qwen2-5_variance_measure_0 / trainer_state.json
gsmyrnis's picture
End of training
59c54bb 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": 5.711534558196169,
"learning_rate": 3.125e-07,
"loss": 0.8336,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 5.925282951884479,
"learning_rate": 6.25e-07,
"loss": 0.8843,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 5.867533277653184,
"learning_rate": 9.375000000000001e-07,
"loss": 0.8479,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 5.824896671631882,
"learning_rate": 1.25e-06,
"loss": 0.8791,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 5.987531686771989,
"learning_rate": 1.5625e-06,
"loss": 0.9066,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 5.725917987282305,
"learning_rate": 1.8750000000000003e-06,
"loss": 0.8942,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 4.432509194712245,
"learning_rate": 2.1875000000000002e-06,
"loss": 0.8506,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.040144860527513,
"learning_rate": 2.5e-06,
"loss": 0.81,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 2.4488107864819852,
"learning_rate": 2.8125e-06,
"loss": 0.7746,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.2606800524253234,
"learning_rate": 3.125e-06,
"loss": 0.7643,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 1.951751229859671,
"learning_rate": 3.4375e-06,
"loss": 0.7635,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.3819192730318512,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.7142,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.8248142834310235,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.7219,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.7317214667554794,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.7085,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 3.7770220178446126,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.7252,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 3.0287370005760663,
"learning_rate": 5e-06,
"loss": 0.6709,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.6514038139395355,
"learning_rate": 5.3125e-06,
"loss": 0.6686,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.2373701343744914,
"learning_rate": 5.625e-06,
"loss": 0.668,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.5941560486608044,
"learning_rate": 5.9375e-06,
"loss": 0.6901,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.2731396760499303,
"learning_rate": 6.25e-06,
"loss": 0.6457,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.4184513474698528,
"learning_rate": 6.5625e-06,
"loss": 0.6476,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.6505836501747209,
"learning_rate": 6.875e-06,
"loss": 0.6561,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.521206243181705,
"learning_rate": 7.1875e-06,
"loss": 0.6412,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.1064476959153147,
"learning_rate": 7.500000000000001e-06,
"loss": 0.584,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 0.9045330461833524,
"learning_rate": 7.8125e-06,
"loss": 0.6152,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.203701597224554,
"learning_rate": 8.125000000000001e-06,
"loss": 0.6118,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.1271022288577037,
"learning_rate": 8.4375e-06,
"loss": 0.5845,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 0.7945307912659143,
"learning_rate": 8.750000000000001e-06,
"loss": 0.6021,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 0.7914084554497669,
"learning_rate": 9.0625e-06,
"loss": 0.6172,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 0.9767683472544576,
"learning_rate": 9.375000000000001e-06,
"loss": 0.5681,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.990971759550561,
"learning_rate": 9.6875e-06,
"loss": 0.6093,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.7668567462345751,
"learning_rate": 1e-05,
"loss": 0.6072,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.7540337567092248,
"learning_rate": 9.999685283773504e-06,
"loss": 0.5768,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.8800077580779977,
"learning_rate": 9.998741174712534e-06,
"loss": 0.5687,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.6930011804022109,
"learning_rate": 9.997167791667668e-06,
"loss": 0.5418,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.6196818050766172,
"learning_rate": 9.994965332706574e-06,
"loss": 0.5794,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.7286548777709642,
"learning_rate": 9.992134075089085e-06,
"loss": 0.542,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.6399643955662926,
"learning_rate": 9.98867437523228e-06,
"loss": 0.5781,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.5450611147899289,
"learning_rate": 9.984586668665641e-06,
"loss": 0.563,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.617884326802846,
"learning_rate": 9.979871469976197e-06,
"loss": 0.5282,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.5840918090166053,
"learning_rate": 9.974529372743762e-06,
"loss": 0.5291,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.5349911975240627,
"learning_rate": 9.968561049466214e-06,
"loss": 0.5578,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.51839396059655,
"learning_rate": 9.961967251474823e-06,
"loss": 0.5703,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.8109692188301675,
"learning_rate": 9.954748808839675e-06,
"loss": 0.5642,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.6147342989253928,
"learning_rate": 9.946906630265184e-06,
"loss": 0.5619,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.567019625664884,
"learning_rate": 9.938441702975689e-06,
"loss": 0.545,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.4953287105454508,
"learning_rate": 9.92935509259118e-06,
"loss": 0.5599,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.6016445676709195,
"learning_rate": 9.91964794299315e-06,
"loss": 0.5329,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.5908631595962877,
"learning_rate": 9.909321476180594e-06,
"loss": 0.5219,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.5652677102305135,
"learning_rate": 9.898376992116179e-06,
"loss": 0.5566,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.5439139083795868,
"learning_rate": 9.886815868562596e-06,
"loss": 0.5538,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.6677077478397748,
"learning_rate": 9.874639560909118e-06,
"loss": 0.5811,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.5523597374932737,
"learning_rate": 9.861849601988384e-06,
"loss": 0.5387,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.4171165369861334,
"learning_rate": 9.848447601883436e-06,
"loss": 0.5299,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.5280446786279971,
"learning_rate": 9.834435247725032e-06,
"loss": 0.5186,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.5716593439908328,
"learning_rate": 9.819814303479268e-06,
"loss": 0.5105,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.45849128243454307,
"learning_rate": 9.804586609725499e-06,
"loss": 0.5413,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.5318426646671739,
"learning_rate": 9.788754083424654e-06,
"loss": 0.5783,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.5214711459350322,
"learning_rate": 9.772318717677905e-06,
"loss": 0.5452,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.5323612364764054,
"learning_rate": 9.755282581475769e-06,
"loss": 0.547,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.4363510956741478,
"learning_rate": 9.737647819437645e-06,
"loss": 0.5324,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.5192909663558944,
"learning_rate": 9.719416651541839e-06,
"loss": 0.5045,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.5178487871491405,
"learning_rate": 9.700591372846096e-06,
"loss": 0.5208,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.48318786093602634,
"learning_rate": 9.681174353198687e-06,
"loss": 0.5442,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.44642576811069634,
"learning_rate": 9.661168036940071e-06,
"loss": 0.5372,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.4528303828927131,
"learning_rate": 9.640574942595195e-06,
"loss": 0.5292,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.45725722304534117,
"learning_rate": 9.619397662556434e-06,
"loss": 0.5331,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.46938129632173714,
"learning_rate": 9.597638862757255e-06,
"loss": 0.525,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.44279436276775747,
"learning_rate": 9.5753012823366e-06,
"loss": 0.5147,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.5340379664891574,
"learning_rate": 9.552387733294081e-06,
"loss": 0.5142,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.47440462325400606,
"learning_rate": 9.528901100135971e-06,
"loss": 0.5372,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.4706549828594795,
"learning_rate": 9.504844339512096e-06,
"loss": 0.5374,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.4836872053058255,
"learning_rate": 9.480220479843627e-06,
"loss": 0.5041,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.4805451555376812,
"learning_rate": 9.45503262094184e-06,
"loss": 0.5334,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.48153061086649585,
"learning_rate": 9.4292839336179e-06,
"loss": 0.5324,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.4564123471644034,
"learning_rate": 9.40297765928369e-06,
"loss": 0.5101,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.45989508234603055,
"learning_rate": 9.376117109543769e-06,
"loss": 0.5052,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.43554532242825844,
"learning_rate": 9.348705665778479e-06,
"loss": 0.5336,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.4920664902702373,
"learning_rate": 9.320746778718274e-06,
"loss": 0.5642,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.4377154200225534,
"learning_rate": 9.292243968009332e-06,
"loss": 0.5025,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.449983132094906,
"learning_rate": 9.263200821770462e-06,
"loss": 0.5143,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.48756185902516125,
"learning_rate": 9.233620996141421e-06,
"loss": 0.5276,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.4380577552900679,
"learning_rate": 9.203508214822652e-06,
"loss": 0.5452,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.44923056355002305,
"learning_rate": 9.172866268606514e-06,
"loss": 0.5232,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.4538618524086431,
"learning_rate": 9.141699014900084e-06,
"loss": 0.5275,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.44221144691670766,
"learning_rate": 9.110010377239552e-06,
"loss": 0.4884,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.4581860478381589,
"learning_rate": 9.077804344796302e-06,
"loss": 0.5079,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.44405705818130814,
"learning_rate": 9.045084971874738e-06,
"loss": 0.5192,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.4146483166679835,
"learning_rate": 9.011856377401891e-06,
"loss": 0.522,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.5142109335190371,
"learning_rate": 8.978122744408905e-06,
"loss": 0.5027,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.438873064356511,
"learning_rate": 8.943888319504456e-06,
"loss": 0.5204,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.5355977125710618,
"learning_rate": 8.90915741234015e-06,
"loss": 0.5556,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.5939525877330505,
"learning_rate": 8.873934395068006e-06,
"loss": 0.5216,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.4553088502604275,
"learning_rate": 8.838223701790057e-06,
"loss": 0.5124,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.4580259422775178,
"learning_rate": 8.802029828000157e-06,
"loss": 0.5155,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.4669945570550204,
"learning_rate": 8.765357330018056e-06,
"loss": 0.5302,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.4461089704075184,
"learning_rate": 8.728210824415829e-06,
"loss": 0.5393,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.43534225291195927,
"learning_rate": 8.690594987436705e-06,
"loss": 0.4877,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.46128857566180115,
"learning_rate": 8.652514554406388e-06,
"loss": 0.4977,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.5054507605871718,
"learning_rate": 8.613974319136959e-06,
"loss": 0.5307,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.4244849701263305,
"learning_rate": 8.574979133323378e-06,
"loss": 0.5435,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.5046344323516175,
"learning_rate": 8.535533905932739e-06,
"loss": 0.5451,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.4516628408807449,
"learning_rate": 8.495643602586287e-06,
"loss": 0.4971,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.44289189374586174,
"learning_rate": 8.455313244934324e-06,
"loss": 0.5168,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.8390630076462685,
"learning_rate": 8.414547910024035e-06,
"loss": 0.8212,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.5400734738209363,
"learning_rate": 8.373352729660373e-06,
"loss": 0.5147,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.4697979170979528,
"learning_rate": 8.331732889760021e-06,
"loss": 0.463,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.4761130079621333,
"learning_rate": 8.289693629698564e-06,
"loss": 0.5075,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.41129736663353383,
"learning_rate": 8.247240241650918e-06,
"loss": 0.4537,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.5484873390314843,
"learning_rate": 8.204378069925121e-06,
"loss": 0.487,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.5003598615952572,
"learning_rate": 8.16111251028955e-06,
"loss": 0.499,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.451679239878114,
"learning_rate": 8.117449009293668e-06,
"loss": 0.4675,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.4680619668865558,
"learning_rate": 8.073393063582386e-06,
"loss": 0.487,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.4583142078805796,
"learning_rate": 8.0289502192041e-06,
"loss": 0.4809,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.44354454381350605,
"learning_rate": 7.984126070912519e-06,
"loss": 0.4797,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.356596712133377,
"learning_rate": 7.938926261462366e-06,
"loss": 0.4553,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.5064232405227002,
"learning_rate": 7.89335648089903e-06,
"loss": 0.5246,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.4295523351208256,
"learning_rate": 7.84742246584226e-06,
"loss": 0.5059,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.45566522336720117,
"learning_rate": 7.801129998764014e-06,
"loss": 0.4389,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.47560827727981786,
"learning_rate": 7.754484907260513e-06,
"loss": 0.5038,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.42649981368791234,
"learning_rate": 7.70749306331863e-06,
"loss": 0.4856,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.49206190561165525,
"learning_rate": 7.660160382576683e-06,
"loss": 0.4565,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.4539863272125339,
"learning_rate": 7.612492823579744e-06,
"loss": 0.4615,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.4020138020329491,
"learning_rate": 7.564496387029532e-06,
"loss": 0.4716,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.42475576131519654,
"learning_rate": 7.516177115029002e-06,
"loss": 0.5057,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.40988551818922375,
"learning_rate": 7.467541090321735e-06,
"loss": 0.4624,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.48126905147229293,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.48,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.3638269900541516,
"learning_rate": 7.369343312364994e-06,
"loss": 0.4534,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.36866860753975594,
"learning_rate": 7.319793920889171e-06,
"loss": 0.4668,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.4439824336459749,
"learning_rate": 7.269952498697734e-06,
"loss": 0.4722,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.4347005501645301,
"learning_rate": 7.219825320152411e-06,
"loss": 0.4618,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.41295647546044356,
"learning_rate": 7.169418695587791e-06,
"loss": 0.4765,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.44610189289809943,
"learning_rate": 7.118738970516944e-06,
"loss": 0.454,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.4262764899323212,
"learning_rate": 7.067792524832604e-06,
"loss": 0.4903,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.40269113880980567,
"learning_rate": 7.016585772004026e-06,
"loss": 0.4955,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.42949137944112714,
"learning_rate": 6.965125158269619e-06,
"loss": 0.4592,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.3971832787127952,
"learning_rate": 6.913417161825449e-06,
"loss": 0.4769,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.42706894747315516,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.4825,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.37622008658478756,
"learning_rate": 6.809285088483361e-06,
"loss": 0.4148,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.4512819589988797,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.5317,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.3357834141811991,
"learning_rate": 6.704241985612625e-06,
"loss": 0.4046,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.41497879631319207,
"learning_rate": 6.651395309775837e-06,
"loss": 0.5236,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.3766914951069415,
"learning_rate": 6.598340745578908e-06,
"loss": 0.4589,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.40289608234241914,
"learning_rate": 6.545084971874738e-06,
"loss": 0.4665,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.4407104768846786,
"learning_rate": 6.491634692845781e-06,
"loss": 0.4966,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.3733212661834858,
"learning_rate": 6.437996637160086e-06,
"loss": 0.4549,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.42693563095719544,
"learning_rate": 6.384177557124247e-06,
"loss": 0.4918,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.34906774289947445,
"learning_rate": 6.330184227833376e-06,
"loss": 0.4557,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.4286903426024423,
"learning_rate": 6.276023446318214e-06,
"loss": 0.4714,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.39525622917708014,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.4373,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.38091233728988383,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.4537,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.41099078733781397,
"learning_rate": 6.112604669781572e-06,
"loss": 0.4812,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.4062062686056688,
"learning_rate": 6.057842458386315e-06,
"loss": 0.4647,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.38402739439580924,
"learning_rate": 6.002947078916365e-06,
"loss": 0.4607,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.3809246909640802,
"learning_rate": 5.947925441958393e-06,
"loss": 0.4657,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.42830874041082445,
"learning_rate": 5.892784473993184e-06,
"loss": 0.4544,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.38494534295058547,
"learning_rate": 5.837531116523683e-06,
"loss": 0.5099,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.37656045947612327,
"learning_rate": 5.782172325201155e-06,
"loss": 0.4625,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.4100777640712403,
"learning_rate": 5.726715068949564e-06,
"loss": 0.4555,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.3784629821633463,
"learning_rate": 5.671166329088278e-06,
"loss": 0.4564,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.43014870440926317,
"learning_rate": 5.615533098453215e-06,
"loss": 0.4739,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.3692585194037088,
"learning_rate": 5.559822380516539e-06,
"loss": 0.4535,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.398883626029488,
"learning_rate": 5.504041188505022e-06,
"loss": 0.4459,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.3770970450675236,
"learning_rate": 5.448196544517168e-06,
"loss": 0.4796,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.4135848032232121,
"learning_rate": 5.392295478639226e-06,
"loss": 0.4888,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.3826566799869738,
"learning_rate": 5.336345028060199e-06,
"loss": 0.457,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.4048333976493676,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.4909,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.41111108004219327,
"learning_rate": 5.224324151752575e-06,
"loss": 0.4872,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.3855082499889585,
"learning_rate": 5.168267827938971e-06,
"loss": 0.472,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.3554732146526522,
"learning_rate": 5.112190321479026e-06,
"loss": 0.4735,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.46672855289753856,
"learning_rate": 5.05609869177323e-06,
"loss": 0.4788,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.3825673822619758,
"learning_rate": 5e-06,
"loss": 0.437,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.39070539372615254,
"learning_rate": 4.943901308226771e-06,
"loss": 0.4995,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.39163210255200964,
"learning_rate": 4.887809678520976e-06,
"loss": 0.4655,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.3941515972591125,
"learning_rate": 4.831732172061032e-06,
"loss": 0.4415,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.39290747495729883,
"learning_rate": 4.775675848247427e-06,
"loss": 0.5085,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.3857601625654789,
"learning_rate": 4.719647763814041e-06,
"loss": 0.4774,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.38481645443193224,
"learning_rate": 4.663654971939802e-06,
"loss": 0.4992,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.34651704553815316,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.4303,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.35017519326595875,
"learning_rate": 4.551803455482833e-06,
"loss": 0.4481,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.38239804367141184,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.4761,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.37774880386113624,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.4857,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.37874747443035606,
"learning_rate": 4.384466901546786e-06,
"loss": 0.475,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.3771003831572188,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.4637,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.36733752406572745,
"learning_rate": 4.273284931050438e-06,
"loss": 0.4744,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.3710249259592041,
"learning_rate": 4.217827674798845e-06,
"loss": 0.4732,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.3380084120342326,
"learning_rate": 4.162468883476319e-06,
"loss": 0.4689,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.3646580392640309,
"learning_rate": 4.107215526006818e-06,
"loss": 0.4563,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.3889956131963747,
"learning_rate": 4.052074558041608e-06,
"loss": 0.468,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.3689674242656601,
"learning_rate": 3.997052921083637e-06,
"loss": 0.4558,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.36384058878217945,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.4732,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.37483910236221657,
"learning_rate": 3.887395330218429e-06,
"loss": 0.4676,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.32663096863266505,
"learning_rate": 3.832773180720475e-06,
"loss": 0.4503,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.38548410227297764,
"learning_rate": 3.778297969310529e-06,
"loss": 0.4734,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.4031775178190752,
"learning_rate": 3.723976553681787e-06,
"loss": 0.4644,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.3614330140662186,
"learning_rate": 3.669815772166625e-06,
"loss": 0.4421,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.3693804220370989,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.461,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.3624094029713892,
"learning_rate": 3.562003362839914e-06,
"loss": 0.4538,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.34164249561672144,
"learning_rate": 3.50836530715422e-06,
"loss": 0.4759,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.3849506500642619,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.4893,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.3638839131466772,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.5331,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.3368889298501968,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.4231,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.38053335340125116,
"learning_rate": 3.295758014387375e-06,
"loss": 0.458,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.38423028847150537,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.5021,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.4065770090089661,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.4666,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.33386690427880855,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.4245,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.3638100237179528,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.5228,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.372182458712491,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.452,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.8736804071574767,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.7553,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.3817438966363151,
"learning_rate": 2.932207475167398e-06,
"loss": 0.439,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.37256472735848656,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.4201,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.4189989056856489,
"learning_rate": 2.83058130441221e-06,
"loss": 0.4403,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.41085626942381753,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.4257,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.4017437968831196,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.4814,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.35100875773519224,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.4192,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.419853825313074,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.4745,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.3824793477734069,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.4126,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.3620315253856857,
"learning_rate": 2.532458909678266e-06,
"loss": 0.4108,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.34002506173757857,
"learning_rate": 2.483822884971e-06,
"loss": 0.4361,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.3472392939730272,
"learning_rate": 2.43550361297047e-06,
"loss": 0.4589,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.3436442030182686,
"learning_rate": 2.387507176420256e-06,
"loss": 0.4226,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.3610174485497869,
"learning_rate": 2.339839617423318e-06,
"loss": 0.4247,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.40007490957341496,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.4689,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.3245352194112737,
"learning_rate": 2.245515092739488e-06,
"loss": 0.3744,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.36316460331991257,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.4915,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.3311112124777016,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.4004,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.3785765804837821,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.4747,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.31834343972432916,
"learning_rate": 2.061073738537635e-06,
"loss": 0.4299,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.3458553992955153,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.4668,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.36658370995927336,
"learning_rate": 1.971049780795901e-06,
"loss": 0.438,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.36438688999529567,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.4681,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.3786998128899529,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.4499,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.32131883547899753,
"learning_rate": 1.838887489710452e-06,
"loss": 0.3835,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.3382048118958599,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.4125,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.33776439068892333,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.4245,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.3429170347972964,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.4129,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.34077892562419876,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.475,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.3479902951665374,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.4661,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.31868234028331355,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.4025,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.3958434061564218,
"learning_rate": 1.544686755065677e-06,
"loss": 0.4695,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.33368979825061673,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.4259,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.3518344473791682,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.4557,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.3484117977456733,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.4422,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.3152773034224358,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.4024,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.3137021817873724,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.4136,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.3542042973538173,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.4279,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.33964339581216657,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.3795,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.3151131635307428,
"learning_rate": 1.234642669981946e-06,
"loss": 0.4425,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.330671622019116,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.4407,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.3113481102759853,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.4214,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.3365661589134693,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.4644,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.3403105867561746,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.4208,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.34251654207818516,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.4555,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.3178171974549021,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.4147,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.3130631549769216,
"learning_rate": 9.881436225981107e-07,
"loss": 0.4178,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.30922082983608534,
"learning_rate": 9.549150281252633e-07,
"loss": 0.4433,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.31687438968039944,
"learning_rate": 9.221956552036992e-07,
"loss": 0.4622,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.3507597399301423,
"learning_rate": 8.899896227604509e-07,
"loss": 0.424,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.3230652672131586,
"learning_rate": 8.58300985099918e-07,
"loss": 0.4062,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.3201645081924483,
"learning_rate": 8.271337313934869e-07,
"loss": 0.4414,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.3364316373242383,
"learning_rate": 7.964917851773496e-07,
"loss": 0.3993,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.3173389284350239,
"learning_rate": 7.663790038585794e-07,
"loss": 0.4328,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.3205198719050877,
"learning_rate": 7.367991782295392e-07,
"loss": 0.4348,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.3272319740109694,
"learning_rate": 7.077560319906696e-07,
"loss": 0.4569,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.3233194352344347,
"learning_rate": 6.792532212817271e-07,
"loss": 0.4453,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.3158270218890542,
"learning_rate": 6.512943342215234e-07,
"loss": 0.411,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.32645433619438335,
"learning_rate": 6.238828904562316e-07,
"loss": 0.3993,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.3435642184896803,
"learning_rate": 5.9702234071631e-07,
"loss": 0.4594,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.32515207990243716,
"learning_rate": 5.707160663821009e-07,
"loss": 0.4441,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.32638466412222183,
"learning_rate": 5.449673790581611e-07,
"loss": 0.4228,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.30412496698307606,
"learning_rate": 5.197795201563744e-07,
"loss": 0.4092,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.301176395549818,
"learning_rate": 4.951556604879049e-07,
"loss": 0.4137,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.33909930292661994,
"learning_rate": 4.710988998640298e-07,
"loss": 0.4425,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.30734446384712466,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.4516,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.30884613503314146,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.4209,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.3206146560053614,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.407,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.3326175566683035,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.4381,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.3287180597990721,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.4339,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.3123380745565648,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.4254,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.3319138655850274,
"learning_rate": 3.18825646801314e-07,
"loss": 0.4441,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.32080207290138363,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.4311,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.31747337463979347,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.4414,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.2877252479808375,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.4236,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.3244669226907146,
"learning_rate": 2.447174185242324e-07,
"loss": 0.4604,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.3281677844105852,
"learning_rate": 2.276812823220964e-07,
"loss": 0.3892,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.3172652188139629,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.462,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.3263782841520062,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.4321,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.3034614465818679,
"learning_rate": 1.801856965207338e-07,
"loss": 0.3963,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.4567496587275747,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.4193,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.33071907590705435,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.4479,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.3150662983134378,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.4045,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.3005624056657451,
"learning_rate": 1.253604390908819e-07,
"loss": 0.384,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.35309209620213133,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.483,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.3107378909890752,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.4147,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.31227258120302026,
"learning_rate": 9.0678523819408e-08,
"loss": 0.3937,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.32100905902813515,
"learning_rate": 8.035205700685167e-08,
"loss": 0.4567,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.3134685730739442,
"learning_rate": 7.064490740882057e-08,
"loss": 0.4206,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.30730491229563306,
"learning_rate": 6.15582970243117e-08,
"loss": 0.4292,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.3301265005689028,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.4813,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.3229907557435683,
"learning_rate": 4.52511911603265e-08,
"loss": 0.4648,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.3052686914879135,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.3767,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.2956196678493814,
"learning_rate": 3.143895053378698e-08,
"loss": 0.4534,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.30192590339023084,
"learning_rate": 2.547062725623828e-08,
"loss": 0.3992,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.31197717952726994,
"learning_rate": 2.012853002380466e-08,
"loss": 0.4262,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.3240864253326664,
"learning_rate": 1.541333133436018e-08,
"loss": 0.459,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.32655239358176297,
"learning_rate": 1.132562476771959e-08,
"loss": 0.451,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.30842483537006643,
"learning_rate": 7.865924910916977e-09,
"loss": 0.4189,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.32928312247676134,
"learning_rate": 5.034667293427053e-09,
"loss": 0.3956,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.2936408752359236,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.4033,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.31673803021283686,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.4176,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.29092686913094873,
"learning_rate": 3.147162264971471e-10,
"loss": 0.3897,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.32443894231187104,
"learning_rate": 0.0,
"loss": 0.4432,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 401928723890176.0,
"train_loss": 0.49936966349681217,
"train_runtime": 5327.0055,
"train_samples_per_second": 5.632,
"train_steps_per_second": 0.059
}
],
"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": 401928723890176.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}