{
  "best_metric": 0.03418846055865288,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold1/checkpoint-1300",
  "epoch": 4.994228549442093,
  "eval_steps": 50,
  "global_step": 3245,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015390534821085032,
      "grad_norm": 5.894165992736816,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.412,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 6.348907470703125,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.361,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 4.138282775878906,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.1151,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 2.88796067237854,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.7911,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 1.7659647464752197,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.4401,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.316586971282959,
      "eval_runtime": 156.6478,
      "eval_samples_per_second": 7.38,
      "eval_steps_per_second": 7.38,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 1.5039206743240356,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.2682,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 0.9816625714302063,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.2,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 0.8088498115539551,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1568,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 1.8118078708648682,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.167,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.810931921005249,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.1222,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.08822520077228546,
      "eval_runtime": 157.2324,
      "eval_samples_per_second": 7.352,
      "eval_steps_per_second": 7.352,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 1.2906358242034912,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0838,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.725613534450531,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0937,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.1534507274627686,
      "learning_rate": 4e-05,
      "loss": 0.0797,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 1.34184730052948,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.091,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.9567771553993225,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0887,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.06686875224113464,
      "eval_runtime": 157.5122,
      "eval_samples_per_second": 7.339,
      "eval_steps_per_second": 7.339,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 1.1859939098358154,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0833,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 0.8370753526687622,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0848,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.9880083203315735,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.061,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.7187602519989014,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0522,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.8273035883903503,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0749,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.057053424417972565,
      "eval_runtime": 157.5757,
      "eval_samples_per_second": 7.336,
      "eval_steps_per_second": 7.336,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.7401017546653748,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0588,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.8715997934341431,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0619,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.704896867275238,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0463,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.9513587355613708,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0572,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.4892939329147339,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0607,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.05361827090382576,
      "eval_runtime": 157.6666,
      "eval_samples_per_second": 7.332,
      "eval_steps_per_second": 7.332,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.5411283373832703,
      "learning_rate": 8e-05,
      "loss": 0.0566,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.7738116979598999,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0572,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.5578853487968445,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0563,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.3322557508945465,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.062,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.9492285251617432,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0595,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.05384283512830734,
      "eval_runtime": 157.7542,
      "eval_samples_per_second": 7.328,
      "eval_steps_per_second": 7.328,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.4770812690258026,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0574,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.6283535957336426,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0569,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.44413498044013977,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0557,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.4896267354488373,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0556,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.5086399912834167,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0569,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04792175814509392,
      "eval_runtime": 157.8864,
      "eval_samples_per_second": 7.322,
      "eval_steps_per_second": 7.322,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.46917974948883057,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.067,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.546509325504303,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.053,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.4638884663581848,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0445,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.4855862557888031,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0414,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.6737373471260071,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0692,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.05169215798377991,
      "eval_runtime": 157.7323,
      "eval_samples_per_second": 7.329,
      "eval_steps_per_second": 7.329,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.34504202008247375,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0445,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.735038161277771,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0497,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.8294709324836731,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0547,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.4231754243373871,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0713,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.4925121068954468,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0384,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.04761970415711403,
      "eval_runtime": 157.7716,
      "eval_samples_per_second": 7.327,
      "eval_steps_per_second": 7.327,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.3430332839488983,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0453,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.19870778918266296,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0384,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.661488950252533,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0428,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.24697484076023102,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0505,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.3326261639595032,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0339,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.04450162500143051,
      "eval_runtime": 158.0208,
      "eval_samples_per_second": 7.315,
      "eval_steps_per_second": 7.315,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.46648862957954407,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0328,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.3364703059196472,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0587,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.3591529428958893,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0694,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.4767170548439026,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0492,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.26863181591033936,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0515,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.04122591018676758,
      "eval_runtime": 157.8174,
      "eval_samples_per_second": 7.325,
      "eval_steps_per_second": 7.325,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.33508747816085815,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0449,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.31792694330215454,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0327,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.3093491196632385,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0431,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.2710738480091095,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0599,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.4037732183933258,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0423,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.04028555378317833,
      "eval_runtime": 158.1211,
      "eval_samples_per_second": 7.311,
      "eval_steps_per_second": 7.311,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.27133461833000183,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0397,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.3921193778514862,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0479,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.6487473845481873,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0491,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.3198828101158142,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0371,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.38300368189811707,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.031,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.04116838425397873,
      "eval_runtime": 158.0818,
      "eval_samples_per_second": 7.313,
      "eval_steps_per_second": 7.313,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.494517058134079,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0379,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.29125815629959106,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0372,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.2907956838607788,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0367,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.5117995142936707,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0512,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.3068409264087677,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0415,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.040468961000442505,
      "eval_runtime": 158.1428,
      "eval_samples_per_second": 7.31,
      "eval_steps_per_second": 7.31,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.27428820729255676,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0302,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.30620571970939636,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0338,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.3344297409057617,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0346,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.5765268206596375,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0495,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.20199866592884064,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0308,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.038386326283216476,
      "eval_runtime": 158.1395,
      "eval_samples_per_second": 7.31,
      "eval_steps_per_second": 7.31,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.21272775530815125,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0415,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.29667800664901733,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0407,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.3070436120033264,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0381,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.3689572811126709,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0349,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.2321677803993225,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0306,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.03756831958889961,
      "eval_runtime": 158.1521,
      "eval_samples_per_second": 7.309,
      "eval_steps_per_second": 7.309,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.14289599657058716,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0354,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.34113964438438416,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0312,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.32071739435195923,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0348,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.7445811629295349,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0283,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.36859065294265747,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0376,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.03910450264811516,
      "eval_runtime": 158.1567,
      "eval_samples_per_second": 7.309,
      "eval_steps_per_second": 7.309,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.2233244776725769,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0359,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.3060639202594757,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0317,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.14547188580036163,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0257,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.21049383282661438,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0411,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.40515780448913574,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.024,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.038377635180950165,
      "eval_runtime": 158.1377,
      "eval_samples_per_second": 7.31,
      "eval_steps_per_second": 7.31,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.32427331805229187,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0327,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.31314197182655334,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0432,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.18320870399475098,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0315,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.4675890803337097,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0345,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.09710101783275604,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0406,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03921637311577797,
      "eval_runtime": 158.2264,
      "eval_samples_per_second": 7.306,
      "eval_steps_per_second": 7.306,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.2644425332546234,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0304,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.14567822217941284,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0262,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.4217154383659363,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0346,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.29085230827331543,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0401,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.3565951883792877,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0398,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.037606630474328995,
      "eval_runtime": 158.3197,
      "eval_samples_per_second": 7.302,
      "eval_steps_per_second": 7.302,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.23342475295066833,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0385,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.29356491565704346,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0387,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.22889961302280426,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0323,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.21156692504882812,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0349,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.27868443727493286,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0325,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03563035652041435,
      "eval_runtime": 158.4284,
      "eval_samples_per_second": 7.297,
      "eval_steps_per_second": 7.297,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.13841097056865692,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0271,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.17964451014995575,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0363,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.161866694688797,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0244,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.38038918375968933,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0276,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.19593781232833862,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0396,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.03936422988772392,
      "eval_runtime": 158.4145,
      "eval_samples_per_second": 7.297,
      "eval_steps_per_second": 7.297,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.4134126305580139,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0387,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.21383151412010193,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0301,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.29743513464927673,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0399,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.18911777436733246,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0509,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.2681870460510254,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0266,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.03880546987056732,
      "eval_runtime": 158.4161,
      "eval_samples_per_second": 7.297,
      "eval_steps_per_second": 7.297,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.20182037353515625,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0345,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.20083127915859222,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0267,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.2567434012889862,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0305,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.41531264781951904,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0386,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.11381297558546066,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.023,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.039177533239126205,
      "eval_runtime": 158.4521,
      "eval_samples_per_second": 7.296,
      "eval_steps_per_second": 7.296,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.30389973521232605,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.034,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.12882469594478607,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0287,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.13934579491615295,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0341,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.2897407114505768,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0404,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.3756007254123688,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0315,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.03887965902686119,
      "eval_runtime": 158.4722,
      "eval_samples_per_second": 7.295,
      "eval_steps_per_second": 7.295,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.3847171366214752,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0351,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.28197699785232544,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0349,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.20573636889457703,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0242,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.8939650654792786,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0377,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.08158174902200699,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0238,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.03418846055865288,
      "eval_runtime": 158.3336,
      "eval_samples_per_second": 7.301,
      "eval_steps_per_second": 7.301,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.18735063076019287,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0221,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.3361535966396332,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0221,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.26995670795440674,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0223,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.21700310707092285,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0224,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.2816634178161621,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0189,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.036083854734897614,
      "eval_runtime": 158.5099,
      "eval_samples_per_second": 7.293,
      "eval_steps_per_second": 7.293,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.10387425124645233,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0218,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.6028366088867188,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0178,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.5149375796318054,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0138,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.20839239656925201,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0182,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.37215471267700195,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0293,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.03665517643094063,
      "eval_runtime": 158.5363,
      "eval_samples_per_second": 7.292,
      "eval_steps_per_second": 7.292,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.2517184317111969,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0237,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.31830158829689026,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0243,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.1488017588853836,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0126,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.21085503697395325,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0217,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.13404421508312225,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0128,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.04199405014514923,
      "eval_runtime": 158.5079,
      "eval_samples_per_second": 7.293,
      "eval_steps_per_second": 7.293,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.20932886004447937,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0215,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.12694035470485687,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0265,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.22968275845050812,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0194,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.22519809007644653,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0243,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.35927438735961914,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0195,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.038543544709682465,
      "eval_runtime": 158.556,
      "eval_samples_per_second": 7.291,
      "eval_steps_per_second": 7.291,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.3260902166366577,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0219,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.17589454352855682,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0249,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.1324448436498642,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0185,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.33059003949165344,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0202,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.2055257111787796,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0174,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.03829577937722206,
      "eval_runtime": 158.5448,
      "eval_samples_per_second": 7.291,
      "eval_steps_per_second": 7.291,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.2823956608772278,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0136,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.22851237654685974,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0212,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.3365919291973114,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0225,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.20184405148029327,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0195,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.3508014976978302,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0143,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.04149091616272926,
      "eval_runtime": 158.7251,
      "eval_samples_per_second": 7.283,
      "eval_steps_per_second": 7.283,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.23150603473186493,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0152,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.25979942083358765,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0245,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.43624642491340637,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0192,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.40092676877975464,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0214,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.22205431759357452,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0249,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.04037570580840111,
      "eval_runtime": 158.6755,
      "eval_samples_per_second": 7.285,
      "eval_steps_per_second": 7.285,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.25793302059173584,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0171,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.5910273194313049,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0233,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.29524514079093933,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0259,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.19323395192623138,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0208,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.4049088954925537,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0195,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03831124305725098,
      "eval_runtime": 158.7249,
      "eval_samples_per_second": 7.283,
      "eval_steps_per_second": 7.283,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.27759331464767456,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0189,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.24512633681297302,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0305,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.3918320834636688,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0224,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.25709059834480286,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0171,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.3009372651576996,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0266,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.037615735083818436,
      "eval_runtime": 158.6899,
      "eval_samples_per_second": 7.285,
      "eval_steps_per_second": 7.285,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.27946242690086365,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.015,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.7809072136878967,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.023,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.476836621761322,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0269,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.28281018137931824,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0199,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.3204270005226135,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0216,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03652812913060188,
      "eval_runtime": 158.7438,
      "eval_samples_per_second": 7.282,
      "eval_steps_per_second": 7.282,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.31815871596336365,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0204,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.3212921917438507,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0256,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.4266219139099121,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0215,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.2512921988964081,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0238,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.1962926983833313,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0236,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.03656553104519844,
      "eval_runtime": 158.3077,
      "eval_samples_per_second": 7.302,
      "eval_steps_per_second": 7.302,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.3850853741168976,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0173,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.3224900960922241,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0281,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.4167907238006592,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0121,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.26658695936203003,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0173,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.22370237112045288,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0198,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.036889951676130295,
      "eval_runtime": 158.3162,
      "eval_samples_per_second": 7.302,
      "eval_steps_per_second": 7.302,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.19981063902378082,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0219,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.4523862302303314,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0172,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.06757289171218872,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0148,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.2735544741153717,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0174,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.20762528479099274,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0311,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.03702723979949951,
      "eval_runtime": 158.5727,
      "eval_samples_per_second": 7.29,
      "eval_steps_per_second": 7.29,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.12028948962688446,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.007,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.19118699431419373,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0087,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.1755441278219223,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0099,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.18532782793045044,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0058,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.15551483631134033,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0088,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.04241878166794777,
      "eval_runtime": 158.6023,
      "eval_samples_per_second": 7.289,
      "eval_steps_per_second": 7.289,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 1.033755898475647,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0072,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.22541406750679016,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0093,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.18234965205192566,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0111,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.03136371448636055,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0088,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.5648914575576782,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0126,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.046655021607875824,
      "eval_runtime": 158.5464,
      "eval_samples_per_second": 7.291,
      "eval_steps_per_second": 7.291,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.3313479423522949,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0061,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.4383080303668976,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0094,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.2923309803009033,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0115,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.31625160574913025,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0063,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.4579697847366333,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0085,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.0463048480451107,
      "eval_runtime": 158.623,
      "eval_samples_per_second": 7.288,
      "eval_steps_per_second": 7.288,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.33424171805381775,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0068,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.39984947443008423,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0104,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.4942300319671631,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0085,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.13812139630317688,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0061,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.4164913296699524,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0083,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.045313119888305664,
      "eval_runtime": 158.5556,
      "eval_samples_per_second": 7.291,
      "eval_steps_per_second": 7.291,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.061174582690000534,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0068,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.29132723808288574,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0112,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.2071104645729065,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0085,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.7489820718765259,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0085,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.18367169797420502,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0164,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.046990569680929184,
      "eval_runtime": 158.4962,
      "eval_samples_per_second": 7.294,
      "eval_steps_per_second": 7.294,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.15213659405708313,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0101,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.36414727568626404,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0078,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.3981834650039673,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0137,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.3077724874019623,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0097,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.14465726912021637,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0115,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.04648184776306152,
      "eval_runtime": 158.5192,
      "eval_samples_per_second": 7.292,
      "eval_steps_per_second": 7.292,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.030671967193484306,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0051,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.28764620423316956,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0048,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.4497350752353668,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0093,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.19971530139446259,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0068,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.0892835259437561,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0134,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.046868808567523956,
      "eval_runtime": 158.3898,
      "eval_samples_per_second": 7.298,
      "eval_steps_per_second": 7.298,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.10751787573099136,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0078,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.4460418224334717,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0135,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.21748164296150208,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0071,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.28973084688186646,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.011,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.030565926805138588,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0052,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.04699597507715225,
      "eval_runtime": 158.3644,
      "eval_samples_per_second": 7.3,
      "eval_steps_per_second": 7.3,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.3756799101829529,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0136,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.3436908721923828,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0084,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.26337122917175293,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0135,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.4950447082519531,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0102,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.028451379388570786,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0104,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.044791873544454575,
      "eval_runtime": 158.5238,
      "eval_samples_per_second": 7.292,
      "eval_steps_per_second": 7.292,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.10130444914102554,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0085,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.2004706859588623,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0069,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.14130496978759766,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0093,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 1.0703330039978027,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0069,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.6593865156173706,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0075,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.04586057737469673,
      "eval_runtime": 158.2726,
      "eval_samples_per_second": 7.304,
      "eval_steps_per_second": 7.304,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.027158798649907112,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0067,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.2729403078556061,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0048,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.32945534586906433,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0098,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.14923302829265594,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0091,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.29195690155029297,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0068,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.04852357506752014,
      "eval_runtime": 157.7578,
      "eval_samples_per_second": 7.328,
      "eval_steps_per_second": 7.328,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.021099422127008438,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.009,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.4982509911060333,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0036,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.020435772836208344,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0071,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.0502375103533268,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0034,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.31526339054107666,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0089,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.0494471974670887,
      "eval_runtime": 157.2684,
      "eval_samples_per_second": 7.35,
      "eval_steps_per_second": 7.35,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.30607834458351135,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0064,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.6153056621551514,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0082,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.02956816367805004,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0109,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.34335827827453613,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0075,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.06555047631263733,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0091,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.04757598415017128,
      "eval_runtime": 156.7228,
      "eval_samples_per_second": 7.376,
      "eval_steps_per_second": 7.376,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.10410808771848679,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0046,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.179936021566391,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0017,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.14201240241527557,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.002,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.043476756662130356,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0034,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.06336613744497299,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0021,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.04980561137199402,
      "eval_runtime": 156.3369,
      "eval_samples_per_second": 7.394,
      "eval_steps_per_second": 7.394,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.29756295680999756,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0059,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.49036499857902527,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0025,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.15332582592964172,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0023,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.05230104550719261,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0033,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.13933300971984863,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0061,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.05291583761572838,
      "eval_runtime": 156.0003,
      "eval_samples_per_second": 7.41,
      "eval_steps_per_second": 7.41,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.05018053576350212,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0027,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.0987987071275711,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0064,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.11932341754436493,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0024,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.19836074113845825,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0025,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.03341757133603096,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0011,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.054074157029390335,
      "eval_runtime": 155.7811,
      "eval_samples_per_second": 7.421,
      "eval_steps_per_second": 7.421,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.039698828011751175,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0019,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.018008965998888016,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0024,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.0988951176404953,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0026,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.028782740235328674,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0026,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.046109676361083984,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0025,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.05490599945187569,
      "eval_runtime": 155.6615,
      "eval_samples_per_second": 7.426,
      "eval_steps_per_second": 7.426,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.07521144300699234,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0012,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.022351477295160294,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.003,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.09315711259841919,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0023,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.03538105636835098,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0031,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.015765981748700142,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0029,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.055953964591026306,
      "eval_runtime": 155.4436,
      "eval_samples_per_second": 7.437,
      "eval_steps_per_second": 7.437,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.32189977169036865,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0016,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.044547632336616516,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0014,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.025402750819921494,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0052,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.016883034259080887,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0016,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.2613152861595154,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0027,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.05703258886933327,
      "eval_runtime": 155.4074,
      "eval_samples_per_second": 7.439,
      "eval_steps_per_second": 7.439,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.056421246379613876,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0056,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.03421706706285477,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0034,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.042870502918958664,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0023,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.3700352907180786,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0043,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.015034668147563934,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0017,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.05716672167181969,
      "eval_runtime": 155.3485,
      "eval_samples_per_second": 7.441,
      "eval_steps_per_second": 7.441,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.15378257632255554,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0016,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 1.56369948387146,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0065,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.01959718018770218,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0027,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.22856013476848602,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0033,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.020138591527938843,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0019,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.05735043063759804,
      "eval_runtime": 155.2844,
      "eval_samples_per_second": 7.444,
      "eval_steps_per_second": 7.444,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.08202099800109863,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0021,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.2669055461883545,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0035,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.127384215593338,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0027,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.0757393091917038,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0025,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.014194714836776257,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.005,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.057465746998786926,
      "eval_runtime": 155.2164,
      "eval_samples_per_second": 7.448,
      "eval_steps_per_second": 7.448,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.04634212702512741,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0047,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.478548139333725,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0021,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.6174399256706238,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.008,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.1303272694349289,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0018,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.06364283710718155,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0033,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.057286251336336136,
      "eval_runtime": 155.2078,
      "eval_samples_per_second": 7.448,
      "eval_steps_per_second": 7.448,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.06751379370689392,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.004,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.02918647788465023,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0015,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.01717446744441986,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0032,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.07885854691267014,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0016,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.32916271686553955,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.005,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.05758997052907944,
      "eval_runtime": 155.1613,
      "eval_samples_per_second": 7.45,
      "eval_steps_per_second": 7.45,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.05061342194676399,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0031,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.2501945495605469,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0022,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.03239763155579567,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0012,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.0456206314265728,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0025,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.027485232800245285,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0019,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.0574953556060791,
      "eval_runtime": 155.0472,
      "eval_samples_per_second": 7.456,
      "eval_steps_per_second": 7.456,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.13778692483901978,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.004,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.147592693567276,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0027,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.19019150733947754,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0034,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.10471858084201813,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0015,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 8.1511813636463e+17,
      "train_loss": 0.041532339387284865,
      "train_runtime": 35427.3978,
      "train_samples_per_second": 1.467,
      "train_steps_per_second": 0.092
    }
  ],
  "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
}