{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.46793746342279074,
  "eval_steps": 500,
  "global_step": 5597,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000836050497450046,
      "grad_norm": 5.8952789306640625,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 1.5346,
      "step": 10
    },
    {
      "epoch": 0.001672100994900092,
      "grad_norm": 2.239838123321533,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 1.3887,
      "step": 20
    },
    {
      "epoch": 0.002508151492350138,
      "grad_norm": 1.8720221519470215,
      "learning_rate": 3.342134831460675e-06,
      "loss": 1.2524,
      "step": 30
    },
    {
      "epoch": 0.003344201989800184,
      "grad_norm": 1.1449085474014282,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 1.0681,
      "step": 40
    },
    {
      "epoch": 0.0041802524872502295,
      "grad_norm": 1.360563039779663,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.8906,
      "step": 50
    },
    {
      "epoch": 0.005016302984700276,
      "grad_norm": 0.8737388849258423,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.8052,
      "step": 60
    },
    {
      "epoch": 0.005852353482150322,
      "grad_norm": 0.50184565782547,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.7827,
      "step": 70
    },
    {
      "epoch": 0.006688403979600368,
      "grad_norm": 0.4593908190727234,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.7758,
      "step": 80
    },
    {
      "epoch": 0.007524454477050414,
      "grad_norm": 0.5259537100791931,
      "learning_rate": 1.005e-05,
      "loss": 0.7545,
      "step": 90
    },
    {
      "epoch": 0.008360504974500459,
      "grad_norm": 0.5117310285568237,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.7779,
      "step": 100
    },
    {
      "epoch": 0.009196555471950505,
      "grad_norm": 0.46054765582084656,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.7623,
      "step": 110
    },
    {
      "epoch": 0.010032605969400551,
      "grad_norm": 0.6909846067428589,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.7352,
      "step": 120
    },
    {
      "epoch": 0.010868656466850597,
      "grad_norm": 0.4003996253013611,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.7429,
      "step": 130
    },
    {
      "epoch": 0.011704706964300643,
      "grad_norm": 0.47065919637680054,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.7416,
      "step": 140
    },
    {
      "epoch": 0.01254075746175069,
      "grad_norm": 0.5040171146392822,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.7821,
      "step": 150
    },
    {
      "epoch": 0.013376807959200735,
      "grad_norm": 0.396039217710495,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.6167,
      "step": 160
    },
    {
      "epoch": 0.014212858456650782,
      "grad_norm": 0.42775487899780273,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.703,
      "step": 170
    },
    {
      "epoch": 0.015048908954100828,
      "grad_norm": 0.5157128572463989,
      "learning_rate": 2e-05,
      "loss": 0.7219,
      "step": 180
    },
    {
      "epoch": 0.015884959451550874,
      "grad_norm": 0.5191335082054138,
      "learning_rate": 2e-05,
      "loss": 0.6895,
      "step": 190
    },
    {
      "epoch": 0.016721009949000918,
      "grad_norm": 0.44552481174468994,
      "learning_rate": 2e-05,
      "loss": 0.7773,
      "step": 200
    },
    {
      "epoch": 0.017557060446450966,
      "grad_norm": 0.3954372704029083,
      "learning_rate": 2e-05,
      "loss": 0.7441,
      "step": 210
    },
    {
      "epoch": 0.01839311094390101,
      "grad_norm": 0.45830389857292175,
      "learning_rate": 2e-05,
      "loss": 0.7504,
      "step": 220
    },
    {
      "epoch": 0.019229161441351058,
      "grad_norm": 0.4249721169471741,
      "learning_rate": 2e-05,
      "loss": 0.7186,
      "step": 230
    },
    {
      "epoch": 0.020065211938801102,
      "grad_norm": 0.4513215720653534,
      "learning_rate": 2e-05,
      "loss": 0.745,
      "step": 240
    },
    {
      "epoch": 0.02090126243625115,
      "grad_norm": 0.43894580006599426,
      "learning_rate": 2e-05,
      "loss": 0.76,
      "step": 250
    },
    {
      "epoch": 0.021737312933701194,
      "grad_norm": 7.123316287994385,
      "learning_rate": 2e-05,
      "loss": 0.7635,
      "step": 260
    },
    {
      "epoch": 0.022573363431151242,
      "grad_norm": 0.4213659465312958,
      "learning_rate": 2e-05,
      "loss": 0.7474,
      "step": 270
    },
    {
      "epoch": 0.023409413928601287,
      "grad_norm": 0.4028798043727875,
      "learning_rate": 2e-05,
      "loss": 0.7168,
      "step": 280
    },
    {
      "epoch": 0.024245464426051334,
      "grad_norm": 0.3978917598724365,
      "learning_rate": 2e-05,
      "loss": 0.7272,
      "step": 290
    },
    {
      "epoch": 0.02508151492350138,
      "grad_norm": 0.45915454626083374,
      "learning_rate": 2e-05,
      "loss": 0.7791,
      "step": 300
    },
    {
      "epoch": 0.025917565420951427,
      "grad_norm": 0.44984596967697144,
      "learning_rate": 2e-05,
      "loss": 0.7934,
      "step": 310
    },
    {
      "epoch": 0.02675361591840147,
      "grad_norm": 0.43878060579299927,
      "learning_rate": 2e-05,
      "loss": 0.7644,
      "step": 320
    },
    {
      "epoch": 0.02758966641585152,
      "grad_norm": 0.4639083743095398,
      "learning_rate": 2e-05,
      "loss": 0.7801,
      "step": 330
    },
    {
      "epoch": 0.028425716913301563,
      "grad_norm": 0.43591466546058655,
      "learning_rate": 2e-05,
      "loss": 0.7261,
      "step": 340
    },
    {
      "epoch": 0.02926176741075161,
      "grad_norm": 0.5230687260627747,
      "learning_rate": 2e-05,
      "loss": 0.7732,
      "step": 350
    },
    {
      "epoch": 0.030097817908201655,
      "grad_norm": 0.4230197072029114,
      "learning_rate": 2e-05,
      "loss": 0.7576,
      "step": 360
    },
    {
      "epoch": 0.030933868405651703,
      "grad_norm": 0.44197097420692444,
      "learning_rate": 2e-05,
      "loss": 0.7232,
      "step": 370
    },
    {
      "epoch": 0.03176991890310175,
      "grad_norm": 0.4144206643104553,
      "learning_rate": 2e-05,
      "loss": 0.7316,
      "step": 380
    },
    {
      "epoch": 0.03260596940055179,
      "grad_norm": 0.4325674772262573,
      "learning_rate": 2e-05,
      "loss": 0.7103,
      "step": 390
    },
    {
      "epoch": 0.033442019898001836,
      "grad_norm": 0.3876762390136719,
      "learning_rate": 2e-05,
      "loss": 0.7396,
      "step": 400
    },
    {
      "epoch": 0.03427807039545189,
      "grad_norm": 0.4181121587753296,
      "learning_rate": 2e-05,
      "loss": 0.7415,
      "step": 410
    },
    {
      "epoch": 0.03511412089290193,
      "grad_norm": 0.6723706722259521,
      "learning_rate": 2e-05,
      "loss": 0.7408,
      "step": 420
    },
    {
      "epoch": 0.035950171390351976,
      "grad_norm": 0.42723605036735535,
      "learning_rate": 2e-05,
      "loss": 0.7381,
      "step": 430
    },
    {
      "epoch": 0.03678622188780202,
      "grad_norm": 0.40303024649620056,
      "learning_rate": 2e-05,
      "loss": 0.7108,
      "step": 440
    },
    {
      "epoch": 0.03762227238525207,
      "grad_norm": 0.4335564076900482,
      "learning_rate": 2e-05,
      "loss": 0.7368,
      "step": 450
    },
    {
      "epoch": 0.038458322882702116,
      "grad_norm": 0.4370777904987335,
      "learning_rate": 2e-05,
      "loss": 0.7813,
      "step": 460
    },
    {
      "epoch": 0.03929437338015216,
      "grad_norm": 0.4104174077510834,
      "learning_rate": 2e-05,
      "loss": 0.7359,
      "step": 470
    },
    {
      "epoch": 0.040130423877602205,
      "grad_norm": 0.4259282946586609,
      "learning_rate": 2e-05,
      "loss": 0.7206,
      "step": 480
    },
    {
      "epoch": 0.040966474375052256,
      "grad_norm": 0.39744478464126587,
      "learning_rate": 2e-05,
      "loss": 0.7245,
      "step": 490
    },
    {
      "epoch": 0.0418025248725023,
      "grad_norm": 0.42015591263771057,
      "learning_rate": 2e-05,
      "loss": 0.7416,
      "step": 500
    },
    {
      "epoch": 0.042638575369952345,
      "grad_norm": 0.379894882440567,
      "learning_rate": 2e-05,
      "loss": 0.7315,
      "step": 510
    },
    {
      "epoch": 0.04347462586740239,
      "grad_norm": 0.4095825254917145,
      "learning_rate": 2e-05,
      "loss": 0.7185,
      "step": 520
    },
    {
      "epoch": 0.04431067636485244,
      "grad_norm": 0.4044616222381592,
      "learning_rate": 2e-05,
      "loss": 0.7443,
      "step": 530
    },
    {
      "epoch": 0.045146726862302484,
      "grad_norm": 0.4145281910896301,
      "learning_rate": 2e-05,
      "loss": 0.7009,
      "step": 540
    },
    {
      "epoch": 0.04598277735975253,
      "grad_norm": 0.6660082936286926,
      "learning_rate": 2e-05,
      "loss": 0.7139,
      "step": 550
    },
    {
      "epoch": 0.04681882785720257,
      "grad_norm": 0.39196470379829407,
      "learning_rate": 2e-05,
      "loss": 0.6958,
      "step": 560
    },
    {
      "epoch": 0.047654878354652624,
      "grad_norm": 0.39314568042755127,
      "learning_rate": 2e-05,
      "loss": 0.7562,
      "step": 570
    },
    {
      "epoch": 0.04849092885210267,
      "grad_norm": 0.4184809923171997,
      "learning_rate": 2e-05,
      "loss": 0.7281,
      "step": 580
    },
    {
      "epoch": 0.04932697934955271,
      "grad_norm": 0.4390700161457062,
      "learning_rate": 2e-05,
      "loss": 0.7206,
      "step": 590
    },
    {
      "epoch": 0.05016302984700276,
      "grad_norm": 0.39948198199272156,
      "learning_rate": 2e-05,
      "loss": 0.7311,
      "step": 600
    },
    {
      "epoch": 0.0509990803444528,
      "grad_norm": 0.3739895820617676,
      "learning_rate": 2e-05,
      "loss": 0.7089,
      "step": 610
    },
    {
      "epoch": 0.05183513084190285,
      "grad_norm": 0.37271761894226074,
      "learning_rate": 2e-05,
      "loss": 0.7092,
      "step": 620
    },
    {
      "epoch": 0.0526711813393529,
      "grad_norm": 0.404062420129776,
      "learning_rate": 2e-05,
      "loss": 0.6953,
      "step": 630
    },
    {
      "epoch": 0.05350723183680294,
      "grad_norm": 0.4164421260356903,
      "learning_rate": 2e-05,
      "loss": 0.7431,
      "step": 640
    },
    {
      "epoch": 0.054343282334252986,
      "grad_norm": 0.4303136169910431,
      "learning_rate": 2e-05,
      "loss": 0.7299,
      "step": 650
    },
    {
      "epoch": 0.05517933283170304,
      "grad_norm": 0.4234829843044281,
      "learning_rate": 2e-05,
      "loss": 0.7442,
      "step": 660
    },
    {
      "epoch": 0.05601538332915308,
      "grad_norm": 0.3876648545265198,
      "learning_rate": 2e-05,
      "loss": 0.7146,
      "step": 670
    },
    {
      "epoch": 0.056851433826603126,
      "grad_norm": 0.4167863726615906,
      "learning_rate": 2e-05,
      "loss": 0.7098,
      "step": 680
    },
    {
      "epoch": 0.05768748432405317,
      "grad_norm": 0.44938138127326965,
      "learning_rate": 2e-05,
      "loss": 0.7457,
      "step": 690
    },
    {
      "epoch": 0.05852353482150322,
      "grad_norm": 0.4265012741088867,
      "learning_rate": 2e-05,
      "loss": 0.7091,
      "step": 700
    },
    {
      "epoch": 0.059359585318953266,
      "grad_norm": 0.42332151532173157,
      "learning_rate": 2e-05,
      "loss": 0.7499,
      "step": 710
    },
    {
      "epoch": 0.06019563581640331,
      "grad_norm": 0.4157474637031555,
      "learning_rate": 2e-05,
      "loss": 0.6786,
      "step": 720
    },
    {
      "epoch": 0.061031686313853355,
      "grad_norm": 0.370686799287796,
      "learning_rate": 2e-05,
      "loss": 0.7323,
      "step": 730
    },
    {
      "epoch": 0.061867736811303406,
      "grad_norm": 0.3953830599784851,
      "learning_rate": 2e-05,
      "loss": 0.7309,
      "step": 740
    },
    {
      "epoch": 0.06270378730875345,
      "grad_norm": 0.40635451674461365,
      "learning_rate": 2e-05,
      "loss": 0.732,
      "step": 750
    },
    {
      "epoch": 0.0635398378062035,
      "grad_norm": 0.4116203486919403,
      "learning_rate": 2e-05,
      "loss": 0.718,
      "step": 760
    },
    {
      "epoch": 0.06437588830365354,
      "grad_norm": 0.3634946048259735,
      "learning_rate": 2e-05,
      "loss": 0.679,
      "step": 770
    },
    {
      "epoch": 0.06521193880110358,
      "grad_norm": 0.4082890450954437,
      "learning_rate": 2e-05,
      "loss": 0.687,
      "step": 780
    },
    {
      "epoch": 0.06604798929855363,
      "grad_norm": 0.4118865728378296,
      "learning_rate": 2e-05,
      "loss": 0.7137,
      "step": 790
    },
    {
      "epoch": 0.06688403979600367,
      "grad_norm": 0.40019503235816956,
      "learning_rate": 2e-05,
      "loss": 0.7234,
      "step": 800
    },
    {
      "epoch": 0.06772009029345373,
      "grad_norm": 0.41497135162353516,
      "learning_rate": 2e-05,
      "loss": 0.7026,
      "step": 810
    },
    {
      "epoch": 0.06855614079090377,
      "grad_norm": 0.37976041436195374,
      "learning_rate": 2e-05,
      "loss": 0.6984,
      "step": 820
    },
    {
      "epoch": 0.06939219128835382,
      "grad_norm": 0.4239822030067444,
      "learning_rate": 2e-05,
      "loss": 0.7123,
      "step": 830
    },
    {
      "epoch": 0.07022824178580386,
      "grad_norm": 0.40763989090919495,
      "learning_rate": 2e-05,
      "loss": 0.6923,
      "step": 840
    },
    {
      "epoch": 0.07106429228325391,
      "grad_norm": 0.40808385610580444,
      "learning_rate": 2e-05,
      "loss": 0.7095,
      "step": 850
    },
    {
      "epoch": 0.07190034278070395,
      "grad_norm": 0.44947150349617004,
      "learning_rate": 2e-05,
      "loss": 0.7188,
      "step": 860
    },
    {
      "epoch": 0.072736393278154,
      "grad_norm": 0.3837183713912964,
      "learning_rate": 2e-05,
      "loss": 0.7442,
      "step": 870
    },
    {
      "epoch": 0.07357244377560404,
      "grad_norm": 0.4867953956127167,
      "learning_rate": 2e-05,
      "loss": 0.717,
      "step": 880
    },
    {
      "epoch": 0.0744084942730541,
      "grad_norm": 0.4862518310546875,
      "learning_rate": 2e-05,
      "loss": 0.6994,
      "step": 890
    },
    {
      "epoch": 0.07524454477050414,
      "grad_norm": 0.4381377398967743,
      "learning_rate": 2e-05,
      "loss": 0.7172,
      "step": 900
    },
    {
      "epoch": 0.07608059526795419,
      "grad_norm": 0.44232502579689026,
      "learning_rate": 2e-05,
      "loss": 0.7078,
      "step": 910
    },
    {
      "epoch": 0.07691664576540423,
      "grad_norm": 0.39795035123825073,
      "learning_rate": 2e-05,
      "loss": 0.7007,
      "step": 920
    },
    {
      "epoch": 0.07775269626285428,
      "grad_norm": 0.3873278796672821,
      "learning_rate": 2e-05,
      "loss": 0.7302,
      "step": 930
    },
    {
      "epoch": 0.07858874676030432,
      "grad_norm": 0.4043973684310913,
      "learning_rate": 2e-05,
      "loss": 0.7497,
      "step": 940
    },
    {
      "epoch": 0.07942479725775436,
      "grad_norm": 0.43075886368751526,
      "learning_rate": 2e-05,
      "loss": 0.726,
      "step": 950
    },
    {
      "epoch": 0.08026084775520441,
      "grad_norm": 0.4056244492530823,
      "learning_rate": 2e-05,
      "loss": 0.6807,
      "step": 960
    },
    {
      "epoch": 0.08109689825265445,
      "grad_norm": 0.4083726406097412,
      "learning_rate": 2e-05,
      "loss": 0.718,
      "step": 970
    },
    {
      "epoch": 0.08193294875010451,
      "grad_norm": 0.4576243460178375,
      "learning_rate": 2e-05,
      "loss": 0.6996,
      "step": 980
    },
    {
      "epoch": 0.08276899924755456,
      "grad_norm": 0.41480258107185364,
      "learning_rate": 2e-05,
      "loss": 0.6787,
      "step": 990
    },
    {
      "epoch": 0.0836050497450046,
      "grad_norm": 0.38216501474380493,
      "learning_rate": 2e-05,
      "loss": 0.7211,
      "step": 1000
    },
    {
      "epoch": 0.08444110024245464,
      "grad_norm": 0.37291425466537476,
      "learning_rate": 2e-05,
      "loss": 0.7013,
      "step": 1010
    },
    {
      "epoch": 0.08527715073990469,
      "grad_norm": 0.44588473439216614,
      "learning_rate": 2e-05,
      "loss": 0.7362,
      "step": 1020
    },
    {
      "epoch": 0.08611320123735473,
      "grad_norm": 0.38740336894989014,
      "learning_rate": 2e-05,
      "loss": 0.7268,
      "step": 1030
    },
    {
      "epoch": 0.08694925173480478,
      "grad_norm": 0.3775883913040161,
      "learning_rate": 2e-05,
      "loss": 0.693,
      "step": 1040
    },
    {
      "epoch": 0.08778530223225482,
      "grad_norm": 0.4709722399711609,
      "learning_rate": 2e-05,
      "loss": 0.731,
      "step": 1050
    },
    {
      "epoch": 0.08862135272970488,
      "grad_norm": 0.4146833121776581,
      "learning_rate": 2e-05,
      "loss": 0.7448,
      "step": 1060
    },
    {
      "epoch": 0.08945740322715492,
      "grad_norm": 0.39623764157295227,
      "learning_rate": 2e-05,
      "loss": 0.6927,
      "step": 1070
    },
    {
      "epoch": 0.09029345372460497,
      "grad_norm": 0.3726375102996826,
      "learning_rate": 2e-05,
      "loss": 0.7264,
      "step": 1080
    },
    {
      "epoch": 0.09112950422205501,
      "grad_norm": 0.3898238241672516,
      "learning_rate": 2e-05,
      "loss": 0.6968,
      "step": 1090
    },
    {
      "epoch": 0.09196555471950506,
      "grad_norm": 0.3660300076007843,
      "learning_rate": 2e-05,
      "loss": 0.7464,
      "step": 1100
    },
    {
      "epoch": 0.0928016052169551,
      "grad_norm": 0.43982595205307007,
      "learning_rate": 2e-05,
      "loss": 0.7031,
      "step": 1110
    },
    {
      "epoch": 0.09363765571440515,
      "grad_norm": 0.41887497901916504,
      "learning_rate": 2e-05,
      "loss": 0.6931,
      "step": 1120
    },
    {
      "epoch": 0.09447370621185519,
      "grad_norm": 0.3669204115867615,
      "learning_rate": 2e-05,
      "loss": 0.7256,
      "step": 1130
    },
    {
      "epoch": 0.09530975670930525,
      "grad_norm": 0.39729389548301697,
      "learning_rate": 2e-05,
      "loss": 0.7253,
      "step": 1140
    },
    {
      "epoch": 0.0961458072067553,
      "grad_norm": 0.3818066418170929,
      "learning_rate": 2e-05,
      "loss": 0.6842,
      "step": 1150
    },
    {
      "epoch": 0.09698185770420534,
      "grad_norm": 0.3896878659725189,
      "learning_rate": 2e-05,
      "loss": 0.6928,
      "step": 1160
    },
    {
      "epoch": 0.09781790820165538,
      "grad_norm": 0.37976863980293274,
      "learning_rate": 2e-05,
      "loss": 0.7184,
      "step": 1170
    },
    {
      "epoch": 0.09865395869910543,
      "grad_norm": 0.37471631169319153,
      "learning_rate": 2e-05,
      "loss": 0.6839,
      "step": 1180
    },
    {
      "epoch": 0.09949000919655547,
      "grad_norm": 0.3750368356704712,
      "learning_rate": 2e-05,
      "loss": 0.693,
      "step": 1190
    },
    {
      "epoch": 0.10032605969400551,
      "grad_norm": 0.37220221757888794,
      "learning_rate": 2e-05,
      "loss": 0.7175,
      "step": 1200
    },
    {
      "epoch": 0.10116211019145556,
      "grad_norm": 0.407367080450058,
      "learning_rate": 2e-05,
      "loss": 0.706,
      "step": 1210
    },
    {
      "epoch": 0.1019981606889056,
      "grad_norm": 0.35999026894569397,
      "learning_rate": 2e-05,
      "loss": 0.7197,
      "step": 1220
    },
    {
      "epoch": 0.10283421118635566,
      "grad_norm": 0.36594730615615845,
      "learning_rate": 2e-05,
      "loss": 0.6944,
      "step": 1230
    },
    {
      "epoch": 0.1036702616838057,
      "grad_norm": 0.3911300301551819,
      "learning_rate": 2e-05,
      "loss": 0.7117,
      "step": 1240
    },
    {
      "epoch": 0.10450631218125575,
      "grad_norm": 0.4042055010795593,
      "learning_rate": 2e-05,
      "loss": 0.7157,
      "step": 1250
    },
    {
      "epoch": 0.1053423626787058,
      "grad_norm": 0.4203055500984192,
      "learning_rate": 2e-05,
      "loss": 0.7426,
      "step": 1260
    },
    {
      "epoch": 0.10617841317615584,
      "grad_norm": 0.3811381459236145,
      "learning_rate": 2e-05,
      "loss": 0.699,
      "step": 1270
    },
    {
      "epoch": 0.10701446367360588,
      "grad_norm": 0.35855522751808167,
      "learning_rate": 2e-05,
      "loss": 0.7526,
      "step": 1280
    },
    {
      "epoch": 0.10785051417105593,
      "grad_norm": 0.3886844217777252,
      "learning_rate": 2e-05,
      "loss": 0.7052,
      "step": 1290
    },
    {
      "epoch": 0.10868656466850597,
      "grad_norm": 0.3775457739830017,
      "learning_rate": 2e-05,
      "loss": 0.71,
      "step": 1300
    },
    {
      "epoch": 0.10952261516595603,
      "grad_norm": 0.4011971652507782,
      "learning_rate": 2e-05,
      "loss": 0.6869,
      "step": 1310
    },
    {
      "epoch": 0.11035866566340607,
      "grad_norm": 0.40073534846305847,
      "learning_rate": 2e-05,
      "loss": 0.7294,
      "step": 1320
    },
    {
      "epoch": 0.11119471616085612,
      "grad_norm": 0.6666777729988098,
      "learning_rate": 2e-05,
      "loss": 0.7135,
      "step": 1330
    },
    {
      "epoch": 0.11203076665830616,
      "grad_norm": 0.40671786665916443,
      "learning_rate": 2e-05,
      "loss": 0.7055,
      "step": 1340
    },
    {
      "epoch": 0.11286681715575621,
      "grad_norm": 0.35212790966033936,
      "learning_rate": 2e-05,
      "loss": 0.7068,
      "step": 1350
    },
    {
      "epoch": 0.11370286765320625,
      "grad_norm": 0.39304518699645996,
      "learning_rate": 2e-05,
      "loss": 0.7132,
      "step": 1360
    },
    {
      "epoch": 0.1145389181506563,
      "grad_norm": 0.40921252965927124,
      "learning_rate": 2e-05,
      "loss": 0.7149,
      "step": 1370
    },
    {
      "epoch": 0.11537496864810634,
      "grad_norm": 0.3641948401927948,
      "learning_rate": 2e-05,
      "loss": 0.6947,
      "step": 1380
    },
    {
      "epoch": 0.11621101914555639,
      "grad_norm": 0.3776053786277771,
      "learning_rate": 2e-05,
      "loss": 0.7098,
      "step": 1390
    },
    {
      "epoch": 0.11704706964300644,
      "grad_norm": 0.40454384684562683,
      "learning_rate": 2e-05,
      "loss": 0.6876,
      "step": 1400
    },
    {
      "epoch": 0.11788312014045649,
      "grad_norm": 0.3694128692150116,
      "learning_rate": 2e-05,
      "loss": 0.7112,
      "step": 1410
    },
    {
      "epoch": 0.11871917063790653,
      "grad_norm": 0.4031735956668854,
      "learning_rate": 2e-05,
      "loss": 0.6743,
      "step": 1420
    },
    {
      "epoch": 0.11955522113535658,
      "grad_norm": 0.36599618196487427,
      "learning_rate": 2e-05,
      "loss": 0.7177,
      "step": 1430
    },
    {
      "epoch": 0.12039127163280662,
      "grad_norm": 0.3794446587562561,
      "learning_rate": 2e-05,
      "loss": 0.7672,
      "step": 1440
    },
    {
      "epoch": 0.12122732213025667,
      "grad_norm": 0.42689165472984314,
      "learning_rate": 2e-05,
      "loss": 0.6943,
      "step": 1450
    },
    {
      "epoch": 0.12206337262770671,
      "grad_norm": 0.39176973700523376,
      "learning_rate": 2e-05,
      "loss": 0.7454,
      "step": 1460
    },
    {
      "epoch": 0.12289942312515675,
      "grad_norm": 0.38585641980171204,
      "learning_rate": 2e-05,
      "loss": 0.6796,
      "step": 1470
    },
    {
      "epoch": 0.12373547362260681,
      "grad_norm": 0.3828059434890747,
      "learning_rate": 2e-05,
      "loss": 0.7162,
      "step": 1480
    },
    {
      "epoch": 0.12457152412005686,
      "grad_norm": 0.36190828680992126,
      "learning_rate": 2e-05,
      "loss": 0.7189,
      "step": 1490
    },
    {
      "epoch": 0.1254075746175069,
      "grad_norm": 0.35898563265800476,
      "learning_rate": 2e-05,
      "loss": 0.6823,
      "step": 1500
    },
    {
      "epoch": 0.12624362511495693,
      "grad_norm": 0.3842126429080963,
      "learning_rate": 2e-05,
      "loss": 0.7445,
      "step": 1510
    },
    {
      "epoch": 0.127079675612407,
      "grad_norm": 0.38555908203125,
      "learning_rate": 2e-05,
      "loss": 0.6886,
      "step": 1520
    },
    {
      "epoch": 0.12791572610985705,
      "grad_norm": 0.41499239206314087,
      "learning_rate": 2e-05,
      "loss": 0.7005,
      "step": 1530
    },
    {
      "epoch": 0.12875177660730708,
      "grad_norm": 0.3923279345035553,
      "learning_rate": 2e-05,
      "loss": 0.6807,
      "step": 1540
    },
    {
      "epoch": 0.12958782710475714,
      "grad_norm": 0.38119590282440186,
      "learning_rate": 2e-05,
      "loss": 0.7245,
      "step": 1550
    },
    {
      "epoch": 0.13042387760220717,
      "grad_norm": 0.44171908497810364,
      "learning_rate": 2e-05,
      "loss": 0.711,
      "step": 1560
    },
    {
      "epoch": 0.13125992809965722,
      "grad_norm": 0.3561314642429352,
      "learning_rate": 2e-05,
      "loss": 0.7106,
      "step": 1570
    },
    {
      "epoch": 0.13209597859710726,
      "grad_norm": 0.44831010699272156,
      "learning_rate": 2e-05,
      "loss": 0.7118,
      "step": 1580
    },
    {
      "epoch": 0.1329320290945573,
      "grad_norm": 0.42015114426612854,
      "learning_rate": 2e-05,
      "loss": 0.7072,
      "step": 1590
    },
    {
      "epoch": 0.13376807959200734,
      "grad_norm": 0.4430758059024811,
      "learning_rate": 2e-05,
      "loss": 0.7187,
      "step": 1600
    },
    {
      "epoch": 0.1346041300894574,
      "grad_norm": 0.37224727869033813,
      "learning_rate": 2e-05,
      "loss": 0.7045,
      "step": 1610
    },
    {
      "epoch": 0.13544018058690746,
      "grad_norm": 0.3727637529373169,
      "learning_rate": 2e-05,
      "loss": 0.7151,
      "step": 1620
    },
    {
      "epoch": 0.1362762310843575,
      "grad_norm": 0.36329200863838196,
      "learning_rate": 2e-05,
      "loss": 0.7359,
      "step": 1630
    },
    {
      "epoch": 0.13711228158180755,
      "grad_norm": 0.36346569657325745,
      "learning_rate": 2e-05,
      "loss": 0.689,
      "step": 1640
    },
    {
      "epoch": 0.13794833207925758,
      "grad_norm": 0.3709527254104614,
      "learning_rate": 2e-05,
      "loss": 0.7124,
      "step": 1650
    },
    {
      "epoch": 0.13878438257670764,
      "grad_norm": 0.41566887497901917,
      "learning_rate": 2e-05,
      "loss": 0.708,
      "step": 1660
    },
    {
      "epoch": 0.13962043307415767,
      "grad_norm": 0.3450981378555298,
      "learning_rate": 2e-05,
      "loss": 0.7007,
      "step": 1670
    },
    {
      "epoch": 0.14045648357160773,
      "grad_norm": 0.3517451584339142,
      "learning_rate": 2e-05,
      "loss": 0.7031,
      "step": 1680
    },
    {
      "epoch": 0.14129253406905778,
      "grad_norm": 0.400630384683609,
      "learning_rate": 2e-05,
      "loss": 0.7124,
      "step": 1690
    },
    {
      "epoch": 0.14212858456650782,
      "grad_norm": 0.4013160765171051,
      "learning_rate": 2e-05,
      "loss": 0.7085,
      "step": 1700
    },
    {
      "epoch": 0.14296463506395787,
      "grad_norm": 0.3802119791507721,
      "learning_rate": 2e-05,
      "loss": 0.7056,
      "step": 1710
    },
    {
      "epoch": 0.1438006855614079,
      "grad_norm": 0.4005148112773895,
      "learning_rate": 2e-05,
      "loss": 0.7065,
      "step": 1720
    },
    {
      "epoch": 0.14463673605885796,
      "grad_norm": 0.36928942799568176,
      "learning_rate": 2e-05,
      "loss": 0.6813,
      "step": 1730
    },
    {
      "epoch": 0.145472786556308,
      "grad_norm": 0.37812256813049316,
      "learning_rate": 2e-05,
      "loss": 0.7049,
      "step": 1740
    },
    {
      "epoch": 0.14630883705375805,
      "grad_norm": 0.3583225607872009,
      "learning_rate": 2e-05,
      "loss": 0.7184,
      "step": 1750
    },
    {
      "epoch": 0.14714488755120808,
      "grad_norm": 0.3970085680484772,
      "learning_rate": 2e-05,
      "loss": 0.6873,
      "step": 1760
    },
    {
      "epoch": 0.14798093804865814,
      "grad_norm": 0.38219332695007324,
      "learning_rate": 2e-05,
      "loss": 0.6963,
      "step": 1770
    },
    {
      "epoch": 0.1488169885461082,
      "grad_norm": 0.362723171710968,
      "learning_rate": 2e-05,
      "loss": 0.7162,
      "step": 1780
    },
    {
      "epoch": 0.14965303904355823,
      "grad_norm": 0.43519818782806396,
      "learning_rate": 2e-05,
      "loss": 0.7274,
      "step": 1790
    },
    {
      "epoch": 0.1504890895410083,
      "grad_norm": 0.36672765016555786,
      "learning_rate": 2e-05,
      "loss": 0.746,
      "step": 1800
    },
    {
      "epoch": 0.15132514003845832,
      "grad_norm": 0.34433141350746155,
      "learning_rate": 2e-05,
      "loss": 0.6847,
      "step": 1810
    },
    {
      "epoch": 0.15216119053590837,
      "grad_norm": 0.3767019510269165,
      "learning_rate": 2e-05,
      "loss": 0.7217,
      "step": 1820
    },
    {
      "epoch": 0.1529972410333584,
      "grad_norm": 0.3689115047454834,
      "learning_rate": 2e-05,
      "loss": 0.7378,
      "step": 1830
    },
    {
      "epoch": 0.15383329153080846,
      "grad_norm": 0.36576899886131287,
      "learning_rate": 2e-05,
      "loss": 0.7198,
      "step": 1840
    },
    {
      "epoch": 0.1546693420282585,
      "grad_norm": 0.4920649230480194,
      "learning_rate": 2e-05,
      "loss": 0.6827,
      "step": 1850
    },
    {
      "epoch": 0.15550539252570855,
      "grad_norm": 0.35931605100631714,
      "learning_rate": 2e-05,
      "loss": 0.7097,
      "step": 1860
    },
    {
      "epoch": 0.1563414430231586,
      "grad_norm": 0.36113813519477844,
      "learning_rate": 2e-05,
      "loss": 0.7307,
      "step": 1870
    },
    {
      "epoch": 0.15717749352060864,
      "grad_norm": 0.3809521496295929,
      "learning_rate": 2e-05,
      "loss": 0.7038,
      "step": 1880
    },
    {
      "epoch": 0.1580135440180587,
      "grad_norm": 0.4045011103153229,
      "learning_rate": 2e-05,
      "loss": 0.7031,
      "step": 1890
    },
    {
      "epoch": 0.15884959451550873,
      "grad_norm": 0.3887285590171814,
      "learning_rate": 2e-05,
      "loss": 0.6739,
      "step": 1900
    },
    {
      "epoch": 0.1596856450129588,
      "grad_norm": 0.4463697671890259,
      "learning_rate": 2e-05,
      "loss": 0.7147,
      "step": 1910
    },
    {
      "epoch": 0.16052169551040882,
      "grad_norm": 0.3879319131374359,
      "learning_rate": 2e-05,
      "loss": 0.7375,
      "step": 1920
    },
    {
      "epoch": 0.16135774600785888,
      "grad_norm": 0.35314497351646423,
      "learning_rate": 2e-05,
      "loss": 0.6811,
      "step": 1930
    },
    {
      "epoch": 0.1621937965053089,
      "grad_norm": 0.4394693970680237,
      "learning_rate": 2e-05,
      "loss": 0.7156,
      "step": 1940
    },
    {
      "epoch": 0.16302984700275897,
      "grad_norm": 0.3700949549674988,
      "learning_rate": 2e-05,
      "loss": 0.7054,
      "step": 1950
    },
    {
      "epoch": 0.16386589750020902,
      "grad_norm": 0.3599170446395874,
      "learning_rate": 2e-05,
      "loss": 0.7122,
      "step": 1960
    },
    {
      "epoch": 0.16470194799765905,
      "grad_norm": 0.41550520062446594,
      "learning_rate": 2e-05,
      "loss": 0.6874,
      "step": 1970
    },
    {
      "epoch": 0.1655379984951091,
      "grad_norm": 0.4028865694999695,
      "learning_rate": 2e-05,
      "loss": 0.7222,
      "step": 1980
    },
    {
      "epoch": 0.16637404899255914,
      "grad_norm": 0.3849325478076935,
      "learning_rate": 2e-05,
      "loss": 0.7122,
      "step": 1990
    },
    {
      "epoch": 0.1672100994900092,
      "grad_norm": 0.34337013959884644,
      "learning_rate": 2e-05,
      "loss": 0.6604,
      "step": 2000
    },
    {
      "epoch": 0.16804614998745923,
      "grad_norm": 0.3867923319339752,
      "learning_rate": 2e-05,
      "loss": 0.6965,
      "step": 2010
    },
    {
      "epoch": 0.1688822004849093,
      "grad_norm": 0.3923274874687195,
      "learning_rate": 2e-05,
      "loss": 0.7098,
      "step": 2020
    },
    {
      "epoch": 0.16971825098235935,
      "grad_norm": 0.3679618239402771,
      "learning_rate": 2e-05,
      "loss": 0.6978,
      "step": 2030
    },
    {
      "epoch": 0.17055430147980938,
      "grad_norm": 0.39310866594314575,
      "learning_rate": 2e-05,
      "loss": 0.71,
      "step": 2040
    },
    {
      "epoch": 0.17139035197725944,
      "grad_norm": 0.3967050313949585,
      "learning_rate": 2e-05,
      "loss": 0.7243,
      "step": 2050
    },
    {
      "epoch": 0.17222640247470947,
      "grad_norm": 0.40121224522590637,
      "learning_rate": 2e-05,
      "loss": 0.6967,
      "step": 2060
    },
    {
      "epoch": 0.17306245297215953,
      "grad_norm": 0.38088318705558777,
      "learning_rate": 2e-05,
      "loss": 0.6401,
      "step": 2070
    },
    {
      "epoch": 0.17389850346960956,
      "grad_norm": 0.3435222804546356,
      "learning_rate": 2e-05,
      "loss": 0.6647,
      "step": 2080
    },
    {
      "epoch": 0.1747345539670596,
      "grad_norm": 0.4220814108848572,
      "learning_rate": 2e-05,
      "loss": 0.7141,
      "step": 2090
    },
    {
      "epoch": 0.17557060446450964,
      "grad_norm": 0.35260170698165894,
      "learning_rate": 2e-05,
      "loss": 0.6713,
      "step": 2100
    },
    {
      "epoch": 0.1764066549619597,
      "grad_norm": 0.40899789333343506,
      "learning_rate": 2e-05,
      "loss": 0.7129,
      "step": 2110
    },
    {
      "epoch": 0.17724270545940976,
      "grad_norm": 0.4075130522251129,
      "learning_rate": 2e-05,
      "loss": 0.6915,
      "step": 2120
    },
    {
      "epoch": 0.1780787559568598,
      "grad_norm": 0.4192414879798889,
      "learning_rate": 2e-05,
      "loss": 0.6905,
      "step": 2130
    },
    {
      "epoch": 0.17891480645430985,
      "grad_norm": 0.38824033737182617,
      "learning_rate": 2e-05,
      "loss": 0.7058,
      "step": 2140
    },
    {
      "epoch": 0.17975085695175988,
      "grad_norm": 0.3785799443721771,
      "learning_rate": 2e-05,
      "loss": 0.7003,
      "step": 2150
    },
    {
      "epoch": 0.18058690744920994,
      "grad_norm": 0.3589932322502136,
      "learning_rate": 2e-05,
      "loss": 0.6703,
      "step": 2160
    },
    {
      "epoch": 0.18142295794665997,
      "grad_norm": 0.36105579137802124,
      "learning_rate": 2e-05,
      "loss": 0.6972,
      "step": 2170
    },
    {
      "epoch": 0.18225900844411003,
      "grad_norm": 0.3652353584766388,
      "learning_rate": 2e-05,
      "loss": 0.6906,
      "step": 2180
    },
    {
      "epoch": 0.18309505894156006,
      "grad_norm": 0.40131404995918274,
      "learning_rate": 2e-05,
      "loss": 0.7082,
      "step": 2190
    },
    {
      "epoch": 0.18393110943901012,
      "grad_norm": 0.40471386909484863,
      "learning_rate": 2e-05,
      "loss": 0.696,
      "step": 2200
    },
    {
      "epoch": 0.18476715993646017,
      "grad_norm": 0.3407021462917328,
      "learning_rate": 2e-05,
      "loss": 0.653,
      "step": 2210
    },
    {
      "epoch": 0.1856032104339102,
      "grad_norm": 0.37151047587394714,
      "learning_rate": 2e-05,
      "loss": 0.6928,
      "step": 2220
    },
    {
      "epoch": 0.18643926093136026,
      "grad_norm": 0.3697320818901062,
      "learning_rate": 2e-05,
      "loss": 0.692,
      "step": 2230
    },
    {
      "epoch": 0.1872753114288103,
      "grad_norm": 0.3968718349933624,
      "learning_rate": 2e-05,
      "loss": 0.6909,
      "step": 2240
    },
    {
      "epoch": 0.18811136192626035,
      "grad_norm": 0.3623242676258087,
      "learning_rate": 2e-05,
      "loss": 0.7195,
      "step": 2250
    },
    {
      "epoch": 0.18894741242371038,
      "grad_norm": 0.3716718852519989,
      "learning_rate": 2e-05,
      "loss": 0.7327,
      "step": 2260
    },
    {
      "epoch": 0.18978346292116044,
      "grad_norm": 0.38747113943099976,
      "learning_rate": 2e-05,
      "loss": 0.7033,
      "step": 2270
    },
    {
      "epoch": 0.1906195134186105,
      "grad_norm": 0.3699060380458832,
      "learning_rate": 2e-05,
      "loss": 0.7035,
      "step": 2280
    },
    {
      "epoch": 0.19145556391606053,
      "grad_norm": 0.3485587239265442,
      "learning_rate": 2e-05,
      "loss": 0.6982,
      "step": 2290
    },
    {
      "epoch": 0.1922916144135106,
      "grad_norm": 0.35775020718574524,
      "learning_rate": 2e-05,
      "loss": 0.6977,
      "step": 2300
    },
    {
      "epoch": 0.19312766491096062,
      "grad_norm": 0.386114239692688,
      "learning_rate": 2e-05,
      "loss": 0.6901,
      "step": 2310
    },
    {
      "epoch": 0.19396371540841068,
      "grad_norm": 0.4111344814300537,
      "learning_rate": 2e-05,
      "loss": 0.7134,
      "step": 2320
    },
    {
      "epoch": 0.1947997659058607,
      "grad_norm": 0.3693266212940216,
      "learning_rate": 2e-05,
      "loss": 0.6793,
      "step": 2330
    },
    {
      "epoch": 0.19563581640331076,
      "grad_norm": 0.3926129639148712,
      "learning_rate": 2e-05,
      "loss": 0.6856,
      "step": 2340
    },
    {
      "epoch": 0.1964718669007608,
      "grad_norm": 0.38753512501716614,
      "learning_rate": 2e-05,
      "loss": 0.7137,
      "step": 2350
    },
    {
      "epoch": 0.19730791739821085,
      "grad_norm": 0.3788638710975647,
      "learning_rate": 2e-05,
      "loss": 0.6795,
      "step": 2360
    },
    {
      "epoch": 0.1981439678956609,
      "grad_norm": 0.3338569104671478,
      "learning_rate": 2e-05,
      "loss": 0.6775,
      "step": 2370
    },
    {
      "epoch": 0.19898001839311094,
      "grad_norm": 0.366208016872406,
      "learning_rate": 2e-05,
      "loss": 0.7282,
      "step": 2380
    },
    {
      "epoch": 0.199816068890561,
      "grad_norm": 0.379518985748291,
      "learning_rate": 2e-05,
      "loss": 0.6781,
      "step": 2390
    },
    {
      "epoch": 0.20065211938801103,
      "grad_norm": 0.3841893970966339,
      "learning_rate": 2e-05,
      "loss": 0.7026,
      "step": 2400
    },
    {
      "epoch": 0.2014881698854611,
      "grad_norm": 0.3684847950935364,
      "learning_rate": 2e-05,
      "loss": 0.6626,
      "step": 2410
    },
    {
      "epoch": 0.20232422038291112,
      "grad_norm": 0.3933838903903961,
      "learning_rate": 2e-05,
      "loss": 0.7178,
      "step": 2420
    },
    {
      "epoch": 0.20316027088036118,
      "grad_norm": 0.3518088757991791,
      "learning_rate": 2e-05,
      "loss": 0.6663,
      "step": 2430
    },
    {
      "epoch": 0.2039963213778112,
      "grad_norm": 0.383081316947937,
      "learning_rate": 2e-05,
      "loss": 0.6947,
      "step": 2440
    },
    {
      "epoch": 0.20483237187526127,
      "grad_norm": 0.34842392802238464,
      "learning_rate": 2e-05,
      "loss": 0.6893,
      "step": 2450
    },
    {
      "epoch": 0.20566842237271132,
      "grad_norm": 0.36848798394203186,
      "learning_rate": 2e-05,
      "loss": 0.6873,
      "step": 2460
    },
    {
      "epoch": 0.20650447287016135,
      "grad_norm": 0.36847376823425293,
      "learning_rate": 2e-05,
      "loss": 0.73,
      "step": 2470
    },
    {
      "epoch": 0.2073405233676114,
      "grad_norm": 0.4002387523651123,
      "learning_rate": 2e-05,
      "loss": 0.6958,
      "step": 2480
    },
    {
      "epoch": 0.20817657386506144,
      "grad_norm": 0.383114755153656,
      "learning_rate": 2e-05,
      "loss": 0.704,
      "step": 2490
    },
    {
      "epoch": 0.2090126243625115,
      "grad_norm": 0.3491422235965729,
      "learning_rate": 2e-05,
      "loss": 0.6831,
      "step": 2500
    },
    {
      "epoch": 0.20984867485996153,
      "grad_norm": 0.3453516364097595,
      "learning_rate": 2e-05,
      "loss": 0.6988,
      "step": 2510
    },
    {
      "epoch": 0.2106847253574116,
      "grad_norm": 0.3652738332748413,
      "learning_rate": 2e-05,
      "loss": 0.7116,
      "step": 2520
    },
    {
      "epoch": 0.21152077585486162,
      "grad_norm": 0.3410079777240753,
      "learning_rate": 2e-05,
      "loss": 0.7418,
      "step": 2530
    },
    {
      "epoch": 0.21235682635231168,
      "grad_norm": 0.38150179386138916,
      "learning_rate": 2e-05,
      "loss": 0.6719,
      "step": 2540
    },
    {
      "epoch": 0.21319287684976174,
      "grad_norm": 0.33772507309913635,
      "learning_rate": 2e-05,
      "loss": 0.6822,
      "step": 2550
    },
    {
      "epoch": 0.21402892734721177,
      "grad_norm": 0.40987348556518555,
      "learning_rate": 2e-05,
      "loss": 0.7091,
      "step": 2560
    },
    {
      "epoch": 0.21486497784466183,
      "grad_norm": 0.3767969012260437,
      "learning_rate": 2e-05,
      "loss": 0.7025,
      "step": 2570
    },
    {
      "epoch": 0.21570102834211186,
      "grad_norm": 0.3843465745449066,
      "learning_rate": 2e-05,
      "loss": 0.7031,
      "step": 2580
    },
    {
      "epoch": 0.21653707883956191,
      "grad_norm": 0.363743394613266,
      "learning_rate": 2e-05,
      "loss": 0.6946,
      "step": 2590
    },
    {
      "epoch": 0.21737312933701194,
      "grad_norm": 0.3590858280658722,
      "learning_rate": 2e-05,
      "loss": 0.683,
      "step": 2600
    },
    {
      "epoch": 0.218209179834462,
      "grad_norm": 0.5465121865272522,
      "learning_rate": 2e-05,
      "loss": 0.7204,
      "step": 2610
    },
    {
      "epoch": 0.21904523033191206,
      "grad_norm": 0.37804701924324036,
      "learning_rate": 2e-05,
      "loss": 0.6918,
      "step": 2620
    },
    {
      "epoch": 0.2198812808293621,
      "grad_norm": 0.37326765060424805,
      "learning_rate": 2e-05,
      "loss": 0.6885,
      "step": 2630
    },
    {
      "epoch": 0.22071733132681215,
      "grad_norm": 0.42487889528274536,
      "learning_rate": 2e-05,
      "loss": 0.6697,
      "step": 2640
    },
    {
      "epoch": 0.22155338182426218,
      "grad_norm": 0.3364788293838501,
      "learning_rate": 2e-05,
      "loss": 0.7001,
      "step": 2650
    },
    {
      "epoch": 0.22238943232171224,
      "grad_norm": 0.34788456559181213,
      "learning_rate": 2e-05,
      "loss": 0.6872,
      "step": 2660
    },
    {
      "epoch": 0.22322548281916227,
      "grad_norm": 0.4006475508213043,
      "learning_rate": 2e-05,
      "loss": 0.6791,
      "step": 2670
    },
    {
      "epoch": 0.22406153331661233,
      "grad_norm": 0.3868175446987152,
      "learning_rate": 2e-05,
      "loss": 0.669,
      "step": 2680
    },
    {
      "epoch": 0.22489758381406236,
      "grad_norm": 0.38875290751457214,
      "learning_rate": 2e-05,
      "loss": 0.6633,
      "step": 2690
    },
    {
      "epoch": 0.22573363431151242,
      "grad_norm": 0.3829600214958191,
      "learning_rate": 2e-05,
      "loss": 0.6843,
      "step": 2700
    },
    {
      "epoch": 0.22656968480896247,
      "grad_norm": 0.4310683310031891,
      "learning_rate": 2e-05,
      "loss": 0.7004,
      "step": 2710
    },
    {
      "epoch": 0.2274057353064125,
      "grad_norm": 0.36083126068115234,
      "learning_rate": 2e-05,
      "loss": 0.6841,
      "step": 2720
    },
    {
      "epoch": 0.22824178580386256,
      "grad_norm": 0.4124043583869934,
      "learning_rate": 2e-05,
      "loss": 0.6971,
      "step": 2730
    },
    {
      "epoch": 0.2290778363013126,
      "grad_norm": 0.4539124667644501,
      "learning_rate": 2e-05,
      "loss": 0.6848,
      "step": 2740
    },
    {
      "epoch": 0.22991388679876265,
      "grad_norm": 0.37063753604888916,
      "learning_rate": 2e-05,
      "loss": 0.6925,
      "step": 2750
    },
    {
      "epoch": 0.23074993729621268,
      "grad_norm": 0.3623632788658142,
      "learning_rate": 2e-05,
      "loss": 0.702,
      "step": 2760
    },
    {
      "epoch": 0.23158598779366274,
      "grad_norm": 0.44299641251564026,
      "learning_rate": 2e-05,
      "loss": 0.6755,
      "step": 2770
    },
    {
      "epoch": 0.23242203829111277,
      "grad_norm": 0.356118381023407,
      "learning_rate": 2e-05,
      "loss": 0.7021,
      "step": 2780
    },
    {
      "epoch": 0.23325808878856283,
      "grad_norm": 0.38731667399406433,
      "learning_rate": 2e-05,
      "loss": 0.6911,
      "step": 2790
    },
    {
      "epoch": 0.2340941392860129,
      "grad_norm": 0.34089842438697815,
      "learning_rate": 2e-05,
      "loss": 0.6782,
      "step": 2800
    },
    {
      "epoch": 0.23493018978346292,
      "grad_norm": 0.3597925007343292,
      "learning_rate": 2e-05,
      "loss": 0.6557,
      "step": 2810
    },
    {
      "epoch": 0.23576624028091298,
      "grad_norm": 0.36081627011299133,
      "learning_rate": 2e-05,
      "loss": 0.6539,
      "step": 2820
    },
    {
      "epoch": 0.236602290778363,
      "grad_norm": 0.3887976408004761,
      "learning_rate": 2e-05,
      "loss": 0.6868,
      "step": 2830
    },
    {
      "epoch": 0.23743834127581306,
      "grad_norm": 0.37405726313591003,
      "learning_rate": 2e-05,
      "loss": 0.6438,
      "step": 2840
    },
    {
      "epoch": 0.2382743917732631,
      "grad_norm": 0.37012699246406555,
      "learning_rate": 2e-05,
      "loss": 0.6773,
      "step": 2850
    },
    {
      "epoch": 0.23911044227071315,
      "grad_norm": 0.39837878942489624,
      "learning_rate": 2e-05,
      "loss": 0.6733,
      "step": 2860
    },
    {
      "epoch": 0.2399464927681632,
      "grad_norm": 0.3539638817310333,
      "learning_rate": 2e-05,
      "loss": 0.6752,
      "step": 2870
    },
    {
      "epoch": 0.24078254326561324,
      "grad_norm": 0.4221721291542053,
      "learning_rate": 2e-05,
      "loss": 0.6896,
      "step": 2880
    },
    {
      "epoch": 0.2416185937630633,
      "grad_norm": 0.35411736369132996,
      "learning_rate": 2e-05,
      "loss": 0.6895,
      "step": 2890
    },
    {
      "epoch": 0.24245464426051333,
      "grad_norm": 0.3807287812232971,
      "learning_rate": 2e-05,
      "loss": 0.6683,
      "step": 2900
    },
    {
      "epoch": 0.2432906947579634,
      "grad_norm": 0.33664026856422424,
      "learning_rate": 2e-05,
      "loss": 0.6926,
      "step": 2910
    },
    {
      "epoch": 0.24412674525541342,
      "grad_norm": 0.3574134111404419,
      "learning_rate": 2e-05,
      "loss": 0.702,
      "step": 2920
    },
    {
      "epoch": 0.24496279575286348,
      "grad_norm": 0.34514501690864563,
      "learning_rate": 2e-05,
      "loss": 0.6632,
      "step": 2930
    },
    {
      "epoch": 0.2457988462503135,
      "grad_norm": 0.3853206932544708,
      "learning_rate": 2e-05,
      "loss": 0.6684,
      "step": 2940
    },
    {
      "epoch": 0.24663489674776357,
      "grad_norm": 0.37129130959510803,
      "learning_rate": 2e-05,
      "loss": 0.6754,
      "step": 2950
    },
    {
      "epoch": 0.24747094724521362,
      "grad_norm": 0.36231255531311035,
      "learning_rate": 2e-05,
      "loss": 0.6557,
      "step": 2960
    },
    {
      "epoch": 0.24830699774266365,
      "grad_norm": 0.36850860714912415,
      "learning_rate": 2e-05,
      "loss": 0.6721,
      "step": 2970
    },
    {
      "epoch": 0.2491430482401137,
      "grad_norm": 0.3462553024291992,
      "learning_rate": 2e-05,
      "loss": 0.6848,
      "step": 2980
    },
    {
      "epoch": 0.24997909873756374,
      "grad_norm": 0.3916807770729065,
      "learning_rate": 2e-05,
      "loss": 0.6859,
      "step": 2990
    },
    {
      "epoch": 0.2508151492350138,
      "grad_norm": 0.37022486329078674,
      "learning_rate": 2e-05,
      "loss": 0.7,
      "step": 3000
    },
    {
      "epoch": 0.25165119973246386,
      "grad_norm": 0.3758009970188141,
      "learning_rate": 2e-05,
      "loss": 0.6982,
      "step": 3010
    },
    {
      "epoch": 0.25248725022991386,
      "grad_norm": 0.3388315439224243,
      "learning_rate": 2e-05,
      "loss": 0.6803,
      "step": 3020
    },
    {
      "epoch": 0.2533233007273639,
      "grad_norm": 0.3719688057899475,
      "learning_rate": 2e-05,
      "loss": 0.701,
      "step": 3030
    },
    {
      "epoch": 0.254159351224814,
      "grad_norm": 0.38503506779670715,
      "learning_rate": 2e-05,
      "loss": 0.6778,
      "step": 3040
    },
    {
      "epoch": 0.25499540172226404,
      "grad_norm": 0.408718466758728,
      "learning_rate": 2e-05,
      "loss": 0.6707,
      "step": 3050
    },
    {
      "epoch": 0.2558314522197141,
      "grad_norm": 0.3707769215106964,
      "learning_rate": 2e-05,
      "loss": 0.6685,
      "step": 3060
    },
    {
      "epoch": 0.2566675027171641,
      "grad_norm": 0.35878700017929077,
      "learning_rate": 2e-05,
      "loss": 0.6771,
      "step": 3070
    },
    {
      "epoch": 0.25750355321461416,
      "grad_norm": 0.40199294686317444,
      "learning_rate": 2e-05,
      "loss": 0.7029,
      "step": 3080
    },
    {
      "epoch": 0.2583396037120642,
      "grad_norm": 0.39858362078666687,
      "learning_rate": 2e-05,
      "loss": 0.6961,
      "step": 3090
    },
    {
      "epoch": 0.2591756542095143,
      "grad_norm": 0.4109138548374176,
      "learning_rate": 2e-05,
      "loss": 0.7061,
      "step": 3100
    },
    {
      "epoch": 0.2600117047069643,
      "grad_norm": 0.34975865483283997,
      "learning_rate": 2e-05,
      "loss": 0.6675,
      "step": 3110
    },
    {
      "epoch": 0.26084775520441433,
      "grad_norm": 0.35346266627311707,
      "learning_rate": 2e-05,
      "loss": 0.6397,
      "step": 3120
    },
    {
      "epoch": 0.2616838057018644,
      "grad_norm": 0.37450355291366577,
      "learning_rate": 2e-05,
      "loss": 0.6871,
      "step": 3130
    },
    {
      "epoch": 0.26251985619931445,
      "grad_norm": 0.33483919501304626,
      "learning_rate": 2e-05,
      "loss": 0.6629,
      "step": 3140
    },
    {
      "epoch": 0.2633559066967645,
      "grad_norm": 0.3381412625312805,
      "learning_rate": 2e-05,
      "loss": 0.6839,
      "step": 3150
    },
    {
      "epoch": 0.2641919571942145,
      "grad_norm": 0.34557151794433594,
      "learning_rate": 2e-05,
      "loss": 0.6498,
      "step": 3160
    },
    {
      "epoch": 0.26502800769166457,
      "grad_norm": 0.3589693009853363,
      "learning_rate": 2e-05,
      "loss": 0.6742,
      "step": 3170
    },
    {
      "epoch": 0.2658640581891146,
      "grad_norm": 0.5422878861427307,
      "learning_rate": 2e-05,
      "loss": 0.7182,
      "step": 3180
    },
    {
      "epoch": 0.2667001086865647,
      "grad_norm": 0.39269575476646423,
      "learning_rate": 2e-05,
      "loss": 0.6678,
      "step": 3190
    },
    {
      "epoch": 0.2675361591840147,
      "grad_norm": 0.3926292359828949,
      "learning_rate": 2e-05,
      "loss": 0.6907,
      "step": 3200
    },
    {
      "epoch": 0.26837220968146475,
      "grad_norm": 0.36348679661750793,
      "learning_rate": 2e-05,
      "loss": 0.6649,
      "step": 3210
    },
    {
      "epoch": 0.2692082601789148,
      "grad_norm": 0.36078062653541565,
      "learning_rate": 2e-05,
      "loss": 0.6617,
      "step": 3220
    },
    {
      "epoch": 0.27004431067636486,
      "grad_norm": 0.43532776832580566,
      "learning_rate": 2e-05,
      "loss": 0.7007,
      "step": 3230
    },
    {
      "epoch": 0.2708803611738149,
      "grad_norm": 0.3712572753429413,
      "learning_rate": 2e-05,
      "loss": 0.6804,
      "step": 3240
    },
    {
      "epoch": 0.2717164116712649,
      "grad_norm": 0.3930246829986572,
      "learning_rate": 2e-05,
      "loss": 0.6852,
      "step": 3250
    },
    {
      "epoch": 0.272552462168715,
      "grad_norm": 0.3693724572658539,
      "learning_rate": 2e-05,
      "loss": 0.681,
      "step": 3260
    },
    {
      "epoch": 0.27338851266616504,
      "grad_norm": 0.3628208637237549,
      "learning_rate": 2e-05,
      "loss": 0.672,
      "step": 3270
    },
    {
      "epoch": 0.2742245631636151,
      "grad_norm": 0.3537019193172455,
      "learning_rate": 2e-05,
      "loss": 0.6722,
      "step": 3280
    },
    {
      "epoch": 0.2750606136610651,
      "grad_norm": 0.40035539865493774,
      "learning_rate": 2e-05,
      "loss": 0.6546,
      "step": 3290
    },
    {
      "epoch": 0.27589666415851516,
      "grad_norm": 0.38971859216690063,
      "learning_rate": 2e-05,
      "loss": 0.6566,
      "step": 3300
    },
    {
      "epoch": 0.2767327146559652,
      "grad_norm": 0.6136935949325562,
      "learning_rate": 2e-05,
      "loss": 0.6877,
      "step": 3310
    },
    {
      "epoch": 0.2775687651534153,
      "grad_norm": 0.3618673086166382,
      "learning_rate": 2e-05,
      "loss": 0.6393,
      "step": 3320
    },
    {
      "epoch": 0.27840481565086533,
      "grad_norm": 0.3465752899646759,
      "learning_rate": 2e-05,
      "loss": 0.636,
      "step": 3330
    },
    {
      "epoch": 0.27924086614831534,
      "grad_norm": 0.3752216696739197,
      "learning_rate": 2e-05,
      "loss": 0.6406,
      "step": 3340
    },
    {
      "epoch": 0.2800769166457654,
      "grad_norm": 0.37954431772232056,
      "learning_rate": 2e-05,
      "loss": 0.689,
      "step": 3350
    },
    {
      "epoch": 0.28091296714321545,
      "grad_norm": 0.39894333481788635,
      "learning_rate": 2e-05,
      "loss": 0.6845,
      "step": 3360
    },
    {
      "epoch": 0.2817490176406655,
      "grad_norm": 0.33840516209602356,
      "learning_rate": 2e-05,
      "loss": 0.6856,
      "step": 3370
    },
    {
      "epoch": 0.28258506813811557,
      "grad_norm": 0.3931964039802551,
      "learning_rate": 2e-05,
      "loss": 0.6862,
      "step": 3380
    },
    {
      "epoch": 0.28342111863556557,
      "grad_norm": 0.40239232778549194,
      "learning_rate": 2e-05,
      "loss": 0.6439,
      "step": 3390
    },
    {
      "epoch": 0.28425716913301563,
      "grad_norm": 0.6196666359901428,
      "learning_rate": 2e-05,
      "loss": 0.6669,
      "step": 3400
    },
    {
      "epoch": 0.2850932196304657,
      "grad_norm": 0.4078761041164398,
      "learning_rate": 2e-05,
      "loss": 0.6737,
      "step": 3410
    },
    {
      "epoch": 0.28592927012791575,
      "grad_norm": 0.35943377017974854,
      "learning_rate": 2e-05,
      "loss": 0.641,
      "step": 3420
    },
    {
      "epoch": 0.28676532062536575,
      "grad_norm": 0.39068886637687683,
      "learning_rate": 2e-05,
      "loss": 0.6867,
      "step": 3430
    },
    {
      "epoch": 0.2876013711228158,
      "grad_norm": 1.7716305255889893,
      "learning_rate": 2e-05,
      "loss": 0.674,
      "step": 3440
    },
    {
      "epoch": 0.28843742162026587,
      "grad_norm": 0.38933104276657104,
      "learning_rate": 2e-05,
      "loss": 0.6772,
      "step": 3450
    },
    {
      "epoch": 0.2892734721177159,
      "grad_norm": 0.3752393424510956,
      "learning_rate": 2e-05,
      "loss": 0.6573,
      "step": 3460
    },
    {
      "epoch": 0.290109522615166,
      "grad_norm": 0.35931167006492615,
      "learning_rate": 2e-05,
      "loss": 0.6881,
      "step": 3470
    },
    {
      "epoch": 0.290945573112616,
      "grad_norm": 0.38396894931793213,
      "learning_rate": 2e-05,
      "loss": 0.6869,
      "step": 3480
    },
    {
      "epoch": 0.29178162361006604,
      "grad_norm": 0.37433430552482605,
      "learning_rate": 2e-05,
      "loss": 0.697,
      "step": 3490
    },
    {
      "epoch": 0.2926176741075161,
      "grad_norm": 0.3965727984905243,
      "learning_rate": 2e-05,
      "loss": 0.6878,
      "step": 3500
    },
    {
      "epoch": 0.29345372460496616,
      "grad_norm": 0.327019602060318,
      "learning_rate": 2e-05,
      "loss": 0.6601,
      "step": 3510
    },
    {
      "epoch": 0.29428977510241616,
      "grad_norm": 0.37980222702026367,
      "learning_rate": 2e-05,
      "loss": 0.6974,
      "step": 3520
    },
    {
      "epoch": 0.2951258255998662,
      "grad_norm": 0.4086945652961731,
      "learning_rate": 2e-05,
      "loss": 0.6966,
      "step": 3530
    },
    {
      "epoch": 0.2959618760973163,
      "grad_norm": 0.362612783908844,
      "learning_rate": 2e-05,
      "loss": 0.6942,
      "step": 3540
    },
    {
      "epoch": 0.29679792659476634,
      "grad_norm": 0.37178340554237366,
      "learning_rate": 2e-05,
      "loss": 0.6664,
      "step": 3550
    },
    {
      "epoch": 0.2976339770922164,
      "grad_norm": 0.37645795941352844,
      "learning_rate": 2e-05,
      "loss": 0.6761,
      "step": 3560
    },
    {
      "epoch": 0.2984700275896664,
      "grad_norm": 0.347457617521286,
      "learning_rate": 2e-05,
      "loss": 0.6669,
      "step": 3570
    },
    {
      "epoch": 0.29930607808711646,
      "grad_norm": 0.3348453640937805,
      "learning_rate": 2e-05,
      "loss": 0.6509,
      "step": 3580
    },
    {
      "epoch": 0.3001421285845665,
      "grad_norm": 0.35900646448135376,
      "learning_rate": 2e-05,
      "loss": 0.6732,
      "step": 3590
    },
    {
      "epoch": 0.3009781790820166,
      "grad_norm": 0.35713493824005127,
      "learning_rate": 2e-05,
      "loss": 0.6206,
      "step": 3600
    },
    {
      "epoch": 0.3018142295794666,
      "grad_norm": 0.3648741841316223,
      "learning_rate": 2e-05,
      "loss": 0.6575,
      "step": 3610
    },
    {
      "epoch": 0.30265028007691663,
      "grad_norm": 0.3970864415168762,
      "learning_rate": 2e-05,
      "loss": 0.6909,
      "step": 3620
    },
    {
      "epoch": 0.3034863305743667,
      "grad_norm": 0.38486719131469727,
      "learning_rate": 2e-05,
      "loss": 0.6768,
      "step": 3630
    },
    {
      "epoch": 0.30432238107181675,
      "grad_norm": 0.38870275020599365,
      "learning_rate": 2e-05,
      "loss": 0.6524,
      "step": 3640
    },
    {
      "epoch": 0.3051584315692668,
      "grad_norm": 0.4234541356563568,
      "learning_rate": 2e-05,
      "loss": 0.6823,
      "step": 3650
    },
    {
      "epoch": 0.3059944820667168,
      "grad_norm": 0.37357190251350403,
      "learning_rate": 2e-05,
      "loss": 0.6846,
      "step": 3660
    },
    {
      "epoch": 0.30683053256416687,
      "grad_norm": 0.39218512177467346,
      "learning_rate": 2e-05,
      "loss": 0.6627,
      "step": 3670
    },
    {
      "epoch": 0.3076665830616169,
      "grad_norm": 0.38100025057792664,
      "learning_rate": 2e-05,
      "loss": 0.6931,
      "step": 3680
    },
    {
      "epoch": 0.308502633559067,
      "grad_norm": 0.3262445330619812,
      "learning_rate": 2e-05,
      "loss": 0.6824,
      "step": 3690
    },
    {
      "epoch": 0.309338684056517,
      "grad_norm": 0.36331281065940857,
      "learning_rate": 2e-05,
      "loss": 0.6591,
      "step": 3700
    },
    {
      "epoch": 0.31017473455396705,
      "grad_norm": 0.35214877128601074,
      "learning_rate": 2e-05,
      "loss": 0.7103,
      "step": 3710
    },
    {
      "epoch": 0.3110107850514171,
      "grad_norm": 0.38968539237976074,
      "learning_rate": 2e-05,
      "loss": 0.6981,
      "step": 3720
    },
    {
      "epoch": 0.31184683554886716,
      "grad_norm": 0.3726252317428589,
      "learning_rate": 2e-05,
      "loss": 0.6803,
      "step": 3730
    },
    {
      "epoch": 0.3126828860463172,
      "grad_norm": 0.3784383535385132,
      "learning_rate": 2e-05,
      "loss": 0.6678,
      "step": 3740
    },
    {
      "epoch": 0.3135189365437672,
      "grad_norm": 0.3368676006793976,
      "learning_rate": 2e-05,
      "loss": 0.6521,
      "step": 3750
    },
    {
      "epoch": 0.3143549870412173,
      "grad_norm": 0.3492288589477539,
      "learning_rate": 2e-05,
      "loss": 0.6623,
      "step": 3760
    },
    {
      "epoch": 0.31519103753866734,
      "grad_norm": 0.35406845808029175,
      "learning_rate": 2e-05,
      "loss": 0.6549,
      "step": 3770
    },
    {
      "epoch": 0.3160270880361174,
      "grad_norm": 0.39807578921318054,
      "learning_rate": 2e-05,
      "loss": 0.6746,
      "step": 3780
    },
    {
      "epoch": 0.3168631385335674,
      "grad_norm": 0.3588227927684784,
      "learning_rate": 2e-05,
      "loss": 0.6718,
      "step": 3790
    },
    {
      "epoch": 0.31769918903101746,
      "grad_norm": 0.36940997838974,
      "learning_rate": 2e-05,
      "loss": 0.6615,
      "step": 3800
    },
    {
      "epoch": 0.3185352395284675,
      "grad_norm": 0.3693230450153351,
      "learning_rate": 2e-05,
      "loss": 0.6823,
      "step": 3810
    },
    {
      "epoch": 0.3193712900259176,
      "grad_norm": 0.3649313449859619,
      "learning_rate": 2e-05,
      "loss": 0.6929,
      "step": 3820
    },
    {
      "epoch": 0.32020734052336763,
      "grad_norm": 0.3412187099456787,
      "learning_rate": 2e-05,
      "loss": 0.7131,
      "step": 3830
    },
    {
      "epoch": 0.32104339102081764,
      "grad_norm": 0.4318503737449646,
      "learning_rate": 2e-05,
      "loss": 0.7043,
      "step": 3840
    },
    {
      "epoch": 0.3218794415182677,
      "grad_norm": 0.35563552379608154,
      "learning_rate": 2e-05,
      "loss": 0.6621,
      "step": 3850
    },
    {
      "epoch": 0.32271549201571775,
      "grad_norm": 0.3535933494567871,
      "learning_rate": 2e-05,
      "loss": 0.6662,
      "step": 3860
    },
    {
      "epoch": 0.3235515425131678,
      "grad_norm": 0.3648977279663086,
      "learning_rate": 2e-05,
      "loss": 0.6853,
      "step": 3870
    },
    {
      "epoch": 0.3243875930106178,
      "grad_norm": 0.3543550372123718,
      "learning_rate": 2e-05,
      "loss": 0.6848,
      "step": 3880
    },
    {
      "epoch": 0.3252236435080679,
      "grad_norm": 0.38484376668930054,
      "learning_rate": 2e-05,
      "loss": 0.6893,
      "step": 3890
    },
    {
      "epoch": 0.32605969400551793,
      "grad_norm": 0.37139758467674255,
      "learning_rate": 2e-05,
      "loss": 0.6962,
      "step": 3900
    },
    {
      "epoch": 0.326895744502968,
      "grad_norm": 0.38717278838157654,
      "learning_rate": 2e-05,
      "loss": 0.7069,
      "step": 3910
    },
    {
      "epoch": 0.32773179500041805,
      "grad_norm": 0.3783646523952484,
      "learning_rate": 2e-05,
      "loss": 0.6738,
      "step": 3920
    },
    {
      "epoch": 0.32856784549786805,
      "grad_norm": 0.3423800766468048,
      "learning_rate": 2e-05,
      "loss": 0.6651,
      "step": 3930
    },
    {
      "epoch": 0.3294038959953181,
      "grad_norm": 0.3589499294757843,
      "learning_rate": 2e-05,
      "loss": 0.6574,
      "step": 3940
    },
    {
      "epoch": 0.33023994649276817,
      "grad_norm": 0.40661120414733887,
      "learning_rate": 2e-05,
      "loss": 0.6873,
      "step": 3950
    },
    {
      "epoch": 0.3310759969902182,
      "grad_norm": 0.3802325129508972,
      "learning_rate": 2e-05,
      "loss": 0.6759,
      "step": 3960
    },
    {
      "epoch": 0.3319120474876683,
      "grad_norm": 0.3785724639892578,
      "learning_rate": 2e-05,
      "loss": 0.6507,
      "step": 3970
    },
    {
      "epoch": 0.3327480979851183,
      "grad_norm": 0.36652204394340515,
      "learning_rate": 2e-05,
      "loss": 0.7036,
      "step": 3980
    },
    {
      "epoch": 0.33358414848256834,
      "grad_norm": 0.33603420853614807,
      "learning_rate": 2e-05,
      "loss": 0.7145,
      "step": 3990
    },
    {
      "epoch": 0.3344201989800184,
      "grad_norm": 0.3536396026611328,
      "learning_rate": 2e-05,
      "loss": 0.6736,
      "step": 4000
    },
    {
      "epoch": 0.33525624947746846,
      "grad_norm": 0.4215162396430969,
      "learning_rate": 2e-05,
      "loss": 0.6499,
      "step": 4010
    },
    {
      "epoch": 0.33609229997491846,
      "grad_norm": 0.36844271421432495,
      "learning_rate": 2e-05,
      "loss": 0.6881,
      "step": 4020
    },
    {
      "epoch": 0.3369283504723685,
      "grad_norm": 0.4516617953777313,
      "learning_rate": 2e-05,
      "loss": 0.6588,
      "step": 4030
    },
    {
      "epoch": 0.3377644009698186,
      "grad_norm": 0.3380790054798126,
      "learning_rate": 2e-05,
      "loss": 0.6262,
      "step": 4040
    },
    {
      "epoch": 0.33860045146726864,
      "grad_norm": 0.3842700719833374,
      "learning_rate": 2e-05,
      "loss": 0.6276,
      "step": 4050
    },
    {
      "epoch": 0.3394365019647187,
      "grad_norm": 0.34872040152549744,
      "learning_rate": 2e-05,
      "loss": 0.653,
      "step": 4060
    },
    {
      "epoch": 0.3402725524621687,
      "grad_norm": 0.3866327702999115,
      "learning_rate": 2e-05,
      "loss": 0.6716,
      "step": 4070
    },
    {
      "epoch": 0.34110860295961876,
      "grad_norm": 0.3686073124408722,
      "learning_rate": 2e-05,
      "loss": 0.6701,
      "step": 4080
    },
    {
      "epoch": 0.3419446534570688,
      "grad_norm": 0.42053115367889404,
      "learning_rate": 2e-05,
      "loss": 0.6859,
      "step": 4090
    },
    {
      "epoch": 0.3427807039545189,
      "grad_norm": 0.3751194477081299,
      "learning_rate": 2e-05,
      "loss": 0.658,
      "step": 4100
    },
    {
      "epoch": 0.3436167544519689,
      "grad_norm": 0.375248521566391,
      "learning_rate": 2e-05,
      "loss": 0.6791,
      "step": 4110
    },
    {
      "epoch": 0.34445280494941893,
      "grad_norm": 0.32522687315940857,
      "learning_rate": 2e-05,
      "loss": 0.6833,
      "step": 4120
    },
    {
      "epoch": 0.345288855446869,
      "grad_norm": 0.34771233797073364,
      "learning_rate": 2e-05,
      "loss": 0.6234,
      "step": 4130
    },
    {
      "epoch": 0.34612490594431905,
      "grad_norm": 0.3351838290691376,
      "learning_rate": 2e-05,
      "loss": 0.6574,
      "step": 4140
    },
    {
      "epoch": 0.3469609564417691,
      "grad_norm": 0.3469686210155487,
      "learning_rate": 2e-05,
      "loss": 0.6474,
      "step": 4150
    },
    {
      "epoch": 0.3477970069392191,
      "grad_norm": 0.3836895823478699,
      "learning_rate": 2e-05,
      "loss": 0.6388,
      "step": 4160
    },
    {
      "epoch": 0.34863305743666917,
      "grad_norm": 0.3830288350582123,
      "learning_rate": 2e-05,
      "loss": 0.6745,
      "step": 4170
    },
    {
      "epoch": 0.3494691079341192,
      "grad_norm": 0.37637174129486084,
      "learning_rate": 2e-05,
      "loss": 0.6619,
      "step": 4180
    },
    {
      "epoch": 0.3503051584315693,
      "grad_norm": 0.4341719150543213,
      "learning_rate": 2e-05,
      "loss": 0.6545,
      "step": 4190
    },
    {
      "epoch": 0.3511412089290193,
      "grad_norm": 0.42583167552948,
      "learning_rate": 2e-05,
      "loss": 0.6617,
      "step": 4200
    },
    {
      "epoch": 0.35197725942646935,
      "grad_norm": 0.3464331030845642,
      "learning_rate": 2e-05,
      "loss": 0.659,
      "step": 4210
    },
    {
      "epoch": 0.3528133099239194,
      "grad_norm": 0.3504962921142578,
      "learning_rate": 2e-05,
      "loss": 0.6594,
      "step": 4220
    },
    {
      "epoch": 0.35364936042136946,
      "grad_norm": 0.3957189619541168,
      "learning_rate": 2e-05,
      "loss": 0.6585,
      "step": 4230
    },
    {
      "epoch": 0.3544854109188195,
      "grad_norm": 0.35942843556404114,
      "learning_rate": 2e-05,
      "loss": 0.6663,
      "step": 4240
    },
    {
      "epoch": 0.3553214614162695,
      "grad_norm": 0.36357617378234863,
      "learning_rate": 2e-05,
      "loss": 0.6762,
      "step": 4250
    },
    {
      "epoch": 0.3561575119137196,
      "grad_norm": 0.3632410764694214,
      "learning_rate": 2e-05,
      "loss": 0.6707,
      "step": 4260
    },
    {
      "epoch": 0.35699356241116964,
      "grad_norm": 0.40080034732818604,
      "learning_rate": 2e-05,
      "loss": 0.6864,
      "step": 4270
    },
    {
      "epoch": 0.3578296129086197,
      "grad_norm": 0.37345850467681885,
      "learning_rate": 2e-05,
      "loss": 0.6821,
      "step": 4280
    },
    {
      "epoch": 0.3586656634060697,
      "grad_norm": 0.3551454246044159,
      "learning_rate": 2e-05,
      "loss": 0.6923,
      "step": 4290
    },
    {
      "epoch": 0.35950171390351976,
      "grad_norm": 0.41821813583374023,
      "learning_rate": 2e-05,
      "loss": 0.6756,
      "step": 4300
    },
    {
      "epoch": 0.3603377644009698,
      "grad_norm": 0.35665106773376465,
      "learning_rate": 2e-05,
      "loss": 0.6719,
      "step": 4310
    },
    {
      "epoch": 0.3611738148984199,
      "grad_norm": 0.32754582166671753,
      "learning_rate": 2e-05,
      "loss": 0.6623,
      "step": 4320
    },
    {
      "epoch": 0.36200986539586993,
      "grad_norm": 0.36815065145492554,
      "learning_rate": 2e-05,
      "loss": 0.6808,
      "step": 4330
    },
    {
      "epoch": 0.36284591589331994,
      "grad_norm": 0.41788262128829956,
      "learning_rate": 2e-05,
      "loss": 0.6529,
      "step": 4340
    },
    {
      "epoch": 0.36368196639077,
      "grad_norm": 0.3677462041378021,
      "learning_rate": 2e-05,
      "loss": 0.6819,
      "step": 4350
    },
    {
      "epoch": 0.36451801688822005,
      "grad_norm": 0.3276795446872711,
      "learning_rate": 2e-05,
      "loss": 0.6403,
      "step": 4360
    },
    {
      "epoch": 0.3653540673856701,
      "grad_norm": 0.3425849378108978,
      "learning_rate": 2e-05,
      "loss": 0.6592,
      "step": 4370
    },
    {
      "epoch": 0.3661901178831201,
      "grad_norm": 0.34393274784088135,
      "learning_rate": 2e-05,
      "loss": 0.6464,
      "step": 4380
    },
    {
      "epoch": 0.3670261683805702,
      "grad_norm": 0.36734405159950256,
      "learning_rate": 2e-05,
      "loss": 0.6722,
      "step": 4390
    },
    {
      "epoch": 0.36786221887802023,
      "grad_norm": 0.3715624213218689,
      "learning_rate": 2e-05,
      "loss": 0.6579,
      "step": 4400
    },
    {
      "epoch": 0.3686982693754703,
      "grad_norm": 0.4003128111362457,
      "learning_rate": 2e-05,
      "loss": 0.7004,
      "step": 4410
    },
    {
      "epoch": 0.36953431987292035,
      "grad_norm": 0.3671753406524658,
      "learning_rate": 2e-05,
      "loss": 0.6795,
      "step": 4420
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.3676854372024536,
      "learning_rate": 2e-05,
      "loss": 0.6676,
      "step": 4430
    },
    {
      "epoch": 0.3712064208678204,
      "grad_norm": 0.34763482213020325,
      "learning_rate": 2e-05,
      "loss": 0.6921,
      "step": 4440
    },
    {
      "epoch": 0.37204247136527047,
      "grad_norm": 0.38204512000083923,
      "learning_rate": 2e-05,
      "loss": 0.6733,
      "step": 4450
    },
    {
      "epoch": 0.3728785218627205,
      "grad_norm": 0.36626923084259033,
      "learning_rate": 2e-05,
      "loss": 0.6402,
      "step": 4460
    },
    {
      "epoch": 0.3737145723601705,
      "grad_norm": 0.3997361361980438,
      "learning_rate": 2e-05,
      "loss": 0.6596,
      "step": 4470
    },
    {
      "epoch": 0.3745506228576206,
      "grad_norm": 0.3861956298351288,
      "learning_rate": 2e-05,
      "loss": 0.6494,
      "step": 4480
    },
    {
      "epoch": 0.37538667335507064,
      "grad_norm": 0.3729952871799469,
      "learning_rate": 2e-05,
      "loss": 0.6949,
      "step": 4490
    },
    {
      "epoch": 0.3762227238525207,
      "grad_norm": 0.37843218445777893,
      "learning_rate": 2e-05,
      "loss": 0.6501,
      "step": 4500
    },
    {
      "epoch": 0.37705877434997076,
      "grad_norm": 0.3461994230747223,
      "learning_rate": 2e-05,
      "loss": 0.6882,
      "step": 4510
    },
    {
      "epoch": 0.37789482484742076,
      "grad_norm": 0.37727099657058716,
      "learning_rate": 2e-05,
      "loss": 0.6639,
      "step": 4520
    },
    {
      "epoch": 0.3787308753448708,
      "grad_norm": 0.3723204433917999,
      "learning_rate": 2e-05,
      "loss": 0.6924,
      "step": 4530
    },
    {
      "epoch": 0.3795669258423209,
      "grad_norm": 0.34893015027046204,
      "learning_rate": 2e-05,
      "loss": 0.6693,
      "step": 4540
    },
    {
      "epoch": 0.38040297633977094,
      "grad_norm": 0.37652260065078735,
      "learning_rate": 2e-05,
      "loss": 0.6659,
      "step": 4550
    },
    {
      "epoch": 0.381239026837221,
      "grad_norm": 0.3943141996860504,
      "learning_rate": 2e-05,
      "loss": 0.6502,
      "step": 4560
    },
    {
      "epoch": 0.382075077334671,
      "grad_norm": 0.39593082666397095,
      "learning_rate": 2e-05,
      "loss": 0.6285,
      "step": 4570
    },
    {
      "epoch": 0.38291112783212106,
      "grad_norm": 0.3677999675273895,
      "learning_rate": 2e-05,
      "loss": 0.6215,
      "step": 4580
    },
    {
      "epoch": 0.3837471783295711,
      "grad_norm": 0.44067198038101196,
      "learning_rate": 2e-05,
      "loss": 0.659,
      "step": 4590
    },
    {
      "epoch": 0.3845832288270212,
      "grad_norm": 0.3503727316856384,
      "learning_rate": 2e-05,
      "loss": 0.6523,
      "step": 4600
    },
    {
      "epoch": 0.3854192793244712,
      "grad_norm": 0.3658464252948761,
      "learning_rate": 2e-05,
      "loss": 0.6668,
      "step": 4610
    },
    {
      "epoch": 0.38625532982192123,
      "grad_norm": 0.379948228597641,
      "learning_rate": 2e-05,
      "loss": 0.6483,
      "step": 4620
    },
    {
      "epoch": 0.3870913803193713,
      "grad_norm": 0.3512934446334839,
      "learning_rate": 2e-05,
      "loss": 0.6856,
      "step": 4630
    },
    {
      "epoch": 0.38792743081682135,
      "grad_norm": 0.3343440592288971,
      "learning_rate": 2e-05,
      "loss": 0.6485,
      "step": 4640
    },
    {
      "epoch": 0.3887634813142714,
      "grad_norm": 0.3770824372768402,
      "learning_rate": 2e-05,
      "loss": 0.6745,
      "step": 4650
    },
    {
      "epoch": 0.3895995318117214,
      "grad_norm": 0.3582046627998352,
      "learning_rate": 2e-05,
      "loss": 0.6691,
      "step": 4660
    },
    {
      "epoch": 0.39043558230917147,
      "grad_norm": 0.3880390226840973,
      "learning_rate": 2e-05,
      "loss": 0.6501,
      "step": 4670
    },
    {
      "epoch": 0.3912716328066215,
      "grad_norm": 0.35768410563468933,
      "learning_rate": 2e-05,
      "loss": 0.6634,
      "step": 4680
    },
    {
      "epoch": 0.3921076833040716,
      "grad_norm": 0.3596102297306061,
      "learning_rate": 2e-05,
      "loss": 0.6683,
      "step": 4690
    },
    {
      "epoch": 0.3929437338015216,
      "grad_norm": 0.3419645428657532,
      "learning_rate": 2e-05,
      "loss": 0.6551,
      "step": 4700
    },
    {
      "epoch": 0.39377978429897165,
      "grad_norm": 0.3907841145992279,
      "learning_rate": 2e-05,
      "loss": 0.6705,
      "step": 4710
    },
    {
      "epoch": 0.3946158347964217,
      "grad_norm": 0.35143643617630005,
      "learning_rate": 2e-05,
      "loss": 0.6407,
      "step": 4720
    },
    {
      "epoch": 0.39545188529387176,
      "grad_norm": 0.3897535800933838,
      "learning_rate": 2e-05,
      "loss": 0.6844,
      "step": 4730
    },
    {
      "epoch": 0.3962879357913218,
      "grad_norm": 0.3715405762195587,
      "learning_rate": 2e-05,
      "loss": 0.6429,
      "step": 4740
    },
    {
      "epoch": 0.3971239862887718,
      "grad_norm": 0.3395850360393524,
      "learning_rate": 2e-05,
      "loss": 0.6814,
      "step": 4750
    },
    {
      "epoch": 0.3979600367862219,
      "grad_norm": 0.3568040728569031,
      "learning_rate": 2e-05,
      "loss": 0.666,
      "step": 4760
    },
    {
      "epoch": 0.39879608728367194,
      "grad_norm": 0.6625901460647583,
      "learning_rate": 2e-05,
      "loss": 0.6961,
      "step": 4770
    },
    {
      "epoch": 0.399632137781122,
      "grad_norm": 0.41476544737815857,
      "learning_rate": 2e-05,
      "loss": 0.6464,
      "step": 4780
    },
    {
      "epoch": 0.400468188278572,
      "grad_norm": 0.3590489327907562,
      "learning_rate": 2e-05,
      "loss": 0.613,
      "step": 4790
    },
    {
      "epoch": 0.40130423877602206,
      "grad_norm": 0.37327614426612854,
      "learning_rate": 2e-05,
      "loss": 0.6438,
      "step": 4800
    },
    {
      "epoch": 0.4021402892734721,
      "grad_norm": 0.3871711194515228,
      "learning_rate": 2e-05,
      "loss": 0.6481,
      "step": 4810
    },
    {
      "epoch": 0.4029763397709222,
      "grad_norm": 0.37193727493286133,
      "learning_rate": 2e-05,
      "loss": 0.6478,
      "step": 4820
    },
    {
      "epoch": 0.40381239026837223,
      "grad_norm": 0.3502122163772583,
      "learning_rate": 2e-05,
      "loss": 0.6584,
      "step": 4830
    },
    {
      "epoch": 0.40464844076582224,
      "grad_norm": 0.39542460441589355,
      "learning_rate": 2e-05,
      "loss": 0.6797,
      "step": 4840
    },
    {
      "epoch": 0.4054844912632723,
      "grad_norm": 0.3589038550853729,
      "learning_rate": 2e-05,
      "loss": 0.6514,
      "step": 4850
    },
    {
      "epoch": 0.40632054176072235,
      "grad_norm": 0.38512277603149414,
      "learning_rate": 2e-05,
      "loss": 0.6179,
      "step": 4860
    },
    {
      "epoch": 0.4071565922581724,
      "grad_norm": 0.4002944231033325,
      "learning_rate": 2e-05,
      "loss": 0.6633,
      "step": 4870
    },
    {
      "epoch": 0.4079926427556224,
      "grad_norm": 0.36734187602996826,
      "learning_rate": 2e-05,
      "loss": 0.6271,
      "step": 4880
    },
    {
      "epoch": 0.4088286932530725,
      "grad_norm": 0.3817862868309021,
      "learning_rate": 2e-05,
      "loss": 0.6877,
      "step": 4890
    },
    {
      "epoch": 0.40966474375052253,
      "grad_norm": 0.3771727979183197,
      "learning_rate": 2e-05,
      "loss": 0.6649,
      "step": 4900
    },
    {
      "epoch": 0.4105007942479726,
      "grad_norm": 0.3870513439178467,
      "learning_rate": 2e-05,
      "loss": 0.6546,
      "step": 4910
    },
    {
      "epoch": 0.41133684474542265,
      "grad_norm": 0.34439635276794434,
      "learning_rate": 2e-05,
      "loss": 0.6159,
      "step": 4920
    },
    {
      "epoch": 0.41217289524287265,
      "grad_norm": 0.3439784049987793,
      "learning_rate": 2e-05,
      "loss": 0.6122,
      "step": 4930
    },
    {
      "epoch": 0.4130089457403227,
      "grad_norm": 0.32294797897338867,
      "learning_rate": 2e-05,
      "loss": 0.6493,
      "step": 4940
    },
    {
      "epoch": 0.41384499623777277,
      "grad_norm": 0.3561195433139801,
      "learning_rate": 2e-05,
      "loss": 0.6806,
      "step": 4950
    },
    {
      "epoch": 0.4146810467352228,
      "grad_norm": 0.37559956312179565,
      "learning_rate": 2e-05,
      "loss": 0.6382,
      "step": 4960
    },
    {
      "epoch": 0.4155170972326728,
      "grad_norm": 0.37067362666130066,
      "learning_rate": 2e-05,
      "loss": 0.6345,
      "step": 4970
    },
    {
      "epoch": 0.4163531477301229,
      "grad_norm": 0.37078824639320374,
      "learning_rate": 2e-05,
      "loss": 0.6718,
      "step": 4980
    },
    {
      "epoch": 0.41718919822757294,
      "grad_norm": 0.3785903751850128,
      "learning_rate": 2e-05,
      "loss": 0.6597,
      "step": 4990
    },
    {
      "epoch": 0.418025248725023,
      "grad_norm": 0.3805888295173645,
      "learning_rate": 2e-05,
      "loss": 0.6429,
      "step": 5000
    },
    {
      "epoch": 0.41886129922247306,
      "grad_norm": 0.3531053066253662,
      "learning_rate": 2e-05,
      "loss": 0.6455,
      "step": 5010
    },
    {
      "epoch": 0.41969734971992306,
      "grad_norm": 0.36093053221702576,
      "learning_rate": 2e-05,
      "loss": 0.6551,
      "step": 5020
    },
    {
      "epoch": 0.4205334002173731,
      "grad_norm": 0.35770487785339355,
      "learning_rate": 2e-05,
      "loss": 0.64,
      "step": 5030
    },
    {
      "epoch": 0.4213694507148232,
      "grad_norm": 0.39595308899879456,
      "learning_rate": 2e-05,
      "loss": 0.6729,
      "step": 5040
    },
    {
      "epoch": 0.42220550121227324,
      "grad_norm": 0.3384452760219574,
      "learning_rate": 2e-05,
      "loss": 0.6496,
      "step": 5050
    },
    {
      "epoch": 0.42304155170972324,
      "grad_norm": 0.34772634506225586,
      "learning_rate": 2e-05,
      "loss": 0.6019,
      "step": 5060
    },
    {
      "epoch": 0.4238776022071733,
      "grad_norm": 0.34704867005348206,
      "learning_rate": 2e-05,
      "loss": 0.6495,
      "step": 5070
    },
    {
      "epoch": 0.42471365270462336,
      "grad_norm": 0.3721390664577484,
      "learning_rate": 2e-05,
      "loss": 0.6705,
      "step": 5080
    },
    {
      "epoch": 0.4255497032020734,
      "grad_norm": 0.3732820749282837,
      "learning_rate": 1.9999810479633703e-05,
      "loss": 0.6739,
      "step": 5090
    },
    {
      "epoch": 0.4263857536995235,
      "grad_norm": 0.33602797985076904,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.6554,
      "step": 5100
    },
    {
      "epoch": 0.4272218041969735,
      "grad_norm": 0.3727985620498657,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.6699,
      "step": 5110
    },
    {
      "epoch": 0.42805785469442353,
      "grad_norm": 0.37097159028053284,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.662,
      "step": 5120
    },
    {
      "epoch": 0.4288939051918736,
      "grad_norm": 0.3508121967315674,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.6533,
      "step": 5130
    },
    {
      "epoch": 0.42972995568932365,
      "grad_norm": 0.37132754921913147,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.6514,
      "step": 5140
    },
    {
      "epoch": 0.4305660061867737,
      "grad_norm": 0.38130393624305725,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.6845,
      "step": 5150
    },
    {
      "epoch": 0.4314020566842237,
      "grad_norm": 0.49500593543052673,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.6667,
      "step": 5160
    },
    {
      "epoch": 0.43223810718167377,
      "grad_norm": 0.3573121428489685,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.6604,
      "step": 5170
    },
    {
      "epoch": 0.43307415767912383,
      "grad_norm": 0.35328567028045654,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.63,
      "step": 5180
    },
    {
      "epoch": 0.4339102081765739,
      "grad_norm": 0.3663601279258728,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.65,
      "step": 5190
    },
    {
      "epoch": 0.4347462586740239,
      "grad_norm": 0.3781638443470001,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.6569,
      "step": 5200
    },
    {
      "epoch": 0.43558230917147395,
      "grad_norm": 0.35734984278678894,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.6756,
      "step": 5210
    },
    {
      "epoch": 0.436418359668924,
      "grad_norm": 0.3732764720916748,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.6655,
      "step": 5220
    },
    {
      "epoch": 0.43725441016637406,
      "grad_norm": 0.36329156160354614,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.6586,
      "step": 5230
    },
    {
      "epoch": 0.4380904606638241,
      "grad_norm": 0.3787406086921692,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.6592,
      "step": 5240
    },
    {
      "epoch": 0.4389265111612741,
      "grad_norm": 0.36866825819015503,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.6646,
      "step": 5250
    },
    {
      "epoch": 0.4397625616587242,
      "grad_norm": 0.36877572536468506,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.6437,
      "step": 5260
    },
    {
      "epoch": 0.44059861215617424,
      "grad_norm": 0.3391215205192566,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.6348,
      "step": 5270
    },
    {
      "epoch": 0.4414346626536243,
      "grad_norm": 0.3432348370552063,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.6209,
      "step": 5280
    },
    {
      "epoch": 0.4422707131510743,
      "grad_norm": 0.33547064661979675,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.6761,
      "step": 5290
    },
    {
      "epoch": 0.44310676364852436,
      "grad_norm": 0.32597166299819946,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.6408,
      "step": 5300
    },
    {
      "epoch": 0.4439428141459744,
      "grad_norm": 0.3425397574901581,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.6492,
      "step": 5310
    },
    {
      "epoch": 0.4447788646434245,
      "grad_norm": 0.3789878487586975,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.642,
      "step": 5320
    },
    {
      "epoch": 0.44561491514087453,
      "grad_norm": 0.3498951196670532,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.6404,
      "step": 5330
    },
    {
      "epoch": 0.44645096563832454,
      "grad_norm": 0.35242897272109985,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.6955,
      "step": 5340
    },
    {
      "epoch": 0.4472870161357746,
      "grad_norm": 0.35826271772384644,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.6135,
      "step": 5350
    },
    {
      "epoch": 0.44812306663322465,
      "grad_norm": 0.3751309812068939,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.6564,
      "step": 5360
    },
    {
      "epoch": 0.4489591171306747,
      "grad_norm": 0.34536999464035034,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.6549,
      "step": 5370
    },
    {
      "epoch": 0.4497951676281247,
      "grad_norm": 0.3441609740257263,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.6459,
      "step": 5380
    },
    {
      "epoch": 0.4506312181255748,
      "grad_norm": 0.3223273456096649,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.6566,
      "step": 5390
    },
    {
      "epoch": 0.45146726862302483,
      "grad_norm": 0.36248692870140076,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.6265,
      "step": 5400
    },
    {
      "epoch": 0.4523033191204749,
      "grad_norm": 0.37122297286987305,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.6546,
      "step": 5410
    },
    {
      "epoch": 0.45313936961792495,
      "grad_norm": 0.36266106367111206,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.6395,
      "step": 5420
    },
    {
      "epoch": 0.45397542011537495,
      "grad_norm": 0.38601091504096985,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.6412,
      "step": 5430
    },
    {
      "epoch": 0.454811470612825,
      "grad_norm": 0.36704421043395996,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.6804,
      "step": 5440
    },
    {
      "epoch": 0.45564752111027507,
      "grad_norm": 0.36249980330467224,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.6873,
      "step": 5450
    },
    {
      "epoch": 0.4564835716077251,
      "grad_norm": 0.3475612699985504,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.6424,
      "step": 5460
    },
    {
      "epoch": 0.4573196221051751,
      "grad_norm": 0.35630661249160767,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.6365,
      "step": 5470
    },
    {
      "epoch": 0.4581556726026252,
      "grad_norm": 0.3406435251235962,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.6644,
      "step": 5480
    },
    {
      "epoch": 0.45899172310007524,
      "grad_norm": 0.37329667806625366,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.637,
      "step": 5490
    },
    {
      "epoch": 0.4598277735975253,
      "grad_norm": 0.3776876926422119,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.6738,
      "step": 5500
    },
    {
      "epoch": 0.46066382409497536,
      "grad_norm": 0.3520512878894806,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.6381,
      "step": 5510
    },
    {
      "epoch": 0.46149987459242536,
      "grad_norm": 0.3906896412372589,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.6373,
      "step": 5520
    },
    {
      "epoch": 0.4623359250898754,
      "grad_norm": 0.35087552666664124,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.6392,
      "step": 5530
    },
    {
      "epoch": 0.4631719755873255,
      "grad_norm": 0.34476229548454285,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.6372,
      "step": 5540
    },
    {
      "epoch": 0.46400802608477554,
      "grad_norm": 0.34274494647979736,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.6315,
      "step": 5550
    },
    {
      "epoch": 0.46484407658222554,
      "grad_norm": 0.34277746081352234,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.6244,
      "step": 5560
    },
    {
      "epoch": 0.4656801270796756,
      "grad_norm": 0.34543684124946594,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.6176,
      "step": 5570
    },
    {
      "epoch": 0.46651617757712566,
      "grad_norm": 0.35937270522117615,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.6379,
      "step": 5580
    },
    {
      "epoch": 0.4673522280745757,
      "grad_norm": 0.3285781145095825,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.6353,
      "step": 5590
    },
    {
      "epoch": 0.46793746342279074,
      "step": 5597,
      "total_flos": 2.2737574718397743e+19,
      "train_loss": 0.022054104373053354,
      "train_runtime": 1760.0088,
      "train_samples_per_second": 38.161,
      "train_steps_per_second": 3.18
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 159,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2737574718397743e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}