llama3.2_feedback_1b / trainer_state.json
terry69's picture
Model save
62f00bf verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1656,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006038647342995169,
"grad_norm": 10.485260000044514,
"learning_rate": 6.02409638554217e-08,
"loss": 1.7287,
"step": 1
},
{
"epoch": 0.0030193236714975845,
"grad_norm": 10.032015562071633,
"learning_rate": 3.0120481927710845e-07,
"loss": 1.6893,
"step": 5
},
{
"epoch": 0.006038647342995169,
"grad_norm": 9.612514394248656,
"learning_rate": 6.024096385542169e-07,
"loss": 1.7063,
"step": 10
},
{
"epoch": 0.009057971014492754,
"grad_norm": 7.557592473361227,
"learning_rate": 9.036144578313254e-07,
"loss": 1.6866,
"step": 15
},
{
"epoch": 0.012077294685990338,
"grad_norm": 4.565478227553907,
"learning_rate": 1.2048192771084338e-06,
"loss": 1.6072,
"step": 20
},
{
"epoch": 0.015096618357487922,
"grad_norm": 2.9798309816553026,
"learning_rate": 1.5060240963855425e-06,
"loss": 1.5475,
"step": 25
},
{
"epoch": 0.018115942028985508,
"grad_norm": 2.7008857218275306,
"learning_rate": 1.8072289156626508e-06,
"loss": 1.4782,
"step": 30
},
{
"epoch": 0.021135265700483092,
"grad_norm": 2.2357152537880713,
"learning_rate": 2.1084337349397595e-06,
"loss": 1.3857,
"step": 35
},
{
"epoch": 0.024154589371980676,
"grad_norm": 1.9475475298214286,
"learning_rate": 2.4096385542168676e-06,
"loss": 1.3258,
"step": 40
},
{
"epoch": 0.02717391304347826,
"grad_norm": 1.4792226037165068,
"learning_rate": 2.710843373493976e-06,
"loss": 1.2592,
"step": 45
},
{
"epoch": 0.030193236714975844,
"grad_norm": 1.2792537099768388,
"learning_rate": 3.012048192771085e-06,
"loss": 1.2245,
"step": 50
},
{
"epoch": 0.03321256038647343,
"grad_norm": 1.2523748122313099,
"learning_rate": 3.313253012048193e-06,
"loss": 1.1959,
"step": 55
},
{
"epoch": 0.036231884057971016,
"grad_norm": 1.0864985697164489,
"learning_rate": 3.6144578313253016e-06,
"loss": 1.1785,
"step": 60
},
{
"epoch": 0.0392512077294686,
"grad_norm": 1.142977110301484,
"learning_rate": 3.91566265060241e-06,
"loss": 1.1762,
"step": 65
},
{
"epoch": 0.042270531400966184,
"grad_norm": 1.098649172466519,
"learning_rate": 4.216867469879519e-06,
"loss": 1.1607,
"step": 70
},
{
"epoch": 0.04528985507246377,
"grad_norm": 1.057865362960247,
"learning_rate": 4.518072289156627e-06,
"loss": 1.1589,
"step": 75
},
{
"epoch": 0.04830917874396135,
"grad_norm": 1.054325832096538,
"learning_rate": 4.819277108433735e-06,
"loss": 1.1344,
"step": 80
},
{
"epoch": 0.051328502415458936,
"grad_norm": 1.099271650362997,
"learning_rate": 5.120481927710844e-06,
"loss": 1.1252,
"step": 85
},
{
"epoch": 0.05434782608695652,
"grad_norm": 1.0998416744125088,
"learning_rate": 5.421686746987952e-06,
"loss": 1.1215,
"step": 90
},
{
"epoch": 0.057367149758454104,
"grad_norm": 1.1151773621303298,
"learning_rate": 5.722891566265061e-06,
"loss": 1.1135,
"step": 95
},
{
"epoch": 0.06038647342995169,
"grad_norm": 1.0965263618680452,
"learning_rate": 6.02409638554217e-06,
"loss": 1.1255,
"step": 100
},
{
"epoch": 0.06340579710144928,
"grad_norm": 1.1335389413366264,
"learning_rate": 6.325301204819277e-06,
"loss": 1.0936,
"step": 105
},
{
"epoch": 0.06642512077294686,
"grad_norm": 1.130790372670886,
"learning_rate": 6.626506024096386e-06,
"loss": 1.0998,
"step": 110
},
{
"epoch": 0.06944444444444445,
"grad_norm": 1.1733531035655729,
"learning_rate": 6.927710843373494e-06,
"loss": 1.1035,
"step": 115
},
{
"epoch": 0.07246376811594203,
"grad_norm": 1.3208668126357825,
"learning_rate": 7.228915662650603e-06,
"loss": 1.0642,
"step": 120
},
{
"epoch": 0.07548309178743962,
"grad_norm": 1.3106018142699076,
"learning_rate": 7.530120481927712e-06,
"loss": 1.0809,
"step": 125
},
{
"epoch": 0.0785024154589372,
"grad_norm": 1.1965232933467733,
"learning_rate": 7.83132530120482e-06,
"loss": 1.0739,
"step": 130
},
{
"epoch": 0.08152173913043478,
"grad_norm": 1.196930618773617,
"learning_rate": 8.132530120481928e-06,
"loss": 1.0711,
"step": 135
},
{
"epoch": 0.08454106280193237,
"grad_norm": 1.135659652675276,
"learning_rate": 8.433734939759038e-06,
"loss": 1.0489,
"step": 140
},
{
"epoch": 0.08756038647342995,
"grad_norm": 1.1334631389824636,
"learning_rate": 8.734939759036145e-06,
"loss": 1.0737,
"step": 145
},
{
"epoch": 0.09057971014492754,
"grad_norm": 1.2418129844310752,
"learning_rate": 9.036144578313254e-06,
"loss": 1.0458,
"step": 150
},
{
"epoch": 0.09359903381642512,
"grad_norm": 1.2437853442182127,
"learning_rate": 9.337349397590362e-06,
"loss": 1.036,
"step": 155
},
{
"epoch": 0.0966183574879227,
"grad_norm": 1.1287786590038726,
"learning_rate": 9.63855421686747e-06,
"loss": 1.0394,
"step": 160
},
{
"epoch": 0.09963768115942029,
"grad_norm": 1.1656244356824084,
"learning_rate": 9.93975903614458e-06,
"loss": 1.0357,
"step": 165
},
{
"epoch": 0.10265700483091787,
"grad_norm": 1.2290880544327538,
"learning_rate": 9.999822178354131e-06,
"loss": 1.0407,
"step": 170
},
{
"epoch": 0.10567632850241546,
"grad_norm": 1.1787986003632454,
"learning_rate": 9.999099799595088e-06,
"loss": 1.0202,
"step": 175
},
{
"epoch": 0.10869565217391304,
"grad_norm": 1.3579493877689446,
"learning_rate": 9.997821830092095e-06,
"loss": 0.9995,
"step": 180
},
{
"epoch": 0.11171497584541062,
"grad_norm": 1.236495272199702,
"learning_rate": 9.995988411876328e-06,
"loss": 1.0094,
"step": 185
},
{
"epoch": 0.11473429951690821,
"grad_norm": 1.1891336346777943,
"learning_rate": 9.993599748710505e-06,
"loss": 1.0038,
"step": 190
},
{
"epoch": 0.11775362318840579,
"grad_norm": 1.1457890107370514,
"learning_rate": 9.990656106066257e-06,
"loss": 0.9842,
"step": 195
},
{
"epoch": 0.12077294685990338,
"grad_norm": 1.197983166001558,
"learning_rate": 9.9871578110946e-06,
"loss": 0.9947,
"step": 200
},
{
"epoch": 0.12379227053140096,
"grad_norm": 1.2276842269473731,
"learning_rate": 9.983105252589599e-06,
"loss": 0.9743,
"step": 205
},
{
"epoch": 0.12681159420289856,
"grad_norm": 1.2305159106762475,
"learning_rate": 9.978498880945138e-06,
"loss": 0.9699,
"step": 210
},
{
"epoch": 0.12983091787439613,
"grad_norm": 1.287504685759514,
"learning_rate": 9.97333920810488e-06,
"loss": 0.9774,
"step": 215
},
{
"epoch": 0.13285024154589373,
"grad_norm": 1.2170202059428485,
"learning_rate": 9.967626807505359e-06,
"loss": 0.9723,
"step": 220
},
{
"epoch": 0.1358695652173913,
"grad_norm": 1.1891615466242549,
"learning_rate": 9.961362314012258e-06,
"loss": 0.9634,
"step": 225
},
{
"epoch": 0.1388888888888889,
"grad_norm": 1.259330792562172,
"learning_rate": 9.954546423849842e-06,
"loss": 0.9549,
"step": 230
},
{
"epoch": 0.14190821256038647,
"grad_norm": 1.2632205741447957,
"learning_rate": 9.947179894523594e-06,
"loss": 0.9464,
"step": 235
},
{
"epoch": 0.14492753623188406,
"grad_norm": 1.1804173499957151,
"learning_rate": 9.93926354473601e-06,
"loss": 0.9398,
"step": 240
},
{
"epoch": 0.14794685990338163,
"grad_norm": 1.1537804333636499,
"learning_rate": 9.930798254295628e-06,
"loss": 0.9653,
"step": 245
},
{
"epoch": 0.15096618357487923,
"grad_norm": 1.1421841081498953,
"learning_rate": 9.921784964019234e-06,
"loss": 0.9358,
"step": 250
},
{
"epoch": 0.1539855072463768,
"grad_norm": 1.1449230835912114,
"learning_rate": 9.91222467562731e-06,
"loss": 0.9491,
"step": 255
},
{
"epoch": 0.1570048309178744,
"grad_norm": 1.063437049402578,
"learning_rate": 9.902118451632694e-06,
"loss": 0.9465,
"step": 260
},
{
"epoch": 0.16002415458937197,
"grad_norm": 1.085929837294407,
"learning_rate": 9.891467415222511e-06,
"loss": 0.9275,
"step": 265
},
{
"epoch": 0.16304347826086957,
"grad_norm": 1.1524817196228028,
"learning_rate": 9.880272750133328e-06,
"loss": 0.9213,
"step": 270
},
{
"epoch": 0.16606280193236714,
"grad_norm": 1.1194662690657433,
"learning_rate": 9.868535700519605e-06,
"loss": 0.9315,
"step": 275
},
{
"epoch": 0.16908212560386474,
"grad_norm": 1.0905003429988187,
"learning_rate": 9.856257570815415e-06,
"loss": 0.9282,
"step": 280
},
{
"epoch": 0.1721014492753623,
"grad_norm": 1.1166663888676513,
"learning_rate": 9.843439725589481e-06,
"loss": 0.9281,
"step": 285
},
{
"epoch": 0.1751207729468599,
"grad_norm": 1.1630718200026984,
"learning_rate": 9.83008358939351e-06,
"loss": 0.9088,
"step": 290
},
{
"epoch": 0.17814009661835747,
"grad_norm": 1.1447207666491321,
"learning_rate": 9.81619064660388e-06,
"loss": 0.9035,
"step": 295
},
{
"epoch": 0.18115942028985507,
"grad_norm": 1.1556484647632965,
"learning_rate": 9.801762441256663e-06,
"loss": 0.9182,
"step": 300
},
{
"epoch": 0.18417874396135267,
"grad_norm": 1.091531593613017,
"learning_rate": 9.786800576876026e-06,
"loss": 0.907,
"step": 305
},
{
"epoch": 0.18719806763285024,
"grad_norm": 1.0623938281401049,
"learning_rate": 9.77130671629602e-06,
"loss": 0.9087,
"step": 310
},
{
"epoch": 0.19021739130434784,
"grad_norm": 1.1041557388108325,
"learning_rate": 9.755282581475769e-06,
"loss": 0.9097,
"step": 315
},
{
"epoch": 0.1932367149758454,
"grad_norm": 1.1073654251326603,
"learning_rate": 9.738729953308104e-06,
"loss": 0.9231,
"step": 320
},
{
"epoch": 0.196256038647343,
"grad_norm": 1.0823061373486982,
"learning_rate": 9.72165067142163e-06,
"loss": 0.908,
"step": 325
},
{
"epoch": 0.19927536231884058,
"grad_norm": 1.0339903077159918,
"learning_rate": 9.70404663397628e-06,
"loss": 0.9107,
"step": 330
},
{
"epoch": 0.20229468599033817,
"grad_norm": 1.162588874842489,
"learning_rate": 9.68591979745235e-06,
"loss": 0.8985,
"step": 335
},
{
"epoch": 0.20531400966183574,
"grad_norm": 1.1076782260164195,
"learning_rate": 9.667272176433063e-06,
"loss": 0.8979,
"step": 340
},
{
"epoch": 0.20833333333333334,
"grad_norm": 1.0465799767504245,
"learning_rate": 9.648105843380674e-06,
"loss": 0.9082,
"step": 345
},
{
"epoch": 0.2113526570048309,
"grad_norm": 1.1501313415042025,
"learning_rate": 9.628422928406133e-06,
"loss": 0.9,
"step": 350
},
{
"epoch": 0.2143719806763285,
"grad_norm": 1.0644785823217744,
"learning_rate": 9.608225619032361e-06,
"loss": 0.9029,
"step": 355
},
{
"epoch": 0.21739130434782608,
"grad_norm": 1.0757621081585198,
"learning_rate": 9.587516159951118e-06,
"loss": 0.8902,
"step": 360
},
{
"epoch": 0.22041062801932368,
"grad_norm": 1.0472929816511527,
"learning_rate": 9.566296852773541e-06,
"loss": 0.8921,
"step": 365
},
{
"epoch": 0.22342995169082125,
"grad_norm": 1.0276460031701986,
"learning_rate": 9.544570055774348e-06,
"loss": 0.902,
"step": 370
},
{
"epoch": 0.22644927536231885,
"grad_norm": 1.0354757235612004,
"learning_rate": 9.522338183629737e-06,
"loss": 0.9007,
"step": 375
},
{
"epoch": 0.22946859903381642,
"grad_norm": 1.0665922729683273,
"learning_rate": 9.499603707149035e-06,
"loss": 0.8954,
"step": 380
},
{
"epoch": 0.23248792270531402,
"grad_norm": 1.0107097936080194,
"learning_rate": 9.476369153000076e-06,
"loss": 0.8903,
"step": 385
},
{
"epoch": 0.23550724637681159,
"grad_norm": 1.0771246264884797,
"learning_rate": 9.45263710342842e-06,
"loss": 0.8959,
"step": 390
},
{
"epoch": 0.23852657004830918,
"grad_norm": 1.0423588321422452,
"learning_rate": 9.428410195970337e-06,
"loss": 0.8816,
"step": 395
},
{
"epoch": 0.24154589371980675,
"grad_norm": 1.0678506449777372,
"learning_rate": 9.403691123159707e-06,
"loss": 0.895,
"step": 400
},
{
"epoch": 0.24456521739130435,
"grad_norm": 1.081407003837132,
"learning_rate": 9.378482632228745e-06,
"loss": 0.8995,
"step": 405
},
{
"epoch": 0.24758454106280192,
"grad_norm": 0.9982667658619887,
"learning_rate": 9.352787524802707e-06,
"loss": 0.8921,
"step": 410
},
{
"epoch": 0.2506038647342995,
"grad_norm": 0.9993884156920768,
"learning_rate": 9.326608656588502e-06,
"loss": 0.8697,
"step": 415
},
{
"epoch": 0.2536231884057971,
"grad_norm": 0.9802873427819824,
"learning_rate": 9.299948937057325e-06,
"loss": 0.8783,
"step": 420
},
{
"epoch": 0.25664251207729466,
"grad_norm": 1.070955381882503,
"learning_rate": 9.272811329121305e-06,
"loss": 0.8712,
"step": 425
},
{
"epoch": 0.25966183574879226,
"grad_norm": 0.9929872238415203,
"learning_rate": 9.245198848804197e-06,
"loss": 0.8615,
"step": 430
},
{
"epoch": 0.26268115942028986,
"grad_norm": 1.0161446197547341,
"learning_rate": 9.217114564906208e-06,
"loss": 0.8741,
"step": 435
},
{
"epoch": 0.26570048309178745,
"grad_norm": 1.0336101978655394,
"learning_rate": 9.188561598662921e-06,
"loss": 0.8625,
"step": 440
},
{
"epoch": 0.26871980676328505,
"grad_norm": 0.987163288450998,
"learning_rate": 9.159543123398416e-06,
"loss": 0.8867,
"step": 445
},
{
"epoch": 0.2717391304347826,
"grad_norm": 0.9812717125356297,
"learning_rate": 9.130062364172582e-06,
"loss": 0.8765,
"step": 450
},
{
"epoch": 0.2747584541062802,
"grad_norm": 1.0149030855230246,
"learning_rate": 9.1001225974227e-06,
"loss": 0.8836,
"step": 455
},
{
"epoch": 0.2777777777777778,
"grad_norm": 0.9859822686466229,
"learning_rate": 9.0697271505993e-06,
"loss": 0.8863,
"step": 460
},
{
"epoch": 0.2807971014492754,
"grad_norm": 0.9863559376411042,
"learning_rate": 9.038879401796358e-06,
"loss": 0.8764,
"step": 465
},
{
"epoch": 0.28381642512077293,
"grad_norm": 0.9709580333992508,
"learning_rate": 9.00758277937586e-06,
"loss": 0.8708,
"step": 470
},
{
"epoch": 0.28683574879227053,
"grad_norm": 0.9672345556799925,
"learning_rate": 8.975840761586772e-06,
"loss": 0.8685,
"step": 475
},
{
"epoch": 0.2898550724637681,
"grad_norm": 1.0442405044492378,
"learning_rate": 8.94365687617849e-06,
"loss": 0.8714,
"step": 480
},
{
"epoch": 0.2928743961352657,
"grad_norm": 1.003497060733472,
"learning_rate": 8.911034700008757e-06,
"loss": 0.8723,
"step": 485
},
{
"epoch": 0.29589371980676327,
"grad_norm": 0.9912702297034715,
"learning_rate": 8.87797785864615e-06,
"loss": 0.8755,
"step": 490
},
{
"epoch": 0.29891304347826086,
"grad_norm": 0.9620747196303233,
"learning_rate": 8.844490025967126e-06,
"loss": 0.8674,
"step": 495
},
{
"epoch": 0.30193236714975846,
"grad_norm": 1.0470048364648747,
"learning_rate": 8.810574923747729e-06,
"loss": 0.8555,
"step": 500
},
{
"epoch": 0.30495169082125606,
"grad_norm": 0.9573341316350298,
"learning_rate": 8.776236321249955e-06,
"loss": 0.8806,
"step": 505
},
{
"epoch": 0.3079710144927536,
"grad_norm": 0.9951662459035352,
"learning_rate": 8.741478034802835e-06,
"loss": 0.8621,
"step": 510
},
{
"epoch": 0.3109903381642512,
"grad_norm": 0.9902413122683058,
"learning_rate": 8.706303927378306e-06,
"loss": 0.8673,
"step": 515
},
{
"epoch": 0.3140096618357488,
"grad_norm": 1.062455795370126,
"learning_rate": 8.670717908161878e-06,
"loss": 0.8728,
"step": 520
},
{
"epoch": 0.3170289855072464,
"grad_norm": 1.0382561615088126,
"learning_rate": 8.634723932118184e-06,
"loss": 0.8708,
"step": 525
},
{
"epoch": 0.32004830917874394,
"grad_norm": 0.9807369655332918,
"learning_rate": 8.598325999551425e-06,
"loss": 0.869,
"step": 530
},
{
"epoch": 0.32306763285024154,
"grad_norm": 1.041804239187294,
"learning_rate": 8.56152815566078e-06,
"loss": 0.8603,
"step": 535
},
{
"epoch": 0.32608695652173914,
"grad_norm": 0.98653685926571,
"learning_rate": 8.524334490090848e-06,
"loss": 0.8601,
"step": 540
},
{
"epoch": 0.32910628019323673,
"grad_norm": 1.0824551595193905,
"learning_rate": 8.486749136477112e-06,
"loss": 0.8527,
"step": 545
},
{
"epoch": 0.3321256038647343,
"grad_norm": 1.012559998288933,
"learning_rate": 8.448776271986542e-06,
"loss": 0.8484,
"step": 550
},
{
"epoch": 0.3351449275362319,
"grad_norm": 1.0472603651755694,
"learning_rate": 8.41042011685336e-06,
"loss": 0.8764,
"step": 555
},
{
"epoch": 0.33816425120772947,
"grad_norm": 1.0091109552760384,
"learning_rate": 8.371684933909996e-06,
"loss": 0.8778,
"step": 560
},
{
"epoch": 0.34118357487922707,
"grad_norm": 0.9679690923784716,
"learning_rate": 8.33257502811334e-06,
"loss": 0.8573,
"step": 565
},
{
"epoch": 0.3442028985507246,
"grad_norm": 1.0239715610365467,
"learning_rate": 8.293094746066283e-06,
"loss": 0.8759,
"step": 570
},
{
"epoch": 0.3472222222222222,
"grad_norm": 1.0215620182556282,
"learning_rate": 8.253248475534656e-06,
"loss": 0.8524,
"step": 575
},
{
"epoch": 0.3502415458937198,
"grad_norm": 1.0214058034537257,
"learning_rate": 8.213040644959572e-06,
"loss": 0.8479,
"step": 580
},
{
"epoch": 0.3532608695652174,
"grad_norm": 0.9901787487365526,
"learning_rate": 8.172475722965263e-06,
"loss": 0.8387,
"step": 585
},
{
"epoch": 0.35628019323671495,
"grad_norm": 0.9666966190379256,
"learning_rate": 8.131558217862444e-06,
"loss": 0.8389,
"step": 590
},
{
"epoch": 0.35929951690821255,
"grad_norm": 0.9919380582986378,
"learning_rate": 8.090292677147268e-06,
"loss": 0.8623,
"step": 595
},
{
"epoch": 0.36231884057971014,
"grad_norm": 0.9829820982309616,
"learning_rate": 8.048683686995921e-06,
"loss": 0.8483,
"step": 600
},
{
"epoch": 0.36533816425120774,
"grad_norm": 1.0033573699240617,
"learning_rate": 8.006735871754932e-06,
"loss": 0.8467,
"step": 605
},
{
"epoch": 0.36835748792270534,
"grad_norm": 0.9729124425349844,
"learning_rate": 7.96445389342722e-06,
"loss": 0.8482,
"step": 610
},
{
"epoch": 0.3713768115942029,
"grad_norm": 0.9846164642356084,
"learning_rate": 7.921842451153982e-06,
"loss": 0.8585,
"step": 615
},
{
"epoch": 0.3743961352657005,
"grad_norm": 1.0204270522909766,
"learning_rate": 7.878906280692424e-06,
"loss": 0.8609,
"step": 620
},
{
"epoch": 0.3774154589371981,
"grad_norm": 0.9508580696057062,
"learning_rate": 7.835650153889449e-06,
"loss": 0.8559,
"step": 625
},
{
"epoch": 0.3804347826086957,
"grad_norm": 1.0138140609913775,
"learning_rate": 7.792078878151318e-06,
"loss": 0.8369,
"step": 630
},
{
"epoch": 0.3834541062801932,
"grad_norm": 0.9931878326042032,
"learning_rate": 7.748197295909359e-06,
"loss": 0.8505,
"step": 635
},
{
"epoch": 0.3864734299516908,
"grad_norm": 1.0342608052140974,
"learning_rate": 7.704010284081801e-06,
"loss": 0.849,
"step": 640
},
{
"epoch": 0.3894927536231884,
"grad_norm": 0.9819364059644684,
"learning_rate": 7.65952275353175e-06,
"loss": 0.8368,
"step": 645
},
{
"epoch": 0.392512077294686,
"grad_norm": 1.0187861799083533,
"learning_rate": 7.614739648521412e-06,
"loss": 0.8391,
"step": 650
},
{
"epoch": 0.39553140096618356,
"grad_norm": 1.0055302486721571,
"learning_rate": 7.56966594616259e-06,
"loss": 0.8473,
"step": 655
},
{
"epoch": 0.39855072463768115,
"grad_norm": 1.0213194412236104,
"learning_rate": 7.524306655863544e-06,
"loss": 0.8584,
"step": 660
},
{
"epoch": 0.40157004830917875,
"grad_norm": 1.0457938971980807,
"learning_rate": 7.478666818772252e-06,
"loss": 0.8462,
"step": 665
},
{
"epoch": 0.40458937198067635,
"grad_norm": 0.9848170024807922,
"learning_rate": 7.432751507216146e-06,
"loss": 0.844,
"step": 670
},
{
"epoch": 0.4076086956521739,
"grad_norm": 1.0787598071037892,
"learning_rate": 7.386565824138378e-06,
"loss": 0.8326,
"step": 675
},
{
"epoch": 0.4106280193236715,
"grad_norm": 0.9847580752716315,
"learning_rate": 7.3401149025306995e-06,
"loss": 0.853,
"step": 680
},
{
"epoch": 0.4136473429951691,
"grad_norm": 1.0402445922973875,
"learning_rate": 7.293403904862981e-06,
"loss": 0.849,
"step": 685
},
{
"epoch": 0.4166666666666667,
"grad_norm": 0.952409187764052,
"learning_rate": 7.246438022509465e-06,
"loss": 0.8323,
"step": 690
},
{
"epoch": 0.4196859903381642,
"grad_norm": 1.0019800169099848,
"learning_rate": 7.199222475171812e-06,
"loss": 0.8508,
"step": 695
},
{
"epoch": 0.4227053140096618,
"grad_norm": 0.9868139541797347,
"learning_rate": 7.151762510298985e-06,
"loss": 0.8313,
"step": 700
},
{
"epoch": 0.4257246376811594,
"grad_norm": 0.9834600378684528,
"learning_rate": 7.104063402504065e-06,
"loss": 0.8431,
"step": 705
},
{
"epoch": 0.428743961352657,
"grad_norm": 0.9566341224700101,
"learning_rate": 7.056130452978039e-06,
"loss": 0.8291,
"step": 710
},
{
"epoch": 0.43176328502415456,
"grad_norm": 0.9504357134569014,
"learning_rate": 7.0079689889006275e-06,
"loss": 0.8176,
"step": 715
},
{
"epoch": 0.43478260869565216,
"grad_norm": 0.987828355029499,
"learning_rate": 6.959584362848239e-06,
"loss": 0.8282,
"step": 720
},
{
"epoch": 0.43780193236714976,
"grad_norm": 0.9731535769551426,
"learning_rate": 6.910981952199097e-06,
"loss": 0.8376,
"step": 725
},
{
"epoch": 0.44082125603864736,
"grad_norm": 1.0015295092351129,
"learning_rate": 6.862167158535599e-06,
"loss": 0.828,
"step": 730
},
{
"epoch": 0.4438405797101449,
"grad_norm": 0.9879634179543377,
"learning_rate": 6.813145407044003e-06,
"loss": 0.8194,
"step": 735
},
{
"epoch": 0.4468599033816425,
"grad_norm": 0.9753047696912267,
"learning_rate": 6.763922145911474e-06,
"loss": 0.8336,
"step": 740
},
{
"epoch": 0.4498792270531401,
"grad_norm": 0.9924657298914499,
"learning_rate": 6.714502845720595e-06,
"loss": 0.8329,
"step": 745
},
{
"epoch": 0.4528985507246377,
"grad_norm": 1.0659305402069452,
"learning_rate": 6.664892998841361e-06,
"loss": 0.8224,
"step": 750
},
{
"epoch": 0.4559178743961353,
"grad_norm": 0.9926149466224716,
"learning_rate": 6.61509811882078e-06,
"loss": 0.8204,
"step": 755
},
{
"epoch": 0.45893719806763283,
"grad_norm": 1.0312896682847614,
"learning_rate": 6.565123739770102e-06,
"loss": 0.8205,
"step": 760
},
{
"epoch": 0.46195652173913043,
"grad_norm": 0.9742444674619639,
"learning_rate": 6.5149754157497645e-06,
"loss": 0.8281,
"step": 765
},
{
"epoch": 0.46497584541062803,
"grad_norm": 0.9571473848092585,
"learning_rate": 6.464658720152135e-06,
"loss": 0.8164,
"step": 770
},
{
"epoch": 0.46799516908212563,
"grad_norm": 0.9959898314723186,
"learning_rate": 6.41417924508208e-06,
"loss": 0.843,
"step": 775
},
{
"epoch": 0.47101449275362317,
"grad_norm": 0.9960648835738334,
"learning_rate": 6.363542600735486e-06,
"loss": 0.8167,
"step": 780
},
{
"epoch": 0.47403381642512077,
"grad_norm": 1.0006502116357578,
"learning_rate": 6.312754414775737e-06,
"loss": 0.8306,
"step": 785
},
{
"epoch": 0.47705314009661837,
"grad_norm": 1.0341128294537794,
"learning_rate": 6.261820331708275e-06,
"loss": 0.8404,
"step": 790
},
{
"epoch": 0.48007246376811596,
"grad_norm": 1.013484766876777,
"learning_rate": 6.210746012253277e-06,
"loss": 0.8383,
"step": 795
},
{
"epoch": 0.4830917874396135,
"grad_norm": 1.0048584571797106,
"learning_rate": 6.159537132716532e-06,
"loss": 0.8128,
"step": 800
},
{
"epoch": 0.4861111111111111,
"grad_norm": 1.0066412617634983,
"learning_rate": 6.108199384358595e-06,
"loss": 0.8182,
"step": 805
},
{
"epoch": 0.4891304347826087,
"grad_norm": 0.9873011172531366,
"learning_rate": 6.0567384727622566e-06,
"loss": 0.8206,
"step": 810
},
{
"epoch": 0.4921497584541063,
"grad_norm": 1.0092839597226688,
"learning_rate": 6.005160117198448e-06,
"loss": 0.8345,
"step": 815
},
{
"epoch": 0.49516908212560384,
"grad_norm": 0.9625783040223157,
"learning_rate": 5.953470049990605e-06,
"loss": 0.8195,
"step": 820
},
{
"epoch": 0.49818840579710144,
"grad_norm": 1.0146672682921258,
"learning_rate": 5.90167401587759e-06,
"loss": 0.8222,
"step": 825
},
{
"epoch": 0.501207729468599,
"grad_norm": 0.9750234992861176,
"learning_rate": 5.84977777137523e-06,
"loss": 0.8149,
"step": 830
},
{
"epoch": 0.5042270531400966,
"grad_norm": 1.0388609954354933,
"learning_rate": 5.797787084136556e-06,
"loss": 0.8297,
"step": 835
},
{
"epoch": 0.5072463768115942,
"grad_norm": 1.0092212264949234,
"learning_rate": 5.745707732310781e-06,
"loss": 0.8255,
"step": 840
},
{
"epoch": 0.5102657004830918,
"grad_norm": 1.0014867558076073,
"learning_rate": 5.693545503901149e-06,
"loss": 0.8212,
"step": 845
},
{
"epoch": 0.5132850241545893,
"grad_norm": 0.9955486191780001,
"learning_rate": 5.641306196121643e-06,
"loss": 0.8202,
"step": 850
},
{
"epoch": 0.5163043478260869,
"grad_norm": 1.033006798193324,
"learning_rate": 5.5889956147527156e-06,
"loss": 0.8209,
"step": 855
},
{
"epoch": 0.5193236714975845,
"grad_norm": 1.0429153983324886,
"learning_rate": 5.536619573496027e-06,
"loss": 0.8194,
"step": 860
},
{
"epoch": 0.5223429951690821,
"grad_norm": 1.026989922350979,
"learning_rate": 5.484183893328332e-06,
"loss": 0.8199,
"step": 865
},
{
"epoch": 0.5253623188405797,
"grad_norm": 1.0238509806451253,
"learning_rate": 5.431694401854545e-06,
"loss": 0.7951,
"step": 870
},
{
"epoch": 0.5283816425120773,
"grad_norm": 0.9563995023957641,
"learning_rate": 5.379156932660067e-06,
"loss": 0.8185,
"step": 875
},
{
"epoch": 0.5314009661835749,
"grad_norm": 0.9676712123206674,
"learning_rate": 5.326577324662459e-06,
"loss": 0.8089,
"step": 880
},
{
"epoch": 0.5344202898550725,
"grad_norm": 0.9898039955174832,
"learning_rate": 5.273961421462505e-06,
"loss": 0.8238,
"step": 885
},
{
"epoch": 0.5374396135265701,
"grad_norm": 0.9452218463977323,
"learning_rate": 5.221315070694775e-06,
"loss": 0.8042,
"step": 890
},
{
"epoch": 0.5404589371980676,
"grad_norm": 1.0185049851273638,
"learning_rate": 5.168644123377725e-06,
"loss": 0.8017,
"step": 895
},
{
"epoch": 0.5434782608695652,
"grad_norm": 1.0173971075365176,
"learning_rate": 5.1159544332634256e-06,
"loss": 0.821,
"step": 900
},
{
"epoch": 0.5464975845410628,
"grad_norm": 1.0610338990705965,
"learning_rate": 5.063251856186991e-06,
"loss": 0.802,
"step": 905
},
{
"epoch": 0.5495169082125604,
"grad_norm": 0.978719611658541,
"learning_rate": 5.010542249415761e-06,
"loss": 0.8133,
"step": 910
},
{
"epoch": 0.552536231884058,
"grad_norm": 0.9889441669899723,
"learning_rate": 4.95783147099835e-06,
"loss": 0.8039,
"step": 915
},
{
"epoch": 0.5555555555555556,
"grad_norm": 1.0148896374998926,
"learning_rate": 4.90512537911358e-06,
"loss": 0.8146,
"step": 920
},
{
"epoch": 0.5585748792270532,
"grad_norm": 0.9891668368377164,
"learning_rate": 4.852429831419428e-06,
"loss": 0.8214,
"step": 925
},
{
"epoch": 0.5615942028985508,
"grad_norm": 1.041355982063113,
"learning_rate": 4.799750684402006e-06,
"loss": 0.8041,
"step": 930
},
{
"epoch": 0.5646135265700483,
"grad_norm": 1.0153214084697235,
"learning_rate": 4.747093792724679e-06,
"loss": 0.8121,
"step": 935
},
{
"epoch": 0.5676328502415459,
"grad_norm": 1.0183122535937061,
"learning_rate": 4.6944650085774095e-06,
"loss": 0.8002,
"step": 940
},
{
"epoch": 0.5706521739130435,
"grad_norm": 0.9957864079530732,
"learning_rate": 4.641870181026322e-06,
"loss": 0.8157,
"step": 945
},
{
"epoch": 0.5736714975845411,
"grad_norm": 1.0370461103314468,
"learning_rate": 4.589315155363683e-06,
"loss": 0.811,
"step": 950
},
{
"epoch": 0.5766908212560387,
"grad_norm": 1.004518929806163,
"learning_rate": 4.53680577245824e-06,
"loss": 0.8016,
"step": 955
},
{
"epoch": 0.5797101449275363,
"grad_norm": 0.9969753713374386,
"learning_rate": 4.484347868106097e-06,
"loss": 0.7989,
"step": 960
},
{
"epoch": 0.5827294685990339,
"grad_norm": 1.007321359295188,
"learning_rate": 4.431947272382118e-06,
"loss": 0.8094,
"step": 965
},
{
"epoch": 0.5857487922705314,
"grad_norm": 0.98791492635868,
"learning_rate": 4.379609808992e-06,
"loss": 0.8143,
"step": 970
},
{
"epoch": 0.5887681159420289,
"grad_norm": 0.9922336761644133,
"learning_rate": 4.327341294625019e-06,
"loss": 0.7996,
"step": 975
},
{
"epoch": 0.5917874396135265,
"grad_norm": 0.9725954746289334,
"learning_rate": 4.275147538307594e-06,
"loss": 0.8133,
"step": 980
},
{
"epoch": 0.5948067632850241,
"grad_norm": 0.9717320614269667,
"learning_rate": 4.223034340757666e-06,
"loss": 0.8005,
"step": 985
},
{
"epoch": 0.5978260869565217,
"grad_norm": 0.9389495916799526,
"learning_rate": 4.171007493740023e-06,
"loss": 0.7958,
"step": 990
},
{
"epoch": 0.6008454106280193,
"grad_norm": 0.9810023298322507,
"learning_rate": 4.1190727794226175e-06,
"loss": 0.8165,
"step": 995
},
{
"epoch": 0.6038647342995169,
"grad_norm": 0.9779941495613733,
"learning_rate": 4.067235969733937e-06,
"loss": 0.7976,
"step": 1000
},
{
"epoch": 0.6068840579710145,
"grad_norm": 0.9510093739942304,
"learning_rate": 4.015502825721537e-06,
"loss": 0.8073,
"step": 1005
},
{
"epoch": 0.6099033816425121,
"grad_norm": 0.9936696587970497,
"learning_rate": 3.963879096911751e-06,
"loss": 0.8068,
"step": 1010
},
{
"epoch": 0.6129227053140096,
"grad_norm": 0.9978910530135231,
"learning_rate": 3.91237052067072e-06,
"loss": 0.8062,
"step": 1015
},
{
"epoch": 0.6159420289855072,
"grad_norm": 0.9723677719807516,
"learning_rate": 3.860982821566729e-06,
"loss": 0.8178,
"step": 1020
},
{
"epoch": 0.6189613526570048,
"grad_norm": 0.9657700887809438,
"learning_rate": 3.8097217107340107e-06,
"loss": 0.7836,
"step": 1025
},
{
"epoch": 0.6219806763285024,
"grad_norm": 0.9716901570758927,
"learning_rate": 3.7585928852380025e-06,
"loss": 0.7911,
"step": 1030
},
{
"epoch": 0.625,
"grad_norm": 1.0126975602818444,
"learning_rate": 3.7076020274421996e-06,
"loss": 0.8127,
"step": 1035
},
{
"epoch": 0.6280193236714976,
"grad_norm": 1.0012384413238815,
"learning_rate": 3.6567548043766157e-06,
"loss": 0.7926,
"step": 1040
},
{
"epoch": 0.6310386473429952,
"grad_norm": 1.0533344104817726,
"learning_rate": 3.6060568671079658e-06,
"loss": 0.7971,
"step": 1045
},
{
"epoch": 0.6340579710144928,
"grad_norm": 0.9909170809609505,
"learning_rate": 3.5555138501116247e-06,
"loss": 0.8032,
"step": 1050
},
{
"epoch": 0.6370772946859904,
"grad_norm": 1.0058635009835104,
"learning_rate": 3.5051313706453995e-06,
"loss": 0.8088,
"step": 1055
},
{
"epoch": 0.6400966183574879,
"grad_norm": 0.982286478542063,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.8021,
"step": 1060
},
{
"epoch": 0.6431159420289855,
"grad_norm": 0.9912609730677274,
"learning_rate": 3.40487040350303e-06,
"loss": 0.8048,
"step": 1065
},
{
"epoch": 0.6461352657004831,
"grad_norm": 1.0240454701942718,
"learning_rate": 3.355003058646105e-06,
"loss": 0.8219,
"step": 1070
},
{
"epoch": 0.6491545893719807,
"grad_norm": 0.955787854052142,
"learning_rate": 3.305318535719343e-06,
"loss": 0.7735,
"step": 1075
},
{
"epoch": 0.6521739130434783,
"grad_norm": 0.9556171904961169,
"learning_rate": 3.2558223565691104e-06,
"loss": 0.7955,
"step": 1080
},
{
"epoch": 0.6551932367149759,
"grad_norm": 0.9869405009720019,
"learning_rate": 3.2065200221095905e-06,
"loss": 0.7903,
"step": 1085
},
{
"epoch": 0.6582125603864735,
"grad_norm": 1.0184010512565573,
"learning_rate": 3.1574170117114293e-06,
"loss": 0.8059,
"step": 1090
},
{
"epoch": 0.6612318840579711,
"grad_norm": 0.9615842976570751,
"learning_rate": 3.1085187825927555e-06,
"loss": 0.7919,
"step": 1095
},
{
"epoch": 0.6642512077294686,
"grad_norm": 1.0051644960541348,
"learning_rate": 3.0598307692126904e-06,
"loss": 0.8025,
"step": 1100
},
{
"epoch": 0.6672705314009661,
"grad_norm": 0.9641208944556371,
"learning_rate": 3.0113583826673655e-06,
"loss": 0.799,
"step": 1105
},
{
"epoch": 0.6702898550724637,
"grad_norm": 0.9925774913072852,
"learning_rate": 2.9631070100885373e-06,
"loss": 0.8106,
"step": 1110
},
{
"epoch": 0.6733091787439613,
"grad_norm": 1.0107634066694202,
"learning_rate": 2.915082014044883e-06,
"loss": 0.8057,
"step": 1115
},
{
"epoch": 0.6763285024154589,
"grad_norm": 1.0046756211140373,
"learning_rate": 2.867288731946004e-06,
"loss": 0.8031,
"step": 1120
},
{
"epoch": 0.6793478260869565,
"grad_norm": 1.0053926491433522,
"learning_rate": 2.8197324754492456e-06,
"loss": 0.7921,
"step": 1125
},
{
"epoch": 0.6823671497584541,
"grad_norm": 1.006856299076646,
"learning_rate": 2.7724185298693596e-06,
"loss": 0.792,
"step": 1130
},
{
"epoch": 0.6853864734299517,
"grad_norm": 0.9902733067610593,
"learning_rate": 2.7253521535911144e-06,
"loss": 0.8113,
"step": 1135
},
{
"epoch": 0.6884057971014492,
"grad_norm": 1.022489282390403,
"learning_rate": 2.678538577484871e-06,
"loss": 0.802,
"step": 1140
},
{
"epoch": 0.6914251207729468,
"grad_norm": 1.0032934808608855,
"learning_rate": 2.6319830043252616e-06,
"loss": 0.7909,
"step": 1145
},
{
"epoch": 0.6944444444444444,
"grad_norm": 1.0127829062494287,
"learning_rate": 2.5856906082129313e-06,
"loss": 0.8063,
"step": 1150
},
{
"epoch": 0.697463768115942,
"grad_norm": 1.0334651022516967,
"learning_rate": 2.53966653399952e-06,
"loss": 0.7905,
"step": 1155
},
{
"epoch": 0.7004830917874396,
"grad_norm": 0.9947842946950315,
"learning_rate": 2.4939158967158657e-06,
"loss": 0.8053,
"step": 1160
},
{
"epoch": 0.7035024154589372,
"grad_norm": 1.0205258197932596,
"learning_rate": 2.448443781003527e-06,
"loss": 0.7921,
"step": 1165
},
{
"epoch": 0.7065217391304348,
"grad_norm": 1.016719694781447,
"learning_rate": 2.403255240549693e-06,
"loss": 0.7954,
"step": 1170
},
{
"epoch": 0.7095410628019324,
"grad_norm": 0.9765393946874324,
"learning_rate": 2.3583552975255108e-06,
"loss": 0.7831,
"step": 1175
},
{
"epoch": 0.7125603864734299,
"grad_norm": 1.0013834706558566,
"learning_rate": 2.313748942027956e-06,
"loss": 0.798,
"step": 1180
},
{
"epoch": 0.7155797101449275,
"grad_norm": 0.9832682717375602,
"learning_rate": 2.269441131525213e-06,
"loss": 0.7868,
"step": 1185
},
{
"epoch": 0.7185990338164251,
"grad_norm": 1.012871482321928,
"learning_rate": 2.225436790305733e-06,
"loss": 0.8016,
"step": 1190
},
{
"epoch": 0.7216183574879227,
"grad_norm": 1.0215595382277152,
"learning_rate": 2.181740808930947e-06,
"loss": 0.7958,
"step": 1195
},
{
"epoch": 0.7246376811594203,
"grad_norm": 0.9908102708938528,
"learning_rate": 2.1383580436917452e-06,
"loss": 0.7881,
"step": 1200
},
{
"epoch": 0.7276570048309179,
"grad_norm": 0.9703952014882963,
"learning_rate": 2.0952933160687456e-06,
"loss": 0.7935,
"step": 1205
},
{
"epoch": 0.7306763285024155,
"grad_norm": 1.0107034346751447,
"learning_rate": 2.052551412196456e-06,
"loss": 0.7904,
"step": 1210
},
{
"epoch": 0.7336956521739131,
"grad_norm": 0.9826589600845559,
"learning_rate": 2.010137082331354e-06,
"loss": 0.78,
"step": 1215
},
{
"epoch": 0.7367149758454107,
"grad_norm": 1.040060721769402,
"learning_rate": 1.96805504032393e-06,
"loss": 0.7876,
"step": 1220
},
{
"epoch": 0.7397342995169082,
"grad_norm": 0.9802643555891328,
"learning_rate": 1.9263099630948274e-06,
"loss": 0.7926,
"step": 1225
},
{
"epoch": 0.7427536231884058,
"grad_norm": 0.9845466368440403,
"learning_rate": 1.8849064901150372e-06,
"loss": 0.7882,
"step": 1230
},
{
"epoch": 0.7457729468599034,
"grad_norm": 0.9835806835545442,
"learning_rate": 1.8438492228902893e-06,
"loss": 0.7891,
"step": 1235
},
{
"epoch": 0.748792270531401,
"grad_norm": 1.0076367530435018,
"learning_rate": 1.8031427244496357e-06,
"loss": 0.7867,
"step": 1240
},
{
"epoch": 0.7518115942028986,
"grad_norm": 1.0313420481341204,
"learning_rate": 1.7627915188383382e-06,
"loss": 0.7982,
"step": 1245
},
{
"epoch": 0.7548309178743962,
"grad_norm": 0.9516216668137772,
"learning_rate": 1.7228000906150672e-06,
"loss": 0.7916,
"step": 1250
},
{
"epoch": 0.7578502415458938,
"grad_norm": 0.9955635624833065,
"learning_rate": 1.6831728843534962e-06,
"loss": 0.7871,
"step": 1255
},
{
"epoch": 0.7608695652173914,
"grad_norm": 1.0446368549655927,
"learning_rate": 1.6439143041483352e-06,
"loss": 0.7949,
"step": 1260
},
{
"epoch": 0.7638888888888888,
"grad_norm": 1.0107540263173203,
"learning_rate": 1.6050287131258862e-06,
"loss": 0.7782,
"step": 1265
},
{
"epoch": 0.7669082125603864,
"grad_norm": 1.0090639381521684,
"learning_rate": 1.5665204329591066e-06,
"loss": 0.7733,
"step": 1270
},
{
"epoch": 0.769927536231884,
"grad_norm": 1.017965866931342,
"learning_rate": 1.528393743387328e-06,
"loss": 0.7854,
"step": 1275
},
{
"epoch": 0.7729468599033816,
"grad_norm": 0.9757529990214369,
"learning_rate": 1.4906528817406052e-06,
"loss": 0.7896,
"step": 1280
},
{
"epoch": 0.7759661835748792,
"grad_norm": 0.9627515940764363,
"learning_rate": 1.453302042468786e-06,
"loss": 0.7743,
"step": 1285
},
{
"epoch": 0.7789855072463768,
"grad_norm": 0.9813369690634022,
"learning_rate": 1.4163453766753537e-06,
"loss": 0.7861,
"step": 1290
},
{
"epoch": 0.7820048309178744,
"grad_norm": 0.950962763947152,
"learning_rate": 1.3797869916560692e-06,
"loss": 0.7769,
"step": 1295
},
{
"epoch": 0.785024154589372,
"grad_norm": 1.0035116704753089,
"learning_rate": 1.3436309504425137e-06,
"loss": 0.7842,
"step": 1300
},
{
"epoch": 0.7880434782608695,
"grad_norm": 1.0316977087419814,
"learning_rate": 1.3078812713505079e-06,
"loss": 0.7889,
"step": 1305
},
{
"epoch": 0.7910628019323671,
"grad_norm": 0.9724293464755874,
"learning_rate": 1.2725419275335404e-06,
"loss": 0.792,
"step": 1310
},
{
"epoch": 0.7940821256038647,
"grad_norm": 1.024840109809241,
"learning_rate": 1.237616846541192e-06,
"loss": 0.7868,
"step": 1315
},
{
"epoch": 0.7971014492753623,
"grad_norm": 1.0132634017753674,
"learning_rate": 1.2031099098826376e-06,
"loss": 0.7905,
"step": 1320
},
{
"epoch": 0.8001207729468599,
"grad_norm": 0.9844574186374294,
"learning_rate": 1.1690249525952569e-06,
"loss": 0.7908,
"step": 1325
},
{
"epoch": 0.8031400966183575,
"grad_norm": 1.013907969102555,
"learning_rate": 1.1353657628184217e-06,
"loss": 0.7698,
"step": 1330
},
{
"epoch": 0.8061594202898551,
"grad_norm": 1.0118537572839486,
"learning_rate": 1.1021360813724924e-06,
"loss": 0.7926,
"step": 1335
},
{
"epoch": 0.8091787439613527,
"grad_norm": 0.9789094346489745,
"learning_rate": 1.0693396013430552e-06,
"loss": 0.7882,
"step": 1340
},
{
"epoch": 0.8121980676328503,
"grad_norm": 0.9834676458951446,
"learning_rate": 1.036979967670494e-06,
"loss": 0.7928,
"step": 1345
},
{
"epoch": 0.8152173913043478,
"grad_norm": 1.0061732844801357,
"learning_rate": 1.0050607767448928e-06,
"loss": 0.7744,
"step": 1350
},
{
"epoch": 0.8182367149758454,
"grad_norm": 1.0145249274790882,
"learning_rate": 9.735855760063412e-07,
"loss": 0.7907,
"step": 1355
},
{
"epoch": 0.821256038647343,
"grad_norm": 1.0014057594170862,
"learning_rate": 9.425578635506721e-07,
"loss": 0.7821,
"step": 1360
},
{
"epoch": 0.8242753623188406,
"grad_norm": 0.9863204913817585,
"learning_rate": 9.119810877406998e-07,
"loss": 0.786,
"step": 1365
},
{
"epoch": 0.8272946859903382,
"grad_norm": 0.9959450118029448,
"learning_rate": 8.818586468229695e-07,
"loss": 0.7838,
"step": 1370
},
{
"epoch": 0.8303140096618358,
"grad_norm": 1.0702182133587888,
"learning_rate": 8.521938885500825e-07,
"loss": 0.786,
"step": 1375
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.9968624582335651,
"learning_rate": 8.229901098086335e-07,
"loss": 0.7862,
"step": 1380
},
{
"epoch": 0.836352657004831,
"grad_norm": 0.9863986016846847,
"learning_rate": 7.942505562528024e-07,
"loss": 0.787,
"step": 1385
},
{
"epoch": 0.8393719806763285,
"grad_norm": 0.9545460547627316,
"learning_rate": 7.659784219436373e-07,
"loss": 0.7731,
"step": 1390
},
{
"epoch": 0.842391304347826,
"grad_norm": 1.007290245733153,
"learning_rate": 7.381768489940678e-07,
"loss": 0.7667,
"step": 1395
},
{
"epoch": 0.8454106280193237,
"grad_norm": 0.9929498673353992,
"learning_rate": 7.108489272197089e-07,
"loss": 0.7895,
"step": 1400
},
{
"epoch": 0.8484299516908212,
"grad_norm": 0.9848636431503859,
"learning_rate": 6.839976937954479e-07,
"loss": 0.7891,
"step": 1405
},
{
"epoch": 0.8514492753623188,
"grad_norm": 1.0038021231533398,
"learning_rate": 6.576261329179123e-07,
"loss": 0.7863,
"step": 1410
},
{
"epoch": 0.8544685990338164,
"grad_norm": 1.0021524457556812,
"learning_rate": 6.317371754738044e-07,
"loss": 0.7933,
"step": 1415
},
{
"epoch": 0.857487922705314,
"grad_norm": 0.9768075572854382,
"learning_rate": 6.06333698714171e-07,
"loss": 0.7771,
"step": 1420
},
{
"epoch": 0.8605072463768116,
"grad_norm": 0.9820365924428122,
"learning_rate": 5.814185259346267e-07,
"loss": 0.777,
"step": 1425
},
{
"epoch": 0.8635265700483091,
"grad_norm": 0.9872147557767675,
"learning_rate": 5.56994426161584e-07,
"loss": 0.791,
"step": 1430
},
{
"epoch": 0.8665458937198067,
"grad_norm": 0.9829200122907301,
"learning_rate": 5.330641138445064e-07,
"loss": 0.7764,
"step": 1435
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.9944086706773513,
"learning_rate": 5.096302485542265e-07,
"loss": 0.7891,
"step": 1440
},
{
"epoch": 0.8725845410628019,
"grad_norm": 1.0225322366517011,
"learning_rate": 4.866954346873715e-07,
"loss": 0.7897,
"step": 1445
},
{
"epoch": 0.8756038647342995,
"grad_norm": 1.0049159635351292,
"learning_rate": 4.642622211769099e-07,
"loss": 0.7868,
"step": 1450
},
{
"epoch": 0.8786231884057971,
"grad_norm": 0.9901193471531758,
"learning_rate": 4.4233310120887387e-07,
"loss": 0.7786,
"step": 1455
},
{
"epoch": 0.8816425120772947,
"grad_norm": 0.9957012743951548,
"learning_rate": 4.209105119452628e-07,
"loss": 0.7836,
"step": 1460
},
{
"epoch": 0.8846618357487923,
"grad_norm": 1.0004791381791707,
"learning_rate": 3.999968342531918e-07,
"loss": 0.768,
"step": 1465
},
{
"epoch": 0.8876811594202898,
"grad_norm": 0.9987769951972304,
"learning_rate": 3.7959439244027727e-07,
"loss": 0.7749,
"step": 1470
},
{
"epoch": 0.8907004830917874,
"grad_norm": 0.9862109297591363,
"learning_rate": 3.5970545399632574e-07,
"loss": 0.778,
"step": 1475
},
{
"epoch": 0.893719806763285,
"grad_norm": 0.9791302034502961,
"learning_rate": 3.4033222934131914e-07,
"loss": 0.7771,
"step": 1480
},
{
"epoch": 0.8967391304347826,
"grad_norm": 0.9625302334808536,
"learning_rate": 3.214768715797656e-07,
"loss": 0.7638,
"step": 1485
},
{
"epoch": 0.8997584541062802,
"grad_norm": 0.9748416151171142,
"learning_rate": 3.0314147626139543e-07,
"loss": 0.7845,
"step": 1490
},
{
"epoch": 0.9027777777777778,
"grad_norm": 0.9949692829145934,
"learning_rate": 2.853280811482734e-07,
"loss": 0.7722,
"step": 1495
},
{
"epoch": 0.9057971014492754,
"grad_norm": 0.9744535368779195,
"learning_rate": 2.6803866598832216e-07,
"loss": 0.7794,
"step": 1500
},
{
"epoch": 0.908816425120773,
"grad_norm": 0.992041943581358,
"learning_rate": 2.5127515229529665e-07,
"loss": 0.7836,
"step": 1505
},
{
"epoch": 0.9118357487922706,
"grad_norm": 0.9813156647468022,
"learning_rate": 2.350394031352343e-07,
"loss": 0.7863,
"step": 1510
},
{
"epoch": 0.9148550724637681,
"grad_norm": 0.9963647656973325,
"learning_rate": 2.1933322291938897e-07,
"loss": 0.782,
"step": 1515
},
{
"epoch": 0.9178743961352657,
"grad_norm": 1.0200768363247714,
"learning_rate": 2.041583572037037e-07,
"loss": 0.7924,
"step": 1520
},
{
"epoch": 0.9208937198067633,
"grad_norm": 0.9739653495159714,
"learning_rate": 1.8951649249480287e-07,
"loss": 0.7899,
"step": 1525
},
{
"epoch": 0.9239130434782609,
"grad_norm": 0.995954989126771,
"learning_rate": 1.7540925606256088e-07,
"loss": 0.8044,
"step": 1530
},
{
"epoch": 0.9269323671497585,
"grad_norm": 0.9939242624751221,
"learning_rate": 1.6183821575925186e-07,
"loss": 0.7777,
"step": 1535
},
{
"epoch": 0.9299516908212561,
"grad_norm": 0.9799525062498867,
"learning_rate": 1.4880487984529846e-07,
"loss": 0.7888,
"step": 1540
},
{
"epoch": 0.9329710144927537,
"grad_norm": 0.9911311370863874,
"learning_rate": 1.363106968216482e-07,
"loss": 0.7935,
"step": 1545
},
{
"epoch": 0.9359903381642513,
"grad_norm": 1.003564658664697,
"learning_rate": 1.243570552687895e-07,
"loss": 0.7938,
"step": 1550
},
{
"epoch": 0.9390096618357487,
"grad_norm": 0.9800326260440435,
"learning_rate": 1.1294528369242663e-07,
"loss": 0.7928,
"step": 1555
},
{
"epoch": 0.9420289855072463,
"grad_norm": 0.986565572978078,
"learning_rate": 1.020766503758347e-07,
"loss": 0.7906,
"step": 1560
},
{
"epoch": 0.9450483091787439,
"grad_norm": 0.966200390084136,
"learning_rate": 9.175236323890058e-08,
"loss": 0.7864,
"step": 1565
},
{
"epoch": 0.9480676328502415,
"grad_norm": 0.9843328895814176,
"learning_rate": 8.197356970388148e-08,
"loss": 0.7798,
"step": 1570
},
{
"epoch": 0.9510869565217391,
"grad_norm": 1.0058804827600163,
"learning_rate": 7.274135656787917e-08,
"loss": 0.7904,
"step": 1575
},
{
"epoch": 0.9541062801932367,
"grad_norm": 0.9570446950074213,
"learning_rate": 6.405674988205602e-08,
"loss": 0.7725,
"step": 1580
},
{
"epoch": 0.9571256038647343,
"grad_norm": 1.0097440052548536,
"learning_rate": 5.592071483760397e-08,
"loss": 0.7988,
"step": 1585
},
{
"epoch": 0.9601449275362319,
"grad_norm": 1.0050549132372466,
"learning_rate": 4.833415565847155e-08,
"loss": 0.7843,
"step": 1590
},
{
"epoch": 0.9631642512077294,
"grad_norm": 1.013353688401092,
"learning_rate": 4.1297915500873034e-08,
"loss": 0.7919,
"step": 1595
},
{
"epoch": 0.966183574879227,
"grad_norm": 0.977208847695192,
"learning_rate": 3.481277635957903e-08,
"loss": 0.7908,
"step": 1600
},
{
"epoch": 0.9692028985507246,
"grad_norm": 0.9806969063880059,
"learning_rate": 2.88794589810093e-08,
"loss": 0.7883,
"step": 1605
},
{
"epoch": 0.9722222222222222,
"grad_norm": 0.9884154444375535,
"learning_rate": 2.3498622783128533e-08,
"loss": 0.7756,
"step": 1610
},
{
"epoch": 0.9752415458937198,
"grad_norm": 1.0036601159352383,
"learning_rate": 1.8670865782161042e-08,
"loss": 0.785,
"step": 1615
},
{
"epoch": 0.9782608695652174,
"grad_norm": 1.0066570396591208,
"learning_rate": 1.4396724526127282e-08,
"loss": 0.784,
"step": 1620
},
{
"epoch": 0.981280193236715,
"grad_norm": 1.0110694707000738,
"learning_rate": 1.067667403521433e-08,
"loss": 0.7993,
"step": 1625
},
{
"epoch": 0.9842995169082126,
"grad_norm": 0.9862631356943906,
"learning_rate": 7.51112774898144e-09,
"loss": 0.7716,
"step": 1630
},
{
"epoch": 0.9873188405797102,
"grad_norm": 1.0008914135572442,
"learning_rate": 4.900437480413467e-09,
"loss": 0.7775,
"step": 1635
},
{
"epoch": 0.9903381642512077,
"grad_norm": 0.9753190697055402,
"learning_rate": 2.844893376816593e-09,
"loss": 0.7655,
"step": 1640
},
{
"epoch": 0.9933574879227053,
"grad_norm": 0.9719880320026482,
"learning_rate": 1.3447238875774482e-09,
"loss": 0.7747,
"step": 1645
},
{
"epoch": 0.9963768115942029,
"grad_norm": 0.9866118574069571,
"learning_rate": 4.000957387700899e-10,
"loss": 0.7889,
"step": 1650
},
{
"epoch": 0.9993961352657005,
"grad_norm": 1.0028543954306115,
"learning_rate": 1.1113914626381672e-11,
"loss": 0.7873,
"step": 1655
},
{
"epoch": 1.0,
"eval_runtime": 0.8993,
"eval_samples_per_second": 11.12,
"eval_steps_per_second": 3.336,
"step": 1656
},
{
"epoch": 1.0,
"step": 1656,
"total_flos": 44008382398464.0,
"train_loss": 0.0,
"train_runtime": 0.0097,
"train_samples_per_second": 2738278.668,
"train_steps_per_second": 171187.643
}
],
"logging_steps": 5,
"max_steps": 1656,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 44008382398464.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}