{
  "best_metric": 0.031097887083888054,
  "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold2/checkpoint-1850",
  "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": 3.7184898853302,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.998,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 4.057581901550293,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.926,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 3.969796657562256,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.7163,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 3.344660997390747,
      "learning_rate": 1.230769230769231e-05,
      "loss": 1.3792,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 2.0781354904174805,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.8706,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.6917524933815002,
      "eval_runtime": 162.0629,
      "eval_samples_per_second": 7.133,
      "eval_steps_per_second": 7.133,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 0.8926203846931458,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.5877,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.209885597229004,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.4071,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 0.6038497686386108,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.2133,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 0.9736213088035583,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.218,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.6797245740890503,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.1565,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.10812776535749435,
      "eval_runtime": 161.9404,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8679372072219849,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.1099,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.8212986588478088,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.1095,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.2469704151153564,
      "learning_rate": 4e-05,
      "loss": 0.0783,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.98650723695755,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0829,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.6961676478385925,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0877,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.0727488100528717,
      "eval_runtime": 161.9426,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.7198331356048584,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0827,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 1.3550165891647339,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0753,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.7143179178237915,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0592,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.669539749622345,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0518,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5632102489471436,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0728,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.055795446038246155,
      "eval_runtime": 161.9055,
      "eval_samples_per_second": 7.14,
      "eval_steps_per_second": 7.14,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.6978083848953247,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0642,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.756537139415741,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0619,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.33150988817214966,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0461,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.6388887763023376,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0587,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.5019749402999878,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0558,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.05286071076989174,
      "eval_runtime": 161.9827,
      "eval_samples_per_second": 7.137,
      "eval_steps_per_second": 7.137,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.3597353398799896,
      "learning_rate": 8e-05,
      "loss": 0.052,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.7532400488853455,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0522,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.5048686861991882,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0591,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.3708929717540741,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0618,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.4449922740459442,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0519,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04912148788571358,
      "eval_runtime": 161.9227,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.4721928536891937,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0572,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.9835398197174072,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0615,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.3343799114227295,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0557,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.4256531298160553,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0526,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.3974859118461609,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0457,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04665379226207733,
      "eval_runtime": 162.0019,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.3047736883163452,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0591,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.4831086993217468,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0438,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.4846143126487732,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0459,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.5660770535469055,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0411,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.25281986594200134,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0611,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.0438636876642704,
      "eval_runtime": 161.9246,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.18269330263137817,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0422,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.45976439118385315,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0531,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.5809056162834167,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0484,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.3213350772857666,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0607,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.2857457399368286,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0352,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.043765343725681305,
      "eval_runtime": 161.8962,
      "eval_samples_per_second": 7.14,
      "eval_steps_per_second": 7.14,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.35135766863822937,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0362,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.24557457864284515,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0296,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.5367558002471924,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0357,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.2384008914232254,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0402,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.24920910596847534,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0297,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.03962478041648865,
      "eval_runtime": 161.9666,
      "eval_samples_per_second": 7.137,
      "eval_steps_per_second": 7.137,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.5029212832450867,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0314,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.1468925029039383,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0487,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.426734983921051,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0619,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.3422389626502991,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0416,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.23161952197551727,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0474,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.03801177069544792,
      "eval_runtime": 161.9671,
      "eval_samples_per_second": 7.137,
      "eval_steps_per_second": 7.137,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.21084405481815338,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0314,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.3454536199569702,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0318,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.3396990895271301,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0381,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.26614364981651306,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0536,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.2947111129760742,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.038,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.03785459324717522,
      "eval_runtime": 161.9251,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.25751546025276184,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0352,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.18564197421073914,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0393,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.36892443895339966,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0439,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.8133302927017212,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0354,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.7524241209030151,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0326,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03658038005232811,
      "eval_runtime": 161.9544,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.443203866481781,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0396,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.35536855459213257,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0382,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.2834513485431671,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0327,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.457702100276947,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0482,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.21063891053199768,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0356,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.03831103816628456,
      "eval_runtime": 161.8865,
      "eval_samples_per_second": 7.141,
      "eval_steps_per_second": 7.141,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.32350003719329834,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0321,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.18803557753562927,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0265,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.275654673576355,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0312,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.5381351113319397,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0381,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.2770777642726898,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0307,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.035575151443481445,
      "eval_runtime": 161.923,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.21906712651252747,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.038,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.2827393412590027,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0393,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.21216118335723877,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0325,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.2924606502056122,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0345,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.2744523882865906,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0301,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.034924864768981934,
      "eval_runtime": 162.0264,
      "eval_samples_per_second": 7.135,
      "eval_steps_per_second": 7.135,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.15679049491882324,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0352,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.2103583663702011,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0266,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.298764169216156,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0309,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.4625924527645111,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0291,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.5786860585212708,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0339,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.03563137724995613,
      "eval_runtime": 162.0186,
      "eval_samples_per_second": 7.135,
      "eval_steps_per_second": 7.135,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.20404283702373505,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0336,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.32495617866516113,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0292,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.11683793365955353,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0278,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.17084576189517975,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.036,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.24109672009944916,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0208,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.034224171191453934,
      "eval_runtime": 162.0536,
      "eval_samples_per_second": 7.133,
      "eval_steps_per_second": 7.133,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.20868462324142456,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0298,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.2536265254020691,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0349,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.24942819774150848,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0278,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.4909574091434479,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0369,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.08471255004405975,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0414,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03562645986676216,
      "eval_runtime": 161.9588,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.30817851424217224,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.03,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.07204034179449081,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0233,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.3109855055809021,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.034,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.19040171802043915,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0335,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.22780922055244446,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0324,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.0334600955247879,
      "eval_runtime": 161.9465,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.20537780225276947,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0339,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.24358350038528442,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0398,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.1369139850139618,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0313,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.12342860549688339,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0323,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.2659969627857208,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0327,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03304734453558922,
      "eval_runtime": 161.9286,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.21718044579029083,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0273,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.160589799284935,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0289,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.1429855227470398,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.023,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.5315752625465393,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0305,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.2066251039505005,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0319,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.0352105088531971,
      "eval_runtime": 162.1153,
      "eval_samples_per_second": 7.131,
      "eval_steps_per_second": 7.131,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.4789196252822876,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0379,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.1851641833782196,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0276,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.2565332353115082,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0393,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.15459075570106506,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0467,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.19173301756381989,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0215,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.033734068274497986,
      "eval_runtime": 162.0742,
      "eval_samples_per_second": 7.133,
      "eval_steps_per_second": 7.133,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.16135014593601227,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0316,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.1974264532327652,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0252,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.14705108106136322,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0232,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.3617679178714752,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0367,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.10081601142883301,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0222,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.03507999703288078,
      "eval_runtime": 162.0768,
      "eval_samples_per_second": 7.132,
      "eval_steps_per_second": 7.132,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.28450441360473633,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0307,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.13416704535484314,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0267,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.11310270428657532,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0301,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.23658329248428345,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0366,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.3311021029949188,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0288,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.0339454784989357,
      "eval_runtime": 161.9591,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.34184664487838745,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0313,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.20106428861618042,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0291,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.15832459926605225,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0229,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.22986823320388794,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0336,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.17141278088092804,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0225,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.03173167631030083,
      "eval_runtime": 161.9488,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.264302521944046,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0228,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.27341678738594055,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0228,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.13762882351875305,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0232,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.1359223574399948,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0199,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.2618677318096161,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.017,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.033239927142858505,
      "eval_runtime": 161.9491,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.1168808713555336,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0212,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.4989379942417145,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0185,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.2586815357208252,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0141,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.23399510979652405,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0165,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.6931097507476807,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0296,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.034901056438684464,
      "eval_runtime": 161.9274,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.08135487139225006,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0156,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.3020218312740326,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0241,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.4115012288093567,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0153,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.20426853001117706,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0216,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.10999682545661926,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0141,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.03486805036664009,
      "eval_runtime": 161.9553,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.30394506454467773,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0209,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.1824904829263687,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0212,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.23356309533119202,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0234,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.11600757390260696,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0242,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.23938904702663422,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0188,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.03359838202595711,
      "eval_runtime": 161.9206,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.18546506762504578,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0215,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.1776905059814453,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0228,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.21087220311164856,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0195,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.35192716121673584,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0189,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.20723119378089905,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0154,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.036039017140865326,
      "eval_runtime": 161.9382,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.24183039367198944,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0129,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.33688676357269287,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0265,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.12291789799928665,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0194,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.245294451713562,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0195,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.24481762945652008,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.014,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.03427223861217499,
      "eval_runtime": 162.0458,
      "eval_samples_per_second": 7.134,
      "eval_steps_per_second": 7.134,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.27064210176467896,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0172,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.18759691715240479,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0204,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.28914928436279297,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0194,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.3302218019962311,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0193,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.2863548696041107,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0219,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.03165784105658531,
      "eval_runtime": 162.0531,
      "eval_samples_per_second": 7.133,
      "eval_steps_per_second": 7.133,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.5005291104316711,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0182,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.3212396502494812,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0185,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.2601298391819,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0242,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.2422500103712082,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0234,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.2468331903219223,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0186,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03232479467988014,
      "eval_runtime": 162.0455,
      "eval_samples_per_second": 7.134,
      "eval_steps_per_second": 7.134,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.2208268791437149,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0186,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.21705351769924164,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0293,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.3873547911643982,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0243,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.11156151443719864,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0164,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.17500939965248108,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0261,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.03119010664522648,
      "eval_runtime": 162.0111,
      "eval_samples_per_second": 7.135,
      "eval_steps_per_second": 7.135,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.24699608981609344,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0147,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.3515814542770386,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0218,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.39447590708732605,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0234,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.11628853529691696,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0196,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.3066122233867645,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.024,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03355173394083977,
      "eval_runtime": 161.9958,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.2605317234992981,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0209,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.2621488571166992,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0236,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.6438840627670288,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0258,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.19783122837543488,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0243,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.1960369348526001,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0231,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.031097887083888054,
      "eval_runtime": 162.0725,
      "eval_samples_per_second": 7.133,
      "eval_steps_per_second": 7.133,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.24138881266117096,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0135,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.288493275642395,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0255,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.5271561741828918,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0118,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.1764020323753357,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0182,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.23046904802322388,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.017,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.031723637133836746,
      "eval_runtime": 162.0379,
      "eval_samples_per_second": 7.134,
      "eval_steps_per_second": 7.134,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.23520152270793915,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0225,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.41894492506980896,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0197,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.15018050372600555,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0166,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.20791491866111755,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0147,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.2574230432510376,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0277,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.032280512154102325,
      "eval_runtime": 162.0849,
      "eval_samples_per_second": 7.132,
      "eval_steps_per_second": 7.132,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.11673092097043991,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0083,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.2004876434803009,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.011,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.38148438930511475,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0098,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.31826213002204895,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0098,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.21975873410701752,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0122,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.03427407518029213,
      "eval_runtime": 162.1036,
      "eval_samples_per_second": 7.131,
      "eval_steps_per_second": 7.131,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 1.1208361387252808,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.007,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.050865769386291504,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0107,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.14180181920528412,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0135,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.08250804990530014,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0076,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.2896089255809784,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0096,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.03652224689722061,
      "eval_runtime": 162.1326,
      "eval_samples_per_second": 7.13,
      "eval_steps_per_second": 7.13,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.4725174605846405,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0074,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.48031046986579895,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0083,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.2488190233707428,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0091,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.30023273825645447,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0103,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.29484134912490845,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0077,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.040214501321315765,
      "eval_runtime": 161.9877,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.09078393876552582,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0097,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.22656723856925964,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0088,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.509220540523529,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0075,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.1839645653963089,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0067,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.44757288694381714,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.01,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.03901513293385506,
      "eval_runtime": 161.9901,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.19456231594085693,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0095,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.4218941032886505,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0124,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.09554542601108551,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0093,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.4752233922481537,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0087,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.09902461618185043,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0159,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.0384637750685215,
      "eval_runtime": 162.0516,
      "eval_samples_per_second": 7.134,
      "eval_steps_per_second": 7.134,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.2265639752149582,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0098,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.27344372868537903,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0083,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.20929038524627686,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0125,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.25159597396850586,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0134,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.267764151096344,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0092,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.037459906190633774,
      "eval_runtime": 162.0573,
      "eval_samples_per_second": 7.133,
      "eval_steps_per_second": 7.133,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.16818983852863312,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0109,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.367042601108551,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0063,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.16918085515499115,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0114,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.3930964767932892,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0076,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.15988709032535553,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0113,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.038027454167604446,
      "eval_runtime": 162.0043,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.18698349595069885,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0138,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.509650707244873,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0133,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.18544328212738037,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0108,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.25118619203567505,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0111,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.06452061980962753,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0049,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.03875124081969261,
      "eval_runtime": 162.0036,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.29958564043045044,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0122,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.30901655554771423,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0098,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.23807315528392792,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.01,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.3113029897212982,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0111,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.12216784060001373,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0115,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.03817567601799965,
      "eval_runtime": 161.9938,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.44794878363609314,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0094,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.2763858437538147,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.007,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.1254671961069107,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0087,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.4882674813270569,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0079,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.6019213199615479,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0072,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.03840717300772667,
      "eval_runtime": 161.9697,
      "eval_samples_per_second": 7.137,
      "eval_steps_per_second": 7.137,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.10386688262224197,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0074,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.1666368693113327,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0049,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.16184291243553162,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0108,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.2994827926158905,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0105,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.23739321529865265,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0078,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.039616908878088,
      "eval_runtime": 162.0745,
      "eval_samples_per_second": 7.133,
      "eval_steps_per_second": 7.133,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.033695876598358154,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0084,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.4435981512069702,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0068,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.09426259994506836,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0064,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.3510986566543579,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0064,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.4434868395328522,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0106,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.039574407041072845,
      "eval_runtime": 162.1158,
      "eval_samples_per_second": 7.131,
      "eval_steps_per_second": 7.131,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.31082645058631897,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0058,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.3960266709327698,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0055,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.10288826376199722,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0124,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.4329286515712738,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0072,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.09900029003620148,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0094,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.04026506096124649,
      "eval_runtime": 161.9578,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.10629642009735107,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0043,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.15210218727588654,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0016,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.09583138674497604,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0019,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.056596193462610245,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0051,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.05663219094276428,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.002,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.04172023385763168,
      "eval_runtime": 161.93,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.366218626499176,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0046,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.5057364106178284,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0028,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.0566055066883564,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0027,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.060457080602645874,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0048,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.06777138262987137,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0044,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.04404466226696968,
      "eval_runtime": 162.0836,
      "eval_samples_per_second": 7.132,
      "eval_steps_per_second": 7.132,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.035058751702308655,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0045,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.1338721215724945,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0033,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.10651364177465439,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.003,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.2930299639701843,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0043,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.12897969782352448,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0025,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.045186951756477356,
      "eval_runtime": 161.971,
      "eval_samples_per_second": 7.137,
      "eval_steps_per_second": 7.137,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.06688085943460464,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0019,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.08125784248113632,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0016,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.4488185942173004,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0031,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.1194901391863823,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0044,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.02015116438269615,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0043,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.04616488143801689,
      "eval_runtime": 162.0274,
      "eval_samples_per_second": 7.135,
      "eval_steps_per_second": 7.135,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.018860025331377983,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.002,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.023462194949388504,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0019,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.08702544867992401,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0032,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.4034475088119507,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.004,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.10886283218860626,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0043,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.04664229974150658,
      "eval_runtime": 161.9321,
      "eval_samples_per_second": 7.139,
      "eval_steps_per_second": 7.139,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.1697157770395279,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0022,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.04288719967007637,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0013,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.08654361963272095,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0083,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.03394753485918045,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.002,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.28359848260879517,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0046,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04686031490564346,
      "eval_runtime": 161.9607,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.05516982451081276,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0065,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.059713490307331085,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0042,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.022417467087507248,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0019,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.4301671087741852,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0033,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.012211650609970093,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.002,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.04694652557373047,
      "eval_runtime": 161.6697,
      "eval_samples_per_second": 7.15,
      "eval_steps_per_second": 7.15,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.17931555211544037,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.001,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.8564173579216003,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0035,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.021505998447537422,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0021,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.1402561515569687,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0023,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.2800159752368927,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0031,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.04732413589954376,
      "eval_runtime": 159.9248,
      "eval_samples_per_second": 7.228,
      "eval_steps_per_second": 7.228,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.07902012020349503,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.002,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.04443228244781494,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0031,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.33196526765823364,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0045,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.27408158779144287,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0041,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.0130156259983778,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0055,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.047541093081235886,
      "eval_runtime": 159.3353,
      "eval_samples_per_second": 7.255,
      "eval_steps_per_second": 7.255,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.09641927480697632,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0019,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.4680754840373993,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0021,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.44706758856773376,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0032,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.09566780179738998,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0037,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.08343682438135147,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0024,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.04742332547903061,
      "eval_runtime": 159.0411,
      "eval_samples_per_second": 7.269,
      "eval_steps_per_second": 7.269,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.04475397989153862,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0048,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.20023611187934875,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0018,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.04049602523446083,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0036,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.06195271015167236,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.002,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.2731001675128937,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0059,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.047556012868881226,
      "eval_runtime": 159.1576,
      "eval_samples_per_second": 7.263,
      "eval_steps_per_second": 7.263,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.08143557608127594,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.004,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.3825170695781708,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0023,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.07588217407464981,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0033,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.19663433730602264,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0027,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.03383241593837738,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.002,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.04763248190283775,
      "eval_runtime": 159.1779,
      "eval_samples_per_second": 7.262,
      "eval_steps_per_second": 7.262,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.06472008675336838,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0035,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.026334259659051895,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0023,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.35545289516448975,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0045,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.07686346024274826,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0043,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 7.696760527880847e+17,
      "train_loss": 0.051167534938360494,
      "train_runtime": 38225.4735,
      "train_samples_per_second": 1.36,
      "train_steps_per_second": 0.085
    }
  ],
  "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
}