1360 lines
33 KiB
JSON
1360 lines
33 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 1.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 1881,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.0053177346450412125,
|
||
|
|
"grad_norm": 0.039577700197696686,
|
||
|
|
"learning_rate": 9.523809523809525e-07,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.010635469290082425,
|
||
|
|
"grad_norm": 0.03883662074804306,
|
||
|
|
"learning_rate": 2.0105820105820108e-06,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.015953203935123637,
|
||
|
|
"grad_norm": 0.039891913533210754,
|
||
|
|
"learning_rate": 3.068783068783069e-06,
|
||
|
|
"loss": 0.0066,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.02127093858016485,
|
||
|
|
"grad_norm": 0.03841518983244896,
|
||
|
|
"learning_rate": 4.126984126984127e-06,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.026588673225206062,
|
||
|
|
"grad_norm": 0.038002241402864456,
|
||
|
|
"learning_rate": 5.185185185185185e-06,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.031906407870247275,
|
||
|
|
"grad_norm": 0.05087496340274811,
|
||
|
|
"learning_rate": 6.243386243386243e-06,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03722414251528849,
|
||
|
|
"grad_norm": 0.06635119765996933,
|
||
|
|
"learning_rate": 7.301587301587301e-06,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0425418771603297,
|
||
|
|
"grad_norm": 0.04691258817911148,
|
||
|
|
"learning_rate": 8.35978835978836e-06,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04785961180537091,
|
||
|
|
"grad_norm": 0.047258518636226654,
|
||
|
|
"learning_rate": 9.417989417989418e-06,
|
||
|
|
"loss": 0.0066,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.053177346450412125,
|
||
|
|
"grad_norm": 0.1077064797282219,
|
||
|
|
"learning_rate": 1.0476190476190477e-05,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05849508109545334,
|
||
|
|
"grad_norm": 0.053190864622592926,
|
||
|
|
"learning_rate": 1.1534391534391536e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06381281574049455,
|
||
|
|
"grad_norm": 0.04659568890929222,
|
||
|
|
"learning_rate": 1.2592592592592593e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06913055038553576,
|
||
|
|
"grad_norm": 0.05943409726023674,
|
||
|
|
"learning_rate": 1.3650793650793652e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07444828503057697,
|
||
|
|
"grad_norm": 0.0551825612783432,
|
||
|
|
"learning_rate": 1.470899470899471e-05,
|
||
|
|
"loss": 0.0065,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07976601967561818,
|
||
|
|
"grad_norm": 0.06374579668045044,
|
||
|
|
"learning_rate": 1.576719576719577e-05,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0850837543206594,
|
||
|
|
"grad_norm": 0.0627669170498848,
|
||
|
|
"learning_rate": 1.6825396825396828e-05,
|
||
|
|
"loss": 0.0068,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0904014889657006,
|
||
|
|
"grad_norm": 0.08319643139839172,
|
||
|
|
"learning_rate": 1.7883597883597884e-05,
|
||
|
|
"loss": 0.0068,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09571922361074182,
|
||
|
|
"grad_norm": 0.04758685082197189,
|
||
|
|
"learning_rate": 1.8941798941798943e-05,
|
||
|
|
"loss": 0.0072,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10103695825578303,
|
||
|
|
"grad_norm": 0.04842646047472954,
|
||
|
|
"learning_rate": 2e-05,
|
||
|
|
"loss": 0.0083,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10635469290082425,
|
||
|
|
"grad_norm": 0.061520617455244064,
|
||
|
|
"learning_rate": 1.9998276320345247e-05,
|
||
|
|
"loss": 0.0081,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11167242754586545,
|
||
|
|
"grad_norm": 0.053946252912282944,
|
||
|
|
"learning_rate": 1.999310587559529e-05,
|
||
|
|
"loss": 0.0082,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11699016219090667,
|
||
|
|
"grad_norm": 0.07307784259319305,
|
||
|
|
"learning_rate": 1.998449044818822e-05,
|
||
|
|
"loss": 0.0086,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12230789683594788,
|
||
|
|
"grad_norm": 0.05764337629079819,
|
||
|
|
"learning_rate": 1.9972433008171417e-05,
|
||
|
|
"loss": 0.0084,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1276256314809891,
|
||
|
|
"grad_norm": 0.054092682898044586,
|
||
|
|
"learning_rate": 1.99569377121777e-05,
|
||
|
|
"loss": 0.0083,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1329433661260303,
|
||
|
|
"grad_norm": 0.04634464159607887,
|
||
|
|
"learning_rate": 1.993800990199235e-05,
|
||
|
|
"loss": 0.0081,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1382611007710715,
|
||
|
|
"grad_norm": 0.04918840155005455,
|
||
|
|
"learning_rate": 1.9915656102711635e-05,
|
||
|
|
"loss": 0.0083,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14357883541611274,
|
||
|
|
"grad_norm": 0.062442682683467865,
|
||
|
|
"learning_rate": 1.9889884020493363e-05,
|
||
|
|
"loss": 0.0081,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14889657006115395,
|
||
|
|
"grad_norm": 0.05739056318998337,
|
||
|
|
"learning_rate": 1.9860702539900288e-05,
|
||
|
|
"loss": 0.0078,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15421430470619515,
|
||
|
|
"grad_norm": 0.05813131108880043,
|
||
|
|
"learning_rate": 1.9828121720837288e-05,
|
||
|
|
"loss": 0.0083,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15953203935123636,
|
||
|
|
"grad_norm": 0.053296517580747604,
|
||
|
|
"learning_rate": 1.979215279508335e-05,
|
||
|
|
"loss": 0.008,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1648497739962776,
|
||
|
|
"grad_norm": 0.05386708676815033,
|
||
|
|
"learning_rate": 1.975280816241959e-05,
|
||
|
|
"loss": 0.0079,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1701675086413188,
|
||
|
|
"grad_norm": 0.049940075725317,
|
||
|
|
"learning_rate": 1.9710101386354572e-05,
|
||
|
|
"loss": 0.0078,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17548524328636,
|
||
|
|
"grad_norm": 0.04508732259273529,
|
||
|
|
"learning_rate": 1.9664047189448496e-05,
|
||
|
|
"loss": 0.0079,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1808029779314012,
|
||
|
|
"grad_norm": 0.05223456770181656,
|
||
|
|
"learning_rate": 1.961466144823781e-05,
|
||
|
|
"loss": 0.0076,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18612071257644244,
|
||
|
|
"grad_norm": 0.0454275980591774,
|
||
|
|
"learning_rate": 1.9561961187761987e-05,
|
||
|
|
"loss": 0.0078,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19143844722148365,
|
||
|
|
"grad_norm": 0.06865548342466354,
|
||
|
|
"learning_rate": 1.9505964575694387e-05,
|
||
|
|
"loss": 0.008,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19675618186652485,
|
||
|
|
"grad_norm": 0.04975816607475281,
|
||
|
|
"learning_rate": 1.944669091607919e-05,
|
||
|
|
"loss": 0.0079,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20207391651156606,
|
||
|
|
"grad_norm": 0.04051858186721802,
|
||
|
|
"learning_rate": 1.9384160642676634e-05,
|
||
|
|
"loss": 0.008,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2073916511566073,
|
||
|
|
"grad_norm": 0.05171920359134674,
|
||
|
|
"learning_rate": 1.931839531191873e-05,
|
||
|
|
"loss": 0.0076,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2127093858016485,
|
||
|
|
"grad_norm": 0.04655490070581436,
|
||
|
|
"learning_rate": 1.9249417595478005e-05,
|
||
|
|
"loss": 0.0078,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2180271204466897,
|
||
|
|
"grad_norm": 0.05582937225699425,
|
||
|
|
"learning_rate": 1.9177251272451742e-05,
|
||
|
|
"loss": 0.0075,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2233448550917309,
|
||
|
|
"grad_norm": 0.04188157245516777,
|
||
|
|
"learning_rate": 1.9101921221164502e-05,
|
||
|
|
"loss": 0.0077,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22866258973677214,
|
||
|
|
"grad_norm": 0.045288506895303726,
|
||
|
|
"learning_rate": 1.902345341059164e-05,
|
||
|
|
"loss": 0.0074,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23398032438181335,
|
||
|
|
"grad_norm": 0.042311619967222214,
|
||
|
|
"learning_rate": 1.894187489140688e-05,
|
||
|
|
"loss": 0.0076,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23929805902685455,
|
||
|
|
"grad_norm": 0.04253494367003441,
|
||
|
|
"learning_rate": 1.8857213786656986e-05,
|
||
|
|
"loss": 0.0076,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24461579367189576,
|
||
|
|
"grad_norm": 0.06455250829458237,
|
||
|
|
"learning_rate": 1.8769499282066716e-05,
|
||
|
|
"loss": 0.0074,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.249933528316937,
|
||
|
|
"grad_norm": 0.04057531803846359,
|
||
|
|
"learning_rate": 1.867876161597747e-05,
|
||
|
|
"loss": 0.0074,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2552512629619782,
|
||
|
|
"grad_norm": 0.054350581020116806,
|
||
|
|
"learning_rate": 1.8585032068923034e-05,
|
||
|
|
"loss": 0.0071,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2605689976070194,
|
||
|
|
"grad_norm": 0.06098693609237671,
|
||
|
|
"learning_rate": 1.8488342952846074e-05,
|
||
|
|
"loss": 0.0072,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2658867322520606,
|
||
|
|
"grad_norm": 0.05364970117807388,
|
||
|
|
"learning_rate": 1.8388727599959033e-05,
|
||
|
|
"loss": 0.0074,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2712044668971018,
|
||
|
|
"grad_norm": 0.0433780662715435,
|
||
|
|
"learning_rate": 1.8286220351253324e-05,
|
||
|
|
"loss": 0.0071,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.276522201542143,
|
||
|
|
"grad_norm": 0.06102316454052925,
|
||
|
|
"learning_rate": 1.8180856544660763e-05,
|
||
|
|
"loss": 0.007,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2818399361871843,
|
||
|
|
"grad_norm": 0.04557032510638237,
|
||
|
|
"learning_rate": 1.8072672502871295e-05,
|
||
|
|
"loss": 0.0072,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2871576708322255,
|
||
|
|
"grad_norm": 0.057244665920734406,
|
||
|
|
"learning_rate": 1.796170552081129e-05,
|
||
|
|
"loss": 0.0072,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2924754054772667,
|
||
|
|
"grad_norm": 0.04519479721784592,
|
||
|
|
"learning_rate": 1.7847993852786612e-05,
|
||
|
|
"loss": 0.0071,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2977931401223079,
|
||
|
|
"grad_norm": 0.04738995060324669,
|
||
|
|
"learning_rate": 1.773157669929499e-05,
|
||
|
|
"loss": 0.0073,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3031108747673491,
|
||
|
|
"grad_norm": 0.04299917444586754,
|
||
|
|
"learning_rate": 1.761249419351222e-05,
|
||
|
|
"loss": 0.007,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3084286094123903,
|
||
|
|
"grad_norm": 0.0751781016588211,
|
||
|
|
"learning_rate": 1.749078738745679e-05,
|
||
|
|
"loss": 0.007,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3137463440574315,
|
||
|
|
"grad_norm": 0.04250544309616089,
|
||
|
|
"learning_rate": 1.736649823783779e-05,
|
||
|
|
"loss": 0.0069,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3190640787024727,
|
||
|
|
"grad_norm": 0.042321231216192245,
|
||
|
|
"learning_rate": 1.7239669591590916e-05,
|
||
|
|
"loss": 0.007,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.324381813347514,
|
||
|
|
"grad_norm": 0.058343447744846344,
|
||
|
|
"learning_rate": 1.711034517110761e-05,
|
||
|
|
"loss": 0.0069,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3296995479925552,
|
||
|
|
"grad_norm": 0.0474131815135479,
|
||
|
|
"learning_rate": 1.6978569559162358e-05,
|
||
|
|
"loss": 0.0068,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3350172826375964,
|
||
|
|
"grad_norm": 0.05743961036205292,
|
||
|
|
"learning_rate": 1.6844388183543418e-05,
|
||
|
|
"loss": 0.007,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3403350172826376,
|
||
|
|
"grad_norm": 0.04410571604967117,
|
||
|
|
"learning_rate": 1.6707847301392237e-05,
|
||
|
|
"loss": 0.0071,
|
||
|
|
"step": 640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3456527519276788,
|
||
|
|
"grad_norm": 0.03746718168258667,
|
||
|
|
"learning_rate": 1.656899398325693e-05,
|
||
|
|
"loss": 0.0069,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35097048657272,
|
||
|
|
"grad_norm": 0.04007519781589508,
|
||
|
|
"learning_rate": 1.6427876096865394e-05,
|
||
|
|
"loss": 0.0068,
|
||
|
|
"step": 660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3562882212177612,
|
||
|
|
"grad_norm": 0.04943158105015755,
|
||
|
|
"learning_rate": 1.6284542290623568e-05,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3616059558628024,
|
||
|
|
"grad_norm": 0.054357461631298065,
|
||
|
|
"learning_rate": 1.6139041976844583e-05,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3669236905078437,
|
||
|
|
"grad_norm": 0.05400274693965912,
|
||
|
|
"learning_rate": 1.599142531471456e-05,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3722414251528849,
|
||
|
|
"grad_norm": 0.05498107895255089,
|
||
|
|
"learning_rate": 1.5841743193000945e-05,
|
||
|
|
"loss": 0.0068,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3775591597979261,
|
||
|
|
"grad_norm": 0.04064950346946716,
|
||
|
|
"learning_rate": 1.5690047212509317e-05,
|
||
|
|
"loss": 0.0068,
|
||
|
|
"step": 710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3828768944429673,
|
||
|
|
"grad_norm": 0.04839126765727997,
|
||
|
|
"learning_rate": 1.5536389668294724e-05,
|
||
|
|
"loss": 0.0068,
|
||
|
|
"step": 720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3881946290880085,
|
||
|
|
"grad_norm": 0.05424703657627106,
|
||
|
|
"learning_rate": 1.5380823531633727e-05,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3935123637330497,
|
||
|
|
"grad_norm": 0.04082576930522919,
|
||
|
|
"learning_rate": 1.522340243176327e-05,
|
||
|
|
"loss": 0.0066,
|
||
|
|
"step": 740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3988300983780909,
|
||
|
|
"grad_norm": 0.03836412727832794,
|
||
|
|
"learning_rate": 1.5064180637392765e-05,
|
||
|
|
"loss": 0.0067,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4041478330231321,
|
||
|
|
"grad_norm": 0.04045991972088814,
|
||
|
|
"learning_rate": 1.4903213037995725e-05,
|
||
|
|
"loss": 0.0066,
|
||
|
|
"step": 760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4094655676681734,
|
||
|
|
"grad_norm": 0.036179762333631516,
|
||
|
|
"learning_rate": 1.4740555124887377e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4147833023132146,
|
||
|
|
"grad_norm": 0.043810952454805374,
|
||
|
|
"learning_rate": 1.4576262972094828e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4201010369582558,
|
||
|
|
"grad_norm": 0.057811930775642395,
|
||
|
|
"learning_rate": 1.4410393217026317e-05,
|
||
|
|
"loss": 0.0066,
|
||
|
|
"step": 790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.425418771603297,
|
||
|
|
"grad_norm": 0.040190473198890686,
|
||
|
|
"learning_rate": 1.4243003040946275e-05,
|
||
|
|
"loss": 0.0065,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4307365062483382,
|
||
|
|
"grad_norm": 0.0423622652888298,
|
||
|
|
"learning_rate": 1.4074150149262882e-05,
|
||
|
|
"loss": 0.0065,
|
||
|
|
"step": 810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4360542408933794,
|
||
|
|
"grad_norm": 0.03963879495859146,
|
||
|
|
"learning_rate": 1.3903892751634949e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4413719755384206,
|
||
|
|
"grad_norm": 0.03814619034528732,
|
||
|
|
"learning_rate": 1.3732289541904948e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4466897101834618,
|
||
|
|
"grad_norm": 0.04084085300564766,
|
||
|
|
"learning_rate": 1.355939967786514e-05,
|
||
|
|
"loss": 0.0065,
|
||
|
|
"step": 840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4520074448285031,
|
||
|
|
"grad_norm": 0.04477806016802788,
|
||
|
|
"learning_rate": 1.3385282760863758e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4573251794735443,
|
||
|
|
"grad_norm": 0.03881698101758957,
|
||
|
|
"learning_rate": 1.3209998815258273e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4626429141185855,
|
||
|
|
"grad_norm": 0.06034592166543007,
|
||
|
|
"learning_rate": 1.303360826772286e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4679606487636267,
|
||
|
|
"grad_norm": 0.04441380500793457,
|
||
|
|
"learning_rate": 1.2856171926417134e-05,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4732783834086679,
|
||
|
|
"grad_norm": 0.05011635273694992,
|
||
|
|
"learning_rate": 1.2677750960023396e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4785961180537091,
|
||
|
|
"grad_norm": 0.04159679636359215,
|
||
|
|
"learning_rate": 1.2498406876659599e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4839138526987503,
|
||
|
|
"grad_norm": 0.05039462447166443,
|
||
|
|
"learning_rate": 1.2318201502675285e-05,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4892315873437915,
|
||
|
|
"grad_norm": 0.05377871170639992,
|
||
|
|
"learning_rate": 1.2137196961337811e-05,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4945493219888328,
|
||
|
|
"grad_norm": 0.03828214854001999,
|
||
|
|
"learning_rate": 1.1955455651416246e-05,
|
||
|
|
"loss": 0.0065,
|
||
|
|
"step": 930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.499867056633874,
|
||
|
|
"grad_norm": 0.03937741369009018,
|
||
|
|
"learning_rate": 1.1773040225670257e-05,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5051847912789151,
|
||
|
|
"grad_norm": 0.037178609520196915,
|
||
|
|
"learning_rate": 1.1590013569251457e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5105025259239564,
|
||
|
|
"grad_norm": 0.03845744580030441,
|
||
|
|
"learning_rate": 1.1406438778024635e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5158202605689977,
|
||
|
|
"grad_norm": 0.03919456526637077,
|
||
|
|
"learning_rate": 1.1222379136816347e-05,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5211379952140388,
|
||
|
|
"grad_norm": 0.04106747359037399,
|
||
|
|
"learning_rate": 1.1037898097598353e-05,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5264557298590801,
|
||
|
|
"grad_norm": 0.03811858221888542,
|
||
|
|
"learning_rate": 1.0853059257613447e-05,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5317734645041212,
|
||
|
|
"grad_norm": 0.04202568531036377,
|
||
|
|
"learning_rate": 1.0667926337451217e-05,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5370911991491625,
|
||
|
|
"grad_norm": 0.037443362176418304,
|
||
|
|
"learning_rate": 1.0482563159081238e-05,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5424089337942036,
|
||
|
|
"grad_norm": 0.03632836043834686,
|
||
|
|
"learning_rate": 1.0297033623851368e-05,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5477266684392449,
|
||
|
|
"grad_norm": 0.03794227913022041,
|
||
|
|
"learning_rate": 1.0111401690458655e-05,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.553044403084286,
|
||
|
|
"grad_norm": 0.039649080485105515,
|
||
|
|
"learning_rate": 9.925731352900477e-06,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5583621377293273,
|
||
|
|
"grad_norm": 0.042000215500593185,
|
||
|
|
"learning_rate": 9.740086618413495e-06,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5636798723743686,
|
||
|
|
"grad_norm": 0.05552341789007187,
|
||
|
|
"learning_rate": 9.554531485408078e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5689976070194097,
|
||
|
|
"grad_norm": 0.035325516015291214,
|
||
|
|
"learning_rate": 9.369129921405754e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.574315341664451,
|
||
|
|
"grad_norm": 0.04520582780241966,
|
||
|
|
"learning_rate": 9.183945840987277e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5796330763094921,
|
||
|
|
"grad_norm": 0.03700090944766998,
|
||
|
|
"learning_rate": 8.999043083759016e-06,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5849508109545334,
|
||
|
|
"grad_norm": 0.03953060507774353,
|
||
|
|
"learning_rate": 8.81448539234512e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5902685455995745,
|
||
|
|
"grad_norm": 0.04827453941106796,
|
||
|
|
"learning_rate": 8.630336390413148e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5955862802446158,
|
||
|
|
"grad_norm": 0.04095720499753952,
|
||
|
|
"learning_rate": 8.446659560740718e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6009040148896571,
|
||
|
|
"grad_norm": 0.03602287173271179,
|
||
|
|
"learning_rate": 8.263518223330698e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6062217495346982,
|
||
|
|
"grad_norm": 0.042592573910951614,
|
||
|
|
"learning_rate": 8.08097551358254e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6115394841797395,
|
||
|
|
"grad_norm": 0.03963078185915947,
|
||
|
|
"learning_rate": 7.89909436052722e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6168572188247806,
|
||
|
|
"grad_norm": 0.03793001174926758,
|
||
|
|
"learning_rate": 7.71793746513337e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6221749534698219,
|
||
|
|
"grad_norm": 0.043642308562994,
|
||
|
|
"learning_rate": 7.5375672786919655e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.627492688114863,
|
||
|
|
"grad_norm": 0.04830528423190117,
|
||
|
|
"learning_rate": 7.358045981287141e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6328104227599043,
|
||
|
|
"grad_norm": 0.040264688432216644,
|
||
|
|
"learning_rate": 7.179435460360491e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6381281574049454,
|
||
|
|
"grad_norm": 0.03443283215165138,
|
||
|
|
"learning_rate": 7.0017972893762176e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6434458920499867,
|
||
|
|
"grad_norm": 0.034863632172346115,
|
||
|
|
"learning_rate": 6.8251927065945755e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.648763626695028,
|
||
|
|
"grad_norm": 0.05887424945831299,
|
||
|
|
"learning_rate": 6.649682593960813e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6540813613400691,
|
||
|
|
"grad_norm": 0.06200723350048065,
|
||
|
|
"learning_rate": 6.475327456117005e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6593990959851104,
|
||
|
|
"grad_norm": 0.05243268609046936,
|
||
|
|
"learning_rate": 6.302187399543911e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6647168306301515,
|
||
|
|
"grad_norm": 0.03554186969995499,
|
||
|
|
"learning_rate": 6.130322111840114e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6700345652751928,
|
||
|
|
"grad_norm": 0.034715306013822556,
|
||
|
|
"learning_rate": 5.959790841145577e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6753522999202339,
|
||
|
|
"grad_norm": 0.04114481806755066,
|
||
|
|
"learning_rate": 5.790652375716653e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6806700345652752,
|
||
|
|
"grad_norm": 0.03605886921286583,
|
||
|
|
"learning_rate": 5.622965023659684e-06,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6859877692103165,
|
||
|
|
"grad_norm": 0.04023313522338867,
|
||
|
|
"learning_rate": 5.4567865928300835e-06,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6913055038553576,
|
||
|
|
"grad_norm": 0.03677726909518242,
|
||
|
|
"learning_rate": 5.292174370903919e-06,
|
||
|
|
"loss": 0.0057,
|
||
|
|
"step": 1300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6966232385003989,
|
||
|
|
"grad_norm": 0.03900739923119545,
|
||
|
|
"learning_rate": 5.1291851056287555e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.70194097314544,
|
||
|
|
"grad_norm": 0.05076342821121216,
|
||
|
|
"learning_rate": 4.96787498526069e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7072587077904813,
|
||
|
|
"grad_norm": 0.03732963278889656,
|
||
|
|
"learning_rate": 4.808299619194251e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7125764424355224,
|
||
|
|
"grad_norm": 0.03477802500128746,
|
||
|
|
"learning_rate": 4.6505140187918e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7178941770805637,
|
||
|
|
"grad_norm": 0.0560050792992115,
|
||
|
|
"learning_rate": 4.494572578419194e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7232119117256048,
|
||
|
|
"grad_norm": 0.04002802073955536,
|
||
|
|
"learning_rate": 4.3405290566940475e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7285296463706461,
|
||
|
|
"grad_norm": 0.03611234575510025,
|
||
|
|
"learning_rate": 4.188436557953235e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7338473810156874,
|
||
|
|
"grad_norm": 0.034957099705934525,
|
||
|
|
"learning_rate": 4.0383475139458985e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7391651156607285,
|
||
|
|
"grad_norm": 0.04099763557314873,
|
||
|
|
"learning_rate": 3.890313665758348e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7444828503057698,
|
||
|
|
"grad_norm": 0.036687929183244705,
|
||
|
|
"learning_rate": 3.744386045977052e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7498005849508109,
|
||
|
|
"grad_norm": 0.0366140753030777,
|
||
|
|
"learning_rate": 3.600614961095863e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7551183195958522,
|
||
|
|
"grad_norm": 0.03608519583940506,
|
||
|
|
"learning_rate": 3.4590499741735796e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7604360542408933,
|
||
|
|
"grad_norm": 0.04826873913407326,
|
||
|
|
"learning_rate": 3.3197398877477528e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7657537888859346,
|
||
|
|
"grad_norm": 0.03552957624197006,
|
||
|
|
"learning_rate": 3.1827327270107157e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7710715235309759,
|
||
|
|
"grad_norm": 0.04187224805355072,
|
||
|
|
"learning_rate": 3.0480757232535773e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.776389258176017,
|
||
|
|
"grad_norm": 0.04442904144525528,
|
||
|
|
"learning_rate": 2.915815297583886e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7817069928210583,
|
||
|
|
"grad_norm": 0.03516370803117752,
|
||
|
|
"learning_rate": 2.7859970449226103e-06,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7870247274660994,
|
||
|
|
"grad_norm": 0.03615553304553032,
|
||
|
|
"learning_rate": 2.658665718285934e-06,
|
||
|
|
"loss": 0.0057,
|
||
|
|
"step": 1480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7923424621111407,
|
||
|
|
"grad_norm": 0.045939717441797256,
|
||
|
|
"learning_rate": 2.5338652133572915e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7976601967561818,
|
||
|
|
"grad_norm": 0.03305557742714882,
|
||
|
|
"learning_rate": 2.411638553354928e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8029779314012231,
|
||
|
|
"grad_norm": 0.03413030132651329,
|
||
|
|
"learning_rate": 2.2920278742002677e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8082956660462642,
|
||
|
|
"grad_norm": 0.03630566969513893,
|
||
|
|
"learning_rate": 2.1750744099921394e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8136134006913055,
|
||
|
|
"grad_norm": 0.03545108065009117,
|
||
|
|
"learning_rate": 2.0608184787919027e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 1530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8189311353363468,
|
||
|
|
"grad_norm": 0.03896397352218628,
|
||
|
|
"learning_rate": 1.9492994687243715e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8242488699813879,
|
||
|
|
"grad_norm": 0.041112203150987625,
|
||
|
|
"learning_rate": 1.840555824399296e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8295666046264292,
|
||
|
|
"grad_norm": 0.03573285788297653,
|
||
|
|
"learning_rate": 1.7346250336581426e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8348843392714703,
|
||
|
|
"grad_norm": 0.03390691056847572,
|
||
|
|
"learning_rate": 1.6315436146506702e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8402020739165116,
|
||
|
|
"grad_norm": 0.03588733822107315,
|
||
|
|
"learning_rate": 1.531347103245825e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8455198085615527,
|
||
|
|
"grad_norm": 0.03812939673662186,
|
||
|
|
"learning_rate": 1.4340700407812436e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.850837543206594,
|
||
|
|
"grad_norm": 0.0381702184677124,
|
||
|
|
"learning_rate": 1.339745962155613e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8561552778516353,
|
||
|
|
"grad_norm": 0.03450675308704376,
|
||
|
|
"learning_rate": 1.2484073842679945e-06,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8614730124966764,
|
||
|
|
"grad_norm": 0.03307728469371796,
|
||
|
|
"learning_rate": 1.1600857948080623e-06,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8667907471417177,
|
||
|
|
"grad_norm": 0.03805455192923546,
|
||
|
|
"learning_rate": 1.074811641401189e-06,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8721084817867588,
|
||
|
|
"grad_norm": 0.039794716984033585,
|
||
|
|
"learning_rate": 9.926143211120288e-07,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8774262164318001,
|
||
|
|
"grad_norm": 0.036838822066783905,
|
||
|
|
"learning_rate": 9.135221703103137e-07,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8827439510768412,
|
||
|
|
"grad_norm": 0.034202683717012405,
|
||
|
|
"learning_rate": 8.375624549022854e-07,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8880616857218825,
|
||
|
|
"grad_norm": 0.0360797755420208,
|
||
|
|
"learning_rate": 7.647613609311455e-07,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8933794203669236,
|
||
|
|
"grad_norm": 0.03642485290765762,
|
||
|
|
"learning_rate": 6.951439855498021e-07,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8986971550119649,
|
||
|
|
"grad_norm": 0.03886672109365463,
|
||
|
|
"learning_rate": 6.287343283689662e-07,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9040148896570062,
|
||
|
|
"grad_norm": 0.03337244689464569,
|
||
|
|
"learning_rate": 5.655552831836297e-07,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 1700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9093326243020473,
|
||
|
|
"grad_norm": 0.0369742251932621,
|
||
|
|
"learning_rate": 5.056286300807511e-07,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9146503589470886,
|
||
|
|
"grad_norm": 0.03584656864404678,
|
||
|
|
"learning_rate": 4.4897502793087576e-07,
|
||
|
|
"loss": 0.0059,
|
||
|
|
"step": 1720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9199680935921297,
|
||
|
|
"grad_norm": 0.03646034002304077,
|
||
|
|
"learning_rate": 3.956140072662851e-07,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.925285828237171,
|
||
|
|
"grad_norm": 0.03360486403107643,
|
||
|
|
"learning_rate": 3.455639635481123e-07,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 1740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9306035628822121,
|
||
|
|
"grad_norm": 0.03404919430613518,
|
||
|
|
"learning_rate": 2.988421508247741e-07,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9359212975272534,
|
||
|
|
"grad_norm": 0.033960774540901184,
|
||
|
|
"learning_rate": 2.554646757838763e-07,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 1760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9412390321722947,
|
||
|
|
"grad_norm": 0.03541607782244682,
|
||
|
|
"learning_rate": 2.1544649219965574e-07,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9465567668173358,
|
||
|
|
"grad_norm": 0.030705399811267853,
|
||
|
|
"learning_rate": 1.7880139577788779e-07,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9518745014623771,
|
||
|
|
"grad_norm": 0.036013998091220856,
|
||
|
|
"learning_rate": 1.4554201940000123e-07,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9571922361074182,
|
||
|
|
"grad_norm": 0.034899793565273285,
|
||
|
|
"learning_rate": 1.1567982876807604e-07,
|
||
|
|
"loss": 0.0064,
|
||
|
|
"step": 1800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9625099707524595,
|
||
|
|
"grad_norm": 0.0332038588821888,
|
||
|
|
"learning_rate": 8.922511845219972e-08,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9678277053975006,
|
||
|
|
"grad_norm": 0.03491925820708275,
|
||
|
|
"learning_rate": 6.618700834155944e-08,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9731454400425419,
|
||
|
|
"grad_norm": 0.035845331847667694,
|
||
|
|
"learning_rate": 4.657344050049251e-08,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 1830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.978463174687583,
|
||
|
|
"grad_norm": 0.03505609929561615,
|
||
|
|
"learning_rate": 3.039117643056755e-08,
|
||
|
|
"loss": 0.0063,
|
||
|
|
"step": 1840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9837809093326243,
|
||
|
|
"grad_norm": 0.033112842589616776,
|
||
|
|
"learning_rate": 1.7645794739654665e-08,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9890986439776656,
|
||
|
|
"grad_norm": 0.0354820154607296,
|
||
|
|
"learning_rate": 8.341689218775984e-09,
|
||
|
|
"loss": 0.0061,
|
||
|
|
"step": 1860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9944163786227067,
|
||
|
|
"grad_norm": 0.03354404866695404,
|
||
|
|
"learning_rate": 2.4820673274095207e-09,
|
||
|
|
"loss": 0.006,
|
||
|
|
"step": 1870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.999734113267748,
|
||
|
|
"grad_norm": 0.03434652462601662,
|
||
|
|
"learning_rate": 6.894908776255982e-11,
|
||
|
|
"loss": 0.0062,
|
||
|
|
"step": 1880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"step": 1881,
|
||
|
|
"total_flos": 3.4440299923195822e+19,
|
||
|
|
"train_loss": 0.006512725361979852,
|
||
|
|
"train_runtime": 112821.1839,
|
||
|
|
"train_samples_per_second": 8.534,
|
||
|
|
"train_steps_per_second": 0.017
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 10,
|
||
|
|
"max_steps": 1881,
|
||
|
|
"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": 3.4440299923195822e+19,
|
||
|
|
"train_batch_size": 8,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|