{
  "best_metric": 0.030947599560022354,
  "best_model_checkpoint": "saves/psy-course/Ministral-8B-Instruct-2410/train/fold3/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": 2.7129294872283936,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.3361,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 2.7998530864715576,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.2705,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 3.176865339279175,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.0344,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 2.2590904235839844,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.5543,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 1.1336504220962524,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.2581,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.24142955243587494,
      "eval_runtime": 204.2818,
      "eval_samples_per_second": 5.659,
      "eval_steps_per_second": 5.659,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 1.488216519355774,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.1591,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.2044562101364136,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.1277,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 1.2177847623825073,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1016,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 1.4868992567062378,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.1181,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.7344977855682373,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0852,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.06964194774627686,
      "eval_runtime": 206.1842,
      "eval_samples_per_second": 5.607,
      "eval_steps_per_second": 5.607,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8094958662986755,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0648,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.799035370349884,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0724,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.0943729877471924,
      "learning_rate": 4e-05,
      "loss": 0.0693,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.7663387060165405,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.079,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.9013597965240479,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0612,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.05835791304707527,
      "eval_runtime": 206.1691,
      "eval_samples_per_second": 5.607,
      "eval_steps_per_second": 5.607,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.7569472193717957,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0614,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 0.7749213576316833,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0611,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.8102966547012329,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0527,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.3785383999347687,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0446,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5616209506988525,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0579,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.05373537540435791,
      "eval_runtime": 206.2593,
      "eval_samples_per_second": 5.605,
      "eval_steps_per_second": 5.605,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.7150264382362366,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0547,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.8464581370353699,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0499,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.361303448677063,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0378,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.5058801770210266,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0458,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.4638279378414154,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0436,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.04330739751458168,
      "eval_runtime": 206.3367,
      "eval_samples_per_second": 5.602,
      "eval_steps_per_second": 5.602,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.6122944951057434,
      "learning_rate": 8e-05,
      "loss": 0.0406,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.43344515562057495,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0416,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.5257235169410706,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0542,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.43893206119537354,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0505,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.39863696694374084,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0395,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04695376381278038,
      "eval_runtime": 206.5581,
      "eval_samples_per_second": 5.596,
      "eval_steps_per_second": 5.596,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.48693934082984924,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0513,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.3248462975025177,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0486,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.4264102578163147,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0478,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.40055567026138306,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0448,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.436410129070282,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0436,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04537563398480415,
      "eval_runtime": 206.6595,
      "eval_samples_per_second": 5.594,
      "eval_steps_per_second": 5.594,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.3384298086166382,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0556,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.41121906042099,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0382,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.6204984784126282,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0389,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.5442774891853333,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0342,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.33074888586997986,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0487,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.043601877987384796,
      "eval_runtime": 206.598,
      "eval_samples_per_second": 5.595,
      "eval_steps_per_second": 5.595,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.29724112153053284,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0416,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.492950975894928,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0467,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.5168049931526184,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0441,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.4878815710544586,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.06,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.486400842666626,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0302,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.0376572348177433,
      "eval_runtime": 206.2438,
      "eval_samples_per_second": 5.605,
      "eval_steps_per_second": 5.605,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.19189871847629547,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0345,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.35765111446380615,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0309,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.6570756435394287,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0423,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.22649993002414703,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0444,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.3664339780807495,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0301,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.03770829364657402,
      "eval_runtime": 206.2208,
      "eval_samples_per_second": 5.606,
      "eval_steps_per_second": 5.606,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.41719621419906616,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0312,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.30601200461387634,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0428,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.3862694203853607,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0528,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.3709644377231598,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0436,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.2840009331703186,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0422,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.035304248332977295,
      "eval_runtime": 206.2335,
      "eval_samples_per_second": 5.605,
      "eval_steps_per_second": 5.605,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.2940672039985657,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0354,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.26936912536621094,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.027,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.3261650800704956,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0349,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.2601899206638336,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0522,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.35210129618644714,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0352,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.034105055034160614,
      "eval_runtime": 206.1896,
      "eval_samples_per_second": 5.606,
      "eval_steps_per_second": 5.606,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.5016776323318481,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0329,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.5350625514984131,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0439,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.3558119237422943,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0434,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.2524898648262024,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0323,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.2947213351726532,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0327,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03460729122161865,
      "eval_runtime": 206.0695,
      "eval_samples_per_second": 5.61,
      "eval_steps_per_second": 5.61,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.2889857590198517,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0328,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.2618435323238373,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0319,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.2214561402797699,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.033,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.5059773921966553,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0459,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.4088204801082611,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0328,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.03613020479679108,
      "eval_runtime": 206.0875,
      "eval_samples_per_second": 5.609,
      "eval_steps_per_second": 5.609,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.3483586013317108,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0304,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.26023438572883606,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0296,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.3373538553714752,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0271,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.5894502997398376,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0373,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.4273133873939514,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0278,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.03469457849860191,
      "eval_runtime": 206.0887,
      "eval_samples_per_second": 5.609,
      "eval_steps_per_second": 5.609,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.2676852345466614,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0332,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.2201552540063858,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0366,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.3264544904232025,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0351,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.3948807716369629,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0346,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.28096306324005127,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0277,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.03360264375805855,
      "eval_runtime": 206.0768,
      "eval_samples_per_second": 5.61,
      "eval_steps_per_second": 5.61,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.1512415111064911,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0295,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.38847845792770386,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0266,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.2812429368495941,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0322,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.3665665090084076,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0263,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.31587252020835876,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0278,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.034673381596803665,
      "eval_runtime": 206.167,
      "eval_samples_per_second": 5.607,
      "eval_steps_per_second": 5.607,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.22272425889968872,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0262,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.9581472873687744,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0268,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.17257864773273468,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0273,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.24850012362003326,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0377,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.2794986665248871,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0208,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.034112174063920975,
      "eval_runtime": 206.049,
      "eval_samples_per_second": 5.61,
      "eval_steps_per_second": 5.61,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.29584696888923645,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0307,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.2715909779071808,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0305,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.27279144525527954,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0282,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.532991349697113,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0292,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.12468016147613525,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.037,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03454824537038803,
      "eval_runtime": 205.9458,
      "eval_samples_per_second": 5.613,
      "eval_steps_per_second": 5.613,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.2593000531196594,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0293,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.12138005346059799,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0233,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.38476723432540894,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0304,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.2522941827774048,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0301,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.41724494099617004,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0335,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.035657770931720734,
      "eval_runtime": 205.9581,
      "eval_samples_per_second": 5.613,
      "eval_steps_per_second": 5.613,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.3103076219558716,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0346,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.3300725519657135,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0328,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.29734182357788086,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0286,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.39744675159454346,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0258,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.24222847819328308,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0305,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.032191626727581024,
      "eval_runtime": 206.0598,
      "eval_samples_per_second": 5.61,
      "eval_steps_per_second": 5.61,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.1691523939371109,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0236,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.2633664608001709,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0288,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.5212844014167786,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.023,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.904237687587738,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0308,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.3433602452278137,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0337,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.03766747564077377,
      "eval_runtime": 205.8986,
      "eval_samples_per_second": 5.614,
      "eval_steps_per_second": 5.614,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.3405746519565582,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0318,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.2209967076778412,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0226,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.40576061606407166,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0344,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.2081703245639801,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.039,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.15152692794799805,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0221,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.03250744193792343,
      "eval_runtime": 206.0014,
      "eval_samples_per_second": 5.612,
      "eval_steps_per_second": 5.612,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.1951758712530136,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0328,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.27299606800079346,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0236,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.3067050278186798,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0245,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.3050411343574524,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0294,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.14683391153812408,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0192,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.03784427046775818,
      "eval_runtime": 205.7569,
      "eval_samples_per_second": 5.618,
      "eval_steps_per_second": 5.618,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.40251582860946655,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.027,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.171850323677063,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0234,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.123468779027462,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0278,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.2736228406429291,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0338,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.5321407914161682,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0282,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.03249868378043175,
      "eval_runtime": 205.8007,
      "eval_samples_per_second": 5.617,
      "eval_steps_per_second": 5.617,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.7650699615478516,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.03,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.2135242223739624,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0323,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.23481760919094086,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.024,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.6037285923957825,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0316,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.1242920458316803,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0216,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.030947599560022354,
      "eval_runtime": 205.7097,
      "eval_samples_per_second": 5.62,
      "eval_steps_per_second": 5.62,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.2552458941936493,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0169,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.45093226432800293,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0174,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.30306553840637207,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.018,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.07415292412042618,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0205,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.19454999268054962,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0172,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.031199408695101738,
      "eval_runtime": 205.533,
      "eval_samples_per_second": 5.624,
      "eval_steps_per_second": 5.624,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.2673622965812683,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0189,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.5548039078712463,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0162,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.15172356367111206,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.014,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.10359642654657364,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0139,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.5064147710800171,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0238,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.03423206880688667,
      "eval_runtime": 206.0002,
      "eval_samples_per_second": 5.612,
      "eval_steps_per_second": 5.612,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.0905536487698555,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0174,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.21097618341445923,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0209,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.14064981043338776,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0118,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.30104926228523254,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0193,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.12663201987743378,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0118,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.037876494228839874,
      "eval_runtime": 205.7156,
      "eval_samples_per_second": 5.619,
      "eval_steps_per_second": 5.619,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.30306276679039,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0173,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.13030338287353516,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0219,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.33335408568382263,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0199,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.1504828780889511,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0218,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.7764264345169067,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.02,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.03489841893315315,
      "eval_runtime": 205.6949,
      "eval_samples_per_second": 5.62,
      "eval_steps_per_second": 5.62,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.15162287652492523,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0218,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.14409463107585907,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0202,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.08051624149084091,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0132,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.23142850399017334,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.02,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.41714704036712646,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0162,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.03894202038645744,
      "eval_runtime": 205.7496,
      "eval_samples_per_second": 5.618,
      "eval_steps_per_second": 5.618,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.23085202276706696,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0139,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.22396957874298096,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0178,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.16986897587776184,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0183,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.2747359573841095,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0164,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.2951468527317047,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0138,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.036692582070827484,
      "eval_runtime": 205.3121,
      "eval_samples_per_second": 5.63,
      "eval_steps_per_second": 5.63,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.3755318522453308,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0171,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.189901202917099,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0189,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.2906784117221832,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0154,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.35252946615219116,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0226,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.182037815451622,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0193,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.03480008244514465,
      "eval_runtime": 204.9736,
      "eval_samples_per_second": 5.64,
      "eval_steps_per_second": 5.64,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.3194112479686737,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0126,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.3529907763004303,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0174,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.4833955466747284,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0181,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.18235287070274353,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.018,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.45806336402893066,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0208,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03558473661541939,
      "eval_runtime": 205.2271,
      "eval_samples_per_second": 5.633,
      "eval_steps_per_second": 5.633,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.4757392108440399,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0169,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.38991662859916687,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0263,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.5427097082138062,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0213,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.16979289054870605,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0132,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.12156710773706436,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0228,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.03264903649687767,
      "eval_runtime": 205.233,
      "eval_samples_per_second": 5.633,
      "eval_steps_per_second": 5.633,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.24176715314388275,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0148,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.3845118284225464,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0191,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.4394952952861786,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0231,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.1571224182844162,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0167,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.33347785472869873,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0195,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03232903033494949,
      "eval_runtime": 205.4246,
      "eval_samples_per_second": 5.627,
      "eval_steps_per_second": 5.627,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.3520554304122925,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0178,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.361346572637558,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0222,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.9539710283279419,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0209,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.2285640388727188,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0188,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.30473828315734863,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0219,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.03171970695257187,
      "eval_runtime": 205.2034,
      "eval_samples_per_second": 5.633,
      "eval_steps_per_second": 5.633,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.30735838413238525,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.014,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.31664836406707764,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0245,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.2992251217365265,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0134,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.16290029883384705,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0155,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.23914092779159546,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0169,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.032893672585487366,
      "eval_runtime": 205.2876,
      "eval_samples_per_second": 5.631,
      "eval_steps_per_second": 5.631,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.30798497796058655,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0212,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.3921068012714386,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0205,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.10858126729726791,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0139,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.3226618766784668,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0137,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.16719426214694977,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0235,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.03400113061070442,
      "eval_runtime": 205.1805,
      "eval_samples_per_second": 5.634,
      "eval_steps_per_second": 5.634,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.19277110695838928,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0062,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.04823404923081398,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0099,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.14269518852233887,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0085,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.2702989876270294,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0062,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.1252523958683014,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0092,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.03769548237323761,
      "eval_runtime": 205.727,
      "eval_samples_per_second": 5.619,
      "eval_steps_per_second": 5.619,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 0.4280714988708496,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0052,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.038806311786174774,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0093,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.10537301003932953,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.01,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.15586970746517181,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0095,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.6365472674369812,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0107,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.04130587726831436,
      "eval_runtime": 205.1703,
      "eval_samples_per_second": 5.634,
      "eval_steps_per_second": 5.634,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.06414748728275299,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0075,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.3724590539932251,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0076,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.1347227543592453,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0104,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.4249352216720581,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0081,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.2213268131017685,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0093,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.039757389575242996,
      "eval_runtime": 205.1175,
      "eval_samples_per_second": 5.636,
      "eval_steps_per_second": 5.636,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.07781395316123962,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0093,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.3686499297618866,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0098,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.2970055341720581,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0077,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.16210752725601196,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0067,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.6002753376960754,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0076,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.04058561474084854,
      "eval_runtime": 204.9772,
      "eval_samples_per_second": 5.64,
      "eval_steps_per_second": 5.64,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.19426089525222778,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0076,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.1623201072216034,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0086,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.24628621339797974,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0102,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.22403481602668762,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0057,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.13181528449058533,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0115,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.03798670321702957,
      "eval_runtime": 204.6984,
      "eval_samples_per_second": 5.647,
      "eval_steps_per_second": 5.647,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.03959172964096069,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0063,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.22117912769317627,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0074,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.27058205008506775,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0117,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.19432082772254944,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0116,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.12404157966375351,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0065,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.037050120532512665,
      "eval_runtime": 205.0649,
      "eval_samples_per_second": 5.637,
      "eval_steps_per_second": 5.637,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.07123508304357529,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.007,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.20804937183856964,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0048,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.1703631430864334,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0085,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.23363769054412842,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.006,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.08716625720262527,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0115,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.03938056528568268,
      "eval_runtime": 205.03,
      "eval_samples_per_second": 5.638,
      "eval_steps_per_second": 5.638,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.05090276524424553,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0083,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.42657580971717834,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0088,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.2217244654893875,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0075,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.6814695596694946,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0131,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.19453896582126617,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.006,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.03990740329027176,
      "eval_runtime": 204.8245,
      "eval_samples_per_second": 5.644,
      "eval_steps_per_second": 5.644,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.3497713804244995,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.015,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.23189005255699158,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0087,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.22898951172828674,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0101,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.2536865472793579,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.01,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.2650982141494751,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0119,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.036597561091184616,
      "eval_runtime": 204.8254,
      "eval_samples_per_second": 5.644,
      "eval_steps_per_second": 5.644,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.11519203335046768,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0082,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.045235008001327515,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0059,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.6607848405838013,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0095,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.12694579362869263,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.006,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.031129347160458565,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0068,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.038730423897504807,
      "eval_runtime": 204.9684,
      "eval_samples_per_second": 5.64,
      "eval_steps_per_second": 5.64,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.34668412804603577,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0099,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.4107019901275635,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0071,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.09503304958343506,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0116,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.11153826117515564,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0086,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.20469005405902863,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0079,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.0393652580678463,
      "eval_runtime": 204.6532,
      "eval_samples_per_second": 5.649,
      "eval_steps_per_second": 5.649,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.023741349577903748,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0097,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.2665390074253082,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0051,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.0571470707654953,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0065,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.14969134330749512,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0039,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.3763895034790039,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0092,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.04046928137540817,
      "eval_runtime": 204.8731,
      "eval_samples_per_second": 5.643,
      "eval_steps_per_second": 5.643,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 1.1178613901138306,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0088,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.28649458289146423,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0061,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.018256502225995064,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0118,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.2700015604496002,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0081,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.03348098695278168,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0088,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.039251428097486496,
      "eval_runtime": 204.8037,
      "eval_samples_per_second": 5.644,
      "eval_steps_per_second": 5.644,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.0350678451359272,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0036,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.1188267171382904,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0023,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.13259093463420868,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0024,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.016152262687683105,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0058,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.03424251452088356,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0017,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.0415111780166626,
      "eval_runtime": 203.6968,
      "eval_samples_per_second": 5.675,
      "eval_steps_per_second": 5.675,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.17463412880897522,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0056,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.07210592180490494,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0029,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.0655309185385704,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0028,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.013929404318332672,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.006,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.06032650172710419,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0076,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.04455247521400452,
      "eval_runtime": 204.5549,
      "eval_samples_per_second": 5.651,
      "eval_steps_per_second": 5.651,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.09175076335668564,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0046,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.0822143703699112,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0087,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.038291268050670624,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0023,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.05717266723513603,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0039,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.03294439613819122,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0017,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.04526090621948242,
      "eval_runtime": 204.6019,
      "eval_samples_per_second": 5.65,
      "eval_steps_per_second": 5.65,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.05905228480696678,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0018,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.10082878172397614,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0026,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.12022367864847183,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0028,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.021185602992773056,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0037,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.07566625624895096,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0027,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.04690178483724594,
      "eval_runtime": 177.4562,
      "eval_samples_per_second": 6.514,
      "eval_steps_per_second": 6.514,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.040234968066215515,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0026,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.021735873073339462,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0017,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.04729858413338661,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0022,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.015267541632056236,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0028,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.02810192108154297,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.003,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.04846663400530815,
      "eval_runtime": 203.6935,
      "eval_samples_per_second": 5.675,
      "eval_steps_per_second": 5.675,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.19460941851139069,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.002,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.02065412327647209,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0011,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.023317057639360428,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0067,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.017580850049853325,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0022,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.4506520628929138,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0047,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04926421493291855,
      "eval_runtime": 203.5851,
      "eval_samples_per_second": 5.678,
      "eval_steps_per_second": 5.678,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.032597895711660385,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0076,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.07329336553812027,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0038,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.04605919495224953,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0033,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.2866140604019165,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0061,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.017162030562758446,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0021,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.04840932786464691,
      "eval_runtime": 204.0265,
      "eval_samples_per_second": 5.666,
      "eval_steps_per_second": 5.666,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.0646035224199295,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0013,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.15958349406719208,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0072,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.03168458864092827,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0027,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.18651697039604187,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.002,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.04747346416115761,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0031,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.04847216233611107,
      "eval_runtime": 204.0849,
      "eval_samples_per_second": 5.664,
      "eval_steps_per_second": 5.664,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.06841609627008438,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.004,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.261078804731369,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0027,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.3438866436481476,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0044,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.1622624397277832,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0032,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.008925117552280426,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0036,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.0487653985619545,
      "eval_runtime": 204.2189,
      "eval_samples_per_second": 5.661,
      "eval_steps_per_second": 5.661,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.05633935704827309,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0033,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.6485021114349365,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0053,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.3753652274608612,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0075,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.26364409923553467,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0031,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.10415954887866974,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0028,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.04879259690642357,
      "eval_runtime": 204.5165,
      "eval_samples_per_second": 5.652,
      "eval_steps_per_second": 5.652,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.05293230712413788,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0077,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.061115507036447525,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0029,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.003671360434964299,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0048,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.03889927268028259,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0019,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.04646418243646622,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0031,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.048677533864974976,
      "eval_runtime": 203.9457,
      "eval_samples_per_second": 5.668,
      "eval_steps_per_second": 5.668,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.03158125653862953,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0028,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.16477619111537933,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0023,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.049345288425683975,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.002,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.01945136860013008,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0043,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.023310184478759766,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0035,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.04867711290717125,
      "eval_runtime": 204.1792,
      "eval_samples_per_second": 5.662,
      "eval_steps_per_second": 5.662,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.05658973753452301,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0047,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.04455265402793884,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0016,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.05317055061459541,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0053,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.07499335706233978,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0025,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 8.169046237996646e+17,
      "train_loss": 0.03533345094547594,
      "train_runtime": 44857.0674,
      "train_samples_per_second": 1.159,
      "train_steps_per_second": 0.072
    }
  ],
  "logging_steps": 10,
  "max_steps": 3245,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.169046237996646e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}