Qwen-sft-ls-v0.1 / trainer_state.json
sci-m-wang's picture
Upload 16 files
eabb226 verified
raw
history blame contribute delete
No virus
72.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 9.987995198079231,
"eval_steps": 500,
"global_step": 2080,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.024009603841536616,
"grad_norm": 0.48156169056892395,
"learning_rate": 4.9999287112445194e-05,
"loss": 1.2537,
"step": 5
},
{
"epoch": 0.04801920768307323,
"grad_norm": 0.48818066716194153,
"learning_rate": 4.999714849043745e-05,
"loss": 1.1695,
"step": 10
},
{
"epoch": 0.07202881152460984,
"grad_norm": 0.42254194617271423,
"learning_rate": 4.999358425594454e-05,
"loss": 1.1879,
"step": 15
},
{
"epoch": 0.09603841536614646,
"grad_norm": 0.4787527024745941,
"learning_rate": 4.9988594612238336e-05,
"loss": 1.0137,
"step": 20
},
{
"epoch": 0.12004801920768307,
"grad_norm": 0.5006477236747742,
"learning_rate": 4.9982179843883225e-05,
"loss": 1.1109,
"step": 25
},
{
"epoch": 0.14405762304921968,
"grad_norm": 0.4857625961303711,
"learning_rate": 4.9974340316719906e-05,
"loss": 1.04,
"step": 30
},
{
"epoch": 0.16806722689075632,
"grad_norm": 0.5294004678726196,
"learning_rate": 4.996507647784446e-05,
"loss": 1.0158,
"step": 35
},
{
"epoch": 0.19207683073229292,
"grad_norm": 0.5924685597419739,
"learning_rate": 4.995438885558294e-05,
"loss": 0.9004,
"step": 40
},
{
"epoch": 0.21608643457382953,
"grad_norm": 0.5567044615745544,
"learning_rate": 4.9942278059461175e-05,
"loss": 0.9626,
"step": 45
},
{
"epoch": 0.24009603841536614,
"grad_norm": 0.5250746607780457,
"learning_rate": 4.992874478017003e-05,
"loss": 1.0787,
"step": 50
},
{
"epoch": 0.26410564225690275,
"grad_norm": 0.5340147018432617,
"learning_rate": 4.9913789789526023e-05,
"loss": 0.8533,
"step": 55
},
{
"epoch": 0.28811524609843936,
"grad_norm": 0.5907275080680847,
"learning_rate": 4.989741394042727e-05,
"loss": 0.9406,
"step": 60
},
{
"epoch": 0.31212484993997597,
"grad_norm": 0.6365830898284912,
"learning_rate": 4.987961816680492e-05,
"loss": 1.0017,
"step": 65
},
{
"epoch": 0.33613445378151263,
"grad_norm": 0.6287215948104858,
"learning_rate": 4.9860403483569805e-05,
"loss": 0.8902,
"step": 70
},
{
"epoch": 0.36014405762304924,
"grad_norm": 0.6473432183265686,
"learning_rate": 4.983977098655461e-05,
"loss": 0.907,
"step": 75
},
{
"epoch": 0.38415366146458585,
"grad_norm": 0.663551926612854,
"learning_rate": 4.981772185245135e-05,
"loss": 1.0197,
"step": 80
},
{
"epoch": 0.40816326530612246,
"grad_norm": 0.5929960012435913,
"learning_rate": 4.979425733874431e-05,
"loss": 0.8131,
"step": 85
},
{
"epoch": 0.43217286914765907,
"grad_norm": 0.6948341131210327,
"learning_rate": 4.9769378783638255e-05,
"loss": 0.8665,
"step": 90
},
{
"epoch": 0.4561824729891957,
"grad_norm": 0.8430640697479248,
"learning_rate": 4.974308760598218e-05,
"loss": 0.9219,
"step": 95
},
{
"epoch": 0.4801920768307323,
"grad_norm": 0.7752465009689331,
"learning_rate": 4.971538530518836e-05,
"loss": 0.9425,
"step": 100
},
{
"epoch": 0.5042016806722689,
"grad_norm": 0.7160285711288452,
"learning_rate": 4.968627346114681e-05,
"loss": 0.8781,
"step": 105
},
{
"epoch": 0.5282112845138055,
"grad_norm": 0.7385575175285339,
"learning_rate": 4.965575373413527e-05,
"loss": 0.8446,
"step": 110
},
{
"epoch": 0.5522208883553421,
"grad_norm": 0.8385459184646606,
"learning_rate": 4.9623827864724394e-05,
"loss": 0.9453,
"step": 115
},
{
"epoch": 0.5762304921968787,
"grad_norm": 0.8416388630867004,
"learning_rate": 4.959049767367859e-05,
"loss": 0.8086,
"step": 120
},
{
"epoch": 0.6002400960384153,
"grad_norm": 0.7977305054664612,
"learning_rate": 4.955576506185213e-05,
"loss": 0.8657,
"step": 125
},
{
"epoch": 0.6242496998799519,
"grad_norm": 0.8073853850364685,
"learning_rate": 4.951963201008076e-05,
"loss": 0.9168,
"step": 130
},
{
"epoch": 0.6482593037214885,
"grad_norm": 0.825240433216095,
"learning_rate": 4.9482100579068706e-05,
"loss": 0.9192,
"step": 135
},
{
"epoch": 0.6722689075630253,
"grad_norm": 0.7907525300979614,
"learning_rate": 4.944317290927117e-05,
"loss": 0.7927,
"step": 140
},
{
"epoch": 0.6962785114045619,
"grad_norm": 0.7800418138504028,
"learning_rate": 4.9402851220772274e-05,
"loss": 0.8486,
"step": 145
},
{
"epoch": 0.7202881152460985,
"grad_norm": 0.8340911269187927,
"learning_rate": 4.93611378131584e-05,
"loss": 0.7937,
"step": 150
},
{
"epoch": 0.7442977190876351,
"grad_norm": 0.9103155732154846,
"learning_rate": 4.931803506538707e-05,
"loss": 0.834,
"step": 155
},
{
"epoch": 0.7683073229291717,
"grad_norm": 0.8660081028938293,
"learning_rate": 4.92735454356513e-05,
"loss": 0.7799,
"step": 160
},
{
"epoch": 0.7923169267707083,
"grad_norm": 0.7390795350074768,
"learning_rate": 4.9227671461239354e-05,
"loss": 0.8608,
"step": 165
},
{
"epoch": 0.8163265306122449,
"grad_norm": 0.8534154891967773,
"learning_rate": 4.918041575839007e-05,
"loss": 0.8872,
"step": 170
},
{
"epoch": 0.8403361344537815,
"grad_norm": 1.0927950143814087,
"learning_rate": 4.913178102214363e-05,
"loss": 0.8681,
"step": 175
},
{
"epoch": 0.8643457382953181,
"grad_norm": 0.8660680055618286,
"learning_rate": 4.9081770026187914e-05,
"loss": 0.8451,
"step": 180
},
{
"epoch": 0.8883553421368547,
"grad_norm": 0.7952775955200195,
"learning_rate": 4.9030385622700225e-05,
"loss": 0.8662,
"step": 185
},
{
"epoch": 0.9123649459783914,
"grad_norm": 0.8808633685112,
"learning_rate": 4.897763074218472e-05,
"loss": 0.8183,
"step": 190
},
{
"epoch": 0.936374549819928,
"grad_norm": 0.7455188632011414,
"learning_rate": 4.892350839330522e-05,
"loss": 0.8887,
"step": 195
},
{
"epoch": 0.9603841536614646,
"grad_norm": 0.7687739729881287,
"learning_rate": 4.886802166271364e-05,
"loss": 0.8505,
"step": 200
},
{
"epoch": 0.9843937575030012,
"grad_norm": 0.7754830121994019,
"learning_rate": 4.881117371487396e-05,
"loss": 0.835,
"step": 205
},
{
"epoch": 1.0084033613445378,
"grad_norm": 0.7828541398048401,
"learning_rate": 4.875296779188173e-05,
"loss": 0.8531,
"step": 210
},
{
"epoch": 1.0324129651860745,
"grad_norm": 1.015205979347229,
"learning_rate": 4.8693407213279206e-05,
"loss": 0.8679,
"step": 215
},
{
"epoch": 1.056422569027611,
"grad_norm": 0.8347631692886353,
"learning_rate": 4.8632495375866004e-05,
"loss": 0.8432,
"step": 220
},
{
"epoch": 1.0804321728691477,
"grad_norm": 0.9165553450584412,
"learning_rate": 4.8570235753505406e-05,
"loss": 0.8362,
"step": 225
},
{
"epoch": 1.1044417767106842,
"grad_norm": 0.954313337802887,
"learning_rate": 4.850663189692619e-05,
"loss": 0.7704,
"step": 230
},
{
"epoch": 1.128451380552221,
"grad_norm": 1.096909761428833,
"learning_rate": 4.844168743352019e-05,
"loss": 0.8357,
"step": 235
},
{
"epoch": 1.1524609843937574,
"grad_norm": 0.9319111108779907,
"learning_rate": 4.837540606713538e-05,
"loss": 0.7461,
"step": 240
},
{
"epoch": 1.1764705882352942,
"grad_norm": 1.0731585025787354,
"learning_rate": 4.830779157786465e-05,
"loss": 0.8821,
"step": 245
},
{
"epoch": 1.2004801920768307,
"grad_norm": 0.8962486982345581,
"learning_rate": 4.823884782183023e-05,
"loss": 0.889,
"step": 250
},
{
"epoch": 1.2244897959183674,
"grad_norm": 0.8899649381637573,
"learning_rate": 4.8168578730963804e-05,
"loss": 0.7407,
"step": 255
},
{
"epoch": 1.2484993997599039,
"grad_norm": 0.9980818033218384,
"learning_rate": 4.8096988312782174e-05,
"loss": 0.7822,
"step": 260
},
{
"epoch": 1.2725090036014406,
"grad_norm": 0.9607000350952148,
"learning_rate": 4.80240806501588e-05,
"loss": 0.9008,
"step": 265
},
{
"epoch": 1.296518607442977,
"grad_norm": 1.0252315998077393,
"learning_rate": 4.7949859901090896e-05,
"loss": 0.7477,
"step": 270
},
{
"epoch": 1.3205282112845138,
"grad_norm": 0.8092382550239563,
"learning_rate": 4.787433029846236e-05,
"loss": 0.7846,
"step": 275
},
{
"epoch": 1.3445378151260505,
"grad_norm": 0.9593423008918762,
"learning_rate": 4.7797496149802256e-05,
"loss": 0.8544,
"step": 280
},
{
"epoch": 1.368547418967587,
"grad_norm": 0.9516417980194092,
"learning_rate": 4.771936183703927e-05,
"loss": 0.7706,
"step": 285
},
{
"epoch": 1.3925570228091235,
"grad_norm": 1.2291992902755737,
"learning_rate": 4.763993181625174e-05,
"loss": 0.8256,
"step": 290
},
{
"epoch": 1.4165666266506602,
"grad_norm": 0.9861960411071777,
"learning_rate": 4.7559210617413514e-05,
"loss": 0.8185,
"step": 295
},
{
"epoch": 1.440576230492197,
"grad_norm": 0.8679934144020081,
"learning_rate": 4.7477202844135646e-05,
"loss": 0.7356,
"step": 300
},
{
"epoch": 1.4645858343337335,
"grad_norm": 0.754160463809967,
"learning_rate": 4.739391317340383e-05,
"loss": 0.8015,
"step": 305
},
{
"epoch": 1.4885954381752702,
"grad_norm": 1.0975936651229858,
"learning_rate": 4.730934635531161e-05,
"loss": 0.8103,
"step": 310
},
{
"epoch": 1.5126050420168067,
"grad_norm": 0.8794222474098206,
"learning_rate": 4.722350721278958e-05,
"loss": 0.6949,
"step": 315
},
{
"epoch": 1.5366146458583434,
"grad_norm": 1.009118676185608,
"learning_rate": 4.713640064133025e-05,
"loss": 0.779,
"step": 320
},
{
"epoch": 1.5606242496998801,
"grad_norm": 1.1100735664367676,
"learning_rate": 4.7048031608708876e-05,
"loss": 0.8016,
"step": 325
},
{
"epoch": 1.5846338535414166,
"grad_norm": 0.8312719464302063,
"learning_rate": 4.6958405154700154e-05,
"loss": 0.746,
"step": 330
},
{
"epoch": 1.6086434573829531,
"grad_norm": 0.9220231175422668,
"learning_rate": 4.686752639079076e-05,
"loss": 0.7439,
"step": 335
},
{
"epoch": 1.6326530612244898,
"grad_norm": 1.1262476444244385,
"learning_rate": 4.677540049988789e-05,
"loss": 0.7969,
"step": 340
},
{
"epoch": 1.6566626650660266,
"grad_norm": 1.1132248640060425,
"learning_rate": 4.668203273602363e-05,
"loss": 0.858,
"step": 345
},
{
"epoch": 1.680672268907563,
"grad_norm": 1.0028260946273804,
"learning_rate": 4.6587428424055326e-05,
"loss": 0.8538,
"step": 350
},
{
"epoch": 1.7046818727490995,
"grad_norm": 1.057077407836914,
"learning_rate": 4.649159295936191e-05,
"loss": 0.8241,
"step": 355
},
{
"epoch": 1.7286914765906363,
"grad_norm": 0.8777754902839661,
"learning_rate": 4.639453180753619e-05,
"loss": 0.7388,
"step": 360
},
{
"epoch": 1.752701080432173,
"grad_norm": 0.9972108006477356,
"learning_rate": 4.6296250504073145e-05,
"loss": 0.7924,
"step": 365
},
{
"epoch": 1.7767106842737095,
"grad_norm": 1.0971965789794922,
"learning_rate": 4.6196754654054216e-05,
"loss": 0.7719,
"step": 370
},
{
"epoch": 1.800720288115246,
"grad_norm": 1.1859172582626343,
"learning_rate": 4.609604993182767e-05,
"loss": 0.8365,
"step": 375
},
{
"epoch": 1.8247298919567827,
"grad_norm": 0.9028448462486267,
"learning_rate": 4.599414208068495e-05,
"loss": 0.8118,
"step": 380
},
{
"epoch": 1.8487394957983194,
"grad_norm": 0.8329624533653259,
"learning_rate": 4.589103691253317e-05,
"loss": 0.7881,
"step": 385
},
{
"epoch": 1.872749099639856,
"grad_norm": 0.985603928565979,
"learning_rate": 4.5786740307563636e-05,
"loss": 0.7155,
"step": 390
},
{
"epoch": 1.8967587034813924,
"grad_norm": 0.9097521901130676,
"learning_rate": 4.568125821391647e-05,
"loss": 0.8555,
"step": 395
},
{
"epoch": 1.9207683073229291,
"grad_norm": 1.0277526378631592,
"learning_rate": 4.557459664734141e-05,
"loss": 0.8685,
"step": 400
},
{
"epoch": 1.9447779111644659,
"grad_norm": 0.8617818355560303,
"learning_rate": 4.5466761690854746e-05,
"loss": 0.7708,
"step": 405
},
{
"epoch": 1.9687875150060024,
"grad_norm": 0.8086485266685486,
"learning_rate": 4.535775949439235e-05,
"loss": 0.8448,
"step": 410
},
{
"epoch": 1.9927971188475389,
"grad_norm": 1.0069591999053955,
"learning_rate": 4.5247596274458956e-05,
"loss": 0.736,
"step": 415
},
{
"epoch": 2.0168067226890756,
"grad_norm": 0.7122059464454651,
"learning_rate": 4.513627831377365e-05,
"loss": 0.7223,
"step": 420
},
{
"epoch": 2.0408163265306123,
"grad_norm": 0.9053510427474976,
"learning_rate": 4.502381196091154e-05,
"loss": 0.7504,
"step": 425
},
{
"epoch": 2.064825930372149,
"grad_norm": 1.0357030630111694,
"learning_rate": 4.491020362994168e-05,
"loss": 0.8011,
"step": 430
},
{
"epoch": 2.0888355342136853,
"grad_norm": 0.9885731339454651,
"learning_rate": 4.47954598000613e-05,
"loss": 0.8665,
"step": 435
},
{
"epoch": 2.112845138055222,
"grad_norm": 0.9989228844642639,
"learning_rate": 4.4679587015226253e-05,
"loss": 0.749,
"step": 440
},
{
"epoch": 2.1368547418967587,
"grad_norm": 0.8704085350036621,
"learning_rate": 4.456259188377786e-05,
"loss": 0.7962,
"step": 445
},
{
"epoch": 2.1608643457382954,
"grad_norm": 0.9389162063598633,
"learning_rate": 4.444448107806596e-05,
"loss": 0.7964,
"step": 450
},
{
"epoch": 2.184873949579832,
"grad_norm": 0.9259791970252991,
"learning_rate": 4.4325261334068426e-05,
"loss": 0.7516,
"step": 455
},
{
"epoch": 2.2088835534213684,
"grad_norm": 1.4041215181350708,
"learning_rate": 4.420493945100702e-05,
"loss": 0.8173,
"step": 460
},
{
"epoch": 2.232893157262905,
"grad_norm": 0.9259509444236755,
"learning_rate": 4.4083522290959564e-05,
"loss": 0.8038,
"step": 465
},
{
"epoch": 2.256902761104442,
"grad_norm": 1.0205409526824951,
"learning_rate": 4.396101677846866e-05,
"loss": 0.8055,
"step": 470
},
{
"epoch": 2.280912364945978,
"grad_norm": 0.9302276372909546,
"learning_rate": 4.383742990014671e-05,
"loss": 0.7459,
"step": 475
},
{
"epoch": 2.304921968787515,
"grad_norm": 0.9929239749908447,
"learning_rate": 4.371276870427753e-05,
"loss": 0.7959,
"step": 480
},
{
"epoch": 2.3289315726290516,
"grad_norm": 1.0315582752227783,
"learning_rate": 4.358704030041432e-05,
"loss": 0.8528,
"step": 485
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.9557288289070129,
"learning_rate": 4.346025185897424e-05,
"loss": 0.759,
"step": 490
},
{
"epoch": 2.376950780312125,
"grad_norm": 1.080580472946167,
"learning_rate": 4.333241061082944e-05,
"loss": 0.7499,
"step": 495
},
{
"epoch": 2.4009603841536613,
"grad_norm": 0.96761155128479,
"learning_rate": 4.320352384689471e-05,
"loss": 0.7279,
"step": 500
},
{
"epoch": 2.424969987995198,
"grad_norm": 1.0945556163787842,
"learning_rate": 4.307359891771165e-05,
"loss": 0.7453,
"step": 505
},
{
"epoch": 2.4489795918367347,
"grad_norm": 0.9983144998550415,
"learning_rate": 4.294264323302946e-05,
"loss": 0.8311,
"step": 510
},
{
"epoch": 2.4729891956782715,
"grad_norm": 0.8600877523422241,
"learning_rate": 4.2810664261382375e-05,
"loss": 0.7172,
"step": 515
},
{
"epoch": 2.4969987995198077,
"grad_norm": 1.0828715562820435,
"learning_rate": 4.267766952966369e-05,
"loss": 0.8084,
"step": 520
},
{
"epoch": 2.5210084033613445,
"grad_norm": 0.9153307676315308,
"learning_rate": 4.254366662269655e-05,
"loss": 0.702,
"step": 525
},
{
"epoch": 2.545018007202881,
"grad_norm": 0.8633860349655151,
"learning_rate": 4.240866318280132e-05,
"loss": 0.7261,
"step": 530
},
{
"epoch": 2.569027611044418,
"grad_norm": 0.9197015762329102,
"learning_rate": 4.227266690935978e-05,
"loss": 0.754,
"step": 535
},
{
"epoch": 2.593037214885954,
"grad_norm": 0.9729764461517334,
"learning_rate": 4.2135685558375994e-05,
"loss": 0.7855,
"step": 540
},
{
"epoch": 2.617046818727491,
"grad_norm": 0.981951117515564,
"learning_rate": 4.199772694203399e-05,
"loss": 0.7584,
"step": 545
},
{
"epoch": 2.6410564225690276,
"grad_norm": 0.8585065603256226,
"learning_rate": 4.185879892825222e-05,
"loss": 0.8196,
"step": 550
},
{
"epoch": 2.6650660264105643,
"grad_norm": 0.8955056071281433,
"learning_rate": 4.1718909440234853e-05,
"loss": 0.6836,
"step": 555
},
{
"epoch": 2.689075630252101,
"grad_norm": 0.9751344919204712,
"learning_rate": 4.157806645601988e-05,
"loss": 0.7656,
"step": 560
},
{
"epoch": 2.7130852340936373,
"grad_norm": 0.9694618582725525,
"learning_rate": 4.143627800802417e-05,
"loss": 0.7787,
"step": 565
},
{
"epoch": 2.737094837935174,
"grad_norm": 1.0527385473251343,
"learning_rate": 4.1293552182585307e-05,
"loss": 0.7665,
"step": 570
},
{
"epoch": 2.7611044417767108,
"grad_norm": 0.9396669864654541,
"learning_rate": 4.114989711950047e-05,
"loss": 0.748,
"step": 575
},
{
"epoch": 2.785114045618247,
"grad_norm": 1.0898234844207764,
"learning_rate": 4.1005321011562206e-05,
"loss": 0.7761,
"step": 580
},
{
"epoch": 2.8091236494597838,
"grad_norm": 1.0972914695739746,
"learning_rate": 4.085983210409114e-05,
"loss": 0.7873,
"step": 585
},
{
"epoch": 2.8331332533013205,
"grad_norm": 1.1150754690170288,
"learning_rate": 4.0713438694465806e-05,
"loss": 0.8288,
"step": 590
},
{
"epoch": 2.857142857142857,
"grad_norm": 0.9612060785293579,
"learning_rate": 4.056614913164938e-05,
"loss": 0.7587,
"step": 595
},
{
"epoch": 2.881152460984394,
"grad_norm": 0.8540961146354675,
"learning_rate": 4.0417971815713584e-05,
"loss": 0.7096,
"step": 600
},
{
"epoch": 2.90516206482593,
"grad_norm": 1.1552797555923462,
"learning_rate": 4.026891519735955e-05,
"loss": 0.7177,
"step": 605
},
{
"epoch": 2.929171668667467,
"grad_norm": 0.8366991281509399,
"learning_rate": 4.011898777743594e-05,
"loss": 0.7244,
"step": 610
},
{
"epoch": 2.9531812725090036,
"grad_norm": 1.1516789197921753,
"learning_rate": 3.99681981064541e-05,
"loss": 0.7627,
"step": 615
},
{
"epoch": 2.9771908763505404,
"grad_norm": 1.119287133216858,
"learning_rate": 3.981655478410043e-05,
"loss": 0.7562,
"step": 620
},
{
"epoch": 3.0012004801920766,
"grad_norm": 0.9897541403770447,
"learning_rate": 3.966406645874589e-05,
"loss": 0.8271,
"step": 625
},
{
"epoch": 3.0252100840336134,
"grad_norm": 0.979117214679718,
"learning_rate": 3.951074182695284e-05,
"loss": 0.7639,
"step": 630
},
{
"epoch": 3.04921968787515,
"grad_norm": 1.234833836555481,
"learning_rate": 3.935658963297902e-05,
"loss": 0.7117,
"step": 635
},
{
"epoch": 3.073229291716687,
"grad_norm": 1.0178768634796143,
"learning_rate": 3.920161866827889e-05,
"loss": 0.7358,
"step": 640
},
{
"epoch": 3.097238895558223,
"grad_norm": 0.9386712312698364,
"learning_rate": 3.904583777100223e-05,
"loss": 0.7409,
"step": 645
},
{
"epoch": 3.12124849939976,
"grad_norm": 1.140830397605896,
"learning_rate": 3.888925582549006e-05,
"loss": 0.7052,
"step": 650
},
{
"epoch": 3.1452581032412965,
"grad_norm": 0.9676099419593811,
"learning_rate": 3.8731881761768e-05,
"loss": 0.7602,
"step": 655
},
{
"epoch": 3.1692677070828332,
"grad_norm": 1.1593329906463623,
"learning_rate": 3.857372455503697e-05,
"loss": 0.8514,
"step": 660
},
{
"epoch": 3.19327731092437,
"grad_norm": 1.1297277212142944,
"learning_rate": 3.8414793225161325e-05,
"loss": 0.7699,
"step": 665
},
{
"epoch": 3.2172869147659062,
"grad_norm": 0.9413173794746399,
"learning_rate": 3.825509683615442e-05,
"loss": 0.8067,
"step": 670
},
{
"epoch": 3.241296518607443,
"grad_norm": 0.9704225063323975,
"learning_rate": 3.809464449566175e-05,
"loss": 0.6783,
"step": 675
},
{
"epoch": 3.2653061224489797,
"grad_norm": 0.9347226023674011,
"learning_rate": 3.793344535444142e-05,
"loss": 0.7315,
"step": 680
},
{
"epoch": 3.2893157262905164,
"grad_norm": 0.9760938882827759,
"learning_rate": 3.777150860584237e-05,
"loss": 0.6999,
"step": 685
},
{
"epoch": 3.3133253301320527,
"grad_norm": 1.114785075187683,
"learning_rate": 3.760884348528002e-05,
"loss": 0.7182,
"step": 690
},
{
"epoch": 3.3373349339735894,
"grad_norm": 1.0659857988357544,
"learning_rate": 3.744545926970957e-05,
"loss": 0.7577,
"step": 695
},
{
"epoch": 3.361344537815126,
"grad_norm": 1.0988335609436035,
"learning_rate": 3.728136527709694e-05,
"loss": 0.7825,
"step": 700
},
{
"epoch": 3.385354141656663,
"grad_norm": 0.9283490180969238,
"learning_rate": 3.711657086588733e-05,
"loss": 0.7414,
"step": 705
},
{
"epoch": 3.409363745498199,
"grad_norm": 1.1617555618286133,
"learning_rate": 3.695108543447154e-05,
"loss": 0.7144,
"step": 710
},
{
"epoch": 3.433373349339736,
"grad_norm": 1.1011674404144287,
"learning_rate": 3.678491842064995e-05,
"loss": 0.7516,
"step": 715
},
{
"epoch": 3.4573829531812725,
"grad_norm": 0.9130650758743286,
"learning_rate": 3.6618079301094216e-05,
"loss": 0.7484,
"step": 720
},
{
"epoch": 3.4813925570228093,
"grad_norm": 1.1056820154190063,
"learning_rate": 3.645057759080692e-05,
"loss": 0.7775,
"step": 725
},
{
"epoch": 3.505402160864346,
"grad_norm": 1.219381332397461,
"learning_rate": 3.6282422842578845e-05,
"loss": 0.7086,
"step": 730
},
{
"epoch": 3.5294117647058822,
"grad_norm": 1.2623860836029053,
"learning_rate": 3.611362464644415e-05,
"loss": 0.7165,
"step": 735
},
{
"epoch": 3.553421368547419,
"grad_norm": 1.0355645418167114,
"learning_rate": 3.594419262913351e-05,
"loss": 0.7668,
"step": 740
},
{
"epoch": 3.5774309723889557,
"grad_norm": 0.9813769459724426,
"learning_rate": 3.577413645352506e-05,
"loss": 0.6595,
"step": 745
},
{
"epoch": 3.601440576230492,
"grad_norm": 1.0704172849655151,
"learning_rate": 3.560346581809328e-05,
"loss": 0.7469,
"step": 750
},
{
"epoch": 3.6254501800720287,
"grad_norm": 1.0033190250396729,
"learning_rate": 3.543219045635593e-05,
"loss": 0.7063,
"step": 755
},
{
"epoch": 3.6494597839135654,
"grad_norm": 0.9432305097579956,
"learning_rate": 3.526032013631893e-05,
"loss": 0.6694,
"step": 760
},
{
"epoch": 3.673469387755102,
"grad_norm": 1.0187017917633057,
"learning_rate": 3.508786465991923e-05,
"loss": 0.7789,
"step": 765
},
{
"epoch": 3.697478991596639,
"grad_norm": 1.076158881187439,
"learning_rate": 3.491483386246588e-05,
"loss": 0.7934,
"step": 770
},
{
"epoch": 3.721488595438175,
"grad_norm": 0.8803985714912415,
"learning_rate": 3.474123761207905e-05,
"loss": 0.7295,
"step": 775
},
{
"epoch": 3.745498199279712,
"grad_norm": 0.9059305787086487,
"learning_rate": 3.456708580912725e-05,
"loss": 0.7474,
"step": 780
},
{
"epoch": 3.7695078031212486,
"grad_norm": 1.1186120510101318,
"learning_rate": 3.4392388385662714e-05,
"loss": 0.7711,
"step": 785
},
{
"epoch": 3.7935174069627853,
"grad_norm": 1.0779385566711426,
"learning_rate": 3.4217155304854976e-05,
"loss": 0.7363,
"step": 790
},
{
"epoch": 3.817527010804322,
"grad_norm": 1.0094141960144043,
"learning_rate": 3.4041396560422624e-05,
"loss": 0.8531,
"step": 795
},
{
"epoch": 3.8415366146458583,
"grad_norm": 0.9442546963691711,
"learning_rate": 3.386512217606339e-05,
"loss": 0.8025,
"step": 800
},
{
"epoch": 3.865546218487395,
"grad_norm": 1.0685694217681885,
"learning_rate": 3.3688342204882466e-05,
"loss": 0.6634,
"step": 805
},
{
"epoch": 3.8895558223289317,
"grad_norm": 0.9987934827804565,
"learning_rate": 3.351106672881915e-05,
"loss": 0.7664,
"step": 810
},
{
"epoch": 3.913565426170468,
"grad_norm": 1.0503917932510376,
"learning_rate": 3.33333058580719e-05,
"loss": 0.865,
"step": 815
},
{
"epoch": 3.9375750300120047,
"grad_norm": 0.9366332292556763,
"learning_rate": 3.3155069730521735e-05,
"loss": 0.7096,
"step": 820
},
{
"epoch": 3.9615846338535414,
"grad_norm": 1.173298954963684,
"learning_rate": 3.2976368511153996e-05,
"loss": 0.8476,
"step": 825
},
{
"epoch": 3.985594237695078,
"grad_norm": 1.0985863208770752,
"learning_rate": 3.2797212391478724e-05,
"loss": 0.705,
"step": 830
},
{
"epoch": 4.009603841536615,
"grad_norm": 1.124514102935791,
"learning_rate": 3.261761158894937e-05,
"loss": 0.7086,
"step": 835
},
{
"epoch": 4.033613445378151,
"grad_norm": 0.9647889733314514,
"learning_rate": 3.243757634638008e-05,
"loss": 0.6899,
"step": 840
},
{
"epoch": 4.057623049219688,
"grad_norm": 0.9771691560745239,
"learning_rate": 3.225711693136156e-05,
"loss": 0.8088,
"step": 845
},
{
"epoch": 4.081632653061225,
"grad_norm": 1.1176905632019043,
"learning_rate": 3.2076243635675513e-05,
"loss": 0.7096,
"step": 850
},
{
"epoch": 4.105642256902761,
"grad_norm": 1.0890402793884277,
"learning_rate": 3.189496677470765e-05,
"loss": 0.7646,
"step": 855
},
{
"epoch": 4.129651860744298,
"grad_norm": 1.099631905555725,
"learning_rate": 3.1713296686859426e-05,
"loss": 0.7525,
"step": 860
},
{
"epoch": 4.153661464585834,
"grad_norm": 0.995556652545929,
"learning_rate": 3.153124373295841e-05,
"loss": 0.7231,
"step": 865
},
{
"epoch": 4.177671068427371,
"grad_norm": 1.1333503723144531,
"learning_rate": 3.1348818295667424e-05,
"loss": 0.7431,
"step": 870
},
{
"epoch": 4.201680672268908,
"grad_norm": 1.0261584520339966,
"learning_rate": 3.116603077889238e-05,
"loss": 0.7619,
"step": 875
},
{
"epoch": 4.225690276110444,
"grad_norm": 0.9810305833816528,
"learning_rate": 3.098289160718895e-05,
"loss": 0.7806,
"step": 880
},
{
"epoch": 4.249699879951981,
"grad_norm": 1.160054087638855,
"learning_rate": 3.079941122516803e-05,
"loss": 0.701,
"step": 885
},
{
"epoch": 4.2737094837935174,
"grad_norm": 0.9885622262954712,
"learning_rate": 3.061560009690011e-05,
"loss": 0.7114,
"step": 890
},
{
"epoch": 4.297719087635054,
"grad_norm": 1.092362880706787,
"learning_rate": 3.0431468705318424e-05,
"loss": 0.7407,
"step": 895
},
{
"epoch": 4.321728691476591,
"grad_norm": 1.3226685523986816,
"learning_rate": 3.024702755162119e-05,
"loss": 0.756,
"step": 900
},
{
"epoch": 4.345738295318127,
"grad_norm": 1.164993166923523,
"learning_rate": 3.0062287154672658e-05,
"loss": 0.6897,
"step": 905
},
{
"epoch": 4.369747899159664,
"grad_norm": 1.1299794912338257,
"learning_rate": 2.9877258050403212e-05,
"loss": 0.7212,
"step": 910
},
{
"epoch": 4.393757503001201,
"grad_norm": 1.1793334484100342,
"learning_rate": 2.9691950791208502e-05,
"loss": 0.8578,
"step": 915
},
{
"epoch": 4.417767106842737,
"grad_norm": 1.2607135772705078,
"learning_rate": 2.950637594534765e-05,
"loss": 0.7116,
"step": 920
},
{
"epoch": 4.441776710684274,
"grad_norm": 1.309200644493103,
"learning_rate": 2.9320544096340493e-05,
"loss": 0.7693,
"step": 925
},
{
"epoch": 4.46578631452581,
"grad_norm": 0.9688151478767395,
"learning_rate": 2.9134465842364035e-05,
"loss": 0.6641,
"step": 930
},
{
"epoch": 4.489795918367347,
"grad_norm": 1.155928373336792,
"learning_rate": 2.8948151795647993e-05,
"loss": 0.7488,
"step": 935
},
{
"epoch": 4.513805522208884,
"grad_norm": 1.2800368070602417,
"learning_rate": 2.876161258186958e-05,
"loss": 0.6626,
"step": 940
},
{
"epoch": 4.53781512605042,
"grad_norm": 1.2294648885726929,
"learning_rate": 2.8574858839547512e-05,
"loss": 0.7064,
"step": 945
},
{
"epoch": 4.561824729891956,
"grad_norm": 1.1497923135757446,
"learning_rate": 2.83879012194353e-05,
"loss": 0.703,
"step": 950
},
{
"epoch": 4.5858343337334935,
"grad_norm": 1.0689631700515747,
"learning_rate": 2.8200750383913776e-05,
"loss": 0.8085,
"step": 955
},
{
"epoch": 4.60984393757503,
"grad_norm": 1.038040041923523,
"learning_rate": 2.8013417006383076e-05,
"loss": 0.7251,
"step": 960
},
{
"epoch": 4.633853541416567,
"grad_norm": 0.9959046244621277,
"learning_rate": 2.782591177065388e-05,
"loss": 0.7358,
"step": 965
},
{
"epoch": 4.657863145258103,
"grad_norm": 1.1039036512374878,
"learning_rate": 2.763824537033809e-05,
"loss": 0.7625,
"step": 970
},
{
"epoch": 4.6818727490996395,
"grad_norm": 0.98404860496521,
"learning_rate": 2.7450428508239024e-05,
"loss": 0.8123,
"step": 975
},
{
"epoch": 4.705882352941177,
"grad_norm": 1.1559010744094849,
"learning_rate": 2.726247189574095e-05,
"loss": 0.7068,
"step": 980
},
{
"epoch": 4.729891956782713,
"grad_norm": 1.0589702129364014,
"learning_rate": 2.707438625219827e-05,
"loss": 0.7431,
"step": 985
},
{
"epoch": 4.75390156062425,
"grad_norm": 1.207343339920044,
"learning_rate": 2.6886182304324153e-05,
"loss": 0.7161,
"step": 990
},
{
"epoch": 4.777911164465786,
"grad_norm": 1.023870825767517,
"learning_rate": 2.669787078557876e-05,
"loss": 0.7311,
"step": 995
},
{
"epoch": 4.801920768307323,
"grad_norm": 0.9353678226470947,
"learning_rate": 2.6509462435557152e-05,
"loss": 0.6689,
"step": 1000
},
{
"epoch": 4.82593037214886,
"grad_norm": 0.8823081851005554,
"learning_rate": 2.6320967999376767e-05,
"loss": 0.6888,
"step": 1005
},
{
"epoch": 4.849939975990396,
"grad_norm": 1.1103065013885498,
"learning_rate": 2.6132398227064615e-05,
"loss": 0.7879,
"step": 1010
},
{
"epoch": 4.873949579831933,
"grad_norm": 1.061865210533142,
"learning_rate": 2.5943763872944206e-05,
"loss": 0.6958,
"step": 1015
},
{
"epoch": 4.8979591836734695,
"grad_norm": 1.1248444318771362,
"learning_rate": 2.5755075695022224e-05,
"loss": 0.6858,
"step": 1020
},
{
"epoch": 4.921968787515006,
"grad_norm": 1.1603715419769287,
"learning_rate": 2.5566344454374968e-05,
"loss": 0.6799,
"step": 1025
},
{
"epoch": 4.945978391356543,
"grad_norm": 1.1303590536117554,
"learning_rate": 2.5377580914534647e-05,
"loss": 0.767,
"step": 1030
},
{
"epoch": 4.969987995198079,
"grad_norm": 1.2461163997650146,
"learning_rate": 2.5188795840875544e-05,
"loss": 0.7453,
"step": 1035
},
{
"epoch": 4.9939975990396155,
"grad_norm": 1.0747658014297485,
"learning_rate": 2.5e-05,
"loss": 0.7559,
"step": 1040
},
{
"epoch": 5.018007202881153,
"grad_norm": 1.0154122114181519,
"learning_rate": 2.481120415912446e-05,
"loss": 0.7729,
"step": 1045
},
{
"epoch": 5.042016806722689,
"grad_norm": 1.195665717124939,
"learning_rate": 2.4622419085465355e-05,
"loss": 0.6956,
"step": 1050
},
{
"epoch": 5.066026410564226,
"grad_norm": 1.1738536357879639,
"learning_rate": 2.4433655545625038e-05,
"loss": 0.7124,
"step": 1055
},
{
"epoch": 5.090036014405762,
"grad_norm": 1.158319354057312,
"learning_rate": 2.4244924304977785e-05,
"loss": 0.7472,
"step": 1060
},
{
"epoch": 5.114045618247299,
"grad_norm": 1.1242644786834717,
"learning_rate": 2.40562361270558e-05,
"loss": 0.6805,
"step": 1065
},
{
"epoch": 5.138055222088836,
"grad_norm": 1.0289393663406372,
"learning_rate": 2.3867601772935397e-05,
"loss": 0.6898,
"step": 1070
},
{
"epoch": 5.162064825930372,
"grad_norm": 1.0357887744903564,
"learning_rate": 2.367903200062324e-05,
"loss": 0.6664,
"step": 1075
},
{
"epoch": 5.186074429771908,
"grad_norm": 1.1436580419540405,
"learning_rate": 2.3490537564442847e-05,
"loss": 0.7125,
"step": 1080
},
{
"epoch": 5.2100840336134455,
"grad_norm": 1.1259691715240479,
"learning_rate": 2.3302129214421242e-05,
"loss": 0.7797,
"step": 1085
},
{
"epoch": 5.234093637454982,
"grad_norm": 1.1516937017440796,
"learning_rate": 2.3113817695675853e-05,
"loss": 0.7587,
"step": 1090
},
{
"epoch": 5.258103241296519,
"grad_norm": 1.1358476877212524,
"learning_rate": 2.292561374780173e-05,
"loss": 0.7414,
"step": 1095
},
{
"epoch": 5.282112845138055,
"grad_norm": 1.2753323316574097,
"learning_rate": 2.2737528104259056e-05,
"loss": 0.6972,
"step": 1100
},
{
"epoch": 5.3061224489795915,
"grad_norm": 1.2500908374786377,
"learning_rate": 2.2549571491760986e-05,
"loss": 0.7168,
"step": 1105
},
{
"epoch": 5.330132052821129,
"grad_norm": 1.157196283340454,
"learning_rate": 2.236175462966192e-05,
"loss": 0.6441,
"step": 1110
},
{
"epoch": 5.354141656662665,
"grad_norm": 1.1226704120635986,
"learning_rate": 2.217408822934613e-05,
"loss": 0.7473,
"step": 1115
},
{
"epoch": 5.378151260504202,
"grad_norm": 1.0768887996673584,
"learning_rate": 2.1986582993616926e-05,
"loss": 0.6849,
"step": 1120
},
{
"epoch": 5.402160864345738,
"grad_norm": 1.2554682493209839,
"learning_rate": 2.179924961608623e-05,
"loss": 0.6715,
"step": 1125
},
{
"epoch": 5.426170468187275,
"grad_norm": 0.9506689310073853,
"learning_rate": 2.1612098780564714e-05,
"loss": 0.7139,
"step": 1130
},
{
"epoch": 5.450180072028812,
"grad_norm": 1.2209597826004028,
"learning_rate": 2.1425141160452494e-05,
"loss": 0.8102,
"step": 1135
},
{
"epoch": 5.474189675870348,
"grad_norm": 0.9822332262992859,
"learning_rate": 2.1238387418130422e-05,
"loss": 0.7772,
"step": 1140
},
{
"epoch": 5.498199279711884,
"grad_norm": 1.1269121170043945,
"learning_rate": 2.1051848204352013e-05,
"loss": 0.7506,
"step": 1145
},
{
"epoch": 5.5222088835534215,
"grad_norm": 1.1985368728637695,
"learning_rate": 2.0865534157635967e-05,
"loss": 0.7174,
"step": 1150
},
{
"epoch": 5.546218487394958,
"grad_norm": 1.1794227361679077,
"learning_rate": 2.0679455903659513e-05,
"loss": 0.7339,
"step": 1155
},
{
"epoch": 5.570228091236495,
"grad_norm": 1.1553770303726196,
"learning_rate": 2.0493624054652357e-05,
"loss": 0.7076,
"step": 1160
},
{
"epoch": 5.594237695078031,
"grad_norm": 1.0076508522033691,
"learning_rate": 2.0308049208791507e-05,
"loss": 0.7758,
"step": 1165
},
{
"epoch": 5.6182472989195675,
"grad_norm": 1.0329455137252808,
"learning_rate": 2.0122741949596797e-05,
"loss": 0.7103,
"step": 1170
},
{
"epoch": 5.642256902761105,
"grad_norm": 1.0466042757034302,
"learning_rate": 1.9937712845327345e-05,
"loss": 0.747,
"step": 1175
},
{
"epoch": 5.666266506602641,
"grad_norm": 1.1813960075378418,
"learning_rate": 1.9752972448378814e-05,
"loss": 0.7325,
"step": 1180
},
{
"epoch": 5.690276110444177,
"grad_norm": 1.0979856252670288,
"learning_rate": 1.9568531294681586e-05,
"loss": 0.7336,
"step": 1185
},
{
"epoch": 5.714285714285714,
"grad_norm": 1.1668668985366821,
"learning_rate": 1.938439990309991e-05,
"loss": 0.7238,
"step": 1190
},
{
"epoch": 5.738295318127251,
"grad_norm": 1.1428868770599365,
"learning_rate": 1.9200588774831975e-05,
"loss": 0.6934,
"step": 1195
},
{
"epoch": 5.762304921968788,
"grad_norm": 1.1208099126815796,
"learning_rate": 1.9017108392811065e-05,
"loss": 0.7573,
"step": 1200
},
{
"epoch": 5.786314525810324,
"grad_norm": 1.1175333261489868,
"learning_rate": 1.8833969221107622e-05,
"loss": 0.7232,
"step": 1205
},
{
"epoch": 5.81032412965186,
"grad_norm": 0.9412729740142822,
"learning_rate": 1.8651181704332578e-05,
"loss": 0.7221,
"step": 1210
},
{
"epoch": 5.834333733493398,
"grad_norm": 1.0848782062530518,
"learning_rate": 1.8468756267041595e-05,
"loss": 0.755,
"step": 1215
},
{
"epoch": 5.858343337334934,
"grad_norm": 1.2409135103225708,
"learning_rate": 1.828670331314058e-05,
"loss": 0.6962,
"step": 1220
},
{
"epoch": 5.882352941176471,
"grad_norm": 1.05846107006073,
"learning_rate": 1.810503322529236e-05,
"loss": 0.7443,
"step": 1225
},
{
"epoch": 5.906362545018007,
"grad_norm": 1.0066033601760864,
"learning_rate": 1.7923756364324492e-05,
"loss": 0.6654,
"step": 1230
},
{
"epoch": 5.930372148859544,
"grad_norm": 0.9819132685661316,
"learning_rate": 1.7742883068638447e-05,
"loss": 0.7271,
"step": 1235
},
{
"epoch": 5.954381752701081,
"grad_norm": 1.143143892288208,
"learning_rate": 1.756242365361993e-05,
"loss": 0.7538,
"step": 1240
},
{
"epoch": 5.978391356542617,
"grad_norm": 1.1252750158309937,
"learning_rate": 1.7382388411050638e-05,
"loss": 0.6795,
"step": 1245
},
{
"epoch": 6.002400960384153,
"grad_norm": 1.3320536613464355,
"learning_rate": 1.7202787608521278e-05,
"loss": 0.6446,
"step": 1250
},
{
"epoch": 6.02641056422569,
"grad_norm": 1.1098029613494873,
"learning_rate": 1.7023631488846006e-05,
"loss": 0.6994,
"step": 1255
},
{
"epoch": 6.050420168067227,
"grad_norm": 1.1677411794662476,
"learning_rate": 1.6844930269478274e-05,
"loss": 0.6886,
"step": 1260
},
{
"epoch": 6.074429771908764,
"grad_norm": 1.1150180101394653,
"learning_rate": 1.6666694141928096e-05,
"loss": 0.7485,
"step": 1265
},
{
"epoch": 6.0984393757503,
"grad_norm": 1.2847915887832642,
"learning_rate": 1.6488933271180845e-05,
"loss": 0.6399,
"step": 1270
},
{
"epoch": 6.122448979591836,
"grad_norm": 1.1303415298461914,
"learning_rate": 1.631165779511754e-05,
"loss": 0.7799,
"step": 1275
},
{
"epoch": 6.146458583433374,
"grad_norm": 1.2571803331375122,
"learning_rate": 1.613487782393661e-05,
"loss": 0.7589,
"step": 1280
},
{
"epoch": 6.17046818727491,
"grad_norm": 1.0048649311065674,
"learning_rate": 1.595860343957738e-05,
"loss": 0.6969,
"step": 1285
},
{
"epoch": 6.194477791116446,
"grad_norm": 1.129724383354187,
"learning_rate": 1.5782844695145033e-05,
"loss": 0.7108,
"step": 1290
},
{
"epoch": 6.218487394957983,
"grad_norm": 1.0512356758117676,
"learning_rate": 1.5607611614337292e-05,
"loss": 0.7478,
"step": 1295
},
{
"epoch": 6.24249699879952,
"grad_norm": 1.0714658498764038,
"learning_rate": 1.5432914190872757e-05,
"loss": 0.7225,
"step": 1300
},
{
"epoch": 6.266506602641057,
"grad_norm": 1.2147399187088013,
"learning_rate": 1.5258762387920956e-05,
"loss": 0.6964,
"step": 1305
},
{
"epoch": 6.290516206482593,
"grad_norm": 1.1545813083648682,
"learning_rate": 1.5085166137534123e-05,
"loss": 0.7673,
"step": 1310
},
{
"epoch": 6.314525810324129,
"grad_norm": 1.092599630355835,
"learning_rate": 1.4912135340080774e-05,
"loss": 0.7194,
"step": 1315
},
{
"epoch": 6.3385354141656665,
"grad_norm": 1.2281053066253662,
"learning_rate": 1.4739679863681086e-05,
"loss": 0.6869,
"step": 1320
},
{
"epoch": 6.362545018007203,
"grad_norm": 1.1293562650680542,
"learning_rate": 1.4567809543644076e-05,
"loss": 0.6851,
"step": 1325
},
{
"epoch": 6.38655462184874,
"grad_norm": 1.459465742111206,
"learning_rate": 1.4396534181906725e-05,
"loss": 0.7242,
"step": 1330
},
{
"epoch": 6.410564225690276,
"grad_norm": 1.203008770942688,
"learning_rate": 1.4225863546474943e-05,
"loss": 0.7155,
"step": 1335
},
{
"epoch": 6.4345738295318124,
"grad_norm": 1.0592601299285889,
"learning_rate": 1.4055807370866485e-05,
"loss": 0.6706,
"step": 1340
},
{
"epoch": 6.45858343337335,
"grad_norm": 1.1361570358276367,
"learning_rate": 1.388637535355585e-05,
"loss": 0.6937,
"step": 1345
},
{
"epoch": 6.482593037214886,
"grad_norm": 1.135237455368042,
"learning_rate": 1.3717577157421169e-05,
"loss": 0.8316,
"step": 1350
},
{
"epoch": 6.506602641056423,
"grad_norm": 1.0417910814285278,
"learning_rate": 1.3549422409193083e-05,
"loss": 0.7229,
"step": 1355
},
{
"epoch": 6.530612244897959,
"grad_norm": 1.1558960676193237,
"learning_rate": 1.3381920698905787e-05,
"loss": 0.7025,
"step": 1360
},
{
"epoch": 6.554621848739496,
"grad_norm": 1.0858848094940186,
"learning_rate": 1.3215081579350058e-05,
"loss": 0.742,
"step": 1365
},
{
"epoch": 6.578631452581033,
"grad_norm": 1.2004518508911133,
"learning_rate": 1.3048914565528454e-05,
"loss": 0.6737,
"step": 1370
},
{
"epoch": 6.602641056422569,
"grad_norm": 1.1666450500488281,
"learning_rate": 1.2883429134112673e-05,
"loss": 0.7577,
"step": 1375
},
{
"epoch": 6.626650660264105,
"grad_norm": 1.1294101476669312,
"learning_rate": 1.2718634722903073e-05,
"loss": 0.6495,
"step": 1380
},
{
"epoch": 6.6506602641056425,
"grad_norm": 1.2848787307739258,
"learning_rate": 1.2554540730290437e-05,
"loss": 0.7069,
"step": 1385
},
{
"epoch": 6.674669867947179,
"grad_norm": 1.2171602249145508,
"learning_rate": 1.2391156514719984e-05,
"loss": 0.705,
"step": 1390
},
{
"epoch": 6.698679471788715,
"grad_norm": 1.0168397426605225,
"learning_rate": 1.222849139415764e-05,
"loss": 0.671,
"step": 1395
},
{
"epoch": 6.722689075630252,
"grad_norm": 1.2341060638427734,
"learning_rate": 1.2066554645558578e-05,
"loss": 0.7456,
"step": 1400
},
{
"epoch": 6.7466986794717885,
"grad_norm": 1.174575686454773,
"learning_rate": 1.1905355504338248e-05,
"loss": 0.6899,
"step": 1405
},
{
"epoch": 6.770708283313326,
"grad_norm": 1.125368595123291,
"learning_rate": 1.1744903163845577e-05,
"loss": 0.7024,
"step": 1410
},
{
"epoch": 6.794717887154862,
"grad_norm": 1.157349705696106,
"learning_rate": 1.1585206774838683e-05,
"loss": 0.6571,
"step": 1415
},
{
"epoch": 6.818727490996398,
"grad_norm": 1.1240679025650024,
"learning_rate": 1.1426275444963034e-05,
"loss": 0.729,
"step": 1420
},
{
"epoch": 6.842737094837935,
"grad_norm": 1.2598992586135864,
"learning_rate": 1.1268118238232003e-05,
"loss": 0.7971,
"step": 1425
},
{
"epoch": 6.866746698679472,
"grad_norm": 1.2101902961730957,
"learning_rate": 1.1110744174509952e-05,
"loss": 0.6376,
"step": 1430
},
{
"epoch": 6.890756302521009,
"grad_norm": 1.2665272951126099,
"learning_rate": 1.0954162228997777e-05,
"loss": 0.6663,
"step": 1435
},
{
"epoch": 6.914765906362545,
"grad_norm": 1.0040076971054077,
"learning_rate": 1.0798381331721109e-05,
"loss": 0.6898,
"step": 1440
},
{
"epoch": 6.938775510204081,
"grad_norm": 1.180012583732605,
"learning_rate": 1.0643410367020983e-05,
"loss": 0.686,
"step": 1445
},
{
"epoch": 6.9627851140456185,
"grad_norm": 1.1320979595184326,
"learning_rate": 1.048925817304717e-05,
"loss": 0.7803,
"step": 1450
},
{
"epoch": 6.986794717887155,
"grad_norm": 1.141840934753418,
"learning_rate": 1.0335933541254129e-05,
"loss": 0.6809,
"step": 1455
},
{
"epoch": 7.010804321728691,
"grad_norm": 1.0778816938400269,
"learning_rate": 1.0183445215899584e-05,
"loss": 0.6936,
"step": 1460
},
{
"epoch": 7.034813925570228,
"grad_norm": 0.9368105530738831,
"learning_rate": 1.0031801893545895e-05,
"loss": 0.7005,
"step": 1465
},
{
"epoch": 7.0588235294117645,
"grad_norm": 1.0788689851760864,
"learning_rate": 9.881012222564065e-06,
"loss": 0.8484,
"step": 1470
},
{
"epoch": 7.082833133253302,
"grad_norm": 1.1011931896209717,
"learning_rate": 9.731084802640459e-06,
"loss": 0.7897,
"step": 1475
},
{
"epoch": 7.106842737094838,
"grad_norm": 1.22800612449646,
"learning_rate": 9.582028184286423e-06,
"loss": 0.6696,
"step": 1480
},
{
"epoch": 7.130852340936374,
"grad_norm": 1.224184274673462,
"learning_rate": 9.43385086835062e-06,
"loss": 0.6867,
"step": 1485
},
{
"epoch": 7.154861944777911,
"grad_norm": 1.2134716510772705,
"learning_rate": 9.286561305534203e-06,
"loss": 0.7372,
"step": 1490
},
{
"epoch": 7.178871548619448,
"grad_norm": 1.01565682888031,
"learning_rate": 9.140167895908867e-06,
"loss": 0.6718,
"step": 1495
},
{
"epoch": 7.202881152460985,
"grad_norm": 1.176138162612915,
"learning_rate": 8.994678988437802e-06,
"loss": 0.6819,
"step": 1500
},
{
"epoch": 7.226890756302521,
"grad_norm": 1.1396265029907227,
"learning_rate": 8.850102880499531e-06,
"loss": 0.7094,
"step": 1505
},
{
"epoch": 7.250900360144057,
"grad_norm": 1.1229872703552246,
"learning_rate": 8.706447817414696e-06,
"loss": 0.7423,
"step": 1510
},
{
"epoch": 7.2749099639855945,
"grad_norm": 1.0950995683670044,
"learning_rate": 8.563721991975843e-06,
"loss": 0.6957,
"step": 1515
},
{
"epoch": 7.298919567827131,
"grad_norm": 1.322180986404419,
"learning_rate": 8.421933543980126e-06,
"loss": 0.6382,
"step": 1520
},
{
"epoch": 7.322929171668667,
"grad_norm": 1.1946529150009155,
"learning_rate": 8.281090559765156e-06,
"loss": 0.6939,
"step": 1525
},
{
"epoch": 7.346938775510204,
"grad_norm": 1.3446452617645264,
"learning_rate": 8.141201071747784e-06,
"loss": 0.7682,
"step": 1530
},
{
"epoch": 7.3709483793517405,
"grad_norm": 1.063050627708435,
"learning_rate": 8.002273057966011e-06,
"loss": 0.6319,
"step": 1535
},
{
"epoch": 7.394957983193278,
"grad_norm": 1.0960049629211426,
"learning_rate": 7.864314441624004e-06,
"loss": 0.7331,
"step": 1540
},
{
"epoch": 7.418967587034814,
"grad_norm": 1.153906226158142,
"learning_rate": 7.727333090640218e-06,
"loss": 0.7249,
"step": 1545
},
{
"epoch": 7.44297719087635,
"grad_norm": 1.1610156297683716,
"learning_rate": 7.591336817198682e-06,
"loss": 0.6484,
"step": 1550
},
{
"epoch": 7.466986794717887,
"grad_norm": 1.1550188064575195,
"learning_rate": 7.456333377303457e-06,
"loss": 0.7723,
"step": 1555
},
{
"epoch": 7.490996398559424,
"grad_norm": 1.2671056985855103,
"learning_rate": 7.3223304703363135e-06,
"loss": 0.7266,
"step": 1560
},
{
"epoch": 7.515006002400961,
"grad_norm": 1.234147548675537,
"learning_rate": 7.189335738617633e-06,
"loss": 0.6927,
"step": 1565
},
{
"epoch": 7.539015606242497,
"grad_norm": 1.1650274991989136,
"learning_rate": 7.057356766970541e-06,
"loss": 0.7219,
"step": 1570
},
{
"epoch": 7.563025210084033,
"grad_norm": 1.2067919969558716,
"learning_rate": 6.926401082288359e-06,
"loss": 0.6933,
"step": 1575
},
{
"epoch": 7.5870348139255706,
"grad_norm": 1.1489832401275635,
"learning_rate": 6.796476153105294e-06,
"loss": 0.7651,
"step": 1580
},
{
"epoch": 7.611044417767107,
"grad_norm": 1.24705970287323,
"learning_rate": 6.667589389170562e-06,
"loss": 0.6558,
"step": 1585
},
{
"epoch": 7.635054021608643,
"grad_norm": 1.0954616069793701,
"learning_rate": 6.5397481410257645e-06,
"loss": 0.6601,
"step": 1590
},
{
"epoch": 7.65906362545018,
"grad_norm": 1.272865653038025,
"learning_rate": 6.41295969958568e-06,
"loss": 0.6938,
"step": 1595
},
{
"epoch": 7.6830732292917165,
"grad_norm": 1.2320659160614014,
"learning_rate": 6.28723129572247e-06,
"loss": 0.7301,
"step": 1600
},
{
"epoch": 7.707082833133253,
"grad_norm": 1.2697712182998657,
"learning_rate": 6.16257009985329e-06,
"loss": 0.6848,
"step": 1605
},
{
"epoch": 7.73109243697479,
"grad_norm": 1.1409835815429688,
"learning_rate": 6.038983221531352e-06,
"loss": 0.6372,
"step": 1610
},
{
"epoch": 7.755102040816326,
"grad_norm": 1.1304911375045776,
"learning_rate": 5.916477709040444e-06,
"loss": 0.6736,
"step": 1615
},
{
"epoch": 7.779111644657863,
"grad_norm": 1.1968624591827393,
"learning_rate": 5.79506054899299e-06,
"loss": 0.6915,
"step": 1620
},
{
"epoch": 7.8031212484994,
"grad_norm": 1.0843942165374756,
"learning_rate": 5.674738665931575e-06,
"loss": 0.696,
"step": 1625
},
{
"epoch": 7.827130852340936,
"grad_norm": 1.0682406425476074,
"learning_rate": 5.555518921934047e-06,
"loss": 0.7064,
"step": 1630
},
{
"epoch": 7.851140456182473,
"grad_norm": 1.13326096534729,
"learning_rate": 5.437408116222148e-06,
"loss": 0.7497,
"step": 1635
},
{
"epoch": 7.875150060024009,
"grad_norm": 1.1618019342422485,
"learning_rate": 5.320412984773748e-06,
"loss": 0.7025,
"step": 1640
},
{
"epoch": 7.899159663865547,
"grad_norm": 1.2662303447723389,
"learning_rate": 5.204540199938707e-06,
"loss": 0.6823,
"step": 1645
},
{
"epoch": 7.923169267707083,
"grad_norm": 1.1508044004440308,
"learning_rate": 5.089796370058325e-06,
"loss": 0.8013,
"step": 1650
},
{
"epoch": 7.947178871548619,
"grad_norm": 1.1236162185668945,
"learning_rate": 4.9761880390884694e-06,
"loss": 0.7497,
"step": 1655
},
{
"epoch": 7.971188475390156,
"grad_norm": 1.239221453666687,
"learning_rate": 4.86372168622635e-06,
"loss": 0.7817,
"step": 1660
},
{
"epoch": 7.995198079231693,
"grad_norm": 1.1400153636932373,
"learning_rate": 4.7524037255410434e-06,
"loss": 0.615,
"step": 1665
},
{
"epoch": 8.01920768307323,
"grad_norm": 1.1075925827026367,
"learning_rate": 4.642240505607659e-06,
"loss": 0.7037,
"step": 1670
},
{
"epoch": 8.043217286914766,
"grad_norm": 1.244471788406372,
"learning_rate": 4.533238309145258e-06,
"loss": 0.6747,
"step": 1675
},
{
"epoch": 8.067226890756302,
"grad_norm": 1.114261269569397,
"learning_rate": 4.425403352658591e-06,
"loss": 0.7154,
"step": 1680
},
{
"epoch": 8.091236494597839,
"grad_norm": 1.2810643911361694,
"learning_rate": 4.318741786083538e-06,
"loss": 0.7838,
"step": 1685
},
{
"epoch": 8.115246098439377,
"grad_norm": 1.1281261444091797,
"learning_rate": 4.213259692436367e-06,
"loss": 0.7525,
"step": 1690
},
{
"epoch": 8.139255702280913,
"grad_norm": 1.111745834350586,
"learning_rate": 4.1089630874668325e-06,
"loss": 0.6624,
"step": 1695
},
{
"epoch": 8.16326530612245,
"grad_norm": 0.9732471108436584,
"learning_rate": 4.0058579193150535e-06,
"loss": 0.6821,
"step": 1700
},
{
"epoch": 8.187274909963985,
"grad_norm": 1.1068557500839233,
"learning_rate": 3.903950068172338e-06,
"loss": 0.6932,
"step": 1705
},
{
"epoch": 8.211284513805522,
"grad_norm": 1.1466325521469116,
"learning_rate": 3.8032453459457884e-06,
"loss": 0.7308,
"step": 1710
},
{
"epoch": 8.235294117647058,
"grad_norm": 1.076535940170288,
"learning_rate": 3.7037494959268644e-06,
"loss": 0.7136,
"step": 1715
},
{
"epoch": 8.259303721488596,
"grad_norm": 1.234830617904663,
"learning_rate": 3.605468192463815e-06,
"loss": 0.6595,
"step": 1720
},
{
"epoch": 8.283313325330132,
"grad_norm": 1.2779568433761597,
"learning_rate": 3.5084070406380897e-06,
"loss": 0.7671,
"step": 1725
},
{
"epoch": 8.307322929171669,
"grad_norm": 1.0826342105865479,
"learning_rate": 3.4125715759446785e-06,
"loss": 0.6561,
"step": 1730
},
{
"epoch": 8.331332533013205,
"grad_norm": 1.292898416519165,
"learning_rate": 3.317967263976374e-06,
"loss": 0.6876,
"step": 1735
},
{
"epoch": 8.355342136854741,
"grad_norm": 1.2453337907791138,
"learning_rate": 3.2245995001121106e-06,
"loss": 0.6819,
"step": 1740
},
{
"epoch": 8.37935174069628,
"grad_norm": 1.109753131866455,
"learning_rate": 3.1324736092092412e-06,
"loss": 0.7044,
"step": 1745
},
{
"epoch": 8.403361344537815,
"grad_norm": 1.1907482147216797,
"learning_rate": 3.0415948452998557e-06,
"loss": 0.6317,
"step": 1750
},
{
"epoch": 8.427370948379352,
"grad_norm": 1.036994218826294,
"learning_rate": 2.9519683912911266e-06,
"loss": 0.6521,
"step": 1755
},
{
"epoch": 8.451380552220888,
"grad_norm": 1.1481200456619263,
"learning_rate": 2.8635993586697553e-06,
"loss": 0.7572,
"step": 1760
},
{
"epoch": 8.475390156062424,
"grad_norm": 1.2365052700042725,
"learning_rate": 2.776492787210425e-06,
"loss": 0.7184,
"step": 1765
},
{
"epoch": 8.499399759903962,
"grad_norm": 1.1394840478897095,
"learning_rate": 2.690653644688393e-06,
"loss": 0.6881,
"step": 1770
},
{
"epoch": 8.523409363745499,
"grad_norm": 1.2600808143615723,
"learning_rate": 2.6060868265961822e-06,
"loss": 0.6575,
"step": 1775
},
{
"epoch": 8.547418967587035,
"grad_norm": 1.09065842628479,
"learning_rate": 2.5227971558643537e-06,
"loss": 0.7457,
"step": 1780
},
{
"epoch": 8.571428571428571,
"grad_norm": 1.245969533920288,
"learning_rate": 2.4407893825864892e-06,
"loss": 0.7127,
"step": 1785
},
{
"epoch": 8.595438175270107,
"grad_norm": 1.3175572156906128,
"learning_rate": 2.360068183748268e-06,
"loss": 0.7442,
"step": 1790
},
{
"epoch": 8.619447779111646,
"grad_norm": 1.220615267753601,
"learning_rate": 2.2806381629607327e-06,
"loss": 0.6743,
"step": 1795
},
{
"epoch": 8.643457382953182,
"grad_norm": 1.1195836067199707,
"learning_rate": 2.2025038501977486e-06,
"loss": 0.7338,
"step": 1800
},
{
"epoch": 8.667466986794718,
"grad_norm": 1.075049638748169,
"learning_rate": 2.125669701537647e-06,
"loss": 0.7514,
"step": 1805
},
{
"epoch": 8.691476590636254,
"grad_norm": 1.1793599128723145,
"learning_rate": 2.0501400989091036e-06,
"loss": 0.7774,
"step": 1810
},
{
"epoch": 8.71548619447779,
"grad_norm": 1.165880560874939,
"learning_rate": 1.97591934984121e-06,
"loss": 0.6895,
"step": 1815
},
{
"epoch": 8.739495798319329,
"grad_norm": 1.101431131362915,
"learning_rate": 1.9030116872178316e-06,
"loss": 0.7218,
"step": 1820
},
{
"epoch": 8.763505402160865,
"grad_norm": 1.2261079549789429,
"learning_rate": 1.8314212690361987e-06,
"loss": 0.7553,
"step": 1825
},
{
"epoch": 8.787515006002401,
"grad_norm": 1.1773276329040527,
"learning_rate": 1.7611521781697644e-06,
"loss": 0.6779,
"step": 1830
},
{
"epoch": 8.811524609843937,
"grad_norm": 1.2577153444290161,
"learning_rate": 1.6922084221353607e-06,
"loss": 0.6946,
"step": 1835
},
{
"epoch": 8.835534213685474,
"grad_norm": 1.1246775388717651,
"learning_rate": 1.624593932864632e-06,
"loss": 0.7149,
"step": 1840
},
{
"epoch": 8.85954381752701,
"grad_norm": 1.1089775562286377,
"learning_rate": 1.5583125664798165e-06,
"loss": 0.6681,
"step": 1845
},
{
"epoch": 8.883553421368548,
"grad_norm": 1.138650894165039,
"learning_rate": 1.4933681030738138e-06,
"loss": 0.6519,
"step": 1850
},
{
"epoch": 8.907563025210084,
"grad_norm": 1.1409391164779663,
"learning_rate": 1.429764246494597e-06,
"loss": 0.6957,
"step": 1855
},
{
"epoch": 8.93157262905162,
"grad_norm": 1.2397470474243164,
"learning_rate": 1.3675046241339918e-06,
"loss": 0.6942,
"step": 1860
},
{
"epoch": 8.955582232893157,
"grad_norm": 1.310257077217102,
"learning_rate": 1.306592786720795e-06,
"loss": 0.6578,
"step": 1865
},
{
"epoch": 8.979591836734693,
"grad_norm": 1.057267189025879,
"learning_rate": 1.2470322081182761e-06,
"loss": 0.6994,
"step": 1870
},
{
"epoch": 9.003601440576231,
"grad_norm": 1.2477996349334717,
"learning_rate": 1.1888262851260462e-06,
"loss": 0.643,
"step": 1875
},
{
"epoch": 9.027611044417768,
"grad_norm": 1.2775191068649292,
"learning_rate": 1.1319783372863602e-06,
"loss": 0.6307,
"step": 1880
},
{
"epoch": 9.051620648259304,
"grad_norm": 1.1136633157730103,
"learning_rate": 1.0764916066947794e-06,
"loss": 0.7088,
"step": 1885
},
{
"epoch": 9.07563025210084,
"grad_norm": 1.2463817596435547,
"learning_rate": 1.0223692578152782e-06,
"loss": 0.6603,
"step": 1890
},
{
"epoch": 9.099639855942376,
"grad_norm": 1.140369176864624,
"learning_rate": 9.696143772997768e-07,
"loss": 0.8255,
"step": 1895
},
{
"epoch": 9.123649459783914,
"grad_norm": 1.0478204488754272,
"learning_rate": 9.182299738120931e-07,
"loss": 0.6834,
"step": 1900
},
{
"epoch": 9.14765906362545,
"grad_norm": 1.243027925491333,
"learning_rate": 8.682189778563693e-07,
"loss": 0.6365,
"step": 1905
},
{
"epoch": 9.171668667466987,
"grad_norm": 1.1335963010787964,
"learning_rate": 8.195842416099359e-07,
"loss": 0.7317,
"step": 1910
},
{
"epoch": 9.195678271308523,
"grad_norm": 1.2012873888015747,
"learning_rate": 7.723285387606471e-07,
"loss": 0.741,
"step": 1915
},
{
"epoch": 9.21968787515006,
"grad_norm": 1.098880648612976,
"learning_rate": 7.264545643486997e-07,
"loss": 0.6881,
"step": 1920
},
{
"epoch": 9.243697478991596,
"grad_norm": 1.138432502746582,
"learning_rate": 6.819649346129304e-07,
"loss": 0.735,
"step": 1925
},
{
"epoch": 9.267707082833134,
"grad_norm": 1.163900375366211,
"learning_rate": 6.3886218684161e-07,
"loss": 0.7227,
"step": 1930
},
{
"epoch": 9.29171668667467,
"grad_norm": 1.163901448249817,
"learning_rate": 5.971487792277297e-07,
"loss": 0.681,
"step": 1935
},
{
"epoch": 9.315726290516206,
"grad_norm": 1.3683992624282837,
"learning_rate": 5.568270907288287e-07,
"loss": 0.7081,
"step": 1940
},
{
"epoch": 9.339735894357743,
"grad_norm": 1.2360917329788208,
"learning_rate": 5.178994209312948e-07,
"loss": 0.7064,
"step": 1945
},
{
"epoch": 9.363745498199279,
"grad_norm": 1.0632774829864502,
"learning_rate": 4.803679899192392e-07,
"loss": 0.6829,
"step": 1950
},
{
"epoch": 9.387755102040817,
"grad_norm": 1.200908899307251,
"learning_rate": 4.4423493814786667e-07,
"loss": 0.6814,
"step": 1955
},
{
"epoch": 9.411764705882353,
"grad_norm": 1.2517386674880981,
"learning_rate": 4.095023263214121e-07,
"loss": 0.769,
"step": 1960
},
{
"epoch": 9.43577430972389,
"grad_norm": 1.0798803567886353,
"learning_rate": 3.761721352756098e-07,
"loss": 0.7539,
"step": 1965
},
{
"epoch": 9.459783913565426,
"grad_norm": 0.9635175466537476,
"learning_rate": 3.4424626586473385e-07,
"loss": 0.6506,
"step": 1970
},
{
"epoch": 9.483793517406962,
"grad_norm": 1.1478654146194458,
"learning_rate": 3.1372653885318736e-07,
"loss": 0.6763,
"step": 1975
},
{
"epoch": 9.5078031212485,
"grad_norm": 1.2677192687988281,
"learning_rate": 2.846146948116468e-07,
"loss": 0.6776,
"step": 1980
},
{
"epoch": 9.531812725090036,
"grad_norm": 1.1099594831466675,
"learning_rate": 2.569123940178192e-07,
"loss": 0.7292,
"step": 1985
},
{
"epoch": 9.555822328931573,
"grad_norm": 1.1103848218917847,
"learning_rate": 2.3062121636174826e-07,
"loss": 0.8111,
"step": 1990
},
{
"epoch": 9.579831932773109,
"grad_norm": 1.1750984191894531,
"learning_rate": 2.0574266125569509e-07,
"loss": 0.6473,
"step": 1995
},
{
"epoch": 9.603841536614645,
"grad_norm": 1.0524730682373047,
"learning_rate": 1.8227814754865068e-07,
"loss": 0.6774,
"step": 2000
},
{
"epoch": 9.627851140456183,
"grad_norm": 1.1044491529464722,
"learning_rate": 1.6022901344539543e-07,
"loss": 0.6782,
"step": 2005
},
{
"epoch": 9.65186074429772,
"grad_norm": 1.2015644311904907,
"learning_rate": 1.39596516430196e-07,
"loss": 0.6965,
"step": 2010
},
{
"epoch": 9.675870348139256,
"grad_norm": 1.066968560218811,
"learning_rate": 1.2038183319507955e-07,
"loss": 0.6496,
"step": 2015
},
{
"epoch": 9.699879951980792,
"grad_norm": 1.1246190071105957,
"learning_rate": 1.0258605957272627e-07,
"loss": 0.6511,
"step": 2020
},
{
"epoch": 9.723889555822328,
"grad_norm": 1.017182469367981,
"learning_rate": 8.621021047398314e-08,
"loss": 0.7524,
"step": 2025
},
{
"epoch": 9.747899159663866,
"grad_norm": 1.2211177349090576,
"learning_rate": 7.125521982997152e-08,
"loss": 0.6653,
"step": 2030
},
{
"epoch": 9.771908763505403,
"grad_norm": 1.2134476900100708,
"learning_rate": 5.772194053882962e-08,
"loss": 0.6718,
"step": 2035
},
{
"epoch": 9.795918367346939,
"grad_norm": 1.3641300201416016,
"learning_rate": 4.56111444170626e-08,
"loss": 0.7114,
"step": 2040
},
{
"epoch": 9.819927971188475,
"grad_norm": 1.1759401559829712,
"learning_rate": 3.4923522155544394e-08,
"loss": 0.7125,
"step": 2045
},
{
"epoch": 9.843937575030012,
"grad_norm": 1.0511598587036133,
"learning_rate": 2.5659683280102044e-08,
"loss": 0.6236,
"step": 2050
},
{
"epoch": 9.867947178871548,
"grad_norm": 1.1891964673995972,
"learning_rate": 1.782015611677401e-08,
"loss": 0.7353,
"step": 2055
},
{
"epoch": 9.891956782713086,
"grad_norm": 1.116445541381836,
"learning_rate": 1.1405387761664887e-08,
"loss": 0.7489,
"step": 2060
},
{
"epoch": 9.915966386554622,
"grad_norm": 1.1506844758987427,
"learning_rate": 6.415744055460193e-09,
"loss": 0.7743,
"step": 2065
},
{
"epoch": 9.939975990396158,
"grad_norm": 1.2120825052261353,
"learning_rate": 2.8515095625514244e-09,
"loss": 0.6435,
"step": 2070
},
{
"epoch": 9.963985594237695,
"grad_norm": 1.1971231698989868,
"learning_rate": 7.128875548101377e-10,
"loss": 0.7375,
"step": 2075
},
{
"epoch": 9.987995198079231,
"grad_norm": 1.2055232524871826,
"learning_rate": 0.0,
"loss": 0.7238,
"step": 2080
},
{
"epoch": 9.987995198079231,
"step": 2080,
"total_flos": 8.185270082173747e+17,
"train_loss": 0.7521515275423344,
"train_runtime": 19400.1091,
"train_samples_per_second": 1.717,
"train_steps_per_second": 0.107
}
],
"logging_steps": 5,
"max_steps": 2080,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 100,
"total_flos": 8.185270082173747e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}