{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1812,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0055206680008281,
      "grad_norm": 2.724377425142952,
      "learning_rate": 4.945054945054946e-07,
      "loss": 0.8242,
      "step": 10
    },
    {
      "epoch": 0.0110413360016562,
      "grad_norm": 1.284291775194619,
      "learning_rate": 1.0439560439560442e-06,
      "loss": 0.8091,
      "step": 20
    },
    {
      "epoch": 0.0165620040024843,
      "grad_norm": 1.0454805966753062,
      "learning_rate": 1.5934065934065933e-06,
      "loss": 0.7388,
      "step": 30
    },
    {
      "epoch": 0.0220826720033124,
      "grad_norm": 0.7048614585560524,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 0.679,
      "step": 40
    },
    {
      "epoch": 0.0276033400041405,
      "grad_norm": 0.6074503593651366,
      "learning_rate": 2.6923076923076923e-06,
      "loss": 0.6528,
      "step": 50
    },
    {
      "epoch": 0.0331240080049686,
      "grad_norm": 0.5264552266794984,
      "learning_rate": 3.2417582417582424e-06,
      "loss": 0.6198,
      "step": 60
    },
    {
      "epoch": 0.0386446760057967,
      "grad_norm": 0.4920838737445388,
      "learning_rate": 3.7912087912087915e-06,
      "loss": 0.615,
      "step": 70
    },
    {
      "epoch": 0.0441653440066248,
      "grad_norm": 0.5309319835672981,
      "learning_rate": 4.340659340659341e-06,
      "loss": 0.5839,
      "step": 80
    },
    {
      "epoch": 0.0496860120074529,
      "grad_norm": 0.4786509013122027,
      "learning_rate": 4.890109890109891e-06,
      "loss": 0.5818,
      "step": 90
    },
    {
      "epoch": 0.055206680008281,
      "grad_norm": 0.5092441443476972,
      "learning_rate": 5.43956043956044e-06,
      "loss": 0.5811,
      "step": 100
    },
    {
      "epoch": 0.0607273480091091,
      "grad_norm": 0.4755408759173669,
      "learning_rate": 5.989010989010989e-06,
      "loss": 0.5661,
      "step": 110
    },
    {
      "epoch": 0.0662480160099372,
      "grad_norm": 0.5265777994225805,
      "learning_rate": 6.538461538461539e-06,
      "loss": 0.5648,
      "step": 120
    },
    {
      "epoch": 0.0717686840107653,
      "grad_norm": 0.5040136371199962,
      "learning_rate": 7.087912087912089e-06,
      "loss": 0.568,
      "step": 130
    },
    {
      "epoch": 0.0772893520115934,
      "grad_norm": 0.47714376645910456,
      "learning_rate": 7.637362637362638e-06,
      "loss": 0.5625,
      "step": 140
    },
    {
      "epoch": 0.0828100200124215,
      "grad_norm": 0.5243701736230444,
      "learning_rate": 8.186813186813188e-06,
      "loss": 0.5519,
      "step": 150
    },
    {
      "epoch": 0.0883306880132496,
      "grad_norm": 0.542262405438888,
      "learning_rate": 8.736263736263737e-06,
      "loss": 0.5435,
      "step": 160
    },
    {
      "epoch": 0.09385135601407771,
      "grad_norm": 0.4969530650283355,
      "learning_rate": 9.285714285714288e-06,
      "loss": 0.55,
      "step": 170
    },
    {
      "epoch": 0.0993720240149058,
      "grad_norm": 0.5287227416455338,
      "learning_rate": 9.835164835164835e-06,
      "loss": 0.5518,
      "step": 180
    },
    {
      "epoch": 0.10489269201573391,
      "grad_norm": 0.5054251237982458,
      "learning_rate": 9.999544955323775e-06,
      "loss": 0.5433,
      "step": 190
    },
    {
      "epoch": 0.110413360016562,
      "grad_norm": 0.5558917676912626,
      "learning_rate": 9.997316364460749e-06,
      "loss": 0.5481,
      "step": 200
    },
    {
      "epoch": 0.1159340280173901,
      "grad_norm": 0.5546800536713287,
      "learning_rate": 9.993231474572122e-06,
      "loss": 0.5358,
      "step": 210
    },
    {
      "epoch": 0.1214546960182182,
      "grad_norm": 0.6944982840440856,
      "learning_rate": 9.98729180302773e-06,
      "loss": 0.5483,
      "step": 220
    },
    {
      "epoch": 0.1269753640190463,
      "grad_norm": 0.526570568361616,
      "learning_rate": 9.97949955617308e-06,
      "loss": 0.5398,
      "step": 230
    },
    {
      "epoch": 0.1324960320198744,
      "grad_norm": 0.5570988628708096,
      "learning_rate": 9.969857628509778e-06,
      "loss": 0.5314,
      "step": 240
    },
    {
      "epoch": 0.13801670002070252,
      "grad_norm": 0.5543346041148294,
      "learning_rate": 9.958369601620341e-06,
      "loss": 0.5333,
      "step": 250
    },
    {
      "epoch": 0.1435373680215306,
      "grad_norm": 0.5620784833366843,
      "learning_rate": 9.945039742837788e-06,
      "loss": 0.5214,
      "step": 260
    },
    {
      "epoch": 0.1490580360223587,
      "grad_norm": 0.5584481213945615,
      "learning_rate": 9.929873003660493e-06,
      "loss": 0.5326,
      "step": 270
    },
    {
      "epoch": 0.1545787040231868,
      "grad_norm": 0.5034275806160109,
      "learning_rate": 9.912875017912917e-06,
      "loss": 0.5258,
      "step": 280
    },
    {
      "epoch": 0.16009937202401492,
      "grad_norm": 0.5144781408990675,
      "learning_rate": 9.89405209965286e-06,
      "loss": 0.5335,
      "step": 290
    },
    {
      "epoch": 0.165620040024843,
      "grad_norm": 0.5424733474681244,
      "learning_rate": 9.873411240826056e-06,
      "loss": 0.5258,
      "step": 300
    },
    {
      "epoch": 0.1711407080256711,
      "grad_norm": 0.5085645633449801,
      "learning_rate": 9.850960108668942e-06,
      "loss": 0.521,
      "step": 310
    },
    {
      "epoch": 0.1766613760264992,
      "grad_norm": 0.5053765687453395,
      "learning_rate": 9.8267070428606e-06,
      "loss": 0.5311,
      "step": 320
    },
    {
      "epoch": 0.1821820440273273,
      "grad_norm": 0.5099335256756267,
      "learning_rate": 9.800661052424903e-06,
      "loss": 0.5281,
      "step": 330
    },
    {
      "epoch": 0.18770271202815542,
      "grad_norm": 0.5528207771333934,
      "learning_rate": 9.77283181238403e-06,
      "loss": 0.5241,
      "step": 340
    },
    {
      "epoch": 0.1932233800289835,
      "grad_norm": 0.5054045853448398,
      "learning_rate": 9.743229660164594e-06,
      "loss": 0.526,
      "step": 350
    },
    {
      "epoch": 0.1987440480298116,
      "grad_norm": 0.5117420818270064,
      "learning_rate": 9.711865591757704e-06,
      "loss": 0.5103,
      "step": 360
    },
    {
      "epoch": 0.2042647160306397,
      "grad_norm": 0.5674187067979708,
      "learning_rate": 9.67875125763441e-06,
      "loss": 0.5191,
      "step": 370
    },
    {
      "epoch": 0.20978538403146782,
      "grad_norm": 0.49172722872211405,
      "learning_rate": 9.643898958418023e-06,
      "loss": 0.5241,
      "step": 380
    },
    {
      "epoch": 0.2153060520322959,
      "grad_norm": 0.5231158042562101,
      "learning_rate": 9.607321640314935e-06,
      "loss": 0.5128,
      "step": 390
    },
    {
      "epoch": 0.220826720033124,
      "grad_norm": 0.511491870556463,
      "learning_rate": 9.569032890305642e-06,
      "loss": 0.5227,
      "step": 400
    },
    {
      "epoch": 0.2263473880339521,
      "grad_norm": 0.5387512115687196,
      "learning_rate": 9.52904693109772e-06,
      "loss": 0.513,
      "step": 410
    },
    {
      "epoch": 0.2318680560347802,
      "grad_norm": 0.5834101449520662,
      "learning_rate": 9.487378615842683e-06,
      "loss": 0.5134,
      "step": 420
    },
    {
      "epoch": 0.23738872403560832,
      "grad_norm": 0.530021582877069,
      "learning_rate": 9.444043422618635e-06,
      "loss": 0.5083,
      "step": 430
    },
    {
      "epoch": 0.2429093920364364,
      "grad_norm": 0.5210909851563422,
      "learning_rate": 9.399057448680801e-06,
      "loss": 0.5125,
      "step": 440
    },
    {
      "epoch": 0.2484300600372645,
      "grad_norm": 0.5450664959987903,
      "learning_rate": 9.352437404482053e-06,
      "loss": 0.5149,
      "step": 450
    },
    {
      "epoch": 0.2539507280380926,
      "grad_norm": 0.5341386020136629,
      "learning_rate": 9.304200607465655e-06,
      "loss": 0.5146,
      "step": 460
    },
    {
      "epoch": 0.2594713960389207,
      "grad_norm": 0.5614390219052757,
      "learning_rate": 9.254364975632541e-06,
      "loss": 0.5056,
      "step": 470
    },
    {
      "epoch": 0.2649920640397488,
      "grad_norm": 0.4912341420917925,
      "learning_rate": 9.202949020885519e-06,
      "loss": 0.5047,
      "step": 480
    },
    {
      "epoch": 0.2705127320405769,
      "grad_norm": 0.632044527316854,
      "learning_rate": 9.149971842152836e-06,
      "loss": 0.5135,
      "step": 490
    },
    {
      "epoch": 0.27603340004140503,
      "grad_norm": 0.5064243623627217,
      "learning_rate": 9.095453118293724e-06,
      "loss": 0.5117,
      "step": 500
    },
    {
      "epoch": 0.2815540680422331,
      "grad_norm": 0.5246241705598543,
      "learning_rate": 9.039413100788495e-06,
      "loss": 0.4981,
      "step": 510
    },
    {
      "epoch": 0.2870747360430612,
      "grad_norm": 0.5582603492173178,
      "learning_rate": 8.98187260621595e-06,
      "loss": 0.5075,
      "step": 520
    },
    {
      "epoch": 0.2925954040438893,
      "grad_norm": 0.5167783592826254,
      "learning_rate": 8.922853008520867e-06,
      "loss": 0.5137,
      "step": 530
    },
    {
      "epoch": 0.2981160720447174,
      "grad_norm": 0.5770311356676618,
      "learning_rate": 8.862376231074452e-06,
      "loss": 0.5093,
      "step": 540
    },
    {
      "epoch": 0.30363674004554553,
      "grad_norm": 0.4977794925278661,
      "learning_rate": 8.800464738530701e-06,
      "loss": 0.5053,
      "step": 550
    },
    {
      "epoch": 0.3091574080463736,
      "grad_norm": 0.4995994131771849,
      "learning_rate": 8.737141528481705e-06,
      "loss": 0.4961,
      "step": 560
    },
    {
      "epoch": 0.3146780760472017,
      "grad_norm": 0.5542691102649394,
      "learning_rate": 8.67243012291497e-06,
      "loss": 0.5078,
      "step": 570
    },
    {
      "epoch": 0.32019874404802984,
      "grad_norm": 0.5561889774458496,
      "learning_rate": 8.606354559475975e-06,
      "loss": 0.5076,
      "step": 580
    },
    {
      "epoch": 0.3257194120488579,
      "grad_norm": 0.7114887483746155,
      "learning_rate": 8.538939382539155e-06,
      "loss": 0.511,
      "step": 590
    },
    {
      "epoch": 0.331240080049686,
      "grad_norm": 0.502243172326384,
      "learning_rate": 8.470209634090676e-06,
      "loss": 0.5071,
      "step": 600
    },
    {
      "epoch": 0.3367607480505141,
      "grad_norm": 0.6263183318774282,
      "learning_rate": 8.400190844426355e-06,
      "loss": 0.5003,
      "step": 610
    },
    {
      "epoch": 0.3422814160513422,
      "grad_norm": 0.5526970089501734,
      "learning_rate": 8.328909022668196e-06,
      "loss": 0.5077,
      "step": 620
    },
    {
      "epoch": 0.34780208405217033,
      "grad_norm": 0.5165347278260365,
      "learning_rate": 8.256390647103072e-06,
      "loss": 0.5047,
      "step": 630
    },
    {
      "epoch": 0.3533227520529984,
      "grad_norm": 0.5312278037705919,
      "learning_rate": 8.182662655347108e-06,
      "loss": 0.4994,
      "step": 640
    },
    {
      "epoch": 0.3588434200538265,
      "grad_norm": 0.5371347701765647,
      "learning_rate": 8.107752434339469e-06,
      "loss": 0.5017,
      "step": 650
    },
    {
      "epoch": 0.3643640880546546,
      "grad_norm": 0.5185838646119547,
      "learning_rate": 8.031687810169224e-06,
      "loss": 0.4958,
      "step": 660
    },
    {
      "epoch": 0.3698847560554827,
      "grad_norm": 0.5331361169549013,
      "learning_rate": 7.95449703773909e-06,
      "loss": 0.5078,
      "step": 670
    },
    {
      "epoch": 0.37540542405631083,
      "grad_norm": 0.5215524976839387,
      "learning_rate": 7.876208790269894e-06,
      "loss": 0.5049,
      "step": 680
    },
    {
      "epoch": 0.3809260920571389,
      "grad_norm": 0.5034213183694672,
      "learning_rate": 7.796852148649636e-06,
      "loss": 0.4977,
      "step": 690
    },
    {
      "epoch": 0.386446760057967,
      "grad_norm": 0.5156963187091642,
      "learning_rate": 7.716456590631125e-06,
      "loss": 0.5006,
      "step": 700
    },
    {
      "epoch": 0.3919674280587951,
      "grad_norm": 0.5285560243266483,
      "learning_rate": 7.635051979882205e-06,
      "loss": 0.4923,
      "step": 710
    },
    {
      "epoch": 0.3974880960596232,
      "grad_norm": 0.5311023206382333,
      "learning_rate": 7.552668554892614e-06,
      "loss": 0.5006,
      "step": 720
    },
    {
      "epoch": 0.40300876406045133,
      "grad_norm": 0.5186749759329228,
      "learning_rate": 7.469336917741627e-06,
      "loss": 0.4963,
      "step": 730
    },
    {
      "epoch": 0.4085294320612794,
      "grad_norm": 0.5235502510741408,
      "learning_rate": 7.385088022730624e-06,
      "loss": 0.4965,
      "step": 740
    },
    {
      "epoch": 0.4140501000621075,
      "grad_norm": 0.5353791732901545,
      "learning_rate": 7.299953164884839e-06,
      "loss": 0.4951,
      "step": 750
    },
    {
      "epoch": 0.41957076806293564,
      "grad_norm": 0.5295328402639815,
      "learning_rate": 7.213963968328526e-06,
      "loss": 0.4892,
      "step": 760
    },
    {
      "epoch": 0.4250914360637637,
      "grad_norm": 0.6110821094395759,
      "learning_rate": 7.1271523745379e-06,
      "loss": 0.5004,
      "step": 770
    },
    {
      "epoch": 0.4306121040645918,
      "grad_norm": 0.5291473583910687,
      "learning_rate": 7.039550630476171e-06,
      "loss": 0.4931,
      "step": 780
    },
    {
      "epoch": 0.4361327720654199,
      "grad_norm": 0.5267918375166477,
      "learning_rate": 6.951191276615123e-06,
      "loss": 0.4921,
      "step": 790
    },
    {
      "epoch": 0.441653440066248,
      "grad_norm": 0.5649494943921666,
      "learning_rate": 6.8621071348476485e-06,
      "loss": 0.5004,
      "step": 800
    },
    {
      "epoch": 0.44717410806707614,
      "grad_norm": 0.5360516130970742,
      "learning_rate": 6.772331296295763e-06,
      "loss": 0.4891,
      "step": 810
    },
    {
      "epoch": 0.4526947760679042,
      "grad_norm": 0.6109807883069402,
      "learning_rate": 6.681897109018599e-06,
      "loss": 0.4911,
      "step": 820
    },
    {
      "epoch": 0.4582154440687323,
      "grad_norm": 0.5019964557525745,
      "learning_rate": 6.590838165624966e-06,
      "loss": 0.495,
      "step": 830
    },
    {
      "epoch": 0.4637361120695604,
      "grad_norm": 0.5966909096605543,
      "learning_rate": 6.4991882907950624e-06,
      "loss": 0.4917,
      "step": 840
    },
    {
      "epoch": 0.4692567800703885,
      "grad_norm": 0.523483010734906,
      "learning_rate": 6.4069815287159975e-06,
      "loss": 0.4906,
      "step": 850
    },
    {
      "epoch": 0.47477744807121663,
      "grad_norm": 0.5316609125694963,
      "learning_rate": 6.314252130435758e-06,
      "loss": 0.4896,
      "step": 860
    },
    {
      "epoch": 0.4802981160720447,
      "grad_norm": 0.6821071669028048,
      "learning_rate": 6.2210345411403416e-06,
      "loss": 0.4866,
      "step": 870
    },
    {
      "epoch": 0.4858187840728728,
      "grad_norm": 0.5393653005940481,
      "learning_rate": 6.127363387358782e-06,
      "loss": 0.4917,
      "step": 880
    },
    {
      "epoch": 0.49133945207370094,
      "grad_norm": 0.5929416546820472,
      "learning_rate": 6.033273464100801e-06,
      "loss": 0.4986,
      "step": 890
    },
    {
      "epoch": 0.496860120074529,
      "grad_norm": 0.5639784561570959,
      "learning_rate": 5.938799721931889e-06,
      "loss": 0.4911,
      "step": 900
    },
    {
      "epoch": 0.5023807880753571,
      "grad_norm": 0.5535736820999778,
      "learning_rate": 5.84397725399059e-06,
      "loss": 0.4876,
      "step": 910
    },
    {
      "epoch": 0.5079014560761852,
      "grad_norm": 0.5048317080894295,
      "learning_rate": 5.748841282952837e-06,
      "loss": 0.4869,
      "step": 920
    },
    {
      "epoch": 0.5134221240770133,
      "grad_norm": 0.5869718228532016,
      "learning_rate": 5.653427147948167e-06,
      "loss": 0.4914,
      "step": 930
    },
    {
      "epoch": 0.5189427920778414,
      "grad_norm": 0.6006831300391656,
      "learning_rate": 5.557770291432678e-06,
      "loss": 0.4852,
      "step": 940
    },
    {
      "epoch": 0.5244634600786695,
      "grad_norm": 0.5182188267121967,
      "learning_rate": 5.461906246023592e-06,
      "loss": 0.4905,
      "step": 950
    },
    {
      "epoch": 0.5299841280794976,
      "grad_norm": 0.5252415500292095,
      "learning_rate": 5.365870621300354e-06,
      "loss": 0.4764,
      "step": 960
    },
    {
      "epoch": 0.5355047960803258,
      "grad_norm": 0.5702450851595816,
      "learning_rate": 5.2696990905771196e-06,
      "loss": 0.4853,
      "step": 970
    },
    {
      "epoch": 0.5410254640811538,
      "grad_norm": 0.5361466303695557,
      "learning_rate": 5.173427377651572e-06,
      "loss": 0.4927,
      "step": 980
    },
    {
      "epoch": 0.5465461320819819,
      "grad_norm": 0.5125434509749757,
      "learning_rate": 5.077091243534996e-06,
      "loss": 0.4814,
      "step": 990
    },
    {
      "epoch": 0.5520668000828101,
      "grad_norm": 0.5471748712593393,
      "learning_rate": 4.9807264731685174e-06,
      "loss": 0.4863,
      "step": 1000
    },
    {
      "epoch": 0.5575874680836381,
      "grad_norm": 0.6057904832088499,
      "learning_rate": 4.884368862130459e-06,
      "loss": 0.4784,
      "step": 1010
    },
    {
      "epoch": 0.5631081360844662,
      "grad_norm": 0.5329865888930896,
      "learning_rate": 4.7880542033397466e-06,
      "loss": 0.4944,
      "step": 1020
    },
    {
      "epoch": 0.5686288040852944,
      "grad_norm": 0.6261704093513056,
      "learning_rate": 4.691818273760302e-06,
      "loss": 0.4855,
      "step": 1030
    },
    {
      "epoch": 0.5741494720861224,
      "grad_norm": 0.5634689358774165,
      "learning_rate": 4.595696821111354e-06,
      "loss": 0.4826,
      "step": 1040
    },
    {
      "epoch": 0.5796701400869505,
      "grad_norm": 0.5014920955789893,
      "learning_rate": 4.499725550588626e-06,
      "loss": 0.4847,
      "step": 1050
    },
    {
      "epoch": 0.5851908080877786,
      "grad_norm": 0.5386330962726379,
      "learning_rate": 4.4039401116013195e-06,
      "loss": 0.4781,
      "step": 1060
    },
    {
      "epoch": 0.5907114760886067,
      "grad_norm": 0.5760448552179569,
      "learning_rate": 4.3083760845297956e-06,
      "loss": 0.4892,
      "step": 1070
    },
    {
      "epoch": 0.5962321440894348,
      "grad_norm": 0.5015790796578021,
      "learning_rate": 4.213068967508932e-06,
      "loss": 0.481,
      "step": 1080
    },
    {
      "epoch": 0.6017528120902629,
      "grad_norm": 0.5655663599339508,
      "learning_rate": 4.118054163242015e-06,
      "loss": 0.4803,
      "step": 1090
    },
    {
      "epoch": 0.6072734800910911,
      "grad_norm": 0.5302180266189818,
      "learning_rate": 4.023366965850071e-06,
      "loss": 0.4883,
      "step": 1100
    },
    {
      "epoch": 0.6127941480919191,
      "grad_norm": 0.5781982106959161,
      "learning_rate": 3.929042547761544e-06,
      "loss": 0.4782,
      "step": 1110
    },
    {
      "epoch": 0.6183148160927472,
      "grad_norm": 0.5245772049540094,
      "learning_rate": 3.835115946647182e-06,
      "loss": 0.4843,
      "step": 1120
    },
    {
      "epoch": 0.6238354840935754,
      "grad_norm": 0.5269556201169964,
      "learning_rate": 3.741622052404964e-06,
      "loss": 0.4843,
      "step": 1130
    },
    {
      "epoch": 0.6293561520944034,
      "grad_norm": 0.5445796305907685,
      "learning_rate": 3.6485955941999286e-06,
      "loss": 0.4744,
      "step": 1140
    },
    {
      "epoch": 0.6348768200952315,
      "grad_norm": 0.5175498662025362,
      "learning_rate": 3.5560711275637137e-06,
      "loss": 0.4814,
      "step": 1150
    },
    {
      "epoch": 0.6403974880960597,
      "grad_norm": 0.5391040858276955,
      "learning_rate": 3.4640830215585786e-06,
      "loss": 0.4782,
      "step": 1160
    },
    {
      "epoch": 0.6459181560968877,
      "grad_norm": 0.5270072708977098,
      "learning_rate": 3.372665446010701e-06,
      "loss": 0.4781,
      "step": 1170
    },
    {
      "epoch": 0.6514388240977158,
      "grad_norm": 0.583247698056478,
      "learning_rate": 3.2818523588174876e-06,
      "loss": 0.4773,
      "step": 1180
    },
    {
      "epoch": 0.6569594920985439,
      "grad_norm": 0.6274598734421784,
      "learning_rate": 3.191677493333603e-06,
      "loss": 0.4773,
      "step": 1190
    },
    {
      "epoch": 0.662480160099372,
      "grad_norm": 0.5118669003244579,
      "learning_rate": 3.1021743458403907e-06,
      "loss": 0.4842,
      "step": 1200
    },
    {
      "epoch": 0.6680008281002001,
      "grad_norm": 0.5443705254850117,
      "learning_rate": 3.0133761631033887e-06,
      "loss": 0.4759,
      "step": 1210
    },
    {
      "epoch": 0.6735214961010282,
      "grad_norm": 0.5390220352615862,
      "learning_rate": 2.9253159300225067e-06,
      "loss": 0.4858,
      "step": 1220
    },
    {
      "epoch": 0.6790421641018564,
      "grad_norm": 0.6953460123555169,
      "learning_rate": 2.8380263573794775e-06,
      "loss": 0.482,
      "step": 1230
    },
    {
      "epoch": 0.6845628321026844,
      "grad_norm": 0.5352590150971323,
      "learning_rate": 2.7515398696871452e-06,
      "loss": 0.4752,
      "step": 1240
    },
    {
      "epoch": 0.6900835001035125,
      "grad_norm": 0.5423898570170415,
      "learning_rate": 2.66588859314508e-06,
      "loss": 0.4803,
      "step": 1250
    },
    {
      "epoch": 0.6956041681043407,
      "grad_norm": 0.5522146186825717,
      "learning_rate": 2.5811043437060023e-06,
      "loss": 0.4747,
      "step": 1260
    },
    {
      "epoch": 0.7011248361051687,
      "grad_norm": 0.533222796025025,
      "learning_rate": 2.497218615257458e-06,
      "loss": 0.4791,
      "step": 1270
    },
    {
      "epoch": 0.7066455041059968,
      "grad_norm": 0.5227694810914811,
      "learning_rate": 2.4142625679231267e-06,
      "loss": 0.4754,
      "step": 1280
    },
    {
      "epoch": 0.712166172106825,
      "grad_norm": 0.5269824409460636,
      "learning_rate": 2.332267016488109e-06,
      "loss": 0.4751,
      "step": 1290
    },
    {
      "epoch": 0.717686840107653,
      "grad_norm": 0.5215977920809132,
      "learning_rate": 2.251262418952482e-06,
      "loss": 0.4893,
      "step": 1300
    },
    {
      "epoch": 0.7232075081084811,
      "grad_norm": 0.5482175622357263,
      "learning_rate": 2.171278865217409e-06,
      "loss": 0.4754,
      "step": 1310
    },
    {
      "epoch": 0.7287281761093092,
      "grad_norm": 0.5090995315514565,
      "learning_rate": 2.0923460659079587e-06,
      "loss": 0.4784,
      "step": 1320
    },
    {
      "epoch": 0.7342488441101374,
      "grad_norm": 0.5207395373957621,
      "learning_rate": 2.0144933413368254e-06,
      "loss": 0.4775,
      "step": 1330
    },
    {
      "epoch": 0.7397695121109654,
      "grad_norm": 0.5272338452800517,
      "learning_rate": 1.9377496106130357e-06,
      "loss": 0.4688,
      "step": 1340
    },
    {
      "epoch": 0.7452901801117935,
      "grad_norm": 0.531795809840052,
      "learning_rate": 1.8621433808996641e-06,
      "loss": 0.4833,
      "step": 1350
    },
    {
      "epoch": 0.7508108481126217,
      "grad_norm": 0.5974740783570726,
      "learning_rate": 1.7877027368246048e-06,
      "loss": 0.4804,
      "step": 1360
    },
    {
      "epoch": 0.7563315161134497,
      "grad_norm": 0.5445997748333807,
      "learning_rate": 1.7144553300482659e-06,
      "loss": 0.4837,
      "step": 1370
    },
    {
      "epoch": 0.7618521841142778,
      "grad_norm": 0.5069032715629829,
      "learning_rate": 1.6424283689921089e-06,
      "loss": 0.4769,
      "step": 1380
    },
    {
      "epoch": 0.767372852115106,
      "grad_norm": 0.5712897928998661,
      "learning_rate": 1.5716486087318323e-06,
      "loss": 0.4764,
      "step": 1390
    },
    {
      "epoch": 0.772893520115934,
      "grad_norm": 0.5382111895501923,
      "learning_rate": 1.5021423410589564e-06,
      "loss": 0.4758,
      "step": 1400
    },
    {
      "epoch": 0.7784141881167621,
      "grad_norm": 0.5266690807886757,
      "learning_rate": 1.4339353847144894e-06,
      "loss": 0.4739,
      "step": 1410
    },
    {
      "epoch": 0.7839348561175902,
      "grad_norm": 0.5164853294081637,
      "learning_rate": 1.3670530757983226e-06,
      "loss": 0.4743,
      "step": 1420
    },
    {
      "epoch": 0.7894555241184184,
      "grad_norm": 0.5447585865471612,
      "learning_rate": 1.301520258357903e-06,
      "loss": 0.4794,
      "step": 1430
    },
    {
      "epoch": 0.7949761921192464,
      "grad_norm": 0.5660764549380585,
      "learning_rate": 1.2373612751596859e-06,
      "loss": 0.477,
      "step": 1440
    },
    {
      "epoch": 0.8004968601200745,
      "grad_norm": 0.5798095805828375,
      "learning_rate": 1.1745999586467903e-06,
      "loss": 0.48,
      "step": 1450
    },
    {
      "epoch": 0.8060175281209027,
      "grad_norm": 0.5234681928353129,
      "learning_rate": 1.1132596220862136e-06,
      "loss": 0.4698,
      "step": 1460
    },
    {
      "epoch": 0.8115381961217307,
      "grad_norm": 0.5153955411762714,
      "learning_rate": 1.0533630509089144e-06,
      "loss": 0.4769,
      "step": 1470
    },
    {
      "epoch": 0.8170588641225588,
      "grad_norm": 0.9088724418581628,
      "learning_rate": 9.949324942459455e-07,
      "loss": 0.4723,
      "step": 1480
    },
    {
      "epoch": 0.822579532123387,
      "grad_norm": 0.739958257902892,
      "learning_rate": 9.379896566638153e-07,
      "loss": 0.4756,
      "step": 1490
    },
    {
      "epoch": 0.828100200124215,
      "grad_norm": 0.5175483705269135,
      "learning_rate": 8.825556901021282e-07,
      "loss": 0.4733,
      "step": 1500
    },
    {
      "epoch": 0.8336208681250431,
      "grad_norm": 0.530633488925471,
      "learning_rate": 8.28651186016498e-07,
      "loss": 0.472,
      "step": 1510
    },
    {
      "epoch": 0.8391415361258713,
      "grad_norm": 0.5147000387470615,
      "learning_rate": 7.762961677296677e-07,
      "loss": 0.4712,
      "step": 1520
    },
    {
      "epoch": 0.8446622041266993,
      "grad_norm": 0.5225729683409204,
      "learning_rate": 7.255100829936606e-07,
      "loss": 0.4771,
      "step": 1530
    },
    {
      "epoch": 0.8501828721275274,
      "grad_norm": 0.6438825817296967,
      "learning_rate": 6.763117967657307e-07,
      "loss": 0.4709,
      "step": 1540
    },
    {
      "epoch": 0.8557035401283555,
      "grad_norm": 0.5261153410217088,
      "learning_rate": 6.287195842008054e-07,
      "loss": 0.4789,
      "step": 1550
    },
    {
      "epoch": 0.8612242081291837,
      "grad_norm": 0.608087588319129,
      "learning_rate": 5.82751123863014e-07,
      "loss": 0.469,
      "step": 1560
    },
    {
      "epoch": 0.8667448761300117,
      "grad_norm": 0.5522679086935413,
      "learning_rate": 5.384234911588154e-07,
      "loss": 0.4787,
      "step": 1570
    },
    {
      "epoch": 0.8722655441308398,
      "grad_norm": 0.5508297291855024,
      "learning_rate": 4.957531519941855e-07,
      "loss": 0.4709,
      "step": 1580
    },
    {
      "epoch": 0.877786212131668,
      "grad_norm": 0.522903531538663,
      "learning_rate": 4.5475595665820995e-07,
      "loss": 0.4726,
      "step": 1590
    },
    {
      "epoch": 0.883306880132496,
      "grad_norm": 0.5150594352732931,
      "learning_rate": 4.154471339353378e-07,
      "loss": 0.4753,
      "step": 1600
    },
    {
      "epoch": 0.8888275481333241,
      "grad_norm": 0.5136068316306784,
      "learning_rate": 3.7784128544852084e-07,
      "loss": 0.4746,
      "step": 1610
    },
    {
      "epoch": 0.8943482161341523,
      "grad_norm": 0.5352167000548326,
      "learning_rate": 3.4195238023530254e-07,
      "loss": 0.4756,
      "step": 1620
    },
    {
      "epoch": 0.8998688841349803,
      "grad_norm": 0.5310346870141006,
      "learning_rate": 3.077937495588862e-07,
      "loss": 0.4749,
      "step": 1630
    },
    {
      "epoch": 0.9053895521358084,
      "grad_norm": 0.5223577826342829,
      "learning_rate": 2.7537808195612306e-07,
      "loss": 0.4749,
      "step": 1640
    },
    {
      "epoch": 0.9109102201366366,
      "grad_norm": 0.520159607290227,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.475,
      "step": 1650
    },
    {
      "epoch": 0.9164308881374646,
      "grad_norm": 0.5310673368872487,
      "learning_rate": 2.1582314844802997e-07,
      "loss": 0.4687,
      "step": 1660
    },
    {
      "epoch": 0.9219515561382927,
      "grad_norm": 0.5102531578109037,
      "learning_rate": 1.8870600476930678e-07,
      "loss": 0.4698,
      "step": 1670
    },
    {
      "epoch": 0.9274722241391208,
      "grad_norm": 0.5005757503772048,
      "learning_rate": 1.6337606039994625e-07,
      "loss": 0.4757,
      "step": 1680
    },
    {
      "epoch": 0.932992892139949,
      "grad_norm": 0.5381072393037725,
      "learning_rate": 1.3984272438024526e-07,
      "loss": 0.4809,
      "step": 1690
    },
    {
      "epoch": 0.938513560140777,
      "grad_norm": 0.5179279274433923,
      "learning_rate": 1.1811473838384214e-07,
      "loss": 0.4765,
      "step": 1700
    },
    {
      "epoch": 0.9440342281416051,
      "grad_norm": 0.5085578726821878,
      "learning_rate": 9.8200173470539e-08,
      "loss": 0.4645,
      "step": 1710
    },
    {
      "epoch": 0.9495548961424333,
      "grad_norm": 0.5279637024084495,
      "learning_rate": 8.010642708823623e-08,
      "loss": 0.4769,
      "step": 1720
    },
    {
      "epoch": 0.9550755641432613,
      "grad_norm": 0.5416761249540766,
      "learning_rate": 6.38402203250793e-08,
      "loss": 0.4784,
      "step": 1730
    },
    {
      "epoch": 0.9605962321440894,
      "grad_norm": 0.5624479312474986,
      "learning_rate": 4.940759541285145e-08,
      "loss": 0.4812,
      "step": 1740
    },
    {
      "epoch": 0.9661169001449176,
      "grad_norm": 0.5436378053154044,
      "learning_rate": 3.6813913482528495e-08,
      "loss": 0.4716,
      "step": 1750
    },
    {
      "epoch": 0.9716375681457456,
      "grad_norm": 0.5274588778397284,
      "learning_rate": 2.606385257284072e-08,
      "loss": 0.4738,
      "step": 1760
    },
    {
      "epoch": 0.9771582361465737,
      "grad_norm": 0.5368327855089428,
      "learning_rate": 1.7161405892568983e-08,
      "loss": 0.4696,
      "step": 1770
    },
    {
      "epoch": 0.9826789041474019,
      "grad_norm": 0.4844347591390851,
      "learning_rate": 1.0109880337234036e-08,
      "loss": 0.4753,
      "step": 1780
    },
    {
      "epoch": 0.98819957214823,
      "grad_norm": 0.5082694682505529,
      "learning_rate": 4.9118952607168926e-09,
      "loss": 0.476,
      "step": 1790
    },
    {
      "epoch": 0.993720240149058,
      "grad_norm": 0.5315552809684106,
      "learning_rate": 1.5693815022788105e-09,
      "loss": 0.4785,
      "step": 1800
    },
    {
      "epoch": 0.9992409081498861,
      "grad_norm": 0.5808324665139376,
      "learning_rate": 8.358066933000786e-11,
      "loss": 0.4757,
      "step": 1810
    },
    {
      "epoch": 1.0,
      "step": 1812,
      "total_flos": 4.5944668897526415e+18,
      "train_loss": 0.5047884532486893,
      "train_runtime": 92781.3825,
      "train_samples_per_second": 2.499,
      "train_steps_per_second": 0.02
    }
  ],
  "logging_steps": 10,
  "max_steps": 1812,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.5944668897526415e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}