{
  "best_metric": 0.03181542828679085,
  "best_model_checkpoint": "saves/psy-course/Ministral-8B-Instruct-2410/train/fold2/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": 2.714885711669922,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.3364,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 2.8060686588287354,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.2705,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 3.170583486557007,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.0348,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 2.260378122329712,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.555,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 1.1415810585021973,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.2583,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.24182522296905518,
      "eval_runtime": 181.7987,
      "eval_samples_per_second": 6.359,
      "eval_steps_per_second": 6.359,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 1.5008163452148438,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.1594,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.2160450220108032,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.1275,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 1.2109729051589966,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1016,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 1.4886703491210938,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.1185,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.7382891774177551,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0854,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.06951398402452469,
      "eval_runtime": 181.7285,
      "eval_samples_per_second": 6.361,
      "eval_steps_per_second": 6.361,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8113016486167908,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0648,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.8134452700614929,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0724,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.111953616142273,
      "learning_rate": 4e-05,
      "loss": 0.0692,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.7792268991470337,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.079,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.9049783945083618,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.061,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.0585860013961792,
      "eval_runtime": 181.7435,
      "eval_samples_per_second": 6.361,
      "eval_steps_per_second": 6.361,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.7568258047103882,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0617,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 0.7794573307037354,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0615,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.8195641040802002,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0526,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.3618631064891815,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.044,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5989110469818115,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0581,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.05444643646478653,
      "eval_runtime": 181.7053,
      "eval_samples_per_second": 6.362,
      "eval_steps_per_second": 6.362,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.7792056798934937,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0558,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.8352409601211548,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0499,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.3525795042514801,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0384,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.4654476046562195,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0458,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.45548224449157715,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0443,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.04309243708848953,
      "eval_runtime": 181.7049,
      "eval_samples_per_second": 6.362,
      "eval_steps_per_second": 6.362,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.5583269596099854,
      "learning_rate": 8e-05,
      "loss": 0.0404,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.41081997752189636,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0411,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.513261616230011,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0541,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.43006983399391174,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0506,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.38600465655326843,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0398,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04688544571399689,
      "eval_runtime": 181.771,
      "eval_samples_per_second": 6.36,
      "eval_steps_per_second": 6.36,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.5189172625541687,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0522,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.31192007660865784,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0464,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.47563448548316956,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.048,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.3781372308731079,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0445,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.43737003207206726,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0431,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.045772019773721695,
      "eval_runtime": 181.7468,
      "eval_samples_per_second": 6.36,
      "eval_steps_per_second": 6.36,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.36792123317718506,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0571,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.4085187613964081,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0369,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.5380191206932068,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.038,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.4832334518432617,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0351,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.26285237073898315,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0482,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.04355475306510925,
      "eval_runtime": 181.7461,
      "eval_samples_per_second": 6.361,
      "eval_steps_per_second": 6.361,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.2608589828014374,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.042,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.5077489018440247,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0454,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.4572131037712097,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0437,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.4632183015346527,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0581,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.3846943974494934,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.029,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.038769323378801346,
      "eval_runtime": 181.7543,
      "eval_samples_per_second": 6.36,
      "eval_steps_per_second": 6.36,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.21216456592082977,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0351,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.3389206528663635,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0308,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.6215451955795288,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0427,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.23567695915699005,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0453,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.32285988330841064,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0294,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.03677697852253914,
      "eval_runtime": 181.7747,
      "eval_samples_per_second": 6.36,
      "eval_steps_per_second": 6.36,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.44437527656555176,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0299,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.32935965061187744,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0423,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.3649428188800812,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0525,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.33312538266181946,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0413,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.2349109649658203,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0415,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.034638144075870514,
      "eval_runtime": 181.7519,
      "eval_samples_per_second": 6.36,
      "eval_steps_per_second": 6.36,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.29396870732307434,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0355,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.23835968971252441,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0259,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.32715070247650146,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0346,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.2577593922615051,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0499,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.2869316339492798,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0341,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.034780871123075485,
      "eval_runtime": 181.7801,
      "eval_samples_per_second": 6.359,
      "eval_steps_per_second": 6.359,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.3402882516384125,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0313,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.27556324005126953,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0435,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.3360099196434021,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0427,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.2659825384616852,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0327,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.37849897146224976,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0303,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03528154641389847,
      "eval_runtime": 181.8918,
      "eval_samples_per_second": 6.355,
      "eval_steps_per_second": 6.355,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.3455275595188141,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0339,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.23953555524349213,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0311,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.28274527192115784,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.033,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.526482343673706,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0442,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.3312353491783142,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0353,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.037871893495321274,
      "eval_runtime": 181.8407,
      "eval_samples_per_second": 6.357,
      "eval_steps_per_second": 6.357,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.5539495348930359,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0325,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.21886028349399567,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0287,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.20140358805656433,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.029,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.4935223162174225,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0361,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.3583446443080902,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.026,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.03458985313773155,
      "eval_runtime": 181.8591,
      "eval_samples_per_second": 6.357,
      "eval_steps_per_second": 6.357,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.3114105463027954,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.033,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.18866032361984253,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0344,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.37387722730636597,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0347,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.39051756262779236,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0346,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.3173379600048065,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0289,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.0333932600915432,
      "eval_runtime": 181.7815,
      "eval_samples_per_second": 6.359,
      "eval_steps_per_second": 6.359,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.21952079236507416,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0287,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.8045993447303772,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0286,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.28386226296424866,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0341,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.35170432925224304,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0269,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.3447723984718323,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0274,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.03453436121344566,
      "eval_runtime": 181.7653,
      "eval_samples_per_second": 6.36,
      "eval_steps_per_second": 6.36,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.21927620470523834,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0265,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.26884472370147705,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0266,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.17841775715351105,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0288,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.28215092420578003,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0383,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.2890825569629669,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0206,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.032716140151023865,
      "eval_runtime": 181.7567,
      "eval_samples_per_second": 6.36,
      "eval_steps_per_second": 6.36,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.2569309175014496,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0294,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.28140658140182495,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0312,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.21810609102249146,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0266,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.5112712383270264,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0283,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.14848096668720245,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0362,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.0331537090241909,
      "eval_runtime": 181.9323,
      "eval_samples_per_second": 6.354,
      "eval_steps_per_second": 6.354,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.255219966173172,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0283,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.09283377230167389,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0229,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.363229364156723,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0295,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.26844626665115356,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0295,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.42271193861961365,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0305,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.035451073199510574,
      "eval_runtime": 181.8826,
      "eval_samples_per_second": 6.356,
      "eval_steps_per_second": 6.356,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.3385646939277649,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0346,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.35555073618888855,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0316,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.26448431611061096,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0297,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.3541312515735626,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0274,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.6882481575012207,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.033,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03250161185860634,
      "eval_runtime": 181.7856,
      "eval_samples_per_second": 6.359,
      "eval_steps_per_second": 6.359,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.19301781058311462,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0226,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.24382206797599792,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0258,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.4981677234172821,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0231,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.8556720614433289,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0287,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.28823643922805786,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.035,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.03675594925880432,
      "eval_runtime": 181.8802,
      "eval_samples_per_second": 6.356,
      "eval_steps_per_second": 6.356,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.4210245907306671,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0307,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.2030617594718933,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0241,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.39251893758773804,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0338,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.21648012101650238,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0391,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.21202810108661652,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0219,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.03331456333398819,
      "eval_runtime": 181.9008,
      "eval_samples_per_second": 6.355,
      "eval_steps_per_second": 6.355,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.18095539510250092,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0336,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.22352856397628784,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0242,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.28671175241470337,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0259,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.26024141907691956,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0299,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.13861975073814392,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0199,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.037526000291109085,
      "eval_runtime": 181.8816,
      "eval_samples_per_second": 6.356,
      "eval_steps_per_second": 6.356,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.5025237798690796,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0295,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.18757639825344086,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0233,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.13187754154205322,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0276,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.3836573362350464,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0347,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.4797174632549286,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0279,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.032986849546432495,
      "eval_runtime": 181.8359,
      "eval_samples_per_second": 6.357,
      "eval_steps_per_second": 6.357,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.6321930289268494,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0302,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.22899775207042694,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0309,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.27863240242004395,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0225,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.1779979020357132,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0315,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.18828840553760529,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0226,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.03224512189626694,
      "eval_runtime": 181.8611,
      "eval_samples_per_second": 6.357,
      "eval_steps_per_second": 6.357,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.2461148500442505,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0176,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.35559847950935364,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.018,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.17443691194057465,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0192,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.0687423050403595,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0213,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.12578760087490082,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0178,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.03181542828679085,
      "eval_runtime": 181.7239,
      "eval_samples_per_second": 6.361,
      "eval_steps_per_second": 6.361,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.1363806426525116,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0185,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.43135547637939453,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0164,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.2357529252767563,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0134,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.08881539851427078,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0142,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.5172821283340454,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.022,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.03445231914520264,
      "eval_runtime": 181.7197,
      "eval_samples_per_second": 6.361,
      "eval_steps_per_second": 6.361,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.06512246280908585,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0168,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.24386030435562134,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0219,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.17258454859256744,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0111,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.24131588637828827,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.021,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.12388203293085098,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0098,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.03824140876531601,
      "eval_runtime": 181.8235,
      "eval_samples_per_second": 6.358,
      "eval_steps_per_second": 6.358,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.4268186092376709,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0188,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.2414432018995285,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0211,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.3145079016685486,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0186,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.12285123765468597,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0204,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.8541750907897949,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0209,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.03401584550738335,
      "eval_runtime": 181.7355,
      "eval_samples_per_second": 6.361,
      "eval_steps_per_second": 6.361,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.13757294416427612,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0216,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.1696965992450714,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0219,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.11982149630784988,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0145,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.2645229399204254,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0184,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.2876879572868347,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.015,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.0370011031627655,
      "eval_runtime": 181.7426,
      "eval_samples_per_second": 6.361,
      "eval_steps_per_second": 6.361,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.3021048903465271,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0137,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.18024735152721405,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0177,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.17211011052131653,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0177,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.2713499069213867,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.018,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.18026965856552124,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0139,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.036044903099536896,
      "eval_runtime": 181.9632,
      "eval_samples_per_second": 6.353,
      "eval_steps_per_second": 6.353,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.28000470995903015,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0166,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.2058064043521881,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0211,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.4643082022666931,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0168,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.3830045163631439,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0213,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.1851435899734497,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0196,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.0336180217564106,
      "eval_runtime": 181.8677,
      "eval_samples_per_second": 6.356,
      "eval_steps_per_second": 6.356,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.25386324524879456,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0111,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.41750213503837585,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0175,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.5525423288345337,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0195,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.18864604830741882,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.021,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.4163510799407959,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0205,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03435565531253815,
      "eval_runtime": 181.8924,
      "eval_samples_per_second": 6.355,
      "eval_steps_per_second": 6.355,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.46013107895851135,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0192,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.35217469930648804,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0268,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.5294527411460876,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0201,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.15640372037887573,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0152,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.08065128326416016,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0229,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.032898351550102234,
      "eval_runtime": 181.9287,
      "eval_samples_per_second": 6.354,
      "eval_steps_per_second": 6.354,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.2963849604129791,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0129,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.35389959812164307,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0173,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.5031420588493347,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0221,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.2417978197336197,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0175,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.2386423647403717,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0184,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03297585994005203,
      "eval_runtime": 181.9888,
      "eval_samples_per_second": 6.352,
      "eval_steps_per_second": 6.352,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.424748957157135,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0186,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.2975512742996216,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0217,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.6139042377471924,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0205,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.20860710740089417,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0204,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.1627204269170761,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0192,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.03292803466320038,
      "eval_runtime": 182.25,
      "eval_samples_per_second": 6.343,
      "eval_steps_per_second": 6.343,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.31320932507514954,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0146,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.2864499092102051,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0245,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.16330762207508087,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0137,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.1614055633544922,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0169,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.18817143142223358,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0176,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.03322253003716469,
      "eval_runtime": 181.8996,
      "eval_samples_per_second": 6.355,
      "eval_steps_per_second": 6.355,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.5733834505081177,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.023,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.383525550365448,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0187,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.1702577918767929,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0155,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.25431495904922485,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0134,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.1699695587158203,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0235,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.034495554864406586,
      "eval_runtime": 182.1707,
      "eval_samples_per_second": 6.346,
      "eval_steps_per_second": 6.346,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.06585952639579773,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0068,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.06236855685710907,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0107,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.19067367911338806,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0084,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.1599722057580948,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0061,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.11424776166677475,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0092,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.03834828734397888,
      "eval_runtime": 182.1435,
      "eval_samples_per_second": 6.347,
      "eval_steps_per_second": 6.347,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 0.2723098695278168,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0046,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.053860701620578766,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0095,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.13399919867515564,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0094,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.07494858652353287,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.009,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.6289196014404297,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.012,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.041652098298072815,
      "eval_runtime": 181.992,
      "eval_samples_per_second": 6.352,
      "eval_steps_per_second": 6.352,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.10614260286092758,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0061,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.3607775568962097,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0071,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.48487016558647156,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0089,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.44034343957901,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0102,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.18452033400535583,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0077,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.03836505860090256,
      "eval_runtime": 181.9754,
      "eval_samples_per_second": 6.353,
      "eval_steps_per_second": 6.353,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.06003500148653984,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0082,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.354106605052948,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0089,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.26153624057769775,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0086,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.0856068953871727,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0065,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.4745621383190155,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0068,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.039253655821084976,
      "eval_runtime": 182.2155,
      "eval_samples_per_second": 6.344,
      "eval_steps_per_second": 6.344,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.23959848284721375,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0079,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.606198251247406,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0108,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.24547742307186127,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0091,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.21174629032611847,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.006,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.05651284009218216,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0121,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.03834627568721771,
      "eval_runtime": 182.1278,
      "eval_samples_per_second": 6.347,
      "eval_steps_per_second": 6.347,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.16370394825935364,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0096,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.20642130076885223,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0064,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.07156592607498169,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0107,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.21420058608055115,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.011,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.10118517279624939,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0077,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.03682588040828705,
      "eval_runtime": 182.308,
      "eval_samples_per_second": 6.341,
      "eval_steps_per_second": 6.341,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.0672723799943924,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0072,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.18549731373786926,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0052,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.12499016523361206,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0076,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.22891055047512054,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0056,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.11049254238605499,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0115,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.03893190994858742,
      "eval_runtime": 182.6193,
      "eval_samples_per_second": 6.33,
      "eval_steps_per_second": 6.33,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.05207858979701996,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0097,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.4066181480884552,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0093,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.569721519947052,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0074,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.8917023539543152,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0134,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.18456467986106873,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0061,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.03987393528223038,
      "eval_runtime": 182.164,
      "eval_samples_per_second": 6.346,
      "eval_steps_per_second": 6.346,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.2530272603034973,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0145,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.28562411665916443,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0088,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.18049819767475128,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0095,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.21345478296279907,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0118,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.13870681822299957,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0132,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.036747824400663376,
      "eval_runtime": 182.5195,
      "eval_samples_per_second": 6.334,
      "eval_steps_per_second": 6.334,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.03900235891342163,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0083,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.07937924563884735,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0068,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.03925955668091774,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0097,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.33661210536956787,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.006,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.024305246770381927,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0066,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.03801007568836212,
      "eval_runtime": 182.2954,
      "eval_samples_per_second": 6.341,
      "eval_steps_per_second": 6.341,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.2657100558280945,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0075,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.4867688715457916,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0051,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.18266838788986206,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0117,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.9398632049560547,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0092,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.09116644412279129,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0076,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.03867095336318016,
      "eval_runtime": 182.6346,
      "eval_samples_per_second": 6.33,
      "eval_steps_per_second": 6.33,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.4438709318637848,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0104,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.2803715169429779,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0043,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.05800764635205269,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0057,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.08257286995649338,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0039,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.6039398312568665,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0087,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.0398300401866436,
      "eval_runtime": 182.752,
      "eval_samples_per_second": 6.326,
      "eval_steps_per_second": 6.326,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.4885454773902893,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0072,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.6458340287208557,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0065,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.018503481522202492,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0118,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.2634695768356323,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0082,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.08771725744009018,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0094,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.03990135341882706,
      "eval_runtime": 182.5063,
      "eval_samples_per_second": 6.334,
      "eval_steps_per_second": 6.334,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.041795965284109116,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0036,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.15982815623283386,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0024,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.14116846024990082,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.002,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.029664216563105583,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0054,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.02512684278190136,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0024,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.041614893823862076,
      "eval_runtime": 182.6617,
      "eval_samples_per_second": 6.329,
      "eval_steps_per_second": 6.329,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.17406991124153137,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0051,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.08610302209854126,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0024,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.04503106325864792,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0029,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.06606682389974594,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0051,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.06268336623907089,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0077,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.04621283337473869,
      "eval_runtime": 182.3492,
      "eval_samples_per_second": 6.339,
      "eval_steps_per_second": 6.339,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.050343628972768784,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0037,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.10917084664106369,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.008,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.09494730085134506,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0039,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.06835534423589706,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.004,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.09156963229179382,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0018,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.04548865929245949,
      "eval_runtime": 183.0184,
      "eval_samples_per_second": 6.316,
      "eval_steps_per_second": 6.316,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.09536194801330566,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0019,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.11849187314510345,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0021,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.14611715078353882,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0027,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.1069348007440567,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0034,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.06568971276283264,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0021,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.04689283296465874,
      "eval_runtime": 182.6389,
      "eval_samples_per_second": 6.329,
      "eval_steps_per_second": 6.329,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.08502360433340073,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0029,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.020981520414352417,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0015,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.06051165983080864,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0021,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.03838418796658516,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0024,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.02361064776778221,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0026,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.048040326684713364,
      "eval_runtime": 182.6191,
      "eval_samples_per_second": 6.33,
      "eval_steps_per_second": 6.33,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.21955329179763794,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0022,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.04029189795255661,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0013,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.02981954626739025,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0074,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.017013270407915115,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0039,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.2801615595817566,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0043,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04885905236005783,
      "eval_runtime": 182.3966,
      "eval_samples_per_second": 6.338,
      "eval_steps_per_second": 6.338,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.03359270840883255,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0073,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.12972494959831238,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0041,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.05984209477901459,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0033,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.4022550880908966,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0058,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.05884170159697533,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0018,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.047888416796922684,
      "eval_runtime": 182.5315,
      "eval_samples_per_second": 6.333,
      "eval_steps_per_second": 6.333,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.28109657764434814,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0011,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.15115970373153687,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0072,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.08594682067632675,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0023,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.48176631331443787,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0043,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.0299384705722332,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0034,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.04794705659151077,
      "eval_runtime": 182.5327,
      "eval_samples_per_second": 6.333,
      "eval_steps_per_second": 6.333,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.1159643605351448,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0026,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.31304869055747986,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0031,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.2677801847457886,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0051,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.09320933371782303,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0036,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.0045815082266926765,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0037,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.048097576946020126,
      "eval_runtime": 182.8076,
      "eval_samples_per_second": 6.324,
      "eval_steps_per_second": 6.324,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.06252841651439667,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0021,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.497509628534317,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0031,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.37336939573287964,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0078,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.21054156124591827,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0028,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.05313187092542648,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0037,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.048036374151706696,
      "eval_runtime": 182.2974,
      "eval_samples_per_second": 6.341,
      "eval_steps_per_second": 6.341,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.10948553681373596,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0073,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.09476953744888306,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0039,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.005633699707686901,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.006,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.024644823744893074,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.002,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.07547037303447723,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0036,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.04801381751894951,
      "eval_runtime": 180.6529,
      "eval_samples_per_second": 6.399,
      "eval_steps_per_second": 6.399,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.027425028383731842,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.003,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.15399925410747528,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.003,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.17350825667381287,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0026,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.01795775815844536,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0045,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.02166036330163479,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0027,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.04795505106449127,
      "eval_runtime": 180.5439,
      "eval_samples_per_second": 6.403,
      "eval_steps_per_second": 6.403,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.02288454957306385,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0031,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.04133681207895279,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0019,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.041048236191272736,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0055,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.051259420812129974,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0017,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 8.169046237996646e+17,
      "train_loss": 0.03529967672029404,
      "train_runtime": 40405.575,
      "train_samples_per_second": 1.286,
      "train_steps_per_second": 0.08
    }
  ],
  "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
}