EtashGuha's picture
Upload folder using huggingface_hub
f67a907 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": 10.1846284866333,
"learning_rate": 3.125e-07,
"loss": 1.2855,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 11.885759353637695,
"learning_rate": 6.25e-07,
"loss": 1.3909,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 9.910966873168945,
"learning_rate": 9.375000000000001e-07,
"loss": 1.2643,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 9.593328475952148,
"learning_rate": 1.25e-06,
"loss": 1.2583,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 9.58667278289795,
"learning_rate": 1.5625e-06,
"loss": 1.2324,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 9.016371726989746,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.2277,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 7.2298994064331055,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.1731,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 6.3173675537109375,
"learning_rate": 2.5e-06,
"loss": 1.1552,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 3.8943889141082764,
"learning_rate": 2.8125e-06,
"loss": 1.0752,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 4.622074127197266,
"learning_rate": 3.125e-06,
"loss": 1.1635,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 3.461489677429199,
"learning_rate": 3.4375e-06,
"loss": 1.0916,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.6233980655670166,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.0307,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.684861660003662,
"learning_rate": 4.0625000000000005e-06,
"loss": 1.0262,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.1212050914764404,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9923,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 2.8368988037109375,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.9784,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.5900933742523193,
"learning_rate": 5e-06,
"loss": 0.965,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.7696807384490967,
"learning_rate": 5.3125e-06,
"loss": 0.9421,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.772247791290283,
"learning_rate": 5.625e-06,
"loss": 0.9252,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 2.127742052078247,
"learning_rate": 5.9375e-06,
"loss": 0.8913,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 2.2174060344696045,
"learning_rate": 6.25e-06,
"loss": 0.9142,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.467130184173584,
"learning_rate": 6.5625e-06,
"loss": 0.8945,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.8869190216064453,
"learning_rate": 6.875e-06,
"loss": 0.8571,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.8495758771896362,
"learning_rate": 7.1875e-06,
"loss": 0.8635,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.550546407699585,
"learning_rate": 7.500000000000001e-06,
"loss": 0.8469,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.2359421253204346,
"learning_rate": 7.8125e-06,
"loss": 0.858,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.334065556526184,
"learning_rate": 8.125000000000001e-06,
"loss": 0.8227,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.1901493072509766,
"learning_rate": 8.4375e-06,
"loss": 0.8211,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.0726540088653564,
"learning_rate": 8.750000000000001e-06,
"loss": 0.8087,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 0.9990488290786743,
"learning_rate": 9.0625e-06,
"loss": 0.8213,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 1.1288069486618042,
"learning_rate": 9.375000000000001e-06,
"loss": 0.8121,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 1.0612502098083496,
"learning_rate": 9.6875e-06,
"loss": 0.7779,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 3.408714532852173,
"learning_rate": 1e-05,
"loss": 0.8001,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 1.1763397455215454,
"learning_rate": 9.999685283773504e-06,
"loss": 0.7777,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 1.1109788417816162,
"learning_rate": 9.998741174712534e-06,
"loss": 0.8028,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.9961385726928711,
"learning_rate": 9.997167791667668e-06,
"loss": 0.7922,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 1.1096503734588623,
"learning_rate": 9.994965332706574e-06,
"loss": 0.775,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 1.0257437229156494,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7878,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.9336013197898865,
"learning_rate": 9.98867437523228e-06,
"loss": 0.8034,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 1.7581638097763062,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7604,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 1.1630841493606567,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7719,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 1.2277592420578003,
"learning_rate": 9.974529372743762e-06,
"loss": 0.7509,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 1.0433387756347656,
"learning_rate": 9.968561049466214e-06,
"loss": 0.774,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.9740708470344543,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7583,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 1.0013972520828247,
"learning_rate": 9.954748808839675e-06,
"loss": 0.7475,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 1.0041651725769043,
"learning_rate": 9.946906630265184e-06,
"loss": 0.7276,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 1.0584808588027954,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7818,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.8802754878997803,
"learning_rate": 9.92935509259118e-06,
"loss": 0.7489,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 1.0446633100509644,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7751,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.9814447164535522,
"learning_rate": 9.909321476180594e-06,
"loss": 0.7611,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.9204258322715759,
"learning_rate": 9.898376992116179e-06,
"loss": 0.7697,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 1.0312919616699219,
"learning_rate": 9.886815868562596e-06,
"loss": 0.7501,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.9704650044441223,
"learning_rate": 9.874639560909118e-06,
"loss": 0.7584,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.9472712278366089,
"learning_rate": 9.861849601988384e-06,
"loss": 0.7651,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 1.081247329711914,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7754,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.9737085103988647,
"learning_rate": 9.834435247725032e-06,
"loss": 0.7737,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.8868014216423035,
"learning_rate": 9.819814303479268e-06,
"loss": 0.7656,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.9546142816543579,
"learning_rate": 9.804586609725499e-06,
"loss": 0.7717,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 1.0966267585754395,
"learning_rate": 9.788754083424654e-06,
"loss": 0.7579,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 1.167971134185791,
"learning_rate": 9.772318717677905e-06,
"loss": 0.766,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.8990470767021179,
"learning_rate": 9.755282581475769e-06,
"loss": 0.7289,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 1.0461461544036865,
"learning_rate": 9.737647819437645e-06,
"loss": 0.7365,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.9309303760528564,
"learning_rate": 9.719416651541839e-06,
"loss": 0.755,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 1.0776662826538086,
"learning_rate": 9.700591372846096e-06,
"loss": 0.7286,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.9998989105224609,
"learning_rate": 9.681174353198687e-06,
"loss": 0.7302,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 1.0955569744110107,
"learning_rate": 9.661168036940071e-06,
"loss": 0.7277,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 1.2573649883270264,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7484,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.9250706434249878,
"learning_rate": 9.619397662556434e-06,
"loss": 0.7356,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 1.050192952156067,
"learning_rate": 9.597638862757255e-06,
"loss": 0.7652,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.9782637357711792,
"learning_rate": 9.5753012823366e-06,
"loss": 0.7355,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 1.0369902849197388,
"learning_rate": 9.552387733294081e-06,
"loss": 0.7262,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.9345138072967529,
"learning_rate": 9.528901100135971e-06,
"loss": 0.755,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.9086159467697144,
"learning_rate": 9.504844339512096e-06,
"loss": 0.7174,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.9462155699729919,
"learning_rate": 9.480220479843627e-06,
"loss": 0.7354,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.8859170079231262,
"learning_rate": 9.45503262094184e-06,
"loss": 0.7489,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.9321438670158386,
"learning_rate": 9.4292839336179e-06,
"loss": 0.7295,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.9737212061882019,
"learning_rate": 9.40297765928369e-06,
"loss": 0.7583,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.9324398040771484,
"learning_rate": 9.376117109543769e-06,
"loss": 0.7403,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.9065219163894653,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7243,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 1.0223456621170044,
"learning_rate": 9.320746778718274e-06,
"loss": 0.7171,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.9862157106399536,
"learning_rate": 9.292243968009332e-06,
"loss": 0.7044,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.9068260192871094,
"learning_rate": 9.263200821770462e-06,
"loss": 0.7193,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.8513049483299255,
"learning_rate": 9.233620996141421e-06,
"loss": 0.735,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 1.0724048614501953,
"learning_rate": 9.203508214822652e-06,
"loss": 0.7389,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.8821753263473511,
"learning_rate": 9.172866268606514e-06,
"loss": 0.704,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.9746748805046082,
"learning_rate": 9.141699014900084e-06,
"loss": 0.7029,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.8271751999855042,
"learning_rate": 9.110010377239552e-06,
"loss": 0.7254,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.886039674282074,
"learning_rate": 9.077804344796302e-06,
"loss": 0.7394,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.890937328338623,
"learning_rate": 9.045084971874738e-06,
"loss": 0.7283,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.7532746195793152,
"learning_rate": 9.011856377401891e-06,
"loss": 0.713,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.9271005392074585,
"learning_rate": 8.978122744408905e-06,
"loss": 0.7109,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.9248338937759399,
"learning_rate": 8.943888319504456e-06,
"loss": 0.733,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.7954618334770203,
"learning_rate": 8.90915741234015e-06,
"loss": 0.7418,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.902425229549408,
"learning_rate": 8.873934395068006e-06,
"loss": 0.7202,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.9635447859764099,
"learning_rate": 8.838223701790057e-06,
"loss": 0.7049,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 1.0547822713851929,
"learning_rate": 8.802029828000157e-06,
"loss": 0.7171,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 1.0897992849349976,
"learning_rate": 8.765357330018056e-06,
"loss": 0.7222,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.8879277110099792,
"learning_rate": 8.728210824415829e-06,
"loss": 0.7488,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 1.0495959520339966,
"learning_rate": 8.690594987436705e-06,
"loss": 0.7237,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.963057279586792,
"learning_rate": 8.652514554406388e-06,
"loss": 0.7622,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.9057732224464417,
"learning_rate": 8.613974319136959e-06,
"loss": 0.7147,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.8921683430671692,
"learning_rate": 8.574979133323378e-06,
"loss": 0.7139,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.8844946026802063,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6896,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.8735024929046631,
"learning_rate": 8.495643602586287e-06,
"loss": 0.7148,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 1.0184752941131592,
"learning_rate": 8.455313244934324e-06,
"loss": 0.7251,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.8756181597709656,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6595,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.9014113545417786,
"learning_rate": 8.373352729660373e-06,
"loss": 0.644,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.9087943434715271,
"learning_rate": 8.331732889760021e-06,
"loss": 0.6423,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.8448387384414673,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6398,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.8832601308822632,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6514,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.942776620388031,
"learning_rate": 8.204378069925121e-06,
"loss": 0.6533,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.9780378937721252,
"learning_rate": 8.16111251028955e-06,
"loss": 0.62,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.957817792892456,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6431,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.8933224081993103,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6474,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.9795970916748047,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6291,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.9787712693214417,
"learning_rate": 7.984126070912519e-06,
"loss": 0.6173,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.9803339838981628,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6465,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.8933205008506775,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6348,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.8859388828277588,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6162,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.8237574696540833,
"learning_rate": 7.801129998764014e-06,
"loss": 0.6014,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.8303595781326294,
"learning_rate": 7.754484907260513e-06,
"loss": 0.637,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.805557131767273,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6475,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.8250300884246826,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6325,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.927651047706604,
"learning_rate": 7.612492823579744e-06,
"loss": 0.6245,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.7403294444084167,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6222,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.7962580323219299,
"learning_rate": 7.516177115029002e-06,
"loss": 0.6057,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.8116708993911743,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6161,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.8612444996833801,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.63,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.853906512260437,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6701,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 1.1103078126907349,
"learning_rate": 7.319793920889171e-06,
"loss": 0.6482,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.840362548828125,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6338,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.9150324463844299,
"learning_rate": 7.219825320152411e-06,
"loss": 0.5999,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.999043881893158,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6472,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.8423965573310852,
"learning_rate": 7.118738970516944e-06,
"loss": 0.6029,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.8174357414245605,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6156,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.9509724378585815,
"learning_rate": 7.016585772004026e-06,
"loss": 0.6375,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.7809696197509766,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6353,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.8257637023925781,
"learning_rate": 6.913417161825449e-06,
"loss": 0.6233,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.9169142246246338,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.6179,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.8016465306282043,
"learning_rate": 6.809285088483361e-06,
"loss": 0.6299,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.8083800077438354,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6663,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.8169625401496887,
"learning_rate": 6.704241985612625e-06,
"loss": 0.6005,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.8032822608947754,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6148,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.8095478415489197,
"learning_rate": 6.598340745578908e-06,
"loss": 0.6342,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.8488128781318665,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6248,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.7461427450180054,
"learning_rate": 6.491634692845781e-06,
"loss": 0.6334,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.7666280269622803,
"learning_rate": 6.437996637160086e-06,
"loss": 0.6128,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.9193455576896667,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6091,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.8365340828895569,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6355,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 1.0293723344802856,
"learning_rate": 6.276023446318214e-06,
"loss": 0.6351,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.9114359617233276,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6149,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.8085458278656006,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.6254,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.8302372097969055,
"learning_rate": 6.112604669781572e-06,
"loss": 0.5897,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.7839798331260681,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6172,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.8091347813606262,
"learning_rate": 6.002947078916365e-06,
"loss": 0.6216,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.781509280204773,
"learning_rate": 5.947925441958393e-06,
"loss": 0.6188,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.8232189416885376,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6249,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.7533282041549683,
"learning_rate": 5.837531116523683e-06,
"loss": 0.6327,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.778330385684967,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6346,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.7863406538963318,
"learning_rate": 5.726715068949564e-06,
"loss": 0.6125,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.7328237295150757,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6344,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.7579127550125122,
"learning_rate": 5.615533098453215e-06,
"loss": 0.6026,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.8167097568511963,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6192,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.775786817073822,
"learning_rate": 5.504041188505022e-06,
"loss": 0.6342,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.7578753232955933,
"learning_rate": 5.448196544517168e-06,
"loss": 0.6394,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.7454594969749451,
"learning_rate": 5.392295478639226e-06,
"loss": 0.6187,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.7795440554618835,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6384,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.730388879776001,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6435,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.7352913022041321,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6179,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.7331161499023438,
"learning_rate": 5.168267827938971e-06,
"loss": 0.594,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.7361404299736023,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6206,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.7892236113548279,
"learning_rate": 5.05609869177323e-06,
"loss": 0.6326,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.7411230802536011,
"learning_rate": 5e-06,
"loss": 0.6325,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.7337684035301208,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6098,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.7578845620155334,
"learning_rate": 4.887809678520976e-06,
"loss": 0.628,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.7919145226478577,
"learning_rate": 4.831732172061032e-06,
"loss": 0.6097,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.8074067831039429,
"learning_rate": 4.775675848247427e-06,
"loss": 0.6305,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.7526893019676208,
"learning_rate": 4.719647763814041e-06,
"loss": 0.6194,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.7639228105545044,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6156,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.7965607643127441,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.6364,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.7543295621871948,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6247,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.7019674777984619,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.6261,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.7598593831062317,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6167,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.7225008010864258,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6456,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.7532970309257507,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6107,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.7489970326423645,
"learning_rate": 4.273284931050438e-06,
"loss": 0.6107,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.7544114589691162,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6017,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.7361477613449097,
"learning_rate": 4.162468883476319e-06,
"loss": 0.6239,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.7545927166938782,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6235,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.714535653591156,
"learning_rate": 4.052074558041608e-06,
"loss": 0.6591,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.7137702703475952,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6251,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.7074740529060364,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.6,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.7791515588760376,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6399,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.9246359467506409,
"learning_rate": 3.832773180720475e-06,
"loss": 0.619,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.7348170876502991,
"learning_rate": 3.778297969310529e-06,
"loss": 0.629,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.7313644289970398,
"learning_rate": 3.723976553681787e-06,
"loss": 0.6311,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.7346382737159729,
"learning_rate": 3.669815772166625e-06,
"loss": 0.6005,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.7484610080718994,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.621,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.6898188591003418,
"learning_rate": 3.562003362839914e-06,
"loss": 0.5986,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.774992048740387,
"learning_rate": 3.50836530715422e-06,
"loss": 0.6078,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.6841826438903809,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.5921,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.737432062625885,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.6068,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.7477750182151794,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6056,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.725186824798584,
"learning_rate": 3.295758014387375e-06,
"loss": 0.6262,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.7065021395683289,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6523,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.7657485604286194,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.6194,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.7598536610603333,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6095,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.6754752993583679,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.6468,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.6686747074127197,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6121,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.7546200752258301,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5609,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.7604199051856995,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5585,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.681459903717041,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5649,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.6781082153320312,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5252,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.712261974811554,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5467,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.8066717982292175,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.536,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.8354761600494385,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5513,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.7632562518119812,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5039,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.718770444393158,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5497,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.6983654499053955,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5125,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.6584542989730835,
"learning_rate": 2.483822884971e-06,
"loss": 0.5633,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.8151694536209106,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5327,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.7218419909477234,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5566,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.6894270777702332,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5321,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.673761785030365,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5137,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.7555160522460938,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5384,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.6827072501182556,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5687,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.7825382351875305,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5151,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.7007843852043152,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5314,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.6600434184074402,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5313,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.6345268487930298,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5473,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.6742668151855469,
"learning_rate": 1.971049780795901e-06,
"loss": 0.548,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.6980031728744507,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5352,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.692200243473053,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5293,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.6861526966094971,
"learning_rate": 1.838887489710452e-06,
"loss": 0.533,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.694861650466919,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5171,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.672132670879364,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5225,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.6298666596412659,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5314,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.7913171052932739,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5654,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.758754312992096,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5453,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.6853262782096863,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5489,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.6726512908935547,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5467,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.646966278553009,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5243,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.6261551380157471,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5452,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.6790262460708618,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5341,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.6597793698310852,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5567,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.6485868096351624,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5781,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.6337335705757141,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5161,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.6472348570823669,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5259,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.66475909948349,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5322,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.645846962928772,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5332,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.6724075675010681,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5259,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.6482080817222595,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5411,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.6637566089630127,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5631,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.6474559307098389,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5295,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.6283987164497375,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.501,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.6833587288856506,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5509,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.7471374869346619,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5392,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.6412124037742615,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5469,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.6492853760719299,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5476,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.647890031337738,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5233,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.6412302255630493,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5309,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.6383857727050781,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5415,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.6454289555549622,
"learning_rate": 7.663790038585794e-07,
"loss": 0.545,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.6530759930610657,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5278,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.6237668991088867,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5495,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.6286064386367798,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5267,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.6250666379928589,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5342,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.6225396990776062,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5433,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.6211515665054321,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5439,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.6376200914382935,
"learning_rate": 5.707160663821009e-07,
"loss": 0.5387,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.6262568235397339,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5604,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.6484907269477844,
"learning_rate": 5.197795201563744e-07,
"loss": 0.544,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.6246148347854614,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5331,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.6578778624534607,
"learning_rate": 4.710988998640298e-07,
"loss": 0.5369,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.6562095880508423,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5317,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.6582750082015991,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.5553,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.6188002228736877,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5321,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.6145332455635071,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.543,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.590064287185669,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.56,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.5917049050331116,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.5618,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.5938456058502197,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5458,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.6525412797927856,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.517,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.6126443147659302,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5375,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.5874183177947998,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.5658,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.6362209320068359,
"learning_rate": 2.447174185242324e-07,
"loss": 0.4995,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.6762949228286743,
"learning_rate": 2.276812823220964e-07,
"loss": 0.5082,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.6433969140052795,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5372,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.6176011562347412,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5187,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.620287299156189,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5261,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.6410089135169983,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5458,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.6426790356636047,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5043,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.6398621797561646,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5466,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.6380493640899658,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5599,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.6233465671539307,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5397,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.6533657908439636,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5171,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.6348655223846436,
"learning_rate": 9.0678523819408e-08,
"loss": 0.539,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.5884503722190857,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5661,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.585837185382843,
"learning_rate": 7.064490740882057e-08,
"loss": 0.5586,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.6132030487060547,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5474,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.6571253538131714,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5283,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.645630955696106,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5153,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.6425459980964661,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5316,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.6377753019332886,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5318,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.5929200649261475,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5563,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.6763033270835876,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5398,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.6499728560447693,
"learning_rate": 1.541333133436018e-08,
"loss": 0.4992,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.6668295860290527,
"learning_rate": 1.132562476771959e-08,
"loss": 0.4918,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.6175950765609741,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5105,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.6248342394828796,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5696,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.6357370615005493,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5259,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.5905089378356934,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5496,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.6271190643310547,
"learning_rate": 3.147162264971471e-10,
"loss": 0.5213,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.6295599341392517,
"learning_rate": 0.0,
"loss": 0.5098,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 1.8386307631588966e+17,
"train_loss": 0.0,
"train_runtime": 9.6901,
"train_samples_per_second": 3095.953,
"train_steps_per_second": 32.198
}
],
"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": 1.8386307631588966e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}