{
  "best_metric": 0.03381282836198807,
  "best_model_checkpoint": "saves/psy-course/Llama-3.1-8B-Instruct/train/fold1/checkpoint-1350",
  "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": 5.0106048583984375,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.6199,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 5.837756633758545,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.5918,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 7.694819927215576,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.3184,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 2.596924066543579,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.8797,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 1.5033406019210815,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.5562,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.40699678659439087,
      "eval_runtime": 185.4981,
      "eval_samples_per_second": 6.232,
      "eval_steps_per_second": 6.232,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 0.904315173625946,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.3409,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.0325497388839722,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.2488,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 0.5230876803398132,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1659,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 0.9011468887329102,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.167,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.6379432678222656,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.1226,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.09147929400205612,
      "eval_runtime": 188.019,
      "eval_samples_per_second": 6.148,
      "eval_steps_per_second": 6.148,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8800907731056213,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0863,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.41934528946876526,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0894,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 0.8273314237594604,
      "learning_rate": 4e-05,
      "loss": 0.0754,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.9328588247299194,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0912,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.8604061007499695,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0859,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.0644315704703331,
      "eval_runtime": 188.1359,
      "eval_samples_per_second": 6.144,
      "eval_steps_per_second": 6.144,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 1.4581396579742432,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0727,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 0.7286003232002258,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0806,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.9489079713821411,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0615,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.828267514705658,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0521,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.6866675615310669,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0714,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.052990105003118515,
      "eval_runtime": 188.5543,
      "eval_samples_per_second": 6.131,
      "eval_steps_per_second": 6.131,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.725680410861969,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0609,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.8611897826194763,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0594,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.34665313363075256,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0502,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.6683058738708496,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0525,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.4202400743961334,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.054,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.049264274537563324,
      "eval_runtime": 188.1154,
      "eval_samples_per_second": 6.145,
      "eval_steps_per_second": 6.145,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.3903276324272156,
      "learning_rate": 8e-05,
      "loss": 0.0509,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.7589438557624817,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0538,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.5696157217025757,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0605,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.3944014608860016,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0605,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.588414192199707,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0553,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04779207333922386,
      "eval_runtime": 187.659,
      "eval_samples_per_second": 6.16,
      "eval_steps_per_second": 6.16,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.46967023611068726,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0552,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.4400937259197235,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0587,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.7571660876274109,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0608,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.356673002243042,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0566,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.4959634244441986,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.053,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.0478888563811779,
      "eval_runtime": 187.3914,
      "eval_samples_per_second": 6.169,
      "eval_steps_per_second": 6.169,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.3542782664299011,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0679,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.49819979071617126,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0511,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.37595507502555847,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0499,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.8225634694099426,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0364,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.4043976962566376,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0645,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.048657890409231186,
      "eval_runtime": 187.2004,
      "eval_samples_per_second": 6.175,
      "eval_steps_per_second": 6.175,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.29372572898864746,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0462,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.5222019553184509,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0509,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.8109437227249146,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0529,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.4393615424633026,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0658,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.3185001015663147,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0391,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.04509444162249565,
      "eval_runtime": 187.6444,
      "eval_samples_per_second": 6.161,
      "eval_steps_per_second": 6.161,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.4155886173248291,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0422,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.31224191188812256,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0336,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.4808642566204071,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0403,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.18449532985687256,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0452,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.48674625158309937,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0339,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.04046137258410454,
      "eval_runtime": 187.9753,
      "eval_samples_per_second": 6.15,
      "eval_steps_per_second": 6.15,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.5263739228248596,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0328,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.24675846099853516,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0519,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.3642294704914093,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0656,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.487305223941803,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0456,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.2354636788368225,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0513,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.04050951078534126,
      "eval_runtime": 188.1628,
      "eval_samples_per_second": 6.144,
      "eval_steps_per_second": 6.144,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.18946993350982666,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0417,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.40360134840011597,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0333,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.316399484872818,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0362,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.4593469798564911,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0579,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.28380537033081055,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0374,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.03845710679888725,
      "eval_runtime": 187.7867,
      "eval_samples_per_second": 6.156,
      "eval_steps_per_second": 6.156,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.2812202274799347,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0377,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.2947665750980377,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.053,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.33196261525154114,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0484,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.27342891693115234,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0355,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.8010798692703247,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0334,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03811522200703621,
      "eval_runtime": 187.6177,
      "eval_samples_per_second": 6.161,
      "eval_steps_per_second": 6.161,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.4056728780269623,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0398,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.2418406456708908,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.038,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.28666359186172485,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0351,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.41955092549324036,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0486,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.4770854413509369,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0405,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.04033925011754036,
      "eval_runtime": 187.98,
      "eval_samples_per_second": 6.15,
      "eval_steps_per_second": 6.15,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.4085399806499481,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.033,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.24918586015701294,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0338,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.21423755586147308,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0323,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.7340275049209595,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.05,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.25578582286834717,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0337,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.0360797680914402,
      "eval_runtime": 188.046,
      "eval_samples_per_second": 6.147,
      "eval_steps_per_second": 6.147,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.23202097415924072,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0407,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.2221558541059494,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0407,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.29214197397232056,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.038,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.4598447382450104,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0358,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.24903126060962677,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0344,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.03588108345866203,
      "eval_runtime": 187.57,
      "eval_samples_per_second": 6.163,
      "eval_steps_per_second": 6.163,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.15016677975654602,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.033,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.3766383230686188,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0315,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.3017776310443878,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.038,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.37091585993766785,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0279,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.32512232661247253,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0344,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.037309836596250534,
      "eval_runtime": 187.42,
      "eval_samples_per_second": 6.168,
      "eval_steps_per_second": 6.168,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.19161005318164825,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0331,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.41218462586402893,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.033,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.151007279753685,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0266,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.19659636914730072,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0434,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.1985592097043991,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0221,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.03570735454559326,
      "eval_runtime": 187.0359,
      "eval_samples_per_second": 6.181,
      "eval_steps_per_second": 6.181,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.30656376481056213,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0304,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.5850927233695984,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0366,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.21166619658470154,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0328,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.4661455452442169,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0377,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.11176864802837372,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0457,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03746819868683815,
      "eval_runtime": 187.3821,
      "eval_samples_per_second": 6.169,
      "eval_steps_per_second": 6.169,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.2294084131717682,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0286,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.09593243896961212,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.027,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.3172210454940796,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0345,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.21809054911136627,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0398,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.36311158537864685,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0357,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.03651273250579834,
      "eval_runtime": 187.9276,
      "eval_samples_per_second": 6.151,
      "eval_steps_per_second": 6.151,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.2572426199913025,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0398,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.2838153541088104,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0374,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.22808720171451569,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0357,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.25805601477622986,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.033,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.2337527722120285,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.034,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03555304929614067,
      "eval_runtime": 188.0492,
      "eval_samples_per_second": 6.147,
      "eval_steps_per_second": 6.147,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.25864410400390625,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0285,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.20299604535102844,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0318,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.2811971604824066,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0271,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.32877394556999207,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0262,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.15650756657123566,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0394,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.03640196472406387,
      "eval_runtime": 187.9016,
      "eval_samples_per_second": 6.152,
      "eval_steps_per_second": 6.152,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.41813358664512634,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.041,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.18765102326869965,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0315,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.2128918319940567,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0396,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.12499304115772247,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.049,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.19719086587429047,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0244,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.0359814316034317,
      "eval_runtime": 187.8554,
      "eval_samples_per_second": 6.154,
      "eval_steps_per_second": 6.154,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.1700337678194046,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0349,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.2521379292011261,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0263,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.26141706109046936,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0283,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.4066651165485382,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0356,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.1369960904121399,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.023,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.03705717623233795,
      "eval_runtime": 187.2707,
      "eval_samples_per_second": 6.173,
      "eval_steps_per_second": 6.173,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.243990957736969,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0346,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.14930765330791473,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.026,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.10069471597671509,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0338,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.24041235446929932,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0417,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.28064167499542236,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0339,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.035844240337610245,
      "eval_runtime": 187.6727,
      "eval_samples_per_second": 6.16,
      "eval_steps_per_second": 6.16,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.3073214292526245,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0342,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.26703837513923645,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0364,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.21204885840415955,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.025,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.3635454475879669,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0372,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.11565925180912018,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0252,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.0344332717359066,
      "eval_runtime": 187.9094,
      "eval_samples_per_second": 6.152,
      "eval_steps_per_second": 6.152,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.19626887142658234,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0227,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.34589436650276184,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0237,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.21786868572235107,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0255,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.15011264383792877,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0248,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.18167394399642944,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.019,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.03381282836198807,
      "eval_runtime": 187.8437,
      "eval_samples_per_second": 6.154,
      "eval_steps_per_second": 6.154,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.10537901520729065,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0227,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.5727980136871338,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0218,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.24711571633815765,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0153,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.18777436017990112,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0208,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.3703112304210663,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0297,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.03612199053168297,
      "eval_runtime": 187.1523,
      "eval_samples_per_second": 6.177,
      "eval_steps_per_second": 6.177,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.11688963323831558,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.024,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.27680692076683044,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.03,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.1767677664756775,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0178,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.21000392735004425,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0261,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.10033946484327316,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0142,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.03711491823196411,
      "eval_runtime": 186.5296,
      "eval_samples_per_second": 6.197,
      "eval_steps_per_second": 6.197,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.18027913570404053,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0233,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.16498275101184845,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0268,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.16076700389385223,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0233,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.12392139434814453,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0253,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.30897045135498047,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0227,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.0353684276342392,
      "eval_runtime": 186.308,
      "eval_samples_per_second": 6.205,
      "eval_steps_per_second": 6.205,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.1789022982120514,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0244,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.15649132430553436,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0214,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.1704787164926529,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0201,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.22701707482337952,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0199,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.21540381014347076,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0179,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.04112153872847557,
      "eval_runtime": 186.6486,
      "eval_samples_per_second": 6.193,
      "eval_steps_per_second": 6.193,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.5612645745277405,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0155,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.20288218557834625,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0246,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.2309695929288864,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0217,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.23696504533290863,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0262,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.38717809319496155,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0189,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.036341600120067596,
      "eval_runtime": 186.7956,
      "eval_samples_per_second": 6.189,
      "eval_steps_per_second": 6.189,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.22446145117282867,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0182,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.2164313942193985,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0236,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.14063049852848053,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0208,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.3135584890842438,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0236,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.16733896732330322,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0246,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.0360361784696579,
      "eval_runtime": 186.9536,
      "eval_samples_per_second": 6.183,
      "eval_steps_per_second": 6.183,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.3761216402053833,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0202,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.37585002183914185,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0231,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.5329094529151917,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0261,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.28828349709510803,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.025,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.5588506460189819,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0244,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03538386896252632,
      "eval_runtime": 186.8111,
      "eval_samples_per_second": 6.188,
      "eval_steps_per_second": 6.188,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.22283612191677094,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0197,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.27689316868782043,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0327,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.3931240141391754,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0252,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.19020158052444458,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.018,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.11216674000024796,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0295,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.03553741052746773,
      "eval_runtime": 186.4672,
      "eval_samples_per_second": 6.199,
      "eval_steps_per_second": 6.199,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.19189698994159698,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0174,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.37652498483657837,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0274,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.4836426377296448,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0286,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.20131860673427582,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0228,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.3384079337120056,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0257,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03536887839436531,
      "eval_runtime": 186.2101,
      "eval_samples_per_second": 6.208,
      "eval_steps_per_second": 6.208,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.2767975330352783,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0223,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.2422277182340622,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0244,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.5190377831459045,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0207,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.2180313616991043,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0256,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.2425866425037384,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0313,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.03393126651644707,
      "eval_runtime": 186.1971,
      "eval_samples_per_second": 6.208,
      "eval_steps_per_second": 6.208,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.2784096598625183,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0183,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.3296208381652832,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0286,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.5275260806083679,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0176,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.15682271122932434,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0193,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.2102474868297577,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0208,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.034816108644008636,
      "eval_runtime": 186.8118,
      "eval_samples_per_second": 6.188,
      "eval_steps_per_second": 6.188,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.21557755768299103,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0228,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.4167330861091614,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0201,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.14496108889579773,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.017,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.327414870262146,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0187,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.1764715611934662,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0314,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.034862760454416275,
      "eval_runtime": 186.7003,
      "eval_samples_per_second": 6.192,
      "eval_steps_per_second": 6.192,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.10091342031955719,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.01,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.11954168230295181,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.012,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.4728318750858307,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0127,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.22921116650104523,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0078,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.17478808760643005,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0119,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.0392003133893013,
      "eval_runtime": 186.2797,
      "eval_samples_per_second": 6.206,
      "eval_steps_per_second": 6.206,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 0.536690890789032,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0087,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.08325193077325821,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0102,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.3025301694869995,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0162,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.0437934547662735,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0094,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.6723917126655579,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0147,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.04251823201775551,
      "eval_runtime": 186.0299,
      "eval_samples_per_second": 6.214,
      "eval_steps_per_second": 6.214,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.06159785017371178,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0085,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.3324078321456909,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0112,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.4389561414718628,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0177,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.30180397629737854,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0112,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.18360494077205658,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.011,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.040228795260190964,
      "eval_runtime": 186.1617,
      "eval_samples_per_second": 6.21,
      "eval_steps_per_second": 6.21,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.1583905965089798,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0104,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.18586666882038116,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0107,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.389637291431427,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0168,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.093821682035923,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0079,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.5508767366409302,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.012,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.04161189869046211,
      "eval_runtime": 186.4024,
      "eval_samples_per_second": 6.202,
      "eval_steps_per_second": 6.202,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.08765236288309097,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0131,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.38336998224258423,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0135,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.2137378603219986,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.01,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.6786463856697083,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0108,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.2550009489059448,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0205,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.03981276601552963,
      "eval_runtime": 186.4982,
      "eval_samples_per_second": 6.198,
      "eval_steps_per_second": 6.198,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.531782865524292,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0139,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.30775323510169983,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0108,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.45309802889823914,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0166,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.1881212443113327,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0135,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.1511940211057663,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0127,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.039891570806503296,
      "eval_runtime": 186.7336,
      "eval_samples_per_second": 6.191,
      "eval_steps_per_second": 6.191,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.049151744693517685,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0073,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.42421895265579224,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0108,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.3423459529876709,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0148,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.324720561504364,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0086,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.1445930153131485,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0141,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.0418756864964962,
      "eval_runtime": 186.4389,
      "eval_samples_per_second": 6.2,
      "eval_steps_per_second": 6.2,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.27503931522369385,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0127,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.46346715092658997,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0145,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.11502953618764877,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0085,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.23123833537101746,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0151,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.17446036636829376,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0081,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.043476805090904236,
      "eval_runtime": 185.8129,
      "eval_samples_per_second": 6.221,
      "eval_steps_per_second": 6.221,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.21171227097511292,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0159,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.20531825721263885,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0094,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.20440693199634552,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.013,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.4047004282474518,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0153,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.08073116838932037,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0132,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.04118052124977112,
      "eval_runtime": 185.9242,
      "eval_samples_per_second": 6.218,
      "eval_steps_per_second": 6.218,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.36109527945518494,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.011,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.29550692439079285,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0091,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.6761743426322937,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0147,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.38532111048698425,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0086,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.2847141921520233,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0111,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.04333437979221344,
      "eval_runtime": 185.5961,
      "eval_samples_per_second": 6.229,
      "eval_steps_per_second": 6.229,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.22558213770389557,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0113,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.13944610953330994,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0076,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.32348164916038513,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.012,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.556498646736145,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0137,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.19787055253982544,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0069,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.044614341109991074,
      "eval_runtime": 184.6686,
      "eval_samples_per_second": 6.26,
      "eval_steps_per_second": 6.26,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.23233471810817719,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0141,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.8727635741233826,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0062,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.01985669508576393,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0093,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.19380433857440948,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0059,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.30975109338760376,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0135,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.04540552943944931,
      "eval_runtime": 183.9066,
      "eval_samples_per_second": 6.286,
      "eval_steps_per_second": 6.286,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.36627084016799927,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0127,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.3993094861507416,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.008,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.0302486140280962,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.014,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.4973717927932739,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.009,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.07611800730228424,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0093,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.04523477330803871,
      "eval_runtime": 182.4267,
      "eval_samples_per_second": 6.337,
      "eval_steps_per_second": 6.337,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.04856440797448158,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0061,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.335890531539917,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0027,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.3047681748867035,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0032,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.0912618488073349,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0058,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.07931726425886154,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0024,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.04883251711726189,
      "eval_runtime": 180.2492,
      "eval_samples_per_second": 6.413,
      "eval_steps_per_second": 6.413,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.5609334111213684,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0065,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.7794962525367737,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0035,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.09867623448371887,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0035,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.08890987932682037,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0042,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.17334403097629547,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0082,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.05498456954956055,
      "eval_runtime": 177.9616,
      "eval_samples_per_second": 6.496,
      "eval_steps_per_second": 6.496,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.4304162561893463,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0058,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.22083640098571777,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0107,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.32587379217147827,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0059,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.9534060955047607,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0066,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.18303728103637695,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.004,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.05399877950549126,
      "eval_runtime": 176.8479,
      "eval_samples_per_second": 6.537,
      "eval_steps_per_second": 6.537,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.08475390076637268,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.002,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.046780169010162354,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0034,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.37187516689300537,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0029,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.22455869615077972,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0036,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.1463429480791092,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0044,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.0570867583155632,
      "eval_runtime": 176.453,
      "eval_samples_per_second": 6.551,
      "eval_steps_per_second": 6.551,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.4433397054672241,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0045,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.1657050997018814,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0031,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.11662856489419937,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0026,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.02125265821814537,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0035,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.05928665027022362,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0027,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.059105079621076584,
      "eval_runtime": 174.6544,
      "eval_samples_per_second": 6.619,
      "eval_steps_per_second": 6.619,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.26842960715293884,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0024,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.06822570413351059,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0018,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.06828371435403824,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0095,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.03534894064068794,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0033,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.36536815762519836,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0044,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.059850871562957764,
      "eval_runtime": 175.0039,
      "eval_samples_per_second": 6.606,
      "eval_steps_per_second": 6.606,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.12761013209819794,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0113,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.09695135056972504,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0043,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.29140061140060425,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0029,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.38758528232574463,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0065,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.011333267204463482,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0017,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.05833715572953224,
      "eval_runtime": 174.9986,
      "eval_samples_per_second": 6.606,
      "eval_steps_per_second": 6.606,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.10328371077775955,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0027,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.07473993301391602,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0074,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.0183301642537117,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0054,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.2049105167388916,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0055,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.04020959138870239,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0034,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.05813543125987053,
      "eval_runtime": 174.7446,
      "eval_samples_per_second": 6.615,
      "eval_steps_per_second": 6.615,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.20188899338245392,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0026,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.2812189757823944,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0069,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.8022979497909546,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0059,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.02248234860599041,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0091,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.00878127757459879,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0048,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.05782487243413925,
      "eval_runtime": 174.7306,
      "eval_samples_per_second": 6.616,
      "eval_steps_per_second": 6.616,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.05645116791129112,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0032,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.3131750822067261,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0033,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.44069525599479675,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0107,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.14901208877563477,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0038,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.11920231580734253,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0051,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.0577433742582798,
      "eval_runtime": 174.4067,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.17919094860553741,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0095,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.10345309972763062,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.002,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.03707447648048401,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.006,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.24922269582748413,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0026,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.47316935658454895,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0053,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.057770997285842896,
      "eval_runtime": 174.5817,
      "eval_samples_per_second": 6.622,
      "eval_steps_per_second": 6.622,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.2129555642604828,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0058,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.31666478514671326,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0041,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.07376430183649063,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0024,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.031059809029102325,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0053,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.037241578102111816,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0033,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.057740841060876846,
      "eval_runtime": 174.3938,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 6.629,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.2316117137670517,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0061,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.10855795443058014,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0039,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.48546239733695984,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0058,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.07362986356019974,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0044,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 8.1511813636463e+17,
      "train_loss": 0.04558033469306715,
      "train_runtime": 40942.202,
      "train_samples_per_second": 1.27,
      "train_steps_per_second": 0.079
    }
  ],
  "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": 8.1511813636463e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}