EtashGuha's picture
Upload folder using huggingface_hub
6094d3d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9904153354632586,
"eval_steps": 500,
"global_step": 312,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009584664536741214,
"grad_norm": 6.522148609161377,
"learning_rate": 3.125e-07,
"loss": 1.0912,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 6.512198448181152,
"learning_rate": 6.25e-07,
"loss": 1.0863,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 6.517064571380615,
"learning_rate": 9.375000000000001e-07,
"loss": 1.0755,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 6.4605488777160645,
"learning_rate": 1.25e-06,
"loss": 1.0498,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 6.430959701538086,
"learning_rate": 1.5625e-06,
"loss": 1.0655,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 6.215013027191162,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.0356,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 4.995906829833984,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.0335,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.521610260009766,
"learning_rate": 2.5e-06,
"loss": 1.0104,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 2.387202024459839,
"learning_rate": 2.8125e-06,
"loss": 0.9163,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.42122745513916,
"learning_rate": 3.125e-06,
"loss": 0.9514,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.1006500720977783,
"learning_rate": 3.4375e-06,
"loss": 0.9227,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.755831718444824,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.9271,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 4.204206466674805,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.9457,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 4.184535026550293,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9294,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 3.9500532150268555,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.9384,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.8082499504089355,
"learning_rate": 5e-06,
"loss": 0.8787,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.364701986312866,
"learning_rate": 5.3125e-06,
"loss": 0.8882,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.1051669120788574,
"learning_rate": 5.625e-06,
"loss": 0.8866,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.8516038656234741,
"learning_rate": 5.9375e-06,
"loss": 0.8262,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.711443543434143,
"learning_rate": 6.25e-06,
"loss": 0.8258,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.5168342590332031,
"learning_rate": 6.5625e-06,
"loss": 0.8182,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.422421932220459,
"learning_rate": 6.875e-06,
"loss": 0.8159,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.4139771461486816,
"learning_rate": 7.1875e-06,
"loss": 0.8328,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.3288089036941528,
"learning_rate": 7.500000000000001e-06,
"loss": 0.7934,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.2513765096664429,
"learning_rate": 7.8125e-06,
"loss": 0.7647,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.3500066995620728,
"learning_rate": 8.125000000000001e-06,
"loss": 0.7962,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.1734682321548462,
"learning_rate": 8.4375e-06,
"loss": 0.7936,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.1593419313430786,
"learning_rate": 8.750000000000001e-06,
"loss": 0.7623,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 1.1951789855957031,
"learning_rate": 9.0625e-06,
"loss": 0.8203,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 1.0117515325546265,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7291,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.9888120293617249,
"learning_rate": 9.6875e-06,
"loss": 0.787,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 1.0005769729614258,
"learning_rate": 1e-05,
"loss": 0.7805,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.9033569097518921,
"learning_rate": 9.999685283773504e-06,
"loss": 0.769,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.7813165187835693,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7464,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.8380439877510071,
"learning_rate": 9.997167791667668e-06,
"loss": 0.758,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.9050424695014954,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7619,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.7175437808036804,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7579,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.9002037644386292,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7772,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.7795717716217041,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7375,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.6820951104164124,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7492,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.9073566794395447,
"learning_rate": 9.974529372743762e-06,
"loss": 0.7192,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.981580913066864,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7411,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.6163010001182556,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7201,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.9524637460708618,
"learning_rate": 9.954748808839675e-06,
"loss": 0.7263,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.8435295820236206,
"learning_rate": 9.946906630265184e-06,
"loss": 0.7683,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.6974939703941345,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7381,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.8612231612205505,
"learning_rate": 9.92935509259118e-06,
"loss": 0.7024,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.8264601826667786,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7402,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.7052453756332397,
"learning_rate": 9.909321476180594e-06,
"loss": 0.7646,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.7098926305770874,
"learning_rate": 9.898376992116179e-06,
"loss": 0.7406,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.7754364609718323,
"learning_rate": 9.886815868562596e-06,
"loss": 0.7336,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.6379909515380859,
"learning_rate": 9.874639560909118e-06,
"loss": 0.7182,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.7664453387260437,
"learning_rate": 9.861849601988384e-06,
"loss": 0.7149,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.6363523006439209,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7311,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.6430462598800659,
"learning_rate": 9.834435247725032e-06,
"loss": 0.7285,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.723187267780304,
"learning_rate": 9.819814303479268e-06,
"loss": 0.7571,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.7131056785583496,
"learning_rate": 9.804586609725499e-06,
"loss": 0.753,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.8157994151115417,
"learning_rate": 9.788754083424654e-06,
"loss": 0.7367,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.6841234564781189,
"learning_rate": 9.772318717677905e-06,
"loss": 0.7403,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.7338866591453552,
"learning_rate": 9.755282581475769e-06,
"loss": 0.7519,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.7053446173667908,
"learning_rate": 9.737647819437645e-06,
"loss": 0.7482,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.7356135845184326,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7263,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.7298708558082581,
"learning_rate": 9.700591372846096e-06,
"loss": 0.7062,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.7238677740097046,
"learning_rate": 9.681174353198687e-06,
"loss": 0.7567,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.6636947989463806,
"learning_rate": 9.661168036940071e-06,
"loss": 0.7153,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.6508307456970215,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7306,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.7953429222106934,
"learning_rate": 9.619397662556434e-06,
"loss": 0.7341,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.6882739067077637,
"learning_rate": 9.597638862757255e-06,
"loss": 0.7347,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.6194055080413818,
"learning_rate": 9.5753012823366e-06,
"loss": 0.7061,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.6052485704421997,
"learning_rate": 9.552387733294081e-06,
"loss": 0.6964,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.7361829876899719,
"learning_rate": 9.528901100135971e-06,
"loss": 0.6991,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.715272843837738,
"learning_rate": 9.504844339512096e-06,
"loss": 0.755,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.62732994556427,
"learning_rate": 9.480220479843627e-06,
"loss": 0.6806,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.6641494631767273,
"learning_rate": 9.45503262094184e-06,
"loss": 0.7107,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.8504228591918945,
"learning_rate": 9.4292839336179e-06,
"loss": 0.7253,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.6230931282043457,
"learning_rate": 9.40297765928369e-06,
"loss": 0.7186,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.6660647988319397,
"learning_rate": 9.376117109543769e-06,
"loss": 0.7081,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.751896858215332,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7173,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.7081503868103027,
"learning_rate": 9.320746778718274e-06,
"loss": 0.7204,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.6379644870758057,
"learning_rate": 9.292243968009332e-06,
"loss": 0.7084,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.6575475931167603,
"learning_rate": 9.263200821770462e-06,
"loss": 0.7059,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.5726634860038757,
"learning_rate": 9.233620996141421e-06,
"loss": 0.7039,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.6370197534561157,
"learning_rate": 9.203508214822652e-06,
"loss": 0.7164,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.5949355959892273,
"learning_rate": 9.172866268606514e-06,
"loss": 0.7383,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.6636143922805786,
"learning_rate": 9.141699014900084e-06,
"loss": 0.7029,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.6890977025032043,
"learning_rate": 9.110010377239552e-06,
"loss": 0.7332,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.6502646207809448,
"learning_rate": 9.077804344796302e-06,
"loss": 0.708,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.5691036581993103,
"learning_rate": 9.045084971874738e-06,
"loss": 0.6873,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.6357378363609314,
"learning_rate": 9.011856377401891e-06,
"loss": 0.6985,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.6481894254684448,
"learning_rate": 8.978122744408905e-06,
"loss": 0.733,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.5811349749565125,
"learning_rate": 8.943888319504456e-06,
"loss": 0.6713,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.6648602485656738,
"learning_rate": 8.90915741234015e-06,
"loss": 0.7035,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.681630551815033,
"learning_rate": 8.873934395068006e-06,
"loss": 0.7115,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.66861891746521,
"learning_rate": 8.838223701790057e-06,
"loss": 0.7198,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.7022870779037476,
"learning_rate": 8.802029828000157e-06,
"loss": 0.7223,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.5813931822776794,
"learning_rate": 8.765357330018056e-06,
"loss": 0.7168,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.6184871792793274,
"learning_rate": 8.728210824415829e-06,
"loss": 0.7028,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.5462105870246887,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6982,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.7070772051811218,
"learning_rate": 8.652514554406388e-06,
"loss": 0.7226,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.630681574344635,
"learning_rate": 8.613974319136959e-06,
"loss": 0.7254,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.6217668652534485,
"learning_rate": 8.574979133323378e-06,
"loss": 0.723,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.6539044380187988,
"learning_rate": 8.535533905932739e-06,
"loss": 0.7028,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.6340758800506592,
"learning_rate": 8.495643602586287e-06,
"loss": 0.723,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.6706730723381042,
"learning_rate": 8.455313244934324e-06,
"loss": 0.7124,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.680267333984375,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6709,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.6726241111755371,
"learning_rate": 8.373352729660373e-06,
"loss": 0.6179,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.7150067687034607,
"learning_rate": 8.331732889760021e-06,
"loss": 0.6241,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.56097811460495,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6308,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.7451176047325134,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6663,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.7421462535858154,
"learning_rate": 8.204378069925121e-06,
"loss": 0.619,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.6038477420806885,
"learning_rate": 8.16111251028955e-06,
"loss": 0.6176,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.7240898013114929,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6356,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.6246081590652466,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6252,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.5648534893989563,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6171,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.5854071378707886,
"learning_rate": 7.984126070912519e-06,
"loss": 0.6111,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.5936861038208008,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6313,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.565150797367096,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6269,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.6269688606262207,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6285,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.6298412680625916,
"learning_rate": 7.801129998764014e-06,
"loss": 0.6023,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.6648682951927185,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6389,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.5603852272033691,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6331,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.58585524559021,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6213,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.6422139406204224,
"learning_rate": 7.612492823579744e-06,
"loss": 0.6301,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.6292246580123901,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6339,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.6058092713356018,
"learning_rate": 7.516177115029002e-06,
"loss": 0.6158,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.6214110851287842,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6161,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.5520999431610107,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.6206,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.6907647848129272,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6053,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.5952545404434204,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5911,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.5405617356300354,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6274,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.5691115260124207,
"learning_rate": 7.219825320152411e-06,
"loss": 0.6187,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.6347238421440125,
"learning_rate": 7.169418695587791e-06,
"loss": 0.62,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.5302215814590454,
"learning_rate": 7.118738970516944e-06,
"loss": 0.6159,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.5417319536209106,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6148,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.6301731467247009,
"learning_rate": 7.016585772004026e-06,
"loss": 0.6116,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.552765965461731,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6191,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.5948565602302551,
"learning_rate": 6.913417161825449e-06,
"loss": 0.6404,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.6133633255958557,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.5985,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.557786762714386,
"learning_rate": 6.809285088483361e-06,
"loss": 0.6312,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.5906978249549866,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6248,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.6574854850769043,
"learning_rate": 6.704241985612625e-06,
"loss": 0.6207,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.5244562029838562,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6263,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.5198882222175598,
"learning_rate": 6.598340745578908e-06,
"loss": 0.6139,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.6142861247062683,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6349,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.5821816325187683,
"learning_rate": 6.491634692845781e-06,
"loss": 0.6108,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.5869280695915222,
"learning_rate": 6.437996637160086e-06,
"loss": 0.6087,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.5360026955604553,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6139,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.5599037408828735,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6216,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.6269216537475586,
"learning_rate": 6.276023446318214e-06,
"loss": 0.616,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.5310914516448975,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6017,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.5873135924339294,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.6213,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.5658697485923767,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6148,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.550531268119812,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6132,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.5577810406684875,
"learning_rate": 6.002947078916365e-06,
"loss": 0.642,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.5472114086151123,
"learning_rate": 5.947925441958393e-06,
"loss": 0.6545,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.591293454170227,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6183,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.5645745396614075,
"learning_rate": 5.837531116523683e-06,
"loss": 0.627,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.5754922032356262,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6099,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.6357850432395935,
"learning_rate": 5.726715068949564e-06,
"loss": 0.6201,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.5578529238700867,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6211,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.581805408000946,
"learning_rate": 5.615533098453215e-06,
"loss": 0.619,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 6.261775016784668,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6533,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.6721850633621216,
"learning_rate": 5.504041188505022e-06,
"loss": 0.6313,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.5541729927062988,
"learning_rate": 5.448196544517168e-06,
"loss": 0.6263,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.5610416531562805,
"learning_rate": 5.392295478639226e-06,
"loss": 0.6275,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.6091299057006836,
"learning_rate": 5.336345028060199e-06,
"loss": 0.5998,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.6036515831947327,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6388,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.5713905692100525,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6189,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.5874393582344055,
"learning_rate": 5.168267827938971e-06,
"loss": 0.6271,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.588840663433075,
"learning_rate": 5.112190321479026e-06,
"loss": 0.5981,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.5443650484085083,
"learning_rate": 5.05609869177323e-06,
"loss": 0.6152,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.5509213209152222,
"learning_rate": 5e-06,
"loss": 0.5995,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.5613495111465454,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6183,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.514927327632904,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6116,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.5181869268417358,
"learning_rate": 4.831732172061032e-06,
"loss": 0.6238,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.6072043180465698,
"learning_rate": 4.775675848247427e-06,
"loss": 0.624,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.5527750849723816,
"learning_rate": 4.719647763814041e-06,
"loss": 0.6161,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.5072109699249268,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6051,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.49356722831726074,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.6005,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.5539764165878296,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6409,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.5382592082023621,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.6361,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.5737978219985962,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6038,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.5617745518684387,
"learning_rate": 4.384466901546786e-06,
"loss": 0.62,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.528666079044342,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.608,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.48081570863723755,
"learning_rate": 4.273284931050438e-06,
"loss": 0.5945,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5572990775108337,
"learning_rate": 4.217827674798845e-06,
"loss": 0.5987,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.6292819380760193,
"learning_rate": 4.162468883476319e-06,
"loss": 0.6274,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.5033635497093201,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6075,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.5088822841644287,
"learning_rate": 4.052074558041608e-06,
"loss": 0.5961,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5283027291297913,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6286,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.4841823875904083,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.6133,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.4581049382686615,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6126,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.5011346936225891,
"learning_rate": 3.832773180720475e-06,
"loss": 0.6089,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.5460132956504822,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6028,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.5823938250541687,
"learning_rate": 3.723976553681787e-06,
"loss": 0.6097,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.528449535369873,
"learning_rate": 3.669815772166625e-06,
"loss": 0.6454,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.5251087546348572,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.6153,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.5823463201522827,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6089,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.47486257553100586,
"learning_rate": 3.50836530715422e-06,
"loss": 0.611,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.45398715138435364,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.5828,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.5514724850654602,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.602,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.6049546599388123,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6171,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.5199345350265503,
"learning_rate": 3.295758014387375e-06,
"loss": 0.6049,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.5168394446372986,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6088,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.5167913436889648,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.6092,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.47773662209510803,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.5918,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.5226515531539917,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.6125,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.560547947883606,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6221,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.5965922474861145,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.574,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.6111416220664978,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5479,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.5487639307975769,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.55,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.4956718981266022,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5704,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.5371181964874268,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5697,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.5468514561653137,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5517,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.607031524181366,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5699,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.4856856167316437,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5468,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.5080529451370239,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5447,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.5684489607810974,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5431,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.5614061951637268,
"learning_rate": 2.483822884971e-06,
"loss": 0.5776,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.4923718273639679,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5456,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.5106306672096252,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5529,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.5100261569023132,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5655,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.4807545244693756,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5411,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.5379756093025208,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5544,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.46299222111701965,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.54,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.47700613737106323,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5362,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.5008731484413147,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5334,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.45344123244285583,
"learning_rate": 2.061073738537635e-06,
"loss": 0.568,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.4419834315776825,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5417,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.519332230091095,
"learning_rate": 1.971049780795901e-06,
"loss": 0.516,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.46591806411743164,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5668,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.46662262082099915,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.546,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.4593885838985443,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5563,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.4612986743450165,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5291,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.4624190330505371,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.567,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.5169050693511963,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5396,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.41065356135368347,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5677,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.47851431369781494,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5638,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.4506586194038391,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5541,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.4581591486930847,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5363,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.4875186085700989,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5535,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.44540533423423767,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5564,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.43508046865463257,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5415,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.45627349615097046,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5365,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.45240721106529236,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5548,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.45987048745155334,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5704,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.46185731887817383,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5404,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.4758492112159729,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5547,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.4201287627220154,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5558,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.43883946537971497,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.581,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.4413963854312897,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5221,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.45473068952560425,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5145,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.4768509566783905,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5632,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.45426684617996216,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5364,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.47609567642211914,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5653,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.45595109462738037,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5571,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.48026198148727417,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5667,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.4823807179927826,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5371,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.4506840109825134,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5293,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.4451153576374054,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5635,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.45654985308647156,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5779,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.43412479758262634,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5461,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.409091979265213,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5267,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.3990032374858856,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5555,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.49492818117141724,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5402,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.45726269483566284,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5048,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.47563958168029785,
"learning_rate": 6.238828904562316e-07,
"loss": 0.526,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.4424806237220764,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5792,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.45100167393684387,
"learning_rate": 5.707160663821009e-07,
"loss": 0.567,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.45940259099006653,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5669,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.44366738200187683,
"learning_rate": 5.197795201563744e-07,
"loss": 0.5634,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.43361055850982666,
"learning_rate": 4.951556604879049e-07,
"loss": 0.532,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.4700828790664673,
"learning_rate": 4.710988998640298e-07,
"loss": 0.5463,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.44592180848121643,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5545,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.4467045068740845,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.5433,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.4569363296031952,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5766,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.400199294090271,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.5398,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.451200008392334,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5358,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.47678378224372864,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.562,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.4619581997394562,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5676,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.4295126497745514,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.5546,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.4408467710018158,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5563,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.43506360054016113,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.5361,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.4526090919971466,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5345,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.4317993223667145,
"learning_rate": 2.276812823220964e-07,
"loss": 0.5436,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.43400490283966064,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5585,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.47704049944877625,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5498,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.42226728796958923,
"learning_rate": 1.801856965207338e-07,
"loss": 0.536,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.4588082432746887,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5648,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.42655742168426514,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.566,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.3996799886226654,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5512,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.41889339685440063,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5445,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.42305031418800354,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5436,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.45049145817756653,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.541,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.388206422328949,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5477,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.4722599685192108,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5568,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.43938127160072327,
"learning_rate": 7.064490740882057e-08,
"loss": 0.5696,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.4504903256893158,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5688,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.4146743714809418,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5562,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.42245328426361084,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5558,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.4283214211463928,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5449,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.44432875514030457,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5579,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.4334840178489685,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5811,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.4303494393825531,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5438,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.4944767951965332,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5513,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.44507452845573425,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5536,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.43108487129211426,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5147,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.4232444167137146,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5723,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.42380115389823914,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5316,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.44843682646751404,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5481,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.4626663625240326,
"learning_rate": 3.147162264971471e-10,
"loss": 0.536,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.38271576166152954,
"learning_rate": 0.0,
"loss": 0.5586,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 4.449990764932956e+17,
"train_loss": 0.0,
"train_runtime": 4.0505,
"train_samples_per_second": 7406.477,
"train_steps_per_second": 77.027
}
],
"logging_steps": 1,
"max_steps": 312,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.449990764932956e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}