EtashGuha's picture
Upload folder using huggingface_hub
00e666a verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9904153354632586,
"eval_steps": 500,
"global_step": 312,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009584664536741214,
"grad_norm": 6.384647846221924,
"learning_rate": 3.125e-07,
"loss": 1.0698,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 6.609872817993164,
"learning_rate": 6.25e-07,
"loss": 1.0636,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 6.251129150390625,
"learning_rate": 9.375000000000001e-07,
"loss": 1.0254,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 6.678479194641113,
"learning_rate": 1.25e-06,
"loss": 1.1069,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 6.002686023712158,
"learning_rate": 1.5625e-06,
"loss": 1.0133,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 6.0887370109558105,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.0772,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 4.826687335968018,
"learning_rate": 2.1875000000000002e-06,
"loss": 0.9895,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.832592010498047,
"learning_rate": 2.5e-06,
"loss": 1.0408,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 2.666738271713257,
"learning_rate": 2.8125e-06,
"loss": 0.9769,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.3659074306488037,
"learning_rate": 3.125e-06,
"loss": 0.9635,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.1485908031463623,
"learning_rate": 3.4375e-06,
"loss": 0.9435,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.6152048110961914,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.9127,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 4.0634236335754395,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.919,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.8888704776763916,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9266,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 3.6467268466949463,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.8877,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.771181344985962,
"learning_rate": 5e-06,
"loss": 0.8665,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.460860252380371,
"learning_rate": 5.3125e-06,
"loss": 0.8846,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.1305549144744873,
"learning_rate": 5.625e-06,
"loss": 0.8958,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.9240392446517944,
"learning_rate": 5.9375e-06,
"loss": 0.8334,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.7840166091918945,
"learning_rate": 6.25e-06,
"loss": 0.8437,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.680340051651001,
"learning_rate": 6.5625e-06,
"loss": 0.838,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.7359209060668945,
"learning_rate": 6.875e-06,
"loss": 0.8606,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.575268030166626,
"learning_rate": 7.1875e-06,
"loss": 0.8025,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.440421462059021,
"learning_rate": 7.500000000000001e-06,
"loss": 0.7837,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.2794287204742432,
"learning_rate": 7.8125e-06,
"loss": 0.7693,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.1461139917373657,
"learning_rate": 8.125000000000001e-06,
"loss": 0.829,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.2778260707855225,
"learning_rate": 8.4375e-06,
"loss": 0.809,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.2392792701721191,
"learning_rate": 8.750000000000001e-06,
"loss": 0.7858,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 1.1554087400436401,
"learning_rate": 9.0625e-06,
"loss": 0.8247,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 1.0011378526687622,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7572,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.9859980344772339,
"learning_rate": 9.6875e-06,
"loss": 0.7591,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 1.0400274991989136,
"learning_rate": 1e-05,
"loss": 0.7461,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 1.0274598598480225,
"learning_rate": 9.999685283773504e-06,
"loss": 0.8076,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.7825862765312195,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7362,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.987974226474762,
"learning_rate": 9.997167791667668e-06,
"loss": 0.7522,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 5.671663284301758,
"learning_rate": 9.994965332706574e-06,
"loss": 0.8798,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 1.1712332963943481,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7711,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.7852103114128113,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7244,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.8656781911849976,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7421,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.9151356816291809,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7601,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.6975906491279602,
"learning_rate": 9.974529372743762e-06,
"loss": 0.7301,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.9235749840736389,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7676,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.8292255401611328,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7443,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.7527896165847778,
"learning_rate": 9.954748808839675e-06,
"loss": 0.7466,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.7226002216339111,
"learning_rate": 9.946906630265184e-06,
"loss": 0.7469,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.8879698514938354,
"learning_rate": 9.938441702975689e-06,
"loss": 0.736,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.74617999792099,
"learning_rate": 9.92935509259118e-06,
"loss": 0.7281,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.7005394101142883,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7598,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.8603487610816956,
"learning_rate": 9.909321476180594e-06,
"loss": 0.7363,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.6847314834594727,
"learning_rate": 9.898376992116179e-06,
"loss": 0.7226,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.7452696561813354,
"learning_rate": 9.886815868562596e-06,
"loss": 0.7514,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.7947917580604553,
"learning_rate": 9.874639560909118e-06,
"loss": 0.7379,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.6648988127708435,
"learning_rate": 9.861849601988384e-06,
"loss": 0.7049,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.6720185875892639,
"learning_rate": 9.848447601883436e-06,
"loss": 0.732,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.6735864877700806,
"learning_rate": 9.834435247725032e-06,
"loss": 0.7102,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.7485708594322205,
"learning_rate": 9.819814303479268e-06,
"loss": 0.7165,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.6619908213615417,
"learning_rate": 9.804586609725499e-06,
"loss": 0.7425,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.6923986673355103,
"learning_rate": 9.788754083424654e-06,
"loss": 0.7354,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.6507387757301331,
"learning_rate": 9.772318717677905e-06,
"loss": 0.6848,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.616255521774292,
"learning_rate": 9.755282581475769e-06,
"loss": 0.7163,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.7321062684059143,
"learning_rate": 9.737647819437645e-06,
"loss": 0.7105,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.6692952513694763,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7283,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.7782991528511047,
"learning_rate": 9.700591372846096e-06,
"loss": 0.7068,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.7885490655899048,
"learning_rate": 9.681174353198687e-06,
"loss": 0.7236,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.6257630586624146,
"learning_rate": 9.661168036940071e-06,
"loss": 0.6918,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.7561619281768799,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7016,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.8016512989997864,
"learning_rate": 9.619397662556434e-06,
"loss": 0.7334,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.6697183847427368,
"learning_rate": 9.597638862757255e-06,
"loss": 0.6953,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.8066216111183167,
"learning_rate": 9.5753012823366e-06,
"loss": 0.7068,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.6715505719184875,
"learning_rate": 9.552387733294081e-06,
"loss": 0.6805,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.9659311175346375,
"learning_rate": 9.528901100135971e-06,
"loss": 0.7615,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.678727924823761,
"learning_rate": 9.504844339512096e-06,
"loss": 0.6942,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.7602061033248901,
"learning_rate": 9.480220479843627e-06,
"loss": 0.7242,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.7977614998817444,
"learning_rate": 9.45503262094184e-06,
"loss": 0.7243,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.7053202986717224,
"learning_rate": 9.4292839336179e-06,
"loss": 0.6979,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.6573786735534668,
"learning_rate": 9.40297765928369e-06,
"loss": 0.7081,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.62590491771698,
"learning_rate": 9.376117109543769e-06,
"loss": 0.6977,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.7059029340744019,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7033,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.8483073711395264,
"learning_rate": 9.320746778718274e-06,
"loss": 0.7179,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.6268270015716553,
"learning_rate": 9.292243968009332e-06,
"loss": 0.701,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.6996360421180725,
"learning_rate": 9.263200821770462e-06,
"loss": 0.7265,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.7423107028007507,
"learning_rate": 9.233620996141421e-06,
"loss": 0.7025,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.6664867997169495,
"learning_rate": 9.203508214822652e-06,
"loss": 0.7429,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.605104923248291,
"learning_rate": 9.172866268606514e-06,
"loss": 0.6735,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.5932936072349548,
"learning_rate": 9.141699014900084e-06,
"loss": 0.7267,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.6191230416297913,
"learning_rate": 9.110010377239552e-06,
"loss": 0.6776,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.5926156640052795,
"learning_rate": 9.077804344796302e-06,
"loss": 0.6631,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.6015069484710693,
"learning_rate": 9.045084971874738e-06,
"loss": 0.719,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.5310649871826172,
"learning_rate": 9.011856377401891e-06,
"loss": 0.6741,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.6096251606941223,
"learning_rate": 8.978122744408905e-06,
"loss": 0.6985,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.6101433634757996,
"learning_rate": 8.943888319504456e-06,
"loss": 0.6716,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.6175541281700134,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6948,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.6703361868858337,
"learning_rate": 8.873934395068006e-06,
"loss": 0.7445,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.6667814254760742,
"learning_rate": 8.838223701790057e-06,
"loss": 0.7033,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.6830708384513855,
"learning_rate": 8.802029828000157e-06,
"loss": 0.6971,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.5849157571792603,
"learning_rate": 8.765357330018056e-06,
"loss": 0.6982,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.6008332967758179,
"learning_rate": 8.728210824415829e-06,
"loss": 0.669,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.6959962844848633,
"learning_rate": 8.690594987436705e-06,
"loss": 0.7144,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.6326403021812439,
"learning_rate": 8.652514554406388e-06,
"loss": 0.695,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.6814602017402649,
"learning_rate": 8.613974319136959e-06,
"loss": 0.706,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.8258033990859985,
"learning_rate": 8.574979133323378e-06,
"loss": 0.7155,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.6495648622512817,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6934,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.613713264465332,
"learning_rate": 8.495643602586287e-06,
"loss": 0.6738,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.689940333366394,
"learning_rate": 8.455313244934324e-06,
"loss": 0.6645,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.6975083947181702,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6966,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.5884633660316467,
"learning_rate": 8.373352729660373e-06,
"loss": 0.6461,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.6319261193275452,
"learning_rate": 8.331732889760021e-06,
"loss": 0.6193,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.6757458448410034,
"learning_rate": 8.289693629698564e-06,
"loss": 0.647,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.6059986352920532,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6401,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.64361172914505,
"learning_rate": 8.204378069925121e-06,
"loss": 0.6139,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.6705755591392517,
"learning_rate": 8.16111251028955e-06,
"loss": 0.6079,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.6351562738418579,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6101,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.7013677954673767,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6261,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.6416124105453491,
"learning_rate": 8.0289502192041e-06,
"loss": 0.623,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.5879318118095398,
"learning_rate": 7.984126070912519e-06,
"loss": 0.6338,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.6589182019233704,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6025,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.614884078502655,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6056,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.5856279134750366,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6105,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.6292409896850586,
"learning_rate": 7.801129998764014e-06,
"loss": 0.642,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.5748828649520874,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6362,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.5927382111549377,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6228,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.5776206254959106,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6139,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.5544201731681824,
"learning_rate": 7.612492823579744e-06,
"loss": 0.5939,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.5677115321159363,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6249,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.6009501218795776,
"learning_rate": 7.516177115029002e-06,
"loss": 0.6163,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.6300259828567505,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6293,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.5929584503173828,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.6217,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.5914409160614014,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6348,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.5696073174476624,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5984,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.5948671102523804,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6247,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.5504754781723022,
"learning_rate": 7.219825320152411e-06,
"loss": 0.6142,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.5680699944496155,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6139,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.6490772366523743,
"learning_rate": 7.118738970516944e-06,
"loss": 0.6205,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.535884439945221,
"learning_rate": 7.067792524832604e-06,
"loss": 0.5988,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.5742437839508057,
"learning_rate": 7.016585772004026e-06,
"loss": 0.6205,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.6046292781829834,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6053,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.5980097651481628,
"learning_rate": 6.913417161825449e-06,
"loss": 0.616,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.586821436882019,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.6114,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.6167811751365662,
"learning_rate": 6.809285088483361e-06,
"loss": 0.614,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.6146674156188965,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.617,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.5550600290298462,
"learning_rate": 6.704241985612625e-06,
"loss": 0.6355,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.553723156452179,
"learning_rate": 6.651395309775837e-06,
"loss": 0.5918,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.5631532669067383,
"learning_rate": 6.598340745578908e-06,
"loss": 0.6085,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.5185543298721313,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6156,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.49711281061172485,
"learning_rate": 6.491634692845781e-06,
"loss": 0.6161,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.6072052121162415,
"learning_rate": 6.437996637160086e-06,
"loss": 0.5895,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.5126143097877502,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6041,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.5021834969520569,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6102,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.5503814220428467,
"learning_rate": 6.276023446318214e-06,
"loss": 0.6378,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 1.8309805393218994,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6238,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.6785635352134705,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.6062,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.6122629642486572,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6375,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.5528141260147095,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6365,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.6375743746757507,
"learning_rate": 6.002947078916365e-06,
"loss": 0.6459,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.6211557984352112,
"learning_rate": 5.947925441958393e-06,
"loss": 0.6112,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.6069737076759338,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6318,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.6636558175086975,
"learning_rate": 5.837531116523683e-06,
"loss": 0.6237,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.5873900055885315,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6267,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.6024633646011353,
"learning_rate": 5.726715068949564e-06,
"loss": 0.6067,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.5689429640769958,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6036,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.6224725842475891,
"learning_rate": 5.615533098453215e-06,
"loss": 0.6027,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.5541941523551941,
"learning_rate": 5.559822380516539e-06,
"loss": 0.5911,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.5280975699424744,
"learning_rate": 5.504041188505022e-06,
"loss": 0.6031,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.5941200852394104,
"learning_rate": 5.448196544517168e-06,
"loss": 0.6203,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.6535778045654297,
"learning_rate": 5.392295478639226e-06,
"loss": 0.6045,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.5731402635574341,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6232,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.6163918375968933,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6251,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.563120424747467,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6278,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.5305071473121643,
"learning_rate": 5.168267827938971e-06,
"loss": 0.5822,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.6606442332267761,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6428,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.57990562915802,
"learning_rate": 5.05609869177323e-06,
"loss": 0.5998,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.5651428699493408,
"learning_rate": 5e-06,
"loss": 0.6178,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.5303675532341003,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6121,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.5384981632232666,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6046,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.5476791858673096,
"learning_rate": 4.831732172061032e-06,
"loss": 0.586,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.5967368483543396,
"learning_rate": 4.775675848247427e-06,
"loss": 0.6192,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.5415930151939392,
"learning_rate": 4.719647763814041e-06,
"loss": 0.6222,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.553633451461792,
"learning_rate": 4.663654971939802e-06,
"loss": 0.5977,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.4556650221347809,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.6005,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.5191108584403992,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6217,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.5628607273101807,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.6219,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.5499104857444763,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.5996,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.5642635822296143,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6303,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.554854691028595,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6013,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.5956094861030579,
"learning_rate": 4.273284931050438e-06,
"loss": 0.6151,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5540267825126648,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6236,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.48829886317253113,
"learning_rate": 4.162468883476319e-06,
"loss": 0.5918,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.5904005765914917,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6255,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.5287757515907288,
"learning_rate": 4.052074558041608e-06,
"loss": 0.6255,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5756903886795044,
"learning_rate": 3.997052921083637e-06,
"loss": 0.5892,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.5119880437850952,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.5898,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.5788736939430237,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6169,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.6066918969154358,
"learning_rate": 3.832773180720475e-06,
"loss": 0.5975,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.47923052310943604,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6063,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.5256167054176331,
"learning_rate": 3.723976553681787e-06,
"loss": 0.5898,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.5218703150749207,
"learning_rate": 3.669815772166625e-06,
"loss": 0.6163,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.5827054381370544,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.6411,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.5353463888168335,
"learning_rate": 3.562003362839914e-06,
"loss": 0.5828,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.5629212856292725,
"learning_rate": 3.50836530715422e-06,
"loss": 0.606,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.5529426336288452,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6047,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.5595647692680359,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.6007,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.510316789150238,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.605,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.5082827806472778,
"learning_rate": 3.295758014387375e-06,
"loss": 0.6044,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.5071288347244263,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.5941,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.5056788325309753,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.6328,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.5091270208358765,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.628,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.5131690502166748,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.6153,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.5017776489257812,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6145,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.5688852667808533,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5952,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.5297237634658813,
"learning_rate": 2.932207475167398e-06,
"loss": 0.564,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.5823726654052734,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5464,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.5303489565849304,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5374,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.5290647745132446,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5458,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.6019632816314697,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5666,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.642267107963562,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5417,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.554168701171875,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5479,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.5158907771110535,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5463,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.5869964957237244,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5542,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.5414969325065613,
"learning_rate": 2.483822884971e-06,
"loss": 0.5804,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.6054401993751526,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5412,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.5310540795326233,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5464,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.46688178181648254,
"learning_rate": 2.339839617423318e-06,
"loss": 0.577,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.5174608826637268,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5485,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.5395613312721252,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5748,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.5993642807006836,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5491,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.5429391860961914,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5544,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.4957495331764221,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.53,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.5186243057250977,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5357,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.504743754863739,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5474,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.4906250834465027,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5213,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.49817684292793274,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5511,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.4939645528793335,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5645,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.46551719307899475,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5608,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.48018112778663635,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5662,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.44103774428367615,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5538,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.495027095079422,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.54,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.45933493971824646,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5363,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.5015433430671692,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.539,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.4601951539516449,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.551,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.4475500285625458,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5746,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.4685775637626648,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5339,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.45527276396751404,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5826,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.48752960562705994,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5389,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.46335822343826294,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5368,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.43882399797439575,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5513,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.4834885597229004,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.53,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.4527883529663086,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5352,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.4326450526714325,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5582,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.4567236602306366,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5473,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.4383062720298767,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5466,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.46677571535110474,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5546,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.4352693259716034,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5753,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.5085779428482056,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5585,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.4980652928352356,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5228,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.4570230543613434,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5279,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.4729381501674652,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5104,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.44161003828048706,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5309,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.44213470816612244,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5414,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.4700586795806885,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5303,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.46862950921058655,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5397,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.4479934275150299,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5549,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.4277068078517914,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5159,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.43517524003982544,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5507,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.43408796191215515,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5689,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.4458410143852234,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5417,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.454603374004364,
"learning_rate": 6.512943342215234e-07,
"loss": 0.548,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.42746439576148987,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5358,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.45722904801368713,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5342,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.4474881589412689,
"learning_rate": 5.707160663821009e-07,
"loss": 0.5414,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.44120532274246216,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5559,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.45479756593704224,
"learning_rate": 5.197795201563744e-07,
"loss": 0.5674,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.44058603048324585,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5455,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.4413107931613922,
"learning_rate": 4.710988998640298e-07,
"loss": 0.548,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.3991833031177521,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5328,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.41719958186149597,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.5584,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.5134970545768738,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5362,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.4417891502380371,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.5535,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.40042635798454285,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5137,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.42440158128738403,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.5437,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.4304046332836151,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5412,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.41541606187820435,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.5699,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.44152793288230896,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.548,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.43628525733947754,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.5541,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.4218094050884247,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5181,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.4643074870109558,
"learning_rate": 2.276812823220964e-07,
"loss": 0.5489,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 2.1665196418762207,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5795,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.48348286747932434,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5422,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.4040493667125702,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5576,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.4414437413215637,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5454,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.4265543222427368,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5349,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.39996179938316345,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5641,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.42793670296669006,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5449,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.463965505361557,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5581,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.47273877263069153,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5541,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.4276675581932068,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5481,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.4636850953102112,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5302,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.42180106043815613,
"learning_rate": 7.064490740882057e-08,
"loss": 0.547,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.3987959623336792,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5379,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.41314876079559326,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.555,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.4317784607410431,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5445,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.41900503635406494,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5458,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.4481179118156433,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5538,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.43400096893310547,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5618,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.440768837928772,
"learning_rate": 2.012853002380466e-08,
"loss": 0.543,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.4923563301563263,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5306,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.4089133143424988,
"learning_rate": 1.132562476771959e-08,
"loss": 0.55,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.4582107365131378,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5245,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.4360067844390869,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5658,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.44387513399124146,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5424,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.39601054787635803,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5515,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.4209800660610199,
"learning_rate": 3.147162264971471e-10,
"loss": 0.5456,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.4246397614479065,
"learning_rate": 0.0,
"loss": 0.5526,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 4.516070283014963e+17,
"train_loss": 0.0,
"train_runtime": 11.9218,
"train_samples_per_second": 2516.4,
"train_steps_per_second": 26.171
}
],
"logging_steps": 1,
"max_steps": 312,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.516070283014963e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}