{
  "best_metric": 0.40588063,
  "best_model_checkpoint": "/home/ubuntu/output/v0-20250315-052746/checkpoint-800",
  "epoch": 0.9549388242315726,
  "eval_steps": 100,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001193673530289466,
      "grad_norm": 22.298568401591066,
      "learning_rate": 7.936507936507937e-08,
      "loss": 1.0442615747451782,
      "memory(GiB)": 30.7,
      "step": 1,
      "token_acc": 0.7699836867862969,
      "train_speed(iter/s)": 0.093757
    },
    {
      "epoch": 0.005968367651447329,
      "grad_norm": 19.63376949197587,
      "learning_rate": 3.9682539682539683e-07,
      "loss": 0.9844925403594971,
      "memory(GiB)": 36.92,
      "step": 5,
      "token_acc": 0.7549315386400557,
      "train_speed(iter/s)": 0.149789
    },
    {
      "epoch": 0.011936735302894658,
      "grad_norm": 16.540823173599176,
      "learning_rate": 7.936507936507937e-07,
      "loss": 0.9816521644592285,
      "memory(GiB)": 36.92,
      "step": 10,
      "token_acc": 0.7844917012448133,
      "train_speed(iter/s)": 0.162277
    },
    {
      "epoch": 0.017905102954341987,
      "grad_norm": 7.069008652065498,
      "learning_rate": 1.1904761904761906e-06,
      "loss": 0.8435724258422852,
      "memory(GiB)": 36.92,
      "step": 15,
      "token_acc": 0.7823455233291299,
      "train_speed(iter/s)": 0.176636
    },
    {
      "epoch": 0.023873470605789315,
      "grad_norm": 5.623338707216517,
      "learning_rate": 1.5873015873015873e-06,
      "loss": 0.7347106456756591,
      "memory(GiB)": 36.92,
      "step": 20,
      "token_acc": 0.8210757409440176,
      "train_speed(iter/s)": 0.177914
    },
    {
      "epoch": 0.029841838257236644,
      "grad_norm": 4.003256651554691,
      "learning_rate": 1.984126984126984e-06,
      "loss": 0.654999828338623,
      "memory(GiB)": 36.92,
      "step": 25,
      "token_acc": 0.7765267826680314,
      "train_speed(iter/s)": 0.183726
    },
    {
      "epoch": 0.03581020590868397,
      "grad_norm": 3.4287088973952624,
      "learning_rate": 2.380952380952381e-06,
      "loss": 0.5874819755554199,
      "memory(GiB)": 36.92,
      "step": 30,
      "token_acc": 0.8138706921105098,
      "train_speed(iter/s)": 0.184018
    },
    {
      "epoch": 0.0417785735601313,
      "grad_norm": 2.8169959397852256,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.6081454753875732,
      "memory(GiB)": 36.92,
      "step": 35,
      "token_acc": 0.81794500723589,
      "train_speed(iter/s)": 0.183529
    },
    {
      "epoch": 0.04774694121157863,
      "grad_norm": 3.325157673054176,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 0.5828543663024902,
      "memory(GiB)": 36.92,
      "step": 40,
      "token_acc": 0.8394230769230769,
      "train_speed(iter/s)": 0.184431
    },
    {
      "epoch": 0.05371530886302596,
      "grad_norm": 2.7879997870731166,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 0.5575875282287598,
      "memory(GiB)": 36.92,
      "step": 45,
      "token_acc": 0.8293939393939394,
      "train_speed(iter/s)": 0.187841
    },
    {
      "epoch": 0.05968367651447329,
      "grad_norm": 2.778252157541432,
      "learning_rate": 3.968253968253968e-06,
      "loss": 0.5551129341125488,
      "memory(GiB)": 36.92,
      "step": 50,
      "token_acc": 0.8260184559981995,
      "train_speed(iter/s)": 0.189026
    },
    {
      "epoch": 0.06565204416592062,
      "grad_norm": 3.0281694802961816,
      "learning_rate": 4.365079365079366e-06,
      "loss": 0.5619981765747071,
      "memory(GiB)": 36.92,
      "step": 55,
      "token_acc": 0.8104107766505904,
      "train_speed(iter/s)": 0.188889
    },
    {
      "epoch": 0.07162041181736795,
      "grad_norm": 3.102265765306523,
      "learning_rate": 4.761904761904762e-06,
      "loss": 0.5332321166992188,
      "memory(GiB)": 36.92,
      "step": 60,
      "token_acc": 0.8470640768028578,
      "train_speed(iter/s)": 0.189797
    },
    {
      "epoch": 0.07758877946881527,
      "grad_norm": 2.8144373694536444,
      "learning_rate": 5.15873015873016e-06,
      "loss": 0.5349865436553956,
      "memory(GiB)": 36.92,
      "step": 65,
      "token_acc": 0.8722838137472284,
      "train_speed(iter/s)": 0.190181
    },
    {
      "epoch": 0.0835571471202626,
      "grad_norm": 3.337016219899744,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.5452562808990479,
      "memory(GiB)": 36.92,
      "step": 70,
      "token_acc": 0.8432369942196531,
      "train_speed(iter/s)": 0.189336
    },
    {
      "epoch": 0.08952551477170993,
      "grad_norm": 3.06136632501639,
      "learning_rate": 5.9523809523809525e-06,
      "loss": 0.49980897903442384,
      "memory(GiB)": 36.92,
      "step": 75,
      "token_acc": 0.8338361568809468,
      "train_speed(iter/s)": 0.190084
    },
    {
      "epoch": 0.09549388242315726,
      "grad_norm": 2.4877794848384633,
      "learning_rate": 6.349206349206349e-06,
      "loss": 0.539669418334961,
      "memory(GiB)": 36.92,
      "step": 80,
      "token_acc": 0.8455654331197023,
      "train_speed(iter/s)": 0.189397
    },
    {
      "epoch": 0.10146225007460459,
      "grad_norm": 3.221405214526254,
      "learning_rate": 6.746031746031747e-06,
      "loss": 0.5160573959350586,
      "memory(GiB)": 36.92,
      "step": 85,
      "token_acc": 0.8462370242214533,
      "train_speed(iter/s)": 0.190557
    },
    {
      "epoch": 0.10743061772605192,
      "grad_norm": 3.2868066256101085,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.4913814067840576,
      "memory(GiB)": 36.92,
      "step": 90,
      "token_acc": 0.8452227659026526,
      "train_speed(iter/s)": 0.190372
    },
    {
      "epoch": 0.11339898537749925,
      "grad_norm": 3.051253318495505,
      "learning_rate": 7.53968253968254e-06,
      "loss": 0.49474325180053713,
      "memory(GiB)": 36.92,
      "step": 95,
      "token_acc": 0.8567275747508306,
      "train_speed(iter/s)": 0.190208
    },
    {
      "epoch": 0.11936735302894658,
      "grad_norm": 2.912617922195808,
      "learning_rate": 7.936507936507936e-06,
      "loss": 0.4725308418273926,
      "memory(GiB)": 36.92,
      "step": 100,
      "token_acc": 0.8608458390177354,
      "train_speed(iter/s)": 0.190477
    },
    {
      "epoch": 0.11936735302894658,
      "eval_loss": 0.45056208968162537,
      "eval_runtime": 10.9299,
      "eval_samples_per_second": 24.611,
      "eval_steps_per_second": 3.111,
      "eval_token_acc": 0.8447598692022433,
      "step": 100
    },
    {
      "epoch": 0.12533572068039392,
      "grad_norm": 2.8012333103883647,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.5016227722167969,
      "memory(GiB)": 36.92,
      "step": 105,
      "token_acc": 0.842248243559719,
      "train_speed(iter/s)": 0.17126
    },
    {
      "epoch": 0.13130408833184123,
      "grad_norm": 2.912876340051396,
      "learning_rate": 8.730158730158731e-06,
      "loss": 0.518134593963623,
      "memory(GiB)": 36.92,
      "step": 110,
      "token_acc": 0.8509636604384287,
      "train_speed(iter/s)": 0.172461
    },
    {
      "epoch": 0.13727245598328858,
      "grad_norm": 3.3516293509261117,
      "learning_rate": 9.126984126984127e-06,
      "loss": 0.5215555191040039,
      "memory(GiB)": 36.92,
      "step": 115,
      "token_acc": 0.8062340503098797,
      "train_speed(iter/s)": 0.172888
    },
    {
      "epoch": 0.1432408236347359,
      "grad_norm": 3.137187651305806,
      "learning_rate": 9.523809523809525e-06,
      "loss": 0.4938058853149414,
      "memory(GiB)": 36.92,
      "step": 120,
      "token_acc": 0.8689788053949904,
      "train_speed(iter/s)": 0.173312
    },
    {
      "epoch": 0.14920919128618323,
      "grad_norm": 2.9778302322354255,
      "learning_rate": 9.920634920634922e-06,
      "loss": 0.47763543128967284,
      "memory(GiB)": 36.92,
      "step": 125,
      "token_acc": 0.8223684210526315,
      "train_speed(iter/s)": 0.173984
    },
    {
      "epoch": 0.15517755893763055,
      "grad_norm": 2.454176368796001,
      "learning_rate": 9.999930596405254e-06,
      "loss": 0.5025428771972656,
      "memory(GiB)": 36.92,
      "step": 130,
      "token_acc": 0.8584961515689757,
      "train_speed(iter/s)": 0.17494
    },
    {
      "epoch": 0.1611459265890779,
      "grad_norm": 2.308752936530914,
      "learning_rate": 9.999648647603774e-06,
      "loss": 0.4561060905456543,
      "memory(GiB)": 36.92,
      "step": 135,
      "token_acc": 0.8763222131814483,
      "train_speed(iter/s)": 0.175709
    },
    {
      "epoch": 0.1671142942405252,
      "grad_norm": 3.2296962246861276,
      "learning_rate": 9.999149828091632e-06,
      "loss": 0.5205905437469482,
      "memory(GiB)": 36.92,
      "step": 140,
      "token_acc": 0.8126463700234192,
      "train_speed(iter/s)": 0.176096
    },
    {
      "epoch": 0.17308266189197255,
      "grad_norm": 2.8523931115518915,
      "learning_rate": 9.998434159506211e-06,
      "loss": 0.4669060230255127,
      "memory(GiB)": 36.92,
      "step": 145,
      "token_acc": 0.8612167300380228,
      "train_speed(iter/s)": 0.176748
    },
    {
      "epoch": 0.17905102954341987,
      "grad_norm": 2.689699959591659,
      "learning_rate": 9.997501672891208e-06,
      "loss": 0.4870173454284668,
      "memory(GiB)": 36.92,
      "step": 150,
      "token_acc": 0.8250571369208394,
      "train_speed(iter/s)": 0.177191
    },
    {
      "epoch": 0.1850193971948672,
      "grad_norm": 2.8278119407117517,
      "learning_rate": 9.99635240869527e-06,
      "loss": 0.47814245223999025,
      "memory(GiB)": 36.92,
      "step": 155,
      "token_acc": 0.819718309859155,
      "train_speed(iter/s)": 0.177751
    },
    {
      "epoch": 0.19098776484631452,
      "grad_norm": 3.0394183749078887,
      "learning_rate": 9.99498641677025e-06,
      "loss": 0.5187320232391357,
      "memory(GiB)": 36.92,
      "step": 160,
      "token_acc": 0.8540501094624179,
      "train_speed(iter/s)": 0.178072
    },
    {
      "epoch": 0.19695613249776187,
      "grad_norm": 2.4913040205595696,
      "learning_rate": 9.993403756369037e-06,
      "loss": 0.471418571472168,
      "memory(GiB)": 36.92,
      "step": 165,
      "token_acc": 0.8507351108896087,
      "train_speed(iter/s)": 0.178675
    },
    {
      "epoch": 0.20292450014920918,
      "grad_norm": 2.5344653445390937,
      "learning_rate": 9.991604496142997e-06,
      "loss": 0.5218185901641845,
      "memory(GiB)": 36.92,
      "step": 170,
      "token_acc": 0.8207423580786026,
      "train_speed(iter/s)": 0.179206
    },
    {
      "epoch": 0.20889286780065652,
      "grad_norm": 3.1475346404756737,
      "learning_rate": 9.989588714138977e-06,
      "loss": 0.4809536933898926,
      "memory(GiB)": 36.92,
      "step": 175,
      "token_acc": 0.8179710144927537,
      "train_speed(iter/s)": 0.179609
    },
    {
      "epoch": 0.21486123545210384,
      "grad_norm": 2.9443708297446927,
      "learning_rate": 9.987356497795944e-06,
      "loss": 0.5137897491455078,
      "memory(GiB)": 36.92,
      "step": 180,
      "token_acc": 0.838412017167382,
      "train_speed(iter/s)": 0.179743
    },
    {
      "epoch": 0.22082960310355118,
      "grad_norm": 2.3691979165448864,
      "learning_rate": 9.984907943941164e-06,
      "loss": 0.47942285537719725,
      "memory(GiB)": 36.92,
      "step": 185,
      "token_acc": 0.8178681677864537,
      "train_speed(iter/s)": 0.179758
    },
    {
      "epoch": 0.2267979707549985,
      "grad_norm": 2.7655733596568806,
      "learning_rate": 9.98224315878603e-06,
      "loss": 0.4850442886352539,
      "memory(GiB)": 36.92,
      "step": 190,
      "token_acc": 0.8557343020238714,
      "train_speed(iter/s)": 0.180113
    },
    {
      "epoch": 0.23276633840644584,
      "grad_norm": 2.9415771183832837,
      "learning_rate": 9.979362257921428e-06,
      "loss": 0.4999836921691895,
      "memory(GiB)": 36.92,
      "step": 195,
      "token_acc": 0.851006381934217,
      "train_speed(iter/s)": 0.180236
    },
    {
      "epoch": 0.23873470605789315,
      "grad_norm": 3.1074760910693797,
      "learning_rate": 9.976265366312746e-06,
      "loss": 0.5033563137054443,
      "memory(GiB)": 36.92,
      "step": 200,
      "token_acc": 0.8293048128342246,
      "train_speed(iter/s)": 0.180618
    },
    {
      "epoch": 0.23873470605789315,
      "eval_loss": 0.4415110647678375,
      "eval_runtime": 10.926,
      "eval_samples_per_second": 24.62,
      "eval_steps_per_second": 3.112,
      "eval_token_acc": 0.8473173672384502,
      "step": 200
    },
    {
      "epoch": 0.2447030737093405,
      "grad_norm": 2.7742402422517016,
      "learning_rate": 9.972952618294442e-06,
      "loss": 0.48658447265625,
      "memory(GiB)": 36.92,
      "step": 205,
      "token_acc": 0.8399616256759114,
      "train_speed(iter/s)": 0.171547
    },
    {
      "epoch": 0.25067144136078784,
      "grad_norm": 2.9146485975442946,
      "learning_rate": 9.969424157564215e-06,
      "loss": 0.48202037811279297,
      "memory(GiB)": 36.92,
      "step": 210,
      "token_acc": 0.8229777256740914,
      "train_speed(iter/s)": 0.172058
    },
    {
      "epoch": 0.25663980901223515,
      "grad_norm": 2.6037700192849007,
      "learning_rate": 9.965680137176778e-06,
      "loss": 0.4780398368835449,
      "memory(GiB)": 36.92,
      "step": 215,
      "token_acc": 0.8451862602806,
      "train_speed(iter/s)": 0.172776
    },
    {
      "epoch": 0.26260817666368247,
      "grad_norm": 2.4624431066871555,
      "learning_rate": 9.961720719537217e-06,
      "loss": 0.46450080871582033,
      "memory(GiB)": 36.92,
      "step": 220,
      "token_acc": 0.8089250493096647,
      "train_speed(iter/s)": 0.173186
    },
    {
      "epoch": 0.26857654431512984,
      "grad_norm": 2.6192496099911624,
      "learning_rate": 9.957546076393944e-06,
      "loss": 0.44403810501098634,
      "memory(GiB)": 36.92,
      "step": 225,
      "token_acc": 0.8560982743492249,
      "train_speed(iter/s)": 0.173308
    },
    {
      "epoch": 0.27454491196657715,
      "grad_norm": 2.5789044914565227,
      "learning_rate": 9.953156388831246e-06,
      "loss": 0.4940804481506348,
      "memory(GiB)": 36.92,
      "step": 230,
      "token_acc": 0.8385935769656699,
      "train_speed(iter/s)": 0.173656
    },
    {
      "epoch": 0.28051327961802447,
      "grad_norm": 2.3813674364243984,
      "learning_rate": 9.948551847261439e-06,
      "loss": 0.4587420463562012,
      "memory(GiB)": 36.92,
      "step": 235,
      "token_acc": 0.8549975381585426,
      "train_speed(iter/s)": 0.173976
    },
    {
      "epoch": 0.2864816472694718,
      "grad_norm": 2.7610173702743865,
      "learning_rate": 9.943732651416597e-06,
      "loss": 0.4972860336303711,
      "memory(GiB)": 36.92,
      "step": 240,
      "token_acc": 0.8406979379107183,
      "train_speed(iter/s)": 0.174337
    },
    {
      "epoch": 0.29245001492091915,
      "grad_norm": 2.390894612477832,
      "learning_rate": 9.938699010339898e-06,
      "loss": 0.4903904438018799,
      "memory(GiB)": 36.92,
      "step": 245,
      "token_acc": 0.8545253863134658,
      "train_speed(iter/s)": 0.174579
    },
    {
      "epoch": 0.29841838257236647,
      "grad_norm": 2.3711824949447546,
      "learning_rate": 9.933451142376545e-06,
      "loss": 0.4524253845214844,
      "memory(GiB)": 37.05,
      "step": 250,
      "token_acc": 0.8489612577203818,
      "train_speed(iter/s)": 0.174973
    },
    {
      "epoch": 0.3043867502238138,
      "grad_norm": 2.2514671634568364,
      "learning_rate": 9.927989275164305e-06,
      "loss": 0.48909597396850585,
      "memory(GiB)": 37.05,
      "step": 255,
      "token_acc": 0.8518639633747548,
      "train_speed(iter/s)": 0.175028
    },
    {
      "epoch": 0.3103551178752611,
      "grad_norm": 2.3714755110814005,
      "learning_rate": 9.922313645623634e-06,
      "loss": 0.4785162448883057,
      "memory(GiB)": 37.05,
      "step": 260,
      "token_acc": 0.8465215082315454,
      "train_speed(iter/s)": 0.175714
    },
    {
      "epoch": 0.31632348552670847,
      "grad_norm": 2.648679383955696,
      "learning_rate": 9.916424499947395e-06,
      "loss": 0.46675701141357423,
      "memory(GiB)": 37.05,
      "step": 265,
      "token_acc": 0.8571428571428571,
      "train_speed(iter/s)": 0.175927
    },
    {
      "epoch": 0.3222918531781558,
      "grad_norm": 2.579144815458166,
      "learning_rate": 9.910322093590177e-06,
      "loss": 0.47339348793029784,
      "memory(GiB)": 37.05,
      "step": 270,
      "token_acc": 0.8505902192242834,
      "train_speed(iter/s)": 0.176471
    },
    {
      "epoch": 0.3282602208296031,
      "grad_norm": 2.2898701347032793,
      "learning_rate": 9.904006691257224e-06,
      "loss": 0.49665226936340334,
      "memory(GiB)": 37.05,
      "step": 275,
      "token_acc": 0.8427124366910523,
      "train_speed(iter/s)": 0.17689
    },
    {
      "epoch": 0.3342285884810504,
      "grad_norm": 1.9441720928034771,
      "learning_rate": 9.897478566892942e-06,
      "loss": 0.44453701972961424,
      "memory(GiB)": 37.05,
      "step": 280,
      "token_acc": 0.8629363449691991,
      "train_speed(iter/s)": 0.177368
    },
    {
      "epoch": 0.3401969561324978,
      "grad_norm": 2.4637260658165,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.4563939094543457,
      "memory(GiB)": 37.05,
      "step": 285,
      "token_acc": 0.8230596456201648,
      "train_speed(iter/s)": 0.1776
    },
    {
      "epoch": 0.3461653237839451,
      "grad_norm": 2.287302517723748,
      "learning_rate": 9.883785293972175e-06,
      "loss": 0.504718017578125,
      "memory(GiB)": 37.05,
      "step": 290,
      "token_acc": 0.7899543378995434,
      "train_speed(iter/s)": 0.177582
    },
    {
      "epoch": 0.3521336914353924,
      "grad_norm": 2.328908891504034,
      "learning_rate": 9.87662073939139e-06,
      "loss": 0.4355961799621582,
      "memory(GiB)": 37.05,
      "step": 295,
      "token_acc": 0.8636019960683502,
      "train_speed(iter/s)": 0.177798
    },
    {
      "epoch": 0.35810205908683973,
      "grad_norm": 2.444070696496546,
      "learning_rate": 9.869244650704924e-06,
      "loss": 0.4655925750732422,
      "memory(GiB)": 37.05,
      "step": 300,
      "token_acc": 0.8573033707865169,
      "train_speed(iter/s)": 0.177836
    },
    {
      "epoch": 0.35810205908683973,
      "eval_loss": 0.4284290373325348,
      "eval_runtime": 10.9831,
      "eval_samples_per_second": 24.492,
      "eval_steps_per_second": 3.096,
      "eval_token_acc": 0.8515189711550757,
      "step": 300
    },
    {
      "epoch": 0.3640704267382871,
      "grad_norm": 2.659611192736946,
      "learning_rate": 9.861657347866778e-06,
      "loss": 0.5253509521484375,
      "memory(GiB)": 37.06,
      "step": 305,
      "token_acc": 0.828113750899928,
      "train_speed(iter/s)": 0.171888
    },
    {
      "epoch": 0.3700387943897344,
      "grad_norm": 2.6984676971627226,
      "learning_rate": 9.853859159992831e-06,
      "loss": 0.47617392539978026,
      "memory(GiB)": 37.06,
      "step": 310,
      "token_acc": 0.8316008316008316,
      "train_speed(iter/s)": 0.172231
    },
    {
      "epoch": 0.37600716204118173,
      "grad_norm": 2.195598140600359,
      "learning_rate": 9.845850425346563e-06,
      "loss": 0.4360311508178711,
      "memory(GiB)": 37.06,
      "step": 315,
      "token_acc": 0.848318462594372,
      "train_speed(iter/s)": 0.172652
    },
    {
      "epoch": 0.38197552969262905,
      "grad_norm": 2.4976869303898597,
      "learning_rate": 9.837631491324379e-06,
      "loss": 0.46515851020812987,
      "memory(GiB)": 37.06,
      "step": 320,
      "token_acc": 0.8522144522144522,
      "train_speed(iter/s)": 0.172786
    },
    {
      "epoch": 0.3879438973440764,
      "grad_norm": 3.017469180784894,
      "learning_rate": 9.829202714440544e-06,
      "loss": 0.5420156478881836,
      "memory(GiB)": 37.06,
      "step": 325,
      "token_acc": 0.8376052027543994,
      "train_speed(iter/s)": 0.17318
    },
    {
      "epoch": 0.39391226499552373,
      "grad_norm": 2.5730220119384297,
      "learning_rate": 9.820564460311719e-06,
      "loss": 0.4916552543640137,
      "memory(GiB)": 37.06,
      "step": 330,
      "token_acc": 0.8207920792079207,
      "train_speed(iter/s)": 0.173365
    },
    {
      "epoch": 0.39988063264697105,
      "grad_norm": 2.798903385122773,
      "learning_rate": 9.811717103641096e-06,
      "loss": 0.4587296485900879,
      "memory(GiB)": 37.06,
      "step": 335,
      "token_acc": 0.8592551001310126,
      "train_speed(iter/s)": 0.173592
    },
    {
      "epoch": 0.40584900029841836,
      "grad_norm": 2.6409823275058653,
      "learning_rate": 9.802661028202147e-06,
      "loss": 0.48290514945983887,
      "memory(GiB)": 37.06,
      "step": 340,
      "token_acc": 0.823793194407808,
      "train_speed(iter/s)": 0.173952
    },
    {
      "epoch": 0.41181736794986573,
      "grad_norm": 3.0285812809146635,
      "learning_rate": 9.79339662682198e-06,
      "loss": 0.46567506790161134,
      "memory(GiB)": 37.06,
      "step": 345,
      "token_acc": 0.8304556354916067,
      "train_speed(iter/s)": 0.174192
    },
    {
      "epoch": 0.41778573560131305,
      "grad_norm": 2.4611578793858486,
      "learning_rate": 9.783924301364297e-06,
      "loss": 0.4647653579711914,
      "memory(GiB)": 37.06,
      "step": 350,
      "token_acc": 0.8199260286638927,
      "train_speed(iter/s)": 0.17443
    },
    {
      "epoch": 0.42375410325276036,
      "grad_norm": 2.154896994755901,
      "learning_rate": 9.774244462711962e-06,
      "loss": 0.4952418327331543,
      "memory(GiB)": 37.06,
      "step": 355,
      "token_acc": 0.8217054263565892,
      "train_speed(iter/s)": 0.174757
    },
    {
      "epoch": 0.4297224709042077,
      "grad_norm": 2.005838047714932,
      "learning_rate": 9.764357530749178e-06,
      "loss": 0.4674674034118652,
      "memory(GiB)": 37.06,
      "step": 360,
      "token_acc": 0.841979596266551,
      "train_speed(iter/s)": 0.174828
    },
    {
      "epoch": 0.43569083855565505,
      "grad_norm": 2.292609923640767,
      "learning_rate": 9.754263934343272e-06,
      "loss": 0.44636335372924807,
      "memory(GiB)": 37.06,
      "step": 365,
      "token_acc": 0.8596112311015118,
      "train_speed(iter/s)": 0.175118
    },
    {
      "epoch": 0.44165920620710236,
      "grad_norm": 2.477107058493794,
      "learning_rate": 9.743964111326098e-06,
      "loss": 0.4866192817687988,
      "memory(GiB)": 37.06,
      "step": 370,
      "token_acc": 0.809440252675908,
      "train_speed(iter/s)": 0.175357
    },
    {
      "epoch": 0.4476275738585497,
      "grad_norm": 2.3446291196746922,
      "learning_rate": 9.733458508475038e-06,
      "loss": 0.4887577533721924,
      "memory(GiB)": 37.06,
      "step": 375,
      "token_acc": 0.8332948510736551,
      "train_speed(iter/s)": 0.175371
    },
    {
      "epoch": 0.453595941509997,
      "grad_norm": 2.29799169108157,
      "learning_rate": 9.722747581493625e-06,
      "loss": 0.49045257568359374,
      "memory(GiB)": 37.06,
      "step": 380,
      "token_acc": 0.8406266882766072,
      "train_speed(iter/s)": 0.175414
    },
    {
      "epoch": 0.45956430916144436,
      "grad_norm": 2.563802674403576,
      "learning_rate": 9.711831794991777e-06,
      "loss": 0.4675490379333496,
      "memory(GiB)": 37.06,
      "step": 385,
      "token_acc": 0.847358529964502,
      "train_speed(iter/s)": 0.175567
    },
    {
      "epoch": 0.4655326768128917,
      "grad_norm": 2.480776446284018,
      "learning_rate": 9.700711622465645e-06,
      "loss": 0.4845867156982422,
      "memory(GiB)": 37.06,
      "step": 390,
      "token_acc": 0.8422996998383745,
      "train_speed(iter/s)": 0.17572
    },
    {
      "epoch": 0.471501044464339,
      "grad_norm": 2.721044012538843,
      "learning_rate": 9.689387546277062e-06,
      "loss": 0.46145071983337405,
      "memory(GiB)": 37.06,
      "step": 395,
      "token_acc": 0.8513663630304377,
      "train_speed(iter/s)": 0.175882
    },
    {
      "epoch": 0.4774694121157863,
      "grad_norm": 2.580126202957563,
      "learning_rate": 9.677860057632642e-06,
      "loss": 0.5093360424041748,
      "memory(GiB)": 37.06,
      "step": 400,
      "token_acc": 0.8206378986866791,
      "train_speed(iter/s)": 0.175987
    },
    {
      "epoch": 0.4774694121157863,
      "eval_loss": 0.42347872257232666,
      "eval_runtime": 10.9358,
      "eval_samples_per_second": 24.598,
      "eval_steps_per_second": 3.109,
      "eval_token_acc": 0.8527429166438318,
      "step": 400
    },
    {
      "epoch": 0.4834377797672337,
      "grad_norm": 2.355447977882308,
      "learning_rate": 9.66612965656245e-06,
      "loss": 0.48992347717285156,
      "memory(GiB)": 37.06,
      "step": 405,
      "token_acc": 0.8608419645840294,
      "train_speed(iter/s)": 0.171561
    },
    {
      "epoch": 0.489406147418681,
      "grad_norm": 2.0174115419967773,
      "learning_rate": 9.654196851898325e-06,
      "loss": 0.4750755786895752,
      "memory(GiB)": 37.06,
      "step": 410,
      "token_acc": 0.8274902615470228,
      "train_speed(iter/s)": 0.171858
    },
    {
      "epoch": 0.4953745150701283,
      "grad_norm": 2.155026242929759,
      "learning_rate": 9.642062161251807e-06,
      "loss": 0.46627135276794435,
      "memory(GiB)": 37.06,
      "step": 415,
      "token_acc": 0.8661600496277916,
      "train_speed(iter/s)": 0.17197
    },
    {
      "epoch": 0.5013428827215757,
      "grad_norm": 2.8519922687228174,
      "learning_rate": 9.62972611099168e-06,
      "loss": 0.4620970726013184,
      "memory(GiB)": 37.06,
      "step": 420,
      "token_acc": 0.8595988538681948,
      "train_speed(iter/s)": 0.172268
    },
    {
      "epoch": 0.5073112503730229,
      "grad_norm": 2.5658438134794324,
      "learning_rate": 9.617189236221143e-06,
      "loss": 0.45318241119384767,
      "memory(GiB)": 37.06,
      "step": 425,
      "token_acc": 0.8252274866645748,
      "train_speed(iter/s)": 0.172438
    },
    {
      "epoch": 0.5132796180244703,
      "grad_norm": 2.2980368916312206,
      "learning_rate": 9.604452080754601e-06,
      "loss": 0.46477622985839845,
      "memory(GiB)": 37.06,
      "step": 430,
      "token_acc": 0.8681318681318682,
      "train_speed(iter/s)": 0.17271
    },
    {
      "epoch": 0.5192479856759177,
      "grad_norm": 2.3920351806796925,
      "learning_rate": 9.591515197094064e-06,
      "loss": 0.43802127838134763,
      "memory(GiB)": 37.06,
      "step": 435,
      "token_acc": 0.8632865550022635,
      "train_speed(iter/s)": 0.172963
    },
    {
      "epoch": 0.5252163533273649,
      "grad_norm": 2.3926322888936196,
      "learning_rate": 9.578379146405202e-06,
      "loss": 0.4414364814758301,
      "memory(GiB)": 37.06,
      "step": 440,
      "token_acc": 0.8378196500672948,
      "train_speed(iter/s)": 0.173049
    },
    {
      "epoch": 0.5311847209788123,
      "grad_norm": 2.5309415862721787,
      "learning_rate": 9.565044498492984e-06,
      "loss": 0.4737836837768555,
      "memory(GiB)": 37.06,
      "step": 445,
      "token_acc": 0.8400094809196492,
      "train_speed(iter/s)": 0.173413
    },
    {
      "epoch": 0.5371530886302597,
      "grad_norm": 2.574732220606661,
      "learning_rate": 9.551511831776966e-06,
      "loss": 0.4299252986907959,
      "memory(GiB)": 37.06,
      "step": 450,
      "token_acc": 0.8394777265745008,
      "train_speed(iter/s)": 0.173639
    },
    {
      "epoch": 0.5431214562817069,
      "grad_norm": 2.209862389780888,
      "learning_rate": 9.53778173326621e-06,
      "loss": 0.44927520751953126,
      "memory(GiB)": 37.06,
      "step": 455,
      "token_acc": 0.8641338013627916,
      "train_speed(iter/s)": 0.173751
    },
    {
      "epoch": 0.5490898239331543,
      "grad_norm": 2.524639918389781,
      "learning_rate": 9.523854798533814e-06,
      "loss": 0.44107656478881835,
      "memory(GiB)": 37.06,
      "step": 460,
      "token_acc": 0.8868033496967946,
      "train_speed(iter/s)": 0.174216
    },
    {
      "epoch": 0.5550581915846016,
      "grad_norm": 2.1182849441153215,
      "learning_rate": 9.509731631691071e-06,
      "loss": 0.43174285888671876,
      "memory(GiB)": 37.06,
      "step": 465,
      "token_acc": 0.855464759959142,
      "train_speed(iter/s)": 0.174365
    },
    {
      "epoch": 0.5610265592360489,
      "grad_norm": 2.2926487255366688,
      "learning_rate": 9.495412845361279e-06,
      "loss": 0.48258438110351565,
      "memory(GiB)": 37.06,
      "step": 470,
      "token_acc": 0.8603872818551279,
      "train_speed(iter/s)": 0.174664
    },
    {
      "epoch": 0.5669949268874963,
      "grad_norm": 2.192746026976168,
      "learning_rate": 9.480899060653154e-06,
      "loss": 0.4563854217529297,
      "memory(GiB)": 37.06,
      "step": 475,
      "token_acc": 0.8394289067083904,
      "train_speed(iter/s)": 0.17502
    },
    {
      "epoch": 0.5729632945389436,
      "grad_norm": 2.014209866578747,
      "learning_rate": 9.466190907133901e-06,
      "loss": 0.4754791259765625,
      "memory(GiB)": 37.06,
      "step": 480,
      "token_acc": 0.8577712609970675,
      "train_speed(iter/s)": 0.175025
    },
    {
      "epoch": 0.5789316621903909,
      "grad_norm": 2.559320864210838,
      "learning_rate": 9.451289022801894e-06,
      "loss": 0.47232685089111326,
      "memory(GiB)": 37.06,
      "step": 485,
      "token_acc": 0.8380402225074882,
      "train_speed(iter/s)": 0.175186
    },
    {
      "epoch": 0.5849000298418383,
      "grad_norm": 2.2053676509330433,
      "learning_rate": 9.436194054058998e-06,
      "loss": 0.4336155891418457,
      "memory(GiB)": 37.06,
      "step": 490,
      "token_acc": 0.8529990167158309,
      "train_speed(iter/s)": 0.175216
    },
    {
      "epoch": 0.5908683974932856,
      "grad_norm": 2.46940001428622,
      "learning_rate": 9.420906655682553e-06,
      "loss": 0.45275249481201174,
      "memory(GiB)": 37.06,
      "step": 495,
      "token_acc": 0.8271080928126768,
      "train_speed(iter/s)": 0.175432
    },
    {
      "epoch": 0.5968367651447329,
      "grad_norm": 2.3675730058319293,
      "learning_rate": 9.405427490796941e-06,
      "loss": 0.48803205490112306,
      "memory(GiB)": 37.06,
      "step": 500,
      "token_acc": 0.8432593011741406,
      "train_speed(iter/s)": 0.175539
    },
    {
      "epoch": 0.5968367651447329,
      "eval_loss": 0.4169776141643524,
      "eval_runtime": 10.9599,
      "eval_samples_per_second": 24.544,
      "eval_steps_per_second": 3.102,
      "eval_token_acc": 0.8532361484079575,
      "step": 500
    },
    {
      "epoch": 0.6028051327961802,
      "grad_norm": 2.1414646330001217,
      "learning_rate": 9.389757230844845e-06,
      "loss": 0.46323652267456056,
      "memory(GiB)": 37.06,
      "step": 505,
      "token_acc": 0.8552877345904119,
      "train_speed(iter/s)": 0.159112
    },
    {
      "epoch": 0.6087735004476276,
      "grad_norm": 2.5503273386919667,
      "learning_rate": 9.373896555558113e-06,
      "loss": 0.4701972961425781,
      "memory(GiB)": 37.06,
      "step": 510,
      "token_acc": 0.8592652620205294,
      "train_speed(iter/s)": 0.159422
    },
    {
      "epoch": 0.6147418680990749,
      "grad_norm": 2.6125713791079996,
      "learning_rate": 9.357846152928275e-06,
      "loss": 0.4990544319152832,
      "memory(GiB)": 37.06,
      "step": 515,
      "token_acc": 0.824811732065002,
      "train_speed(iter/s)": 0.159707
    },
    {
      "epoch": 0.6207102357505222,
      "grad_norm": 1.9353177630019818,
      "learning_rate": 9.341606719176695e-06,
      "loss": 0.4381883144378662,
      "memory(GiB)": 37.06,
      "step": 520,
      "token_acc": 0.867666063582321,
      "train_speed(iter/s)": 0.159909
    },
    {
      "epoch": 0.6266786034019696,
      "grad_norm": 2.3284686918748667,
      "learning_rate": 9.325178958724387e-06,
      "loss": 0.45581645965576173,
      "memory(GiB)": 37.06,
      "step": 525,
      "token_acc": 0.8706395348837209,
      "train_speed(iter/s)": 0.160206
    },
    {
      "epoch": 0.6326469710534169,
      "grad_norm": 2.2369421417810926,
      "learning_rate": 9.308563584161439e-06,
      "loss": 0.4688922882080078,
      "memory(GiB)": 37.06,
      "step": 530,
      "token_acc": 0.8338983050847457,
      "train_speed(iter/s)": 0.160549
    },
    {
      "epoch": 0.6386153387048642,
      "grad_norm": 2.4187058758316202,
      "learning_rate": 9.291761316216115e-06,
      "loss": 0.43785710334777833,
      "memory(GiB)": 37.06,
      "step": 535,
      "token_acc": 0.8175961715442666,
      "train_speed(iter/s)": 0.160901
    },
    {
      "epoch": 0.6445837063563116,
      "grad_norm": 2.11230034988461,
      "learning_rate": 9.274772883723587e-06,
      "loss": 0.4285177707672119,
      "memory(GiB)": 37.06,
      "step": 540,
      "token_acc": 0.8522423025435074,
      "train_speed(iter/s)": 0.161093
    },
    {
      "epoch": 0.6505520740077588,
      "grad_norm": 2.340278397663115,
      "learning_rate": 9.257599023594326e-06,
      "loss": 0.4503736972808838,
      "memory(GiB)": 37.06,
      "step": 545,
      "token_acc": 0.8704713049054184,
      "train_speed(iter/s)": 0.161286
    },
    {
      "epoch": 0.6565204416592062,
      "grad_norm": 2.3913667503479705,
      "learning_rate": 9.24024048078213e-06,
      "loss": 0.42584834098815916,
      "memory(GiB)": 37.06,
      "step": 550,
      "token_acc": 0.8828032979976443,
      "train_speed(iter/s)": 0.161464
    },
    {
      "epoch": 0.6624888093106536,
      "grad_norm": 2.2991966974662628,
      "learning_rate": 9.222698008251814e-06,
      "loss": 0.48091468811035154,
      "memory(GiB)": 37.06,
      "step": 555,
      "token_acc": 0.8286792452830188,
      "train_speed(iter/s)": 0.161689
    },
    {
      "epoch": 0.6684571769621008,
      "grad_norm": 2.083499198931165,
      "learning_rate": 9.204972366946546e-06,
      "loss": 0.4586004734039307,
      "memory(GiB)": 37.06,
      "step": 560,
      "token_acc": 0.8503009027081244,
      "train_speed(iter/s)": 0.16188
    },
    {
      "epoch": 0.6744255446135482,
      "grad_norm": 2.475812664409812,
      "learning_rate": 9.187064325754838e-06,
      "loss": 0.4561641693115234,
      "memory(GiB)": 37.06,
      "step": 565,
      "token_acc": 0.8384485031067596,
      "train_speed(iter/s)": 0.162054
    },
    {
      "epoch": 0.6803939122649956,
      "grad_norm": 2.4413316196832984,
      "learning_rate": 9.168974661477206e-06,
      "loss": 0.43843851089477537,
      "memory(GiB)": 37.06,
      "step": 570,
      "token_acc": 0.839965019676432,
      "train_speed(iter/s)": 0.162185
    },
    {
      "epoch": 0.6863622799164428,
      "grad_norm": 2.1737549301105075,
      "learning_rate": 9.150704158792456e-06,
      "loss": 0.4771718502044678,
      "memory(GiB)": 37.06,
      "step": 575,
      "token_acc": 0.8196035642844154,
      "train_speed(iter/s)": 0.162359
    },
    {
      "epoch": 0.6923306475678902,
      "grad_norm": 2.1356874443108342,
      "learning_rate": 9.13225361022366e-06,
      "loss": 0.48221721649169924,
      "memory(GiB)": 37.06,
      "step": 580,
      "token_acc": 0.8299897993879632,
      "train_speed(iter/s)": 0.162445
    },
    {
      "epoch": 0.6982990152193375,
      "grad_norm": 2.3220256859553077,
      "learning_rate": 9.113623816103775e-06,
      "loss": 0.4806779384613037,
      "memory(GiB)": 37.06,
      "step": 585,
      "token_acc": 0.8411007545494895,
      "train_speed(iter/s)": 0.162682
    },
    {
      "epoch": 0.7042673828707848,
      "grad_norm": 2.069813477739464,
      "learning_rate": 9.094815584540922e-06,
      "loss": 0.4947704792022705,
      "memory(GiB)": 37.06,
      "step": 590,
      "token_acc": 0.862796833773087,
      "train_speed(iter/s)": 0.162845
    },
    {
      "epoch": 0.7102357505222322,
      "grad_norm": 2.252802103709778,
      "learning_rate": 9.075829731383342e-06,
      "loss": 0.4306300163269043,
      "memory(GiB)": 37.06,
      "step": 595,
      "token_acc": 0.8425353797089894,
      "train_speed(iter/s)": 0.163154
    },
    {
      "epoch": 0.7162041181736795,
      "grad_norm": 2.241419478853809,
      "learning_rate": 9.056667080184004e-06,
      "loss": 0.4567378520965576,
      "memory(GiB)": 37.06,
      "step": 600,
      "token_acc": 0.8388354561996361,
      "train_speed(iter/s)": 0.163286
    },
    {
      "epoch": 0.7162041181736795,
      "eval_loss": 0.41334930062294006,
      "eval_runtime": 10.9312,
      "eval_samples_per_second": 24.608,
      "eval_steps_per_second": 3.11,
      "eval_token_acc": 0.8542591476224403,
      "step": 600
    },
    {
      "epoch": 0.7221724858251268,
      "grad_norm": 2.1208660287310384,
      "learning_rate": 9.037328462164866e-06,
      "loss": 0.44713678359985354,
      "memory(GiB)": 37.06,
      "step": 605,
      "token_acc": 0.8356246777796872,
      "train_speed(iter/s)": 0.151305
    },
    {
      "epoch": 0.7281408534765742,
      "grad_norm": 1.9420061515865858,
      "learning_rate": 9.01781471618085e-06,
      "loss": 0.45147147178649905,
      "memory(GiB)": 37.06,
      "step": 610,
      "token_acc": 0.8882771277816013,
      "train_speed(iter/s)": 0.151579
    },
    {
      "epoch": 0.7341092211280215,
      "grad_norm": 2.370549361627338,
      "learning_rate": 8.998126688683423e-06,
      "loss": 0.4287998199462891,
      "memory(GiB)": 37.06,
      "step": 615,
      "token_acc": 0.8318122555410691,
      "train_speed(iter/s)": 0.15183
    },
    {
      "epoch": 0.7400775887794688,
      "grad_norm": 2.003208951467392,
      "learning_rate": 8.978265233683903e-06,
      "loss": 0.4494300842285156,
      "memory(GiB)": 37.06,
      "step": 620,
      "token_acc": 0.8252328878088295,
      "train_speed(iter/s)": 0.15205
    },
    {
      "epoch": 0.7460459564309161,
      "grad_norm": 2.602367805333985,
      "learning_rate": 8.9582312127164e-06,
      "loss": 0.46652889251708984,
      "memory(GiB)": 37.06,
      "step": 625,
      "token_acc": 0.8474077428118633,
      "train_speed(iter/s)": 0.152311
    },
    {
      "epoch": 0.7520143240823635,
      "grad_norm": 2.3007477614457765,
      "learning_rate": 8.938025494800454e-06,
      "loss": 0.46235361099243166,
      "memory(GiB)": 37.06,
      "step": 630,
      "token_acc": 0.8234998744664825,
      "train_speed(iter/s)": 0.152632
    },
    {
      "epoch": 0.7579826917338108,
      "grad_norm": 2.403260011722763,
      "learning_rate": 8.917648956403338e-06,
      "loss": 0.4329329490661621,
      "memory(GiB)": 37.06,
      "step": 635,
      "token_acc": 0.8512756689483509,
      "train_speed(iter/s)": 0.152969
    },
    {
      "epoch": 0.7639510593852581,
      "grad_norm": 1.8459463363591184,
      "learning_rate": 8.897102481402031e-06,
      "loss": 0.45981664657592775,
      "memory(GiB)": 37.06,
      "step": 640,
      "token_acc": 0.8598321614878657,
      "train_speed(iter/s)": 0.153182
    },
    {
      "epoch": 0.7699194270367055,
      "grad_norm": 2.0204814112895044,
      "learning_rate": 8.876386961044892e-06,
      "loss": 0.46657752990722656,
      "memory(GiB)": 37.06,
      "step": 645,
      "token_acc": 0.8745874587458746,
      "train_speed(iter/s)": 0.153345
    },
    {
      "epoch": 0.7758877946881528,
      "grad_norm": 1.8481808083298177,
      "learning_rate": 8.855503293912987e-06,
      "loss": 0.4649078369140625,
      "memory(GiB)": 37.06,
      "step": 650,
      "token_acc": 0.8592820512820513,
      "train_speed(iter/s)": 0.153498
    },
    {
      "epoch": 0.7818561623396001,
      "grad_norm": 2.2884914044841698,
      "learning_rate": 8.834452385881121e-06,
      "loss": 0.4653633117675781,
      "memory(GiB)": 37.06,
      "step": 655,
      "token_acc": 0.8515602216389618,
      "train_speed(iter/s)": 0.153659
    },
    {
      "epoch": 0.7878245299910475,
      "grad_norm": 2.173340273942357,
      "learning_rate": 8.813235150078532e-06,
      "loss": 0.46648712158203126,
      "memory(GiB)": 37.06,
      "step": 660,
      "token_acc": 0.8156269959548648,
      "train_speed(iter/s)": 0.153953
    },
    {
      "epoch": 0.7937928976424948,
      "grad_norm": 2.2191296614587563,
      "learning_rate": 8.791852506849301e-06,
      "loss": 0.45751609802246096,
      "memory(GiB)": 37.06,
      "step": 665,
      "token_acc": 0.8260312580066616,
      "train_speed(iter/s)": 0.154161
    },
    {
      "epoch": 0.7997612652939421,
      "grad_norm": 2.2870388856485335,
      "learning_rate": 8.770305383712407e-06,
      "loss": 0.4709470748901367,
      "memory(GiB)": 37.06,
      "step": 670,
      "token_acc": 0.842337607735968,
      "train_speed(iter/s)": 0.154453
    },
    {
      "epoch": 0.8057296329453895,
      "grad_norm": 2.3046312751781866,
      "learning_rate": 8.748594715321512e-06,
      "loss": 0.44265017509460447,
      "memory(GiB)": 37.06,
      "step": 675,
      "token_acc": 0.8602195071443363,
      "train_speed(iter/s)": 0.154677
    },
    {
      "epoch": 0.8116980005968367,
      "grad_norm": 2.2464744707673985,
      "learning_rate": 8.726721443424409e-06,
      "loss": 0.4592324733734131,
      "memory(GiB)": 37.06,
      "step": 680,
      "token_acc": 0.8654945054945055,
      "train_speed(iter/s)": 0.154905
    },
    {
      "epoch": 0.8176663682482841,
      "grad_norm": 2.194092144648434,
      "learning_rate": 8.704686516822177e-06,
      "loss": 0.43160429000854494,
      "memory(GiB)": 37.06,
      "step": 685,
      "token_acc": 0.8649193548387096,
      "train_speed(iter/s)": 0.155078
    },
    {
      "epoch": 0.8236347358997315,
      "grad_norm": 2.247411516392796,
      "learning_rate": 8.682490891328016e-06,
      "loss": 0.45626983642578123,
      "memory(GiB)": 37.06,
      "step": 690,
      "token_acc": 0.8643364928909952,
      "train_speed(iter/s)": 0.155279
    },
    {
      "epoch": 0.8296031035511787,
      "grad_norm": 2.035754411138357,
      "learning_rate": 8.660135529725799e-06,
      "loss": 0.4315452575683594,
      "memory(GiB)": 37.06,
      "step": 695,
      "token_acc": 0.8554044380816035,
      "train_speed(iter/s)": 0.155502
    },
    {
      "epoch": 0.8355714712026261,
      "grad_norm": 2.292286762424394,
      "learning_rate": 8.6376214017283e-06,
      "loss": 0.4535685539245605,
      "memory(GiB)": 37.06,
      "step": 700,
      "token_acc": 0.833079268292683,
      "train_speed(iter/s)": 0.155636
    },
    {
      "epoch": 0.8355714712026261,
      "eval_loss": 0.4100053906440735,
      "eval_runtime": 10.9163,
      "eval_samples_per_second": 24.642,
      "eval_steps_per_second": 3.115,
      "eval_token_acc": 0.8548802542883762,
      "step": 700
    },
    {
      "epoch": 0.8415398388540735,
      "grad_norm": 2.6314360636405714,
      "learning_rate": 8.61494948393513e-06,
      "loss": 0.4539949417114258,
      "memory(GiB)": 37.06,
      "step": 705,
      "token_acc": 0.8583042973286876,
      "train_speed(iter/s)": 0.146478
    },
    {
      "epoch": 0.8475082065055207,
      "grad_norm": 2.1848010999728715,
      "learning_rate": 8.592120759790383e-06,
      "loss": 0.46171207427978517,
      "memory(GiB)": 37.06,
      "step": 710,
      "token_acc": 0.8417105263157895,
      "train_speed(iter/s)": 0.146671
    },
    {
      "epoch": 0.8534765741569681,
      "grad_norm": 2.447774461275868,
      "learning_rate": 8.56913621953997e-06,
      "loss": 0.4798592567443848,
      "memory(GiB)": 37.06,
      "step": 715,
      "token_acc": 0.8562048588312541,
      "train_speed(iter/s)": 0.146953
    },
    {
      "epoch": 0.8594449418084154,
      "grad_norm": 2.596951485691162,
      "learning_rate": 8.545996860188668e-06,
      "loss": 0.4231537342071533,
      "memory(GiB)": 37.06,
      "step": 720,
      "token_acc": 0.831799700406591,
      "train_speed(iter/s)": 0.147232
    },
    {
      "epoch": 0.8654133094598627,
      "grad_norm": 2.0232163854750027,
      "learning_rate": 8.522703685456866e-06,
      "loss": 0.44301156997680663,
      "memory(GiB)": 37.06,
      "step": 725,
      "token_acc": 0.8794139744552968,
      "train_speed(iter/s)": 0.1475
    },
    {
      "epoch": 0.8713816771113101,
      "grad_norm": 2.281907577430269,
      "learning_rate": 8.49925770573704e-06,
      "loss": 0.46319947242736814,
      "memory(GiB)": 37.06,
      "step": 730,
      "token_acc": 0.8430570505920344,
      "train_speed(iter/s)": 0.147765
    },
    {
      "epoch": 0.8773500447627574,
      "grad_norm": 2.190179810988922,
      "learning_rate": 8.475659938049912e-06,
      "loss": 0.4825079917907715,
      "memory(GiB)": 37.06,
      "step": 735,
      "token_acc": 0.839588377723971,
      "train_speed(iter/s)": 0.147996
    },
    {
      "epoch": 0.8833184124142047,
      "grad_norm": 2.014804370593861,
      "learning_rate": 8.45191140600034e-06,
      "loss": 0.454302978515625,
      "memory(GiB)": 37.06,
      "step": 740,
      "token_acc": 0.8007774538386784,
      "train_speed(iter/s)": 0.148279
    },
    {
      "epoch": 0.8892867800656521,
      "grad_norm": 2.1256355584342077,
      "learning_rate": 8.42801313973292e-06,
      "loss": 0.4445801258087158,
      "memory(GiB)": 37.06,
      "step": 745,
      "token_acc": 0.846286205907657,
      "train_speed(iter/s)": 0.148536
    },
    {
      "epoch": 0.8952551477170994,
      "grad_norm": 2.6544295779283575,
      "learning_rate": 8.403966175887293e-06,
      "loss": 0.4630784511566162,
      "memory(GiB)": 37.06,
      "step": 750,
      "token_acc": 0.8537764350453172,
      "train_speed(iter/s)": 0.148704
    },
    {
      "epoch": 0.9012235153685467,
      "grad_norm": 2.4745309667627255,
      "learning_rate": 8.379771557553184e-06,
      "loss": 0.43903446197509766,
      "memory(GiB)": 37.06,
      "step": 755,
      "token_acc": 0.8682237600922722,
      "train_speed(iter/s)": 0.148945
    },
    {
      "epoch": 0.907191883019994,
      "grad_norm": 2.167884085714607,
      "learning_rate": 8.355430334225159e-06,
      "loss": 0.445455265045166,
      "memory(GiB)": 37.06,
      "step": 760,
      "token_acc": 0.852589641434263,
      "train_speed(iter/s)": 0.149189
    },
    {
      "epoch": 0.9131602506714414,
      "grad_norm": 2.3516013470748116,
      "learning_rate": 8.330943561757092e-06,
      "loss": 0.44769630432128904,
      "memory(GiB)": 37.06,
      "step": 765,
      "token_acc": 0.8217955651703623,
      "train_speed(iter/s)": 0.149338
    },
    {
      "epoch": 0.9191286183228887,
      "grad_norm": 2.0619205640970506,
      "learning_rate": 8.30631230231637e-06,
      "loss": 0.46817874908447266,
      "memory(GiB)": 37.06,
      "step": 770,
      "token_acc": 0.8363870967741935,
      "train_speed(iter/s)": 0.149487
    },
    {
      "epoch": 0.925096985974336,
      "grad_norm": 2.3440589362137993,
      "learning_rate": 8.281537624337823e-06,
      "loss": 0.4982964038848877,
      "memory(GiB)": 37.06,
      "step": 775,
      "token_acc": 0.8594432314410481,
      "train_speed(iter/s)": 0.149779
    },
    {
      "epoch": 0.9310653536257834,
      "grad_norm": 2.0757541904974097,
      "learning_rate": 8.256620602477372e-06,
      "loss": 0.4509378433227539,
      "memory(GiB)": 37.06,
      "step": 780,
      "token_acc": 0.8259721555448872,
      "train_speed(iter/s)": 0.149971
    },
    {
      "epoch": 0.9370337212772307,
      "grad_norm": 2.086378932611534,
      "learning_rate": 8.231562317565412e-06,
      "loss": 0.43694629669189455,
      "memory(GiB)": 37.06,
      "step": 785,
      "token_acc": 0.856384262611634,
      "train_speed(iter/s)": 0.150204
    },
    {
      "epoch": 0.943002088928678,
      "grad_norm": 2.308538899901496,
      "learning_rate": 8.206363856559935e-06,
      "loss": 0.4430408477783203,
      "memory(GiB)": 37.06,
      "step": 790,
      "token_acc": 0.8422222222222222,
      "train_speed(iter/s)": 0.15035
    },
    {
      "epoch": 0.9489704565801254,
      "grad_norm": 1.8314796079076852,
      "learning_rate": 8.181026312499383e-06,
      "loss": 0.44437146186828613,
      "memory(GiB)": 37.06,
      "step": 795,
      "token_acc": 0.8529804865009356,
      "train_speed(iter/s)": 0.150549
    },
    {
      "epoch": 0.9549388242315726,
      "grad_norm": 2.2397424826021792,
      "learning_rate": 8.155550784455224e-06,
      "loss": 0.4815809726715088,
      "memory(GiB)": 37.06,
      "step": 800,
      "token_acc": 0.8588266107909901,
      "train_speed(iter/s)": 0.150753
    },
    {
      "epoch": 0.9549388242315726,
      "eval_loss": 0.4058806300163269,
      "eval_runtime": 11.0737,
      "eval_samples_per_second": 24.292,
      "eval_steps_per_second": 3.07,
      "eval_token_acc": 0.8572368060503096,
      "step": 800
    }
  ],
  "logging_steps": 5,
  "max_steps": 2511,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 88119181914112.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}