DCFT-KodCode-V1-SFT-R1_10k-etash / trainer_state.json
neginr's picture
Upload folder using huggingface_hub
e04f06b 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": 7.44984028007181,
"learning_rate": 3.125e-07,
"loss": 1.1921,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 7.865277920559293,
"learning_rate": 6.25e-07,
"loss": 1.2325,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 7.4012919143382385,
"learning_rate": 9.375000000000001e-07,
"loss": 1.173,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 7.272440406613324,
"learning_rate": 1.25e-06,
"loss": 1.1684,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 7.533063861500159,
"learning_rate": 1.5625e-06,
"loss": 1.1835,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 6.757655793149794,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.1548,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 5.369015160279249,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.1197,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 5.225036916176714,
"learning_rate": 2.5e-06,
"loss": 1.1023,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 3.2917751973712868,
"learning_rate": 2.8125e-06,
"loss": 1.0457,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 3.306558507085019,
"learning_rate": 3.125e-06,
"loss": 1.0506,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.9985478877493352,
"learning_rate": 3.4375e-06,
"loss": 1.0267,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 4.2151456119871975,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.9883,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 4.6842766871834955,
"learning_rate": 4.0625000000000005e-06,
"loss": 1.0024,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 4.42164212867909,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.012,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 4.020436168662707,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.9894,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 3.0150401026784452,
"learning_rate": 5e-06,
"loss": 0.9185,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.7964597700858733,
"learning_rate": 5.3125e-06,
"loss": 0.9306,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.397064809639588,
"learning_rate": 5.625e-06,
"loss": 0.896,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.8752178678402174,
"learning_rate": 5.9375e-06,
"loss": 0.8776,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.9447771250823462,
"learning_rate": 6.25e-06,
"loss": 0.9055,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.823769549756628,
"learning_rate": 6.5625e-06,
"loss": 0.8557,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.3648797978532374,
"learning_rate": 6.875e-06,
"loss": 0.8408,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.4540540958662285,
"learning_rate": 7.1875e-06,
"loss": 0.8417,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.5387069205489259,
"learning_rate": 7.500000000000001e-06,
"loss": 0.8169,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.245603658810282,
"learning_rate": 7.8125e-06,
"loss": 0.7888,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.1863895363443724,
"learning_rate": 8.125000000000001e-06,
"loss": 0.8382,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.0796307948904094,
"learning_rate": 8.4375e-06,
"loss": 0.7996,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.1949084029840293,
"learning_rate": 8.750000000000001e-06,
"loss": 0.7754,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 1.0941131450020283,
"learning_rate": 9.0625e-06,
"loss": 0.8055,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 1.0626570184362158,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7941,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 1.0422732831528572,
"learning_rate": 9.6875e-06,
"loss": 0.7909,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 1.0333659686740362,
"learning_rate": 1e-05,
"loss": 0.8121,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.8273052105178655,
"learning_rate": 9.999685283773504e-06,
"loss": 0.7771,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.9787498179943497,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7856,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 1.0078623650822598,
"learning_rate": 9.997167791667668e-06,
"loss": 0.7796,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.8273496459125615,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7468,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.9797003164673342,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7981,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.8059888339121687,
"learning_rate": 9.98867437523228e-06,
"loss": 0.742,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.8107743556171683,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7749,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.9383585038575013,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7835,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.7618237855317876,
"learning_rate": 9.974529372743762e-06,
"loss": 0.7264,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.840790903370474,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7322,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.8501692517168098,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7436,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.6975416233345055,
"learning_rate": 9.954748808839675e-06,
"loss": 0.7334,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.7258190004965743,
"learning_rate": 9.946906630265184e-06,
"loss": 0.7492,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.7227894066019654,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7656,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.8368564448662366,
"learning_rate": 9.92935509259118e-06,
"loss": 0.7499,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.8000120382600069,
"learning_rate": 9.91964794299315e-06,
"loss": 0.6898,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.6478237133446474,
"learning_rate": 9.909321476180594e-06,
"loss": 0.7554,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.7993689460245798,
"learning_rate": 9.898376992116179e-06,
"loss": 0.7331,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.6740024540358694,
"learning_rate": 9.886815868562596e-06,
"loss": 0.7593,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.7875927414401914,
"learning_rate": 9.874639560909118e-06,
"loss": 0.7218,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.9204495532754047,
"learning_rate": 9.861849601988384e-06,
"loss": 0.7606,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.8515543697844061,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7508,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.8793539425883056,
"learning_rate": 9.834435247725032e-06,
"loss": 0.7446,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.7125855791521011,
"learning_rate": 9.819814303479268e-06,
"loss": 0.7198,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.8601414197039233,
"learning_rate": 9.804586609725499e-06,
"loss": 0.7584,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.8978963063652711,
"learning_rate": 9.788754083424654e-06,
"loss": 0.7581,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.7578688005644455,
"learning_rate": 9.772318717677905e-06,
"loss": 0.731,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.7915714328743959,
"learning_rate": 9.755282581475769e-06,
"loss": 0.7178,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.7113147632201263,
"learning_rate": 9.737647819437645e-06,
"loss": 0.6756,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.7897266858083429,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7194,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.7340558779757722,
"learning_rate": 9.700591372846096e-06,
"loss": 0.7238,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.7335400220915098,
"learning_rate": 9.681174353198687e-06,
"loss": 0.746,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.7853590853939908,
"learning_rate": 9.661168036940071e-06,
"loss": 0.7472,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.7994729327397367,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7383,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.7110696865363189,
"learning_rate": 9.619397662556434e-06,
"loss": 0.725,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.7149598613139251,
"learning_rate": 9.597638862757255e-06,
"loss": 0.7176,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.7290816732525277,
"learning_rate": 9.5753012823366e-06,
"loss": 0.6935,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.8542940855675434,
"learning_rate": 9.552387733294081e-06,
"loss": 0.7064,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.7108044776852782,
"learning_rate": 9.528901100135971e-06,
"loss": 0.7178,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.75180937438623,
"learning_rate": 9.504844339512096e-06,
"loss": 0.7083,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.6823258471944705,
"learning_rate": 9.480220479843627e-06,
"loss": 0.7378,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.7254584425824749,
"learning_rate": 9.45503262094184e-06,
"loss": 0.7337,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.754473840877677,
"learning_rate": 9.4292839336179e-06,
"loss": 0.7023,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.8598760005004826,
"learning_rate": 9.40297765928369e-06,
"loss": 0.709,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.7113125643994377,
"learning_rate": 9.376117109543769e-06,
"loss": 0.7131,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.885836860505212,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7125,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.7392713682575284,
"learning_rate": 9.320746778718274e-06,
"loss": 0.7348,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.6673731239845452,
"learning_rate": 9.292243968009332e-06,
"loss": 0.7132,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 1.0711559982941317,
"learning_rate": 9.263200821770462e-06,
"loss": 0.7072,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.6627998597983672,
"learning_rate": 9.233620996141421e-06,
"loss": 0.7116,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.7827123520204692,
"learning_rate": 9.203508214822652e-06,
"loss": 0.7127,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.7864623760995096,
"learning_rate": 9.172866268606514e-06,
"loss": 0.7231,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.7531910221475312,
"learning_rate": 9.141699014900084e-06,
"loss": 0.7031,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.6939852510853312,
"learning_rate": 9.110010377239552e-06,
"loss": 0.7083,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.740110423358011,
"learning_rate": 9.077804344796302e-06,
"loss": 0.736,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.6962717245943012,
"learning_rate": 9.045084971874738e-06,
"loss": 0.725,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.7282758181393519,
"learning_rate": 9.011856377401891e-06,
"loss": 0.6616,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.6591364614232718,
"learning_rate": 8.978122744408905e-06,
"loss": 0.7201,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.7687794176181812,
"learning_rate": 8.943888319504456e-06,
"loss": 0.7141,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.7025109214373477,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6786,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.7934438924739007,
"learning_rate": 8.873934395068006e-06,
"loss": 0.696,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.7286914028970188,
"learning_rate": 8.838223701790057e-06,
"loss": 0.6979,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.8523995640187266,
"learning_rate": 8.802029828000157e-06,
"loss": 0.7152,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.7405848382433704,
"learning_rate": 8.765357330018056e-06,
"loss": 0.7339,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.7570257564883277,
"learning_rate": 8.728210824415829e-06,
"loss": 0.7076,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.6508583056719521,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6953,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.6618913080055969,
"learning_rate": 8.652514554406388e-06,
"loss": 0.6811,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.8406583650567178,
"learning_rate": 8.613974319136959e-06,
"loss": 0.7228,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.6897544216899391,
"learning_rate": 8.574979133323378e-06,
"loss": 0.7336,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.7583095432170773,
"learning_rate": 8.535533905932739e-06,
"loss": 0.7027,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.7562540241623603,
"learning_rate": 8.495643602586287e-06,
"loss": 0.7045,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.649305216312411,
"learning_rate": 8.455313244934324e-06,
"loss": 0.73,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.7936824470510633,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6708,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.6141815005105694,
"learning_rate": 8.373352729660373e-06,
"loss": 0.6686,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.8417773076963088,
"learning_rate": 8.331732889760021e-06,
"loss": 0.6873,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.7923376427315969,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6464,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.760422564767531,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6917,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.8257382873134871,
"learning_rate": 8.204378069925121e-06,
"loss": 0.6415,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.6933773407473951,
"learning_rate": 8.16111251028955e-06,
"loss": 0.6574,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.8403906810689268,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6171,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.7491119510874427,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6599,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.7027348672649106,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6536,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.8641747947407908,
"learning_rate": 7.984126070912519e-06,
"loss": 0.6697,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.7591042759192601,
"learning_rate": 7.938926261462366e-06,
"loss": 0.657,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.6005010585667181,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6272,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.7173517518038414,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6083,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.80224449265041,
"learning_rate": 7.801129998764014e-06,
"loss": 0.661,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.6058908763716326,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6841,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.633164195810777,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6862,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.719953118777752,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6666,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.6983099150509602,
"learning_rate": 7.612492823579744e-06,
"loss": 0.6641,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.6891277146866401,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6742,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.6623484163517424,
"learning_rate": 7.516177115029002e-06,
"loss": 0.6251,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.6225047215183483,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6405,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.6605434081670686,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.6573,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.6235409038723179,
"learning_rate": 7.369343312364994e-06,
"loss": 0.661,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.6976078528660998,
"learning_rate": 7.319793920889171e-06,
"loss": 0.6165,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.7739825597786174,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6806,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.6425735978356044,
"learning_rate": 7.219825320152411e-06,
"loss": 0.6481,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.8665823860581879,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6952,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.7542562951285996,
"learning_rate": 7.118738970516944e-06,
"loss": 0.6297,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.6090926426794891,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6243,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.9185040416097427,
"learning_rate": 7.016585772004026e-06,
"loss": 0.6502,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.7628525553326614,
"learning_rate": 6.965125158269619e-06,
"loss": 0.634,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.6994978812780258,
"learning_rate": 6.913417161825449e-06,
"loss": 0.677,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.7539031255853901,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.6692,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.8076957468511414,
"learning_rate": 6.809285088483361e-06,
"loss": 0.6875,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.6727346394508487,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6532,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.7867357450146949,
"learning_rate": 6.704241985612625e-06,
"loss": 0.6282,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.7727798049035731,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6904,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.6287673093258204,
"learning_rate": 6.598340745578908e-06,
"loss": 0.6224,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.6345136167386021,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6254,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.7365144040502278,
"learning_rate": 6.491634692845781e-06,
"loss": 0.6198,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.6414360303854832,
"learning_rate": 6.437996637160086e-06,
"loss": 0.6616,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.6909904036898891,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6749,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.6242562261324572,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6249,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.569316835732891,
"learning_rate": 6.276023446318214e-06,
"loss": 0.675,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.6910501984303238,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6287,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.6084409092238928,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.6321,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.5892970964768208,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6642,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.6183641236892907,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6246,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.6855201051512687,
"learning_rate": 6.002947078916365e-06,
"loss": 0.672,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.6555434258531646,
"learning_rate": 5.947925441958393e-06,
"loss": 0.6675,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.6424648103064591,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6363,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.7028703454661197,
"learning_rate": 5.837531116523683e-06,
"loss": 0.6631,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.7469237946254669,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6315,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.5598595675062865,
"learning_rate": 5.726715068949564e-06,
"loss": 0.6485,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.7559714835816086,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6434,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.8079343224021268,
"learning_rate": 5.615533098453215e-06,
"loss": 0.6862,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.5664039055158315,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6827,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.6680902358411386,
"learning_rate": 5.504041188505022e-06,
"loss": 0.6428,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.7485310487169867,
"learning_rate": 5.448196544517168e-06,
"loss": 0.6204,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.6391370911110529,
"learning_rate": 5.392295478639226e-06,
"loss": 0.6448,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.5909960012560049,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6262,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.6071177374378659,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6602,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.6076003529371118,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6669,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.5874267580607758,
"learning_rate": 5.168267827938971e-06,
"loss": 0.6532,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.5756288630830325,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6298,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.693092296385674,
"learning_rate": 5.05609869177323e-06,
"loss": 0.6747,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.6030547849167742,
"learning_rate": 5e-06,
"loss": 0.6854,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.555374738880788,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6416,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.64154124016582,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6323,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.5940097258452632,
"learning_rate": 4.831732172061032e-06,
"loss": 0.6048,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.6361734381993863,
"learning_rate": 4.775675848247427e-06,
"loss": 0.6455,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.5307667489975789,
"learning_rate": 4.719647763814041e-06,
"loss": 0.615,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.5874701890326913,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6324,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.5881609560151907,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.6516,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.5646786407999248,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6252,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.5821953417333238,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.6757,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.5595924149249945,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6788,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.5886870438006275,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6396,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.625717167232356,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6189,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.527551327269441,
"learning_rate": 4.273284931050438e-06,
"loss": 0.6542,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.6835177896296061,
"learning_rate": 4.217827674798845e-06,
"loss": 0.645,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.5538221530384875,
"learning_rate": 4.162468883476319e-06,
"loss": 0.6316,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.565954892821739,
"learning_rate": 4.107215526006818e-06,
"loss": 0.664,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.5469891398130653,
"learning_rate": 4.052074558041608e-06,
"loss": 0.6742,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5478822613454761,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6337,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.5199639432837908,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.6148,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.6016156259123874,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6291,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.5953815799044481,
"learning_rate": 3.832773180720475e-06,
"loss": 0.6694,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.5504868262038806,
"learning_rate": 3.778297969310529e-06,
"loss": 0.642,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.6002534805634625,
"learning_rate": 3.723976553681787e-06,
"loss": 0.6242,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.6302837327348573,
"learning_rate": 3.669815772166625e-06,
"loss": 0.692,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.536552683367648,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.6507,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.5846052000495897,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6188,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.55086292257183,
"learning_rate": 3.50836530715422e-06,
"loss": 0.6247,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.5329170832298599,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6298,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.5265290419183813,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.6419,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.5167574376703518,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6608,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.5993454231731594,
"learning_rate": 3.295758014387375e-06,
"loss": 0.6339,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.6106478169414713,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6471,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.5471732117426183,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.6499,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.5326074120233216,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6604,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.5334694027203158,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.6302,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.5258328967338055,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6383,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.583575379731641,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5924,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.6334724517267839,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5834,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.5470104846749575,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5915,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.6916968809007217,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5918,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.5376671268932444,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.6028,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.5760504350202456,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.6175,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.5897233986516864,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5767,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.5015083503628716,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.6098,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.5931784979747177,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.6481,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.5743437527925073,
"learning_rate": 2.532458909678266e-06,
"loss": 0.6209,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.5183546803452326,
"learning_rate": 2.483822884971e-06,
"loss": 0.628,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.5039524125197884,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5573,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.5731238533515997,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5847,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.5072609427229656,
"learning_rate": 2.339839617423318e-06,
"loss": 0.6146,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.5160986054749295,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5693,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.5600848904688962,
"learning_rate": 2.245515092739488e-06,
"loss": 0.6508,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.5140594142075502,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.6122,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.5097718572359252,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.6145,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.5820544609319552,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5888,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.5054920129958911,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5991,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.5425705672401173,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5828,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.48937109556522007,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5937,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.5320254513879936,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5789,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.52295002817346,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.6155,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.5529251336318773,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5797,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.4700063553063849,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5951,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.46480448343390307,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.6167,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.5009979923754798,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5749,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.47706698702026823,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5752,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.5430264460486801,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.6095,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.5113530294573172,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.6118,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.4584779595320786,
"learning_rate": 1.544686755065677e-06,
"loss": 0.6374,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.4764932130659342,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.6034,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.5210278901348432,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.6356,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.4632087603938993,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5993,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.5132316523590685,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5792,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.4852832053425941,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.574,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.5252116643480881,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5939,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.4862240053746051,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.6142,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.47476446962456964,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5722,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.4643451858023064,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.6013,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.5090403257670568,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5736,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.5343013731891096,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.6341,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.5124202103318586,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5764,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.5001664535960665,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5923,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.4710253650026844,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5897,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.4695656441623169,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5675,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.45039137095213694,
"learning_rate": 9.549150281252633e-07,
"loss": 0.6086,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.4931944867908003,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5912,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.5112019849727703,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5932,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.501232515150562,
"learning_rate": 8.58300985099918e-07,
"loss": 0.6169,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.4776038944829597,
"learning_rate": 8.271337313934869e-07,
"loss": 0.6469,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.48736619407431087,
"learning_rate": 7.964917851773496e-07,
"loss": 0.6145,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.465398644065999,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5844,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.45296467477789154,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5763,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.49231006784034054,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5941,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.44881879154609666,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5895,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.46281821350929553,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5862,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.4890132883996681,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5577,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.47076830565670497,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5845,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.4459028576425142,
"learning_rate": 5.707160663821009e-07,
"loss": 0.6054,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.4545708994316169,
"learning_rate": 5.449673790581611e-07,
"loss": 0.6178,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.4607783672624146,
"learning_rate": 5.197795201563744e-07,
"loss": 0.5954,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.4663798066663291,
"learning_rate": 4.951556604879049e-07,
"loss": 0.6075,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.4836733591759614,
"learning_rate": 4.710988998640298e-07,
"loss": 0.5836,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.4382086381296098,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.6038,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.4610253988991969,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.6313,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.4330949510800422,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.6316,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.47102519708025403,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.572,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.4799417721877224,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5571,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.45879670001187994,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.6145,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.4350578308941833,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5641,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.43043028295865776,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.6158,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.455553680110645,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5973,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.4342219734244717,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.6175,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.4539427376241072,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5891,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.44737300766498533,
"learning_rate": 2.276812823220964e-07,
"loss": 0.6051,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.45349217341687514,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.6238,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.4332142034176807,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.6076,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.417905496727454,
"learning_rate": 1.801856965207338e-07,
"loss": 0.6148,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.440111747609688,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.6001,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.43149680493546944,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.6155,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.4829484321047708,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.6156,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.45094657343557,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5845,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.4362652177529693,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.6423,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.445117707632346,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.6439,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.43088707983720037,
"learning_rate": 9.0678523819408e-08,
"loss": 0.6205,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.4405008670029747,
"learning_rate": 8.035205700685167e-08,
"loss": 0.6198,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.44795332162569246,
"learning_rate": 7.064490740882057e-08,
"loss": 0.6161,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.4415848521804436,
"learning_rate": 6.15582970243117e-08,
"loss": 0.6267,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.4702963639169449,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.6028,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.4441322996088177,
"learning_rate": 4.52511911603265e-08,
"loss": 0.6208,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.47309111217468974,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5744,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.46779719091735655,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5772,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.47176088504664054,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5986,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.4477069928642587,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5938,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.4530767864386601,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5637,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.4563153629114904,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5932,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.4542550230147794,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5953,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.4736451425527256,
"learning_rate": 5.034667293427053e-09,
"loss": 0.6022,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.430854585941986,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5723,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.46149848243823854,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5711,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.4751338844609642,
"learning_rate": 3.147162264971471e-10,
"loss": 0.6233,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.4401166291606131,
"learning_rate": 0.0,
"loss": 0.5777,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 2.984816118121431e+17,
"train_loss": 0.0,
"train_runtime": 14.4713,
"train_samples_per_second": 2073.074,
"train_steps_per_second": 21.56
}
],
"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": 2.984816118121431e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}