{
  "best_metric": 0.0316387340426445,
  "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold10/checkpoint-1300",
  "epoch": 4.994228549442093,
  "eval_steps": 50,
  "global_step": 3245,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015390534821085032,
      "grad_norm": 3.6962974071502686,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.9983,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 4.087631702423096,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.9269,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 4.001797676086426,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.7178,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 3.338139533996582,
      "learning_rate": 1.230769230769231e-05,
      "loss": 1.3795,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 2.086989402770996,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.8737,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.6946070790290833,
      "eval_runtime": 153.6303,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 0.8905605673789978,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.5902,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.2118861675262451,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.4088,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 0.6014496088027954,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.2134,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 0.9638709425926208,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.2183,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.6695938110351562,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.1557,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.10775511711835861,
      "eval_runtime": 153.6519,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8546701073646545,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.1099,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.8118547797203064,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.1089,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.2977904081344604,
      "learning_rate": 4e-05,
      "loss": 0.0782,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.9515172839164734,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0842,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.6907873749732971,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0875,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.07312605530023575,
      "eval_runtime": 153.4915,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.740244448184967,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0828,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 1.3892215490341187,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0754,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.6738784313201904,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0596,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.6430587768554688,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0516,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5701011419296265,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0735,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.0560503825545311,
      "eval_runtime": 153.5905,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.7298381924629211,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0647,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.7854873538017273,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0618,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.3517865836620331,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0465,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.6518716812133789,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0595,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.4971441924571991,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0547,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.05295779928565025,
      "eval_runtime": 153.5885,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.37159463763237,
      "learning_rate": 8e-05,
      "loss": 0.0521,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.7185795903205872,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.052,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.502921462059021,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0583,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.3665381968021393,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0616,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.45875951647758484,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.052,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04994189739227295,
      "eval_runtime": 153.6637,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.4589117765426636,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0581,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.4827307462692261,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0616,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.3521598279476166,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0559,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.4430323839187622,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0529,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.41543933749198914,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.047,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04689747840166092,
      "eval_runtime": 153.6218,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.28220847249031067,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0573,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.4786580502986908,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0445,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.39624959230422974,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.045,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.6487065553665161,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0415,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.24142639338970184,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0618,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.044152989983558655,
      "eval_runtime": 153.5066,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.1686740517616272,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0427,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.46816080808639526,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0537,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.5836164355278015,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0482,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.36026954650878906,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0601,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.30692824721336365,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0357,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.04477749764919281,
      "eval_runtime": 153.5428,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.3506256341934204,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0367,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.24528156220912933,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0302,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.5186636447906494,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0357,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.24547439813613892,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0397,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.35081154108047485,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0314,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.04018177092075348,
      "eval_runtime": 153.5682,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.49510422348976135,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0335,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.17423120141029358,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0499,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.3895149827003479,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0645,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.36647558212280273,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.043,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.23674258589744568,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0476,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.03882215917110443,
      "eval_runtime": 153.4132,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.22855113446712494,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0335,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.3441568613052368,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0309,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.3535885810852051,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0392,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.2783699631690979,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0524,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.3052879571914673,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0367,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.0375247560441494,
      "eval_runtime": 153.4345,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.2565286457538605,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0369,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.16643978655338287,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0402,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.3817373216152191,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0449,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.8698011636734009,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0366,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.6110942363739014,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.031,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03648454695940018,
      "eval_runtime": 153.6282,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.40813663601875305,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0352,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.638588011264801,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0381,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.26830026507377625,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0348,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.4589786231517792,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0467,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.2720070481300354,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0368,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.0376083143055439,
      "eval_runtime": 153.4749,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.28107619285583496,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0315,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.19602063298225403,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0264,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.3360397517681122,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0319,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.5052377581596375,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0392,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.2521992623806,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0299,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.03559281677007675,
      "eval_runtime": 153.4849,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.2175040990114212,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0392,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.25719285011291504,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0399,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.20776422321796417,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0324,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.2892984449863434,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0336,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.29296332597732544,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0296,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.03480331972241402,
      "eval_runtime": 153.592,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.1793573647737503,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0359,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.19759151339530945,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0268,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.287117600440979,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0313,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.45344942808151245,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0296,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.6343822479248047,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0345,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.03453903645277023,
      "eval_runtime": 153.5339,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.18368108570575714,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0321,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.3269821107387543,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0285,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.1382218301296234,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0269,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.17753195762634277,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0363,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.2324984222650528,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0203,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.03357860445976257,
      "eval_runtime": 153.5504,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.209682434797287,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0292,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.2549091875553131,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0343,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.2586798369884491,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0268,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.4496294856071472,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0351,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.09680108726024628,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0406,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03410950303077698,
      "eval_runtime": 153.5083,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.332719087600708,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0296,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.08519801497459412,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0236,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.31014642119407654,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0362,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.1737714409828186,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0348,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.21817009150981903,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0333,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.033163826912641525,
      "eval_runtime": 153.4467,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.21000780165195465,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.037,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.24734270572662354,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0383,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.1453760862350464,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0308,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.14932888746261597,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0323,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.25502777099609375,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0327,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.032797425985336304,
      "eval_runtime": 153.4831,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.22471189498901367,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0272,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.16413453221321106,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0298,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.29426315426826477,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.025,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.36057248711586,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.03,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.1969657987356186,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0329,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.0344463549554348,
      "eval_runtime": 153.5439,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.4524608850479126,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0375,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.18868717551231384,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0269,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.28245416283607483,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0412,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.13972032070159912,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0479,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.21419501304626465,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.021,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.033020198345184326,
      "eval_runtime": 153.5004,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.1563420295715332,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0302,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.15720856189727783,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0238,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.13855260610580444,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.024,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.3510202169418335,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0373,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.09290120005607605,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.021,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.03481899946928024,
      "eval_runtime": 153.5322,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.24013583362102509,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.029,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.1155989021062851,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0259,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.1192859634757042,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.031,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.26294198632240295,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0359,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.3431008458137512,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0293,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.0336800292134285,
      "eval_runtime": 153.5675,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.3619416356086731,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0341,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.20910708606243134,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0295,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.1809314638376236,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0227,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.247967928647995,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.032,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.22990968823432922,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0229,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.0316387340426445,
      "eval_runtime": 153.618,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.21826183795928955,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0216,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.24052348732948303,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0241,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.13398176431655884,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0217,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.1350214034318924,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0202,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.2682395875453949,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0163,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.03314452990889549,
      "eval_runtime": 153.5504,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.2269609272480011,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0206,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.589372456073761,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0182,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.22638769447803497,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0144,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.25639820098876953,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0164,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.8363105654716492,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0355,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.03446889668703079,
      "eval_runtime": 153.5688,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.07887092232704163,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0157,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.31042855978012085,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0254,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.2983793616294861,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0161,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.20418977737426758,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0229,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.1082405298948288,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0129,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.03640338405966759,
      "eval_runtime": 153.5065,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.17766191065311432,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0209,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.49748969078063965,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0224,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.26122623682022095,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0225,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.10986757278442383,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0246,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.25621822476387024,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0188,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.03445897623896599,
      "eval_runtime": 153.6793,
      "eval_samples_per_second": 7.522,
      "eval_steps_per_second": 7.522,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.31663352251052856,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0213,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.7980028390884399,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.023,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.1939709484577179,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0187,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.33200180530548096,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.018,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.24923843145370483,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0158,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.036918047815561295,
      "eval_runtime": 153.6598,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.26014432311058044,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.014,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.34731632471084595,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0248,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.13151800632476807,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0214,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.25735166668891907,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0193,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.27362245321273804,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0158,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.0337149016559124,
      "eval_runtime": 153.5923,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.2439916729927063,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0166,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.1753331571817398,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0207,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.1372663527727127,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0194,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.3168756067752838,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0188,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.3263346552848816,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0219,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.032706085592508316,
      "eval_runtime": 153.6438,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.5108525156974792,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0185,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.2747831344604492,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0185,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.2620287835597992,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0235,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.25414949655532837,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0235,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.22579240798950195,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0171,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03209012746810913,
      "eval_runtime": 153.7652,
      "eval_samples_per_second": 7.518,
      "eval_steps_per_second": 7.518,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.22137820720672607,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0185,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.20134028792381287,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.029,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.37715816497802734,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0235,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.13373401761054993,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0176,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.16151902079582214,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0266,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.03183140605688095,
      "eval_runtime": 153.5645,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.14197759330272675,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.014,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.7188109159469604,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0224,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.4266507923603058,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0242,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.15354563295841217,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0194,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.2967778146266937,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0244,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.033598218113183975,
      "eval_runtime": 153.5876,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.26596781611442566,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0185,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.43763554096221924,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0255,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.552191436290741,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0238,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.22136783599853516,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0258,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.21893136203289032,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0231,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.03165159747004509,
      "eval_runtime": 153.5543,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.24700641632080078,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0136,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.3103381097316742,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0254,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.681530237197876,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.013,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.22731877863407135,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0192,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.2536536455154419,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0186,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.03188923001289368,
      "eval_runtime": 153.5489,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.29401111602783203,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0235,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.437948077917099,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0207,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.12666533887386322,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0179,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.2051066905260086,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0156,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.2554575204849243,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0296,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.031849347054958344,
      "eval_runtime": 153.5638,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.10088874399662018,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0066,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.19921386241912842,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.011,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.16983646154403687,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0098,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.3448812961578369,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0093,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.15517745912075043,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0102,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.03518034890294075,
      "eval_runtime": 153.6595,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 0.727115273475647,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0069,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.06432376801967621,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0098,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.12430425733327866,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0111,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.05119750648736954,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0073,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.46323227882385254,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0088,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.03952306881546974,
      "eval_runtime": 153.5967,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.1432286947965622,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0116,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.4693296551704407,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0105,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.2055344581604004,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.01,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.27149972319602966,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0099,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.275809645652771,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0099,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.037611912935972214,
      "eval_runtime": 153.4672,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.159001886844635,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.01,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.2500559091567993,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0115,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.4198713004589081,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0073,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.1822633445262909,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0051,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.36250460147857666,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0088,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.03914394974708557,
      "eval_runtime": 153.6333,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.28773462772369385,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0088,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.3742242157459259,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0146,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.16853833198547363,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0092,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.2965463101863861,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0086,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.1081705242395401,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0138,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.03789239749312401,
      "eval_runtime": 153.5992,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.22400911152362823,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0096,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.2630206048488617,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0066,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.2708906829357147,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0144,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.29934924840927124,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0122,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.2441568374633789,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.008,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.03879847377538681,
      "eval_runtime": 153.5039,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.15279562771320343,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0079,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.38521939516067505,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0055,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.24803970754146576,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.013,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.3747299313545227,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0067,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.14824092388153076,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0112,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.039526548236608505,
      "eval_runtime": 153.5777,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.3587571680545807,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0147,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.5314242839813232,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.014,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.16513949632644653,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0087,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.192780539393425,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0119,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.16656778752803802,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0045,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.038642995059490204,
      "eval_runtime": 153.5728,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.36291611194610596,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0138,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.14250826835632324,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0087,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.22413814067840576,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0081,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.2454422563314438,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0106,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.07044625282287598,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0127,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.03926815837621689,
      "eval_runtime": 153.4563,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.3773225247859955,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0093,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.2959787845611572,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0062,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.2518642246723175,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0094,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.3883044123649597,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0073,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.28949129581451416,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0074,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.039728496223688126,
      "eval_runtime": 153.5898,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.053840313106775284,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0075,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.22710900008678436,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0047,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.16011688113212585,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0123,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.397163063287735,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0115,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.30710774660110474,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0102,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.0399308018386364,
      "eval_runtime": 153.6539,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.04209047183394432,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0089,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.4946348965167999,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0043,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.02418249286711216,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0059,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.4496955871582031,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0058,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.8701761960983276,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0105,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.04102666676044464,
      "eval_runtime": 153.7332,
      "eval_samples_per_second": 7.52,
      "eval_steps_per_second": 7.52,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.36230742931365967,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0069,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.4245879650115967,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0095,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.8481021523475647,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.013,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.4724014699459076,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.008,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.05916439741849899,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0085,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.04120805114507675,
      "eval_runtime": 153.6006,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.15523886680603027,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0046,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.15802671015262604,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0017,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.14251644909381866,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0017,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.03620534390211105,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0053,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.05793361738324165,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.002,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.042593687772750854,
      "eval_runtime": 153.6233,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.23678231239318848,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0045,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.17169508337974548,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0029,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.03504731506109238,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0021,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.055942658334970474,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.005,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.22842493653297424,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0051,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.04528476670384407,
      "eval_runtime": 153.5503,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.04418591037392616,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0041,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.10995133966207504,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0037,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.04861040040850639,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0028,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.36464837193489075,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0042,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.0768180787563324,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0024,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.046782538294792175,
      "eval_runtime": 153.653,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.06458859890699387,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.002,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.015520976856350899,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0018,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.2339668869972229,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0025,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.25470617413520813,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0036,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.020808173343539238,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0022,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.047823019325733185,
      "eval_runtime": 153.599,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.02676706202328205,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0017,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.064998559653759,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0017,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.04566827416419983,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0025,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.49272918701171875,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0032,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.05237194895744324,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0031,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.04885661602020264,
      "eval_runtime": 153.6973,
      "eval_samples_per_second": 7.521,
      "eval_steps_per_second": 7.521,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.1885896921157837,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0024,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.3768438398838043,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0017,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.08688444644212723,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0081,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.030819829553365707,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.003,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.2656801640987396,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0042,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04927871376276016,
      "eval_runtime": 153.5722,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.06223306432366371,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.007,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.08695472031831741,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0042,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.023526107892394066,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0024,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.528073251247406,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.003,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.02643161080777645,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0017,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.04946993291378021,
      "eval_runtime": 153.6774,
      "eval_samples_per_second": 7.522,
      "eval_steps_per_second": 7.522,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.15514503419399261,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.001,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.1566014587879181,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0031,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.011410780251026154,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0018,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.4070422053337097,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0028,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.2733380198478699,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0025,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.04990413412451744,
      "eval_runtime": 153.5767,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.07123268395662308,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0018,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.07872439920902252,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0032,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.11000199615955353,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0061,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 1.2715568542480469,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0051,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.003793353447690606,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0025,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.04986204952001572,
      "eval_runtime": 153.5437,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.040509939193725586,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0036,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.5129521489143372,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0026,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.4752088189125061,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.003,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.031645532697439194,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0029,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.03806203231215477,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0022,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.04996936768293381,
      "eval_runtime": 153.6196,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.13632304966449738,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0049,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.1704360693693161,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0018,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.01616274192929268,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0038,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.02635025791823864,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0029,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.2748010456562042,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0048,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.04997621849179268,
      "eval_runtime": 153.708,
      "eval_samples_per_second": 7.521,
      "eval_steps_per_second": 7.521,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.07322733104228973,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0041,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.6482827067375183,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0034,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.056879185140132904,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0032,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.12976330518722534,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0024,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.06589337438344955,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.002,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.05013520270586014,
      "eval_runtime": 153.5775,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.3519189655780792,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0035,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.03211615979671478,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.002,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.38430172204971313,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0051,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.14816714823246002,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0025,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 7.696760527880847e+17,
      "train_loss": 0.051272692256112036,
      "train_runtime": 36384.7526,
      "train_samples_per_second": 1.429,
      "train_steps_per_second": 0.089
    }
  ],
  "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": 7.696760527880847e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}