dedup_ablation_sim_threshold_10 / trainer_state.json
sedrickkeh's picture
End of training
5ee8204 verified
raw
history blame
51.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9952,
"eval_steps": 500,
"global_step": 312,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0096,
"grad_norm": 5.5126953125,
"learning_rate": 3.125e-07,
"loss": 0.9243,
"step": 1
},
{
"epoch": 0.0192,
"grad_norm": 5.716032981872559,
"learning_rate": 6.25e-07,
"loss": 0.9516,
"step": 2
},
{
"epoch": 0.0288,
"grad_norm": 5.74594783782959,
"learning_rate": 9.375000000000001e-07,
"loss": 0.9504,
"step": 3
},
{
"epoch": 0.0384,
"grad_norm": 5.5603346824646,
"learning_rate": 1.25e-06,
"loss": 0.9333,
"step": 4
},
{
"epoch": 0.048,
"grad_norm": 5.4722819328308105,
"learning_rate": 1.5625e-06,
"loss": 0.9205,
"step": 5
},
{
"epoch": 0.0576,
"grad_norm": 5.219153881072998,
"learning_rate": 1.8750000000000003e-06,
"loss": 0.8936,
"step": 6
},
{
"epoch": 0.0672,
"grad_norm": 4.2324910163879395,
"learning_rate": 2.1875000000000002e-06,
"loss": 0.9097,
"step": 7
},
{
"epoch": 0.0768,
"grad_norm": 3.7411277294158936,
"learning_rate": 2.5e-06,
"loss": 0.8786,
"step": 8
},
{
"epoch": 0.0864,
"grad_norm": 2.112518310546875,
"learning_rate": 2.8125e-06,
"loss": 0.8326,
"step": 9
},
{
"epoch": 0.096,
"grad_norm": 2.0676801204681396,
"learning_rate": 3.125e-06,
"loss": 0.8372,
"step": 10
},
{
"epoch": 0.1056,
"grad_norm": 1.8498632907867432,
"learning_rate": 3.4375e-06,
"loss": 0.8214,
"step": 11
},
{
"epoch": 0.1152,
"grad_norm": 3.1311442852020264,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.8228,
"step": 12
},
{
"epoch": 0.1248,
"grad_norm": 3.837650775909424,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.8353,
"step": 13
},
{
"epoch": 0.1344,
"grad_norm": 3.9779882431030273,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.8535,
"step": 14
},
{
"epoch": 0.144,
"grad_norm": 3.5275862216949463,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.8375,
"step": 15
},
{
"epoch": 0.1536,
"grad_norm": 2.5623204708099365,
"learning_rate": 5e-06,
"loss": 0.7765,
"step": 16
},
{
"epoch": 0.1632,
"grad_norm": 2.1568212509155273,
"learning_rate": 5.3125e-06,
"loss": 0.794,
"step": 17
},
{
"epoch": 0.1728,
"grad_norm": 1.8221849203109741,
"learning_rate": 5.625e-06,
"loss": 0.7912,
"step": 18
},
{
"epoch": 0.1824,
"grad_norm": 1.5951319932937622,
"learning_rate": 5.9375e-06,
"loss": 0.7577,
"step": 19
},
{
"epoch": 0.192,
"grad_norm": 1.4769177436828613,
"learning_rate": 6.25e-06,
"loss": 0.7431,
"step": 20
},
{
"epoch": 0.2016,
"grad_norm": 1.240143060684204,
"learning_rate": 6.5625e-06,
"loss": 0.7311,
"step": 21
},
{
"epoch": 0.2112,
"grad_norm": 1.276717185974121,
"learning_rate": 6.875e-06,
"loss": 0.754,
"step": 22
},
{
"epoch": 0.2208,
"grad_norm": 1.178741216659546,
"learning_rate": 7.1875e-06,
"loss": 0.7501,
"step": 23
},
{
"epoch": 0.2304,
"grad_norm": 1.0051218271255493,
"learning_rate": 7.500000000000001e-06,
"loss": 0.7007,
"step": 24
},
{
"epoch": 0.24,
"grad_norm": 1.0521109104156494,
"learning_rate": 7.8125e-06,
"loss": 0.6908,
"step": 25
},
{
"epoch": 0.2496,
"grad_norm": 1.020888328552246,
"learning_rate": 8.125000000000001e-06,
"loss": 0.7001,
"step": 26
},
{
"epoch": 0.2592,
"grad_norm": 0.9082687497138977,
"learning_rate": 8.4375e-06,
"loss": 0.6936,
"step": 27
},
{
"epoch": 0.2688,
"grad_norm": 0.8129175901412964,
"learning_rate": 8.750000000000001e-06,
"loss": 0.6863,
"step": 28
},
{
"epoch": 0.2784,
"grad_norm": 0.7594968676567078,
"learning_rate": 9.0625e-06,
"loss": 0.6994,
"step": 29
},
{
"epoch": 0.288,
"grad_norm": 0.7986474633216858,
"learning_rate": 9.375000000000001e-06,
"loss": 0.66,
"step": 30
},
{
"epoch": 0.2976,
"grad_norm": 0.7827369570732117,
"learning_rate": 9.6875e-06,
"loss": 0.71,
"step": 31
},
{
"epoch": 0.3072,
"grad_norm": 0.6687161326408386,
"learning_rate": 1e-05,
"loss": 0.6784,
"step": 32
},
{
"epoch": 0.3168,
"grad_norm": 0.655497670173645,
"learning_rate": 9.999685283773504e-06,
"loss": 0.6894,
"step": 33
},
{
"epoch": 0.3264,
"grad_norm": 0.6853446364402771,
"learning_rate": 9.998741174712534e-06,
"loss": 0.6678,
"step": 34
},
{
"epoch": 0.336,
"grad_norm": 0.5890143513679504,
"learning_rate": 9.997167791667668e-06,
"loss": 0.676,
"step": 35
},
{
"epoch": 0.3456,
"grad_norm": 0.6678131818771362,
"learning_rate": 9.994965332706574e-06,
"loss": 0.6666,
"step": 36
},
{
"epoch": 0.3552,
"grad_norm": 0.6165468096733093,
"learning_rate": 9.992134075089085e-06,
"loss": 0.6752,
"step": 37
},
{
"epoch": 0.3648,
"grad_norm": 0.6728239059448242,
"learning_rate": 9.98867437523228e-06,
"loss": 0.701,
"step": 38
},
{
"epoch": 0.3744,
"grad_norm": 0.4962407946586609,
"learning_rate": 9.984586668665641e-06,
"loss": 0.6734,
"step": 39
},
{
"epoch": 0.384,
"grad_norm": 0.6008801460266113,
"learning_rate": 9.979871469976197e-06,
"loss": 0.6707,
"step": 40
},
{
"epoch": 0.3936,
"grad_norm": 0.6295132637023926,
"learning_rate": 9.974529372743762e-06,
"loss": 0.6454,
"step": 41
},
{
"epoch": 0.4032,
"grad_norm": 0.5358794331550598,
"learning_rate": 9.968561049466214e-06,
"loss": 0.6553,
"step": 42
},
{
"epoch": 0.4128,
"grad_norm": 0.5187429785728455,
"learning_rate": 9.961967251474823e-06,
"loss": 0.6392,
"step": 43
},
{
"epoch": 0.4224,
"grad_norm": 0.6720361709594727,
"learning_rate": 9.954748808839675e-06,
"loss": 0.6466,
"step": 44
},
{
"epoch": 0.432,
"grad_norm": 0.5458486676216125,
"learning_rate": 9.946906630265184e-06,
"loss": 0.678,
"step": 45
},
{
"epoch": 0.4416,
"grad_norm": 0.6491691470146179,
"learning_rate": 9.938441702975689e-06,
"loss": 0.6482,
"step": 46
},
{
"epoch": 0.4512,
"grad_norm": 0.6741710305213928,
"learning_rate": 9.92935509259118e-06,
"loss": 0.6289,
"step": 47
},
{
"epoch": 0.4608,
"grad_norm": 0.5362808704376221,
"learning_rate": 9.91964794299315e-06,
"loss": 0.6569,
"step": 48
},
{
"epoch": 0.4704,
"grad_norm": 0.52982497215271,
"learning_rate": 9.909321476180594e-06,
"loss": 0.672,
"step": 49
},
{
"epoch": 0.48,
"grad_norm": 0.5318625569343567,
"learning_rate": 9.898376992116179e-06,
"loss": 0.6555,
"step": 50
},
{
"epoch": 0.4896,
"grad_norm": 0.5790548324584961,
"learning_rate": 9.886815868562596e-06,
"loss": 0.6607,
"step": 51
},
{
"epoch": 0.4992,
"grad_norm": 0.5010600686073303,
"learning_rate": 9.874639560909118e-06,
"loss": 0.637,
"step": 52
},
{
"epoch": 0.5088,
"grad_norm": 0.5037438273429871,
"learning_rate": 9.861849601988384e-06,
"loss": 0.6501,
"step": 53
},
{
"epoch": 0.5184,
"grad_norm": 0.4869060218334198,
"learning_rate": 9.848447601883436e-06,
"loss": 0.6384,
"step": 54
},
{
"epoch": 0.528,
"grad_norm": 0.5368464589118958,
"learning_rate": 9.834435247725032e-06,
"loss": 0.6459,
"step": 55
},
{
"epoch": 0.5376,
"grad_norm": 0.5480874180793762,
"learning_rate": 9.819814303479268e-06,
"loss": 0.6672,
"step": 56
},
{
"epoch": 0.5472,
"grad_norm": 0.4623332917690277,
"learning_rate": 9.804586609725499e-06,
"loss": 0.6531,
"step": 57
},
{
"epoch": 0.5568,
"grad_norm": 0.6133888959884644,
"learning_rate": 9.788754083424654e-06,
"loss": 0.6533,
"step": 58
},
{
"epoch": 0.5664,
"grad_norm": 0.5914374589920044,
"learning_rate": 9.772318717677905e-06,
"loss": 0.6575,
"step": 59
},
{
"epoch": 0.576,
"grad_norm": 0.5839924216270447,
"learning_rate": 9.755282581475769e-06,
"loss": 0.6683,
"step": 60
},
{
"epoch": 0.5856,
"grad_norm": 0.5651989579200745,
"learning_rate": 9.737647819437645e-06,
"loss": 0.6473,
"step": 61
},
{
"epoch": 0.5952,
"grad_norm": 0.6176101565361023,
"learning_rate": 9.719416651541839e-06,
"loss": 0.6355,
"step": 62
},
{
"epoch": 0.6048,
"grad_norm": 0.48143985867500305,
"learning_rate": 9.700591372846096e-06,
"loss": 0.6384,
"step": 63
},
{
"epoch": 0.6144,
"grad_norm": 0.6189300417900085,
"learning_rate": 9.681174353198687e-06,
"loss": 0.6722,
"step": 64
},
{
"epoch": 0.624,
"grad_norm": 0.5594016909599304,
"learning_rate": 9.661168036940071e-06,
"loss": 0.6574,
"step": 65
},
{
"epoch": 0.6336,
"grad_norm": 0.48622775077819824,
"learning_rate": 9.640574942595195e-06,
"loss": 0.6522,
"step": 66
},
{
"epoch": 0.6432,
"grad_norm": 0.6044607162475586,
"learning_rate": 9.619397662556434e-06,
"loss": 0.6514,
"step": 67
},
{
"epoch": 0.6528,
"grad_norm": 0.5377262234687805,
"learning_rate": 9.597638862757255e-06,
"loss": 0.6505,
"step": 68
},
{
"epoch": 0.6624,
"grad_norm": 0.4681186378002167,
"learning_rate": 9.5753012823366e-06,
"loss": 0.6263,
"step": 69
},
{
"epoch": 0.672,
"grad_norm": 0.6148974299430847,
"learning_rate": 9.552387733294081e-06,
"loss": 0.6015,
"step": 70
},
{
"epoch": 0.6816,
"grad_norm": 0.6111718416213989,
"learning_rate": 9.528901100135971e-06,
"loss": 0.6129,
"step": 71
},
{
"epoch": 0.6912,
"grad_norm": 0.47291842103004456,
"learning_rate": 9.504844339512096e-06,
"loss": 0.6538,
"step": 72
},
{
"epoch": 0.7008,
"grad_norm": 0.5861859917640686,
"learning_rate": 9.480220479843627e-06,
"loss": 0.6008,
"step": 73
},
{
"epoch": 0.7104,
"grad_norm": 0.5189301371574402,
"learning_rate": 9.45503262094184e-06,
"loss": 0.6322,
"step": 74
},
{
"epoch": 0.72,
"grad_norm": 0.515916645526886,
"learning_rate": 9.4292839336179e-06,
"loss": 0.642,
"step": 75
},
{
"epoch": 0.7296,
"grad_norm": 0.4605845808982849,
"learning_rate": 9.40297765928369e-06,
"loss": 0.6362,
"step": 76
},
{
"epoch": 0.7392,
"grad_norm": 0.4964100122451782,
"learning_rate": 9.376117109543769e-06,
"loss": 0.6282,
"step": 77
},
{
"epoch": 0.7488,
"grad_norm": 0.5005691647529602,
"learning_rate": 9.348705665778479e-06,
"loss": 0.6414,
"step": 78
},
{
"epoch": 0.7584,
"grad_norm": 0.46636098623275757,
"learning_rate": 9.320746778718274e-06,
"loss": 0.6186,
"step": 79
},
{
"epoch": 0.768,
"grad_norm": 0.4857327938079834,
"learning_rate": 9.292243968009332e-06,
"loss": 0.645,
"step": 80
},
{
"epoch": 0.7776,
"grad_norm": 0.4664401113986969,
"learning_rate": 9.263200821770462e-06,
"loss": 0.6276,
"step": 81
},
{
"epoch": 0.7872,
"grad_norm": 0.48841968178749084,
"learning_rate": 9.233620996141421e-06,
"loss": 0.6276,
"step": 82
},
{
"epoch": 0.7968,
"grad_norm": 0.4317878782749176,
"learning_rate": 9.203508214822652e-06,
"loss": 0.6387,
"step": 83
},
{
"epoch": 0.8064,
"grad_norm": 0.5166445970535278,
"learning_rate": 9.172866268606514e-06,
"loss": 0.6695,
"step": 84
},
{
"epoch": 0.816,
"grad_norm": 0.49552348256111145,
"learning_rate": 9.141699014900084e-06,
"loss": 0.6065,
"step": 85
},
{
"epoch": 0.8256,
"grad_norm": 0.4629752039909363,
"learning_rate": 9.110010377239552e-06,
"loss": 0.6357,
"step": 86
},
{
"epoch": 0.8352,
"grad_norm": 0.5324919819831848,
"learning_rate": 9.077804344796302e-06,
"loss": 0.6426,
"step": 87
},
{
"epoch": 0.8448,
"grad_norm": 0.4791981279850006,
"learning_rate": 9.045084971874738e-06,
"loss": 0.6071,
"step": 88
},
{
"epoch": 0.8544,
"grad_norm": 0.4232598543167114,
"learning_rate": 9.011856377401891e-06,
"loss": 0.63,
"step": 89
},
{
"epoch": 0.864,
"grad_norm": 0.46060553193092346,
"learning_rate": 8.978122744408905e-06,
"loss": 0.6498,
"step": 90
},
{
"epoch": 0.8736,
"grad_norm": 0.4134940207004547,
"learning_rate": 8.943888319504456e-06,
"loss": 0.5998,
"step": 91
},
{
"epoch": 0.8832,
"grad_norm": 0.4399256408214569,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6178,
"step": 92
},
{
"epoch": 0.8928,
"grad_norm": 0.43876218795776367,
"learning_rate": 8.873934395068006e-06,
"loss": 0.6172,
"step": 93
},
{
"epoch": 0.9024,
"grad_norm": 0.4813985824584961,
"learning_rate": 8.838223701790057e-06,
"loss": 0.6378,
"step": 94
},
{
"epoch": 0.912,
"grad_norm": 0.4228210747241974,
"learning_rate": 8.802029828000157e-06,
"loss": 0.6287,
"step": 95
},
{
"epoch": 0.9216,
"grad_norm": 0.42882269620895386,
"learning_rate": 8.765357330018056e-06,
"loss": 0.6357,
"step": 96
},
{
"epoch": 0.9312,
"grad_norm": 0.4681991636753082,
"learning_rate": 8.728210824415829e-06,
"loss": 0.65,
"step": 97
},
{
"epoch": 0.9408,
"grad_norm": 0.49026817083358765,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6235,
"step": 98
},
{
"epoch": 0.9504,
"grad_norm": 0.49928873777389526,
"learning_rate": 8.652514554406388e-06,
"loss": 0.622,
"step": 99
},
{
"epoch": 0.96,
"grad_norm": 0.4295845329761505,
"learning_rate": 8.613974319136959e-06,
"loss": 0.6338,
"step": 100
},
{
"epoch": 0.9696,
"grad_norm": 0.41574129462242126,
"learning_rate": 8.574979133323378e-06,
"loss": 0.6315,
"step": 101
},
{
"epoch": 0.9792,
"grad_norm": 0.48868778347969055,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6188,
"step": 102
},
{
"epoch": 0.9888,
"grad_norm": 0.42378732562065125,
"learning_rate": 8.495643602586287e-06,
"loss": 0.6346,
"step": 103
},
{
"epoch": 0.9984,
"grad_norm": 0.4838070869445801,
"learning_rate": 8.455313244934324e-06,
"loss": 0.6316,
"step": 104
},
{
"epoch": 1.008,
"grad_norm": 1.09999680519104,
"learning_rate": 8.414547910024035e-06,
"loss": 1.1071,
"step": 105
},
{
"epoch": 1.0176,
"grad_norm": 0.5700429081916809,
"learning_rate": 8.373352729660373e-06,
"loss": 0.5876,
"step": 106
},
{
"epoch": 1.0272,
"grad_norm": 0.5152708292007446,
"learning_rate": 8.331732889760021e-06,
"loss": 0.5217,
"step": 107
},
{
"epoch": 1.0368,
"grad_norm": 0.4970182180404663,
"learning_rate": 8.289693629698564e-06,
"loss": 0.5914,
"step": 108
},
{
"epoch": 1.0464,
"grad_norm": 0.6691729426383972,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6299,
"step": 109
},
{
"epoch": 1.056,
"grad_norm": 0.4859987497329712,
"learning_rate": 8.204378069925121e-06,
"loss": 0.5391,
"step": 110
},
{
"epoch": 1.0656,
"grad_norm": 0.5490798950195312,
"learning_rate": 8.16111251028955e-06,
"loss": 0.594,
"step": 111
},
{
"epoch": 1.0752,
"grad_norm": 0.679706871509552,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6051,
"step": 112
},
{
"epoch": 1.0848,
"grad_norm": 0.45486193895339966,
"learning_rate": 8.073393063582386e-06,
"loss": 0.548,
"step": 113
},
{
"epoch": 1.0944,
"grad_norm": 0.6088062524795532,
"learning_rate": 8.0289502192041e-06,
"loss": 0.5724,
"step": 114
},
{
"epoch": 1.104,
"grad_norm": 0.5828673243522644,
"learning_rate": 7.984126070912519e-06,
"loss": 0.5814,
"step": 115
},
{
"epoch": 1.1136,
"grad_norm": 0.41168221831321716,
"learning_rate": 7.938926261462366e-06,
"loss": 0.578,
"step": 116
},
{
"epoch": 1.1232,
"grad_norm": 0.5625458359718323,
"learning_rate": 7.89335648089903e-06,
"loss": 0.587,
"step": 117
},
{
"epoch": 1.1328,
"grad_norm": 0.4607331156730652,
"learning_rate": 7.84742246584226e-06,
"loss": 0.5643,
"step": 118
},
{
"epoch": 1.1424,
"grad_norm": 0.4779949188232422,
"learning_rate": 7.801129998764014e-06,
"loss": 0.5876,
"step": 119
},
{
"epoch": 1.152,
"grad_norm": 0.5013940930366516,
"learning_rate": 7.754484907260513e-06,
"loss": 0.5953,
"step": 120
},
{
"epoch": 1.1616,
"grad_norm": 0.41801536083221436,
"learning_rate": 7.70749306331863e-06,
"loss": 0.5568,
"step": 121
},
{
"epoch": 1.1712,
"grad_norm": 0.45169973373413086,
"learning_rate": 7.660160382576683e-06,
"loss": 0.5861,
"step": 122
},
{
"epoch": 1.1808,
"grad_norm": 0.48225757479667664,
"learning_rate": 7.612492823579744e-06,
"loss": 0.5618,
"step": 123
},
{
"epoch": 1.1904,
"grad_norm": 0.4255233108997345,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6188,
"step": 124
},
{
"epoch": 1.2,
"grad_norm": 0.5242148041725159,
"learning_rate": 7.516177115029002e-06,
"loss": 0.6036,
"step": 125
},
{
"epoch": 1.2096,
"grad_norm": 0.414063423871994,
"learning_rate": 7.467541090321735e-06,
"loss": 0.5282,
"step": 126
},
{
"epoch": 1.2192,
"grad_norm": 0.43281540274620056,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.5724,
"step": 127
},
{
"epoch": 1.2288000000000001,
"grad_norm": 0.6164405345916748,
"learning_rate": 7.369343312364994e-06,
"loss": 0.5786,
"step": 128
},
{
"epoch": 1.2384,
"grad_norm": 0.4729873836040497,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5656,
"step": 129
},
{
"epoch": 1.248,
"grad_norm": 0.3974047303199768,
"learning_rate": 7.269952498697734e-06,
"loss": 0.5664,
"step": 130
},
{
"epoch": 1.2576,
"grad_norm": 0.5228157639503479,
"learning_rate": 7.219825320152411e-06,
"loss": 0.6072,
"step": 131
},
{
"epoch": 1.2671999999999999,
"grad_norm": 0.4351467192173004,
"learning_rate": 7.169418695587791e-06,
"loss": 0.5715,
"step": 132
},
{
"epoch": 1.2768,
"grad_norm": 0.42677679657936096,
"learning_rate": 7.118738970516944e-06,
"loss": 0.5683,
"step": 133
},
{
"epoch": 1.2864,
"grad_norm": 0.4629952311515808,
"learning_rate": 7.067792524832604e-06,
"loss": 0.5773,
"step": 134
},
{
"epoch": 1.296,
"grad_norm": 0.512144923210144,
"learning_rate": 7.016585772004026e-06,
"loss": 0.5741,
"step": 135
},
{
"epoch": 1.3056,
"grad_norm": 0.43901005387306213,
"learning_rate": 6.965125158269619e-06,
"loss": 0.5855,
"step": 136
},
{
"epoch": 1.3152,
"grad_norm": 0.43872442841529846,
"learning_rate": 6.913417161825449e-06,
"loss": 0.586,
"step": 137
},
{
"epoch": 1.3248,
"grad_norm": 0.4614115357398987,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.5578,
"step": 138
},
{
"epoch": 1.3344,
"grad_norm": 0.39648428559303284,
"learning_rate": 6.809285088483361e-06,
"loss": 0.5868,
"step": 139
},
{
"epoch": 1.3439999999999999,
"grad_norm": 0.42903226613998413,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.5604,
"step": 140
},
{
"epoch": 1.3536000000000001,
"grad_norm": 0.43323439359664917,
"learning_rate": 6.704241985612625e-06,
"loss": 0.5594,
"step": 141
},
{
"epoch": 1.3632,
"grad_norm": 0.44281768798828125,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6009,
"step": 142
},
{
"epoch": 1.3728,
"grad_norm": 0.42851829528808594,
"learning_rate": 6.598340745578908e-06,
"loss": 0.5857,
"step": 143
},
{
"epoch": 1.3824,
"grad_norm": 0.42387306690216064,
"learning_rate": 6.545084971874738e-06,
"loss": 0.5891,
"step": 144
},
{
"epoch": 1.392,
"grad_norm": 0.4314691722393036,
"learning_rate": 6.491634692845781e-06,
"loss": 0.5774,
"step": 145
},
{
"epoch": 1.4016,
"grad_norm": 0.43843144178390503,
"learning_rate": 6.437996637160086e-06,
"loss": 0.5657,
"step": 146
},
{
"epoch": 1.4112,
"grad_norm": 0.40325650572776794,
"learning_rate": 6.384177557124247e-06,
"loss": 0.5936,
"step": 147
},
{
"epoch": 1.4208,
"grad_norm": 0.4526993930339813,
"learning_rate": 6.330184227833376e-06,
"loss": 0.5902,
"step": 148
},
{
"epoch": 1.4304000000000001,
"grad_norm": 0.4019985795021057,
"learning_rate": 6.276023446318214e-06,
"loss": 0.5398,
"step": 149
},
{
"epoch": 1.44,
"grad_norm": 0.4319303333759308,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.5577,
"step": 150
},
{
"epoch": 1.4496,
"grad_norm": 0.4279302656650543,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.5424,
"step": 151
},
{
"epoch": 1.4592,
"grad_norm": 0.41032660007476807,
"learning_rate": 6.112604669781572e-06,
"loss": 0.5758,
"step": 152
},
{
"epoch": 1.4687999999999999,
"grad_norm": 0.42172425985336304,
"learning_rate": 6.057842458386315e-06,
"loss": 0.5969,
"step": 153
},
{
"epoch": 1.4784,
"grad_norm": 0.4094562232494354,
"learning_rate": 6.002947078916365e-06,
"loss": 0.5977,
"step": 154
},
{
"epoch": 1.488,
"grad_norm": 0.45888322591781616,
"learning_rate": 5.947925441958393e-06,
"loss": 0.5995,
"step": 155
},
{
"epoch": 1.4976,
"grad_norm": 0.3935939073562622,
"learning_rate": 5.892784473993184e-06,
"loss": 0.5695,
"step": 156
},
{
"epoch": 1.5072,
"grad_norm": 0.41924765706062317,
"learning_rate": 5.837531116523683e-06,
"loss": 0.6031,
"step": 157
},
{
"epoch": 1.5168,
"grad_norm": 0.43053096532821655,
"learning_rate": 5.782172325201155e-06,
"loss": 0.5576,
"step": 158
},
{
"epoch": 1.5264,
"grad_norm": 0.3961893618106842,
"learning_rate": 5.726715068949564e-06,
"loss": 0.5886,
"step": 159
},
{
"epoch": 1.536,
"grad_norm": 0.42763078212738037,
"learning_rate": 5.671166329088278e-06,
"loss": 0.5752,
"step": 160
},
{
"epoch": 1.5455999999999999,
"grad_norm": 0.42291295528411865,
"learning_rate": 5.615533098453215e-06,
"loss": 0.5921,
"step": 161
},
{
"epoch": 1.5552000000000001,
"grad_norm": 0.38611024618148804,
"learning_rate": 5.559822380516539e-06,
"loss": 0.5697,
"step": 162
},
{
"epoch": 1.5648,
"grad_norm": 0.4422116279602051,
"learning_rate": 5.504041188505022e-06,
"loss": 0.5581,
"step": 163
},
{
"epoch": 1.5744,
"grad_norm": 0.4732320308685303,
"learning_rate": 5.448196544517168e-06,
"loss": 0.5994,
"step": 164
},
{
"epoch": 1.584,
"grad_norm": 0.40770646929740906,
"learning_rate": 5.392295478639226e-06,
"loss": 0.5717,
"step": 165
},
{
"epoch": 1.5936,
"grad_norm": 0.4280209541320801,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6037,
"step": 166
},
{
"epoch": 1.6032,
"grad_norm": 0.3919738233089447,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.5224,
"step": 167
},
{
"epoch": 1.6128,
"grad_norm": 0.40155377984046936,
"learning_rate": 5.224324151752575e-06,
"loss": 0.5911,
"step": 168
},
{
"epoch": 1.6223999999999998,
"grad_norm": 0.39139828085899353,
"learning_rate": 5.168267827938971e-06,
"loss": 0.5879,
"step": 169
},
{
"epoch": 1.6320000000000001,
"grad_norm": 0.3596862852573395,
"learning_rate": 5.112190321479026e-06,
"loss": 0.5605,
"step": 170
},
{
"epoch": 1.6416,
"grad_norm": 0.39607909321784973,
"learning_rate": 5.05609869177323e-06,
"loss": 0.5613,
"step": 171
},
{
"epoch": 1.6512,
"grad_norm": 0.3523328900337219,
"learning_rate": 5e-06,
"loss": 0.5281,
"step": 172
},
{
"epoch": 1.6608,
"grad_norm": 0.37867018580436707,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6229,
"step": 173
},
{
"epoch": 1.6703999999999999,
"grad_norm": 0.35012221336364746,
"learning_rate": 4.887809678520976e-06,
"loss": 0.5233,
"step": 174
},
{
"epoch": 1.6800000000000002,
"grad_norm": 0.3891908526420593,
"learning_rate": 4.831732172061032e-06,
"loss": 0.5824,
"step": 175
},
{
"epoch": 1.6896,
"grad_norm": 0.38234302401542664,
"learning_rate": 4.775675848247427e-06,
"loss": 0.5997,
"step": 176
},
{
"epoch": 1.6992,
"grad_norm": 0.3892868757247925,
"learning_rate": 4.719647763814041e-06,
"loss": 0.5559,
"step": 177
},
{
"epoch": 1.7088,
"grad_norm": 0.3620501160621643,
"learning_rate": 4.663654971939802e-06,
"loss": 0.5516,
"step": 178
},
{
"epoch": 1.7184,
"grad_norm": 0.4023611545562744,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.5694,
"step": 179
},
{
"epoch": 1.728,
"grad_norm": 0.4760470986366272,
"learning_rate": 4.551803455482833e-06,
"loss": 0.5964,
"step": 180
},
{
"epoch": 1.7376,
"grad_norm": 0.3798375427722931,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.576,
"step": 181
},
{
"epoch": 1.7471999999999999,
"grad_norm": 0.3852797746658325,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.5491,
"step": 182
},
{
"epoch": 1.7568000000000001,
"grad_norm": 0.3949350118637085,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6095,
"step": 183
},
{
"epoch": 1.7664,
"grad_norm": 0.3639887273311615,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.5486,
"step": 184
},
{
"epoch": 1.776,
"grad_norm": 0.34000834822654724,
"learning_rate": 4.273284931050438e-06,
"loss": 0.5461,
"step": 185
},
{
"epoch": 1.7856,
"grad_norm": 0.37569451332092285,
"learning_rate": 4.217827674798845e-06,
"loss": 0.5604,
"step": 186
},
{
"epoch": 1.7952,
"grad_norm": 0.4080134630203247,
"learning_rate": 4.162468883476319e-06,
"loss": 0.5852,
"step": 187
},
{
"epoch": 1.8048,
"grad_norm": 0.3695431649684906,
"learning_rate": 4.107215526006818e-06,
"loss": 0.5638,
"step": 188
},
{
"epoch": 1.8144,
"grad_norm": 0.3547602593898773,
"learning_rate": 4.052074558041608e-06,
"loss": 0.5277,
"step": 189
},
{
"epoch": 1.8239999999999998,
"grad_norm": 0.4300488531589508,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6041,
"step": 190
},
{
"epoch": 1.8336000000000001,
"grad_norm": 0.3675505220890045,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.5681,
"step": 191
},
{
"epoch": 1.8432,
"grad_norm": 0.3509397804737091,
"learning_rate": 3.887395330218429e-06,
"loss": 0.5469,
"step": 192
},
{
"epoch": 1.8528,
"grad_norm": 0.4187622666358948,
"learning_rate": 3.832773180720475e-06,
"loss": 0.577,
"step": 193
},
{
"epoch": 1.8624,
"grad_norm": 0.43187215924263,
"learning_rate": 3.778297969310529e-06,
"loss": 0.5763,
"step": 194
},
{
"epoch": 1.8719999999999999,
"grad_norm": 0.37899720668792725,
"learning_rate": 3.723976553681787e-06,
"loss": 0.598,
"step": 195
},
{
"epoch": 1.8816000000000002,
"grad_norm": 0.3809452950954437,
"learning_rate": 3.669815772166625e-06,
"loss": 0.5931,
"step": 196
},
{
"epoch": 1.8912,
"grad_norm": 0.39111581444740295,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.5557,
"step": 197
},
{
"epoch": 1.9008,
"grad_norm": 0.3807132840156555,
"learning_rate": 3.562003362839914e-06,
"loss": 0.5698,
"step": 198
},
{
"epoch": 1.9104,
"grad_norm": 0.39497897028923035,
"learning_rate": 3.50836530715422e-06,
"loss": 0.5781,
"step": 199
},
{
"epoch": 1.92,
"grad_norm": 0.36814066767692566,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.5294,
"step": 200
},
{
"epoch": 1.9296,
"grad_norm": 0.38876456022262573,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.5544,
"step": 201
},
{
"epoch": 1.9392,
"grad_norm": 0.43197450041770935,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6093,
"step": 202
},
{
"epoch": 1.9487999999999999,
"grad_norm": 0.3526746332645416,
"learning_rate": 3.295758014387375e-06,
"loss": 0.5674,
"step": 203
},
{
"epoch": 1.9584000000000001,
"grad_norm": 0.38960641622543335,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.569,
"step": 204
},
{
"epoch": 1.968,
"grad_norm": 0.3506081998348236,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.5645,
"step": 205
},
{
"epoch": 1.9776,
"grad_norm": 0.38984987139701843,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.5699,
"step": 206
},
{
"epoch": 1.9872,
"grad_norm": 0.37052249908447266,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.5766,
"step": 207
},
{
"epoch": 1.9968,
"grad_norm": 0.38269490003585815,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.5525,
"step": 208
},
{
"epoch": 2.0064,
"grad_norm": 1.0422611236572266,
"learning_rate": 2.9834142279959754e-06,
"loss": 1.0042,
"step": 209
},
{
"epoch": 2.016,
"grad_norm": 0.4426552355289459,
"learning_rate": 2.932207475167398e-06,
"loss": 0.554,
"step": 210
},
{
"epoch": 2.0256,
"grad_norm": 0.3669895529747009,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.515,
"step": 211
},
{
"epoch": 2.0352,
"grad_norm": 0.4123702347278595,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5396,
"step": 212
},
{
"epoch": 2.0448,
"grad_norm": 0.3744567036628723,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5631,
"step": 213
},
{
"epoch": 2.0544,
"grad_norm": 0.34888023138046265,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5205,
"step": 214
},
{
"epoch": 2.064,
"grad_norm": 0.38978055119514465,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5299,
"step": 215
},
{
"epoch": 2.0736,
"grad_norm": 0.3805748224258423,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5577,
"step": 216
},
{
"epoch": 2.0832,
"grad_norm": 0.384547621011734,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5107,
"step": 217
},
{
"epoch": 2.0928,
"grad_norm": 0.40043580532073975,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5443,
"step": 218
},
{
"epoch": 2.1024,
"grad_norm": 0.3578666150569916,
"learning_rate": 2.483822884971e-06,
"loss": 0.5417,
"step": 219
},
{
"epoch": 2.112,
"grad_norm": 0.3533429801464081,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5455,
"step": 220
},
{
"epoch": 2.1216,
"grad_norm": 0.3117644190788269,
"learning_rate": 2.387507176420256e-06,
"loss": 0.4882,
"step": 221
},
{
"epoch": 2.1312,
"grad_norm": 0.4052339494228363,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5515,
"step": 222
},
{
"epoch": 2.1408,
"grad_norm": 0.3945196270942688,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.527,
"step": 223
},
{
"epoch": 2.1504,
"grad_norm": 0.3631379008293152,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5637,
"step": 224
},
{
"epoch": 2.16,
"grad_norm": 0.342517614364624,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5007,
"step": 225
},
{
"epoch": 2.1696,
"grad_norm": 0.3698350191116333,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.51,
"step": 226
},
{
"epoch": 2.1792,
"grad_norm": 0.3405527174472809,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5016,
"step": 227
},
{
"epoch": 2.1888,
"grad_norm": 0.36109381914138794,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5232,
"step": 228
},
{
"epoch": 2.1984,
"grad_norm": 0.3690553903579712,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5606,
"step": 229
},
{
"epoch": 2.208,
"grad_norm": 0.3747081160545349,
"learning_rate": 1.971049780795901e-06,
"loss": 0.533,
"step": 230
},
{
"epoch": 2.2176,
"grad_norm": 0.3522194027900696,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5567,
"step": 231
},
{
"epoch": 2.2272,
"grad_norm": 0.3159579634666443,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.4776,
"step": 232
},
{
"epoch": 2.2368,
"grad_norm": 0.400075227022171,
"learning_rate": 1.838887489710452e-06,
"loss": 0.542,
"step": 233
},
{
"epoch": 2.2464,
"grad_norm": 0.34253227710723877,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5327,
"step": 234
},
{
"epoch": 2.2560000000000002,
"grad_norm": 0.34481409192085266,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5695,
"step": 235
},
{
"epoch": 2.2656,
"grad_norm": 0.33534330129623413,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5053,
"step": 236
},
{
"epoch": 2.2752,
"grad_norm": 0.3279606103897095,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5478,
"step": 237
},
{
"epoch": 2.2848,
"grad_norm": 0.3384232521057129,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5107,
"step": 238
},
{
"epoch": 2.2944,
"grad_norm": 0.35894083976745605,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5465,
"step": 239
},
{
"epoch": 2.304,
"grad_norm": 0.32638615369796753,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5271,
"step": 240
},
{
"epoch": 2.3136,
"grad_norm": 0.33072298765182495,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5515,
"step": 241
},
{
"epoch": 2.3232,
"grad_norm": 0.32796502113342285,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5396,
"step": 242
},
{
"epoch": 2.3327999999999998,
"grad_norm": 0.3144128918647766,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5181,
"step": 243
},
{
"epoch": 2.3424,
"grad_norm": 0.30645841360092163,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.4946,
"step": 244
},
{
"epoch": 2.352,
"grad_norm": 0.3447537124156952,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5373,
"step": 245
},
{
"epoch": 2.3616,
"grad_norm": 0.32241493463516235,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5381,
"step": 246
},
{
"epoch": 2.3712,
"grad_norm": 0.32736265659332275,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5552,
"step": 247
},
{
"epoch": 2.3808,
"grad_norm": 0.331808477640152,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5274,
"step": 248
},
{
"epoch": 2.3904,
"grad_norm": 0.32642239332199097,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5381,
"step": 249
},
{
"epoch": 2.4,
"grad_norm": 0.33340737223625183,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5336,
"step": 250
},
{
"epoch": 2.4096,
"grad_norm": 0.32736656069755554,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.4979,
"step": 251
},
{
"epoch": 2.4192,
"grad_norm": 0.3242564797401428,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5143,
"step": 252
},
{
"epoch": 2.4288,
"grad_norm": 0.3379511535167694,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.527,
"step": 253
},
{
"epoch": 2.4384,
"grad_norm": 0.3252667188644409,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5412,
"step": 254
},
{
"epoch": 2.448,
"grad_norm": 0.3234214782714844,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5393,
"step": 255
},
{
"epoch": 2.4576000000000002,
"grad_norm": 0.3364090621471405,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5687,
"step": 256
},
{
"epoch": 2.4672,
"grad_norm": 0.34504103660583496,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5732,
"step": 257
},
{
"epoch": 2.4768,
"grad_norm": 0.3204050660133362,
"learning_rate": 8.899896227604509e-07,
"loss": 0.4957,
"step": 258
},
{
"epoch": 2.4864,
"grad_norm": 0.3602452874183655,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5491,
"step": 259
},
{
"epoch": 2.496,
"grad_norm": 0.31696954369544983,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5361,
"step": 260
},
{
"epoch": 2.5056000000000003,
"grad_norm": 0.33390817046165466,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5677,
"step": 261
},
{
"epoch": 2.5152,
"grad_norm": 0.3218343257904053,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5103,
"step": 262
},
{
"epoch": 2.5248,
"grad_norm": 0.2960663437843323,
"learning_rate": 7.367991782295392e-07,
"loss": 0.4911,
"step": 263
},
{
"epoch": 2.5343999999999998,
"grad_norm": 0.3217982053756714,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5322,
"step": 264
},
{
"epoch": 2.544,
"grad_norm": 0.32205691933631897,
"learning_rate": 6.792532212817271e-07,
"loss": 0.4978,
"step": 265
},
{
"epoch": 2.5536,
"grad_norm": 0.30698198080062866,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5047,
"step": 266
},
{
"epoch": 2.5632,
"grad_norm": 0.32447564601898193,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5156,
"step": 267
},
{
"epoch": 2.5728,
"grad_norm": 0.3703906536102295,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5779,
"step": 268
},
{
"epoch": 2.5824,
"grad_norm": 0.3244221806526184,
"learning_rate": 5.707160663821009e-07,
"loss": 0.5195,
"step": 269
},
{
"epoch": 2.592,
"grad_norm": 0.3410571813583374,
"learning_rate": 5.449673790581611e-07,
"loss": 0.536,
"step": 270
},
{
"epoch": 2.6016,
"grad_norm": 0.308290034532547,
"learning_rate": 5.197795201563744e-07,
"loss": 0.5492,
"step": 271
},
{
"epoch": 2.6112,
"grad_norm": 0.3327993154525757,
"learning_rate": 4.951556604879049e-07,
"loss": 0.512,
"step": 272
},
{
"epoch": 2.6208,
"grad_norm": 0.3642774820327759,
"learning_rate": 4.710988998640298e-07,
"loss": 0.5351,
"step": 273
},
{
"epoch": 2.6304,
"grad_norm": 0.3263518810272217,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5445,
"step": 274
},
{
"epoch": 2.64,
"grad_norm": 0.31554028391838074,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.526,
"step": 275
},
{
"epoch": 2.6496,
"grad_norm": 0.3256407082080841,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5306,
"step": 276
},
{
"epoch": 2.6592000000000002,
"grad_norm": 0.31249770522117615,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.549,
"step": 277
},
{
"epoch": 2.6688,
"grad_norm": 0.3235461711883545,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5244,
"step": 278
},
{
"epoch": 2.6784,
"grad_norm": 0.31085437536239624,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.5421,
"step": 279
},
{
"epoch": 2.6879999999999997,
"grad_norm": 0.30151858925819397,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5203,
"step": 280
},
{
"epoch": 2.6976,
"grad_norm": 0.31177854537963867,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.5348,
"step": 281
},
{
"epoch": 2.7072000000000003,
"grad_norm": 0.2945941686630249,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5156,
"step": 282
},
{
"epoch": 2.7168,
"grad_norm": 0.3394186794757843,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.5493,
"step": 283
},
{
"epoch": 2.7264,
"grad_norm": 0.3355710506439209,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5399,
"step": 284
},
{
"epoch": 2.7359999999999998,
"grad_norm": 0.30999863147735596,
"learning_rate": 2.276812823220964e-07,
"loss": 0.5238,
"step": 285
},
{
"epoch": 2.7456,
"grad_norm": 0.31282395124435425,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5372,
"step": 286
},
{
"epoch": 2.7552,
"grad_norm": 0.3219570517539978,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5301,
"step": 287
},
{
"epoch": 2.7648,
"grad_norm": 0.30571168661117554,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5122,
"step": 288
},
{
"epoch": 2.7744,
"grad_norm": 0.3257659375667572,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5436,
"step": 289
},
{
"epoch": 2.784,
"grad_norm": 0.29299965500831604,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5102,
"step": 290
},
{
"epoch": 2.7936,
"grad_norm": 0.2914055585861206,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5171,
"step": 291
},
{
"epoch": 2.8032,
"grad_norm": 0.31332334876060486,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5552,
"step": 292
},
{
"epoch": 2.8128,
"grad_norm": 0.32967719435691833,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5493,
"step": 293
},
{
"epoch": 2.8224,
"grad_norm": 0.2999667823314667,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.4915,
"step": 294
},
{
"epoch": 2.832,
"grad_norm": 0.3241068422794342,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5589,
"step": 295
},
{
"epoch": 2.8416,
"grad_norm": 0.3096008002758026,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5172,
"step": 296
},
{
"epoch": 2.8512,
"grad_norm": 0.33340883255004883,
"learning_rate": 7.064490740882057e-08,
"loss": 0.5614,
"step": 297
},
{
"epoch": 2.8608000000000002,
"grad_norm": 0.3046957850456238,
"learning_rate": 6.15582970243117e-08,
"loss": 0.529,
"step": 298
},
{
"epoch": 2.8704,
"grad_norm": 0.30237892270088196,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5302,
"step": 299
},
{
"epoch": 2.88,
"grad_norm": 0.30969762802124023,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5129,
"step": 300
},
{
"epoch": 2.8895999999999997,
"grad_norm": 0.31994253396987915,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.558,
"step": 301
},
{
"epoch": 2.8992,
"grad_norm": 0.29612258076667786,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5123,
"step": 302
},
{
"epoch": 2.9088000000000003,
"grad_norm": 0.3210684359073639,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5758,
"step": 303
},
{
"epoch": 2.9184,
"grad_norm": 0.3382963538169861,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5496,
"step": 304
},
{
"epoch": 2.928,
"grad_norm": 0.32367751002311707,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5206,
"step": 305
},
{
"epoch": 2.9375999999999998,
"grad_norm": 0.32362955808639526,
"learning_rate": 1.132562476771959e-08,
"loss": 0.562,
"step": 306
},
{
"epoch": 2.9472,
"grad_norm": 0.3154355585575104,
"learning_rate": 7.865924910916977e-09,
"loss": 0.4977,
"step": 307
},
{
"epoch": 2.9568,
"grad_norm": 0.3255595564842224,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5958,
"step": 308
},
{
"epoch": 2.9664,
"grad_norm": 0.29286327958106995,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.4834,
"step": 309
},
{
"epoch": 2.976,
"grad_norm": 0.3221506178379059,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5826,
"step": 310
},
{
"epoch": 2.9856,
"grad_norm": 0.30886587500572205,
"learning_rate": 3.147162264971471e-10,
"loss": 0.5197,
"step": 311
},
{
"epoch": 2.9952,
"grad_norm": 0.29504039883613586,
"learning_rate": 0.0,
"loss": 0.5294,
"step": 312
},
{
"epoch": 2.9952,
"step": 312,
"total_flos": 449468303376384.0,
"train_loss": 0.6024729885543004,
"train_runtime": 34730.5386,
"train_samples_per_second": 0.864,
"train_steps_per_second": 0.009
}
],
"logging_steps": 1.0,
"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": 449468303376384.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}