rqadri's picture
Upload folder using huggingface_hub
df34abe verified
Raw History Blame Contribute Delete
90.9 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.46012824728707663,
"eval_steps": 500,
"global_step": 5597,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008220979940808945,
"grad_norm": 0.7510645985603333,
"learning_rate": 1.1061797752808989e-06,
"loss": 0.5669,
"step": 10
},
{
"epoch": 0.001644195988161789,
"grad_norm": 0.6885573267936707,
"learning_rate": 2.2241573033707866e-06,
"loss": 0.5563,
"step": 20
},
{
"epoch": 0.0024662939822426835,
"grad_norm": 0.6587523221969604,
"learning_rate": 3.342134831460675e-06,
"loss": 0.556,
"step": 30
},
{
"epoch": 0.003288391976323578,
"grad_norm": 0.6111008524894714,
"learning_rate": 4.4601123595505624e-06,
"loss": 0.5686,
"step": 40
},
{
"epoch": 0.004110489970404472,
"grad_norm": 0.6500401496887207,
"learning_rate": 5.5780898876404495e-06,
"loss": 0.5576,
"step": 50
},
{
"epoch": 0.004932587964485367,
"grad_norm": 0.6981553435325623,
"learning_rate": 6.6960674157303374e-06,
"loss": 0.5667,
"step": 60
},
{
"epoch": 0.005754685958566261,
"grad_norm": 0.5952242612838745,
"learning_rate": 7.814044943820226e-06,
"loss": 0.5455,
"step": 70
},
{
"epoch": 0.006576783952647156,
"grad_norm": 0.5906009674072266,
"learning_rate": 8.932022471910112e-06,
"loss": 0.5634,
"step": 80
},
{
"epoch": 0.00739888194672805,
"grad_norm": 0.5970085859298706,
"learning_rate": 1.005e-05,
"loss": 0.5666,
"step": 90
},
{
"epoch": 0.008220979940808944,
"grad_norm": 0.598000168800354,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.5561,
"step": 100
},
{
"epoch": 0.00904307793488984,
"grad_norm": 0.6920904517173767,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.5571,
"step": 110
},
{
"epoch": 0.009865175928970734,
"grad_norm": 0.5882319211959839,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.5831,
"step": 120
},
{
"epoch": 0.010687273923051627,
"grad_norm": 0.5964992642402649,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.5511,
"step": 130
},
{
"epoch": 0.011509371917132522,
"grad_norm": 0.6679021120071411,
"learning_rate": 1.563988764044944e-05,
"loss": 0.5583,
"step": 140
},
{
"epoch": 0.012331469911213417,
"grad_norm": 0.5619528889656067,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.5377,
"step": 150
},
{
"epoch": 0.013153567905294311,
"grad_norm": 0.5494176149368286,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.5539,
"step": 160
},
{
"epoch": 0.013975665899375206,
"grad_norm": 0.5273856520652771,
"learning_rate": 1.89938202247191e-05,
"loss": 0.5493,
"step": 170
},
{
"epoch": 0.0147977638934561,
"grad_norm": 0.5497198104858398,
"learning_rate": 2e-05,
"loss": 0.5684,
"step": 180
},
{
"epoch": 0.015619861887536994,
"grad_norm": 0.5500702261924744,
"learning_rate": 2e-05,
"loss": 0.5553,
"step": 190
},
{
"epoch": 0.01644195988161789,
"grad_norm": 0.5789015889167786,
"learning_rate": 2e-05,
"loss": 0.5661,
"step": 200
},
{
"epoch": 0.017264057875698784,
"grad_norm": 0.44678372144699097,
"learning_rate": 2e-05,
"loss": 0.5739,
"step": 210
},
{
"epoch": 0.01808615586977968,
"grad_norm": 0.4713243246078491,
"learning_rate": 2e-05,
"loss": 0.546,
"step": 220
},
{
"epoch": 0.018908253863860573,
"grad_norm": 0.5556458234786987,
"learning_rate": 2e-05,
"loss": 0.5751,
"step": 230
},
{
"epoch": 0.019730351857941468,
"grad_norm": 0.5001806616783142,
"learning_rate": 2e-05,
"loss": 0.5627,
"step": 240
},
{
"epoch": 0.020552449852022363,
"grad_norm": 0.4647273123264313,
"learning_rate": 2e-05,
"loss": 0.5765,
"step": 250
},
{
"epoch": 0.021374547846103254,
"grad_norm": 0.4200420379638672,
"learning_rate": 2e-05,
"loss": 0.5885,
"step": 260
},
{
"epoch": 0.02219664584018415,
"grad_norm": 0.5070334076881409,
"learning_rate": 2e-05,
"loss": 0.5428,
"step": 270
},
{
"epoch": 0.023018743834265044,
"grad_norm": 0.41507670283317566,
"learning_rate": 2e-05,
"loss": 0.5836,
"step": 280
},
{
"epoch": 0.02384084182834594,
"grad_norm": 0.4536267817020416,
"learning_rate": 2e-05,
"loss": 0.5687,
"step": 290
},
{
"epoch": 0.024662939822426833,
"grad_norm": 0.4120713472366333,
"learning_rate": 2e-05,
"loss": 0.5512,
"step": 300
},
{
"epoch": 0.025485037816507728,
"grad_norm": 0.44302648305892944,
"learning_rate": 2e-05,
"loss": 0.553,
"step": 310
},
{
"epoch": 0.026307135810588623,
"grad_norm": 0.41790035367012024,
"learning_rate": 2e-05,
"loss": 0.5793,
"step": 320
},
{
"epoch": 0.027129233804669518,
"grad_norm": 0.4572747051715851,
"learning_rate": 2e-05,
"loss": 0.5838,
"step": 330
},
{
"epoch": 0.027951331798750412,
"grad_norm": 0.4104056656360626,
"learning_rate": 2e-05,
"loss": 0.5716,
"step": 340
},
{
"epoch": 0.028773429792831307,
"grad_norm": 0.41988691687583923,
"learning_rate": 2e-05,
"loss": 0.5515,
"step": 350
},
{
"epoch": 0.0295955277869122,
"grad_norm": 0.3765335977077484,
"learning_rate": 2e-05,
"loss": 0.5535,
"step": 360
},
{
"epoch": 0.030417625780993093,
"grad_norm": 0.42008137702941895,
"learning_rate": 2e-05,
"loss": 0.544,
"step": 370
},
{
"epoch": 0.031239723775073988,
"grad_norm": 0.3445565402507782,
"learning_rate": 2e-05,
"loss": 0.5757,
"step": 380
},
{
"epoch": 0.03206182176915488,
"grad_norm": 0.4134063124656677,
"learning_rate": 2e-05,
"loss": 0.5484,
"step": 390
},
{
"epoch": 0.03288391976323578,
"grad_norm": 0.400969535112381,
"learning_rate": 2e-05,
"loss": 0.5582,
"step": 400
},
{
"epoch": 0.03370601775731667,
"grad_norm": 0.5791268348693848,
"learning_rate": 2e-05,
"loss": 0.5586,
"step": 410
},
{
"epoch": 0.03452811575139757,
"grad_norm": 0.3688918650150299,
"learning_rate": 2e-05,
"loss": 0.566,
"step": 420
},
{
"epoch": 0.03535021374547846,
"grad_norm": 0.41482818126678467,
"learning_rate": 2e-05,
"loss": 0.5501,
"step": 430
},
{
"epoch": 0.03617231173955936,
"grad_norm": 0.4072619676589966,
"learning_rate": 2e-05,
"loss": 0.5756,
"step": 440
},
{
"epoch": 0.03699440973364025,
"grad_norm": 0.389475554227829,
"learning_rate": 2e-05,
"loss": 0.5472,
"step": 450
},
{
"epoch": 0.037816507727721146,
"grad_norm": 0.3854443430900574,
"learning_rate": 2e-05,
"loss": 0.5475,
"step": 460
},
{
"epoch": 0.03863860572180204,
"grad_norm": 0.35742488503456116,
"learning_rate": 2e-05,
"loss": 0.5758,
"step": 470
},
{
"epoch": 0.039460703715882936,
"grad_norm": 0.3568844795227051,
"learning_rate": 2e-05,
"loss": 0.5463,
"step": 480
},
{
"epoch": 0.04028280170996383,
"grad_norm": 0.3795979917049408,
"learning_rate": 2e-05,
"loss": 0.5436,
"step": 490
},
{
"epoch": 0.041104899704044726,
"grad_norm": 0.37757909297943115,
"learning_rate": 2e-05,
"loss": 0.5663,
"step": 500
},
{
"epoch": 0.04192699769812561,
"grad_norm": 0.35400980710983276,
"learning_rate": 2e-05,
"loss": 0.5598,
"step": 510
},
{
"epoch": 0.04274909569220651,
"grad_norm": 0.3492357134819031,
"learning_rate": 2e-05,
"loss": 0.595,
"step": 520
},
{
"epoch": 0.0435711936862874,
"grad_norm": 0.3667033910751343,
"learning_rate": 2e-05,
"loss": 0.5476,
"step": 530
},
{
"epoch": 0.0443932916803683,
"grad_norm": 0.42593854665756226,
"learning_rate": 2e-05,
"loss": 0.5971,
"step": 540
},
{
"epoch": 0.04521538967444919,
"grad_norm": 0.37045472860336304,
"learning_rate": 2e-05,
"loss": 0.5591,
"step": 550
},
{
"epoch": 0.04603748766853009,
"grad_norm": 0.38231995701789856,
"learning_rate": 2e-05,
"loss": 0.5601,
"step": 560
},
{
"epoch": 0.04685958566261098,
"grad_norm": 0.37826621532440186,
"learning_rate": 2e-05,
"loss": 0.543,
"step": 570
},
{
"epoch": 0.04768168365669188,
"grad_norm": 0.3782760202884674,
"learning_rate": 2e-05,
"loss": 0.5883,
"step": 580
},
{
"epoch": 0.04850378165077277,
"grad_norm": 0.3855355679988861,
"learning_rate": 2e-05,
"loss": 0.561,
"step": 590
},
{
"epoch": 0.04932587964485367,
"grad_norm": 0.40973329544067383,
"learning_rate": 2e-05,
"loss": 0.5722,
"step": 600
},
{
"epoch": 0.05014797763893456,
"grad_norm": 0.3679392337799072,
"learning_rate": 2e-05,
"loss": 0.566,
"step": 610
},
{
"epoch": 0.050970075633015456,
"grad_norm": 0.36318325996398926,
"learning_rate": 2e-05,
"loss": 0.5539,
"step": 620
},
{
"epoch": 0.05179217362709635,
"grad_norm": 0.3448237478733063,
"learning_rate": 2e-05,
"loss": 0.5757,
"step": 630
},
{
"epoch": 0.052614271621177246,
"grad_norm": 0.48463839292526245,
"learning_rate": 2e-05,
"loss": 0.538,
"step": 640
},
{
"epoch": 0.05343636961525814,
"grad_norm": 0.368011474609375,
"learning_rate": 2e-05,
"loss": 0.5423,
"step": 650
},
{
"epoch": 0.054258467609339035,
"grad_norm": 0.3750072419643402,
"learning_rate": 2e-05,
"loss": 0.5709,
"step": 660
},
{
"epoch": 0.05508056560341993,
"grad_norm": 0.39397040009498596,
"learning_rate": 2e-05,
"loss": 0.5442,
"step": 670
},
{
"epoch": 0.055902663597500825,
"grad_norm": 0.38440752029418945,
"learning_rate": 2e-05,
"loss": 0.5653,
"step": 680
},
{
"epoch": 0.05672476159158172,
"grad_norm": 0.44925448298454285,
"learning_rate": 2e-05,
"loss": 0.5776,
"step": 690
},
{
"epoch": 0.057546859585662614,
"grad_norm": 0.4260270297527313,
"learning_rate": 2e-05,
"loss": 0.5737,
"step": 700
},
{
"epoch": 0.0583689575797435,
"grad_norm": 0.39539211988449097,
"learning_rate": 2e-05,
"loss": 0.5613,
"step": 710
},
{
"epoch": 0.0591910555738244,
"grad_norm": 0.35687392950057983,
"learning_rate": 2e-05,
"loss": 0.5827,
"step": 720
},
{
"epoch": 0.06001315356790529,
"grad_norm": 0.6145081520080566,
"learning_rate": 2e-05,
"loss": 0.5687,
"step": 730
},
{
"epoch": 0.06083525156198619,
"grad_norm": 0.4325614273548126,
"learning_rate": 2e-05,
"loss": 0.5366,
"step": 740
},
{
"epoch": 0.06165734955606708,
"grad_norm": 0.35582131147384644,
"learning_rate": 2e-05,
"loss": 0.5812,
"step": 750
},
{
"epoch": 0.062479447550147976,
"grad_norm": 0.38189223408699036,
"learning_rate": 2e-05,
"loss": 0.5593,
"step": 760
},
{
"epoch": 0.06330154554422887,
"grad_norm": 0.4139549434185028,
"learning_rate": 2e-05,
"loss": 0.5186,
"step": 770
},
{
"epoch": 0.06412364353830977,
"grad_norm": 0.39057818055152893,
"learning_rate": 2e-05,
"loss": 0.5161,
"step": 780
},
{
"epoch": 0.06494574153239066,
"grad_norm": 0.46800991892814636,
"learning_rate": 2e-05,
"loss": 0.5564,
"step": 790
},
{
"epoch": 0.06576783952647156,
"grad_norm": 0.3704215884208679,
"learning_rate": 2e-05,
"loss": 0.5404,
"step": 800
},
{
"epoch": 0.06658993752055245,
"grad_norm": 0.3691960871219635,
"learning_rate": 2e-05,
"loss": 0.5429,
"step": 810
},
{
"epoch": 0.06741203551463334,
"grad_norm": 0.35629382729530334,
"learning_rate": 2e-05,
"loss": 0.5053,
"step": 820
},
{
"epoch": 0.06823413350871424,
"grad_norm": 0.45477181673049927,
"learning_rate": 2e-05,
"loss": 0.5396,
"step": 830
},
{
"epoch": 0.06905623150279513,
"grad_norm": 0.36613255739212036,
"learning_rate": 2e-05,
"loss": 0.5621,
"step": 840
},
{
"epoch": 0.06987832949687603,
"grad_norm": 0.38469594717025757,
"learning_rate": 2e-05,
"loss": 0.5734,
"step": 850
},
{
"epoch": 0.07070042749095692,
"grad_norm": 0.38618406653404236,
"learning_rate": 2e-05,
"loss": 0.5549,
"step": 860
},
{
"epoch": 0.07152252548503782,
"grad_norm": 0.3400881588459015,
"learning_rate": 2e-05,
"loss": 0.5348,
"step": 870
},
{
"epoch": 0.07234462347911871,
"grad_norm": 0.36379608511924744,
"learning_rate": 2e-05,
"loss": 0.5436,
"step": 880
},
{
"epoch": 0.07316672147319961,
"grad_norm": 0.4529801309108734,
"learning_rate": 2e-05,
"loss": 0.5772,
"step": 890
},
{
"epoch": 0.0739888194672805,
"grad_norm": 0.37483784556388855,
"learning_rate": 2e-05,
"loss": 0.5383,
"step": 900
},
{
"epoch": 0.0748109174613614,
"grad_norm": 0.42458927631378174,
"learning_rate": 2e-05,
"loss": 0.5492,
"step": 910
},
{
"epoch": 0.07563301545544229,
"grad_norm": 0.3337687849998474,
"learning_rate": 2e-05,
"loss": 0.5676,
"step": 920
},
{
"epoch": 0.07645511344952319,
"grad_norm": 0.40329763293266296,
"learning_rate": 2e-05,
"loss": 0.5502,
"step": 930
},
{
"epoch": 0.07727721144360408,
"grad_norm": 0.37778130173683167,
"learning_rate": 2e-05,
"loss": 0.5855,
"step": 940
},
{
"epoch": 0.07809930943768498,
"grad_norm": 0.4349541962146759,
"learning_rate": 2e-05,
"loss": 0.5457,
"step": 950
},
{
"epoch": 0.07892140743176587,
"grad_norm": 0.33621495962142944,
"learning_rate": 2e-05,
"loss": 0.5314,
"step": 960
},
{
"epoch": 0.07974350542584677,
"grad_norm": 0.3857935965061188,
"learning_rate": 2e-05,
"loss": 0.5672,
"step": 970
},
{
"epoch": 0.08056560341992766,
"grad_norm": 0.3538459539413452,
"learning_rate": 2e-05,
"loss": 0.5577,
"step": 980
},
{
"epoch": 0.08138770141400856,
"grad_norm": 0.3277754783630371,
"learning_rate": 2e-05,
"loss": 0.5736,
"step": 990
},
{
"epoch": 0.08220979940808945,
"grad_norm": 0.33061307668685913,
"learning_rate": 2e-05,
"loss": 0.5463,
"step": 1000
},
{
"epoch": 0.08303189740217033,
"grad_norm": 0.5522162914276123,
"learning_rate": 2e-05,
"loss": 0.5415,
"step": 1010
},
{
"epoch": 0.08385399539625123,
"grad_norm": 0.337895929813385,
"learning_rate": 2e-05,
"loss": 0.5408,
"step": 1020
},
{
"epoch": 0.08467609339033212,
"grad_norm": 0.4023716449737549,
"learning_rate": 2e-05,
"loss": 0.5925,
"step": 1030
},
{
"epoch": 0.08549819138441302,
"grad_norm": 0.35799404978752136,
"learning_rate": 2e-05,
"loss": 0.5532,
"step": 1040
},
{
"epoch": 0.08632028937849391,
"grad_norm": 0.3547460734844208,
"learning_rate": 2e-05,
"loss": 0.5552,
"step": 1050
},
{
"epoch": 0.0871423873725748,
"grad_norm": 0.44388723373413086,
"learning_rate": 2e-05,
"loss": 0.5746,
"step": 1060
},
{
"epoch": 0.0879644853666557,
"grad_norm": 0.3479222059249878,
"learning_rate": 2e-05,
"loss": 0.5638,
"step": 1070
},
{
"epoch": 0.0887865833607366,
"grad_norm": 0.37281978130340576,
"learning_rate": 2e-05,
"loss": 0.5595,
"step": 1080
},
{
"epoch": 0.08960868135481749,
"grad_norm": 0.3608080744743347,
"learning_rate": 2e-05,
"loss": 0.5655,
"step": 1090
},
{
"epoch": 0.09043077934889839,
"grad_norm": 0.3156391680240631,
"learning_rate": 2e-05,
"loss": 0.5364,
"step": 1100
},
{
"epoch": 0.09125287734297928,
"grad_norm": 0.3983137905597687,
"learning_rate": 2e-05,
"loss": 0.5511,
"step": 1110
},
{
"epoch": 0.09207497533706017,
"grad_norm": 0.3987083435058594,
"learning_rate": 2e-05,
"loss": 0.5419,
"step": 1120
},
{
"epoch": 0.09289707333114107,
"grad_norm": 0.3358446955680847,
"learning_rate": 2e-05,
"loss": 0.5497,
"step": 1130
},
{
"epoch": 0.09371917132522196,
"grad_norm": 0.3355688154697418,
"learning_rate": 2e-05,
"loss": 0.5395,
"step": 1140
},
{
"epoch": 0.09454126931930286,
"grad_norm": 0.39273035526275635,
"learning_rate": 2e-05,
"loss": 0.5551,
"step": 1150
},
{
"epoch": 0.09536336731338375,
"grad_norm": 0.37203627824783325,
"learning_rate": 2e-05,
"loss": 0.5894,
"step": 1160
},
{
"epoch": 0.09618546530746465,
"grad_norm": 0.34250086545944214,
"learning_rate": 2e-05,
"loss": 0.5475,
"step": 1170
},
{
"epoch": 0.09700756330154554,
"grad_norm": 0.3525017499923706,
"learning_rate": 2e-05,
"loss": 0.5527,
"step": 1180
},
{
"epoch": 0.09782966129562644,
"grad_norm": 0.37996408343315125,
"learning_rate": 2e-05,
"loss": 0.5454,
"step": 1190
},
{
"epoch": 0.09865175928970733,
"grad_norm": 0.46024736762046814,
"learning_rate": 2e-05,
"loss": 0.5672,
"step": 1200
},
{
"epoch": 0.09947385728378823,
"grad_norm": 0.3837468922138214,
"learning_rate": 2e-05,
"loss": 0.5551,
"step": 1210
},
{
"epoch": 0.10029595527786912,
"grad_norm": 0.40620625019073486,
"learning_rate": 2e-05,
"loss": 0.5468,
"step": 1220
},
{
"epoch": 0.10111805327195002,
"grad_norm": 0.35995301604270935,
"learning_rate": 2e-05,
"loss": 0.5552,
"step": 1230
},
{
"epoch": 0.10194015126603091,
"grad_norm": 0.396985799074173,
"learning_rate": 2e-05,
"loss": 0.5298,
"step": 1240
},
{
"epoch": 0.10276224926011181,
"grad_norm": 0.38154250383377075,
"learning_rate": 2e-05,
"loss": 0.5574,
"step": 1250
},
{
"epoch": 0.1035843472541927,
"grad_norm": 0.3399718403816223,
"learning_rate": 2e-05,
"loss": 0.5873,
"step": 1260
},
{
"epoch": 0.1044064452482736,
"grad_norm": 0.39190253615379333,
"learning_rate": 2e-05,
"loss": 0.523,
"step": 1270
},
{
"epoch": 0.10522854324235449,
"grad_norm": 0.34379804134368896,
"learning_rate": 2e-05,
"loss": 0.5429,
"step": 1280
},
{
"epoch": 0.10605064123643539,
"grad_norm": 0.34028005599975586,
"learning_rate": 2e-05,
"loss": 0.5527,
"step": 1290
},
{
"epoch": 0.10687273923051628,
"grad_norm": 0.3677566349506378,
"learning_rate": 2e-05,
"loss": 0.544,
"step": 1300
},
{
"epoch": 0.10769483722459718,
"grad_norm": 0.3977677524089813,
"learning_rate": 2e-05,
"loss": 0.53,
"step": 1310
},
{
"epoch": 0.10851693521867807,
"grad_norm": 0.33426037430763245,
"learning_rate": 2e-05,
"loss": 0.54,
"step": 1320
},
{
"epoch": 0.10933903321275897,
"grad_norm": 0.3396483361721039,
"learning_rate": 2e-05,
"loss": 0.5681,
"step": 1330
},
{
"epoch": 0.11016113120683986,
"grad_norm": 0.3713797926902771,
"learning_rate": 2e-05,
"loss": 0.5565,
"step": 1340
},
{
"epoch": 0.11098322920092075,
"grad_norm": 0.3870110511779785,
"learning_rate": 2e-05,
"loss": 0.5489,
"step": 1350
},
{
"epoch": 0.11180532719500165,
"grad_norm": 0.4049339294433594,
"learning_rate": 2e-05,
"loss": 0.5656,
"step": 1360
},
{
"epoch": 0.11262742518908254,
"grad_norm": 0.3298855423927307,
"learning_rate": 2e-05,
"loss": 0.5605,
"step": 1370
},
{
"epoch": 0.11344952318316344,
"grad_norm": 0.3444337546825409,
"learning_rate": 2e-05,
"loss": 0.5446,
"step": 1380
},
{
"epoch": 0.11427162117724433,
"grad_norm": 0.329982191324234,
"learning_rate": 2e-05,
"loss": 0.5602,
"step": 1390
},
{
"epoch": 0.11509371917132523,
"grad_norm": 0.3550669252872467,
"learning_rate": 2e-05,
"loss": 0.5293,
"step": 1400
},
{
"epoch": 0.11591581716540611,
"grad_norm": 0.3423064053058624,
"learning_rate": 2e-05,
"loss": 0.5502,
"step": 1410
},
{
"epoch": 0.116737915159487,
"grad_norm": 0.4455831050872803,
"learning_rate": 2e-05,
"loss": 0.5537,
"step": 1420
},
{
"epoch": 0.1175600131535679,
"grad_norm": 0.3592797517776489,
"learning_rate": 2e-05,
"loss": 0.5289,
"step": 1430
},
{
"epoch": 0.1183821111476488,
"grad_norm": 0.3526698648929596,
"learning_rate": 2e-05,
"loss": 0.5458,
"step": 1440
},
{
"epoch": 0.11920420914172969,
"grad_norm": 0.3346903324127197,
"learning_rate": 2e-05,
"loss": 0.5394,
"step": 1450
},
{
"epoch": 0.12002630713581058,
"grad_norm": 0.36272871494293213,
"learning_rate": 2e-05,
"loss": 0.5565,
"step": 1460
},
{
"epoch": 0.12084840512989148,
"grad_norm": 0.38370686769485474,
"learning_rate": 2e-05,
"loss": 0.5616,
"step": 1470
},
{
"epoch": 0.12167050312397237,
"grad_norm": 0.43002524971961975,
"learning_rate": 2e-05,
"loss": 0.5522,
"step": 1480
},
{
"epoch": 0.12249260111805327,
"grad_norm": 0.3865845501422882,
"learning_rate": 2e-05,
"loss": 0.5561,
"step": 1490
},
{
"epoch": 0.12331469911213416,
"grad_norm": 0.3935903012752533,
"learning_rate": 2e-05,
"loss": 0.5519,
"step": 1500
},
{
"epoch": 0.12413679710621506,
"grad_norm": 0.3224612772464752,
"learning_rate": 2e-05,
"loss": 0.5478,
"step": 1510
},
{
"epoch": 0.12495889510029595,
"grad_norm": 0.39488735795021057,
"learning_rate": 2e-05,
"loss": 0.5589,
"step": 1520
},
{
"epoch": 0.12578099309437685,
"grad_norm": 0.366058349609375,
"learning_rate": 2e-05,
"loss": 0.5673,
"step": 1530
},
{
"epoch": 0.12660309108845774,
"grad_norm": 0.40682297945022583,
"learning_rate": 2e-05,
"loss": 0.5565,
"step": 1540
},
{
"epoch": 0.12742518908253864,
"grad_norm": 0.3384777903556824,
"learning_rate": 2e-05,
"loss": 0.5319,
"step": 1550
},
{
"epoch": 0.12824728707661953,
"grad_norm": 0.3520301282405853,
"learning_rate": 2e-05,
"loss": 0.5555,
"step": 1560
},
{
"epoch": 0.12906938507070043,
"grad_norm": 0.37539342045783997,
"learning_rate": 2e-05,
"loss": 0.5639,
"step": 1570
},
{
"epoch": 0.12989148306478132,
"grad_norm": 0.418217271566391,
"learning_rate": 2e-05,
"loss": 0.5483,
"step": 1580
},
{
"epoch": 0.13071358105886222,
"grad_norm": 0.37258896231651306,
"learning_rate": 2e-05,
"loss": 0.6018,
"step": 1590
},
{
"epoch": 0.1315356790529431,
"grad_norm": 0.35443881154060364,
"learning_rate": 2e-05,
"loss": 0.5485,
"step": 1600
},
{
"epoch": 0.132357777047024,
"grad_norm": 0.3421747386455536,
"learning_rate": 2e-05,
"loss": 0.5602,
"step": 1610
},
{
"epoch": 0.1331798750411049,
"grad_norm": 0.3465687036514282,
"learning_rate": 2e-05,
"loss": 0.5542,
"step": 1620
},
{
"epoch": 0.1340019730351858,
"grad_norm": 0.7691576480865479,
"learning_rate": 2e-05,
"loss": 0.5456,
"step": 1630
},
{
"epoch": 0.1348240710292667,
"grad_norm": 0.6803935766220093,
"learning_rate": 2e-05,
"loss": 0.58,
"step": 1640
},
{
"epoch": 0.13564616902334758,
"grad_norm": 0.3570682108402252,
"learning_rate": 2e-05,
"loss": 0.5721,
"step": 1650
},
{
"epoch": 0.13646826701742848,
"grad_norm": 0.5207669734954834,
"learning_rate": 2e-05,
"loss": 0.5423,
"step": 1660
},
{
"epoch": 0.13729036501150937,
"grad_norm": 0.32501304149627686,
"learning_rate": 2e-05,
"loss": 0.527,
"step": 1670
},
{
"epoch": 0.13811246300559027,
"grad_norm": 0.39292895793914795,
"learning_rate": 2e-05,
"loss": 0.5437,
"step": 1680
},
{
"epoch": 0.13893456099967116,
"grad_norm": 0.3334016501903534,
"learning_rate": 2e-05,
"loss": 0.5599,
"step": 1690
},
{
"epoch": 0.13975665899375206,
"grad_norm": 0.31479495763778687,
"learning_rate": 2e-05,
"loss": 0.5658,
"step": 1700
},
{
"epoch": 0.14057875698783295,
"grad_norm": 0.38708022236824036,
"learning_rate": 2e-05,
"loss": 0.5554,
"step": 1710
},
{
"epoch": 0.14140085498191385,
"grad_norm": 0.29850685596466064,
"learning_rate": 2e-05,
"loss": 0.5217,
"step": 1720
},
{
"epoch": 0.14222295297599474,
"grad_norm": 0.3497243821620941,
"learning_rate": 2e-05,
"loss": 0.5571,
"step": 1730
},
{
"epoch": 0.14304505097007564,
"grad_norm": 0.34561386704444885,
"learning_rate": 2e-05,
"loss": 0.5815,
"step": 1740
},
{
"epoch": 0.14386714896415653,
"grad_norm": 0.3378225862979889,
"learning_rate": 2e-05,
"loss": 0.5522,
"step": 1750
},
{
"epoch": 0.14468924695823743,
"grad_norm": 0.3681405782699585,
"learning_rate": 2e-05,
"loss": 0.5257,
"step": 1760
},
{
"epoch": 0.14551134495231832,
"grad_norm": 0.3730759024620056,
"learning_rate": 2e-05,
"loss": 0.5182,
"step": 1770
},
{
"epoch": 0.14633344294639922,
"grad_norm": 0.41251009702682495,
"learning_rate": 2e-05,
"loss": 0.5824,
"step": 1780
},
{
"epoch": 0.1471555409404801,
"grad_norm": 0.34546494483947754,
"learning_rate": 2e-05,
"loss": 0.5333,
"step": 1790
},
{
"epoch": 0.147977638934561,
"grad_norm": 0.3387209177017212,
"learning_rate": 2e-05,
"loss": 0.5125,
"step": 1800
},
{
"epoch": 0.1487997369286419,
"grad_norm": 0.3762285113334656,
"learning_rate": 2e-05,
"loss": 0.5307,
"step": 1810
},
{
"epoch": 0.1496218349227228,
"grad_norm": 0.34216463565826416,
"learning_rate": 2e-05,
"loss": 0.5536,
"step": 1820
},
{
"epoch": 0.1504439329168037,
"grad_norm": 0.3812594711780548,
"learning_rate": 2e-05,
"loss": 0.5356,
"step": 1830
},
{
"epoch": 0.15126603091088459,
"grad_norm": 0.34243831038475037,
"learning_rate": 2e-05,
"loss": 0.5408,
"step": 1840
},
{
"epoch": 0.15208812890496548,
"grad_norm": 0.3886333107948303,
"learning_rate": 2e-05,
"loss": 0.5536,
"step": 1850
},
{
"epoch": 0.15291022689904638,
"grad_norm": 0.37051960825920105,
"learning_rate": 2e-05,
"loss": 0.5584,
"step": 1860
},
{
"epoch": 0.15373232489312727,
"grad_norm": 0.3963000178337097,
"learning_rate": 2e-05,
"loss": 0.5415,
"step": 1870
},
{
"epoch": 0.15455442288720816,
"grad_norm": 0.36416611075401306,
"learning_rate": 2e-05,
"loss": 0.5514,
"step": 1880
},
{
"epoch": 0.15537652088128906,
"grad_norm": 0.765783965587616,
"learning_rate": 2e-05,
"loss": 0.5476,
"step": 1890
},
{
"epoch": 0.15619861887536995,
"grad_norm": 0.3947313129901886,
"learning_rate": 2e-05,
"loss": 0.5351,
"step": 1900
},
{
"epoch": 0.15702071686945085,
"grad_norm": 0.42167967557907104,
"learning_rate": 2e-05,
"loss": 0.584,
"step": 1910
},
{
"epoch": 0.15784281486353174,
"grad_norm": 0.3771551251411438,
"learning_rate": 2e-05,
"loss": 0.5411,
"step": 1920
},
{
"epoch": 0.15866491285761264,
"grad_norm": 0.34432297945022583,
"learning_rate": 2e-05,
"loss": 0.5238,
"step": 1930
},
{
"epoch": 0.15948701085169353,
"grad_norm": 0.3422872722148895,
"learning_rate": 2e-05,
"loss": 0.5669,
"step": 1940
},
{
"epoch": 0.16030910884577443,
"grad_norm": 0.35941794514656067,
"learning_rate": 2e-05,
"loss": 0.5323,
"step": 1950
},
{
"epoch": 0.16113120683985532,
"grad_norm": 0.4349963068962097,
"learning_rate": 2e-05,
"loss": 0.5451,
"step": 1960
},
{
"epoch": 0.16195330483393622,
"grad_norm": 0.49831265211105347,
"learning_rate": 2e-05,
"loss": 0.5511,
"step": 1970
},
{
"epoch": 0.1627754028280171,
"grad_norm": 0.3435792624950409,
"learning_rate": 2e-05,
"loss": 0.5323,
"step": 1980
},
{
"epoch": 0.163597500822098,
"grad_norm": 0.5475247502326965,
"learning_rate": 2e-05,
"loss": 0.5605,
"step": 1990
},
{
"epoch": 0.1644195988161789,
"grad_norm": 0.3261669874191284,
"learning_rate": 2e-05,
"loss": 0.5466,
"step": 2000
},
{
"epoch": 0.16524169681025977,
"grad_norm": 0.3482590615749359,
"learning_rate": 2e-05,
"loss": 0.5449,
"step": 2010
},
{
"epoch": 0.16606379480434066,
"grad_norm": 0.41366517543792725,
"learning_rate": 2e-05,
"loss": 0.5535,
"step": 2020
},
{
"epoch": 0.16688589279842156,
"grad_norm": 0.36402422189712524,
"learning_rate": 2e-05,
"loss": 0.5613,
"step": 2030
},
{
"epoch": 0.16770799079250245,
"grad_norm": 0.347074419260025,
"learning_rate": 2e-05,
"loss": 0.5455,
"step": 2040
},
{
"epoch": 0.16853008878658335,
"grad_norm": 0.35407862067222595,
"learning_rate": 2e-05,
"loss": 0.5549,
"step": 2050
},
{
"epoch": 0.16935218678066424,
"grad_norm": 0.3733304738998413,
"learning_rate": 2e-05,
"loss": 0.5534,
"step": 2060
},
{
"epoch": 0.17017428477474514,
"grad_norm": 0.36466488242149353,
"learning_rate": 2e-05,
"loss": 0.5256,
"step": 2070
},
{
"epoch": 0.17099638276882603,
"grad_norm": 0.3579888641834259,
"learning_rate": 2e-05,
"loss": 0.5472,
"step": 2080
},
{
"epoch": 0.17181848076290693,
"grad_norm": 0.3732098639011383,
"learning_rate": 2e-05,
"loss": 0.5302,
"step": 2090
},
{
"epoch": 0.17264057875698782,
"grad_norm": 0.3745439946651459,
"learning_rate": 2e-05,
"loss": 0.5729,
"step": 2100
},
{
"epoch": 0.17346267675106872,
"grad_norm": 0.33778682351112366,
"learning_rate": 2e-05,
"loss": 0.5414,
"step": 2110
},
{
"epoch": 0.1742847747451496,
"grad_norm": 0.4035145044326782,
"learning_rate": 2e-05,
"loss": 0.5604,
"step": 2120
},
{
"epoch": 0.1751068727392305,
"grad_norm": 0.41023483872413635,
"learning_rate": 2e-05,
"loss": 0.5526,
"step": 2130
},
{
"epoch": 0.1759289707333114,
"grad_norm": 0.3626518249511719,
"learning_rate": 2e-05,
"loss": 0.5384,
"step": 2140
},
{
"epoch": 0.1767510687273923,
"grad_norm": 0.3237241506576538,
"learning_rate": 2e-05,
"loss": 0.5577,
"step": 2150
},
{
"epoch": 0.1775731667214732,
"grad_norm": 0.37217530608177185,
"learning_rate": 2e-05,
"loss": 0.5607,
"step": 2160
},
{
"epoch": 0.17839526471555409,
"grad_norm": 0.3166368305683136,
"learning_rate": 2e-05,
"loss": 0.5184,
"step": 2170
},
{
"epoch": 0.17921736270963498,
"grad_norm": 4.73347282409668,
"learning_rate": 2e-05,
"loss": 0.5469,
"step": 2180
},
{
"epoch": 0.18003946070371588,
"grad_norm": 0.3166395127773285,
"learning_rate": 2e-05,
"loss": 0.548,
"step": 2190
},
{
"epoch": 0.18086155869779677,
"grad_norm": 0.37149322032928467,
"learning_rate": 2e-05,
"loss": 0.5374,
"step": 2200
},
{
"epoch": 0.18168365669187767,
"grad_norm": 0.3564207851886749,
"learning_rate": 2e-05,
"loss": 0.5534,
"step": 2210
},
{
"epoch": 0.18250575468595856,
"grad_norm": 0.35720905661582947,
"learning_rate": 2e-05,
"loss": 0.5442,
"step": 2220
},
{
"epoch": 0.18332785268003945,
"grad_norm": 0.3910946249961853,
"learning_rate": 2e-05,
"loss": 0.5389,
"step": 2230
},
{
"epoch": 0.18414995067412035,
"grad_norm": 0.35082006454467773,
"learning_rate": 2e-05,
"loss": 0.5641,
"step": 2240
},
{
"epoch": 0.18497204866820124,
"grad_norm": 0.352914035320282,
"learning_rate": 2e-05,
"loss": 0.572,
"step": 2250
},
{
"epoch": 0.18579414666228214,
"grad_norm": 0.3582814335823059,
"learning_rate": 2e-05,
"loss": 0.549,
"step": 2260
},
{
"epoch": 0.18661624465636303,
"grad_norm": 0.4184710681438446,
"learning_rate": 2e-05,
"loss": 0.5692,
"step": 2270
},
{
"epoch": 0.18743834265044393,
"grad_norm": 0.5157349109649658,
"learning_rate": 2e-05,
"loss": 0.5233,
"step": 2280
},
{
"epoch": 0.18826044064452482,
"grad_norm": 0.438728928565979,
"learning_rate": 2e-05,
"loss": 0.5388,
"step": 2290
},
{
"epoch": 0.18908253863860572,
"grad_norm": 0.31124523282051086,
"learning_rate": 2e-05,
"loss": 0.539,
"step": 2300
},
{
"epoch": 0.1899046366326866,
"grad_norm": 0.35619640350341797,
"learning_rate": 2e-05,
"loss": 0.5229,
"step": 2310
},
{
"epoch": 0.1907267346267675,
"grad_norm": 0.3216730058193207,
"learning_rate": 2e-05,
"loss": 0.5358,
"step": 2320
},
{
"epoch": 0.1915488326208484,
"grad_norm": 0.38024744391441345,
"learning_rate": 2e-05,
"loss": 0.562,
"step": 2330
},
{
"epoch": 0.1923709306149293,
"grad_norm": 0.3398594856262207,
"learning_rate": 2e-05,
"loss": 0.5448,
"step": 2340
},
{
"epoch": 0.1931930286090102,
"grad_norm": 0.3224997818470001,
"learning_rate": 2e-05,
"loss": 0.549,
"step": 2350
},
{
"epoch": 0.1940151266030911,
"grad_norm": 0.43699830770492554,
"learning_rate": 2e-05,
"loss": 0.5309,
"step": 2360
},
{
"epoch": 0.19483722459717198,
"grad_norm": 0.7130771279335022,
"learning_rate": 2e-05,
"loss": 0.5407,
"step": 2370
},
{
"epoch": 0.19565932259125288,
"grad_norm": 0.345954567193985,
"learning_rate": 2e-05,
"loss": 0.5396,
"step": 2380
},
{
"epoch": 0.19648142058533377,
"grad_norm": 0.3121144473552704,
"learning_rate": 2e-05,
"loss": 0.5435,
"step": 2390
},
{
"epoch": 0.19730351857941467,
"grad_norm": 0.47650909423828125,
"learning_rate": 2e-05,
"loss": 0.542,
"step": 2400
},
{
"epoch": 0.19812561657349556,
"grad_norm": 0.3822428286075592,
"learning_rate": 2e-05,
"loss": 0.5577,
"step": 2410
},
{
"epoch": 0.19894771456757646,
"grad_norm": 0.3568475842475891,
"learning_rate": 2e-05,
"loss": 0.5418,
"step": 2420
},
{
"epoch": 0.19976981256165735,
"grad_norm": 0.3287544846534729,
"learning_rate": 2e-05,
"loss": 0.5381,
"step": 2430
},
{
"epoch": 0.20059191055573825,
"grad_norm": 0.3641667664051056,
"learning_rate": 2e-05,
"loss": 0.5694,
"step": 2440
},
{
"epoch": 0.20141400854981914,
"grad_norm": 0.3377435803413391,
"learning_rate": 2e-05,
"loss": 0.5552,
"step": 2450
},
{
"epoch": 0.20223610654390003,
"grad_norm": 0.41057026386260986,
"learning_rate": 2e-05,
"loss": 0.5421,
"step": 2460
},
{
"epoch": 0.20305820453798093,
"grad_norm": 0.3275510370731354,
"learning_rate": 2e-05,
"loss": 0.5646,
"step": 2470
},
{
"epoch": 0.20388030253206182,
"grad_norm": 0.31972360610961914,
"learning_rate": 2e-05,
"loss": 0.5204,
"step": 2480
},
{
"epoch": 0.20470240052614272,
"grad_norm": 0.37606051564216614,
"learning_rate": 2e-05,
"loss": 0.5597,
"step": 2490
},
{
"epoch": 0.20552449852022361,
"grad_norm": 0.3712795674800873,
"learning_rate": 2e-05,
"loss": 0.5722,
"step": 2500
},
{
"epoch": 0.2063465965143045,
"grad_norm": 0.4065472483634949,
"learning_rate": 2e-05,
"loss": 0.5362,
"step": 2510
},
{
"epoch": 0.2071686945083854,
"grad_norm": 0.34219712018966675,
"learning_rate": 2e-05,
"loss": 0.5525,
"step": 2520
},
{
"epoch": 0.2079907925024663,
"grad_norm": 0.3456852436065674,
"learning_rate": 2e-05,
"loss": 0.5529,
"step": 2530
},
{
"epoch": 0.2088128904965472,
"grad_norm": 0.7063271999359131,
"learning_rate": 2e-05,
"loss": 0.5781,
"step": 2540
},
{
"epoch": 0.2096349884906281,
"grad_norm": 0.3616321086883545,
"learning_rate": 2e-05,
"loss": 0.5553,
"step": 2550
},
{
"epoch": 0.21045708648470898,
"grad_norm": 0.34892845153808594,
"learning_rate": 2e-05,
"loss": 0.5239,
"step": 2560
},
{
"epoch": 0.21127918447878988,
"grad_norm": 0.33791273832321167,
"learning_rate": 2e-05,
"loss": 0.5362,
"step": 2570
},
{
"epoch": 0.21210128247287077,
"grad_norm": 0.34417805075645447,
"learning_rate": 2e-05,
"loss": 0.5361,
"step": 2580
},
{
"epoch": 0.21292338046695167,
"grad_norm": 0.37011924386024475,
"learning_rate": 2e-05,
"loss": 0.5126,
"step": 2590
},
{
"epoch": 0.21374547846103256,
"grad_norm": 0.39199256896972656,
"learning_rate": 2e-05,
"loss": 0.5637,
"step": 2600
},
{
"epoch": 0.21456757645511346,
"grad_norm": 0.326111763715744,
"learning_rate": 2e-05,
"loss": 0.5169,
"step": 2610
},
{
"epoch": 0.21538967444919435,
"grad_norm": 0.33117803931236267,
"learning_rate": 2e-05,
"loss": 0.5498,
"step": 2620
},
{
"epoch": 0.21621177244327525,
"grad_norm": 0.39731863141059875,
"learning_rate": 2e-05,
"loss": 0.5821,
"step": 2630
},
{
"epoch": 0.21703387043735614,
"grad_norm": 0.33802589774131775,
"learning_rate": 2e-05,
"loss": 0.5496,
"step": 2640
},
{
"epoch": 0.21785596843143704,
"grad_norm": 0.36961087584495544,
"learning_rate": 2e-05,
"loss": 0.5138,
"step": 2650
},
{
"epoch": 0.21867806642551793,
"grad_norm": 0.35770997405052185,
"learning_rate": 2e-05,
"loss": 0.5267,
"step": 2660
},
{
"epoch": 0.21950016441959883,
"grad_norm": 0.337510347366333,
"learning_rate": 2e-05,
"loss": 0.5712,
"step": 2670
},
{
"epoch": 0.22032226241367972,
"grad_norm": 0.3274226486682892,
"learning_rate": 2e-05,
"loss": 0.5355,
"step": 2680
},
{
"epoch": 0.22114436040776062,
"grad_norm": 0.41182374954223633,
"learning_rate": 2e-05,
"loss": 0.5395,
"step": 2690
},
{
"epoch": 0.2219664584018415,
"grad_norm": 0.3141595721244812,
"learning_rate": 2e-05,
"loss": 0.5567,
"step": 2700
},
{
"epoch": 0.2227885563959224,
"grad_norm": 0.35677820444107056,
"learning_rate": 2e-05,
"loss": 0.5467,
"step": 2710
},
{
"epoch": 0.2236106543900033,
"grad_norm": 0.45076557993888855,
"learning_rate": 2e-05,
"loss": 0.5495,
"step": 2720
},
{
"epoch": 0.2244327523840842,
"grad_norm": 0.3377240002155304,
"learning_rate": 2e-05,
"loss": 0.5301,
"step": 2730
},
{
"epoch": 0.2252548503781651,
"grad_norm": 0.36233970522880554,
"learning_rate": 2e-05,
"loss": 0.5674,
"step": 2740
},
{
"epoch": 0.22607694837224598,
"grad_norm": 0.38470005989074707,
"learning_rate": 2e-05,
"loss": 0.5442,
"step": 2750
},
{
"epoch": 0.22689904636632688,
"grad_norm": 0.4208364486694336,
"learning_rate": 2e-05,
"loss": 0.5489,
"step": 2760
},
{
"epoch": 0.22772114436040777,
"grad_norm": 0.38245609402656555,
"learning_rate": 2e-05,
"loss": 0.5351,
"step": 2770
},
{
"epoch": 0.22854324235448867,
"grad_norm": 0.3385438919067383,
"learning_rate": 2e-05,
"loss": 0.5315,
"step": 2780
},
{
"epoch": 0.22936534034856956,
"grad_norm": 0.36279070377349854,
"learning_rate": 2e-05,
"loss": 0.5252,
"step": 2790
},
{
"epoch": 0.23018743834265046,
"grad_norm": 0.3103238344192505,
"learning_rate": 2e-05,
"loss": 0.5225,
"step": 2800
},
{
"epoch": 0.23100953633673132,
"grad_norm": 0.37252289056777954,
"learning_rate": 2e-05,
"loss": 0.511,
"step": 2810
},
{
"epoch": 0.23183163433081222,
"grad_norm": 0.38126927614212036,
"learning_rate": 2e-05,
"loss": 0.5819,
"step": 2820
},
{
"epoch": 0.23265373232489311,
"grad_norm": 0.34690383076667786,
"learning_rate": 2e-05,
"loss": 0.5587,
"step": 2830
},
{
"epoch": 0.233475830318974,
"grad_norm": 0.43841126561164856,
"learning_rate": 2e-05,
"loss": 0.5591,
"step": 2840
},
{
"epoch": 0.2342979283130549,
"grad_norm": 0.3482837975025177,
"learning_rate": 2e-05,
"loss": 0.5495,
"step": 2850
},
{
"epoch": 0.2351200263071358,
"grad_norm": 0.34099289774894714,
"learning_rate": 2e-05,
"loss": 0.5371,
"step": 2860
},
{
"epoch": 0.2359421243012167,
"grad_norm": 0.3340098559856415,
"learning_rate": 2e-05,
"loss": 0.5801,
"step": 2870
},
{
"epoch": 0.2367642222952976,
"grad_norm": 0.3647443354129791,
"learning_rate": 2e-05,
"loss": 0.5344,
"step": 2880
},
{
"epoch": 0.23758632028937848,
"grad_norm": 0.34515708684921265,
"learning_rate": 2e-05,
"loss": 0.5279,
"step": 2890
},
{
"epoch": 0.23840841828345938,
"grad_norm": 0.31294670701026917,
"learning_rate": 2e-05,
"loss": 0.5585,
"step": 2900
},
{
"epoch": 0.23923051627754027,
"grad_norm": 0.32953768968582153,
"learning_rate": 2e-05,
"loss": 0.5323,
"step": 2910
},
{
"epoch": 0.24005261427162117,
"grad_norm": 0.35687941312789917,
"learning_rate": 2e-05,
"loss": 0.5396,
"step": 2920
},
{
"epoch": 0.24087471226570206,
"grad_norm": 0.35154175758361816,
"learning_rate": 2e-05,
"loss": 0.5532,
"step": 2930
},
{
"epoch": 0.24169681025978296,
"grad_norm": 0.3024180829524994,
"learning_rate": 2e-05,
"loss": 0.5115,
"step": 2940
},
{
"epoch": 0.24251890825386385,
"grad_norm": 0.37728214263916016,
"learning_rate": 2e-05,
"loss": 0.5792,
"step": 2950
},
{
"epoch": 0.24334100624794475,
"grad_norm": 0.2942720949649811,
"learning_rate": 2e-05,
"loss": 0.5171,
"step": 2960
},
{
"epoch": 0.24416310424202564,
"grad_norm": 0.29982566833496094,
"learning_rate": 2e-05,
"loss": 0.5597,
"step": 2970
},
{
"epoch": 0.24498520223610654,
"grad_norm": 0.506549596786499,
"learning_rate": 2e-05,
"loss": 0.5301,
"step": 2980
},
{
"epoch": 0.24580730023018743,
"grad_norm": 0.3588046431541443,
"learning_rate": 2e-05,
"loss": 0.5574,
"step": 2990
},
{
"epoch": 0.24662939822426833,
"grad_norm": 0.32193976640701294,
"learning_rate": 2e-05,
"loss": 0.5524,
"step": 3000
},
{
"epoch": 0.24745149621834922,
"grad_norm": 0.386301726102829,
"learning_rate": 2e-05,
"loss": 0.5558,
"step": 3010
},
{
"epoch": 0.24827359421243012,
"grad_norm": 0.33212605118751526,
"learning_rate": 2e-05,
"loss": 0.5639,
"step": 3020
},
{
"epoch": 0.249095692206511,
"grad_norm": 0.3520735204219818,
"learning_rate": 2e-05,
"loss": 0.5247,
"step": 3030
},
{
"epoch": 0.2499177902005919,
"grad_norm": 0.4220942258834839,
"learning_rate": 2e-05,
"loss": 0.5788,
"step": 3040
},
{
"epoch": 0.2507398881946728,
"grad_norm": 0.3613930344581604,
"learning_rate": 2e-05,
"loss": 0.5356,
"step": 3050
},
{
"epoch": 0.2515619861887537,
"grad_norm": 0.35459303855895996,
"learning_rate": 2e-05,
"loss": 0.5566,
"step": 3060
},
{
"epoch": 0.2523840841828346,
"grad_norm": 0.36353611946105957,
"learning_rate": 2e-05,
"loss": 0.5542,
"step": 3070
},
{
"epoch": 0.2532061821769155,
"grad_norm": 0.35317590832710266,
"learning_rate": 2e-05,
"loss": 0.5784,
"step": 3080
},
{
"epoch": 0.2540282801709964,
"grad_norm": 0.4215488135814667,
"learning_rate": 2e-05,
"loss": 0.5443,
"step": 3090
},
{
"epoch": 0.2548503781650773,
"grad_norm": 0.32683974504470825,
"learning_rate": 2e-05,
"loss": 0.543,
"step": 3100
},
{
"epoch": 0.25567247615915817,
"grad_norm": 0.3655347228050232,
"learning_rate": 2e-05,
"loss": 0.5361,
"step": 3110
},
{
"epoch": 0.25649457415323906,
"grad_norm": 0.37827545404434204,
"learning_rate": 2e-05,
"loss": 0.5533,
"step": 3120
},
{
"epoch": 0.25731667214731996,
"grad_norm": 0.34333717823028564,
"learning_rate": 2e-05,
"loss": 0.5558,
"step": 3130
},
{
"epoch": 0.25813877014140085,
"grad_norm": 0.3396783471107483,
"learning_rate": 2e-05,
"loss": 0.5436,
"step": 3140
},
{
"epoch": 0.25896086813548175,
"grad_norm": 0.37508583068847656,
"learning_rate": 2e-05,
"loss": 0.5169,
"step": 3150
},
{
"epoch": 0.25978296612956264,
"grad_norm": 0.3585703372955322,
"learning_rate": 2e-05,
"loss": 0.5422,
"step": 3160
},
{
"epoch": 0.26060506412364354,
"grad_norm": 0.3634900450706482,
"learning_rate": 2e-05,
"loss": 0.5459,
"step": 3170
},
{
"epoch": 0.26142716211772443,
"grad_norm": 0.34407326579093933,
"learning_rate": 2e-05,
"loss": 0.5208,
"step": 3180
},
{
"epoch": 0.2622492601118053,
"grad_norm": 0.41253870725631714,
"learning_rate": 2e-05,
"loss": 0.5141,
"step": 3190
},
{
"epoch": 0.2630713581058862,
"grad_norm": 0.3736776113510132,
"learning_rate": 2e-05,
"loss": 0.5594,
"step": 3200
},
{
"epoch": 0.2638934560999671,
"grad_norm": 0.3139011859893799,
"learning_rate": 2e-05,
"loss": 0.5385,
"step": 3210
},
{
"epoch": 0.264715554094048,
"grad_norm": 0.3514135777950287,
"learning_rate": 2e-05,
"loss": 0.5434,
"step": 3220
},
{
"epoch": 0.2655376520881289,
"grad_norm": 0.38396161794662476,
"learning_rate": 2e-05,
"loss": 0.521,
"step": 3230
},
{
"epoch": 0.2663597500822098,
"grad_norm": 0.32594597339630127,
"learning_rate": 2e-05,
"loss": 0.5404,
"step": 3240
},
{
"epoch": 0.2671818480762907,
"grad_norm": 0.3474389612674713,
"learning_rate": 2e-05,
"loss": 0.5434,
"step": 3250
},
{
"epoch": 0.2680039460703716,
"grad_norm": 0.4189068675041199,
"learning_rate": 2e-05,
"loss": 0.549,
"step": 3260
},
{
"epoch": 0.2688260440644525,
"grad_norm": 0.5115881562232971,
"learning_rate": 2e-05,
"loss": 0.538,
"step": 3270
},
{
"epoch": 0.2696481420585334,
"grad_norm": 0.31304091215133667,
"learning_rate": 2e-05,
"loss": 0.5662,
"step": 3280
},
{
"epoch": 0.2704702400526143,
"grad_norm": 0.2987343370914459,
"learning_rate": 2e-05,
"loss": 0.5443,
"step": 3290
},
{
"epoch": 0.27129233804669517,
"grad_norm": 0.33361944556236267,
"learning_rate": 2e-05,
"loss": 0.5654,
"step": 3300
},
{
"epoch": 0.27211443604077606,
"grad_norm": 0.330786794424057,
"learning_rate": 2e-05,
"loss": 0.5063,
"step": 3310
},
{
"epoch": 0.27293653403485696,
"grad_norm": 0.33479249477386475,
"learning_rate": 2e-05,
"loss": 0.5536,
"step": 3320
},
{
"epoch": 0.27375863202893785,
"grad_norm": 0.3566909432411194,
"learning_rate": 2e-05,
"loss": 0.5505,
"step": 3330
},
{
"epoch": 0.27458073002301875,
"grad_norm": 0.3387537896633148,
"learning_rate": 2e-05,
"loss": 0.5342,
"step": 3340
},
{
"epoch": 0.27540282801709964,
"grad_norm": 0.36577555537223816,
"learning_rate": 2e-05,
"loss": 0.5352,
"step": 3350
},
{
"epoch": 0.27622492601118054,
"grad_norm": 0.35958272218704224,
"learning_rate": 2e-05,
"loss": 0.5746,
"step": 3360
},
{
"epoch": 0.27704702400526143,
"grad_norm": 0.391726553440094,
"learning_rate": 2e-05,
"loss": 0.5379,
"step": 3370
},
{
"epoch": 0.27786912199934233,
"grad_norm": 0.3955790400505066,
"learning_rate": 2e-05,
"loss": 0.5225,
"step": 3380
},
{
"epoch": 0.2786912199934232,
"grad_norm": 0.36854982376098633,
"learning_rate": 2e-05,
"loss": 0.5418,
"step": 3390
},
{
"epoch": 0.2795133179875041,
"grad_norm": 0.366266131401062,
"learning_rate": 2e-05,
"loss": 0.5826,
"step": 3400
},
{
"epoch": 0.280335415981585,
"grad_norm": 0.4051331579685211,
"learning_rate": 2e-05,
"loss": 0.5352,
"step": 3410
},
{
"epoch": 0.2811575139756659,
"grad_norm": 0.3959217369556427,
"learning_rate": 2e-05,
"loss": 0.5248,
"step": 3420
},
{
"epoch": 0.2819796119697468,
"grad_norm": 0.3575945496559143,
"learning_rate": 2e-05,
"loss": 0.5547,
"step": 3430
},
{
"epoch": 0.2828017099638277,
"grad_norm": 0.35974130034446716,
"learning_rate": 2e-05,
"loss": 0.5492,
"step": 3440
},
{
"epoch": 0.2836238079579086,
"grad_norm": 0.31665098667144775,
"learning_rate": 2e-05,
"loss": 0.5389,
"step": 3450
},
{
"epoch": 0.2844459059519895,
"grad_norm": 0.3459745943546295,
"learning_rate": 2e-05,
"loss": 0.5666,
"step": 3460
},
{
"epoch": 0.2852680039460704,
"grad_norm": 0.47385168075561523,
"learning_rate": 2e-05,
"loss": 0.5415,
"step": 3470
},
{
"epoch": 0.2860901019401513,
"grad_norm": 0.29525819420814514,
"learning_rate": 2e-05,
"loss": 0.5201,
"step": 3480
},
{
"epoch": 0.28691219993423217,
"grad_norm": 0.32592856884002686,
"learning_rate": 2e-05,
"loss": 0.535,
"step": 3490
},
{
"epoch": 0.28773429792831307,
"grad_norm": 0.3753909766674042,
"learning_rate": 2e-05,
"loss": 0.5414,
"step": 3500
},
{
"epoch": 0.28855639592239396,
"grad_norm": 0.35471951961517334,
"learning_rate": 2e-05,
"loss": 0.5548,
"step": 3510
},
{
"epoch": 0.28937849391647485,
"grad_norm": 0.31236377358436584,
"learning_rate": 2e-05,
"loss": 0.5608,
"step": 3520
},
{
"epoch": 0.29020059191055575,
"grad_norm": 0.3410734534263611,
"learning_rate": 2e-05,
"loss": 0.5344,
"step": 3530
},
{
"epoch": 0.29102268990463664,
"grad_norm": 0.362771213054657,
"learning_rate": 2e-05,
"loss": 0.5247,
"step": 3540
},
{
"epoch": 0.29184478789871754,
"grad_norm": 0.36238428950309753,
"learning_rate": 2e-05,
"loss": 0.5464,
"step": 3550
},
{
"epoch": 0.29266688589279843,
"grad_norm": 0.38895243406295776,
"learning_rate": 2e-05,
"loss": 0.519,
"step": 3560
},
{
"epoch": 0.29348898388687933,
"grad_norm": 0.32610785961151123,
"learning_rate": 2e-05,
"loss": 0.5282,
"step": 3570
},
{
"epoch": 0.2943110818809602,
"grad_norm": 0.4277496337890625,
"learning_rate": 2e-05,
"loss": 0.5356,
"step": 3580
},
{
"epoch": 0.2951331798750411,
"grad_norm": 0.35077282786369324,
"learning_rate": 2e-05,
"loss": 0.5421,
"step": 3590
},
{
"epoch": 0.295955277869122,
"grad_norm": 0.3360082507133484,
"learning_rate": 2e-05,
"loss": 0.5236,
"step": 3600
},
{
"epoch": 0.2967773758632029,
"grad_norm": 0.41645798087120056,
"learning_rate": 2e-05,
"loss": 0.5403,
"step": 3610
},
{
"epoch": 0.2975994738572838,
"grad_norm": 0.3737342953681946,
"learning_rate": 2e-05,
"loss": 0.5266,
"step": 3620
},
{
"epoch": 0.2984215718513647,
"grad_norm": 0.45222207903862,
"learning_rate": 2e-05,
"loss": 0.5524,
"step": 3630
},
{
"epoch": 0.2992436698454456,
"grad_norm": 0.3541415333747864,
"learning_rate": 2e-05,
"loss": 0.5506,
"step": 3640
},
{
"epoch": 0.3000657678395265,
"grad_norm": 0.3760773837566376,
"learning_rate": 2e-05,
"loss": 0.5374,
"step": 3650
},
{
"epoch": 0.3008878658336074,
"grad_norm": 0.374584436416626,
"learning_rate": 2e-05,
"loss": 0.5814,
"step": 3660
},
{
"epoch": 0.3017099638276883,
"grad_norm": 0.35466334223747253,
"learning_rate": 2e-05,
"loss": 0.5587,
"step": 3670
},
{
"epoch": 0.30253206182176917,
"grad_norm": 0.3851785659790039,
"learning_rate": 2e-05,
"loss": 0.5636,
"step": 3680
},
{
"epoch": 0.30335415981585007,
"grad_norm": 0.3262382447719574,
"learning_rate": 2e-05,
"loss": 0.5509,
"step": 3690
},
{
"epoch": 0.30417625780993096,
"grad_norm": 0.35221946239471436,
"learning_rate": 2e-05,
"loss": 0.5395,
"step": 3700
},
{
"epoch": 0.30499835580401186,
"grad_norm": 0.33840522170066833,
"learning_rate": 2e-05,
"loss": 0.4874,
"step": 3710
},
{
"epoch": 0.30582045379809275,
"grad_norm": 0.3443030118942261,
"learning_rate": 2e-05,
"loss": 0.5604,
"step": 3720
},
{
"epoch": 0.30664255179217365,
"grad_norm": 0.34090104699134827,
"learning_rate": 2e-05,
"loss": 0.5368,
"step": 3730
},
{
"epoch": 0.30746464978625454,
"grad_norm": 0.4112664461135864,
"learning_rate": 2e-05,
"loss": 0.5413,
"step": 3740
},
{
"epoch": 0.30828674778033544,
"grad_norm": 0.36865583062171936,
"learning_rate": 2e-05,
"loss": 0.5266,
"step": 3750
},
{
"epoch": 0.30910884577441633,
"grad_norm": 0.2920708954334259,
"learning_rate": 2e-05,
"loss": 0.5206,
"step": 3760
},
{
"epoch": 0.3099309437684972,
"grad_norm": 0.3259790539741516,
"learning_rate": 2e-05,
"loss": 0.5138,
"step": 3770
},
{
"epoch": 0.3107530417625781,
"grad_norm": 0.3496236801147461,
"learning_rate": 2e-05,
"loss": 0.5521,
"step": 3780
},
{
"epoch": 0.311575139756659,
"grad_norm": 0.36148640513420105,
"learning_rate": 2e-05,
"loss": 0.5167,
"step": 3790
},
{
"epoch": 0.3123972377507399,
"grad_norm": 0.3326259255409241,
"learning_rate": 2e-05,
"loss": 0.512,
"step": 3800
},
{
"epoch": 0.3132193357448208,
"grad_norm": 0.3559156656265259,
"learning_rate": 2e-05,
"loss": 0.5413,
"step": 3810
},
{
"epoch": 0.3140414337389017,
"grad_norm": 0.3168695867061615,
"learning_rate": 2e-05,
"loss": 0.5592,
"step": 3820
},
{
"epoch": 0.3148635317329826,
"grad_norm": 0.3359394967556,
"learning_rate": 2e-05,
"loss": 0.5355,
"step": 3830
},
{
"epoch": 0.3156856297270635,
"grad_norm": 0.3557104766368866,
"learning_rate": 2e-05,
"loss": 0.5495,
"step": 3840
},
{
"epoch": 0.3165077277211444,
"grad_norm": 0.3635749816894531,
"learning_rate": 2e-05,
"loss": 0.543,
"step": 3850
},
{
"epoch": 0.3173298257152253,
"grad_norm": 0.35071372985839844,
"learning_rate": 2e-05,
"loss": 0.5198,
"step": 3860
},
{
"epoch": 0.3181519237093062,
"grad_norm": 0.38483187556266785,
"learning_rate": 2e-05,
"loss": 0.572,
"step": 3870
},
{
"epoch": 0.31897402170338707,
"grad_norm": 0.33359649777412415,
"learning_rate": 2e-05,
"loss": 0.5426,
"step": 3880
},
{
"epoch": 0.31979611969746796,
"grad_norm": 0.31382572650909424,
"learning_rate": 2e-05,
"loss": 0.5469,
"step": 3890
},
{
"epoch": 0.32061821769154886,
"grad_norm": 0.35558342933654785,
"learning_rate": 2e-05,
"loss": 0.5293,
"step": 3900
},
{
"epoch": 0.32144031568562975,
"grad_norm": 0.35809141397476196,
"learning_rate": 2e-05,
"loss": 0.5275,
"step": 3910
},
{
"epoch": 0.32226241367971065,
"grad_norm": 0.34361934661865234,
"learning_rate": 2e-05,
"loss": 0.5106,
"step": 3920
},
{
"epoch": 0.32308451167379154,
"grad_norm": 0.338867723941803,
"learning_rate": 2e-05,
"loss": 0.5193,
"step": 3930
},
{
"epoch": 0.32390660966787244,
"grad_norm": 0.34359851479530334,
"learning_rate": 2e-05,
"loss": 0.547,
"step": 3940
},
{
"epoch": 0.32472870766195333,
"grad_norm": 0.3797745406627655,
"learning_rate": 2e-05,
"loss": 0.5256,
"step": 3950
},
{
"epoch": 0.3255508056560342,
"grad_norm": 0.3485299348831177,
"learning_rate": 2e-05,
"loss": 0.5275,
"step": 3960
},
{
"epoch": 0.3263729036501151,
"grad_norm": 0.33082035183906555,
"learning_rate": 2e-05,
"loss": 0.5398,
"step": 3970
},
{
"epoch": 0.327195001644196,
"grad_norm": 0.3318018913269043,
"learning_rate": 2e-05,
"loss": 0.5226,
"step": 3980
},
{
"epoch": 0.3280170996382769,
"grad_norm": 0.3452273905277252,
"learning_rate": 2e-05,
"loss": 0.5526,
"step": 3990
},
{
"epoch": 0.3288391976323578,
"grad_norm": 0.32671037316322327,
"learning_rate": 2e-05,
"loss": 0.539,
"step": 4000
},
{
"epoch": 0.32966129562643864,
"grad_norm": 0.44799134135246277,
"learning_rate": 2e-05,
"loss": 0.5391,
"step": 4010
},
{
"epoch": 0.33048339362051954,
"grad_norm": 0.3295125663280487,
"learning_rate": 2e-05,
"loss": 0.5346,
"step": 4020
},
{
"epoch": 0.33130549161460043,
"grad_norm": 0.32488885521888733,
"learning_rate": 2e-05,
"loss": 0.531,
"step": 4030
},
{
"epoch": 0.33212758960868133,
"grad_norm": 0.3428230881690979,
"learning_rate": 2e-05,
"loss": 0.5293,
"step": 4040
},
{
"epoch": 0.3329496876027622,
"grad_norm": 0.34698739647865295,
"learning_rate": 2e-05,
"loss": 0.5167,
"step": 4050
},
{
"epoch": 0.3337717855968431,
"grad_norm": 0.47139301896095276,
"learning_rate": 2e-05,
"loss": 0.5732,
"step": 4060
},
{
"epoch": 0.334593883590924,
"grad_norm": 0.3707517683506012,
"learning_rate": 2e-05,
"loss": 0.5125,
"step": 4070
},
{
"epoch": 0.3354159815850049,
"grad_norm": 0.3662269115447998,
"learning_rate": 2e-05,
"loss": 0.5531,
"step": 4080
},
{
"epoch": 0.3362380795790858,
"grad_norm": 0.34573638439178467,
"learning_rate": 2e-05,
"loss": 0.4987,
"step": 4090
},
{
"epoch": 0.3370601775731667,
"grad_norm": 0.35743385553359985,
"learning_rate": 2e-05,
"loss": 0.544,
"step": 4100
},
{
"epoch": 0.3378822755672476,
"grad_norm": 0.3641563057899475,
"learning_rate": 2e-05,
"loss": 0.5418,
"step": 4110
},
{
"epoch": 0.3387043735613285,
"grad_norm": 0.3505076467990875,
"learning_rate": 2e-05,
"loss": 0.5492,
"step": 4120
},
{
"epoch": 0.3395264715554094,
"grad_norm": 0.35745880007743835,
"learning_rate": 2e-05,
"loss": 0.5515,
"step": 4130
},
{
"epoch": 0.3403485695494903,
"grad_norm": 0.41468533873558044,
"learning_rate": 2e-05,
"loss": 0.52,
"step": 4140
},
{
"epoch": 0.34117066754357117,
"grad_norm": 0.3421972692012787,
"learning_rate": 2e-05,
"loss": 0.5282,
"step": 4150
},
{
"epoch": 0.34199276553765207,
"grad_norm": 0.3675749897956848,
"learning_rate": 2e-05,
"loss": 0.5501,
"step": 4160
},
{
"epoch": 0.34281486353173296,
"grad_norm": 0.39348721504211426,
"learning_rate": 2e-05,
"loss": 0.5454,
"step": 4170
},
{
"epoch": 0.34363696152581386,
"grad_norm": 0.34833019971847534,
"learning_rate": 2e-05,
"loss": 0.53,
"step": 4180
},
{
"epoch": 0.34445905951989475,
"grad_norm": 0.36116981506347656,
"learning_rate": 2e-05,
"loss": 0.5231,
"step": 4190
},
{
"epoch": 0.34528115751397565,
"grad_norm": 0.3307689130306244,
"learning_rate": 2e-05,
"loss": 0.516,
"step": 4200
},
{
"epoch": 0.34610325550805654,
"grad_norm": 0.3135853409767151,
"learning_rate": 2e-05,
"loss": 0.5475,
"step": 4210
},
{
"epoch": 0.34692535350213743,
"grad_norm": 0.3765367269515991,
"learning_rate": 2e-05,
"loss": 0.5044,
"step": 4220
},
{
"epoch": 0.34774745149621833,
"grad_norm": 0.3693373501300812,
"learning_rate": 2e-05,
"loss": 0.5423,
"step": 4230
},
{
"epoch": 0.3485695494902992,
"grad_norm": 0.37614792585372925,
"learning_rate": 2e-05,
"loss": 0.5231,
"step": 4240
},
{
"epoch": 0.3493916474843801,
"grad_norm": 0.32215288281440735,
"learning_rate": 2e-05,
"loss": 0.5654,
"step": 4250
},
{
"epoch": 0.350213745478461,
"grad_norm": 0.4110385775566101,
"learning_rate": 2e-05,
"loss": 0.5266,
"step": 4260
},
{
"epoch": 0.3510358434725419,
"grad_norm": 0.34171709418296814,
"learning_rate": 2e-05,
"loss": 0.5149,
"step": 4270
},
{
"epoch": 0.3518579414666228,
"grad_norm": 0.35466277599334717,
"learning_rate": 2e-05,
"loss": 0.5064,
"step": 4280
},
{
"epoch": 0.3526800394607037,
"grad_norm": 0.3796570897102356,
"learning_rate": 2e-05,
"loss": 0.5486,
"step": 4290
},
{
"epoch": 0.3535021374547846,
"grad_norm": 0.32320407032966614,
"learning_rate": 2e-05,
"loss": 0.5647,
"step": 4300
},
{
"epoch": 0.3543242354488655,
"grad_norm": 0.3509526550769806,
"learning_rate": 2e-05,
"loss": 0.5017,
"step": 4310
},
{
"epoch": 0.3551463334429464,
"grad_norm": 0.34390702843666077,
"learning_rate": 2e-05,
"loss": 0.5591,
"step": 4320
},
{
"epoch": 0.3559684314370273,
"grad_norm": 0.3451758623123169,
"learning_rate": 2e-05,
"loss": 0.5305,
"step": 4330
},
{
"epoch": 0.35679052943110817,
"grad_norm": 0.40349626541137695,
"learning_rate": 2e-05,
"loss": 0.5279,
"step": 4340
},
{
"epoch": 0.35761262742518907,
"grad_norm": 0.32908904552459717,
"learning_rate": 2e-05,
"loss": 0.5327,
"step": 4350
},
{
"epoch": 0.35843472541926996,
"grad_norm": 0.3217417597770691,
"learning_rate": 2e-05,
"loss": 0.5414,
"step": 4360
},
{
"epoch": 0.35925682341335086,
"grad_norm": 0.37074434757232666,
"learning_rate": 2e-05,
"loss": 0.5376,
"step": 4370
},
{
"epoch": 0.36007892140743175,
"grad_norm": 0.5618411302566528,
"learning_rate": 2e-05,
"loss": 0.5039,
"step": 4380
},
{
"epoch": 0.36090101940151265,
"grad_norm": 0.3570129871368408,
"learning_rate": 2e-05,
"loss": 0.5209,
"step": 4390
},
{
"epoch": 0.36172311739559354,
"grad_norm": 0.34965282678604126,
"learning_rate": 2e-05,
"loss": 0.5073,
"step": 4400
},
{
"epoch": 0.36254521538967444,
"grad_norm": 0.3438706696033478,
"learning_rate": 2e-05,
"loss": 0.512,
"step": 4410
},
{
"epoch": 0.36336731338375533,
"grad_norm": 0.3440127670764923,
"learning_rate": 2e-05,
"loss": 0.5234,
"step": 4420
},
{
"epoch": 0.3641894113778362,
"grad_norm": 0.3571926951408386,
"learning_rate": 2e-05,
"loss": 0.5249,
"step": 4430
},
{
"epoch": 0.3650115093719171,
"grad_norm": 0.4331763684749603,
"learning_rate": 2e-05,
"loss": 0.5334,
"step": 4440
},
{
"epoch": 0.365833607365998,
"grad_norm": 0.36819732189178467,
"learning_rate": 2e-05,
"loss": 0.5538,
"step": 4450
},
{
"epoch": 0.3666557053600789,
"grad_norm": 0.36837688088417053,
"learning_rate": 2e-05,
"loss": 0.5046,
"step": 4460
},
{
"epoch": 0.3674778033541598,
"grad_norm": 0.35506975650787354,
"learning_rate": 2e-05,
"loss": 0.5339,
"step": 4470
},
{
"epoch": 0.3682999013482407,
"grad_norm": 0.3211697041988373,
"learning_rate": 2e-05,
"loss": 0.5151,
"step": 4480
},
{
"epoch": 0.3691219993423216,
"grad_norm": 0.32097750902175903,
"learning_rate": 2e-05,
"loss": 0.5478,
"step": 4490
},
{
"epoch": 0.3699440973364025,
"grad_norm": 0.35200196504592896,
"learning_rate": 2e-05,
"loss": 0.5249,
"step": 4500
},
{
"epoch": 0.3707661953304834,
"grad_norm": 0.3686334192752838,
"learning_rate": 2e-05,
"loss": 0.5445,
"step": 4510
},
{
"epoch": 0.3715882933245643,
"grad_norm": 0.3223293423652649,
"learning_rate": 2e-05,
"loss": 0.5355,
"step": 4520
},
{
"epoch": 0.3724103913186452,
"grad_norm": 0.3505318760871887,
"learning_rate": 2e-05,
"loss": 0.5244,
"step": 4530
},
{
"epoch": 0.37323248931272607,
"grad_norm": 0.37446868419647217,
"learning_rate": 2e-05,
"loss": 0.5585,
"step": 4540
},
{
"epoch": 0.37405458730680696,
"grad_norm": 0.3326321542263031,
"learning_rate": 2e-05,
"loss": 0.541,
"step": 4550
},
{
"epoch": 0.37487668530088786,
"grad_norm": 0.3477605879306793,
"learning_rate": 2e-05,
"loss": 0.5146,
"step": 4560
},
{
"epoch": 0.37569878329496875,
"grad_norm": 0.36050575971603394,
"learning_rate": 2e-05,
"loss": 0.5205,
"step": 4570
},
{
"epoch": 0.37652088128904965,
"grad_norm": 0.33603909611701965,
"learning_rate": 2e-05,
"loss": 0.5254,
"step": 4580
},
{
"epoch": 0.37734297928313054,
"grad_norm": 0.3777899742126465,
"learning_rate": 2e-05,
"loss": 0.5236,
"step": 4590
},
{
"epoch": 0.37816507727721144,
"grad_norm": 0.3817446827888489,
"learning_rate": 2e-05,
"loss": 0.5453,
"step": 4600
},
{
"epoch": 0.37898717527129233,
"grad_norm": 0.42952242493629456,
"learning_rate": 2e-05,
"loss": 0.5271,
"step": 4610
},
{
"epoch": 0.3798092732653732,
"grad_norm": 0.3556438386440277,
"learning_rate": 2e-05,
"loss": 0.5129,
"step": 4620
},
{
"epoch": 0.3806313712594541,
"grad_norm": 0.3687664866447449,
"learning_rate": 2e-05,
"loss": 0.5171,
"step": 4630
},
{
"epoch": 0.381453469253535,
"grad_norm": 0.33991408348083496,
"learning_rate": 2e-05,
"loss": 0.5303,
"step": 4640
},
{
"epoch": 0.3822755672476159,
"grad_norm": 0.3397250175476074,
"learning_rate": 2e-05,
"loss": 0.5286,
"step": 4650
},
{
"epoch": 0.3830976652416968,
"grad_norm": 0.34458306431770325,
"learning_rate": 2e-05,
"loss": 0.5626,
"step": 4660
},
{
"epoch": 0.3839197632357777,
"grad_norm": 0.32431814074516296,
"learning_rate": 2e-05,
"loss": 0.5157,
"step": 4670
},
{
"epoch": 0.3847418612298586,
"grad_norm": 0.3668116331100464,
"learning_rate": 2e-05,
"loss": 0.5457,
"step": 4680
},
{
"epoch": 0.3855639592239395,
"grad_norm": 0.36735913157463074,
"learning_rate": 2e-05,
"loss": 0.535,
"step": 4690
},
{
"epoch": 0.3863860572180204,
"grad_norm": 0.35322412848472595,
"learning_rate": 2e-05,
"loss": 0.5197,
"step": 4700
},
{
"epoch": 0.3872081552121013,
"grad_norm": 0.3773740828037262,
"learning_rate": 2e-05,
"loss": 0.5211,
"step": 4710
},
{
"epoch": 0.3880302532061822,
"grad_norm": 0.335239052772522,
"learning_rate": 2e-05,
"loss": 0.4863,
"step": 4720
},
{
"epoch": 0.38885235120026307,
"grad_norm": 0.43967121839523315,
"learning_rate": 2e-05,
"loss": 0.4973,
"step": 4730
},
{
"epoch": 0.38967444919434396,
"grad_norm": 0.33292362093925476,
"learning_rate": 2e-05,
"loss": 0.5264,
"step": 4740
},
{
"epoch": 0.39049654718842486,
"grad_norm": 0.3542301058769226,
"learning_rate": 2e-05,
"loss": 0.543,
"step": 4750
},
{
"epoch": 0.39131864518250575,
"grad_norm": 0.3485114872455597,
"learning_rate": 2e-05,
"loss": 0.5239,
"step": 4760
},
{
"epoch": 0.39214074317658665,
"grad_norm": 0.35498690605163574,
"learning_rate": 2e-05,
"loss": 0.5293,
"step": 4770
},
{
"epoch": 0.39296284117066754,
"grad_norm": 0.3507038950920105,
"learning_rate": 2e-05,
"loss": 0.5295,
"step": 4780
},
{
"epoch": 0.39378493916474844,
"grad_norm": 0.32928594946861267,
"learning_rate": 2e-05,
"loss": 0.5537,
"step": 4790
},
{
"epoch": 0.39460703715882933,
"grad_norm": 0.3561805486679077,
"learning_rate": 2e-05,
"loss": 0.521,
"step": 4800
},
{
"epoch": 0.3954291351529102,
"grad_norm": 0.3253518342971802,
"learning_rate": 2e-05,
"loss": 0.5484,
"step": 4810
},
{
"epoch": 0.3962512331469911,
"grad_norm": 0.3671138882637024,
"learning_rate": 2e-05,
"loss": 0.5726,
"step": 4820
},
{
"epoch": 0.397073331141072,
"grad_norm": 0.3315047025680542,
"learning_rate": 2e-05,
"loss": 0.5463,
"step": 4830
},
{
"epoch": 0.3978954291351529,
"grad_norm": 0.33220726251602173,
"learning_rate": 2e-05,
"loss": 0.485,
"step": 4840
},
{
"epoch": 0.3987175271292338,
"grad_norm": 0.3390117287635803,
"learning_rate": 2e-05,
"loss": 0.5407,
"step": 4850
},
{
"epoch": 0.3995396251233147,
"grad_norm": 0.34227901697158813,
"learning_rate": 2e-05,
"loss": 0.5372,
"step": 4860
},
{
"epoch": 0.4003617231173956,
"grad_norm": 0.36416247487068176,
"learning_rate": 2e-05,
"loss": 0.5212,
"step": 4870
},
{
"epoch": 0.4011838211114765,
"grad_norm": 0.31804022192955017,
"learning_rate": 2e-05,
"loss": 0.5,
"step": 4880
},
{
"epoch": 0.4020059191055574,
"grad_norm": 0.350694864988327,
"learning_rate": 2e-05,
"loss": 0.5213,
"step": 4890
},
{
"epoch": 0.4028280170996383,
"grad_norm": 0.3451733887195587,
"learning_rate": 2e-05,
"loss": 0.5365,
"step": 4900
},
{
"epoch": 0.4036501150937192,
"grad_norm": 0.3070090711116791,
"learning_rate": 2e-05,
"loss": 0.5215,
"step": 4910
},
{
"epoch": 0.40447221308780007,
"grad_norm": 0.33814436197280884,
"learning_rate": 2e-05,
"loss": 0.5272,
"step": 4920
},
{
"epoch": 0.40529431108188096,
"grad_norm": 0.3543502986431122,
"learning_rate": 2e-05,
"loss": 0.5344,
"step": 4930
},
{
"epoch": 0.40611640907596186,
"grad_norm": 0.33732470870018005,
"learning_rate": 2e-05,
"loss": 0.5044,
"step": 4940
},
{
"epoch": 0.40693850707004275,
"grad_norm": 0.34642818570137024,
"learning_rate": 2e-05,
"loss": 0.5282,
"step": 4950
},
{
"epoch": 0.40776060506412365,
"grad_norm": 0.34865498542785645,
"learning_rate": 2e-05,
"loss": 0.5013,
"step": 4960
},
{
"epoch": 0.40858270305820454,
"grad_norm": 0.34857216477394104,
"learning_rate": 2e-05,
"loss": 0.5233,
"step": 4970
},
{
"epoch": 0.40940480105228544,
"grad_norm": 0.3917066752910614,
"learning_rate": 2e-05,
"loss": 0.5359,
"step": 4980
},
{
"epoch": 0.41022689904636633,
"grad_norm": 0.4312105178833008,
"learning_rate": 2e-05,
"loss": 0.5726,
"step": 4990
},
{
"epoch": 0.41104899704044723,
"grad_norm": 0.40886908769607544,
"learning_rate": 2e-05,
"loss": 0.4993,
"step": 5000
},
{
"epoch": 0.4118710950345281,
"grad_norm": 0.7453745603561401,
"learning_rate": 2e-05,
"loss": 0.5222,
"step": 5010
},
{
"epoch": 0.412693193028609,
"grad_norm": 0.32858678698539734,
"learning_rate": 2e-05,
"loss": 0.5316,
"step": 5020
},
{
"epoch": 0.4135152910226899,
"grad_norm": 0.3431444764137268,
"learning_rate": 2e-05,
"loss": 0.5494,
"step": 5030
},
{
"epoch": 0.4143373890167708,
"grad_norm": 0.3860490024089813,
"learning_rate": 2e-05,
"loss": 0.5413,
"step": 5040
},
{
"epoch": 0.4151594870108517,
"grad_norm": 0.34055987000465393,
"learning_rate": 2e-05,
"loss": 0.5274,
"step": 5050
},
{
"epoch": 0.4159815850049326,
"grad_norm": 0.3648587763309479,
"learning_rate": 2e-05,
"loss": 0.5389,
"step": 5060
},
{
"epoch": 0.4168036829990135,
"grad_norm": 0.4195209741592407,
"learning_rate": 2e-05,
"loss": 0.5467,
"step": 5070
},
{
"epoch": 0.4176257809930944,
"grad_norm": 0.36065131425857544,
"learning_rate": 2e-05,
"loss": 0.5558,
"step": 5080
},
{
"epoch": 0.4184478789871753,
"grad_norm": 0.3311932384967804,
"learning_rate": 1.9999810479633703e-05,
"loss": 0.5286,
"step": 5090
},
{
"epoch": 0.4192699769812562,
"grad_norm": 0.37973183393478394,
"learning_rate": 1.997707677020786e-05,
"loss": 0.5855,
"step": 5100
},
{
"epoch": 0.42009207497533707,
"grad_norm": 0.4070264399051666,
"learning_rate": 1.9916538196124637e-05,
"loss": 0.5198,
"step": 5110
},
{
"epoch": 0.42091417296941797,
"grad_norm": 0.34926316142082214,
"learning_rate": 1.9818425303863402e-05,
"loss": 0.5181,
"step": 5120
},
{
"epoch": 0.42173627096349886,
"grad_norm": 0.3183152675628662,
"learning_rate": 1.9683111732582887e-05,
"loss": 0.5616,
"step": 5130
},
{
"epoch": 0.42255836895757976,
"grad_norm": 0.3100976049900055,
"learning_rate": 1.951111279120706e-05,
"loss": 0.5064,
"step": 5140
},
{
"epoch": 0.42338046695166065,
"grad_norm": 0.35486313700675964,
"learning_rate": 1.9303083495996028e-05,
"loss": 0.5181,
"step": 5150
},
{
"epoch": 0.42420256494574154,
"grad_norm": 0.33848389983177185,
"learning_rate": 1.9059816076075482e-05,
"loss": 0.5355,
"step": 5160
},
{
"epoch": 0.42502466293982244,
"grad_norm": 0.3523194193840027,
"learning_rate": 1.8782236956424142e-05,
"loss": 0.5467,
"step": 5170
},
{
"epoch": 0.42584676093390333,
"grad_norm": 0.38473793864250183,
"learning_rate": 1.8471403229808897e-05,
"loss": 0.5423,
"step": 5180
},
{
"epoch": 0.42666885892798423,
"grad_norm": 0.3128930628299713,
"learning_rate": 1.8128498631103333e-05,
"loss": 0.5271,
"step": 5190
},
{
"epoch": 0.4274909569220651,
"grad_norm": 0.3259393870830536,
"learning_rate": 1.7754829029320507e-05,
"loss": 0.5354,
"step": 5200
},
{
"epoch": 0.428313054916146,
"grad_norm": 0.3782612979412079,
"learning_rate": 1.7351817454527473e-05,
"loss": 0.5055,
"step": 5210
},
{
"epoch": 0.4291351529102269,
"grad_norm": 0.3767496943473816,
"learning_rate": 1.6920998678580345e-05,
"loss": 0.5581,
"step": 5220
},
{
"epoch": 0.4299572509043078,
"grad_norm": 0.36680933833122253,
"learning_rate": 1.6464013370317854e-05,
"loss": 0.5326,
"step": 5230
},
{
"epoch": 0.4307793488983887,
"grad_norm": 0.3721514940261841,
"learning_rate": 1.5982601847471915e-05,
"loss": 0.5325,
"step": 5240
},
{
"epoch": 0.4316014468924696,
"grad_norm": 0.3875768184661865,
"learning_rate": 1.5478597449089553e-05,
"loss": 0.5341,
"step": 5250
},
{
"epoch": 0.4324235448865505,
"grad_norm": 0.3299482464790344,
"learning_rate": 1.4953919553705725e-05,
"loss": 0.5621,
"step": 5260
},
{
"epoch": 0.4332456428806314,
"grad_norm": 0.35813918709754944,
"learning_rate": 1.441056626985559e-05,
"loss": 0.5166,
"step": 5270
},
{
"epoch": 0.4340677408747123,
"grad_norm": 0.3345077931880951,
"learning_rate": 1.3850606826762668e-05,
"loss": 0.4928,
"step": 5280
},
{
"epoch": 0.4348898388687932,
"grad_norm": 0.3183812201023102,
"learning_rate": 1.3276173694181128e-05,
"loss": 0.5266,
"step": 5290
},
{
"epoch": 0.43571193686287407,
"grad_norm": 0.38698452711105347,
"learning_rate": 1.2689454461401822e-05,
"loss": 0.533,
"step": 5300
},
{
"epoch": 0.43653403485695497,
"grad_norm": 0.33036819100379944,
"learning_rate": 1.2092683506348939e-05,
"loss": 0.5034,
"step": 5310
},
{
"epoch": 0.43735613285103586,
"grad_norm": 0.31868958473205566,
"learning_rate": 1.14881334864935e-05,
"loss": 0.5014,
"step": 5320
},
{
"epoch": 0.43817823084511676,
"grad_norm": 0.3324577510356903,
"learning_rate": 1.0878106683988481e-05,
"loss": 0.5006,
"step": 5330
},
{
"epoch": 0.43900032883919765,
"grad_norm": 0.3703984022140503,
"learning_rate": 1.026492623798551e-05,
"loss": 0.5305,
"step": 5340
},
{
"epoch": 0.43982242683327855,
"grad_norm": 0.36006996035575867,
"learning_rate": 9.650927297522743e-06,
"loss": 0.5396,
"step": 5350
},
{
"epoch": 0.44064452482735944,
"grad_norm": 0.3693901002407074,
"learning_rate": 9.038448128676012e-06,
"loss": 0.5206,
"step": 5360
},
{
"epoch": 0.44146662282144034,
"grad_norm": 0.4380963146686554,
"learning_rate": 8.42982120983949e-06,
"loss": 0.563,
"step": 5370
},
{
"epoch": 0.44228872081552123,
"grad_norm": 0.3041839003562927,
"learning_rate": 7.827364349047303e-06,
"loss": 0.5226,
"step": 5380
},
{
"epoch": 0.4431108188096021,
"grad_norm": 0.34294480085372925,
"learning_rate": 7.233371857163706e-06,
"loss": 0.5202,
"step": 5390
},
{
"epoch": 0.443932916803683,
"grad_norm": 0.3337588310241699,
"learning_rate": 6.650105810556492e-06,
"loss": 0.5735,
"step": 5400
},
{
"epoch": 0.4447550147977639,
"grad_norm": 0.3277808129787445,
"learning_rate": 6.079787436527681e-06,
"loss": 0.5081,
"step": 5410
},
{
"epoch": 0.4455771127918448,
"grad_norm": 0.399967223405838,
"learning_rate": 5.524588654307961e-06,
"loss": 0.5554,
"step": 5420
},
{
"epoch": 0.4463992107859257,
"grad_norm": 0.37750962376594543,
"learning_rate": 4.986623803828985e-06,
"loss": 0.517,
"step": 5430
},
{
"epoch": 0.4472213087800066,
"grad_norm": 0.3405945599079132,
"learning_rate": 4.467941593772423e-06,
"loss": 0.5625,
"step": 5440
},
{
"epoch": 0.4480434067740875,
"grad_norm": 0.3132422864437103,
"learning_rate": 3.9705172995597606e-06,
"loss": 0.5334,
"step": 5450
},
{
"epoch": 0.4488655047681684,
"grad_norm": 0.42094117403030396,
"learning_rate": 3.4962452409948346e-06,
"loss": 0.4846,
"step": 5460
},
{
"epoch": 0.4496876027622493,
"grad_norm": 0.3482487201690674,
"learning_rate": 3.046931568206211e-06,
"loss": 0.5184,
"step": 5470
},
{
"epoch": 0.4505097007563302,
"grad_norm": 0.3529168665409088,
"learning_rate": 2.624287383362404e-06,
"loss": 0.5302,
"step": 5480
},
{
"epoch": 0.4513317987504111,
"grad_norm": 0.3500068783760071,
"learning_rate": 2.2299222243541845e-06,
"loss": 0.5162,
"step": 5490
},
{
"epoch": 0.45215389674449197,
"grad_norm": 0.3623697757720947,
"learning_rate": 1.8653379352598246e-06,
"loss": 0.5366,
"step": 5500
},
{
"epoch": 0.45297599473857286,
"grad_norm": 0.31660425662994385,
"learning_rate": 1.5319229469361075e-06,
"loss": 0.5572,
"step": 5510
},
{
"epoch": 0.45379809273265376,
"grad_norm": 0.3756289780139923,
"learning_rate": 1.2309469895160917e-06,
"loss": 0.5115,
"step": 5520
},
{
"epoch": 0.45462019072673465,
"grad_norm": 0.4156130254268646,
"learning_rate": 9.635562569497974e-07,
"loss": 0.5237,
"step": 5530
},
{
"epoch": 0.45544228872081555,
"grad_norm": 0.3490898013114929,
"learning_rate": 7.307690420024867e-07,
"loss": 0.5342,
"step": 5540
},
{
"epoch": 0.45626438671489644,
"grad_norm": 0.3284478485584259,
"learning_rate": 5.334718583335942e-07,
"loss": 0.5213,
"step": 5550
},
{
"epoch": 0.45708648470897734,
"grad_norm": 0.36004093289375305,
"learning_rate": 3.7241606442442763e-07,
"loss": 0.536,
"step": 5560
},
{
"epoch": 0.45790858270305823,
"grad_norm": 0.3448970913887024,
"learning_rate": 2.482150022115879e-07,
"loss": 0.5099,
"step": 5570
},
{
"epoch": 0.4587306806971391,
"grad_norm": 0.34658676385879517,
"learning_rate": 1.6134166132293174e-07,
"loss": 0.5204,
"step": 5580
},
{
"epoch": 0.45955277869122,
"grad_norm": 0.3773450553417206,
"learning_rate": 1.1212687781127357e-07,
"loss": 0.5121,
"step": 5590
},
{
"epoch": 0.46012824728707663,
"step": 5597,
"total_flos": 2.4985159616803373e+19,
"train_loss": 0.04806772112441698,
"train_runtime": 5679.2258,
"train_samples_per_second": 11.826,
"train_steps_per_second": 0.986
}
],
"logging_steps": 10,
"max_steps": 5597,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 636,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.4985159616803373e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}