{
  "best_metric": 0.036797065287828445,
  "best_model_checkpoint": "saves/psy-course/Llama-3.1-8B-Instruct/train/fold9/checkpoint-1300",
  "epoch": 4.997121473805412,
  "eval_steps": 50,
  "global_step": 3255,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01535213970447131,
      "grad_norm": 4.642497539520264,
      "learning_rate": 3.067484662576687e-06,
      "loss": 1.6312,
      "step": 10
    },
    {
      "epoch": 0.03070427940894262,
      "grad_norm": 5.409369945526123,
      "learning_rate": 6.134969325153374e-06,
      "loss": 1.4429,
      "step": 20
    },
    {
      "epoch": 0.04605641911341393,
      "grad_norm": 5.949030876159668,
      "learning_rate": 9.202453987730062e-06,
      "loss": 1.4142,
      "step": 30
    },
    {
      "epoch": 0.06140855881788524,
      "grad_norm": 2.751120090484619,
      "learning_rate": 1.2269938650306748e-05,
      "loss": 0.9213,
      "step": 40
    },
    {
      "epoch": 0.07676069852235655,
      "grad_norm": 1.7110086679458618,
      "learning_rate": 1.5337423312883436e-05,
      "loss": 0.6165,
      "step": 50
    },
    {
      "epoch": 0.07676069852235655,
      "eval_loss": 0.3982980251312256,
      "eval_runtime": 157.3623,
      "eval_samples_per_second": 7.365,
      "eval_steps_per_second": 7.365,
      "step": 50
    },
    {
      "epoch": 0.09211283822682786,
      "grad_norm": 1.144059181213379,
      "learning_rate": 1.8404907975460123e-05,
      "loss": 0.4066,
      "step": 60
    },
    {
      "epoch": 0.10746497793129918,
      "grad_norm": 1.527937412261963,
      "learning_rate": 2.1472392638036813e-05,
      "loss": 0.276,
      "step": 70
    },
    {
      "epoch": 0.12281711763577048,
      "grad_norm": 1.0224796533584595,
      "learning_rate": 2.4539877300613496e-05,
      "loss": 0.1687,
      "step": 80
    },
    {
      "epoch": 0.1381692573402418,
      "grad_norm": 0.9973428249359131,
      "learning_rate": 2.7607361963190186e-05,
      "loss": 0.1319,
      "step": 90
    },
    {
      "epoch": 0.1535213970447131,
      "grad_norm": 0.7085102200508118,
      "learning_rate": 3.067484662576687e-05,
      "loss": 0.1035,
      "step": 100
    },
    {
      "epoch": 0.1535213970447131,
      "eval_loss": 0.08296720683574677,
      "eval_runtime": 157.3193,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 100
    },
    {
      "epoch": 0.16887353674918443,
      "grad_norm": 0.830489993095398,
      "learning_rate": 3.3742331288343556e-05,
      "loss": 0.0736,
      "step": 110
    },
    {
      "epoch": 0.18422567645365573,
      "grad_norm": 0.8759797811508179,
      "learning_rate": 3.6809815950920246e-05,
      "loss": 0.0691,
      "step": 120
    },
    {
      "epoch": 0.19957781615812703,
      "grad_norm": 1.1546653509140015,
      "learning_rate": 3.987730061349693e-05,
      "loss": 0.0809,
      "step": 130
    },
    {
      "epoch": 0.21492995586259836,
      "grad_norm": 0.631569504737854,
      "learning_rate": 4.2944785276073626e-05,
      "loss": 0.0557,
      "step": 140
    },
    {
      "epoch": 0.23028209556706966,
      "grad_norm": 1.1803288459777832,
      "learning_rate": 4.601226993865031e-05,
      "loss": 0.0746,
      "step": 150
    },
    {
      "epoch": 0.23028209556706966,
      "eval_loss": 0.06817680597305298,
      "eval_runtime": 157.2581,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 150
    },
    {
      "epoch": 0.24563423527154096,
      "grad_norm": 0.8749077916145325,
      "learning_rate": 4.907975460122699e-05,
      "loss": 0.0854,
      "step": 160
    },
    {
      "epoch": 0.2609863749760123,
      "grad_norm": 0.7158986926078796,
      "learning_rate": 5.214723926380368e-05,
      "loss": 0.0763,
      "step": 170
    },
    {
      "epoch": 0.2763385146804836,
      "grad_norm": 0.7949298024177551,
      "learning_rate": 5.521472392638037e-05,
      "loss": 0.0714,
      "step": 180
    },
    {
      "epoch": 0.2916906543849549,
      "grad_norm": 2.1728086471557617,
      "learning_rate": 5.8282208588957056e-05,
      "loss": 0.0591,
      "step": 190
    },
    {
      "epoch": 0.3070427940894262,
      "grad_norm": 0.8605216145515442,
      "learning_rate": 6.134969325153375e-05,
      "loss": 0.0591,
      "step": 200
    },
    {
      "epoch": 0.3070427940894262,
      "eval_loss": 0.059281475841999054,
      "eval_runtime": 157.2933,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 200
    },
    {
      "epoch": 0.3223949337938975,
      "grad_norm": 0.9615307450294495,
      "learning_rate": 6.441717791411042e-05,
      "loss": 0.0519,
      "step": 210
    },
    {
      "epoch": 0.33774707349836885,
      "grad_norm": 0.60662442445755,
      "learning_rate": 6.748466257668711e-05,
      "loss": 0.0436,
      "step": 220
    },
    {
      "epoch": 0.35309921320284016,
      "grad_norm": 0.3369896411895752,
      "learning_rate": 7.055214723926382e-05,
      "loss": 0.082,
      "step": 230
    },
    {
      "epoch": 0.36845135290731146,
      "grad_norm": 0.5379114747047424,
      "learning_rate": 7.361963190184049e-05,
      "loss": 0.0583,
      "step": 240
    },
    {
      "epoch": 0.38380349261178276,
      "grad_norm": 1.2060985565185547,
      "learning_rate": 7.668711656441718e-05,
      "loss": 0.0716,
      "step": 250
    },
    {
      "epoch": 0.38380349261178276,
      "eval_loss": 0.052977994084358215,
      "eval_runtime": 157.3133,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 250
    },
    {
      "epoch": 0.39915563231625406,
      "grad_norm": 0.752123236656189,
      "learning_rate": 7.975460122699386e-05,
      "loss": 0.0559,
      "step": 260
    },
    {
      "epoch": 0.41450777202072536,
      "grad_norm": 1.1011897325515747,
      "learning_rate": 8.282208588957055e-05,
      "loss": 0.0611,
      "step": 270
    },
    {
      "epoch": 0.4298599117251967,
      "grad_norm": 0.8125905394554138,
      "learning_rate": 8.588957055214725e-05,
      "loss": 0.0614,
      "step": 280
    },
    {
      "epoch": 0.445212051429668,
      "grad_norm": 0.48761671781539917,
      "learning_rate": 8.895705521472393e-05,
      "loss": 0.0472,
      "step": 290
    },
    {
      "epoch": 0.4605641911341393,
      "grad_norm": 0.3559521436691284,
      "learning_rate": 9.202453987730062e-05,
      "loss": 0.0571,
      "step": 300
    },
    {
      "epoch": 0.4605641911341393,
      "eval_loss": 0.04749785736203194,
      "eval_runtime": 157.325,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 300
    },
    {
      "epoch": 0.4759163308386106,
      "grad_norm": 0.600727915763855,
      "learning_rate": 9.50920245398773e-05,
      "loss": 0.0457,
      "step": 310
    },
    {
      "epoch": 0.4912684705430819,
      "grad_norm": 0.4633491635322571,
      "learning_rate": 9.815950920245399e-05,
      "loss": 0.0464,
      "step": 320
    },
    {
      "epoch": 0.5066206102475532,
      "grad_norm": 0.6039092540740967,
      "learning_rate": 9.999953982785432e-05,
      "loss": 0.0532,
      "step": 330
    },
    {
      "epoch": 0.5219727499520246,
      "grad_norm": 0.35319024324417114,
      "learning_rate": 9.999436298849151e-05,
      "loss": 0.0476,
      "step": 340
    },
    {
      "epoch": 0.5373248896564958,
      "grad_norm": 0.2034178227186203,
      "learning_rate": 9.998343469212352e-05,
      "loss": 0.0499,
      "step": 350
    },
    {
      "epoch": 0.5373248896564958,
      "eval_loss": 0.0483204647898674,
      "eval_runtime": 157.3109,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 350
    },
    {
      "epoch": 0.5526770293609672,
      "grad_norm": 0.7975025773048401,
      "learning_rate": 9.996675619596465e-05,
      "loss": 0.0486,
      "step": 360
    },
    {
      "epoch": 0.5680291690654385,
      "grad_norm": 0.7012478709220886,
      "learning_rate": 9.99443294187443e-05,
      "loss": 0.0545,
      "step": 370
    },
    {
      "epoch": 0.5833813087699098,
      "grad_norm": 0.17911651730537415,
      "learning_rate": 9.991615694048621e-05,
      "loss": 0.0449,
      "step": 380
    },
    {
      "epoch": 0.5987334484743811,
      "grad_norm": 0.5984811782836914,
      "learning_rate": 9.988224200221172e-05,
      "loss": 0.0606,
      "step": 390
    },
    {
      "epoch": 0.6140855881788524,
      "grad_norm": 0.37629756331443787,
      "learning_rate": 9.984258850556693e-05,
      "loss": 0.0421,
      "step": 400
    },
    {
      "epoch": 0.6140855881788524,
      "eval_loss": 0.047019653022289276,
      "eval_runtime": 157.366,
      "eval_samples_per_second": 7.365,
      "eval_steps_per_second": 7.365,
      "step": 400
    },
    {
      "epoch": 0.6294377278833237,
      "grad_norm": 0.7881872653961182,
      "learning_rate": 9.979720101237375e-05,
      "loss": 0.0445,
      "step": 410
    },
    {
      "epoch": 0.644789867587795,
      "grad_norm": 0.32935699820518494,
      "learning_rate": 9.974608474410512e-05,
      "loss": 0.0492,
      "step": 420
    },
    {
      "epoch": 0.6601420072922664,
      "grad_norm": 0.6060999631881714,
      "learning_rate": 9.968924558128445e-05,
      "loss": 0.0482,
      "step": 430
    },
    {
      "epoch": 0.6754941469967377,
      "grad_norm": 0.3332858979701996,
      "learning_rate": 9.962669006280894e-05,
      "loss": 0.0408,
      "step": 440
    },
    {
      "epoch": 0.690846286701209,
      "grad_norm": 0.4833207428455353,
      "learning_rate": 9.95584253851974e-05,
      "loss": 0.0514,
      "step": 450
    },
    {
      "epoch": 0.690846286701209,
      "eval_loss": 0.047262296080589294,
      "eval_runtime": 157.2723,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 450
    },
    {
      "epoch": 0.7061984264056803,
      "grad_norm": 0.20673757791519165,
      "learning_rate": 9.948445940176243e-05,
      "loss": 0.0515,
      "step": 460
    },
    {
      "epoch": 0.7215505661101516,
      "grad_norm": 0.23464085161685944,
      "learning_rate": 9.940480062170679e-05,
      "loss": 0.0435,
      "step": 470
    },
    {
      "epoch": 0.7369027058146229,
      "grad_norm": 0.3727859854698181,
      "learning_rate": 9.931945820914462e-05,
      "loss": 0.0544,
      "step": 480
    },
    {
      "epoch": 0.7522548455190943,
      "grad_norm": 0.4665268361568451,
      "learning_rate": 9.922844198204715e-05,
      "loss": 0.0407,
      "step": 490
    },
    {
      "epoch": 0.7676069852235655,
      "grad_norm": 0.4792843461036682,
      "learning_rate": 9.913176241111319e-05,
      "loss": 0.042,
      "step": 500
    },
    {
      "epoch": 0.7676069852235655,
      "eval_loss": 0.04632759094238281,
      "eval_runtime": 157.3047,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 500
    },
    {
      "epoch": 0.7829591249280369,
      "grad_norm": 0.6618539094924927,
      "learning_rate": 9.902943061856456e-05,
      "loss": 0.0374,
      "step": 510
    },
    {
      "epoch": 0.7983112646325081,
      "grad_norm": 0.3555477559566498,
      "learning_rate": 9.892145837686657e-05,
      "loss": 0.0651,
      "step": 520
    },
    {
      "epoch": 0.8136634043369795,
      "grad_norm": 0.46410855650901794,
      "learning_rate": 9.880785810737378e-05,
      "loss": 0.0585,
      "step": 530
    },
    {
      "epoch": 0.8290155440414507,
      "grad_norm": 0.20220302045345306,
      "learning_rate": 9.868864287890083e-05,
      "loss": 0.0438,
      "step": 540
    },
    {
      "epoch": 0.8443676837459221,
      "grad_norm": 0.275446355342865,
      "learning_rate": 9.856382640621917e-05,
      "loss": 0.0672,
      "step": 550
    },
    {
      "epoch": 0.8443676837459221,
      "eval_loss": 0.04763461649417877,
      "eval_runtime": 157.2892,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 550
    },
    {
      "epoch": 0.8597198234503934,
      "grad_norm": 0.31557002663612366,
      "learning_rate": 9.84334230484792e-05,
      "loss": 0.0402,
      "step": 560
    },
    {
      "epoch": 0.8750719631548647,
      "grad_norm": 0.22945792973041534,
      "learning_rate": 9.82974478075583e-05,
      "loss": 0.0381,
      "step": 570
    },
    {
      "epoch": 0.890424102859336,
      "grad_norm": 0.28530406951904297,
      "learning_rate": 9.815591632633509e-05,
      "loss": 0.0354,
      "step": 580
    },
    {
      "epoch": 0.9057762425638073,
      "grad_norm": 0.3434598743915558,
      "learning_rate": 9.800884488688985e-05,
      "loss": 0.0443,
      "step": 590
    },
    {
      "epoch": 0.9211283822682786,
      "grad_norm": 0.33944979310035706,
      "learning_rate": 9.785625040863124e-05,
      "loss": 0.0388,
      "step": 600
    },
    {
      "epoch": 0.9211283822682786,
      "eval_loss": 0.04456296190619469,
      "eval_runtime": 157.234,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 600
    },
    {
      "epoch": 0.93648052197275,
      "grad_norm": 0.3150840401649475,
      "learning_rate": 9.769815044635005e-05,
      "loss": 0.0352,
      "step": 610
    },
    {
      "epoch": 0.9518326616772212,
      "grad_norm": 0.35916006565093994,
      "learning_rate": 9.753456318819946e-05,
      "loss": 0.0485,
      "step": 620
    },
    {
      "epoch": 0.9671848013816926,
      "grad_norm": 0.5761755704879761,
      "learning_rate": 9.736550745360292e-05,
      "loss": 0.0413,
      "step": 630
    },
    {
      "epoch": 0.9825369410861639,
      "grad_norm": 0.2672175467014313,
      "learning_rate": 9.719100269108872e-05,
      "loss": 0.0399,
      "step": 640
    },
    {
      "epoch": 0.9978890807906352,
      "grad_norm": 0.2594519853591919,
      "learning_rate": 9.701106897605304e-05,
      "loss": 0.0391,
      "step": 650
    },
    {
      "epoch": 0.9978890807906352,
      "eval_loss": 0.04029757156968117,
      "eval_runtime": 157.2496,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 650
    },
    {
      "epoch": 1.0132412204951065,
      "grad_norm": 0.3294517397880554,
      "learning_rate": 9.682572700845006e-05,
      "loss": 0.032,
      "step": 660
    },
    {
      "epoch": 1.0285933601995778,
      "grad_norm": 0.3584151566028595,
      "learning_rate": 9.663499811041082e-05,
      "loss": 0.0361,
      "step": 670
    },
    {
      "epoch": 1.0439454999040492,
      "grad_norm": 0.24276219308376312,
      "learning_rate": 9.643890422379018e-05,
      "loss": 0.0341,
      "step": 680
    },
    {
      "epoch": 1.0592976396085205,
      "grad_norm": 0.25595176219940186,
      "learning_rate": 9.623746790764261e-05,
      "loss": 0.0323,
      "step": 690
    },
    {
      "epoch": 1.0746497793129917,
      "grad_norm": 0.33263134956359863,
      "learning_rate": 9.603071233562695e-05,
      "loss": 0.0304,
      "step": 700
    },
    {
      "epoch": 1.0746497793129917,
      "eval_loss": 0.042355190962553024,
      "eval_runtime": 157.2545,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 700
    },
    {
      "epoch": 1.090001919017463,
      "grad_norm": 0.11574736982584,
      "learning_rate": 9.581866129334044e-05,
      "loss": 0.0416,
      "step": 710
    },
    {
      "epoch": 1.1053540587219344,
      "grad_norm": 0.304192453622818,
      "learning_rate": 9.560133917558242e-05,
      "loss": 0.048,
      "step": 720
    },
    {
      "epoch": 1.1207061984264057,
      "grad_norm": 0.21460862457752228,
      "learning_rate": 9.537877098354786e-05,
      "loss": 0.0355,
      "step": 730
    },
    {
      "epoch": 1.136058338130877,
      "grad_norm": 0.3124927878379822,
      "learning_rate": 9.51509823219512e-05,
      "loss": 0.0301,
      "step": 740
    },
    {
      "epoch": 1.1514104778353482,
      "grad_norm": 0.3640657067298889,
      "learning_rate": 9.491799939608065e-05,
      "loss": 0.0396,
      "step": 750
    },
    {
      "epoch": 1.1514104778353482,
      "eval_loss": 0.042958181351423264,
      "eval_runtime": 157.2635,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 750
    },
    {
      "epoch": 1.1667626175398196,
      "grad_norm": 0.33257851004600525,
      "learning_rate": 9.467984900878364e-05,
      "loss": 0.0336,
      "step": 760
    },
    {
      "epoch": 1.182114757244291,
      "grad_norm": 0.21673721075057983,
      "learning_rate": 9.443655855738321e-05,
      "loss": 0.029,
      "step": 770
    },
    {
      "epoch": 1.1974668969487623,
      "grad_norm": 0.2433970421552658,
      "learning_rate": 9.41881560305262e-05,
      "loss": 0.0335,
      "step": 780
    },
    {
      "epoch": 1.2128190366532334,
      "grad_norm": 0.3440137207508087,
      "learning_rate": 9.393467000496344e-05,
      "loss": 0.0409,
      "step": 790
    },
    {
      "epoch": 1.2281711763577048,
      "grad_norm": 0.2682274580001831,
      "learning_rate": 9.367612964226218e-05,
      "loss": 0.0347,
      "step": 800
    },
    {
      "epoch": 1.2281711763577048,
      "eval_loss": 0.039533697068691254,
      "eval_runtime": 157.2692,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 800
    },
    {
      "epoch": 1.2435233160621761,
      "grad_norm": 0.12573248147964478,
      "learning_rate": 9.341256468545122e-05,
      "loss": 0.0246,
      "step": 810
    },
    {
      "epoch": 1.2588754557666475,
      "grad_norm": 0.5845211744308472,
      "learning_rate": 9.314400545559934e-05,
      "loss": 0.0287,
      "step": 820
    },
    {
      "epoch": 1.2742275954711189,
      "grad_norm": 0.43409937620162964,
      "learning_rate": 9.287048284832698e-05,
      "loss": 0.0358,
      "step": 830
    },
    {
      "epoch": 1.28957973517559,
      "grad_norm": 0.47061261534690857,
      "learning_rate": 9.2592028330252e-05,
      "loss": 0.0339,
      "step": 840
    },
    {
      "epoch": 1.3049318748800613,
      "grad_norm": 0.21692904829978943,
      "learning_rate": 9.230867393536972e-05,
      "loss": 0.0356,
      "step": 850
    },
    {
      "epoch": 1.3049318748800613,
      "eval_loss": 0.04097624123096466,
      "eval_runtime": 157.2932,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 850
    },
    {
      "epoch": 1.3202840145845327,
      "grad_norm": 0.4102625846862793,
      "learning_rate": 9.202045226136757e-05,
      "loss": 0.0315,
      "step": 860
    },
    {
      "epoch": 1.335636154289004,
      "grad_norm": 0.10215727984905243,
      "learning_rate": 9.172739646587509e-05,
      "loss": 0.0242,
      "step": 870
    },
    {
      "epoch": 1.3509882939934754,
      "grad_norm": 0.14424508810043335,
      "learning_rate": 9.142954026264931e-05,
      "loss": 0.0313,
      "step": 880
    },
    {
      "epoch": 1.3663404336979466,
      "grad_norm": 0.21271869540214539,
      "learning_rate": 9.112691791769634e-05,
      "loss": 0.0296,
      "step": 890
    },
    {
      "epoch": 1.381692573402418,
      "grad_norm": 0.39837774634361267,
      "learning_rate": 9.081956424532926e-05,
      "loss": 0.0333,
      "step": 900
    },
    {
      "epoch": 1.381692573402418,
      "eval_loss": 0.03842702880501747,
      "eval_runtime": 157.3229,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 900
    },
    {
      "epoch": 1.3970447131068893,
      "grad_norm": 0.22184543311595917,
      "learning_rate": 9.050751460416305e-05,
      "loss": 0.0334,
      "step": 910
    },
    {
      "epoch": 1.4123968528113606,
      "grad_norm": 0.191727876663208,
      "learning_rate": 9.019080489304685e-05,
      "loss": 0.0367,
      "step": 920
    },
    {
      "epoch": 1.427748992515832,
      "grad_norm": 0.39098262786865234,
      "learning_rate": 8.986947154693408e-05,
      "loss": 0.0487,
      "step": 930
    },
    {
      "epoch": 1.4431011322203031,
      "grad_norm": 0.3450808823108673,
      "learning_rate": 8.954355153269088e-05,
      "loss": 0.0427,
      "step": 940
    },
    {
      "epoch": 1.4584532719247745,
      "grad_norm": 0.1870778352022171,
      "learning_rate": 8.921308234484336e-05,
      "loss": 0.0318,
      "step": 950
    },
    {
      "epoch": 1.4584532719247745,
      "eval_loss": 0.037589143961668015,
      "eval_runtime": 157.2893,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 950
    },
    {
      "epoch": 1.4738054116292458,
      "grad_norm": 0.18667995929718018,
      "learning_rate": 8.887810200126419e-05,
      "loss": 0.0352,
      "step": 960
    },
    {
      "epoch": 1.4891575513337172,
      "grad_norm": 0.40204545855522156,
      "learning_rate": 8.853864903879889e-05,
      "loss": 0.0383,
      "step": 970
    },
    {
      "epoch": 1.5045096910381885,
      "grad_norm": 0.15286822617053986,
      "learning_rate": 8.81947625088325e-05,
      "loss": 0.0345,
      "step": 980
    },
    {
      "epoch": 1.5198618307426597,
      "grad_norm": 0.17312301695346832,
      "learning_rate": 8.784648197279701e-05,
      "loss": 0.032,
      "step": 990
    },
    {
      "epoch": 1.535213970447131,
      "grad_norm": 0.31317198276519775,
      "learning_rate": 8.749384749762015e-05,
      "loss": 0.0309,
      "step": 1000
    },
    {
      "epoch": 1.535213970447131,
      "eval_loss": 0.03832419961690903,
      "eval_runtime": 157.292,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 1000
    },
    {
      "epoch": 1.5505661101516024,
      "grad_norm": 0.2550368010997772,
      "learning_rate": 8.713689965111602e-05,
      "loss": 0.0246,
      "step": 1010
    },
    {
      "epoch": 1.5659182498560735,
      "grad_norm": 0.2712564468383789,
      "learning_rate": 8.677567949731801e-05,
      "loss": 0.0372,
      "step": 1020
    },
    {
      "epoch": 1.581270389560545,
      "grad_norm": 0.3656397759914398,
      "learning_rate": 8.64102285917548e-05,
      "loss": 0.0502,
      "step": 1030
    },
    {
      "epoch": 1.5966225292650162,
      "grad_norm": 0.1517850011587143,
      "learning_rate": 8.604058897666962e-05,
      "loss": 0.0333,
      "step": 1040
    },
    {
      "epoch": 1.6119746689694876,
      "grad_norm": 0.24108999967575073,
      "learning_rate": 8.566680317618377e-05,
      "loss": 0.0216,
      "step": 1050
    },
    {
      "epoch": 1.6119746689694876,
      "eval_loss": 0.037316225469112396,
      "eval_runtime": 157.2218,
      "eval_samples_per_second": 7.372,
      "eval_steps_per_second": 7.372,
      "step": 1050
    },
    {
      "epoch": 1.627326808673959,
      "grad_norm": 0.21492698788642883,
      "learning_rate": 8.528891419140438e-05,
      "loss": 0.0387,
      "step": 1060
    },
    {
      "epoch": 1.64267894837843,
      "grad_norm": 0.2957633137702942,
      "learning_rate": 8.490696549547761e-05,
      "loss": 0.0272,
      "step": 1070
    },
    {
      "epoch": 1.6580310880829017,
      "grad_norm": 0.3371316194534302,
      "learning_rate": 8.452100102858734e-05,
      "loss": 0.0401,
      "step": 1080
    },
    {
      "epoch": 1.6733832277873728,
      "grad_norm": 0.28173157572746277,
      "learning_rate": 8.413106519290023e-05,
      "loss": 0.0322,
      "step": 1090
    },
    {
      "epoch": 1.6887353674918442,
      "grad_norm": 0.2843840718269348,
      "learning_rate": 8.373720284745757e-05,
      "loss": 0.0315,
      "step": 1100
    },
    {
      "epoch": 1.6887353674918442,
      "eval_loss": 0.03704118728637695,
      "eval_runtime": 157.1462,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1100
    },
    {
      "epoch": 1.7040875071963155,
      "grad_norm": 0.18807817995548248,
      "learning_rate": 8.333945930301459e-05,
      "loss": 0.0312,
      "step": 1110
    },
    {
      "epoch": 1.7194396469007867,
      "grad_norm": 0.199060320854187,
      "learning_rate": 8.293788031682789e-05,
      "loss": 0.0305,
      "step": 1120
    },
    {
      "epoch": 1.7347917866052582,
      "grad_norm": 0.22268791496753693,
      "learning_rate": 8.253251208739137e-05,
      "loss": 0.0303,
      "step": 1130
    },
    {
      "epoch": 1.7501439263097294,
      "grad_norm": 0.3476749658584595,
      "learning_rate": 8.21234012491215e-05,
      "loss": 0.0292,
      "step": 1140
    },
    {
      "epoch": 1.7654960660142007,
      "grad_norm": 0.32021111249923706,
      "learning_rate": 8.171059486699224e-05,
      "loss": 0.0273,
      "step": 1150
    },
    {
      "epoch": 1.7654960660142007,
      "eval_loss": 0.03893882781267166,
      "eval_runtime": 157.1622,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1150
    },
    {
      "epoch": 1.780848205718672,
      "grad_norm": 0.20445989072322845,
      "learning_rate": 8.129414043112087e-05,
      "loss": 0.035,
      "step": 1160
    },
    {
      "epoch": 1.7962003454231432,
      "grad_norm": 0.2090442031621933,
      "learning_rate": 8.087408585130438e-05,
      "loss": 0.0283,
      "step": 1170
    },
    {
      "epoch": 1.8115524851276148,
      "grad_norm": 0.16279439628124237,
      "learning_rate": 8.045047945150796e-05,
      "loss": 0.0348,
      "step": 1180
    },
    {
      "epoch": 1.826904624832086,
      "grad_norm": 0.277883380651474,
      "learning_rate": 8.002336996430561e-05,
      "loss": 0.0309,
      "step": 1190
    },
    {
      "epoch": 1.8422567645365573,
      "grad_norm": 0.10432474315166473,
      "learning_rate": 7.959280652527394e-05,
      "loss": 0.0188,
      "step": 1200
    },
    {
      "epoch": 1.8422567645365573,
      "eval_loss": 0.03792406991124153,
      "eval_runtime": 157.189,
      "eval_samples_per_second": 7.373,
      "eval_steps_per_second": 7.373,
      "step": 1200
    },
    {
      "epoch": 1.8576089042410286,
      "grad_norm": 0.1943598836660385,
      "learning_rate": 7.915883866733946e-05,
      "loss": 0.0348,
      "step": 1210
    },
    {
      "epoch": 1.8729610439454998,
      "grad_norm": 0.36124810576438904,
      "learning_rate": 7.872151631508022e-05,
      "loss": 0.0311,
      "step": 1220
    },
    {
      "epoch": 1.8883131836499714,
      "grad_norm": 0.24344752728939056,
      "learning_rate": 7.828088977898234e-05,
      "loss": 0.0361,
      "step": 1230
    },
    {
      "epoch": 1.9036653233544425,
      "grad_norm": 0.2729787826538086,
      "learning_rate": 7.783700974965225e-05,
      "loss": 0.043,
      "step": 1240
    },
    {
      "epoch": 1.9190174630589139,
      "grad_norm": 0.20694006979465485,
      "learning_rate": 7.738992729198511e-05,
      "loss": 0.0362,
      "step": 1250
    },
    {
      "epoch": 1.9190174630589139,
      "eval_loss": 0.037059471011161804,
      "eval_runtime": 157.1426,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1250
    },
    {
      "epoch": 1.9343696027633852,
      "grad_norm": 0.13363595306873322,
      "learning_rate": 7.693969383929017e-05,
      "loss": 0.0247,
      "step": 1260
    },
    {
      "epoch": 1.9497217424678563,
      "grad_norm": 0.4264666438102722,
      "learning_rate": 7.648636118737385e-05,
      "loss": 0.0375,
      "step": 1270
    },
    {
      "epoch": 1.965073882172328,
      "grad_norm": 0.2813757359981537,
      "learning_rate": 7.602998148858089e-05,
      "loss": 0.0302,
      "step": 1280
    },
    {
      "epoch": 1.980426021876799,
      "grad_norm": 0.3236190974712372,
      "learning_rate": 7.557060724579484e-05,
      "loss": 0.0325,
      "step": 1290
    },
    {
      "epoch": 1.9957781615812704,
      "grad_norm": 0.4948330223560333,
      "learning_rate": 7.51082913063978e-05,
      "loss": 0.0435,
      "step": 1300
    },
    {
      "epoch": 1.9957781615812704,
      "eval_loss": 0.036797065287828445,
      "eval_runtime": 157.0855,
      "eval_samples_per_second": 7.378,
      "eval_steps_per_second": 7.378,
      "step": 1300
    },
    {
      "epoch": 2.0111303012857418,
      "grad_norm": 0.22444044053554535,
      "learning_rate": 7.464308685619099e-05,
      "loss": 0.0272,
      "step": 1310
    },
    {
      "epoch": 2.026482440990213,
      "grad_norm": 0.3434648811817169,
      "learning_rate": 7.417504741327587e-05,
      "loss": 0.0168,
      "step": 1320
    },
    {
      "epoch": 2.0418345806946845,
      "grad_norm": 0.4413999617099762,
      "learning_rate": 7.370422682189755e-05,
      "loss": 0.0209,
      "step": 1330
    },
    {
      "epoch": 2.0571867203991556,
      "grad_norm": 0.16911853849887848,
      "learning_rate": 7.323067924625024e-05,
      "loss": 0.0172,
      "step": 1340
    },
    {
      "epoch": 2.0725388601036268,
      "grad_norm": 0.21051056683063507,
      "learning_rate": 7.275445916424627e-05,
      "loss": 0.0212,
      "step": 1350
    },
    {
      "epoch": 2.0725388601036268,
      "eval_loss": 0.03962073475122452,
      "eval_runtime": 157.106,
      "eval_samples_per_second": 7.377,
      "eval_steps_per_second": 7.377,
      "step": 1350
    },
    {
      "epoch": 2.0878909998080983,
      "grad_norm": 0.4296096861362457,
      "learning_rate": 7.227562136124864e-05,
      "loss": 0.0179,
      "step": 1360
    },
    {
      "epoch": 2.1032431395125695,
      "grad_norm": 0.23325330018997192,
      "learning_rate": 7.179422092376856e-05,
      "loss": 0.02,
      "step": 1370
    },
    {
      "epoch": 2.118595279217041,
      "grad_norm": 0.4717545211315155,
      "learning_rate": 7.13103132331281e-05,
      "loss": 0.0223,
      "step": 1380
    },
    {
      "epoch": 2.133947418921512,
      "grad_norm": 0.42749759554862976,
      "learning_rate": 7.082395395908903e-05,
      "loss": 0.0249,
      "step": 1390
    },
    {
      "epoch": 2.1492995586259833,
      "grad_norm": 0.20575480163097382,
      "learning_rate": 7.033519905344846e-05,
      "loss": 0.0184,
      "step": 1400
    },
    {
      "epoch": 2.1492995586259833,
      "eval_loss": 0.03931669518351555,
      "eval_runtime": 157.1133,
      "eval_samples_per_second": 7.377,
      "eval_steps_per_second": 7.377,
      "step": 1400
    },
    {
      "epoch": 2.164651698330455,
      "grad_norm": 0.22173267602920532,
      "learning_rate": 6.984410474360195e-05,
      "loss": 0.0243,
      "step": 1410
    },
    {
      "epoch": 2.180003838034926,
      "grad_norm": 0.2206946164369583,
      "learning_rate": 6.935072752607511e-05,
      "loss": 0.0271,
      "step": 1420
    },
    {
      "epoch": 2.1953559777393976,
      "grad_norm": 0.21722865104675293,
      "learning_rate": 6.885512416002412e-05,
      "loss": 0.0308,
      "step": 1430
    },
    {
      "epoch": 2.2107081174438687,
      "grad_norm": 0.264813631772995,
      "learning_rate": 6.835735166070587e-05,
      "loss": 0.0205,
      "step": 1440
    },
    {
      "epoch": 2.22606025714834,
      "grad_norm": 0.23824357986450195,
      "learning_rate": 6.785746729291897e-05,
      "loss": 0.0205,
      "step": 1450
    },
    {
      "epoch": 2.22606025714834,
      "eval_loss": 0.03809254616498947,
      "eval_runtime": 157.1629,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1450
    },
    {
      "epoch": 2.2414123968528115,
      "grad_norm": 0.23905223608016968,
      "learning_rate": 6.735552856441585e-05,
      "loss": 0.0232,
      "step": 1460
    },
    {
      "epoch": 2.2567645365572826,
      "grad_norm": 0.32694312930107117,
      "learning_rate": 6.685159321928691e-05,
      "loss": 0.0216,
      "step": 1470
    },
    {
      "epoch": 2.272116676261754,
      "grad_norm": 0.2926463186740875,
      "learning_rate": 6.634571923131756e-05,
      "loss": 0.0239,
      "step": 1480
    },
    {
      "epoch": 2.2874688159662253,
      "grad_norm": 0.154023677110672,
      "learning_rate": 6.583796479731872e-05,
      "loss": 0.0175,
      "step": 1490
    },
    {
      "epoch": 2.3028209556706964,
      "grad_norm": 0.10085402429103851,
      "learning_rate": 6.532838833043189e-05,
      "loss": 0.0316,
      "step": 1500
    },
    {
      "epoch": 2.3028209556706964,
      "eval_loss": 0.03903723508119583,
      "eval_runtime": 157.1873,
      "eval_samples_per_second": 7.373,
      "eval_steps_per_second": 7.373,
      "step": 1500
    },
    {
      "epoch": 2.318173095375168,
      "grad_norm": 0.36862310767173767,
      "learning_rate": 6.481704845340894e-05,
      "loss": 0.025,
      "step": 1510
    },
    {
      "epoch": 2.333525235079639,
      "grad_norm": 0.3234424591064453,
      "learning_rate": 6.43040039918683e-05,
      "loss": 0.0219,
      "step": 1520
    },
    {
      "epoch": 2.3488773747841103,
      "grad_norm": 0.5369458198547363,
      "learning_rate": 6.37893139675273e-05,
      "loss": 0.019,
      "step": 1530
    },
    {
      "epoch": 2.364229514488582,
      "grad_norm": 0.31320443749427795,
      "learning_rate": 6.327303759141235e-05,
      "loss": 0.0188,
      "step": 1540
    },
    {
      "epoch": 2.379581654193053,
      "grad_norm": 0.35849952697753906,
      "learning_rate": 6.275523425704707e-05,
      "loss": 0.0234,
      "step": 1550
    },
    {
      "epoch": 2.379581654193053,
      "eval_loss": 0.03996282070875168,
      "eval_runtime": 157.1594,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "step": 1550
    },
    {
      "epoch": 2.3949337938975246,
      "grad_norm": 0.32893258333206177,
      "learning_rate": 6.223596353361961e-05,
      "loss": 0.0226,
      "step": 1560
    },
    {
      "epoch": 2.4102859336019957,
      "grad_norm": 0.03046131506562233,
      "learning_rate": 6.171528515912965e-05,
      "loss": 0.0162,
      "step": 1570
    },
    {
      "epoch": 2.425638073306467,
      "grad_norm": 0.3565883934497833,
      "learning_rate": 6.119325903351599e-05,
      "loss": 0.0253,
      "step": 1580
    },
    {
      "epoch": 2.4409902130109384,
      "grad_norm": 0.3316241204738617,
      "learning_rate": 6.0669945211765585e-05,
      "loss": 0.0312,
      "step": 1590
    },
    {
      "epoch": 2.4563423527154096,
      "grad_norm": 0.338368684053421,
      "learning_rate": 6.0145403897004696e-05,
      "loss": 0.0258,
      "step": 1600
    },
    {
      "epoch": 2.4563423527154096,
      "eval_loss": 0.038748133927583694,
      "eval_runtime": 157.2369,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 1600
    },
    {
      "epoch": 2.471694492419881,
      "grad_norm": 0.45592939853668213,
      "learning_rate": 5.961969543357292e-05,
      "loss": 0.0247,
      "step": 1610
    },
    {
      "epoch": 2.4870466321243523,
      "grad_norm": 0.17753665149211884,
      "learning_rate": 5.9092880300081123e-05,
      "loss": 0.0277,
      "step": 1620
    },
    {
      "epoch": 2.5023987718288234,
      "grad_norm": 0.4086242914199829,
      "learning_rate": 5.8565019102453844e-05,
      "loss": 0.0237,
      "step": 1630
    },
    {
      "epoch": 2.517750911533295,
      "grad_norm": 0.24414750933647156,
      "learning_rate": 5.8036172566957006e-05,
      "loss": 0.0231,
      "step": 1640
    },
    {
      "epoch": 2.533103051237766,
      "grad_norm": 0.44709134101867676,
      "learning_rate": 5.750640153321194e-05,
      "loss": 0.0273,
      "step": 1650
    },
    {
      "epoch": 2.533103051237766,
      "eval_loss": 0.0392879992723465,
      "eval_runtime": 157.2456,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 1650
    },
    {
      "epoch": 2.5484551909422377,
      "grad_norm": 0.3048468232154846,
      "learning_rate": 5.697576694719616e-05,
      "loss": 0.0215,
      "step": 1660
    },
    {
      "epoch": 2.563807330646709,
      "grad_norm": 0.22297337651252747,
      "learning_rate": 5.644432985423206e-05,
      "loss": 0.0229,
      "step": 1670
    },
    {
      "epoch": 2.57915947035118,
      "grad_norm": 0.3140508830547333,
      "learning_rate": 5.591215139196414e-05,
      "loss": 0.0242,
      "step": 1680
    },
    {
      "epoch": 2.5945116100556516,
      "grad_norm": 0.22138510644435883,
      "learning_rate": 5.5379292783325585e-05,
      "loss": 0.0286,
      "step": 1690
    },
    {
      "epoch": 2.6098637497601227,
      "grad_norm": 0.3354733884334564,
      "learning_rate": 5.4845815329495054e-05,
      "loss": 0.0199,
      "step": 1700
    },
    {
      "epoch": 2.6098637497601227,
      "eval_loss": 0.03853138908743858,
      "eval_runtime": 157.254,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 1700
    },
    {
      "epoch": 2.6252158894645943,
      "grad_norm": 0.283342182636261,
      "learning_rate": 5.431178040284446e-05,
      "loss": 0.0223,
      "step": 1710
    },
    {
      "epoch": 2.6405680291690654,
      "grad_norm": 1.3571743965148926,
      "learning_rate": 5.377724943987855e-05,
      "loss": 0.0181,
      "step": 1720
    },
    {
      "epoch": 2.6559201688735365,
      "grad_norm": 0.10672520101070404,
      "learning_rate": 5.324228393416718e-05,
      "loss": 0.0181,
      "step": 1730
    },
    {
      "epoch": 2.671272308578008,
      "grad_norm": 0.5193530321121216,
      "learning_rate": 5.270694542927088e-05,
      "loss": 0.0257,
      "step": 1740
    },
    {
      "epoch": 2.6866244482824793,
      "grad_norm": 0.22108089923858643,
      "learning_rate": 5.21712955116608e-05,
      "loss": 0.0167,
      "step": 1750
    },
    {
      "epoch": 2.6866244482824793,
      "eval_loss": 0.03936908766627312,
      "eval_runtime": 157.2427,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 1750
    },
    {
      "epoch": 2.701976587986951,
      "grad_norm": 0.09635429084300995,
      "learning_rate": 5.1635395803633666e-05,
      "loss": 0.0197,
      "step": 1760
    },
    {
      "epoch": 2.717328727691422,
      "grad_norm": 0.14154723286628723,
      "learning_rate": 5.109930795622265e-05,
      "loss": 0.0153,
      "step": 1770
    },
    {
      "epoch": 2.732680867395893,
      "grad_norm": 0.19964846968650818,
      "learning_rate": 5.056309364210483e-05,
      "loss": 0.0191,
      "step": 1780
    },
    {
      "epoch": 2.7480330071003647,
      "grad_norm": 0.32814568281173706,
      "learning_rate": 5.002681454850632e-05,
      "loss": 0.024,
      "step": 1790
    },
    {
      "epoch": 2.763385146804836,
      "grad_norm": 0.3342536985874176,
      "learning_rate": 4.949053237010554e-05,
      "loss": 0.0288,
      "step": 1800
    },
    {
      "epoch": 2.763385146804836,
      "eval_loss": 0.042679354548454285,
      "eval_runtime": 157.2517,
      "eval_samples_per_second": 7.37,
      "eval_steps_per_second": 7.37,
      "step": 1800
    },
    {
      "epoch": 2.7787372865093074,
      "grad_norm": 0.3160741329193115,
      "learning_rate": 4.89543088019359e-05,
      "loss": 0.0172,
      "step": 1810
    },
    {
      "epoch": 2.7940894262137785,
      "grad_norm": 0.3922727108001709,
      "learning_rate": 4.841820553228805e-05,
      "loss": 0.0284,
      "step": 1820
    },
    {
      "epoch": 2.8094415659182497,
      "grad_norm": 0.3402051031589508,
      "learning_rate": 4.7882284235613324e-05,
      "loss": 0.0281,
      "step": 1830
    },
    {
      "epoch": 2.8247937056227213,
      "grad_norm": 0.25749653577804565,
      "learning_rate": 4.734660656542846e-05,
      "loss": 0.0231,
      "step": 1840
    },
    {
      "epoch": 2.8401458453271924,
      "grad_norm": 0.2327001392841339,
      "learning_rate": 4.681123414722291e-05,
      "loss": 0.022,
      "step": 1850
    },
    {
      "epoch": 2.8401458453271924,
      "eval_loss": 0.03755514323711395,
      "eval_runtime": 157.276,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 1850
    },
    {
      "epoch": 2.855497985031664,
      "grad_norm": 0.2501806616783142,
      "learning_rate": 4.627622857136929e-05,
      "loss": 0.0193,
      "step": 1860
    },
    {
      "epoch": 2.870850124736135,
      "grad_norm": 0.7793471217155457,
      "learning_rate": 4.5741651386037883e-05,
      "loss": 0.0298,
      "step": 1870
    },
    {
      "epoch": 2.8862022644406062,
      "grad_norm": 0.41012799739837646,
      "learning_rate": 4.5207564090116064e-05,
      "loss": 0.0183,
      "step": 1880
    },
    {
      "epoch": 2.901554404145078,
      "grad_norm": 0.3157999813556671,
      "learning_rate": 4.467402812613323e-05,
      "loss": 0.0182,
      "step": 1890
    },
    {
      "epoch": 2.916906543849549,
      "grad_norm": 0.3505517542362213,
      "learning_rate": 4.414110487319244e-05,
      "loss": 0.0237,
      "step": 1900
    },
    {
      "epoch": 2.916906543849549,
      "eval_loss": 0.03842131420969963,
      "eval_runtime": 157.2876,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 1900
    },
    {
      "epoch": 2.9322586835540205,
      "grad_norm": 0.3128899931907654,
      "learning_rate": 4.360885563990919e-05,
      "loss": 0.0226,
      "step": 1910
    },
    {
      "epoch": 2.9476108232584917,
      "grad_norm": 0.2808375656604767,
      "learning_rate": 4.307734165735829e-05,
      "loss": 0.0165,
      "step": 1920
    },
    {
      "epoch": 2.962962962962963,
      "grad_norm": 0.22949250042438507,
      "learning_rate": 4.254662407202976e-05,
      "loss": 0.022,
      "step": 1930
    },
    {
      "epoch": 2.9783151026674344,
      "grad_norm": 0.3448737859725952,
      "learning_rate": 4.201676393879446e-05,
      "loss": 0.021,
      "step": 1940
    },
    {
      "epoch": 2.9936672423719055,
      "grad_norm": 0.13228139281272888,
      "learning_rate": 4.148782221388007e-05,
      "loss": 0.0176,
      "step": 1950
    },
    {
      "epoch": 2.9936672423719055,
      "eval_loss": 0.03724433481693268,
      "eval_runtime": 157.321,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 1950
    },
    {
      "epoch": 3.009019382076377,
      "grad_norm": 0.19390782713890076,
      "learning_rate": 4.0959859747858706e-05,
      "loss": 0.0113,
      "step": 1960
    },
    {
      "epoch": 3.0243715217808482,
      "grad_norm": 0.4046991467475891,
      "learning_rate": 4.043293727864644e-05,
      "loss": 0.0108,
      "step": 1970
    },
    {
      "epoch": 3.0397236614853194,
      "grad_norm": 0.3703497052192688,
      "learning_rate": 3.990711542451591e-05,
      "loss": 0.0097,
      "step": 1980
    },
    {
      "epoch": 3.055075801189791,
      "grad_norm": 0.2644294202327728,
      "learning_rate": 3.9382454677122704e-05,
      "loss": 0.0127,
      "step": 1990
    },
    {
      "epoch": 3.070427940894262,
      "grad_norm": 0.15409328043460846,
      "learning_rate": 3.885901539454623e-05,
      "loss": 0.0057,
      "step": 2000
    },
    {
      "epoch": 3.070427940894262,
      "eval_loss": 0.04431070759892464,
      "eval_runtime": 157.3002,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 2000
    },
    {
      "epoch": 3.0857800805987337,
      "grad_norm": 0.4345476031303406,
      "learning_rate": 3.833685779434597e-05,
      "loss": 0.0172,
      "step": 2010
    },
    {
      "epoch": 3.101132220303205,
      "grad_norm": 0.23631352186203003,
      "learning_rate": 3.7816041946634024e-05,
      "loss": 0.0111,
      "step": 2020
    },
    {
      "epoch": 3.116484360007676,
      "grad_norm": 0.4313735365867615,
      "learning_rate": 3.729662776716439e-05,
      "loss": 0.0126,
      "step": 2030
    },
    {
      "epoch": 3.1318364997121475,
      "grad_norm": 0.13873319327831268,
      "learning_rate": 3.677867501044029e-05,
      "loss": 0.0059,
      "step": 2040
    },
    {
      "epoch": 3.1471886394166186,
      "grad_norm": 0.2149595320224762,
      "learning_rate": 3.6262243262839654e-05,
      "loss": 0.0148,
      "step": 2050
    },
    {
      "epoch": 3.1471886394166186,
      "eval_loss": 0.04357267543673515,
      "eval_runtime": 157.232,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 2050
    },
    {
      "epoch": 3.16254077912109,
      "grad_norm": 0.07435044646263123,
      "learning_rate": 3.574739193576042e-05,
      "loss": 0.007,
      "step": 2060
    },
    {
      "epoch": 3.1778929188255614,
      "grad_norm": 0.41588830947875977,
      "learning_rate": 3.5234180258785554e-05,
      "loss": 0.0091,
      "step": 2070
    },
    {
      "epoch": 3.1932450585300325,
      "grad_norm": 0.038579829037189484,
      "learning_rate": 3.472266727286928e-05,
      "loss": 0.0073,
      "step": 2080
    },
    {
      "epoch": 3.208597198234504,
      "grad_norm": 0.15435343980789185,
      "learning_rate": 3.4212911823544746e-05,
      "loss": 0.0123,
      "step": 2090
    },
    {
      "epoch": 3.223949337938975,
      "grad_norm": 3.7131805419921875,
      "learning_rate": 3.370497255415443e-05,
      "loss": 0.0153,
      "step": 2100
    },
    {
      "epoch": 3.223949337938975,
      "eval_loss": 0.04993755370378494,
      "eval_runtime": 157.2762,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 2100
    },
    {
      "epoch": 3.239301477643447,
      "grad_norm": 0.7524594664573669,
      "learning_rate": 3.319890789910364e-05,
      "loss": 0.0171,
      "step": 2110
    },
    {
      "epoch": 3.254653617347918,
      "grad_norm": 0.08906435966491699,
      "learning_rate": 3.269477607713802e-05,
      "loss": 0.0155,
      "step": 2120
    },
    {
      "epoch": 3.270005757052389,
      "grad_norm": 0.5883983373641968,
      "learning_rate": 3.219263508464614e-05,
      "loss": 0.0125,
      "step": 2130
    },
    {
      "epoch": 3.2853578967568606,
      "grad_norm": 0.8002910017967224,
      "learning_rate": 3.169254268898725e-05,
      "loss": 0.0113,
      "step": 2140
    },
    {
      "epoch": 3.3007100364613318,
      "grad_norm": 0.023618843406438828,
      "learning_rate": 3.119455642184572e-05,
      "loss": 0.0116,
      "step": 2150
    },
    {
      "epoch": 3.3007100364613318,
      "eval_loss": 0.044418033212423325,
      "eval_runtime": 157.2945,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 2150
    },
    {
      "epoch": 3.3160621761658033,
      "grad_norm": 0.24659119546413422,
      "learning_rate": 3.069873357261249e-05,
      "loss": 0.0135,
      "step": 2160
    },
    {
      "epoch": 3.3314143158702745,
      "grad_norm": 0.2751818597316742,
      "learning_rate": 3.020513118179428e-05,
      "loss": 0.0156,
      "step": 2170
    },
    {
      "epoch": 3.3467664555747456,
      "grad_norm": 0.17336240410804749,
      "learning_rate": 2.9713806034451652e-05,
      "loss": 0.0107,
      "step": 2180
    },
    {
      "epoch": 3.362118595279217,
      "grad_norm": 0.03696625307202339,
      "learning_rate": 2.9224814653666242e-05,
      "loss": 0.0107,
      "step": 2190
    },
    {
      "epoch": 3.3774707349836883,
      "grad_norm": 0.14558713138103485,
      "learning_rate": 2.8738213294038212e-05,
      "loss": 0.0116,
      "step": 2200
    },
    {
      "epoch": 3.3774707349836883,
      "eval_loss": 0.04727363586425781,
      "eval_runtime": 157.2356,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 2200
    },
    {
      "epoch": 3.39282287468816,
      "grad_norm": 0.4054368734359741,
      "learning_rate": 2.8254057935214735e-05,
      "loss": 0.0118,
      "step": 2210
    },
    {
      "epoch": 3.408175014392631,
      "grad_norm": 0.3659217059612274,
      "learning_rate": 2.7772404275449825e-05,
      "loss": 0.0126,
      "step": 2220
    },
    {
      "epoch": 3.423527154097102,
      "grad_norm": 0.2420692890882492,
      "learning_rate": 2.7293307725196793e-05,
      "loss": 0.0084,
      "step": 2230
    },
    {
      "epoch": 3.4388792938015738,
      "grad_norm": 0.11345157027244568,
      "learning_rate": 2.6816823400733625e-05,
      "loss": 0.008,
      "step": 2240
    },
    {
      "epoch": 3.454231433506045,
      "grad_norm": 0.6473270654678345,
      "learning_rate": 2.6343006117822437e-05,
      "loss": 0.0133,
      "step": 2250
    },
    {
      "epoch": 3.454231433506045,
      "eval_loss": 0.05019335821270943,
      "eval_runtime": 157.2449,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 2250
    },
    {
      "epoch": 3.469583573210516,
      "grad_norm": 0.18109719455242157,
      "learning_rate": 2.587191038540317e-05,
      "loss": 0.0106,
      "step": 2260
    },
    {
      "epoch": 3.4849357129149876,
      "grad_norm": 0.05215310677886009,
      "learning_rate": 2.5403590399322886e-05,
      "loss": 0.0048,
      "step": 2270
    },
    {
      "epoch": 3.5002878526194587,
      "grad_norm": 0.39558789134025574,
      "learning_rate": 2.493810003610092e-05,
      "loss": 0.0134,
      "step": 2280
    },
    {
      "epoch": 3.51563999232393,
      "grad_norm": 0.2711808979511261,
      "learning_rate": 2.4475492846730835e-05,
      "loss": 0.0158,
      "step": 2290
    },
    {
      "epoch": 3.5309921320284015,
      "grad_norm": 0.31566399335861206,
      "learning_rate": 2.4015822050519794e-05,
      "loss": 0.0095,
      "step": 2300
    },
    {
      "epoch": 3.5309921320284015,
      "eval_loss": 0.05009521171450615,
      "eval_runtime": 157.2158,
      "eval_samples_per_second": 7.372,
      "eval_steps_per_second": 7.372,
      "step": 2300
    },
    {
      "epoch": 3.546344271732873,
      "grad_norm": 0.49842846393585205,
      "learning_rate": 2.3559140528966074e-05,
      "loss": 0.0067,
      "step": 2310
    },
    {
      "epoch": 3.561696411437344,
      "grad_norm": 0.45083844661712646,
      "learning_rate": 2.3105500819675495e-05,
      "loss": 0.0129,
      "step": 2320
    },
    {
      "epoch": 3.5770485511418153,
      "grad_norm": 0.521081268787384,
      "learning_rate": 2.265495511031742e-05,
      "loss": 0.0151,
      "step": 2330
    },
    {
      "epoch": 3.5924006908462864,
      "grad_norm": 0.4393942356109619,
      "learning_rate": 2.2207555232620893e-05,
      "loss": 0.0124,
      "step": 2340
    },
    {
      "epoch": 3.607752830550758,
      "grad_norm": 0.13999144732952118,
      "learning_rate": 2.1763352656411785e-05,
      "loss": 0.011,
      "step": 2350
    },
    {
      "epoch": 3.607752830550758,
      "eval_loss": 0.046690210700035095,
      "eval_runtime": 157.2409,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 2350
    },
    {
      "epoch": 3.6231049702552296,
      "grad_norm": 0.3251853585243225,
      "learning_rate": 2.1322398483691787e-05,
      "loss": 0.0069,
      "step": 2360
    },
    {
      "epoch": 3.6384571099597007,
      "grad_norm": 0.30056822299957275,
      "learning_rate": 2.08847434427593e-05,
      "loss": 0.0144,
      "step": 2370
    },
    {
      "epoch": 3.653809249664172,
      "grad_norm": 0.10957108438014984,
      "learning_rate": 2.0450437882373697e-05,
      "loss": 0.0115,
      "step": 2380
    },
    {
      "epoch": 3.669161389368643,
      "grad_norm": 0.2679850459098816,
      "learning_rate": 2.0019531765962995e-05,
      "loss": 0.0087,
      "step": 2390
    },
    {
      "epoch": 3.6845135290731146,
      "grad_norm": 0.5390302538871765,
      "learning_rate": 1.9592074665876026e-05,
      "loss": 0.0136,
      "step": 2400
    },
    {
      "epoch": 3.6845135290731146,
      "eval_loss": 0.05131568759679794,
      "eval_runtime": 157.2039,
      "eval_samples_per_second": 7.373,
      "eval_steps_per_second": 7.373,
      "step": 2400
    },
    {
      "epoch": 3.6998656687775857,
      "grad_norm": 0.13512513041496277,
      "learning_rate": 1.9168115757679535e-05,
      "loss": 0.0155,
      "step": 2410
    },
    {
      "epoch": 3.7152178084820573,
      "grad_norm": 0.1011972650885582,
      "learning_rate": 1.8747703814500866e-05,
      "loss": 0.0095,
      "step": 2420
    },
    {
      "epoch": 3.7305699481865284,
      "grad_norm": 0.4893917739391327,
      "learning_rate": 1.833088720141698e-05,
      "loss": 0.0099,
      "step": 2430
    },
    {
      "epoch": 3.7459220878909996,
      "grad_norm": 0.725928544998169,
      "learning_rate": 1.7917713869890557e-05,
      "loss": 0.0121,
      "step": 2440
    },
    {
      "epoch": 3.761274227595471,
      "grad_norm": 0.11228631436824799,
      "learning_rate": 1.7508231352253435e-05,
      "loss": 0.0074,
      "step": 2450
    },
    {
      "epoch": 3.761274227595471,
      "eval_loss": 0.04784177616238594,
      "eval_runtime": 157.3439,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 2450
    },
    {
      "epoch": 3.7766263672999423,
      "grad_norm": 0.11749975383281708,
      "learning_rate": 1.7102486756238435e-05,
      "loss": 0.0115,
      "step": 2460
    },
    {
      "epoch": 3.791978507004414,
      "grad_norm": 0.3067283630371094,
      "learning_rate": 1.6700526759560002e-05,
      "loss": 0.0094,
      "step": 2470
    },
    {
      "epoch": 3.807330646708885,
      "grad_norm": 0.4429052472114563,
      "learning_rate": 1.6302397604544257e-05,
      "loss": 0.0142,
      "step": 2480
    },
    {
      "epoch": 3.822682786413356,
      "grad_norm": 0.26236897706985474,
      "learning_rate": 1.5908145092809272e-05,
      "loss": 0.0074,
      "step": 2490
    },
    {
      "epoch": 3.8380349261178277,
      "grad_norm": 0.14888080954551697,
      "learning_rate": 1.551781457999586e-05,
      "loss": 0.0104,
      "step": 2500
    },
    {
      "epoch": 3.8380349261178277,
      "eval_loss": 0.04920686408877373,
      "eval_runtime": 157.3187,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2500
    },
    {
      "epoch": 3.853387065822299,
      "grad_norm": 0.155548557639122,
      "learning_rate": 1.513145097054977e-05,
      "loss": 0.0083,
      "step": 2510
    },
    {
      "epoch": 3.8687392055267704,
      "grad_norm": 0.4206964075565338,
      "learning_rate": 1.4749098712555854e-05,
      "loss": 0.0116,
      "step": 2520
    },
    {
      "epoch": 3.8840913452312416,
      "grad_norm": 0.49253013730049133,
      "learning_rate": 1.4370801792624656e-05,
      "loss": 0.0063,
      "step": 2530
    },
    {
      "epoch": 3.8994434849357127,
      "grad_norm": 0.3531017601490021,
      "learning_rate": 1.399660373083203e-05,
      "loss": 0.0098,
      "step": 2540
    },
    {
      "epoch": 3.9147956246401843,
      "grad_norm": 0.17188264429569244,
      "learning_rate": 1.3626547575712545e-05,
      "loss": 0.0131,
      "step": 2550
    },
    {
      "epoch": 3.9147956246401843,
      "eval_loss": 0.051388129591941833,
      "eval_runtime": 157.2859,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 2550
    },
    {
      "epoch": 3.9301477643446554,
      "grad_norm": 0.5859571695327759,
      "learning_rate": 1.3260675899307096e-05,
      "loss": 0.0177,
      "step": 2560
    },
    {
      "epoch": 3.945499904049127,
      "grad_norm": 0.1914072185754776,
      "learning_rate": 1.2899030792265349e-05,
      "loss": 0.0107,
      "step": 2570
    },
    {
      "epoch": 3.960852043753598,
      "grad_norm": 0.21298187971115112,
      "learning_rate": 1.2541653859003437e-05,
      "loss": 0.0124,
      "step": 2580
    },
    {
      "epoch": 3.9762041834580693,
      "grad_norm": 0.3145734667778015,
      "learning_rate": 1.2188586212917846e-05,
      "loss": 0.0121,
      "step": 2590
    },
    {
      "epoch": 3.991556323162541,
      "grad_norm": 0.27779558300971985,
      "learning_rate": 1.1839868471655523e-05,
      "loss": 0.0113,
      "step": 2600
    },
    {
      "epoch": 3.991556323162541,
      "eval_loss": 0.04836370795965195,
      "eval_runtime": 157.3072,
      "eval_samples_per_second": 7.368,
      "eval_steps_per_second": 7.368,
      "step": 2600
    },
    {
      "epoch": 4.006908462867012,
      "grad_norm": 0.08331400901079178,
      "learning_rate": 1.1495540752441235e-05,
      "loss": 0.0093,
      "step": 2610
    },
    {
      "epoch": 4.0222606025714835,
      "grad_norm": 0.275341272354126,
      "learning_rate": 1.1155642667462318e-05,
      "loss": 0.0047,
      "step": 2620
    },
    {
      "epoch": 4.037612742275955,
      "grad_norm": 0.5107333064079285,
      "learning_rate": 1.082021331931164e-05,
      "loss": 0.0039,
      "step": 2630
    },
    {
      "epoch": 4.052964881980426,
      "grad_norm": 0.2175411880016327,
      "learning_rate": 1.0489291296489152e-05,
      "loss": 0.0046,
      "step": 2640
    },
    {
      "epoch": 4.068317021684897,
      "grad_norm": 0.031360745429992676,
      "learning_rate": 1.0162914668962631e-05,
      "loss": 0.0028,
      "step": 2650
    },
    {
      "epoch": 4.068317021684897,
      "eval_loss": 0.05359194427728653,
      "eval_runtime": 157.3201,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2650
    },
    {
      "epoch": 4.083669161389369,
      "grad_norm": 0.12836819887161255,
      "learning_rate": 9.841120983787915e-06,
      "loss": 0.0034,
      "step": 2660
    },
    {
      "epoch": 4.09902130109384,
      "grad_norm": 0.2958119809627533,
      "learning_rate": 9.523947260789546e-06,
      "loss": 0.0043,
      "step": 2670
    },
    {
      "epoch": 4.114373440798311,
      "grad_norm": 0.2515159249305725,
      "learning_rate": 9.211429988301823e-06,
      "loss": 0.0035,
      "step": 2680
    },
    {
      "epoch": 4.129725580502782,
      "grad_norm": 0.7020996809005737,
      "learning_rate": 8.90360511897121e-06,
      "loss": 0.0065,
      "step": 2690
    },
    {
      "epoch": 4.1450777202072535,
      "grad_norm": 0.1990860551595688,
      "learning_rate": 8.600508065620161e-06,
      "loss": 0.0046,
      "step": 2700
    },
    {
      "epoch": 4.1450777202072535,
      "eval_loss": 0.057626817375421524,
      "eval_runtime": 157.3149,
      "eval_samples_per_second": 7.367,
      "eval_steps_per_second": 7.367,
      "step": 2700
    },
    {
      "epoch": 4.1604298599117255,
      "grad_norm": 0.21370138227939606,
      "learning_rate": 8.302173697173226e-06,
      "loss": 0.0036,
      "step": 2710
    },
    {
      "epoch": 4.175781999616197,
      "grad_norm": 0.1672574132680893,
      "learning_rate": 8.008636334645631e-06,
      "loss": 0.0033,
      "step": 2720
    },
    {
      "epoch": 4.191134139320668,
      "grad_norm": 0.128431499004364,
      "learning_rate": 7.71992974719491e-06,
      "loss": 0.0024,
      "step": 2730
    },
    {
      "epoch": 4.206486279025139,
      "grad_norm": 0.04721586033701897,
      "learning_rate": 7.436087148236054e-06,
      "loss": 0.0019,
      "step": 2740
    },
    {
      "epoch": 4.22183841872961,
      "grad_norm": 0.1882028728723526,
      "learning_rate": 7.157141191620548e-06,
      "loss": 0.0038,
      "step": 2750
    },
    {
      "epoch": 4.22183841872961,
      "eval_loss": 0.06159433349967003,
      "eval_runtime": 157.3463,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 2750
    },
    {
      "epoch": 4.237190558434082,
      "grad_norm": 0.08508176356554031,
      "learning_rate": 6.883123967879796e-06,
      "loss": 0.0052,
      "step": 2760
    },
    {
      "epoch": 4.252542698138553,
      "grad_norm": 0.07488391548395157,
      "learning_rate": 6.6140670005334136e-06,
      "loss": 0.0036,
      "step": 2770
    },
    {
      "epoch": 4.267894837843024,
      "grad_norm": 0.30203863978385925,
      "learning_rate": 6.350001242462617e-06,
      "loss": 0.0048,
      "step": 2780
    },
    {
      "epoch": 4.2832469775474955,
      "grad_norm": 0.3173889219760895,
      "learning_rate": 6.090957072349385e-06,
      "loss": 0.0045,
      "step": 2790
    },
    {
      "epoch": 4.298599117251967,
      "grad_norm": 0.03237370029091835,
      "learning_rate": 5.836964291181624e-06,
      "loss": 0.0029,
      "step": 2800
    },
    {
      "epoch": 4.298599117251967,
      "eval_loss": 0.062100403010845184,
      "eval_runtime": 157.3803,
      "eval_samples_per_second": 7.364,
      "eval_steps_per_second": 7.364,
      "step": 2800
    },
    {
      "epoch": 4.313951256956439,
      "grad_norm": 0.2693499028682709,
      "learning_rate": 5.588052118824804e-06,
      "loss": 0.0033,
      "step": 2810
    },
    {
      "epoch": 4.32930339666091,
      "grad_norm": 0.23804733157157898,
      "learning_rate": 5.344249190660428e-06,
      "loss": 0.0021,
      "step": 2820
    },
    {
      "epoch": 4.344655536365381,
      "grad_norm": 0.39224836230278015,
      "learning_rate": 5.105583554291765e-06,
      "loss": 0.0036,
      "step": 2830
    },
    {
      "epoch": 4.360007676069852,
      "grad_norm": 0.03341009467840195,
      "learning_rate": 4.872082666317207e-06,
      "loss": 0.0027,
      "step": 2840
    },
    {
      "epoch": 4.375359815774323,
      "grad_norm": 0.10064135491847992,
      "learning_rate": 4.6437733891715905e-06,
      "loss": 0.0025,
      "step": 2850
    },
    {
      "epoch": 4.375359815774323,
      "eval_loss": 0.06398806720972061,
      "eval_runtime": 157.3861,
      "eval_samples_per_second": 7.364,
      "eval_steps_per_second": 7.364,
      "step": 2850
    },
    {
      "epoch": 4.390711955478795,
      "grad_norm": 0.1537330597639084,
      "learning_rate": 4.420681988035891e-06,
      "loss": 0.0042,
      "step": 2860
    },
    {
      "epoch": 4.406064095183266,
      "grad_norm": 0.08775684237480164,
      "learning_rate": 4.2028341278156026e-06,
      "loss": 0.0055,
      "step": 2870
    },
    {
      "epoch": 4.4214162348877375,
      "grad_norm": 0.15035293996334076,
      "learning_rate": 3.990254870188221e-06,
      "loss": 0.0052,
      "step": 2880
    },
    {
      "epoch": 4.436768374592209,
      "grad_norm": 0.026623625308275223,
      "learning_rate": 3.7829686707200827e-06,
      "loss": 0.0046,
      "step": 2890
    },
    {
      "epoch": 4.45212051429668,
      "grad_norm": 0.39092618227005005,
      "learning_rate": 3.580999376052946e-06,
      "loss": 0.0037,
      "step": 2900
    },
    {
      "epoch": 4.45212051429668,
      "eval_loss": 0.06381876021623611,
      "eval_runtime": 157.3824,
      "eval_samples_per_second": 7.364,
      "eval_steps_per_second": 7.364,
      "step": 2900
    },
    {
      "epoch": 4.467472654001152,
      "grad_norm": 0.3707815110683441,
      "learning_rate": 3.3843702211606153e-06,
      "loss": 0.0031,
      "step": 2910
    },
    {
      "epoch": 4.482824793705623,
      "grad_norm": 0.05645956099033356,
      "learning_rate": 3.193103826675947e-06,
      "loss": 0.0058,
      "step": 2920
    },
    {
      "epoch": 4.498176933410094,
      "grad_norm": 0.10445626080036163,
      "learning_rate": 3.007222196288545e-06,
      "loss": 0.002,
      "step": 2930
    },
    {
      "epoch": 4.513529073114565,
      "grad_norm": 0.05766968056559563,
      "learning_rate": 2.8267467142133687e-06,
      "loss": 0.0025,
      "step": 2940
    },
    {
      "epoch": 4.528881212819036,
      "grad_norm": 3.2273781299591064,
      "learning_rate": 2.651698142730674e-06,
      "loss": 0.0071,
      "step": 2950
    },
    {
      "epoch": 4.528881212819036,
      "eval_loss": 0.06478563696146011,
      "eval_runtime": 157.419,
      "eval_samples_per_second": 7.363,
      "eval_steps_per_second": 7.363,
      "step": 2950
    },
    {
      "epoch": 4.544233352523508,
      "grad_norm": 0.0076272995211184025,
      "learning_rate": 2.4820966197974748e-06,
      "loss": 0.0053,
      "step": 2960
    },
    {
      "epoch": 4.5595854922279795,
      "grad_norm": 0.1376052349805832,
      "learning_rate": 2.3179616567308216e-06,
      "loss": 0.002,
      "step": 2970
    },
    {
      "epoch": 4.574937631932451,
      "grad_norm": 0.10553725808858871,
      "learning_rate": 2.1593121359631873e-06,
      "loss": 0.0026,
      "step": 2980
    },
    {
      "epoch": 4.590289771636922,
      "grad_norm": 0.043689433485269547,
      "learning_rate": 2.006166308870189e-06,
      "loss": 0.0045,
      "step": 2990
    },
    {
      "epoch": 4.605641911341393,
      "grad_norm": 0.04369067773222923,
      "learning_rate": 1.8585417936709038e-06,
      "loss": 0.0015,
      "step": 3000
    },
    {
      "epoch": 4.605641911341393,
      "eval_loss": 0.0653165951371193,
      "eval_runtime": 157.3862,
      "eval_samples_per_second": 7.364,
      "eval_steps_per_second": 7.364,
      "step": 3000
    },
    {
      "epoch": 4.620994051045864,
      "grad_norm": 0.09847624599933624,
      "learning_rate": 1.7164555734010545e-06,
      "loss": 0.0093,
      "step": 3010
    },
    {
      "epoch": 4.636346190750336,
      "grad_norm": 0.10092262178659439,
      "learning_rate": 1.5799239939592204e-06,
      "loss": 0.0031,
      "step": 3020
    },
    {
      "epoch": 4.651698330454807,
      "grad_norm": 0.11192231625318527,
      "learning_rate": 1.4489627622263747e-06,
      "loss": 0.0034,
      "step": 3030
    },
    {
      "epoch": 4.667050470159278,
      "grad_norm": 0.49182823300361633,
      "learning_rate": 1.3235869442589255e-06,
      "loss": 0.0083,
      "step": 3040
    },
    {
      "epoch": 4.6824026098637495,
      "grad_norm": 0.25427430868148804,
      "learning_rate": 1.2038109635555406e-06,
      "loss": 0.002,
      "step": 3050
    },
    {
      "epoch": 4.6824026098637495,
      "eval_loss": 0.06511291116476059,
      "eval_runtime": 157.4295,
      "eval_samples_per_second": 7.362,
      "eval_steps_per_second": 7.362,
      "step": 3050
    },
    {
      "epoch": 4.697754749568221,
      "grad_norm": 0.004702554550021887,
      "learning_rate": 1.0896485993977467e-06,
      "loss": 0.005,
      "step": 3060
    },
    {
      "epoch": 4.713106889272693,
      "grad_norm": 0.1602880358695984,
      "learning_rate": 9.811129852647982e-07,
      "loss": 0.0044,
      "step": 3070
    },
    {
      "epoch": 4.728459028977164,
      "grad_norm": 0.5000692009925842,
      "learning_rate": 8.782166073227515e-07,
      "loss": 0.0104,
      "step": 3080
    },
    {
      "epoch": 4.743811168681635,
      "grad_norm": 0.015439675189554691,
      "learning_rate": 7.809713029880428e-07,
      "loss": 0.0048,
      "step": 3090
    },
    {
      "epoch": 4.759163308386106,
      "grad_norm": 0.3154544532299042,
      "learning_rate": 6.893882595656598e-07,
      "loss": 0.0073,
      "step": 3100
    },
    {
      "epoch": 4.759163308386106,
      "eval_loss": 0.0645485520362854,
      "eval_runtime": 157.4614,
      "eval_samples_per_second": 7.361,
      "eval_steps_per_second": 7.361,
      "step": 3100
    },
    {
      "epoch": 4.774515448090577,
      "grad_norm": 0.36479201912879944,
      "learning_rate": 6.034780129621664e-07,
      "loss": 0.008,
      "step": 3110
    },
    {
      "epoch": 4.789867587795049,
      "grad_norm": 0.5649594068527222,
      "learning_rate": 5.232504464735833e-07,
      "loss": 0.0069,
      "step": 3120
    },
    {
      "epoch": 4.80521972749952,
      "grad_norm": 0.4218672513961792,
      "learning_rate": 4.487147896484523e-07,
      "loss": 0.005,
      "step": 3130
    },
    {
      "epoch": 4.8205718672039914,
      "grad_norm": 0.02274581603705883,
      "learning_rate": 3.7987961722599773e-07,
      "loss": 0.0077,
      "step": 3140
    },
    {
      "epoch": 4.835924006908463,
      "grad_norm": 0.11879457533359528,
      "learning_rate": 3.167528481496984e-07,
      "loss": 0.0061,
      "step": 3150
    },
    {
      "epoch": 4.835924006908463,
      "eval_loss": 0.06395623832941055,
      "eval_runtime": 157.4124,
      "eval_samples_per_second": 7.363,
      "eval_steps_per_second": 7.363,
      "step": 3150
    },
    {
      "epoch": 4.851276146612934,
      "grad_norm": 0.07862385362386703,
      "learning_rate": 2.593417446562607e-07,
      "loss": 0.0103,
      "step": 3160
    },
    {
      "epoch": 4.866628286317406,
      "grad_norm": 0.10271923989057541,
      "learning_rate": 2.0765291144016486e-07,
      "loss": 0.0017,
      "step": 3170
    },
    {
      "epoch": 4.881980426021877,
      "grad_norm": 0.023111067712306976,
      "learning_rate": 1.6169229489385595e-07,
      "loss": 0.0041,
      "step": 3180
    },
    {
      "epoch": 4.897332565726348,
      "grad_norm": 0.38597050309181213,
      "learning_rate": 1.2146518242363014e-07,
      "loss": 0.003,
      "step": 3190
    },
    {
      "epoch": 4.912684705430819,
      "grad_norm": 0.10105849802494049,
      "learning_rate": 8.697620184138222e-08,
      "loss": 0.0024,
      "step": 3200
    },
    {
      "epoch": 4.912684705430819,
      "eval_loss": 0.0639420673251152,
      "eval_runtime": 157.414,
      "eval_samples_per_second": 7.363,
      "eval_steps_per_second": 7.363,
      "step": 3200
    },
    {
      "epoch": 4.92803684513529,
      "grad_norm": 0.06087774783372879,
      "learning_rate": 5.822932083221488e-08,
      "loss": 0.0098,
      "step": 3210
    },
    {
      "epoch": 4.943388984839762,
      "grad_norm": 0.1292206346988678,
      "learning_rate": 3.5227846497970504e-08,
      "loss": 0.0037,
      "step": 3220
    },
    {
      "epoch": 4.958741124544233,
      "grad_norm": 0.042202211916446686,
      "learning_rate": 1.7974424976796577e-08,
      "loss": 0.0014,
      "step": 3230
    },
    {
      "epoch": 4.974093264248705,
      "grad_norm": 0.06265482306480408,
      "learning_rate": 6.47104113870034e-09,
      "loss": 0.0055,
      "step": 3240
    },
    {
      "epoch": 4.989445403953176,
      "grad_norm": 0.3988359868526459,
      "learning_rate": 7.190183572314269e-10,
      "loss": 0.0054,
      "step": 3250
    },
    {
      "epoch": 4.989445403953176,
      "eval_loss": 0.06391575932502747,
      "eval_runtime": 157.4119,
      "eval_samples_per_second": 7.363,
      "eval_steps_per_second": 7.363,
      "step": 3250
    },
    {
      "epoch": 4.997121473805412,
      "step": 3255,
      "total_flos": 8.175027015212728e+17,
      "train_loss": 0.04520614844061629,
      "train_runtime": 35604.7112,
      "train_samples_per_second": 1.464,
      "train_steps_per_second": 0.091
    }
  ],
  "logging_steps": 10,
  "max_steps": 3255,
  "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": 8.175027015212728e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}