random_oXC0MqfHQb2vnfrT / trainer_state.json
cutelemonlili's picture
Add files using upload-large-folder tool
c00442f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 412,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0048543689320388345,
"grad_norm": 13.42069128193306,
"learning_rate": 9.999854640567861e-06,
"loss": 0.7668,
"step": 1
},
{
"epoch": 0.009708737864077669,
"grad_norm": 6.53274659931177,
"learning_rate": 9.999418570723189e-06,
"loss": 0.6273,
"step": 2
},
{
"epoch": 0.014563106796116505,
"grad_norm": 8.380401805740549,
"learning_rate": 9.998691815820732e-06,
"loss": 0.801,
"step": 3
},
{
"epoch": 0.019417475728155338,
"grad_norm": 6.604033590737317,
"learning_rate": 9.997674418116759e-06,
"loss": 0.7125,
"step": 4
},
{
"epoch": 0.024271844660194174,
"grad_norm": 5.146672711385003,
"learning_rate": 9.996366436766612e-06,
"loss": 0.5931,
"step": 5
},
{
"epoch": 0.02912621359223301,
"grad_norm": 4.87244867404994,
"learning_rate": 9.994767947821261e-06,
"loss": 0.6197,
"step": 6
},
{
"epoch": 0.03398058252427184,
"grad_norm": 4.930406125100111,
"learning_rate": 9.992879044222887e-06,
"loss": 0.6025,
"step": 7
},
{
"epoch": 0.038834951456310676,
"grad_norm": 4.180850446735125,
"learning_rate": 9.99069983579947e-06,
"loss": 0.5438,
"step": 8
},
{
"epoch": 0.043689320388349516,
"grad_norm": 5.0503615853624755,
"learning_rate": 9.988230449258409e-06,
"loss": 0.6569,
"step": 9
},
{
"epoch": 0.04854368932038835,
"grad_norm": 3.4186297543712767,
"learning_rate": 9.985471028179155e-06,
"loss": 0.4081,
"step": 10
},
{
"epoch": 0.05339805825242718,
"grad_norm": 5.110382531237502,
"learning_rate": 9.982421733004857e-06,
"loss": 0.628,
"step": 11
},
{
"epoch": 0.05825242718446602,
"grad_norm": 4.853720420662286,
"learning_rate": 9.979082741033047e-06,
"loss": 0.6117,
"step": 12
},
{
"epoch": 0.06310679611650485,
"grad_norm": 4.948980666401182,
"learning_rate": 9.975454246405312e-06,
"loss": 0.6344,
"step": 13
},
{
"epoch": 0.06796116504854369,
"grad_norm": 5.636871326973096,
"learning_rate": 9.971536460096021e-06,
"loss": 0.7293,
"step": 14
},
{
"epoch": 0.07281553398058252,
"grad_norm": 4.360243957668438,
"learning_rate": 9.96732960990005e-06,
"loss": 0.5801,
"step": 15
},
{
"epoch": 0.07766990291262135,
"grad_norm": 4.577797258437923,
"learning_rate": 9.96283394041954e-06,
"loss": 0.6937,
"step": 16
},
{
"epoch": 0.0825242718446602,
"grad_norm": 5.891030536882447,
"learning_rate": 9.95804971304968e-06,
"loss": 0.6804,
"step": 17
},
{
"epoch": 0.08737864077669903,
"grad_norm": 4.120736342853116,
"learning_rate": 9.952977205963496e-06,
"loss": 0.5768,
"step": 18
},
{
"epoch": 0.09223300970873786,
"grad_norm": 4.075403602987412,
"learning_rate": 9.94761671409569e-06,
"loss": 0.6568,
"step": 19
},
{
"epoch": 0.0970873786407767,
"grad_norm": 4.474362492464127,
"learning_rate": 9.941968549125481e-06,
"loss": 0.6182,
"step": 20
},
{
"epoch": 0.10194174757281553,
"grad_norm": 4.267133262124595,
"learning_rate": 9.936033039458494e-06,
"loss": 0.6353,
"step": 21
},
{
"epoch": 0.10679611650485436,
"grad_norm": 3.6429603871494067,
"learning_rate": 9.929810530207651e-06,
"loss": 0.4694,
"step": 22
},
{
"epoch": 0.11165048543689321,
"grad_norm": 3.3599700740906253,
"learning_rate": 9.923301383173119e-06,
"loss": 0.5403,
"step": 23
},
{
"epoch": 0.11650485436893204,
"grad_norm": 4.150849592127758,
"learning_rate": 9.916505976821262e-06,
"loss": 0.6295,
"step": 24
},
{
"epoch": 0.12135922330097088,
"grad_norm": 3.967697671113486,
"learning_rate": 9.909424706262647e-06,
"loss": 0.5767,
"step": 25
},
{
"epoch": 0.1262135922330097,
"grad_norm": 3.915338910316096,
"learning_rate": 9.902057983229059e-06,
"loss": 0.5087,
"step": 26
},
{
"epoch": 0.13106796116504854,
"grad_norm": 3.882530112432383,
"learning_rate": 9.894406236049569e-06,
"loss": 0.47,
"step": 27
},
{
"epoch": 0.13592233009708737,
"grad_norm": 4.025075453651069,
"learning_rate": 9.886469909625624e-06,
"loss": 0.5271,
"step": 28
},
{
"epoch": 0.1407766990291262,
"grad_norm": 4.873882283974574,
"learning_rate": 9.87824946540519e-06,
"loss": 0.6163,
"step": 29
},
{
"epoch": 0.14563106796116504,
"grad_norm": 4.06872395535626,
"learning_rate": 9.869745381355906e-06,
"loss": 0.5674,
"step": 30
},
{
"epoch": 0.15048543689320387,
"grad_norm": 4.260632171969729,
"learning_rate": 9.860958151937303e-06,
"loss": 0.59,
"step": 31
},
{
"epoch": 0.1553398058252427,
"grad_norm": 3.927021858743994,
"learning_rate": 9.851888288072053e-06,
"loss": 0.536,
"step": 32
},
{
"epoch": 0.16019417475728157,
"grad_norm": 4.853465734568144,
"learning_rate": 9.842536317116262e-06,
"loss": 0.6542,
"step": 33
},
{
"epoch": 0.1650485436893204,
"grad_norm": 4.438919899646417,
"learning_rate": 9.832902782828801e-06,
"loss": 0.6539,
"step": 34
},
{
"epoch": 0.16990291262135923,
"grad_norm": 4.56990257837148,
"learning_rate": 9.822988245339701e-06,
"loss": 0.6313,
"step": 35
},
{
"epoch": 0.17475728155339806,
"grad_norm": 5.31270622098839,
"learning_rate": 9.81279328111758e-06,
"loss": 0.6952,
"step": 36
},
{
"epoch": 0.1796116504854369,
"grad_norm": 4.060284941507083,
"learning_rate": 9.802318482936121e-06,
"loss": 0.5672,
"step": 37
},
{
"epoch": 0.18446601941747573,
"grad_norm": 4.120387062256013,
"learning_rate": 9.791564459839609e-06,
"loss": 0.5523,
"step": 38
},
{
"epoch": 0.18932038834951456,
"grad_norm": 4.5570100336739765,
"learning_rate": 9.780531837107519e-06,
"loss": 0.6585,
"step": 39
},
{
"epoch": 0.1941747572815534,
"grad_norm": 5.076174414072437,
"learning_rate": 9.769221256218165e-06,
"loss": 0.6861,
"step": 40
},
{
"epoch": 0.19902912621359223,
"grad_norm": 4.604754361920019,
"learning_rate": 9.75763337481139e-06,
"loss": 0.6471,
"step": 41
},
{
"epoch": 0.20388349514563106,
"grad_norm": 4.605000469357057,
"learning_rate": 9.745768866650339e-06,
"loss": 0.6841,
"step": 42
},
{
"epoch": 0.2087378640776699,
"grad_norm": 3.313092749775837,
"learning_rate": 9.73362842158228e-06,
"loss": 0.5443,
"step": 43
},
{
"epoch": 0.21359223300970873,
"grad_norm": 3.78982744975658,
"learning_rate": 9.721212745498493e-06,
"loss": 0.5832,
"step": 44
},
{
"epoch": 0.21844660194174756,
"grad_norm": 4.103293783804162,
"learning_rate": 9.70852256029323e-06,
"loss": 0.5643,
"step": 45
},
{
"epoch": 0.22330097087378642,
"grad_norm": 4.233210716366976,
"learning_rate": 9.695558603821735e-06,
"loss": 0.5775,
"step": 46
},
{
"epoch": 0.22815533980582525,
"grad_norm": 3.8390019031497755,
"learning_rate": 9.682321629857348e-06,
"loss": 0.5539,
"step": 47
},
{
"epoch": 0.23300970873786409,
"grad_norm": 4.222790396911806,
"learning_rate": 9.66881240804768e-06,
"loss": 0.5506,
"step": 48
},
{
"epoch": 0.23786407766990292,
"grad_norm": 4.225277268416746,
"learning_rate": 9.655031723869848e-06,
"loss": 0.5728,
"step": 49
},
{
"epoch": 0.24271844660194175,
"grad_norm": 5.521035926327871,
"learning_rate": 9.64098037858483e-06,
"loss": 0.6491,
"step": 50
},
{
"epoch": 0.24757281553398058,
"grad_norm": 3.9750487775451346,
"learning_rate": 9.626659189190852e-06,
"loss": 0.5119,
"step": 51
},
{
"epoch": 0.2524271844660194,
"grad_norm": 3.6146721282279572,
"learning_rate": 9.612068988375898e-06,
"loss": 0.5388,
"step": 52
},
{
"epoch": 0.25728155339805825,
"grad_norm": 4.7300244435664585,
"learning_rate": 9.597210624469288e-06,
"loss": 0.5571,
"step": 53
},
{
"epoch": 0.2621359223300971,
"grad_norm": 3.914750222619553,
"learning_rate": 9.582084961392358e-06,
"loss": 0.5721,
"step": 54
},
{
"epoch": 0.2669902912621359,
"grad_norm": 3.8597782256667323,
"learning_rate": 9.566692878608229e-06,
"loss": 0.5476,
"step": 55
},
{
"epoch": 0.27184466019417475,
"grad_norm": 4.1890058209844865,
"learning_rate": 9.551035271070665e-06,
"loss": 0.5918,
"step": 56
},
{
"epoch": 0.2766990291262136,
"grad_norm": 4.607339029787459,
"learning_rate": 9.53511304917204e-06,
"loss": 0.5651,
"step": 57
},
{
"epoch": 0.2815533980582524,
"grad_norm": 3.351426145638587,
"learning_rate": 9.51892713869041e-06,
"loss": 0.482,
"step": 58
},
{
"epoch": 0.28640776699029125,
"grad_norm": 4.540672723924717,
"learning_rate": 9.502478480735678e-06,
"loss": 0.6103,
"step": 59
},
{
"epoch": 0.2912621359223301,
"grad_norm": 4.299900945913631,
"learning_rate": 9.485768031694872e-06,
"loss": 0.5882,
"step": 60
},
{
"epoch": 0.2961165048543689,
"grad_norm": 4.480436960387508,
"learning_rate": 9.468796763176549e-06,
"loss": 0.6752,
"step": 61
},
{
"epoch": 0.30097087378640774,
"grad_norm": 4.137496460348275,
"learning_rate": 9.45156566195429e-06,
"loss": 0.5841,
"step": 62
},
{
"epoch": 0.3058252427184466,
"grad_norm": 4.516278897177943,
"learning_rate": 9.43407572990933e-06,
"loss": 0.544,
"step": 63
},
{
"epoch": 0.3106796116504854,
"grad_norm": 4.002681652824515,
"learning_rate": 9.416327983972304e-06,
"loss": 0.5967,
"step": 64
},
{
"epoch": 0.3155339805825243,
"grad_norm": 4.2126750589949005,
"learning_rate": 9.398323456064124e-06,
"loss": 0.5991,
"step": 65
},
{
"epoch": 0.32038834951456313,
"grad_norm": 4.487778420822117,
"learning_rate": 9.380063193035968e-06,
"loss": 0.6408,
"step": 66
},
{
"epoch": 0.32524271844660196,
"grad_norm": 4.314965739730904,
"learning_rate": 9.361548256608421e-06,
"loss": 0.5741,
"step": 67
},
{
"epoch": 0.3300970873786408,
"grad_norm": 3.690718412716262,
"learning_rate": 9.342779723309746e-06,
"loss": 0.5432,
"step": 68
},
{
"epoch": 0.33495145631067963,
"grad_norm": 4.236384345136889,
"learning_rate": 9.323758684413272e-06,
"loss": 0.5336,
"step": 69
},
{
"epoch": 0.33980582524271846,
"grad_norm": 4.150724448051698,
"learning_rate": 9.304486245873973e-06,
"loss": 0.6689,
"step": 70
},
{
"epoch": 0.3446601941747573,
"grad_norm": 3.9771657271854823,
"learning_rate": 9.284963528264133e-06,
"loss": 0.5521,
"step": 71
},
{
"epoch": 0.34951456310679613,
"grad_norm": 4.261812079477349,
"learning_rate": 9.26519166670821e-06,
"loss": 0.5784,
"step": 72
},
{
"epoch": 0.35436893203883496,
"grad_norm": 4.247964250572644,
"learning_rate": 9.24517181081683e-06,
"loss": 0.5794,
"step": 73
},
{
"epoch": 0.3592233009708738,
"grad_norm": 4.319209622266621,
"learning_rate": 9.22490512461995e-06,
"loss": 0.6215,
"step": 74
},
{
"epoch": 0.3640776699029126,
"grad_norm": 4.674980763505609,
"learning_rate": 9.204392786499168e-06,
"loss": 0.5635,
"step": 75
},
{
"epoch": 0.36893203883495146,
"grad_norm": 3.7640016156897644,
"learning_rate": 9.183635989119211e-06,
"loss": 0.5085,
"step": 76
},
{
"epoch": 0.3737864077669903,
"grad_norm": 4.25878408255199,
"learning_rate": 9.162635939358593e-06,
"loss": 0.5295,
"step": 77
},
{
"epoch": 0.3786407766990291,
"grad_norm": 4.355454623965196,
"learning_rate": 9.141393858239435e-06,
"loss": 0.5415,
"step": 78
},
{
"epoch": 0.38349514563106796,
"grad_norm": 4.035239583214188,
"learning_rate": 9.119910980856477e-06,
"loss": 0.571,
"step": 79
},
{
"epoch": 0.3883495145631068,
"grad_norm": 3.802426065366751,
"learning_rate": 9.098188556305262e-06,
"loss": 0.5465,
"step": 80
},
{
"epoch": 0.3932038834951456,
"grad_norm": 4.025264517049104,
"learning_rate": 9.076227847609513e-06,
"loss": 0.5858,
"step": 81
},
{
"epoch": 0.39805825242718446,
"grad_norm": 3.425408896418311,
"learning_rate": 9.054030131647682e-06,
"loss": 0.5257,
"step": 82
},
{
"epoch": 0.4029126213592233,
"grad_norm": 4.257974072995988,
"learning_rate": 9.031596699078727e-06,
"loss": 0.695,
"step": 83
},
{
"epoch": 0.4077669902912621,
"grad_norm": 4.8875577459201365,
"learning_rate": 9.008928854267054e-06,
"loss": 0.6674,
"step": 84
},
{
"epoch": 0.41262135922330095,
"grad_norm": 3.718045056842805,
"learning_rate": 8.986027915206686e-06,
"loss": 0.4824,
"step": 85
},
{
"epoch": 0.4174757281553398,
"grad_norm": 4.716766004872851,
"learning_rate": 8.962895213444618e-06,
"loss": 0.5735,
"step": 86
},
{
"epoch": 0.4223300970873786,
"grad_norm": 3.379567851194754,
"learning_rate": 8.939532094003409e-06,
"loss": 0.4145,
"step": 87
},
{
"epoch": 0.42718446601941745,
"grad_norm": 4.091168231013593,
"learning_rate": 8.91593991530297e-06,
"loss": 0.6006,
"step": 88
},
{
"epoch": 0.4320388349514563,
"grad_norm": 4.754523553359645,
"learning_rate": 8.892120049081577e-06,
"loss": 0.6244,
"step": 89
},
{
"epoch": 0.4368932038834951,
"grad_norm": 4.295708758518802,
"learning_rate": 8.868073880316125e-06,
"loss": 0.6034,
"step": 90
},
{
"epoch": 0.441747572815534,
"grad_norm": 4.407143475652452,
"learning_rate": 8.843802807141584e-06,
"loss": 0.5916,
"step": 91
},
{
"epoch": 0.44660194174757284,
"grad_norm": 3.973756766361989,
"learning_rate": 8.819308240769726e-06,
"loss": 0.5373,
"step": 92
},
{
"epoch": 0.45145631067961167,
"grad_norm": 3.822485896919711,
"learning_rate": 8.794591605407047e-06,
"loss": 0.65,
"step": 93
},
{
"epoch": 0.4563106796116505,
"grad_norm": 5.01686152621343,
"learning_rate": 8.769654338171986e-06,
"loss": 0.7032,
"step": 94
},
{
"epoch": 0.46116504854368934,
"grad_norm": 4.246037513099975,
"learning_rate": 8.744497889011344e-06,
"loss": 0.6067,
"step": 95
},
{
"epoch": 0.46601941747572817,
"grad_norm": 3.787512331559668,
"learning_rate": 8.71912372061598e-06,
"loss": 0.5974,
"step": 96
},
{
"epoch": 0.470873786407767,
"grad_norm": 3.7492885311977715,
"learning_rate": 8.693533308335786e-06,
"loss": 0.5424,
"step": 97
},
{
"epoch": 0.47572815533980584,
"grad_norm": 3.7803809691255834,
"learning_rate": 8.667728140093876e-06,
"loss": 0.5585,
"step": 98
},
{
"epoch": 0.48058252427184467,
"grad_norm": 5.014122066260363,
"learning_rate": 8.641709716300092e-06,
"loss": 0.6658,
"step": 99
},
{
"epoch": 0.4854368932038835,
"grad_norm": 4.054759455439137,
"learning_rate": 8.615479549763756e-06,
"loss": 0.4973,
"step": 100
},
{
"epoch": 0.49029126213592233,
"grad_norm": 3.6211948283358146,
"learning_rate": 8.589039165605716e-06,
"loss": 0.532,
"step": 101
},
{
"epoch": 0.49514563106796117,
"grad_norm": 3.672666895752865,
"learning_rate": 8.56239010116966e-06,
"loss": 0.5033,
"step": 102
},
{
"epoch": 0.5,
"grad_norm": 3.9518779216894595,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6054,
"step": 103
},
{
"epoch": 0.5048543689320388,
"grad_norm": 3.9380228019776395,
"learning_rate": 8.508472141415468e-06,
"loss": 0.5214,
"step": 104
},
{
"epoch": 0.5097087378640777,
"grad_norm": 4.020582581327899,
"learning_rate": 8.481206381090934e-06,
"loss": 0.5069,
"step": 105
},
{
"epoch": 0.5145631067961165,
"grad_norm": 4.045839417590654,
"learning_rate": 8.453738210293316e-06,
"loss": 0.6215,
"step": 106
},
{
"epoch": 0.5194174757281553,
"grad_norm": 4.30226993140095,
"learning_rate": 8.426069226125695e-06,
"loss": 0.49,
"step": 107
},
{
"epoch": 0.5242718446601942,
"grad_norm": 4.035919439908701,
"learning_rate": 8.398201037367202e-06,
"loss": 0.5294,
"step": 108
},
{
"epoch": 0.529126213592233,
"grad_norm": 3.81867859642456,
"learning_rate": 8.370135264379475e-06,
"loss": 0.6611,
"step": 109
},
{
"epoch": 0.5339805825242718,
"grad_norm": 4.107255513497325,
"learning_rate": 8.341873539012443e-06,
"loss": 0.6004,
"step": 110
},
{
"epoch": 0.5388349514563107,
"grad_norm": 4.483104869193575,
"learning_rate": 8.313417504509446e-06,
"loss": 0.4746,
"step": 111
},
{
"epoch": 0.5436893203883495,
"grad_norm": 3.771794440022536,
"learning_rate": 8.284768815411693e-06,
"loss": 0.5521,
"step": 112
},
{
"epoch": 0.5485436893203883,
"grad_norm": 5.397767910444214,
"learning_rate": 8.255929137462049e-06,
"loss": 0.6007,
"step": 113
},
{
"epoch": 0.5533980582524272,
"grad_norm": 4.244521514959644,
"learning_rate": 8.226900147508205e-06,
"loss": 0.6208,
"step": 114
},
{
"epoch": 0.558252427184466,
"grad_norm": 4.007749368236666,
"learning_rate": 8.197683533405156e-06,
"loss": 0.5277,
"step": 115
},
{
"epoch": 0.5631067961165048,
"grad_norm": 4.1435563304476375,
"learning_rate": 8.168280993917078e-06,
"loss": 0.6164,
"step": 116
},
{
"epoch": 0.5679611650485437,
"grad_norm": 4.15612210895854,
"learning_rate": 8.138694238618543e-06,
"loss": 0.5875,
"step": 117
},
{
"epoch": 0.5728155339805825,
"grad_norm": 4.2612877696175095,
"learning_rate": 8.108924987795137e-06,
"loss": 0.5852,
"step": 118
},
{
"epoch": 0.5776699029126213,
"grad_norm": 4.028796607567781,
"learning_rate": 8.078974972343414e-06,
"loss": 0.5878,
"step": 119
},
{
"epoch": 0.5825242718446602,
"grad_norm": 4.661021644423876,
"learning_rate": 8.048845933670274e-06,
"loss": 0.6077,
"step": 120
},
{
"epoch": 0.587378640776699,
"grad_norm": 3.602712179040846,
"learning_rate": 8.01853962359169e-06,
"loss": 0.4644,
"step": 121
},
{
"epoch": 0.5922330097087378,
"grad_norm": 4.113312607833202,
"learning_rate": 7.988057804230878e-06,
"loss": 0.5156,
"step": 122
},
{
"epoch": 0.5970873786407767,
"grad_norm": 4.151401377380039,
"learning_rate": 7.957402247915817e-06,
"loss": 0.5185,
"step": 123
},
{
"epoch": 0.6019417475728155,
"grad_norm": 4.05964571043831,
"learning_rate": 7.92657473707621e-06,
"loss": 0.5716,
"step": 124
},
{
"epoch": 0.6067961165048543,
"grad_norm": 3.8642392522415396,
"learning_rate": 7.895577064139847e-06,
"loss": 0.5274,
"step": 125
},
{
"epoch": 0.6116504854368932,
"grad_norm": 4.3129713686545506,
"learning_rate": 7.864411031428379e-06,
"loss": 0.5757,
"step": 126
},
{
"epoch": 0.616504854368932,
"grad_norm": 3.4743733049902112,
"learning_rate": 7.833078451052537e-06,
"loss": 0.4332,
"step": 127
},
{
"epoch": 0.6213592233009708,
"grad_norm": 4.226981090702445,
"learning_rate": 7.801581144806752e-06,
"loss": 0.5838,
"step": 128
},
{
"epoch": 0.6262135922330098,
"grad_norm": 4.160080199924846,
"learning_rate": 7.769920944063244e-06,
"loss": 0.5121,
"step": 129
},
{
"epoch": 0.6310679611650486,
"grad_norm": 3.808974242898024,
"learning_rate": 7.73809968966554e-06,
"loss": 0.4741,
"step": 130
},
{
"epoch": 0.6359223300970874,
"grad_norm": 4.486007726481835,
"learning_rate": 7.706119231821423e-06,
"loss": 0.6164,
"step": 131
},
{
"epoch": 0.6407766990291263,
"grad_norm": 3.4894103075414304,
"learning_rate": 7.673981429995372e-06,
"loss": 0.4562,
"step": 132
},
{
"epoch": 0.6456310679611651,
"grad_norm": 3.6250386690091423,
"learning_rate": 7.641688152800433e-06,
"loss": 0.4774,
"step": 133
},
{
"epoch": 0.6504854368932039,
"grad_norm": 4.143069849010099,
"learning_rate": 7.609241277889583e-06,
"loss": 0.6065,
"step": 134
},
{
"epoch": 0.6553398058252428,
"grad_norm": 5.210714099312508,
"learning_rate": 7.5766426918465455e-06,
"loss": 0.7482,
"step": 135
},
{
"epoch": 0.6601941747572816,
"grad_norm": 3.8571308280714125,
"learning_rate": 7.5438942900761035e-06,
"loss": 0.4617,
"step": 136
},
{
"epoch": 0.6650485436893204,
"grad_norm": 4.583184662697303,
"learning_rate": 7.51099797669389e-06,
"loss": 0.6032,
"step": 137
},
{
"epoch": 0.6699029126213593,
"grad_norm": 4.232180099193374,
"learning_rate": 7.477955664415678e-06,
"loss": 0.6256,
"step": 138
},
{
"epoch": 0.6747572815533981,
"grad_norm": 4.589131272826703,
"learning_rate": 7.444769274446168e-06,
"loss": 0.5908,
"step": 139
},
{
"epoch": 0.6796116504854369,
"grad_norm": 4.101642857538624,
"learning_rate": 7.411440736367281e-06,
"loss": 0.6028,
"step": 140
},
{
"epoch": 0.6844660194174758,
"grad_norm": 4.044133991141443,
"learning_rate": 7.377971988025964e-06,
"loss": 0.5572,
"step": 141
},
{
"epoch": 0.6893203883495146,
"grad_norm": 4.046147176686249,
"learning_rate": 7.3443649754215175e-06,
"loss": 0.5152,
"step": 142
},
{
"epoch": 0.6941747572815534,
"grad_norm": 4.191229657073093,
"learning_rate": 7.310621652592449e-06,
"loss": 0.5797,
"step": 143
},
{
"epoch": 0.6990291262135923,
"grad_norm": 4.364766718908912,
"learning_rate": 7.276743981502856e-06,
"loss": 0.6351,
"step": 144
},
{
"epoch": 0.7038834951456311,
"grad_norm": 3.986203506930155,
"learning_rate": 7.242733931928352e-06,
"loss": 0.525,
"step": 145
},
{
"epoch": 0.7087378640776699,
"grad_norm": 3.3279061504390612,
"learning_rate": 7.208593481341536e-06,
"loss": 0.5427,
"step": 146
},
{
"epoch": 0.7135922330097088,
"grad_norm": 3.705577210435045,
"learning_rate": 7.1743246147970095e-06,
"loss": 0.5769,
"step": 147
},
{
"epoch": 0.7184466019417476,
"grad_norm": 4.38543196191147,
"learning_rate": 7.139929324815965e-06,
"loss": 0.6507,
"step": 148
},
{
"epoch": 0.7233009708737864,
"grad_norm": 3.873388951697851,
"learning_rate": 7.105409611270332e-06,
"loss": 0.5066,
"step": 149
},
{
"epoch": 0.7281553398058253,
"grad_norm": 4.059913100738543,
"learning_rate": 7.070767481266493e-06,
"loss": 0.5658,
"step": 150
},
{
"epoch": 0.7330097087378641,
"grad_norm": 4.397781674662709,
"learning_rate": 7.036004949028587e-06,
"loss": 0.6359,
"step": 151
},
{
"epoch": 0.7378640776699029,
"grad_norm": 3.954432531191293,
"learning_rate": 7.00112403578139e-06,
"loss": 0.5753,
"step": 152
},
{
"epoch": 0.7427184466019418,
"grad_norm": 3.9397256843389603,
"learning_rate": 6.9661267696328015e-06,
"loss": 0.4909,
"step": 153
},
{
"epoch": 0.7475728155339806,
"grad_norm": 4.130473967546563,
"learning_rate": 6.931015185455915e-06,
"loss": 0.5,
"step": 154
},
{
"epoch": 0.7524271844660194,
"grad_norm": 4.412084164682045,
"learning_rate": 6.895791324770702e-06,
"loss": 0.6293,
"step": 155
},
{
"epoch": 0.7572815533980582,
"grad_norm": 4.487617569979081,
"learning_rate": 6.860457235625322e-06,
"loss": 0.5884,
"step": 156
},
{
"epoch": 0.7621359223300971,
"grad_norm": 3.798641697958784,
"learning_rate": 6.825014972477024e-06,
"loss": 0.609,
"step": 157
},
{
"epoch": 0.7669902912621359,
"grad_norm": 4.296339268501908,
"learning_rate": 6.7894665960727105e-06,
"loss": 0.5429,
"step": 158
},
{
"epoch": 0.7718446601941747,
"grad_norm": 4.1883940194959175,
"learning_rate": 6.7538141733291e-06,
"loss": 0.581,
"step": 159
},
{
"epoch": 0.7766990291262136,
"grad_norm": 4.147529387997262,
"learning_rate": 6.7180597772125665e-06,
"loss": 0.5571,
"step": 160
},
{
"epoch": 0.7815533980582524,
"grad_norm": 3.948400976638983,
"learning_rate": 6.682205486618592e-06,
"loss": 0.6422,
"step": 161
},
{
"epoch": 0.7864077669902912,
"grad_norm": 4.007431423310787,
"learning_rate": 6.646253386250909e-06,
"loss": 0.5357,
"step": 162
},
{
"epoch": 0.7912621359223301,
"grad_norm": 3.927412628183711,
"learning_rate": 6.610205566500272e-06,
"loss": 0.5898,
"step": 163
},
{
"epoch": 0.7961165048543689,
"grad_norm": 3.8261520472551545,
"learning_rate": 6.574064123322925e-06,
"loss": 0.5469,
"step": 164
},
{
"epoch": 0.8009708737864077,
"grad_norm": 3.4872036890968934,
"learning_rate": 6.537831158118733e-06,
"loss": 0.4888,
"step": 165
},
{
"epoch": 0.8058252427184466,
"grad_norm": 3.905172841501351,
"learning_rate": 6.50150877760899e-06,
"loss": 0.5187,
"step": 166
},
{
"epoch": 0.8106796116504854,
"grad_norm": 4.16703492428348,
"learning_rate": 6.465099093713944e-06,
"loss": 0.5767,
"step": 167
},
{
"epoch": 0.8155339805825242,
"grad_norm": 3.821105213060246,
"learning_rate": 6.42860422342998e-06,
"loss": 0.5761,
"step": 168
},
{
"epoch": 0.8203883495145631,
"grad_norm": 4.196199504604192,
"learning_rate": 6.392026288706549e-06,
"loss": 0.5413,
"step": 169
},
{
"epoch": 0.8252427184466019,
"grad_norm": 3.5765900163175273,
"learning_rate": 6.3553674163227786e-06,
"loss": 0.4195,
"step": 170
},
{
"epoch": 0.8300970873786407,
"grad_norm": 3.5316777829501618,
"learning_rate": 6.318629737763818e-06,
"loss": 0.4771,
"step": 171
},
{
"epoch": 0.8349514563106796,
"grad_norm": 4.760420855989642,
"learning_rate": 6.281815389096903e-06,
"loss": 0.6044,
"step": 172
},
{
"epoch": 0.8398058252427184,
"grad_norm": 4.232747714861395,
"learning_rate": 6.244926510847162e-06,
"loss": 0.6208,
"step": 173
},
{
"epoch": 0.8446601941747572,
"grad_norm": 4.49881673114576,
"learning_rate": 6.207965247873151e-06,
"loss": 0.5503,
"step": 174
},
{
"epoch": 0.8495145631067961,
"grad_norm": 4.263280050278279,
"learning_rate": 6.1709337492421515e-06,
"loss": 0.5607,
"step": 175
},
{
"epoch": 0.8543689320388349,
"grad_norm": 3.8056134974631517,
"learning_rate": 6.133834168105206e-06,
"loss": 0.5422,
"step": 176
},
{
"epoch": 0.8592233009708737,
"grad_norm": 3.7159982075739535,
"learning_rate": 6.096668661571934e-06,
"loss": 0.4761,
"step": 177
},
{
"epoch": 0.8640776699029126,
"grad_norm": 4.433523608656481,
"learning_rate": 6.0594393905851065e-06,
"loss": 0.5745,
"step": 178
},
{
"epoch": 0.8689320388349514,
"grad_norm": 4.250437810740844,
"learning_rate": 6.0221485197949995e-06,
"loss": 0.6156,
"step": 179
},
{
"epoch": 0.8737864077669902,
"grad_norm": 3.937956508041183,
"learning_rate": 5.9847982174335314e-06,
"loss": 0.5776,
"step": 180
},
{
"epoch": 0.8786407766990292,
"grad_norm": 4.137209446521738,
"learning_rate": 5.9473906551881985e-06,
"loss": 0.5569,
"step": 181
},
{
"epoch": 0.883495145631068,
"grad_norm": 3.7925563025814255,
"learning_rate": 5.9099280080758085e-06,
"loss": 0.5022,
"step": 182
},
{
"epoch": 0.8883495145631068,
"grad_norm": 3.947797576669721,
"learning_rate": 5.872412454315999e-06,
"loss": 0.5922,
"step": 183
},
{
"epoch": 0.8932038834951457,
"grad_norm": 4.142592958840424,
"learning_rate": 5.834846175204612e-06,
"loss": 0.6236,
"step": 184
},
{
"epoch": 0.8980582524271845,
"grad_norm": 4.322397369381492,
"learning_rate": 5.797231354986842e-06,
"loss": 0.5771,
"step": 185
},
{
"epoch": 0.9029126213592233,
"grad_norm": 3.097717171690413,
"learning_rate": 5.759570180730255e-06,
"loss": 0.4874,
"step": 186
},
{
"epoch": 0.9077669902912622,
"grad_norm": 4.202731429329119,
"learning_rate": 5.721864842197612e-06,
"loss": 0.5089,
"step": 187
},
{
"epoch": 0.912621359223301,
"grad_norm": 4.101905378840448,
"learning_rate": 5.684117531719552e-06,
"loss": 0.5834,
"step": 188
},
{
"epoch": 0.9174757281553398,
"grad_norm": 4.220209460616022,
"learning_rate": 5.646330444067121e-06,
"loss": 0.5278,
"step": 189
},
{
"epoch": 0.9223300970873787,
"grad_norm": 3.6407941327169646,
"learning_rate": 5.608505776324158e-06,
"loss": 0.5416,
"step": 190
},
{
"epoch": 0.9271844660194175,
"grad_norm": 4.04423159887254,
"learning_rate": 5.570645727759558e-06,
"loss": 0.5512,
"step": 191
},
{
"epoch": 0.9320388349514563,
"grad_norm": 3.784272870175967,
"learning_rate": 5.532752499699381e-06,
"loss": 0.5451,
"step": 192
},
{
"epoch": 0.9368932038834952,
"grad_norm": 3.5359594145199185,
"learning_rate": 5.494828295398874e-06,
"loss": 0.5077,
"step": 193
},
{
"epoch": 0.941747572815534,
"grad_norm": 3.3730422330235936,
"learning_rate": 5.456875319914355e-06,
"loss": 0.4393,
"step": 194
},
{
"epoch": 0.9466019417475728,
"grad_norm": 3.830301378511958,
"learning_rate": 5.4188957799750145e-06,
"loss": 0.4505,
"step": 195
},
{
"epoch": 0.9514563106796117,
"grad_norm": 3.778636044638277,
"learning_rate": 5.380891883854591e-06,
"loss": 0.511,
"step": 196
},
{
"epoch": 0.9563106796116505,
"grad_norm": 3.4350894159519196,
"learning_rate": 5.34286584124299e-06,
"loss": 0.4249,
"step": 197
},
{
"epoch": 0.9611650485436893,
"grad_norm": 4.4692953750497155,
"learning_rate": 5.304819863117796e-06,
"loss": 0.5872,
"step": 198
},
{
"epoch": 0.9660194174757282,
"grad_norm": 4.031250356279573,
"learning_rate": 5.266756161615719e-06,
"loss": 0.5529,
"step": 199
},
{
"epoch": 0.970873786407767,
"grad_norm": 4.1745049180996014,
"learning_rate": 5.228676949903974e-06,
"loss": 0.6264,
"step": 200
},
{
"epoch": 0.9757281553398058,
"grad_norm": 4.01912731604719,
"learning_rate": 5.190584442051594e-06,
"loss": 0.5768,
"step": 201
},
{
"epoch": 0.9805825242718447,
"grad_norm": 3.6247949853458308,
"learning_rate": 5.1524808529007075e-06,
"loss": 0.556,
"step": 202
},
{
"epoch": 0.9854368932038835,
"grad_norm": 3.9142637838055756,
"learning_rate": 5.114368397937744e-06,
"loss": 0.5823,
"step": 203
},
{
"epoch": 0.9902912621359223,
"grad_norm": 3.5814741714879275,
"learning_rate": 5.07624929316463e-06,
"loss": 0.5492,
"step": 204
},
{
"epoch": 0.9951456310679612,
"grad_norm": 4.173978940529795,
"learning_rate": 5.038125754969933e-06,
"loss": 0.5574,
"step": 205
},
{
"epoch": 1.0,
"grad_norm": 4.0084672163268635,
"learning_rate": 5e-06,
"loss": 0.4134,
"step": 206
},
{
"epoch": 1.0048543689320388,
"grad_norm": 3.496605105902824,
"learning_rate": 4.9618742450300675e-06,
"loss": 0.3843,
"step": 207
},
{
"epoch": 1.0097087378640777,
"grad_norm": 3.7902712180485945,
"learning_rate": 4.923750706835371e-06,
"loss": 0.4444,
"step": 208
},
{
"epoch": 1.0145631067961165,
"grad_norm": 3.727987334348961,
"learning_rate": 4.8856316020622564e-06,
"loss": 0.4436,
"step": 209
},
{
"epoch": 1.0194174757281553,
"grad_norm": 2.631746304441649,
"learning_rate": 4.847519147099294e-06,
"loss": 0.3251,
"step": 210
},
{
"epoch": 1.0242718446601942,
"grad_norm": 3.7303118359118415,
"learning_rate": 4.809415557948407e-06,
"loss": 0.3962,
"step": 211
},
{
"epoch": 1.029126213592233,
"grad_norm": 2.7371328883886714,
"learning_rate": 4.771323050096028e-06,
"loss": 0.3591,
"step": 212
},
{
"epoch": 1.0339805825242718,
"grad_norm": 4.0908488142094575,
"learning_rate": 4.733243838384282e-06,
"loss": 0.4258,
"step": 213
},
{
"epoch": 1.0388349514563107,
"grad_norm": 3.684061911684185,
"learning_rate": 4.6951801368822055e-06,
"loss": 0.4439,
"step": 214
},
{
"epoch": 1.0436893203883495,
"grad_norm": 4.277830044538694,
"learning_rate": 4.6571341587570114e-06,
"loss": 0.4019,
"step": 215
},
{
"epoch": 1.0485436893203883,
"grad_norm": 3.017553350723136,
"learning_rate": 4.619108116145411e-06,
"loss": 0.3154,
"step": 216
},
{
"epoch": 1.0533980582524272,
"grad_norm": 3.8114352902805013,
"learning_rate": 4.581104220024988e-06,
"loss": 0.4494,
"step": 217
},
{
"epoch": 1.058252427184466,
"grad_norm": 3.573465696111325,
"learning_rate": 4.5431246800856455e-06,
"loss": 0.3623,
"step": 218
},
{
"epoch": 1.0631067961165048,
"grad_norm": 3.3846553881044414,
"learning_rate": 4.505171704601128e-06,
"loss": 0.3439,
"step": 219
},
{
"epoch": 1.0679611650485437,
"grad_norm": 3.802252845245479,
"learning_rate": 4.467247500300621e-06,
"loss": 0.3804,
"step": 220
},
{
"epoch": 1.0728155339805825,
"grad_norm": 3.992833646518992,
"learning_rate": 4.4293542722404435e-06,
"loss": 0.4043,
"step": 221
},
{
"epoch": 1.0776699029126213,
"grad_norm": 5.263590963330377,
"learning_rate": 4.391494223675843e-06,
"loss": 0.4186,
"step": 222
},
{
"epoch": 1.0825242718446602,
"grad_norm": 4.249756141920356,
"learning_rate": 4.3536695559328816e-06,
"loss": 0.3495,
"step": 223
},
{
"epoch": 1.087378640776699,
"grad_norm": 4.739841082847628,
"learning_rate": 4.31588246828045e-06,
"loss": 0.4242,
"step": 224
},
{
"epoch": 1.0922330097087378,
"grad_norm": 4.294969521268887,
"learning_rate": 4.278135157802389e-06,
"loss": 0.4646,
"step": 225
},
{
"epoch": 1.0970873786407767,
"grad_norm": 4.201585664309024,
"learning_rate": 4.240429819269746e-06,
"loss": 0.4105,
"step": 226
},
{
"epoch": 1.1019417475728155,
"grad_norm": 3.8196683565687435,
"learning_rate": 4.20276864501316e-06,
"loss": 0.3925,
"step": 227
},
{
"epoch": 1.1067961165048543,
"grad_norm": 4.446057887609339,
"learning_rate": 4.165153824795391e-06,
"loss": 0.3904,
"step": 228
},
{
"epoch": 1.1116504854368932,
"grad_norm": 3.767740266078369,
"learning_rate": 4.127587545684002e-06,
"loss": 0.391,
"step": 229
},
{
"epoch": 1.116504854368932,
"grad_norm": 3.272986147861369,
"learning_rate": 4.090071991924194e-06,
"loss": 0.3223,
"step": 230
},
{
"epoch": 1.1213592233009708,
"grad_norm": 3.4867657680648163,
"learning_rate": 4.052609344811802e-06,
"loss": 0.3472,
"step": 231
},
{
"epoch": 1.1262135922330097,
"grad_norm": 3.4311847620578004,
"learning_rate": 4.015201782566471e-06,
"loss": 0.337,
"step": 232
},
{
"epoch": 1.1310679611650485,
"grad_norm": 4.384639352526652,
"learning_rate": 3.977851480205003e-06,
"loss": 0.4596,
"step": 233
},
{
"epoch": 1.1359223300970873,
"grad_norm": 3.597532259556013,
"learning_rate": 3.940560609414894e-06,
"loss": 0.2823,
"step": 234
},
{
"epoch": 1.1407766990291262,
"grad_norm": 3.500219417381477,
"learning_rate": 3.903331338428067e-06,
"loss": 0.3789,
"step": 235
},
{
"epoch": 1.145631067961165,
"grad_norm": 3.675107189705643,
"learning_rate": 3.866165831894796e-06,
"loss": 0.3771,
"step": 236
},
{
"epoch": 1.1504854368932038,
"grad_norm": 3.3300467953923425,
"learning_rate": 3.829066250757851e-06,
"loss": 0.3019,
"step": 237
},
{
"epoch": 1.1553398058252426,
"grad_norm": 4.251324179761842,
"learning_rate": 3.7920347521268514e-06,
"loss": 0.4328,
"step": 238
},
{
"epoch": 1.1601941747572815,
"grad_norm": 2.921640821148963,
"learning_rate": 3.7550734891528413e-06,
"loss": 0.3385,
"step": 239
},
{
"epoch": 1.1650485436893203,
"grad_norm": 3.5714425790471553,
"learning_rate": 3.7181846109031007e-06,
"loss": 0.3602,
"step": 240
},
{
"epoch": 1.1699029126213591,
"grad_norm": 3.651206945750236,
"learning_rate": 3.6813702622361858e-06,
"loss": 0.354,
"step": 241
},
{
"epoch": 1.174757281553398,
"grad_norm": 4.548151987056661,
"learning_rate": 3.6446325836772244e-06,
"loss": 0.4343,
"step": 242
},
{
"epoch": 1.1796116504854368,
"grad_norm": 4.037140100225124,
"learning_rate": 3.6079737112934533e-06,
"loss": 0.4024,
"step": 243
},
{
"epoch": 1.1844660194174756,
"grad_norm": 3.878145136960394,
"learning_rate": 3.5713957765700224e-06,
"loss": 0.3472,
"step": 244
},
{
"epoch": 1.1893203883495145,
"grad_norm": 3.523053214989574,
"learning_rate": 3.5349009062860586e-06,
"loss": 0.4294,
"step": 245
},
{
"epoch": 1.1941747572815533,
"grad_norm": 4.158187259776681,
"learning_rate": 3.4984912223910105e-06,
"loss": 0.4077,
"step": 246
},
{
"epoch": 1.1990291262135921,
"grad_norm": 3.644466837721695,
"learning_rate": 3.46216884188127e-06,
"loss": 0.3064,
"step": 247
},
{
"epoch": 1.203883495145631,
"grad_norm": 3.7949333994527152,
"learning_rate": 3.425935876677077e-06,
"loss": 0.4356,
"step": 248
},
{
"epoch": 1.2087378640776698,
"grad_norm": 4.29761710488119,
"learning_rate": 3.38979443349973e-06,
"loss": 0.4547,
"step": 249
},
{
"epoch": 1.2135922330097086,
"grad_norm": 3.8023373204574105,
"learning_rate": 3.3537466137490937e-06,
"loss": 0.3415,
"step": 250
},
{
"epoch": 1.2184466019417475,
"grad_norm": 4.030137677070585,
"learning_rate": 3.3177945133814093e-06,
"loss": 0.4555,
"step": 251
},
{
"epoch": 1.2233009708737863,
"grad_norm": 4.164408617866339,
"learning_rate": 3.2819402227874364e-06,
"loss": 0.3341,
"step": 252
},
{
"epoch": 1.2281553398058254,
"grad_norm": 3.7914695128165374,
"learning_rate": 3.2461858266709017e-06,
"loss": 0.3716,
"step": 253
},
{
"epoch": 1.233009708737864,
"grad_norm": 3.167816846899686,
"learning_rate": 3.2105334039272924e-06,
"loss": 0.3482,
"step": 254
},
{
"epoch": 1.237864077669903,
"grad_norm": 3.106878527556992,
"learning_rate": 3.1749850275229777e-06,
"loss": 0.3295,
"step": 255
},
{
"epoch": 1.2427184466019416,
"grad_norm": 3.5266503004314838,
"learning_rate": 3.1395427643746802e-06,
"loss": 0.323,
"step": 256
},
{
"epoch": 1.2475728155339807,
"grad_norm": 3.612460833210493,
"learning_rate": 3.1042086752292995e-06,
"loss": 0.3653,
"step": 257
},
{
"epoch": 1.2524271844660193,
"grad_norm": 3.2630301729147155,
"learning_rate": 3.068984814544087e-06,
"loss": 0.3806,
"step": 258
},
{
"epoch": 1.2572815533980584,
"grad_norm": 5.011331004941775,
"learning_rate": 3.0338732303671993e-06,
"loss": 0.4043,
"step": 259
},
{
"epoch": 1.262135922330097,
"grad_norm": 3.680870440115493,
"learning_rate": 2.99887596421861e-06,
"loss": 0.3987,
"step": 260
},
{
"epoch": 1.266990291262136,
"grad_norm": 4.349360393471288,
"learning_rate": 2.9639950509714138e-06,
"loss": 0.4425,
"step": 261
},
{
"epoch": 1.2718446601941746,
"grad_norm": 3.8827103846414994,
"learning_rate": 2.929232518733507e-06,
"loss": 0.4829,
"step": 262
},
{
"epoch": 1.2766990291262137,
"grad_norm": 3.9702425284434306,
"learning_rate": 2.8945903887296686e-06,
"loss": 0.3523,
"step": 263
},
{
"epoch": 1.2815533980582523,
"grad_norm": 3.834178934738047,
"learning_rate": 2.860070675184036e-06,
"loss": 0.3874,
"step": 264
},
{
"epoch": 1.2864077669902914,
"grad_norm": 3.5351170944001002,
"learning_rate": 2.8256753852029917e-06,
"loss": 0.3936,
"step": 265
},
{
"epoch": 1.29126213592233,
"grad_norm": 3.950582302242211,
"learning_rate": 2.7914065186584637e-06,
"loss": 0.3978,
"step": 266
},
{
"epoch": 1.296116504854369,
"grad_norm": 4.0713639287929295,
"learning_rate": 2.757266068071648e-06,
"loss": 0.4088,
"step": 267
},
{
"epoch": 1.3009708737864076,
"grad_norm": 4.100332519310959,
"learning_rate": 2.7232560184971437e-06,
"loss": 0.3779,
"step": 268
},
{
"epoch": 1.3058252427184467,
"grad_norm": 3.7337329435040187,
"learning_rate": 2.689378347407553e-06,
"loss": 0.3994,
"step": 269
},
{
"epoch": 1.3106796116504853,
"grad_norm": 3.792259934851255,
"learning_rate": 2.6556350245784833e-06,
"loss": 0.3408,
"step": 270
},
{
"epoch": 1.3155339805825244,
"grad_norm": 3.5099418964063074,
"learning_rate": 2.6220280119740376e-06,
"loss": 0.2759,
"step": 271
},
{
"epoch": 1.3203883495145632,
"grad_norm": 4.469284032133799,
"learning_rate": 2.588559263632719e-06,
"loss": 0.5543,
"step": 272
},
{
"epoch": 1.325242718446602,
"grad_norm": 4.426616455207192,
"learning_rate": 2.555230725553832e-06,
"loss": 0.3267,
"step": 273
},
{
"epoch": 1.3300970873786409,
"grad_norm": 3.622623794447689,
"learning_rate": 2.522044335584322e-06,
"loss": 0.4094,
"step": 274
},
{
"epoch": 1.3349514563106797,
"grad_norm": 4.570982863821399,
"learning_rate": 2.489002023306112e-06,
"loss": 0.5116,
"step": 275
},
{
"epoch": 1.3398058252427185,
"grad_norm": 3.987317364219781,
"learning_rate": 2.4561057099238973e-06,
"loss": 0.4175,
"step": 276
},
{
"epoch": 1.3446601941747574,
"grad_norm": 4.380860733212577,
"learning_rate": 2.423357308153454e-06,
"loss": 0.4041,
"step": 277
},
{
"epoch": 1.3495145631067962,
"grad_norm": 3.2752217697947406,
"learning_rate": 2.390758722110418e-06,
"loss": 0.3615,
"step": 278
},
{
"epoch": 1.354368932038835,
"grad_norm": 3.8417595443552255,
"learning_rate": 2.358311847199567e-06,
"loss": 0.3952,
"step": 279
},
{
"epoch": 1.3592233009708738,
"grad_norm": 3.5643975822384903,
"learning_rate": 2.3260185700046295e-06,
"loss": 0.3373,
"step": 280
},
{
"epoch": 1.3640776699029127,
"grad_norm": 3.605860128653127,
"learning_rate": 2.2938807681785764e-06,
"loss": 0.4037,
"step": 281
},
{
"epoch": 1.3689320388349515,
"grad_norm": 4.225685654517484,
"learning_rate": 2.2619003103344607e-06,
"loss": 0.4018,
"step": 282
},
{
"epoch": 1.3737864077669903,
"grad_norm": 3.845865695831952,
"learning_rate": 2.2300790559367553e-06,
"loss": 0.3598,
"step": 283
},
{
"epoch": 1.3786407766990292,
"grad_norm": 4.292354780245367,
"learning_rate": 2.1984188551932513e-06,
"loss": 0.3907,
"step": 284
},
{
"epoch": 1.383495145631068,
"grad_norm": 3.873410505724975,
"learning_rate": 2.166921548947466e-06,
"loss": 0.372,
"step": 285
},
{
"epoch": 1.3883495145631068,
"grad_norm": 4.34595859540743,
"learning_rate": 2.1355889685716225e-06,
"loss": 0.463,
"step": 286
},
{
"epoch": 1.3932038834951457,
"grad_norm": 4.658967796353035,
"learning_rate": 2.1044229358601543e-06,
"loss": 0.4594,
"step": 287
},
{
"epoch": 1.3980582524271845,
"grad_norm": 3.730083545740747,
"learning_rate": 2.0734252629237892e-06,
"loss": 0.3517,
"step": 288
},
{
"epoch": 1.4029126213592233,
"grad_norm": 3.6893395131508573,
"learning_rate": 2.0425977520841837e-06,
"loss": 0.3345,
"step": 289
},
{
"epoch": 1.4077669902912622,
"grad_norm": 3.2326491411528715,
"learning_rate": 2.011942195769122e-06,
"loss": 0.3402,
"step": 290
},
{
"epoch": 1.412621359223301,
"grad_norm": 4.341968906331775,
"learning_rate": 1.9814603764083112e-06,
"loss": 0.4244,
"step": 291
},
{
"epoch": 1.4174757281553398,
"grad_norm": 4.638374628949702,
"learning_rate": 1.9511540663297284e-06,
"loss": 0.428,
"step": 292
},
{
"epoch": 1.4223300970873787,
"grad_norm": 3.8746258196888017,
"learning_rate": 1.921025027656587e-06,
"loss": 0.3443,
"step": 293
},
{
"epoch": 1.4271844660194175,
"grad_norm": 4.4526655613630695,
"learning_rate": 1.8910750122048638e-06,
"loss": 0.4191,
"step": 294
},
{
"epoch": 1.4320388349514563,
"grad_norm": 4.192734287016373,
"learning_rate": 1.8613057613814584e-06,
"loss": 0.4304,
"step": 295
},
{
"epoch": 1.4368932038834952,
"grad_norm": 3.916006256696625,
"learning_rate": 1.8317190060829242e-06,
"loss": 0.4808,
"step": 296
},
{
"epoch": 1.441747572815534,
"grad_norm": 4.171527243642291,
"learning_rate": 1.8023164665948455e-06,
"loss": 0.4255,
"step": 297
},
{
"epoch": 1.4466019417475728,
"grad_norm": 3.963819376454969,
"learning_rate": 1.773099852491796e-06,
"loss": 0.3999,
"step": 298
},
{
"epoch": 1.4514563106796117,
"grad_norm": 3.2106436226036723,
"learning_rate": 1.7440708625379503e-06,
"loss": 0.3352,
"step": 299
},
{
"epoch": 1.4563106796116505,
"grad_norm": 4.155391926087341,
"learning_rate": 1.7152311845883096e-06,
"loss": 0.3465,
"step": 300
},
{
"epoch": 1.4611650485436893,
"grad_norm": 3.8279845137959323,
"learning_rate": 1.686582495490554e-06,
"loss": 0.4383,
"step": 301
},
{
"epoch": 1.4660194174757282,
"grad_norm": 3.6446053398247154,
"learning_rate": 1.658126460987558e-06,
"loss": 0.3517,
"step": 302
},
{
"epoch": 1.470873786407767,
"grad_norm": 4.278232584198821,
"learning_rate": 1.6298647356205255e-06,
"loss": 0.4239,
"step": 303
},
{
"epoch": 1.4757281553398058,
"grad_norm": 3.553632609839044,
"learning_rate": 1.601798962632799e-06,
"loss": 0.3224,
"step": 304
},
{
"epoch": 1.4805825242718447,
"grad_norm": 4.490197897574275,
"learning_rate": 1.573930773874306e-06,
"loss": 0.387,
"step": 305
},
{
"epoch": 1.4854368932038835,
"grad_norm": 3.804873355604368,
"learning_rate": 1.5462617897066863e-06,
"loss": 0.3654,
"step": 306
},
{
"epoch": 1.4902912621359223,
"grad_norm": 3.177625767465489,
"learning_rate": 1.5187936189090668e-06,
"loss": 0.347,
"step": 307
},
{
"epoch": 1.4951456310679612,
"grad_norm": 3.813427856557575,
"learning_rate": 1.491527858584535e-06,
"loss": 0.3681,
"step": 308
},
{
"epoch": 1.5,
"grad_norm": 3.6284839940706752,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.4381,
"step": 309
},
{
"epoch": 1.5048543689320388,
"grad_norm": 3.810598592463692,
"learning_rate": 1.4376098988303406e-06,
"loss": 0.332,
"step": 310
},
{
"epoch": 1.5097087378640777,
"grad_norm": 3.63488700987213,
"learning_rate": 1.4109608343942855e-06,
"loss": 0.3336,
"step": 311
},
{
"epoch": 1.5145631067961165,
"grad_norm": 3.850419264016209,
"learning_rate": 1.3845204502362442e-06,
"loss": 0.406,
"step": 312
},
{
"epoch": 1.5194174757281553,
"grad_norm": 4.412242124984995,
"learning_rate": 1.35829028369991e-06,
"loss": 0.4057,
"step": 313
},
{
"epoch": 1.5242718446601942,
"grad_norm": 3.5851483721870387,
"learning_rate": 1.3322718599061252e-06,
"loss": 0.3848,
"step": 314
},
{
"epoch": 1.529126213592233,
"grad_norm": 4.505744106720642,
"learning_rate": 1.306466691664216e-06,
"loss": 0.4659,
"step": 315
},
{
"epoch": 1.5339805825242718,
"grad_norm": 4.084664225108395,
"learning_rate": 1.28087627938402e-06,
"loss": 0.4441,
"step": 316
},
{
"epoch": 1.5388349514563107,
"grad_norm": 3.4922984464639666,
"learning_rate": 1.2555021109886589e-06,
"loss": 0.3918,
"step": 317
},
{
"epoch": 1.5436893203883495,
"grad_norm": 4.16388125743772,
"learning_rate": 1.2303456618280141e-06,
"loss": 0.3978,
"step": 318
},
{
"epoch": 1.5485436893203883,
"grad_norm": 3.620132639015818,
"learning_rate": 1.2054083945929534e-06,
"loss": 0.3301,
"step": 319
},
{
"epoch": 1.5533980582524272,
"grad_norm": 3.182519671999271,
"learning_rate": 1.1806917592302763e-06,
"loss": 0.344,
"step": 320
},
{
"epoch": 1.558252427184466,
"grad_norm": 4.1356750212302815,
"learning_rate": 1.1561971928584158e-06,
"loss": 0.4009,
"step": 321
},
{
"epoch": 1.5631067961165048,
"grad_norm": 3.762778213476317,
"learning_rate": 1.1319261196838782e-06,
"loss": 0.344,
"step": 322
},
{
"epoch": 1.5679611650485437,
"grad_norm": 4.77670450513713,
"learning_rate": 1.1078799509184246e-06,
"loss": 0.4363,
"step": 323
},
{
"epoch": 1.5728155339805825,
"grad_norm": 4.97992620031663,
"learning_rate": 1.0840600846970333e-06,
"loss": 0.4247,
"step": 324
},
{
"epoch": 1.5776699029126213,
"grad_norm": 3.6214172269077562,
"learning_rate": 1.0604679059965923e-06,
"loss": 0.3682,
"step": 325
},
{
"epoch": 1.5825242718446602,
"grad_norm": 4.038128806113305,
"learning_rate": 1.0371047865553847e-06,
"loss": 0.4836,
"step": 326
},
{
"epoch": 1.587378640776699,
"grad_norm": 4.218674648713838,
"learning_rate": 1.0139720847933166e-06,
"loss": 0.4054,
"step": 327
},
{
"epoch": 1.5922330097087378,
"grad_norm": 3.4524660941213448,
"learning_rate": 9.91071145732948e-07,
"loss": 0.3966,
"step": 328
},
{
"epoch": 1.5970873786407767,
"grad_norm": 3.8418211425667175,
"learning_rate": 9.684033009212752e-07,
"loss": 0.4099,
"step": 329
},
{
"epoch": 1.6019417475728155,
"grad_norm": 4.518230081457923,
"learning_rate": 9.459698683523205e-07,
"loss": 0.3864,
"step": 330
},
{
"epoch": 1.6067961165048543,
"grad_norm": 4.444081075097612,
"learning_rate": 9.237721523904891e-07,
"loss": 0.4827,
"step": 331
},
{
"epoch": 1.6116504854368932,
"grad_norm": 2.994214298809699,
"learning_rate": 9.018114436947373e-07,
"loss": 0.3293,
"step": 332
},
{
"epoch": 1.616504854368932,
"grad_norm": 3.719084184765044,
"learning_rate": 8.80089019143524e-07,
"loss": 0.3616,
"step": 333
},
{
"epoch": 1.6213592233009708,
"grad_norm": 3.6989856224639963,
"learning_rate": 8.586061417605668e-07,
"loss": 0.3793,
"step": 334
},
{
"epoch": 1.6262135922330097,
"grad_norm": 3.8235469463286216,
"learning_rate": 8.373640606414097e-07,
"loss": 0.3305,
"step": 335
},
{
"epoch": 1.6310679611650487,
"grad_norm": 3.654609458651861,
"learning_rate": 8.163640108807897e-07,
"loss": 0.39,
"step": 336
},
{
"epoch": 1.6359223300970873,
"grad_norm": 3.4550101567686897,
"learning_rate": 7.956072135008336e-07,
"loss": 0.3956,
"step": 337
},
{
"epoch": 1.6407766990291264,
"grad_norm": 3.6284063410127234,
"learning_rate": 7.750948753800508e-07,
"loss": 0.3376,
"step": 338
},
{
"epoch": 1.645631067961165,
"grad_norm": 4.338443444747467,
"learning_rate": 7.548281891831715e-07,
"loss": 0.443,
"step": 339
},
{
"epoch": 1.650485436893204,
"grad_norm": 3.855988692161462,
"learning_rate": 7.348083332917927e-07,
"loss": 0.3337,
"step": 340
},
{
"epoch": 1.6553398058252426,
"grad_norm": 4.897930316305917,
"learning_rate": 7.150364717358699e-07,
"loss": 0.3827,
"step": 341
},
{
"epoch": 1.6601941747572817,
"grad_norm": 3.6497831827277003,
"learning_rate": 6.955137541260287e-07,
"loss": 0.3707,
"step": 342
},
{
"epoch": 1.6650485436893203,
"grad_norm": 4.211838424144729,
"learning_rate": 6.762413155867276e-07,
"loss": 0.3556,
"step": 343
},
{
"epoch": 1.6699029126213594,
"grad_norm": 4.513690726677611,
"learning_rate": 6.572202766902569e-07,
"loss": 0.5142,
"step": 344
},
{
"epoch": 1.674757281553398,
"grad_norm": 4.00506710748267,
"learning_rate": 6.384517433915794e-07,
"loss": 0.4298,
"step": 345
},
{
"epoch": 1.679611650485437,
"grad_norm": 3.708974626189533,
"learning_rate": 6.199368069640343e-07,
"loss": 0.3991,
"step": 346
},
{
"epoch": 1.6844660194174756,
"grad_norm": 3.8803373599498525,
"learning_rate": 6.016765439358774e-07,
"loss": 0.4265,
"step": 347
},
{
"epoch": 1.6893203883495147,
"grad_norm": 4.243738248525009,
"learning_rate": 5.836720160276971e-07,
"loss": 0.4084,
"step": 348
},
{
"epoch": 1.6941747572815533,
"grad_norm": 3.471552501208322,
"learning_rate": 5.659242700906719e-07,
"loss": 0.3506,
"step": 349
},
{
"epoch": 1.6990291262135924,
"grad_norm": 3.6224163318043305,
"learning_rate": 5.484343380457124e-07,
"loss": 0.3986,
"step": 350
},
{
"epoch": 1.703883495145631,
"grad_norm": 3.5561959644452426,
"learning_rate": 5.312032368234527e-07,
"loss": 0.3746,
"step": 351
},
{
"epoch": 1.70873786407767,
"grad_norm": 3.890277485479215,
"learning_rate": 5.1423196830513e-07,
"loss": 0.3543,
"step": 352
},
{
"epoch": 1.7135922330097086,
"grad_norm": 3.97276022754447,
"learning_rate": 4.975215192643246e-07,
"loss": 0.3562,
"step": 353
},
{
"epoch": 1.7184466019417477,
"grad_norm": 4.167194231694843,
"learning_rate": 4.81072861309591e-07,
"loss": 0.34,
"step": 354
},
{
"epoch": 1.7233009708737863,
"grad_norm": 3.816302512326562,
"learning_rate": 4.648869508279613e-07,
"loss": 0.4056,
"step": 355
},
{
"epoch": 1.7281553398058254,
"grad_norm": 3.589002492007345,
"learning_rate": 4.4896472892933693e-07,
"loss": 0.3196,
"step": 356
},
{
"epoch": 1.733009708737864,
"grad_norm": 3.788504484004072,
"learning_rate": 4.333071213917722e-07,
"loss": 0.3666,
"step": 357
},
{
"epoch": 1.737864077669903,
"grad_norm": 3.7373395503876097,
"learning_rate": 4.179150386076425e-07,
"loss": 0.4119,
"step": 358
},
{
"epoch": 1.7427184466019416,
"grad_norm": 4.118560675638611,
"learning_rate": 4.027893755307144e-07,
"loss": 0.3885,
"step": 359
},
{
"epoch": 1.7475728155339807,
"grad_norm": 4.214811979508283,
"learning_rate": 3.8793101162410417e-07,
"loss": 0.4314,
"step": 360
},
{
"epoch": 1.7524271844660193,
"grad_norm": 4.083658384432838,
"learning_rate": 3.733408108091485e-07,
"loss": 0.3515,
"step": 361
},
{
"epoch": 1.7572815533980584,
"grad_norm": 4.533971137542316,
"learning_rate": 3.5901962141516975e-07,
"loss": 0.4136,
"step": 362
},
{
"epoch": 1.762135922330097,
"grad_norm": 3.8477578380636372,
"learning_rate": 3.4496827613015206e-07,
"loss": 0.4237,
"step": 363
},
{
"epoch": 1.766990291262136,
"grad_norm": 3.5799317137924436,
"learning_rate": 3.3118759195232273e-07,
"loss": 0.3869,
"step": 364
},
{
"epoch": 1.7718446601941746,
"grad_norm": 3.7650628981044116,
"learning_rate": 3.176783701426528e-07,
"loss": 0.4454,
"step": 365
},
{
"epoch": 1.7766990291262137,
"grad_norm": 3.949215927517889,
"learning_rate": 3.0444139617826605e-07,
"loss": 0.3893,
"step": 366
},
{
"epoch": 1.7815533980582523,
"grad_norm": 4.040866385983115,
"learning_rate": 2.91477439706771e-07,
"loss": 0.4147,
"step": 367
},
{
"epoch": 1.7864077669902914,
"grad_norm": 3.813616771397068,
"learning_rate": 2.787872545015069e-07,
"loss": 0.3311,
"step": 368
},
{
"epoch": 1.79126213592233,
"grad_norm": 3.7762362750478884,
"learning_rate": 2.663715784177201e-07,
"loss": 0.3649,
"step": 369
},
{
"epoch": 1.796116504854369,
"grad_norm": 5.360237076084266,
"learning_rate": 2.542311333496622e-07,
"loss": 0.4443,
"step": 370
},
{
"epoch": 1.8009708737864076,
"grad_norm": 4.264947081745297,
"learning_rate": 2.423666251886114e-07,
"loss": 0.4609,
"step": 371
},
{
"epoch": 1.8058252427184467,
"grad_norm": 4.097010679962042,
"learning_rate": 2.307787437818365e-07,
"loss": 0.3758,
"step": 372
},
{
"epoch": 1.8106796116504853,
"grad_norm": 3.8713575580421873,
"learning_rate": 2.1946816289248163e-07,
"loss": 0.4887,
"step": 373
},
{
"epoch": 1.8155339805825244,
"grad_norm": 3.8942355876784123,
"learning_rate": 2.0843554016039326e-07,
"loss": 0.4069,
"step": 374
},
{
"epoch": 1.820388349514563,
"grad_norm": 3.6942656116086567,
"learning_rate": 1.9768151706388016e-07,
"loss": 0.4312,
"step": 375
},
{
"epoch": 1.825242718446602,
"grad_norm": 4.3141992575654875,
"learning_rate": 1.8720671888242058e-07,
"loss": 0.3957,
"step": 376
},
{
"epoch": 1.8300970873786406,
"grad_norm": 4.281607466801436,
"learning_rate": 1.7701175466029895e-07,
"loss": 0.4216,
"step": 377
},
{
"epoch": 1.8349514563106797,
"grad_norm": 4.064660344235405,
"learning_rate": 1.6709721717120042e-07,
"loss": 0.4566,
"step": 378
},
{
"epoch": 1.8398058252427183,
"grad_norm": 3.304394890132256,
"learning_rate": 1.574636828837395e-07,
"loss": 0.3435,
"step": 379
},
{
"epoch": 1.8446601941747574,
"grad_norm": 3.7411847131143987,
"learning_rate": 1.4811171192794628e-07,
"loss": 0.4241,
"step": 380
},
{
"epoch": 1.849514563106796,
"grad_norm": 3.768761878129938,
"learning_rate": 1.3904184806269705e-07,
"loss": 0.3789,
"step": 381
},
{
"epoch": 1.854368932038835,
"grad_norm": 3.5827717745983096,
"learning_rate": 1.3025461864409395e-07,
"loss": 0.3644,
"step": 382
},
{
"epoch": 1.8592233009708736,
"grad_norm": 4.244439889594276,
"learning_rate": 1.2175053459481e-07,
"loss": 0.4193,
"step": 383
},
{
"epoch": 1.8640776699029127,
"grad_norm": 4.599340721540073,
"learning_rate": 1.1353009037437523e-07,
"loss": 0.4277,
"step": 384
},
{
"epoch": 1.8689320388349513,
"grad_norm": 4.101506699054163,
"learning_rate": 1.0559376395043285e-07,
"loss": 0.4106,
"step": 385
},
{
"epoch": 1.8737864077669903,
"grad_norm": 3.71635245861999,
"learning_rate": 9.794201677094162e-08,
"loss": 0.3746,
"step": 386
},
{
"epoch": 1.8786407766990292,
"grad_norm": 3.989534729025634,
"learning_rate": 9.05752937373533e-08,
"loss": 0.3584,
"step": 387
},
{
"epoch": 1.883495145631068,
"grad_norm": 3.9585534274269762,
"learning_rate": 8.34940231787379e-08,
"loss": 0.3845,
"step": 388
},
{
"epoch": 1.8883495145631068,
"grad_norm": 3.814761006523607,
"learning_rate": 7.66986168268824e-08,
"loss": 0.3677,
"step": 389
},
{
"epoch": 1.8932038834951457,
"grad_norm": 4.054845824130783,
"learning_rate": 7.018946979234997e-08,
"loss": 0.3727,
"step": 390
},
{
"epoch": 1.8980582524271845,
"grad_norm": 4.067971393105176,
"learning_rate": 6.396696054150719e-08,
"loss": 0.3255,
"step": 391
},
{
"epoch": 1.9029126213592233,
"grad_norm": 4.191540621130695,
"learning_rate": 5.803145087451945e-08,
"loss": 0.4115,
"step": 392
},
{
"epoch": 1.9077669902912622,
"grad_norm": 4.74293215790585,
"learning_rate": 5.238328590431163e-08,
"loss": 0.4783,
"step": 393
},
{
"epoch": 1.912621359223301,
"grad_norm": 3.77410325549763,
"learning_rate": 4.702279403650534e-08,
"loss": 0.3517,
"step": 394
},
{
"epoch": 1.9174757281553398,
"grad_norm": 4.311940113885601,
"learning_rate": 4.195028695032133e-08,
"loss": 0.4141,
"step": 395
},
{
"epoch": 1.9223300970873787,
"grad_norm": 4.579804678462541,
"learning_rate": 3.716605958046071e-08,
"loss": 0.3849,
"step": 396
},
{
"epoch": 1.9271844660194175,
"grad_norm": 3.914230952884989,
"learning_rate": 3.2670390099951985e-08,
"loss": 0.346,
"step": 397
},
{
"epoch": 1.9320388349514563,
"grad_norm": 4.1100647312486425,
"learning_rate": 2.846353990398065e-08,
"loss": 0.4072,
"step": 398
},
{
"epoch": 1.9368932038834952,
"grad_norm": 3.6078645346574927,
"learning_rate": 2.4545753594688582e-08,
"loss": 0.3719,
"step": 399
},
{
"epoch": 1.941747572815534,
"grad_norm": 4.147236568941717,
"learning_rate": 2.0917258966953735e-08,
"loss": 0.4069,
"step": 400
},
{
"epoch": 1.9466019417475728,
"grad_norm": 3.55427554003578,
"learning_rate": 1.757826699514298e-08,
"loss": 0.3629,
"step": 401
},
{
"epoch": 1.9514563106796117,
"grad_norm": 3.734626654364938,
"learning_rate": 1.4528971820846894e-08,
"loss": 0.3173,
"step": 402
},
{
"epoch": 1.9563106796116505,
"grad_norm": 3.5211230545408907,
"learning_rate": 1.176955074159214e-08,
"loss": 0.3455,
"step": 403
},
{
"epoch": 1.9611650485436893,
"grad_norm": 4.142334626863241,
"learning_rate": 9.300164200530815e-09,
"loss": 0.4251,
"step": 404
},
{
"epoch": 1.9660194174757282,
"grad_norm": 3.40413114892008,
"learning_rate": 7.120955777112915e-09,
"loss": 0.343,
"step": 405
},
{
"epoch": 1.970873786407767,
"grad_norm": 3.667483870121202,
"learning_rate": 5.232052178738567e-09,
"loss": 0.3065,
"step": 406
},
{
"epoch": 1.9757281553398058,
"grad_norm": 4.150885828781062,
"learning_rate": 3.633563233388926e-09,
"loss": 0.3697,
"step": 407
},
{
"epoch": 1.9805825242718447,
"grad_norm": 4.025412485385259,
"learning_rate": 2.3255818832423894e-09,
"loss": 0.3903,
"step": 408
},
{
"epoch": 1.9854368932038835,
"grad_norm": 4.233175242592799,
"learning_rate": 1.3081841792694783e-09,
"loss": 0.3698,
"step": 409
},
{
"epoch": 1.9902912621359223,
"grad_norm": 3.587695193162535,
"learning_rate": 5.814292768108187e-10,
"loss": 0.5008,
"step": 410
},
{
"epoch": 1.9951456310679612,
"grad_norm": 3.9996526032023407,
"learning_rate": 1.453594321393359e-10,
"loss": 0.39,
"step": 411
},
{
"epoch": 2.0,
"grad_norm": 3.061876894624852,
"learning_rate": 0.0,
"loss": 0.304,
"step": 412
},
{
"epoch": 2.0,
"step": 412,
"total_flos": 1056725729280.0,
"train_loss": 0.4808927903910285,
"train_runtime": 257.5102,
"train_samples_per_second": 12.745,
"train_steps_per_second": 1.6
}
],
"logging_steps": 1,
"max_steps": 412,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 50000000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1056725729280.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}