DCFT-seed_code_r1_magicoder-etash / trainer_state.json
EtashGuha's picture
Upload folder using huggingface_hub
e84f18e 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": 8.619830131530762,
"learning_rate": 3.125e-07,
"loss": 1.1477,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 8.286690711975098,
"learning_rate": 6.25e-07,
"loss": 1.148,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 8.450620651245117,
"learning_rate": 9.375000000000001e-07,
"loss": 1.138,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 8.70119571685791,
"learning_rate": 1.25e-06,
"loss": 1.162,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 7.683633804321289,
"learning_rate": 1.5625e-06,
"loss": 1.0968,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 7.421406269073486,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.0936,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 5.629795551300049,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.0751,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 5.259838581085205,
"learning_rate": 2.5e-06,
"loss": 1.0505,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 3.384989023208618,
"learning_rate": 2.8125e-06,
"loss": 1.0067,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 3.083430528640747,
"learning_rate": 3.125e-06,
"loss": 0.9975,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 3.0232350826263428,
"learning_rate": 3.4375e-06,
"loss": 0.9979,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.6285958290100098,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.948,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.827526807785034,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.9445,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.6110482215881348,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9462,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 2.8511009216308594,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.9123,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.0414063930511475,
"learning_rate": 5e-06,
"loss": 0.8783,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.499544620513916,
"learning_rate": 5.3125e-06,
"loss": 0.8782,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.2455239295959473,
"learning_rate": 5.625e-06,
"loss": 0.8458,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.6272025108337402,
"learning_rate": 5.9375e-06,
"loss": 0.8261,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.4080849885940552,
"learning_rate": 6.25e-06,
"loss": 0.8295,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.5042369365692139,
"learning_rate": 6.5625e-06,
"loss": 0.8013,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.4783216714859009,
"learning_rate": 6.875e-06,
"loss": 0.7976,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.564474105834961,
"learning_rate": 7.1875e-06,
"loss": 0.796,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.3670371770858765,
"learning_rate": 7.500000000000001e-06,
"loss": 0.7872,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.2051235437393188,
"learning_rate": 7.8125e-06,
"loss": 0.7655,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.0404964685440063,
"learning_rate": 8.125000000000001e-06,
"loss": 0.7502,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.0301539897918701,
"learning_rate": 8.4375e-06,
"loss": 0.7479,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 0.98060142993927,
"learning_rate": 8.750000000000001e-06,
"loss": 0.743,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 0.9372722506523132,
"learning_rate": 9.0625e-06,
"loss": 0.7447,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 0.9625725150108337,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7324,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.9534512758255005,
"learning_rate": 9.6875e-06,
"loss": 0.732,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.9993299245834351,
"learning_rate": 1e-05,
"loss": 0.728,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.8795299530029297,
"learning_rate": 9.999685283773504e-06,
"loss": 0.7388,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.9067621827125549,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7295,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.7828402519226074,
"learning_rate": 9.997167791667668e-06,
"loss": 0.7061,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.7652558088302612,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7133,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.7585594058036804,
"learning_rate": 9.992134075089085e-06,
"loss": 0.721,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.8700973391532898,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7143,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.8631901144981384,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7068,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.8145694732666016,
"learning_rate": 9.979871469976197e-06,
"loss": 0.6981,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.7626639604568481,
"learning_rate": 9.974529372743762e-06,
"loss": 0.6896,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.7362940907478333,
"learning_rate": 9.968561049466214e-06,
"loss": 0.6919,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.7744758129119873,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7049,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.7759642601013184,
"learning_rate": 9.954748808839675e-06,
"loss": 0.6905,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.8709149956703186,
"learning_rate": 9.946906630265184e-06,
"loss": 0.701,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.8436416387557983,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7015,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.6881698966026306,
"learning_rate": 9.92935509259118e-06,
"loss": 0.688,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.8486517071723938,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7034,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.819234311580658,
"learning_rate": 9.909321476180594e-06,
"loss": 0.6808,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.7803587913513184,
"learning_rate": 9.898376992116179e-06,
"loss": 0.6665,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.7330102920532227,
"learning_rate": 9.886815868562596e-06,
"loss": 0.6927,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.6954732537269592,
"learning_rate": 9.874639560909118e-06,
"loss": 0.673,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.787542462348938,
"learning_rate": 9.861849601988384e-06,
"loss": 0.6833,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.8558406829833984,
"learning_rate": 9.848447601883436e-06,
"loss": 0.6806,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.6984784007072449,
"learning_rate": 9.834435247725032e-06,
"loss": 0.6717,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.8145155310630798,
"learning_rate": 9.819814303479268e-06,
"loss": 0.6466,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.9503493905067444,
"learning_rate": 9.804586609725499e-06,
"loss": 0.6732,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.7588793039321899,
"learning_rate": 9.788754083424654e-06,
"loss": 0.6902,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.8395458459854126,
"learning_rate": 9.772318717677905e-06,
"loss": 0.6849,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.8674512505531311,
"learning_rate": 9.755282581475769e-06,
"loss": 0.6728,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.8832570910453796,
"learning_rate": 9.737647819437645e-06,
"loss": 0.6884,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.9429735541343689,
"learning_rate": 9.719416651541839e-06,
"loss": 0.6719,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 11.26850700378418,
"learning_rate": 9.700591372846096e-06,
"loss": 0.6742,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 1.137119174003601,
"learning_rate": 9.681174353198687e-06,
"loss": 0.6649,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.9743894338607788,
"learning_rate": 9.661168036940071e-06,
"loss": 0.6831,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.9408482313156128,
"learning_rate": 9.640574942595195e-06,
"loss": 0.6618,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 1.032985806465149,
"learning_rate": 9.619397662556434e-06,
"loss": 0.6768,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.9230046272277832,
"learning_rate": 9.597638862757255e-06,
"loss": 0.6723,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.994785726070404,
"learning_rate": 9.5753012823366e-06,
"loss": 0.6855,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.930493950843811,
"learning_rate": 9.552387733294081e-06,
"loss": 0.6805,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.9931144714355469,
"learning_rate": 9.528901100135971e-06,
"loss": 0.6567,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.9111120700836182,
"learning_rate": 9.504844339512096e-06,
"loss": 0.6635,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.8690065741539001,
"learning_rate": 9.480220479843627e-06,
"loss": 0.6685,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.7638235688209534,
"learning_rate": 9.45503262094184e-06,
"loss": 0.6745,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.8946619629859924,
"learning_rate": 9.4292839336179e-06,
"loss": 0.6699,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.9409710168838501,
"learning_rate": 9.40297765928369e-06,
"loss": 0.6596,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.7191086411476135,
"learning_rate": 9.376117109543769e-06,
"loss": 0.6467,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.8330948948860168,
"learning_rate": 9.348705665778479e-06,
"loss": 0.6571,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.8344268798828125,
"learning_rate": 9.320746778718274e-06,
"loss": 0.6495,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.7885737419128418,
"learning_rate": 9.292243968009332e-06,
"loss": 0.6706,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.7269691824913025,
"learning_rate": 9.263200821770462e-06,
"loss": 0.652,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.7536022067070007,
"learning_rate": 9.233620996141421e-06,
"loss": 0.6652,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.7258433699607849,
"learning_rate": 9.203508214822652e-06,
"loss": 0.6662,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.7775776982307434,
"learning_rate": 9.172866268606514e-06,
"loss": 0.6607,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.6933174729347229,
"learning_rate": 9.141699014900084e-06,
"loss": 0.6513,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.6570079326629639,
"learning_rate": 9.110010377239552e-06,
"loss": 0.6605,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.637382984161377,
"learning_rate": 9.077804344796302e-06,
"loss": 0.6556,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.7673482298851013,
"learning_rate": 9.045084971874738e-06,
"loss": 0.6701,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.7024659514427185,
"learning_rate": 9.011856377401891e-06,
"loss": 0.6668,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.7080751061439514,
"learning_rate": 8.978122744408905e-06,
"loss": 0.6723,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.7323607802391052,
"learning_rate": 8.943888319504456e-06,
"loss": 0.6742,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.8182899951934814,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6419,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.731065034866333,
"learning_rate": 8.873934395068006e-06,
"loss": 0.6638,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.7267597317695618,
"learning_rate": 8.838223701790057e-06,
"loss": 0.6603,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.7526478171348572,
"learning_rate": 8.802029828000157e-06,
"loss": 0.6572,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.642124354839325,
"learning_rate": 8.765357330018056e-06,
"loss": 0.6601,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.9729664921760559,
"learning_rate": 8.728210824415829e-06,
"loss": 0.6461,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.7368820309638977,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6726,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.7143698334693909,
"learning_rate": 8.652514554406388e-06,
"loss": 0.6621,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.7394453287124634,
"learning_rate": 8.613974319136959e-06,
"loss": 0.646,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.737785816192627,
"learning_rate": 8.574979133323378e-06,
"loss": 0.6535,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.7278648018836975,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6436,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.6332582235336304,
"learning_rate": 8.495643602586287e-06,
"loss": 0.6524,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.6592235565185547,
"learning_rate": 8.455313244934324e-06,
"loss": 0.6488,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.7093728184700012,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6052,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.7073942422866821,
"learning_rate": 8.373352729660373e-06,
"loss": 0.607,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.6666473150253296,
"learning_rate": 8.331732889760021e-06,
"loss": 0.594,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.695756733417511,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6008,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.7600190043449402,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6002,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.6950688362121582,
"learning_rate": 8.204378069925121e-06,
"loss": 0.586,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.7344626784324646,
"learning_rate": 8.16111251028955e-06,
"loss": 0.5892,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.8006746768951416,
"learning_rate": 8.117449009293668e-06,
"loss": 0.5845,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.6858304142951965,
"learning_rate": 8.073393063582386e-06,
"loss": 0.5934,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.7899407744407654,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6039,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.6752017736434937,
"learning_rate": 7.984126070912519e-06,
"loss": 0.5839,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.7475123405456543,
"learning_rate": 7.938926261462366e-06,
"loss": 0.5879,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.7280110716819763,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6254,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.7622445821762085,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6012,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.8279544115066528,
"learning_rate": 7.801129998764014e-06,
"loss": 0.5838,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.6798257827758789,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6083,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.7504909634590149,
"learning_rate": 7.70749306331863e-06,
"loss": 0.5772,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.7014206051826477,
"learning_rate": 7.660160382576683e-06,
"loss": 0.5882,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.7289438247680664,
"learning_rate": 7.612492823579744e-06,
"loss": 0.5896,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.795145571231842,
"learning_rate": 7.564496387029532e-06,
"loss": 0.5753,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.6939886808395386,
"learning_rate": 7.516177115029002e-06,
"loss": 0.5923,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.8253046870231628,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6032,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.8045299053192139,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.5981,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.7394763827323914,
"learning_rate": 7.369343312364994e-06,
"loss": 0.5985,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.7608659863471985,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5823,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.6254313588142395,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6009,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.6663405895233154,
"learning_rate": 7.219825320152411e-06,
"loss": 0.5934,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.636886477470398,
"learning_rate": 7.169418695587791e-06,
"loss": 0.5955,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.6905349493026733,
"learning_rate": 7.118738970516944e-06,
"loss": 0.5962,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.6985878348350525,
"learning_rate": 7.067792524832604e-06,
"loss": 0.5945,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.6564013957977295,
"learning_rate": 7.016585772004026e-06,
"loss": 0.5812,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.5902811288833618,
"learning_rate": 6.965125158269619e-06,
"loss": 0.5749,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.6935226917266846,
"learning_rate": 6.913417161825449e-06,
"loss": 0.5686,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.7379180192947388,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.5862,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.6640008687973022,
"learning_rate": 6.809285088483361e-06,
"loss": 0.5785,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.639917254447937,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.5869,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.6701115965843201,
"learning_rate": 6.704241985612625e-06,
"loss": 0.5753,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.6772593259811401,
"learning_rate": 6.651395309775837e-06,
"loss": 0.5777,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.608108401298523,
"learning_rate": 6.598340745578908e-06,
"loss": 0.5859,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.7229862809181213,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6077,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.6466304659843445,
"learning_rate": 6.491634692845781e-06,
"loss": 0.5882,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.646152138710022,
"learning_rate": 6.437996637160086e-06,
"loss": 0.5976,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.6978784799575806,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6027,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.6528997421264648,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6017,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.7351871132850647,
"learning_rate": 6.276023446318214e-06,
"loss": 0.5759,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.7768481969833374,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.5971,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.6240535974502563,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.5921,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.6948301196098328,
"learning_rate": 6.112604669781572e-06,
"loss": 0.5819,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.6924031972885132,
"learning_rate": 6.057842458386315e-06,
"loss": 0.57,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.6463066935539246,
"learning_rate": 6.002947078916365e-06,
"loss": 0.5954,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.6304520964622498,
"learning_rate": 5.947925441958393e-06,
"loss": 0.5811,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.7478348016738892,
"learning_rate": 5.892784473993184e-06,
"loss": 0.5818,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.6953530311584473,
"learning_rate": 5.837531116523683e-06,
"loss": 0.5733,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.6256540417671204,
"learning_rate": 5.782172325201155e-06,
"loss": 0.5946,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.6258138418197632,
"learning_rate": 5.726715068949564e-06,
"loss": 0.5628,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.6524773836135864,
"learning_rate": 5.671166329088278e-06,
"loss": 0.586,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.597075343132019,
"learning_rate": 5.615533098453215e-06,
"loss": 0.5902,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.607329249382019,
"learning_rate": 5.559822380516539e-06,
"loss": 0.583,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.7235519289970398,
"learning_rate": 5.504041188505022e-06,
"loss": 0.5833,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.627850353717804,
"learning_rate": 5.448196544517168e-06,
"loss": 0.5651,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.6454402208328247,
"learning_rate": 5.392295478639226e-06,
"loss": 0.602,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.7224230170249939,
"learning_rate": 5.336345028060199e-06,
"loss": 0.584,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.6418601870536804,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.5857,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.6513985991477966,
"learning_rate": 5.224324151752575e-06,
"loss": 0.5793,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.677497923374176,
"learning_rate": 5.168267827938971e-06,
"loss": 0.5762,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.6658104658126831,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6099,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.8325548768043518,
"learning_rate": 5.05609869177323e-06,
"loss": 0.5886,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.6375129818916321,
"learning_rate": 5e-06,
"loss": 0.5776,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.7483816146850586,
"learning_rate": 4.943901308226771e-06,
"loss": 0.5722,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.638129472732544,
"learning_rate": 4.887809678520976e-06,
"loss": 0.5503,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.5901349782943726,
"learning_rate": 4.831732172061032e-06,
"loss": 0.577,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.6541542410850525,
"learning_rate": 4.775675848247427e-06,
"loss": 0.5782,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.637847900390625,
"learning_rate": 4.719647763814041e-06,
"loss": 0.5887,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.6435571908950806,
"learning_rate": 4.663654971939802e-06,
"loss": 0.5896,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.6529213190078735,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.5636,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.6160262823104858,
"learning_rate": 4.551803455482833e-06,
"loss": 0.5713,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.6255659461021423,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.5779,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.6023534536361694,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.5859,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.6693056225776672,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6084,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.5382462739944458,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.5894,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.6022966504096985,
"learning_rate": 4.273284931050438e-06,
"loss": 0.564,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.6305966377258301,
"learning_rate": 4.217827674798845e-06,
"loss": 0.5964,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.6237593293190002,
"learning_rate": 4.162468883476319e-06,
"loss": 0.595,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.6263700127601624,
"learning_rate": 4.107215526006818e-06,
"loss": 0.575,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.7057296633720398,
"learning_rate": 4.052074558041608e-06,
"loss": 0.5907,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5898810029029846,
"learning_rate": 3.997052921083637e-06,
"loss": 0.5737,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.6338618397712708,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.5817,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.6271618604660034,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6014,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.6142144203186035,
"learning_rate": 3.832773180720475e-06,
"loss": 0.6027,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.5756183862686157,
"learning_rate": 3.778297969310529e-06,
"loss": 0.5751,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.5941733121871948,
"learning_rate": 3.723976553681787e-06,
"loss": 0.5884,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.5747978687286377,
"learning_rate": 3.669815772166625e-06,
"loss": 0.5731,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.6479167342185974,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.5721,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.6165737509727478,
"learning_rate": 3.562003362839914e-06,
"loss": 0.5747,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.5749819874763489,
"learning_rate": 3.50836530715422e-06,
"loss": 0.5709,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.5728811621665955,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.585,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.5578553676605225,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.5985,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.6435189843177795,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.5857,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.5662099123001099,
"learning_rate": 3.295758014387375e-06,
"loss": 0.594,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.5592925548553467,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.5873,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.6612251996994019,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.5842,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.6291359066963196,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.5938,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.5411191582679749,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.5945,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.5789668560028076,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.5675,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.5404688715934753,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5702,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.6695454120635986,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5605,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.5605763792991638,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5316,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.5510132908821106,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5256,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.5348244309425354,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5512,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.5787271857261658,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5618,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.5774601101875305,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5243,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.5650174021720886,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5161,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.5298798084259033,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5334,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.5396589040756226,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5474,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.5267285108566284,
"learning_rate": 2.483822884971e-06,
"loss": 0.5363,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.5327050685882568,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5206,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.5457900166511536,
"learning_rate": 2.387507176420256e-06,
"loss": 0.517,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.5315083861351013,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5257,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.5180266499519348,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5379,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.5602002143859863,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5242,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.5138322114944458,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5238,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.5235224366188049,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5394,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.5572905540466309,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5148,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.5326873660087585,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5309,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.5680348873138428,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.513,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.5308465957641602,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5196,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.531053900718689,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5229,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.5372322797775269,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5187,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.5155448317527771,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5327,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.5087056159973145,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5141,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.5122584700584412,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5194,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.5604382753372192,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5301,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.49856239557266235,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5453,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.48079735040664673,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5253,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.5152550339698792,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5113,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.49119043350219727,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5356,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.4969671964645386,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5446,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.5082017183303833,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5028,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.5095709562301636,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5407,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.4851570725440979,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5185,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.47453898191452026,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5296,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.5277453660964966,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.513,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.4782722592353821,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5083,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.4986642301082611,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5169,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.4988004267215729,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.547,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.5005159974098206,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5273,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.525696337223053,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5369,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.4685957729816437,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5443,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.5020018219947815,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5473,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.46969881653785706,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5156,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.4854971766471863,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5305,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.4787842631340027,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5253,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.5075872540473938,
"learning_rate": 9.221956552036992e-07,
"loss": 0.4995,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.477522611618042,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5227,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.47175735235214233,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5203,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.5393279194831848,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5098,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.4977039694786072,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5177,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.4892188310623169,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5432,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.5188687443733215,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5325,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.48087698221206665,
"learning_rate": 7.077560319906696e-07,
"loss": 0.52,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.5087722539901733,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5339,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.49504950642585754,
"learning_rate": 6.512943342215234e-07,
"loss": 0.534,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.4981255829334259,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5078,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.4903486371040344,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5381,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.498220294713974,
"learning_rate": 5.707160663821009e-07,
"loss": 0.4945,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.49345526099205017,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5375,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.4738101363182068,
"learning_rate": 5.197795201563744e-07,
"loss": 0.5153,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.4820142984390259,
"learning_rate": 4.951556604879049e-07,
"loss": 0.4957,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.48794591426849365,
"learning_rate": 4.710988998640298e-07,
"loss": 0.5214,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.4865429699420929,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5302,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.5159239172935486,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.5252,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.4868268072605133,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5287,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.47113215923309326,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.5327,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.44318968057632446,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5453,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.47545257210731506,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.5068,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.4889649450778961,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5247,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.4773139953613281,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.4949,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.47189274430274963,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5169,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.4697064757347107,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.5396,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.4627130925655365,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5407,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.4732663333415985,
"learning_rate": 2.276812823220964e-07,
"loss": 0.4842,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.4764968454837799,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5369,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.4686462879180908,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.504,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.4672069847583771,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5387,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.4739827811717987,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5386,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.4638729989528656,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5404,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.46186888217926025,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5286,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.44833967089653015,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5396,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.46380218863487244,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5427,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.4559478163719177,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5081,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.46556219458580017,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5263,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.4788520932197571,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5331,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.46646323800086975,
"learning_rate": 7.064490740882057e-08,
"loss": 0.5542,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.4761679172515869,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5254,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.4787537157535553,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5211,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.4544961154460907,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5333,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.4399887025356293,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5322,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.45263734459877014,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5505,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.4458680748939514,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5342,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.4879269599914551,
"learning_rate": 2.012853002380466e-08,
"loss": 0.519,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.4802978038787842,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5134,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.4840865731239319,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5167,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.4477526843547821,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5369,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.5075541138648987,
"learning_rate": 5.034667293427053e-09,
"loss": 0.516,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.43588176369667053,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5192,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.44260454177856445,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5331,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.45163494348526,
"learning_rate": 3.147162264971471e-10,
"loss": 0.5255,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.4906923770904541,
"learning_rate": 0.0,
"loss": 0.5394,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 3.000164095335137e+17,
"train_loss": 0.6203194147883317,
"train_runtime": 8286.036,
"train_samples_per_second": 3.621,
"train_steps_per_second": 0.038
}
],
"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": 3.000164095335137e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}