.7564, + "step": 222 + }, + { + "epoch": 0.21948818897637795, + "grad_norm": 17.775938034057617, + "learning_rate": 2.1769383697813126e-07, + "loss": 1.1071, + "step": 223 + }, + { + "epoch": 0.2204724409448819, + "grad_norm": 18.539440155029297, + "learning_rate": 2.186878727634195e-07, + "loss": 1.2011, + "step": 224 + }, + { + "epoch": 0.22145669291338582, + "grad_norm": 18.234455108642578, + "learning_rate": 2.196819085487078e-07, + "loss": 1.2303, + "step": 225 + }, + { + "epoch": 0.22244094488188976, + "grad_norm": 20.291522979736328, + "learning_rate": 2.2067594433399604e-07, + "loss": 1.212, + "step": 226 + }, + { + "epoch": 0.2234251968503937, + "grad_norm": 15.60973072052002, + "learning_rate": 2.2166998011928432e-07, + "loss": 1.0117, + "step": 227 + }, + { + "epoch": 0.22440944881889763, + "grad_norm": 20.482995986938477, + "learning_rate": 2.2266401590457263e-07, + "loss": 1.1907, + "step": 228 + }, + { + "epoch": 0.22539370078740156, + "grad_norm": 24.126340866088867, + "learning_rate": 2.2365805168986088e-07, + "loss": 2.1293, + "step": 229 + }, + { + "epoch": 0.2263779527559055, + "grad_norm": 21.279150009155273, + "learning_rate": 2.2465208747514916e-07, + "loss": 1.3063, + "step": 230 + }, + { + "epoch": 0.22736220472440946, + "grad_norm": 20.62582778930664, + "learning_rate": 2.256461232604374e-07, + "loss": 1.2841, + "step": 231 + }, + { + "epoch": 0.2283464566929134, + "grad_norm": 20.63606834411621, + "learning_rate": 2.266401590457257e-07, + "loss": 1.3778, + "step": 232 + }, + { + "epoch": 0.22933070866141733, + "grad_norm": 18.099950790405273, + "learning_rate": 2.2763419483101394e-07, + "loss": 1.2242, + "step": 233 + }, + { + "epoch": 0.23031496062992127, + "grad_norm": 18.96858024597168, + "learning_rate": 2.2862823061630222e-07, + "loss": 0.9227, + "step": 234 + }, + { + "epoch": 0.2312992125984252, + "grad_norm": 22.201169967651367, + "learning_rate": 2.296222664015905e-07, + "loss": 1.2221, + "step": 235 + }, + { + "epoch": 0.23228346456692914, + "grad_norm": 26.657480239868164, + "learning_rate": 2.3061630218687878e-07, + "loss": 2.1041, + "step": 236 + }, + { + "epoch": 0.23326771653543307, + "grad_norm": 25.558757781982422, + "learning_rate": 2.3161033797216703e-07, + "loss": 1.3341, + "step": 237 + }, + { + "epoch": 0.234251968503937, + "grad_norm": 22.0543212890625, + "learning_rate": 2.326043737574553e-07, + "loss": 1.0876, + "step": 238 + }, + { + "epoch": 0.23523622047244094, + "grad_norm": 22.258333206176758, + "learning_rate": 2.335984095427436e-07, + "loss": 1.3328, + "step": 239 + }, + { + "epoch": 0.23622047244094488, + "grad_norm": 21.519908905029297, + "learning_rate": 2.3459244532803184e-07, + "loss": 1.2958, + "step": 240 + }, + { + "epoch": 0.2372047244094488, + "grad_norm": 23.345388412475586, + "learning_rate": 2.3558648111332012e-07, + "loss": 1.1522, + "step": 241 + }, + { + "epoch": 0.23818897637795275, + "grad_norm": 25.306140899658203, + "learning_rate": 2.365805168986084e-07, + "loss": 1.7942, + "step": 242 + }, + { + "epoch": 0.23917322834645668, + "grad_norm": 21.840364456176758, + "learning_rate": 2.3757455268389668e-07, + "loss": 1.1325, + "step": 243 + }, + { + "epoch": 0.24015748031496062, + "grad_norm": 26.5495662689209, + "learning_rate": 2.385685884691849e-07, + "loss": 1.6466, + "step": 244 + }, + { + "epoch": 0.24114173228346455, + "grad_norm": 21.928937911987305, + "learning_rate": 2.395626242544732e-07, + "loss": 1.4608, + "step": 245 + }, + { + "epoch": 0.2421259842519685, + "grad_norm": 16.23383903503418, + "learning_rate": 2.4055666003976146e-07, + "loss": 0.6375, + "step": 246 + }, + { + "epoch": 0.24311023622047245, + "grad_norm": 26.922739028930664, + "learning_rate": 2.4155069582504976e-07, + "loss": 2.0177, + "step": 247 + }, + { + "epoch": 0.2440944881889764, + "grad_norm": 23.482742309570312, + "learning_rate": 2.42544731610338e-07, + "loss": 1.2069, + "step": 248 + }, + { + "epoch": 0.24507874015748032, + "grad_norm": 14.740228652954102, + "learning_rate": 2.4353876739562627e-07, + "loss": 0.7639, + "step": 249 + }, + { + "epoch": 0.24606299212598426, + "grad_norm": 20.238948822021484, + "learning_rate": 2.445328031809146e-07, + "loss": 1.3465, + "step": 250 + }, + { + "epoch": 0.2470472440944882, + "grad_norm": 16.91303253173828, + "learning_rate": 2.4552683896620283e-07, + "loss": 1.064, + "step": 251 + }, + { + "epoch": 0.24803149606299213, + "grad_norm": 22.081308364868164, + "learning_rate": 2.4652087475149113e-07, + "loss": 1.3757, + "step": 252 + }, + { + "epoch": 0.24901574803149606, + "grad_norm": 21.756267547607422, + "learning_rate": 2.475149105367794e-07, + "loss": 1.612, + "step": 253 + }, + { + "epoch": 0.25, + "grad_norm": 16.87566566467285, + "learning_rate": 2.4850894632206764e-07, + "loss": 0.7917, + "step": 254 + }, + { + "epoch": 0.25098425196850394, + "grad_norm": 22.13374900817871, + "learning_rate": 2.495029821073559e-07, + "loss": 1.5515, + "step": 255 + }, + { + "epoch": 0.25196850393700787, + "grad_norm": 18.96138572692871, + "learning_rate": 2.5049701789264414e-07, + "loss": 0.799, + "step": 256 + }, + { + "epoch": 0.2529527559055118, + "grad_norm": 17.10991859436035, + "learning_rate": 2.5149105367793245e-07, + "loss": 0.9882, + "step": 257 + }, + { + "epoch": 0.25393700787401574, + "grad_norm": 15.347769737243652, + "learning_rate": 2.524850894632207e-07, + "loss": 1.1814, + "step": 258 + }, + { + "epoch": 0.2549212598425197, + "grad_norm": 11.857462882995605, + "learning_rate": 2.53479125248509e-07, + "loss": 0.6394, + "step": 259 + }, + { + "epoch": 0.2559055118110236, + "grad_norm": 20.00450325012207, + "learning_rate": 2.5447316103379726e-07, + "loss": 1.4756, + "step": 260 + }, + { + "epoch": 0.25688976377952755, + "grad_norm": 12.062335968017578, + "learning_rate": 2.554671968190855e-07, + "loss": 0.5338, + "step": 261 + }, + { + "epoch": 0.2578740157480315, + "grad_norm": 16.912168502807617, + "learning_rate": 2.564612326043738e-07, + "loss": 0.9779, + "step": 262 + }, + { + "epoch": 0.2588582677165354, + "grad_norm": 22.713825225830078, + "learning_rate": 2.5745526838966207e-07, + "loss": 1.5307, + "step": 263 + }, + { + "epoch": 0.25984251968503935, + "grad_norm": 22.88834571838379, + "learning_rate": 2.5844930417495037e-07, + "loss": 1.1213, + "step": 264 + }, + { + "epoch": 0.2608267716535433, + "grad_norm": 18.133310317993164, + "learning_rate": 2.5944333996023857e-07, + "loss": 0.9482, + "step": 265 + }, + { + "epoch": 0.2618110236220472, + "grad_norm": 19.092018127441406, + "learning_rate": 2.604373757455269e-07, + "loss": 0.9599, + "step": 266 + }, + { + "epoch": 0.26279527559055116, + "grad_norm": 25.283647537231445, + "learning_rate": 2.614314115308152e-07, + "loss": 1.4455, + "step": 267 + }, + { + "epoch": 0.2637795275590551, + "grad_norm": 28.519742965698242, + "learning_rate": 2.6242544731610343e-07, + "loss": 1.6496, + "step": 268 + }, + { + "epoch": 0.26476377952755903, + "grad_norm": 15.56734561920166, + "learning_rate": 2.634194831013917e-07, + "loss": 0.7402, + "step": 269 + }, + { + "epoch": 0.265748031496063, + "grad_norm": 16.450769424438477, + "learning_rate": 2.6441351888667994e-07, + "loss": 0.7835, + "step": 270 + }, + { + "epoch": 0.26673228346456695, + "grad_norm": 18.447402954101562, + "learning_rate": 2.6540755467196824e-07, + "loss": 0.7821, + "step": 271 + }, + { + "epoch": 0.2677165354330709, + "grad_norm": 22.634441375732422, + "learning_rate": 2.664015904572565e-07, + "loss": 1.5422, + "step": 272 + }, + { + "epoch": 0.2687007874015748, + "grad_norm": 20.091428756713867, + "learning_rate": 2.6739562624254475e-07, + "loss": 1.0995, + "step": 273 + }, + { + "epoch": 0.26968503937007876, + "grad_norm": 22.89579200744629, + "learning_rate": 2.6838966202783305e-07, + "loss": 1.378, + "step": 274 + }, + { + "epoch": 0.2706692913385827, + "grad_norm": 25.36764144897461, + "learning_rate": 2.693836978131213e-07, + "loss": 1.3562, + "step": 275 + }, + { + "epoch": 0.27165354330708663, + "grad_norm": 15.717765808105469, + "learning_rate": 2.703777335984096e-07, + "loss": 0.7376, + "step": 276 + }, + { + "epoch": 0.27263779527559057, + "grad_norm": 21.739999771118164, + "learning_rate": 2.7137176938369786e-07, + "loss": 1.1678, + "step": 277 + }, + { + "epoch": 0.2736220472440945, + "grad_norm": 21.12672233581543, + "learning_rate": 2.723658051689861e-07, + "loss": 1.2989, + "step": 278 + }, + { + "epoch": 0.27460629921259844, + "grad_norm": 22.614158630371094, + "learning_rate": 2.7335984095427437e-07, + "loss": 1.9559, + "step": 279 + }, + { + "epoch": 0.2755905511811024, + "grad_norm": 18.742107391357422, + "learning_rate": 2.743538767395627e-07, + "loss": 1.1237, + "step": 280 + }, + { + "epoch": 0.2765748031496063, + "grad_norm": 17.969554901123047, + "learning_rate": 2.75347912524851e-07, + "loss": 0.952, + "step": 281 + }, + { + "epoch": 0.27755905511811024, + "grad_norm": 24.01414680480957, + "learning_rate": 2.763419483101392e-07, + "loss": 1.6629, + "step": 282 + }, + { + "epoch": 0.2785433070866142, + "grad_norm": 21.362796783447266, + "learning_rate": 2.773359840954275e-07, + "loss": 1.871, + "step": 283 + }, + { + "epoch": 0.2795275590551181, + "grad_norm": 19.05644989013672, + "learning_rate": 2.7833001988071574e-07, + "loss": 1.5946, + "step": 284 + }, + { + "epoch": 0.28051181102362205, + "grad_norm": 18.40645408630371, + "learning_rate": 2.7932405566600404e-07, + "loss": 1.4456, + "step": 285 + }, + { + "epoch": 0.281496062992126, + "grad_norm": 18.53467559814453, + "learning_rate": 2.803180914512923e-07, + "loss": 1.4085, + "step": 286 + }, + { + "epoch": 0.2824803149606299, + "grad_norm": 18.1446533203125, + "learning_rate": 2.8131212723658055e-07, + "loss": 1.1394, + "step": 287 + }, + { + "epoch": 0.28346456692913385, + "grad_norm": 15.109747886657715, + "learning_rate": 2.8230616302186885e-07, + "loss": 1.0315, + "step": 288 + }, + { + "epoch": 0.2844488188976378, + "grad_norm": 22.63469696044922, + "learning_rate": 2.833001988071571e-07, + "loss": 1.488, + "step": 289 + }, + { + "epoch": 0.2854330708661417, + "grad_norm": 18.841909408569336, + "learning_rate": 2.842942345924454e-07, + "loss": 1.4006, + "step": 290 + }, + { + "epoch": 0.28641732283464566, + "grad_norm": 20.073745727539062, + "learning_rate": 2.852882703777336e-07, + "loss": 0.9237, + "step": 291 + }, + { + "epoch": 0.2874015748031496, + "grad_norm": 18.586259841918945, + "learning_rate": 2.862823061630219e-07, + "loss": 1.163, + "step": 292 + }, + { + "epoch": 0.28838582677165353, + "grad_norm": 23.599292755126953, + "learning_rate": 2.8727634194831017e-07, + "loss": 1.7037, + "step": 293 + }, + { + "epoch": 0.28937007874015747, + "grad_norm": 16.384347915649414, + "learning_rate": 2.8827037773359847e-07, + "loss": 0.8715, + "step": 294 + }, + { + "epoch": 0.2903543307086614, + "grad_norm": 20.26093101501465, + "learning_rate": 2.892644135188867e-07, + "loss": 1.2101, + "step": 295 + }, + { + "epoch": 0.29133858267716534, + "grad_norm": 19.598831176757812, + "learning_rate": 2.90258449304175e-07, + "loss": 1.1179, + "step": 296 + }, + { + "epoch": 0.29232283464566927, + "grad_norm": 22.65572738647461, + "learning_rate": 2.912524850894633e-07, + "loss": 1.3986, + "step": 297 + }, + { + "epoch": 0.2933070866141732, + "grad_norm": 22.658931732177734, + "learning_rate": 2.9224652087475153e-07, + "loss": 1.7068, + "step": 298 + }, + { + "epoch": 0.29429133858267714, + "grad_norm": 16.94671058654785, + "learning_rate": 2.9324055666003984e-07, + "loss": 0.8695, + "step": 299 + }, + { + "epoch": 0.2952755905511811, + "grad_norm": 20.13766860961914, + "learning_rate": 2.9423459244532804e-07, + "loss": 1.3778, + "step": 300 + }, + { + "epoch": 0.296259842519685, + "grad_norm": 16.939062118530273, + "learning_rate": 2.9522862823061634e-07, + "loss": 1.2834, + "step": 301 + }, + { + "epoch": 0.297244094488189, + "grad_norm": 14.732866287231445, + "learning_rate": 2.9622266401590465e-07, + "loss": 0.8123, + "step": 302 + }, + { + "epoch": 0.29822834645669294, + "grad_norm": 23.4341983795166, + "learning_rate": 2.972166998011929e-07, + "loss": 1.6521, + "step": 303 + }, + { + "epoch": 0.2992125984251969, + "grad_norm": 15.455023765563965, + "learning_rate": 2.9821073558648115e-07, + "loss": 1.1064, + "step": 304 + }, + { + "epoch": 0.3001968503937008, + "grad_norm": 15.683412551879883, + "learning_rate": 2.992047713717694e-07, + "loss": 0.9578, + "step": 305 + }, + { + "epoch": 0.30118110236220474, + "grad_norm": 16.594762802124023, + "learning_rate": 3.001988071570577e-07, + "loss": 0.9254, + "step": 306 + }, + { + "epoch": 0.30118110236220474, + "eval_Qnli-dev_cosine_accuracy": 0.6796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8338999152183533, + "eval_Qnli-dev_cosine_ap": 0.7040139627388139, + "eval_Qnli-dev_cosine_f1": 0.6716697936210132, + "eval_Qnli-dev_cosine_f1_threshold": 0.7830559015274048, + "eval_Qnli-dev_cosine_precision": 0.6026936026936027, + "eval_Qnli-dev_cosine_recall": 0.7584745762711864, + "eval_Qnli-dev_dot_accuracy": 0.6796875, + "eval_Qnli-dev_dot_accuracy_threshold": 640.5870361328125, + "eval_Qnli-dev_dot_ap": 0.7040479892892765, + "eval_Qnli-dev_dot_f1": 0.6716697936210132, + "eval_Qnli-dev_dot_f1_threshold": 601.5180053710938, + "eval_Qnli-dev_dot_precision": 0.6026936026936027, + "eval_Qnli-dev_dot_recall": 0.7584745762711864, + "eval_Qnli-dev_euclidean_accuracy": 0.6796875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.974668502807617, + "eval_Qnli-dev_euclidean_ap": 0.7040139627388139, + "eval_Qnli-dev_euclidean_f1": 0.6716697936210132, + "eval_Qnli-dev_euclidean_f1_threshold": 18.256603240966797, + "eval_Qnli-dev_euclidean_precision": 0.6026936026936027, + "eval_Qnli-dev_euclidean_recall": 0.7584745762711864, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 354.99566650390625, + "eval_Qnli-dev_manhattan_ap": 0.7020916363435745, + "eval_Qnli-dev_manhattan_f1": 0.6715328467153285, + "eval_Qnli-dev_manhattan_f1_threshold": 406.69915771484375, + "eval_Qnli-dev_manhattan_precision": 0.5897435897435898, + "eval_Qnli-dev_manhattan_recall": 0.7796610169491526, + "eval_Qnli-dev_max_accuracy": 0.6796875, + "eval_Qnli-dev_max_accuracy_threshold": 640.5870361328125, + "eval_Qnli-dev_max_ap": 0.7040479892892765, + "eval_Qnli-dev_max_f1": 0.6716697936210132, + "eval_Qnli-dev_max_f1_threshold": 601.5180053710938, + "eval_Qnli-dev_max_precision": 0.6026936026936027, + "eval_Qnli-dev_max_recall": 0.7796610169491526, + "eval_allNLI-dev_cosine_accuracy": 0.685546875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.931303083896637, + "eval_allNLI-dev_cosine_ap": 0.4777228569875479, + "eval_allNLI-dev_cosine_f1": 0.5344467640918581, + "eval_allNLI-dev_cosine_f1_threshold": 0.8382494449615479, + "eval_allNLI-dev_cosine_precision": 0.41830065359477125, + "eval_allNLI-dev_cosine_recall": 0.7398843930635838, + "eval_allNLI-dev_dot_accuracy": 0.685546875, + "eval_allNLI-dev_dot_accuracy_threshold": 715.4099731445312, + "eval_allNLI-dev_dot_ap": 0.47773833073270383, + "eval_allNLI-dev_dot_f1": 0.5344467640918581, + "eval_allNLI-dev_dot_f1_threshold": 643.9212646484375, + "eval_allNLI-dev_dot_precision": 0.41830065359477125, + "eval_allNLI-dev_dot_recall": 0.7398843930635838, + "eval_allNLI-dev_euclidean_accuracy": 0.685546875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.273423194885254, + "eval_allNLI-dev_euclidean_ap": 0.477722856987548, + "eval_allNLI-dev_euclidean_f1": 0.5344467640918581, + "eval_allNLI-dev_euclidean_f1_threshold": 15.764041900634766, + "eval_allNLI-dev_euclidean_precision": 0.41830065359477125, + "eval_allNLI-dev_euclidean_recall": 0.7398843930635838, + "eval_allNLI-dev_manhattan_accuracy": 0.685546875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 236.41517639160156, + "eval_allNLI-dev_manhattan_ap": 0.47890307996210313, + "eval_allNLI-dev_manhattan_f1": 0.5341880341880343, + "eval_allNLI-dev_manhattan_f1_threshold": 339.87042236328125, + "eval_allNLI-dev_manhattan_precision": 0.423728813559322, + "eval_allNLI-dev_manhattan_recall": 0.7225433526011561, + "eval_allNLI-dev_max_accuracy": 0.685546875, + "eval_allNLI-dev_max_accuracy_threshold": 715.4099731445312, + "eval_allNLI-dev_max_ap": 0.47890307996210313, + "eval_allNLI-dev_max_f1": 0.5344467640918581, + "eval_allNLI-dev_max_f1_threshold": 643.9212646484375, + "eval_allNLI-dev_max_precision": 0.423728813559322, + "eval_allNLI-dev_max_recall": 0.7398843930635838, + "eval_runtime": 1.2301, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.7040479892892765, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.4665812233506641, + "eval_sts-test_pearson_dot": 0.4665227418822839, + "eval_sts-test_pearson_euclidean": 0.47993570552371345, + "eval_sts-test_pearson_manhattan": 0.4795106797562281, + "eval_sts-test_pearson_max": 0.47993570552371345, + "eval_sts-test_spearman_cosine": 0.48884197010310126, + "eval_sts-test_spearman_dot": 0.4886054596866477, + "eval_sts-test_spearman_euclidean": 0.4888477430158208, + "eval_sts-test_spearman_manhattan": 0.487254956120378, + "eval_sts-test_spearman_max": 0.4888477430158208, + "step": 306 + }, + { + "epoch": 0.3021653543307087, + "grad_norm": 17.415273666381836, + "learning_rate": 3.01192842942346e-07, + "loss": 0.7541, + "step": 307 + }, + { + "epoch": 0.3031496062992126, + "grad_norm": 15.18093204498291, + "learning_rate": 3.021868787276342e-07, + "loss": 0.7324, + "step": 308 + }, + { + "epoch": 0.30413385826771655, + "grad_norm": 14.747346878051758, + "learning_rate": 3.031809145129225e-07, + "loss": 0.5974, + "step": 309 + }, + { + "epoch": 0.3051181102362205, + "grad_norm": 23.3188533782959, + "learning_rate": 3.0417495029821077e-07, + "loss": 1.1481, + "step": 310 + }, + { + "epoch": 0.3061023622047244, + "grad_norm": 28.704936981201172, + "learning_rate": 3.051689860834991e-07, + "loss": 1.6179, + "step": 311 + }, + { + "epoch": 0.30708661417322836, + "grad_norm": 26.603492736816406, + "learning_rate": 3.0616302186878733e-07, + "loss": 1.4641, + "step": 312 + }, + { + "epoch": 0.3080708661417323, + "grad_norm": 30.99028968811035, + "learning_rate": 3.071570576540756e-07, + "loss": 1.7185, + "step": 313 + }, + { + "epoch": 0.3090551181102362, + "grad_norm": 21.485858917236328, + "learning_rate": 3.0815109343936384e-07, + "loss": 0.9328, + "step": 314 + }, + { + "epoch": 0.31003937007874016, + "grad_norm": 20.982534408569336, + "learning_rate": 3.0914512922465214e-07, + "loss": 0.742, + "step": 315 + }, + { + "epoch": 0.3110236220472441, + "grad_norm": 24.897323608398438, + "learning_rate": 3.1013916500994045e-07, + "loss": 1.4173, + "step": 316 + }, + { + "epoch": 0.31200787401574803, + "grad_norm": 18.03276252746582, + "learning_rate": 3.1113320079522865e-07, + "loss": 0.7267, + "step": 317 + }, + { + "epoch": 0.31299212598425197, + "grad_norm": 18.415945053100586, + "learning_rate": 3.1212723658051695e-07, + "loss": 0.9494, + "step": 318 + }, + { + "epoch": 0.3139763779527559, + "grad_norm": 23.052213668823242, + "learning_rate": 3.131212723658052e-07, + "loss": 1.5111, + "step": 319 + }, + { + "epoch": 0.31496062992125984, + "grad_norm": 21.024145126342773, + "learning_rate": 3.141153081510935e-07, + "loss": 1.6949, + "step": 320 + }, + { + "epoch": 0.3159448818897638, + "grad_norm": 24.442323684692383, + "learning_rate": 3.151093439363818e-07, + "loss": 1.7562, + "step": 321 + }, + { + "epoch": 0.3169291338582677, + "grad_norm": 16.658889770507812, + "learning_rate": 3.1610337972167e-07, + "loss": 1.2532, + "step": 322 + }, + { + "epoch": 0.31791338582677164, + "grad_norm": 12.664291381835938, + "learning_rate": 3.170974155069583e-07, + "loss": 1.1086, + "step": 323 + }, + { + "epoch": 0.3188976377952756, + "grad_norm": 16.310880661010742, + "learning_rate": 3.1809145129224657e-07, + "loss": 0.7377, + "step": 324 + }, + { + "epoch": 0.3198818897637795, + "grad_norm": 18.340065002441406, + "learning_rate": 3.190854870775349e-07, + "loss": 1.085, + "step": 325 + }, + { + "epoch": 0.32086614173228345, + "grad_norm": 14.548378944396973, + "learning_rate": 3.200795228628231e-07, + "loss": 0.7767, + "step": 326 + }, + { + "epoch": 0.3218503937007874, + "grad_norm": 19.989688873291016, + "learning_rate": 3.210735586481114e-07, + "loss": 1.4441, + "step": 327 + }, + { + "epoch": 0.3228346456692913, + "grad_norm": 16.6571102142334, + "learning_rate": 3.220675944333997e-07, + "loss": 0.8146, + "step": 328 + }, + { + "epoch": 0.32381889763779526, + "grad_norm": 16.355384826660156, + "learning_rate": 3.2306163021868794e-07, + "loss": 0.7403, + "step": 329 + }, + { + "epoch": 0.3248031496062992, + "grad_norm": 19.589183807373047, + "learning_rate": 3.240556660039762e-07, + "loss": 0.8476, + "step": 330 + }, + { + "epoch": 0.3257874015748031, + "grad_norm": 16.7521915435791, + "learning_rate": 3.2504970178926444e-07, + "loss": 0.7323, + "step": 331 + }, + { + "epoch": 0.32677165354330706, + "grad_norm": 22.762434005737305, + "learning_rate": 3.2604373757455275e-07, + "loss": 1.2241, + "step": 332 + }, + { + "epoch": 0.327755905511811, + "grad_norm": 21.42218017578125, + "learning_rate": 3.27037773359841e-07, + "loss": 1.5065, + "step": 333 + }, + { + "epoch": 0.328740157480315, + "grad_norm": 12.703171730041504, + "learning_rate": 3.280318091451293e-07, + "loss": 0.5259, + "step": 334 + }, + { + "epoch": 0.3297244094488189, + "grad_norm": 22.40658187866211, + "learning_rate": 3.2902584493041756e-07, + "loss": 1.3103, + "step": 335 + }, + { + "epoch": 0.33070866141732286, + "grad_norm": 19.746475219726562, + "learning_rate": 3.300198807157058e-07, + "loss": 0.8655, + "step": 336 + }, + { + "epoch": 0.3316929133858268, + "grad_norm": 19.323755264282227, + "learning_rate": 3.310139165009941e-07, + "loss": 0.7575, + "step": 337 + }, + { + "epoch": 0.33267716535433073, + "grad_norm": 24.301485061645508, + "learning_rate": 3.3200795228628237e-07, + "loss": 1.968, + "step": 338 + }, + { + "epoch": 0.33366141732283466, + "grad_norm": 20.01712417602539, + "learning_rate": 3.330019880715706e-07, + "loss": 1.317, + "step": 339 + }, + { + "epoch": 0.3346456692913386, + "grad_norm": 17.13825035095215, + "learning_rate": 3.3399602385685887e-07, + "loss": 1.1972, + "step": 340 + }, + { + "epoch": 0.33562992125984253, + "grad_norm": 22.386058807373047, + "learning_rate": 3.349900596421472e-07, + "loss": 1.6323, + "step": 341 + }, + { + "epoch": 0.33661417322834647, + "grad_norm": 16.26036834716797, + "learning_rate": 3.359840954274355e-07, + "loss": 1.0469, + "step": 342 + }, + { + "epoch": 0.3375984251968504, + "grad_norm": 16.711883544921875, + "learning_rate": 3.369781312127237e-07, + "loss": 1.3349, + "step": 343 + }, + { + "epoch": 0.33858267716535434, + "grad_norm": 14.21274471282959, + "learning_rate": 3.37972166998012e-07, + "loss": 0.9544, + "step": 344 + }, + { + "epoch": 0.3395669291338583, + "grad_norm": 18.23252296447754, + "learning_rate": 3.3896620278330024e-07, + "loss": 1.1894, + "step": 345 + }, + { + "epoch": 0.3405511811023622, + "grad_norm": 13.04709243774414, + "learning_rate": 3.3996023856858855e-07, + "loss": 0.7717, + "step": 346 + }, + { + "epoch": 0.34153543307086615, + "grad_norm": 20.66889190673828, + "learning_rate": 3.409542743538768e-07, + "loss": 1.2563, + "step": 347 + }, + { + "epoch": 0.3425196850393701, + "grad_norm": 19.415496826171875, + "learning_rate": 3.4194831013916505e-07, + "loss": 1.2437, + "step": 348 + }, + { + "epoch": 0.343503937007874, + "grad_norm": 18.92940330505371, + "learning_rate": 3.4294234592445336e-07, + "loss": 0.7806, + "step": 349 + }, + { + "epoch": 0.34448818897637795, + "grad_norm": 18.86058235168457, + "learning_rate": 3.439363817097416e-07, + "loss": 0.8303, + "step": 350 + }, + { + "epoch": 0.3454724409448819, + "grad_norm": 22.05217933654785, + "learning_rate": 3.449304174950299e-07, + "loss": 1.0926, + "step": 351 + }, + { + "epoch": 0.3464566929133858, + "grad_norm": 18.41050910949707, + "learning_rate": 3.459244532803181e-07, + "loss": 0.6654, + "step": 352 + }, + { + "epoch": 0.34744094488188976, + "grad_norm": 20.6577091217041, + "learning_rate": 3.469184890656064e-07, + "loss": 1.1087, + "step": 353 + }, + { + "epoch": 0.3484251968503937, + "grad_norm": 21.558391571044922, + "learning_rate": 3.4791252485089467e-07, + "loss": 1.1525, + "step": 354 + }, + { + "epoch": 0.3494094488188976, + "grad_norm": 17.46851921081543, + "learning_rate": 3.48906560636183e-07, + "loss": 1.1127, + "step": 355 + }, + { + "epoch": 0.35039370078740156, + "grad_norm": 19.606983184814453, + "learning_rate": 3.499005964214713e-07, + "loss": 1.4267, + "step": 356 + }, + { + "epoch": 0.3513779527559055, + "grad_norm": 16.4560604095459, + "learning_rate": 3.508946322067595e-07, + "loss": 0.6148, + "step": 357 + }, + { + "epoch": 0.35236220472440943, + "grad_norm": 18.274593353271484, + "learning_rate": 3.518886679920478e-07, + "loss": 1.0123, + "step": 358 + }, + { + "epoch": 0.35334645669291337, + "grad_norm": 26.427845001220703, + "learning_rate": 3.5288270377733604e-07, + "loss": 1.9682, + "step": 359 + }, + { + "epoch": 0.3543307086614173, + "grad_norm": 15.270173072814941, + "learning_rate": 3.5387673956262434e-07, + "loss": 0.8487, + "step": 360 + }, + { + "epoch": 0.35531496062992124, + "grad_norm": 15.928024291992188, + "learning_rate": 3.5487077534791254e-07, + "loss": 1.0412, + "step": 361 + }, + { + "epoch": 0.3562992125984252, + "grad_norm": 16.021299362182617, + "learning_rate": 3.5586481113320085e-07, + "loss": 1.0902, + "step": 362 + }, + { + "epoch": 0.3572834645669291, + "grad_norm": 15.06041431427002, + "learning_rate": 3.5685884691848915e-07, + "loss": 0.9606, + "step": 363 + }, + { + "epoch": 0.35826771653543305, + "grad_norm": 17.597152709960938, + "learning_rate": 3.578528827037774e-07, + "loss": 0.9206, + "step": 364 + }, + { + "epoch": 0.359251968503937, + "grad_norm": 19.437129974365234, + "learning_rate": 3.5884691848906566e-07, + "loss": 1.4727, + "step": 365 + }, + { + "epoch": 0.36023622047244097, + "grad_norm": 16.03767204284668, + "learning_rate": 3.598409542743539e-07, + "loss": 0.9379, + "step": 366 + }, + { + "epoch": 0.3612204724409449, + "grad_norm": 16.83603286743164, + "learning_rate": 3.608349900596422e-07, + "loss": 0.8387, + "step": 367 + }, + { + "epoch": 0.36220472440944884, + "grad_norm": 18.7794132232666, + "learning_rate": 3.6182902584493047e-07, + "loss": 0.9692, + "step": 368 + }, + { + "epoch": 0.3631889763779528, + "grad_norm": 21.516565322875977, + "learning_rate": 3.6282306163021877e-07, + "loss": 1.6298, + "step": 369 + }, + { + "epoch": 0.3641732283464567, + "grad_norm": 14.46429443359375, + "learning_rate": 3.63817097415507e-07, + "loss": 1.0882, + "step": 370 + }, + { + "epoch": 0.36515748031496065, + "grad_norm": 18.343765258789062, + "learning_rate": 3.648111332007953e-07, + "loss": 1.1558, + "step": 371 + }, + { + "epoch": 0.3661417322834646, + "grad_norm": 16.055957794189453, + "learning_rate": 3.658051689860836e-07, + "loss": 0.9546, + "step": 372 + }, + { + "epoch": 0.3671259842519685, + "grad_norm": 15.39949893951416, + "learning_rate": 3.6679920477137184e-07, + "loss": 1.0124, + "step": 373 + }, + { + "epoch": 0.36811023622047245, + "grad_norm": 15.101058006286621, + "learning_rate": 3.677932405566601e-07, + "loss": 1.3916, + "step": 374 + }, + { + "epoch": 0.3690944881889764, + "grad_norm": 11.735250473022461, + "learning_rate": 3.6878727634194834e-07, + "loss": 0.527, + "step": 375 + }, + { + "epoch": 0.3700787401574803, + "grad_norm": 13.659126281738281, + "learning_rate": 3.6978131212723665e-07, + "loss": 0.6387, + "step": 376 + }, + { + "epoch": 0.37106299212598426, + "grad_norm": 18.6676025390625, + "learning_rate": 3.7077534791252495e-07, + "loss": 1.1445, + "step": 377 + }, + { + "epoch": 0.3720472440944882, + "grad_norm": 23.352149963378906, + "learning_rate": 3.7176938369781315e-07, + "loss": 1.3309, + "step": 378 + }, + { + "epoch": 0.37303149606299213, + "grad_norm": 25.64788818359375, + "learning_rate": 3.7276341948310145e-07, + "loss": 1.5888, + "step": 379 + }, + { + "epoch": 0.37401574803149606, + "grad_norm": 23.087392807006836, + "learning_rate": 3.737574552683897e-07, + "loss": 1.4422, + "step": 380 + }, + { + "epoch": 0.375, + "grad_norm": 21.51338005065918, + "learning_rate": 3.74751491053678e-07, + "loss": 1.7044, + "step": 381 + }, + { + "epoch": 0.37598425196850394, + "grad_norm": 13.316896438598633, + "learning_rate": 3.7574552683896626e-07, + "loss": 0.7913, + "step": 382 + }, + { + "epoch": 0.37696850393700787, + "grad_norm": 19.068180084228516, + "learning_rate": 3.767395626242545e-07, + "loss": 1.3241, + "step": 383 + }, + { + "epoch": 0.3779527559055118, + "grad_norm": 11.564128875732422, + "learning_rate": 3.777335984095428e-07, + "loss": 0.6473, + "step": 384 + }, + { + "epoch": 0.37893700787401574, + "grad_norm": 18.001808166503906, + "learning_rate": 3.787276341948311e-07, + "loss": 1.221, + "step": 385 + }, + { + "epoch": 0.3799212598425197, + "grad_norm": 13.824004173278809, + "learning_rate": 3.797216699801194e-07, + "loss": 0.7773, + "step": 386 + }, + { + "epoch": 0.3809055118110236, + "grad_norm": 14.364906311035156, + "learning_rate": 3.807157057654076e-07, + "loss": 1.054, + "step": 387 + }, + { + "epoch": 0.38188976377952755, + "grad_norm": 16.183347702026367, + "learning_rate": 3.817097415506959e-07, + "loss": 0.9862, + "step": 388 + }, + { + "epoch": 0.3828740157480315, + "grad_norm": 16.849651336669922, + "learning_rate": 3.8270377733598414e-07, + "loss": 0.9684, + "step": 389 + }, + { + "epoch": 0.3838582677165354, + "grad_norm": 17.902210235595703, + "learning_rate": 3.8369781312127244e-07, + "loss": 1.3244, + "step": 390 + }, + { + "epoch": 0.38484251968503935, + "grad_norm": 20.95703887939453, + "learning_rate": 3.8469184890656075e-07, + "loss": 1.1787, + "step": 391 + }, + { + "epoch": 0.3858267716535433, + "grad_norm": 22.145349502563477, + "learning_rate": 3.8568588469184895e-07, + "loss": 1.4698, + "step": 392 + }, + { + "epoch": 0.3868110236220472, + "grad_norm": 21.477500915527344, + "learning_rate": 3.8667992047713725e-07, + "loss": 1.0961, + "step": 393 + }, + { + "epoch": 0.38779527559055116, + "grad_norm": 17.14502716064453, + "learning_rate": 3.876739562624255e-07, + "loss": 1.1364, + "step": 394 + }, + { + "epoch": 0.3887795275590551, + "grad_norm": 18.223770141601562, + "learning_rate": 3.886679920477138e-07, + "loss": 0.9368, + "step": 395 + }, + { + "epoch": 0.38976377952755903, + "grad_norm": 20.13445472717285, + "learning_rate": 3.89662027833002e-07, + "loss": 1.1731, + "step": 396 + }, + { + "epoch": 0.390748031496063, + "grad_norm": 17.56426429748535, + "learning_rate": 3.906560636182903e-07, + "loss": 0.8686, + "step": 397 + }, + { + "epoch": 0.39173228346456695, + "grad_norm": 15.231781005859375, + "learning_rate": 3.916500994035786e-07, + "loss": 0.7481, + "step": 398 + }, + { + "epoch": 0.3927165354330709, + "grad_norm": 17.00189781188965, + "learning_rate": 3.9264413518886687e-07, + "loss": 0.7261, + "step": 399 + }, + { + "epoch": 0.3937007874015748, + "grad_norm": 17.566858291625977, + "learning_rate": 3.936381709741551e-07, + "loss": 1.2062, + "step": 400 + }, + { + "epoch": 0.39468503937007876, + "grad_norm": 15.011639595031738, + "learning_rate": 3.946322067594434e-07, + "loss": 0.7462, + "step": 401 + }, + { + "epoch": 0.3956692913385827, + "grad_norm": 19.905258178710938, + "learning_rate": 3.956262425447317e-07, + "loss": 1.0318, + "step": 402 + }, + { + "epoch": 0.39665354330708663, + "grad_norm": 17.95429229736328, + "learning_rate": 3.9662027833001993e-07, + "loss": 1.105, + "step": 403 + }, + { + "epoch": 0.39763779527559057, + "grad_norm": 18.217300415039062, + "learning_rate": 3.9761431411530824e-07, + "loss": 1.009, + "step": 404 + }, + { + "epoch": 0.3986220472440945, + "grad_norm": 13.397478103637695, + "learning_rate": 3.986083499005965e-07, + "loss": 0.5941, + "step": 405 + }, + { + "epoch": 0.39960629921259844, + "grad_norm": 21.784292221069336, + "learning_rate": 3.9960238568588474e-07, + "loss": 1.7972, + "step": 406 + }, + { + "epoch": 0.4005905511811024, + "grad_norm": 21.588985443115234, + "learning_rate": 4.0059642147117305e-07, + "loss": 1.0544, + "step": 407 + }, + { + "epoch": 0.4015748031496063, + "grad_norm": 18.62330436706543, + "learning_rate": 4.015904572564613e-07, + "loss": 1.3912, + "step": 408 + }, + { + "epoch": 0.40255905511811024, + "grad_norm": 14.669224739074707, + "learning_rate": 4.0258449304174955e-07, + "loss": 0.8305, + "step": 409 + }, + { + "epoch": 0.4035433070866142, + "grad_norm": 15.314811706542969, + "learning_rate": 4.035785288270378e-07, + "loss": 0.8688, + "step": 410 + }, + { + "epoch": 0.4045275590551181, + "grad_norm": 17.41103172302246, + "learning_rate": 4.045725646123261e-07, + "loss": 1.0069, + "step": 411 + }, + { + "epoch": 0.40551181102362205, + "grad_norm": 18.243221282958984, + "learning_rate": 4.055666003976144e-07, + "loss": 1.3141, + "step": 412 + }, + { + "epoch": 0.406496062992126, + "grad_norm": 14.436189651489258, + "learning_rate": 4.065606361829026e-07, + "loss": 1.1042, + "step": 413 + }, + { + "epoch": 0.4074803149606299, + "grad_norm": 18.251113891601562, + "learning_rate": 4.075546719681909e-07, + "loss": 1.1011, + "step": 414 + }, + { + "epoch": 0.40846456692913385, + "grad_norm": 18.68360710144043, + "learning_rate": 4.085487077534792e-07, + "loss": 1.1192, + "step": 415 + }, + { + "epoch": 0.4094488188976378, + "grad_norm": 23.807735443115234, + "learning_rate": 4.095427435387675e-07, + "loss": 1.5957, + "step": 416 + }, + { + "epoch": 0.4104330708661417, + "grad_norm": 17.419734954833984, + "learning_rate": 4.1053677932405573e-07, + "loss": 1.164, + "step": 417 + }, + { + "epoch": 0.41141732283464566, + "grad_norm": 13.43248462677002, + "learning_rate": 4.11530815109344e-07, + "loss": 0.6425, + "step": 418 + }, + { + "epoch": 0.4124015748031496, + "grad_norm": 11.816337585449219, + "learning_rate": 4.125248508946323e-07, + "loss": 0.6068, + "step": 419 + }, + { + "epoch": 0.41338582677165353, + "grad_norm": 16.29872703552246, + "learning_rate": 4.1351888667992054e-07, + "loss": 0.9275, + "step": 420 + }, + { + "epoch": 0.41437007874015747, + "grad_norm": 16.47966957092285, + "learning_rate": 4.1451292246520885e-07, + "loss": 0.8836, + "step": 421 + }, + { + "epoch": 0.4153543307086614, + "grad_norm": 18.686519622802734, + "learning_rate": 4.1550695825049705e-07, + "loss": 1.2115, + "step": 422 + }, + { + "epoch": 0.41633858267716534, + "grad_norm": 17.829578399658203, + "learning_rate": 4.1650099403578535e-07, + "loss": 0.8367, + "step": 423 + }, + { + "epoch": 0.41732283464566927, + "grad_norm": 17.083707809448242, + "learning_rate": 4.174950298210736e-07, + "loss": 1.0595, + "step": 424 + }, + { + "epoch": 0.4183070866141732, + "grad_norm": 17.380428314208984, + "learning_rate": 4.184890656063619e-07, + "loss": 0.826, + "step": 425 + }, + { + "epoch": 0.41929133858267714, + "grad_norm": 14.618813514709473, + "learning_rate": 4.194831013916502e-07, + "loss": 0.707, + "step": 426 + }, + { + "epoch": 0.4202755905511811, + "grad_norm": 13.27505111694336, + "learning_rate": 4.204771371769384e-07, + "loss": 0.6235, + "step": 427 + }, + { + "epoch": 0.421259842519685, + "grad_norm": 17.46605682373047, + "learning_rate": 4.214711729622267e-07, + "loss": 0.7719, + "step": 428 + }, + { + "epoch": 0.422244094488189, + "grad_norm": 19.197139739990234, + "learning_rate": 4.2246520874751497e-07, + "loss": 1.0862, + "step": 429 + }, + { + "epoch": 0.42322834645669294, + "grad_norm": 19.14566421508789, + "learning_rate": 4.234592445328033e-07, + "loss": 0.9311, + "step": 430 + }, + { + "epoch": 0.4242125984251969, + "grad_norm": 23.42700958251953, + "learning_rate": 4.244532803180915e-07, + "loss": 1.2339, + "step": 431 + }, + { + "epoch": 0.4251968503937008, + "grad_norm": 19.94841766357422, + "learning_rate": 4.254473161033798e-07, + "loss": 0.9891, + "step": 432 + }, + { + "epoch": 0.42618110236220474, + "grad_norm": 22.30583953857422, + "learning_rate": 4.264413518886681e-07, + "loss": 1.8443, + "step": 433 + }, + { + "epoch": 0.4271653543307087, + "grad_norm": 19.260499954223633, + "learning_rate": 4.2743538767395634e-07, + "loss": 1.1799, + "step": 434 + }, + { + "epoch": 0.4281496062992126, + "grad_norm": 16.3272762298584, + "learning_rate": 4.284294234592446e-07, + "loss": 0.759, + "step": 435 + }, + { + "epoch": 0.42913385826771655, + "grad_norm": 17.50089454650879, + "learning_rate": 4.2942345924453284e-07, + "loss": 1.1002, + "step": 436 + }, + { + "epoch": 0.4301181102362205, + "grad_norm": 14.367104530334473, + "learning_rate": 4.3041749502982115e-07, + "loss": 0.9141, + "step": 437 + }, + { + "epoch": 0.4311023622047244, + "grad_norm": 11.757801055908203, + "learning_rate": 4.3141153081510945e-07, + "loss": 0.5467, + "step": 438 + }, + { + "epoch": 0.43208661417322836, + "grad_norm": 16.255727767944336, + "learning_rate": 4.324055666003977e-07, + "loss": 0.7476, + "step": 439 + }, + { + "epoch": 0.4330708661417323, + "grad_norm": 20.517709732055664, + "learning_rate": 4.3339960238568596e-07, + "loss": 1.14, + "step": 440 + }, + { + "epoch": 0.4340551181102362, + "grad_norm": 18.24642562866211, + "learning_rate": 4.343936381709742e-07, + "loss": 1.1504, + "step": 441 + }, + { + "epoch": 0.43503937007874016, + "grad_norm": 23.892919540405273, + "learning_rate": 4.353876739562625e-07, + "loss": 1.26, + "step": 442 + }, + { + "epoch": 0.4360236220472441, + "grad_norm": 17.60642433166504, + "learning_rate": 4.3638170974155077e-07, + "loss": 1.0311, + "step": 443 + }, + { + "epoch": 0.43700787401574803, + "grad_norm": 18.21918487548828, + "learning_rate": 4.37375745526839e-07, + "loss": 1.0646, + "step": 444 + }, + { + "epoch": 0.43799212598425197, + "grad_norm": 17.054752349853516, + "learning_rate": 4.383697813121273e-07, + "loss": 0.8687, + "step": 445 + }, + { + "epoch": 0.4389763779527559, + "grad_norm": 15.23765754699707, + "learning_rate": 4.393638170974156e-07, + "loss": 0.6839, + "step": 446 + }, + { + "epoch": 0.43996062992125984, + "grad_norm": 17.00933837890625, + "learning_rate": 4.403578528827039e-07, + "loss": 1.1376, + "step": 447 + }, + { + "epoch": 0.4409448818897638, + "grad_norm": 15.645007133483887, + "learning_rate": 4.413518886679921e-07, + "loss": 0.9759, + "step": 448 + }, + { + "epoch": 0.4419291338582677, + "grad_norm": 17.53651237487793, + "learning_rate": 4.423459244532804e-07, + "loss": 0.7971, + "step": 449 + }, + { + "epoch": 0.44291338582677164, + "grad_norm": 19.482398986816406, + "learning_rate": 4.4333996023856864e-07, + "loss": 0.9708, + "step": 450 + }, + { + "epoch": 0.4438976377952756, + "grad_norm": 14.216598510742188, + "learning_rate": 4.4433399602385695e-07, + "loss": 0.8217, + "step": 451 + }, + { + "epoch": 0.4448818897637795, + "grad_norm": 18.865297317504883, + "learning_rate": 4.4532803180914525e-07, + "loss": 1.3728, + "step": 452 + }, + { + "epoch": 0.44586614173228345, + "grad_norm": 16.348478317260742, + "learning_rate": 4.4632206759443345e-07, + "loss": 0.9119, + "step": 453 + }, + { + "epoch": 0.4468503937007874, + "grad_norm": 18.980161666870117, + "learning_rate": 4.4731610337972176e-07, + "loss": 1.012, + "step": 454 + }, + { + "epoch": 0.4478346456692913, + "grad_norm": 20.558147430419922, + "learning_rate": 4.4831013916501e-07, + "loss": 1.3738, + "step": 455 + }, + { + "epoch": 0.44881889763779526, + "grad_norm": 16.78011703491211, + "learning_rate": 4.493041749502983e-07, + "loss": 0.8219, + "step": 456 + }, + { + "epoch": 0.4498031496062992, + "grad_norm": 19.52912712097168, + "learning_rate": 4.502982107355865e-07, + "loss": 1.2558, + "step": 457 + }, + { + "epoch": 0.4507874015748031, + "grad_norm": 15.149005889892578, + "learning_rate": 4.512922465208748e-07, + "loss": 0.6247, + "step": 458 + }, + { + "epoch": 0.45177165354330706, + "grad_norm": 18.364900588989258, + "learning_rate": 4.522862823061631e-07, + "loss": 0.7295, + "step": 459 + }, + { + "epoch": 0.45177165354330706, + "eval_Qnli-dev_cosine_accuracy": 0.693359375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8319265842437744, + "eval_Qnli-dev_cosine_ap": 0.6873625888187367, + "eval_Qnli-dev_cosine_f1": 0.685337726523888, + "eval_Qnli-dev_cosine_f1_threshold": 0.74552983045578, + "eval_Qnli-dev_cosine_precision": 0.5606469002695418, + "eval_Qnli-dev_cosine_recall": 0.8813559322033898, + "eval_Qnli-dev_dot_accuracy": 0.693359375, + "eval_Qnli-dev_dot_accuracy_threshold": 639.0776977539062, + "eval_Qnli-dev_dot_ap": 0.6878718449643791, + "eval_Qnli-dev_dot_f1": 0.685337726523888, + "eval_Qnli-dev_dot_f1_threshold": 572.7136840820312, + "eval_Qnli-dev_dot_precision": 0.5606469002695418, + "eval_Qnli-dev_dot_recall": 0.8813559322033898, + "eval_Qnli-dev_euclidean_accuracy": 0.693359375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.06937026977539, + "eval_Qnli-dev_euclidean_ap": 0.6873686687008952, + "eval_Qnli-dev_euclidean_f1": 0.685337726523888, + "eval_Qnli-dev_euclidean_f1_threshold": 19.772865295410156, + "eval_Qnli-dev_euclidean_precision": 0.5606469002695418, + "eval_Qnli-dev_euclidean_recall": 0.8813559322033898, + "eval_Qnli-dev_manhattan_accuracy": 0.69921875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 362.1485900878906, + "eval_Qnli-dev_manhattan_ap": 0.6874910715870401, + "eval_Qnli-dev_manhattan_f1": 0.6857142857142857, + "eval_Qnli-dev_manhattan_f1_threshold": 430.38519287109375, + "eval_Qnli-dev_manhattan_precision": 0.5682451253481894, + "eval_Qnli-dev_manhattan_recall": 0.864406779661017, + "eval_Qnli-dev_max_accuracy": 0.69921875, + "eval_Qnli-dev_max_accuracy_threshold": 639.0776977539062, + "eval_Qnli-dev_max_ap": 0.6878718449643791, + "eval_Qnli-dev_max_f1": 0.6857142857142857, + "eval_Qnli-dev_max_f1_threshold": 572.7136840820312, + "eval_Qnli-dev_max_precision": 0.5682451253481894, + "eval_Qnli-dev_max_recall": 0.8813559322033898, + "eval_allNLI-dev_cosine_accuracy": 0.68359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9088386297225952, + "eval_allNLI-dev_cosine_ap": 0.48873606015680937, + "eval_allNLI-dev_cosine_f1": 0.5350553505535056, + "eval_allNLI-dev_cosine_f1_threshold": 0.8140230178833008, + "eval_allNLI-dev_cosine_precision": 0.39295392953929537, + "eval_allNLI-dev_cosine_recall": 0.838150289017341, + "eval_allNLI-dev_dot_accuracy": 0.68359375, + "eval_allNLI-dev_dot_accuracy_threshold": 699.0950927734375, + "eval_allNLI-dev_dot_ap": 0.48885724782911755, + "eval_allNLI-dev_dot_f1": 0.5350553505535056, + "eval_allNLI-dev_dot_f1_threshold": 625.3240356445312, + "eval_allNLI-dev_dot_precision": 0.39295392953929537, + "eval_allNLI-dev_dot_recall": 0.838150289017341, + "eval_allNLI-dev_euclidean_accuracy": 0.68359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.834823608398438, + "eval_allNLI-dev_euclidean_ap": 0.4887203371983184, + "eval_allNLI-dev_euclidean_f1": 0.5350553505535056, + "eval_allNLI-dev_euclidean_f1_threshold": 16.90357780456543, + "eval_allNLI-dev_euclidean_precision": 0.39295392953929537, + "eval_allNLI-dev_euclidean_recall": 0.838150289017341, + "eval_allNLI-dev_manhattan_accuracy": 0.68359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 256.45477294921875, + "eval_allNLI-dev_manhattan_ap": 0.4920209563997524, + "eval_allNLI-dev_manhattan_f1": 0.5396145610278372, + "eval_allNLI-dev_manhattan_f1_threshold": 339.225830078125, + "eval_allNLI-dev_manhattan_precision": 0.42857142857142855, + "eval_allNLI-dev_manhattan_recall": 0.7283236994219653, + "eval_allNLI-dev_max_accuracy": 0.68359375, + "eval_allNLI-dev_max_accuracy_threshold": 699.0950927734375, + "eval_allNLI-dev_max_ap": 0.4920209563997524, + "eval_allNLI-dev_max_f1": 0.5396145610278372, + "eval_allNLI-dev_max_f1_threshold": 625.3240356445312, + "eval_allNLI-dev_max_precision": 0.42857142857142855, + "eval_allNLI-dev_max_recall": 0.838150289017341, + "eval_runtime": 1.2229, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.6878718449643791, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.5377382226514003, + "eval_sts-test_pearson_dot": 0.5376496659087263, + "eval_sts-test_pearson_euclidean": 0.5469897917607747, + "eval_sts-test_pearson_manhattan": 0.5464293120330461, + "eval_sts-test_pearson_max": 0.5469897917607747, + "eval_sts-test_spearman_cosine": 0.5410237309359288, + "eval_sts-test_spearman_dot": 0.5408811086658744, + "eval_sts-test_spearman_euclidean": 0.5409800984560722, + "eval_sts-test_spearman_manhattan": 0.5401021234588343, + "eval_sts-test_spearman_max": 0.5410237309359288, + "step": 459 + }, + { + "epoch": 0.452755905511811, + "grad_norm": 17.561704635620117, + "learning_rate": 4.532803180914514e-07, + "loss": 0.8154, + "step": 460 + }, + { + "epoch": 0.453740157480315, + "grad_norm": 21.252965927124023, + "learning_rate": 4.542743538767397e-07, + "loss": 1.1392, + "step": 461 + }, + { + "epoch": 0.4547244094488189, + "grad_norm": 18.453880310058594, + "learning_rate": 4.552683896620279e-07, + "loss": 0.8618, + "step": 462 + }, + { + "epoch": 0.45570866141732286, + "grad_norm": 17.269075393676758, + "learning_rate": 4.562624254473162e-07, + "loss": 0.9669, + "step": 463 + }, + { + "epoch": 0.4566929133858268, + "grad_norm": 20.30898094177246, + "learning_rate": 4.5725646123260444e-07, + "loss": 0.8804, + "step": 464 + }, + { + "epoch": 0.45767716535433073, + "grad_norm": 17.70629119873047, + "learning_rate": 4.5825049701789274e-07, + "loss": 0.8479, + "step": 465 + }, + { + "epoch": 0.45866141732283466, + "grad_norm": 17.85552215576172, + "learning_rate": 4.59244532803181e-07, + "loss": 0.6296, + "step": 466 + }, + { + "epoch": 0.4596456692913386, + "grad_norm": 20.20493507385254, + "learning_rate": 4.6023856858846925e-07, + "loss": 0.8449, + "step": 467 + }, + { + "epoch": 0.46062992125984253, + "grad_norm": 17.319948196411133, + "learning_rate": 4.6123260437375755e-07, + "loss": 0.9772, + "step": 468 + }, + { + "epoch": 0.46161417322834647, + "grad_norm": 14.453144073486328, + "learning_rate": 4.622266401590458e-07, + "loss": 0.6424, + "step": 469 + }, + { + "epoch": 0.4625984251968504, + "grad_norm": 19.411174774169922, + "learning_rate": 4.6322067594433406e-07, + "loss": 0.9169, + "step": 470 + }, + { + "epoch": 0.46358267716535434, + "grad_norm": 15.519804954528809, + "learning_rate": 4.642147117296223e-07, + "loss": 0.7599, + "step": 471 + }, + { + "epoch": 0.4645669291338583, + "grad_norm": 18.444625854492188, + "learning_rate": 4.652087475149106e-07, + "loss": 0.8943, + "step": 472 + }, + { + "epoch": 0.4655511811023622, + "grad_norm": 15.110048294067383, + "learning_rate": 4.662027833001989e-07, + "loss": 0.9475, + "step": 473 + }, + { + "epoch": 0.46653543307086615, + "grad_norm": 21.10722541809082, + "learning_rate": 4.671968190854872e-07, + "loss": 1.4518, + "step": 474 + }, + { + "epoch": 0.4675196850393701, + "grad_norm": 20.026628494262695, + "learning_rate": 4.681908548707754e-07, + "loss": 1.274, + "step": 475 + }, + { + "epoch": 0.468503937007874, + "grad_norm": 14.985383987426758, + "learning_rate": 4.691848906560637e-07, + "loss": 0.7306, + "step": 476 + }, + { + "epoch": 0.46948818897637795, + "grad_norm": 17.787303924560547, + "learning_rate": 4.70178926441352e-07, + "loss": 0.9238, + "step": 477 + }, + { + "epoch": 0.4704724409448819, + "grad_norm": 12.555164337158203, + "learning_rate": 4.7117296222664024e-07, + "loss": 0.6593, + "step": 478 + }, + { + "epoch": 0.4714566929133858, + "grad_norm": 17.015419006347656, + "learning_rate": 4.721669980119285e-07, + "loss": 1.0183, + "step": 479 + }, + { + "epoch": 0.47244094488188976, + "grad_norm": 18.000446319580078, + "learning_rate": 4.731610337972168e-07, + "loss": 1.2577, + "step": 480 + }, + { + "epoch": 0.4734251968503937, + "grad_norm": 17.278921127319336, + "learning_rate": 4.7415506958250505e-07, + "loss": 0.8738, + "step": 481 + }, + { + "epoch": 0.4744094488188976, + "grad_norm": 17.930349349975586, + "learning_rate": 4.7514910536779335e-07, + "loss": 1.1416, + "step": 482 + }, + { + "epoch": 0.47539370078740156, + "grad_norm": 13.973302841186523, + "learning_rate": 4.7614314115308155e-07, + "loss": 0.7135, + "step": 483 + }, + { + "epoch": 0.4763779527559055, + "grad_norm": 18.695253372192383, + "learning_rate": 4.771371769383698e-07, + "loss": 1.2587, + "step": 484 + }, + { + "epoch": 0.47736220472440943, + "grad_norm": 13.285400390625, + "learning_rate": 4.781312127236581e-07, + "loss": 0.8823, + "step": 485 + }, + { + "epoch": 0.47834645669291337, + "grad_norm": 15.625781059265137, + "learning_rate": 4.791252485089464e-07, + "loss": 0.8423, + "step": 486 + }, + { + "epoch": 0.4793307086614173, + "grad_norm": 15.811039924621582, + "learning_rate": 4.801192842942347e-07, + "loss": 0.7704, + "step": 487 + }, + { + "epoch": 0.48031496062992124, + "grad_norm": 17.256977081298828, + "learning_rate": 4.811133200795229e-07, + "loss": 0.7049, + "step": 488 + }, + { + "epoch": 0.4812992125984252, + "grad_norm": 20.418750762939453, + "learning_rate": 4.821073558648112e-07, + "loss": 1.1893, + "step": 489 + }, + { + "epoch": 0.4822834645669291, + "grad_norm": 21.882749557495117, + "learning_rate": 4.831013916500995e-07, + "loss": 1.3985, + "step": 490 + }, + { + "epoch": 0.48326771653543305, + "grad_norm": 21.558752059936523, + "learning_rate": 4.840954274353877e-07, + "loss": 1.3567, + "step": 491 + }, + { + "epoch": 0.484251968503937, + "grad_norm": 23.495107650756836, + "learning_rate": 4.85089463220676e-07, + "loss": 1.2573, + "step": 492 + }, + { + "epoch": 0.48523622047244097, + "grad_norm": 14.981607437133789, + "learning_rate": 4.860834990059642e-07, + "loss": 0.7671, + "step": 493 + }, + { + "epoch": 0.4862204724409449, + "grad_norm": 17.21627426147461, + "learning_rate": 4.870775347912525e-07, + "loss": 0.5425, + "step": 494 + }, + { + "epoch": 0.48720472440944884, + "grad_norm": 20.916128158569336, + "learning_rate": 4.880715705765408e-07, + "loss": 0.9372, + "step": 495 + }, + { + "epoch": 0.4881889763779528, + "grad_norm": 15.274470329284668, + "learning_rate": 4.890656063618291e-07, + "loss": 0.799, + "step": 496 + }, + { + "epoch": 0.4891732283464567, + "grad_norm": 18.584993362426758, + "learning_rate": 4.900596421471173e-07, + "loss": 0.9548, + "step": 497 + }, + { + "epoch": 0.49015748031496065, + "grad_norm": 18.775136947631836, + "learning_rate": 4.910536779324057e-07, + "loss": 1.0855, + "step": 498 + }, + { + "epoch": 0.4911417322834646, + "grad_norm": 24.180419921875, + "learning_rate": 4.92047713717694e-07, + "loss": 1.0465, + "step": 499 + }, + { + "epoch": 0.4921259842519685, + "grad_norm": 18.497556686401367, + "learning_rate": 4.930417495029823e-07, + "loss": 1.1004, + "step": 500 + }, + { + "epoch": 0.49311023622047245, + "grad_norm": 14.301048278808594, + "learning_rate": 4.940357852882705e-07, + "loss": 0.6392, + "step": 501 + }, + { + "epoch": 0.4940944881889764, + "grad_norm": 14.273008346557617, + "learning_rate": 4.950298210735588e-07, + "loss": 0.7102, + "step": 502 + }, + { + "epoch": 0.4950787401574803, + "grad_norm": 18.05272102355957, + "learning_rate": 4.96023856858847e-07, + "loss": 1.3242, + "step": 503 + }, + { + "epoch": 0.49606299212598426, + "grad_norm": 13.768959045410156, + "learning_rate": 4.970178926441353e-07, + "loss": 0.6861, + "step": 504 + }, + { + "epoch": 0.4970472440944882, + "grad_norm": 14.028704643249512, + "learning_rate": 4.980119284294235e-07, + "loss": 0.9291, + "step": 505 + }, + { + "epoch": 0.49803149606299213, + "grad_norm": 14.870085716247559, + "learning_rate": 4.990059642147118e-07, + "loss": 0.8592, + "step": 506 + }, + { + "epoch": 0.49901574803149606, + "grad_norm": 15.56464672088623, + "learning_rate": 5.000000000000001e-07, + "loss": 0.9462, + "step": 507 + }, + { + "epoch": 0.5, + "grad_norm": 15.741423606872559, + "learning_rate": 5.009940357852883e-07, + "loss": 1.0167, + "step": 508 + }, + { + "epoch": 0.5009842519685039, + "grad_norm": 18.02831268310547, + "learning_rate": 5.019880715705767e-07, + "loss": 1.0118, + "step": 509 + }, + { + "epoch": 0.5019685039370079, + "grad_norm": 15.293427467346191, + "learning_rate": 5.029821073558649e-07, + "loss": 0.6741, + "step": 510 + }, + { + "epoch": 0.5029527559055118, + "grad_norm": 19.62667465209961, + "learning_rate": 5.039761431411532e-07, + "loss": 1.4578, + "step": 511 + }, + { + "epoch": 0.5039370078740157, + "grad_norm": 20.828519821166992, + "learning_rate": 5.049701789264414e-07, + "loss": 1.2959, + "step": 512 + }, + { + "epoch": 0.5049212598425197, + "grad_norm": 14.50117015838623, + "learning_rate": 5.059642147117297e-07, + "loss": 0.8533, + "step": 513 + }, + { + "epoch": 0.5059055118110236, + "grad_norm": 15.186738967895508, + "learning_rate": 5.06958250497018e-07, + "loss": 0.6685, + "step": 514 + }, + { + "epoch": 0.5068897637795275, + "grad_norm": 17.811168670654297, + "learning_rate": 5.079522862823062e-07, + "loss": 1.1556, + "step": 515 + }, + { + "epoch": 0.5078740157480315, + "grad_norm": 17.15772247314453, + "learning_rate": 5.089463220675945e-07, + "loss": 0.8177, + "step": 516 + }, + { + "epoch": 0.5088582677165354, + "grad_norm": 12.814756393432617, + "learning_rate": 5.099403578528828e-07, + "loss": 0.6296, + "step": 517 + }, + { + "epoch": 0.5098425196850394, + "grad_norm": 14.35216999053955, + "learning_rate": 5.10934393638171e-07, + "loss": 0.8407, + "step": 518 + }, + { + "epoch": 0.5108267716535433, + "grad_norm": 14.875484466552734, + "learning_rate": 5.119284294234593e-07, + "loss": 0.6987, + "step": 519 + }, + { + "epoch": 0.5118110236220472, + "grad_norm": 15.191465377807617, + "learning_rate": 5.129224652087476e-07, + "loss": 0.9888, + "step": 520 + }, + { + "epoch": 0.5127952755905512, + "grad_norm": 15.37837028503418, + "learning_rate": 5.139165009940358e-07, + "loss": 0.8938, + "step": 521 + }, + { + "epoch": 0.5137795275590551, + "grad_norm": 13.611766815185547, + "learning_rate": 5.149105367793241e-07, + "loss": 0.582, + "step": 522 + }, + { + "epoch": 0.514763779527559, + "grad_norm": 17.264280319213867, + "learning_rate": 5.159045725646124e-07, + "loss": 0.6596, + "step": 523 + }, + { + "epoch": 0.515748031496063, + "grad_norm": 15.050413131713867, + "learning_rate": 5.168986083499007e-07, + "loss": 0.6029, + "step": 524 + }, + { + "epoch": 0.5167322834645669, + "grad_norm": 19.290075302124023, + "learning_rate": 5.178926441351889e-07, + "loss": 0.9806, + "step": 525 + }, + { + "epoch": 0.5177165354330708, + "grad_norm": 15.048296928405762, + "learning_rate": 5.188866799204771e-07, + "loss": 0.9463, + "step": 526 + }, + { + "epoch": 0.5187007874015748, + "grad_norm": 15.788368225097656, + "learning_rate": 5.198807157057656e-07, + "loss": 0.7088, + "step": 527 + }, + { + "epoch": 0.5196850393700787, + "grad_norm": 15.568428039550781, + "learning_rate": 5.208747514910538e-07, + "loss": 0.7525, + "step": 528 + }, + { + "epoch": 0.5206692913385826, + "grad_norm": 13.9373197555542, + "learning_rate": 5.21868787276342e-07, + "loss": 0.7625, + "step": 529 + }, + { + "epoch": 0.5216535433070866, + "grad_norm": 16.379072189331055, + "learning_rate": 5.228628230616304e-07, + "loss": 0.8271, + "step": 530 + }, + { + "epoch": 0.5226377952755905, + "grad_norm": 17.776195526123047, + "learning_rate": 5.238568588469186e-07, + "loss": 0.6129, + "step": 531 + }, + { + "epoch": 0.5236220472440944, + "grad_norm": 24.5928897857666, + "learning_rate": 5.248508946322069e-07, + "loss": 1.1563, + "step": 532 + }, + { + "epoch": 0.5246062992125984, + "grad_norm": 17.05551528930664, + "learning_rate": 5.258449304174951e-07, + "loss": 0.8131, + "step": 533 + }, + { + "epoch": 0.5255905511811023, + "grad_norm": 14.337945938110352, + "learning_rate": 5.268389662027834e-07, + "loss": 0.5363, + "step": 534 + }, + { + "epoch": 0.5265748031496063, + "grad_norm": 20.1900634765625, + "learning_rate": 5.278330019880717e-07, + "loss": 0.8819, + "step": 535 + }, + { + "epoch": 0.5275590551181102, + "grad_norm": 24.67251205444336, + "learning_rate": 5.288270377733599e-07, + "loss": 0.9772, + "step": 536 + }, + { + "epoch": 0.5285433070866141, + "grad_norm": 22.329355239868164, + "learning_rate": 5.298210735586483e-07, + "loss": 1.2102, + "step": 537 + }, + { + "epoch": 0.5295275590551181, + "grad_norm": 22.19415855407715, + "learning_rate": 5.308151093439365e-07, + "loss": 1.1234, + "step": 538 + }, + { + "epoch": 0.530511811023622, + "grad_norm": 23.826335906982422, + "learning_rate": 5.318091451292247e-07, + "loss": 1.1857, + "step": 539 + }, + { + "epoch": 0.531496062992126, + "grad_norm": 16.586576461791992, + "learning_rate": 5.32803180914513e-07, + "loss": 0.7873, + "step": 540 + }, + { + "epoch": 0.53248031496063, + "grad_norm": 12.48427677154541, + "learning_rate": 5.337972166998013e-07, + "loss": 0.5034, + "step": 541 + }, + { + "epoch": 0.5334645669291339, + "grad_norm": 20.41586685180664, + "learning_rate": 5.347912524850895e-07, + "loss": 1.3305, + "step": 542 + }, + { + "epoch": 0.5344488188976378, + "grad_norm": 22.552961349487305, + "learning_rate": 5.357852882703778e-07, + "loss": 1.1727, + "step": 543 + }, + { + "epoch": 0.5354330708661418, + "grad_norm": 21.429880142211914, + "learning_rate": 5.367793240556661e-07, + "loss": 1.2825, + "step": 544 + }, + { + "epoch": 0.5364173228346457, + "grad_norm": 19.160961151123047, + "learning_rate": 5.377733598409544e-07, + "loss": 1.0446, + "step": 545 + }, + { + "epoch": 0.5374015748031497, + "grad_norm": 17.107141494750977, + "learning_rate": 5.387673956262426e-07, + "loss": 0.9838, + "step": 546 + }, + { + "epoch": 0.5383858267716536, + "grad_norm": 18.039453506469727, + "learning_rate": 5.397614314115308e-07, + "loss": 1.2194, + "step": 547 + }, + { + "epoch": 0.5393700787401575, + "grad_norm": 13.434783935546875, + "learning_rate": 5.407554671968192e-07, + "loss": 0.7709, + "step": 548 + }, + { + "epoch": 0.5403543307086615, + "grad_norm": 14.461050987243652, + "learning_rate": 5.417495029821074e-07, + "loss": 0.748, + "step": 549 + }, + { + "epoch": 0.5413385826771654, + "grad_norm": 18.0071964263916, + "learning_rate": 5.427435387673957e-07, + "loss": 1.0948, + "step": 550 + }, + { + "epoch": 0.5423228346456693, + "grad_norm": 15.916511535644531, + "learning_rate": 5.43737574552684e-07, + "loss": 0.915, + "step": 551 + }, + { + "epoch": 0.5433070866141733, + "grad_norm": 22.280460357666016, + "learning_rate": 5.447316103379722e-07, + "loss": 1.537, + "step": 552 + }, + { + "epoch": 0.5442913385826772, + "grad_norm": 9.682397842407227, + "learning_rate": 5.457256461232605e-07, + "loss": 0.3239, + "step": 553 + }, + { + "epoch": 0.5452755905511811, + "grad_norm": 14.292939186096191, + "learning_rate": 5.467196819085487e-07, + "loss": 0.9592, + "step": 554 + }, + { + "epoch": 0.5462598425196851, + "grad_norm": 22.54702377319336, + "learning_rate": 5.477137176938371e-07, + "loss": 0.7737, + "step": 555 + }, + { + "epoch": 0.547244094488189, + "grad_norm": 14.62065315246582, + "learning_rate": 5.487077534791253e-07, + "loss": 0.613, + "step": 556 + }, + { + "epoch": 0.5482283464566929, + "grad_norm": 22.537317276000977, + "learning_rate": 5.497017892644135e-07, + "loss": 1.3646, + "step": 557 + }, + { + "epoch": 0.5492125984251969, + "grad_norm": 15.06298542022705, + "learning_rate": 5.50695825049702e-07, + "loss": 0.6659, + "step": 558 + }, + { + "epoch": 0.5501968503937008, + "grad_norm": 13.530902862548828, + "learning_rate": 5.516898608349902e-07, + "loss": 0.5207, + "step": 559 + }, + { + "epoch": 0.5511811023622047, + "grad_norm": 19.478429794311523, + "learning_rate": 5.526838966202784e-07, + "loss": 0.9467, + "step": 560 + }, + { + "epoch": 0.5521653543307087, + "grad_norm": 10.947720527648926, + "learning_rate": 5.536779324055667e-07, + "loss": 0.5692, + "step": 561 + }, + { + "epoch": 0.5531496062992126, + "grad_norm": 33.76875686645508, + "learning_rate": 5.54671968190855e-07, + "loss": 1.5855, + "step": 562 + }, + { + "epoch": 0.5541338582677166, + "grad_norm": 18.264493942260742, + "learning_rate": 5.556660039761433e-07, + "loss": 0.8855, + "step": 563 + }, + { + "epoch": 0.5551181102362205, + "grad_norm": 25.19763946533203, + "learning_rate": 5.566600397614315e-07, + "loss": 1.1829, + "step": 564 + }, + { + "epoch": 0.5561023622047244, + "grad_norm": 22.214569091796875, + "learning_rate": 5.576540755467198e-07, + "loss": 0.978, + "step": 565 + }, + { + "epoch": 0.5570866141732284, + "grad_norm": 15.043540000915527, + "learning_rate": 5.586481113320081e-07, + "loss": 1.1818, + "step": 566 + }, + { + "epoch": 0.5580708661417323, + "grad_norm": 15.76466178894043, + "learning_rate": 5.596421471172963e-07, + "loss": 0.701, + "step": 567 + }, + { + "epoch": 0.5590551181102362, + "grad_norm": 16.75844955444336, + "learning_rate": 5.606361829025846e-07, + "loss": 1.0226, + "step": 568 + }, + { + "epoch": 0.5600393700787402, + "grad_norm": 10.848587989807129, + "learning_rate": 5.616302186878729e-07, + "loss": 0.5937, + "step": 569 + }, + { + "epoch": 0.5610236220472441, + "grad_norm": 15.039306640625, + "learning_rate": 5.626242544731611e-07, + "loss": 0.8095, + "step": 570 + }, + { + "epoch": 0.562007874015748, + "grad_norm": 17.843490600585938, + "learning_rate": 5.636182902584494e-07, + "loss": 1.174, + "step": 571 + }, + { + "epoch": 0.562992125984252, + "grad_norm": 15.040814399719238, + "learning_rate": 5.646123260437377e-07, + "loss": 0.96, + "step": 572 + }, + { + "epoch": 0.5639763779527559, + "grad_norm": 17.238296508789062, + "learning_rate": 5.656063618290259e-07, + "loss": 0.8339, + "step": 573 + }, + { + "epoch": 0.5649606299212598, + "grad_norm": 13.707849502563477, + "learning_rate": 5.666003976143142e-07, + "loss": 0.717, + "step": 574 + }, + { + "epoch": 0.5659448818897638, + "grad_norm": 14.237223625183105, + "learning_rate": 5.675944333996024e-07, + "loss": 0.5938, + "step": 575 + }, + { + "epoch": 0.5669291338582677, + "grad_norm": 12.927653312683105, + "learning_rate": 5.685884691848908e-07, + "loss": 0.6501, + "step": 576 + }, + { + "epoch": 0.5679133858267716, + "grad_norm": 15.383443832397461, + "learning_rate": 5.69582504970179e-07, + "loss": 0.7003, + "step": 577 + }, + { + "epoch": 0.5688976377952756, + "grad_norm": 14.056544303894043, + "learning_rate": 5.705765407554672e-07, + "loss": 0.5525, + "step": 578 + }, + { + "epoch": 0.5698818897637795, + "grad_norm": 17.352069854736328, + "learning_rate": 5.715705765407556e-07, + "loss": 0.7003, + "step": 579 + }, + { + "epoch": 0.5708661417322834, + "grad_norm": 22.5910587310791, + "learning_rate": 5.725646123260438e-07, + "loss": 1.059, + "step": 580 + }, + { + "epoch": 0.5718503937007874, + "grad_norm": 19.49325942993164, + "learning_rate": 5.735586481113321e-07, + "loss": 0.8625, + "step": 581 + }, + { + "epoch": 0.5728346456692913, + "grad_norm": 14.92470645904541, + "learning_rate": 5.745526838966203e-07, + "loss": 0.5862, + "step": 582 + }, + { + "epoch": 0.5738188976377953, + "grad_norm": 16.41839027404785, + "learning_rate": 5.755467196819086e-07, + "loss": 0.9162, + "step": 583 + }, + { + "epoch": 0.5748031496062992, + "grad_norm": 20.332698822021484, + "learning_rate": 5.765407554671969e-07, + "loss": 0.926, + "step": 584 + }, + { + "epoch": 0.5757874015748031, + "grad_norm": 22.728179931640625, + "learning_rate": 5.775347912524851e-07, + "loss": 1.2729, + "step": 585 + }, + { + "epoch": 0.5767716535433071, + "grad_norm": 16.058364868164062, + "learning_rate": 5.785288270377734e-07, + "loss": 0.8935, + "step": 586 + }, + { + "epoch": 0.577755905511811, + "grad_norm": 14.835709571838379, + "learning_rate": 5.795228628230618e-07, + "loss": 0.541, + "step": 587 + }, + { + "epoch": 0.5787401574803149, + "grad_norm": 19.8588809967041, + "learning_rate": 5.8051689860835e-07, + "loss": 1.1455, + "step": 588 + }, + { + "epoch": 0.5797244094488189, + "grad_norm": 14.425910949707031, + "learning_rate": 5.815109343936383e-07, + "loss": 0.7306, + "step": 589 + }, + { + "epoch": 0.5807086614173228, + "grad_norm": 17.056396484375, + "learning_rate": 5.825049701789266e-07, + "loss": 0.9088, + "step": 590 + }, + { + "epoch": 0.5816929133858267, + "grad_norm": 14.791122436523438, + "learning_rate": 5.834990059642148e-07, + "loss": 0.9166, + "step": 591 + }, + { + "epoch": 0.5826771653543307, + "grad_norm": 14.80226993560791, + "learning_rate": 5.844930417495031e-07, + "loss": 0.8679, + "step": 592 + }, + { + "epoch": 0.5836614173228346, + "grad_norm": 17.1209716796875, + "learning_rate": 5.854870775347914e-07, + "loss": 0.9329, + "step": 593 + }, + { + "epoch": 0.5846456692913385, + "grad_norm": 17.293729782104492, + "learning_rate": 5.864811133200797e-07, + "loss": 1.1201, + "step": 594 + }, + { + "epoch": 0.5856299212598425, + "grad_norm": 15.181586265563965, + "learning_rate": 5.874751491053679e-07, + "loss": 0.6418, + "step": 595 + }, + { + "epoch": 0.5866141732283464, + "grad_norm": 18.143863677978516, + "learning_rate": 5.884691848906561e-07, + "loss": 1.145, + "step": 596 + }, + { + "epoch": 0.5875984251968503, + "grad_norm": 19.976341247558594, + "learning_rate": 5.894632206759445e-07, + "loss": 1.4041, + "step": 597 + }, + { + "epoch": 0.5885826771653543, + "grad_norm": 13.67269515991211, + "learning_rate": 5.904572564612327e-07, + "loss": 0.6954, + "step": 598 + }, + { + "epoch": 0.5895669291338582, + "grad_norm": 14.636340141296387, + "learning_rate": 5.914512922465209e-07, + "loss": 0.4567, + "step": 599 + }, + { + "epoch": 0.5905511811023622, + "grad_norm": 16.68194580078125, + "learning_rate": 5.924453280318093e-07, + "loss": 1.1305, + "step": 600 + }, + { + "epoch": 0.5915354330708661, + "grad_norm": 17.648611068725586, + "learning_rate": 5.934393638170975e-07, + "loss": 0.8077, + "step": 601 + }, + { + "epoch": 0.59251968503937, + "grad_norm": 14.298154830932617, + "learning_rate": 5.944333996023858e-07, + "loss": 0.6143, + "step": 602 + }, + { + "epoch": 0.593503937007874, + "grad_norm": 19.324697494506836, + "learning_rate": 5.95427435387674e-07, + "loss": 1.3139, + "step": 603 + }, + { + "epoch": 0.594488188976378, + "grad_norm": 14.576811790466309, + "learning_rate": 5.964214711729623e-07, + "loss": 0.7694, + "step": 604 + }, + { + "epoch": 0.5954724409448819, + "grad_norm": 15.335044860839844, + "learning_rate": 5.974155069582506e-07, + "loss": 0.9622, + "step": 605 + }, + { + "epoch": 0.5964566929133859, + "grad_norm": 16.537038803100586, + "learning_rate": 5.984095427435388e-07, + "loss": 0.91, + "step": 606 + }, + { + "epoch": 0.5974409448818898, + "grad_norm": 21.826786041259766, + "learning_rate": 5.994035785288272e-07, + "loss": 1.3125, + "step": 607 + }, + { + "epoch": 0.5984251968503937, + "grad_norm": 13.646114349365234, + "learning_rate": 6.003976143141154e-07, + "loss": 1.0153, + "step": 608 + }, + { + "epoch": 0.5994094488188977, + "grad_norm": 11.858122825622559, + "learning_rate": 6.013916500994036e-07, + "loss": 0.8468, + "step": 609 + }, + { + "epoch": 0.6003937007874016, + "grad_norm": 16.532007217407227, + "learning_rate": 6.02385685884692e-07, + "loss": 1.1026, + "step": 610 + }, + { + "epoch": 0.6013779527559056, + "grad_norm": 13.020824432373047, + "learning_rate": 6.033797216699802e-07, + "loss": 0.8291, + "step": 611 + }, + { + "epoch": 0.6023622047244095, + "grad_norm": 13.50063705444336, + "learning_rate": 6.043737574552684e-07, + "loss": 0.7235, + "step": 612 + }, + { + "epoch": 0.6023622047244095, + "eval_Qnli-dev_cosine_accuracy": 0.66015625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8440727591514587, + "eval_Qnli-dev_cosine_ap": 0.6897570972593343, + "eval_Qnli-dev_cosine_f1": 0.6816720257234726, + "eval_Qnli-dev_cosine_f1_threshold": 0.786305844783783, + "eval_Qnli-dev_cosine_precision": 0.5492227979274611, + "eval_Qnli-dev_cosine_recall": 0.8983050847457628, + "eval_Qnli-dev_dot_accuracy": 0.66015625, + "eval_Qnli-dev_dot_accuracy_threshold": 648.5560913085938, + "eval_Qnli-dev_dot_ap": 0.6896647379697238, + "eval_Qnli-dev_dot_f1": 0.680577849117175, + "eval_Qnli-dev_dot_f1_threshold": 604.0833129882812, + "eval_Qnli-dev_dot_precision": 0.5478036175710594, + "eval_Qnli-dev_dot_recall": 0.8983050847457628, + "eval_Qnli-dev_euclidean_accuracy": 0.66015625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.479604721069336, + "eval_Qnli-dev_euclidean_ap": 0.6897476247610781, + "eval_Qnli-dev_euclidean_f1": 0.6816720257234726, + "eval_Qnli-dev_euclidean_f1_threshold": 18.121519088745117, + "eval_Qnli-dev_euclidean_precision": 0.5492227979274611, + "eval_Qnli-dev_euclidean_recall": 0.8983050847457628, + "eval_Qnli-dev_manhattan_accuracy": 0.662109375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 319.4159851074219, + "eval_Qnli-dev_manhattan_ap": 0.6900139538809252, + "eval_Qnli-dev_manhattan_f1": 0.6774193548387097, + "eval_Qnli-dev_manhattan_f1_threshold": 392.64697265625, + "eval_Qnli-dev_manhattan_precision": 0.546875, + "eval_Qnli-dev_manhattan_recall": 0.8898305084745762, + "eval_Qnli-dev_max_accuracy": 0.662109375, + "eval_Qnli-dev_max_accuracy_threshold": 648.5560913085938, + "eval_Qnli-dev_max_ap": 0.6900139538809252, + "eval_Qnli-dev_max_f1": 0.6816720257234726, + "eval_Qnli-dev_max_f1_threshold": 604.0833129882812, + "eval_Qnli-dev_max_precision": 0.5492227979274611, + "eval_Qnli-dev_max_recall": 0.8983050847457628, + "eval_allNLI-dev_cosine_accuracy": 0.68359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9587128162384033, + "eval_allNLI-dev_cosine_ap": 0.5125323722619199, + "eval_allNLI-dev_cosine_f1": 0.5523809523809524, + "eval_allNLI-dev_cosine_f1_threshold": 0.8827621340751648, + "eval_allNLI-dev_cosine_precision": 0.46963562753036436, + "eval_allNLI-dev_cosine_recall": 0.6705202312138728, + "eval_allNLI-dev_dot_accuracy": 0.68359375, + "eval_allNLI-dev_dot_accuracy_threshold": 736.6729736328125, + "eval_allNLI-dev_dot_ap": 0.5127016128181208, + "eval_allNLI-dev_dot_f1": 0.5523809523809524, + "eval_allNLI-dev_dot_f1_threshold": 678.336181640625, + "eval_allNLI-dev_dot_precision": 0.46963562753036436, + "eval_allNLI-dev_dot_recall": 0.6705202312138728, + "eval_allNLI-dev_euclidean_accuracy": 0.68359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 7.965542793273926, + "eval_allNLI-dev_euclidean_ap": 0.5125323722619199, + "eval_allNLI-dev_euclidean_f1": 0.5523809523809524, + "eval_allNLI-dev_euclidean_f1_threshold": 13.422992706298828, + "eval_allNLI-dev_euclidean_precision": 0.46963562753036436, + "eval_allNLI-dev_euclidean_recall": 0.6705202312138728, + "eval_allNLI-dev_manhattan_accuracy": 0.685546875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 174.88226318359375, + "eval_allNLI-dev_manhattan_ap": 0.5174962030025715, + "eval_allNLI-dev_manhattan_f1": 0.549618320610687, + "eval_allNLI-dev_manhattan_f1_threshold": 282.00616455078125, + "eval_allNLI-dev_manhattan_precision": 0.4909090909090909, + "eval_allNLI-dev_manhattan_recall": 0.6242774566473989, + "eval_allNLI-dev_max_accuracy": 0.685546875, + "eval_allNLI-dev_max_accuracy_threshold": 736.6729736328125, + "eval_allNLI-dev_max_ap": 0.5174962030025715, + "eval_allNLI-dev_max_f1": 0.5523809523809524, + "eval_allNLI-dev_max_f1_threshold": 678.336181640625, + "eval_allNLI-dev_max_precision": 0.4909090909090909, + "eval_allNLI-dev_max_recall": 0.6705202312138728, + "eval_runtime": 1.2365, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.6900139538809252, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.5501902426288274, + "eval_sts-test_pearson_dot": 0.5500321508853105, + "eval_sts-test_pearson_euclidean": 0.5690634649114448, + "eval_sts-test_pearson_manhattan": 0.5686044657697955, + "eval_sts-test_pearson_max": 0.5690634649114448, + "eval_sts-test_spearman_cosine": 0.5679896775175899, + "eval_sts-test_spearman_dot": 0.5679785792047649, + "eval_sts-test_spearman_euclidean": 0.5679722245256472, + "eval_sts-test_spearman_manhattan": 0.567807674137511, + "eval_sts-test_spearman_max": 0.5679896775175899, + "step": 612 + }, + { + "epoch": 0.6033464566929134, + "grad_norm": 15.621910095214844, + "learning_rate": 6.053677932405567e-07, + "loss": 0.9613, + "step": 613 + }, + { + "epoch": 0.6043307086614174, + "grad_norm": 14.314550399780273, + "learning_rate": 6.06361829025845e-07, + "loss": 0.7124, + "step": 614 + }, + { + "epoch": 0.6053149606299213, + "grad_norm": 17.164976119995117, + "learning_rate": 6.073558648111333e-07, + "loss": 1.0719, + "step": 615 + }, + { + "epoch": 0.6062992125984252, + "grad_norm": 13.165125846862793, + "learning_rate": 6.083499005964215e-07, + "loss": 0.7233, + "step": 616 + }, + { + "epoch": 0.6072834645669292, + "grad_norm": 24.77128791809082, + "learning_rate": 6.093439363817097e-07, + "loss": 1.6863, + "step": 617 + }, + { + "epoch": 0.6082677165354331, + "grad_norm": 17.56781578063965, + "learning_rate": 6.103379721669982e-07, + "loss": 0.8665, + "step": 618 + }, + { + "epoch": 0.609251968503937, + "grad_norm": 24.854677200317383, + "learning_rate": 6.113320079522864e-07, + "loss": 1.6432, + "step": 619 + }, + { + "epoch": 0.610236220472441, + "grad_norm": 15.149316787719727, + "learning_rate": 6.123260437375747e-07, + "loss": 0.771, + "step": 620 + }, + { + "epoch": 0.6112204724409449, + "grad_norm": 11.395992279052734, + "learning_rate": 6.13320079522863e-07, + "loss": 0.6755, + "step": 621 + }, + { + "epoch": 0.6122047244094488, + "grad_norm": 16.920135498046875, + "learning_rate": 6.143141153081512e-07, + "loss": 0.6809, + "step": 622 + }, + { + "epoch": 0.6131889763779528, + "grad_norm": 12.814380645751953, + "learning_rate": 6.153081510934395e-07, + "loss": 0.5626, + "step": 623 + }, + { + "epoch": 0.6141732283464567, + "grad_norm": 16.69883155822754, + "learning_rate": 6.163021868787277e-07, + "loss": 0.6287, + "step": 624 + }, + { + "epoch": 0.6151574803149606, + "grad_norm": 11.835079193115234, + "learning_rate": 6.172962226640161e-07, + "loss": 0.5478, + "step": 625 + }, + { + "epoch": 0.6161417322834646, + "grad_norm": 19.864343643188477, + "learning_rate": 6.182902584493043e-07, + "loss": 0.9155, + "step": 626 + }, + { + "epoch": 0.6171259842519685, + "grad_norm": 23.956178665161133, + "learning_rate": 6.192842942345925e-07, + "loss": 1.8947, + "step": 627 + }, + { + "epoch": 0.6181102362204725, + "grad_norm": 19.554405212402344, + "learning_rate": 6.202783300198809e-07, + "loss": 1.1943, + "step": 628 + }, + { + "epoch": 0.6190944881889764, + "grad_norm": 20.88629913330078, + "learning_rate": 6.212723658051691e-07, + "loss": 1.2465, + "step": 629 + }, + { + "epoch": 0.6200787401574803, + "grad_norm": 11.647974967956543, + "learning_rate": 6.222664015904573e-07, + "loss": 0.4045, + "step": 630 + }, + { + "epoch": 0.6210629921259843, + "grad_norm": 14.859945297241211, + "learning_rate": 6.232604373757457e-07, + "loss": 0.5688, + "step": 631 + }, + { + "epoch": 0.6220472440944882, + "grad_norm": 19.23811149597168, + "learning_rate": 6.242544731610339e-07, + "loss": 0.8682, + "step": 632 + }, + { + "epoch": 0.6230314960629921, + "grad_norm": 20.735319137573242, + "learning_rate": 6.252485089463222e-07, + "loss": 0.901, + "step": 633 + }, + { + "epoch": 0.6240157480314961, + "grad_norm": 16.829877853393555, + "learning_rate": 6.262425447316104e-07, + "loss": 0.8978, + "step": 634 + }, + { + "epoch": 0.625, + "grad_norm": 13.868021965026855, + "learning_rate": 6.272365805168987e-07, + "loss": 0.6592, + "step": 635 + }, + { + "epoch": 0.6259842519685039, + "grad_norm": 13.756629943847656, + "learning_rate": 6.28230616302187e-07, + "loss": 0.6828, + "step": 636 + }, + { + "epoch": 0.6269685039370079, + "grad_norm": 9.567875862121582, + "learning_rate": 6.292246520874752e-07, + "loss": 0.3995, + "step": 637 + }, + { + "epoch": 0.6279527559055118, + "grad_norm": 15.121262550354004, + "learning_rate": 6.302186878727636e-07, + "loss": 0.8283, + "step": 638 + }, + { + "epoch": 0.6289370078740157, + "grad_norm": 11.696405410766602, + "learning_rate": 6.312127236580518e-07, + "loss": 0.5938, + "step": 639 + }, + { + "epoch": 0.6299212598425197, + "grad_norm": 18.15774917602539, + "learning_rate": 6.3220675944334e-07, + "loss": 1.256, + "step": 640 + }, + { + "epoch": 0.6309055118110236, + "grad_norm": 16.493196487426758, + "learning_rate": 6.332007952286283e-07, + "loss": 0.7801, + "step": 641 + }, + { + "epoch": 0.6318897637795275, + "grad_norm": 15.939420700073242, + "learning_rate": 6.341948310139166e-07, + "loss": 0.8975, + "step": 642 + }, + { + "epoch": 0.6328740157480315, + "grad_norm": 15.438959121704102, + "learning_rate": 6.351888667992048e-07, + "loss": 1.255, + "step": 643 + }, + { + "epoch": 0.6338582677165354, + "grad_norm": 18.454252243041992, + "learning_rate": 6.361829025844931e-07, + "loss": 1.1252, + "step": 644 + }, + { + "epoch": 0.6348425196850394, + "grad_norm": 18.03512954711914, + "learning_rate": 6.371769383697813e-07, + "loss": 1.2394, + "step": 645 + }, + { + "epoch": 0.6358267716535433, + "grad_norm": 14.261012077331543, + "learning_rate": 6.381709741550698e-07, + "loss": 0.9579, + "step": 646 + }, + { + "epoch": 0.6368110236220472, + "grad_norm": 18.299516677856445, + "learning_rate": 6.39165009940358e-07, + "loss": 0.8664, + "step": 647 + }, + { + "epoch": 0.6377952755905512, + "grad_norm": 8.959807395935059, + "learning_rate": 6.401590457256462e-07, + "loss": 0.4087, + "step": 648 + }, + { + "epoch": 0.6387795275590551, + "grad_norm": 11.289334297180176, + "learning_rate": 6.411530815109346e-07, + "loss": 0.4084, + "step": 649 + }, + { + "epoch": 0.639763779527559, + "grad_norm": 15.420426368713379, + "learning_rate": 6.421471172962228e-07, + "loss": 0.6214, + "step": 650 + }, + { + "epoch": 0.640748031496063, + "grad_norm": 16.837846755981445, + "learning_rate": 6.431411530815111e-07, + "loss": 0.7237, + "step": 651 + }, + { + "epoch": 0.6417322834645669, + "grad_norm": 13.567400932312012, + "learning_rate": 6.441351888667994e-07, + "loss": 0.6294, + "step": 652 + }, + { + "epoch": 0.6427165354330708, + "grad_norm": 15.608287811279297, + "learning_rate": 6.451292246520876e-07, + "loss": 0.8264, + "step": 653 + }, + { + "epoch": 0.6437007874015748, + "grad_norm": 15.987324714660645, + "learning_rate": 6.461232604373759e-07, + "loss": 0.8258, + "step": 654 + }, + { + "epoch": 0.6446850393700787, + "grad_norm": 15.27828598022461, + "learning_rate": 6.471172962226641e-07, + "loss": 0.6865, + "step": 655 + }, + { + "epoch": 0.6456692913385826, + "grad_norm": 15.643537521362305, + "learning_rate": 6.481113320079524e-07, + "loss": 0.7408, + "step": 656 + }, + { + "epoch": 0.6466535433070866, + "grad_norm": 14.839326858520508, + "learning_rate": 6.491053677932407e-07, + "loss": 0.6196, + "step": 657 + }, + { + "epoch": 0.6476377952755905, + "grad_norm": 13.28332805633545, + "learning_rate": 6.500994035785289e-07, + "loss": 0.5659, + "step": 658 + }, + { + "epoch": 0.6486220472440944, + "grad_norm": 19.671979904174805, + "learning_rate": 6.510934393638173e-07, + "loss": 0.6353, + "step": 659 + }, + { + "epoch": 0.6496062992125984, + "grad_norm": 19.528797149658203, + "learning_rate": 6.520874751491055e-07, + "loss": 0.8432, + "step": 660 + }, + { + "epoch": 0.6505905511811023, + "grad_norm": 14.360482215881348, + "learning_rate": 6.530815109343937e-07, + "loss": 0.6573, + "step": 661 + }, + { + "epoch": 0.6515748031496063, + "grad_norm": 13.865690231323242, + "learning_rate": 6.54075546719682e-07, + "loss": 0.4918, + "step": 662 + }, + { + "epoch": 0.6525590551181102, + "grad_norm": 16.34115219116211, + "learning_rate": 6.550695825049703e-07, + "loss": 0.8571, + "step": 663 + }, + { + "epoch": 0.6535433070866141, + "grad_norm": 16.546566009521484, + "learning_rate": 6.560636182902586e-07, + "loss": 0.9105, + "step": 664 + }, + { + "epoch": 0.6545275590551181, + "grad_norm": 18.405078887939453, + "learning_rate": 6.570576540755468e-07, + "loss": 0.7499, + "step": 665 + }, + { + "epoch": 0.655511811023622, + "grad_norm": 14.086722373962402, + "learning_rate": 6.580516898608351e-07, + "loss": 0.5277, + "step": 666 + }, + { + "epoch": 0.656496062992126, + "grad_norm": 19.698976516723633, + "learning_rate": 6.590457256461234e-07, + "loss": 1.3046, + "step": 667 + }, + { + "epoch": 0.65748031496063, + "grad_norm": 14.89286994934082, + "learning_rate": 6.600397614314116e-07, + "loss": 0.6112, + "step": 668 + }, + { + "epoch": 0.6584645669291339, + "grad_norm": 17.876667022705078, + "learning_rate": 6.610337972166998e-07, + "loss": 0.754, + "step": 669 + }, + { + "epoch": 0.6594488188976378, + "grad_norm": 25.529808044433594, + "learning_rate": 6.620278330019882e-07, + "loss": 1.3611, + "step": 670 + }, + { + "epoch": 0.6604330708661418, + "grad_norm": 22.641721725463867, + "learning_rate": 6.630218687872764e-07, + "loss": 1.3654, + "step": 671 + }, + { + "epoch": 0.6614173228346457, + "grad_norm": 16.472440719604492, + "learning_rate": 6.640159045725647e-07, + "loss": 0.6479, + "step": 672 + }, + { + "epoch": 0.6624015748031497, + "grad_norm": 16.357345581054688, + "learning_rate": 6.65009940357853e-07, + "loss": 0.9422, + "step": 673 + }, + { + "epoch": 0.6633858267716536, + "grad_norm": 14.273353576660156, + "learning_rate": 6.660039761431412e-07, + "loss": 0.5007, + "step": 674 + }, + { + "epoch": 0.6643700787401575, + "grad_norm": 12.19189167022705, + "learning_rate": 6.669980119284295e-07, + "loss": 0.4789, + "step": 675 + }, + { + "epoch": 0.6653543307086615, + "grad_norm": 14.178848266601562, + "learning_rate": 6.679920477137177e-07, + "loss": 0.3962, + "step": 676 + }, + { + "epoch": 0.6663385826771654, + "grad_norm": 16.41002655029297, + "learning_rate": 6.689860834990062e-07, + "loss": 0.8648, + "step": 677 + }, + { + "epoch": 0.6673228346456693, + "grad_norm": 15.098052978515625, + "learning_rate": 6.699801192842944e-07, + "loss": 0.8241, + "step": 678 + }, + { + "epoch": 0.6683070866141733, + "grad_norm": 16.379117965698242, + "learning_rate": 6.709741550695826e-07, + "loss": 0.7923, + "step": 679 + }, + { + "epoch": 0.6692913385826772, + "grad_norm": 13.851663589477539, + "learning_rate": 6.71968190854871e-07, + "loss": 0.5799, + "step": 680 + }, + { + "epoch": 0.6702755905511811, + "grad_norm": 17.2502498626709, + "learning_rate": 6.729622266401592e-07, + "loss": 0.8422, + "step": 681 + }, + { + "epoch": 0.6712598425196851, + "grad_norm": 15.975495338439941, + "learning_rate": 6.739562624254474e-07, + "loss": 0.5638, + "step": 682 + }, + { + "epoch": 0.672244094488189, + "grad_norm": 14.308259010314941, + "learning_rate": 6.749502982107357e-07, + "loss": 0.9627, + "step": 683 + }, + { + "epoch": 0.6732283464566929, + "grad_norm": 16.338415145874023, + "learning_rate": 6.75944333996024e-07, + "loss": 0.7585, + "step": 684 + }, + { + "epoch": 0.6742125984251969, + "grad_norm": 11.060367584228516, + "learning_rate": 6.769383697813123e-07, + "loss": 0.488, + "step": 685 + }, + { + "epoch": 0.6751968503937008, + "grad_norm": 20.239194869995117, + "learning_rate": 6.779324055666005e-07, + "loss": 1.0991, + "step": 686 + }, + { + "epoch": 0.6761811023622047, + "grad_norm": 13.97426986694336, + "learning_rate": 6.789264413518888e-07, + "loss": 0.4808, + "step": 687 + }, + { + "epoch": 0.6771653543307087, + "grad_norm": 25.122629165649414, + "learning_rate": 6.799204771371771e-07, + "loss": 1.3717, + "step": 688 + }, + { + "epoch": 0.6781496062992126, + "grad_norm": 17.241931915283203, + "learning_rate": 6.809145129224653e-07, + "loss": 0.8225, + "step": 689 + }, + { + "epoch": 0.6791338582677166, + "grad_norm": 17.039987564086914, + "learning_rate": 6.819085487077536e-07, + "loss": 0.9377, + "step": 690 + }, + { + "epoch": 0.6801181102362205, + "grad_norm": 19.283510208129883, + "learning_rate": 6.829025844930419e-07, + "loss": 0.8016, + "step": 691 + }, + { + "epoch": 0.6811023622047244, + "grad_norm": 16.10774040222168, + "learning_rate": 6.838966202783301e-07, + "loss": 0.6257, + "step": 692 + }, + { + "epoch": 0.6820866141732284, + "grad_norm": 16.3678035736084, + "learning_rate": 6.848906560636184e-07, + "loss": 1.0941, + "step": 693 + }, + { + "epoch": 0.6830708661417323, + "grad_norm": 18.00971031188965, + "learning_rate": 6.858846918489067e-07, + "loss": 1.0878, + "step": 694 + }, + { + "epoch": 0.6840551181102362, + "grad_norm": 14.44545841217041, + "learning_rate": 6.86878727634195e-07, + "loss": 0.9231, + "step": 695 + }, + { + "epoch": 0.6850393700787402, + "grad_norm": 23.54404067993164, + "learning_rate": 6.878727634194832e-07, + "loss": 1.3284, + "step": 696 + }, + { + "epoch": 0.6860236220472441, + "grad_norm": 16.13181495666504, + "learning_rate": 6.888667992047714e-07, + "loss": 0.7274, + "step": 697 + }, + { + "epoch": 0.687007874015748, + "grad_norm": 16.25248908996582, + "learning_rate": 6.898608349900598e-07, + "loss": 0.7851, + "step": 698 + }, + { + "epoch": 0.687992125984252, + "grad_norm": 13.41655158996582, + "learning_rate": 6.90854870775348e-07, + "loss": 0.6807, + "step": 699 + }, + { + "epoch": 0.6889763779527559, + "grad_norm": 17.609289169311523, + "learning_rate": 6.918489065606362e-07, + "loss": 1.1592, + "step": 700 + }, + { + "epoch": 0.6899606299212598, + "grad_norm": 12.81274127960205, + "learning_rate": 6.928429423459246e-07, + "loss": 0.6347, + "step": 701 + }, + { + "epoch": 0.6909448818897638, + "grad_norm": 18.530323028564453, + "learning_rate": 6.938369781312128e-07, + "loss": 1.1326, + "step": 702 + }, + { + "epoch": 0.6919291338582677, + "grad_norm": 19.012773513793945, + "learning_rate": 6.948310139165011e-07, + "loss": 1.3797, + "step": 703 + }, + { + "epoch": 0.6929133858267716, + "grad_norm": 11.886448860168457, + "learning_rate": 6.958250497017893e-07, + "loss": 0.7782, + "step": 704 + }, + { + "epoch": 0.6938976377952756, + "grad_norm": 16.526399612426758, + "learning_rate": 6.968190854870776e-07, + "loss": 1.1746, + "step": 705 + }, + { + "epoch": 0.6948818897637795, + "grad_norm": 16.093982696533203, + "learning_rate": 6.97813121272366e-07, + "loss": 0.7624, + "step": 706 + }, + { + "epoch": 0.6958661417322834, + "grad_norm": 13.874664306640625, + "learning_rate": 6.988071570576542e-07, + "loss": 0.7216, + "step": 707 + }, + { + "epoch": 0.6968503937007874, + "grad_norm": 13.424209594726562, + "learning_rate": 6.998011928429426e-07, + "loss": 0.4666, + "step": 708 + }, + { + "epoch": 0.6978346456692913, + "grad_norm": 14.339217185974121, + "learning_rate": 7.007952286282308e-07, + "loss": 0.9398, + "step": 709 + }, + { + "epoch": 0.6988188976377953, + "grad_norm": 14.998201370239258, + "learning_rate": 7.01789264413519e-07, + "loss": 0.9532, + "step": 710 + }, + { + "epoch": 0.6998031496062992, + "grad_norm": 19.089515686035156, + "learning_rate": 7.027833001988073e-07, + "loss": 0.8285, + "step": 711 + }, + { + "epoch": 0.7007874015748031, + "grad_norm": 15.774718284606934, + "learning_rate": 7.037773359840956e-07, + "loss": 0.7042, + "step": 712 + }, + { + "epoch": 0.7017716535433071, + "grad_norm": 19.596515655517578, + "learning_rate": 7.047713717693838e-07, + "loss": 1.0057, + "step": 713 + }, + { + "epoch": 0.702755905511811, + "grad_norm": 13.490047454833984, + "learning_rate": 7.057654075546721e-07, + "loss": 0.6696, + "step": 714 + }, + { + "epoch": 0.7037401574803149, + "grad_norm": 13.74376106262207, + "learning_rate": 7.067594433399604e-07, + "loss": 0.5701, + "step": 715 + }, + { + "epoch": 0.7047244094488189, + "grad_norm": 19.08216094970703, + "learning_rate": 7.077534791252487e-07, + "loss": 0.7894, + "step": 716 + }, + { + "epoch": 0.7057086614173228, + "grad_norm": 17.408754348754883, + "learning_rate": 7.087475149105369e-07, + "loss": 0.9327, + "step": 717 + }, + { + "epoch": 0.7066929133858267, + "grad_norm": 19.364654541015625, + "learning_rate": 7.097415506958251e-07, + "loss": 1.0283, + "step": 718 + }, + { + "epoch": 0.7076771653543307, + "grad_norm": 14.826738357543945, + "learning_rate": 7.107355864811135e-07, + "loss": 0.6419, + "step": 719 + }, + { + "epoch": 0.7086614173228346, + "grad_norm": 12.711971282958984, + "learning_rate": 7.117296222664017e-07, + "loss": 0.7716, + "step": 720 + }, + { + "epoch": 0.7096456692913385, + "grad_norm": 12.659597396850586, + "learning_rate": 7.1272365805169e-07, + "loss": 0.5968, + "step": 721 + }, + { + "epoch": 0.7106299212598425, + "grad_norm": 15.052842140197754, + "learning_rate": 7.137176938369783e-07, + "loss": 0.6547, + "step": 722 + }, + { + "epoch": 0.7116141732283464, + "grad_norm": 15.427705764770508, + "learning_rate": 7.147117296222665e-07, + "loss": 0.8344, + "step": 723 + }, + { + "epoch": 0.7125984251968503, + "grad_norm": 12.529552459716797, + "learning_rate": 7.157057654075548e-07, + "loss": 0.5448, + "step": 724 + }, + { + "epoch": 0.7135826771653543, + "grad_norm": 15.054535865783691, + "learning_rate": 7.16699801192843e-07, + "loss": 0.6075, + "step": 725 + }, + { + "epoch": 0.7145669291338582, + "grad_norm": 13.75622844696045, + "learning_rate": 7.176938369781313e-07, + "loss": 0.6142, + "step": 726 + }, + { + "epoch": 0.7155511811023622, + "grad_norm": 15.974542617797852, + "learning_rate": 7.186878727634196e-07, + "loss": 0.681, + "step": 727 + }, + { + "epoch": 0.7165354330708661, + "grad_norm": 15.134488105773926, + "learning_rate": 7.196819085487078e-07, + "loss": 0.6137, + "step": 728 + }, + { + "epoch": 0.71751968503937, + "grad_norm": 24.855924606323242, + "learning_rate": 7.206759443339962e-07, + "loss": 1.4516, + "step": 729 + }, + { + "epoch": 0.718503937007874, + "grad_norm": 20.611099243164062, + "learning_rate": 7.216699801192844e-07, + "loss": 0.9083, + "step": 730 + }, + { + "epoch": 0.719488188976378, + "grad_norm": 13.90792179107666, + "learning_rate": 7.226640159045726e-07, + "loss": 0.6369, + "step": 731 + }, + { + "epoch": 0.7204724409448819, + "grad_norm": 14.253960609436035, + "learning_rate": 7.236580516898609e-07, + "loss": 0.8102, + "step": 732 + }, + { + "epoch": 0.7214566929133859, + "grad_norm": 14.44489574432373, + "learning_rate": 7.246520874751492e-07, + "loss": 0.7808, + "step": 733 + }, + { + "epoch": 0.7224409448818898, + "grad_norm": 12.419161796569824, + "learning_rate": 7.256461232604375e-07, + "loss": 0.512, + "step": 734 + }, + { + "epoch": 0.7234251968503937, + "grad_norm": 23.816450119018555, + "learning_rate": 7.266401590457257e-07, + "loss": 1.3589, + "step": 735 + }, + { + "epoch": 0.7244094488188977, + "grad_norm": 15.650845527648926, + "learning_rate": 7.27634194831014e-07, + "loss": 0.6555, + "step": 736 + }, + { + "epoch": 0.7253937007874016, + "grad_norm": 19.33920669555664, + "learning_rate": 7.286282306163024e-07, + "loss": 0.8732, + "step": 737 + }, + { + "epoch": 0.7263779527559056, + "grad_norm": 17.49276351928711, + "learning_rate": 7.296222664015906e-07, + "loss": 0.7591, + "step": 738 + }, + { + "epoch": 0.7273622047244095, + "grad_norm": 18.234067916870117, + "learning_rate": 7.306163021868788e-07, + "loss": 0.8237, + "step": 739 + }, + { + "epoch": 0.7283464566929134, + "grad_norm": 18.252256393432617, + "learning_rate": 7.316103379721672e-07, + "loss": 1.1272, + "step": 740 + }, + { + "epoch": 0.7293307086614174, + "grad_norm": 13.679298400878906, + "learning_rate": 7.326043737574554e-07, + "loss": 0.5353, + "step": 741 + }, + { + "epoch": 0.7303149606299213, + "grad_norm": 18.588396072387695, + "learning_rate": 7.335984095427437e-07, + "loss": 1.1713, + "step": 742 + }, + { + "epoch": 0.7312992125984252, + "grad_norm": 13.681812286376953, + "learning_rate": 7.34592445328032e-07, + "loss": 0.7034, + "step": 743 + }, + { + "epoch": 0.7322834645669292, + "grad_norm": 12.749622344970703, + "learning_rate": 7.355864811133202e-07, + "loss": 0.5118, + "step": 744 + }, + { + "epoch": 0.7332677165354331, + "grad_norm": 18.917062759399414, + "learning_rate": 7.365805168986085e-07, + "loss": 0.7816, + "step": 745 + }, + { + "epoch": 0.734251968503937, + "grad_norm": 16.785316467285156, + "learning_rate": 7.375745526838967e-07, + "loss": 0.7935, + "step": 746 + }, + { + "epoch": 0.735236220472441, + "grad_norm": 17.046855926513672, + "learning_rate": 7.385685884691851e-07, + "loss": 1.0736, + "step": 747 + }, + { + "epoch": 0.7362204724409449, + "grad_norm": 14.675033569335938, + "learning_rate": 7.395626242544733e-07, + "loss": 0.6528, + "step": 748 + }, + { + "epoch": 0.7372047244094488, + "grad_norm": 13.068770408630371, + "learning_rate": 7.405566600397615e-07, + "loss": 0.5556, + "step": 749 + }, + { + "epoch": 0.7381889763779528, + "grad_norm": 11.902694702148438, + "learning_rate": 7.415506958250499e-07, + "loss": 0.3339, + "step": 750 + }, + { + "epoch": 0.7391732283464567, + "grad_norm": 17.953121185302734, + "learning_rate": 7.425447316103381e-07, + "loss": 1.0371, + "step": 751 + }, + { + "epoch": 0.7401574803149606, + "grad_norm": 21.42251968383789, + "learning_rate": 7.435387673956263e-07, + "loss": 1.3273, + "step": 752 + }, + { + "epoch": 0.7411417322834646, + "grad_norm": 16.906721115112305, + "learning_rate": 7.445328031809146e-07, + "loss": 0.5307, + "step": 753 + }, + { + "epoch": 0.7421259842519685, + "grad_norm": 17.6983642578125, + "learning_rate": 7.455268389662029e-07, + "loss": 0.908, + "step": 754 + }, + { + "epoch": 0.7431102362204725, + "grad_norm": 15.194808959960938, + "learning_rate": 7.465208747514912e-07, + "loss": 0.5248, + "step": 755 + }, + { + "epoch": 0.7440944881889764, + "grad_norm": 17.43975067138672, + "learning_rate": 7.475149105367794e-07, + "loss": 0.6128, + "step": 756 + }, + { + "epoch": 0.7450787401574803, + "grad_norm": 23.67591667175293, + "learning_rate": 7.485089463220677e-07, + "loss": 1.5065, + "step": 757 + }, + { + "epoch": 0.7460629921259843, + "grad_norm": 20.08376693725586, + "learning_rate": 7.49502982107356e-07, + "loss": 1.2071, + "step": 758 + }, + { + "epoch": 0.7470472440944882, + "grad_norm": 19.50629425048828, + "learning_rate": 7.504970178926442e-07, + "loss": 0.9983, + "step": 759 + }, + { + "epoch": 0.7480314960629921, + "grad_norm": 18.14684295654297, + "learning_rate": 7.514910536779325e-07, + "loss": 1.0378, + "step": 760 + }, + { + "epoch": 0.7490157480314961, + "grad_norm": 16.399993896484375, + "learning_rate": 7.524850894632208e-07, + "loss": 1.0732, + "step": 761 + }, + { + "epoch": 0.75, + "grad_norm": 19.110761642456055, + "learning_rate": 7.53479125248509e-07, + "loss": 0.8046, + "step": 762 + }, + { + "epoch": 0.7509842519685039, + "grad_norm": 17.772491455078125, + "learning_rate": 7.544731610337973e-07, + "loss": 1.1666, + "step": 763 + }, + { + "epoch": 0.7519685039370079, + "grad_norm": 17.930280685424805, + "learning_rate": 7.554671968190856e-07, + "loss": 0.7042, + "step": 764 + }, + { + "epoch": 0.7529527559055118, + "grad_norm": 20.83265495300293, + "learning_rate": 7.56461232604374e-07, + "loss": 1.4759, + "step": 765 + }, + { + "epoch": 0.7529527559055118, + "eval_Qnli-dev_cosine_accuracy": 0.67578125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8426380157470703, + "eval_Qnli-dev_cosine_ap": 0.6951783685178275, + "eval_Qnli-dev_cosine_f1": 0.6866197183098591, + "eval_Qnli-dev_cosine_f1_threshold": 0.8136929869651794, + "eval_Qnli-dev_cosine_precision": 0.5873493975903614, + "eval_Qnli-dev_cosine_recall": 0.826271186440678, + "eval_Qnli-dev_dot_accuracy": 0.67578125, + "eval_Qnli-dev_dot_accuracy_threshold": 647.48828125, + "eval_Qnli-dev_dot_ap": 0.6953081100663845, + "eval_Qnli-dev_dot_f1": 0.6866197183098591, + "eval_Qnli-dev_dot_f1_threshold": 625.2723388671875, + "eval_Qnli-dev_dot_precision": 0.5873493975903614, + "eval_Qnli-dev_dot_recall": 0.826271186440678, + "eval_Qnli-dev_euclidean_accuracy": 0.67578125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.551056861877441, + "eval_Qnli-dev_euclidean_ap": 0.695195292717627, + "eval_Qnli-dev_euclidean_f1": 0.6866197183098591, + "eval_Qnli-dev_euclidean_f1_threshold": 16.921308517456055, + "eval_Qnli-dev_euclidean_precision": 0.5873493975903614, + "eval_Qnli-dev_euclidean_recall": 0.826271186440678, + "eval_Qnli-dev_manhattan_accuracy": 0.673828125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 337.1568603515625, + "eval_Qnli-dev_manhattan_ap": 0.6981557028078741, + "eval_Qnli-dev_manhattan_f1": 0.6913996627318718, + "eval_Qnli-dev_manhattan_f1_threshold": 378.724609375, + "eval_Qnli-dev_manhattan_precision": 0.5742296918767507, + "eval_Qnli-dev_manhattan_recall": 0.8686440677966102, + "eval_Qnli-dev_max_accuracy": 0.67578125, + "eval_Qnli-dev_max_accuracy_threshold": 647.48828125, + "eval_Qnli-dev_max_ap": 0.6981557028078741, + "eval_Qnli-dev_max_f1": 0.6913996627318718, + "eval_Qnli-dev_max_f1_threshold": 625.2723388671875, + "eval_Qnli-dev_max_precision": 0.5873493975903614, + "eval_Qnli-dev_max_recall": 0.8686440677966102, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9270462989807129, + "eval_allNLI-dev_cosine_ap": 0.5065548770111167, + "eval_allNLI-dev_cosine_f1": 0.5446808510638298, + "eval_allNLI-dev_cosine_f1_threshold": 0.8589708805084229, + "eval_allNLI-dev_cosine_precision": 0.43097643097643096, + "eval_allNLI-dev_cosine_recall": 0.7398843930635838, + "eval_allNLI-dev_dot_accuracy": 0.689453125, + "eval_allNLI-dev_dot_accuracy_threshold": 712.409912109375, + "eval_allNLI-dev_dot_ap": 0.5065544970851109, + "eval_allNLI-dev_dot_f1": 0.5446808510638298, + "eval_allNLI-dev_dot_f1_threshold": 660.105224609375, + "eval_allNLI-dev_dot_precision": 0.43097643097643096, + "eval_allNLI-dev_dot_recall": 0.7398843930635838, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.588932037353516, + "eval_allNLI-dev_euclidean_ap": 0.5065548770111167, + "eval_allNLI-dev_euclidean_f1": 0.5446808510638298, + "eval_allNLI-dev_euclidean_f1_threshold": 14.722663879394531, + "eval_allNLI-dev_euclidean_precision": 0.43097643097643096, + "eval_allNLI-dev_euclidean_recall": 0.7398843930635838, + "eval_allNLI-dev_manhattan_accuracy": 0.693359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 261.6011962890625, + "eval_allNLI-dev_manhattan_ap": 0.5104540022958912, + "eval_allNLI-dev_manhattan_f1": 0.5442764578833693, + "eval_allNLI-dev_manhattan_f1_threshold": 316.1423645019531, + "eval_allNLI-dev_manhattan_precision": 0.43448275862068964, + "eval_allNLI-dev_manhattan_recall": 0.7283236994219653, + "eval_allNLI-dev_max_accuracy": 0.693359375, + "eval_allNLI-dev_max_accuracy_threshold": 712.409912109375, + "eval_allNLI-dev_max_ap": 0.5104540022958912, + "eval_allNLI-dev_max_f1": 0.5446808510638298, + "eval_allNLI-dev_max_f1_threshold": 660.105224609375, + "eval_allNLI-dev_max_precision": 0.43448275862068964, + "eval_allNLI-dev_max_recall": 0.7398843930635838, + "eval_runtime": 1.228, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.6981557028078741, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.5639496482103532, + "eval_sts-test_pearson_dot": 0.5637779615119842, + "eval_sts-test_pearson_euclidean": 0.5732787731326285, + "eval_sts-test_pearson_manhattan": 0.5739911399908657, + "eval_sts-test_pearson_max": 0.5739911399908657, + "eval_sts-test_spearman_cosine": 0.5729838735376087, + "eval_sts-test_spearman_dot": 0.5726842638425932, + "eval_sts-test_spearman_euclidean": 0.5730641573005449, + "eval_sts-test_spearman_manhattan": 0.5731718735867141, + "eval_sts-test_spearman_max": 0.5731718735867141, + "step": 765 + }, + { + "epoch": 0.7539370078740157, + "grad_norm": 12.599635124206543, + "learning_rate": 7.574552683896621e-07, + "loss": 0.5595, + "step": 766 + }, + { + "epoch": 0.7549212598425197, + "grad_norm": 11.732769966125488, + "learning_rate": 7.584493041749503e-07, + "loss": 0.3646, + "step": 767 + }, + { + "epoch": 0.7559055118110236, + "grad_norm": 15.522697448730469, + "learning_rate": 7.594433399602388e-07, + "loss": 0.6791, + "step": 768 + }, + { + "epoch": 0.7568897637795275, + "grad_norm": 14.634027481079102, + "learning_rate": 7.60437375745527e-07, + "loss": 0.6889, + "step": 769 + }, + { + "epoch": 0.7578740157480315, + "grad_norm": 18.199993133544922, + "learning_rate": 7.614314115308152e-07, + "loss": 0.9715, + "step": 770 + }, + { + "epoch": 0.7588582677165354, + "grad_norm": 19.96280288696289, + "learning_rate": 7.624254473161036e-07, + "loss": 1.2218, + "step": 771 + }, + { + "epoch": 0.7598425196850394, + "grad_norm": 11.620733261108398, + "learning_rate": 7.634194831013918e-07, + "loss": 0.3957, + "step": 772 + }, + { + "epoch": 0.7608267716535433, + "grad_norm": 22.017301559448242, + "learning_rate": 7.644135188866801e-07, + "loss": 1.5635, + "step": 773 + }, + { + "epoch": 0.7618110236220472, + "grad_norm": 14.363546371459961, + "learning_rate": 7.654075546719683e-07, + "loss": 0.9081, + "step": 774 + }, + { + "epoch": 0.7627952755905512, + "grad_norm": 15.683711051940918, + "learning_rate": 7.664015904572566e-07, + "loss": 0.8633, + "step": 775 + }, + { + "epoch": 0.7637795275590551, + "grad_norm": 15.376205444335938, + "learning_rate": 7.673956262425449e-07, + "loss": 0.9482, + "step": 776 + }, + { + "epoch": 0.764763779527559, + "grad_norm": 14.532920837402344, + "learning_rate": 7.683896620278331e-07, + "loss": 0.8021, + "step": 777 + }, + { + "epoch": 0.765748031496063, + "grad_norm": 18.542530059814453, + "learning_rate": 7.693836978131215e-07, + "loss": 0.9083, + "step": 778 + }, + { + "epoch": 0.7667322834645669, + "grad_norm": 17.333229064941406, + "learning_rate": 7.703777335984097e-07, + "loss": 0.9599, + "step": 779 + }, + { + "epoch": 0.7677165354330708, + "grad_norm": 17.525917053222656, + "learning_rate": 7.713717693836979e-07, + "loss": 0.798, + "step": 780 + }, + { + "epoch": 0.7687007874015748, + "grad_norm": 12.102128982543945, + "learning_rate": 7.723658051689862e-07, + "loss": 0.5817, + "step": 781 + }, + { + "epoch": 0.7696850393700787, + "grad_norm": 11.137146949768066, + "learning_rate": 7.733598409542745e-07, + "loss": 0.5287, + "step": 782 + }, + { + "epoch": 0.7706692913385826, + "grad_norm": 17.296815872192383, + "learning_rate": 7.743538767395627e-07, + "loss": 1.0251, + "step": 783 + }, + { + "epoch": 0.7716535433070866, + "grad_norm": 12.617779731750488, + "learning_rate": 7.75347912524851e-07, + "loss": 0.3995, + "step": 784 + }, + { + "epoch": 0.7726377952755905, + "grad_norm": 18.820236206054688, + "learning_rate": 7.763419483101393e-07, + "loss": 0.8753, + "step": 785 + }, + { + "epoch": 0.7736220472440944, + "grad_norm": 26.430797576904297, + "learning_rate": 7.773359840954276e-07, + "loss": 1.315, + "step": 786 + }, + { + "epoch": 0.7746062992125984, + "grad_norm": 15.11793327331543, + "learning_rate": 7.783300198807158e-07, + "loss": 0.8053, + "step": 787 + }, + { + "epoch": 0.7755905511811023, + "grad_norm": 16.853519439697266, + "learning_rate": 7.79324055666004e-07, + "loss": 0.6032, + "step": 788 + }, + { + "epoch": 0.7765748031496063, + "grad_norm": 14.197515487670898, + "learning_rate": 7.803180914512924e-07, + "loss": 0.8059, + "step": 789 + }, + { + "epoch": 0.7775590551181102, + "grad_norm": 17.37537956237793, + "learning_rate": 7.813121272365806e-07, + "loss": 0.9387, + "step": 790 + }, + { + "epoch": 0.7785433070866141, + "grad_norm": 12.913000106811523, + "learning_rate": 7.823061630218689e-07, + "loss": 0.8321, + "step": 791 + }, + { + "epoch": 0.7795275590551181, + "grad_norm": 14.940260887145996, + "learning_rate": 7.833001988071572e-07, + "loss": 0.5788, + "step": 792 + }, + { + "epoch": 0.780511811023622, + "grad_norm": 13.924427032470703, + "learning_rate": 7.842942345924454e-07, + "loss": 0.7874, + "step": 793 + }, + { + "epoch": 0.781496062992126, + "grad_norm": 17.241674423217773, + "learning_rate": 7.852882703777337e-07, + "loss": 0.5145, + "step": 794 + }, + { + "epoch": 0.78248031496063, + "grad_norm": 26.6125431060791, + "learning_rate": 7.862823061630219e-07, + "loss": 1.5847, + "step": 795 + }, + { + "epoch": 0.7834645669291339, + "grad_norm": 20.913875579833984, + "learning_rate": 7.872763419483102e-07, + "loss": 1.1473, + "step": 796 + }, + { + "epoch": 0.7844488188976378, + "grad_norm": 18.444318771362305, + "learning_rate": 7.882703777335986e-07, + "loss": 0.968, + "step": 797 + }, + { + "epoch": 0.7854330708661418, + "grad_norm": 18.83793830871582, + "learning_rate": 7.892644135188868e-07, + "loss": 1.2691, + "step": 798 + }, + { + "epoch": 0.7864173228346457, + "grad_norm": 13.278544425964355, + "learning_rate": 7.902584493041752e-07, + "loss": 0.7046, + "step": 799 + }, + { + "epoch": 0.7874015748031497, + "grad_norm": 15.995448112487793, + "learning_rate": 7.912524850894634e-07, + "loss": 0.9358, + "step": 800 + }, + { + "epoch": 0.7883858267716536, + "grad_norm": 11.438610076904297, + "learning_rate": 7.922465208747516e-07, + "loss": 0.3945, + "step": 801 + }, + { + "epoch": 0.7893700787401575, + "grad_norm": 17.80826187133789, + "learning_rate": 7.932405566600399e-07, + "loss": 0.861, + "step": 802 + }, + { + "epoch": 0.7903543307086615, + "grad_norm": 17.29227066040039, + "learning_rate": 7.942345924453282e-07, + "loss": 1.0736, + "step": 803 + }, + { + "epoch": 0.7913385826771654, + "grad_norm": 17.638120651245117, + "learning_rate": 7.952286282306165e-07, + "loss": 1.0279, + "step": 804 + }, + { + "epoch": 0.7923228346456693, + "grad_norm": 13.98741340637207, + "learning_rate": 7.962226640159047e-07, + "loss": 0.6902, + "step": 805 + }, + { + "epoch": 0.7933070866141733, + "grad_norm": 13.206335067749023, + "learning_rate": 7.97216699801193e-07, + "loss": 0.7344, + "step": 806 + }, + { + "epoch": 0.7942913385826772, + "grad_norm": 14.008211135864258, + "learning_rate": 7.982107355864813e-07, + "loss": 0.8184, + "step": 807 + }, + { + "epoch": 0.7952755905511811, + "grad_norm": 14.538393020629883, + "learning_rate": 7.992047713717695e-07, + "loss": 0.8121, + "step": 808 + }, + { + "epoch": 0.7962598425196851, + "grad_norm": 10.556940078735352, + "learning_rate": 8.001988071570577e-07, + "loss": 0.3772, + "step": 809 + }, + { + "epoch": 0.797244094488189, + "grad_norm": 8.442699432373047, + "learning_rate": 8.011928429423461e-07, + "loss": 0.2367, + "step": 810 + }, + { + "epoch": 0.7982283464566929, + "grad_norm": 15.54948902130127, + "learning_rate": 8.021868787276343e-07, + "loss": 0.5584, + "step": 811 + }, + { + "epoch": 0.7992125984251969, + "grad_norm": 11.8131103515625, + "learning_rate": 8.031809145129226e-07, + "loss": 0.3465, + "step": 812 + }, + { + "epoch": 0.8001968503937008, + "grad_norm": 20.774011611938477, + "learning_rate": 8.041749502982109e-07, + "loss": 0.9229, + "step": 813 + }, + { + "epoch": 0.8011811023622047, + "grad_norm": 16.44454574584961, + "learning_rate": 8.051689860834991e-07, + "loss": 0.7351, + "step": 814 + }, + { + "epoch": 0.8021653543307087, + "grad_norm": 17.405271530151367, + "learning_rate": 8.061630218687874e-07, + "loss": 0.9024, + "step": 815 + }, + { + "epoch": 0.8031496062992126, + "grad_norm": 15.097607612609863, + "learning_rate": 8.071570576540756e-07, + "loss": 0.4958, + "step": 816 + }, + { + "epoch": 0.8041338582677166, + "grad_norm": 14.987422943115234, + "learning_rate": 8.08151093439364e-07, + "loss": 0.4321, + "step": 817 + }, + { + "epoch": 0.8051181102362205, + "grad_norm": 16.28684425354004, + "learning_rate": 8.091451292246522e-07, + "loss": 0.6945, + "step": 818 + }, + { + "epoch": 0.8061023622047244, + "grad_norm": 17.50418472290039, + "learning_rate": 8.101391650099404e-07, + "loss": 0.6415, + "step": 819 + }, + { + "epoch": 0.8070866141732284, + "grad_norm": 11.575599670410156, + "learning_rate": 8.111332007952288e-07, + "loss": 0.5615, + "step": 820 + }, + { + "epoch": 0.8080708661417323, + "grad_norm": 18.075096130371094, + "learning_rate": 8.12127236580517e-07, + "loss": 0.9732, + "step": 821 + }, + { + "epoch": 0.8090551181102362, + "grad_norm": 19.075334548950195, + "learning_rate": 8.131212723658052e-07, + "loss": 1.018, + "step": 822 + }, + { + "epoch": 0.8100393700787402, + "grad_norm": 17.008920669555664, + "learning_rate": 8.141153081510935e-07, + "loss": 1.1392, + "step": 823 + }, + { + "epoch": 0.8110236220472441, + "grad_norm": 15.066267967224121, + "learning_rate": 8.151093439363818e-07, + "loss": 0.8539, + "step": 824 + }, + { + "epoch": 0.812007874015748, + "grad_norm": 14.156815528869629, + "learning_rate": 8.161033797216701e-07, + "loss": 0.7081, + "step": 825 + }, + { + "epoch": 0.812992125984252, + "grad_norm": 18.120996475219727, + "learning_rate": 8.170974155069583e-07, + "loss": 1.3176, + "step": 826 + }, + { + "epoch": 0.8139763779527559, + "grad_norm": 14.92929458618164, + "learning_rate": 8.180914512922467e-07, + "loss": 0.7387, + "step": 827 + }, + { + "epoch": 0.8149606299212598, + "grad_norm": 16.15532875061035, + "learning_rate": 8.19085487077535e-07, + "loss": 1.0975, + "step": 828 + }, + { + "epoch": 0.8159448818897638, + "grad_norm": 12.352117538452148, + "learning_rate": 8.200795228628232e-07, + "loss": 0.5883, + "step": 829 + }, + { + "epoch": 0.8169291338582677, + "grad_norm": 14.640236854553223, + "learning_rate": 8.210735586481115e-07, + "loss": 0.9677, + "step": 830 + }, + { + "epoch": 0.8179133858267716, + "grad_norm": 8.980406761169434, + "learning_rate": 8.220675944333998e-07, + "loss": 0.3869, + "step": 831 + }, + { + "epoch": 0.8188976377952756, + "grad_norm": 13.586002349853516, + "learning_rate": 8.23061630218688e-07, + "loss": 0.9083, + "step": 832 + }, + { + "epoch": 0.8198818897637795, + "grad_norm": 16.51270294189453, + "learning_rate": 8.240556660039763e-07, + "loss": 0.6939, + "step": 833 + }, + { + "epoch": 0.8208661417322834, + "grad_norm": 12.254718780517578, + "learning_rate": 8.250497017892646e-07, + "loss": 0.5046, + "step": 834 + }, + { + "epoch": 0.8218503937007874, + "grad_norm": 10.314390182495117, + "learning_rate": 8.260437375745529e-07, + "loss": 0.3594, + "step": 835 + }, + { + "epoch": 0.8228346456692913, + "grad_norm": 12.390287399291992, + "learning_rate": 8.270377733598411e-07, + "loss": 0.4609, + "step": 836 + }, + { + "epoch": 0.8238188976377953, + "grad_norm": 19.40741729736328, + "learning_rate": 8.280318091451293e-07, + "loss": 0.8132, + "step": 837 + }, + { + "epoch": 0.8248031496062992, + "grad_norm": 14.05292797088623, + "learning_rate": 8.290258449304177e-07, + "loss": 0.4493, + "step": 838 + }, + { + "epoch": 0.8257874015748031, + "grad_norm": 17.19973373413086, + "learning_rate": 8.300198807157059e-07, + "loss": 0.8566, + "step": 839 + }, + { + "epoch": 0.8267716535433071, + "grad_norm": 19.417970657348633, + "learning_rate": 8.310139165009941e-07, + "loss": 0.9958, + "step": 840 + }, + { + "epoch": 0.827755905511811, + "grad_norm": 21.66157341003418, + "learning_rate": 8.320079522862825e-07, + "loss": 0.936, + "step": 841 + }, + { + "epoch": 0.8287401574803149, + "grad_norm": 23.779054641723633, + "learning_rate": 8.330019880715707e-07, + "loss": 1.4087, + "step": 842 + }, + { + "epoch": 0.8297244094488189, + "grad_norm": 23.371971130371094, + "learning_rate": 8.33996023856859e-07, + "loss": 1.2214, + "step": 843 + }, + { + "epoch": 0.8307086614173228, + "grad_norm": 11.387801170349121, + "learning_rate": 8.349900596421472e-07, + "loss": 0.7135, + "step": 844 + }, + { + "epoch": 0.8316929133858267, + "grad_norm": 20.820316314697266, + "learning_rate": 8.359840954274355e-07, + "loss": 1.1881, + "step": 845 + }, + { + "epoch": 0.8326771653543307, + "grad_norm": 15.91108512878418, + "learning_rate": 8.369781312127238e-07, + "loss": 1.1807, + "step": 846 + }, + { + "epoch": 0.8336614173228346, + "grad_norm": 11.777990341186523, + "learning_rate": 8.37972166998012e-07, + "loss": 0.5896, + "step": 847 + }, + { + "epoch": 0.8346456692913385, + "grad_norm": 11.666973114013672, + "learning_rate": 8.389662027833004e-07, + "loss": 0.8483, + "step": 848 + }, + { + "epoch": 0.8356299212598425, + "grad_norm": 9.575151443481445, + "learning_rate": 8.399602385685886e-07, + "loss": 0.5409, + "step": 849 + }, + { + "epoch": 0.8366141732283464, + "grad_norm": 11.666224479675293, + "learning_rate": 8.409542743538768e-07, + "loss": 0.8286, + "step": 850 + }, + { + "epoch": 0.8375984251968503, + "grad_norm": 17.450098037719727, + "learning_rate": 8.419483101391651e-07, + "loss": 1.136, + "step": 851 + }, + { + "epoch": 0.8385826771653543, + "grad_norm": 12.592039108276367, + "learning_rate": 8.429423459244534e-07, + "loss": 0.819, + "step": 852 + }, + { + "epoch": 0.8395669291338582, + "grad_norm": 9.238852500915527, + "learning_rate": 8.439363817097416e-07, + "loss": 0.3029, + "step": 853 + }, + { + "epoch": 0.8405511811023622, + "grad_norm": 11.89981460571289, + "learning_rate": 8.449304174950299e-07, + "loss": 0.5538, + "step": 854 + }, + { + "epoch": 0.8415354330708661, + "grad_norm": 12.593090057373047, + "learning_rate": 8.459244532803182e-07, + "loss": 0.4572, + "step": 855 + }, + { + "epoch": 0.84251968503937, + "grad_norm": 17.559059143066406, + "learning_rate": 8.469184890656066e-07, + "loss": 1.1516, + "step": 856 + }, + { + "epoch": 0.843503937007874, + "grad_norm": 13.354970932006836, + "learning_rate": 8.479125248508948e-07, + "loss": 0.4768, + "step": 857 + }, + { + "epoch": 0.844488188976378, + "grad_norm": 13.012131690979004, + "learning_rate": 8.48906560636183e-07, + "loss": 0.5357, + "step": 858 + }, + { + "epoch": 0.8454724409448819, + "grad_norm": 15.363044738769531, + "learning_rate": 8.499005964214714e-07, + "loss": 0.8879, + "step": 859 + }, + { + "epoch": 0.8464566929133859, + "grad_norm": 14.82031536102295, + "learning_rate": 8.508946322067596e-07, + "loss": 0.9121, + "step": 860 + }, + { + "epoch": 0.8474409448818898, + "grad_norm": 20.62348175048828, + "learning_rate": 8.518886679920479e-07, + "loss": 1.052, + "step": 861 + }, + { + "epoch": 0.8484251968503937, + "grad_norm": 18.42710304260254, + "learning_rate": 8.528827037773362e-07, + "loss": 0.8009, + "step": 862 + }, + { + "epoch": 0.8494094488188977, + "grad_norm": 8.109525680541992, + "learning_rate": 8.538767395626244e-07, + "loss": 0.2474, + "step": 863 + }, + { + "epoch": 0.8503937007874016, + "grad_norm": 16.07292938232422, + "learning_rate": 8.548707753479127e-07, + "loss": 0.8517, + "step": 864 + }, + { + "epoch": 0.8513779527559056, + "grad_norm": 12.144257545471191, + "learning_rate": 8.558648111332009e-07, + "loss": 0.5678, + "step": 865 + }, + { + "epoch": 0.8523622047244095, + "grad_norm": 13.008956909179688, + "learning_rate": 8.568588469184892e-07, + "loss": 0.4882, + "step": 866 + }, + { + "epoch": 0.8533464566929134, + "grad_norm": 13.153145790100098, + "learning_rate": 8.578528827037775e-07, + "loss": 0.638, + "step": 867 + }, + { + "epoch": 0.8543307086614174, + "grad_norm": 17.879867553710938, + "learning_rate": 8.588469184890657e-07, + "loss": 0.8012, + "step": 868 + }, + { + "epoch": 0.8553149606299213, + "grad_norm": 12.189448356628418, + "learning_rate": 8.598409542743541e-07, + "loss": 0.4172, + "step": 869 + }, + { + "epoch": 0.8562992125984252, + "grad_norm": 15.825995445251465, + "learning_rate": 8.608349900596423e-07, + "loss": 0.6727, + "step": 870 + }, + { + "epoch": 0.8572834645669292, + "grad_norm": 13.850029945373535, + "learning_rate": 8.618290258449305e-07, + "loss": 0.6383, + "step": 871 + }, + { + "epoch": 0.8582677165354331, + "grad_norm": 18.910564422607422, + "learning_rate": 8.628230616302189e-07, + "loss": 1.0728, + "step": 872 + }, + { + "epoch": 0.859251968503937, + "grad_norm": 23.591352462768555, + "learning_rate": 8.638170974155071e-07, + "loss": 1.3336, + "step": 873 + }, + { + "epoch": 0.860236220472441, + "grad_norm": 24.322397232055664, + "learning_rate": 8.648111332007954e-07, + "loss": 1.217, + "step": 874 + }, + { + "epoch": 0.8612204724409449, + "grad_norm": 17.652233123779297, + "learning_rate": 8.658051689860836e-07, + "loss": 0.8018, + "step": 875 + }, + { + "epoch": 0.8622047244094488, + "grad_norm": 23.8310546875, + "learning_rate": 8.667992047713719e-07, + "loss": 1.2326, + "step": 876 + }, + { + "epoch": 0.8631889763779528, + "grad_norm": 24.26620864868164, + "learning_rate": 8.677932405566602e-07, + "loss": 1.0898, + "step": 877 + }, + { + "epoch": 0.8641732283464567, + "grad_norm": 21.123672485351562, + "learning_rate": 8.687872763419484e-07, + "loss": 0.7192, + "step": 878 + }, + { + "epoch": 0.8651574803149606, + "grad_norm": 19.14971923828125, + "learning_rate": 8.697813121272366e-07, + "loss": 0.7779, + "step": 879 + }, + { + "epoch": 0.8661417322834646, + "grad_norm": 16.506956100463867, + "learning_rate": 8.70775347912525e-07, + "loss": 0.9291, + "step": 880 + }, + { + "epoch": 0.8671259842519685, + "grad_norm": 15.134740829467773, + "learning_rate": 8.717693836978132e-07, + "loss": 0.6333, + "step": 881 + }, + { + "epoch": 0.8681102362204725, + "grad_norm": 15.866471290588379, + "learning_rate": 8.727634194831015e-07, + "loss": 0.9028, + "step": 882 + }, + { + "epoch": 0.8690944881889764, + "grad_norm": 19.800067901611328, + "learning_rate": 8.737574552683898e-07, + "loss": 0.9904, + "step": 883 + }, + { + "epoch": 0.8700787401574803, + "grad_norm": 13.108041763305664, + "learning_rate": 8.74751491053678e-07, + "loss": 0.5775, + "step": 884 + }, + { + "epoch": 0.8710629921259843, + "grad_norm": 23.06117820739746, + "learning_rate": 8.757455268389663e-07, + "loss": 0.9783, + "step": 885 + }, + { + "epoch": 0.8720472440944882, + "grad_norm": 16.812610626220703, + "learning_rate": 8.767395626242545e-07, + "loss": 1.111, + "step": 886 + }, + { + "epoch": 0.8730314960629921, + "grad_norm": 18.184879302978516, + "learning_rate": 8.77733598409543e-07, + "loss": 1.0398, + "step": 887 + }, + { + "epoch": 0.8740157480314961, + "grad_norm": 16.649301528930664, + "learning_rate": 8.787276341948312e-07, + "loss": 0.742, + "step": 888 + }, + { + "epoch": 0.875, + "grad_norm": 14.830896377563477, + "learning_rate": 8.797216699801194e-07, + "loss": 0.7606, + "step": 889 + }, + { + "epoch": 0.8759842519685039, + "grad_norm": 13.571001052856445, + "learning_rate": 8.807157057654078e-07, + "loss": 0.6941, + "step": 890 + }, + { + "epoch": 0.8769685039370079, + "grad_norm": 15.108964920043945, + "learning_rate": 8.81709741550696e-07, + "loss": 1.0219, + "step": 891 + }, + { + "epoch": 0.8779527559055118, + "grad_norm": 14.814958572387695, + "learning_rate": 8.827037773359842e-07, + "loss": 0.8949, + "step": 892 + }, + { + "epoch": 0.8789370078740157, + "grad_norm": 16.229421615600586, + "learning_rate": 8.836978131212726e-07, + "loss": 1.0013, + "step": 893 + }, + { + "epoch": 0.8799212598425197, + "grad_norm": 17.339427947998047, + "learning_rate": 8.846918489065608e-07, + "loss": 1.1122, + "step": 894 + }, + { + "epoch": 0.8809055118110236, + "grad_norm": 17.463390350341797, + "learning_rate": 8.856858846918491e-07, + "loss": 1.1711, + "step": 895 + }, + { + "epoch": 0.8818897637795275, + "grad_norm": 14.053199768066406, + "learning_rate": 8.866799204771373e-07, + "loss": 0.7457, + "step": 896 + }, + { + "epoch": 0.8828740157480315, + "grad_norm": 13.493846893310547, + "learning_rate": 8.876739562624256e-07, + "loss": 0.5431, + "step": 897 + }, + { + "epoch": 0.8838582677165354, + "grad_norm": 15.312845230102539, + "learning_rate": 8.886679920477139e-07, + "loss": 0.8159, + "step": 898 + }, + { + "epoch": 0.8848425196850394, + "grad_norm": 20.35736083984375, + "learning_rate": 8.896620278330021e-07, + "loss": 1.6812, + "step": 899 + }, + { + "epoch": 0.8858267716535433, + "grad_norm": 17.09781265258789, + "learning_rate": 8.906560636182905e-07, + "loss": 0.8163, + "step": 900 + }, + { + "epoch": 0.8868110236220472, + "grad_norm": 15.29171371459961, + "learning_rate": 8.916500994035787e-07, + "loss": 0.9315, + "step": 901 + }, + { + "epoch": 0.8877952755905512, + "grad_norm": 12.617769241333008, + "learning_rate": 8.926441351888669e-07, + "loss": 0.883, + "step": 902 + }, + { + "epoch": 0.8887795275590551, + "grad_norm": 15.941761016845703, + "learning_rate": 8.936381709741552e-07, + "loss": 0.8014, + "step": 903 + }, + { + "epoch": 0.889763779527559, + "grad_norm": 17.682308197021484, + "learning_rate": 8.946322067594435e-07, + "loss": 1.4277, + "step": 904 + }, + { + "epoch": 0.890748031496063, + "grad_norm": 12.226950645446777, + "learning_rate": 8.956262425447318e-07, + "loss": 0.7507, + "step": 905 + }, + { + "epoch": 0.8917322834645669, + "grad_norm": 14.347071647644043, + "learning_rate": 8.9662027833002e-07, + "loss": 0.9757, + "step": 906 + }, + { + "epoch": 0.8927165354330708, + "grad_norm": 12.406197547912598, + "learning_rate": 8.976143141153082e-07, + "loss": 0.6366, + "step": 907 + }, + { + "epoch": 0.8937007874015748, + "grad_norm": 14.094864845275879, + "learning_rate": 8.986083499005966e-07, + "loss": 0.7567, + "step": 908 + }, + { + "epoch": 0.8946850393700787, + "grad_norm": 14.407698631286621, + "learning_rate": 8.996023856858848e-07, + "loss": 0.7853, + "step": 909 + }, + { + "epoch": 0.8956692913385826, + "grad_norm": 12.980844497680664, + "learning_rate": 9.00596421471173e-07, + "loss": 0.7611, + "step": 910 + }, + { + "epoch": 0.8966535433070866, + "grad_norm": 14.911482810974121, + "learning_rate": 9.015904572564614e-07, + "loss": 0.8417, + "step": 911 + }, + { + "epoch": 0.8976377952755905, + "grad_norm": 11.791304588317871, + "learning_rate": 9.025844930417496e-07, + "loss": 0.5825, + "step": 912 + }, + { + "epoch": 0.8986220472440944, + "grad_norm": 18.612478256225586, + "learning_rate": 9.035785288270379e-07, + "loss": 1.0816, + "step": 913 + }, + { + "epoch": 0.8996062992125984, + "grad_norm": 17.988069534301758, + "learning_rate": 9.045725646123262e-07, + "loss": 1.1768, + "step": 914 + }, + { + "epoch": 0.9005905511811023, + "grad_norm": 15.855648040771484, + "learning_rate": 9.055666003976144e-07, + "loss": 0.4894, + "step": 915 + }, + { + "epoch": 0.9015748031496063, + "grad_norm": 18.538938522338867, + "learning_rate": 9.065606361829028e-07, + "loss": 0.8739, + "step": 916 + }, + { + "epoch": 0.9025590551181102, + "grad_norm": 19.68818473815918, + "learning_rate": 9.07554671968191e-07, + "loss": 1.0047, + "step": 917 + }, + { + "epoch": 0.9035433070866141, + "grad_norm": 17.291460037231445, + "learning_rate": 9.085487077534794e-07, + "loss": 0.7733, + "step": 918 + }, + { + "epoch": 0.9035433070866141, + "eval_Qnli-dev_cosine_accuracy": 0.681640625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8324013948440552, + "eval_Qnli-dev_cosine_ap": 0.7067569103812171, + "eval_Qnli-dev_cosine_f1": 0.6714801444043321, + "eval_Qnli-dev_cosine_f1_threshold": 0.7745975255966187, + "eval_Qnli-dev_cosine_precision": 0.5849056603773585, + "eval_Qnli-dev_cosine_recall": 0.788135593220339, + "eval_Qnli-dev_dot_accuracy": 0.681640625, + "eval_Qnli-dev_dot_accuracy_threshold": 639.7288818359375, + "eval_Qnli-dev_dot_ap": 0.7069071577235138, + "eval_Qnli-dev_dot_f1": 0.6714801444043321, + "eval_Qnli-dev_dot_f1_threshold": 595.2959594726562, + "eval_Qnli-dev_dot_precision": 0.5849056603773585, + "eval_Qnli-dev_dot_recall": 0.788135593220339, + "eval_Qnli-dev_euclidean_accuracy": 0.681640625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.050247192382812, + "eval_Qnli-dev_euclidean_ap": 0.7067598336380498, + "eval_Qnli-dev_euclidean_f1": 0.6714801444043321, + "eval_Qnli-dev_euclidean_f1_threshold": 18.613271713256836, + "eval_Qnli-dev_euclidean_precision": 0.5849056603773585, + "eval_Qnli-dev_euclidean_recall": 0.788135593220339, + "eval_Qnli-dev_manhattan_accuracy": 0.68359375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 351.75384521484375, + "eval_Qnli-dev_manhattan_ap": 0.70447128312057, + "eval_Qnli-dev_manhattan_f1": 0.6678082191780822, + "eval_Qnli-dev_manhattan_f1_threshold": 417.1152038574219, + "eval_Qnli-dev_manhattan_precision": 0.5603448275862069, + "eval_Qnli-dev_manhattan_recall": 0.826271186440678, + "eval_Qnli-dev_max_accuracy": 0.68359375, + "eval_Qnli-dev_max_accuracy_threshold": 639.7288818359375, + "eval_Qnli-dev_max_ap": 0.7069071577235138, + "eval_Qnli-dev_max_f1": 0.6714801444043321, + "eval_Qnli-dev_max_f1_threshold": 595.2959594726562, + "eval_Qnli-dev_max_precision": 0.5849056603773585, + "eval_Qnli-dev_max_recall": 0.826271186440678, + "eval_allNLI-dev_cosine_accuracy": 0.69921875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9117722511291504, + "eval_allNLI-dev_cosine_ap": 0.5121141282155157, + "eval_allNLI-dev_cosine_f1": 0.5527426160337552, + "eval_allNLI-dev_cosine_f1_threshold": 0.8372529745101929, + "eval_allNLI-dev_cosine_precision": 0.43521594684385384, + "eval_allNLI-dev_cosine_recall": 0.7572254335260116, + "eval_allNLI-dev_dot_accuracy": 0.69921875, + "eval_allNLI-dev_dot_accuracy_threshold": 700.7924194335938, + "eval_allNLI-dev_dot_ap": 0.5123382967410339, + "eval_allNLI-dev_dot_f1": 0.5521472392638037, + "eval_allNLI-dev_dot_f1_threshold": 640.079345703125, + "eval_allNLI-dev_dot_precision": 0.4272151898734177, + "eval_allNLI-dev_dot_recall": 0.7803468208092486, + "eval_allNLI-dev_euclidean_accuracy": 0.69921875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.64576244354248, + "eval_allNLI-dev_euclidean_ap": 0.5121089719747234, + "eval_allNLI-dev_euclidean_f1": 0.5527426160337552, + "eval_allNLI-dev_euclidean_f1_threshold": 15.815890312194824, + "eval_allNLI-dev_euclidean_precision": 0.43521594684385384, + "eval_allNLI-dev_euclidean_recall": 0.7572254335260116, + "eval_allNLI-dev_manhattan_accuracy": 0.6953125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 251.29330444335938, + "eval_allNLI-dev_manhattan_ap": 0.5106978709472474, + "eval_allNLI-dev_manhattan_f1": 0.5523012552301255, + "eval_allNLI-dev_manhattan_f1_threshold": 345.16973876953125, + "eval_allNLI-dev_manhattan_precision": 0.43278688524590164, + "eval_allNLI-dev_manhattan_recall": 0.7630057803468208, + "eval_allNLI-dev_max_accuracy": 0.69921875, + "eval_allNLI-dev_max_accuracy_threshold": 700.7924194335938, + "eval_allNLI-dev_max_ap": 0.5123382967410339, + "eval_allNLI-dev_max_f1": 0.5527426160337552, + "eval_allNLI-dev_max_f1_threshold": 640.079345703125, + "eval_allNLI-dev_max_precision": 0.43521594684385384, + "eval_allNLI-dev_max_recall": 0.7803468208092486, + "eval_runtime": 1.2191, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.7069071577235138, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.5987576250918593, + "eval_sts-test_pearson_dot": 0.5986941531022862, + "eval_sts-test_pearson_euclidean": 0.6164655142144211, + "eval_sts-test_pearson_manhattan": 0.616736762468741, + "eval_sts-test_pearson_max": 0.616736762468741, + "eval_sts-test_spearman_cosine": 0.6115360111960261, + "eval_sts-test_spearman_dot": 0.6112811080111406, + "eval_sts-test_spearman_euclidean": 0.6115426343827118, + "eval_sts-test_spearman_manhattan": 0.6125215234743874, + "eval_sts-test_spearman_max": 0.6125215234743874, + "step": 918 + }, + { + "epoch": 0.9045275590551181, + "grad_norm": 23.673276901245117, + "learning_rate": 9.095427435387676e-07, + "loss": 1.1274, + "step": 919 + }, + { + "epoch": 0.905511811023622, + "grad_norm": 21.368350982666016, + "learning_rate": 9.105367793240558e-07, + "loss": 1.5414, + "step": 920 + }, + { + "epoch": 0.906496062992126, + "grad_norm": 20.24589729309082, + "learning_rate": 9.115308151093442e-07, + "loss": 1.0839, + "step": 921 + }, + { + "epoch": 0.90748031496063, + "grad_norm": 14.09628677368164, + "learning_rate": 9.125248508946324e-07, + "loss": 0.5004, + "step": 922 + }, + { + "epoch": 0.9084645669291339, + "grad_norm": 13.60596752166748, + "learning_rate": 9.135188866799206e-07, + "loss": 0.8282, + "step": 923 + }, + { + "epoch": 0.9094488188976378, + "grad_norm": 13.115971565246582, + "learning_rate": 9.145129224652089e-07, + "loss": 0.8041, + "step": 924 + }, + { + "epoch": 0.9104330708661418, + "grad_norm": 13.023757934570312, + "learning_rate": 9.155069582504972e-07, + "loss": 0.6129, + "step": 925 + }, + { + "epoch": 0.9114173228346457, + "grad_norm": 18.587055206298828, + "learning_rate": 9.165009940357855e-07, + "loss": 0.8697, + "step": 926 + }, + { + "epoch": 0.9124015748031497, + "grad_norm": 15.160517692565918, + "learning_rate": 9.174950298210737e-07, + "loss": 0.7138, + "step": 927 + }, + { + "epoch": 0.9133858267716536, + "grad_norm": 15.338627815246582, + "learning_rate": 9.18489065606362e-07, + "loss": 1.0387, + "step": 928 + }, + { + "epoch": 0.9143700787401575, + "grad_norm": 13.919590950012207, + "learning_rate": 9.194831013916503e-07, + "loss": 0.6024, + "step": 929 + }, + { + "epoch": 0.9153543307086615, + "grad_norm": 12.949687957763672, + "learning_rate": 9.204771371769385e-07, + "loss": 0.6641, + "step": 930 + }, + { + "epoch": 0.9163385826771654, + "grad_norm": 17.4031925201416, + "learning_rate": 9.214711729622268e-07, + "loss": 0.8028, + "step": 931 + }, + { + "epoch": 0.9173228346456693, + "grad_norm": 16.457807540893555, + "learning_rate": 9.224652087475151e-07, + "loss": 0.8097, + "step": 932 + }, + { + "epoch": 0.9183070866141733, + "grad_norm": 15.084158897399902, + "learning_rate": 9.234592445328033e-07, + "loss": 0.6162, + "step": 933 + }, + { + "epoch": 0.9192913385826772, + "grad_norm": 11.633761405944824, + "learning_rate": 9.244532803180916e-07, + "loss": 0.4264, + "step": 934 + }, + { + "epoch": 0.9202755905511811, + "grad_norm": 13.683588027954102, + "learning_rate": 9.254473161033799e-07, + "loss": 0.3983, + "step": 935 + }, + { + "epoch": 0.9212598425196851, + "grad_norm": 16.877418518066406, + "learning_rate": 9.264413518886681e-07, + "loss": 0.7519, + "step": 936 + }, + { + "epoch": 0.922244094488189, + "grad_norm": 17.437131881713867, + "learning_rate": 9.274353876739564e-07, + "loss": 1.0682, + "step": 937 + }, + { + "epoch": 0.9232283464566929, + "grad_norm": 14.597518920898438, + "learning_rate": 9.284294234592446e-07, + "loss": 0.6805, + "step": 938 + }, + { + "epoch": 0.9242125984251969, + "grad_norm": 21.670108795166016, + "learning_rate": 9.29423459244533e-07, + "loss": 1.1305, + "step": 939 + }, + { + "epoch": 0.9251968503937008, + "grad_norm": 21.508169174194336, + "learning_rate": 9.304174950298212e-07, + "loss": 1.5805, + "step": 940 + }, + { + "epoch": 0.9261811023622047, + "grad_norm": 14.561578750610352, + "learning_rate": 9.314115308151094e-07, + "loss": 0.7031, + "step": 941 + }, + { + "epoch": 0.9271653543307087, + "grad_norm": 16.739912033081055, + "learning_rate": 9.324055666003978e-07, + "loss": 0.6847, + "step": 942 + }, + { + "epoch": 0.9281496062992126, + "grad_norm": 9.454686164855957, + "learning_rate": 9.33399602385686e-07, + "loss": 0.3777, + "step": 943 + }, + { + "epoch": 0.9291338582677166, + "grad_norm": 13.690561294555664, + "learning_rate": 9.343936381709743e-07, + "loss": 0.7438, + "step": 944 + }, + { + "epoch": 0.9301181102362205, + "grad_norm": 16.093313217163086, + "learning_rate": 9.353876739562625e-07, + "loss": 1.1754, + "step": 945 + }, + { + "epoch": 0.9311023622047244, + "grad_norm": 14.520867347717285, + "learning_rate": 9.363817097415509e-07, + "loss": 0.6602, + "step": 946 + }, + { + "epoch": 0.9320866141732284, + "grad_norm": 17.03354263305664, + "learning_rate": 9.373757455268392e-07, + "loss": 1.2893, + "step": 947 + }, + { + "epoch": 0.9330708661417323, + "grad_norm": 12.835516929626465, + "learning_rate": 9.383697813121274e-07, + "loss": 0.8315, + "step": 948 + }, + { + "epoch": 0.9340551181102362, + "grad_norm": 16.730058670043945, + "learning_rate": 9.393638170974157e-07, + "loss": 1.0961, + "step": 949 + }, + { + "epoch": 0.9350393700787402, + "grad_norm": 13.361641883850098, + "learning_rate": 9.40357852882704e-07, + "loss": 0.7657, + "step": 950 + }, + { + "epoch": 0.9360236220472441, + "grad_norm": 15.310640335083008, + "learning_rate": 9.413518886679922e-07, + "loss": 0.8915, + "step": 951 + }, + { + "epoch": 0.937007874015748, + "grad_norm": 12.999738693237305, + "learning_rate": 9.423459244532805e-07, + "loss": 0.6784, + "step": 952 + }, + { + "epoch": 0.937992125984252, + "grad_norm": 14.18407154083252, + "learning_rate": 9.433399602385688e-07, + "loss": 0.6772, + "step": 953 + }, + { + "epoch": 0.9389763779527559, + "grad_norm": 15.605731964111328, + "learning_rate": 9.44333996023857e-07, + "loss": 0.9102, + "step": 954 + }, + { + "epoch": 0.9399606299212598, + "grad_norm": 15.650123596191406, + "learning_rate": 9.453280318091453e-07, + "loss": 0.8083, + "step": 955 + }, + { + "epoch": 0.9409448818897638, + "grad_norm": 12.516120910644531, + "learning_rate": 9.463220675944336e-07, + "loss": 0.5966, + "step": 956 + }, + { + "epoch": 0.9419291338582677, + "grad_norm": 15.975485801696777, + "learning_rate": 9.473161033797219e-07, + "loss": 1.06, + "step": 957 + }, + { + "epoch": 0.9429133858267716, + "grad_norm": 16.759889602661133, + "learning_rate": 9.483101391650101e-07, + "loss": 1.0412, + "step": 958 + }, + { + "epoch": 0.9438976377952756, + "grad_norm": 16.401609420776367, + "learning_rate": 9.493041749502983e-07, + "loss": 0.9467, + "step": 959 + }, + { + "epoch": 0.9448818897637795, + "grad_norm": 15.220511436462402, + "learning_rate": 9.502982107355867e-07, + "loss": 0.847, + "step": 960 + }, + { + "epoch": 0.9458661417322834, + "grad_norm": 13.838984489440918, + "learning_rate": 9.512922465208749e-07, + "loss": 0.875, + "step": 961 + }, + { + "epoch": 0.9468503937007874, + "grad_norm": 15.328348159790039, + "learning_rate": 9.522862823061631e-07, + "loss": 0.8303, + "step": 962 + }, + { + "epoch": 0.9478346456692913, + "grad_norm": 13.95258903503418, + "learning_rate": 9.532803180914515e-07, + "loss": 0.8641, + "step": 963 + }, + { + "epoch": 0.9488188976377953, + "grad_norm": 10.344456672668457, + "learning_rate": 9.542743538767396e-07, + "loss": 0.6343, + "step": 964 + }, + { + "epoch": 0.9498031496062992, + "grad_norm": 11.78313159942627, + "learning_rate": 9.55268389662028e-07, + "loss": 0.7566, + "step": 965 + }, + { + "epoch": 0.9507874015748031, + "grad_norm": 15.270023345947266, + "learning_rate": 9.562624254473162e-07, + "loss": 0.9767, + "step": 966 + }, + { + "epoch": 0.9517716535433071, + "grad_norm": 12.159958839416504, + "learning_rate": 9.572564612326044e-07, + "loss": 0.7518, + "step": 967 + }, + { + "epoch": 0.952755905511811, + "grad_norm": 14.247990608215332, + "learning_rate": 9.582504970178928e-07, + "loss": 0.5381, + "step": 968 + }, + { + "epoch": 0.9537401574803149, + "grad_norm": 16.251956939697266, + "learning_rate": 9.59244532803181e-07, + "loss": 0.9959, + "step": 969 + }, + { + "epoch": 0.9547244094488189, + "grad_norm": 18.743146896362305, + "learning_rate": 9.602385685884694e-07, + "loss": 0.8374, + "step": 970 + }, + { + "epoch": 0.9557086614173228, + "grad_norm": 11.866987228393555, + "learning_rate": 9.612326043737576e-07, + "loss": 0.5459, + "step": 971 + }, + { + "epoch": 0.9566929133858267, + "grad_norm": 14.472687721252441, + "learning_rate": 9.622266401590458e-07, + "loss": 0.6633, + "step": 972 + }, + { + "epoch": 0.9576771653543307, + "grad_norm": 15.333280563354492, + "learning_rate": 9.632206759443342e-07, + "loss": 0.9526, + "step": 973 + }, + { + "epoch": 0.9586614173228346, + "grad_norm": 15.25340461730957, + "learning_rate": 9.642147117296224e-07, + "loss": 0.8487, + "step": 974 + }, + { + "epoch": 0.9596456692913385, + "grad_norm": 11.816839218139648, + "learning_rate": 9.652087475149109e-07, + "loss": 0.6572, + "step": 975 + }, + { + "epoch": 0.9606299212598425, + "grad_norm": 17.762727737426758, + "learning_rate": 9.66202783300199e-07, + "loss": 0.8039, + "step": 976 + }, + { + "epoch": 0.9616141732283464, + "grad_norm": 13.454944610595703, + "learning_rate": 9.671968190854873e-07, + "loss": 0.6213, + "step": 977 + }, + { + "epoch": 0.9625984251968503, + "grad_norm": 12.26315689086914, + "learning_rate": 9.681908548707755e-07, + "loss": 0.5483, + "step": 978 + }, + { + "epoch": 0.9635826771653543, + "grad_norm": 10.695785522460938, + "learning_rate": 9.691848906560639e-07, + "loss": 0.516, + "step": 979 + }, + { + "epoch": 0.9645669291338582, + "grad_norm": 13.537649154663086, + "learning_rate": 9.70178926441352e-07, + "loss": 0.8891, + "step": 980 + }, + { + "epoch": 0.9655511811023622, + "grad_norm": 14.947914123535156, + "learning_rate": 9.711729622266403e-07, + "loss": 0.7904, + "step": 981 + }, + { + "epoch": 0.9665354330708661, + "grad_norm": 19.540485382080078, + "learning_rate": 9.721669980119285e-07, + "loss": 1.2282, + "step": 982 + }, + { + "epoch": 0.96751968503937, + "grad_norm": 20.05003547668457, + "learning_rate": 9.731610337972169e-07, + "loss": 1.0609, + "step": 983 + }, + { + "epoch": 0.968503937007874, + "grad_norm": 15.105903625488281, + "learning_rate": 9.74155069582505e-07, + "loss": 0.8063, + "step": 984 + }, + { + "epoch": 0.969488188976378, + "grad_norm": 11.5243501663208, + "learning_rate": 9.751491053677933e-07, + "loss": 0.5294, + "step": 985 + }, + { + "epoch": 0.9704724409448819, + "grad_norm": 11.790663719177246, + "learning_rate": 9.761431411530817e-07, + "loss": 0.7394, + "step": 986 + }, + { + "epoch": 0.9714566929133859, + "grad_norm": 13.63713264465332, + "learning_rate": 9.771371769383699e-07, + "loss": 0.5749, + "step": 987 + }, + { + "epoch": 0.9724409448818898, + "grad_norm": 16.178760528564453, + "learning_rate": 9.781312127236583e-07, + "loss": 0.9125, + "step": 988 + }, + { + "epoch": 0.9734251968503937, + "grad_norm": 17.45037078857422, + "learning_rate": 9.791252485089465e-07, + "loss": 0.961, + "step": 989 + }, + { + "epoch": 0.9744094488188977, + "grad_norm": 15.806575775146484, + "learning_rate": 9.801192842942347e-07, + "loss": 0.5947, + "step": 990 + }, + { + "epoch": 0.9753937007874016, + "grad_norm": 15.190784454345703, + "learning_rate": 9.811133200795231e-07, + "loss": 0.6246, + "step": 991 + }, + { + "epoch": 0.9763779527559056, + "grad_norm": 13.295624732971191, + "learning_rate": 9.821073558648113e-07, + "loss": 0.6492, + "step": 992 + }, + { + "epoch": 0.9773622047244095, + "grad_norm": 18.87274932861328, + "learning_rate": 9.831013916500995e-07, + "loss": 1.0508, + "step": 993 + }, + { + "epoch": 0.9783464566929134, + "grad_norm": 18.1697940826416, + "learning_rate": 9.84095427435388e-07, + "loss": 0.7613, + "step": 994 + }, + { + "epoch": 0.9793307086614174, + "grad_norm": 14.127570152282715, + "learning_rate": 9.850894632206761e-07, + "loss": 0.706, + "step": 995 + }, + { + "epoch": 0.9803149606299213, + "grad_norm": 15.612168312072754, + "learning_rate": 9.860834990059645e-07, + "loss": 0.9458, + "step": 996 + }, + { + "epoch": 0.9812992125984252, + "grad_norm": 13.476499557495117, + "learning_rate": 9.870775347912527e-07, + "loss": 0.836, + "step": 997 + }, + { + "epoch": 0.9822834645669292, + "grad_norm": 11.830811500549316, + "learning_rate": 9.88071570576541e-07, + "loss": 0.7472, + "step": 998 + }, + { + "epoch": 0.9832677165354331, + "grad_norm": 15.934710502624512, + "learning_rate": 9.890656063618291e-07, + "loss": 0.9071, + "step": 999 + }, + { + "epoch": 0.984251968503937, + "grad_norm": 14.55718994140625, + "learning_rate": 9.900596421471175e-07, + "loss": 1.0935, + "step": 1000 + }, + { + "epoch": 0.985236220472441, + "grad_norm": 10.328267097473145, + "learning_rate": 9.910536779324057e-07, + "loss": 0.5822, + "step": 1001 + }, + { + "epoch": 0.9862204724409449, + "grad_norm": 13.852025032043457, + "learning_rate": 9.92047713717694e-07, + "loss": 0.6596, + "step": 1002 + }, + { + "epoch": 0.9872047244094488, + "grad_norm": 9.049901962280273, + "learning_rate": 9.930417495029821e-07, + "loss": 0.5536, + "step": 1003 + }, + { + "epoch": 0.9881889763779528, + "grad_norm": 16.871662139892578, + "learning_rate": 9.940357852882705e-07, + "loss": 1.0741, + "step": 1004 + }, + { + "epoch": 0.9891732283464567, + "grad_norm": 10.545391082763672, + "learning_rate": 9.950298210735587e-07, + "loss": 0.4586, + "step": 1005 + }, + { + "epoch": 0.9901574803149606, + "grad_norm": 17.480083465576172, + "learning_rate": 9.96023856858847e-07, + "loss": 0.9633, + "step": 1006 + }, + { + "epoch": 0.9911417322834646, + "grad_norm": 12.648763656616211, + "learning_rate": 9.970178926441354e-07, + "loss": 0.8739, + "step": 1007 + }, + { + "epoch": 0.9921259842519685, + "grad_norm": 15.744301795959473, + "learning_rate": 9.980119284294236e-07, + "loss": 0.665, + "step": 1008 + }, + { + "epoch": 0.9931102362204725, + "grad_norm": 17.54332160949707, + "learning_rate": 9.99005964214712e-07, + "loss": 0.9486, + "step": 1009 + }, + { + "epoch": 0.9940944881889764, + "grad_norm": 13.077927589416504, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.5323, + "step": 1010 + }, + { + "epoch": 0.9950787401574803, + "grad_norm": 12.237645149230957, + "learning_rate": 9.999994102361802e-07, + "loss": 0.6006, + "step": 1011 + }, + { + "epoch": 0.9960629921259843, + "grad_norm": 10.984200477600098, + "learning_rate": 9.999976409461164e-07, + "loss": 0.4568, + "step": 1012 + }, + { + "epoch": 0.9970472440944882, + "grad_norm": 14.41800308227539, + "learning_rate": 9.999946921339963e-07, + "loss": 0.7544, + "step": 1013 + }, + { + "epoch": 0.9980314960629921, + "grad_norm": 16.392780303955078, + "learning_rate": 9.999905638068e-07, + "loss": 1.2437, + "step": 1014 + }, + { + "epoch": 0.9990157480314961, + "grad_norm": 19.318571090698242, + "learning_rate": 9.999852559742985e-07, + "loss": 0.8985, + "step": 1015 + }, + { + "epoch": 1.0, + "grad_norm": 11.829497337341309, + "learning_rate": 9.999787686490555e-07, + "loss": 0.3179, + "step": 1016 + }, + { + "epoch": 1.000984251968504, + "grad_norm": 17.870615005493164, + "learning_rate": 9.99971101846426e-07, + "loss": 1.2519, + "step": 1017 + }, + { + "epoch": 1.0019685039370079, + "grad_norm": 16.654064178466797, + "learning_rate": 9.999622555845573e-07, + "loss": 0.9775, + "step": 1018 + }, + { + "epoch": 1.0029527559055118, + "grad_norm": 14.995962142944336, + "learning_rate": 9.999522298843875e-07, + "loss": 0.8967, + "step": 1019 + }, + { + "epoch": 1.0039370078740157, + "grad_norm": 11.28644847869873, + "learning_rate": 9.99941024769647e-07, + "loss": 0.5918, + "step": 1020 + }, + { + "epoch": 1.0049212598425197, + "grad_norm": 12.760509490966797, + "learning_rate": 9.999286402668578e-07, + "loss": 0.5798, + "step": 1021 + }, + { + "epoch": 1.0059055118110236, + "grad_norm": 12.091622352600098, + "learning_rate": 9.999150764053334e-07, + "loss": 0.5934, + "step": 1022 + }, + { + "epoch": 1.0068897637795275, + "grad_norm": 16.093765258789062, + "learning_rate": 9.999003332171784e-07, + "loss": 0.6872, + "step": 1023 + }, + { + "epoch": 1.0078740157480315, + "grad_norm": 19.56390380859375, + "learning_rate": 9.998844107372894e-07, + "loss": 0.9535, + "step": 1024 + }, + { + "epoch": 1.0088582677165354, + "grad_norm": 13.822117805480957, + "learning_rate": 9.99867309003354e-07, + "loss": 0.4967, + "step": 1025 + }, + { + "epoch": 1.0098425196850394, + "grad_norm": 19.477834701538086, + "learning_rate": 9.998490280558511e-07, + "loss": 1.116, + "step": 1026 + }, + { + "epoch": 1.0108267716535433, + "grad_norm": 18.705852508544922, + "learning_rate": 9.998295679380506e-07, + "loss": 1.0446, + "step": 1027 + }, + { + "epoch": 1.0118110236220472, + "grad_norm": 11.78742790222168, + "learning_rate": 9.998089286960135e-07, + "loss": 0.5189, + "step": 1028 + }, + { + "epoch": 1.0127952755905512, + "grad_norm": 17.845172882080078, + "learning_rate": 9.99787110378592e-07, + "loss": 0.8272, + "step": 1029 + }, + { + "epoch": 1.013779527559055, + "grad_norm": 10.446756362915039, + "learning_rate": 9.997641130374284e-07, + "loss": 0.4901, + "step": 1030 + }, + { + "epoch": 1.014763779527559, + "grad_norm": 15.8695650100708, + "learning_rate": 9.997399367269565e-07, + "loss": 0.8926, + "step": 1031 + }, + { + "epoch": 1.015748031496063, + "grad_norm": 11.739765167236328, + "learning_rate": 9.997145815044003e-07, + "loss": 0.5543, + "step": 1032 + }, + { + "epoch": 1.016732283464567, + "grad_norm": 7.6339945793151855, + "learning_rate": 9.99688047429774e-07, + "loss": 0.2579, + "step": 1033 + }, + { + "epoch": 1.0177165354330708, + "grad_norm": 14.852460861206055, + "learning_rate": 9.996603345658824e-07, + "loss": 0.865, + "step": 1034 + }, + { + "epoch": 1.0187007874015748, + "grad_norm": 15.955395698547363, + "learning_rate": 9.996314429783204e-07, + "loss": 1.0559, + "step": 1035 + }, + { + "epoch": 1.0196850393700787, + "grad_norm": 13.856797218322754, + "learning_rate": 9.99601372735473e-07, + "loss": 0.714, + "step": 1036 + }, + { + "epoch": 1.0206692913385826, + "grad_norm": 13.131881713867188, + "learning_rate": 9.995701239085143e-07, + "loss": 0.5185, + "step": 1037 + }, + { + "epoch": 1.0216535433070866, + "grad_norm": 20.843055725097656, + "learning_rate": 9.99537696571409e-07, + "loss": 1.3209, + "step": 1038 + }, + { + "epoch": 1.0226377952755905, + "grad_norm": 10.4310302734375, + "learning_rate": 9.995040908009107e-07, + "loss": 0.3696, + "step": 1039 + }, + { + "epoch": 1.0236220472440944, + "grad_norm": 12.129494667053223, + "learning_rate": 9.994693066765622e-07, + "loss": 0.4839, + "step": 1040 + }, + { + "epoch": 1.0246062992125984, + "grad_norm": 13.931997299194336, + "learning_rate": 9.994333442806959e-07, + "loss": 0.7694, + "step": 1041 + }, + { + "epoch": 1.0255905511811023, + "grad_norm": 22.863691329956055, + "learning_rate": 9.993962036984325e-07, + "loss": 1.3417, + "step": 1042 + }, + { + "epoch": 1.0265748031496063, + "grad_norm": 16.12286376953125, + "learning_rate": 9.99357885017682e-07, + "loss": 0.8292, + "step": 1043 + }, + { + "epoch": 1.0275590551181102, + "grad_norm": 9.64384651184082, + "learning_rate": 9.993183883291425e-07, + "loss": 0.3835, + "step": 1044 + }, + { + "epoch": 1.0285433070866141, + "grad_norm": 10.787782669067383, + "learning_rate": 9.992777137263007e-07, + "loss": 0.3755, + "step": 1045 + }, + { + "epoch": 1.029527559055118, + "grad_norm": 15.431944847106934, + "learning_rate": 9.99235861305431e-07, + "loss": 0.6396, + "step": 1046 + }, + { + "epoch": 1.030511811023622, + "grad_norm": 8.80448055267334, + "learning_rate": 9.991928311655956e-07, + "loss": 0.364, + "step": 1047 + }, + { + "epoch": 1.031496062992126, + "grad_norm": 17.40666961669922, + "learning_rate": 9.991486234086447e-07, + "loss": 0.9232, + "step": 1048 + }, + { + "epoch": 1.0324803149606299, + "grad_norm": 14.322542190551758, + "learning_rate": 9.991032381392158e-07, + "loss": 0.6706, + "step": 1049 + }, + { + "epoch": 1.0334645669291338, + "grad_norm": 15.621203422546387, + "learning_rate": 9.990566754647328e-07, + "loss": 0.8329, + "step": 1050 + }, + { + "epoch": 1.0344488188976377, + "grad_norm": 13.89022445678711, + "learning_rate": 9.990089354954073e-07, + "loss": 0.4093, + "step": 1051 + }, + { + "epoch": 1.0354330708661417, + "grad_norm": 17.091297149658203, + "learning_rate": 9.989600183442373e-07, + "loss": 0.7251, + "step": 1052 + }, + { + "epoch": 1.0364173228346456, + "grad_norm": 15.276006698608398, + "learning_rate": 9.98909924127007e-07, + "loss": 0.9254, + "step": 1053 + }, + { + "epoch": 1.0374015748031495, + "grad_norm": 16.77952766418457, + "learning_rate": 9.988586529622864e-07, + "loss": 0.6365, + "step": 1054 + }, + { + "epoch": 1.0383858267716535, + "grad_norm": 16.019168853759766, + "learning_rate": 9.98806204971432e-07, + "loss": 0.7136, + "step": 1055 + }, + { + "epoch": 1.0393700787401574, + "grad_norm": 13.023910522460938, + "learning_rate": 9.987525802785847e-07, + "loss": 0.6278, + "step": 1056 + }, + { + "epoch": 1.0403543307086613, + "grad_norm": 12.696486473083496, + "learning_rate": 9.986977790106715e-07, + "loss": 0.6769, + "step": 1057 + }, + { + "epoch": 1.0413385826771653, + "grad_norm": 15.542061805725098, + "learning_rate": 9.986418012974042e-07, + "loss": 0.7704, + "step": 1058 + }, + { + "epoch": 1.0423228346456692, + "grad_norm": 18.10808753967285, + "learning_rate": 9.985846472712786e-07, + "loss": 0.7701, + "step": 1059 + }, + { + "epoch": 1.0433070866141732, + "grad_norm": 16.275318145751953, + "learning_rate": 9.985263170675756e-07, + "loss": 0.9639, + "step": 1060 + }, + { + "epoch": 1.044291338582677, + "grad_norm": 16.48155975341797, + "learning_rate": 9.98466810824359e-07, + "loss": 0.7518, + "step": 1061 + }, + { + "epoch": 1.045275590551181, + "grad_norm": 15.806731224060059, + "learning_rate": 9.984061286824774e-07, + "loss": 0.7474, + "step": 1062 + }, + { + "epoch": 1.046259842519685, + "grad_norm": 19.146682739257812, + "learning_rate": 9.983442707855616e-07, + "loss": 1.2196, + "step": 1063 + }, + { + "epoch": 1.047244094488189, + "grad_norm": 16.093191146850586, + "learning_rate": 9.982812372800262e-07, + "loss": 1.1476, + "step": 1064 + }, + { + "epoch": 1.0482283464566928, + "grad_norm": 13.356393814086914, + "learning_rate": 9.982170283150678e-07, + "loss": 0.7139, + "step": 1065 + }, + { + "epoch": 1.0492125984251968, + "grad_norm": 18.03761100769043, + "learning_rate": 9.981516440426659e-07, + "loss": 1.0979, + "step": 1066 + }, + { + "epoch": 1.0501968503937007, + "grad_norm": 12.974353790283203, + "learning_rate": 9.98085084617581e-07, + "loss": 0.7113, + "step": 1067 + }, + { + "epoch": 1.0511811023622046, + "grad_norm": 11.472173690795898, + "learning_rate": 9.980173501973558e-07, + "loss": 0.6796, + "step": 1068 + }, + { + "epoch": 1.0521653543307086, + "grad_norm": 17.4272403717041, + "learning_rate": 9.97948440942314e-07, + "loss": 1.1551, + "step": 1069 + }, + { + "epoch": 1.0531496062992125, + "grad_norm": 12.593327522277832, + "learning_rate": 9.978783570155597e-07, + "loss": 0.7947, + "step": 1070 + }, + { + "epoch": 1.0541338582677164, + "grad_norm": 13.903247833251953, + "learning_rate": 9.978070985829782e-07, + "loss": 0.6783, + "step": 1071 + }, + { + "epoch": 1.0541338582677164, + "eval_Qnli-dev_cosine_accuracy": 0.666015625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8842042684555054, + "eval_Qnli-dev_cosine_ap": 0.7052019976296955, + "eval_Qnli-dev_cosine_f1": 0.6793103448275861, + "eval_Qnli-dev_cosine_f1_threshold": 0.8030812740325928, + "eval_Qnli-dev_cosine_precision": 0.5726744186046512, + "eval_Qnli-dev_cosine_recall": 0.8347457627118644, + "eval_Qnli-dev_dot_accuracy": 0.66796875, + "eval_Qnli-dev_dot_accuracy_threshold": 679.375244140625, + "eval_Qnli-dev_dot_ap": 0.7054354866961744, + "eval_Qnli-dev_dot_f1": 0.6794425087108014, + "eval_Qnli-dev_dot_f1_threshold": 618.5501708984375, + "eval_Qnli-dev_dot_precision": 0.5769230769230769, + "eval_Qnli-dev_dot_recall": 0.826271186440678, + "eval_Qnli-dev_euclidean_accuracy": 0.666015625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 13.342350006103516, + "eval_Qnli-dev_euclidean_ap": 0.7052194558904903, + "eval_Qnli-dev_euclidean_f1": 0.6793103448275861, + "eval_Qnli-dev_euclidean_f1_threshold": 17.398447036743164, + "eval_Qnli-dev_euclidean_precision": 0.5726744186046512, + "eval_Qnli-dev_euclidean_recall": 0.8347457627118644, + "eval_Qnli-dev_manhattan_accuracy": 0.669921875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 288.81317138671875, + "eval_Qnli-dev_manhattan_ap": 0.7053588266129154, + "eval_Qnli-dev_manhattan_f1": 0.6791862284820032, + "eval_Qnli-dev_manhattan_f1_threshold": 405.0858459472656, + "eval_Qnli-dev_manhattan_precision": 0.5384615384615384, + "eval_Qnli-dev_manhattan_recall": 0.9194915254237288, + "eval_Qnli-dev_max_accuracy": 0.669921875, + "eval_Qnli-dev_max_accuracy_threshold": 679.375244140625, + "eval_Qnli-dev_max_ap": 0.7054354866961744, + "eval_Qnli-dev_max_f1": 0.6794425087108014, + "eval_Qnli-dev_max_f1_threshold": 618.5501708984375, + "eval_Qnli-dev_max_precision": 0.5769230769230769, + "eval_Qnli-dev_max_recall": 0.9194915254237288, + "eval_allNLI-dev_cosine_accuracy": 0.6796875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9540110230445862, + "eval_allNLI-dev_cosine_ap": 0.4867883737362577, + "eval_allNLI-dev_cosine_f1": 0.5454545454545455, + "eval_allNLI-dev_cosine_f1_threshold": 0.8630210161209106, + "eval_allNLI-dev_cosine_precision": 0.44244604316546765, + "eval_allNLI-dev_cosine_recall": 0.7109826589595376, + "eval_allNLI-dev_dot_accuracy": 0.6796875, + "eval_allNLI-dev_dot_accuracy_threshold": 733.4241943359375, + "eval_allNLI-dev_dot_ap": 0.4866316260057906, + "eval_allNLI-dev_dot_f1": 0.5454545454545455, + "eval_allNLI-dev_dot_f1_threshold": 663.4332885742188, + "eval_allNLI-dev_dot_precision": 0.44244604316546765, + "eval_allNLI-dev_dot_recall": 0.7109826589595376, + "eval_allNLI-dev_euclidean_accuracy": 0.6796875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 8.408852577209473, + "eval_allNLI-dev_euclidean_ap": 0.48671939529879626, + "eval_allNLI-dev_euclidean_f1": 0.5454545454545455, + "eval_allNLI-dev_euclidean_f1_threshold": 14.51208782196045, + "eval_allNLI-dev_euclidean_precision": 0.44244604316546765, + "eval_allNLI-dev_euclidean_recall": 0.7109826589595376, + "eval_allNLI-dev_manhattan_accuracy": 0.677734375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 181.60472106933594, + "eval_allNLI-dev_manhattan_ap": 0.4882823795400136, + "eval_allNLI-dev_manhattan_f1": 0.550537634408602, + "eval_allNLI-dev_manhattan_f1_threshold": 319.623046875, + "eval_allNLI-dev_manhattan_precision": 0.4383561643835616, + "eval_allNLI-dev_manhattan_recall": 0.7398843930635838, + "eval_allNLI-dev_max_accuracy": 0.6796875, + "eval_allNLI-dev_max_accuracy_threshold": 733.4241943359375, + "eval_allNLI-dev_max_ap": 0.4882823795400136, + "eval_allNLI-dev_max_f1": 0.550537634408602, + "eval_allNLI-dev_max_f1_threshold": 663.4332885742188, + "eval_allNLI-dev_max_precision": 0.44244604316546765, + "eval_allNLI-dev_max_recall": 0.7398843930635838, + "eval_runtime": 1.2206, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.7054354866961744, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.6261543137722272, + "eval_sts-test_pearson_dot": 0.6259993945400585, + "eval_sts-test_pearson_euclidean": 0.6433597379437007, + "eval_sts-test_pearson_manhattan": 0.64545238834986, + "eval_sts-test_pearson_max": 0.64545238834986, + "eval_sts-test_spearman_cosine": 0.6380566379755814, + "eval_sts-test_spearman_dot": 0.6378358352518761, + "eval_sts-test_spearman_euclidean": 0.6380140347747367, + "eval_sts-test_spearman_manhattan": 0.638675279413056, + "eval_sts-test_spearman_max": 0.638675279413056, + "step": 1071 + }, + { + "epoch": 1.0551181102362204, + "grad_norm": 13.063305854797363, + "learning_rate": 9.97734665813234e-07, + "loss": 0.7295, + "step": 1072 + }, + { + "epoch": 1.0561023622047243, + "grad_norm": 12.699029922485352, + "learning_rate": 9.976610588777717e-07, + "loss": 0.4529, + "step": 1073 + }, + { + "epoch": 1.0570866141732282, + "grad_norm": 8.52369499206543, + "learning_rate": 9.975862779508149e-07, + "loss": 0.2625, + "step": 1074 + }, + { + "epoch": 1.0580708661417322, + "grad_norm": 17.706161499023438, + "learning_rate": 9.975103232093656e-07, + "loss": 1.2557, + "step": 1075 + }, + { + "epoch": 1.0590551181102361, + "grad_norm": 11.824196815490723, + "learning_rate": 9.974331948332048e-07, + "loss": 0.4826, + "step": 1076 + }, + { + "epoch": 1.06003937007874, + "grad_norm": 16.63976287841797, + "learning_rate": 9.97354893004891e-07, + "loss": 1.2309, + "step": 1077 + }, + { + "epoch": 1.061023622047244, + "grad_norm": 11.964570999145508, + "learning_rate": 9.972754179097602e-07, + "loss": 0.4545, + "step": 1078 + }, + { + "epoch": 1.0620078740157481, + "grad_norm": 14.151701927185059, + "learning_rate": 9.971947697359257e-07, + "loss": 0.6658, + "step": 1079 + }, + { + "epoch": 1.0629921259842519, + "grad_norm": 15.346922874450684, + "learning_rate": 9.971129486742777e-07, + "loss": 0.8651, + "step": 1080 + }, + { + "epoch": 1.063976377952756, + "grad_norm": 24.134613037109375, + "learning_rate": 9.970299549184814e-07, + "loss": 1.0702, + "step": 1081 + }, + { + "epoch": 1.0649606299212597, + "grad_norm": 17.18680191040039, + "learning_rate": 9.969457886649791e-07, + "loss": 1.0225, + "step": 1082 + }, + { + "epoch": 1.0659448818897639, + "grad_norm": 13.291850090026855, + "learning_rate": 9.968604501129877e-07, + "loss": 0.6952, + "step": 1083 + }, + { + "epoch": 1.0669291338582678, + "grad_norm": 15.94408893585205, + "learning_rate": 9.967739394644982e-07, + "loss": 1.0918, + "step": 1084 + }, + { + "epoch": 1.0679133858267718, + "grad_norm": 11.357185363769531, + "learning_rate": 9.966862569242773e-07, + "loss": 0.5222, + "step": 1085 + }, + { + "epoch": 1.0688976377952757, + "grad_norm": 18.47614860534668, + "learning_rate": 9.965974026998642e-07, + "loss": 1.1311, + "step": 1086 + }, + { + "epoch": 1.0698818897637796, + "grad_norm": 13.027080535888672, + "learning_rate": 9.965073770015722e-07, + "loss": 0.6464, + "step": 1087 + }, + { + "epoch": 1.0708661417322836, + "grad_norm": 13.948921203613281, + "learning_rate": 9.964161800424873e-07, + "loss": 0.801, + "step": 1088 + }, + { + "epoch": 1.0718503937007875, + "grad_norm": 10.788043022155762, + "learning_rate": 9.963238120384677e-07, + "loss": 0.3743, + "step": 1089 + }, + { + "epoch": 1.0728346456692914, + "grad_norm": 9.750899314880371, + "learning_rate": 9.962302732081432e-07, + "loss": 0.6204, + "step": 1090 + }, + { + "epoch": 1.0738188976377954, + "grad_norm": 14.213951110839844, + "learning_rate": 9.961355637729153e-07, + "loss": 0.5369, + "step": 1091 + }, + { + "epoch": 1.0748031496062993, + "grad_norm": 11.047940254211426, + "learning_rate": 9.960396839569557e-07, + "loss": 0.6895, + "step": 1092 + }, + { + "epoch": 1.0757874015748032, + "grad_norm": 15.831133842468262, + "learning_rate": 9.95942633987207e-07, + "loss": 1.2127, + "step": 1093 + }, + { + "epoch": 1.0767716535433072, + "grad_norm": 14.451669692993164, + "learning_rate": 9.95844414093381e-07, + "loss": 0.8397, + "step": 1094 + }, + { + "epoch": 1.077755905511811, + "grad_norm": 11.084488868713379, + "learning_rate": 9.957450245079588e-07, + "loss": 0.3694, + "step": 1095 + }, + { + "epoch": 1.078740157480315, + "grad_norm": 13.861403465270996, + "learning_rate": 9.9564446546619e-07, + "loss": 0.7576, + "step": 1096 + }, + { + "epoch": 1.079724409448819, + "grad_norm": 16.21560287475586, + "learning_rate": 9.955427372060928e-07, + "loss": 0.9494, + "step": 1097 + }, + { + "epoch": 1.080708661417323, + "grad_norm": 18.243650436401367, + "learning_rate": 9.954398399684515e-07, + "loss": 0.8477, + "step": 1098 + }, + { + "epoch": 1.0816929133858268, + "grad_norm": 10.780070304870605, + "learning_rate": 9.953357739968192e-07, + "loss": 0.4869, + "step": 1099 + }, + { + "epoch": 1.0826771653543308, + "grad_norm": 13.332379341125488, + "learning_rate": 9.952305395375137e-07, + "loss": 0.6753, + "step": 1100 + }, + { + "epoch": 1.0836614173228347, + "grad_norm": 13.206958770751953, + "learning_rate": 9.951241368396194e-07, + "loss": 0.7351, + "step": 1101 + }, + { + "epoch": 1.0846456692913387, + "grad_norm": 16.770225524902344, + "learning_rate": 9.950165661549856e-07, + "loss": 0.8257, + "step": 1102 + }, + { + "epoch": 1.0856299212598426, + "grad_norm": 17.56561851501465, + "learning_rate": 9.949078277382262e-07, + "loss": 1.144, + "step": 1103 + }, + { + "epoch": 1.0866141732283465, + "grad_norm": 15.408824920654297, + "learning_rate": 9.947979218467193e-07, + "loss": 0.5526, + "step": 1104 + }, + { + "epoch": 1.0875984251968505, + "grad_norm": 15.738862037658691, + "learning_rate": 9.946868487406057e-07, + "loss": 0.6686, + "step": 1105 + }, + { + "epoch": 1.0885826771653544, + "grad_norm": 15.481057167053223, + "learning_rate": 9.945746086827896e-07, + "loss": 0.7683, + "step": 1106 + }, + { + "epoch": 1.0895669291338583, + "grad_norm": 10.804634094238281, + "learning_rate": 9.94461201938937e-07, + "loss": 0.3337, + "step": 1107 + }, + { + "epoch": 1.0905511811023623, + "grad_norm": 15.124865531921387, + "learning_rate": 9.943466287774756e-07, + "loss": 0.8686, + "step": 1108 + }, + { + "epoch": 1.0915354330708662, + "grad_norm": 13.105254173278809, + "learning_rate": 9.942308894695932e-07, + "loss": 0.66, + "step": 1109 + }, + { + "epoch": 1.0925196850393701, + "grad_norm": 15.874667167663574, + "learning_rate": 9.941139842892391e-07, + "loss": 0.9329, + "step": 1110 + }, + { + "epoch": 1.093503937007874, + "grad_norm": 10.224324226379395, + "learning_rate": 9.939959135131214e-07, + "loss": 0.5129, + "step": 1111 + }, + { + "epoch": 1.094488188976378, + "grad_norm": 18.19011688232422, + "learning_rate": 9.938766774207064e-07, + "loss": 0.8907, + "step": 1112 + }, + { + "epoch": 1.095472440944882, + "grad_norm": 12.274703979492188, + "learning_rate": 9.937562762942202e-07, + "loss": 0.6305, + "step": 1113 + }, + { + "epoch": 1.0964566929133859, + "grad_norm": 8.143845558166504, + "learning_rate": 9.936347104186452e-07, + "loss": 0.3746, + "step": 1114 + }, + { + "epoch": 1.0974409448818898, + "grad_norm": 14.044692039489746, + "learning_rate": 9.935119800817215e-07, + "loss": 0.6791, + "step": 1115 + }, + { + "epoch": 1.0984251968503937, + "grad_norm": 17.596525192260742, + "learning_rate": 9.933880855739446e-07, + "loss": 1.4172, + "step": 1116 + }, + { + "epoch": 1.0994094488188977, + "grad_norm": 11.362157821655273, + "learning_rate": 9.932630271885662e-07, + "loss": 0.5184, + "step": 1117 + }, + { + "epoch": 1.1003937007874016, + "grad_norm": 9.954078674316406, + "learning_rate": 9.93136805221593e-07, + "loss": 0.4803, + "step": 1118 + }, + { + "epoch": 1.1013779527559056, + "grad_norm": 10.72042465209961, + "learning_rate": 9.930094199717848e-07, + "loss": 0.647, + "step": 1119 + }, + { + "epoch": 1.1023622047244095, + "grad_norm": 11.251803398132324, + "learning_rate": 9.92880871740656e-07, + "loss": 0.3867, + "step": 1120 + }, + { + "epoch": 1.1033464566929134, + "grad_norm": 13.162117004394531, + "learning_rate": 9.927511608324732e-07, + "loss": 0.4965, + "step": 1121 + }, + { + "epoch": 1.1043307086614174, + "grad_norm": 14.909209251403809, + "learning_rate": 9.926202875542546e-07, + "loss": 0.9508, + "step": 1122 + }, + { + "epoch": 1.1053149606299213, + "grad_norm": 17.707162857055664, + "learning_rate": 9.924882522157706e-07, + "loss": 0.709, + "step": 1123 + }, + { + "epoch": 1.1062992125984252, + "grad_norm": 18.453153610229492, + "learning_rate": 9.923550551295411e-07, + "loss": 0.9625, + "step": 1124 + }, + { + "epoch": 1.1072834645669292, + "grad_norm": 20.41254997253418, + "learning_rate": 9.922206966108367e-07, + "loss": 0.7724, + "step": 1125 + }, + { + "epoch": 1.108267716535433, + "grad_norm": 13.927225112915039, + "learning_rate": 9.920851769776761e-07, + "loss": 0.4318, + "step": 1126 + }, + { + "epoch": 1.109251968503937, + "grad_norm": 17.073375701904297, + "learning_rate": 9.919484965508276e-07, + "loss": 0.6653, + "step": 1127 + }, + { + "epoch": 1.110236220472441, + "grad_norm": 11.740160942077637, + "learning_rate": 9.918106556538057e-07, + "loss": 0.781, + "step": 1128 + }, + { + "epoch": 1.111220472440945, + "grad_norm": 14.772197723388672, + "learning_rate": 9.91671654612872e-07, + "loss": 0.6985, + "step": 1129 + }, + { + "epoch": 1.1122047244094488, + "grad_norm": 20.382341384887695, + "learning_rate": 9.91531493757035e-07, + "loss": 1.2949, + "step": 1130 + }, + { + "epoch": 1.1131889763779528, + "grad_norm": 14.498468399047852, + "learning_rate": 9.913901734180476e-07, + "loss": 0.6845, + "step": 1131 + }, + { + "epoch": 1.1141732283464567, + "grad_norm": 13.584673881530762, + "learning_rate": 9.912476939304069e-07, + "loss": 0.7022, + "step": 1132 + }, + { + "epoch": 1.1151574803149606, + "grad_norm": 15.283327102661133, + "learning_rate": 9.911040556313543e-07, + "loss": 0.8645, + "step": 1133 + }, + { + "epoch": 1.1161417322834646, + "grad_norm": 10.71398639678955, + "learning_rate": 9.909592588608738e-07, + "loss": 0.5331, + "step": 1134 + }, + { + "epoch": 1.1171259842519685, + "grad_norm": 11.05366039276123, + "learning_rate": 9.908133039616911e-07, + "loss": 0.5636, + "step": 1135 + }, + { + "epoch": 1.1181102362204725, + "grad_norm": 11.281898498535156, + "learning_rate": 9.90666191279274e-07, + "loss": 0.4511, + "step": 1136 + }, + { + "epoch": 1.1190944881889764, + "grad_norm": 15.47266674041748, + "learning_rate": 9.905179211618296e-07, + "loss": 0.605, + "step": 1137 + }, + { + "epoch": 1.1200787401574803, + "grad_norm": 15.861746788024902, + "learning_rate": 9.903684939603054e-07, + "loss": 0.6367, + "step": 1138 + }, + { + "epoch": 1.1210629921259843, + "grad_norm": 16.15589141845703, + "learning_rate": 9.902179100283875e-07, + "loss": 1.1988, + "step": 1139 + }, + { + "epoch": 1.1220472440944882, + "grad_norm": 12.1558256149292, + "learning_rate": 9.900661697224995e-07, + "loss": 0.5973, + "step": 1140 + }, + { + "epoch": 1.1230314960629921, + "grad_norm": 13.000415802001953, + "learning_rate": 9.899132734018024e-07, + "loss": 0.8703, + "step": 1141 + }, + { + "epoch": 1.124015748031496, + "grad_norm": 12.513956069946289, + "learning_rate": 9.897592214281936e-07, + "loss": 0.48, + "step": 1142 + }, + { + "epoch": 1.125, + "grad_norm": 13.1904296875, + "learning_rate": 9.896040141663054e-07, + "loss": 0.8903, + "step": 1143 + }, + { + "epoch": 1.125984251968504, + "grad_norm": 12.2304048538208, + "learning_rate": 9.894476519835052e-07, + "loss": 0.6978, + "step": 1144 + }, + { + "epoch": 1.1269685039370079, + "grad_norm": 13.895735740661621, + "learning_rate": 9.892901352498932e-07, + "loss": 0.8008, + "step": 1145 + }, + { + "epoch": 1.1279527559055118, + "grad_norm": 14.124261856079102, + "learning_rate": 9.891314643383037e-07, + "loss": 0.9583, + "step": 1146 + }, + { + "epoch": 1.1289370078740157, + "grad_norm": 12.590577125549316, + "learning_rate": 9.88971639624301e-07, + "loss": 0.5823, + "step": 1147 + }, + { + "epoch": 1.1299212598425197, + "grad_norm": 13.695793151855469, + "learning_rate": 9.888106614861823e-07, + "loss": 0.974, + "step": 1148 + }, + { + "epoch": 1.1309055118110236, + "grad_norm": 15.616353988647461, + "learning_rate": 9.886485303049734e-07, + "loss": 1.1149, + "step": 1149 + }, + { + "epoch": 1.1318897637795275, + "grad_norm": 13.464898109436035, + "learning_rate": 9.884852464644305e-07, + "loss": 0.5558, + "step": 1150 + }, + { + "epoch": 1.1328740157480315, + "grad_norm": 13.883283615112305, + "learning_rate": 9.883208103510367e-07, + "loss": 0.9106, + "step": 1151 + }, + { + "epoch": 1.1338582677165354, + "grad_norm": 12.547784805297852, + "learning_rate": 9.88155222354004e-07, + "loss": 0.671, + "step": 1152 + }, + { + "epoch": 1.1348425196850394, + "grad_norm": 14.91753101348877, + "learning_rate": 9.879884828652698e-07, + "loss": 0.9424, + "step": 1153 + }, + { + "epoch": 1.1358267716535433, + "grad_norm": 10.684596061706543, + "learning_rate": 9.878205922794972e-07, + "loss": 0.6605, + "step": 1154 + }, + { + "epoch": 1.1368110236220472, + "grad_norm": 11.036165237426758, + "learning_rate": 9.876515509940742e-07, + "loss": 0.536, + "step": 1155 + }, + { + "epoch": 1.1377952755905512, + "grad_norm": 9.091116905212402, + "learning_rate": 9.874813594091122e-07, + "loss": 0.38, + "step": 1156 + }, + { + "epoch": 1.138779527559055, + "grad_norm": 15.718974113464355, + "learning_rate": 9.873100179274453e-07, + "loss": 0.6255, + "step": 1157 + }, + { + "epoch": 1.139763779527559, + "grad_norm": 20.45486831665039, + "learning_rate": 9.871375269546292e-07, + "loss": 1.435, + "step": 1158 + }, + { + "epoch": 1.140748031496063, + "grad_norm": 13.843032836914062, + "learning_rate": 9.869638868989407e-07, + "loss": 0.5903, + "step": 1159 + }, + { + "epoch": 1.141732283464567, + "grad_norm": 10.708097457885742, + "learning_rate": 9.867890981713764e-07, + "loss": 0.9517, + "step": 1160 + }, + { + "epoch": 1.1427165354330708, + "grad_norm": 9.400053024291992, + "learning_rate": 9.866131611856514e-07, + "loss": 0.4636, + "step": 1161 + }, + { + "epoch": 1.1437007874015748, + "grad_norm": 10.441662788391113, + "learning_rate": 9.864360763581994e-07, + "loss": 0.3721, + "step": 1162 + }, + { + "epoch": 1.1446850393700787, + "grad_norm": 13.340433120727539, + "learning_rate": 9.862578441081699e-07, + "loss": 0.6832, + "step": 1163 + }, + { + "epoch": 1.1456692913385826, + "grad_norm": 10.741105079650879, + "learning_rate": 9.860784648574288e-07, + "loss": 0.5893, + "step": 1164 + }, + { + "epoch": 1.1466535433070866, + "grad_norm": 12.039066314697266, + "learning_rate": 9.858979390305573e-07, + "loss": 0.6226, + "step": 1165 + }, + { + "epoch": 1.1476377952755905, + "grad_norm": 15.493182182312012, + "learning_rate": 9.857162670548502e-07, + "loss": 0.5715, + "step": 1166 + }, + { + "epoch": 1.1486220472440944, + "grad_norm": 18.187973022460938, + "learning_rate": 9.855334493603147e-07, + "loss": 1.0958, + "step": 1167 + }, + { + "epoch": 1.1496062992125984, + "grad_norm": 15.315258979797363, + "learning_rate": 9.853494863796705e-07, + "loss": 0.689, + "step": 1168 + }, + { + "epoch": 1.1505905511811023, + "grad_norm": 17.01951026916504, + "learning_rate": 9.851643785483478e-07, + "loss": 0.6443, + "step": 1169 + }, + { + "epoch": 1.1515748031496063, + "grad_norm": 14.690400123596191, + "learning_rate": 9.849781263044867e-07, + "loss": 0.9278, + "step": 1170 + }, + { + "epoch": 1.1525590551181102, + "grad_norm": 14.312512397766113, + "learning_rate": 9.847907300889358e-07, + "loss": 0.7009, + "step": 1171 + }, + { + "epoch": 1.1535433070866141, + "grad_norm": 11.296528816223145, + "learning_rate": 9.846021903452519e-07, + "loss": 0.6254, + "step": 1172 + }, + { + "epoch": 1.154527559055118, + "grad_norm": 12.49747085571289, + "learning_rate": 9.844125075196983e-07, + "loss": 0.7537, + "step": 1173 + }, + { + "epoch": 1.155511811023622, + "grad_norm": 12.261817932128906, + "learning_rate": 9.842216820612438e-07, + "loss": 0.7559, + "step": 1174 + }, + { + "epoch": 1.156496062992126, + "grad_norm": 16.43539810180664, + "learning_rate": 9.840297144215615e-07, + "loss": 0.8148, + "step": 1175 + }, + { + "epoch": 1.1574803149606299, + "grad_norm": 10.178466796875, + "learning_rate": 9.838366050550287e-07, + "loss": 0.2994, + "step": 1176 + }, + { + "epoch": 1.1584645669291338, + "grad_norm": 7.314722061157227, + "learning_rate": 9.836423544187245e-07, + "loss": 0.2148, + "step": 1177 + }, + { + "epoch": 1.1594488188976377, + "grad_norm": 19.49077796936035, + "learning_rate": 9.834469629724293e-07, + "loss": 1.0492, + "step": 1178 + }, + { + "epoch": 1.1604330708661417, + "grad_norm": 13.28373908996582, + "learning_rate": 9.832504311786238e-07, + "loss": 0.5191, + "step": 1179 + }, + { + "epoch": 1.1614173228346456, + "grad_norm": 21.17142105102539, + "learning_rate": 9.830527595024885e-07, + "loss": 1.3508, + "step": 1180 + }, + { + "epoch": 1.1624015748031495, + "grad_norm": 18.57430076599121, + "learning_rate": 9.82853948411901e-07, + "loss": 1.0396, + "step": 1181 + }, + { + "epoch": 1.1633858267716535, + "grad_norm": 11.258069038391113, + "learning_rate": 9.826539983774365e-07, + "loss": 0.5003, + "step": 1182 + }, + { + "epoch": 1.1643700787401574, + "grad_norm": 15.882991790771484, + "learning_rate": 9.824529098723657e-07, + "loss": 0.7736, + "step": 1183 + }, + { + "epoch": 1.1653543307086613, + "grad_norm": 11.641609191894531, + "learning_rate": 9.822506833726537e-07, + "loss": 0.3263, + "step": 1184 + }, + { + "epoch": 1.1663385826771653, + "grad_norm": 21.703527450561523, + "learning_rate": 9.8204731935696e-07, + "loss": 1.2059, + "step": 1185 + }, + { + "epoch": 1.1673228346456692, + "grad_norm": 23.207040786743164, + "learning_rate": 9.818428183066356e-07, + "loss": 1.3514, + "step": 1186 + }, + { + "epoch": 1.1683070866141732, + "grad_norm": 17.124378204345703, + "learning_rate": 9.816371807057237e-07, + "loss": 0.9086, + "step": 1187 + }, + { + "epoch": 1.169291338582677, + "grad_norm": 12.32595443725586, + "learning_rate": 9.81430407040957e-07, + "loss": 0.7245, + "step": 1188 + }, + { + "epoch": 1.170275590551181, + "grad_norm": 15.552436828613281, + "learning_rate": 9.812224978017573e-07, + "loss": 0.717, + "step": 1189 + }, + { + "epoch": 1.171259842519685, + "grad_norm": 12.38658618927002, + "learning_rate": 9.810134534802346e-07, + "loss": 0.6741, + "step": 1190 + }, + { + "epoch": 1.172244094488189, + "grad_norm": 17.0478458404541, + "learning_rate": 9.80803274571185e-07, + "loss": 1.2695, + "step": 1191 + }, + { + "epoch": 1.1732283464566928, + "grad_norm": 9.946022033691406, + "learning_rate": 9.80591961572091e-07, + "loss": 0.406, + "step": 1192 + }, + { + "epoch": 1.1742125984251968, + "grad_norm": 9.98388957977295, + "learning_rate": 9.803795149831183e-07, + "loss": 0.5408, + "step": 1193 + }, + { + "epoch": 1.1751968503937007, + "grad_norm": 9.720131874084473, + "learning_rate": 9.801659353071165e-07, + "loss": 0.4256, + "step": 1194 + }, + { + "epoch": 1.1761811023622046, + "grad_norm": 10.603318214416504, + "learning_rate": 9.799512230496172e-07, + "loss": 0.4827, + "step": 1195 + }, + { + "epoch": 1.1771653543307086, + "grad_norm": 10.514772415161133, + "learning_rate": 9.79735378718832e-07, + "loss": 0.5621, + "step": 1196 + }, + { + "epoch": 1.1781496062992125, + "grad_norm": 13.1632719039917, + "learning_rate": 9.795184028256534e-07, + "loss": 0.6734, + "step": 1197 + }, + { + "epoch": 1.1791338582677164, + "grad_norm": 15.983120918273926, + "learning_rate": 9.793002958836506e-07, + "loss": 0.4685, + "step": 1198 + }, + { + "epoch": 1.1801181102362204, + "grad_norm": 16.66213607788086, + "learning_rate": 9.790810584090712e-07, + "loss": 0.859, + "step": 1199 + }, + { + "epoch": 1.1811023622047245, + "grad_norm": 23.754188537597656, + "learning_rate": 9.78860690920838e-07, + "loss": 1.4388, + "step": 1200 + }, + { + "epoch": 1.1820866141732282, + "grad_norm": 25.50564956665039, + "learning_rate": 9.78639193940549e-07, + "loss": 1.1846, + "step": 1201 + }, + { + "epoch": 1.1830708661417324, + "grad_norm": 17.544904708862305, + "learning_rate": 9.784165679924752e-07, + "loss": 0.6824, + "step": 1202 + }, + { + "epoch": 1.1840551181102361, + "grad_norm": 19.121761322021484, + "learning_rate": 9.781928136035604e-07, + "loss": 1.2064, + "step": 1203 + }, + { + "epoch": 1.1850393700787403, + "grad_norm": 13.858695983886719, + "learning_rate": 9.779679313034182e-07, + "loss": 0.5668, + "step": 1204 + }, + { + "epoch": 1.186023622047244, + "grad_norm": 10.765403747558594, + "learning_rate": 9.777419216243333e-07, + "loss": 0.5286, + "step": 1205 + }, + { + "epoch": 1.1870078740157481, + "grad_norm": 14.333579063415527, + "learning_rate": 9.77514785101258e-07, + "loss": 0.6316, + "step": 1206 + }, + { + "epoch": 1.1879921259842519, + "grad_norm": 15.135106086730957, + "learning_rate": 9.77286522271812e-07, + "loss": 0.8011, + "step": 1207 + }, + { + "epoch": 1.188976377952756, + "grad_norm": 14.610520362854004, + "learning_rate": 9.77057133676281e-07, + "loss": 0.66, + "step": 1208 + }, + { + "epoch": 1.1899606299212597, + "grad_norm": 9.35317325592041, + "learning_rate": 9.768266198576147e-07, + "loss": 0.3735, + "step": 1209 + }, + { + "epoch": 1.1909448818897639, + "grad_norm": 13.126273155212402, + "learning_rate": 9.765949813614273e-07, + "loss": 0.7461, + "step": 1210 + }, + { + "epoch": 1.1919291338582676, + "grad_norm": 12.053367614746094, + "learning_rate": 9.763622187359942e-07, + "loss": 0.5508, + "step": 1211 + }, + { + "epoch": 1.1929133858267718, + "grad_norm": 14.605379104614258, + "learning_rate": 9.761283325322515e-07, + "loss": 0.6379, + "step": 1212 + }, + { + "epoch": 1.1938976377952755, + "grad_norm": 13.75188159942627, + "learning_rate": 9.758933233037952e-07, + "loss": 0.5204, + "step": 1213 + }, + { + "epoch": 1.1948818897637796, + "grad_norm": 14.337552070617676, + "learning_rate": 9.756571916068793e-07, + "loss": 0.7522, + "step": 1214 + }, + { + "epoch": 1.1958661417322836, + "grad_norm": 14.529984474182129, + "learning_rate": 9.754199380004145e-07, + "loss": 0.9415, + "step": 1215 + }, + { + "epoch": 1.1968503937007875, + "grad_norm": 15.994912147521973, + "learning_rate": 9.75181563045967e-07, + "loss": 0.6351, + "step": 1216 + }, + { + "epoch": 1.1978346456692914, + "grad_norm": 9.06320858001709, + "learning_rate": 9.749420673077574e-07, + "loss": 0.262, + "step": 1217 + }, + { + "epoch": 1.1988188976377954, + "grad_norm": 20.326435089111328, + "learning_rate": 9.747014513526588e-07, + "loss": 0.9048, + "step": 1218 + }, + { + "epoch": 1.1998031496062993, + "grad_norm": 18.30372428894043, + "learning_rate": 9.744597157501957e-07, + "loss": 0.6562, + "step": 1219 + }, + { + "epoch": 1.2007874015748032, + "grad_norm": 13.799022674560547, + "learning_rate": 9.742168610725437e-07, + "loss": 0.4502, + "step": 1220 + }, + { + "epoch": 1.2017716535433072, + "grad_norm": 11.906234741210938, + "learning_rate": 9.73972887894526e-07, + "loss": 0.7093, + "step": 1221 + }, + { + "epoch": 1.202755905511811, + "grad_norm": 9.324673652648926, + "learning_rate": 9.737277967936137e-07, + "loss": 0.3464, + "step": 1222 + }, + { + "epoch": 1.203740157480315, + "grad_norm": 13.872148513793945, + "learning_rate": 9.734815883499241e-07, + "loss": 0.7775, + "step": 1223 + }, + { + "epoch": 1.204724409448819, + "grad_norm": 18.56669807434082, + "learning_rate": 9.73234263146219e-07, + "loss": 0.8652, + "step": 1224 + }, + { + "epoch": 1.204724409448819, + "eval_Qnli-dev_cosine_accuracy": 0.65625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8837510347366333, + "eval_Qnli-dev_cosine_ap": 0.6685117731691971, + "eval_Qnli-dev_cosine_f1": 0.6776859504132232, + "eval_Qnli-dev_cosine_f1_threshold": 0.7688855528831482, + "eval_Qnli-dev_cosine_precision": 0.5555555555555556, + "eval_Qnli-dev_cosine_recall": 0.8686440677966102, + "eval_Qnli-dev_dot_accuracy": 0.65625, + "eval_Qnli-dev_dot_accuracy_threshold": 679.20361328125, + "eval_Qnli-dev_dot_ap": 0.6690677625612276, + "eval_Qnli-dev_dot_f1": 0.6776859504132232, + "eval_Qnli-dev_dot_f1_threshold": 591.0426025390625, + "eval_Qnli-dev_dot_precision": 0.5555555555555556, + "eval_Qnli-dev_dot_recall": 0.8686440677966102, + "eval_Qnli-dev_euclidean_accuracy": 0.65625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 13.367326736450195, + "eval_Qnli-dev_euclidean_ap": 0.668516764002818, + "eval_Qnli-dev_euclidean_f1": 0.6776315789473685, + "eval_Qnli-dev_euclidean_f1_threshold": 18.889358520507812, + "eval_Qnli-dev_euclidean_precision": 0.553763440860215, + "eval_Qnli-dev_euclidean_recall": 0.8728813559322034, + "eval_Qnli-dev_manhattan_accuracy": 0.65234375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 284.21575927734375, + "eval_Qnli-dev_manhattan_ap": 0.6673675935294234, + "eval_Qnli-dev_manhattan_f1": 0.678513731825525, + "eval_Qnli-dev_manhattan_f1_threshold": 413.3231201171875, + "eval_Qnli-dev_manhattan_precision": 0.5483028720626631, + "eval_Qnli-dev_manhattan_recall": 0.8898305084745762, + "eval_Qnli-dev_max_accuracy": 0.65625, + "eval_Qnli-dev_max_accuracy_threshold": 679.20361328125, + "eval_Qnli-dev_max_ap": 0.6690677625612276, + "eval_Qnli-dev_max_f1": 0.678513731825525, + "eval_Qnli-dev_max_f1_threshold": 591.0426025390625, + "eval_Qnli-dev_max_precision": 0.5555555555555556, + "eval_Qnli-dev_max_recall": 0.8898305084745762, + "eval_allNLI-dev_cosine_accuracy": 0.6796875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9548401832580566, + "eval_allNLI-dev_cosine_ap": 0.49736507062359686, + "eval_allNLI-dev_cosine_f1": 0.5496828752642706, + "eval_allNLI-dev_cosine_f1_threshold": 0.8550333380699158, + "eval_allNLI-dev_cosine_precision": 0.43333333333333335, + "eval_allNLI-dev_cosine_recall": 0.7514450867052023, + "eval_allNLI-dev_dot_accuracy": 0.681640625, + "eval_allNLI-dev_dot_accuracy_threshold": 725.6923217773438, + "eval_allNLI-dev_dot_ap": 0.4979534945535311, + "eval_allNLI-dev_dot_f1": 0.5496828752642706, + "eval_allNLI-dev_dot_f1_threshold": 657.20263671875, + "eval_allNLI-dev_dot_precision": 0.43333333333333335, + "eval_allNLI-dev_dot_recall": 0.7514450867052023, + "eval_allNLI-dev_euclidean_accuracy": 0.6796875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 8.332578659057617, + "eval_allNLI-dev_euclidean_ap": 0.49737231136383475, + "eval_allNLI-dev_euclidean_f1": 0.5496828752642706, + "eval_allNLI-dev_euclidean_f1_threshold": 14.928186416625977, + "eval_allNLI-dev_euclidean_precision": 0.43333333333333335, + "eval_allNLI-dev_euclidean_recall": 0.7514450867052023, + "eval_allNLI-dev_manhattan_accuracy": 0.68359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 206.03817749023438, + "eval_allNLI-dev_manhattan_ap": 0.499044679219493, + "eval_allNLI-dev_manhattan_f1": 0.5465838509316769, + "eval_allNLI-dev_manhattan_f1_threshold": 323.8904724121094, + "eval_allNLI-dev_manhattan_precision": 0.4258064516129032, + "eval_allNLI-dev_manhattan_recall": 0.7630057803468208, + "eval_allNLI-dev_max_accuracy": 0.68359375, + "eval_allNLI-dev_max_accuracy_threshold": 725.6923217773438, + "eval_allNLI-dev_max_ap": 0.499044679219493, + "eval_allNLI-dev_max_f1": 0.5496828752642706, + "eval_allNLI-dev_max_f1_threshold": 657.20263671875, + "eval_allNLI-dev_max_precision": 0.43333333333333335, + "eval_allNLI-dev_max_recall": 0.7630057803468208, + "eval_runtime": 1.2169, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.6690677625612276, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.6312669055201392, + "eval_sts-test_pearson_dot": 0.6310885086064295, + "eval_sts-test_pearson_euclidean": 0.6545458372257014, + "eval_sts-test_pearson_manhattan": 0.6553441555056533, + "eval_sts-test_pearson_max": 0.6553441555056533, + "eval_sts-test_spearman_cosine": 0.6533221435101542, + "eval_sts-test_spearman_dot": 0.6530519353938729, + "eval_sts-test_spearman_euclidean": 0.6533345396095596, + "eval_sts-test_spearman_manhattan": 0.6554110428893819, + "eval_sts-test_spearman_max": 0.6554110428893819, + "step": 1224 + }, + { + "epoch": 1.205708661417323, + "grad_norm": 10.5484619140625, + "learning_rate": 9.72985821767904e-07, + "loss": 0.3028, + "step": 1225 + }, + { + "epoch": 1.2066929133858268, + "grad_norm": 14.177131652832031, + "learning_rate": 9.727362648030254e-07, + "loss": 0.5766, + "step": 1226 + }, + { + "epoch": 1.2076771653543308, + "grad_norm": 13.425129890441895, + "learning_rate": 9.724855928422714e-07, + "loss": 0.6506, + "step": 1227 + }, + { + "epoch": 1.2086614173228347, + "grad_norm": 14.001228332519531, + "learning_rate": 9.722338064789688e-07, + "loss": 0.7059, + "step": 1228 + }, + { + "epoch": 1.2096456692913387, + "grad_norm": 11.762188911437988, + "learning_rate": 9.719809063090818e-07, + "loss": 0.6542, + "step": 1229 + }, + { + "epoch": 1.2106299212598426, + "grad_norm": 14.125282287597656, + "learning_rate": 9.717268929312112e-07, + "loss": 0.6426, + "step": 1230 + }, + { + "epoch": 1.2116141732283465, + "grad_norm": 18.986549377441406, + "learning_rate": 9.714717669465932e-07, + "loss": 0.5838, + "step": 1231 + }, + { + "epoch": 1.2125984251968505, + "grad_norm": 6.771852493286133, + "learning_rate": 9.712155289590964e-07, + "loss": 0.198, + "step": 1232 + }, + { + "epoch": 1.2135826771653544, + "grad_norm": 15.633471488952637, + "learning_rate": 9.709581795752225e-07, + "loss": 0.5243, + "step": 1233 + }, + { + "epoch": 1.2145669291338583, + "grad_norm": 15.752259254455566, + "learning_rate": 9.70699719404103e-07, + "loss": 0.9086, + "step": 1234 + }, + { + "epoch": 1.2155511811023623, + "grad_norm": 19.11642837524414, + "learning_rate": 9.704401490574993e-07, + "loss": 0.769, + "step": 1235 + }, + { + "epoch": 1.2165354330708662, + "grad_norm": 13.604310035705566, + "learning_rate": 9.701794691497997e-07, + "loss": 0.5532, + "step": 1236 + }, + { + "epoch": 1.2175196850393701, + "grad_norm": 14.573720932006836, + "learning_rate": 9.699176802980199e-07, + "loss": 0.8527, + "step": 1237 + }, + { + "epoch": 1.218503937007874, + "grad_norm": 10.164519309997559, + "learning_rate": 9.69654783121799e-07, + "loss": 0.3036, + "step": 1238 + }, + { + "epoch": 1.219488188976378, + "grad_norm": 12.259843826293945, + "learning_rate": 9.693907782434008e-07, + "loss": 0.5552, + "step": 1239 + }, + { + "epoch": 1.220472440944882, + "grad_norm": 12.603322982788086, + "learning_rate": 9.691256662877099e-07, + "loss": 0.6249, + "step": 1240 + }, + { + "epoch": 1.2214566929133859, + "grad_norm": 11.567940711975098, + "learning_rate": 9.688594478822317e-07, + "loss": 0.44, + "step": 1241 + }, + { + "epoch": 1.2224409448818898, + "grad_norm": 13.615235328674316, + "learning_rate": 9.685921236570914e-07, + "loss": 0.7128, + "step": 1242 + }, + { + "epoch": 1.2234251968503937, + "grad_norm": 8.880663871765137, + "learning_rate": 9.683236942450294e-07, + "loss": 0.3144, + "step": 1243 + }, + { + "epoch": 1.2244094488188977, + "grad_norm": 13.167153358459473, + "learning_rate": 9.680541602814047e-07, + "loss": 0.6626, + "step": 1244 + }, + { + "epoch": 1.2253937007874016, + "grad_norm": 22.988262176513672, + "learning_rate": 9.677835224041885e-07, + "loss": 1.4887, + "step": 1245 + }, + { + "epoch": 1.2263779527559056, + "grad_norm": 17.363365173339844, + "learning_rate": 9.67511781253966e-07, + "loss": 0.5861, + "step": 1246 + }, + { + "epoch": 1.2273622047244095, + "grad_norm": 13.736028671264648, + "learning_rate": 9.67238937473934e-07, + "loss": 0.5891, + "step": 1247 + }, + { + "epoch": 1.2283464566929134, + "grad_norm": 11.614546775817871, + "learning_rate": 9.669649917098983e-07, + "loss": 0.4475, + "step": 1248 + }, + { + "epoch": 1.2293307086614174, + "grad_norm": 13.978675842285156, + "learning_rate": 9.666899446102736e-07, + "loss": 0.6157, + "step": 1249 + }, + { + "epoch": 1.2303149606299213, + "grad_norm": 13.936750411987305, + "learning_rate": 9.664137968260814e-07, + "loss": 0.7397, + "step": 1250 + }, + { + "epoch": 1.2312992125984252, + "grad_norm": 12.804847717285156, + "learning_rate": 9.661365490109482e-07, + "loss": 0.3894, + "step": 1251 + }, + { + "epoch": 1.2322834645669292, + "grad_norm": 18.680809020996094, + "learning_rate": 9.658582018211045e-07, + "loss": 0.925, + "step": 1252 + }, + { + "epoch": 1.233267716535433, + "grad_norm": 11.71372127532959, + "learning_rate": 9.655787559153824e-07, + "loss": 0.5181, + "step": 1253 + }, + { + "epoch": 1.234251968503937, + "grad_norm": 15.968767166137695, + "learning_rate": 9.652982119552154e-07, + "loss": 0.8689, + "step": 1254 + }, + { + "epoch": 1.235236220472441, + "grad_norm": 13.950754165649414, + "learning_rate": 9.650165706046359e-07, + "loss": 0.6174, + "step": 1255 + }, + { + "epoch": 1.236220472440945, + "grad_norm": 18.690776824951172, + "learning_rate": 9.64733832530273e-07, + "loss": 0.891, + "step": 1256 + }, + { + "epoch": 1.2372047244094488, + "grad_norm": 7.422048568725586, + "learning_rate": 9.644499984013525e-07, + "loss": 0.2898, + "step": 1257 + }, + { + "epoch": 1.2381889763779528, + "grad_norm": 14.230087280273438, + "learning_rate": 9.64165068889694e-07, + "loss": 0.7739, + "step": 1258 + }, + { + "epoch": 1.2391732283464567, + "grad_norm": 10.740768432617188, + "learning_rate": 9.6387904466971e-07, + "loss": 0.3945, + "step": 1259 + }, + { + "epoch": 1.2401574803149606, + "grad_norm": 16.949390411376953, + "learning_rate": 9.635919264184043e-07, + "loss": 0.7109, + "step": 1260 + }, + { + "epoch": 1.2411417322834646, + "grad_norm": 14.32327938079834, + "learning_rate": 9.6330371481537e-07, + "loss": 0.7379, + "step": 1261 + }, + { + "epoch": 1.2421259842519685, + "grad_norm": 11.93581485748291, + "learning_rate": 9.630144105427883e-07, + "loss": 0.4401, + "step": 1262 + }, + { + "epoch": 1.2431102362204725, + "grad_norm": 20.528095245361328, + "learning_rate": 9.627240142854262e-07, + "loss": 1.2032, + "step": 1263 + }, + { + "epoch": 1.2440944881889764, + "grad_norm": 12.662689208984375, + "learning_rate": 9.624325267306357e-07, + "loss": 0.344, + "step": 1264 + }, + { + "epoch": 1.2450787401574803, + "grad_norm": 12.233474731445312, + "learning_rate": 9.621399485683523e-07, + "loss": 0.3768, + "step": 1265 + }, + { + "epoch": 1.2460629921259843, + "grad_norm": 12.87982177734375, + "learning_rate": 9.618462804910917e-07, + "loss": 0.5961, + "step": 1266 + }, + { + "epoch": 1.2470472440944882, + "grad_norm": 14.836060523986816, + "learning_rate": 9.61551523193951e-07, + "loss": 0.9748, + "step": 1267 + }, + { + "epoch": 1.2480314960629921, + "grad_norm": 9.906158447265625, + "learning_rate": 9.612556773746038e-07, + "loss": 0.3169, + "step": 1268 + }, + { + "epoch": 1.249015748031496, + "grad_norm": 14.505841255187988, + "learning_rate": 9.609587437333011e-07, + "loss": 1.0254, + "step": 1269 + }, + { + "epoch": 1.25, + "grad_norm": 6.632061958312988, + "learning_rate": 9.606607229728688e-07, + "loss": 0.2742, + "step": 1270 + }, + { + "epoch": 1.250984251968504, + "grad_norm": 11.823864936828613, + "learning_rate": 9.603616157987054e-07, + "loss": 0.7498, + "step": 1271 + }, + { + "epoch": 1.2519685039370079, + "grad_norm": 8.056884765625, + "learning_rate": 9.600614229187812e-07, + "loss": 0.2447, + "step": 1272 + }, + { + "epoch": 1.2529527559055118, + "grad_norm": 10.888479232788086, + "learning_rate": 9.597601450436364e-07, + "loss": 0.424, + "step": 1273 + }, + { + "epoch": 1.2539370078740157, + "grad_norm": 12.182381629943848, + "learning_rate": 9.594577828863794e-07, + "loss": 0.5414, + "step": 1274 + }, + { + "epoch": 1.2549212598425197, + "grad_norm": 8.81839370727539, + "learning_rate": 9.591543371626842e-07, + "loss": 0.5313, + "step": 1275 + }, + { + "epoch": 1.2559055118110236, + "grad_norm": 17.04354476928711, + "learning_rate": 9.588498085907905e-07, + "loss": 0.8376, + "step": 1276 + }, + { + "epoch": 1.2568897637795275, + "grad_norm": 9.12378215789795, + "learning_rate": 9.585441978915008e-07, + "loss": 0.2828, + "step": 1277 + }, + { + "epoch": 1.2578740157480315, + "grad_norm": 10.7446928024292, + "learning_rate": 9.582375057881785e-07, + "loss": 0.3722, + "step": 1278 + }, + { + "epoch": 1.2588582677165354, + "grad_norm": 20.579914093017578, + "learning_rate": 9.579297330067476e-07, + "loss": 1.2595, + "step": 1279 + }, + { + "epoch": 1.2598425196850394, + "grad_norm": 17.499500274658203, + "learning_rate": 9.576208802756888e-07, + "loss": 0.6573, + "step": 1280 + }, + { + "epoch": 1.2608267716535433, + "grad_norm": 15.187433242797852, + "learning_rate": 9.5731094832604e-07, + "loss": 0.6975, + "step": 1281 + }, + { + "epoch": 1.2618110236220472, + "grad_norm": 12.296855926513672, + "learning_rate": 9.569999378913925e-07, + "loss": 0.4378, + "step": 1282 + }, + { + "epoch": 1.2627952755905512, + "grad_norm": 17.853626251220703, + "learning_rate": 9.566878497078914e-07, + "loss": 0.8803, + "step": 1283 + }, + { + "epoch": 1.263779527559055, + "grad_norm": 18.216079711914062, + "learning_rate": 9.563746845142326e-07, + "loss": 0.7752, + "step": 1284 + }, + { + "epoch": 1.264763779527559, + "grad_norm": 15.150053977966309, + "learning_rate": 9.5606044305166e-07, + "loss": 0.5218, + "step": 1285 + }, + { + "epoch": 1.265748031496063, + "grad_norm": 10.312692642211914, + "learning_rate": 9.557451260639665e-07, + "loss": 0.3086, + "step": 1286 + }, + { + "epoch": 1.266732283464567, + "grad_norm": 14.471681594848633, + "learning_rate": 9.5542873429749e-07, + "loss": 0.6541, + "step": 1287 + }, + { + "epoch": 1.2677165354330708, + "grad_norm": 19.451095581054688, + "learning_rate": 9.551112685011127e-07, + "loss": 1.1188, + "step": 1288 + }, + { + "epoch": 1.2687007874015748, + "grad_norm": 11.578512191772461, + "learning_rate": 9.547927294262583e-07, + "loss": 0.6281, + "step": 1289 + }, + { + "epoch": 1.2696850393700787, + "grad_norm": 19.947269439697266, + "learning_rate": 9.544731178268913e-07, + "loss": 0.9196, + "step": 1290 + }, + { + "epoch": 1.2706692913385826, + "grad_norm": 15.87443733215332, + "learning_rate": 9.54152434459515e-07, + "loss": 0.7273, + "step": 1291 + }, + { + "epoch": 1.2716535433070866, + "grad_norm": 12.08711051940918, + "learning_rate": 9.538306800831695e-07, + "loss": 0.6554, + "step": 1292 + }, + { + "epoch": 1.2726377952755905, + "grad_norm": 12.68822956085205, + "learning_rate": 9.535078554594293e-07, + "loss": 0.5857, + "step": 1293 + }, + { + "epoch": 1.2736220472440944, + "grad_norm": 11.92469310760498, + "learning_rate": 9.531839613524029e-07, + "loss": 0.7422, + "step": 1294 + }, + { + "epoch": 1.2746062992125984, + "grad_norm": 12.925695419311523, + "learning_rate": 9.528589985287297e-07, + "loss": 1.1301, + "step": 1295 + }, + { + "epoch": 1.2755905511811023, + "grad_norm": 11.139888763427734, + "learning_rate": 9.525329677575787e-07, + "loss": 0.8172, + "step": 1296 + }, + { + "epoch": 1.2765748031496063, + "grad_norm": 8.039813995361328, + "learning_rate": 9.522058698106475e-07, + "loss": 0.3944, + "step": 1297 + }, + { + "epoch": 1.2775590551181102, + "grad_norm": 12.37278938293457, + "learning_rate": 9.518777054621582e-07, + "loss": 0.7966, + "step": 1298 + }, + { + "epoch": 1.2785433070866141, + "grad_norm": 15.561352729797363, + "learning_rate": 9.515484754888582e-07, + "loss": 0.9283, + "step": 1299 + }, + { + "epoch": 1.279527559055118, + "grad_norm": 11.318450927734375, + "learning_rate": 9.512181806700167e-07, + "loss": 0.7327, + "step": 1300 + }, + { + "epoch": 1.280511811023622, + "grad_norm": 13.565536499023438, + "learning_rate": 9.508868217874233e-07, + "loss": 0.4587, + "step": 1301 + }, + { + "epoch": 1.281496062992126, + "grad_norm": 11.128378868103027, + "learning_rate": 9.505543996253865e-07, + "loss": 0.5131, + "step": 1302 + }, + { + "epoch": 1.2824803149606299, + "grad_norm": 14.159026145935059, + "learning_rate": 9.502209149707312e-07, + "loss": 0.6072, + "step": 1303 + }, + { + "epoch": 1.2834645669291338, + "grad_norm": 11.408098220825195, + "learning_rate": 9.498863686127971e-07, + "loss": 0.6406, + "step": 1304 + }, + { + "epoch": 1.2844488188976377, + "grad_norm": 18.186187744140625, + "learning_rate": 9.495507613434374e-07, + "loss": 0.9822, + "step": 1305 + }, + { + "epoch": 1.2854330708661417, + "grad_norm": 15.517589569091797, + "learning_rate": 9.492140939570159e-07, + "loss": 0.7562, + "step": 1306 + }, + { + "epoch": 1.2864173228346456, + "grad_norm": 16.00613021850586, + "learning_rate": 9.488763672504058e-07, + "loss": 0.7944, + "step": 1307 + }, + { + "epoch": 1.2874015748031495, + "grad_norm": 16.746559143066406, + "learning_rate": 9.485375820229876e-07, + "loss": 0.9248, + "step": 1308 + }, + { + "epoch": 1.2883858267716535, + "grad_norm": 22.098588943481445, + "learning_rate": 9.481977390766475e-07, + "loss": 1.5515, + "step": 1309 + }, + { + "epoch": 1.2893700787401574, + "grad_norm": 9.19296932220459, + "learning_rate": 9.47856839215775e-07, + "loss": 0.5007, + "step": 1310 + }, + { + "epoch": 1.2903543307086613, + "grad_norm": 15.379034996032715, + "learning_rate": 9.475148832472613e-07, + "loss": 0.6497, + "step": 1311 + }, + { + "epoch": 1.2913385826771653, + "grad_norm": 12.293156623840332, + "learning_rate": 9.471718719804975e-07, + "loss": 0.6082, + "step": 1312 + }, + { + "epoch": 1.2923228346456692, + "grad_norm": 15.737980842590332, + "learning_rate": 9.468278062273726e-07, + "loss": 1.0414, + "step": 1313 + }, + { + "epoch": 1.2933070866141732, + "grad_norm": 15.347722053527832, + "learning_rate": 9.464826868022709e-07, + "loss": 0.9682, + "step": 1314 + }, + { + "epoch": 1.294291338582677, + "grad_norm": 11.852689743041992, + "learning_rate": 9.461365145220715e-07, + "loss": 0.661, + "step": 1315 + }, + { + "epoch": 1.295275590551181, + "grad_norm": 14.368483543395996, + "learning_rate": 9.457892902061452e-07, + "loss": 0.7459, + "step": 1316 + }, + { + "epoch": 1.296259842519685, + "grad_norm": 12.618541717529297, + "learning_rate": 9.454410146763525e-07, + "loss": 0.732, + "step": 1317 + }, + { + "epoch": 1.297244094488189, + "grad_norm": 7.708657264709473, + "learning_rate": 9.450916887570429e-07, + "loss": 0.2937, + "step": 1318 + }, + { + "epoch": 1.298228346456693, + "grad_norm": 10.866966247558594, + "learning_rate": 9.447413132750515e-07, + "loss": 0.5876, + "step": 1319 + }, + { + "epoch": 1.2992125984251968, + "grad_norm": 8.800849914550781, + "learning_rate": 9.443898890596977e-07, + "loss": 0.5333, + "step": 1320 + }, + { + "epoch": 1.300196850393701, + "grad_norm": 10.204243659973145, + "learning_rate": 9.440374169427834e-07, + "loss": 0.3876, + "step": 1321 + }, + { + "epoch": 1.3011811023622046, + "grad_norm": 10.551318168640137, + "learning_rate": 9.43683897758591e-07, + "loss": 0.3384, + "step": 1322 + }, + { + "epoch": 1.3021653543307088, + "grad_norm": 9.43662166595459, + "learning_rate": 9.433293323438807e-07, + "loss": 0.3552, + "step": 1323 + }, + { + "epoch": 1.3031496062992125, + "grad_norm": 9.835858345031738, + "learning_rate": 9.429737215378894e-07, + "loss": 0.3188, + "step": 1324 + }, + { + "epoch": 1.3041338582677167, + "grad_norm": 9.521871566772461, + "learning_rate": 9.426170661823285e-07, + "loss": 0.369, + "step": 1325 + }, + { + "epoch": 1.3051181102362204, + "grad_norm": 14.87257194519043, + "learning_rate": 9.422593671213816e-07, + "loss": 0.6256, + "step": 1326 + }, + { + "epoch": 1.3061023622047245, + "grad_norm": 17.019819259643555, + "learning_rate": 9.419006252017027e-07, + "loss": 1.1555, + "step": 1327 + }, + { + "epoch": 1.3070866141732282, + "grad_norm": 17.319644927978516, + "learning_rate": 9.415408412724141e-07, + "loss": 1.1316, + "step": 1328 + }, + { + "epoch": 1.3080708661417324, + "grad_norm": 14.21795654296875, + "learning_rate": 9.411800161851048e-07, + "loss": 0.7225, + "step": 1329 + }, + { + "epoch": 1.3090551181102361, + "grad_norm": 12.0934476852417, + "learning_rate": 9.408181507938279e-07, + "loss": 0.5129, + "step": 1330 + }, + { + "epoch": 1.3100393700787403, + "grad_norm": 9.549371719360352, + "learning_rate": 9.404552459550988e-07, + "loss": 0.2852, + "step": 1331 + }, + { + "epoch": 1.311023622047244, + "grad_norm": 17.59976577758789, + "learning_rate": 9.400913025278937e-07, + "loss": 0.8099, + "step": 1332 + }, + { + "epoch": 1.3120078740157481, + "grad_norm": 10.90012264251709, + "learning_rate": 9.397263213736464e-07, + "loss": 0.3671, + "step": 1333 + }, + { + "epoch": 1.3129921259842519, + "grad_norm": 13.29249382019043, + "learning_rate": 9.393603033562471e-07, + "loss": 0.6209, + "step": 1334 + }, + { + "epoch": 1.313976377952756, + "grad_norm": 12.090213775634766, + "learning_rate": 9.389932493420407e-07, + "loss": 0.583, + "step": 1335 + }, + { + "epoch": 1.3149606299212597, + "grad_norm": 15.204630851745605, + "learning_rate": 9.386251601998239e-07, + "loss": 1.0555, + "step": 1336 + }, + { + "epoch": 1.3159448818897639, + "grad_norm": 18.04181480407715, + "learning_rate": 9.38256036800843e-07, + "loss": 0.9641, + "step": 1337 + }, + { + "epoch": 1.3169291338582676, + "grad_norm": 15.318765640258789, + "learning_rate": 9.378858800187935e-07, + "loss": 0.7796, + "step": 1338 + }, + { + "epoch": 1.3179133858267718, + "grad_norm": 7.988604545593262, + "learning_rate": 9.375146907298157e-07, + "loss": 0.2689, + "step": 1339 + }, + { + "epoch": 1.3188976377952755, + "grad_norm": 10.125104904174805, + "learning_rate": 9.371424698124945e-07, + "loss": 0.3622, + "step": 1340 + }, + { + "epoch": 1.3198818897637796, + "grad_norm": 13.12625503540039, + "learning_rate": 9.367692181478564e-07, + "loss": 0.5872, + "step": 1341 + }, + { + "epoch": 1.3208661417322833, + "grad_norm": 14.257974624633789, + "learning_rate": 9.363949366193673e-07, + "loss": 0.7294, + "step": 1342 + }, + { + "epoch": 1.3218503937007875, + "grad_norm": 15.67703628540039, + "learning_rate": 9.360196261129313e-07, + "loss": 0.659, + "step": 1343 + }, + { + "epoch": 1.3228346456692912, + "grad_norm": 10.205487251281738, + "learning_rate": 9.356432875168876e-07, + "loss": 0.2835, + "step": 1344 + }, + { + "epoch": 1.3238188976377954, + "grad_norm": 13.3869047164917, + "learning_rate": 9.352659217220091e-07, + "loss": 0.3833, + "step": 1345 + }, + { + "epoch": 1.324803149606299, + "grad_norm": 9.782673835754395, + "learning_rate": 9.348875296215001e-07, + "loss": 0.3645, + "step": 1346 + }, + { + "epoch": 1.3257874015748032, + "grad_norm": 9.724504470825195, + "learning_rate": 9.345081121109935e-07, + "loss": 0.2585, + "step": 1347 + }, + { + "epoch": 1.326771653543307, + "grad_norm": 11.042606353759766, + "learning_rate": 9.341276700885501e-07, + "loss": 0.4054, + "step": 1348 + }, + { + "epoch": 1.327755905511811, + "grad_norm": 15.967182159423828, + "learning_rate": 9.337462044546548e-07, + "loss": 0.8016, + "step": 1349 + }, + { + "epoch": 1.328740157480315, + "grad_norm": 12.706576347351074, + "learning_rate": 9.333637161122166e-07, + "loss": 0.3908, + "step": 1350 + }, + { + "epoch": 1.329724409448819, + "grad_norm": 19.711381912231445, + "learning_rate": 9.329802059665637e-07, + "loss": 0.626, + "step": 1351 + }, + { + "epoch": 1.330708661417323, + "grad_norm": 9.85055923461914, + "learning_rate": 9.325956749254437e-07, + "loss": 0.3939, + "step": 1352 + }, + { + "epoch": 1.3316929133858268, + "grad_norm": 14.548029899597168, + "learning_rate": 9.322101238990206e-07, + "loss": 0.5573, + "step": 1353 + }, + { + "epoch": 1.3326771653543308, + "grad_norm": 17.656423568725586, + "learning_rate": 9.318235537998726e-07, + "loss": 0.8803, + "step": 1354 + }, + { + "epoch": 1.3336614173228347, + "grad_norm": 22.868518829345703, + "learning_rate": 9.314359655429897e-07, + "loss": 1.4102, + "step": 1355 + }, + { + "epoch": 1.3346456692913387, + "grad_norm": 15.359719276428223, + "learning_rate": 9.310473600457721e-07, + "loss": 0.7607, + "step": 1356 + }, + { + "epoch": 1.3356299212598426, + "grad_norm": 14.475640296936035, + "learning_rate": 9.306577382280277e-07, + "loss": 0.5631, + "step": 1357 + }, + { + "epoch": 1.3366141732283465, + "grad_norm": 7.226040363311768, + "learning_rate": 9.3026710101197e-07, + "loss": 0.3118, + "step": 1358 + }, + { + "epoch": 1.3375984251968505, + "grad_norm": 14.177290916442871, + "learning_rate": 9.298754493222155e-07, + "loss": 0.7924, + "step": 1359 + }, + { + "epoch": 1.3385826771653544, + "grad_norm": 9.641798973083496, + "learning_rate": 9.294827840857827e-07, + "loss": 0.4304, + "step": 1360 + }, + { + "epoch": 1.3395669291338583, + "grad_norm": 11.06660270690918, + "learning_rate": 9.290891062320884e-07, + "loss": 0.5641, + "step": 1361 + }, + { + "epoch": 1.3405511811023623, + "grad_norm": 7.934324264526367, + "learning_rate": 9.286944166929466e-07, + "loss": 0.3679, + "step": 1362 + }, + { + "epoch": 1.3415354330708662, + "grad_norm": 9.094664573669434, + "learning_rate": 9.282987164025658e-07, + "loss": 0.4429, + "step": 1363 + }, + { + "epoch": 1.3425196850393701, + "grad_norm": 14.20646858215332, + "learning_rate": 9.279020062975465e-07, + "loss": 0.7248, + "step": 1364 + }, + { + "epoch": 1.343503937007874, + "grad_norm": 14.210354804992676, + "learning_rate": 9.275042873168802e-07, + "loss": 0.4378, + "step": 1365 + }, + { + "epoch": 1.344488188976378, + "grad_norm": 12.429130554199219, + "learning_rate": 9.271055604019458e-07, + "loss": 0.7427, + "step": 1366 + }, + { + "epoch": 1.345472440944882, + "grad_norm": 16.509376525878906, + "learning_rate": 9.267058264965076e-07, + "loss": 0.7427, + "step": 1367 + }, + { + "epoch": 1.3464566929133859, + "grad_norm": 8.723173141479492, + "learning_rate": 9.263050865467143e-07, + "loss": 0.2993, + "step": 1368 + }, + { + "epoch": 1.3474409448818898, + "grad_norm": 12.15352725982666, + "learning_rate": 9.259033415010952e-07, + "loss": 0.5475, + "step": 1369 + }, + { + "epoch": 1.3484251968503937, + "grad_norm": 16.92938804626465, + "learning_rate": 9.255005923105587e-07, + "loss": 0.8495, + "step": 1370 + }, + { + "epoch": 1.3494094488188977, + "grad_norm": 9.949490547180176, + "learning_rate": 9.250968399283902e-07, + "loss": 0.6147, + "step": 1371 + }, + { + "epoch": 1.3503937007874016, + "grad_norm": 15.532437324523926, + "learning_rate": 9.24692085310249e-07, + "loss": 0.67, + "step": 1372 + }, + { + "epoch": 1.3513779527559056, + "grad_norm": 12.590180397033691, + "learning_rate": 9.242863294141675e-07, + "loss": 0.3701, + "step": 1373 + }, + { + "epoch": 1.3523622047244095, + "grad_norm": 6.725234031677246, + "learning_rate": 9.238795732005475e-07, + "loss": 0.2513, + "step": 1374 + }, + { + "epoch": 1.3533464566929134, + "grad_norm": 20.934463500976562, + "learning_rate": 9.234718176321586e-07, + "loss": 0.9747, + "step": 1375 + }, + { + "epoch": 1.3543307086614174, + "grad_norm": 15.258929252624512, + "learning_rate": 9.230630636741358e-07, + "loss": 0.5877, + "step": 1376 + }, + { + "epoch": 1.3553149606299213, + "grad_norm": 17.185503005981445, + "learning_rate": 9.226533122939772e-07, + "loss": 0.8615, + "step": 1377 + }, + { + "epoch": 1.3553149606299213, + "eval_Qnli-dev_cosine_accuracy": 0.669921875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8494552373886108, + "eval_Qnli-dev_cosine_ap": 0.7106783528902046, + "eval_Qnli-dev_cosine_f1": 0.6891651865008882, + "eval_Qnli-dev_cosine_f1_threshold": 0.7761175632476807, + "eval_Qnli-dev_cosine_precision": 0.5932721712538226, + "eval_Qnli-dev_cosine_recall": 0.8220338983050848, + "eval_Qnli-dev_dot_accuracy": 0.671875, + "eval_Qnli-dev_dot_accuracy_threshold": 648.5075073242188, + "eval_Qnli-dev_dot_ap": 0.7109007448621125, + "eval_Qnli-dev_dot_f1": 0.6914893617021277, + "eval_Qnli-dev_dot_f1_threshold": 596.2176513671875, + "eval_Qnli-dev_dot_precision": 0.5945121951219512, + "eval_Qnli-dev_dot_recall": 0.826271186440678, + "eval_Qnli-dev_euclidean_accuracy": 0.669921875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.209634780883789, + "eval_Qnli-dev_euclidean_ap": 0.7107551206516709, + "eval_Qnli-dev_euclidean_f1": 0.6891651865008882, + "eval_Qnli-dev_euclidean_f1_threshold": 18.54657554626465, + "eval_Qnli-dev_euclidean_precision": 0.5932721712538226, + "eval_Qnli-dev_euclidean_recall": 0.8220338983050848, + "eval_Qnli-dev_manhattan_accuracy": 0.669921875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 388.1427307128906, + "eval_Qnli-dev_manhattan_ap": 0.7093026459304114, + "eval_Qnli-dev_manhattan_f1": 0.6901408450704225, + "eval_Qnli-dev_manhattan_f1_threshold": 403.2428894042969, + "eval_Qnli-dev_manhattan_precision": 0.5903614457831325, + "eval_Qnli-dev_manhattan_recall": 0.8305084745762712, + "eval_Qnli-dev_max_accuracy": 0.671875, + "eval_Qnli-dev_max_accuracy_threshold": 648.5075073242188, + "eval_Qnli-dev_max_ap": 0.7109007448621125, + "eval_Qnli-dev_max_f1": 0.6914893617021277, + "eval_Qnli-dev_max_f1_threshold": 596.2176513671875, + "eval_Qnli-dev_max_precision": 0.5945121951219512, + "eval_Qnli-dev_max_recall": 0.8305084745762712, + "eval_allNLI-dev_cosine_accuracy": 0.6875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9168648719787598, + "eval_allNLI-dev_cosine_ap": 0.49590687320528504, + "eval_allNLI-dev_cosine_f1": 0.5425742574257426, + "eval_allNLI-dev_cosine_f1_threshold": 0.8390330672264099, + "eval_allNLI-dev_cosine_precision": 0.4126506024096386, + "eval_allNLI-dev_cosine_recall": 0.791907514450867, + "eval_allNLI-dev_dot_accuracy": 0.6875, + "eval_allNLI-dev_dot_accuracy_threshold": 704.442626953125, + "eval_allNLI-dev_dot_ap": 0.49590643017797076, + "eval_allNLI-dev_dot_f1": 0.5425742574257426, + "eval_allNLI-dev_dot_f1_threshold": 644.679443359375, + "eval_allNLI-dev_dot_precision": 0.4126506024096386, + "eval_allNLI-dev_dot_recall": 0.791907514450867, + "eval_allNLI-dev_euclidean_accuracy": 0.6875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.302576065063477, + "eval_allNLI-dev_euclidean_ap": 0.49598705668299703, + "eval_allNLI-dev_euclidean_f1": 0.5425742574257426, + "eval_allNLI-dev_euclidean_f1_threshold": 15.727437019348145, + "eval_allNLI-dev_euclidean_precision": 0.4126506024096386, + "eval_allNLI-dev_euclidean_recall": 0.791907514450867, + "eval_allNLI-dev_manhattan_accuracy": 0.6875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 222.78009033203125, + "eval_allNLI-dev_manhattan_ap": 0.4978551774755373, + "eval_allNLI-dev_manhattan_f1": 0.546875, + "eval_allNLI-dev_manhattan_f1_threshold": 340.3675537109375, + "eval_allNLI-dev_manhattan_precision": 0.41297935103244837, + "eval_allNLI-dev_manhattan_recall": 0.8092485549132948, + "eval_allNLI-dev_max_accuracy": 0.6875, + "eval_allNLI-dev_max_accuracy_threshold": 704.442626953125, + "eval_allNLI-dev_max_ap": 0.4978551774755373, + "eval_allNLI-dev_max_f1": 0.546875, + "eval_allNLI-dev_max_f1_threshold": 644.679443359375, + "eval_allNLI-dev_max_precision": 0.41297935103244837, + "eval_allNLI-dev_max_recall": 0.8092485549132948, + "eval_runtime": 1.2186, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.7109007448621125, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.6434248115922252, + "eval_sts-test_pearson_dot": 0.6433495406242706, + "eval_sts-test_pearson_euclidean": 0.6741894701523308, + "eval_sts-test_pearson_manhattan": 0.6747840288011084, + "eval_sts-test_pearson_max": 0.6747840288011084, + "eval_sts-test_spearman_cosine": 0.6700466744198709, + "eval_sts-test_spearman_dot": 0.6699834856387862, + "eval_sts-test_spearman_euclidean": 0.6700878753539228, + "eval_sts-test_spearman_manhattan": 0.6708056110618075, + "eval_sts-test_spearman_max": 0.6708056110618075, + "step": 1377 + }, + { + "epoch": 1.3562992125984252, + "grad_norm": 14.39145278930664, + "learning_rate": 9.222425644615421e-07, + "loss": 0.6374, + "step": 1378 + }, + { + "epoch": 1.3572834645669292, + "grad_norm": 11.03903865814209, + "learning_rate": 9.218308211490479e-07, + "loss": 0.3782, + "step": 1379 + }, + { + "epoch": 1.358267716535433, + "grad_norm": 7.740208625793457, + "learning_rate": 9.214180833310684e-07, + "loss": 0.1647, + "step": 1380 + }, + { + "epoch": 1.359251968503937, + "grad_norm": 18.132917404174805, + "learning_rate": 9.210043519845313e-07, + "loss": 1.1326, + "step": 1381 + }, + { + "epoch": 1.360236220472441, + "grad_norm": 12.381155967712402, + "learning_rate": 9.205896280887163e-07, + "loss": 0.6177, + "step": 1382 + }, + { + "epoch": 1.361220472440945, + "grad_norm": 11.164826393127441, + "learning_rate": 9.201739126252518e-07, + "loss": 0.4708, + "step": 1383 + }, + { + "epoch": 1.3622047244094488, + "grad_norm": 12.87879753112793, + "learning_rate": 9.197572065781133e-07, + "loss": 0.75, + "step": 1384 + }, + { + "epoch": 1.3631889763779528, + "grad_norm": 19.596181869506836, + "learning_rate": 9.193395109336216e-07, + "loss": 0.985, + "step": 1385 + }, + { + "epoch": 1.3641732283464567, + "grad_norm": 10.506561279296875, + "learning_rate": 9.189208266804389e-07, + "loss": 0.5775, + "step": 1386 + }, + { + "epoch": 1.3651574803149606, + "grad_norm": 14.01667308807373, + "learning_rate": 9.185011548095682e-07, + "loss": 0.738, + "step": 1387 + }, + { + "epoch": 1.3661417322834646, + "grad_norm": 9.905497550964355, + "learning_rate": 9.180804963143496e-07, + "loss": 0.6335, + "step": 1388 + }, + { + "epoch": 1.3671259842519685, + "grad_norm": 7.986133575439453, + "learning_rate": 9.176588521904588e-07, + "loss": 0.3397, + "step": 1389 + }, + { + "epoch": 1.3681102362204725, + "grad_norm": 10.701656341552734, + "learning_rate": 9.172362234359041e-07, + "loss": 0.5236, + "step": 1390 + }, + { + "epoch": 1.3690944881889764, + "grad_norm": 7.33452033996582, + "learning_rate": 9.168126110510246e-07, + "loss": 0.3185, + "step": 1391 + }, + { + "epoch": 1.3700787401574803, + "grad_norm": 10.925915718078613, + "learning_rate": 9.163880160384877e-07, + "loss": 0.4717, + "step": 1392 + }, + { + "epoch": 1.3710629921259843, + "grad_norm": 11.257904052734375, + "learning_rate": 9.159624394032864e-07, + "loss": 0.6093, + "step": 1393 + }, + { + "epoch": 1.3720472440944882, + "grad_norm": 11.066444396972656, + "learning_rate": 9.155358821527371e-07, + "loss": 0.3521, + "step": 1394 + }, + { + "epoch": 1.3730314960629921, + "grad_norm": 17.658069610595703, + "learning_rate": 9.151083452964775e-07, + "loss": 1.0617, + "step": 1395 + }, + { + "epoch": 1.374015748031496, + "grad_norm": 15.779825210571289, + "learning_rate": 9.146798298464642e-07, + "loss": 0.5693, + "step": 1396 + }, + { + "epoch": 1.375, + "grad_norm": 14.255114555358887, + "learning_rate": 9.14250336816969e-07, + "loss": 0.8272, + "step": 1397 + }, + { + "epoch": 1.375984251968504, + "grad_norm": 8.256271362304688, + "learning_rate": 9.13819867224579e-07, + "loss": 0.3983, + "step": 1398 + }, + { + "epoch": 1.3769685039370079, + "grad_norm": 16.68880844116211, + "learning_rate": 9.133884220881917e-07, + "loss": 1.1205, + "step": 1399 + }, + { + "epoch": 1.3779527559055118, + "grad_norm": 7.657689094543457, + "learning_rate": 9.129560024290145e-07, + "loss": 0.2393, + "step": 1400 + }, + { + "epoch": 1.3789370078740157, + "grad_norm": 13.306052207946777, + "learning_rate": 9.125226092705603e-07, + "loss": 0.465, + "step": 1401 + }, + { + "epoch": 1.3799212598425197, + "grad_norm": 8.191863059997559, + "learning_rate": 9.120882436386474e-07, + "loss": 0.457, + "step": 1402 + }, + { + "epoch": 1.3809055118110236, + "grad_norm": 15.662034034729004, + "learning_rate": 9.116529065613954e-07, + "loss": 0.921, + "step": 1403 + }, + { + "epoch": 1.3818897637795275, + "grad_norm": 10.389616966247559, + "learning_rate": 9.112165990692231e-07, + "loss": 0.4724, + "step": 1404 + }, + { + "epoch": 1.3828740157480315, + "grad_norm": 13.96307373046875, + "learning_rate": 9.107793221948464e-07, + "loss": 0.598, + "step": 1405 + }, + { + "epoch": 1.3838582677165354, + "grad_norm": 13.452620506286621, + "learning_rate": 9.103410769732755e-07, + "loss": 0.8975, + "step": 1406 + }, + { + "epoch": 1.3848425196850394, + "grad_norm": 14.932463645935059, + "learning_rate": 9.099018644418133e-07, + "loss": 0.6914, + "step": 1407 + }, + { + "epoch": 1.3858267716535433, + "grad_norm": 13.74275016784668, + "learning_rate": 9.094616856400512e-07, + "loss": 0.6997, + "step": 1408 + }, + { + "epoch": 1.3868110236220472, + "grad_norm": 14.333803176879883, + "learning_rate": 9.090205416098685e-07, + "loss": 0.6738, + "step": 1409 + }, + { + "epoch": 1.3877952755905512, + "grad_norm": 9.94694995880127, + "learning_rate": 9.08578433395429e-07, + "loss": 0.3945, + "step": 1410 + }, + { + "epoch": 1.388779527559055, + "grad_norm": 12.054594993591309, + "learning_rate": 9.081353620431783e-07, + "loss": 0.5746, + "step": 1411 + }, + { + "epoch": 1.389763779527559, + "grad_norm": 10.242097854614258, + "learning_rate": 9.07691328601842e-07, + "loss": 0.3389, + "step": 1412 + }, + { + "epoch": 1.390748031496063, + "grad_norm": 13.18433952331543, + "learning_rate": 9.072463341224232e-07, + "loss": 0.5674, + "step": 1413 + }, + { + "epoch": 1.391732283464567, + "grad_norm": 10.200501441955566, + "learning_rate": 9.068003796581991e-07, + "loss": 0.4476, + "step": 1414 + }, + { + "epoch": 1.3927165354330708, + "grad_norm": 10.765254020690918, + "learning_rate": 9.063534662647195e-07, + "loss": 0.4308, + "step": 1415 + }, + { + "epoch": 1.3937007874015748, + "grad_norm": 16.95307731628418, + "learning_rate": 9.059055949998039e-07, + "loss": 1.2597, + "step": 1416 + }, + { + "epoch": 1.3946850393700787, + "grad_norm": 11.753730773925781, + "learning_rate": 9.05456766923539e-07, + "loss": 0.6179, + "step": 1417 + }, + { + "epoch": 1.3956692913385826, + "grad_norm": 12.892511367797852, + "learning_rate": 9.05006983098276e-07, + "loss": 0.4679, + "step": 1418 + }, + { + "epoch": 1.3966535433070866, + "grad_norm": 9.340267181396484, + "learning_rate": 9.045562445886289e-07, + "loss": 0.5284, + "step": 1419 + }, + { + "epoch": 1.3976377952755905, + "grad_norm": 15.766507148742676, + "learning_rate": 9.041045524614707e-07, + "loss": 1.0125, + "step": 1420 + }, + { + "epoch": 1.3986220472440944, + "grad_norm": 10.811077117919922, + "learning_rate": 9.03651907785932e-07, + "loss": 0.4006, + "step": 1421 + }, + { + "epoch": 1.3996062992125984, + "grad_norm": 15.585694313049316, + "learning_rate": 9.03198311633398e-07, + "loss": 0.7757, + "step": 1422 + }, + { + "epoch": 1.4005905511811023, + "grad_norm": 12.539312362670898, + "learning_rate": 9.027437650775055e-07, + "loss": 0.6065, + "step": 1423 + }, + { + "epoch": 1.4015748031496063, + "grad_norm": 13.817112922668457, + "learning_rate": 9.022882691941417e-07, + "loss": 0.5428, + "step": 1424 + }, + { + "epoch": 1.4025590551181102, + "grad_norm": 10.922581672668457, + "learning_rate": 9.018318250614402e-07, + "loss": 0.32, + "step": 1425 + }, + { + "epoch": 1.4035433070866141, + "grad_norm": 14.805595397949219, + "learning_rate": 9.01374433759779e-07, + "loss": 0.5435, + "step": 1426 + }, + { + "epoch": 1.404527559055118, + "grad_norm": 13.373017311096191, + "learning_rate": 9.009160963717784e-07, + "loss": 0.3498, + "step": 1427 + }, + { + "epoch": 1.405511811023622, + "grad_norm": 12.346015930175781, + "learning_rate": 9.004568139822978e-07, + "loss": 0.5562, + "step": 1428 + }, + { + "epoch": 1.406496062992126, + "grad_norm": 16.152240753173828, + "learning_rate": 8.999965876784333e-07, + "loss": 0.7056, + "step": 1429 + }, + { + "epoch": 1.4074803149606299, + "grad_norm": 10.077422142028809, + "learning_rate": 8.995354185495155e-07, + "loss": 0.3809, + "step": 1430 + }, + { + "epoch": 1.4084645669291338, + "grad_norm": 16.427684783935547, + "learning_rate": 8.990733076871063e-07, + "loss": 0.9798, + "step": 1431 + }, + { + "epoch": 1.4094488188976377, + "grad_norm": 19.359325408935547, + "learning_rate": 8.986102561849967e-07, + "loss": 1.0254, + "step": 1432 + }, + { + "epoch": 1.4104330708661417, + "grad_norm": 13.302583694458008, + "learning_rate": 8.981462651392041e-07, + "loss": 0.7483, + "step": 1433 + }, + { + "epoch": 1.4114173228346456, + "grad_norm": 12.176793098449707, + "learning_rate": 8.9768133564797e-07, + "loss": 0.3944, + "step": 1434 + }, + { + "epoch": 1.4124015748031495, + "grad_norm": 14.52035140991211, + "learning_rate": 8.972154688117568e-07, + "loss": 0.5626, + "step": 1435 + }, + { + "epoch": 1.4133858267716535, + "grad_norm": 10.754803657531738, + "learning_rate": 8.967486657332462e-07, + "loss": 0.4553, + "step": 1436 + }, + { + "epoch": 1.4143700787401574, + "grad_norm": 12.303003311157227, + "learning_rate": 8.962809275173347e-07, + "loss": 0.4601, + "step": 1437 + }, + { + "epoch": 1.4153543307086613, + "grad_norm": 15.030244827270508, + "learning_rate": 8.958122552711336e-07, + "loss": 0.734, + "step": 1438 + }, + { + "epoch": 1.4163385826771653, + "grad_norm": 11.638476371765137, + "learning_rate": 8.953426501039641e-07, + "loss": 0.3654, + "step": 1439 + }, + { + "epoch": 1.4173228346456692, + "grad_norm": 12.736907958984375, + "learning_rate": 8.94872113127356e-07, + "loss": 0.8148, + "step": 1440 + }, + { + "epoch": 1.4183070866141732, + "grad_norm": 14.538094520568848, + "learning_rate": 8.944006454550443e-07, + "loss": 0.4998, + "step": 1441 + }, + { + "epoch": 1.419291338582677, + "grad_norm": 13.657557487487793, + "learning_rate": 8.939282482029671e-07, + "loss": 0.4505, + "step": 1442 + }, + { + "epoch": 1.420275590551181, + "grad_norm": 11.35772705078125, + "learning_rate": 8.934549224892629e-07, + "loss": 0.4764, + "step": 1443 + }, + { + "epoch": 1.421259842519685, + "grad_norm": 12.238612174987793, + "learning_rate": 8.929806694342677e-07, + "loss": 0.4486, + "step": 1444 + }, + { + "epoch": 1.422244094488189, + "grad_norm": 11.179261207580566, + "learning_rate": 8.925054901605124e-07, + "loss": 0.3752, + "step": 1445 + }, + { + "epoch": 1.423228346456693, + "grad_norm": 12.104653358459473, + "learning_rate": 8.9202938579272e-07, + "loss": 0.428, + "step": 1446 + }, + { + "epoch": 1.4242125984251968, + "grad_norm": 10.585807800292969, + "learning_rate": 8.915523574578038e-07, + "loss": 0.3955, + "step": 1447 + }, + { + "epoch": 1.425196850393701, + "grad_norm": 15.985635757446289, + "learning_rate": 8.910744062848633e-07, + "loss": 0.6988, + "step": 1448 + }, + { + "epoch": 1.4261811023622046, + "grad_norm": 14.4647855758667, + "learning_rate": 8.90595533405183e-07, + "loss": 0.7233, + "step": 1449 + }, + { + "epoch": 1.4271653543307088, + "grad_norm": 19.723024368286133, + "learning_rate": 8.901157399522286e-07, + "loss": 1.2082, + "step": 1450 + }, + { + "epoch": 1.4281496062992125, + "grad_norm": 11.812784194946289, + "learning_rate": 8.896350270616449e-07, + "loss": 0.4325, + "step": 1451 + }, + { + "epoch": 1.4291338582677167, + "grad_norm": 15.073848724365234, + "learning_rate": 8.891533958712529e-07, + "loss": 0.8255, + "step": 1452 + }, + { + "epoch": 1.4301181102362204, + "grad_norm": 13.357765197753906, + "learning_rate": 8.88670847521047e-07, + "loss": 0.6323, + "step": 1453 + }, + { + "epoch": 1.4311023622047245, + "grad_norm": 9.829086303710938, + "learning_rate": 8.881873831531928e-07, + "loss": 0.2209, + "step": 1454 + }, + { + "epoch": 1.4320866141732282, + "grad_norm": 13.290863037109375, + "learning_rate": 8.877030039120242e-07, + "loss": 0.4348, + "step": 1455 + }, + { + "epoch": 1.4330708661417324, + "grad_norm": 15.914552688598633, + "learning_rate": 8.872177109440398e-07, + "loss": 0.6173, + "step": 1456 + }, + { + "epoch": 1.4340551181102361, + "grad_norm": 11.102739334106445, + "learning_rate": 8.867315053979016e-07, + "loss": 0.5231, + "step": 1457 + }, + { + "epoch": 1.4350393700787403, + "grad_norm": 19.192537307739258, + "learning_rate": 8.862443884244314e-07, + "loss": 0.9232, + "step": 1458 + }, + { + "epoch": 1.436023622047244, + "grad_norm": 15.967985153198242, + "learning_rate": 8.857563611766085e-07, + "loss": 0.6264, + "step": 1459 + }, + { + "epoch": 1.4370078740157481, + "grad_norm": 14.293570518493652, + "learning_rate": 8.852674248095663e-07, + "loss": 0.5722, + "step": 1460 + }, + { + "epoch": 1.4379921259842519, + "grad_norm": 10.654985427856445, + "learning_rate": 8.847775804805907e-07, + "loss": 0.5074, + "step": 1461 + }, + { + "epoch": 1.438976377952756, + "grad_norm": 11.394760131835938, + "learning_rate": 8.84286829349116e-07, + "loss": 0.3808, + "step": 1462 + }, + { + "epoch": 1.4399606299212597, + "grad_norm": 11.593530654907227, + "learning_rate": 8.837951725767234e-07, + "loss": 0.5676, + "step": 1463 + }, + { + "epoch": 1.4409448818897639, + "grad_norm": 11.316069602966309, + "learning_rate": 8.833026113271374e-07, + "loss": 0.4172, + "step": 1464 + }, + { + "epoch": 1.4419291338582676, + "grad_norm": 12.25304126739502, + "learning_rate": 8.828091467662234e-07, + "loss": 0.4354, + "step": 1465 + }, + { + "epoch": 1.4429133858267718, + "grad_norm": 15.181941032409668, + "learning_rate": 8.82314780061985e-07, + "loss": 0.6341, + "step": 1466 + }, + { + "epoch": 1.4438976377952755, + "grad_norm": 13.739705085754395, + "learning_rate": 8.81819512384561e-07, + "loss": 0.5613, + "step": 1467 + }, + { + "epoch": 1.4448818897637796, + "grad_norm": 14.740485191345215, + "learning_rate": 8.813233449062229e-07, + "loss": 0.6067, + "step": 1468 + }, + { + "epoch": 1.4458661417322833, + "grad_norm": 12.778326034545898, + "learning_rate": 8.808262788013718e-07, + "loss": 0.5303, + "step": 1469 + }, + { + "epoch": 1.4468503937007875, + "grad_norm": 19.6373348236084, + "learning_rate": 8.803283152465357e-07, + "loss": 0.7982, + "step": 1470 + }, + { + "epoch": 1.4478346456692912, + "grad_norm": 15.062620162963867, + "learning_rate": 8.798294554203674e-07, + "loss": 0.8807, + "step": 1471 + }, + { + "epoch": 1.4488188976377954, + "grad_norm": 9.258832931518555, + "learning_rate": 8.793297005036402e-07, + "loss": 0.3024, + "step": 1472 + }, + { + "epoch": 1.449803149606299, + "grad_norm": 8.651670455932617, + "learning_rate": 8.788290516792471e-07, + "loss": 0.3424, + "step": 1473 + }, + { + "epoch": 1.4507874015748032, + "grad_norm": 13.530675888061523, + "learning_rate": 8.783275101321962e-07, + "loss": 0.5359, + "step": 1474 + }, + { + "epoch": 1.451771653543307, + "grad_norm": 10.47048568725586, + "learning_rate": 8.778250770496084e-07, + "loss": 0.315, + "step": 1475 + }, + { + "epoch": 1.452755905511811, + "grad_norm": 16.513668060302734, + "learning_rate": 8.773217536207158e-07, + "loss": 0.8719, + "step": 1476 + }, + { + "epoch": 1.453740157480315, + "grad_norm": 15.08060073852539, + "learning_rate": 8.76817541036857e-07, + "loss": 0.6459, + "step": 1477 + }, + { + "epoch": 1.454724409448819, + "grad_norm": 11.214242935180664, + "learning_rate": 8.763124404914754e-07, + "loss": 0.2313, + "step": 1478 + }, + { + "epoch": 1.455708661417323, + "grad_norm": 16.405960083007812, + "learning_rate": 8.758064531801166e-07, + "loss": 0.7386, + "step": 1479 + }, + { + "epoch": 1.4566929133858268, + "grad_norm": 13.755053520202637, + "learning_rate": 8.752995803004247e-07, + "loss": 0.7812, + "step": 1480 + }, + { + "epoch": 1.4576771653543308, + "grad_norm": 10.799671173095703, + "learning_rate": 8.747918230521398e-07, + "loss": 0.3368, + "step": 1481 + }, + { + "epoch": 1.4586614173228347, + "grad_norm": 16.855546951293945, + "learning_rate": 8.742831826370956e-07, + "loss": 0.7843, + "step": 1482 + }, + { + "epoch": 1.4596456692913387, + "grad_norm": 9.733506202697754, + "learning_rate": 8.737736602592158e-07, + "loss": 0.3082, + "step": 1483 + }, + { + "epoch": 1.4606299212598426, + "grad_norm": 10.591129302978516, + "learning_rate": 8.732632571245121e-07, + "loss": 0.465, + "step": 1484 + }, + { + "epoch": 1.4616141732283465, + "grad_norm": 7.218691349029541, + "learning_rate": 8.727519744410806e-07, + "loss": 0.2456, + "step": 1485 + }, + { + "epoch": 1.4625984251968505, + "grad_norm": 9.774059295654297, + "learning_rate": 8.722398134190995e-07, + "loss": 0.3764, + "step": 1486 + }, + { + "epoch": 1.4635826771653544, + "grad_norm": 9.469609260559082, + "learning_rate": 8.717267752708257e-07, + "loss": 0.2452, + "step": 1487 + }, + { + "epoch": 1.4645669291338583, + "grad_norm": 12.692319869995117, + "learning_rate": 8.712128612105924e-07, + "loss": 0.3893, + "step": 1488 + }, + { + "epoch": 1.4655511811023623, + "grad_norm": 12.059755325317383, + "learning_rate": 8.706980724548058e-07, + "loss": 0.5935, + "step": 1489 + }, + { + "epoch": 1.4665354330708662, + "grad_norm": 14.06823444366455, + "learning_rate": 8.701824102219428e-07, + "loss": 0.6697, + "step": 1490 + }, + { + "epoch": 1.4675196850393701, + "grad_norm": 11.525129318237305, + "learning_rate": 8.696658757325474e-07, + "loss": 0.4737, + "step": 1491 + }, + { + "epoch": 1.468503937007874, + "grad_norm": 11.31100845336914, + "learning_rate": 8.691484702092286e-07, + "loss": 0.5339, + "step": 1492 + }, + { + "epoch": 1.469488188976378, + "grad_norm": 6.359795093536377, + "learning_rate": 8.686301948766568e-07, + "loss": 0.1327, + "step": 1493 + }, + { + "epoch": 1.470472440944882, + "grad_norm": 16.80344009399414, + "learning_rate": 8.681110509615612e-07, + "loss": 0.7115, + "step": 1494 + }, + { + "epoch": 1.4714566929133859, + "grad_norm": 18.322765350341797, + "learning_rate": 8.675910396927269e-07, + "loss": 1.0257, + "step": 1495 + }, + { + "epoch": 1.4724409448818898, + "grad_norm": 15.705452919006348, + "learning_rate": 8.670701623009921e-07, + "loss": 0.6444, + "step": 1496 + }, + { + "epoch": 1.4734251968503937, + "grad_norm": 10.868653297424316, + "learning_rate": 8.66548420019245e-07, + "loss": 0.2697, + "step": 1497 + }, + { + "epoch": 1.4744094488188977, + "grad_norm": 16.409481048583984, + "learning_rate": 8.66025814082421e-07, + "loss": 0.7672, + "step": 1498 + }, + { + "epoch": 1.4753937007874016, + "grad_norm": 16.953397750854492, + "learning_rate": 8.655023457274994e-07, + "loss": 0.6761, + "step": 1499 + }, + { + "epoch": 1.4763779527559056, + "grad_norm": 21.2070255279541, + "learning_rate": 8.649780161935012e-07, + "loss": 1.3402, + "step": 1500 + }, + { + "epoch": 1.4773622047244095, + "grad_norm": 15.560661315917969, + "learning_rate": 8.644528267214859e-07, + "loss": 0.7029, + "step": 1501 + }, + { + "epoch": 1.4783464566929134, + "grad_norm": 14.812433242797852, + "learning_rate": 8.639267785545478e-07, + "loss": 0.6435, + "step": 1502 + }, + { + "epoch": 1.4793307086614174, + "grad_norm": 10.85482120513916, + "learning_rate": 8.63399872937814e-07, + "loss": 0.5042, + "step": 1503 + }, + { + "epoch": 1.4803149606299213, + "grad_norm": 10.917243003845215, + "learning_rate": 8.628721111184414e-07, + "loss": 0.3626, + "step": 1504 + }, + { + "epoch": 1.4812992125984252, + "grad_norm": 13.62775707244873, + "learning_rate": 8.623434943456132e-07, + "loss": 0.6935, + "step": 1505 + }, + { + "epoch": 1.4822834645669292, + "grad_norm": 15.856972694396973, + "learning_rate": 8.61814023870536e-07, + "loss": 0.7811, + "step": 1506 + }, + { + "epoch": 1.483267716535433, + "grad_norm": 9.752819061279297, + "learning_rate": 8.612837009464377e-07, + "loss": 0.4106, + "step": 1507 + }, + { + "epoch": 1.484251968503937, + "grad_norm": 12.476628303527832, + "learning_rate": 8.607525268285634e-07, + "loss": 0.7377, + "step": 1508 + }, + { + "epoch": 1.485236220472441, + "grad_norm": 9.569382667541504, + "learning_rate": 8.602205027741731e-07, + "loss": 0.4939, + "step": 1509 + }, + { + "epoch": 1.486220472440945, + "grad_norm": 6.528772830963135, + "learning_rate": 8.596876300425384e-07, + "loss": 0.1617, + "step": 1510 + }, + { + "epoch": 1.4872047244094488, + "grad_norm": 11.75894546508789, + "learning_rate": 8.591539098949401e-07, + "loss": 0.6942, + "step": 1511 + }, + { + "epoch": 1.4881889763779528, + "grad_norm": 13.178714752197266, + "learning_rate": 8.586193435946641e-07, + "loss": 0.7478, + "step": 1512 + }, + { + "epoch": 1.4891732283464567, + "grad_norm": 12.073831558227539, + "learning_rate": 8.580839324069998e-07, + "loss": 0.4271, + "step": 1513 + }, + { + "epoch": 1.4901574803149606, + "grad_norm": 12.120266914367676, + "learning_rate": 8.57547677599236e-07, + "loss": 0.5239, + "step": 1514 + }, + { + "epoch": 1.4911417322834646, + "grad_norm": 15.169259071350098, + "learning_rate": 8.570105804406583e-07, + "loss": 0.8726, + "step": 1515 + }, + { + "epoch": 1.4921259842519685, + "grad_norm": 17.439067840576172, + "learning_rate": 8.564726422025462e-07, + "loss": 0.6608, + "step": 1516 + }, + { + "epoch": 1.4931102362204725, + "grad_norm": 10.32081127166748, + "learning_rate": 8.5593386415817e-07, + "loss": 0.4105, + "step": 1517 + }, + { + "epoch": 1.4940944881889764, + "grad_norm": 11.749876022338867, + "learning_rate": 8.553942475827879e-07, + "loss": 0.5149, + "step": 1518 + }, + { + "epoch": 1.4950787401574803, + "grad_norm": 15.706226348876953, + "learning_rate": 8.548537937536425e-07, + "loss": 0.7481, + "step": 1519 + }, + { + "epoch": 1.4960629921259843, + "grad_norm": 12.266777038574219, + "learning_rate": 8.543125039499582e-07, + "loss": 0.4546, + "step": 1520 + }, + { + "epoch": 1.4970472440944882, + "grad_norm": 16.569011688232422, + "learning_rate": 8.537703794529387e-07, + "loss": 0.5731, + "step": 1521 + }, + { + "epoch": 1.4980314960629921, + "grad_norm": 18.642976760864258, + "learning_rate": 8.532274215457625e-07, + "loss": 0.8964, + "step": 1522 + }, + { + "epoch": 1.499015748031496, + "grad_norm": 18.387258529663086, + "learning_rate": 8.526836315135815e-07, + "loss": 0.882, + "step": 1523 + }, + { + "epoch": 1.5, + "grad_norm": 13.095115661621094, + "learning_rate": 8.521390106435168e-07, + "loss": 0.5065, + "step": 1524 + }, + { + "epoch": 1.500984251968504, + "grad_norm": 16.557979583740234, + "learning_rate": 8.515935602246558e-07, + "loss": 0.6484, + "step": 1525 + }, + { + "epoch": 1.5019685039370079, + "grad_norm": 12.26683235168457, + "learning_rate": 8.510472815480498e-07, + "loss": 0.4722, + "step": 1526 + }, + { + "epoch": 1.5029527559055118, + "grad_norm": 17.509347915649414, + "learning_rate": 8.505001759067106e-07, + "loss": 0.8768, + "step": 1527 + }, + { + "epoch": 1.5039370078740157, + "grad_norm": 16.46385383605957, + "learning_rate": 8.499522445956071e-07, + "loss": 0.8905, + "step": 1528 + }, + { + "epoch": 1.5049212598425197, + "grad_norm": 9.140239715576172, + "learning_rate": 8.494034889116626e-07, + "loss": 0.4441, + "step": 1529 + }, + { + "epoch": 1.5059055118110236, + "grad_norm": 12.472856521606445, + "learning_rate": 8.488539101537518e-07, + "loss": 0.4498, + "step": 1530 + }, + { + "epoch": 1.5059055118110236, + "eval_Qnli-dev_cosine_accuracy": 0.673828125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8457307815551758, + "eval_Qnli-dev_cosine_ap": 0.7008836391956376, + "eval_Qnli-dev_cosine_f1": 0.6863711001642037, + "eval_Qnli-dev_cosine_f1_threshold": 0.778075098991394, + "eval_Qnli-dev_cosine_precision": 0.5603217158176944, + "eval_Qnli-dev_cosine_recall": 0.885593220338983, + "eval_Qnli-dev_dot_accuracy": 0.673828125, + "eval_Qnli-dev_dot_accuracy_threshold": 649.8969116210938, + "eval_Qnli-dev_dot_ap": 0.7010672058803974, + "eval_Qnli-dev_dot_f1": 0.6863711001642037, + "eval_Qnli-dev_dot_f1_threshold": 598.0010986328125, + "eval_Qnli-dev_dot_precision": 0.5603217158176944, + "eval_Qnli-dev_dot_recall": 0.885593220338983, + "eval_Qnli-dev_euclidean_accuracy": 0.673828125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.398126602172852, + "eval_Qnli-dev_euclidean_ap": 0.7009127489013454, + "eval_Qnli-dev_euclidean_f1": 0.6852459016393443, + "eval_Qnli-dev_euclidean_f1_threshold": 18.485998153686523, + "eval_Qnli-dev_euclidean_precision": 0.5588235294117647, + "eval_Qnli-dev_euclidean_recall": 0.885593220338983, + "eval_Qnli-dev_manhattan_accuracy": 0.673828125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 332.34075927734375, + "eval_Qnli-dev_manhattan_ap": 0.7004204683248673, + "eval_Qnli-dev_manhattan_f1": 0.6873977086743045, + "eval_Qnli-dev_manhattan_f1_threshold": 399.0382995605469, + "eval_Qnli-dev_manhattan_precision": 0.56, + "eval_Qnli-dev_manhattan_recall": 0.8898305084745762, + "eval_Qnli-dev_max_accuracy": 0.673828125, + "eval_Qnli-dev_max_accuracy_threshold": 649.8969116210938, + "eval_Qnli-dev_max_ap": 0.7010672058803974, + "eval_Qnli-dev_max_f1": 0.6873977086743045, + "eval_Qnli-dev_max_f1_threshold": 598.0010986328125, + "eval_Qnli-dev_max_precision": 0.5603217158176944, + "eval_Qnli-dev_max_recall": 0.8898305084745762, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9398372769355774, + "eval_allNLI-dev_cosine_ap": 0.5227118628137193, + "eval_allNLI-dev_cosine_f1": 0.5627376425855514, + "eval_allNLI-dev_cosine_f1_threshold": 0.8421008586883545, + "eval_allNLI-dev_cosine_precision": 0.4192634560906516, + "eval_allNLI-dev_cosine_recall": 0.8554913294797688, + "eval_allNLI-dev_dot_accuracy": 0.69140625, + "eval_allNLI-dev_dot_accuracy_threshold": 721.468505859375, + "eval_allNLI-dev_dot_ap": 0.5220388994433939, + "eval_allNLI-dev_dot_f1": 0.5627376425855514, + "eval_allNLI-dev_dot_f1_threshold": 647.170166015625, + "eval_allNLI-dev_dot_precision": 0.4192634560906516, + "eval_allNLI-dev_dot_recall": 0.8554913294797688, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.617508888244629, + "eval_allNLI-dev_euclidean_ap": 0.5226874374434805, + "eval_allNLI-dev_euclidean_f1": 0.5627376425855514, + "eval_allNLI-dev_euclidean_f1_threshold": 15.578730583190918, + "eval_allNLI-dev_euclidean_precision": 0.4192634560906516, + "eval_allNLI-dev_euclidean_recall": 0.8554913294797688, + "eval_allNLI-dev_manhattan_accuracy": 0.6875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 181.53054809570312, + "eval_allNLI-dev_manhattan_ap": 0.5259334673346208, + "eval_allNLI-dev_manhattan_f1": 0.5666666666666667, + "eval_allNLI-dev_manhattan_f1_threshold": 294.49053955078125, + "eval_allNLI-dev_manhattan_precision": 0.4817813765182186, + "eval_allNLI-dev_manhattan_recall": 0.6878612716763006, + "eval_allNLI-dev_max_accuracy": 0.69140625, + "eval_allNLI-dev_max_accuracy_threshold": 721.468505859375, + "eval_allNLI-dev_max_ap": 0.5259334673346208, + "eval_allNLI-dev_max_f1": 0.5666666666666667, + "eval_allNLI-dev_max_f1_threshold": 647.170166015625, + "eval_allNLI-dev_max_precision": 0.4817813765182186, + "eval_allNLI-dev_max_recall": 0.8554913294797688, + "eval_runtime": 1.2293, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.7010672058803974, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.6718254920457604, + "eval_sts-test_pearson_dot": 0.6716092283507179, + "eval_sts-test_pearson_euclidean": 0.6944319948043691, + "eval_sts-test_pearson_manhattan": 0.6978467973295288, + "eval_sts-test_pearson_max": 0.6978467973295288, + "eval_sts-test_spearman_cosine": 0.6834985903352957, + "eval_sts-test_spearman_dot": 0.683002746359079, + "eval_sts-test_spearman_euclidean": 0.6834394739189977, + "eval_sts-test_spearman_manhattan": 0.6862889209855723, + "eval_sts-test_spearman_max": 0.6862889209855723, + "step": 1530 + }, + { + "epoch": 1.5068897637795275, + "grad_norm": 12.038389205932617, + "learning_rate": 8.483035096226975e-07, + "loss": 0.7657, + "step": 1531 + }, + { + "epoch": 1.5078740157480315, + "grad_norm": 15.592117309570312, + "learning_rate": 8.477522886212673e-07, + "loss": 0.544, + "step": 1532 + }, + { + "epoch": 1.5088582677165354, + "grad_norm": 11.37441635131836, + "learning_rate": 8.472002484541711e-07, + "loss": 0.35, + "step": 1533 + }, + { + "epoch": 1.5098425196850394, + "grad_norm": 9.020292282104492, + "learning_rate": 8.466473904280579e-07, + "loss": 0.3211, + "step": 1534 + }, + { + "epoch": 1.5108267716535433, + "grad_norm": 11.615494728088379, + "learning_rate": 8.46093715851512e-07, + "loss": 0.4795, + "step": 1535 + }, + { + "epoch": 1.5118110236220472, + "grad_norm": 9.721772193908691, + "learning_rate": 8.455392260350507e-07, + "loss": 0.4776, + "step": 1536 + }, + { + "epoch": 1.5127952755905512, + "grad_norm": 10.600958824157715, + "learning_rate": 8.449839222911214e-07, + "loss": 0.4259, + "step": 1537 + }, + { + "epoch": 1.513779527559055, + "grad_norm": 7.062549114227295, + "learning_rate": 8.444278059340971e-07, + "loss": 0.1436, + "step": 1538 + }, + { + "epoch": 1.514763779527559, + "grad_norm": 13.764479637145996, + "learning_rate": 8.438708782802748e-07, + "loss": 0.5271, + "step": 1539 + }, + { + "epoch": 1.515748031496063, + "grad_norm": 12.343178749084473, + "learning_rate": 8.43313140647872e-07, + "loss": 0.4446, + "step": 1540 + }, + { + "epoch": 1.516732283464567, + "grad_norm": 14.525405883789062, + "learning_rate": 8.427545943570226e-07, + "loss": 0.3741, + "step": 1541 + }, + { + "epoch": 1.5177165354330708, + "grad_norm": 10.17180347442627, + "learning_rate": 8.421952407297752e-07, + "loss": 0.5234, + "step": 1542 + }, + { + "epoch": 1.5187007874015748, + "grad_norm": 10.825239181518555, + "learning_rate": 8.416350810900892e-07, + "loss": 0.3489, + "step": 1543 + }, + { + "epoch": 1.5196850393700787, + "grad_norm": 12.134411811828613, + "learning_rate": 8.410741167638317e-07, + "loss": 0.3911, + "step": 1544 + }, + { + "epoch": 1.5206692913385826, + "grad_norm": 12.819040298461914, + "learning_rate": 8.405123490787741e-07, + "loss": 0.5522, + "step": 1545 + }, + { + "epoch": 1.5216535433070866, + "grad_norm": 7.938567638397217, + "learning_rate": 8.3994977936459e-07, + "loss": 0.2651, + "step": 1546 + }, + { + "epoch": 1.5226377952755905, + "grad_norm": 11.559355735778809, + "learning_rate": 8.393864089528514e-07, + "loss": 0.4473, + "step": 1547 + }, + { + "epoch": 1.5236220472440944, + "grad_norm": 13.801612854003906, + "learning_rate": 8.388222391770245e-07, + "loss": 0.4625, + "step": 1548 + }, + { + "epoch": 1.5246062992125984, + "grad_norm": 14.334702491760254, + "learning_rate": 8.382572713724683e-07, + "loss": 0.6767, + "step": 1549 + }, + { + "epoch": 1.5255905511811023, + "grad_norm": 12.202241897583008, + "learning_rate": 8.376915068764308e-07, + "loss": 0.4444, + "step": 1550 + }, + { + "epoch": 1.5265748031496063, + "grad_norm": 11.767269134521484, + "learning_rate": 8.371249470280453e-07, + "loss": 0.3958, + "step": 1551 + }, + { + "epoch": 1.5275590551181102, + "grad_norm": 12.902238845825195, + "learning_rate": 8.36557593168328e-07, + "loss": 0.3662, + "step": 1552 + }, + { + "epoch": 1.5285433070866141, + "grad_norm": 14.114974975585938, + "learning_rate": 8.359894466401743e-07, + "loss": 0.6872, + "step": 1553 + }, + { + "epoch": 1.529527559055118, + "grad_norm": 15.278046607971191, + "learning_rate": 8.354205087883556e-07, + "loss": 0.6356, + "step": 1554 + }, + { + "epoch": 1.530511811023622, + "grad_norm": 14.643278121948242, + "learning_rate": 8.348507809595169e-07, + "loss": 0.5265, + "step": 1555 + }, + { + "epoch": 1.531496062992126, + "grad_norm": 15.246484756469727, + "learning_rate": 8.342802645021727e-07, + "loss": 0.596, + "step": 1556 + }, + { + "epoch": 1.53248031496063, + "grad_norm": 7.831085205078125, + "learning_rate": 8.337089607667039e-07, + "loss": 0.1691, + "step": 1557 + }, + { + "epoch": 1.5334645669291338, + "grad_norm": 19.259780883789062, + "learning_rate": 8.331368711053552e-07, + "loss": 1.0575, + "step": 1558 + }, + { + "epoch": 1.534448818897638, + "grad_norm": 18.77946662902832, + "learning_rate": 8.325639968722313e-07, + "loss": 1.1115, + "step": 1559 + }, + { + "epoch": 1.5354330708661417, + "grad_norm": 18.560611724853516, + "learning_rate": 8.319903394232942e-07, + "loss": 0.7607, + "step": 1560 + }, + { + "epoch": 1.5364173228346458, + "grad_norm": 13.974424362182617, + "learning_rate": 8.314159001163596e-07, + "loss": 0.5262, + "step": 1561 + }, + { + "epoch": 1.5374015748031495, + "grad_norm": 16.191749572753906, + "learning_rate": 8.308406803110936e-07, + "loss": 0.7648, + "step": 1562 + }, + { + "epoch": 1.5383858267716537, + "grad_norm": 11.78024673461914, + "learning_rate": 8.302646813690101e-07, + "loss": 0.5331, + "step": 1563 + }, + { + "epoch": 1.5393700787401574, + "grad_norm": 11.094069480895996, + "learning_rate": 8.296879046534668e-07, + "loss": 0.4568, + "step": 1564 + }, + { + "epoch": 1.5403543307086616, + "grad_norm": 10.832902908325195, + "learning_rate": 8.291103515296628e-07, + "loss": 0.4017, + "step": 1565 + }, + { + "epoch": 1.5413385826771653, + "grad_norm": 15.592805862426758, + "learning_rate": 8.285320233646345e-07, + "loss": 0.891, + "step": 1566 + }, + { + "epoch": 1.5423228346456694, + "grad_norm": 8.908652305603027, + "learning_rate": 8.279529215272526e-07, + "loss": 0.3974, + "step": 1567 + }, + { + "epoch": 1.5433070866141732, + "grad_norm": 13.945334434509277, + "learning_rate": 8.273730473882198e-07, + "loss": 0.8785, + "step": 1568 + }, + { + "epoch": 1.5442913385826773, + "grad_norm": 8.680691719055176, + "learning_rate": 8.267924023200662e-07, + "loss": 0.2573, + "step": 1569 + }, + { + "epoch": 1.545275590551181, + "grad_norm": 13.301336288452148, + "learning_rate": 8.262109876971466e-07, + "loss": 0.6737, + "step": 1570 + }, + { + "epoch": 1.5462598425196852, + "grad_norm": 10.233235359191895, + "learning_rate": 8.256288048956379e-07, + "loss": 0.3374, + "step": 1571 + }, + { + "epoch": 1.547244094488189, + "grad_norm": 8.64468765258789, + "learning_rate": 8.250458552935343e-07, + "loss": 0.2978, + "step": 1572 + }, + { + "epoch": 1.548228346456693, + "grad_norm": 14.032349586486816, + "learning_rate": 8.244621402706462e-07, + "loss": 0.6462, + "step": 1573 + }, + { + "epoch": 1.5492125984251968, + "grad_norm": 6.8755059242248535, + "learning_rate": 8.238776612085943e-07, + "loss": 0.1323, + "step": 1574 + }, + { + "epoch": 1.550196850393701, + "grad_norm": 9.783378601074219, + "learning_rate": 8.232924194908088e-07, + "loss": 0.2862, + "step": 1575 + }, + { + "epoch": 1.5511811023622046, + "grad_norm": 13.311777114868164, + "learning_rate": 8.227064165025247e-07, + "loss": 0.4854, + "step": 1576 + }, + { + "epoch": 1.5521653543307088, + "grad_norm": 13.517067909240723, + "learning_rate": 8.221196536307788e-07, + "loss": 0.4033, + "step": 1577 + }, + { + "epoch": 1.5531496062992125, + "grad_norm": 21.332433700561523, + "learning_rate": 8.215321322644067e-07, + "loss": 0.7233, + "step": 1578 + }, + { + "epoch": 1.5541338582677167, + "grad_norm": 16.345102310180664, + "learning_rate": 8.209438537940393e-07, + "loss": 0.5184, + "step": 1579 + }, + { + "epoch": 1.5551181102362204, + "grad_norm": 24.670753479003906, + "learning_rate": 8.203548196120994e-07, + "loss": 0.9382, + "step": 1580 + }, + { + "epoch": 1.5561023622047245, + "grad_norm": 18.061687469482422, + "learning_rate": 8.197650311127981e-07, + "loss": 0.6416, + "step": 1581 + }, + { + "epoch": 1.5570866141732282, + "grad_norm": 17.43946075439453, + "learning_rate": 8.191744896921331e-07, + "loss": 0.6853, + "step": 1582 + }, + { + "epoch": 1.5580708661417324, + "grad_norm": 19.403583526611328, + "learning_rate": 8.185831967478831e-07, + "loss": 0.604, + "step": 1583 + }, + { + "epoch": 1.5590551181102361, + "grad_norm": 13.832499504089355, + "learning_rate": 8.179911536796062e-07, + "loss": 0.498, + "step": 1584 + }, + { + "epoch": 1.5600393700787403, + "grad_norm": 8.03169059753418, + "learning_rate": 8.173983618886358e-07, + "loss": 0.3023, + "step": 1585 + }, + { + "epoch": 1.561023622047244, + "grad_norm": 10.473389625549316, + "learning_rate": 8.168048227780775e-07, + "loss": 0.2496, + "step": 1586 + }, + { + "epoch": 1.5620078740157481, + "grad_norm": 18.29000473022461, + "learning_rate": 8.162105377528057e-07, + "loss": 0.8291, + "step": 1587 + }, + { + "epoch": 1.5629921259842519, + "grad_norm": 12.872581481933594, + "learning_rate": 8.156155082194607e-07, + "loss": 0.4926, + "step": 1588 + }, + { + "epoch": 1.563976377952756, + "grad_norm": 17.28192710876465, + "learning_rate": 8.150197355864446e-07, + "loss": 0.6343, + "step": 1589 + }, + { + "epoch": 1.5649606299212597, + "grad_norm": 17.518335342407227, + "learning_rate": 8.144232212639184e-07, + "loss": 0.569, + "step": 1590 + }, + { + "epoch": 1.5659448818897639, + "grad_norm": 10.588553428649902, + "learning_rate": 8.138259666637987e-07, + "loss": 0.21, + "step": 1591 + }, + { + "epoch": 1.5669291338582676, + "grad_norm": 11.502038955688477, + "learning_rate": 8.132279731997544e-07, + "loss": 0.3462, + "step": 1592 + }, + { + "epoch": 1.5679133858267718, + "grad_norm": 15.024426460266113, + "learning_rate": 8.126292422872034e-07, + "loss": 0.6391, + "step": 1593 + }, + { + "epoch": 1.5688976377952755, + "grad_norm": 9.6860933303833, + "learning_rate": 8.120297753433087e-07, + "loss": 0.2757, + "step": 1594 + }, + { + "epoch": 1.5698818897637796, + "grad_norm": 12.575701713562012, + "learning_rate": 8.114295737869759e-07, + "loss": 0.4453, + "step": 1595 + }, + { + "epoch": 1.5708661417322833, + "grad_norm": 11.9063720703125, + "learning_rate": 8.108286390388487e-07, + "loss": 0.3759, + "step": 1596 + }, + { + "epoch": 1.5718503937007875, + "grad_norm": 11.738770484924316, + "learning_rate": 8.102269725213071e-07, + "loss": 0.5071, + "step": 1597 + }, + { + "epoch": 1.5728346456692912, + "grad_norm": 7.347543239593506, + "learning_rate": 8.096245756584625e-07, + "loss": 0.1958, + "step": 1598 + }, + { + "epoch": 1.5738188976377954, + "grad_norm": 9.17516803741455, + "learning_rate": 8.090214498761551e-07, + "loss": 0.3417, + "step": 1599 + }, + { + "epoch": 1.574803149606299, + "grad_norm": 14.779362678527832, + "learning_rate": 8.084175966019507e-07, + "loss": 0.4496, + "step": 1600 + }, + { + "epoch": 1.5757874015748032, + "grad_norm": 13.17792797088623, + "learning_rate": 8.078130172651368e-07, + "loss": 0.5615, + "step": 1601 + }, + { + "epoch": 1.576771653543307, + "grad_norm": 10.993749618530273, + "learning_rate": 8.072077132967193e-07, + "loss": 0.4346, + "step": 1602 + }, + { + "epoch": 1.577755905511811, + "grad_norm": 9.44127082824707, + "learning_rate": 8.066016861294198e-07, + "loss": 0.3385, + "step": 1603 + }, + { + "epoch": 1.5787401574803148, + "grad_norm": 13.825623512268066, + "learning_rate": 8.059949371976711e-07, + "loss": 0.6684, + "step": 1604 + }, + { + "epoch": 1.579724409448819, + "grad_norm": 13.196465492248535, + "learning_rate": 8.053874679376147e-07, + "loss": 0.3634, + "step": 1605 + }, + { + "epoch": 1.5807086614173227, + "grad_norm": 16.361576080322266, + "learning_rate": 8.047792797870968e-07, + "loss": 0.6762, + "step": 1606 + }, + { + "epoch": 1.5816929133858268, + "grad_norm": 11.868363380432129, + "learning_rate": 8.041703741856657e-07, + "loss": 0.3477, + "step": 1607 + }, + { + "epoch": 1.5826771653543306, + "grad_norm": 12.077712059020996, + "learning_rate": 8.035607525745672e-07, + "loss": 0.5679, + "step": 1608 + }, + { + "epoch": 1.5836614173228347, + "grad_norm": 16.465709686279297, + "learning_rate": 8.029504163967423e-07, + "loss": 0.7458, + "step": 1609 + }, + { + "epoch": 1.5846456692913384, + "grad_norm": 16.391393661499023, + "learning_rate": 8.023393670968235e-07, + "loss": 0.7871, + "step": 1610 + }, + { + "epoch": 1.5856299212598426, + "grad_norm": 11.737066268920898, + "learning_rate": 8.017276061211303e-07, + "loss": 0.3641, + "step": 1611 + }, + { + "epoch": 1.5866141732283463, + "grad_norm": 16.41512107849121, + "learning_rate": 8.011151349176681e-07, + "loss": 0.6641, + "step": 1612 + }, + { + "epoch": 1.5875984251968505, + "grad_norm": 21.22294044494629, + "learning_rate": 8.005019549361222e-07, + "loss": 1.1934, + "step": 1613 + }, + { + "epoch": 1.5885826771653542, + "grad_norm": 11.653447151184082, + "learning_rate": 7.998880676278558e-07, + "loss": 0.3024, + "step": 1614 + }, + { + "epoch": 1.5895669291338583, + "grad_norm": 10.337379455566406, + "learning_rate": 7.992734744459068e-07, + "loss": 0.3846, + "step": 1615 + }, + { + "epoch": 1.590551181102362, + "grad_norm": 15.340450286865234, + "learning_rate": 7.986581768449833e-07, + "loss": 0.6288, + "step": 1616 + }, + { + "epoch": 1.5915354330708662, + "grad_norm": 12.190918922424316, + "learning_rate": 7.98042176281461e-07, + "loss": 0.4733, + "step": 1617 + }, + { + "epoch": 1.59251968503937, + "grad_norm": 9.841829299926758, + "learning_rate": 7.974254742133795e-07, + "loss": 0.2775, + "step": 1618 + }, + { + "epoch": 1.593503937007874, + "grad_norm": 15.595701217651367, + "learning_rate": 7.968080721004386e-07, + "loss": 0.6975, + "step": 1619 + }, + { + "epoch": 1.594488188976378, + "grad_norm": 9.914538383483887, + "learning_rate": 7.961899714039954e-07, + "loss": 0.317, + "step": 1620 + }, + { + "epoch": 1.595472440944882, + "grad_norm": 12.646726608276367, + "learning_rate": 7.955711735870601e-07, + "loss": 0.5374, + "step": 1621 + }, + { + "epoch": 1.5964566929133859, + "grad_norm": 17.816314697265625, + "learning_rate": 7.949516801142932e-07, + "loss": 0.8305, + "step": 1622 + }, + { + "epoch": 1.5974409448818898, + "grad_norm": 13.830299377441406, + "learning_rate": 7.943314924520017e-07, + "loss": 0.56, + "step": 1623 + }, + { + "epoch": 1.5984251968503937, + "grad_norm": 11.648980140686035, + "learning_rate": 7.93710612068136e-07, + "loss": 0.5452, + "step": 1624 + }, + { + "epoch": 1.5994094488188977, + "grad_norm": 13.336215019226074, + "learning_rate": 7.930890404322857e-07, + "loss": 0.602, + "step": 1625 + }, + { + "epoch": 1.6003937007874016, + "grad_norm": 14.171172142028809, + "learning_rate": 7.924667790156766e-07, + "loss": 0.7525, + "step": 1626 + }, + { + "epoch": 1.6013779527559056, + "grad_norm": 11.074429512023926, + "learning_rate": 7.918438292911674e-07, + "loss": 0.3503, + "step": 1627 + }, + { + "epoch": 1.6023622047244095, + "grad_norm": 9.00805950164795, + "learning_rate": 7.912201927332461e-07, + "loss": 0.3675, + "step": 1628 + }, + { + "epoch": 1.6033464566929134, + "grad_norm": 13.804633140563965, + "learning_rate": 7.905958708180258e-07, + "loss": 0.7333, + "step": 1629 + }, + { + "epoch": 1.6043307086614174, + "grad_norm": 13.033793449401855, + "learning_rate": 7.899708650232426e-07, + "loss": 0.4482, + "step": 1630 + }, + { + "epoch": 1.6053149606299213, + "grad_norm": 10.6852388381958, + "learning_rate": 7.893451768282508e-07, + "loss": 0.315, + "step": 1631 + }, + { + "epoch": 1.6062992125984252, + "grad_norm": 8.711305618286133, + "learning_rate": 7.8871880771402e-07, + "loss": 0.2173, + "step": 1632 + }, + { + "epoch": 1.6072834645669292, + "grad_norm": 19.138879776000977, + "learning_rate": 7.880917591631312e-07, + "loss": 1.0407, + "step": 1633 + }, + { + "epoch": 1.608267716535433, + "grad_norm": 12.271173477172852, + "learning_rate": 7.874640326597741e-07, + "loss": 0.5099, + "step": 1634 + }, + { + "epoch": 1.609251968503937, + "grad_norm": 16.982139587402344, + "learning_rate": 7.868356296897431e-07, + "loss": 0.8723, + "step": 1635 + }, + { + "epoch": 1.610236220472441, + "grad_norm": 9.335005760192871, + "learning_rate": 7.862065517404334e-07, + "loss": 0.2828, + "step": 1636 + }, + { + "epoch": 1.611220472440945, + "grad_norm": 8.913496971130371, + "learning_rate": 7.855768003008376e-07, + "loss": 0.3407, + "step": 1637 + }, + { + "epoch": 1.6122047244094488, + "grad_norm": 9.617837905883789, + "learning_rate": 7.849463768615432e-07, + "loss": 0.2632, + "step": 1638 + }, + { + "epoch": 1.6131889763779528, + "grad_norm": 8.818242073059082, + "learning_rate": 7.843152829147278e-07, + "loss": 0.2201, + "step": 1639 + }, + { + "epoch": 1.6141732283464567, + "grad_norm": 12.556653022766113, + "learning_rate": 7.836835199541562e-07, + "loss": 0.3298, + "step": 1640 + }, + { + "epoch": 1.6151574803149606, + "grad_norm": 8.544206619262695, + "learning_rate": 7.830510894751764e-07, + "loss": 0.2171, + "step": 1641 + }, + { + "epoch": 1.6161417322834646, + "grad_norm": 12.767942428588867, + "learning_rate": 7.824179929747168e-07, + "loss": 0.418, + "step": 1642 + }, + { + "epoch": 1.6171259842519685, + "grad_norm": 15.299103736877441, + "learning_rate": 7.817842319512818e-07, + "loss": 0.8353, + "step": 1643 + }, + { + "epoch": 1.6181102362204725, + "grad_norm": 14.117609977722168, + "learning_rate": 7.811498079049492e-07, + "loss": 0.4935, + "step": 1644 + }, + { + "epoch": 1.6190944881889764, + "grad_norm": 14.078787803649902, + "learning_rate": 7.805147223373658e-07, + "loss": 0.5569, + "step": 1645 + }, + { + "epoch": 1.6200787401574803, + "grad_norm": 9.347455024719238, + "learning_rate": 7.798789767517444e-07, + "loss": 0.2752, + "step": 1646 + }, + { + "epoch": 1.6210629921259843, + "grad_norm": 12.384529113769531, + "learning_rate": 7.792425726528596e-07, + "loss": 0.3603, + "step": 1647 + }, + { + "epoch": 1.6220472440944882, + "grad_norm": 10.888153076171875, + "learning_rate": 7.786055115470454e-07, + "loss": 0.4289, + "step": 1648 + }, + { + "epoch": 1.6230314960629921, + "grad_norm": 13.588855743408203, + "learning_rate": 7.779677949421902e-07, + "loss": 0.6225, + "step": 1649 + }, + { + "epoch": 1.624015748031496, + "grad_norm": 13.192461967468262, + "learning_rate": 7.77329424347734e-07, + "loss": 0.3835, + "step": 1650 + }, + { + "epoch": 1.625, + "grad_norm": 10.339590072631836, + "learning_rate": 7.766904012746653e-07, + "loss": 0.1686, + "step": 1651 + }, + { + "epoch": 1.625984251968504, + "grad_norm": 13.078779220581055, + "learning_rate": 7.760507272355165e-07, + "loss": 0.5061, + "step": 1652 + }, + { + "epoch": 1.6269685039370079, + "grad_norm": 8.69745922088623, + "learning_rate": 7.754104037443609e-07, + "loss": 0.2085, + "step": 1653 + }, + { + "epoch": 1.6279527559055118, + "grad_norm": 11.973641395568848, + "learning_rate": 7.747694323168091e-07, + "loss": 0.3367, + "step": 1654 + }, + { + "epoch": 1.6289370078740157, + "grad_norm": 14.082365036010742, + "learning_rate": 7.741278144700052e-07, + "loss": 0.5719, + "step": 1655 + }, + { + "epoch": 1.6299212598425197, + "grad_norm": 19.6208438873291, + "learning_rate": 7.734855517226234e-07, + "loss": 1.073, + "step": 1656 + }, + { + "epoch": 1.6309055118110236, + "grad_norm": 9.179506301879883, + "learning_rate": 7.728426455948644e-07, + "loss": 0.1355, + "step": 1657 + }, + { + "epoch": 1.6318897637795275, + "grad_norm": 9.421241760253906, + "learning_rate": 7.721990976084517e-07, + "loss": 0.4165, + "step": 1658 + }, + { + "epoch": 1.6328740157480315, + "grad_norm": 15.605162620544434, + "learning_rate": 7.715549092866282e-07, + "loss": 0.831, + "step": 1659 + }, + { + "epoch": 1.6338582677165354, + "grad_norm": 11.147396087646484, + "learning_rate": 7.70910082154152e-07, + "loss": 0.4592, + "step": 1660 + }, + { + "epoch": 1.6348425196850394, + "grad_norm": 10.895313262939453, + "learning_rate": 7.702646177372938e-07, + "loss": 0.4908, + "step": 1661 + }, + { + "epoch": 1.6358267716535433, + "grad_norm": 14.701689720153809, + "learning_rate": 7.696185175638325e-07, + "loss": 0.7708, + "step": 1662 + }, + { + "epoch": 1.6368110236220472, + "grad_norm": 16.108003616333008, + "learning_rate": 7.689717831630513e-07, + "loss": 0.5747, + "step": 1663 + }, + { + "epoch": 1.6377952755905512, + "grad_norm": 5.823051929473877, + "learning_rate": 7.683244160657353e-07, + "loss": 0.2344, + "step": 1664 + }, + { + "epoch": 1.638779527559055, + "grad_norm": 11.863076210021973, + "learning_rate": 7.676764178041668e-07, + "loss": 0.4998, + "step": 1665 + }, + { + "epoch": 1.639763779527559, + "grad_norm": 7.952276706695557, + "learning_rate": 7.670277899121223e-07, + "loss": 0.1832, + "step": 1666 + }, + { + "epoch": 1.640748031496063, + "grad_norm": 14.000992774963379, + "learning_rate": 7.663785339248683e-07, + "loss": 0.4653, + "step": 1667 + }, + { + "epoch": 1.641732283464567, + "grad_norm": 8.010464668273926, + "learning_rate": 7.657286513791578e-07, + "loss": 0.2594, + "step": 1668 + }, + { + "epoch": 1.6427165354330708, + "grad_norm": 11.359638214111328, + "learning_rate": 7.650781438132276e-07, + "loss": 0.4378, + "step": 1669 + }, + { + "epoch": 1.6437007874015748, + "grad_norm": 12.349568367004395, + "learning_rate": 7.644270127667932e-07, + "loss": 0.3691, + "step": 1670 + }, + { + "epoch": 1.6446850393700787, + "grad_norm": 10.976561546325684, + "learning_rate": 7.637752597810456e-07, + "loss": 0.2519, + "step": 1671 + }, + { + "epoch": 1.6456692913385826, + "grad_norm": 13.04942512512207, + "learning_rate": 7.631228863986491e-07, + "loss": 0.5589, + "step": 1672 + }, + { + "epoch": 1.6466535433070866, + "grad_norm": 10.682896614074707, + "learning_rate": 7.624698941637352e-07, + "loss": 0.2041, + "step": 1673 + }, + { + "epoch": 1.6476377952755905, + "grad_norm": 12.731783866882324, + "learning_rate": 7.618162846219006e-07, + "loss": 0.3355, + "step": 1674 + }, + { + "epoch": 1.6486220472440944, + "grad_norm": 18.489307403564453, + "learning_rate": 7.611620593202037e-07, + "loss": 0.6531, + "step": 1675 + }, + { + "epoch": 1.6496062992125984, + "grad_norm": 8.746728897094727, + "learning_rate": 7.605072198071592e-07, + "loss": 0.2044, + "step": 1676 + }, + { + "epoch": 1.6505905511811023, + "grad_norm": 13.30772590637207, + "learning_rate": 7.598517676327368e-07, + "loss": 0.4636, + "step": 1677 + }, + { + "epoch": 1.6515748031496063, + "grad_norm": 17.488759994506836, + "learning_rate": 7.591957043483555e-07, + "loss": 0.4883, + "step": 1678 + }, + { + "epoch": 1.6525590551181102, + "grad_norm": 16.78904151916504, + "learning_rate": 7.585390315068812e-07, + "loss": 1.1288, + "step": 1679 + }, + { + "epoch": 1.6535433070866141, + "grad_norm": 15.560281753540039, + "learning_rate": 7.578817506626224e-07, + "loss": 0.5687, + "step": 1680 + }, + { + "epoch": 1.654527559055118, + "grad_norm": 15.419073104858398, + "learning_rate": 7.572238633713268e-07, + "loss": 0.6162, + "step": 1681 + }, + { + "epoch": 1.655511811023622, + "grad_norm": 10.971380233764648, + "learning_rate": 7.565653711901778e-07, + "loss": 0.2443, + "step": 1682 + }, + { + "epoch": 1.656496062992126, + "grad_norm": 15.768486976623535, + "learning_rate": 7.559062756777898e-07, + "loss": 0.9008, + "step": 1683 + }, + { + "epoch": 1.656496062992126, + "eval_Qnli-dev_cosine_accuracy": 0.6796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8366684317588806, + "eval_Qnli-dev_cosine_ap": 0.7025711036957332, + "eval_Qnli-dev_cosine_f1": 0.6749611197511665, + "eval_Qnli-dev_cosine_f1_threshold": 0.7199565172195435, + "eval_Qnli-dev_cosine_precision": 0.5331695331695332, + "eval_Qnli-dev_cosine_recall": 0.9194915254237288, + "eval_Qnli-dev_dot_accuracy": 0.6796875, + "eval_Qnli-dev_dot_accuracy_threshold": 643.223388671875, + "eval_Qnli-dev_dot_ap": 0.7028877754827524, + "eval_Qnli-dev_dot_f1": 0.673913043478261, + "eval_Qnli-dev_dot_f1_threshold": 552.9798583984375, + "eval_Qnli-dev_dot_precision": 0.5318627450980392, + "eval_Qnli-dev_dot_recall": 0.9194915254237288, + "eval_Qnli-dev_euclidean_accuracy": 0.6796875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.844499588012695, + "eval_Qnli-dev_euclidean_ap": 0.7026118478627446, + "eval_Qnli-dev_euclidean_f1": 0.6749611197511665, + "eval_Qnli-dev_euclidean_f1_threshold": 20.740978240966797, + "eval_Qnli-dev_euclidean_precision": 0.5331695331695332, + "eval_Qnli-dev_euclidean_recall": 0.9194915254237288, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 335.95025634765625, + "eval_Qnli-dev_manhattan_ap": 0.7026678687600421, + "eval_Qnli-dev_manhattan_f1": 0.6774716369529984, + "eval_Qnli-dev_manhattan_f1_threshold": 431.2140808105469, + "eval_Qnli-dev_manhattan_precision": 0.5485564304461942, + "eval_Qnli-dev_manhattan_recall": 0.885593220338983, + "eval_Qnli-dev_max_accuracy": 0.6796875, + "eval_Qnli-dev_max_accuracy_threshold": 643.223388671875, + "eval_Qnli-dev_max_ap": 0.7028877754827524, + "eval_Qnli-dev_max_f1": 0.6774716369529984, + "eval_Qnli-dev_max_f1_threshold": 552.9798583984375, + "eval_Qnli-dev_max_precision": 0.5485564304461942, + "eval_Qnli-dev_max_recall": 0.9194915254237288, + "eval_allNLI-dev_cosine_accuracy": 0.697265625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9135359525680542, + "eval_allNLI-dev_cosine_ap": 0.5215030688158616, + "eval_allNLI-dev_cosine_f1": 0.5375854214123007, + "eval_allNLI-dev_cosine_f1_threshold": 0.8361715078353882, + "eval_allNLI-dev_cosine_precision": 0.44360902255639095, + "eval_allNLI-dev_cosine_recall": 0.6820809248554913, + "eval_allNLI-dev_dot_accuracy": 0.697265625, + "eval_allNLI-dev_dot_accuracy_threshold": 702.328125, + "eval_allNLI-dev_dot_ap": 0.5215320356910601, + "eval_allNLI-dev_dot_f1": 0.5375854214123007, + "eval_allNLI-dev_dot_f1_threshold": 642.5977783203125, + "eval_allNLI-dev_dot_precision": 0.44360902255639095, + "eval_allNLI-dev_dot_recall": 0.6820809248554913, + "eval_allNLI-dev_euclidean_accuracy": 0.697265625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.530282020568848, + "eval_allNLI-dev_euclidean_ap": 0.5215763814962693, + "eval_allNLI-dev_euclidean_f1": 0.5375854214123007, + "eval_allNLI-dev_euclidean_f1_threshold": 15.868341445922852, + "eval_allNLI-dev_euclidean_precision": 0.44360902255639095, + "eval_allNLI-dev_euclidean_recall": 0.6820809248554913, + "eval_allNLI-dev_manhattan_accuracy": 0.705078125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 246.65103149414062, + "eval_allNLI-dev_manhattan_ap": 0.5262241457144412, + "eval_allNLI-dev_manhattan_f1": 0.5364705882352941, + "eval_allNLI-dev_manhattan_f1_threshold": 332.9952392578125, + "eval_allNLI-dev_manhattan_precision": 0.4523809523809524, + "eval_allNLI-dev_manhattan_recall": 0.6589595375722543, + "eval_allNLI-dev_max_accuracy": 0.705078125, + "eval_allNLI-dev_max_accuracy_threshold": 702.328125, + "eval_allNLI-dev_max_ap": 0.5262241457144412, + "eval_allNLI-dev_max_f1": 0.5375854214123007, + "eval_allNLI-dev_max_f1_threshold": 642.5977783203125, + "eval_allNLI-dev_max_precision": 0.4523809523809524, + "eval_allNLI-dev_max_recall": 0.6820809248554913, + "eval_runtime": 1.2192, + "eval_samples_per_second": 0.0, + "eval_sequential_score": 0.7028877754827524, + "eval_steps_per_second": 0.0, + "eval_sts-test_pearson_cosine": 0.6764054032395426, + "eval_sts-test_pearson_dot": 0.676046362499036, + "eval_sts-test_pearson_euclidean": 0.6974498133985872, + "eval_sts-test_pearson_manhattan": 0.700820653979197, + "eval_sts-test_pearson_max": 0.700820653979197, + "eval_sts-test_spearman_cosine": 0.6929200286447658, + "eval_sts-test_spearman_dot": 0.692721109287881, + "eval_sts-test_spearman_euclidean": 0.6929374816367083, + "eval_sts-test_spearman_manhattan": 0.6972355270345225, + "eval_sts-test_spearman_max": 0.6972355270345225, + "step": 1683 + }, + { + "epoch": 1.65748031496063, + "grad_norm": 11.8477201461792, + "learning_rate": 7.552465783942058e-07, + "loss": 0.344, + "step": 1684 + }, + { + "epoch": 1.6584645669291338, + "grad_norm": 9.94332504272461, + "learning_rate": 7.545862809008931e-07, + "loss": 0.2586, + "step": 1685 + }, + { + "epoch": 1.659448818897638, + "grad_norm": 15.873083114624023, + "learning_rate": 7.539253847607397e-07, + "loss": 0.6468, + "step": 1686 + }, + { + "epoch": 1.6604330708661417, + "grad_norm": 15.39318561553955, + "learning_rate": 7.532638915380503e-07, + "loss": 0.8279, + "step": 1687 + }, + { + "epoch": 1.6614173228346458, + "grad_norm": 8.56396770477295, + "learning_rate": 7.526018027985434e-07, + "loss": 0.3112, + "step": 1688 + }, + { + "epoch": 1.6624015748031495, + "grad_norm": 14.693928718566895, + "learning_rate": 7.519391201093462e-07, + "loss": 0.6875, + "step": 1689 + }, + { + "epoch": 1.6633858267716537, + "grad_norm": 11.19882583618164, + "learning_rate": 7.512758450389925e-07, + "loss": 0.231, + "step": 1690 + }, + { + "epoch": 1.6643700787401574, + "grad_norm": 7.3683319091796875, + "learning_rate": 7.50611979157418e-07, + "loss": 0.235, + "step": 1691 + }, + { + "epoch": 1.6653543307086616, + "grad_norm": 11.351162910461426, + "learning_rate": 7.499475240359567e-07, + "loss": 0.347, + "step": 1692 + }, + { + "epoch": 1.6663385826771653, + "grad_norm": 15.09561538696289, + "learning_rate": 7.492824812473376e-07, + "loss": 0.5593, + "step": 1693 + }, + { + "epoch": 1.6673228346456694, + "grad_norm": 13.831792831420898, + "learning_rate": 7.4861685236568e-07, + "loss": 0.4908, + "step": 1694 + }, + { + "epoch": 1.6683070866141732, + "grad_norm": 11.684814453125, + "learning_rate": 7.479506389664914e-07, + "loss": 0.3652, + "step": 1695 + }, + { + "epoch": 1.6692913385826773, + "grad_norm": 10.993673324584961, + "learning_rate": 7.47283842626662e-07, + "loss": 0.3771, + "step": 1696 + }, + { + "epoch": 1.670275590551181, + "grad_norm": 9.902570724487305, + "learning_rate": 7.466164649244625e-07, + "loss": 0.3547, + "step": 1697 + }, + { + "epoch": 1.6712598425196852, + "grad_norm": 11.589768409729004, + "learning_rate": 7.45948507439539e-07, + "loss": 0.3422, + "step": 1698 + }, + { + "epoch": 1.672244094488189, + "grad_norm": 11.524438858032227, + "learning_rate": 7.452799717529102e-07, + "loss": 0.5911, + "step": 1699 + }, + { + "epoch": 1.673228346456693, + "grad_norm": 14.690237045288086, + "learning_rate": 7.446108594469634e-07, + "loss": 0.5544, + "step": 1700 + }, + { + "epoch": 1.6742125984251968, + "grad_norm": 9.880528450012207, + "learning_rate": 7.439411721054508e-07, + "loss": 0.295, + "step": 1701 + }, + { + "epoch": 1.675196850393701, + "grad_norm": 13.177597999572754, + "learning_rate": 7.432709113134856e-07, + "loss": 0.4211, + "step": 1702 + }, + { + "epoch": 1.6761811023622046, + "grad_norm": 11.107061386108398, + "learning_rate": 7.42600078657538e-07, + "loss": 0.3966, + "step": 1703 + }, + { + "epoch": 1.6771653543307088, + "grad_norm": 14.556212425231934, + "learning_rate": 7.419286757254324e-07, + "loss": 0.7783, + "step": 1704 + }, + { + "epoch": 1.6781496062992125, + "grad_norm": 15.074748039245605, + "learning_rate": 7.412567041063424e-07, + "loss": 0.7513, + "step": 1705 + }, + { + "epoch": 1.6791338582677167, + "grad_norm": 13.13559627532959, + "learning_rate": 7.40584165390788e-07, + "loss": 0.7145, + "step": 1706 + }, + { + "epoch": 1.6801181102362204, + "grad_norm": 16.61490249633789, + "learning_rate": 7.399110611706316e-07, + "loss": 0.8779, + "step": 1707 + }, + { + "epoch": 1.6811023622047245, + "grad_norm": 10.710002899169922, + "learning_rate": 7.392373930390738e-07, + "loss": 0.2505, + "step": 1708 + }, + { + "epoch": 1.6820866141732282, + "grad_norm": 15.57400894165039, + "learning_rate": 7.3856316259065e-07, + "loss": 0.7892, + "step": 1709 + }, + { + "epoch": 1.6830708661417324, + "grad_norm": 11.692428588867188, + "learning_rate": 7.378883714212268e-07, + "loss": 0.4353, + "step": 1710 + }, + { + "epoch": 1.6840551181102361, + "grad_norm": 10.260077476501465, + "learning_rate": 7.372130211279977e-07, + "loss": 0.5166, + "step": 1711 + }, + { + "epoch": 1.6850393700787403, + "grad_norm": 17.070507049560547, + "learning_rate": 7.365371133094799e-07, + "loss": 0.946, + "step": 1712 + }, + { + "epoch": 1.686023622047244, + "grad_norm": 8.439482688903809, + "learning_rate": 7.3586064956551e-07, + "loss": 0.3011, + "step": 1713 + }, + { + "epoch": 1.6870078740157481, + "grad_norm": 16.81289291381836, + "learning_rate": 7.351836314972408e-07, + "loss": 0.7152, + "step": 1714 + }, + { + "epoch": 1.6879921259842519, + "grad_norm": 7.518153190612793, + "learning_rate": 7.345060607071365e-07, + "loss": 0.1964, + "step": 1715 + }, + { + "epoch": 1.688976377952756, + "grad_norm": 11.653175354003906, + "learning_rate": 7.338279387989703e-07, + "loss": 0.4533, + "step": 1716 + }, + { + "epoch": 1.6899606299212597, + "grad_norm": 7.579158782958984, + "learning_rate": 7.331492673778193e-07, + "loss": 0.2566, + "step": 1717 + }, + { + "epoch": 1.6909448818897639, + "grad_norm": 15.213968276977539, + "learning_rate": 7.324700480500615e-07, + "loss": 0.5991, + "step": 1718 + }, + { + "epoch": 1.6919291338582676, + "grad_norm": 14.772518157958984, + "learning_rate": 7.317902824233719e-07, + "loss": 0.8419, + "step": 1719 + }, + { + "epoch": 1.6929133858267718, + "grad_norm": 9.246028900146484, + "learning_rate": 7.311099721067183e-07, + "loss": 0.3322, + "step": 1720 + }, + { + "epoch": 1.6938976377952755, + "grad_norm": 14.696106910705566, + "learning_rate": 7.304291187103578e-07, + "loss": 0.9387, + "step": 1721 + }, + { + "epoch": 1.6948818897637796, + "grad_norm": 13.2029447555542, + "learning_rate": 7.297477238458333e-07, + "loss": 0.6228, + "step": 1722 + }, + { + "epoch": 1.6958661417322833, + "grad_norm": 11.302486419677734, + "learning_rate": 7.290657891259688e-07, + "loss": 0.4324, + "step": 1723 + }, + { + "epoch": 1.6968503937007875, + "grad_norm": 6.261458396911621, + "learning_rate": 7.283833161648664e-07, + "loss": 0.1489, + "step": 1724 + }, + { + "epoch": 1.6978346456692912, + "grad_norm": 13.963261604309082, + "learning_rate": 7.277003065779022e-07, + "loss": 0.6311, + "step": 1725 + }, + { + "epoch": 1.6988188976377954, + "grad_norm": 17.969913482666016, + "learning_rate": 7.270167619817222e-07, + "loss": 0.8488, + "step": 1726 + }, + { + "epoch": 1.699803149606299, + "grad_norm": 10.89698314666748, + "learning_rate": 7.26332683994239e-07, + "loss": 0.2451, + "step": 1727 + }, + { + "epoch": 1.7007874015748032, + "grad_norm": 16.147218704223633, + "learning_rate": 7.256480742346278e-07, + "loss": 0.6744, + "step": 1728 + }, + { + "epoch": 1.701771653543307, + "grad_norm": 12.214776992797852, + "learning_rate": 7.249629343233221e-07, + "loss": 0.3383, + "step": 1729 + }, + { + "epoch": 1.702755905511811, + "grad_norm": 10.016944885253906, + "learning_rate": 7.242772658820105e-07, + "loss": 0.3949, + "step": 1730 + }, + { + "epoch": 1.7037401574803148, + "grad_norm": 9.061641693115234, + "learning_rate": 7.235910705336327e-07, + "loss": 0.4743, + "step": 1731 + }, + { + "epoch": 1.704724409448819, + "grad_norm": 8.589599609375, + "learning_rate": 7.229043499023753e-07, + "loss": 0.3058, + "step": 1732 + }, + { + "epoch": 1.7057086614173227, + "grad_norm": 13.188594818115234, + "learning_rate": 7.222171056136683e-07, + "loss": 0.5883, + "step": 1733 + }, + { + "epoch": 1.7066929133858268, + "grad_norm": 17.603893280029297, + "learning_rate": 7.215293392941814e-07, + "loss": 0.8702, + "step": 1734 + }, + { + "epoch": 1.7076771653543306, + "grad_norm": 11.972126007080078, + "learning_rate": 7.208410525718195e-07, + "loss": 0.3621, + "step": 1735 + }, + { + "epoch": 1.7086614173228347, + "grad_norm": 15.21142864227295, + "learning_rate": 7.201522470757196e-07, + "loss": 0.7945, + "step": 1736 + }, + { + "epoch": 1.7096456692913384, + "grad_norm": 6.827094078063965, + "learning_rate": 7.194629244362462e-07, + "loss": 0.1308, + "step": 1737 + }, + { + "epoch": 1.7106299212598426, + "grad_norm": 9.049776077270508, + "learning_rate": 7.187730862849886e-07, + "loss": 0.2788, + "step": 1738 + }, + { + "epoch": 1.7116141732283463, + "grad_norm": 17.770721435546875, + "learning_rate": 7.180827342547556e-07, + "loss": 0.7254, + "step": 1739 + }, + { + "epoch": 1.7125984251968505, + "grad_norm": 7.357541084289551, + "learning_rate": 7.173918699795726e-07, + "loss": 0.2491, + "step": 1740 + }, + { + "epoch": 1.7135826771653542, + "grad_norm": 12.618313789367676, + "learning_rate": 7.167004950946774e-07, + "loss": 0.4208, + "step": 1741 + }, + { + "epoch": 1.7145669291338583, + "grad_norm": 11.27997875213623, + "learning_rate": 7.160086112365163e-07, + "loss": 0.4208, + "step": 1742 + }, + { + "epoch": 1.715551181102362, + "grad_norm": 13.064844131469727, + "learning_rate": 7.153162200427404e-07, + "loss": 0.4741, + "step": 1743 + }, + { + "epoch": 1.7165354330708662, + "grad_norm": 11.412449836730957, + "learning_rate": 7.146233231522018e-07, + "loss": 0.4124, + "step": 1744 + }, + { + "epoch": 1.71751968503937, + "grad_norm": 14.893802642822266, + "learning_rate": 7.139299222049493e-07, + "loss": 0.5402, + "step": 1745 + }, + { + "epoch": 1.718503937007874, + "grad_norm": 12.713740348815918, + "learning_rate": 7.132360188422249e-07, + "loss": 0.4638, + "step": 1746 + }, + { + "epoch": 1.719488188976378, + "grad_norm": 9.179837226867676, + "learning_rate": 7.125416147064597e-07, + "loss": 0.2356, + "step": 1747 + }, + { + "epoch": 1.720472440944882, + "grad_norm": 10.115601539611816, + "learning_rate": 7.118467114412702e-07, + "loss": 0.4948, + "step": 1748 + }, + { + "epoch": 1.7214566929133859, + "grad_norm": 9.06737232208252, + "learning_rate": 7.111513106914544e-07, + "loss": 0.3384, + "step": 1749 + }, + { + "epoch": 1.7224409448818898, + "grad_norm": 8.449541091918945, + "learning_rate": 7.104554141029874e-07, + "loss": 0.2449, + "step": 1750 + }, + { + "epoch": 1.7234251968503937, + "grad_norm": 12.327445030212402, + "learning_rate": 7.097590233230186e-07, + "loss": 0.6048, + "step": 1751 + }, + { + "epoch": 1.7244094488188977, + "grad_norm": 8.402880668640137, + "learning_rate": 7.090621399998665e-07, + "loss": 0.3199, + "step": 1752 + }, + { + "epoch": 1.7253937007874016, + "grad_norm": 9.642117500305176, + "learning_rate": 7.083647657830155e-07, + "loss": 0.4052, + "step": 1753 + }, + { + "epoch": 1.7263779527559056, + "grad_norm": 10.416515350341797, + "learning_rate": 7.076669023231126e-07, + "loss": 0.2473, + "step": 1754 + }, + { + "epoch": 1.7273622047244095, + "grad_norm": 16.807960510253906, + "learning_rate": 7.069685512719619e-07, + "loss": 0.6658, + "step": 1755 + }, + { + "epoch": 1.7283464566929134, + "grad_norm": 11.707610130310059, + "learning_rate": 7.06269714282522e-07, + "loss": 0.4531, + "step": 1756 + }, + { + "epoch": 1.7293307086614174, + "grad_norm": 12.599430084228516, + "learning_rate": 7.055703930089018e-07, + "loss": 0.6038, + "step": 1757 + }, + { + "epoch": 1.7303149606299213, + "grad_norm": 17.63587188720703, + "learning_rate": 7.048705891063562e-07, + "loss": 0.6887, + "step": 1758 + }, + { + "epoch": 1.7312992125984252, + "grad_norm": 12.01590347290039, + "learning_rate": 7.04170304231283e-07, + "loss": 0.4222, + "step": 1759 + }, + { + "epoch": 1.7322834645669292, + "grad_norm": 4.163928508758545, + "learning_rate": 7.034695400412177e-07, + "loss": 0.068, + "step": 1760 + }, + { + "epoch": 1.733267716535433, + "grad_norm": 11.864555358886719, + "learning_rate": 7.027682981948306e-07, + "loss": 0.4014, + "step": 1761 + }, + { + "epoch": 1.734251968503937, + "grad_norm": 9.201814651489258, + "learning_rate": 7.020665803519229e-07, + "loss": 0.2613, + "step": 1762 + }, + { + "epoch": 1.735236220472441, + "grad_norm": 14.464801788330078, + "learning_rate": 7.013643881734222e-07, + "loss": 0.6287, + "step": 1763 + }, + { + "epoch": 1.736220472440945, + "grad_norm": 12.518234252929688, + "learning_rate": 7.006617233213787e-07, + "loss": 0.462, + "step": 1764 + }, + { + "epoch": 1.7372047244094488, + "grad_norm": 12.262781143188477, + "learning_rate": 6.999585874589616e-07, + "loss": 0.3476, + "step": 1765 + }, + { + "epoch": 1.7381889763779528, + "grad_norm": 9.538678169250488, + "learning_rate": 6.992549822504548e-07, + "loss": 0.2576, + "step": 1766 + }, + { + "epoch": 1.7391732283464567, + "grad_norm": 15.303235054016113, + "learning_rate": 6.985509093612534e-07, + "loss": 0.8844, + "step": 1767 + }, + { + "epoch": 1.7401574803149606, + "grad_norm": 15.228886604309082, + "learning_rate": 6.978463704578589e-07, + "loss": 0.6081, + "step": 1768 + }, + { + "epoch": 1.7411417322834646, + "grad_norm": 7.154201030731201, + "learning_rate": 6.971413672078766e-07, + "loss": 0.1292, + "step": 1769 + }, + { + "epoch": 1.7421259842519685, + "grad_norm": 18.896364212036133, + "learning_rate": 6.9643590128001e-07, + "loss": 0.6448, + "step": 1770 + }, + { + "epoch": 1.7431102362204725, + "grad_norm": 10.694674491882324, + "learning_rate": 6.957299743440589e-07, + "loss": 0.2812, + "step": 1771 + }, + { + "epoch": 1.7440944881889764, + "grad_norm": 9.676472663879395, + "learning_rate": 6.950235880709131e-07, + "loss": 0.2312, + "step": 1772 + }, + { + "epoch": 1.7450787401574803, + "grad_norm": 19.950380325317383, + "learning_rate": 6.943167441325504e-07, + "loss": 1.1008, + "step": 1773 + }, + { + "epoch": 1.7460629921259843, + "grad_norm": 15.241005897521973, + "learning_rate": 6.936094442020314e-07, + "loss": 0.7085, + "step": 1774 + }, + { + "epoch": 1.7470472440944882, + "grad_norm": 10.210464477539062, + "learning_rate": 6.929016899534963e-07, + "loss": 0.4462, + "step": 1775 + }, + { + "epoch": 1.7480314960629921, + "grad_norm": 12.710392951965332, + "learning_rate": 6.921934830621603e-07, + "loss": 0.528, + "step": 1776 + }, + { + "epoch": 1.749015748031496, + "grad_norm": 15.613578796386719, + "learning_rate": 6.914848252043106e-07, + "loss": 0.7114, + "step": 1777 + }, + { + "epoch": 1.75, + "grad_norm": 9.568310737609863, + "learning_rate": 6.907757180573013e-07, + "loss": 0.2463, + "step": 1778 + }, + { + "epoch": 1.750984251968504, + "grad_norm": 14.719171524047852, + "learning_rate": 6.9006616329955e-07, + "loss": 0.4123, + "step": 1779 + }, + { + "epoch": 1.7519685039370079, + "grad_norm": 14.342561721801758, + "learning_rate": 6.893561626105338e-07, + "loss": 0.383, + "step": 1780 + }, + { + "epoch": 1.7529527559055118, + "grad_norm": 19.22394561767578, + "learning_rate": 6.886457176707855e-07, + "loss": 1.4599, + "step": 1781 + }, + { + "epoch": 1.7539370078740157, + "grad_norm": 9.642818450927734, + "learning_rate": 6.879348301618894e-07, + "loss": 0.2778, + "step": 1782 + }, + { + "epoch": 1.7549212598425197, + "grad_norm": 10.987318992614746, + "learning_rate": 6.872235017664769e-07, + "loss": 0.2371, + "step": 1783 + }, + { + "epoch": 1.7559055118110236, + "grad_norm": 8.0023193359375, + "learning_rate": 6.865117341682234e-07, + "loss": 0.2167, + "step": 1784 + }, + { + "epoch": 1.7568897637795275, + "grad_norm": 15.754701614379883, + "learning_rate": 6.857995290518439e-07, + "loss": 0.3909, + "step": 1785 + }, + { + "epoch": 1.7578740157480315, + "grad_norm": 14.946754455566406, + "learning_rate": 6.850868881030883e-07, + "loss": 0.5294, + "step": 1786 + }, + { + "epoch": 1.7588582677165354, + "grad_norm": 15.894475936889648, + "learning_rate": 6.843738130087393e-07, + "loss": 0.5687, + "step": 1787 + }, + { + "epoch": 1.7598425196850394, + "grad_norm": 6.537062168121338, + "learning_rate": 6.836603054566059e-07, + "loss": 0.1999, + "step": 1788 + }, + { + "epoch": 1.7608267716535433, + "grad_norm": 15.89072322845459, + "learning_rate": 6.829463671355216e-07, + "loss": 0.5501, + "step": 1789 + }, + { + "epoch": 1.7618110236220472, + "grad_norm": 15.467384338378906, + "learning_rate": 6.822319997353391e-07, + "loss": 0.71, + "step": 1790 + }, + { + "epoch": 1.7627952755905512, + "grad_norm": 13.845283508300781, + "learning_rate": 6.815172049469268e-07, + "loss": 0.3877, + "step": 1791 + }, + { + "epoch": 1.763779527559055, + "grad_norm": 10.2904634475708, + "learning_rate": 6.808019844621649e-07, + "loss": 0.4649, + "step": 1792 + }, + { + "epoch": 1.764763779527559, + "grad_norm": 17.6174373626709, + "learning_rate": 6.800863399739409e-07, + "loss": 0.719, + "step": 1793 + }, + { + "epoch": 1.765748031496063, + "grad_norm": 10.321569442749023, + "learning_rate": 6.793702731761459e-07, + "loss": 0.2317, + "step": 1794 + }, + { + "epoch": 1.766732283464567, + "grad_norm": 17.85763168334961, + "learning_rate": 6.786537857636707e-07, + "loss": 0.7701, + "step": 1795 + }, + { + "epoch": 1.7677165354330708, + "grad_norm": 7.668783187866211, + "learning_rate": 6.77936879432402e-07, + "loss": 0.1972, + "step": 1796 + }, + { + "epoch": 1.7687007874015748, + "grad_norm": 4.630298614501953, + "learning_rate": 6.772195558792175e-07, + "loss": 0.0844, + "step": 1797 + }, + { + "epoch": 1.7696850393700787, + "grad_norm": 9.893360137939453, + "learning_rate": 6.765018168019825e-07, + "loss": 0.2223, + "step": 1798 + }, + { + "epoch": 1.7706692913385826, + "grad_norm": 15.764335632324219, + "learning_rate": 6.757836638995461e-07, + "loss": 0.6739, + "step": 1799 + }, + { + "epoch": 1.7716535433070866, + "grad_norm": 13.348563194274902, + "learning_rate": 6.75065098871737e-07, + "loss": 0.4762, + "step": 1800 + }, + { + "epoch": 1.7726377952755905, + "grad_norm": 10.99781322479248, + "learning_rate": 6.74346123419359e-07, + "loss": 0.4405, + "step": 1801 + }, + { + "epoch": 1.7736220472440944, + "grad_norm": 22.611328125, + "learning_rate": 6.736267392441874e-07, + "loss": 0.7204, + "step": 1802 + }, + { + "epoch": 1.7746062992125984, + "grad_norm": 14.618666648864746, + "learning_rate": 6.729069480489652e-07, + "loss": 0.5104, + "step": 1803 + }, + { + "epoch": 1.7755905511811023, + "grad_norm": 10.856362342834473, + "learning_rate": 6.721867515373987e-07, + "loss": 0.2602, + "step": 1804 + }, + { + "epoch": 1.7765748031496063, + "grad_norm": 12.097258567810059, + "learning_rate": 6.714661514141535e-07, + "loss": 0.3529, + "step": 1805 + }, + { + "epoch": 1.7775590551181102, + "grad_norm": 12.390385627746582, + "learning_rate": 6.707451493848502e-07, + "loss": 0.3906, + "step": 1806 + }, + { + "epoch": 1.7785433070866141, + "grad_norm": 11.214163780212402, + "learning_rate": 6.700237471560614e-07, + "loss": 0.4859, + "step": 1807 + }, + { + "epoch": 1.779527559055118, + "grad_norm": 15.350358009338379, + "learning_rate": 6.693019464353065e-07, + "loss": 0.5454, + "step": 1808 + }, + { + "epoch": 1.780511811023622, + "grad_norm": 11.093783378601074, + "learning_rate": 6.685797489310481e-07, + "loss": 0.4257, + "step": 1809 + }, + { + "epoch": 1.781496062992126, + "grad_norm": 12.085372924804688, + "learning_rate": 6.67857156352688e-07, + "loss": 0.4074, + "step": 1810 + }, + { + "epoch": 1.78248031496063, + "grad_norm": 16.36724281311035, + "learning_rate": 6.671341704105632e-07, + "loss": 0.6397, + "step": 1811 + }, + { + "epoch": 1.7834645669291338, + "grad_norm": 9.571917533874512, + "learning_rate": 6.664107928159419e-07, + "loss": 0.5058, + "step": 1812 + }, + { + "epoch": 1.784448818897638, + "grad_norm": 12.441268920898438, + "learning_rate": 6.656870252810187e-07, + "loss": 0.6833, + "step": 1813 + }, + { + "epoch": 1.7854330708661417, + "grad_norm": 13.076436996459961, + "learning_rate": 6.649628695189123e-07, + "loss": 0.6613, + "step": 1814 + }, + { + "epoch": 1.7864173228346458, + "grad_norm": 12.561076164245605, + "learning_rate": 6.642383272436589e-07, + "loss": 0.6993, + "step": 1815 + }, + { + "epoch": 1.7874015748031495, + "grad_norm": 10.327812194824219, + "learning_rate": 6.635134001702107e-07, + "loss": 0.4406, + "step": 1816 + }, + { + "epoch": 1.7883858267716537, + "grad_norm": 6.885199546813965, + "learning_rate": 6.627880900144303e-07, + "loss": 0.2914, + "step": 1817 + }, + { + "epoch": 1.7893700787401574, + "grad_norm": 10.133977890014648, + "learning_rate": 6.620623984930869e-07, + "loss": 0.4076, + "step": 1818 + }, + { + "epoch": 1.7903543307086616, + "grad_norm": 12.960695266723633, + "learning_rate": 6.613363273238527e-07, + "loss": 0.8532, + "step": 1819 + }, + { + "epoch": 1.7913385826771653, + "grad_norm": 12.55600643157959, + "learning_rate": 6.606098782252978e-07, + "loss": 0.5586, + "step": 1820 + }, + { + "epoch": 1.7923228346456694, + "grad_norm": 12.925490379333496, + "learning_rate": 6.598830529168877e-07, + "loss": 0.629, + "step": 1821 + }, + { + "epoch": 1.7933070866141732, + "grad_norm": 11.820432662963867, + "learning_rate": 6.591558531189781e-07, + "loss": 0.5267, + "step": 1822 + }, + { + "epoch": 1.7942913385826773, + "grad_norm": 11.783539772033691, + "learning_rate": 6.584282805528106e-07, + "loss": 0.4557, + "step": 1823 + }, + { + "epoch": 1.795275590551181, + "grad_norm": 6.3824052810668945, + "learning_rate": 6.577003369405099e-07, + "loss": 0.2006, + "step": 1824 + }, + { + "epoch": 1.7962598425196852, + "grad_norm": 9.544827461242676, + "learning_rate": 6.569720240050784e-07, + "loss": 0.3406, + "step": 1825 + }, + { + "epoch": 1.797244094488189, + "grad_norm": 6.162224769592285, + "learning_rate": 6.562433434703928e-07, + "loss": 0.1683, + "step": 1826 + }, + { + "epoch": 1.798228346456693, + "grad_norm": 7.8058037757873535, + "learning_rate": 6.555142970612001e-07, + "loss": 0.2141, + "step": 1827 + }, + { + "epoch": 1.7992125984251968, + "grad_norm": 5.146012306213379, + "learning_rate": 6.54784886503113e-07, + "loss": 0.0945, + "step": 1828 + }, + { + "epoch": 1.800196850393701, + "grad_norm": 16.08049964904785, + "learning_rate": 6.540551135226061e-07, + "loss": 0.4701, + "step": 1829 + }, + { + "epoch": 1.8011811023622046, + "grad_norm": 12.357492446899414, + "learning_rate": 6.533249798470123e-07, + "loss": 0.4771, + "step": 1830 + } + ], + "logging_steps": 1, + "max_steps": 3048, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 305, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}