926 lines
22 KiB
JSON
926 lines
22 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 5.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 630,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.039860488290981565,
|
||
|
|
"grad_norm": 11.324122796709977,
|
||
|
|
"learning_rate": 2.53968253968254e-06,
|
||
|
|
"loss": 1.6286,
|
||
|
|
"step": 5
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07972097658196313,
|
||
|
|
"grad_norm": 1.991457650057207,
|
||
|
|
"learning_rate": 5.7142857142857145e-06,
|
||
|
|
"loss": 1.4713,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11958146487294469,
|
||
|
|
"grad_norm": 1.0610273535036991,
|
||
|
|
"learning_rate": 8.888888888888888e-06,
|
||
|
|
"loss": 1.3152,
|
||
|
|
"step": 15
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15944195316392626,
|
||
|
|
"grad_norm": 0.7999727467798167,
|
||
|
|
"learning_rate": 1.2063492063492064e-05,
|
||
|
|
"loss": 1.2185,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19930244145490783,
|
||
|
|
"grad_norm": 0.6462858961182688,
|
||
|
|
"learning_rate": 1.523809523809524e-05,
|
||
|
|
"loss": 1.1579,
|
||
|
|
"step": 25
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23916292974588937,
|
||
|
|
"grad_norm": 0.47561233028025485,
|
||
|
|
"learning_rate": 1.8412698412698415e-05,
|
||
|
|
"loss": 1.1295,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.279023418036871,
|
||
|
|
"grad_norm": 0.4460008015131473,
|
||
|
|
"learning_rate": 2.158730158730159e-05,
|
||
|
|
"loss": 1.1069,
|
||
|
|
"step": 35
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3188839063278525,
|
||
|
|
"grad_norm": 0.35932407714480935,
|
||
|
|
"learning_rate": 2.4761904761904766e-05,
|
||
|
|
"loss": 1.0887,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35874439461883406,
|
||
|
|
"grad_norm": 0.3717286127467533,
|
||
|
|
"learning_rate": 2.7936507936507936e-05,
|
||
|
|
"loss": 1.0869,
|
||
|
|
"step": 45
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.39860488290981566,
|
||
|
|
"grad_norm": 0.38261468570309465,
|
||
|
|
"learning_rate": 3.111111111111112e-05,
|
||
|
|
"loss": 1.0774,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4384653712007972,
|
||
|
|
"grad_norm": 0.383323190556911,
|
||
|
|
"learning_rate": 3.4285714285714284e-05,
|
||
|
|
"loss": 1.0636,
|
||
|
|
"step": 55
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.47832585949177875,
|
||
|
|
"grad_norm": 0.4100491937186789,
|
||
|
|
"learning_rate": 3.7460317460317464e-05,
|
||
|
|
"loss": 1.0535,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5181863477827603,
|
||
|
|
"grad_norm": 0.3975249651869064,
|
||
|
|
"learning_rate": 3.9999693004141615e-05,
|
||
|
|
"loss": 1.0297,
|
||
|
|
"step": 65
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.558046836073742,
|
||
|
|
"grad_norm": 0.3819095926536229,
|
||
|
|
"learning_rate": 3.998894913865352e-05,
|
||
|
|
"loss": 1.0337,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5979073243647235,
|
||
|
|
"grad_norm": 0.36490844056440797,
|
||
|
|
"learning_rate": 3.9962864903591375e-05,
|
||
|
|
"loss": 1.0063,
|
||
|
|
"step": 75
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.637767812655705,
|
||
|
|
"grad_norm": 0.4076298707032981,
|
||
|
|
"learning_rate": 3.992146031710637e-05,
|
||
|
|
"loss": 1.0237,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6776283009466866,
|
||
|
|
"grad_norm": 0.37566110161128824,
|
||
|
|
"learning_rate": 3.9864767154838864e-05,
|
||
|
|
"loss": 1.0145,
|
||
|
|
"step": 85
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7174887892376681,
|
||
|
|
"grad_norm": 0.34530643882424805,
|
||
|
|
"learning_rate": 3.9792828925532376e-05,
|
||
|
|
"loss": 1.0296,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7573492775286498,
|
||
|
|
"grad_norm": 0.3432558970267317,
|
||
|
|
"learning_rate": 3.970570083764316e-05,
|
||
|
|
"loss": 1.0059,
|
||
|
|
"step": 95
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7972097658196313,
|
||
|
|
"grad_norm": 0.3408833181353496,
|
||
|
|
"learning_rate": 3.9603449756970877e-05,
|
||
|
|
"loss": 1.004,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8370702541106129,
|
||
|
|
"grad_norm": 0.3589480700028462,
|
||
|
|
"learning_rate": 3.948615415534294e-05,
|
||
|
|
"loss": 0.9936,
|
||
|
|
"step": 105
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8769307424015944,
|
||
|
|
"grad_norm": 0.33854687709123976,
|
||
|
|
"learning_rate": 3.9353904050391874e-05,
|
||
|
|
"loss": 1.006,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.916791230692576,
|
||
|
|
"grad_norm": 0.3606947276733311,
|
||
|
|
"learning_rate": 3.9206800936472e-05,
|
||
|
|
"loss": 1.0033,
|
||
|
|
"step": 115
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9566517189835575,
|
||
|
|
"grad_norm": 0.3524202550527317,
|
||
|
|
"learning_rate": 3.904495770676831e-05,
|
||
|
|
"loss": 0.9917,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9965122072745392,
|
||
|
|
"grad_norm": 0.3695008401673435,
|
||
|
|
"learning_rate": 3.886849856665746e-05,
|
||
|
|
"loss": 1.0137,
|
||
|
|
"step": 125
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0318883906327851,
|
||
|
|
"grad_norm": 0.5103901242908977,
|
||
|
|
"learning_rate": 3.8677558938387276e-05,
|
||
|
|
"loss": 0.885,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0717488789237668,
|
||
|
|
"grad_norm": 0.39411785206718664,
|
||
|
|
"learning_rate": 3.8472285357147966e-05,
|
||
|
|
"loss": 0.8679,
|
||
|
|
"step": 135
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1116093672147485,
|
||
|
|
"grad_norm": 0.3898193416463423,
|
||
|
|
"learning_rate": 3.825283535861476e-05,
|
||
|
|
"loss": 0.8733,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.15146985550573,
|
||
|
|
"grad_norm": 0.7348802698342378,
|
||
|
|
"learning_rate": 3.801937735804838e-05,
|
||
|
|
"loss": 0.8434,
|
||
|
|
"step": 145
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1913303437967115,
|
||
|
|
"grad_norm": 0.36945102723915507,
|
||
|
|
"learning_rate": 3.777209052104598e-05,
|
||
|
|
"loss": 0.8461,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2311908320876932,
|
||
|
|
"grad_norm": 0.4602742420015167,
|
||
|
|
"learning_rate": 3.7511164626041823e-05,
|
||
|
|
"loss": 0.8606,
|
||
|
|
"step": 155
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2710513203786746,
|
||
|
|
"grad_norm": 0.35621072409911864,
|
||
|
|
"learning_rate": 3.7236799918663284e-05,
|
||
|
|
"loss": 0.8555,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.310911808669656,
|
||
|
|
"grad_norm": 0.42801215480497146,
|
||
|
|
"learning_rate": 3.6949206958053825e-05,
|
||
|
|
"loss": 0.8437,
|
||
|
|
"step": 165
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3507722969606377,
|
||
|
|
"grad_norm": 0.36223537090099234,
|
||
|
|
"learning_rate": 3.6648606455280944e-05,
|
||
|
|
"loss": 0.8566,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3906327852516194,
|
||
|
|
"grad_norm": 0.702392888184491,
|
||
|
|
"learning_rate": 3.633522910395314e-05,
|
||
|
|
"loss": 0.8665,
|
||
|
|
"step": 175
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4304932735426008,
|
||
|
|
"grad_norm": 0.36020937464475294,
|
||
|
|
"learning_rate": 3.6009315403175786e-05,
|
||
|
|
"loss": 0.8363,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4703537618335825,
|
||
|
|
"grad_norm": 0.3500105740239205,
|
||
|
|
"learning_rate": 3.567111547298194e-05,
|
||
|
|
"loss": 0.853,
|
||
|
|
"step": 185
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5102142501245641,
|
||
|
|
"grad_norm": 0.4768600668120408,
|
||
|
|
"learning_rate": 3.532088886237956e-05,
|
||
|
|
"loss": 0.8496,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5500747384155455,
|
||
|
|
"grad_norm": 0.36399360324025654,
|
||
|
|
"learning_rate": 3.495890435016258e-05,
|
||
|
|
"loss": 0.8636,
|
||
|
|
"step": 195
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5899352267065272,
|
||
|
|
"grad_norm": 0.3380619981423237,
|
||
|
|
"learning_rate": 3.458543973863859e-05,
|
||
|
|
"loss": 0.8538,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6297957149975089,
|
||
|
|
"grad_norm": 0.3389261022528899,
|
||
|
|
"learning_rate": 3.420078164043161e-05,
|
||
|
|
"loss": 0.8591,
|
||
|
|
"step": 205
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6696562032884903,
|
||
|
|
"grad_norm": 0.395644963130336,
|
||
|
|
"learning_rate": 3.38052252585233e-05,
|
||
|
|
"loss": 0.8401,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7095166915794717,
|
||
|
|
"grad_norm": 0.30378900619983906,
|
||
|
|
"learning_rate": 3.339907415970168e-05,
|
||
|
|
"loss": 0.8476,
|
||
|
|
"step": 215
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7493771798704534,
|
||
|
|
"grad_norm": 0.3900044033629726,
|
||
|
|
"learning_rate": 3.298264004159104e-05,
|
||
|
|
"loss": 0.8413,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.789237668161435,
|
||
|
|
"grad_norm": 0.36925152268695366,
|
||
|
|
"learning_rate": 3.255624249344198e-05,
|
||
|
|
"loss": 0.8534,
|
||
|
|
"step": 225
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8290981564524165,
|
||
|
|
"grad_norm": 0.3334498107167973,
|
||
|
|
"learning_rate": 3.212020875086495e-05,
|
||
|
|
"loss": 0.8734,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8689586447433981,
|
||
|
|
"grad_norm": 0.3231836736531515,
|
||
|
|
"learning_rate": 3.1674873444695804e-05,
|
||
|
|
"loss": 0.8619,
|
||
|
|
"step": 235
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9088191330343798,
|
||
|
|
"grad_norm": 0.3352794024716405,
|
||
|
|
"learning_rate": 3.122057834418582e-05,
|
||
|
|
"loss": 0.8604,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9486796213253612,
|
||
|
|
"grad_norm": 0.34760623558396486,
|
||
|
|
"learning_rate": 3.075767209471345e-05,
|
||
|
|
"loss": 0.8712,
|
||
|
|
"step": 245
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9885401096163426,
|
||
|
|
"grad_norm": 0.3494983937948561,
|
||
|
|
"learning_rate": 3.0286509950219077e-05,
|
||
|
|
"loss": 0.8449,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.023916292974589,
|
||
|
|
"grad_norm": 0.6438619381464517,
|
||
|
|
"learning_rate": 2.9807453500567937e-05,
|
||
|
|
"loss": 0.7382,
|
||
|
|
"step": 255
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0637767812655703,
|
||
|
|
"grad_norm": 0.48837947056765113,
|
||
|
|
"learning_rate": 2.9320870394050783e-05,
|
||
|
|
"loss": 0.6794,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.103637269556552,
|
||
|
|
"grad_norm": 0.4249085778688401,
|
||
|
|
"learning_rate": 2.8827134055234883e-05,
|
||
|
|
"loss": 0.6878,
|
||
|
|
"step": 265
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1434977578475336,
|
||
|
|
"grad_norm": 0.4112164304088742,
|
||
|
|
"learning_rate": 2.8326623398382174e-05,
|
||
|
|
"loss": 0.6895,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.183358246138515,
|
||
|
|
"grad_norm": 0.4393590314899133,
|
||
|
|
"learning_rate": 2.781972253665431e-05,
|
||
|
|
"loss": 0.6684,
|
||
|
|
"step": 275
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.223218734429497,
|
||
|
|
"grad_norm": 0.37274870892195433,
|
||
|
|
"learning_rate": 2.7306820487327906e-05,
|
||
|
|
"loss": 0.6719,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2630792227204783,
|
||
|
|
"grad_norm": 0.38062444497520725,
|
||
|
|
"learning_rate": 2.6788310873246133e-05,
|
||
|
|
"loss": 0.6735,
|
||
|
|
"step": 285
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.30293971101146,
|
||
|
|
"grad_norm": 0.3754103984568679,
|
||
|
|
"learning_rate": 2.62645916207358e-05,
|
||
|
|
"loss": 0.6757,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3428001993024417,
|
||
|
|
"grad_norm": 0.38094693535399177,
|
||
|
|
"learning_rate": 2.5736064654221808e-05,
|
||
|
|
"loss": 0.6544,
|
||
|
|
"step": 295
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.382660687593423,
|
||
|
|
"grad_norm": 0.37267334400373975,
|
||
|
|
"learning_rate": 2.5203135587773196e-05,
|
||
|
|
"loss": 0.6612,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4225211758844045,
|
||
|
|
"grad_norm": 0.3511334496003159,
|
||
|
|
"learning_rate": 2.4666213413817696e-05,
|
||
|
|
"loss": 0.6763,
|
||
|
|
"step": 305
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4623816641753864,
|
||
|
|
"grad_norm": 0.368576714561806,
|
||
|
|
"learning_rate": 2.4125710189263555e-05,
|
||
|
|
"loss": 0.6563,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.502242152466368,
|
||
|
|
"grad_norm": 0.35196724604768753,
|
||
|
|
"learning_rate": 2.3582040719269504e-05,
|
||
|
|
"loss": 0.65,
|
||
|
|
"step": 315
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5421026407573493,
|
||
|
|
"grad_norm": 0.3540473065272315,
|
||
|
|
"learning_rate": 2.3035622238905694e-05,
|
||
|
|
"loss": 0.6679,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5819631290483307,
|
||
|
|
"grad_norm": 0.3661596143769676,
|
||
|
|
"learning_rate": 2.2486874092949708e-05,
|
||
|
|
"loss": 0.6738,
|
||
|
|
"step": 325
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.621823617339312,
|
||
|
|
"grad_norm": 0.35913068756682465,
|
||
|
|
"learning_rate": 2.1936217414063584e-05,
|
||
|
|
"loss": 0.6887,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.661684105630294,
|
||
|
|
"grad_norm": 0.36799867292080646,
|
||
|
|
"learning_rate": 2.138407479959869e-05,
|
||
|
|
"loss": 0.6709,
|
||
|
|
"step": 335
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7015445939212754,
|
||
|
|
"grad_norm": 0.35691876781155074,
|
||
|
|
"learning_rate": 2.0830869987276537e-05,
|
||
|
|
"loss": 0.665,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.741405082212257,
|
||
|
|
"grad_norm": 0.38987523468576574,
|
||
|
|
"learning_rate": 2.027702752999444e-05,
|
||
|
|
"loss": 0.6528,
|
||
|
|
"step": 345
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7812655705032387,
|
||
|
|
"grad_norm": 0.38505381731754873,
|
||
|
|
"learning_rate": 1.9722972470005573e-05,
|
||
|
|
"loss": 0.6771,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.82112605879422,
|
||
|
|
"grad_norm": 0.35842950010166197,
|
||
|
|
"learning_rate": 1.916913001272347e-05,
|
||
|
|
"loss": 0.6638,
|
||
|
|
"step": 355
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8609865470852016,
|
||
|
|
"grad_norm": 0.37422580701088903,
|
||
|
|
"learning_rate": 1.8615925200401318e-05,
|
||
|
|
"loss": 0.6753,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9008470353761835,
|
||
|
|
"grad_norm": 0.36860495212005495,
|
||
|
|
"learning_rate": 1.806378258593642e-05,
|
||
|
|
"loss": 0.6681,
|
||
|
|
"step": 365
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.940707523667165,
|
||
|
|
"grad_norm": 0.3647903293380729,
|
||
|
|
"learning_rate": 1.7513125907050302e-05,
|
||
|
|
"loss": 0.6658,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9805680119581464,
|
||
|
|
"grad_norm": 0.35870478733795147,
|
||
|
|
"learning_rate": 1.6964377761094313e-05,
|
||
|
|
"loss": 0.667,
|
||
|
|
"step": 375
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0159441953163926,
|
||
|
|
"grad_norm": 0.6451047782511159,
|
||
|
|
"learning_rate": 1.6417959280730506e-05,
|
||
|
|
"loss": 0.612,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.055804683607374,
|
||
|
|
"grad_norm": 0.7456786595663745,
|
||
|
|
"learning_rate": 1.5874289810736452e-05,
|
||
|
|
"loss": 0.5282,
|
||
|
|
"step": 385
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.095665171898356,
|
||
|
|
"grad_norm": 0.4611479200453808,
|
||
|
|
"learning_rate": 1.5333786586182308e-05,
|
||
|
|
"loss": 0.4945,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1355256601893373,
|
||
|
|
"grad_norm": 0.4624423710225065,
|
||
|
|
"learning_rate": 1.4796864412226812e-05,
|
||
|
|
"loss": 0.5178,
|
||
|
|
"step": 395
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1753861484803187,
|
||
|
|
"grad_norm": 0.3988674416095005,
|
||
|
|
"learning_rate": 1.4263935345778202e-05,
|
||
|
|
"loss": 0.5015,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2152466367713006,
|
||
|
|
"grad_norm": 0.4171540538418279,
|
||
|
|
"learning_rate": 1.37354083792642e-05,
|
||
|
|
"loss": 0.4988,
|
||
|
|
"step": 405
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.255107125062282,
|
||
|
|
"grad_norm": 0.4043531142190362,
|
||
|
|
"learning_rate": 1.3211689126753879e-05,
|
||
|
|
"loss": 0.4966,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2949676133532635,
|
||
|
|
"grad_norm": 0.406846754086433,
|
||
|
|
"learning_rate": 1.26931795126721e-05,
|
||
|
|
"loss": 0.5081,
|
||
|
|
"step": 415
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.334828101644245,
|
||
|
|
"grad_norm": 0.42193109699205417,
|
||
|
|
"learning_rate": 1.2180277463345697e-05,
|
||
|
|
"loss": 0.5088,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.374688589935227,
|
||
|
|
"grad_norm": 0.4039970787325296,
|
||
|
|
"learning_rate": 1.167337660161783e-05,
|
||
|
|
"loss": 0.5023,
|
||
|
|
"step": 425
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4145490782262082,
|
||
|
|
"grad_norm": 0.3891294704118612,
|
||
|
|
"learning_rate": 1.1172865944765122e-05,
|
||
|
|
"loss": 0.5054,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4544095665171897,
|
||
|
|
"grad_norm": 0.4123628750841747,
|
||
|
|
"learning_rate": 1.067912960594923e-05,
|
||
|
|
"loss": 0.5078,
|
||
|
|
"step": 435
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4942700548081715,
|
||
|
|
"grad_norm": 0.4116103758152132,
|
||
|
|
"learning_rate": 1.0192546499432066e-05,
|
||
|
|
"loss": 0.5008,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.534130543099153,
|
||
|
|
"grad_norm": 0.39334581361661547,
|
||
|
|
"learning_rate": 9.713490049780931e-06,
|
||
|
|
"loss": 0.5117,
|
||
|
|
"step": 445
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5739910313901344,
|
||
|
|
"grad_norm": 0.3977163931160735,
|
||
|
|
"learning_rate": 9.242327905286552e-06,
|
||
|
|
"loss": 0.5004,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6138515196811163,
|
||
|
|
"grad_norm": 0.41291346064789947,
|
||
|
|
"learning_rate": 8.779421655814189e-06,
|
||
|
|
"loss": 0.4899,
|
||
|
|
"step": 455
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6537120079720977,
|
||
|
|
"grad_norm": 0.39864981854696013,
|
||
|
|
"learning_rate": 8.325126555304208e-06,
|
||
|
|
"loss": 0.4925,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.693572496263079,
|
||
|
|
"grad_norm": 0.3783439845572638,
|
||
|
|
"learning_rate": 7.879791249135059e-06,
|
||
|
|
"loss": 0.4936,
|
||
|
|
"step": 465
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.733432984554061,
|
||
|
|
"grad_norm": 0.36028247570266225,
|
||
|
|
"learning_rate": 7.443757506558033e-06,
|
||
|
|
"loss": 0.5071,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7732934728450425,
|
||
|
|
"grad_norm": 0.36974308586061955,
|
||
|
|
"learning_rate": 7.0173599584089625e-06,
|
||
|
|
"loss": 0.4902,
|
||
|
|
"step": 475
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.813153961136024,
|
||
|
|
"grad_norm": 0.3709767672807715,
|
||
|
|
"learning_rate": 6.600925840298331e-06,
|
||
|
|
"loss": 0.505,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8530144494270053,
|
||
|
|
"grad_norm": 0.3610583139391416,
|
||
|
|
"learning_rate": 6.1947747414767035e-06,
|
||
|
|
"loss": 0.5208,
|
||
|
|
"step": 485
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.892874937717987,
|
||
|
|
"grad_norm": 0.3791524633277678,
|
||
|
|
"learning_rate": 5.799218359568395e-06,
|
||
|
|
"loss": 0.5301,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9327354260089686,
|
||
|
|
"grad_norm": 0.363724718810471,
|
||
|
|
"learning_rate": 5.414560261361415e-06,
|
||
|
|
"loss": 0.4855,
|
||
|
|
"step": 495
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.97259591429995,
|
||
|
|
"grad_norm": 0.3867971355298049,
|
||
|
|
"learning_rate": 5.041095649837429e-06,
|
||
|
|
"loss": 0.4872,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.007972097658197,
|
||
|
|
"grad_norm": 0.6922635714147061,
|
||
|
|
"learning_rate": 4.679111137620442e-06,
|
||
|
|
"loss": 0.4901,
|
||
|
|
"step": 505
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.047832585949178,
|
||
|
|
"grad_norm": 0.5464640833763156,
|
||
|
|
"learning_rate": 4.328884527018067e-06,
|
||
|
|
"loss": 0.4086,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.08769307424016,
|
||
|
|
"grad_norm": 0.5514248166968356,
|
||
|
|
"learning_rate": 3.990684596824219e-06,
|
||
|
|
"loss": 0.4042,
|
||
|
|
"step": 515
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.127553562531141,
|
||
|
|
"grad_norm": 0.46194805338366773,
|
||
|
|
"learning_rate": 3.6647708960468696e-06,
|
||
|
|
"loss": 0.4028,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.1674140508221225,
|
||
|
|
"grad_norm": 0.45065104707045,
|
||
|
|
"learning_rate": 3.3513935447190595e-06,
|
||
|
|
"loss": 0.3937,
|
||
|
|
"step": 525
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.207274539113104,
|
||
|
|
"grad_norm": 0.3884400001553067,
|
||
|
|
"learning_rate": 3.050793041946183e-06,
|
||
|
|
"loss": 0.3941,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.247135027404085,
|
||
|
|
"grad_norm": 0.4057662200999109,
|
||
|
|
"learning_rate": 2.763200081336721e-06,
|
||
|
|
"loss": 0.3913,
|
||
|
|
"step": 535
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.286995515695067,
|
||
|
|
"grad_norm": 0.3927959847718352,
|
||
|
|
"learning_rate": 2.488835373958185e-06,
|
||
|
|
"loss": 0.3993,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.326856003986049,
|
||
|
|
"grad_norm": 0.3915532354537384,
|
||
|
|
"learning_rate": 2.2279094789540244e-06,
|
||
|
|
"loss": 0.4047,
|
||
|
|
"step": 545
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.36671649227703,
|
||
|
|
"grad_norm": 0.37576385131138085,
|
||
|
|
"learning_rate": 1.9806226419516195e-06,
|
||
|
|
"loss": 0.402,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.406576980568012,
|
||
|
|
"grad_norm": 0.3589030046466532,
|
||
|
|
"learning_rate": 1.7471646413852439e-06,
|
||
|
|
"loss": 0.4037,
|
||
|
|
"step": 555
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.446437468858994,
|
||
|
|
"grad_norm": 0.38802398319508147,
|
||
|
|
"learning_rate": 1.527714642852045e-06,
|
||
|
|
"loss": 0.3975,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.486297957149975,
|
||
|
|
"grad_norm": 0.3781055143033295,
|
||
|
|
"learning_rate": 1.3224410616127292e-06,
|
||
|
|
"loss": 0.3902,
|
||
|
|
"step": 565
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.526158445440957,
|
||
|
|
"grad_norm": 0.380556572113873,
|
||
|
|
"learning_rate": 1.1315014333425455e-06,
|
||
|
|
"loss": 0.3955,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.566018933731939,
|
||
|
|
"grad_norm": 0.37627984573745005,
|
||
|
|
"learning_rate": 9.550422932316938e-07,
|
||
|
|
"loss": 0.3925,
|
||
|
|
"step": 575
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.60587942202292,
|
||
|
|
"grad_norm": 0.36605761395168335,
|
||
|
|
"learning_rate": 7.931990635280052e-07,
|
||
|
|
"loss": 0.3929,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.645739910313901,
|
||
|
|
"grad_norm": 0.37403610200301257,
|
||
|
|
"learning_rate": 6.460959496081276e-07,
|
||
|
|
"loss": 0.4068,
|
||
|
|
"step": 585
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.685600398604883,
|
||
|
|
"grad_norm": 0.3673645131484265,
|
||
|
|
"learning_rate": 5.13845844657066e-07,
|
||
|
|
"loss": 0.3962,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.725460886895864,
|
||
|
|
"grad_norm": 0.39800543909133185,
|
||
|
|
"learning_rate": 3.965502430291235e-07,
|
||
|
|
"loss": 0.3912,
|
||
|
|
"step": 595
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.765321375186846,
|
||
|
|
"grad_norm": 0.37969719371341065,
|
||
|
|
"learning_rate": 2.942991623568436e-07,
|
||
|
|
"loss": 0.3801,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.805181863477827,
|
||
|
|
"grad_norm": 0.3777219824944034,
|
||
|
|
"learning_rate": 2.0717107446762696e-07,
|
||
|
|
"loss": 0.369,
|
||
|
|
"step": 605
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.845042351768809,
|
||
|
|
"grad_norm": 0.3913050535737085,
|
||
|
|
"learning_rate": 1.3523284516113955e-07,
|
||
|
|
"loss": 0.3982,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.884902840059791,
|
||
|
|
"grad_norm": 0.3712369916416077,
|
||
|
|
"learning_rate": 7.853968289363245e-08,
|
||
|
|
"loss": 0.3916,
|
||
|
|
"step": 615
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.924763328350773,
|
||
|
|
"grad_norm": 0.38101325204255376,
|
||
|
|
"learning_rate": 3.7135096408631443e-08,
|
||
|
|
"loss": 0.3978,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.964623816641754,
|
||
|
|
"grad_norm": 0.37989319704275204,
|
||
|
|
"learning_rate": 1.1050861346488806e-08,
|
||
|
|
"loss": 0.389,
|
||
|
|
"step": 625
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"grad_norm": 0.5514474703259893,
|
||
|
|
"learning_rate": 3.069958583856725e-10,
|
||
|
|
"loss": 0.4095,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"step": 630,
|
||
|
|
"total_flos": 1235947560108032.0,
|
||
|
|
"train_loss": 0.7053156269921197,
|
||
|
|
"train_runtime": 61423.1886,
|
||
|
|
"train_samples_per_second": 1.307,
|
||
|
|
"train_steps_per_second": 0.01
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 5,
|
||
|
|
"max_steps": 630,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 5,
|
||
|
|
"save_steps": 500,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": false,
|
||
|
|
"should_training_stop": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 1235947560108032.0,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|