terry69's picture
Model save
040e261 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2032,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0004921259842519685,
"grad_norm": 23.289178620841536,
"learning_rate": 4.901960784313726e-08,
"loss": 1.3042,
"step": 1
},
{
"epoch": 0.0024606299212598425,
"grad_norm": 22.578339014503392,
"learning_rate": 2.4509803921568627e-07,
"loss": 1.3271,
"step": 5
},
{
"epoch": 0.004921259842519685,
"grad_norm": 15.254145113240074,
"learning_rate": 4.901960784313725e-07,
"loss": 1.2787,
"step": 10
},
{
"epoch": 0.0073818897637795275,
"grad_norm": 11.623341812407018,
"learning_rate": 7.352941176470589e-07,
"loss": 1.1561,
"step": 15
},
{
"epoch": 0.00984251968503937,
"grad_norm": 9.64376871387551,
"learning_rate": 9.80392156862745e-07,
"loss": 1.034,
"step": 20
},
{
"epoch": 0.012303149606299213,
"grad_norm": 3.4638336177505513,
"learning_rate": 1.2254901960784314e-06,
"loss": 0.9177,
"step": 25
},
{
"epoch": 0.014763779527559055,
"grad_norm": 3.5862986630527347,
"learning_rate": 1.4705882352941177e-06,
"loss": 0.8909,
"step": 30
},
{
"epoch": 0.0172244094488189,
"grad_norm": 2.9675977602993884,
"learning_rate": 1.715686274509804e-06,
"loss": 0.8673,
"step": 35
},
{
"epoch": 0.01968503937007874,
"grad_norm": 3.0016020610328322,
"learning_rate": 1.96078431372549e-06,
"loss": 0.8523,
"step": 40
},
{
"epoch": 0.02214566929133858,
"grad_norm": 3.044795051743997,
"learning_rate": 2.2058823529411767e-06,
"loss": 0.836,
"step": 45
},
{
"epoch": 0.024606299212598427,
"grad_norm": 2.9272556512306047,
"learning_rate": 2.450980392156863e-06,
"loss": 0.8255,
"step": 50
},
{
"epoch": 0.027066929133858268,
"grad_norm": 2.9326611094737216,
"learning_rate": 2.696078431372549e-06,
"loss": 0.8163,
"step": 55
},
{
"epoch": 0.02952755905511811,
"grad_norm": 3.1751190158414553,
"learning_rate": 2.9411764705882355e-06,
"loss": 0.8098,
"step": 60
},
{
"epoch": 0.03198818897637795,
"grad_norm": 3.267001091193129,
"learning_rate": 3.1862745098039216e-06,
"loss": 0.7963,
"step": 65
},
{
"epoch": 0.0344488188976378,
"grad_norm": 3.2037792452122074,
"learning_rate": 3.431372549019608e-06,
"loss": 0.7904,
"step": 70
},
{
"epoch": 0.036909448818897635,
"grad_norm": 3.1335298386999817,
"learning_rate": 3.6764705882352946e-06,
"loss": 0.7767,
"step": 75
},
{
"epoch": 0.03937007874015748,
"grad_norm": 3.0748371474571066,
"learning_rate": 3.92156862745098e-06,
"loss": 0.7566,
"step": 80
},
{
"epoch": 0.041830708661417325,
"grad_norm": 3.168808058148803,
"learning_rate": 4.166666666666667e-06,
"loss": 0.7642,
"step": 85
},
{
"epoch": 0.04429133858267716,
"grad_norm": 3.099061017093863,
"learning_rate": 4.411764705882353e-06,
"loss": 0.7567,
"step": 90
},
{
"epoch": 0.04675196850393701,
"grad_norm": 3.1397025998475545,
"learning_rate": 4.65686274509804e-06,
"loss": 0.7489,
"step": 95
},
{
"epoch": 0.04921259842519685,
"grad_norm": 3.267009036532827,
"learning_rate": 4.901960784313726e-06,
"loss": 0.7428,
"step": 100
},
{
"epoch": 0.05167322834645669,
"grad_norm": 3.0814329808676137,
"learning_rate": 5.147058823529411e-06,
"loss": 0.739,
"step": 105
},
{
"epoch": 0.054133858267716536,
"grad_norm": 2.922079677455855,
"learning_rate": 5.392156862745098e-06,
"loss": 0.7285,
"step": 110
},
{
"epoch": 0.056594488188976375,
"grad_norm": 3.1140432595060723,
"learning_rate": 5.637254901960784e-06,
"loss": 0.7279,
"step": 115
},
{
"epoch": 0.05905511811023622,
"grad_norm": 2.890134615014165,
"learning_rate": 5.882352941176471e-06,
"loss": 0.7061,
"step": 120
},
{
"epoch": 0.061515748031496065,
"grad_norm": 3.003266954139478,
"learning_rate": 6.1274509803921575e-06,
"loss": 0.7124,
"step": 125
},
{
"epoch": 0.0639763779527559,
"grad_norm": 3.490944793138708,
"learning_rate": 6.372549019607843e-06,
"loss": 0.7,
"step": 130
},
{
"epoch": 0.06643700787401575,
"grad_norm": 2.957015579169823,
"learning_rate": 6.61764705882353e-06,
"loss": 0.7109,
"step": 135
},
{
"epoch": 0.0688976377952756,
"grad_norm": 2.9802255807893343,
"learning_rate": 6.862745098039216e-06,
"loss": 0.7307,
"step": 140
},
{
"epoch": 0.07135826771653543,
"grad_norm": 2.979885982433863,
"learning_rate": 7.107843137254903e-06,
"loss": 0.6999,
"step": 145
},
{
"epoch": 0.07381889763779527,
"grad_norm": 3.4724625938807483,
"learning_rate": 7.352941176470589e-06,
"loss": 0.7018,
"step": 150
},
{
"epoch": 0.07627952755905512,
"grad_norm": 2.8756072791555822,
"learning_rate": 7.598039215686275e-06,
"loss": 0.6946,
"step": 155
},
{
"epoch": 0.07874015748031496,
"grad_norm": 2.888280542800706,
"learning_rate": 7.84313725490196e-06,
"loss": 0.6953,
"step": 160
},
{
"epoch": 0.0812007874015748,
"grad_norm": 3.0409331177011176,
"learning_rate": 8.088235294117648e-06,
"loss": 0.693,
"step": 165
},
{
"epoch": 0.08366141732283465,
"grad_norm": 2.905841883072287,
"learning_rate": 8.333333333333334e-06,
"loss": 0.6995,
"step": 170
},
{
"epoch": 0.08612204724409449,
"grad_norm": 2.8022565504484476,
"learning_rate": 8.57843137254902e-06,
"loss": 0.6941,
"step": 175
},
{
"epoch": 0.08858267716535433,
"grad_norm": 2.7594498957096523,
"learning_rate": 8.823529411764707e-06,
"loss": 0.6852,
"step": 180
},
{
"epoch": 0.09104330708661418,
"grad_norm": 2.928739510451242,
"learning_rate": 9.068627450980392e-06,
"loss": 0.6849,
"step": 185
},
{
"epoch": 0.09350393700787402,
"grad_norm": 2.922241571597126,
"learning_rate": 9.31372549019608e-06,
"loss": 0.6961,
"step": 190
},
{
"epoch": 0.09596456692913385,
"grad_norm": 2.733962541795604,
"learning_rate": 9.558823529411766e-06,
"loss": 0.6964,
"step": 195
},
{
"epoch": 0.0984251968503937,
"grad_norm": 3.3436619097327585,
"learning_rate": 9.803921568627451e-06,
"loss": 0.6903,
"step": 200
},
{
"epoch": 0.10088582677165354,
"grad_norm": 2.819756078395206,
"learning_rate": 9.999992616075212e-06,
"loss": 0.6871,
"step": 205
},
{
"epoch": 0.10334645669291338,
"grad_norm": 2.8245769395745217,
"learning_rate": 9.999734180997554e-06,
"loss": 0.7001,
"step": 210
},
{
"epoch": 0.10580708661417323,
"grad_norm": 2.8711422702525207,
"learning_rate": 9.999106571489132e-06,
"loss": 0.6851,
"step": 215
},
{
"epoch": 0.10826771653543307,
"grad_norm": 3.0388643695545077,
"learning_rate": 9.998109833891883e-06,
"loss": 0.6927,
"step": 220
},
{
"epoch": 0.11072834645669291,
"grad_norm": 2.7895631269121393,
"learning_rate": 9.996744041803731e-06,
"loss": 0.6716,
"step": 225
},
{
"epoch": 0.11318897637795275,
"grad_norm": 2.686288810148222,
"learning_rate": 9.995009296073138e-06,
"loss": 0.6725,
"step": 230
},
{
"epoch": 0.1156496062992126,
"grad_norm": 2.800082841358338,
"learning_rate": 9.992905724791669e-06,
"loss": 0.6717,
"step": 235
},
{
"epoch": 0.11811023622047244,
"grad_norm": 2.566573753796988,
"learning_rate": 9.990433483284527e-06,
"loss": 0.6666,
"step": 240
},
{
"epoch": 0.12057086614173228,
"grad_norm": 2.6155377853742467,
"learning_rate": 9.987592754099086e-06,
"loss": 0.6724,
"step": 245
},
{
"epoch": 0.12303149606299213,
"grad_norm": 2.6609874435959795,
"learning_rate": 9.984383746991416e-06,
"loss": 0.6756,
"step": 250
},
{
"epoch": 0.12549212598425197,
"grad_norm": 2.651450027647874,
"learning_rate": 9.980806698910787e-06,
"loss": 0.6764,
"step": 255
},
{
"epoch": 0.1279527559055118,
"grad_norm": 2.738239456171724,
"learning_rate": 9.976861873982177e-06,
"loss": 0.6858,
"step": 260
},
{
"epoch": 0.13041338582677164,
"grad_norm": 2.5749543275223767,
"learning_rate": 9.972549563486776e-06,
"loss": 0.681,
"step": 265
},
{
"epoch": 0.1328740157480315,
"grad_norm": 2.54868336015626,
"learning_rate": 9.967870085840463e-06,
"loss": 0.6822,
"step": 270
},
{
"epoch": 0.13533464566929135,
"grad_norm": 2.348509038708982,
"learning_rate": 9.962823786570306e-06,
"loss": 0.6701,
"step": 275
},
{
"epoch": 0.1377952755905512,
"grad_norm": 2.545785789782642,
"learning_rate": 9.95741103828905e-06,
"loss": 0.6672,
"step": 280
},
{
"epoch": 0.14025590551181102,
"grad_norm": 2.724173815762578,
"learning_rate": 9.951632240667592e-06,
"loss": 0.6659,
"step": 285
},
{
"epoch": 0.14271653543307086,
"grad_norm": 2.6494992216672197,
"learning_rate": 9.945487820405487e-06,
"loss": 0.6642,
"step": 290
},
{
"epoch": 0.1451771653543307,
"grad_norm": 2.477475682520284,
"learning_rate": 9.938978231199419e-06,
"loss": 0.6794,
"step": 295
},
{
"epoch": 0.14763779527559054,
"grad_norm": 2.589636626095672,
"learning_rate": 9.932103953709724e-06,
"loss": 0.6583,
"step": 300
},
{
"epoch": 0.1500984251968504,
"grad_norm": 2.53566392464488,
"learning_rate": 9.924865495524884e-06,
"loss": 0.6737,
"step": 305
},
{
"epoch": 0.15255905511811024,
"grad_norm": 3.013667820537959,
"learning_rate": 9.917263391124046e-06,
"loss": 0.6475,
"step": 310
},
{
"epoch": 0.15501968503937008,
"grad_norm": 2.5986666259102837,
"learning_rate": 9.90929820183757e-06,
"loss": 0.6496,
"step": 315
},
{
"epoch": 0.15748031496062992,
"grad_norm": 3.1918308775019715,
"learning_rate": 9.900970515805564e-06,
"loss": 0.6527,
"step": 320
},
{
"epoch": 0.15994094488188976,
"grad_norm": 2.58199571141749,
"learning_rate": 9.892280947934472e-06,
"loss": 0.6507,
"step": 325
},
{
"epoch": 0.1624015748031496,
"grad_norm": 2.5096275929429828,
"learning_rate": 9.883230139851656e-06,
"loss": 0.6493,
"step": 330
},
{
"epoch": 0.16486220472440946,
"grad_norm": 2.543717236052826,
"learning_rate": 9.873818759858034e-06,
"loss": 0.6398,
"step": 335
},
{
"epoch": 0.1673228346456693,
"grad_norm": 2.3455715291925987,
"learning_rate": 9.864047502878717e-06,
"loss": 0.6509,
"step": 340
},
{
"epoch": 0.16978346456692914,
"grad_norm": 2.7443847653291606,
"learning_rate": 9.853917090411713e-06,
"loss": 0.6587,
"step": 345
},
{
"epoch": 0.17224409448818898,
"grad_norm": 2.574575364550669,
"learning_rate": 9.84342827047464e-06,
"loss": 0.6351,
"step": 350
},
{
"epoch": 0.1747047244094488,
"grad_norm": 2.555747928874768,
"learning_rate": 9.832581817549497e-06,
"loss": 0.6223,
"step": 355
},
{
"epoch": 0.17716535433070865,
"grad_norm": 2.3868178771545834,
"learning_rate": 9.821378532525479e-06,
"loss": 0.6423,
"step": 360
},
{
"epoch": 0.1796259842519685,
"grad_norm": 2.3906560211167434,
"learning_rate": 9.809819242639841e-06,
"loss": 0.6203,
"step": 365
},
{
"epoch": 0.18208661417322836,
"grad_norm": 2.613672178779019,
"learning_rate": 9.79790480141681e-06,
"loss": 0.6397,
"step": 370
},
{
"epoch": 0.1845472440944882,
"grad_norm": 2.704435062795677,
"learning_rate": 9.785636088604571e-06,
"loss": 0.6401,
"step": 375
},
{
"epoch": 0.18700787401574803,
"grad_norm": 2.4110784601265935,
"learning_rate": 9.773014010110298e-06,
"loss": 0.6506,
"step": 380
},
{
"epoch": 0.18946850393700787,
"grad_norm": 2.3808900476483434,
"learning_rate": 9.760039497933266e-06,
"loss": 0.6427,
"step": 385
},
{
"epoch": 0.1919291338582677,
"grad_norm": 2.443382230868993,
"learning_rate": 9.74671351009604e-06,
"loss": 0.6536,
"step": 390
},
{
"epoch": 0.19438976377952755,
"grad_norm": 2.53346201637983,
"learning_rate": 9.733037030573725e-06,
"loss": 0.629,
"step": 395
},
{
"epoch": 0.1968503937007874,
"grad_norm": 2.7306008040110776,
"learning_rate": 9.719011069221316e-06,
"loss": 0.6395,
"step": 400
},
{
"epoch": 0.19931102362204725,
"grad_norm": 2.5885107627509316,
"learning_rate": 9.704636661699133e-06,
"loss": 0.6142,
"step": 405
},
{
"epoch": 0.2017716535433071,
"grad_norm": 2.5758279906754993,
"learning_rate": 9.68991486939635e-06,
"loss": 0.6332,
"step": 410
},
{
"epoch": 0.20423228346456693,
"grad_norm": 2.527125806186406,
"learning_rate": 9.674846779352613e-06,
"loss": 0.6155,
"step": 415
},
{
"epoch": 0.20669291338582677,
"grad_norm": 2.352822836622023,
"learning_rate": 9.659433504177786e-06,
"loss": 0.6108,
"step": 420
},
{
"epoch": 0.2091535433070866,
"grad_norm": 2.428790076631395,
"learning_rate": 9.643676181969792e-06,
"loss": 0.6204,
"step": 425
},
{
"epoch": 0.21161417322834647,
"grad_norm": 2.5614397854394055,
"learning_rate": 9.62757597623058e-06,
"loss": 0.6088,
"step": 430
},
{
"epoch": 0.2140748031496063,
"grad_norm": 2.4898996021687227,
"learning_rate": 9.611134075780209e-06,
"loss": 0.6211,
"step": 435
},
{
"epoch": 0.21653543307086615,
"grad_norm": 2.4088704164520927,
"learning_rate": 9.59435169466907e-06,
"loss": 0.6242,
"step": 440
},
{
"epoch": 0.21899606299212598,
"grad_norm": 2.5289526680832113,
"learning_rate": 9.577230072088246e-06,
"loss": 0.613,
"step": 445
},
{
"epoch": 0.22145669291338582,
"grad_norm": 2.40778858369484,
"learning_rate": 9.559770472277996e-06,
"loss": 0.6106,
"step": 450
},
{
"epoch": 0.22391732283464566,
"grad_norm": 3.00357065325349,
"learning_rate": 9.541974184434426e-06,
"loss": 0.6086,
"step": 455
},
{
"epoch": 0.2263779527559055,
"grad_norm": 2.420620535949469,
"learning_rate": 9.523842522614285e-06,
"loss": 0.604,
"step": 460
},
{
"epoch": 0.22883858267716536,
"grad_norm": 2.8983682401439936,
"learning_rate": 9.505376825637933e-06,
"loss": 0.5962,
"step": 465
},
{
"epoch": 0.2312992125984252,
"grad_norm": 2.6058841172777334,
"learning_rate": 9.486578456990494e-06,
"loss": 0.6022,
"step": 470
},
{
"epoch": 0.23375984251968504,
"grad_norm": 2.420955967093701,
"learning_rate": 9.467448804721171e-06,
"loss": 0.5849,
"step": 475
},
{
"epoch": 0.23622047244094488,
"grad_norm": 2.557055641112582,
"learning_rate": 9.447989281340753e-06,
"loss": 0.6052,
"step": 480
},
{
"epoch": 0.23868110236220472,
"grad_norm": 2.6358527915325625,
"learning_rate": 9.428201323717327e-06,
"loss": 0.6066,
"step": 485
},
{
"epoch": 0.24114173228346455,
"grad_norm": 2.514164645393038,
"learning_rate": 9.408086392970167e-06,
"loss": 0.6057,
"step": 490
},
{
"epoch": 0.24360236220472442,
"grad_norm": 2.386864670354864,
"learning_rate": 9.387645974361858e-06,
"loss": 0.5921,
"step": 495
},
{
"epoch": 0.24606299212598426,
"grad_norm": 2.572659612274251,
"learning_rate": 9.36688157718862e-06,
"loss": 0.5766,
"step": 500
},
{
"epoch": 0.2485236220472441,
"grad_norm": 2.4584516295055194,
"learning_rate": 9.345794734668866e-06,
"loss": 0.5922,
"step": 505
},
{
"epoch": 0.25098425196850394,
"grad_norm": 2.27813169588504,
"learning_rate": 9.324387003829993e-06,
"loss": 0.6002,
"step": 510
},
{
"epoch": 0.2534448818897638,
"grad_norm": 2.513354918132864,
"learning_rate": 9.302659965393404e-06,
"loss": 0.5863,
"step": 515
},
{
"epoch": 0.2559055118110236,
"grad_norm": 2.57065216369518,
"learning_rate": 9.280615223657801e-06,
"loss": 0.5827,
"step": 520
},
{
"epoch": 0.25836614173228345,
"grad_norm": 2.508285807949247,
"learning_rate": 9.258254406380718e-06,
"loss": 0.6003,
"step": 525
},
{
"epoch": 0.2608267716535433,
"grad_norm": 2.393181758359751,
"learning_rate": 9.23557916465833e-06,
"loss": 0.5876,
"step": 530
},
{
"epoch": 0.2632874015748031,
"grad_norm": 2.45141675908852,
"learning_rate": 9.212591172803541e-06,
"loss": 0.5782,
"step": 535
},
{
"epoch": 0.265748031496063,
"grad_norm": 2.602808329250647,
"learning_rate": 9.189292128222355e-06,
"loss": 0.5817,
"step": 540
},
{
"epoch": 0.26820866141732286,
"grad_norm": 2.6264024238291332,
"learning_rate": 9.165683751288537e-06,
"loss": 0.5987,
"step": 545
},
{
"epoch": 0.2706692913385827,
"grad_norm": 2.5206627459795086,
"learning_rate": 9.141767785216585e-06,
"loss": 0.5773,
"step": 550
},
{
"epoch": 0.27312992125984253,
"grad_norm": 2.538664530429439,
"learning_rate": 9.117545995933015e-06,
"loss": 0.5725,
"step": 555
},
{
"epoch": 0.2755905511811024,
"grad_norm": 2.541257403182122,
"learning_rate": 9.093020171945966e-06,
"loss": 0.5667,
"step": 560
},
{
"epoch": 0.2780511811023622,
"grad_norm": 2.3883496422885937,
"learning_rate": 9.068192124213135e-06,
"loss": 0.5681,
"step": 565
},
{
"epoch": 0.28051181102362205,
"grad_norm": 2.4768611856821128,
"learning_rate": 9.043063686008066e-06,
"loss": 0.5693,
"step": 570
},
{
"epoch": 0.2829724409448819,
"grad_norm": 2.531509664449831,
"learning_rate": 9.017636712784776e-06,
"loss": 0.5712,
"step": 575
},
{
"epoch": 0.2854330708661417,
"grad_norm": 2.4497026467822134,
"learning_rate": 8.991913082040752e-06,
"loss": 0.5711,
"step": 580
},
{
"epoch": 0.28789370078740156,
"grad_norm": 2.3627002230344027,
"learning_rate": 8.96589469317832e-06,
"loss": 0.5746,
"step": 585
},
{
"epoch": 0.2903543307086614,
"grad_norm": 2.5848492576021207,
"learning_rate": 8.9395834673644e-06,
"loss": 0.5638,
"step": 590
},
{
"epoch": 0.29281496062992124,
"grad_norm": 2.7943897726154026,
"learning_rate": 8.912981347388634e-06,
"loss": 0.5716,
"step": 595
},
{
"epoch": 0.2952755905511811,
"grad_norm": 2.4233634727846547,
"learning_rate": 8.886090297519956e-06,
"loss": 0.5589,
"step": 600
},
{
"epoch": 0.29773622047244097,
"grad_norm": 2.406617785286132,
"learning_rate": 8.85891230336153e-06,
"loss": 0.5743,
"step": 605
},
{
"epoch": 0.3001968503937008,
"grad_norm": 2.375668869002511,
"learning_rate": 8.83144937170415e-06,
"loss": 0.5536,
"step": 610
},
{
"epoch": 0.30265748031496065,
"grad_norm": 2.3455722754718416,
"learning_rate": 8.803703530378059e-06,
"loss": 0.5516,
"step": 615
},
{
"epoch": 0.3051181102362205,
"grad_norm": 2.5237377829623107,
"learning_rate": 8.775676828103205e-06,
"loss": 0.5565,
"step": 620
},
{
"epoch": 0.3075787401574803,
"grad_norm": 2.287022202884204,
"learning_rate": 8.747371334337983e-06,
"loss": 0.5725,
"step": 625
},
{
"epoch": 0.31003937007874016,
"grad_norm": 2.291688895957761,
"learning_rate": 8.718789139126417e-06,
"loss": 0.545,
"step": 630
},
{
"epoch": 0.3125,
"grad_norm": 2.527855069790082,
"learning_rate": 8.689932352943837e-06,
"loss": 0.543,
"step": 635
},
{
"epoch": 0.31496062992125984,
"grad_norm": 2.3987364116312535,
"learning_rate": 8.660803106541044e-06,
"loss": 0.5475,
"step": 640
},
{
"epoch": 0.3174212598425197,
"grad_norm": 2.61276055348719,
"learning_rate": 8.631403550786979e-06,
"loss": 0.5488,
"step": 645
},
{
"epoch": 0.3198818897637795,
"grad_norm": 2.419050184531339,
"learning_rate": 8.601735856509903e-06,
"loss": 0.5476,
"step": 650
},
{
"epoch": 0.32234251968503935,
"grad_norm": 2.3768631155511537,
"learning_rate": 8.571802214337107e-06,
"loss": 0.5637,
"step": 655
},
{
"epoch": 0.3248031496062992,
"grad_norm": 2.4228616659745987,
"learning_rate": 8.541604834533159e-06,
"loss": 0.5305,
"step": 660
},
{
"epoch": 0.32726377952755903,
"grad_norm": 2.3754663831523875,
"learning_rate": 8.511145946836704e-06,
"loss": 0.5331,
"step": 665
},
{
"epoch": 0.3297244094488189,
"grad_norm": 2.3398125333991535,
"learning_rate": 8.48042780029581e-06,
"loss": 0.5357,
"step": 670
},
{
"epoch": 0.33218503937007876,
"grad_norm": 2.4983820414217663,
"learning_rate": 8.449452663101918e-06,
"loss": 0.5332,
"step": 675
},
{
"epoch": 0.3346456692913386,
"grad_norm": 2.3721219410260397,
"learning_rate": 8.418222822422348e-06,
"loss": 0.5417,
"step": 680
},
{
"epoch": 0.33710629921259844,
"grad_norm": 2.360196232033206,
"learning_rate": 8.386740584231431e-06,
"loss": 0.5461,
"step": 685
},
{
"epoch": 0.3395669291338583,
"grad_norm": 2.466049773095662,
"learning_rate": 8.355008273140222e-06,
"loss": 0.5253,
"step": 690
},
{
"epoch": 0.3420275590551181,
"grad_norm": 2.348269078433154,
"learning_rate": 8.323028232224863e-06,
"loss": 0.5304,
"step": 695
},
{
"epoch": 0.34448818897637795,
"grad_norm": 2.476352965234454,
"learning_rate": 8.290802822853576e-06,
"loss": 0.5349,
"step": 700
},
{
"epoch": 0.3469488188976378,
"grad_norm": 2.279222616967455,
"learning_rate": 8.258334424512293e-06,
"loss": 0.546,
"step": 705
},
{
"epoch": 0.3494094488188976,
"grad_norm": 2.3637014928251623,
"learning_rate": 8.22562543462897e-06,
"loss": 0.5371,
"step": 710
},
{
"epoch": 0.35187007874015747,
"grad_norm": 2.619852828253986,
"learning_rate": 8.192678268396545e-06,
"loss": 0.5237,
"step": 715
},
{
"epoch": 0.3543307086614173,
"grad_norm": 2.5571753512444912,
"learning_rate": 8.159495358594627e-06,
"loss": 0.5296,
"step": 720
},
{
"epoch": 0.35679133858267714,
"grad_norm": 2.312802552942875,
"learning_rate": 8.126079155409845e-06,
"loss": 0.5235,
"step": 725
},
{
"epoch": 0.359251968503937,
"grad_norm": 2.5951586860371374,
"learning_rate": 8.092432126254933e-06,
"loss": 0.5031,
"step": 730
},
{
"epoch": 0.3617125984251969,
"grad_norm": 2.325097293624387,
"learning_rate": 8.058556755586537e-06,
"loss": 0.5233,
"step": 735
},
{
"epoch": 0.3641732283464567,
"grad_norm": 2.2732745100546725,
"learning_rate": 8.024455544721778e-06,
"loss": 0.5247,
"step": 740
},
{
"epoch": 0.36663385826771655,
"grad_norm": 2.363542176910271,
"learning_rate": 7.990131011653545e-06,
"loss": 0.5218,
"step": 745
},
{
"epoch": 0.3690944881889764,
"grad_norm": 2.3003464358723127,
"learning_rate": 7.955585690864567e-06,
"loss": 0.5116,
"step": 750
},
{
"epoch": 0.3715551181102362,
"grad_norm": 2.3293766926582213,
"learning_rate": 7.920822133140285e-06,
"loss": 0.5263,
"step": 755
},
{
"epoch": 0.37401574803149606,
"grad_norm": 2.334188982780053,
"learning_rate": 7.88584290538049e-06,
"loss": 0.4941,
"step": 760
},
{
"epoch": 0.3764763779527559,
"grad_norm": 2.351749135985343,
"learning_rate": 7.850650590409795e-06,
"loss": 0.5116,
"step": 765
},
{
"epoch": 0.37893700787401574,
"grad_norm": 2.296567696173685,
"learning_rate": 7.815247786786919e-06,
"loss": 0.5225,
"step": 770
},
{
"epoch": 0.3813976377952756,
"grad_norm": 2.4483931002237975,
"learning_rate": 7.779637108612813e-06,
"loss": 0.5188,
"step": 775
},
{
"epoch": 0.3838582677165354,
"grad_norm": 2.372343104732536,
"learning_rate": 7.743821185337634e-06,
"loss": 0.5138,
"step": 780
},
{
"epoch": 0.38631889763779526,
"grad_norm": 2.561786660857743,
"learning_rate": 7.7078026615666e-06,
"loss": 0.5005,
"step": 785
},
{
"epoch": 0.3887795275590551,
"grad_norm": 2.2678689682364017,
"learning_rate": 7.671584196864703e-06,
"loss": 0.5144,
"step": 790
},
{
"epoch": 0.391240157480315,
"grad_norm": 2.4118009586518223,
"learning_rate": 7.635168465560343e-06,
"loss": 0.507,
"step": 795
},
{
"epoch": 0.3937007874015748,
"grad_norm": 2.3699190289261742,
"learning_rate": 7.598558156547842e-06,
"loss": 0.5058,
"step": 800
},
{
"epoch": 0.39616141732283466,
"grad_norm": 2.3834254512275894,
"learning_rate": 7.561755973088917e-06,
"loss": 0.5098,
"step": 805
},
{
"epoch": 0.3986220472440945,
"grad_norm": 2.3484210670312304,
"learning_rate": 7.52476463261306e-06,
"loss": 0.5101,
"step": 810
},
{
"epoch": 0.40108267716535434,
"grad_norm": 2.259743181885874,
"learning_rate": 7.487586866516897e-06,
"loss": 0.4957,
"step": 815
},
{
"epoch": 0.4035433070866142,
"grad_norm": 2.459196143941421,
"learning_rate": 7.450225419962498e-06,
"loss": 0.5018,
"step": 820
},
{
"epoch": 0.406003937007874,
"grad_norm": 2.3987573923118326,
"learning_rate": 7.412683051674681e-06,
"loss": 0.505,
"step": 825
},
{
"epoch": 0.40846456692913385,
"grad_norm": 2.3945193465931527,
"learning_rate": 7.374962533737306e-06,
"loss": 0.5021,
"step": 830
},
{
"epoch": 0.4109251968503937,
"grad_norm": 2.450543787256226,
"learning_rate": 7.3370666513885965e-06,
"loss": 0.5045,
"step": 835
},
{
"epoch": 0.41338582677165353,
"grad_norm": 2.3773634574014624,
"learning_rate": 7.298998202815474e-06,
"loss": 0.4843,
"step": 840
},
{
"epoch": 0.41584645669291337,
"grad_norm": 2.3218222930991312,
"learning_rate": 7.260759998946945e-06,
"loss": 0.4999,
"step": 845
},
{
"epoch": 0.4183070866141732,
"grad_norm": 2.2812480048868045,
"learning_rate": 7.2223548632465424e-06,
"loss": 0.4903,
"step": 850
},
{
"epoch": 0.42076771653543305,
"grad_norm": 2.3409310514361246,
"learning_rate": 7.183785631503851e-06,
"loss": 0.4816,
"step": 855
},
{
"epoch": 0.42322834645669294,
"grad_norm": 2.3088182610498897,
"learning_rate": 7.145055151625113e-06,
"loss": 0.4754,
"step": 860
},
{
"epoch": 0.4256889763779528,
"grad_norm": 2.3963782413431325,
"learning_rate": 7.1061662834229375e-06,
"loss": 0.4918,
"step": 865
},
{
"epoch": 0.4281496062992126,
"grad_norm": 2.375549548917402,
"learning_rate": 7.0671218984051385e-06,
"loss": 0.4797,
"step": 870
},
{
"epoch": 0.43061023622047245,
"grad_norm": 2.340240253487199,
"learning_rate": 7.0279248795627156e-06,
"loss": 0.4804,
"step": 875
},
{
"epoch": 0.4330708661417323,
"grad_norm": 2.5999252908460564,
"learning_rate": 6.988578121156956e-06,
"loss": 0.4975,
"step": 880
},
{
"epoch": 0.43553149606299213,
"grad_norm": 2.412860995827959,
"learning_rate": 6.9490845285057505e-06,
"loss": 0.4757,
"step": 885
},
{
"epoch": 0.43799212598425197,
"grad_norm": 2.3723415230376053,
"learning_rate": 6.909447017769047e-06,
"loss": 0.485,
"step": 890
},
{
"epoch": 0.4404527559055118,
"grad_norm": 2.2444505424132357,
"learning_rate": 6.869668515733536e-06,
"loss": 0.4791,
"step": 895
},
{
"epoch": 0.44291338582677164,
"grad_norm": 2.4018284029434627,
"learning_rate": 6.829751959596544e-06,
"loss": 0.4781,
"step": 900
},
{
"epoch": 0.4453740157480315,
"grad_norm": 2.2719247540147873,
"learning_rate": 6.789700296749141e-06,
"loss": 0.4664,
"step": 905
},
{
"epoch": 0.4478346456692913,
"grad_norm": 2.5647531601957296,
"learning_rate": 6.749516484558518e-06,
"loss": 0.4771,
"step": 910
},
{
"epoch": 0.45029527559055116,
"grad_norm": 2.209823332066905,
"learning_rate": 6.709203490149615e-06,
"loss": 0.4775,
"step": 915
},
{
"epoch": 0.452755905511811,
"grad_norm": 2.3889046698682734,
"learning_rate": 6.668764290186039e-06,
"loss": 0.4632,
"step": 920
},
{
"epoch": 0.4552165354330709,
"grad_norm": 2.3497617502204142,
"learning_rate": 6.628201870650262e-06,
"loss": 0.4609,
"step": 925
},
{
"epoch": 0.45767716535433073,
"grad_norm": 2.286479842578479,
"learning_rate": 6.587519226623137e-06,
"loss": 0.4584,
"step": 930
},
{
"epoch": 0.46013779527559057,
"grad_norm": 2.346679839886123,
"learning_rate": 6.546719362062763e-06,
"loss": 0.4627,
"step": 935
},
{
"epoch": 0.4625984251968504,
"grad_norm": 2.372517236495747,
"learning_rate": 6.50580528958265e-06,
"loss": 0.4627,
"step": 940
},
{
"epoch": 0.46505905511811024,
"grad_norm": 2.352894112351414,
"learning_rate": 6.464780030229297e-06,
"loss": 0.4613,
"step": 945
},
{
"epoch": 0.4675196850393701,
"grad_norm": 2.3155134840690073,
"learning_rate": 6.423646613259103e-06,
"loss": 0.4643,
"step": 950
},
{
"epoch": 0.4699803149606299,
"grad_norm": 2.2802677404420604,
"learning_rate": 6.382408075914698e-06,
"loss": 0.4675,
"step": 955
},
{
"epoch": 0.47244094488188976,
"grad_norm": 2.3261149198375577,
"learning_rate": 6.341067463200678e-06,
"loss": 0.4661,
"step": 960
},
{
"epoch": 0.4749015748031496,
"grad_norm": 2.3024057597107723,
"learning_rate": 6.299627827658757e-06,
"loss": 0.4695,
"step": 965
},
{
"epoch": 0.47736220472440943,
"grad_norm": 2.3579401662222805,
"learning_rate": 6.258092229142383e-06,
"loss": 0.4557,
"step": 970
},
{
"epoch": 0.47982283464566927,
"grad_norm": 2.4039007796045624,
"learning_rate": 6.216463734590797e-06,
"loss": 0.4733,
"step": 975
},
{
"epoch": 0.4822834645669291,
"grad_norm": 2.350383170709395,
"learning_rate": 6.174745417802563e-06,
"loss": 0.4542,
"step": 980
},
{
"epoch": 0.484744094488189,
"grad_norm": 2.3727437081826195,
"learning_rate": 6.132940359208625e-06,
"loss": 0.4599,
"step": 985
},
{
"epoch": 0.48720472440944884,
"grad_norm": 2.2638903752102877,
"learning_rate": 6.09105164564483e-06,
"loss": 0.4581,
"step": 990
},
{
"epoch": 0.4896653543307087,
"grad_norm": 2.391666707671801,
"learning_rate": 6.049082370124011e-06,
"loss": 0.4335,
"step": 995
},
{
"epoch": 0.4921259842519685,
"grad_norm": 2.285287290642185,
"learning_rate": 6.007035631607605e-06,
"loss": 0.4322,
"step": 1000
},
{
"epoch": 0.49458661417322836,
"grad_norm": 2.3527234691645886,
"learning_rate": 5.964914534776814e-06,
"loss": 0.4452,
"step": 1005
},
{
"epoch": 0.4970472440944882,
"grad_norm": 2.388047521619243,
"learning_rate": 5.9227221898033785e-06,
"loss": 0.4452,
"step": 1010
},
{
"epoch": 0.49950787401574803,
"grad_norm": 2.330957204189204,
"learning_rate": 5.880461712119913e-06,
"loss": 0.4509,
"step": 1015
},
{
"epoch": 0.5019685039370079,
"grad_norm": 2.3767662326352146,
"learning_rate": 5.838136222189874e-06,
"loss": 0.4417,
"step": 1020
},
{
"epoch": 0.5044291338582677,
"grad_norm": 2.3296456305613553,
"learning_rate": 5.795748845277143e-06,
"loss": 0.4456,
"step": 1025
},
{
"epoch": 0.5068897637795275,
"grad_norm": 2.3238395450456317,
"learning_rate": 5.75330271121526e-06,
"loss": 0.4483,
"step": 1030
},
{
"epoch": 0.5093503937007874,
"grad_norm": 2.271731748541768,
"learning_rate": 5.710800954176326e-06,
"loss": 0.4439,
"step": 1035
},
{
"epoch": 0.5118110236220472,
"grad_norm": 2.2730785805154716,
"learning_rate": 5.668246712439579e-06,
"loss": 0.438,
"step": 1040
},
{
"epoch": 0.5142716535433071,
"grad_norm": 2.2984158447258856,
"learning_rate": 5.625643128159658e-06,
"loss": 0.4371,
"step": 1045
},
{
"epoch": 0.5167322834645669,
"grad_norm": 2.138298543480362,
"learning_rate": 5.582993347134604e-06,
"loss": 0.4358,
"step": 1050
},
{
"epoch": 0.5191929133858267,
"grad_norm": 2.2656933007145326,
"learning_rate": 5.540300518573564e-06,
"loss": 0.4456,
"step": 1055
},
{
"epoch": 0.5216535433070866,
"grad_norm": 2.4687742042434437,
"learning_rate": 5.4975677948642704e-06,
"loss": 0.4292,
"step": 1060
},
{
"epoch": 0.5241141732283464,
"grad_norm": 2.2023440824897538,
"learning_rate": 5.454798331340261e-06,
"loss": 0.4448,
"step": 1065
},
{
"epoch": 0.5265748031496063,
"grad_norm": 2.3820756070238844,
"learning_rate": 5.4119952860479e-06,
"loss": 0.4322,
"step": 1070
},
{
"epoch": 0.5290354330708661,
"grad_norm": 2.307966181110738,
"learning_rate": 5.369161819513189e-06,
"loss": 0.4345,
"step": 1075
},
{
"epoch": 0.531496062992126,
"grad_norm": 2.4566374524002006,
"learning_rate": 5.3263010945083994e-06,
"loss": 0.4342,
"step": 1080
},
{
"epoch": 0.5339566929133859,
"grad_norm": 2.304141656628352,
"learning_rate": 5.283416275818531e-06,
"loss": 0.4304,
"step": 1085
},
{
"epoch": 0.5364173228346457,
"grad_norm": 2.3743587538000974,
"learning_rate": 5.240510530007641e-06,
"loss": 0.4248,
"step": 1090
},
{
"epoch": 0.5388779527559056,
"grad_norm": 2.275764809511964,
"learning_rate": 5.1975870251850105e-06,
"loss": 0.4294,
"step": 1095
},
{
"epoch": 0.5413385826771654,
"grad_norm": 2.4755101283685823,
"learning_rate": 5.1546489307712345e-06,
"loss": 0.4248,
"step": 1100
},
{
"epoch": 0.5437992125984252,
"grad_norm": 2.1985695933129406,
"learning_rate": 5.111699417264177e-06,
"loss": 0.423,
"step": 1105
},
{
"epoch": 0.5462598425196851,
"grad_norm": 2.2532688868144537,
"learning_rate": 5.06874165600488e-06,
"loss": 0.4247,
"step": 1110
},
{
"epoch": 0.5487204724409449,
"grad_norm": 2.5064255810867553,
"learning_rate": 5.025778818943391e-06,
"loss": 0.4171,
"step": 1115
},
{
"epoch": 0.5511811023622047,
"grad_norm": 2.223795201431944,
"learning_rate": 4.982814078404543e-06,
"loss": 0.4136,
"step": 1120
},
{
"epoch": 0.5536417322834646,
"grad_norm": 2.321073710905454,
"learning_rate": 4.939850606853724e-06,
"loss": 0.4128,
"step": 1125
},
{
"epoch": 0.5561023622047244,
"grad_norm": 2.352196407309558,
"learning_rate": 4.89689157666262e-06,
"loss": 0.4225,
"step": 1130
},
{
"epoch": 0.5585629921259843,
"grad_norm": 2.3165609074323386,
"learning_rate": 4.853940159874972e-06,
"loss": 0.4137,
"step": 1135
},
{
"epoch": 0.5610236220472441,
"grad_norm": 2.249863447456596,
"learning_rate": 4.8109995279723556e-06,
"loss": 0.4124,
"step": 1140
},
{
"epoch": 0.5634842519685039,
"grad_norm": 2.3324932344927927,
"learning_rate": 4.768072851640006e-06,
"loss": 0.4092,
"step": 1145
},
{
"epoch": 0.5659448818897638,
"grad_norm": 2.175622171626361,
"learning_rate": 4.7251633005326935e-06,
"loss": 0.4044,
"step": 1150
},
{
"epoch": 0.5684055118110236,
"grad_norm": 2.244989030279403,
"learning_rate": 4.682274043040682e-06,
"loss": 0.4102,
"step": 1155
},
{
"epoch": 0.5708661417322834,
"grad_norm": 2.3655854821968583,
"learning_rate": 4.639408246055781e-06,
"loss": 0.3998,
"step": 1160
},
{
"epoch": 0.5733267716535433,
"grad_norm": 2.3662328414472147,
"learning_rate": 4.596569074737501e-06,
"loss": 0.4209,
"step": 1165
},
{
"epoch": 0.5757874015748031,
"grad_norm": 2.2898685582377114,
"learning_rate": 4.5537596922793475e-06,
"loss": 0.3884,
"step": 1170
},
{
"epoch": 0.578248031496063,
"grad_norm": 2.175371768003494,
"learning_rate": 4.510983259675252e-06,
"loss": 0.413,
"step": 1175
},
{
"epoch": 0.5807086614173228,
"grad_norm": 2.174940506600783,
"learning_rate": 4.468242935486164e-06,
"loss": 0.3977,
"step": 1180
},
{
"epoch": 0.5831692913385826,
"grad_norm": 2.2645627809171205,
"learning_rate": 4.425541875606837e-06,
"loss": 0.4,
"step": 1185
},
{
"epoch": 0.5856299212598425,
"grad_norm": 2.3434052194119097,
"learning_rate": 4.38288323303279e-06,
"loss": 0.416,
"step": 1190
},
{
"epoch": 0.5880905511811023,
"grad_norm": 2.1589090436361027,
"learning_rate": 4.340270157627496e-06,
"loss": 0.3975,
"step": 1195
},
{
"epoch": 0.5905511811023622,
"grad_norm": 2.2770509922795226,
"learning_rate": 4.29770579588981e-06,
"loss": 0.4133,
"step": 1200
},
{
"epoch": 0.593011811023622,
"grad_norm": 2.133861488496685,
"learning_rate": 4.255193290721626e-06,
"loss": 0.3965,
"step": 1205
},
{
"epoch": 0.5954724409448819,
"grad_norm": 2.0592654419552474,
"learning_rate": 4.2127357811958006e-06,
"loss": 0.3944,
"step": 1210
},
{
"epoch": 0.5979330708661418,
"grad_norm": 2.3061861279285756,
"learning_rate": 4.170336402324393e-06,
"loss": 0.3989,
"step": 1215
},
{
"epoch": 0.6003937007874016,
"grad_norm": 2.1897218075138665,
"learning_rate": 4.127998284827148e-06,
"loss": 0.3922,
"step": 1220
},
{
"epoch": 0.6028543307086615,
"grad_norm": 2.273569046487088,
"learning_rate": 4.085724554900359e-06,
"loss": 0.3919,
"step": 1225
},
{
"epoch": 0.6053149606299213,
"grad_norm": 2.316647904273548,
"learning_rate": 4.043518333986009e-06,
"loss": 0.4004,
"step": 1230
},
{
"epoch": 0.6077755905511811,
"grad_norm": 2.139410821479213,
"learning_rate": 4.001382738541291e-06,
"loss": 0.3868,
"step": 1235
},
{
"epoch": 0.610236220472441,
"grad_norm": 2.4158380958610266,
"learning_rate": 3.9593208798085094e-06,
"loss": 0.3968,
"step": 1240
},
{
"epoch": 0.6126968503937008,
"grad_norm": 2.2664180097106934,
"learning_rate": 3.9173358635853285e-06,
"loss": 0.3977,
"step": 1245
},
{
"epoch": 0.6151574803149606,
"grad_norm": 2.1617748237173995,
"learning_rate": 3.875430789995454e-06,
"loss": 0.3951,
"step": 1250
},
{
"epoch": 0.6176181102362205,
"grad_norm": 2.3139618808394258,
"learning_rate": 3.833608753259729e-06,
"loss": 0.3883,
"step": 1255
},
{
"epoch": 0.6200787401574803,
"grad_norm": 2.2557729958410646,
"learning_rate": 3.791872841467643e-06,
"loss": 0.3833,
"step": 1260
},
{
"epoch": 0.6225393700787402,
"grad_norm": 2.288978630108445,
"learning_rate": 3.7502261363493346e-06,
"loss": 0.392,
"step": 1265
},
{
"epoch": 0.625,
"grad_norm": 2.400211890621822,
"learning_rate": 3.708671713048017e-06,
"loss": 0.382,
"step": 1270
},
{
"epoch": 0.6274606299212598,
"grad_norm": 2.14291748228325,
"learning_rate": 3.6672126398929273e-06,
"loss": 0.3868,
"step": 1275
},
{
"epoch": 0.6299212598425197,
"grad_norm": 2.189481811128098,
"learning_rate": 3.625851978172765e-06,
"loss": 0.3844,
"step": 1280
},
{
"epoch": 0.6323818897637795,
"grad_norm": 2.17033479220634,
"learning_rate": 3.5845927819096405e-06,
"loss": 0.3767,
"step": 1285
},
{
"epoch": 0.6348425196850394,
"grad_norm": 2.1024337319523285,
"learning_rate": 3.543438097633577e-06,
"loss": 0.385,
"step": 1290
},
{
"epoch": 0.6373031496062992,
"grad_norm": 2.1601959551273584,
"learning_rate": 3.5023909641575604e-06,
"loss": 0.3655,
"step": 1295
},
{
"epoch": 0.639763779527559,
"grad_norm": 2.2147245957921324,
"learning_rate": 3.4614544123531476e-06,
"loss": 0.3846,
"step": 1300
},
{
"epoch": 0.6422244094488189,
"grad_norm": 2.2767016812027037,
"learning_rate": 3.4206314649266813e-06,
"loss": 0.3829,
"step": 1305
},
{
"epoch": 0.6446850393700787,
"grad_norm": 2.189301400471183,
"learning_rate": 3.3799251361960883e-06,
"loss": 0.3793,
"step": 1310
},
{
"epoch": 0.6471456692913385,
"grad_norm": 2.2486478577990328,
"learning_rate": 3.339338431868311e-06,
"loss": 0.3651,
"step": 1315
},
{
"epoch": 0.6496062992125984,
"grad_norm": 2.1239701183150186,
"learning_rate": 3.29887434881737e-06,
"loss": 0.3962,
"step": 1320
},
{
"epoch": 0.6520669291338582,
"grad_norm": 2.0943809822170194,
"learning_rate": 3.2585358748630725e-06,
"loss": 0.3781,
"step": 1325
},
{
"epoch": 0.6545275590551181,
"grad_norm": 2.302998816495133,
"learning_rate": 3.2183259885504003e-06,
"loss": 0.3599,
"step": 1330
},
{
"epoch": 0.656988188976378,
"grad_norm": 2.093137463735972,
"learning_rate": 3.1782476589295803e-06,
"loss": 0.3693,
"step": 1335
},
{
"epoch": 0.6594488188976378,
"grad_norm": 2.0537589165415038,
"learning_rate": 3.138303845336844e-06,
"loss": 0.3767,
"step": 1340
},
{
"epoch": 0.6619094488188977,
"grad_norm": 2.1644873169815018,
"learning_rate": 3.098497497175925e-06,
"loss": 0.3758,
"step": 1345
},
{
"epoch": 0.6643700787401575,
"grad_norm": 2.1158460604452904,
"learning_rate": 3.0588315537002682e-06,
"loss": 0.3715,
"step": 1350
},
{
"epoch": 0.6668307086614174,
"grad_norm": 2.312489752052895,
"learning_rate": 3.0193089437960043e-06,
"loss": 0.3764,
"step": 1355
},
{
"epoch": 0.6692913385826772,
"grad_norm": 2.4373363981556686,
"learning_rate": 2.9799325857656856e-06,
"loss": 0.3639,
"step": 1360
},
{
"epoch": 0.671751968503937,
"grad_norm": 2.051650821654651,
"learning_rate": 2.940705387112798e-06,
"loss": 0.3456,
"step": 1365
},
{
"epoch": 0.6742125984251969,
"grad_norm": 2.162923091354617,
"learning_rate": 2.901630244327075e-06,
"loss": 0.3737,
"step": 1370
},
{
"epoch": 0.6766732283464567,
"grad_norm": 2.1826289176609617,
"learning_rate": 2.862710042670629e-06,
"loss": 0.3777,
"step": 1375
},
{
"epoch": 0.6791338582677166,
"grad_norm": 2.271492512262752,
"learning_rate": 2.8239476559649013e-06,
"loss": 0.3672,
"step": 1380
},
{
"epoch": 0.6815944881889764,
"grad_norm": 2.0841809045493442,
"learning_rate": 2.7853459463784643e-06,
"loss": 0.3658,
"step": 1385
},
{
"epoch": 0.6840551181102362,
"grad_norm": 2.220969106748428,
"learning_rate": 2.7469077642156844e-06,
"loss": 0.3602,
"step": 1390
},
{
"epoch": 0.6865157480314961,
"grad_norm": 2.009625282982446,
"learning_rate": 2.7086359477062542e-06,
"loss": 0.353,
"step": 1395
},
{
"epoch": 0.6889763779527559,
"grad_norm": 2.267920231263409,
"learning_rate": 2.6705333227956304e-06,
"loss": 0.3672,
"step": 1400
},
{
"epoch": 0.6914370078740157,
"grad_norm": 2.099753211120495,
"learning_rate": 2.6326027029363575e-06,
"loss": 0.3657,
"step": 1405
},
{
"epoch": 0.6938976377952756,
"grad_norm": 2.3887622578641046,
"learning_rate": 2.5948468888803323e-06,
"loss": 0.3641,
"step": 1410
},
{
"epoch": 0.6963582677165354,
"grad_norm": 2.2574854591596063,
"learning_rate": 2.557268668472002e-06,
"loss": 0.3681,
"step": 1415
},
{
"epoch": 0.6988188976377953,
"grad_norm": 2.1541998550392285,
"learning_rate": 2.5198708164425046e-06,
"loss": 0.3666,
"step": 1420
},
{
"epoch": 0.7012795275590551,
"grad_norm": 2.2114896772148924,
"learning_rate": 2.482656094204794e-06,
"loss": 0.3721,
"step": 1425
},
{
"epoch": 0.7037401574803149,
"grad_norm": 1.999095279733753,
"learning_rate": 2.445627249649742e-06,
"loss": 0.3533,
"step": 1430
},
{
"epoch": 0.7062007874015748,
"grad_norm": 2.447157001336924,
"learning_rate": 2.4087870169432263e-06,
"loss": 0.3512,
"step": 1435
},
{
"epoch": 0.7086614173228346,
"grad_norm": 2.1383643267190875,
"learning_rate": 2.372138116324254e-06,
"loss": 0.3637,
"step": 1440
},
{
"epoch": 0.7111220472440944,
"grad_norm": 2.2411777344710933,
"learning_rate": 2.3356832539040976e-06,
"loss": 0.3521,
"step": 1445
},
{
"epoch": 0.7135826771653543,
"grad_norm": 2.081326294904533,
"learning_rate": 2.2994251214664754e-06,
"loss": 0.3636,
"step": 1450
},
{
"epoch": 0.7160433070866141,
"grad_norm": 2.0265851050443717,
"learning_rate": 2.263366396268806e-06,
"loss": 0.3631,
"step": 1455
},
{
"epoch": 0.718503937007874,
"grad_norm": 2.016445263024121,
"learning_rate": 2.227509740844508e-06,
"loss": 0.3506,
"step": 1460
},
{
"epoch": 0.7209645669291339,
"grad_norm": 1.9859142840819781,
"learning_rate": 2.191857802806409e-06,
"loss": 0.3492,
"step": 1465
},
{
"epoch": 0.7234251968503937,
"grad_norm": 2.131894669259727,
"learning_rate": 2.1564132146512494e-06,
"loss": 0.361,
"step": 1470
},
{
"epoch": 0.7258858267716536,
"grad_norm": 2.054521134481261,
"learning_rate": 2.1211785935652974e-06,
"loss": 0.353,
"step": 1475
},
{
"epoch": 0.7283464566929134,
"grad_norm": 2.195553612057405,
"learning_rate": 2.086156541231109e-06,
"loss": 0.3541,
"step": 1480
},
{
"epoch": 0.7308070866141733,
"grad_norm": 2.078190944421504,
"learning_rate": 2.05134964363541e-06,
"loss": 0.3532,
"step": 1485
},
{
"epoch": 0.7332677165354331,
"grad_norm": 1.9784288837683923,
"learning_rate": 2.016760470878158e-06,
"loss": 0.3537,
"step": 1490
},
{
"epoch": 0.7357283464566929,
"grad_norm": 2.161906206866742,
"learning_rate": 1.9823915769827672e-06,
"loss": 0.3635,
"step": 1495
},
{
"epoch": 0.7381889763779528,
"grad_norm": 2.0523238713671135,
"learning_rate": 1.948245499707523e-06,
"loss": 0.3486,
"step": 1500
},
{
"epoch": 0.7406496062992126,
"grad_norm": 2.08663663246033,
"learning_rate": 1.9143247603581925e-06,
"loss": 0.3517,
"step": 1505
},
{
"epoch": 0.7431102362204725,
"grad_norm": 2.247439527318004,
"learning_rate": 1.8806318636018666e-06,
"loss": 0.3485,
"step": 1510
},
{
"epoch": 0.7455708661417323,
"grad_norm": 2.2948321577508355,
"learning_rate": 1.8471692972820027e-06,
"loss": 0.3554,
"step": 1515
},
{
"epoch": 0.7480314960629921,
"grad_norm": 2.1604348165019065,
"learning_rate": 1.8139395322347335e-06,
"loss": 0.3489,
"step": 1520
},
{
"epoch": 0.750492125984252,
"grad_norm": 2.125600663304225,
"learning_rate": 1.780945022106424e-06,
"loss": 0.3589,
"step": 1525
},
{
"epoch": 0.7529527559055118,
"grad_norm": 1.9910694715906239,
"learning_rate": 1.7481882031724929e-06,
"loss": 0.3483,
"step": 1530
},
{
"epoch": 0.7554133858267716,
"grad_norm": 2.024386438154311,
"learning_rate": 1.7156714941575292e-06,
"loss": 0.3582,
"step": 1535
},
{
"epoch": 0.7578740157480315,
"grad_norm": 2.236178264159934,
"learning_rate": 1.6833972960566868e-06,
"loss": 0.3535,
"step": 1540
},
{
"epoch": 0.7603346456692913,
"grad_norm": 2.0509148837648876,
"learning_rate": 1.6513679919583975e-06,
"loss": 0.3485,
"step": 1545
},
{
"epoch": 0.7627952755905512,
"grad_norm": 1.9020807079292121,
"learning_rate": 1.6195859468684199e-06,
"loss": 0.3496,
"step": 1550
},
{
"epoch": 0.765255905511811,
"grad_norm": 2.1302050953871694,
"learning_rate": 1.588053507535195e-06,
"loss": 0.3385,
"step": 1555
},
{
"epoch": 0.7677165354330708,
"grad_norm": 2.184409457671906,
"learning_rate": 1.5567730022765753e-06,
"loss": 0.342,
"step": 1560
},
{
"epoch": 0.7701771653543307,
"grad_norm": 2.1109842289764735,
"learning_rate": 1.5257467408078996e-06,
"loss": 0.3461,
"step": 1565
},
{
"epoch": 0.7726377952755905,
"grad_norm": 1.9447642578789646,
"learning_rate": 1.494977014071441e-06,
"loss": 0.3321,
"step": 1570
},
{
"epoch": 0.7750984251968503,
"grad_norm": 2.160598206464441,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.3463,
"step": 1575
},
{
"epoch": 0.7775590551181102,
"grad_norm": 1.9954045529577038,
"learning_rate": 1.434216233685441e-06,
"loss": 0.354,
"step": 1580
},
{
"epoch": 0.78001968503937,
"grad_norm": 1.8705318043615158,
"learning_rate": 1.4042296665397187e-06,
"loss": 0.3354,
"step": 1585
},
{
"epoch": 0.78248031496063,
"grad_norm": 2.131526940157942,
"learning_rate": 1.374508606802586e-06,
"loss": 0.326,
"step": 1590
},
{
"epoch": 0.7849409448818898,
"grad_norm": 1.9854582148960758,
"learning_rate": 1.3450552490417712e-06,
"loss": 0.3286,
"step": 1595
},
{
"epoch": 0.7874015748031497,
"grad_norm": 2.005367208929083,
"learning_rate": 1.3158717680582128e-06,
"loss": 0.3422,
"step": 1600
},
{
"epoch": 0.7898622047244095,
"grad_norm": 2.0697719484600614,
"learning_rate": 1.286960318725471e-06,
"loss": 0.3405,
"step": 1605
},
{
"epoch": 0.7923228346456693,
"grad_norm": 2.1676514262257656,
"learning_rate": 1.2583230358306053e-06,
"loss": 0.3519,
"step": 1610
},
{
"epoch": 0.7947834645669292,
"grad_norm": 2.030399197550805,
"learning_rate": 1.2299620339165607e-06,
"loss": 0.3371,
"step": 1615
},
{
"epoch": 0.797244094488189,
"grad_norm": 1.8800725084378718,
"learning_rate": 1.201879407126012e-06,
"loss": 0.3387,
"step": 1620
},
{
"epoch": 0.7997047244094488,
"grad_norm": 2.1095666405799345,
"learning_rate": 1.1740772290467518e-06,
"loss": 0.343,
"step": 1625
},
{
"epoch": 0.8021653543307087,
"grad_norm": 2.036216759579623,
"learning_rate": 1.1465575525585743e-06,
"loss": 0.3254,
"step": 1630
},
{
"epoch": 0.8046259842519685,
"grad_norm": 1.9725419805321742,
"learning_rate": 1.119322409681689e-06,
"loss": 0.3377,
"step": 1635
},
{
"epoch": 0.8070866141732284,
"grad_norm": 2.261947410750145,
"learning_rate": 1.0923738114266824e-06,
"loss": 0.3414,
"step": 1640
},
{
"epoch": 0.8095472440944882,
"grad_norm": 1.9967677016675829,
"learning_rate": 1.0657137476460272e-06,
"loss": 0.3427,
"step": 1645
},
{
"epoch": 0.812007874015748,
"grad_norm": 1.9329332696972708,
"learning_rate": 1.0393441868871507e-06,
"loss": 0.3309,
"step": 1650
},
{
"epoch": 0.8144685039370079,
"grad_norm": 2.080216309757756,
"learning_rate": 1.0132670762470875e-06,
"loss": 0.3403,
"step": 1655
},
{
"epoch": 0.8169291338582677,
"grad_norm": 1.9963077191898775,
"learning_rate": 9.874843412286994e-07,
"loss": 0.3395,
"step": 1660
},
{
"epoch": 0.8193897637795275,
"grad_norm": 2.1468283823628584,
"learning_rate": 9.619978855985017e-07,
"loss": 0.3365,
"step": 1665
},
{
"epoch": 0.8218503937007874,
"grad_norm": 2.030201170967621,
"learning_rate": 9.368095912460934e-07,
"loss": 0.3343,
"step": 1670
},
{
"epoch": 0.8243110236220472,
"grad_norm": 2.144207392793775,
"learning_rate": 9.119213180451974e-07,
"loss": 0.3365,
"step": 1675
},
{
"epoch": 0.8267716535433071,
"grad_norm": 2.139359174863488,
"learning_rate": 8.87334903716332e-07,
"loss": 0.3344,
"step": 1680
},
{
"epoch": 0.8292322834645669,
"grad_norm": 2.0756053748869028,
"learning_rate": 8.630521636911171e-07,
"loss": 0.3396,
"step": 1685
},
{
"epoch": 0.8316929133858267,
"grad_norm": 1.9417094509672648,
"learning_rate": 8.390748909782204e-07,
"loss": 0.3319,
"step": 1690
},
{
"epoch": 0.8341535433070866,
"grad_norm": 1.904699920697159,
"learning_rate": 8.154048560309669e-07,
"loss": 0.3383,
"step": 1695
},
{
"epoch": 0.8366141732283464,
"grad_norm": 1.9020653291530694,
"learning_rate": 7.920438066166097e-07,
"loss": 0.3349,
"step": 1700
},
{
"epoch": 0.8390748031496063,
"grad_norm": 2.00777989664957,
"learning_rate": 7.689934676872768e-07,
"loss": 0.337,
"step": 1705
},
{
"epoch": 0.8415354330708661,
"grad_norm": 2.1247990851762677,
"learning_rate": 7.462555412526062e-07,
"loss": 0.3299,
"step": 1710
},
{
"epoch": 0.843996062992126,
"grad_norm": 2.0897419693416004,
"learning_rate": 7.238317062540661e-07,
"loss": 0.3261,
"step": 1715
},
{
"epoch": 0.8464566929133859,
"grad_norm": 2.243164262785359,
"learning_rate": 7.017236184409859e-07,
"loss": 0.3378,
"step": 1720
},
{
"epoch": 0.8489173228346457,
"grad_norm": 1.9439608610518924,
"learning_rate": 6.799329102482988e-07,
"loss": 0.3206,
"step": 1725
},
{
"epoch": 0.8513779527559056,
"grad_norm": 2.001456213276007,
"learning_rate": 6.584611906760036e-07,
"loss": 0.3224,
"step": 1730
},
{
"epoch": 0.8538385826771654,
"grad_norm": 1.9357999946640365,
"learning_rate": 6.373100451703601e-07,
"loss": 0.3514,
"step": 1735
},
{
"epoch": 0.8562992125984252,
"grad_norm": 1.8706183982176696,
"learning_rate": 6.164810355068179e-07,
"loss": 0.3355,
"step": 1740
},
{
"epoch": 0.8587598425196851,
"grad_norm": 1.9372386803072519,
"learning_rate": 5.959756996746996e-07,
"loss": 0.3383,
"step": 1745
},
{
"epoch": 0.8612204724409449,
"grad_norm": 1.9978519375008006,
"learning_rate": 5.757955517636365e-07,
"loss": 0.329,
"step": 1750
},
{
"epoch": 0.8636811023622047,
"grad_norm": 2.260582620281468,
"learning_rate": 5.559420818517702e-07,
"loss": 0.3352,
"step": 1755
},
{
"epoch": 0.8661417322834646,
"grad_norm": 1.905722889331543,
"learning_rate": 5.364167558957267e-07,
"loss": 0.3241,
"step": 1760
},
{
"epoch": 0.8686023622047244,
"grad_norm": 2.08290721357846,
"learning_rate": 5.172210156223745e-07,
"loss": 0.3336,
"step": 1765
},
{
"epoch": 0.8710629921259843,
"grad_norm": 2.0264311598023728,
"learning_rate": 4.983562784223645e-07,
"loss": 0.3191,
"step": 1770
},
{
"epoch": 0.8735236220472441,
"grad_norm": 1.9306929816740388,
"learning_rate": 4.798239372454738e-07,
"loss": 0.3256,
"step": 1775
},
{
"epoch": 0.8759842519685039,
"grad_norm": 1.9166412725439652,
"learning_rate": 4.6162536049775387e-07,
"loss": 0.3284,
"step": 1780
},
{
"epoch": 0.8784448818897638,
"grad_norm": 1.9640277798271495,
"learning_rate": 4.437618919404851e-07,
"loss": 0.3233,
"step": 1785
},
{
"epoch": 0.8809055118110236,
"grad_norm": 1.9101328543862819,
"learning_rate": 4.262348505909608e-07,
"loss": 0.326,
"step": 1790
},
{
"epoch": 0.8833661417322834,
"grad_norm": 1.9078892149815414,
"learning_rate": 4.0904553062508677e-07,
"loss": 0.3273,
"step": 1795
},
{
"epoch": 0.8858267716535433,
"grad_norm": 1.92469620169423,
"learning_rate": 3.9219520128182087e-07,
"loss": 0.3222,
"step": 1800
},
{
"epoch": 0.8882874015748031,
"grad_norm": 1.8872478826399433,
"learning_rate": 3.756851067694606e-07,
"loss": 0.3233,
"step": 1805
},
{
"epoch": 0.890748031496063,
"grad_norm": 2.1313117581123366,
"learning_rate": 3.5951646617376603e-07,
"loss": 0.3309,
"step": 1810
},
{
"epoch": 0.8932086614173228,
"grad_norm": 1.857899228482062,
"learning_rate": 3.436904733679436e-07,
"loss": 0.3231,
"step": 1815
},
{
"epoch": 0.8956692913385826,
"grad_norm": 1.9630046166560744,
"learning_rate": 3.2820829692449984e-07,
"loss": 0.3314,
"step": 1820
},
{
"epoch": 0.8981299212598425,
"grad_norm": 1.878464206381759,
"learning_rate": 3.130710800289416e-07,
"loss": 0.315,
"step": 1825
},
{
"epoch": 0.9005905511811023,
"grad_norm": 1.948299936140257,
"learning_rate": 2.982799403953801e-07,
"loss": 0.3245,
"step": 1830
},
{
"epoch": 0.9030511811023622,
"grad_norm": 1.9803613486774017,
"learning_rate": 2.8383597018398876e-07,
"loss": 0.3106,
"step": 1835
},
{
"epoch": 0.905511811023622,
"grad_norm": 1.9228830866585032,
"learning_rate": 2.697402359203638e-07,
"loss": 0.3206,
"step": 1840
},
{
"epoch": 0.9079724409448819,
"grad_norm": 1.8821637790876162,
"learning_rate": 2.559937784167743e-07,
"loss": 0.3158,
"step": 1845
},
{
"epoch": 0.9104330708661418,
"grad_norm": 2.0523446694418364,
"learning_rate": 2.4259761269530667e-07,
"loss": 0.3182,
"step": 1850
},
{
"epoch": 0.9128937007874016,
"grad_norm": 1.9067478250870236,
"learning_rate": 2.2955272791291894e-07,
"loss": 0.33,
"step": 1855
},
{
"epoch": 0.9153543307086615,
"grad_norm": 1.8161748223299183,
"learning_rate": 2.1686008728840301e-07,
"loss": 0.3195,
"step": 1860
},
{
"epoch": 0.9178149606299213,
"grad_norm": 1.9778227776433959,
"learning_rate": 2.0452062803126005e-07,
"loss": 0.3229,
"step": 1865
},
{
"epoch": 0.9202755905511811,
"grad_norm": 1.8169343313243027,
"learning_rate": 1.9253526127249787e-07,
"loss": 0.328,
"step": 1870
},
{
"epoch": 0.922736220472441,
"grad_norm": 1.9674779916821903,
"learning_rate": 1.8090487199735663e-07,
"loss": 0.3235,
"step": 1875
},
{
"epoch": 0.9251968503937008,
"grad_norm": 2.066617055366464,
"learning_rate": 1.6963031897995863e-07,
"loss": 0.3216,
"step": 1880
},
{
"epoch": 0.9276574803149606,
"grad_norm": 1.9662299527781584,
"learning_rate": 1.5871243471990372e-07,
"loss": 0.3279,
"step": 1885
},
{
"epoch": 0.9301181102362205,
"grad_norm": 2.0282199981854414,
"learning_rate": 1.4815202538079e-07,
"loss": 0.314,
"step": 1890
},
{
"epoch": 0.9325787401574803,
"grad_norm": 1.9766928164218986,
"learning_rate": 1.379498707306942e-07,
"loss": 0.3152,
"step": 1895
},
{
"epoch": 0.9350393700787402,
"grad_norm": 1.8773781832541783,
"learning_rate": 1.28106724084594e-07,
"loss": 0.3286,
"step": 1900
},
{
"epoch": 0.9375,
"grad_norm": 1.9756880092340237,
"learning_rate": 1.1862331224873902e-07,
"loss": 0.3211,
"step": 1905
},
{
"epoch": 0.9399606299212598,
"grad_norm": 1.946646764579541,
"learning_rate": 1.0950033546699213e-07,
"loss": 0.3243,
"step": 1910
},
{
"epoch": 0.9424212598425197,
"grad_norm": 1.8898884090335948,
"learning_rate": 1.0073846736911697e-07,
"loss": 0.3274,
"step": 1915
},
{
"epoch": 0.9448818897637795,
"grad_norm": 1.816110689404615,
"learning_rate": 9.233835492104326e-08,
"loss": 0.3157,
"step": 1920
},
{
"epoch": 0.9473425196850394,
"grad_norm": 1.903238646884656,
"learning_rate": 8.430061837709058e-08,
"loss": 0.3158,
"step": 1925
},
{
"epoch": 0.9498031496062992,
"grad_norm": 1.8739160354182305,
"learning_rate": 7.662585123417609e-08,
"loss": 0.3136,
"step": 1930
},
{
"epoch": 0.952263779527559,
"grad_norm": 2.015422876061845,
"learning_rate": 6.931462018798407e-08,
"loss": 0.3274,
"step": 1935
},
{
"epoch": 0.9547244094488189,
"grad_norm": 1.7724513355973053,
"learning_rate": 6.236746509112824e-08,
"loss": 0.3288,
"step": 1940
},
{
"epoch": 0.9571850393700787,
"grad_norm": 1.9490022383605574,
"learning_rate": 5.5784898913284754e-08,
"loss": 0.3268,
"step": 1945
},
{
"epoch": 0.9596456692913385,
"grad_norm": 2.00148527557363,
"learning_rate": 4.9567407703319247e-08,
"loss": 0.327,
"step": 1950
},
{
"epoch": 0.9621062992125984,
"grad_norm": 1.9582200223538588,
"learning_rate": 4.3715450553393765e-08,
"loss": 0.3252,
"step": 1955
},
{
"epoch": 0.9645669291338582,
"grad_norm": 1.837922483605557,
"learning_rate": 3.8229459565070074e-08,
"loss": 0.31,
"step": 1960
},
{
"epoch": 0.9670275590551181,
"grad_norm": 1.910793470177359,
"learning_rate": 3.3109839817404564e-08,
"loss": 0.318,
"step": 1965
},
{
"epoch": 0.969488188976378,
"grad_norm": 1.8816675788362491,
"learning_rate": 2.8356969337035578e-08,
"loss": 0.3262,
"step": 1970
},
{
"epoch": 0.9719488188976378,
"grad_norm": 1.991419593684435,
"learning_rate": 2.3971199070271234e-08,
"loss": 0.3217,
"step": 1975
},
{
"epoch": 0.9744094488188977,
"grad_norm": 1.8312406446386904,
"learning_rate": 1.99528528571763e-08,
"loss": 0.3267,
"step": 1980
},
{
"epoch": 0.9768700787401575,
"grad_norm": 1.9479093178370912,
"learning_rate": 1.6302227407660744e-08,
"loss": 0.3209,
"step": 1985
},
{
"epoch": 0.9793307086614174,
"grad_norm": 1.9735060238944024,
"learning_rate": 1.3019592279569503e-08,
"loss": 0.3243,
"step": 1990
},
{
"epoch": 0.9817913385826772,
"grad_norm": 2.043603259800287,
"learning_rate": 1.0105189858779507e-08,
"loss": 0.323,
"step": 1995
},
{
"epoch": 0.984251968503937,
"grad_norm": 1.852960962616925,
"learning_rate": 7.559235341302872e-09,
"loss": 0.322,
"step": 2000
},
{
"epoch": 0.9867125984251969,
"grad_norm": 1.8539802968962942,
"learning_rate": 5.381916717395186e-09,
"loss": 0.3222,
"step": 2005
},
{
"epoch": 0.9891732283464567,
"grad_norm": 1.9807602796212282,
"learning_rate": 3.573394757676596e-09,
"loss": 0.3203,
"step": 2010
},
{
"epoch": 0.9916338582677166,
"grad_norm": 1.9586954229212994,
"learning_rate": 2.1338030012596488e-09,
"loss": 0.3272,
"step": 2015
},
{
"epoch": 0.9940944881889764,
"grad_norm": 1.966512685382069,
"learning_rate": 1.0632477458888401e-09,
"loss": 0.3285,
"step": 2020
},
{
"epoch": 0.9965551181102362,
"grad_norm": 1.8305164111174446,
"learning_rate": 3.618080400924484e-10,
"loss": 0.3272,
"step": 2025
},
{
"epoch": 0.9990157480314961,
"grad_norm": 1.8345259182578462,
"learning_rate": 2.9535677343872637e-11,
"loss": 0.3234,
"step": 2030
},
{
"epoch": 1.0,
"eval_runtime": 3.3017,
"eval_samples_per_second": 3.029,
"eval_steps_per_second": 0.909,
"step": 2032
},
{
"epoch": 1.0,
"step": 2032,
"total_flos": 212729730170880.0,
"train_loss": 0.4844009231218672,
"train_runtime": 18284.3129,
"train_samples_per_second": 1.778,
"train_steps_per_second": 0.111
}
],
"logging_steps": 5,
"max_steps": 2032,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 212729730170880.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}