EtashGuha's picture
Upload folder using huggingface_hub
0adc498 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": 7.450248718261719,
"learning_rate": 3.125e-07,
"loss": 1.1186,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 7.24557638168335,
"learning_rate": 6.25e-07,
"loss": 1.1121,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 7.313780784606934,
"learning_rate": 9.375000000000001e-07,
"loss": 1.118,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 7.112710952758789,
"learning_rate": 1.25e-06,
"loss": 1.0784,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 6.781392574310303,
"learning_rate": 1.5625e-06,
"loss": 1.0651,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 6.484024524688721,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.0659,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 4.651825904846191,
"learning_rate": 2.1875000000000002e-06,
"loss": 0.9996,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.462416648864746,
"learning_rate": 2.5e-06,
"loss": 1.0189,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 2.916067600250244,
"learning_rate": 2.8125e-06,
"loss": 0.9991,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.8599324226379395,
"learning_rate": 3.125e-06,
"loss": 0.964,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.618966579437256,
"learning_rate": 3.4375e-06,
"loss": 0.9712,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.536947011947632,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.954,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.6508381366729736,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.936,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.4823105335235596,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9108,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 3.1129627227783203,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.8912,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 1.9364891052246094,
"learning_rate": 5e-06,
"loss": 0.8785,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.11167573928833,
"learning_rate": 5.3125e-06,
"loss": 0.8535,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.0362987518310547,
"learning_rate": 5.625e-06,
"loss": 0.8541,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.5982022285461426,
"learning_rate": 5.9375e-06,
"loss": 0.8226,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.2859333753585815,
"learning_rate": 6.25e-06,
"loss": 0.829,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.3401594161987305,
"learning_rate": 6.5625e-06,
"loss": 0.7909,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.4725675582885742,
"learning_rate": 6.875e-06,
"loss": 0.8068,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.3111298084259033,
"learning_rate": 7.1875e-06,
"loss": 0.7714,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.0379825830459595,
"learning_rate": 7.500000000000001e-06,
"loss": 0.7875,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.0249028205871582,
"learning_rate": 7.8125e-06,
"loss": 0.7789,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.0806434154510498,
"learning_rate": 8.125000000000001e-06,
"loss": 0.7464,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.0358896255493164,
"learning_rate": 8.4375e-06,
"loss": 0.7613,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 0.8783509731292725,
"learning_rate": 8.750000000000001e-06,
"loss": 0.7616,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 0.9225448369979858,
"learning_rate": 9.0625e-06,
"loss": 0.7643,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 0.9625920653343201,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7379,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.9467800259590149,
"learning_rate": 9.6875e-06,
"loss": 0.739,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.7917072176933289,
"learning_rate": 1e-05,
"loss": 0.7246,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.79640132188797,
"learning_rate": 9.999685283773504e-06,
"loss": 0.7187,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.8204032778739929,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7152,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.753680944442749,
"learning_rate": 9.997167791667668e-06,
"loss": 0.7251,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.9382561445236206,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7233,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.8106517791748047,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7327,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.7491211295127869,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7107,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.749087929725647,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7223,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.7449676394462585,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7131,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.7968466281890869,
"learning_rate": 9.974529372743762e-06,
"loss": 0.6924,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.8374283909797668,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7023,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.7383466362953186,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7062,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.7412065267562866,
"learning_rate": 9.954748808839675e-06,
"loss": 0.7058,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.8397060632705688,
"learning_rate": 9.946906630265184e-06,
"loss": 0.6863,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.6921831369400024,
"learning_rate": 9.938441702975689e-06,
"loss": 0.6974,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.8667184114456177,
"learning_rate": 9.92935509259118e-06,
"loss": 0.6686,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.6668515801429749,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7036,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.7965338230133057,
"learning_rate": 9.909321476180594e-06,
"loss": 0.6821,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.7476664781570435,
"learning_rate": 9.898376992116179e-06,
"loss": 0.6925,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.7460647225379944,
"learning_rate": 9.886815868562596e-06,
"loss": 0.6938,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.7604838609695435,
"learning_rate": 9.874639560909118e-06,
"loss": 0.676,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.7249597907066345,
"learning_rate": 9.861849601988384e-06,
"loss": 0.6772,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.672968327999115,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7079,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.6397844552993774,
"learning_rate": 9.834435247725032e-06,
"loss": 0.6782,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.7087897062301636,
"learning_rate": 9.819814303479268e-06,
"loss": 0.6812,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.6313179135322571,
"learning_rate": 9.804586609725499e-06,
"loss": 0.6891,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.7048418521881104,
"learning_rate": 9.788754083424654e-06,
"loss": 0.6854,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.6497188210487366,
"learning_rate": 9.772318717677905e-06,
"loss": 0.679,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.6995434761047363,
"learning_rate": 9.755282581475769e-06,
"loss": 0.6709,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.6485216021537781,
"learning_rate": 9.737647819437645e-06,
"loss": 0.689,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.6646871566772461,
"learning_rate": 9.719416651541839e-06,
"loss": 0.6741,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.6768806576728821,
"learning_rate": 9.700591372846096e-06,
"loss": 0.688,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.5700349807739258,
"learning_rate": 9.681174353198687e-06,
"loss": 0.6751,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.6627057790756226,
"learning_rate": 9.661168036940071e-06,
"loss": 0.6676,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.6286057233810425,
"learning_rate": 9.640574942595195e-06,
"loss": 0.6531,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.7126429677009583,
"learning_rate": 9.619397662556434e-06,
"loss": 0.6731,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.6601884365081787,
"learning_rate": 9.597638862757255e-06,
"loss": 0.6588,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.6407045125961304,
"learning_rate": 9.5753012823366e-06,
"loss": 0.6938,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.8652023077011108,
"learning_rate": 9.552387733294081e-06,
"loss": 0.6735,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.687851071357727,
"learning_rate": 9.528901100135971e-06,
"loss": 0.6723,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.8586165308952332,
"learning_rate": 9.504844339512096e-06,
"loss": 0.6507,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.6593796014785767,
"learning_rate": 9.480220479843627e-06,
"loss": 0.6709,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.6938400268554688,
"learning_rate": 9.45503262094184e-06,
"loss": 0.6587,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.7197189331054688,
"learning_rate": 9.4292839336179e-06,
"loss": 0.6727,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.601469874382019,
"learning_rate": 9.40297765928369e-06,
"loss": 0.6678,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.719049870967865,
"learning_rate": 9.376117109543769e-06,
"loss": 0.6553,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.6036619544029236,
"learning_rate": 9.348705665778479e-06,
"loss": 0.6706,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.6838827729225159,
"learning_rate": 9.320746778718274e-06,
"loss": 0.6612,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.6266648173332214,
"learning_rate": 9.292243968009332e-06,
"loss": 0.6619,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.695432186126709,
"learning_rate": 9.263200821770462e-06,
"loss": 0.6793,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.6491926312446594,
"learning_rate": 9.233620996141421e-06,
"loss": 0.6747,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.6380310654640198,
"learning_rate": 9.203508214822652e-06,
"loss": 0.6725,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.7386761903762817,
"learning_rate": 9.172866268606514e-06,
"loss": 0.6639,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.6107885837554932,
"learning_rate": 9.141699014900084e-06,
"loss": 0.6615,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.5989591479301453,
"learning_rate": 9.110010377239552e-06,
"loss": 0.6501,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.6296937465667725,
"learning_rate": 9.077804344796302e-06,
"loss": 0.6656,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.6283934712409973,
"learning_rate": 9.045084971874738e-06,
"loss": 0.6541,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.5972719788551331,
"learning_rate": 9.011856377401891e-06,
"loss": 0.6502,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.6227993965148926,
"learning_rate": 8.978122744408905e-06,
"loss": 0.6532,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.566012442111969,
"learning_rate": 8.943888319504456e-06,
"loss": 0.6683,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.7559008598327637,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6695,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.5750781893730164,
"learning_rate": 8.873934395068006e-06,
"loss": 0.6555,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.630652904510498,
"learning_rate": 8.838223701790057e-06,
"loss": 0.6669,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.6783167123794556,
"learning_rate": 8.802029828000157e-06,
"loss": 0.6494,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.639247715473175,
"learning_rate": 8.765357330018056e-06,
"loss": 0.6479,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.6114720702171326,
"learning_rate": 8.728210824415829e-06,
"loss": 0.6504,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.75978022813797,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6449,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.5664803385734558,
"learning_rate": 8.652514554406388e-06,
"loss": 0.6618,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.8079279661178589,
"learning_rate": 8.613974319136959e-06,
"loss": 0.6557,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.6475933194160461,
"learning_rate": 8.574979133323378e-06,
"loss": 0.6643,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.6174091696739197,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6627,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.7467196583747864,
"learning_rate": 8.495643602586287e-06,
"loss": 0.6499,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.6515322923660278,
"learning_rate": 8.455313244934324e-06,
"loss": 0.6434,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.7229475975036621,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6155,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.7079372406005859,
"learning_rate": 8.373352729660373e-06,
"loss": 0.5973,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.7019097208976746,
"learning_rate": 8.331732889760021e-06,
"loss": 0.5929,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.5983749032020569,
"learning_rate": 8.289693629698564e-06,
"loss": 0.5806,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.6598142981529236,
"learning_rate": 8.247240241650918e-06,
"loss": 0.5974,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.7377691268920898,
"learning_rate": 8.204378069925121e-06,
"loss": 0.612,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.6421460509300232,
"learning_rate": 8.16111251028955e-06,
"loss": 0.6139,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.6638056635856628,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6027,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.5684019327163696,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6188,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.6008623242378235,
"learning_rate": 8.0289502192041e-06,
"loss": 0.5828,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.6357205510139465,
"learning_rate": 7.984126070912519e-06,
"loss": 0.5991,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.5860634446144104,
"learning_rate": 7.938926261462366e-06,
"loss": 0.5793,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.6548704504966736,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6153,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.5683628916740417,
"learning_rate": 7.84742246584226e-06,
"loss": 0.5948,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.6051099896430969,
"learning_rate": 7.801129998764014e-06,
"loss": 0.5902,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.6686009168624878,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6107,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.5977694988250732,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6012,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.6836264729499817,
"learning_rate": 7.660160382576683e-06,
"loss": 0.5743,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.5479535460472107,
"learning_rate": 7.612492823579744e-06,
"loss": 0.6081,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.6493426561355591,
"learning_rate": 7.564496387029532e-06,
"loss": 0.5946,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.6699090600013733,
"learning_rate": 7.516177115029002e-06,
"loss": 0.5831,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.6036891937255859,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6132,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.6010143160820007,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.5936,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.5711460113525391,
"learning_rate": 7.369343312364994e-06,
"loss": 0.5886,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.6079814434051514,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5869,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.5685780644416809,
"learning_rate": 7.269952498697734e-06,
"loss": 0.5899,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.5830773711204529,
"learning_rate": 7.219825320152411e-06,
"loss": 0.5751,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.5980797410011292,
"learning_rate": 7.169418695587791e-06,
"loss": 0.5925,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.5906989574432373,
"learning_rate": 7.118738970516944e-06,
"loss": 0.5935,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.6026638150215149,
"learning_rate": 7.067792524832604e-06,
"loss": 0.5996,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.6075884103775024,
"learning_rate": 7.016585772004026e-06,
"loss": 0.5874,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.6681843400001526,
"learning_rate": 6.965125158269619e-06,
"loss": 0.5823,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.659089982509613,
"learning_rate": 6.913417161825449e-06,
"loss": 0.5929,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.5526909828186035,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.5977,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.6259062886238098,
"learning_rate": 6.809285088483361e-06,
"loss": 0.5876,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.5340954661369324,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6009,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.6012002825737,
"learning_rate": 6.704241985612625e-06,
"loss": 0.5971,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.5407912135124207,
"learning_rate": 6.651395309775837e-06,
"loss": 0.5871,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.5475369691848755,
"learning_rate": 6.598340745578908e-06,
"loss": 0.5888,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.6183061003684998,
"learning_rate": 6.545084971874738e-06,
"loss": 0.5955,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.531151533126831,
"learning_rate": 6.491634692845781e-06,
"loss": 0.577,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.5379141569137573,
"learning_rate": 6.437996637160086e-06,
"loss": 0.5885,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.5697733759880066,
"learning_rate": 6.384177557124247e-06,
"loss": 0.605,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.5374375581741333,
"learning_rate": 6.330184227833376e-06,
"loss": 0.603,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.5371067523956299,
"learning_rate": 6.276023446318214e-06,
"loss": 0.5774,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.5378482341766357,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.5819,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.5637863278388977,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.5812,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.556528627872467,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6075,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.5494619011878967,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6036,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.5456545352935791,
"learning_rate": 6.002947078916365e-06,
"loss": 0.5859,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.5623578429222107,
"learning_rate": 5.947925441958393e-06,
"loss": 0.5953,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.50341796875,
"learning_rate": 5.892784473993184e-06,
"loss": 0.608,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.5696956515312195,
"learning_rate": 5.837531116523683e-06,
"loss": 0.5784,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.4976644515991211,
"learning_rate": 5.782172325201155e-06,
"loss": 0.5913,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.5173296928405762,
"learning_rate": 5.726715068949564e-06,
"loss": 0.5766,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.5494778752326965,
"learning_rate": 5.671166329088278e-06,
"loss": 0.5975,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.6095888018608093,
"learning_rate": 5.615533098453215e-06,
"loss": 0.5734,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.5598102807998657,
"learning_rate": 5.559822380516539e-06,
"loss": 0.5796,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.5714068412780762,
"learning_rate": 5.504041188505022e-06,
"loss": 0.596,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.6388280391693115,
"learning_rate": 5.448196544517168e-06,
"loss": 0.5847,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.5232503414154053,
"learning_rate": 5.392295478639226e-06,
"loss": 0.5876,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.5480726957321167,
"learning_rate": 5.336345028060199e-06,
"loss": 0.5711,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.6289456486701965,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6026,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.5380907654762268,
"learning_rate": 5.224324151752575e-06,
"loss": 0.5814,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.508001983165741,
"learning_rate": 5.168267827938971e-06,
"loss": 0.5831,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.5125857591629028,
"learning_rate": 5.112190321479026e-06,
"loss": 0.5961,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.5560346245765686,
"learning_rate": 5.05609869177323e-06,
"loss": 0.5922,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.48870888352394104,
"learning_rate": 5e-06,
"loss": 0.6035,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.4971839189529419,
"learning_rate": 4.943901308226771e-06,
"loss": 0.5857,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.4851223826408386,
"learning_rate": 4.887809678520976e-06,
"loss": 0.5835,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.5328238606452942,
"learning_rate": 4.831732172061032e-06,
"loss": 0.5793,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.5126102566719055,
"learning_rate": 4.775675848247427e-06,
"loss": 0.5948,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.5017287731170654,
"learning_rate": 4.719647763814041e-06,
"loss": 0.5913,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.5203403234481812,
"learning_rate": 4.663654971939802e-06,
"loss": 0.5903,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.5143176317214966,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.5853,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.5286027789115906,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6007,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.47466132044792175,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.5984,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.5495244264602661,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.5876,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.5817933082580566,
"learning_rate": 4.384466901546786e-06,
"loss": 0.597,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.5417107939720154,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.5906,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.4903656244277954,
"learning_rate": 4.273284931050438e-06,
"loss": 0.5996,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5955662727355957,
"learning_rate": 4.217827674798845e-06,
"loss": 0.5858,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.547533392906189,
"learning_rate": 4.162468883476319e-06,
"loss": 0.598,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.5238702297210693,
"learning_rate": 4.107215526006818e-06,
"loss": 0.5908,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.5798952579498291,
"learning_rate": 4.052074558041608e-06,
"loss": 0.5816,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5304401516914368,
"learning_rate": 3.997052921083637e-06,
"loss": 0.5683,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.5353245735168457,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.6024,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.551025927066803,
"learning_rate": 3.887395330218429e-06,
"loss": 0.5792,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.526214599609375,
"learning_rate": 3.832773180720475e-06,
"loss": 0.5987,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.5091916918754578,
"learning_rate": 3.778297969310529e-06,
"loss": 0.5935,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.5357905626296997,
"learning_rate": 3.723976553681787e-06,
"loss": 0.5929,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.4728222191333771,
"learning_rate": 3.669815772166625e-06,
"loss": 0.5868,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.5328664779663086,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.5622,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.49979132413864136,
"learning_rate": 3.562003362839914e-06,
"loss": 0.5783,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.5512866377830505,
"learning_rate": 3.50836530715422e-06,
"loss": 0.6064,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.49734431505203247,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.5698,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.474129319190979,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.5816,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.4967298209667206,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.5703,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.5029863119125366,
"learning_rate": 3.295758014387375e-06,
"loss": 0.5738,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.4886316955089569,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.592,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.5520089864730835,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.5765,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.49608224630355835,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.5856,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.4677266776561737,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.5664,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.4918242394924164,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6035,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.5218605399131775,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.556,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.47204381227493286,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5401,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.47082096338272095,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5277,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.4481445550918579,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5636,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.4531959593296051,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5454,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.47793692350387573,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5311,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.48525479435920715,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5428,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.4893665909767151,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5406,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.47740867733955383,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5323,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.469469279050827,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5418,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.48711007833480835,
"learning_rate": 2.483822884971e-06,
"loss": 0.5273,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.4613693058490753,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5646,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.4492114782333374,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5302,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.48082685470581055,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5592,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.5394153594970703,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5425,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.4770394265651703,
"learning_rate": 2.245515092739488e-06,
"loss": 0.531,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.45117616653442383,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5322,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.4770371913909912,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5292,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.4566550850868225,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5532,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.4458199441432953,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5392,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.5058304667472839,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5409,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.4507160782814026,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5263,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.46068480610847473,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5218,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.46374309062957764,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5543,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.46915560960769653,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5258,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.483087420463562,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5333,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.5250474214553833,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5229,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.4742869734764099,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5424,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.4359164535999298,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5371,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.44931837916374207,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5214,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.4842599928379059,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5181,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.5459963083267212,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5238,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.4280550479888916,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5387,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.44869115948677063,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5178,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.4132338762283325,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.548,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.4189448952674866,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5402,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.4935535192489624,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5479,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.4658398926258087,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5433,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.49564602971076965,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5466,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.4603484869003296,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5268,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.4429812729358673,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5225,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.42859166860580444,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5453,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.4350889325141907,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5351,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.42965197563171387,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5198,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.46081408858299255,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.564,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.46524178981781006,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5519,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.41897687315940857,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5193,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.443350613117218,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5354,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.4244984984397888,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5247,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.4337330460548401,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5459,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.4532371163368225,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5169,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.4506976306438446,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5165,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.4344449043273926,
"learning_rate": 7.964917851773496e-07,
"loss": 0.528,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.45108360052108765,
"learning_rate": 7.663790038585794e-07,
"loss": 0.513,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.43251749873161316,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5214,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.4088554382324219,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5391,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.43448659777641296,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5349,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.4252714514732361,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5363,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.43106770515441895,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5113,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.4056999981403351,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5529,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.4228290319442749,
"learning_rate": 5.707160663821009e-07,
"loss": 0.5268,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.416354238986969,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5402,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.42219412326812744,
"learning_rate": 5.197795201563744e-07,
"loss": 0.53,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.4102347493171692,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5298,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.40957483649253845,
"learning_rate": 4.710988998640298e-07,
"loss": 0.565,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.406904011964798,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5469,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.45355647802352905,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.5325,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.4218510091304779,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5238,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.4288652837276459,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.5429,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.4251857101917267,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5479,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.40958911180496216,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.539,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.44091689586639404,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5306,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.4396611750125885,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.5335,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.44288763403892517,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5195,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.3969884216785431,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.5232,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.39765551686286926,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5331,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.4320219159126282,
"learning_rate": 2.276812823220964e-07,
"loss": 0.5172,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.41505593061447144,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5238,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.42440736293792725,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5133,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.3800840675830841,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5524,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.43386730551719666,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5324,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.4183042347431183,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5233,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.3804335594177246,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5529,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.4096001982688904,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5472,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.4090404510498047,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5517,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.39008238911628723,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5527,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.43171262741088867,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5472,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.4097076654434204,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5426,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.42787063121795654,
"learning_rate": 7.064490740882057e-08,
"loss": 0.5601,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.39115628600120544,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5433,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.4000435173511505,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5148,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.39617058634757996,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5597,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.4308837354183197,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5412,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.3953721523284912,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5471,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.4228907525539398,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5075,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.4338865876197815,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5292,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.429914653301239,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5372,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.4009269177913666,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5462,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.39386454224586487,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5568,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.41612502932548523,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5342,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.4304322302341461,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5345,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.38909420371055603,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5417,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.41087430715560913,
"learning_rate": 3.147162264971471e-10,
"loss": 0.5323,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.41089069843292236,
"learning_rate": 0.0,
"loss": 0.529,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 3.8305788676695654e+17,
"train_loss": 0.0,
"train_runtime": 8.3215,
"train_samples_per_second": 3605.12,
"train_steps_per_second": 37.493
}
],
"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": 3.8305788676695654e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}