968 lines
23 KiB
JSON
968 lines
23 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 2.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 664,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.015071590052750565,
|
||
|
|
"grad_norm": 63.93290066194459,
|
||
|
|
"learning_rate": 1.0000000000000002e-06,
|
||
|
|
"loss": 1.451,
|
||
|
|
"step": 5
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03014318010550113,
|
||
|
|
"grad_norm": 35.58752674721741,
|
||
|
|
"learning_rate": 2.25e-06,
|
||
|
|
"loss": 1.5943,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0452147701582517,
|
||
|
|
"grad_norm": 14.633818173823233,
|
||
|
|
"learning_rate": 3.5e-06,
|
||
|
|
"loss": 1.4134,
|
||
|
|
"step": 15
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06028636021100226,
|
||
|
|
"grad_norm": 11.36624439106729,
|
||
|
|
"learning_rate": 4.75e-06,
|
||
|
|
"loss": 1.3789,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07535795026375283,
|
||
|
|
"grad_norm": 19.687207881604735,
|
||
|
|
"learning_rate": 4.999524068745182e-06,
|
||
|
|
"loss": 1.4061,
|
||
|
|
"step": 25
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0904295403165034,
|
||
|
|
"grad_norm": 3.4721808953008786,
|
||
|
|
"learning_rate": 4.997590908585508e-06,
|
||
|
|
"loss": 1.3465,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10550113036925396,
|
||
|
|
"grad_norm": 3.7575390023993256,
|
||
|
|
"learning_rate": 4.994171922976349e-06,
|
||
|
|
"loss": 1.3066,
|
||
|
|
"step": 35
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12057272042200452,
|
||
|
|
"grad_norm": 3.93945708920635,
|
||
|
|
"learning_rate": 4.9892691458840415e-06,
|
||
|
|
"loss": 1.3024,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1356443104747551,
|
||
|
|
"grad_norm": 5.745483553248285,
|
||
|
|
"learning_rate": 4.982885493987595e-06,
|
||
|
|
"loss": 1.3222,
|
||
|
|
"step": 45
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15071590052750566,
|
||
|
|
"grad_norm": 3.556951637647764,
|
||
|
|
"learning_rate": 4.975024764943549e-06,
|
||
|
|
"loss": 1.3441,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1657874905802562,
|
||
|
|
"grad_norm": 2.837174059992352,
|
||
|
|
"learning_rate": 4.965691635126737e-06,
|
||
|
|
"loss": 1.2827,
|
||
|
|
"step": 55
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1808590806330068,
|
||
|
|
"grad_norm": 3.041016464130793,
|
||
|
|
"learning_rate": 4.95489165684829e-06,
|
||
|
|
"loss": 1.2503,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19593067068575734,
|
||
|
|
"grad_norm": 2.9883733154892234,
|
||
|
|
"learning_rate": 4.942631255052551e-06,
|
||
|
|
"loss": 1.2252,
|
||
|
|
"step": 65
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21100226073850792,
|
||
|
|
"grad_norm": 3.6954501199574636,
|
||
|
|
"learning_rate": 4.928917723494854e-06,
|
||
|
|
"loss": 1.2734,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22607385079125847,
|
||
|
|
"grad_norm": 8.370545215192237,
|
||
|
|
"learning_rate": 4.913759220402441e-06,
|
||
|
|
"loss": 1.2101,
|
||
|
|
"step": 75
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24114544084400905,
|
||
|
|
"grad_norm": 28.432365965110936,
|
||
|
|
"learning_rate": 4.897164763621104e-06,
|
||
|
|
"loss": 1.2485,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2562170308967596,
|
||
|
|
"grad_norm": 4.715449035187671,
|
||
|
|
"learning_rate": 4.879144225250445e-06,
|
||
|
|
"loss": 1.2243,
|
||
|
|
"step": 85
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2712886209495102,
|
||
|
|
"grad_norm": 2.9636386869143387,
|
||
|
|
"learning_rate": 4.859708325770919e-06,
|
||
|
|
"loss": 1.2052,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2863602110022607,
|
||
|
|
"grad_norm": 3.066981968094732,
|
||
|
|
"learning_rate": 4.838868627666191e-06,
|
||
|
|
"loss": 1.1686,
|
||
|
|
"step": 95
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30143180105501133,
|
||
|
|
"grad_norm": 2.412797888960037,
|
||
|
|
"learning_rate": 4.81663752854457e-06,
|
||
|
|
"loss": 1.1775,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3165033911077619,
|
||
|
|
"grad_norm": 2.5140619073728105,
|
||
|
|
"learning_rate": 4.793028253763633e-06,
|
||
|
|
"loss": 1.1722,
|
||
|
|
"step": 105
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3315749811605124,
|
||
|
|
"grad_norm": 2.493285093391174,
|
||
|
|
"learning_rate": 4.768054848562411e-06,
|
||
|
|
"loss": 1.1695,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.346646571213263,
|
||
|
|
"grad_norm": 3.144530529949511,
|
||
|
|
"learning_rate": 4.741732169705829e-06,
|
||
|
|
"loss": 1.1841,
|
||
|
|
"step": 115
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3617181612660136,
|
||
|
|
"grad_norm": 2.295823857057481,
|
||
|
|
"learning_rate": 4.714075876646365e-06,
|
||
|
|
"loss": 1.1757,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37678975131876413,
|
||
|
|
"grad_norm": 2.643471529827654,
|
||
|
|
"learning_rate": 4.6851024222081905e-06,
|
||
|
|
"loss": 1.1734,
|
||
|
|
"step": 125
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3918613413715147,
|
||
|
|
"grad_norm": 2.9457088109557668,
|
||
|
|
"learning_rate": 4.654829042799323e-06,
|
||
|
|
"loss": 1.127,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4069329314242653,
|
||
|
|
"grad_norm": 3.0712069515731444,
|
||
|
|
"learning_rate": 4.62327374815763e-06,
|
||
|
|
"loss": 1.2514,
|
||
|
|
"step": 135
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.42200452147701584,
|
||
|
|
"grad_norm": 2.6973425645416174,
|
||
|
|
"learning_rate": 4.590455310636778e-06,
|
||
|
|
"loss": 1.1873,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4370761115297664,
|
||
|
|
"grad_norm": 2.6584776025168337,
|
||
|
|
"learning_rate": 4.556393254038486e-06,
|
||
|
|
"loss": 1.1515,
|
||
|
|
"step": 145
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.45214770158251694,
|
||
|
|
"grad_norm": 2.7834304414005753,
|
||
|
|
"learning_rate": 4.521107841997758e-06,
|
||
|
|
"loss": 1.1197,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.46721929163526754,
|
||
|
|
"grad_norm": 3.0654410008265893,
|
||
|
|
"learning_rate": 4.48462006592797e-06,
|
||
|
|
"loss": 1.1494,
|
||
|
|
"step": 155
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4822908816880181,
|
||
|
|
"grad_norm": 2.3584975942076487,
|
||
|
|
"learning_rate": 4.4469516325330055e-06,
|
||
|
|
"loss": 1.1292,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.49736247174076864,
|
||
|
|
"grad_norm": 2.6296807819946766,
|
||
|
|
"learning_rate": 4.408124950893868e-06,
|
||
|
|
"loss": 1.1195,
|
||
|
|
"step": 165
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5124340617935192,
|
||
|
|
"grad_norm": 2.810412992553872,
|
||
|
|
"learning_rate": 4.368163119137437e-06,
|
||
|
|
"loss": 1.1047,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5275056518462697,
|
||
|
|
"grad_norm": 2.7447890069275207,
|
||
|
|
"learning_rate": 4.32708991069531e-06,
|
||
|
|
"loss": 1.1689,
|
||
|
|
"step": 175
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5425772418990203,
|
||
|
|
"grad_norm": 2.6925744340959104,
|
||
|
|
"learning_rate": 4.284929760160917e-06,
|
||
|
|
"loss": 0.9971,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.557648831951771,
|
||
|
|
"grad_norm": 2.604938296467723,
|
||
|
|
"learning_rate": 4.2417077487532835e-06,
|
||
|
|
"loss": 1.1245,
|
||
|
|
"step": 185
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5727204220045214,
|
||
|
|
"grad_norm": 2.9775634900633534,
|
||
|
|
"learning_rate": 4.197449589396131e-06,
|
||
|
|
"loss": 1.1714,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.587792012057272,
|
||
|
|
"grad_norm": 2.7593158692455066,
|
||
|
|
"learning_rate": 4.152181611421179e-06,
|
||
|
|
"loss": 1.1634,
|
||
|
|
"step": 195
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6028636021100227,
|
||
|
|
"grad_norm": 2.6974834987909357,
|
||
|
|
"learning_rate": 4.105930744904721e-06,
|
||
|
|
"loss": 1.1021,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6179351921627732,
|
||
|
|
"grad_norm": 2.872785492959392,
|
||
|
|
"learning_rate": 4.058724504646834e-06,
|
||
|
|
"loss": 1.1172,
|
||
|
|
"step": 205
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6330067822155238,
|
||
|
|
"grad_norm": 2.562056324367801,
|
||
|
|
"learning_rate": 4.010590973802737e-06,
|
||
|
|
"loss": 1.1375,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6480783722682742,
|
||
|
|
"grad_norm": 2.2291085754732713,
|
||
|
|
"learning_rate": 3.961558787176012e-06,
|
||
|
|
"loss": 1.1482,
|
||
|
|
"step": 215
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6631499623210249,
|
||
|
|
"grad_norm": 2.4045628734860687,
|
||
|
|
"learning_rate": 3.91165711418367e-06,
|
||
|
|
"loss": 1.1041,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6782215523737755,
|
||
|
|
"grad_norm": 3.3080421132480717,
|
||
|
|
"learning_rate": 3.860915641503161e-06,
|
||
|
|
"loss": 1.1142,
|
||
|
|
"step": 225
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.693293142426526,
|
||
|
|
"grad_norm": 3.245502012246857,
|
||
|
|
"learning_rate": 3.809364555411667e-06,
|
||
|
|
"loss": 1.185,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7083647324792766,
|
||
|
|
"grad_norm": 3.0888496819111726,
|
||
|
|
"learning_rate": 3.75703452382818e-06,
|
||
|
|
"loss": 1.1719,
|
||
|
|
"step": 235
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7234363225320272,
|
||
|
|
"grad_norm": 3.1440323353252526,
|
||
|
|
"learning_rate": 3.703956678069055e-06,
|
||
|
|
"loss": 1.1238,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7385079125847777,
|
||
|
|
"grad_norm": 2.8116489290906723,
|
||
|
|
"learning_rate": 3.650162594327881e-06,
|
||
|
|
"loss": 1.0788,
|
||
|
|
"step": 245
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7535795026375283,
|
||
|
|
"grad_norm": 2.6368140598692347,
|
||
|
|
"learning_rate": 3.595684274890691e-06,
|
||
|
|
"loss": 1.0756,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7686510926902789,
|
||
|
|
"grad_norm": 2.408558758625417,
|
||
|
|
"learning_rate": 3.5405541290976968e-06,
|
||
|
|
"loss": 1.1189,
|
||
|
|
"step": 255
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7837226827430294,
|
||
|
|
"grad_norm": 2.897272560922028,
|
||
|
|
"learning_rate": 3.4848049540628588e-06,
|
||
|
|
"loss": 1.0719,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.79879427279578,
|
||
|
|
"grad_norm": 2.606250228734567,
|
||
|
|
"learning_rate": 3.4284699151627672e-06,
|
||
|
|
"loss": 1.1073,
|
||
|
|
"step": 265
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8138658628485306,
|
||
|
|
"grad_norm": 2.7496803732493276,
|
||
|
|
"learning_rate": 3.3715825263064493e-06,
|
||
|
|
"loss": 1.0891,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8289374529012811,
|
||
|
|
"grad_norm": 2.5362507404210146,
|
||
|
|
"learning_rate": 3.314176629997825e-06,
|
||
|
|
"loss": 1.0003,
|
||
|
|
"step": 275
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8440090429540317,
|
||
|
|
"grad_norm": 2.930907397969813,
|
||
|
|
"learning_rate": 3.25628637720269e-06,
|
||
|
|
"loss": 1.1259,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8590806330067822,
|
||
|
|
"grad_norm": 2.339195373127568,
|
||
|
|
"learning_rate": 3.1979462070321817e-06,
|
||
|
|
"loss": 1.1505,
|
||
|
|
"step": 285
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8741522230595328,
|
||
|
|
"grad_norm": 2.8349021987346266,
|
||
|
|
"learning_rate": 3.1391908262548382e-06,
|
||
|
|
"loss": 1.0636,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8892238131122834,
|
||
|
|
"grad_norm": 3.2284091736671905,
|
||
|
|
"learning_rate": 3.08005518864942e-06,
|
||
|
|
"loss": 1.1475,
|
||
|
|
"step": 295
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9042954031650339,
|
||
|
|
"grad_norm": 2.6609799290676768,
|
||
|
|
"learning_rate": 3.0205744742107874e-06,
|
||
|
|
"loss": 1.0161,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9193669932177845,
|
||
|
|
"grad_norm": 2.5739173473029417,
|
||
|
|
"learning_rate": 2.9607840682211987e-06,
|
||
|
|
"loss": 1.0239,
|
||
|
|
"step": 305
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9344385832705351,
|
||
|
|
"grad_norm": 2.937887898785162,
|
||
|
|
"learning_rate": 2.900719540199484e-06,
|
||
|
|
"loss": 1.0436,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9495101733232856,
|
||
|
|
"grad_norm": 2.7246531876982116,
|
||
|
|
"learning_rate": 2.840416622740617e-06,
|
||
|
|
"loss": 1.0982,
|
||
|
|
"step": 315
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9645817633760362,
|
||
|
|
"grad_norm": 2.8889715233179456,
|
||
|
|
"learning_rate": 2.7799111902582697e-06,
|
||
|
|
"loss": 1.1161,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9796533534287868,
|
||
|
|
"grad_norm": 3.20559013895907,
|
||
|
|
"learning_rate": 2.7192392376430014e-06,
|
||
|
|
"loss": 1.066,
|
||
|
|
"step": 325
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9947249434815373,
|
||
|
|
"grad_norm": 2.63714269591387,
|
||
|
|
"learning_rate": 2.6584368588487727e-06,
|
||
|
|
"loss": 1.0798,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0090429540316503,
|
||
|
|
"grad_norm": 2.8236476573888636,
|
||
|
|
"learning_rate": 2.597540225420525e-06,
|
||
|
|
"loss": 0.8717,
|
||
|
|
"step": 335
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0241145440844008,
|
||
|
|
"grad_norm": 2.403622878308974,
|
||
|
|
"learning_rate": 2.5365855649756114e-06,
|
||
|
|
"loss": 0.7775,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0391861341371516,
|
||
|
|
"grad_norm": 3.1556330242872828,
|
||
|
|
"learning_rate": 2.475609139651855e-06,
|
||
|
|
"loss": 0.8101,
|
||
|
|
"step": 345
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.054257724189902,
|
||
|
|
"grad_norm": 2.8657604036527373,
|
||
|
|
"learning_rate": 2.4146472245350804e-06,
|
||
|
|
"loss": 0.7067,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0693293142426525,
|
||
|
|
"grad_norm": 2.845940606079971,
|
||
|
|
"learning_rate": 2.353736086078941e-06,
|
||
|
|
"loss": 0.7314,
|
||
|
|
"step": 355
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0844009042954033,
|
||
|
|
"grad_norm": 2.6386252875863856,
|
||
|
|
"learning_rate": 2.2929119605298787e-06,
|
||
|
|
"loss": 0.7813,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0994724943481538,
|
||
|
|
"grad_norm": 2.9858658644648512,
|
||
|
|
"learning_rate": 2.232211032370057e-06,
|
||
|
|
"loss": 0.7792,
|
||
|
|
"step": 365
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1145440844009042,
|
||
|
|
"grad_norm": 3.1704667094741428,
|
||
|
|
"learning_rate": 2.1716694127910863e-06,
|
||
|
|
"loss": 0.8093,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.129615674453655,
|
||
|
|
"grad_norm": 2.438807649007062,
|
||
|
|
"learning_rate": 2.1113231182113557e-06,
|
||
|
|
"loss": 0.7326,
|
||
|
|
"step": 375
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1446872645064055,
|
||
|
|
"grad_norm": 2.766081091730644,
|
||
|
|
"learning_rate": 2.0512080488497353e-06,
|
||
|
|
"loss": 0.7069,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.159758854559156,
|
||
|
|
"grad_norm": 2.749798423974893,
|
||
|
|
"learning_rate": 1.991359967368416e-06,
|
||
|
|
"loss": 0.696,
|
||
|
|
"step": 385
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1748304446119064,
|
||
|
|
"grad_norm": 2.9444307577352937,
|
||
|
|
"learning_rate": 1.9318144775975773e-06,
|
||
|
|
"loss": 0.7416,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1899020346646572,
|
||
|
|
"grad_norm": 3.04433007961317,
|
||
|
|
"learning_rate": 1.8726070033545468e-06,
|
||
|
|
"loss": 0.7579,
|
||
|
|
"step": 395
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2049736247174077,
|
||
|
|
"grad_norm": 2.9541203492627517,
|
||
|
|
"learning_rate": 1.8137727673700476e-06,
|
||
|
|
"loss": 0.7538,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2200452147701584,
|
||
|
|
"grad_norm": 3.5431910354023497,
|
||
|
|
"learning_rate": 1.7553467703340755e-06,
|
||
|
|
"loss": 0.7116,
|
||
|
|
"step": 405
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2351168048229089,
|
||
|
|
"grad_norm": 2.6008674886254486,
|
||
|
|
"learning_rate": 1.697363770073866e-06,
|
||
|
|
"loss": 0.7292,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2501883948756594,
|
||
|
|
"grad_norm": 2.498409384596954,
|
||
|
|
"learning_rate": 1.6398582608763457e-06,
|
||
|
|
"loss": 0.7165,
|
||
|
|
"step": 415
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2652599849284099,
|
||
|
|
"grad_norm": 2.962253311649292,
|
||
|
|
"learning_rate": 1.5828644529673592e-06,
|
||
|
|
"loss": 0.7987,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2803315749811606,
|
||
|
|
"grad_norm": 2.862039009052619,
|
||
|
|
"learning_rate": 1.5264162521598893e-06,
|
||
|
|
"loss": 0.7438,
|
||
|
|
"step": 425
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.295403165033911,
|
||
|
|
"grad_norm": 2.6095190980853773,
|
||
|
|
"learning_rate": 1.470547239683372e-06,
|
||
|
|
"loss": 0.7239,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3104747550866618,
|
||
|
|
"grad_norm": 2.8744839745796575,
|
||
|
|
"learning_rate": 1.415290652206105e-06,
|
||
|
|
"loss": 0.7571,
|
||
|
|
"step": 435
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3255463451394123,
|
||
|
|
"grad_norm": 2.7520999185916333,
|
||
|
|
"learning_rate": 1.3606793620626434e-06,
|
||
|
|
"loss": 0.7234,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3406179351921628,
|
||
|
|
"grad_norm": 2.7897554528675994,
|
||
|
|
"learning_rate": 1.3067458576979305e-06,
|
||
|
|
"loss": 0.7832,
|
||
|
|
"step": 445
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3556895252449133,
|
||
|
|
"grad_norm": 2.5249576185058418,
|
||
|
|
"learning_rate": 1.2535222243398104e-06,
|
||
|
|
"loss": 0.7063,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.370761115297664,
|
||
|
|
"grad_norm": 2.923951654007246,
|
||
|
|
"learning_rate": 1.2010401249114166e-06,
|
||
|
|
"loss": 0.7749,
|
||
|
|
"step": 455
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3858327053504145,
|
||
|
|
"grad_norm": 2.659838061766203,
|
||
|
|
"learning_rate": 1.1493307811947887e-06,
|
||
|
|
"loss": 0.7036,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.400904295403165,
|
||
|
|
"grad_norm": 2.567444585108736,
|
||
|
|
"learning_rate": 1.098424955256929e-06,
|
||
|
|
"loss": 0.8069,
|
||
|
|
"step": 465
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4159758854559157,
|
||
|
|
"grad_norm": 2.9945953913555243,
|
||
|
|
"learning_rate": 1.048352931149338e-06,
|
||
|
|
"loss": 0.7826,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4310474755086662,
|
||
|
|
"grad_norm": 2.4443317776748605,
|
||
|
|
"learning_rate": 9.991444968919318e-07,
|
||
|
|
"loss": 0.7534,
|
||
|
|
"step": 475
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4461190655614167,
|
||
|
|
"grad_norm": 3.0058692222092236,
|
||
|
|
"learning_rate": 9.50828926752044e-07,
|
||
|
|
"loss": 0.7342,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4611906556141672,
|
||
|
|
"grad_norm": 2.474430249938214,
|
||
|
|
"learning_rate": 9.034349638290643e-07,
|
||
|
|
"loss": 0.7724,
|
||
|
|
"step": 485
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4762622456669179,
|
||
|
|
"grad_norm": 2.4876357050013485,
|
||
|
|
"learning_rate": 8.569908029550686e-07,
|
||
|
|
"loss": 0.7172,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4913338357196684,
|
||
|
|
"grad_norm": 2.9193635869573478,
|
||
|
|
"learning_rate": 8.115240739216182e-07,
|
||
|
|
"loss": 0.7634,
|
||
|
|
"step": 495
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.506405425772419,
|
||
|
|
"grad_norm": 2.9037368464731985,
|
||
|
|
"learning_rate": 7.670618250426981e-07,
|
||
|
|
"loss": 0.7434,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5214770158251696,
|
||
|
|
"grad_norm": 3.621843474966251,
|
||
|
|
"learning_rate": 7.236305070635835e-07,
|
||
|
|
"loss": 0.7432,
|
||
|
|
"step": 505
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.53654860587792,
|
||
|
|
"grad_norm": 3.172725821539682,
|
||
|
|
"learning_rate": 6.812559574252006e-07,
|
||
|
|
"loss": 0.6995,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5516201959306706,
|
||
|
|
"grad_norm": 3.127632922550285,
|
||
|
|
"learning_rate": 6.399633848933434e-07,
|
||
|
|
"loss": 0.7365,
|
||
|
|
"step": 515
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5666917859834213,
|
||
|
|
"grad_norm": 3.0207209188886437,
|
||
|
|
"learning_rate": 5.997773545618907e-07,
|
||
|
|
"loss": 0.7731,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5817633760361718,
|
||
|
|
"grad_norm": 3.1229845348222693,
|
||
|
|
"learning_rate": 5.607217732389503e-07,
|
||
|
|
"loss": 0.7675,
|
||
|
|
"step": 525
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5968349660889225,
|
||
|
|
"grad_norm": 2.556530267042756,
|
||
|
|
"learning_rate": 5.228198752246125e-07,
|
||
|
|
"loss": 0.7298,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.611906556141673,
|
||
|
|
"grad_norm": 2.7122390573301134,
|
||
|
|
"learning_rate": 4.860942084887868e-07,
|
||
|
|
"loss": 0.8014,
|
||
|
|
"step": 535
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6269781461944235,
|
||
|
|
"grad_norm": 2.5497358605686826,
|
||
|
|
"learning_rate": 4.505666212573367e-07,
|
||
|
|
"loss": 0.7098,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.642049736247174,
|
||
|
|
"grad_norm": 2.451710015213246,
|
||
|
|
"learning_rate": 4.162582490144948e-07,
|
||
|
|
"loss": 0.7197,
|
||
|
|
"step": 545
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6571213262999245,
|
||
|
|
"grad_norm": 2.4788034308159634,
|
||
|
|
"learning_rate": 3.831895019292897e-07,
|
||
|
|
"loss": 0.7375,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6721929163526752,
|
||
|
|
"grad_norm": 3.07406346872796,
|
||
|
|
"learning_rate": 3.5138005271346643e-07,
|
||
|
|
"loss": 0.7549,
|
||
|
|
"step": 555
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.687264506405426,
|
||
|
|
"grad_norm": 2.6118807822457115,
|
||
|
|
"learning_rate": 3.208488249181216e-07,
|
||
|
|
"loss": 0.7433,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7023360964581764,
|
||
|
|
"grad_norm": 3.0584684036169865,
|
||
|
|
"learning_rate": 2.9161398167602053e-07,
|
||
|
|
"loss": 0.7914,
|
||
|
|
"step": 565
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.717407686510927,
|
||
|
|
"grad_norm": 2.763834415959353,
|
||
|
|
"learning_rate": 2.6369291489628244e-07,
|
||
|
|
"loss": 0.7405,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7324792765636774,
|
||
|
|
"grad_norm": 2.1449947701115684,
|
||
|
|
"learning_rate": 2.3710223491787643e-07,
|
||
|
|
"loss": 0.7149,
|
||
|
|
"step": 575
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.747550866616428,
|
||
|
|
"grad_norm": 2.63033187099936,
|
||
|
|
"learning_rate": 2.1185776062807307e-07,
|
||
|
|
"loss": 0.7274,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7626224566691786,
|
||
|
|
"grad_norm": 2.8862953965332547,
|
||
|
|
"learning_rate": 1.8797451005173384e-07,
|
||
|
|
"loss": 0.7661,
|
||
|
|
"step": 585
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7776940467219293,
|
||
|
|
"grad_norm": 2.7608278671213897,
|
||
|
|
"learning_rate": 1.654666914170361e-07,
|
||
|
|
"loss": 0.7625,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7927656367746798,
|
||
|
|
"grad_norm": 3.3490002122178,
|
||
|
|
"learning_rate": 1.44347694702949e-07,
|
||
|
|
"loss": 0.7985,
|
||
|
|
"step": 595
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8078372268274303,
|
||
|
|
"grad_norm": 2.5636592675505545,
|
||
|
|
"learning_rate": 1.24630083673489e-07,
|
||
|
|
"loss": 0.7436,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8229088168801808,
|
||
|
|
"grad_norm": 2.895653387694223,
|
||
|
|
"learning_rate": 1.0632558840349333e-07,
|
||
|
|
"loss": 0.7608,
|
||
|
|
"step": 605
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8379804069329313,
|
||
|
|
"grad_norm": 2.553919116507747,
|
||
|
|
"learning_rate": 8.944509830035864e-08,
|
||
|
|
"loss": 0.7131,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.853051996985682,
|
||
|
|
"grad_norm": 2.849246844898084,
|
||
|
|
"learning_rate": 7.399865562589315e-08,
|
||
|
|
"loss": 0.7841,
|
||
|
|
"step": 615
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8681235870384325,
|
||
|
|
"grad_norm": 3.01555827900457,
|
||
|
|
"learning_rate": 5.999544952214192e-08,
|
||
|
|
"loss": 0.7597,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8831951770911832,
|
||
|
|
"grad_norm": 2.5969187446302806,
|
||
|
|
"learning_rate": 4.7443810544734456e-08,
|
||
|
|
"loss": 0.731,
|
||
|
|
"step": 625
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8982667671439337,
|
||
|
|
"grad_norm": 2.7609437293219288,
|
||
|
|
"learning_rate": 3.635120570700784e-08,
|
||
|
|
"loss": 0.7767,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9133383571966842,
|
||
|
|
"grad_norm": 2.6128372521247583,
|
||
|
|
"learning_rate": 2.67242340378554e-08,
|
||
|
|
"loss": 0.7445,
|
||
|
|
"step": 635
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9284099472494347,
|
||
|
|
"grad_norm": 3.0201462908095302,
|
||
|
|
"learning_rate": 1.856862265594117e-08,
|
||
|
|
"loss": 0.7317,
|
||
|
|
"step": 640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9434815373021854,
|
||
|
|
"grad_norm": 3.1957975068044098,
|
||
|
|
"learning_rate": 1.1889223362616664e-08,
|
||
|
|
"loss": 0.7479,
|
||
|
|
"step": 645
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.958553127354936,
|
||
|
|
"grad_norm": 2.7277480748536274,
|
||
|
|
"learning_rate": 6.690009755569715e-09,
|
||
|
|
"loss": 0.6665,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9736247174076866,
|
||
|
|
"grad_norm": 2.6562010103303586,
|
||
|
|
"learning_rate": 2.9740748649145778e-09,
|
||
|
|
"loss": 0.7905,
|
||
|
|
"step": 655
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9886963074604371,
|
||
|
|
"grad_norm": 2.719630315699159,
|
||
|
|
"learning_rate": 7.436293131396488e-10,
|
||
|
|
"loss": 0.7328,
|
||
|
|
"step": 660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"step": 664,
|
||
|
|
"total_flos": 25921569423360.0,
|
||
|
|
"train_loss": 0.9619188387709928,
|
||
|
|
"train_runtime": 4046.5288,
|
||
|
|
"train_samples_per_second": 10.494,
|
||
|
|
"train_steps_per_second": 0.164
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 5,
|
||
|
|
"max_steps": 664,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 2,
|
||
|
|
"save_steps": 350,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 25921569423360.0,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|