{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 75,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.5240228176116943,
      "epoch": 0.04,
      "grad_norm": 37.29308319091797,
      "learning_rate": 2e-05,
      "loss": 6.9422,
      "mean_token_accuracy": 0.47347480058670044,
      "num_tokens": 758.0,
      "step": 1
    },
    {
      "entropy": 0.5185598134994507,
      "epoch": 0.08,
      "grad_norm": 37.85886764526367,
      "learning_rate": 2e-05,
      "loss": 6.8565,
      "mean_token_accuracy": 0.47270306944847107,
      "num_tokens": 1513.0,
      "step": 2
    },
    {
      "entropy": 0.5363126993179321,
      "epoch": 0.12,
      "grad_norm": 36.92781066894531,
      "learning_rate": 2e-05,
      "loss": 6.6136,
      "mean_token_accuracy": 0.47562581300735474,
      "num_tokens": 2276.0,
      "step": 3
    },
    {
      "entropy": 0.5541151762008667,
      "epoch": 0.16,
      "grad_norm": 37.46426773071289,
      "learning_rate": 2e-05,
      "loss": 6.4114,
      "mean_token_accuracy": 0.47200000286102295,
      "num_tokens": 3030.0,
      "step": 4
    },
    {
      "entropy": 0.6072134375572205,
      "epoch": 0.2,
      "grad_norm": 36.258392333984375,
      "learning_rate": 2e-05,
      "loss": 5.9563,
      "mean_token_accuracy": 0.4704301059246063,
      "num_tokens": 3778.0,
      "step": 5
    },
    {
      "entropy": 0.6346766948699951,
      "epoch": 0.24,
      "grad_norm": 34.2164421081543,
      "learning_rate": 2e-05,
      "loss": 5.4104,
      "mean_token_accuracy": 0.48868176341056824,
      "num_tokens": 4533.0,
      "step": 6
    },
    {
      "entropy": 0.699142575263977,
      "epoch": 0.28,
      "grad_norm": 31.32328987121582,
      "learning_rate": 2e-05,
      "loss": 4.9817,
      "mean_token_accuracy": 0.48806366324424744,
      "num_tokens": 5291.0,
      "step": 7
    },
    {
      "entropy": 0.7639284133911133,
      "epoch": 0.32,
      "grad_norm": 27.981550216674805,
      "learning_rate": 2e-05,
      "loss": 4.5891,
      "mean_token_accuracy": 0.5134048461914062,
      "num_tokens": 6041.0,
      "step": 8
    },
    {
      "entropy": 0.8184663653373718,
      "epoch": 0.36,
      "grad_norm": 23.888887405395508,
      "learning_rate": 2e-05,
      "loss": 4.1021,
      "mean_token_accuracy": 0.535380482673645,
      "num_tokens": 6794.0,
      "step": 9
    },
    {
      "entropy": 0.8713943362236023,
      "epoch": 0.4,
      "grad_norm": 21.27410125732422,
      "learning_rate": 2e-05,
      "loss": 3.863,
      "mean_token_accuracy": 0.5392809510231018,
      "num_tokens": 7549.0,
      "step": 10
    },
    {
      "entropy": 0.9519356489181519,
      "epoch": 0.44,
      "grad_norm": 18.26158905029297,
      "learning_rate": 2e-05,
      "loss": 3.5874,
      "mean_token_accuracy": 0.5420560836791992,
      "num_tokens": 8302.0,
      "step": 11
    },
    {
      "entropy": 1.025558352470398,
      "epoch": 0.48,
      "grad_norm": 16.051610946655273,
      "learning_rate": 2e-05,
      "loss": 3.3905,
      "mean_token_accuracy": 0.5483443737030029,
      "num_tokens": 9061.0,
      "step": 12
    },
    {
      "entropy": 1.0597870349884033,
      "epoch": 0.52,
      "grad_norm": 13.387225151062012,
      "learning_rate": 2e-05,
      "loss": 3.1501,
      "mean_token_accuracy": 0.5702811479568481,
      "num_tokens": 9812.0,
      "step": 13
    },
    {
      "entropy": 1.1256167888641357,
      "epoch": 0.56,
      "grad_norm": 11.642187118530273,
      "learning_rate": 2e-05,
      "loss": 2.9905,
      "mean_token_accuracy": 0.5778961181640625,
      "num_tokens": 10567.0,
      "step": 14
    },
    {
      "entropy": 1.1595126390457153,
      "epoch": 0.6,
      "grad_norm": 10.657787322998047,
      "learning_rate": 2e-05,
      "loss": 2.8271,
      "mean_token_accuracy": 0.5927419066429138,
      "num_tokens": 11315.0,
      "step": 15
    },
    {
      "entropy": 1.2265969514846802,
      "epoch": 0.64,
      "grad_norm": 9.537993431091309,
      "learning_rate": 2e-05,
      "loss": 2.749,
      "mean_token_accuracy": 0.5964912176132202,
      "num_tokens": 12060.0,
      "step": 16
    },
    {
      "entropy": 1.2593307495117188,
      "epoch": 0.68,
      "grad_norm": 8.315566062927246,
      "learning_rate": 2e-05,
      "loss": 2.5743,
      "mean_token_accuracy": 0.6169934868812561,
      "num_tokens": 12829.0,
      "step": 17
    },
    {
      "entropy": 1.3245295286178589,
      "epoch": 0.72,
      "grad_norm": 7.915526390075684,
      "learning_rate": 2e-05,
      "loss": 2.5367,
      "mean_token_accuracy": 0.6194332242012024,
      "num_tokens": 13574.0,
      "step": 18
    },
    {
      "entropy": 1.3604302406311035,
      "epoch": 0.76,
      "grad_norm": 7.328187942504883,
      "learning_rate": 2e-05,
      "loss": 2.4018,
      "mean_token_accuracy": 0.6255033612251282,
      "num_tokens": 14323.0,
      "step": 19
    },
    {
      "entropy": 1.3951760530471802,
      "epoch": 0.8,
      "grad_norm": 6.733570098876953,
      "learning_rate": 2e-05,
      "loss": 2.317,
      "mean_token_accuracy": 0.635046124458313,
      "num_tokens": 15086.0,
      "step": 20
    },
    {
      "entropy": 1.4354883432388306,
      "epoch": 0.84,
      "grad_norm": 6.525688648223877,
      "learning_rate": 2e-05,
      "loss": 2.2795,
      "mean_token_accuracy": 0.6360052824020386,
      "num_tokens": 15851.0,
      "step": 21
    },
    {
      "entropy": 1.4653663635253906,
      "epoch": 0.88,
      "grad_norm": 6.118790149688721,
      "learning_rate": 2e-05,
      "loss": 2.1621,
      "mean_token_accuracy": 0.6416107416152954,
      "num_tokens": 16600.0,
      "step": 22
    },
    {
      "entropy": 1.4949076175689697,
      "epoch": 0.92,
      "grad_norm": 5.929473400115967,
      "learning_rate": 2e-05,
      "loss": 2.1265,
      "mean_token_accuracy": 0.6526458859443665,
      "num_tokens": 17341.0,
      "step": 23
    },
    {
      "entropy": 1.4969980716705322,
      "epoch": 0.96,
      "grad_norm": 5.519643783569336,
      "learning_rate": 2e-05,
      "loss": 2.0365,
      "mean_token_accuracy": 0.6627296805381775,
      "num_tokens": 18107.0,
      "step": 24
    },
    {
      "entropy": 1.5308914184570312,
      "epoch": 1.0,
      "grad_norm": 5.461964130401611,
      "learning_rate": 2e-05,
      "loss": 1.9552,
      "mean_token_accuracy": 0.6851119995117188,
      "num_tokens": 18870.0,
      "step": 25
    },
    {
      "entropy": 1.531392216682434,
      "epoch": 1.04,
      "grad_norm": 5.313822269439697,
      "learning_rate": 2e-05,
      "loss": 1.8633,
      "mean_token_accuracy": 0.7098515629768372,
      "num_tokens": 19615.0,
      "step": 26
    },
    {
      "entropy": 1.5274267196655273,
      "epoch": 1.08,
      "grad_norm": 5.325898170471191,
      "learning_rate": 2e-05,
      "loss": 1.9049,
      "mean_token_accuracy": 0.7068965435028076,
      "num_tokens": 20373.0,
      "step": 27
    },
    {
      "entropy": 1.5113599300384521,
      "epoch": 1.12,
      "grad_norm": 5.022428512573242,
      "learning_rate": 2e-05,
      "loss": 1.7917,
      "mean_token_accuracy": 0.7334217429161072,
      "num_tokens": 21131.0,
      "step": 28
    },
    {
      "entropy": 1.4889423847198486,
      "epoch": 1.16,
      "grad_norm": 4.645327568054199,
      "learning_rate": 2e-05,
      "loss": 1.7032,
      "mean_token_accuracy": 0.7447090148925781,
      "num_tokens": 21891.0,
      "step": 29
    },
    {
      "entropy": 1.4870576858520508,
      "epoch": 1.2,
      "grad_norm": 4.541719913482666,
      "learning_rate": 2e-05,
      "loss": 1.6655,
      "mean_token_accuracy": 0.7360000014305115,
      "num_tokens": 22645.0,
      "step": 30
    },
    {
      "entropy": 1.4613386392593384,
      "epoch": 1.24,
      "grad_norm": 4.408016204833984,
      "learning_rate": 2e-05,
      "loss": 1.6312,
      "mean_token_accuracy": 0.7432795763015747,
      "num_tokens": 23393.0,
      "step": 31
    },
    {
      "entropy": 1.4358580112457275,
      "epoch": 1.28,
      "grad_norm": 4.297278881072998,
      "learning_rate": 2e-05,
      "loss": 1.5793,
      "mean_token_accuracy": 0.7510094046592712,
      "num_tokens": 24140.0,
      "step": 32
    },
    {
      "entropy": 1.3814220428466797,
      "epoch": 1.32,
      "grad_norm": 4.178844451904297,
      "learning_rate": 2e-05,
      "loss": 1.4675,
      "mean_token_accuracy": 0.7724867463111877,
      "num_tokens": 24900.0,
      "step": 33
    },
    {
      "entropy": 1.3687061071395874,
      "epoch": 1.3599999999999999,
      "grad_norm": 4.218804836273193,
      "learning_rate": 2e-05,
      "loss": 1.4288,
      "mean_token_accuracy": 0.7776280045509338,
      "num_tokens": 25646.0,
      "step": 34
    },
    {
      "entropy": 1.3419899940490723,
      "epoch": 1.4,
      "grad_norm": 4.205630302429199,
      "learning_rate": 2e-05,
      "loss": 1.4188,
      "mean_token_accuracy": 0.7807486653327942,
      "num_tokens": 26398.0,
      "step": 35
    },
    {
      "entropy": 1.2921583652496338,
      "epoch": 1.44,
      "grad_norm": 3.8976969718933105,
      "learning_rate": 2e-05,
      "loss": 1.3454,
      "mean_token_accuracy": 0.7848605513572693,
      "num_tokens": 27155.0,
      "step": 36
    },
    {
      "entropy": 1.2668341398239136,
      "epoch": 1.48,
      "grad_norm": 3.746513843536377,
      "learning_rate": 2e-05,
      "loss": 1.3557,
      "mean_token_accuracy": 0.7797062993049622,
      "num_tokens": 27908.0,
      "step": 37
    },
    {
      "entropy": 1.228274941444397,
      "epoch": 1.52,
      "grad_norm": 3.520526170730591,
      "learning_rate": 2e-05,
      "loss": 1.2668,
      "mean_token_accuracy": 0.8029294013977051,
      "num_tokens": 28663.0,
      "step": 38
    },
    {
      "entropy": 1.1697431802749634,
      "epoch": 1.56,
      "grad_norm": 3.4416074752807617,
      "learning_rate": 2e-05,
      "loss": 1.1742,
      "mean_token_accuracy": 0.8176870942115784,
      "num_tokens": 29402.0,
      "step": 39
    },
    {
      "entropy": 1.130076289176941,
      "epoch": 1.6,
      "grad_norm": 3.223904848098755,
      "learning_rate": 2e-05,
      "loss": 1.1646,
      "mean_token_accuracy": 0.8208168745040894,
      "num_tokens": 30165.0,
      "step": 40
    },
    {
      "entropy": 1.1070514917373657,
      "epoch": 1.6400000000000001,
      "grad_norm": 3.213395118713379,
      "learning_rate": 2e-05,
      "loss": 1.1102,
      "mean_token_accuracy": 0.8348745107650757,
      "num_tokens": 30926.0,
      "step": 41
    },
    {
      "entropy": 1.0644617080688477,
      "epoch": 1.6800000000000002,
      "grad_norm": 3.1568286418914795,
      "learning_rate": 2e-05,
      "loss": 1.0741,
      "mean_token_accuracy": 0.8320000171661377,
      "num_tokens": 31680.0,
      "step": 42
    },
    {
      "entropy": 1.031175971031189,
      "epoch": 1.72,
      "grad_norm": 3.1758995056152344,
      "learning_rate": 2e-05,
      "loss": 1.0527,
      "mean_token_accuracy": 0.8251001238822937,
      "num_tokens": 32433.0,
      "step": 43
    },
    {
      "entropy": 0.9702178835868835,
      "epoch": 1.76,
      "grad_norm": 3.0158236026763916,
      "learning_rate": 2e-05,
      "loss": 0.9569,
      "mean_token_accuracy": 0.8379814028739929,
      "num_tokens": 33190.0,
      "step": 44
    },
    {
      "entropy": 0.9418991208076477,
      "epoch": 1.8,
      "grad_norm": 3.095254898071289,
      "learning_rate": 2e-05,
      "loss": 0.9229,
      "mean_token_accuracy": 0.852150559425354,
      "num_tokens": 33938.0,
      "step": 45
    },
    {
      "entropy": 0.8792616128921509,
      "epoch": 1.8399999999999999,
      "grad_norm": 3.1904547214508057,
      "learning_rate": 2e-05,
      "loss": 0.8814,
      "mean_token_accuracy": 0.8521400690078735,
      "num_tokens": 34713.0,
      "step": 46
    },
    {
      "entropy": 0.8644114136695862,
      "epoch": 1.88,
      "grad_norm": 3.1296632289886475,
      "learning_rate": 2e-05,
      "loss": 0.9185,
      "mean_token_accuracy": 0.8663967847824097,
      "num_tokens": 35458.0,
      "step": 47
    },
    {
      "entropy": 0.8369636535644531,
      "epoch": 1.92,
      "grad_norm": 3.080310106277466,
      "learning_rate": 2e-05,
      "loss": 0.8598,
      "mean_token_accuracy": 0.875,
      "num_tokens": 36214.0,
      "step": 48
    },
    {
      "entropy": 0.7779583930969238,
      "epoch": 1.96,
      "grad_norm": 2.9825854301452637,
      "learning_rate": 2e-05,
      "loss": 0.757,
      "mean_token_accuracy": 0.8845144510269165,
      "num_tokens": 36980.0,
      "step": 49
    },
    {
      "entropy": 0.7366280555725098,
      "epoch": 2.0,
      "grad_norm": 3.0169920921325684,
      "learning_rate": 2e-05,
      "loss": 0.7402,
      "mean_token_accuracy": 0.8928571343421936,
      "num_tokens": 37740.0,
      "step": 50
    },
    {
      "entropy": 0.7010611891746521,
      "epoch": 2.04,
      "grad_norm": 3.399986743927002,
      "learning_rate": 2e-05,
      "loss": 0.7593,
      "mean_token_accuracy": 0.8872870206832886,
      "num_tokens": 38507.0,
      "step": 51
    },
    {
      "entropy": 0.6273115873336792,
      "epoch": 2.08,
      "grad_norm": 3.619697332382202,
      "learning_rate": 2e-05,
      "loss": 0.6621,
      "mean_token_accuracy": 0.9089692234992981,
      "num_tokens": 39258.0,
      "step": 52
    },
    {
      "entropy": 0.5995044708251953,
      "epoch": 2.12,
      "grad_norm": 2.8443078994750977,
      "learning_rate": 2e-05,
      "loss": 0.662,
      "mean_token_accuracy": 0.9167767763137817,
      "num_tokens": 40019.0,
      "step": 53
    },
    {
      "entropy": 0.5483225584030151,
      "epoch": 2.16,
      "grad_norm": 2.8451199531555176,
      "learning_rate": 2e-05,
      "loss": 0.575,
      "mean_token_accuracy": 0.929427444934845,
      "num_tokens": 40774.0,
      "step": 54
    },
    {
      "entropy": 0.5085722208023071,
      "epoch": 2.2,
      "grad_norm": 2.7123143672943115,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "mean_token_accuracy": 0.9348404407501221,
      "num_tokens": 41530.0,
      "step": 55
    },
    {
      "entropy": 0.4458790123462677,
      "epoch": 2.24,
      "grad_norm": 2.668097496032715,
      "learning_rate": 2e-05,
      "loss": 0.5034,
      "mean_token_accuracy": 0.9419702887535095,
      "num_tokens": 42275.0,
      "step": 56
    },
    {
      "entropy": 0.3930673599243164,
      "epoch": 2.2800000000000002,
      "grad_norm": 2.5482311248779297,
      "learning_rate": 2e-05,
      "loss": 0.5078,
      "mean_token_accuracy": 0.9355263113975525,
      "num_tokens": 43039.0,
      "step": 57
    },
    {
      "entropy": 0.37228837609291077,
      "epoch": 2.32,
      "grad_norm": 2.443922281265259,
      "learning_rate": 2e-05,
      "loss": 0.4665,
      "mean_token_accuracy": 0.9482071995735168,
      "num_tokens": 43796.0,
      "step": 58
    },
    {
      "entropy": 0.3517550826072693,
      "epoch": 2.36,
      "grad_norm": 2.3463053703308105,
      "learning_rate": 2e-05,
      "loss": 0.4632,
      "mean_token_accuracy": 0.9378378391265869,
      "num_tokens": 44540.0,
      "step": 59
    },
    {
      "entropy": 0.33203697204589844,
      "epoch": 2.4,
      "grad_norm": 2.6731350421905518,
      "learning_rate": 2e-05,
      "loss": 0.4491,
      "mean_token_accuracy": 0.9372549057006836,
      "num_tokens": 45309.0,
      "step": 60
    },
    {
      "entropy": 0.31520795822143555,
      "epoch": 2.44,
      "grad_norm": 2.3959364891052246,
      "learning_rate": 2e-05,
      "loss": 0.44,
      "mean_token_accuracy": 0.9494680762290955,
      "num_tokens": 46065.0,
      "step": 61
    },
    {
      "entropy": 0.3004280626773834,
      "epoch": 2.48,
      "grad_norm": 2.1749024391174316,
      "learning_rate": 2e-05,
      "loss": 0.4511,
      "mean_token_accuracy": 0.9522546529769897,
      "num_tokens": 46823.0,
      "step": 62
    },
    {
      "entropy": 0.287904292345047,
      "epoch": 2.52,
      "grad_norm": 2.184871196746826,
      "learning_rate": 2e-05,
      "loss": 0.4058,
      "mean_token_accuracy": 0.9521276354789734,
      "num_tokens": 47579.0,
      "step": 63
    },
    {
      "entropy": 0.28218233585357666,
      "epoch": 2.56,
      "grad_norm": 2.27091121673584,
      "learning_rate": 2e-05,
      "loss": 0.3679,
      "mean_token_accuracy": 0.9529569745063782,
      "num_tokens": 48327.0,
      "step": 64
    },
    {
      "entropy": 0.2818812429904938,
      "epoch": 2.6,
      "grad_norm": 2.9275169372558594,
      "learning_rate": 2e-05,
      "loss": 0.3858,
      "mean_token_accuracy": 0.9478609561920166,
      "num_tokens": 49079.0,
      "step": 65
    },
    {
      "entropy": 0.2774568200111389,
      "epoch": 2.64,
      "grad_norm": 2.837167978286743,
      "learning_rate": 2e-05,
      "loss": 0.3639,
      "mean_token_accuracy": 0.9430463314056396,
      "num_tokens": 49838.0,
      "step": 66
    },
    {
      "entropy": 0.2726022005081177,
      "epoch": 2.68,
      "grad_norm": 2.2975914478302,
      "learning_rate": 2e-05,
      "loss": 0.3028,
      "mean_token_accuracy": 0.9518072009086609,
      "num_tokens": 50589.0,
      "step": 67
    },
    {
      "entropy": 0.25494953989982605,
      "epoch": 2.7199999999999998,
      "grad_norm": 2.0497214794158936,
      "learning_rate": 2e-05,
      "loss": 0.2769,
      "mean_token_accuracy": 0.9570469856262207,
      "num_tokens": 51338.0,
      "step": 68
    },
    {
      "entropy": 0.2599942982196808,
      "epoch": 2.76,
      "grad_norm": 2.4737048149108887,
      "learning_rate": 2e-05,
      "loss": 0.336,
      "mean_token_accuracy": 0.9518072009086609,
      "num_tokens": 52089.0,
      "step": 69
    },
    {
      "entropy": 0.24719160795211792,
      "epoch": 2.8,
      "grad_norm": 2.1889803409576416,
      "learning_rate": 2e-05,
      "loss": 0.3006,
      "mean_token_accuracy": 0.9491298794746399,
      "num_tokens": 52840.0,
      "step": 70
    },
    {
      "entropy": 0.24826745688915253,
      "epoch": 2.84,
      "grad_norm": 2.0455644130706787,
      "learning_rate": 2e-05,
      "loss": 0.2918,
      "mean_token_accuracy": 0.948344349861145,
      "num_tokens": 53599.0,
      "step": 71
    },
    {
      "entropy": 0.2431359887123108,
      "epoch": 2.88,
      "grad_norm": 1.8650648593902588,
      "learning_rate": 2e-05,
      "loss": 0.2582,
      "mean_token_accuracy": 0.9523809552192688,
      "num_tokens": 54359.0,
      "step": 72
    },
    {
      "entropy": 0.24409686028957367,
      "epoch": 2.92,
      "grad_norm": 2.3259599208831787,
      "learning_rate": 2e-05,
      "loss": 0.2527,
      "mean_token_accuracy": 0.960106372833252,
      "num_tokens": 55115.0,
      "step": 73
    },
    {
      "entropy": 0.2385331094264984,
      "epoch": 2.96,
      "grad_norm": 1.3533285856246948,
      "learning_rate": 2e-05,
      "loss": 0.2706,
      "mean_token_accuracy": 0.9580514430999756,
      "num_tokens": 55858.0,
      "step": 74
    },
    {
      "entropy": 0.24642708897590637,
      "epoch": 3.0,
      "grad_norm": 1.1475350856781006,
      "learning_rate": 2e-05,
      "loss": 0.2543,
      "mean_token_accuracy": 0.9598930478096008,
      "num_tokens": 56610.0,
      "step": 75
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 36069875939328.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}