{
  "best_metric": 0.03155430406332016,
  "best_model_checkpoint": "saves/psy-course/Ministral-8B-Instruct-2410/train/fold8/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": 2.7219290733337402,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.3354,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 2.799449920654297,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.2712,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 3.1708827018737793,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.0351,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 2.2599470615386963,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.555,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 1.1390929222106934,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.2582,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.24157023429870605,
      "eval_runtime": 172.7344,
      "eval_samples_per_second": 6.692,
      "eval_steps_per_second": 6.692,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 1.454789638519287,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.1589,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.219090223312378,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.1275,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 1.2359507083892822,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1014,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 1.4742939472198486,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.1181,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.7404343485832214,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0851,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.06950391829013824,
      "eval_runtime": 172.8509,
      "eval_samples_per_second": 6.688,
      "eval_steps_per_second": 6.688,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8187854290008545,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0649,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.7971552610397339,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0722,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.101304054260254,
      "learning_rate": 4e-05,
      "loss": 0.069,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.7783849239349365,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0789,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.9054766297340393,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.061,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.058502163738012314,
      "eval_runtime": 172.947,
      "eval_samples_per_second": 6.684,
      "eval_steps_per_second": 6.684,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.7719207406044006,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0618,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 0.7689031362533569,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.061,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.8108643889427185,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0527,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.3555704951286316,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0449,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5617899894714355,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0577,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.05433733016252518,
      "eval_runtime": 172.9378,
      "eval_samples_per_second": 6.684,
      "eval_steps_per_second": 6.684,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.7014524340629578,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0552,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.8417723178863525,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0497,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.3444922864437103,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.038,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.4500199854373932,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0462,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.43585526943206787,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0438,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.043287038803100586,
      "eval_runtime": 172.9223,
      "eval_samples_per_second": 6.685,
      "eval_steps_per_second": 6.685,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.5859360098838806,
      "learning_rate": 8e-05,
      "loss": 0.04,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.4108608365058899,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0417,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.530652642250061,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0546,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.45784997940063477,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0518,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.4010988771915436,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0407,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04635393247008324,
      "eval_runtime": 173.0308,
      "eval_samples_per_second": 6.681,
      "eval_steps_per_second": 6.681,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.4716113209724426,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0513,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.32197678089141846,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0479,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.44838961958885193,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0486,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.4036242961883545,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0452,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.44787314534187317,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0425,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04357096925377846,
      "eval_runtime": 173.0126,
      "eval_samples_per_second": 6.682,
      "eval_steps_per_second": 6.682,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.36068758368492126,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0555,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.4382631182670593,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0374,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.5948066711425781,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0388,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.4754520654678345,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0346,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.3159029185771942,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0485,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.04306018725037575,
      "eval_runtime": 173.1033,
      "eval_samples_per_second": 6.678,
      "eval_steps_per_second": 6.678,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.3288145661354065,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0405,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.5652763247489929,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0462,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.47297507524490356,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.044,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.4473099708557129,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0595,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.5017976760864258,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0308,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.03791685402393341,
      "eval_runtime": 173.1099,
      "eval_samples_per_second": 6.678,
      "eval_steps_per_second": 6.678,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.22264744341373444,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0339,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.3821590840816498,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.032,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.4943896234035492,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0424,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.2355869561433792,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0448,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.3663094639778137,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0281,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.0374106764793396,
      "eval_runtime": 173.2126,
      "eval_samples_per_second": 6.674,
      "eval_steps_per_second": 6.674,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.4152411222457886,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0296,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.31085723638534546,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0452,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.3943304419517517,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0532,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.3337588310241699,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0404,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.28240227699279785,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0406,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.03465297073125839,
      "eval_runtime": 173.4506,
      "eval_samples_per_second": 6.665,
      "eval_steps_per_second": 6.665,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.3319605588912964,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0359,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.28146156668663025,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0243,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.36267679929733276,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0335,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.2347225546836853,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0505,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.32077303528785706,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0335,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.034932758659124374,
      "eval_runtime": 173.1882,
      "eval_samples_per_second": 6.675,
      "eval_steps_per_second": 6.675,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.3694322407245636,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0313,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.37305518984794617,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0443,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.49409234523773193,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0423,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.2959464192390442,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0326,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.2547084093093872,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0294,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.036260321736335754,
      "eval_runtime": 173.264,
      "eval_samples_per_second": 6.672,
      "eval_steps_per_second": 6.672,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.3338225781917572,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0356,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.24516162276268005,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0307,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.20163822174072266,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0339,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.432851105928421,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.043,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.3996201455593109,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0334,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.03677118569612503,
      "eval_runtime": 173.2667,
      "eval_samples_per_second": 6.672,
      "eval_steps_per_second": 6.672,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.3657374083995819,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0312,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.233443945646286,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0309,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.2006950080394745,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0277,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.5221588015556335,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0369,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.38970911502838135,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0262,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.036153245717287064,
      "eval_runtime": 173.3109,
      "eval_samples_per_second": 6.67,
      "eval_steps_per_second": 6.67,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.3251585364341736,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0359,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.21200118958950043,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0342,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.31985801458358765,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0348,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.45090609788894653,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0332,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.28127601742744446,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0278,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.034282486885786057,
      "eval_runtime": 173.297,
      "eval_samples_per_second": 6.671,
      "eval_steps_per_second": 6.671,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.17184391617774963,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0292,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.4126950800418854,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0285,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.2745838165283203,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0322,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.4735434055328369,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0273,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.3560656011104584,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0277,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.03436972200870514,
      "eval_runtime": 173.3402,
      "eval_samples_per_second": 6.669,
      "eval_steps_per_second": 6.669,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.2575756311416626,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.026,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.2788808345794678,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0278,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.16632190346717834,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.029,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.2726566791534424,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0365,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.3006250262260437,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0201,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.03368045389652252,
      "eval_runtime": 173.4008,
      "eval_samples_per_second": 6.667,
      "eval_steps_per_second": 6.667,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.2546687126159668,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0301,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.27090808749198914,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0323,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.35158222913742065,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0288,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.522314727306366,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0298,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.16340717673301697,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0357,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03347551450133324,
      "eval_runtime": 173.4334,
      "eval_samples_per_second": 6.665,
      "eval_steps_per_second": 6.665,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.2648465037345886,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0279,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.1359444260597229,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0234,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.4287537932395935,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0289,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.2626200318336487,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0305,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.3781191110610962,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0314,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.034793462604284286,
      "eval_runtime": 173.4438,
      "eval_samples_per_second": 6.665,
      "eval_steps_per_second": 6.665,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.44384050369262695,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0332,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.32738709449768066,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0319,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.22795370221138,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0289,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.40003785490989685,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0278,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.21056413650512695,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0312,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03284203261137009,
      "eval_runtime": 173.4569,
      "eval_samples_per_second": 6.664,
      "eval_steps_per_second": 6.664,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.1555100679397583,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0227,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.2727252244949341,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0272,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 1.1931912899017334,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0235,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.5616550445556641,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0276,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.2251860350370407,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0342,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.035226643085479736,
      "eval_runtime": 173.4613,
      "eval_samples_per_second": 6.664,
      "eval_steps_per_second": 6.664,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.3998422622680664,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0306,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.35687488317489624,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0226,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.3978603780269623,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0342,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.20571114122867584,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0382,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.18063031136989594,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.023,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.032391078770160675,
      "eval_runtime": 173.4395,
      "eval_samples_per_second": 6.665,
      "eval_steps_per_second": 6.665,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.14317816495895386,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0332,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.29168206453323364,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0231,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.2447410225868225,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0242,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.23979124426841736,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0301,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.1291966736316681,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.021,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.036621782928705215,
      "eval_runtime": 173.492,
      "eval_samples_per_second": 6.663,
      "eval_steps_per_second": 6.663,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.4357064962387085,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0284,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.1802254319190979,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0235,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.15108895301818848,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0279,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.36413756012916565,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0335,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.4576139748096466,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0276,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.033047281205654144,
      "eval_runtime": 173.3891,
      "eval_samples_per_second": 6.667,
      "eval_steps_per_second": 6.667,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.6762644648551941,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0302,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.19850490987300873,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0308,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.2305378019809723,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0229,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.20170681178569794,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0332,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.21552927792072296,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.022,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.032010674476623535,
      "eval_runtime": 173.4748,
      "eval_samples_per_second": 6.664,
      "eval_steps_per_second": 6.664,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.3831847906112671,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0168,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.33277618885040283,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0164,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.22611477971076965,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0181,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.07006730139255524,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0201,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.17755873501300812,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.017,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.03155430406332016,
      "eval_runtime": 173.6542,
      "eval_samples_per_second": 6.657,
      "eval_steps_per_second": 6.657,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.20978456735610962,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0185,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.2573811411857605,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0152,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.18994827568531036,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0128,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.0983426570892334,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0138,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.411379337310791,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0208,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.035186562687158585,
      "eval_runtime": 173.783,
      "eval_samples_per_second": 6.652,
      "eval_steps_per_second": 6.652,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.11019343882799149,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0172,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.23066164553165436,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0228,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.14030225574970245,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0121,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.41205379366874695,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0231,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.12606282532215118,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.011,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.03656512498855591,
      "eval_runtime": 173.7743,
      "eval_samples_per_second": 6.652,
      "eval_steps_per_second": 6.652,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.34628793597221375,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0174,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.182864248752594,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0222,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.34521186351776123,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0186,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.15361788868904114,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0194,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.7452395558357239,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0196,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.03452925756573677,
      "eval_runtime": 173.8356,
      "eval_samples_per_second": 6.65,
      "eval_steps_per_second": 6.65,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.20888769626617432,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0219,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.13797582685947418,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0205,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.13696569204330444,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.014,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.28211092948913574,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0178,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.6226332187652588,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0155,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.038163114339113235,
      "eval_runtime": 173.8292,
      "eval_samples_per_second": 6.65,
      "eval_steps_per_second": 6.65,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.303265780210495,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0139,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.18224796652793884,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0165,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.1443055421113968,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0178,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.297122061252594,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.018,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.193439319729805,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.012,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.037124570459127426,
      "eval_runtime": 173.9241,
      "eval_samples_per_second": 6.647,
      "eval_steps_per_second": 6.647,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.1973581612110138,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0163,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.2399301379919052,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0215,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.13810937106609344,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0161,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.35670533776283264,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0232,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.1930951625108719,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0199,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.03314182162284851,
      "eval_runtime": 173.8104,
      "eval_samples_per_second": 6.651,
      "eval_steps_per_second": 6.651,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.36426177620887756,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0137,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.3521732985973358,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0173,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.4830394387245178,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.02,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.18669484555721283,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.019,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.3846980631351471,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0209,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.035960279405117035,
      "eval_runtime": 173.8213,
      "eval_samples_per_second": 6.651,
      "eval_steps_per_second": 6.651,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.4228716790676117,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0173,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.6368355751037598,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.027,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.4529774785041809,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0195,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.13082724809646606,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0147,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.09245896339416504,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0228,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.03236575424671173,
      "eval_runtime": 173.9654,
      "eval_samples_per_second": 6.645,
      "eval_steps_per_second": 6.645,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.2635294497013092,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0123,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.3114376962184906,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.018,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.5124761462211609,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0216,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.23497699201107025,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.017,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.32194241881370544,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0192,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03174876049160957,
      "eval_runtime": 173.9711,
      "eval_samples_per_second": 6.645,
      "eval_steps_per_second": 6.645,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.3975355923175812,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0188,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.5243850946426392,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0228,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.47261595726013184,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0181,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.29687321186065674,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0186,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.203555628657341,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0204,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.03203918784856796,
      "eval_runtime": 173.9068,
      "eval_samples_per_second": 6.647,
      "eval_steps_per_second": 6.647,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.2872328460216522,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0141,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.3177731931209564,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0238,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.22723254561424255,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0139,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.14254233241081238,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0158,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.19210264086723328,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0165,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.03305741399526596,
      "eval_runtime": 173.9951,
      "eval_samples_per_second": 6.644,
      "eval_steps_per_second": 6.644,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.24568738043308258,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0205,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.43385347723960876,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0211,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.10379448533058167,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0154,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.2628607749938965,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0138,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.1557886153459549,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0225,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.03408072888851166,
      "eval_runtime": 174.0473,
      "eval_samples_per_second": 6.642,
      "eval_steps_per_second": 6.642,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.13551560044288635,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0064,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.09563615918159485,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0097,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.23707009851932526,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.008,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.3056003749370575,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0062,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.1288238763809204,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0083,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.03800658509135246,
      "eval_runtime": 174.1291,
      "eval_samples_per_second": 6.639,
      "eval_steps_per_second": 6.639,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 0.31253793835639954,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0044,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.05842793732881546,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0098,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.11472361534833908,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0094,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.08621902763843536,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0112,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.7524242997169495,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0119,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.03924186900258064,
      "eval_runtime": 174.1488,
      "eval_samples_per_second": 6.638,
      "eval_steps_per_second": 6.638,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.07297563552856445,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0052,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.5778046250343323,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0086,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.22083096206188202,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0093,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.34668534994125366,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0087,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.39718756079673767,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0069,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.03725030645728111,
      "eval_runtime": 174.2116,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 6.636,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.05647918954491615,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0088,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.355755478143692,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0088,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.21908801794052124,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0075,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.7008113265037537,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0073,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.43440404534339905,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.007,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.03963570296764374,
      "eval_runtime": 174.2118,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 6.636,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.15854176878929138,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0084,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.4227380156517029,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0081,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.1825554072856903,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0087,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.18914279341697693,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0063,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.13039569556713104,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0132,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.03769873455166817,
      "eval_runtime": 174.255,
      "eval_samples_per_second": 6.634,
      "eval_steps_per_second": 6.634,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.5920471549034119,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0078,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.40638601779937744,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0076,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.15520386397838593,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0113,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.11067832261323929,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0111,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.21373113989830017,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0073,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.036785051226615906,
      "eval_runtime": 174.214,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 6.636,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.06737236678600311,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.006,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.24792876839637756,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0048,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.15693798661231995,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0091,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.31480202078819275,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0069,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.1157020851969719,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0104,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.03893345966935158,
      "eval_runtime": 174.3244,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 6.631,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.04068097472190857,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0073,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.4247656762599945,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0092,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.14573094248771667,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0078,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.4784613847732544,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0125,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.21929225325584412,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.005,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.0398164764046669,
      "eval_runtime": 174.3532,
      "eval_samples_per_second": 6.63,
      "eval_steps_per_second": 6.63,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.3877716362476349,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0138,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.24963144958019257,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0074,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.19199822843074799,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0082,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.1781242936849594,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0115,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.2099018096923828,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0117,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.03761526942253113,
      "eval_runtime": 174.2421,
      "eval_samples_per_second": 6.634,
      "eval_steps_per_second": 6.634,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.10772652924060822,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0104,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.09071230888366699,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0068,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.047642093151807785,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0095,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.12309250980615616,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0057,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.017464958131313324,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0066,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.0388132780790329,
      "eval_runtime": 174.4286,
      "eval_samples_per_second": 6.627,
      "eval_steps_per_second": 6.627,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.38554105162620544,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0081,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.3859759569168091,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0058,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.15807801485061646,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0113,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.5472670793533325,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0083,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.14712011814117432,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.007,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.03851594030857086,
      "eval_runtime": 174.3898,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 6.629,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.18404144048690796,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0088,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.21822212636470795,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0036,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.11614836752414703,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0055,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.0687367394566536,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0042,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.48854318261146545,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0081,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.04083527997136116,
      "eval_runtime": 174.4269,
      "eval_samples_per_second": 6.627,
      "eval_steps_per_second": 6.627,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.3334047198295593,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0088,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.23079705238342285,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0064,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.0218326635658741,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0108,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.25172847509384155,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0075,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.12312012165784836,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0089,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.03993625193834305,
      "eval_runtime": 174.3805,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 6.629,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.04196947067975998,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.004,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.17466288805007935,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0026,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.12816299498081207,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0021,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.024005437269806862,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0041,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.024433216080069542,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0023,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.041467901319265366,
      "eval_runtime": 174.4306,
      "eval_samples_per_second": 6.627,
      "eval_steps_per_second": 6.627,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.17065927386283875,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0056,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.07094801962375641,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0019,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.04778605327010155,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0028,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.02742954157292843,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0042,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.05647296831011772,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0079,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.04430392012000084,
      "eval_runtime": 174.493,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.18171408772468567,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0051,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.12204571068286896,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0088,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.05209114030003548,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0037,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.3705243170261383,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0045,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.04583051800727844,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0016,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.04461469128727913,
      "eval_runtime": 174.3833,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 6.629,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.1135777160525322,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0019,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.0552319772541523,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0034,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.09149715304374695,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0023,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.04404257610440254,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0038,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.0674886703491211,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0022,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.045434363186359406,
      "eval_runtime": 174.4111,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.1479644626379013,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0022,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.05602571740746498,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0016,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.06363249570131302,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0022,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.033200863748788834,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0023,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.02339135855436325,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0028,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.046483054757118225,
      "eval_runtime": 174.4068,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.22476452589035034,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0023,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.010722499340772629,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0008,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.029050569981336594,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0065,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.011172047816216946,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0022,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.2565000057220459,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.004,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04684334993362427,
      "eval_runtime": 174.3934,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 6.629,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.050229333341121674,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0066,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.16038791835308075,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0045,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.019315680488944054,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0029,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.3864138126373291,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0053,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.06462980806827545,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0024,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.046574667096138,
      "eval_runtime": 174.5719,
      "eval_samples_per_second": 6.622,
      "eval_steps_per_second": 6.622,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.17723701894283295,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0014,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.04783819243311882,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0057,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.02630511485040188,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0021,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.491956889629364,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0041,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.1486896574497223,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0031,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.04662920907139778,
      "eval_runtime": 174.4886,
      "eval_samples_per_second": 6.625,
      "eval_steps_per_second": 6.625,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.0888807401061058,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0025,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.2614567279815674,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0025,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.8806897401809692,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0047,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.5426682829856873,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0039,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.00869567971676588,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0043,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.04692388325929642,
      "eval_runtime": 174.4708,
      "eval_samples_per_second": 6.626,
      "eval_steps_per_second": 6.626,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.24513138830661774,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0028,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.23918958008289337,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0025,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 1.3059184551239014,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.008,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.19193530082702637,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0027,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.0673174262046814,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0035,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.046790238469839096,
      "eval_runtime": 174.5956,
      "eval_samples_per_second": 6.621,
      "eval_steps_per_second": 6.621,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.13142763078212738,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0077,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.31001541018486023,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0039,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.009679196402430534,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0049,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.016995983198285103,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0017,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.06072453036904335,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0032,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.046724770218133926,
      "eval_runtime": 174.5562,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.025896934792399406,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0026,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.14590482413768768,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0032,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.09465014189481735,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0021,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.01945341005921364,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0039,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.022305987775325775,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.003,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.046812571585178375,
      "eval_runtime": 174.51,
      "eval_samples_per_second": 6.624,
      "eval_steps_per_second": 6.624,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.05401395633816719,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0052,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.0411403626203537,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0016,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.5392578840255737,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0053,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.030586998909711838,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0016,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 8.169046237996646e+17,
      "train_loss": 0.03518242226563428,
      "train_runtime": 38556.4473,
      "train_samples_per_second": 1.348,
      "train_steps_per_second": 0.084
    }
  ],
  "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
}