{
  "best_metric": 0.0314556248486042,
  "best_model_checkpoint": "saves/psy-course/Ministral-8B-Instruct-2410/train/fold9/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": 2.718892812728882,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.3361,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 2.797744035720825,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.2708,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 3.1749162673950195,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.0345,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 2.2526276111602783,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.5541,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 1.1467052698135376,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.2581,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.24156680703163147,
      "eval_runtime": 175.3377,
      "eval_samples_per_second": 6.593,
      "eval_steps_per_second": 6.593,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 1.4773415327072144,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.1595,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.1879477500915527,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.1271,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 1.2361923456192017,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1017,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 1.4640657901763916,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.1183,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.7191898226737976,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0851,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.06925127655267715,
      "eval_runtime": 175.2942,
      "eval_samples_per_second": 6.595,
      "eval_steps_per_second": 6.595,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8226053714752197,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0643,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.8247290849685669,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0722,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.1203566789627075,
      "learning_rate": 4e-05,
      "loss": 0.0691,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.7748081088066101,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.079,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.909217119216919,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0612,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.058652956038713455,
      "eval_runtime": 175.2735,
      "eval_samples_per_second": 6.595,
      "eval_steps_per_second": 6.595,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.7968839406967163,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0614,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 0.7895393967628479,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0617,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.8020516633987427,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0526,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.35680174827575684,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0446,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5127408504486084,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0576,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.05370116978883743,
      "eval_runtime": 175.3605,
      "eval_samples_per_second": 6.592,
      "eval_steps_per_second": 6.592,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.740752100944519,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0558,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.8217968344688416,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0502,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.35808926820755005,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0383,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.4695506989955902,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0458,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.4380370080471039,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0438,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.043141331523656845,
      "eval_runtime": 175.4636,
      "eval_samples_per_second": 6.588,
      "eval_steps_per_second": 6.588,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.5864039659500122,
      "learning_rate": 8e-05,
      "loss": 0.0396,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.4064479470252991,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0419,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.5284066200256348,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.054,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.4555165469646454,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0525,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.39548638463020325,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0407,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04591534659266472,
      "eval_runtime": 175.31,
      "eval_samples_per_second": 6.594,
      "eval_steps_per_second": 6.594,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.486291766166687,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0513,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.3238089978694916,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0481,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.4284800887107849,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0482,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.4129464626312256,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0461,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.4707987904548645,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0433,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04509405791759491,
      "eval_runtime": 175.3884,
      "eval_samples_per_second": 6.591,
      "eval_steps_per_second": 6.591,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.3643520176410675,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0566,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.4068044126033783,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0371,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.6159802079200745,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.039,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.47319525480270386,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0347,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.26943686604499817,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0473,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.04295285418629646,
      "eval_runtime": 175.728,
      "eval_samples_per_second": 6.578,
      "eval_steps_per_second": 6.578,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.2973695993423462,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0408,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.4833831191062927,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0453,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.43758249282836914,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0444,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.47562554478645325,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0595,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.33092936873435974,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0303,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.03931599110364914,
      "eval_runtime": 175.7218,
      "eval_samples_per_second": 6.579,
      "eval_steps_per_second": 6.579,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.2749386429786682,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0358,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.3742888867855072,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.031,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.5937685966491699,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0405,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.23685641586780548,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0464,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.3290925621986389,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0293,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.038906279951334,
      "eval_runtime": 175.7197,
      "eval_samples_per_second": 6.579,
      "eval_steps_per_second": 6.579,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.4313240051269531,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0311,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.31886816024780273,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0419,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.4023323059082031,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0517,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.35093745589256287,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0428,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.2715444266796112,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0421,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.03571254014968872,
      "eval_runtime": 175.5787,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.26416295766830444,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.036,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.2766662836074829,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0256,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.30113324522972107,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0348,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.2379077523946762,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0504,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.33010974526405334,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0347,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.035797443240880966,
      "eval_runtime": 175.4673,
      "eval_samples_per_second": 6.588,
      "eval_steps_per_second": 6.588,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.4186369776725769,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0321,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.3265047073364258,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0434,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.44338884949684143,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0427,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.28677836060523987,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0325,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.25564688444137573,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0305,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03563925623893738,
      "eval_runtime": 175.7934,
      "eval_samples_per_second": 6.576,
      "eval_steps_per_second": 6.576,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.34283924102783203,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0345,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.2367202490568161,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0318,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.2096981406211853,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0319,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.4727417826652527,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0437,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.4262826442718506,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0338,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.03800124675035477,
      "eval_runtime": 175.8475,
      "eval_samples_per_second": 6.574,
      "eval_steps_per_second": 6.574,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.36701545119285583,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0304,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.23882395029067993,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0301,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.2359193116426468,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0278,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.580471932888031,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0367,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.3765440285205841,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0278,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.03505924716591835,
      "eval_runtime": 175.6514,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 6.581,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.32123416662216187,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0364,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.20309022068977356,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0351,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.3137691915035248,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0355,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.3466825485229492,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0338,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.4528277516365051,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0283,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.03416485711932182,
      "eval_runtime": 175.662,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 6.581,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.1724155694246292,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0299,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.4145107865333557,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0287,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.28502559661865234,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.032,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.43659815192222595,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0274,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.3568861484527588,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0276,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.035158488899469376,
      "eval_runtime": 175.6336,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.27670103311538696,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.026,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.4315589964389801,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0261,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.1875932365655899,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0288,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.2877761721611023,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0369,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.4872714579105377,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0218,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.03432812914252281,
      "eval_runtime": 175.5175,
      "eval_samples_per_second": 6.586,
      "eval_steps_per_second": 6.586,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.3045959770679474,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0301,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.21601802110671997,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.031,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.20014813542366028,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0299,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.5667252540588379,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0294,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.1364864706993103,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0366,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.034413401037454605,
      "eval_runtime": 175.6256,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.25798943638801575,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0286,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.09673745930194855,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0231,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.30414289236068726,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0311,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.234617680311203,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0309,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.2941848039627075,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0325,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.035791654139757156,
      "eval_runtime": 175.5077,
      "eval_samples_per_second": 6.587,
      "eval_steps_per_second": 6.587,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.35165417194366455,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0322,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.30164629220962524,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0308,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.24873089790344238,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0292,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.3822762966156006,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0262,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.2577270567417145,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0308,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03172922134399414,
      "eval_runtime": 175.5376,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.20669153332710266,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0224,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.2581917643547058,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0272,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.38528966903686523,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0235,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.80202317237854,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0281,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.230235755443573,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0327,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.03630954772233963,
      "eval_runtime": 175.4912,
      "eval_samples_per_second": 6.587,
      "eval_steps_per_second": 6.587,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.4021379351615906,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0292,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.1810339093208313,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0248,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.4089711904525757,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.034,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.2589546740055084,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0384,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.1923990696668625,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0231,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.033109817653894424,
      "eval_runtime": 175.469,
      "eval_samples_per_second": 6.588,
      "eval_steps_per_second": 6.588,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.14907753467559814,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0339,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.26013439893722534,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0249,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.26201820373535156,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0263,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.2704382538795471,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0291,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.12639008462429047,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0195,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.03798627480864525,
      "eval_runtime": 175.4465,
      "eval_samples_per_second": 6.589,
      "eval_steps_per_second": 6.589,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.42108216881752014,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0292,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.1684366911649704,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.024,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.16571079194545746,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0297,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.32478490471839905,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0348,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.4178849756717682,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.028,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.032729510217905045,
      "eval_runtime": 175.585,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.739521861076355,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0295,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.189444437623024,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0293,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.23456431925296783,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0231,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.40063855051994324,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0313,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.13206598162651062,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.021,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.032091300934553146,
      "eval_runtime": 175.5373,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.43381285667419434,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0166,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.49853095412254333,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0184,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.21651127934455872,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.019,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.08318690210580826,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0208,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.1352061927318573,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0169,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.0314556248486042,
      "eval_runtime": 175.5774,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.16456666588783264,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0182,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.23588073253631592,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0167,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.2566874921321869,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0122,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.1678154468536377,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0149,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 1.1714719533920288,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0226,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.03427564352750778,
      "eval_runtime": 175.5992,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.07414636015892029,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0159,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.972678005695343,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0239,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.1999577432870865,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0129,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.27154386043548584,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0203,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.08592633903026581,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.01,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.03968590125441551,
      "eval_runtime": 175.5875,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.46208322048187256,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0175,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.13938766717910767,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0219,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.4152555465698242,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0195,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.1447705775499344,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0202,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.6547629237174988,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.019,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.0353720560669899,
      "eval_runtime": 175.5786,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.24090097844600677,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0217,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.7100735306739807,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0188,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.13864926993846893,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0144,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.2342522293329239,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0184,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.4761083126068115,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0165,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.03859942406415939,
      "eval_runtime": 175.5908,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.2521819770336151,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0134,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.3120094835758209,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0176,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.15116415917873383,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0182,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.2375267893075943,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0179,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.25881603360176086,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0112,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.03789198771119118,
      "eval_runtime": 175.6307,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.3217921257019043,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0152,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.22059328854084015,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.021,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.18697978556156158,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0161,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.40190941095352173,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0226,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.21625593304634094,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0194,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.034770671278238297,
      "eval_runtime": 175.6016,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.2197883278131485,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0125,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.3602195382118225,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0174,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.4390200078487396,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.018,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.13819743692874908,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.019,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.42978397011756897,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0221,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.0364517942070961,
      "eval_runtime": 175.5529,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.29810720682144165,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0168,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.5205846428871155,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0259,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.4699023365974426,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0181,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.1394762098789215,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0134,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.06343324482440948,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0231,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.034239545464515686,
      "eval_runtime": 175.6741,
      "eval_samples_per_second": 6.58,
      "eval_steps_per_second": 6.58,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.28442394733428955,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0132,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.38578760623931885,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0183,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.6168205738067627,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0231,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.15851372480392456,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0178,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.2747803330421448,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0178,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03372257947921753,
      "eval_runtime": 175.6673,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 6.581,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.4894351065158844,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0197,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.3423890173435211,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0207,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.5567525625228882,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0218,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.29117467999458313,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.019,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.16067267954349518,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0201,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.032902106642723083,
      "eval_runtime": 175.5844,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.30854541063308716,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0132,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.279921293258667,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0239,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.2846156060695648,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0148,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.14873768389225006,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0152,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.18597367405891418,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0178,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.03398587554693222,
      "eval_runtime": 175.506,
      "eval_samples_per_second": 6.587,
      "eval_steps_per_second": 6.587,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 1.0201305150985718,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0222,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.43314221501350403,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0186,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.14487038552761078,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0162,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.3032515048980713,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0154,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.1764432191848755,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.025,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.03440512344241142,
      "eval_runtime": 175.6024,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.11448675394058228,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0073,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.0834617018699646,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0095,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.17162995040416718,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0087,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.2549445331096649,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0069,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.13785775005817413,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.009,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.037305619567632675,
      "eval_runtime": 175.5945,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 0.40954622626304626,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0052,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.1325177401304245,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0104,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.1672409623861313,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0092,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.2046295404434204,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0084,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 1.0192298889160156,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0118,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.04204763099551201,
      "eval_runtime": 175.6541,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 6.581,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.038153309375047684,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0056,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.5369210243225098,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0088,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.23393917083740234,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0079,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.4740985631942749,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0084,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.18593557178974152,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0091,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.04006849229335785,
      "eval_runtime": 175.5374,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.04574641212821007,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0081,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.3719145655632019,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0087,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.293404221534729,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0086,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.8891112804412842,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0074,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.5398598909378052,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0076,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.03983517363667488,
      "eval_runtime": 175.6698,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 6.581,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.2526977062225342,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0094,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.2633676528930664,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0101,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.26152530312538147,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0087,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.23602557182312012,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0054,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.06726223230361938,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0127,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.03852129355072975,
      "eval_runtime": 175.6598,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 6.581,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.2062523365020752,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0081,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.25941672921180725,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0074,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.2114010900259018,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0117,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.15474379062652588,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0115,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.14131970703601837,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0088,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.037573616951704025,
      "eval_runtime": 175.5563,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.04161751642823219,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0066,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.2391417920589447,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0048,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.1056637391448021,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0081,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.21117965877056122,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0081,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.18350252509117126,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0096,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.04031885415315628,
      "eval_runtime": 175.6664,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 6.581,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.06849458068609238,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0106,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.46342042088508606,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0086,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.1368333399295807,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0073,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.1783951073884964,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0129,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.2780916392803192,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0065,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.04053264483809471,
      "eval_runtime": 175.7045,
      "eval_samples_per_second": 6.579,
      "eval_steps_per_second": 6.579,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.32572951912879944,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0151,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.22675101459026337,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0081,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.15402339398860931,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0095,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.22837091982364655,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0099,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.275728702545166,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0111,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.0375065952539444,
      "eval_runtime": 175.7441,
      "eval_samples_per_second": 6.578,
      "eval_steps_per_second": 6.578,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.2737589180469513,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0103,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.06734083592891693,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0067,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.04964146763086319,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0088,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.06971968710422516,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0051,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.047497596591711044,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0069,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.039597149938344955,
      "eval_runtime": 175.5545,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.2800331711769104,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0096,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.4802244007587433,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0046,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.11647926270961761,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0109,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.17670533061027527,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.008,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.19920852780342102,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0084,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.041493915021419525,
      "eval_runtime": 175.6211,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.027272000908851624,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0083,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.2026788294315338,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0041,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.02968156337738037,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.006,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.10857419669628143,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0041,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.5583007335662842,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0087,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.04227972775697708,
      "eval_runtime": 175.6729,
      "eval_samples_per_second": 6.58,
      "eval_steps_per_second": 6.58,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.5067748427391052,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0071,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.29275739192962646,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0058,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.08718723058700562,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0121,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.33604612946510315,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0073,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.023664474487304688,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0092,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.04294198751449585,
      "eval_runtime": 175.5632,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.043392013758420944,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0047,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.13135915994644165,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0025,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.13270501792430878,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0021,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.019360952079296112,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0046,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.10098142921924591,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.002,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.044195566326379776,
      "eval_runtime": 175.5121,
      "eval_samples_per_second": 6.586,
      "eval_steps_per_second": 6.586,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.1618228405714035,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.005,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.06022949516773224,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0028,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.07590459287166595,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0026,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.01247763354331255,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0059,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.05250566825270653,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0075,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.04754495620727539,
      "eval_runtime": 175.5959,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.06010911986231804,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.005,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.08446869254112244,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0085,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.047181662172079086,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0035,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.20420198142528534,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0061,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.0807032510638237,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0018,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.046181511133909225,
      "eval_runtime": 175.5541,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.19284464418888092,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.002,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.23279498517513275,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0032,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.12890943884849548,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0023,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.11439457535743713,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0033,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.06536822021007538,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0034,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.04827659949660301,
      "eval_runtime": 175.497,
      "eval_samples_per_second": 6.587,
      "eval_steps_per_second": 6.587,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.031200280413031578,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0026,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.022764835506677628,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0016,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.04334717616438866,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0028,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.023718567565083504,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.003,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.029488736763596535,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0028,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.04923668131232262,
      "eval_runtime": 175.6392,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.2583252787590027,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0025,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.010081259533762932,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0009,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.025453371927142143,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.006,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.01132876891642809,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0029,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.2927170693874359,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.004,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04996582493185997,
      "eval_runtime": 175.5714,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.10230018198490143,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0074,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.1362355500459671,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0041,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.01672428473830223,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0031,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.34220805764198303,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0052,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.013602997176349163,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0018,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.04955550283193588,
      "eval_runtime": 175.5908,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.22792832553386688,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0013,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.018993359059095383,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.006,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.11517757922410965,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0027,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.0787980780005455,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0019,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.07277962565422058,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0039,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.049806851893663406,
      "eval_runtime": 175.5461,
      "eval_samples_per_second": 6.585,
      "eval_steps_per_second": 6.585,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.239015594124794,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0024,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.21216259896755219,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0034,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.14827050268650055,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0044,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.4856173098087311,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0032,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.005365810822695494,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0035,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.05025773122906685,
      "eval_runtime": 175.6359,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.03526800498366356,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.003,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.2582164406776428,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0027,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.33338820934295654,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0066,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.1538686603307724,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0029,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.072846919298172,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0054,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.05029488727450371,
      "eval_runtime": 175.58,
      "eval_samples_per_second": 6.584,
      "eval_steps_per_second": 6.584,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.07733611017465591,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0078,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.1515783667564392,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0031,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.005200848914682865,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0051,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.028695642948150635,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0021,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.058336369693279266,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0028,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.050209518522024155,
      "eval_runtime": 175.6125,
      "eval_samples_per_second": 6.583,
      "eval_steps_per_second": 6.583,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.055045999586582184,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0029,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.16884483397006989,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0031,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.08949501812458038,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0024,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.012743637897074223,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0046,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.02398603782057762,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0041,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.050147101283073425,
      "eval_runtime": 175.536,
      "eval_samples_per_second": 6.586,
      "eval_steps_per_second": 6.586,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.06706839799880981,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0055,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.03772740811109543,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.002,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.05270882695913315,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0058,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.034575220197439194,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0025,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 8.169046237996646e+17,
      "train_loss": 0.03530819658706616,
      "train_runtime": 38910.5252,
      "train_samples_per_second": 1.336,
      "train_steps_per_second": 0.083
    }
  ],
  "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
}