{
  "best_metric": 0.029672790318727493,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold10/checkpoint-1700",
  "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": 1.982912302017212,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.6846,
      "step": 10
    },
    {
      "epoch": 0.030784030784030785,
      "grad_norm": 2.081594467163086,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.7088,
      "step": 20
    },
    {
      "epoch": 0.046176046176046176,
      "grad_norm": 2.1669394969940186,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.6361,
      "step": 30
    },
    {
      "epoch": 0.06156806156806157,
      "grad_norm": 2.7098140716552734,
      "learning_rate": 1.230769230769231e-05,
      "loss": 1.3897,
      "step": 40
    },
    {
      "epoch": 0.07696007696007696,
      "grad_norm": 0.8535429835319519,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.8796,
      "step": 50
    },
    {
      "epoch": 0.07696007696007696,
      "eval_loss": 0.6148930191993713,
      "eval_runtime": 197.6559,
      "eval_samples_per_second": 5.843,
      "eval_steps_per_second": 5.843,
      "step": 50
    },
    {
      "epoch": 0.09235209235209235,
      "grad_norm": 0.8714393973350525,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.6706,
      "step": 60
    },
    {
      "epoch": 0.10774410774410774,
      "grad_norm": 1.0692670345306396,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.4537,
      "step": 70
    },
    {
      "epoch": 0.12313612313612314,
      "grad_norm": 0.6447597742080688,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.3221,
      "step": 80
    },
    {
      "epoch": 0.13852813852813853,
      "grad_norm": 0.5552180409431458,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.1961,
      "step": 90
    },
    {
      "epoch": 0.15392015392015393,
      "grad_norm": 0.47883060574531555,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.1225,
      "step": 100
    },
    {
      "epoch": 0.15392015392015393,
      "eval_loss": 0.10787783563137054,
      "eval_runtime": 197.4679,
      "eval_samples_per_second": 5.849,
      "eval_steps_per_second": 5.849,
      "step": 100
    },
    {
      "epoch": 0.1693121693121693,
      "grad_norm": 0.5129164457321167,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.1266,
      "step": 110
    },
    {
      "epoch": 0.1847041847041847,
      "grad_norm": 0.7035605311393738,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.095,
      "step": 120
    },
    {
      "epoch": 0.2000962000962001,
      "grad_norm": 0.7644175291061401,
      "learning_rate": 4e-05,
      "loss": 0.0803,
      "step": 130
    },
    {
      "epoch": 0.21548821548821548,
      "grad_norm": 1.2586270570755005,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0952,
      "step": 140
    },
    {
      "epoch": 0.23088023088023088,
      "grad_norm": 0.811309814453125,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0896,
      "step": 150
    },
    {
      "epoch": 0.23088023088023088,
      "eval_loss": 0.061511509120464325,
      "eval_runtime": 196.9347,
      "eval_samples_per_second": 5.865,
      "eval_steps_per_second": 5.865,
      "step": 150
    },
    {
      "epoch": 0.24627224627224628,
      "grad_norm": 0.5694608688354492,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0479,
      "step": 160
    },
    {
      "epoch": 0.26166426166426165,
      "grad_norm": 0.6536459922790527,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0745,
      "step": 170
    },
    {
      "epoch": 0.27705627705627706,
      "grad_norm": 0.26454228162765503,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.062,
      "step": 180
    },
    {
      "epoch": 0.29244829244829246,
      "grad_norm": 0.9954192042350769,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0677,
      "step": 190
    },
    {
      "epoch": 0.30784030784030786,
      "grad_norm": 0.3866685628890991,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0656,
      "step": 200
    },
    {
      "epoch": 0.30784030784030786,
      "eval_loss": 0.05193961039185524,
      "eval_runtime": 196.7137,
      "eval_samples_per_second": 5.871,
      "eval_steps_per_second": 5.871,
      "step": 200
    },
    {
      "epoch": 0.32323232323232326,
      "grad_norm": 0.5789927244186401,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0635,
      "step": 210
    },
    {
      "epoch": 0.3386243386243386,
      "grad_norm": 0.884029746055603,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0705,
      "step": 220
    },
    {
      "epoch": 0.354016354016354,
      "grad_norm": 0.9773724675178528,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.058,
      "step": 230
    },
    {
      "epoch": 0.3694083694083694,
      "grad_norm": 0.614799976348877,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0615,
      "step": 240
    },
    {
      "epoch": 0.3848003848003848,
      "grad_norm": 0.492006778717041,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0599,
      "step": 250
    },
    {
      "epoch": 0.3848003848003848,
      "eval_loss": 0.048700157552957535,
      "eval_runtime": 196.4673,
      "eval_samples_per_second": 5.879,
      "eval_steps_per_second": 5.879,
      "step": 250
    },
    {
      "epoch": 0.4001924001924002,
      "grad_norm": 0.32661059498786926,
      "learning_rate": 8e-05,
      "loss": 0.0589,
      "step": 260
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 0.5522665977478027,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0543,
      "step": 270
    },
    {
      "epoch": 0.43097643097643096,
      "grad_norm": 0.27828100323677063,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0418,
      "step": 280
    },
    {
      "epoch": 0.44636844636844636,
      "grad_norm": 0.3976594805717468,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.056,
      "step": 290
    },
    {
      "epoch": 0.46176046176046176,
      "grad_norm": 0.313279390335083,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0559,
      "step": 300
    },
    {
      "epoch": 0.46176046176046176,
      "eval_loss": 0.0453476682305336,
      "eval_runtime": 196.6037,
      "eval_samples_per_second": 5.875,
      "eval_steps_per_second": 5.875,
      "step": 300
    },
    {
      "epoch": 0.47715247715247716,
      "grad_norm": 0.20976056158542633,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.047,
      "step": 310
    },
    {
      "epoch": 0.49254449254449256,
      "grad_norm": 0.3835485577583313,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.048,
      "step": 320
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 0.2070256471633911,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.065,
      "step": 330
    },
    {
      "epoch": 0.5233285233285233,
      "grad_norm": 0.17644135653972626,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0422,
      "step": 340
    },
    {
      "epoch": 0.5387205387205387,
      "grad_norm": 0.4333091378211975,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.047,
      "step": 350
    },
    {
      "epoch": 0.5387205387205387,
      "eval_loss": 0.04271285608410835,
      "eval_runtime": 196.4264,
      "eval_samples_per_second": 5.88,
      "eval_steps_per_second": 5.88,
      "step": 350
    },
    {
      "epoch": 0.5541125541125541,
      "grad_norm": 0.39647218585014343,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0413,
      "step": 360
    },
    {
      "epoch": 0.5695045695045695,
      "grad_norm": 0.4698505699634552,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0399,
      "step": 370
    },
    {
      "epoch": 0.5848965848965849,
      "grad_norm": 0.4131832420825958,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0383,
      "step": 380
    },
    {
      "epoch": 0.6002886002886003,
      "grad_norm": 0.5616227984428406,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0396,
      "step": 390
    },
    {
      "epoch": 0.6156806156806157,
      "grad_norm": 0.1911724954843521,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0386,
      "step": 400
    },
    {
      "epoch": 0.6156806156806157,
      "eval_loss": 0.03978649154305458,
      "eval_runtime": 195.8607,
      "eval_samples_per_second": 5.897,
      "eval_steps_per_second": 5.897,
      "step": 400
    },
    {
      "epoch": 0.6310726310726311,
      "grad_norm": 0.38908883929252625,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.058,
      "step": 410
    },
    {
      "epoch": 0.6464646464646465,
      "grad_norm": 0.562657356262207,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0347,
      "step": 420
    },
    {
      "epoch": 0.6618566618566618,
      "grad_norm": 0.5357773900032043,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0443,
      "step": 430
    },
    {
      "epoch": 0.6772486772486772,
      "grad_norm": 0.2069566398859024,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0424,
      "step": 440
    },
    {
      "epoch": 0.6926406926406926,
      "grad_norm": 0.22095687687397003,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0402,
      "step": 450
    },
    {
      "epoch": 0.6926406926406926,
      "eval_loss": 0.04286464676260948,
      "eval_runtime": 195.5155,
      "eval_samples_per_second": 5.907,
      "eval_steps_per_second": 5.907,
      "step": 450
    },
    {
      "epoch": 0.708032708032708,
      "grad_norm": 0.2970608174800873,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0444,
      "step": 460
    },
    {
      "epoch": 0.7234247234247234,
      "grad_norm": 0.4059205651283264,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0529,
      "step": 470
    },
    {
      "epoch": 0.7388167388167388,
      "grad_norm": 0.3296886682510376,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0435,
      "step": 480
    },
    {
      "epoch": 0.7542087542087542,
      "grad_norm": 0.25292956829071045,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0393,
      "step": 490
    },
    {
      "epoch": 0.7696007696007696,
      "grad_norm": 0.25819361209869385,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0525,
      "step": 500
    },
    {
      "epoch": 0.7696007696007696,
      "eval_loss": 0.03837933763861656,
      "eval_runtime": 195.6586,
      "eval_samples_per_second": 5.903,
      "eval_steps_per_second": 5.903,
      "step": 500
    },
    {
      "epoch": 0.784992784992785,
      "grad_norm": 0.32330358028411865,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0455,
      "step": 510
    },
    {
      "epoch": 0.8003848003848004,
      "grad_norm": 0.34895986318588257,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0424,
      "step": 520
    },
    {
      "epoch": 0.8157768157768158,
      "grad_norm": 0.23877453804016113,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0289,
      "step": 530
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 0.3795028030872345,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0429,
      "step": 540
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 0.1951112151145935,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0374,
      "step": 550
    },
    {
      "epoch": 0.8465608465608465,
      "eval_loss": 0.03806205838918686,
      "eval_runtime": 195.7203,
      "eval_samples_per_second": 5.901,
      "eval_steps_per_second": 5.901,
      "step": 550
    },
    {
      "epoch": 0.8619528619528619,
      "grad_norm": 0.21981170773506165,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0427,
      "step": 560
    },
    {
      "epoch": 0.8773448773448773,
      "grad_norm": 0.4407474100589752,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0427,
      "step": 570
    },
    {
      "epoch": 0.8927368927368927,
      "grad_norm": 0.29930567741394043,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0344,
      "step": 580
    },
    {
      "epoch": 0.9081289081289081,
      "grad_norm": 0.19037345051765442,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0356,
      "step": 590
    },
    {
      "epoch": 0.9235209235209235,
      "grad_norm": 0.16624511778354645,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0479,
      "step": 600
    },
    {
      "epoch": 0.9235209235209235,
      "eval_loss": 0.03358432278037071,
      "eval_runtime": 195.8091,
      "eval_samples_per_second": 5.899,
      "eval_steps_per_second": 5.899,
      "step": 600
    },
    {
      "epoch": 0.9389129389129389,
      "grad_norm": 0.21530045568943024,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0414,
      "step": 610
    },
    {
      "epoch": 0.9543049543049543,
      "grad_norm": 0.21044142544269562,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0283,
      "step": 620
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.21515077352523804,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0494,
      "step": 630
    },
    {
      "epoch": 0.9850889850889851,
      "grad_norm": 0.23108790814876556,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0339,
      "step": 640
    },
    {
      "epoch": 1.0004810004810005,
      "grad_norm": 0.29014286398887634,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0464,
      "step": 650
    },
    {
      "epoch": 1.0004810004810005,
      "eval_loss": 0.03368612006306648,
      "eval_runtime": 195.6861,
      "eval_samples_per_second": 5.902,
      "eval_steps_per_second": 5.902,
      "step": 650
    },
    {
      "epoch": 1.0158730158730158,
      "grad_norm": 0.17055271565914154,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0321,
      "step": 660
    },
    {
      "epoch": 1.0312650312650313,
      "grad_norm": 0.17788831889629364,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0293,
      "step": 670
    },
    {
      "epoch": 1.0466570466570466,
      "grad_norm": 0.13870351016521454,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.028,
      "step": 680
    },
    {
      "epoch": 1.0620490620490621,
      "grad_norm": 0.16923217475414276,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0309,
      "step": 690
    },
    {
      "epoch": 1.0774410774410774,
      "grad_norm": 0.4101562201976776,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0291,
      "step": 700
    },
    {
      "epoch": 1.0774410774410774,
      "eval_loss": 0.0350741408765316,
      "eval_runtime": 195.4945,
      "eval_samples_per_second": 5.908,
      "eval_steps_per_second": 5.908,
      "step": 700
    },
    {
      "epoch": 1.092833092833093,
      "grad_norm": 0.6801550388336182,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0317,
      "step": 710
    },
    {
      "epoch": 1.1082251082251082,
      "grad_norm": 0.3715743124485016,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0328,
      "step": 720
    },
    {
      "epoch": 1.1236171236171235,
      "grad_norm": 0.15498656034469604,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.038,
      "step": 730
    },
    {
      "epoch": 1.139009139009139,
      "grad_norm": 0.28883275389671326,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0345,
      "step": 740
    },
    {
      "epoch": 1.1544011544011543,
      "grad_norm": 0.2838892340660095,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0266,
      "step": 750
    },
    {
      "epoch": 1.1544011544011543,
      "eval_loss": 0.033587317913770676,
      "eval_runtime": 195.2342,
      "eval_samples_per_second": 5.916,
      "eval_steps_per_second": 5.916,
      "step": 750
    },
    {
      "epoch": 1.1697931697931698,
      "grad_norm": 0.10602987557649612,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0333,
      "step": 760
    },
    {
      "epoch": 1.1851851851851851,
      "grad_norm": 0.04328380152583122,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0222,
      "step": 770
    },
    {
      "epoch": 1.2005772005772006,
      "grad_norm": 0.2159237265586853,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0255,
      "step": 780
    },
    {
      "epoch": 1.215969215969216,
      "grad_norm": 0.533365786075592,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0319,
      "step": 790
    },
    {
      "epoch": 1.2313612313612314,
      "grad_norm": 0.19054140150547028,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0268,
      "step": 800
    },
    {
      "epoch": 1.2313612313612314,
      "eval_loss": 0.03231462836265564,
      "eval_runtime": 195.0022,
      "eval_samples_per_second": 5.923,
      "eval_steps_per_second": 5.923,
      "step": 800
    },
    {
      "epoch": 1.2467532467532467,
      "grad_norm": 0.2857094705104828,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0383,
      "step": 810
    },
    {
      "epoch": 1.2621452621452622,
      "grad_norm": 0.16468562185764313,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0323,
      "step": 820
    },
    {
      "epoch": 1.2775372775372775,
      "grad_norm": 0.15043526887893677,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0307,
      "step": 830
    },
    {
      "epoch": 1.2929292929292928,
      "grad_norm": 0.28781983256340027,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0239,
      "step": 840
    },
    {
      "epoch": 1.3083213083213083,
      "grad_norm": 0.4519936144351959,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0257,
      "step": 850
    },
    {
      "epoch": 1.3083213083213083,
      "eval_loss": 0.03252238407731056,
      "eval_runtime": 194.6609,
      "eval_samples_per_second": 5.933,
      "eval_steps_per_second": 5.933,
      "step": 850
    },
    {
      "epoch": 1.3237133237133238,
      "grad_norm": 0.27222883701324463,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0215,
      "step": 860
    },
    {
      "epoch": 1.3391053391053391,
      "grad_norm": 0.2070714831352234,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0348,
      "step": 870
    },
    {
      "epoch": 1.3544973544973544,
      "grad_norm": 0.20227199792861938,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.03,
      "step": 880
    },
    {
      "epoch": 1.36988936988937,
      "grad_norm": 0.13928696513175964,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0289,
      "step": 890
    },
    {
      "epoch": 1.3852813852813852,
      "grad_norm": 0.27322667837142944,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0311,
      "step": 900
    },
    {
      "epoch": 1.3852813852813852,
      "eval_loss": 0.03226852789521217,
      "eval_runtime": 194.4539,
      "eval_samples_per_second": 5.94,
      "eval_steps_per_second": 5.94,
      "step": 900
    },
    {
      "epoch": 1.4006734006734007,
      "grad_norm": 0.2814335823059082,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0384,
      "step": 910
    },
    {
      "epoch": 1.416065416065416,
      "grad_norm": 0.18716764450073242,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0322,
      "step": 920
    },
    {
      "epoch": 1.4314574314574315,
      "grad_norm": 0.19502660632133484,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0288,
      "step": 930
    },
    {
      "epoch": 1.4468494468494468,
      "grad_norm": 0.35453715920448303,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.036,
      "step": 940
    },
    {
      "epoch": 1.4622414622414621,
      "grad_norm": 0.23714230954647064,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0345,
      "step": 950
    },
    {
      "epoch": 1.4622414622414621,
      "eval_loss": 0.03173607587814331,
      "eval_runtime": 194.6257,
      "eval_samples_per_second": 5.934,
      "eval_steps_per_second": 5.934,
      "step": 950
    },
    {
      "epoch": 1.4776334776334776,
      "grad_norm": 0.3544442355632782,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0295,
      "step": 960
    },
    {
      "epoch": 1.4930254930254931,
      "grad_norm": 0.2871958315372467,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.038,
      "step": 970
    },
    {
      "epoch": 1.5084175084175084,
      "grad_norm": 0.3297398090362549,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0352,
      "step": 980
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 0.17722108960151672,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0236,
      "step": 990
    },
    {
      "epoch": 1.5392015392015392,
      "grad_norm": 0.42388418316841125,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0287,
      "step": 1000
    },
    {
      "epoch": 1.5392015392015392,
      "eval_loss": 0.03453391045331955,
      "eval_runtime": 194.8929,
      "eval_samples_per_second": 5.926,
      "eval_steps_per_second": 5.926,
      "step": 1000
    },
    {
      "epoch": 1.5545935545935547,
      "grad_norm": 0.20600707828998566,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0312,
      "step": 1010
    },
    {
      "epoch": 1.56998556998557,
      "grad_norm": 0.178716778755188,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0232,
      "step": 1020
    },
    {
      "epoch": 1.5853775853775853,
      "grad_norm": 0.3199945092201233,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0228,
      "step": 1030
    },
    {
      "epoch": 1.6007696007696008,
      "grad_norm": 0.3925187289714813,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0338,
      "step": 1040
    },
    {
      "epoch": 1.6161616161616161,
      "grad_norm": 0.15340888500213623,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0333,
      "step": 1050
    },
    {
      "epoch": 1.6161616161616161,
      "eval_loss": 0.03127632290124893,
      "eval_runtime": 196.2596,
      "eval_samples_per_second": 5.885,
      "eval_steps_per_second": 5.885,
      "step": 1050
    },
    {
      "epoch": 1.6315536315536314,
      "grad_norm": 0.1792517602443695,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0344,
      "step": 1060
    },
    {
      "epoch": 1.646945646945647,
      "grad_norm": 0.3584337830543518,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.039,
      "step": 1070
    },
    {
      "epoch": 1.6623376623376624,
      "grad_norm": 0.23393818736076355,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0311,
      "step": 1080
    },
    {
      "epoch": 1.6777296777296777,
      "grad_norm": 0.22196727991104126,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0208,
      "step": 1090
    },
    {
      "epoch": 1.693121693121693,
      "grad_norm": 0.1775234192609787,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0278,
      "step": 1100
    },
    {
      "epoch": 1.693121693121693,
      "eval_loss": 0.03136962652206421,
      "eval_runtime": 199.253,
      "eval_samples_per_second": 5.797,
      "eval_steps_per_second": 5.797,
      "step": 1100
    },
    {
      "epoch": 1.7085137085137085,
      "grad_norm": 0.2123710960149765,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0368,
      "step": 1110
    },
    {
      "epoch": 1.723905723905724,
      "grad_norm": 0.3104514181613922,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0223,
      "step": 1120
    },
    {
      "epoch": 1.7392977392977393,
      "grad_norm": 0.3192251920700073,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0227,
      "step": 1130
    },
    {
      "epoch": 1.7546897546897546,
      "grad_norm": 0.22062405943870544,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0297,
      "step": 1140
    },
    {
      "epoch": 1.7700817700817701,
      "grad_norm": 0.2298513799905777,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0239,
      "step": 1150
    },
    {
      "epoch": 1.7700817700817701,
      "eval_loss": 0.030629225075244904,
      "eval_runtime": 197.9017,
      "eval_samples_per_second": 5.836,
      "eval_steps_per_second": 5.836,
      "step": 1150
    },
    {
      "epoch": 1.7854737854737854,
      "grad_norm": 0.3252947926521301,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0294,
      "step": 1160
    },
    {
      "epoch": 1.8008658008658007,
      "grad_norm": 0.2913691997528076,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0263,
      "step": 1170
    },
    {
      "epoch": 1.8162578162578162,
      "grad_norm": 0.22778797149658203,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0292,
      "step": 1180
    },
    {
      "epoch": 1.8316498316498318,
      "grad_norm": 0.25818929076194763,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.03,
      "step": 1190
    },
    {
      "epoch": 1.847041847041847,
      "grad_norm": 0.29820236563682556,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0346,
      "step": 1200
    },
    {
      "epoch": 1.847041847041847,
      "eval_loss": 0.031223421916365623,
      "eval_runtime": 196.3895,
      "eval_samples_per_second": 5.881,
      "eval_steps_per_second": 5.881,
      "step": 1200
    },
    {
      "epoch": 1.8624338624338623,
      "grad_norm": 0.25816982984542847,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0241,
      "step": 1210
    },
    {
      "epoch": 1.8778258778258778,
      "grad_norm": 0.06710444390773773,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0288,
      "step": 1220
    },
    {
      "epoch": 1.8932178932178934,
      "grad_norm": 0.2378709465265274,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0253,
      "step": 1230
    },
    {
      "epoch": 1.9086099086099086,
      "grad_norm": 0.19303959608078003,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0317,
      "step": 1240
    },
    {
      "epoch": 1.924001924001924,
      "grad_norm": 0.4784505367279053,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0289,
      "step": 1250
    },
    {
      "epoch": 1.924001924001924,
      "eval_loss": 0.030128151178359985,
      "eval_runtime": 177.2707,
      "eval_samples_per_second": 6.515,
      "eval_steps_per_second": 6.515,
      "step": 1250
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.13791890442371368,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0265,
      "step": 1260
    },
    {
      "epoch": 1.9547859547859547,
      "grad_norm": 0.15390145778656006,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0236,
      "step": 1270
    },
    {
      "epoch": 1.97017797017797,
      "grad_norm": 0.18494607508182526,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0224,
      "step": 1280
    },
    {
      "epoch": 1.9855699855699855,
      "grad_norm": 0.11450916528701782,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0348,
      "step": 1290
    },
    {
      "epoch": 2.000962000962001,
      "grad_norm": 0.17159050703048706,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0308,
      "step": 1300
    },
    {
      "epoch": 2.000962000962001,
      "eval_loss": 0.03185825049877167,
      "eval_runtime": 177.4192,
      "eval_samples_per_second": 6.51,
      "eval_steps_per_second": 6.51,
      "step": 1300
    },
    {
      "epoch": 2.0163540163540166,
      "grad_norm": 0.1461666226387024,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.025,
      "step": 1310
    },
    {
      "epoch": 2.0317460317460316,
      "grad_norm": 0.1347372680902481,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0192,
      "step": 1320
    },
    {
      "epoch": 2.047138047138047,
      "grad_norm": 0.24139681458473206,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0199,
      "step": 1330
    },
    {
      "epoch": 2.0625300625300627,
      "grad_norm": 0.090448297560215,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0159,
      "step": 1340
    },
    {
      "epoch": 2.0779220779220777,
      "grad_norm": 0.21202856302261353,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0176,
      "step": 1350
    },
    {
      "epoch": 2.0779220779220777,
      "eval_loss": 0.03141428157687187,
      "eval_runtime": 177.4461,
      "eval_samples_per_second": 6.509,
      "eval_steps_per_second": 6.509,
      "step": 1350
    },
    {
      "epoch": 2.0933140933140932,
      "grad_norm": 0.20801958441734314,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0146,
      "step": 1360
    },
    {
      "epoch": 2.1087061087061088,
      "grad_norm": 0.2737131118774414,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0174,
      "step": 1370
    },
    {
      "epoch": 2.1240981240981243,
      "grad_norm": 0.285129189491272,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.019,
      "step": 1380
    },
    {
      "epoch": 2.1394901394901393,
      "grad_norm": 0.13645029067993164,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0144,
      "step": 1390
    },
    {
      "epoch": 2.154882154882155,
      "grad_norm": 0.3433080315589905,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0192,
      "step": 1400
    },
    {
      "epoch": 2.154882154882155,
      "eval_loss": 0.03115396946668625,
      "eval_runtime": 177.6475,
      "eval_samples_per_second": 6.502,
      "eval_steps_per_second": 6.502,
      "step": 1400
    },
    {
      "epoch": 2.1702741702741704,
      "grad_norm": 0.4344891309738159,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0212,
      "step": 1410
    },
    {
      "epoch": 2.185666185666186,
      "grad_norm": 0.031966980546712875,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0153,
      "step": 1420
    },
    {
      "epoch": 2.201058201058201,
      "grad_norm": 0.1055040955543518,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0241,
      "step": 1430
    },
    {
      "epoch": 2.2164502164502164,
      "grad_norm": 0.27203112840652466,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0163,
      "step": 1440
    },
    {
      "epoch": 2.231842231842232,
      "grad_norm": 0.1582086980342865,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0157,
      "step": 1450
    },
    {
      "epoch": 2.231842231842232,
      "eval_loss": 0.031504761427640915,
      "eval_runtime": 177.7803,
      "eval_samples_per_second": 6.497,
      "eval_steps_per_second": 6.497,
      "step": 1450
    },
    {
      "epoch": 2.247234247234247,
      "grad_norm": 0.2671997845172882,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0213,
      "step": 1460
    },
    {
      "epoch": 2.2626262626262625,
      "grad_norm": 0.12549597024917603,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0171,
      "step": 1470
    },
    {
      "epoch": 2.278018278018278,
      "grad_norm": 0.3192961812019348,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0177,
      "step": 1480
    },
    {
      "epoch": 2.2934102934102936,
      "grad_norm": 0.19741778075695038,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0211,
      "step": 1490
    },
    {
      "epoch": 2.3088023088023086,
      "grad_norm": 0.17573250830173492,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0227,
      "step": 1500
    },
    {
      "epoch": 2.3088023088023086,
      "eval_loss": 0.030094966292381287,
      "eval_runtime": 177.8424,
      "eval_samples_per_second": 6.495,
      "eval_steps_per_second": 6.495,
      "step": 1500
    },
    {
      "epoch": 2.324194324194324,
      "grad_norm": 0.09664864838123322,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0237,
      "step": 1510
    },
    {
      "epoch": 2.3395863395863397,
      "grad_norm": 0.13909515738487244,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0156,
      "step": 1520
    },
    {
      "epoch": 2.354978354978355,
      "grad_norm": 0.11331960558891296,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0224,
      "step": 1530
    },
    {
      "epoch": 2.3703703703703702,
      "grad_norm": 0.34505394101142883,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0272,
      "step": 1540
    },
    {
      "epoch": 2.3857623857623858,
      "grad_norm": 0.21384982764720917,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0211,
      "step": 1550
    },
    {
      "epoch": 2.3857623857623858,
      "eval_loss": 0.030527736991643906,
      "eval_runtime": 177.8434,
      "eval_samples_per_second": 6.494,
      "eval_steps_per_second": 6.494,
      "step": 1550
    },
    {
      "epoch": 2.4011544011544013,
      "grad_norm": 0.16733352839946747,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0226,
      "step": 1560
    },
    {
      "epoch": 2.4165464165464163,
      "grad_norm": 0.20128144323825836,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0184,
      "step": 1570
    },
    {
      "epoch": 2.431938431938432,
      "grad_norm": 0.09988749027252197,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0164,
      "step": 1580
    },
    {
      "epoch": 2.4473304473304474,
      "grad_norm": 0.049513984471559525,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.017,
      "step": 1590
    },
    {
      "epoch": 2.462722462722463,
      "grad_norm": 0.22615675628185272,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0192,
      "step": 1600
    },
    {
      "epoch": 2.462722462722463,
      "eval_loss": 0.03266085684299469,
      "eval_runtime": 177.7662,
      "eval_samples_per_second": 6.497,
      "eval_steps_per_second": 6.497,
      "step": 1600
    },
    {
      "epoch": 2.478114478114478,
      "grad_norm": 0.14208638668060303,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0165,
      "step": 1610
    },
    {
      "epoch": 2.4935064935064934,
      "grad_norm": 0.19447220861911774,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0229,
      "step": 1620
    },
    {
      "epoch": 2.508898508898509,
      "grad_norm": 0.2803027629852295,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0246,
      "step": 1630
    },
    {
      "epoch": 2.5242905242905245,
      "grad_norm": 0.20524460077285767,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0227,
      "step": 1640
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 0.28721189498901367,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0215,
      "step": 1650
    },
    {
      "epoch": 2.5396825396825395,
      "eval_loss": 0.031051570549607277,
      "eval_runtime": 177.8644,
      "eval_samples_per_second": 6.494,
      "eval_steps_per_second": 6.494,
      "step": 1650
    },
    {
      "epoch": 2.555074555074555,
      "grad_norm": 0.11950189620256424,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0289,
      "step": 1660
    },
    {
      "epoch": 2.5704665704665706,
      "grad_norm": 0.1526806503534317,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0243,
      "step": 1670
    },
    {
      "epoch": 2.5858585858585856,
      "grad_norm": 0.20460185408592224,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0215,
      "step": 1680
    },
    {
      "epoch": 2.601250601250601,
      "grad_norm": 0.17265936732292175,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0175,
      "step": 1690
    },
    {
      "epoch": 2.6166426166426167,
      "grad_norm": 0.181435689330101,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0235,
      "step": 1700
    },
    {
      "epoch": 2.6166426166426167,
      "eval_loss": 0.029672790318727493,
      "eval_runtime": 178.0424,
      "eval_samples_per_second": 6.487,
      "eval_steps_per_second": 6.487,
      "step": 1700
    },
    {
      "epoch": 2.632034632034632,
      "grad_norm": 0.0935995876789093,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0175,
      "step": 1710
    },
    {
      "epoch": 2.6474266474266477,
      "grad_norm": 0.1314033716917038,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0215,
      "step": 1720
    },
    {
      "epoch": 2.6628186628186628,
      "grad_norm": 0.20978249609470367,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0198,
      "step": 1730
    },
    {
      "epoch": 2.6782106782106783,
      "grad_norm": 0.06089981645345688,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0173,
      "step": 1740
    },
    {
      "epoch": 2.6936026936026938,
      "grad_norm": 0.1842568963766098,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0149,
      "step": 1750
    },
    {
      "epoch": 2.6936026936026938,
      "eval_loss": 0.031217539682984352,
      "eval_runtime": 178.0933,
      "eval_samples_per_second": 6.485,
      "eval_steps_per_second": 6.485,
      "step": 1750
    },
    {
      "epoch": 2.708994708994709,
      "grad_norm": 0.2004362940788269,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0163,
      "step": 1760
    },
    {
      "epoch": 2.7243867243867244,
      "grad_norm": 0.28942927718162537,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0211,
      "step": 1770
    },
    {
      "epoch": 2.73977873977874,
      "grad_norm": 0.1989394575357437,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0232,
      "step": 1780
    },
    {
      "epoch": 2.755170755170755,
      "grad_norm": 0.1604626327753067,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0149,
      "step": 1790
    },
    {
      "epoch": 2.7705627705627704,
      "grad_norm": 0.29249003529548645,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0206,
      "step": 1800
    },
    {
      "epoch": 2.7705627705627704,
      "eval_loss": 0.02991563268005848,
      "eval_runtime": 178.1963,
      "eval_samples_per_second": 6.482,
      "eval_steps_per_second": 6.482,
      "step": 1800
    },
    {
      "epoch": 2.785954785954786,
      "grad_norm": 0.234651580452919,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0211,
      "step": 1810
    },
    {
      "epoch": 2.8013468013468015,
      "grad_norm": 0.15778271853923798,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0185,
      "step": 1820
    },
    {
      "epoch": 2.816738816738817,
      "grad_norm": 0.24058304727077484,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0231,
      "step": 1830
    },
    {
      "epoch": 2.832130832130832,
      "grad_norm": 0.24088242650032043,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0209,
      "step": 1840
    },
    {
      "epoch": 2.8475228475228476,
      "grad_norm": 0.23293456435203552,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0154,
      "step": 1850
    },
    {
      "epoch": 2.8475228475228476,
      "eval_loss": 0.03181726485490799,
      "eval_runtime": 178.2513,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 1850
    },
    {
      "epoch": 2.862914862914863,
      "grad_norm": 0.15455672144889832,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0161,
      "step": 1860
    },
    {
      "epoch": 2.878306878306878,
      "grad_norm": 0.17855042219161987,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0216,
      "step": 1870
    },
    {
      "epoch": 2.8936988936988937,
      "grad_norm": 0.19437535107135773,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.015,
      "step": 1880
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.11707495152950287,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0143,
      "step": 1890
    },
    {
      "epoch": 2.9244829244829242,
      "grad_norm": 0.07547138631343842,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0188,
      "step": 1900
    },
    {
      "epoch": 2.9244829244829242,
      "eval_loss": 0.03009277582168579,
      "eval_runtime": 178.0798,
      "eval_samples_per_second": 6.486,
      "eval_steps_per_second": 6.486,
      "step": 1900
    },
    {
      "epoch": 2.9398749398749398,
      "grad_norm": 0.12523221969604492,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0166,
      "step": 1910
    },
    {
      "epoch": 2.9552669552669553,
      "grad_norm": 0.2055351734161377,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0178,
      "step": 1920
    },
    {
      "epoch": 2.9706589706589708,
      "grad_norm": 0.1336817592382431,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0227,
      "step": 1930
    },
    {
      "epoch": 2.9860509860509863,
      "grad_norm": 0.1597260981798172,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0212,
      "step": 1940
    },
    {
      "epoch": 3.0014430014430014,
      "grad_norm": 0.2565922737121582,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0198,
      "step": 1950
    },
    {
      "epoch": 3.0014430014430014,
      "eval_loss": 0.029989564791321754,
      "eval_runtime": 177.964,
      "eval_samples_per_second": 6.49,
      "eval_steps_per_second": 6.49,
      "step": 1950
    },
    {
      "epoch": 3.016835016835017,
      "grad_norm": 0.36280080676078796,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0158,
      "step": 1960
    },
    {
      "epoch": 3.0322270322270324,
      "grad_norm": 0.18591882288455963,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0102,
      "step": 1970
    },
    {
      "epoch": 3.0476190476190474,
      "grad_norm": 0.21973514556884766,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0103,
      "step": 1980
    },
    {
      "epoch": 3.063011063011063,
      "grad_norm": 0.07260677218437195,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0111,
      "step": 1990
    },
    {
      "epoch": 3.0784030784030785,
      "grad_norm": 0.200120210647583,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0117,
      "step": 2000
    },
    {
      "epoch": 3.0784030784030785,
      "eval_loss": 0.03222152963280678,
      "eval_runtime": 178.1577,
      "eval_samples_per_second": 6.483,
      "eval_steps_per_second": 6.483,
      "step": 2000
    },
    {
      "epoch": 3.093795093795094,
      "grad_norm": 0.14105476438999176,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0086,
      "step": 2010
    },
    {
      "epoch": 3.109187109187109,
      "grad_norm": 0.14028598368167877,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0127,
      "step": 2020
    },
    {
      "epoch": 3.1245791245791246,
      "grad_norm": 0.3382083475589752,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0103,
      "step": 2030
    },
    {
      "epoch": 3.13997113997114,
      "grad_norm": 0.15088129043579102,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0097,
      "step": 2040
    },
    {
      "epoch": 3.1553631553631556,
      "grad_norm": 0.05970817804336548,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0103,
      "step": 2050
    },
    {
      "epoch": 3.1553631553631556,
      "eval_loss": 0.03335327282547951,
      "eval_runtime": 178.2556,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 2050
    },
    {
      "epoch": 3.1707551707551707,
      "grad_norm": 0.17845909297466278,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0095,
      "step": 2060
    },
    {
      "epoch": 3.186147186147186,
      "grad_norm": 0.44921189546585083,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0101,
      "step": 2070
    },
    {
      "epoch": 3.2015392015392017,
      "grad_norm": 0.42118337750434875,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0135,
      "step": 2080
    },
    {
      "epoch": 3.2169312169312168,
      "grad_norm": 0.530854344367981,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0124,
      "step": 2090
    },
    {
      "epoch": 3.2323232323232323,
      "grad_norm": 0.2749970555305481,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0158,
      "step": 2100
    },
    {
      "epoch": 3.2323232323232323,
      "eval_loss": 0.034329500049352646,
      "eval_runtime": 178.2605,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 2100
    },
    {
      "epoch": 3.2477152477152478,
      "grad_norm": 0.0675465315580368,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0074,
      "step": 2110
    },
    {
      "epoch": 3.2631072631072633,
      "grad_norm": 0.3295849859714508,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0177,
      "step": 2120
    },
    {
      "epoch": 3.2784992784992784,
      "grad_norm": 0.24006688594818115,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0126,
      "step": 2130
    },
    {
      "epoch": 3.293891293891294,
      "grad_norm": 0.25042691826820374,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0112,
      "step": 2140
    },
    {
      "epoch": 3.3092833092833094,
      "grad_norm": 0.08620034903287888,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0116,
      "step": 2150
    },
    {
      "epoch": 3.3092833092833094,
      "eval_loss": 0.03303411602973938,
      "eval_runtime": 178.3625,
      "eval_samples_per_second": 6.476,
      "eval_steps_per_second": 6.476,
      "step": 2150
    },
    {
      "epoch": 3.324675324675325,
      "grad_norm": 0.11519394814968109,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0097,
      "step": 2160
    },
    {
      "epoch": 3.34006734006734,
      "grad_norm": 0.1725398600101471,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0077,
      "step": 2170
    },
    {
      "epoch": 3.3554593554593555,
      "grad_norm": 0.18201743066310883,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0119,
      "step": 2180
    },
    {
      "epoch": 3.370851370851371,
      "grad_norm": 0.11471626162528992,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0101,
      "step": 2190
    },
    {
      "epoch": 3.386243386243386,
      "grad_norm": 0.43149736523628235,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0117,
      "step": 2200
    },
    {
      "epoch": 3.386243386243386,
      "eval_loss": 0.034219466149806976,
      "eval_runtime": 178.3447,
      "eval_samples_per_second": 6.476,
      "eval_steps_per_second": 6.476,
      "step": 2200
    },
    {
      "epoch": 3.4016354016354016,
      "grad_norm": 0.15829546749591827,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0088,
      "step": 2210
    },
    {
      "epoch": 3.417027417027417,
      "grad_norm": 0.07393281161785126,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0131,
      "step": 2220
    },
    {
      "epoch": 3.4324194324194326,
      "grad_norm": 0.21448646485805511,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0104,
      "step": 2230
    },
    {
      "epoch": 3.4478114478114477,
      "grad_norm": 0.07296042889356613,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0074,
      "step": 2240
    },
    {
      "epoch": 3.463203463203463,
      "grad_norm": 0.3533247709274292,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.01,
      "step": 2250
    },
    {
      "epoch": 3.463203463203463,
      "eval_loss": 0.03453659638762474,
      "eval_runtime": 178.2346,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 2250
    },
    {
      "epoch": 3.4785954785954787,
      "grad_norm": 0.06865774840116501,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0135,
      "step": 2260
    },
    {
      "epoch": 3.493987493987494,
      "grad_norm": 0.06761554628610611,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0104,
      "step": 2270
    },
    {
      "epoch": 3.5093795093795093,
      "grad_norm": 0.2272021770477295,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0114,
      "step": 2280
    },
    {
      "epoch": 3.5247715247715248,
      "grad_norm": 0.09709617495536804,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0084,
      "step": 2290
    },
    {
      "epoch": 3.5401635401635403,
      "grad_norm": 0.35291028022766113,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0113,
      "step": 2300
    },
    {
      "epoch": 3.5401635401635403,
      "eval_loss": 0.03454257547855377,
      "eval_runtime": 178.2468,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 2300
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.1527557373046875,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0084,
      "step": 2310
    },
    {
      "epoch": 3.570947570947571,
      "grad_norm": 0.3380734622478485,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0125,
      "step": 2320
    },
    {
      "epoch": 3.5863395863395864,
      "grad_norm": 0.3082076907157898,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0129,
      "step": 2330
    },
    {
      "epoch": 3.601731601731602,
      "grad_norm": 0.3630889058113098,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.01,
      "step": 2340
    },
    {
      "epoch": 3.617123617123617,
      "grad_norm": 0.18567804992198944,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0111,
      "step": 2350
    },
    {
      "epoch": 3.617123617123617,
      "eval_loss": 0.03422347456216812,
      "eval_runtime": 178.2604,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 2350
    },
    {
      "epoch": 3.6325156325156325,
      "grad_norm": 0.04647551849484444,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.007,
      "step": 2360
    },
    {
      "epoch": 3.647907647907648,
      "grad_norm": 0.20195652544498444,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0079,
      "step": 2370
    },
    {
      "epoch": 3.6632996632996635,
      "grad_norm": 0.397169291973114,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0079,
      "step": 2380
    },
    {
      "epoch": 3.6786916786916786,
      "grad_norm": 0.15381094813346863,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0082,
      "step": 2390
    },
    {
      "epoch": 3.694083694083694,
      "grad_norm": 0.30508971214294434,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0105,
      "step": 2400
    },
    {
      "epoch": 3.694083694083694,
      "eval_loss": 0.03513110429048538,
      "eval_runtime": 178.3998,
      "eval_samples_per_second": 6.474,
      "eval_steps_per_second": 6.474,
      "step": 2400
    },
    {
      "epoch": 3.7094757094757096,
      "grad_norm": 0.2903805375099182,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0146,
      "step": 2410
    },
    {
      "epoch": 3.7248677248677247,
      "grad_norm": 0.05851426348090172,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0081,
      "step": 2420
    },
    {
      "epoch": 3.74025974025974,
      "grad_norm": 0.17480336129665375,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0106,
      "step": 2430
    },
    {
      "epoch": 3.7556517556517557,
      "grad_norm": 0.29570499062538147,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0117,
      "step": 2440
    },
    {
      "epoch": 3.771043771043771,
      "grad_norm": 0.04345010221004486,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0094,
      "step": 2450
    },
    {
      "epoch": 3.771043771043771,
      "eval_loss": 0.03652167692780495,
      "eval_runtime": 178.4845,
      "eval_samples_per_second": 6.471,
      "eval_steps_per_second": 6.471,
      "step": 2450
    },
    {
      "epoch": 3.7864357864357867,
      "grad_norm": 0.05769900232553482,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0093,
      "step": 2460
    },
    {
      "epoch": 3.801827801827802,
      "grad_norm": 0.21346734464168549,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.011,
      "step": 2470
    },
    {
      "epoch": 3.8172198172198173,
      "grad_norm": 0.3063548803329468,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0112,
      "step": 2480
    },
    {
      "epoch": 3.832611832611833,
      "grad_norm": 0.11381343752145767,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0097,
      "step": 2490
    },
    {
      "epoch": 3.848003848003848,
      "grad_norm": 0.24341605603694916,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0144,
      "step": 2500
    },
    {
      "epoch": 3.848003848003848,
      "eval_loss": 0.03374967724084854,
      "eval_runtime": 178.5013,
      "eval_samples_per_second": 6.471,
      "eval_steps_per_second": 6.471,
      "step": 2500
    },
    {
      "epoch": 3.8633958633958634,
      "grad_norm": 0.23066754639148712,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0159,
      "step": 2510
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.1408952921628952,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0119,
      "step": 2520
    },
    {
      "epoch": 3.894179894179894,
      "grad_norm": 0.5981900095939636,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0073,
      "step": 2530
    },
    {
      "epoch": 3.9095719095719095,
      "grad_norm": 0.25713270902633667,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0075,
      "step": 2540
    },
    {
      "epoch": 3.924963924963925,
      "grad_norm": 0.2724706530570984,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0071,
      "step": 2550
    },
    {
      "epoch": 3.924963924963925,
      "eval_loss": 0.034059759229421616,
      "eval_runtime": 178.4363,
      "eval_samples_per_second": 6.473,
      "eval_steps_per_second": 6.473,
      "step": 2550
    },
    {
      "epoch": 3.9403559403559405,
      "grad_norm": 0.18167531490325928,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0088,
      "step": 2560
    },
    {
      "epoch": 3.955747955747956,
      "grad_norm": 0.27482736110687256,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0097,
      "step": 2570
    },
    {
      "epoch": 3.971139971139971,
      "grad_norm": 0.2618437707424164,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0084,
      "step": 2580
    },
    {
      "epoch": 3.9865319865319866,
      "grad_norm": 0.11441271007061005,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0099,
      "step": 2590
    },
    {
      "epoch": 4.001924001924002,
      "grad_norm": 0.12754091620445251,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0081,
      "step": 2600
    },
    {
      "epoch": 4.001924001924002,
      "eval_loss": 0.03393612429499626,
      "eval_runtime": 178.3441,
      "eval_samples_per_second": 6.476,
      "eval_steps_per_second": 6.476,
      "step": 2600
    },
    {
      "epoch": 4.017316017316017,
      "grad_norm": 0.11633338779211044,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0062,
      "step": 2610
    },
    {
      "epoch": 4.032708032708033,
      "grad_norm": 0.1154623031616211,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0062,
      "step": 2620
    },
    {
      "epoch": 4.048100048100048,
      "grad_norm": 0.2394174486398697,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0061,
      "step": 2630
    },
    {
      "epoch": 4.063492063492063,
      "grad_norm": 0.157700315117836,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0051,
      "step": 2640
    },
    {
      "epoch": 4.078884078884079,
      "grad_norm": 0.1628115326166153,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0051,
      "step": 2650
    },
    {
      "epoch": 4.078884078884079,
      "eval_loss": 0.0355365127325058,
      "eval_runtime": 178.3106,
      "eval_samples_per_second": 6.477,
      "eval_steps_per_second": 6.477,
      "step": 2650
    },
    {
      "epoch": 4.094276094276094,
      "grad_norm": 0.06331827491521835,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0035,
      "step": 2660
    },
    {
      "epoch": 4.109668109668109,
      "grad_norm": 0.19837036728858948,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0045,
      "step": 2670
    },
    {
      "epoch": 4.125060125060125,
      "grad_norm": 0.29709723591804504,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0076,
      "step": 2680
    },
    {
      "epoch": 4.14045214045214,
      "grad_norm": 0.06853421032428741,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0053,
      "step": 2690
    },
    {
      "epoch": 4.1558441558441555,
      "grad_norm": 0.2256442904472351,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0081,
      "step": 2700
    },
    {
      "epoch": 4.1558441558441555,
      "eval_loss": 0.03636639565229416,
      "eval_runtime": 178.3445,
      "eval_samples_per_second": 6.476,
      "eval_steps_per_second": 6.476,
      "step": 2700
    },
    {
      "epoch": 4.171236171236171,
      "grad_norm": 0.19414715468883514,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0042,
      "step": 2710
    },
    {
      "epoch": 4.1866281866281865,
      "grad_norm": 0.1326339840888977,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0038,
      "step": 2720
    },
    {
      "epoch": 4.202020202020202,
      "grad_norm": 0.28847751021385193,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0086,
      "step": 2730
    },
    {
      "epoch": 4.2174122174122175,
      "grad_norm": 0.0937063917517662,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0053,
      "step": 2740
    },
    {
      "epoch": 4.232804232804233,
      "grad_norm": 0.16331104934215546,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0068,
      "step": 2750
    },
    {
      "epoch": 4.232804232804233,
      "eval_loss": 0.03818191587924957,
      "eval_runtime": 178.4744,
      "eval_samples_per_second": 6.472,
      "eval_steps_per_second": 6.472,
      "step": 2750
    },
    {
      "epoch": 4.2481962481962485,
      "grad_norm": 0.09543789178133011,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.006,
      "step": 2760
    },
    {
      "epoch": 4.263588263588264,
      "grad_norm": 0.037738483399152756,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0081,
      "step": 2770
    },
    {
      "epoch": 4.278980278980279,
      "grad_norm": 0.03703128173947334,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0032,
      "step": 2780
    },
    {
      "epoch": 4.294372294372295,
      "grad_norm": 0.11930393427610397,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0042,
      "step": 2790
    },
    {
      "epoch": 4.30976430976431,
      "grad_norm": 0.1955643743276596,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0055,
      "step": 2800
    },
    {
      "epoch": 4.30976430976431,
      "eval_loss": 0.03888539969921112,
      "eval_runtime": 178.5149,
      "eval_samples_per_second": 6.47,
      "eval_steps_per_second": 6.47,
      "step": 2800
    },
    {
      "epoch": 4.325156325156325,
      "grad_norm": 0.2151470184326172,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0039,
      "step": 2810
    },
    {
      "epoch": 4.340548340548341,
      "grad_norm": 0.21471340954303741,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.004,
      "step": 2820
    },
    {
      "epoch": 4.355940355940356,
      "grad_norm": 0.18419770896434784,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0046,
      "step": 2830
    },
    {
      "epoch": 4.371332371332372,
      "grad_norm": 0.3202359080314636,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0051,
      "step": 2840
    },
    {
      "epoch": 4.386724386724387,
      "grad_norm": 0.11451803892850876,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0045,
      "step": 2850
    },
    {
      "epoch": 4.386724386724387,
      "eval_loss": 0.038550637662410736,
      "eval_runtime": 178.6761,
      "eval_samples_per_second": 6.464,
      "eval_steps_per_second": 6.464,
      "step": 2850
    },
    {
      "epoch": 4.402116402116402,
      "grad_norm": 0.17848871648311615,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0039,
      "step": 2860
    },
    {
      "epoch": 4.417508417508418,
      "grad_norm": 0.25352710485458374,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0053,
      "step": 2870
    },
    {
      "epoch": 4.432900432900433,
      "grad_norm": 0.18680106103420258,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0036,
      "step": 2880
    },
    {
      "epoch": 4.448292448292448,
      "grad_norm": 0.09924650192260742,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.005,
      "step": 2890
    },
    {
      "epoch": 4.463684463684464,
      "grad_norm": 0.15375648438930511,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0033,
      "step": 2900
    },
    {
      "epoch": 4.463684463684464,
      "eval_loss": 0.038627784699201584,
      "eval_runtime": 178.5798,
      "eval_samples_per_second": 6.468,
      "eval_steps_per_second": 6.468,
      "step": 2900
    },
    {
      "epoch": 4.479076479076479,
      "grad_norm": 0.2513602674007416,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0038,
      "step": 2910
    },
    {
      "epoch": 4.494468494468494,
      "grad_norm": 0.021214406937360764,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0038,
      "step": 2920
    },
    {
      "epoch": 4.50986050986051,
      "grad_norm": 0.07368022203445435,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0057,
      "step": 2930
    },
    {
      "epoch": 4.525252525252525,
      "grad_norm": 0.07400281727313995,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0038,
      "step": 2940
    },
    {
      "epoch": 4.540644540644541,
      "grad_norm": 0.24110573530197144,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0031,
      "step": 2950
    },
    {
      "epoch": 4.540644540644541,
      "eval_loss": 0.03910013660788536,
      "eval_runtime": 178.4108,
      "eval_samples_per_second": 6.474,
      "eval_steps_per_second": 6.474,
      "step": 2950
    },
    {
      "epoch": 4.556036556036556,
      "grad_norm": 0.11773854494094849,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0052,
      "step": 2960
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.16476194560527802,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0032,
      "step": 2970
    },
    {
      "epoch": 4.586820586820587,
      "grad_norm": 0.15313835442066193,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0072,
      "step": 2980
    },
    {
      "epoch": 4.602212602212602,
      "grad_norm": 0.11742126941680908,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0031,
      "step": 2990
    },
    {
      "epoch": 4.617604617604617,
      "grad_norm": 0.2843521237373352,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0055,
      "step": 3000
    },
    {
      "epoch": 4.617604617604617,
      "eval_loss": 0.03934362530708313,
      "eval_runtime": 178.3915,
      "eval_samples_per_second": 6.475,
      "eval_steps_per_second": 6.475,
      "step": 3000
    },
    {
      "epoch": 4.632996632996633,
      "grad_norm": 0.15215784311294556,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0056,
      "step": 3010
    },
    {
      "epoch": 4.648388648388648,
      "grad_norm": 0.17378070950508118,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0058,
      "step": 3020
    },
    {
      "epoch": 4.663780663780663,
      "grad_norm": 0.4116346836090088,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0051,
      "step": 3030
    },
    {
      "epoch": 4.679172679172679,
      "grad_norm": 0.16805511713027954,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0049,
      "step": 3040
    },
    {
      "epoch": 4.694564694564694,
      "grad_norm": 0.036651384085416794,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0081,
      "step": 3050
    },
    {
      "epoch": 4.694564694564694,
      "eval_loss": 0.03952018544077873,
      "eval_runtime": 178.5387,
      "eval_samples_per_second": 6.469,
      "eval_steps_per_second": 6.469,
      "step": 3050
    },
    {
      "epoch": 4.70995670995671,
      "grad_norm": 0.05314810946583748,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0042,
      "step": 3060
    },
    {
      "epoch": 4.725348725348725,
      "grad_norm": 0.1587986797094345,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0027,
      "step": 3070
    },
    {
      "epoch": 4.7407407407407405,
      "grad_norm": 0.11767754703760147,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.004,
      "step": 3080
    },
    {
      "epoch": 4.756132756132756,
      "grad_norm": 0.10315631330013275,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0068,
      "step": 3090
    },
    {
      "epoch": 4.7715247715247715,
      "grad_norm": 0.01954597793519497,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0034,
      "step": 3100
    },
    {
      "epoch": 4.7715247715247715,
      "eval_loss": 0.03963901102542877,
      "eval_runtime": 178.5843,
      "eval_samples_per_second": 6.468,
      "eval_steps_per_second": 6.468,
      "step": 3100
    },
    {
      "epoch": 4.786916786916787,
      "grad_norm": 0.0237242691218853,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0029,
      "step": 3110
    },
    {
      "epoch": 4.8023088023088025,
      "grad_norm": 0.2297857254743576,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0061,
      "step": 3120
    },
    {
      "epoch": 4.817700817700818,
      "grad_norm": 0.05104142427444458,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0052,
      "step": 3130
    },
    {
      "epoch": 4.833092833092833,
      "grad_norm": 0.037458714097738266,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0066,
      "step": 3140
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.1383890062570572,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0073,
      "step": 3150
    },
    {
      "epoch": 4.848484848484849,
      "eval_loss": 0.03955617547035217,
      "eval_runtime": 178.6658,
      "eval_samples_per_second": 6.465,
      "eval_steps_per_second": 6.465,
      "step": 3150
    },
    {
      "epoch": 4.863876863876864,
      "grad_norm": 0.044027723371982574,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0035,
      "step": 3160
    },
    {
      "epoch": 4.87926887926888,
      "grad_norm": 0.09859632700681686,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0041,
      "step": 3170
    },
    {
      "epoch": 4.894660894660895,
      "grad_norm": 0.163845956325531,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0039,
      "step": 3180
    },
    {
      "epoch": 4.91005291005291,
      "grad_norm": 0.07517090439796448,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.005,
      "step": 3190
    },
    {
      "epoch": 4.925444925444926,
      "grad_norm": 0.027621205896139145,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0043,
      "step": 3200
    },
    {
      "epoch": 4.925444925444926,
      "eval_loss": 0.039594002068042755,
      "eval_runtime": 178.6595,
      "eval_samples_per_second": 6.465,
      "eval_steps_per_second": 6.465,
      "step": 3200
    },
    {
      "epoch": 4.940836940836941,
      "grad_norm": 0.14774192869663239,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0098,
      "step": 3210
    },
    {
      "epoch": 4.956228956228956,
      "grad_norm": 0.08114589750766754,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0052,
      "step": 3220
    },
    {
      "epoch": 4.971620971620972,
      "grad_norm": 0.1118970587849617,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0058,
      "step": 3230
    },
    {
      "epoch": 4.987012987012987,
      "grad_norm": 0.08631182461977005,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0062,
      "step": 3240
    },
    {
      "epoch": 4.994708994708994,
      "step": 3245,
      "total_flos": 8.781251079886602e+17,
      "train_loss": 0.049838472417740864,
      "train_runtime": 38624.8522,
      "train_samples_per_second": 1.346,
      "train_steps_per_second": 0.084
    }
  ],
  "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": 8.781251079886602e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}