{
  "best_metric": 0.03463359922170639,
  "best_model_checkpoint": "saves/psy-course/Llama-3.1-8B-Instruct/train/fold10/checkpoint-1050",
  "epoch": 4.994708994708994,
  "eval_steps": 50,
  "global_step": 3245,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015392015392015393,
      "grad_norm": 4.563803195953369,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.5254,
      "step": 10
    },
    {
      "epoch": 0.030784030784030785,
      "grad_norm": 4.705235958099365,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.5811,
      "step": 20
    },
    {
      "epoch": 0.046176046176046176,
      "grad_norm": 5.674385070800781,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.3863,
      "step": 30
    },
    {
      "epoch": 0.06156806156806157,
      "grad_norm": 2.284376859664917,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.979,
      "step": 40
    },
    {
      "epoch": 0.07696007696007696,
      "grad_norm": 1.3134299516677856,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.6238,
      "step": 50
    },
    {
      "epoch": 0.07696007696007696,
      "eval_loss": 0.39111095666885376,
      "eval_runtime": 156.5798,
      "eval_samples_per_second": 7.376,
      "eval_steps_per_second": 7.376,
      "step": 50
    },
    {
      "epoch": 0.09235209235209235,
      "grad_norm": 1.310923457145691,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.4051,
      "step": 60
    },
    {
      "epoch": 0.10774410774410774,
      "grad_norm": 1.1337850093841553,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.2462,
      "step": 70
    },
    {
      "epoch": 0.12313612313612314,
      "grad_norm": 0.9682414531707764,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1889,
      "step": 80
    },
    {
      "epoch": 0.13852813852813853,
      "grad_norm": 0.8919181823730469,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.1248,
      "step": 90
    },
    {
      "epoch": 0.15392015392015393,
      "grad_norm": 0.7058643102645874,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0941,
      "step": 100
    },
    {
      "epoch": 0.15392015392015393,
      "eval_loss": 0.08446948230266571,
      "eval_runtime": 156.5295,
      "eval_samples_per_second": 7.379,
      "eval_steps_per_second": 7.379,
      "step": 100
    },
    {
      "epoch": 0.1693121693121693,
      "grad_norm": 1.5853610038757324,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0943,
      "step": 110
    },
    {
      "epoch": 0.1847041847041847,
      "grad_norm": 0.9906721711158752,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0741,
      "step": 120
    },
    {
      "epoch": 0.2000962000962001,
      "grad_norm": 1.2157135009765625,
      "learning_rate": 4e-05,
      "loss": 0.0714,
      "step": 130
    },
    {
      "epoch": 0.21548821548821548,
      "grad_norm": 0.8919536471366882,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0727,
      "step": 140
    },
    {
      "epoch": 0.23088023088023088,
      "grad_norm": 0.8456308841705322,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.087,
      "step": 150
    },
    {
      "epoch": 0.23088023088023088,
      "eval_loss": 0.05765018239617348,
      "eval_runtime": 156.5362,
      "eval_samples_per_second": 7.378,
      "eval_steps_per_second": 7.378,
      "step": 150
    },
    {
      "epoch": 0.24627224627224628,
      "grad_norm": 0.7453258633613586,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0417,
      "step": 160
    },
    {
      "epoch": 0.26166426166426165,
      "grad_norm": 0.59873366355896,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0644,
      "step": 170
    },
    {
      "epoch": 0.27705627705627706,
      "grad_norm": 0.5188323259353638,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0595,
      "step": 180
    },
    {
      "epoch": 0.29244829244829246,
      "grad_norm": 1.653311848640442,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.059,
      "step": 190
    },
    {
      "epoch": 0.30784030784030786,
      "grad_norm": 0.5800780653953552,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0643,
      "step": 200
    },
    {
      "epoch": 0.30784030784030786,
      "eval_loss": 0.05398660898208618,
      "eval_runtime": 156.5307,
      "eval_samples_per_second": 7.379,
      "eval_steps_per_second": 7.379,
      "step": 200
    },
    {
      "epoch": 0.32323232323232326,
      "grad_norm": 1.015110731124878,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0604,
      "step": 210
    },
    {
      "epoch": 0.3386243386243386,
      "grad_norm": 0.5345372557640076,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0686,
      "step": 220
    },
    {
      "epoch": 0.354016354016354,
      "grad_norm": 1.940070390701294,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0628,
      "step": 230
    },
    {
      "epoch": 0.3694083694083694,
      "grad_norm": 0.7167956233024597,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0652,
      "step": 240
    },
    {
      "epoch": 0.3848003848003848,
      "grad_norm": 0.8118871450424194,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0718,
      "step": 250
    },
    {
      "epoch": 0.3848003848003848,
      "eval_loss": 0.05503275245428085,
      "eval_runtime": 156.4628,
      "eval_samples_per_second": 7.382,
      "eval_steps_per_second": 7.382,
      "step": 250
    },
    {
      "epoch": 0.4001924001924002,
      "grad_norm": 1.6490212678909302,
      "learning_rate": 8e-05,
      "loss": 0.0569,
      "step": 260
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 0.4983098804950714,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0578,
      "step": 270
    },
    {
      "epoch": 0.43097643097643096,
      "grad_norm": 0.3780038058757782,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0483,
      "step": 280
    },
    {
      "epoch": 0.44636844636844636,
      "grad_norm": 0.40475353598594666,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0568,
      "step": 290
    },
    {
      "epoch": 0.46176046176046176,
      "grad_norm": 0.39775270223617554,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0561,
      "step": 300
    },
    {
      "epoch": 0.46176046176046176,
      "eval_loss": 0.04800727590918541,
      "eval_runtime": 156.4919,
      "eval_samples_per_second": 7.381,
      "eval_steps_per_second": 7.381,
      "step": 300
    },
    {
      "epoch": 0.47715247715247716,
      "grad_norm": 0.40493085980415344,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0445,
      "step": 310
    },
    {
      "epoch": 0.49254449254449256,
      "grad_norm": 0.6562284827232361,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0661,
      "step": 320
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 0.41162601113319397,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0617,
      "step": 330
    },
    {
      "epoch": 0.5233285233285233,
      "grad_norm": 0.36123448610305786,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0478,
      "step": 340
    },
    {
      "epoch": 0.5387205387205387,
      "grad_norm": 0.6412944197654724,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0517,
      "step": 350
    },
    {
      "epoch": 0.5387205387205387,
      "eval_loss": 0.04569457098841667,
      "eval_runtime": 156.5195,
      "eval_samples_per_second": 7.379,
      "eval_steps_per_second": 7.379,
      "step": 350
    },
    {
      "epoch": 0.5541125541125541,
      "grad_norm": 0.6163384318351746,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0455,
      "step": 360
    },
    {
      "epoch": 0.5695045695045695,
      "grad_norm": 0.5534798502922058,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0453,
      "step": 370
    },
    {
      "epoch": 0.5848965848965849,
      "grad_norm": 0.7223376035690308,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.039,
      "step": 380
    },
    {
      "epoch": 0.6002886002886003,
      "grad_norm": 1.0336365699768066,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0411,
      "step": 390
    },
    {
      "epoch": 0.6156806156806157,
      "grad_norm": 0.3568591773509979,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0473,
      "step": 400
    },
    {
      "epoch": 0.6156806156806157,
      "eval_loss": 0.044439896941185,
      "eval_runtime": 156.5607,
      "eval_samples_per_second": 7.377,
      "eval_steps_per_second": 7.377,
      "step": 400
    },
    {
      "epoch": 0.6310726310726311,
      "grad_norm": 0.679098904132843,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.058,
      "step": 410
    },
    {
      "epoch": 0.6464646464646465,
      "grad_norm": 0.5219511985778809,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0384,
      "step": 420
    },
    {
      "epoch": 0.6618566618566618,
      "grad_norm": 0.4515031576156616,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0445,
      "step": 430
    },
    {
      "epoch": 0.6772486772486772,
      "grad_norm": 0.47124460339546204,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0544,
      "step": 440
    },
    {
      "epoch": 0.6926406926406926,
      "grad_norm": 0.3413451015949249,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0486,
      "step": 450
    },
    {
      "epoch": 0.6926406926406926,
      "eval_loss": 0.0480378083884716,
      "eval_runtime": 156.5377,
      "eval_samples_per_second": 7.378,
      "eval_steps_per_second": 7.378,
      "step": 450
    },
    {
      "epoch": 0.708032708032708,
      "grad_norm": 0.603663980960846,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0506,
      "step": 460
    },
    {
      "epoch": 0.7234247234247234,
      "grad_norm": 0.39696162939071655,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0539,
      "step": 470
    },
    {
      "epoch": 0.7388167388167388,
      "grad_norm": 0.5259657502174377,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0436,
      "step": 480
    },
    {
      "epoch": 0.7542087542087542,
      "grad_norm": 0.3372630178928375,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0406,
      "step": 490
    },
    {
      "epoch": 0.7696007696007696,
      "grad_norm": 0.41772475838661194,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0579,
      "step": 500
    },
    {
      "epoch": 0.7696007696007696,
      "eval_loss": 0.044257622212171555,
      "eval_runtime": 156.5563,
      "eval_samples_per_second": 7.378,
      "eval_steps_per_second": 7.378,
      "step": 500
    },
    {
      "epoch": 0.784992784992785,
      "grad_norm": 0.36939969658851624,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0489,
      "step": 510
    },
    {
      "epoch": 0.8003848003848004,
      "grad_norm": 0.4321548640727997,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0452,
      "step": 520
    },
    {
      "epoch": 0.8157768157768158,
      "grad_norm": 0.4121263921260834,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0346,
      "step": 530
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 0.4331793189048767,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0496,
      "step": 540
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 0.23117110133171082,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0433,
      "step": 550
    },
    {
      "epoch": 0.8465608465608465,
      "eval_loss": 0.04219086840748787,
      "eval_runtime": 156.534,
      "eval_samples_per_second": 7.379,
      "eval_steps_per_second": 7.379,
      "step": 550
    },
    {
      "epoch": 0.8619528619528619,
      "grad_norm": 0.2828313410282135,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0544,
      "step": 560
    },
    {
      "epoch": 0.8773448773448773,
      "grad_norm": 0.9485271573066711,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0547,
      "step": 570
    },
    {
      "epoch": 0.8927368927368927,
      "grad_norm": 0.3164115250110626,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0386,
      "step": 580
    },
    {
      "epoch": 0.9081289081289081,
      "grad_norm": 0.20118576288223267,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0453,
      "step": 590
    },
    {
      "epoch": 0.9235209235209235,
      "grad_norm": 0.3359719216823578,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0464,
      "step": 600
    },
    {
      "epoch": 0.9235209235209235,
      "eval_loss": 0.03837878629565239,
      "eval_runtime": 156.5493,
      "eval_samples_per_second": 7.378,
      "eval_steps_per_second": 7.378,
      "step": 600
    },
    {
      "epoch": 0.9389129389129389,
      "grad_norm": 0.31199684739112854,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0456,
      "step": 610
    },
    {
      "epoch": 0.9543049543049543,
      "grad_norm": 0.26259374618530273,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0334,
      "step": 620
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.26771873235702515,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0581,
      "step": 630
    },
    {
      "epoch": 0.9850889850889851,
      "grad_norm": 0.26962846517562866,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0389,
      "step": 640
    },
    {
      "epoch": 1.0004810004810005,
      "grad_norm": 0.2485397458076477,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0542,
      "step": 650
    },
    {
      "epoch": 1.0004810004810005,
      "eval_loss": 0.03729496896266937,
      "eval_runtime": 156.5649,
      "eval_samples_per_second": 7.377,
      "eval_steps_per_second": 7.377,
      "step": 650
    },
    {
      "epoch": 1.0158730158730158,
      "grad_norm": 0.18984000384807587,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0372,
      "step": 660
    },
    {
      "epoch": 1.0312650312650313,
      "grad_norm": 0.20701906085014343,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0332,
      "step": 670
    },
    {
      "epoch": 1.0466570466570466,
      "grad_norm": 0.18550875782966614,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.034,
      "step": 680
    },
    {
      "epoch": 1.0620490620490621,
      "grad_norm": 0.4641982614994049,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0337,
      "step": 690
    },
    {
      "epoch": 1.0774410774410774,
      "grad_norm": 0.3631673753261566,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0365,
      "step": 700
    },
    {
      "epoch": 1.0774410774410774,
      "eval_loss": 0.04193466529250145,
      "eval_runtime": 156.651,
      "eval_samples_per_second": 7.373,
      "eval_steps_per_second": 7.373,
      "step": 700
    },
    {
      "epoch": 1.092833092833093,
      "grad_norm": 1.5316163301467896,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0366,
      "step": 710
    },
    {
      "epoch": 1.1082251082251082,
      "grad_norm": 1.0215851068496704,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0352,
      "step": 720
    },
    {
      "epoch": 1.1236171236171235,
      "grad_norm": 0.15777002274990082,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0456,
      "step": 730
    },
    {
      "epoch": 1.139009139009139,
      "grad_norm": 0.4358789324760437,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0424,
      "step": 740
    },
    {
      "epoch": 1.1544011544011543,
      "grad_norm": 0.36514341831207275,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0303,
      "step": 750
    },
    {
      "epoch": 1.1544011544011543,
      "eval_loss": 0.038281191140413284,
      "eval_runtime": 156.6464,
      "eval_samples_per_second": 7.373,
      "eval_steps_per_second": 7.373,
      "step": 750
    },
    {
      "epoch": 1.1697931697931698,
      "grad_norm": 0.1458541750907898,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0353,
      "step": 760
    },
    {
      "epoch": 1.1851851851851851,
      "grad_norm": 0.07907761633396149,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.027,
      "step": 770
    },
    {
      "epoch": 1.2005772005772006,
      "grad_norm": 0.1761874556541443,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0328,
      "step": 780
    },
    {
      "epoch": 1.215969215969216,
      "grad_norm": 0.40223100781440735,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0426,
      "step": 790
    },
    {
      "epoch": 1.2313612313612314,
      "grad_norm": 0.17954549193382263,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0281,
      "step": 800
    },
    {
      "epoch": 1.2313612313612314,
      "eval_loss": 0.0356939323246479,
      "eval_runtime": 156.6322,
      "eval_samples_per_second": 7.374,
      "eval_steps_per_second": 7.374,
      "step": 800
    },
    {
      "epoch": 1.2467532467532467,
      "grad_norm": 0.4290909469127655,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0447,
      "step": 810
    },
    {
      "epoch": 1.2621452621452622,
      "grad_norm": 0.7351223230361938,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0395,
      "step": 820
    },
    {
      "epoch": 1.2775372775372775,
      "grad_norm": 0.1873934119939804,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0351,
      "step": 830
    },
    {
      "epoch": 1.2929292929292928,
      "grad_norm": 0.313372403383255,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0273,
      "step": 840
    },
    {
      "epoch": 1.3083213083213083,
      "grad_norm": 0.5782054662704468,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0293,
      "step": 850
    },
    {
      "epoch": 1.3083213083213083,
      "eval_loss": 0.03599648177623749,
      "eval_runtime": 156.6059,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 850
    },
    {
      "epoch": 1.3237133237133238,
      "grad_norm": 0.3863123953342438,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0305,
      "step": 860
    },
    {
      "epoch": 1.3391053391053391,
      "grad_norm": 0.34091198444366455,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0367,
      "step": 870
    },
    {
      "epoch": 1.3544973544973544,
      "grad_norm": 0.22626757621765137,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0366,
      "step": 880
    },
    {
      "epoch": 1.36988936988937,
      "grad_norm": 0.21351377665996552,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0326,
      "step": 890
    },
    {
      "epoch": 1.3852813852813852,
      "grad_norm": 0.28559258580207825,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0408,
      "step": 900
    },
    {
      "epoch": 1.3852813852813852,
      "eval_loss": 0.03619895502924919,
      "eval_runtime": 156.5928,
      "eval_samples_per_second": 7.376,
      "eval_steps_per_second": 7.376,
      "step": 900
    },
    {
      "epoch": 1.4006734006734007,
      "grad_norm": 0.367189884185791,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0431,
      "step": 910
    },
    {
      "epoch": 1.416065416065416,
      "grad_norm": 0.2772747576236725,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0376,
      "step": 920
    },
    {
      "epoch": 1.4314574314574315,
      "grad_norm": 0.2921196520328522,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0403,
      "step": 930
    },
    {
      "epoch": 1.4468494468494468,
      "grad_norm": 0.34466344118118286,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0412,
      "step": 940
    },
    {
      "epoch": 1.4622414622414621,
      "grad_norm": 0.3713025748729706,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0358,
      "step": 950
    },
    {
      "epoch": 1.4622414622414621,
      "eval_loss": 0.03789447620511055,
      "eval_runtime": 156.6229,
      "eval_samples_per_second": 7.374,
      "eval_steps_per_second": 7.374,
      "step": 950
    },
    {
      "epoch": 1.4776334776334776,
      "grad_norm": 0.38492417335510254,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0363,
      "step": 960
    },
    {
      "epoch": 1.4930254930254931,
      "grad_norm": 0.4049034118652344,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0368,
      "step": 970
    },
    {
      "epoch": 1.5084175084175084,
      "grad_norm": 0.2996075749397278,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0401,
      "step": 980
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 0.26517441868782043,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0292,
      "step": 990
    },
    {
      "epoch": 1.5392015392015392,
      "grad_norm": 0.38277241587638855,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0277,
      "step": 1000
    },
    {
      "epoch": 1.5392015392015392,
      "eval_loss": 0.040578410029411316,
      "eval_runtime": 156.5084,
      "eval_samples_per_second": 7.38,
      "eval_steps_per_second": 7.38,
      "step": 1000
    },
    {
      "epoch": 1.5545935545935547,
      "grad_norm": 0.23427893221378326,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.035,
      "step": 1010
    },
    {
      "epoch": 1.56998556998557,
      "grad_norm": 0.26060453057289124,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0283,
      "step": 1020
    },
    {
      "epoch": 1.5853775853775853,
      "grad_norm": 0.3062327802181244,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0265,
      "step": 1030
    },
    {
      "epoch": 1.6007696007696008,
      "grad_norm": 0.45874857902526855,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0384,
      "step": 1040
    },
    {
      "epoch": 1.6161616161616161,
      "grad_norm": 0.12160841375589371,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0401,
      "step": 1050
    },
    {
      "epoch": 1.6161616161616161,
      "eval_loss": 0.03463359922170639,
      "eval_runtime": 156.5103,
      "eval_samples_per_second": 7.38,
      "eval_steps_per_second": 7.38,
      "step": 1050
    },
    {
      "epoch": 1.6315536315536314,
      "grad_norm": 0.1412663608789444,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.037,
      "step": 1060
    },
    {
      "epoch": 1.646945646945647,
      "grad_norm": 0.35587480664253235,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0419,
      "step": 1070
    },
    {
      "epoch": 1.6623376623376624,
      "grad_norm": 0.39557698369026184,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0349,
      "step": 1080
    },
    {
      "epoch": 1.6777296777296777,
      "grad_norm": 0.26560649275779724,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0239,
      "step": 1090
    },
    {
      "epoch": 1.693121693121693,
      "grad_norm": 0.39187079668045044,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0366,
      "step": 1100
    },
    {
      "epoch": 1.693121693121693,
      "eval_loss": 0.03791184723377228,
      "eval_runtime": 156.479,
      "eval_samples_per_second": 7.381,
      "eval_steps_per_second": 7.381,
      "step": 1100
    },
    {
      "epoch": 1.7085137085137085,
      "grad_norm": 0.4246823787689209,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0431,
      "step": 1110
    },
    {
      "epoch": 1.723905723905724,
      "grad_norm": 0.3322702646255493,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0274,
      "step": 1120
    },
    {
      "epoch": 1.7392977392977393,
      "grad_norm": 0.2940083146095276,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0259,
      "step": 1130
    },
    {
      "epoch": 1.7546897546897546,
      "grad_norm": 0.347748339176178,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0275,
      "step": 1140
    },
    {
      "epoch": 1.7700817700817701,
      "grad_norm": 0.3886261582374573,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0292,
      "step": 1150
    },
    {
      "epoch": 1.7700817700817701,
      "eval_loss": 0.035687752068042755,
      "eval_runtime": 156.5903,
      "eval_samples_per_second": 7.376,
      "eval_steps_per_second": 7.376,
      "step": 1150
    },
    {
      "epoch": 1.7854737854737854,
      "grad_norm": 0.40326473116874695,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0354,
      "step": 1160
    },
    {
      "epoch": 1.8008658008658007,
      "grad_norm": 0.23142275214195251,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0273,
      "step": 1170
    },
    {
      "epoch": 1.8162578162578162,
      "grad_norm": 0.24625132977962494,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.028,
      "step": 1180
    },
    {
      "epoch": 1.8316498316498318,
      "grad_norm": 0.34031790494918823,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0389,
      "step": 1190
    },
    {
      "epoch": 1.847041847041847,
      "grad_norm": 0.47231894731521606,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0408,
      "step": 1200
    },
    {
      "epoch": 1.847041847041847,
      "eval_loss": 0.03680937737226486,
      "eval_runtime": 156.6039,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1200
    },
    {
      "epoch": 1.8624338624338623,
      "grad_norm": 0.29717183113098145,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0269,
      "step": 1210
    },
    {
      "epoch": 1.8778258778258778,
      "grad_norm": 0.24861152470111847,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.034,
      "step": 1220
    },
    {
      "epoch": 1.8932178932178934,
      "grad_norm": 0.2200372815132141,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0282,
      "step": 1230
    },
    {
      "epoch": 1.9086099086099086,
      "grad_norm": 0.269031286239624,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0373,
      "step": 1240
    },
    {
      "epoch": 1.924001924001924,
      "grad_norm": 0.25449392199516296,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0354,
      "step": 1250
    },
    {
      "epoch": 1.924001924001924,
      "eval_loss": 0.03668263927102089,
      "eval_runtime": 156.5723,
      "eval_samples_per_second": 7.377,
      "eval_steps_per_second": 7.377,
      "step": 1250
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.18996648490428925,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.03,
      "step": 1260
    },
    {
      "epoch": 1.9547859547859547,
      "grad_norm": 0.17437021434307098,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0332,
      "step": 1270
    },
    {
      "epoch": 1.97017797017797,
      "grad_norm": 0.2715587317943573,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.028,
      "step": 1280
    },
    {
      "epoch": 1.9855699855699855,
      "grad_norm": 0.15359388291835785,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0434,
      "step": 1290
    },
    {
      "epoch": 2.000962000962001,
      "grad_norm": 0.19303089380264282,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0374,
      "step": 1300
    },
    {
      "epoch": 2.000962000962001,
      "eval_loss": 0.03782280907034874,
      "eval_runtime": 156.6197,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1300
    },
    {
      "epoch": 2.0163540163540166,
      "grad_norm": 0.21229633688926697,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0312,
      "step": 1310
    },
    {
      "epoch": 2.0317460317460316,
      "grad_norm": 0.1686246544122696,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0235,
      "step": 1320
    },
    {
      "epoch": 2.047138047138047,
      "grad_norm": 0.2748761475086212,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0249,
      "step": 1330
    },
    {
      "epoch": 2.0625300625300627,
      "grad_norm": 0.155365988612175,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0201,
      "step": 1340
    },
    {
      "epoch": 2.0779220779220777,
      "grad_norm": 0.24583998322486877,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0216,
      "step": 1350
    },
    {
      "epoch": 2.0779220779220777,
      "eval_loss": 0.03559431433677673,
      "eval_runtime": 156.627,
      "eval_samples_per_second": 7.374,
      "eval_steps_per_second": 7.374,
      "step": 1350
    },
    {
      "epoch": 2.0933140933140932,
      "grad_norm": 0.180336594581604,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0182,
      "step": 1360
    },
    {
      "epoch": 2.1087061087061088,
      "grad_norm": 0.2246561348438263,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0196,
      "step": 1370
    },
    {
      "epoch": 2.1240981240981243,
      "grad_norm": 0.2355770319700241,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0248,
      "step": 1380
    },
    {
      "epoch": 2.1394901394901393,
      "grad_norm": 0.18477056920528412,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0208,
      "step": 1390
    },
    {
      "epoch": 2.154882154882155,
      "grad_norm": 0.24728438258171082,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0272,
      "step": 1400
    },
    {
      "epoch": 2.154882154882155,
      "eval_loss": 0.035518452525138855,
      "eval_runtime": 156.5964,
      "eval_samples_per_second": 7.376,
      "eval_steps_per_second": 7.376,
      "step": 1400
    },
    {
      "epoch": 2.1702741702741704,
      "grad_norm": 0.3151705265045166,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0236,
      "step": 1410
    },
    {
      "epoch": 2.185666185666186,
      "grad_norm": 0.15011224150657654,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0201,
      "step": 1420
    },
    {
      "epoch": 2.201058201058201,
      "grad_norm": 0.17343637347221375,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0318,
      "step": 1430
    },
    {
      "epoch": 2.2164502164502164,
      "grad_norm": 0.2319108545780182,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0184,
      "step": 1440
    },
    {
      "epoch": 2.231842231842232,
      "grad_norm": 0.17453564703464508,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0207,
      "step": 1450
    },
    {
      "epoch": 2.231842231842232,
      "eval_loss": 0.03596349433064461,
      "eval_runtime": 156.6063,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1450
    },
    {
      "epoch": 2.247234247234247,
      "grad_norm": 0.2566794455051422,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0265,
      "step": 1460
    },
    {
      "epoch": 2.2626262626262625,
      "grad_norm": 0.2601344585418701,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0204,
      "step": 1470
    },
    {
      "epoch": 2.278018278018278,
      "grad_norm": 0.37681546807289124,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0204,
      "step": 1480
    },
    {
      "epoch": 2.2934102934102936,
      "grad_norm": 0.2275158017873764,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0261,
      "step": 1490
    },
    {
      "epoch": 2.3088023088023086,
      "grad_norm": 0.18621137738227844,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0275,
      "step": 1500
    },
    {
      "epoch": 2.3088023088023086,
      "eval_loss": 0.03567279502749443,
      "eval_runtime": 156.5383,
      "eval_samples_per_second": 7.378,
      "eval_steps_per_second": 7.378,
      "step": 1500
    },
    {
      "epoch": 2.324194324194324,
      "grad_norm": 0.2154095321893692,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0244,
      "step": 1510
    },
    {
      "epoch": 2.3395863395863397,
      "grad_norm": 0.25861501693725586,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0194,
      "step": 1520
    },
    {
      "epoch": 2.354978354978355,
      "grad_norm": 0.17612312734127045,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0293,
      "step": 1530
    },
    {
      "epoch": 2.3703703703703702,
      "grad_norm": 0.29637411236763,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0363,
      "step": 1540
    },
    {
      "epoch": 2.3857623857623858,
      "grad_norm": 0.2852693498134613,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0236,
      "step": 1550
    },
    {
      "epoch": 2.3857623857623858,
      "eval_loss": 0.03613707795739174,
      "eval_runtime": 156.7333,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 1550
    },
    {
      "epoch": 2.4011544011544013,
      "grad_norm": 0.2104523479938507,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0263,
      "step": 1560
    },
    {
      "epoch": 2.4165464165464163,
      "grad_norm": 0.11109448969364166,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0217,
      "step": 1570
    },
    {
      "epoch": 2.431938431938432,
      "grad_norm": 0.13861635327339172,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0163,
      "step": 1580
    },
    {
      "epoch": 2.4473304473304474,
      "grad_norm": 0.056167371571063995,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0202,
      "step": 1590
    },
    {
      "epoch": 2.462722462722463,
      "grad_norm": 0.2391810119152069,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0263,
      "step": 1600
    },
    {
      "epoch": 2.462722462722463,
      "eval_loss": 0.03713103383779526,
      "eval_runtime": 156.7397,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 1600
    },
    {
      "epoch": 2.478114478114478,
      "grad_norm": 0.29789450764656067,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0205,
      "step": 1610
    },
    {
      "epoch": 2.4935064935064934,
      "grad_norm": 0.5143710970878601,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0207,
      "step": 1620
    },
    {
      "epoch": 2.508898508898509,
      "grad_norm": 0.46412357687950134,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0304,
      "step": 1630
    },
    {
      "epoch": 2.5242905242905245,
      "grad_norm": 0.24112604558467865,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0278,
      "step": 1640
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 0.3198855519294739,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0306,
      "step": 1650
    },
    {
      "epoch": 2.5396825396825395,
      "eval_loss": 0.03519214317202568,
      "eval_runtime": 156.7156,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 1650
    },
    {
      "epoch": 2.555074555074555,
      "grad_norm": 0.18971195816993713,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0329,
      "step": 1660
    },
    {
      "epoch": 2.5704665704665706,
      "grad_norm": 0.17642809450626373,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0309,
      "step": 1670
    },
    {
      "epoch": 2.5858585858585856,
      "grad_norm": 0.2664608061313629,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0267,
      "step": 1680
    },
    {
      "epoch": 2.601250601250601,
      "grad_norm": 0.2245262712240219,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0172,
      "step": 1690
    },
    {
      "epoch": 2.6166426166426167,
      "grad_norm": 0.34870776534080505,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0264,
      "step": 1700
    },
    {
      "epoch": 2.6166426166426167,
      "eval_loss": 0.03493504226207733,
      "eval_runtime": 156.7115,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 1700
    },
    {
      "epoch": 2.632034632034632,
      "grad_norm": 0.09928081929683685,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0194,
      "step": 1710
    },
    {
      "epoch": 2.6474266474266477,
      "grad_norm": 0.25485026836395264,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.023,
      "step": 1720
    },
    {
      "epoch": 2.6628186628186628,
      "grad_norm": 0.35241320729255676,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0276,
      "step": 1730
    },
    {
      "epoch": 2.6782106782106783,
      "grad_norm": 0.11527998000383377,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0216,
      "step": 1740
    },
    {
      "epoch": 2.6936026936026938,
      "grad_norm": 0.4768126308917999,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0192,
      "step": 1750
    },
    {
      "epoch": 2.6936026936026938,
      "eval_loss": 0.03474767878651619,
      "eval_runtime": 156.6907,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 1750
    },
    {
      "epoch": 2.708994708994709,
      "grad_norm": 0.21746526658535004,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0177,
      "step": 1760
    },
    {
      "epoch": 2.7243867243867244,
      "grad_norm": 0.3157901167869568,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0266,
      "step": 1770
    },
    {
      "epoch": 2.73977873977874,
      "grad_norm": 0.21928253769874573,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.029,
      "step": 1780
    },
    {
      "epoch": 2.755170755170755,
      "grad_norm": 0.1379837542772293,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0166,
      "step": 1790
    },
    {
      "epoch": 2.7705627705627704,
      "grad_norm": 0.32224082946777344,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0239,
      "step": 1800
    },
    {
      "epoch": 2.7705627705627704,
      "eval_loss": 0.03590710833668709,
      "eval_runtime": 156.7143,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 1800
    },
    {
      "epoch": 2.785954785954786,
      "grad_norm": 0.36362284421920776,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0222,
      "step": 1810
    },
    {
      "epoch": 2.8013468013468015,
      "grad_norm": 0.14847242832183838,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0255,
      "step": 1820
    },
    {
      "epoch": 2.816738816738817,
      "grad_norm": 0.2280205339193344,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0293,
      "step": 1830
    },
    {
      "epoch": 2.832130832130832,
      "grad_norm": 0.3507547080516815,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0223,
      "step": 1840
    },
    {
      "epoch": 2.8475228475228476,
      "grad_norm": 0.2469240128993988,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0158,
      "step": 1850
    },
    {
      "epoch": 2.8475228475228476,
      "eval_loss": 0.037423569709062576,
      "eval_runtime": 156.6745,
      "eval_samples_per_second": 7.372,
      "eval_steps_per_second": 7.372,
      "step": 1850
    },
    {
      "epoch": 2.862914862914863,
      "grad_norm": 0.31700509786605835,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0205,
      "step": 1860
    },
    {
      "epoch": 2.878306878306878,
      "grad_norm": 0.44366294145584106,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0265,
      "step": 1870
    },
    {
      "epoch": 2.8936988936988937,
      "grad_norm": 0.2801212668418884,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0192,
      "step": 1880
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.11529967188835144,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0199,
      "step": 1890
    },
    {
      "epoch": 2.9244829244829242,
      "grad_norm": 0.14974604547023773,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.019,
      "step": 1900
    },
    {
      "epoch": 2.9244829244829242,
      "eval_loss": 0.03615748882293701,
      "eval_runtime": 156.7051,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 1900
    },
    {
      "epoch": 2.9398749398749398,
      "grad_norm": 0.23615773022174835,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0186,
      "step": 1910
    },
    {
      "epoch": 2.9552669552669553,
      "grad_norm": 0.33329370617866516,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0232,
      "step": 1920
    },
    {
      "epoch": 2.9706589706589708,
      "grad_norm": 0.2653747797012329,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0277,
      "step": 1930
    },
    {
      "epoch": 2.9860509860509863,
      "grad_norm": 0.2568110227584839,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0236,
      "step": 1940
    },
    {
      "epoch": 3.0014430014430014,
      "grad_norm": 0.40152430534362793,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0218,
      "step": 1950
    },
    {
      "epoch": 3.0014430014430014,
      "eval_loss": 0.03552442789077759,
      "eval_runtime": 156.8117,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 1950
    },
    {
      "epoch": 3.016835016835017,
      "grad_norm": 0.12220560759305954,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0158,
      "step": 1960
    },
    {
      "epoch": 3.0322270322270324,
      "grad_norm": 0.2578781843185425,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0141,
      "step": 1970
    },
    {
      "epoch": 3.0476190476190474,
      "grad_norm": 0.27527111768722534,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0109,
      "step": 1980
    },
    {
      "epoch": 3.063011063011063,
      "grad_norm": 0.07696670293807983,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0126,
      "step": 1990
    },
    {
      "epoch": 3.0784030784030785,
      "grad_norm": 0.20116227865219116,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.014,
      "step": 2000
    },
    {
      "epoch": 3.0784030784030785,
      "eval_loss": 0.04063978046178818,
      "eval_runtime": 156.8035,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 2000
    },
    {
      "epoch": 3.093795093795094,
      "grad_norm": 0.3493059575557709,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0114,
      "step": 2010
    },
    {
      "epoch": 3.109187109187109,
      "grad_norm": 0.2377541959285736,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0182,
      "step": 2020
    },
    {
      "epoch": 3.1245791245791246,
      "grad_norm": 0.5352174043655396,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0146,
      "step": 2030
    },
    {
      "epoch": 3.13997113997114,
      "grad_norm": 0.3826276361942291,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0103,
      "step": 2040
    },
    {
      "epoch": 3.1553631553631556,
      "grad_norm": 0.19637374579906464,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0125,
      "step": 2050
    },
    {
      "epoch": 3.1553631553631556,
      "eval_loss": 0.04274305701255798,
      "eval_runtime": 156.7943,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 2050
    },
    {
      "epoch": 3.1707551707551707,
      "grad_norm": 0.4286232590675354,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0118,
      "step": 2060
    },
    {
      "epoch": 3.186147186147186,
      "grad_norm": 1.5299073457717896,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0146,
      "step": 2070
    },
    {
      "epoch": 3.2015392015392017,
      "grad_norm": 0.5306471586227417,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0141,
      "step": 2080
    },
    {
      "epoch": 3.2169312169312168,
      "grad_norm": 0.21311762928962708,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.015,
      "step": 2090
    },
    {
      "epoch": 3.2323232323232323,
      "grad_norm": 0.3516111969947815,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0163,
      "step": 2100
    },
    {
      "epoch": 3.2323232323232323,
      "eval_loss": 0.04185684770345688,
      "eval_runtime": 156.7849,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2100
    },
    {
      "epoch": 3.2477152477152478,
      "grad_norm": 0.31273728609085083,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0099,
      "step": 2110
    },
    {
      "epoch": 3.2631072631072633,
      "grad_norm": 0.520576000213623,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0211,
      "step": 2120
    },
    {
      "epoch": 3.2784992784992784,
      "grad_norm": 0.1928965449333191,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0168,
      "step": 2130
    },
    {
      "epoch": 3.293891293891294,
      "grad_norm": 0.317068487405777,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0144,
      "step": 2140
    },
    {
      "epoch": 3.3092833092833094,
      "grad_norm": 0.3155690133571625,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0118,
      "step": 2150
    },
    {
      "epoch": 3.3092833092833094,
      "eval_loss": 0.0408143550157547,
      "eval_runtime": 156.8151,
      "eval_samples_per_second": 7.365,
      "eval_steps_per_second": 7.365,
      "step": 2150
    },
    {
      "epoch": 3.324675324675325,
      "grad_norm": 0.11959491670131683,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0124,
      "step": 2160
    },
    {
      "epoch": 3.34006734006734,
      "grad_norm": 0.10188659280538559,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.01,
      "step": 2170
    },
    {
      "epoch": 3.3554593554593555,
      "grad_norm": 0.3930910527706146,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0111,
      "step": 2180
    },
    {
      "epoch": 3.370851370851371,
      "grad_norm": 0.3361473083496094,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0175,
      "step": 2190
    },
    {
      "epoch": 3.386243386243386,
      "grad_norm": 0.7245724201202393,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0119,
      "step": 2200
    },
    {
      "epoch": 3.386243386243386,
      "eval_loss": 0.042220115661621094,
      "eval_runtime": 156.7815,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2200
    },
    {
      "epoch": 3.4016354016354016,
      "grad_norm": 0.13217632472515106,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0109,
      "step": 2210
    },
    {
      "epoch": 3.417027417027417,
      "grad_norm": 0.12962032854557037,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0129,
      "step": 2220
    },
    {
      "epoch": 3.4324194324194326,
      "grad_norm": 0.20555835962295532,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0101,
      "step": 2230
    },
    {
      "epoch": 3.4478114478114477,
      "grad_norm": 0.09805724769830704,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0089,
      "step": 2240
    },
    {
      "epoch": 3.463203463203463,
      "grad_norm": 0.27442285418510437,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0109,
      "step": 2250
    },
    {
      "epoch": 3.463203463203463,
      "eval_loss": 0.04533122479915619,
      "eval_runtime": 156.7825,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2250
    },
    {
      "epoch": 3.4785954785954787,
      "grad_norm": 0.14604948461055756,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.014,
      "step": 2260
    },
    {
      "epoch": 3.493987493987494,
      "grad_norm": 0.22445155680179596,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0124,
      "step": 2270
    },
    {
      "epoch": 3.5093795093795093,
      "grad_norm": 0.4208320081233978,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0177,
      "step": 2280
    },
    {
      "epoch": 3.5247715247715248,
      "grad_norm": 0.3857502341270447,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0098,
      "step": 2290
    },
    {
      "epoch": 3.5401635401635403,
      "grad_norm": 0.46608108282089233,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.013,
      "step": 2300
    },
    {
      "epoch": 3.5401635401635403,
      "eval_loss": 0.04469175264239311,
      "eval_runtime": 156.7734,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2300
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.31333157420158386,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0101,
      "step": 2310
    },
    {
      "epoch": 3.570947570947571,
      "grad_norm": 0.36925584077835083,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0137,
      "step": 2320
    },
    {
      "epoch": 3.5863395863395864,
      "grad_norm": 0.41226068139076233,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0181,
      "step": 2330
    },
    {
      "epoch": 3.601731601731602,
      "grad_norm": 0.38708868622779846,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0126,
      "step": 2340
    },
    {
      "epoch": 3.617123617123617,
      "grad_norm": 0.2930375337600708,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0108,
      "step": 2350
    },
    {
      "epoch": 3.617123617123617,
      "eval_loss": 0.042753271758556366,
      "eval_runtime": 156.7764,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2350
    },
    {
      "epoch": 3.6325156325156325,
      "grad_norm": 0.07423480600118637,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0079,
      "step": 2360
    },
    {
      "epoch": 3.647907647907648,
      "grad_norm": 0.041776180267333984,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0066,
      "step": 2370
    },
    {
      "epoch": 3.6632996632996635,
      "grad_norm": 0.1635269820690155,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0084,
      "step": 2380
    },
    {
      "epoch": 3.6786916786916786,
      "grad_norm": 0.2790094316005707,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0085,
      "step": 2390
    },
    {
      "epoch": 3.694083694083694,
      "grad_norm": 0.36584314703941345,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0132,
      "step": 2400
    },
    {
      "epoch": 3.694083694083694,
      "eval_loss": 0.047103401273489,
      "eval_runtime": 156.7842,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2400
    },
    {
      "epoch": 3.7094757094757096,
      "grad_norm": 0.08974315971136093,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0163,
      "step": 2410
    },
    {
      "epoch": 3.7248677248677247,
      "grad_norm": 0.09695596992969513,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0089,
      "step": 2420
    },
    {
      "epoch": 3.74025974025974,
      "grad_norm": 0.3750528395175934,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0119,
      "step": 2430
    },
    {
      "epoch": 3.7556517556517557,
      "grad_norm": 0.3999182879924774,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0119,
      "step": 2440
    },
    {
      "epoch": 3.771043771043771,
      "grad_norm": 0.2619767487049103,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0098,
      "step": 2450
    },
    {
      "epoch": 3.771043771043771,
      "eval_loss": 0.04452744498848915,
      "eval_runtime": 156.8112,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 2450
    },
    {
      "epoch": 3.7864357864357867,
      "grad_norm": 0.11869633942842484,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.009,
      "step": 2460
    },
    {
      "epoch": 3.801827801827802,
      "grad_norm": 0.06058422476053238,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0129,
      "step": 2470
    },
    {
      "epoch": 3.8172198172198173,
      "grad_norm": 0.6579355597496033,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0161,
      "step": 2480
    },
    {
      "epoch": 3.832611832611833,
      "grad_norm": 0.36685091257095337,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0126,
      "step": 2490
    },
    {
      "epoch": 3.848003848003848,
      "grad_norm": 0.5982416868209839,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.017,
      "step": 2500
    },
    {
      "epoch": 3.848003848003848,
      "eval_loss": 0.04097314924001694,
      "eval_runtime": 156.7852,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2500
    },
    {
      "epoch": 3.8633958633958634,
      "grad_norm": 0.18948298692703247,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0151,
      "step": 2510
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.2794782817363739,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0125,
      "step": 2520
    },
    {
      "epoch": 3.894179894179894,
      "grad_norm": 0.6842507719993591,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.009,
      "step": 2530
    },
    {
      "epoch": 3.9095719095719095,
      "grad_norm": 0.05386317893862724,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0096,
      "step": 2540
    },
    {
      "epoch": 3.924963924963925,
      "grad_norm": 1.3959304094314575,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0115,
      "step": 2550
    },
    {
      "epoch": 3.924963924963925,
      "eval_loss": 0.040816932916641235,
      "eval_runtime": 156.8304,
      "eval_samples_per_second": 7.365,
      "eval_steps_per_second": 7.365,
      "step": 2550
    },
    {
      "epoch": 3.9403559403559405,
      "grad_norm": 0.2845388650894165,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0108,
      "step": 2560
    },
    {
      "epoch": 3.955747955747956,
      "grad_norm": 0.431376576423645,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0134,
      "step": 2570
    },
    {
      "epoch": 3.971139971139971,
      "grad_norm": 0.3010033965110779,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0094,
      "step": 2580
    },
    {
      "epoch": 3.9865319865319866,
      "grad_norm": 0.11758444458246231,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.01,
      "step": 2590
    },
    {
      "epoch": 4.001924001924002,
      "grad_norm": 0.1270321011543274,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0121,
      "step": 2600
    },
    {
      "epoch": 4.001924001924002,
      "eval_loss": 0.04122837632894516,
      "eval_runtime": 156.8762,
      "eval_samples_per_second": 7.362,
      "eval_steps_per_second": 7.362,
      "step": 2600
    },
    {
      "epoch": 4.017316017316017,
      "grad_norm": 0.10859156399965286,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0063,
      "step": 2610
    },
    {
      "epoch": 4.032708032708033,
      "grad_norm": 0.18566158413887024,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0064,
      "step": 2620
    },
    {
      "epoch": 4.048100048100048,
      "grad_norm": 0.29163435101509094,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0076,
      "step": 2630
    },
    {
      "epoch": 4.063492063492063,
      "grad_norm": 0.5837296843528748,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0069,
      "step": 2640
    },
    {
      "epoch": 4.078884078884079,
      "grad_norm": 0.08933176100254059,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0061,
      "step": 2650
    },
    {
      "epoch": 4.078884078884079,
      "eval_loss": 0.04510257765650749,
      "eval_runtime": 156.7929,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 2650
    },
    {
      "epoch": 4.094276094276094,
      "grad_norm": 0.1644277721643448,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0061,
      "step": 2660
    },
    {
      "epoch": 4.109668109668109,
      "grad_norm": 0.8971966505050659,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0051,
      "step": 2670
    },
    {
      "epoch": 4.125060125060125,
      "grad_norm": 0.3398546278476715,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0082,
      "step": 2680
    },
    {
      "epoch": 4.14045214045214,
      "grad_norm": 0.09622970223426819,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0056,
      "step": 2690
    },
    {
      "epoch": 4.1558441558441555,
      "grad_norm": 0.4042757451534271,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0076,
      "step": 2700
    },
    {
      "epoch": 4.1558441558441555,
      "eval_loss": 0.048488400876522064,
      "eval_runtime": 156.8024,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 2700
    },
    {
      "epoch": 4.171236171236171,
      "grad_norm": 0.35338160395622253,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0045,
      "step": 2710
    },
    {
      "epoch": 4.1866281866281865,
      "grad_norm": 0.5565376877784729,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0036,
      "step": 2720
    },
    {
      "epoch": 4.202020202020202,
      "grad_norm": 0.47380882501602173,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0049,
      "step": 2730
    },
    {
      "epoch": 4.2174122174122175,
      "grad_norm": 0.0966838076710701,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0063,
      "step": 2740
    },
    {
      "epoch": 4.232804232804233,
      "grad_norm": 0.31144577264785767,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0066,
      "step": 2750
    },
    {
      "epoch": 4.232804232804233,
      "eval_loss": 0.05244540423154831,
      "eval_runtime": 156.7554,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 2750
    },
    {
      "epoch": 4.2481962481962485,
      "grad_norm": 0.04861610010266304,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0034,
      "step": 2760
    },
    {
      "epoch": 4.263588263588264,
      "grad_norm": 0.056783027946949005,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0037,
      "step": 2770
    },
    {
      "epoch": 4.278980278980279,
      "grad_norm": 0.03775176778435707,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0028,
      "step": 2780
    },
    {
      "epoch": 4.294372294372295,
      "grad_norm": 1.5728226900100708,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0078,
      "step": 2790
    },
    {
      "epoch": 4.30976430976431,
      "grad_norm": 0.6276909112930298,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0053,
      "step": 2800
    },
    {
      "epoch": 4.30976430976431,
      "eval_loss": 0.053448814898729324,
      "eval_runtime": 156.7877,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2800
    },
    {
      "epoch": 4.325156325156325,
      "grad_norm": 0.325838565826416,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0044,
      "step": 2810
    },
    {
      "epoch": 4.340548340548341,
      "grad_norm": 0.03447064757347107,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0034,
      "step": 2820
    },
    {
      "epoch": 4.355940355940356,
      "grad_norm": 0.12056706100702286,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0043,
      "step": 2830
    },
    {
      "epoch": 4.371332371332372,
      "grad_norm": 0.12052331864833832,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0039,
      "step": 2840
    },
    {
      "epoch": 4.386724386724387,
      "grad_norm": 0.14274635910987854,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0073,
      "step": 2850
    },
    {
      "epoch": 4.386724386724387,
      "eval_loss": 0.05407337099313736,
      "eval_runtime": 156.7258,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 2850
    },
    {
      "epoch": 4.402116402116402,
      "grad_norm": 0.4190588593482971,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0036,
      "step": 2860
    },
    {
      "epoch": 4.417508417508418,
      "grad_norm": 0.38834354281425476,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0066,
      "step": 2870
    },
    {
      "epoch": 4.432900432900433,
      "grad_norm": 0.5118524432182312,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.004,
      "step": 2880
    },
    {
      "epoch": 4.448292448292448,
      "grad_norm": 0.0421019047498703,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0054,
      "step": 2890
    },
    {
      "epoch": 4.463684463684464,
      "grad_norm": 0.2646845281124115,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0033,
      "step": 2900
    },
    {
      "epoch": 4.463684463684464,
      "eval_loss": 0.053281258791685104,
      "eval_runtime": 156.7421,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 2900
    },
    {
      "epoch": 4.479076479076479,
      "grad_norm": 0.29298534989356995,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.003,
      "step": 2910
    },
    {
      "epoch": 4.494468494468494,
      "grad_norm": 0.08777303993701935,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0021,
      "step": 2920
    },
    {
      "epoch": 4.50986050986051,
      "grad_norm": 0.09879230707883835,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0078,
      "step": 2930
    },
    {
      "epoch": 4.525252525252525,
      "grad_norm": 0.1705944687128067,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0055,
      "step": 2940
    },
    {
      "epoch": 4.540644540644541,
      "grad_norm": 0.1345970332622528,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.004,
      "step": 2950
    },
    {
      "epoch": 4.540644540644541,
      "eval_loss": 0.05370073392987251,
      "eval_runtime": 156.734,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 2950
    },
    {
      "epoch": 4.556036556036556,
      "grad_norm": 0.12359315156936646,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0049,
      "step": 2960
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.22575363516807556,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0043,
      "step": 2970
    },
    {
      "epoch": 4.586820586820587,
      "grad_norm": 0.08612951636314392,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0072,
      "step": 2980
    },
    {
      "epoch": 4.602212602212602,
      "grad_norm": 0.11356664448976517,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0042,
      "step": 2990
    },
    {
      "epoch": 4.617604617604617,
      "grad_norm": 0.40614110231399536,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0055,
      "step": 3000
    },
    {
      "epoch": 4.617604617604617,
      "eval_loss": 0.05411815643310547,
      "eval_runtime": 156.7496,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 3000
    },
    {
      "epoch": 4.632996632996633,
      "grad_norm": 1.0607129335403442,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0096,
      "step": 3010
    },
    {
      "epoch": 4.648388648388648,
      "grad_norm": 0.19804799556732178,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0077,
      "step": 3020
    },
    {
      "epoch": 4.663780663780663,
      "grad_norm": 0.5745099782943726,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.005,
      "step": 3030
    },
    {
      "epoch": 4.679172679172679,
      "grad_norm": 0.15612445771694183,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0057,
      "step": 3040
    },
    {
      "epoch": 4.694564694564694,
      "grad_norm": 0.02810909040272236,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0106,
      "step": 3050
    },
    {
      "epoch": 4.694564694564694,
      "eval_loss": 0.05451218783855438,
      "eval_runtime": 156.7077,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 3050
    },
    {
      "epoch": 4.70995670995671,
      "grad_norm": 0.01847260072827339,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0065,
      "step": 3060
    },
    {
      "epoch": 4.725348725348725,
      "grad_norm": 0.12344153970479965,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0036,
      "step": 3070
    },
    {
      "epoch": 4.7407407407407405,
      "grad_norm": 0.28908029198646545,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0048,
      "step": 3080
    },
    {
      "epoch": 4.756132756132756,
      "grad_norm": 0.2897339463233948,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0076,
      "step": 3090
    },
    {
      "epoch": 4.7715247715247715,
      "grad_norm": 0.023922229185700417,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0025,
      "step": 3100
    },
    {
      "epoch": 4.7715247715247715,
      "eval_loss": 0.05432325601577759,
      "eval_runtime": 156.7504,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 3100
    },
    {
      "epoch": 4.786916786916787,
      "grad_norm": 0.037943072617053986,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0026,
      "step": 3110
    },
    {
      "epoch": 4.8023088023088025,
      "grad_norm": 0.39157360792160034,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0059,
      "step": 3120
    },
    {
      "epoch": 4.817700817700818,
      "grad_norm": 0.04153268039226532,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0064,
      "step": 3130
    },
    {
      "epoch": 4.833092833092833,
      "grad_norm": 0.01910061202943325,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0086,
      "step": 3140
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.12265989184379578,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0053,
      "step": 3150
    },
    {
      "epoch": 4.848484848484849,
      "eval_loss": 0.054505910724401474,
      "eval_runtime": 156.7521,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 3150
    },
    {
      "epoch": 4.863876863876864,
      "grad_norm": 0.1458783745765686,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.004,
      "step": 3160
    },
    {
      "epoch": 4.87926887926888,
      "grad_norm": 0.09736804664134979,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0058,
      "step": 3170
    },
    {
      "epoch": 4.894660894660895,
      "grad_norm": 0.05497540533542633,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0035,
      "step": 3180
    },
    {
      "epoch": 4.91005291005291,
      "grad_norm": 0.08371862769126892,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0053,
      "step": 3190
    },
    {
      "epoch": 4.925444925444926,
      "grad_norm": 0.029626749455928802,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0033,
      "step": 3200
    },
    {
      "epoch": 4.925444925444926,
      "eval_loss": 0.05447264015674591,
      "eval_runtime": 156.7727,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 3200
    },
    {
      "epoch": 4.940836940836941,
      "grad_norm": 0.014000551775097847,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.01,
      "step": 3210
    },
    {
      "epoch": 4.956228956228956,
      "grad_norm": 0.25183555483818054,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0082,
      "step": 3220
    },
    {
      "epoch": 4.971620971620972,
      "grad_norm": 0.21810802817344666,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0055,
      "step": 3230
    },
    {
      "epoch": 4.987012987012987,
      "grad_norm": 0.0685734823346138,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0042,
      "step": 3240
    },
    {
      "epoch": 4.994708994708994,
      "step": 3245,
      "total_flos": 8.15387262412456e+17,
      "train_loss": 0.04645328171757869,
      "train_runtime": 35336.8306,
      "train_samples_per_second": 1.471,
      "train_steps_per_second": 0.092
    }
  ],
  "logging_steps": 10,
  "max_steps": 3245,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.15387262412456e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}