{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 100,
  "global_step": 1408,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007106057914372002,
      "grad_norm": 19.49246232730451,
      "learning_rate": 1.86046511627907e-06,
      "loss": 2.4934,
      "mean_token_accuracy": 0.5258669804781675,
      "num_tokens": 4763584.0,
      "step": 5
    },
    {
      "epoch": 0.014212115828744005,
      "grad_norm": 12.095302419910517,
      "learning_rate": 4.186046511627907e-06,
      "loss": 2.3185,
      "mean_token_accuracy": 0.5427483215928077,
      "num_tokens": 9531720.0,
      "step": 10
    },
    {
      "epoch": 0.021318173743116006,
      "grad_norm": 6.796780116822856,
      "learning_rate": 6.511627906976745e-06,
      "loss": 1.8082,
      "mean_token_accuracy": 0.5906881660223007,
      "num_tokens": 14272673.0,
      "step": 15
    },
    {
      "epoch": 0.02842423165748801,
      "grad_norm": 2.028130586531004,
      "learning_rate": 8.837209302325582e-06,
      "loss": 1.4468,
      "mean_token_accuracy": 0.639113237708807,
      "num_tokens": 19045659.0,
      "step": 20
    },
    {
      "epoch": 0.03553028957186001,
      "grad_norm": 1.0588899013746833,
      "learning_rate": 1.116279069767442e-05,
      "loss": 1.2276,
      "mean_token_accuracy": 0.6780799143016338,
      "num_tokens": 23811446.0,
      "step": 25
    },
    {
      "epoch": 0.04263634748623201,
      "grad_norm": 0.6488648360367262,
      "learning_rate": 1.3488372093023257e-05,
      "loss": 1.088,
      "mean_token_accuracy": 0.7027405865490437,
      "num_tokens": 28572862.0,
      "step": 30
    },
    {
      "epoch": 0.04974240540060401,
      "grad_norm": 0.5415057756218988,
      "learning_rate": 1.5813953488372095e-05,
      "loss": 1.0242,
      "mean_token_accuracy": 0.7144973143935204,
      "num_tokens": 33349882.0,
      "step": 35
    },
    {
      "epoch": 0.05684846331497602,
      "grad_norm": 0.48382195510047,
      "learning_rate": 1.813953488372093e-05,
      "loss": 0.9712,
      "mean_token_accuracy": 0.7250196196138858,
      "num_tokens": 38137624.0,
      "step": 40
    },
    {
      "epoch": 0.06395452122934801,
      "grad_norm": 0.49727864530874677,
      "learning_rate": 1.99999761632652e-05,
      "loss": 0.9247,
      "mean_token_accuracy": 0.7344170436263084,
      "num_tokens": 42901997.0,
      "step": 45
    },
    {
      "epoch": 0.07106057914372002,
      "grad_norm": 0.4396836808821335,
      "learning_rate": 1.999914189080485e-05,
      "loss": 0.8892,
      "mean_token_accuracy": 0.7411722339689731,
      "num_tokens": 47660479.0,
      "step": 50
    },
    {
      "epoch": 0.07816663705809203,
      "grad_norm": 0.4277947889677102,
      "learning_rate": 1.9997115907865857e-05,
      "loss": 0.8745,
      "mean_token_accuracy": 0.7443610817193985,
      "num_tokens": 52437558.0,
      "step": 55
    },
    {
      "epoch": 0.08527269497246402,
      "grad_norm": 0.4233524880605398,
      "learning_rate": 1.999389848273882e-05,
      "loss": 0.8603,
      "mean_token_accuracy": 0.747463022172451,
      "num_tokens": 57198957.0,
      "step": 60
    },
    {
      "epoch": 0.09237875288683603,
      "grad_norm": 0.409057413495153,
      "learning_rate": 1.998949004149094e-05,
      "loss": 0.8537,
      "mean_token_accuracy": 0.7483336836099624,
      "num_tokens": 61961974.0,
      "step": 65
    },
    {
      "epoch": 0.09948481080120802,
      "grad_norm": 0.46927095285141973,
      "learning_rate": 1.9983891167909617e-05,
      "loss": 0.8375,
      "mean_token_accuracy": 0.7526809796690941,
      "num_tokens": 66725536.0,
      "step": 70
    },
    {
      "epoch": 0.10659086871558003,
      "grad_norm": 0.4106777573300543,
      "learning_rate": 1.9977102603425134e-05,
      "loss": 0.8309,
      "mean_token_accuracy": 0.7542230375111103,
      "num_tokens": 71469095.0,
      "step": 75
    },
    {
      "epoch": 0.11369692662995204,
      "grad_norm": 0.40584561150068,
      "learning_rate": 1.996912524701247e-05,
      "loss": 0.8258,
      "mean_token_accuracy": 0.7558744698762894,
      "num_tokens": 76221430.0,
      "step": 80
    },
    {
      "epoch": 0.12080298454432403,
      "grad_norm": 0.4436779763048366,
      "learning_rate": 1.995996015507227e-05,
      "loss": 0.8152,
      "mean_token_accuracy": 0.7582607261836529,
      "num_tokens": 80991235.0,
      "step": 85
    },
    {
      "epoch": 0.12790904245869603,
      "grad_norm": 0.4011663933394793,
      "learning_rate": 1.9949608541290924e-05,
      "loss": 0.8128,
      "mean_token_accuracy": 0.7592827767133713,
      "num_tokens": 85760262.0,
      "step": 90
    },
    {
      "epoch": 0.13501510037306805,
      "grad_norm": 0.4484975750242541,
      "learning_rate": 1.9938071776479875e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.7621250681579113,
      "num_tokens": 90505979.0,
      "step": 95
    },
    {
      "epoch": 0.14212115828744004,
      "grad_norm": 0.3973118091711915,
      "learning_rate": 1.992535138839406e-05,
      "loss": 0.7956,
      "mean_token_accuracy": 0.7619979940354824,
      "num_tokens": 95259341.0,
      "step": 100
    },
    {
      "epoch": 0.14212115828744004,
      "eval_loss": 0.7710759043693542,
      "eval_mean_token_accuracy": 0.7617696215186203,
      "eval_num_tokens": 95259341.0,
      "eval_runtime": 149.4719,
      "eval_samples_per_second": 24.346,
      "eval_steps_per_second": 0.763,
      "step": 100
    },
    {
      "epoch": 0.14922721620181204,
      "grad_norm": 0.4055637275946486,
      "learning_rate": 1.991144906152962e-05,
      "loss": 0.804,
      "mean_token_accuracy": 0.7599063582718373,
      "num_tokens": 100023224.0,
      "step": 105
    },
    {
      "epoch": 0.15633327411618406,
      "grad_norm": 0.5425556091455714,
      "learning_rate": 1.9896366636900826e-05,
      "loss": 0.7951,
      "mean_token_accuracy": 0.7621362045407295,
      "num_tokens": 104771415.0,
      "step": 110
    },
    {
      "epoch": 0.16343933203055605,
      "grad_norm": 0.4581521538010428,
      "learning_rate": 1.9880106111796266e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.7619842484593391,
      "num_tokens": 109537100.0,
      "step": 115
    },
    {
      "epoch": 0.17054538994492804,
      "grad_norm": 0.42123272880797796,
      "learning_rate": 1.9862669639514382e-05,
      "loss": 0.7886,
      "mean_token_accuracy": 0.7638748176395893,
      "num_tokens": 114304125.0,
      "step": 120
    },
    {
      "epoch": 0.17765144785930007,
      "grad_norm": 0.3957407600605023,
      "learning_rate": 1.9844059529078297e-05,
      "loss": 0.7763,
      "mean_token_accuracy": 0.7664125673472881,
      "num_tokens": 119067772.0,
      "step": 125
    },
    {
      "epoch": 0.18475750577367206,
      "grad_norm": 1.1428359800811156,
      "learning_rate": 1.9824278244930052e-05,
      "loss": 0.7736,
      "mean_token_accuracy": 0.7676285386085511,
      "num_tokens": 123805148.0,
      "step": 130
    },
    {
      "epoch": 0.19186356368804405,
      "grad_norm": 0.4240204964738557,
      "learning_rate": 1.9803328406604252e-05,
      "loss": 0.7841,
      "mean_token_accuracy": 0.763801097869873,
      "num_tokens": 128568367.0,
      "step": 135
    },
    {
      "epoch": 0.19896962160241605,
      "grad_norm": 0.41090851408734014,
      "learning_rate": 1.9781212788381177e-05,
      "loss": 0.7819,
      "mean_token_accuracy": 0.7644710555672646,
      "num_tokens": 133343344.0,
      "step": 140
    },
    {
      "epoch": 0.20607567951678807,
      "grad_norm": 0.38627786124116875,
      "learning_rate": 1.9757934318919386e-05,
      "loss": 0.7586,
      "mean_token_accuracy": 0.7707050330936909,
      "num_tokens": 138078681.0,
      "step": 145
    },
    {
      "epoch": 0.21318173743116006,
      "grad_norm": 0.35413336695126674,
      "learning_rate": 1.973349608086791e-05,
      "loss": 0.7579,
      "mean_token_accuracy": 0.7715410716831684,
      "num_tokens": 142812366.0,
      "step": 150
    },
    {
      "epoch": 0.22028779534553206,
      "grad_norm": 0.44320259589774924,
      "learning_rate": 1.9707901310458017e-05,
      "loss": 0.7649,
      "mean_token_accuracy": 0.7688324272632598,
      "num_tokens": 147567508.0,
      "step": 155
    },
    {
      "epoch": 0.22739385325990408,
      "grad_norm": 0.4272220313623565,
      "learning_rate": 1.9681153397074658e-05,
      "loss": 0.779,
      "mean_token_accuracy": 0.7649676457047463,
      "num_tokens": 152348975.0,
      "step": 160
    },
    {
      "epoch": 0.23449991117427607,
      "grad_norm": 0.36879913627107413,
      "learning_rate": 1.9653255882807625e-05,
      "loss": 0.7547,
      "mean_token_accuracy": 0.7709379114210606,
      "num_tokens": 157094616.0,
      "step": 165
    },
    {
      "epoch": 0.24160596908864806,
      "grad_norm": 0.39000799041219736,
      "learning_rate": 1.9624212461982497e-05,
      "loss": 0.7594,
      "mean_token_accuracy": 0.7707360699772835,
      "num_tokens": 161849805.0,
      "step": 170
    },
    {
      "epoch": 0.2487120270030201,
      "grad_norm": 0.37089862448174604,
      "learning_rate": 1.9594026980671423e-05,
      "loss": 0.7555,
      "mean_token_accuracy": 0.7713063634932041,
      "num_tokens": 166609253.0,
      "step": 175
    },
    {
      "epoch": 0.25581808491739205,
      "grad_norm": 0.3895218613587052,
      "learning_rate": 1.9562703436183783e-05,
      "loss": 0.7641,
      "mean_token_accuracy": 0.7704252451658249,
      "num_tokens": 171374935.0,
      "step": 180
    },
    {
      "epoch": 0.2629241428317641,
      "grad_norm": 0.41062410007633293,
      "learning_rate": 1.953024597653688e-05,
      "loss": 0.7549,
      "mean_token_accuracy": 0.771364139765501,
      "num_tokens": 176152999.0,
      "step": 185
    },
    {
      "epoch": 0.2700302007461361,
      "grad_norm": 0.41614535212567993,
      "learning_rate": 1.9496658899906605e-05,
      "loss": 0.7479,
      "mean_token_accuracy": 0.7711076475679874,
      "num_tokens": 180913997.0,
      "step": 190
    },
    {
      "epoch": 0.27713625866050806,
      "grad_norm": 0.3927226802114006,
      "learning_rate": 1.946194665405828e-05,
      "loss": 0.7563,
      "mean_token_accuracy": 0.7716620303690434,
      "num_tokens": 185674484.0,
      "step": 195
    },
    {
      "epoch": 0.2842423165748801,
      "grad_norm": 0.3794392681437391,
      "learning_rate": 1.9426113835757637e-05,
      "loss": 0.7537,
      "mean_token_accuracy": 0.7706545531749726,
      "num_tokens": 190438181.0,
      "step": 200
    },
    {
      "epoch": 0.2842423165748801,
      "eval_loss": 0.7250556349754333,
      "eval_mean_token_accuracy": 0.7720574731366676,
      "eval_num_tokens": 190438181.0,
      "eval_runtime": 149.9052,
      "eval_samples_per_second": 24.275,
      "eval_steps_per_second": 0.76,
      "step": 200
    },
    {
      "epoch": 0.2913483744892521,
      "grad_norm": 0.40984266753765214,
      "learning_rate": 1.9389165190162114e-05,
      "loss": 0.753,
      "mean_token_accuracy": 0.7713685400784016,
      "num_tokens": 195189498.0,
      "step": 205
    },
    {
      "epoch": 0.29845443240362407,
      "grad_norm": 0.39761042206791,
      "learning_rate": 1.935110561019246e-05,
      "loss": 0.7424,
      "mean_token_accuracy": 0.7745413303375244,
      "num_tokens": 199953059.0,
      "step": 210
    },
    {
      "epoch": 0.3055604903179961,
      "grad_norm": 0.37462688474606465,
      "learning_rate": 1.931194013588481e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7734048135578633,
      "num_tokens": 204732059.0,
      "step": 215
    },
    {
      "epoch": 0.3126665482323681,
      "grad_norm": 0.39514621265905453,
      "learning_rate": 1.927167395372324e-05,
      "loss": 0.746,
      "mean_token_accuracy": 0.7722579926252365,
      "num_tokens": 209507798.0,
      "step": 220
    },
    {
      "epoch": 0.3197726061467401,
      "grad_norm": 0.3734047512754032,
      "learning_rate": 1.9230312395952955e-05,
      "loss": 0.7444,
      "mean_token_accuracy": 0.7728776805102825,
      "num_tokens": 214261827.0,
      "step": 225
    },
    {
      "epoch": 0.3268786640611121,
      "grad_norm": 0.39160079857564095,
      "learning_rate": 1.9187860939874176e-05,
      "loss": 0.7509,
      "mean_token_accuracy": 0.771727342903614,
      "num_tokens": 219027585.0,
      "step": 230
    },
    {
      "epoch": 0.3339847219754841,
      "grad_norm": 0.38214353202928264,
      "learning_rate": 1.9144325207116785e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7766963444650173,
      "num_tokens": 223775141.0,
      "step": 235
    },
    {
      "epoch": 0.3410907798898561,
      "grad_norm": 0.40133338005632113,
      "learning_rate": 1.909971096289591e-05,
      "loss": 0.7454,
      "mean_token_accuracy": 0.7735047489404678,
      "num_tokens": 228541180.0,
      "step": 240
    },
    {
      "epoch": 0.3481968378042281,
      "grad_norm": 0.4094942578953397,
      "learning_rate": 1.9054024115248448e-05,
      "loss": 0.7401,
      "mean_token_accuracy": 0.7752777233719825,
      "num_tokens": 233303417.0,
      "step": 245
    },
    {
      "epoch": 0.35530289571860013,
      "grad_norm": 0.37342705131252163,
      "learning_rate": 1.90072707142507e-05,
      "loss": 0.746,
      "mean_token_accuracy": 0.7732348993420601,
      "num_tokens": 238086815.0,
      "step": 250
    },
    {
      "epoch": 0.3624089536329721,
      "grad_norm": 0.37950791393566236,
      "learning_rate": 1.8959456951217187e-05,
      "loss": 0.7324,
      "mean_token_accuracy": 0.7766066655516625,
      "num_tokens": 242856686.0,
      "step": 255
    },
    {
      "epoch": 0.3695150115473441,
      "grad_norm": 0.36431061688841127,
      "learning_rate": 1.8910589157880766e-05,
      "loss": 0.7389,
      "mean_token_accuracy": 0.7757058747112751,
      "num_tokens": 247606311.0,
      "step": 260
    },
    {
      "epoch": 0.37662106946171614,
      "grad_norm": 0.4264408595860345,
      "learning_rate": 1.8860673805554167e-05,
      "loss": 0.74,
      "mean_token_accuracy": 0.7750592313706874,
      "num_tokens": 252376279.0,
      "step": 265
    },
    {
      "epoch": 0.3837271273760881,
      "grad_norm": 0.3649269356510504,
      "learning_rate": 1.8809717504273e-05,
      "loss": 0.7294,
      "mean_token_accuracy": 0.7773133426904678,
      "num_tokens": 257157622.0,
      "step": 270
    },
    {
      "epoch": 0.39083318529046013,
      "grad_norm": 0.4595703905257225,
      "learning_rate": 1.8757727001920446e-05,
      "loss": 0.7376,
      "mean_token_accuracy": 0.7763620682060719,
      "num_tokens": 261918809.0,
      "step": 275
    },
    {
      "epoch": 0.3979392432048321,
      "grad_norm": 0.5280527224785482,
      "learning_rate": 1.8704709183333653e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7755794525146484,
      "num_tokens": 266684284.0,
      "step": 280
    },
    {
      "epoch": 0.4050453011192041,
      "grad_norm": 0.37695307187112215,
      "learning_rate": 1.8650671069392034e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.776630100607872,
      "num_tokens": 271450936.0,
      "step": 285
    },
    {
      "epoch": 0.41215135903357614,
      "grad_norm": 0.4776029246227445,
      "learning_rate": 1.85956198160875e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7792015597224236,
      "num_tokens": 276196347.0,
      "step": 290
    },
    {
      "epoch": 0.4192574169479481,
      "grad_norm": 0.40416881485925626,
      "learning_rate": 1.853956271357685e-05,
      "loss": 0.7207,
      "mean_token_accuracy": 0.7793174132704734,
      "num_tokens": 280956754.0,
      "step": 295
    },
    {
      "epoch": 0.4263634748623201,
      "grad_norm": 0.36640042813026624,
      "learning_rate": 1.8482507185216365e-05,
      "loss": 0.7417,
      "mean_token_accuracy": 0.7740650460124016,
      "num_tokens": 285730218.0,
      "step": 300
    },
    {
      "epoch": 0.4263634748623201,
      "eval_loss": 0.7050633430480957,
      "eval_mean_token_accuracy": 0.7762745759989086,
      "eval_num_tokens": 285730218.0,
      "eval_runtime": 149.2344,
      "eval_samples_per_second": 24.384,
      "eval_steps_per_second": 0.764,
      "step": 300
    },
    {
      "epoch": 0.43346953277669215,
      "grad_norm": 0.39913292298186365,
      "learning_rate": 1.842446078657877e-05,
      "loss": 0.7328,
      "mean_token_accuracy": 0.7760866671800614,
      "num_tokens": 290497109.0,
      "step": 305
    },
    {
      "epoch": 0.4405755906910641,
      "grad_norm": 1.1471762547137223,
      "learning_rate": 1.8365431204452683e-05,
      "loss": 0.7364,
      "mean_token_accuracy": 0.7759052954614163,
      "num_tokens": 295276239.0,
      "step": 310
    },
    {
      "epoch": 0.44768164860543613,
      "grad_norm": 0.37744675876656947,
      "learning_rate": 1.8305426255824713e-05,
      "loss": 0.7317,
      "mean_token_accuracy": 0.7751947946846485,
      "num_tokens": 300042876.0,
      "step": 315
    },
    {
      "epoch": 0.45478770651980815,
      "grad_norm": 0.47691506104424974,
      "learning_rate": 1.824445388684426e-05,
      "loss": 0.7277,
      "mean_token_accuracy": 0.7777360931038857,
      "num_tokens": 304798068.0,
      "step": 320
    },
    {
      "epoch": 0.4618937644341801,
      "grad_norm": 0.3588696331537949,
      "learning_rate": 1.8182522171771293e-05,
      "loss": 0.726,
      "mean_token_accuracy": 0.7783321216702461,
      "num_tokens": 309546102.0,
      "step": 325
    },
    {
      "epoch": 0.46899982234855214,
      "grad_norm": 0.4952808278830928,
      "learning_rate": 1.8119639311907074e-05,
      "loss": 0.738,
      "mean_token_accuracy": 0.7744948998093605,
      "num_tokens": 314307721.0,
      "step": 330
    },
    {
      "epoch": 0.47610588026292416,
      "grad_norm": 0.37908339080858117,
      "learning_rate": 1.805581363450813e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7773234643042087,
      "num_tokens": 319073411.0,
      "step": 335
    },
    {
      "epoch": 0.48321193817729613,
      "grad_norm": 0.3852822861687584,
      "learning_rate": 1.7991053591683508e-05,
      "loss": 0.731,
      "mean_token_accuracy": 0.7765265628695488,
      "num_tokens": 323831947.0,
      "step": 340
    },
    {
      "epoch": 0.49031799609166815,
      "grad_norm": 0.36165483826370953,
      "learning_rate": 1.7925367759275495e-05,
      "loss": 0.7232,
      "mean_token_accuracy": 0.7792682178318501,
      "num_tokens": 328590613.0,
      "step": 345
    },
    {
      "epoch": 0.4974240540060402,
      "grad_norm": 0.4313963783745435,
      "learning_rate": 1.7858764835723984e-05,
      "loss": 0.7247,
      "mean_token_accuracy": 0.7771173417568207,
      "num_tokens": 333348383.0,
      "step": 350
    },
    {
      "epoch": 0.5045301119204122,
      "grad_norm": 0.4013920878913357,
      "learning_rate": 1.7791253640914566e-05,
      "loss": 0.7236,
      "mean_token_accuracy": 0.778332532197237,
      "num_tokens": 338109943.0,
      "step": 355
    },
    {
      "epoch": 0.5116361698347841,
      "grad_norm": 0.4057100730404507,
      "learning_rate": 1.7722843115010564e-05,
      "loss": 0.7221,
      "mean_token_accuracy": 0.7787548579275608,
      "num_tokens": 342887490.0,
      "step": 360
    },
    {
      "epoch": 0.5187422277491561,
      "grad_norm": 0.3675393081924701,
      "learning_rate": 1.7653542317269134e-05,
      "loss": 0.7171,
      "mean_token_accuracy": 0.7794929854571819,
      "num_tokens": 347628813.0,
      "step": 365
    },
    {
      "epoch": 0.5258482856635281,
      "grad_norm": 0.3635771034527727,
      "learning_rate": 1.7583360424841595e-05,
      "loss": 0.7272,
      "mean_token_accuracy": 0.7774313412606716,
      "num_tokens": 352403713.0,
      "step": 370
    },
    {
      "epoch": 0.5329543435779002,
      "grad_norm": 0.38584219141761933,
      "learning_rate": 1.7512306731558133e-05,
      "loss": 0.7194,
      "mean_token_accuracy": 0.7801453106105327,
      "num_tokens": 357150848.0,
      "step": 375
    },
    {
      "epoch": 0.5400604014922722,
      "grad_norm": 0.36346011493475233,
      "learning_rate": 1.744039064669709e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7780666872859001,
      "num_tokens": 361924581.0,
      "step": 380
    },
    {
      "epoch": 0.5471664594066442,
      "grad_norm": 0.404515192018998,
      "learning_rate": 1.7367621693738917e-05,
      "loss": 0.715,
      "mean_token_accuracy": 0.7817073427140713,
      "num_tokens": 366676773.0,
      "step": 385
    },
    {
      "epoch": 0.5542725173210161,
      "grad_norm": 0.36988377045483584,
      "learning_rate": 1.7294009509105052e-05,
      "loss": 0.7131,
      "mean_token_accuracy": 0.7806239545345306,
      "num_tokens": 371452085.0,
      "step": 390
    },
    {
      "epoch": 0.5613785752353881,
      "grad_norm": 0.36659851484970235,
      "learning_rate": 1.7219563840881783e-05,
      "loss": 0.7116,
      "mean_token_accuracy": 0.782407358288765,
      "num_tokens": 376207953.0,
      "step": 395
    },
    {
      "epoch": 0.5684846331497602,
      "grad_norm": 0.3624183655015318,
      "learning_rate": 1.71442945475294e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.7801115453243256,
      "num_tokens": 380979250.0,
      "step": 400
    },
    {
      "epoch": 0.5684846331497602,
      "eval_loss": 0.6918764114379883,
      "eval_mean_token_accuracy": 0.7797647902840062,
      "eval_num_tokens": 380979250.0,
      "eval_runtime": 150.2483,
      "eval_samples_per_second": 24.22,
      "eval_steps_per_second": 0.759,
      "step": 400
    },
    {
      "epoch": 0.5755906910641322,
      "grad_norm": 0.3779480713941837,
      "learning_rate": 1.7068211596576662e-05,
      "loss": 0.716,
      "mean_token_accuracy": 0.7807160533964634,
      "num_tokens": 385752024.0,
      "step": 405
    },
    {
      "epoch": 0.5826967489785042,
      "grad_norm": 0.3956915407192127,
      "learning_rate": 1.699132506330086e-05,
      "loss": 0.7168,
      "mean_token_accuracy": 0.780977015197277,
      "num_tokens": 390510208.0,
      "step": 410
    },
    {
      "epoch": 0.5898028068928762,
      "grad_norm": 0.37756059351173565,
      "learning_rate": 1.691364512939358e-05,
      "loss": 0.7138,
      "mean_token_accuracy": 0.7802788965404034,
      "num_tokens": 395264854.0,
      "step": 415
    },
    {
      "epoch": 0.5969088648072481,
      "grad_norm": 0.4087318897216221,
      "learning_rate": 1.6835182081612426e-05,
      "loss": 0.7136,
      "mean_token_accuracy": 0.782038314640522,
      "num_tokens": 400017717.0,
      "step": 420
    },
    {
      "epoch": 0.6040149227216202,
      "grad_norm": 0.40966025914497906,
      "learning_rate": 1.6755946310418777e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7809364423155785,
      "num_tokens": 404785855.0,
      "step": 425
    },
    {
      "epoch": 0.6111209806359922,
      "grad_norm": 0.34380470044932104,
      "learning_rate": 1.6675948308601826e-05,
      "loss": 0.7088,
      "mean_token_accuracy": 0.7824217259883881,
      "num_tokens": 409545265.0,
      "step": 430
    },
    {
      "epoch": 0.6182270385503642,
      "grad_norm": 0.3999223329715891,
      "learning_rate": 1.6595198669889086e-05,
      "loss": 0.7178,
      "mean_token_accuracy": 0.7794642865657806,
      "num_tokens": 414313757.0,
      "step": 435
    },
    {
      "epoch": 0.6253330964647362,
      "grad_norm": 0.4093202361120024,
      "learning_rate": 1.6513708087543507e-05,
      "loss": 0.7112,
      "mean_token_accuracy": 0.7812661081552505,
      "num_tokens": 419067741.0,
      "step": 440
    },
    {
      "epoch": 0.6324391543791081,
      "grad_norm": 0.3927913861855057,
      "learning_rate": 1.643148735294744e-05,
      "loss": 0.7085,
      "mean_token_accuracy": 0.7821477875113487,
      "num_tokens": 423849699.0,
      "step": 445
    },
    {
      "epoch": 0.6395452122934802,
      "grad_norm": 0.38988729195797683,
      "learning_rate": 1.634854735417356e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.7806262195110321,
      "num_tokens": 428613216.0,
      "step": 450
    },
    {
      "epoch": 0.6466512702078522,
      "grad_norm": 0.35125767095510474,
      "learning_rate": 1.6264899074543038e-05,
      "loss": 0.7244,
      "mean_token_accuracy": 0.7782423093914985,
      "num_tokens": 433373732.0,
      "step": 455
    },
    {
      "epoch": 0.6537573281222242,
      "grad_norm": 0.3717296312246723,
      "learning_rate": 1.6180553591171064e-05,
      "loss": 0.7134,
      "mean_token_accuracy": 0.7801944658160209,
      "num_tokens": 438144634.0,
      "step": 460
    },
    {
      "epoch": 0.6608633860365962,
      "grad_norm": 0.3514665073580472,
      "learning_rate": 1.6095522073499968e-05,
      "loss": 0.7094,
      "mean_token_accuracy": 0.782074099034071,
      "num_tokens": 442899589.0,
      "step": 465
    },
    {
      "epoch": 0.6679694439509682,
      "grad_norm": 0.3635349909050601,
      "learning_rate": 1.600981578182011e-05,
      "loss": 0.7125,
      "mean_token_accuracy": 0.7808018557727336,
      "num_tokens": 447672633.0,
      "step": 470
    },
    {
      "epoch": 0.6750755018653402,
      "grad_norm": 0.3548670898985585,
      "learning_rate": 1.5923446065778715e-05,
      "loss": 0.7162,
      "mean_token_accuracy": 0.7795430406928062,
      "num_tokens": 452431439.0,
      "step": 475
    },
    {
      "epoch": 0.6821815597797122,
      "grad_norm": 0.34932388566686257,
      "learning_rate": 1.5836424362876933e-05,
      "loss": 0.6984,
      "mean_token_accuracy": 0.7855889156460762,
      "num_tokens": 457177703.0,
      "step": 480
    },
    {
      "epoch": 0.6892876176940842,
      "grad_norm": 0.3981248379617205,
      "learning_rate": 1.5748762196955198e-05,
      "loss": 0.7036,
      "mean_token_accuracy": 0.7827964283525943,
      "num_tokens": 461930774.0,
      "step": 485
    },
    {
      "epoch": 0.6963936756084562,
      "grad_norm": 0.3393463040601756,
      "learning_rate": 1.5660471176667194e-05,
      "loss": 0.7092,
      "mean_token_accuracy": 0.7816402152180671,
      "num_tokens": 466702045.0,
      "step": 490
    },
    {
      "epoch": 0.7034997335228282,
      "grad_norm": 0.3747245118209944,
      "learning_rate": 1.5571562993942594e-05,
      "loss": 0.7063,
      "mean_token_accuracy": 0.7829745762050152,
      "num_tokens": 471461872.0,
      "step": 495
    },
    {
      "epoch": 0.7106057914372003,
      "grad_norm": 0.356673539185162,
      "learning_rate": 1.5482049422438732e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7823217682540416,
      "num_tokens": 476233238.0,
      "step": 500
    },
    {
      "epoch": 0.7106057914372003,
      "eval_loss": 0.6809196472167969,
      "eval_mean_token_accuracy": 0.7825243828589457,
      "eval_num_tokens": 476233238.0,
      "eval_runtime": 150.1867,
      "eval_samples_per_second": 24.23,
      "eval_steps_per_second": 0.759,
      "step": 500
    },
    {
      "epoch": 0.7177118493515722,
      "grad_norm": 0.342268344236882,
      "learning_rate": 1.5391942315981506e-05,
      "loss": 0.7124,
      "mean_token_accuracy": 0.7804363466799259,
      "num_tokens": 481010410.0,
      "step": 505
    },
    {
      "epoch": 0.7248179072659442,
      "grad_norm": 0.41188433212292186,
      "learning_rate": 1.530125360699561e-05,
      "loss": 0.7089,
      "mean_token_accuracy": 0.7815835013985634,
      "num_tokens": 485757825.0,
      "step": 510
    },
    {
      "epoch": 0.7319239651803162,
      "grad_norm": 0.456891730601901,
      "learning_rate": 1.520999530492441e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7851340644061565,
      "num_tokens": 490512360.0,
      "step": 515
    },
    {
      "epoch": 0.7390300230946882,
      "grad_norm": 0.347445377971525,
      "learning_rate": 1.511817949463956e-05,
      "loss": 0.7066,
      "mean_token_accuracy": 0.7829876273870469,
      "num_tokens": 495265025.0,
      "step": 520
    },
    {
      "epoch": 0.7461360810090603,
      "grad_norm": 0.3649440044406707,
      "learning_rate": 1.5025818334840695e-05,
      "loss": 0.7057,
      "mean_token_accuracy": 0.7825053557753563,
      "num_tokens": 500030371.0,
      "step": 525
    },
    {
      "epoch": 0.7532421389234323,
      "grad_norm": 0.3848276798074321,
      "learning_rate": 1.493292405644531e-05,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7862150557339191,
      "num_tokens": 504787581.0,
      "step": 530
    },
    {
      "epoch": 0.7603481968378042,
      "grad_norm": 0.3545112606618106,
      "learning_rate": 1.4839508960969071e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7828620508313179,
      "num_tokens": 509570758.0,
      "step": 535
    },
    {
      "epoch": 0.7674542547521762,
      "grad_norm": 0.37573645816136086,
      "learning_rate": 1.4745585418896799e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7837928868830204,
      "num_tokens": 514321600.0,
      "step": 540
    },
    {
      "epoch": 0.7745603126665482,
      "grad_norm": 0.367683939418967,
      "learning_rate": 1.4651165868044301e-05,
      "loss": 0.6995,
      "mean_token_accuracy": 0.7847208097577095,
      "num_tokens": 519082348.0,
      "step": 545
    },
    {
      "epoch": 0.7816663705809203,
      "grad_norm": 0.37333173963401106,
      "learning_rate": 1.45562628119113e-05,
      "loss": 0.7008,
      "mean_token_accuracy": 0.7843301363289357,
      "num_tokens": 523847671.0,
      "step": 550
    },
    {
      "epoch": 0.7887724284952923,
      "grad_norm": 0.35304534000101323,
      "learning_rate": 1.446088881802566e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.780696228891611,
      "num_tokens": 528620044.0,
      "step": 555
    },
    {
      "epoch": 0.7958784864096642,
      "grad_norm": 0.37158140908427656,
      "learning_rate": 1.4365056516279126e-05,
      "loss": 0.7016,
      "mean_token_accuracy": 0.7839049801230431,
      "num_tokens": 533367563.0,
      "step": 560
    },
    {
      "epoch": 0.8029845443240362,
      "grad_norm": 0.3666636970376427,
      "learning_rate": 1.426877859725482e-05,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7832373000681401,
      "num_tokens": 538117561.0,
      "step": 565
    },
    {
      "epoch": 0.8100906022384082,
      "grad_norm": 0.3480002426515768,
      "learning_rate": 1.4172067810546689e-05,
      "loss": 0.7024,
      "mean_token_accuracy": 0.7843490958213806,
      "num_tokens": 542889289.0,
      "step": 570
    },
    {
      "epoch": 0.8171966601527803,
      "grad_norm": 0.34159899298495605,
      "learning_rate": 1.4074936963071135e-05,
      "loss": 0.7034,
      "mean_token_accuracy": 0.7836663112044334,
      "num_tokens": 547637481.0,
      "step": 575
    },
    {
      "epoch": 0.8243027180671523,
      "grad_norm": 0.346312037596486,
      "learning_rate": 1.3977398917371074e-05,
      "loss": 0.6952,
      "mean_token_accuracy": 0.7860016152262688,
      "num_tokens": 552402659.0,
      "step": 580
    },
    {
      "epoch": 0.8314087759815243,
      "grad_norm": 0.4021455809002062,
      "learning_rate": 1.3879466589912598e-05,
      "loss": 0.6938,
      "mean_token_accuracy": 0.785366540402174,
      "num_tokens": 557156063.0,
      "step": 585
    },
    {
      "epoch": 0.8385148338958962,
      "grad_norm": 0.35612007953198216,
      "learning_rate": 1.3781152949374527e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7838830970227718,
      "num_tokens": 561916767.0,
      "step": 590
    },
    {
      "epoch": 0.8456208918102682,
      "grad_norm": 0.3655479416355732,
      "learning_rate": 1.3682471014931031e-05,
      "loss": 0.7019,
      "mean_token_accuracy": 0.7831911854445934,
      "num_tokens": 566684863.0,
      "step": 595
    },
    {
      "epoch": 0.8527269497246402,
      "grad_norm": 0.35487507828239484,
      "learning_rate": 1.3583433854527557e-05,
      "loss": 0.6967,
      "mean_token_accuracy": 0.7847252510488033,
      "num_tokens": 571452634.0,
      "step": 600
    },
    {
      "epoch": 0.8527269497246402,
      "eval_loss": 0.672866702079773,
      "eval_mean_token_accuracy": 0.784141309951481,
      "eval_num_tokens": 571452634.0,
      "eval_runtime": 149.8383,
      "eval_samples_per_second": 24.286,
      "eval_steps_per_second": 0.761,
      "step": 600
    },
    {
      "epoch": 0.8598330076390123,
      "grad_norm": 0.3456230748389429,
      "learning_rate": 1.3484054583150315e-05,
      "loss": 0.6906,
      "mean_token_accuracy": 0.7867132879793644,
      "num_tokens": 576198167.0,
      "step": 605
    },
    {
      "epoch": 0.8669390655533843,
      "grad_norm": 0.4168654830920555,
      "learning_rate": 1.3384346361089535e-05,
      "loss": 0.6885,
      "mean_token_accuracy": 0.7866604030132294,
      "num_tokens": 580952899.0,
      "step": 610
    },
    {
      "epoch": 0.8740451234677563,
      "grad_norm": 0.3723864301027309,
      "learning_rate": 1.3284322392196703e-05,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7859079904854298,
      "num_tokens": 585731060.0,
      "step": 615
    },
    {
      "epoch": 0.8811511813821282,
      "grad_norm": 0.3470961871922244,
      "learning_rate": 1.3183995922136048e-05,
      "loss": 0.712,
      "mean_token_accuracy": 0.7812197484076023,
      "num_tokens": 590504105.0,
      "step": 620
    },
    {
      "epoch": 0.8882572392965002,
      "grad_norm": 0.36809607687600115,
      "learning_rate": 1.308338023663049e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7837964847683907,
      "num_tokens": 595263948.0,
      "step": 625
    },
    {
      "epoch": 0.8953632972108723,
      "grad_norm": 0.40488523074148874,
      "learning_rate": 1.2982488659702269e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7849378556013107,
      "num_tokens": 600009699.0,
      "step": 630
    },
    {
      "epoch": 0.9024693551252443,
      "grad_norm": 0.36084599960401653,
      "learning_rate": 1.2881334551908524e-05,
      "loss": 0.6932,
      "mean_token_accuracy": 0.785707937926054,
      "num_tokens": 604750827.0,
      "step": 635
    },
    {
      "epoch": 0.9095754130396163,
      "grad_norm": 0.47712658403892855,
      "learning_rate": 1.2779931308572022e-05,
      "loss": 0.6932,
      "mean_token_accuracy": 0.7863863408565521,
      "num_tokens": 609500130.0,
      "step": 640
    },
    {
      "epoch": 0.9166814709539882,
      "grad_norm": 0.3494828834988836,
      "learning_rate": 1.2678292358007274e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7878520257771016,
      "num_tokens": 614261653.0,
      "step": 645
    },
    {
      "epoch": 0.9237875288683602,
      "grad_norm": 0.38391809182329456,
      "learning_rate": 1.2576431159742298e-05,
      "loss": 0.7083,
      "mean_token_accuracy": 0.7823263764381408,
      "num_tokens": 619054673.0,
      "step": 650
    },
    {
      "epoch": 0.9308935867827323,
      "grad_norm": 0.341786835416634,
      "learning_rate": 1.247436120273624e-05,
      "loss": 0.7049,
      "mean_token_accuracy": 0.7822027482092381,
      "num_tokens": 623817276.0,
      "step": 655
    },
    {
      "epoch": 0.9379996446971043,
      "grad_norm": 0.43974221008421865,
      "learning_rate": 1.237209600359311e-05,
      "loss": 0.6935,
      "mean_token_accuracy": 0.7854240909218788,
      "num_tokens": 628574712.0,
      "step": 660
    },
    {
      "epoch": 0.9451057026114763,
      "grad_norm": 0.4040988616310412,
      "learning_rate": 1.226964910477183e-05,
      "loss": 0.6898,
      "mean_token_accuracy": 0.7870114140212536,
      "num_tokens": 633335145.0,
      "step": 665
    },
    {
      "epoch": 0.9522117605258483,
      "grad_norm": 0.38559148131636184,
      "learning_rate": 1.2167034072792887e-05,
      "loss": 0.6937,
      "mean_token_accuracy": 0.7853186056017876,
      "num_tokens": 638086757.0,
      "step": 670
    },
    {
      "epoch": 0.9593178184402202,
      "grad_norm": 0.35182778835643425,
      "learning_rate": 1.2064264496441786e-05,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7859195664525032,
      "num_tokens": 642864800.0,
      "step": 675
    },
    {
      "epoch": 0.9664238763545923,
      "grad_norm": 0.37801611596609275,
      "learning_rate": 1.1961353984969557e-05,
      "loss": 0.689,
      "mean_token_accuracy": 0.7867573000490665,
      "num_tokens": 647632233.0,
      "step": 680
    },
    {
      "epoch": 0.9735299342689643,
      "grad_norm": 0.3676417856657156,
      "learning_rate": 1.1858316166290542e-05,
      "loss": 0.6933,
      "mean_token_accuracy": 0.7860686622560025,
      "num_tokens": 652408527.0,
      "step": 685
    },
    {
      "epoch": 0.9806359921833363,
      "grad_norm": 0.32727128259002763,
      "learning_rate": 1.1755164685177733e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7855750493705272,
      "num_tokens": 657175363.0,
      "step": 690
    },
    {
      "epoch": 0.9877420500977083,
      "grad_norm": 0.346480246739393,
      "learning_rate": 1.1651913201455865e-05,
      "loss": 0.6901,
      "mean_token_accuracy": 0.78620011433959,
      "num_tokens": 661940460.0,
      "step": 695
    },
    {
      "epoch": 0.9948481080120803,
      "grad_norm": 0.3382141590702366,
      "learning_rate": 1.154857538819249e-05,
      "loss": 0.6935,
      "mean_token_accuracy": 0.7859153963625432,
      "num_tokens": 666706092.0,
      "step": 700
    },
    {
      "epoch": 0.9948481080120803,
      "eval_loss": 0.6661998629570007,
      "eval_mean_token_accuracy": 0.7860168757145864,
      "eval_num_tokens": 666706092.0,
      "eval_runtime": 150.5473,
      "eval_samples_per_second": 24.172,
      "eval_steps_per_second": 0.757,
      "step": 700
    },
    {
      "epoch": 1.0014212115828744,
      "grad_norm": 1.0385755552019693,
      "learning_rate": 1.144516492988736e-05,
      "loss": 0.6823,
      "mean_token_accuracy": 0.7878129852784647,
      "num_tokens": 671095094.0,
      "step": 705
    },
    {
      "epoch": 1.0085272694972465,
      "grad_norm": 0.34901946667846845,
      "learning_rate": 1.134169552066023e-05,
      "loss": 0.6613,
      "mean_token_accuracy": 0.7918078258633614,
      "num_tokens": 675850987.0,
      "step": 710
    },
    {
      "epoch": 1.0156333274116185,
      "grad_norm": 0.35371147066198844,
      "learning_rate": 1.1238180862437431e-05,
      "loss": 0.6616,
      "mean_token_accuracy": 0.7930883727967739,
      "num_tokens": 680610147.0,
      "step": 715
    },
    {
      "epoch": 1.0227393853259905,
      "grad_norm": 0.371067048471263,
      "learning_rate": 1.1134634663137373e-05,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7940364375710487,
      "num_tokens": 685353908.0,
      "step": 720
    },
    {
      "epoch": 1.0298454432403623,
      "grad_norm": 0.3481611518266803,
      "learning_rate": 1.1031070634855314e-05,
      "loss": 0.6593,
      "mean_token_accuracy": 0.7930267058312893,
      "num_tokens": 690111045.0,
      "step": 725
    },
    {
      "epoch": 1.0369515011547343,
      "grad_norm": 0.35013978798256634,
      "learning_rate": 1.0927502492047492e-05,
      "loss": 0.6673,
      "mean_token_accuracy": 0.7919997818768024,
      "num_tokens": 694881554.0,
      "step": 730
    },
    {
      "epoch": 1.0440575590691064,
      "grad_norm": 0.33993734586176455,
      "learning_rate": 1.0823943949715022e-05,
      "loss": 0.67,
      "mean_token_accuracy": 0.7910104177892208,
      "num_tokens": 699670214.0,
      "step": 735
    },
    {
      "epoch": 1.0511636169834784,
      "grad_norm": 0.3412860380013659,
      "learning_rate": 1.0720408721587671e-05,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7910432547330857,
      "num_tokens": 704426344.0,
      "step": 740
    },
    {
      "epoch": 1.0582696748978504,
      "grad_norm": 0.3627320420498548,
      "learning_rate": 1.061691051830783e-05,
      "loss": 0.668,
      "mean_token_accuracy": 0.7916376106441021,
      "num_tokens": 709184272.0,
      "step": 745
    },
    {
      "epoch": 1.0653757328122224,
      "grad_norm": 0.3551359244101548,
      "learning_rate": 1.0513463045614873e-05,
      "loss": 0.6732,
      "mean_token_accuracy": 0.7899613387882709,
      "num_tokens": 713964117.0,
      "step": 750
    },
    {
      "epoch": 1.0724817907265944,
      "grad_norm": 0.35291410759778385,
      "learning_rate": 1.0410080002530188e-05,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7933160819113254,
      "num_tokens": 718714498.0,
      "step": 755
    },
    {
      "epoch": 1.0795878486409665,
      "grad_norm": 0.5183974245642026,
      "learning_rate": 1.030677507954307e-05,
      "loss": 0.669,
      "mean_token_accuracy": 0.7922067753970623,
      "num_tokens": 723480315.0,
      "step": 760
    },
    {
      "epoch": 1.0866939065553385,
      "grad_norm": 0.36178536189786653,
      "learning_rate": 1.0203561956797777e-05,
      "loss": 0.6592,
      "mean_token_accuracy": 0.7924857877194882,
      "num_tokens": 728257943.0,
      "step": 765
    },
    {
      "epoch": 1.0937999644697105,
      "grad_norm": 0.350161084492629,
      "learning_rate": 1.0100454302281917e-05,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7903590828180314,
      "num_tokens": 733027792.0,
      "step": 770
    },
    {
      "epoch": 1.1009060223840825,
      "grad_norm": 0.3924504778281434,
      "learning_rate": 9.997465770016488e-06,
      "loss": 0.665,
      "mean_token_accuracy": 0.7942788422107696,
      "num_tokens": 737777094.0,
      "step": 775
    },
    {
      "epoch": 1.1080120802984545,
      "grad_norm": 0.3784959263919085,
      "learning_rate": 9.894609998247735e-06,
      "loss": 0.667,
      "mean_token_accuracy": 0.7912828728556633,
      "num_tokens": 742543159.0,
      "step": 780
    },
    {
      "epoch": 1.1151181382128263,
      "grad_norm": 0.3498944950913003,
      "learning_rate": 9.791900607641104e-06,
      "loss": 0.6635,
      "mean_token_accuracy": 0.7930950812995434,
      "num_tokens": 747308343.0,
      "step": 785
    },
    {
      "epoch": 1.1222241961271984,
      "grad_norm": 0.3820207111204085,
      "learning_rate": 9.68935119947753e-06,
      "loss": 0.668,
      "mean_token_accuracy": 0.7914499528706074,
      "num_tokens": 752079771.0,
      "step": 790
    },
    {
      "epoch": 1.1293302540415704,
      "grad_norm": 0.37200401458209914,
      "learning_rate": 9.586975353852284e-06,
      "loss": 0.6639,
      "mean_token_accuracy": 0.7913541235029697,
      "num_tokens": 756847538.0,
      "step": 795
    },
    {
      "epoch": 1.1364363119559424,
      "grad_norm": 0.3953405586571805,
      "learning_rate": 9.484786627876655e-06,
      "loss": 0.6697,
      "mean_token_accuracy": 0.790704844892025,
      "num_tokens": 761615758.0,
      "step": 800
    },
    {
      "epoch": 1.1364363119559424,
      "eval_loss": 0.6620959043502808,
      "eval_mean_token_accuracy": 0.7867940282612517,
      "eval_num_tokens": 761615758.0,
      "eval_runtime": 149.9826,
      "eval_samples_per_second": 24.263,
      "eval_steps_per_second": 0.76,
      "step": 800
    },
    {
      "epoch": 1.1435423698703144,
      "grad_norm": 0.3501237218712183,
      "learning_rate": 9.382798553882605e-06,
      "loss": 0.6685,
      "mean_token_accuracy": 0.789706601947546,
      "num_tokens": 766396725.0,
      "step": 805
    },
    {
      "epoch": 1.1506484277846865,
      "grad_norm": 0.35257144104430527,
      "learning_rate": 9.281024637630794e-06,
      "loss": 0.656,
      "mean_token_accuracy": 0.7935691051185131,
      "num_tokens": 771153408.0,
      "step": 810
    },
    {
      "epoch": 1.1577544856990585,
      "grad_norm": 0.3686235797828206,
      "learning_rate": 9.179478356522055e-06,
      "loss": 0.6617,
      "mean_token_accuracy": 0.7928701542317868,
      "num_tokens": 775910085.0,
      "step": 815
    },
    {
      "epoch": 1.1648605436134305,
      "grad_norm": 0.35944613382478946,
      "learning_rate": 9.078173157812669e-06,
      "loss": 0.6673,
      "mean_token_accuracy": 0.7925907090306282,
      "num_tokens": 780683650.0,
      "step": 820
    },
    {
      "epoch": 1.1719666015278025,
      "grad_norm": 0.33694316783063644,
      "learning_rate": 8.97712245683359e-06,
      "loss": 0.6686,
      "mean_token_accuracy": 0.7904776819050312,
      "num_tokens": 785473480.0,
      "step": 825
    },
    {
      "epoch": 1.1790726594421745,
      "grad_norm": 0.37743982477075316,
      "learning_rate": 8.876339635213951e-06,
      "loss": 0.6672,
      "mean_token_accuracy": 0.7913396395742893,
      "num_tokens": 790244466.0,
      "step": 830
    },
    {
      "epoch": 1.1861787173565466,
      "grad_norm": 0.3852825297524554,
      "learning_rate": 8.775838039108975e-06,
      "loss": 0.6577,
      "mean_token_accuracy": 0.7940163776278496,
      "num_tokens": 794986608.0,
      "step": 835
    },
    {
      "epoch": 1.1932847752709184,
      "grad_norm": 0.3584263516575822,
      "learning_rate": 8.67563097743263e-06,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7941929534077644,
      "num_tokens": 799743654.0,
      "step": 840
    },
    {
      "epoch": 1.2003908331852904,
      "grad_norm": 0.36222679284602094,
      "learning_rate": 8.575731720095194e-06,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7949050404131413,
      "num_tokens": 804510301.0,
      "step": 845
    },
    {
      "epoch": 1.2074968910996624,
      "grad_norm": 0.36601316470081835,
      "learning_rate": 8.476153496245978e-06,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7888801738619804,
      "num_tokens": 809295294.0,
      "step": 850
    },
    {
      "epoch": 1.2146029490140344,
      "grad_norm": 0.35840290876863273,
      "learning_rate": 8.376909492521465e-06,
      "loss": 0.6651,
      "mean_token_accuracy": 0.7920581080019474,
      "num_tokens": 814063402.0,
      "step": 855
    },
    {
      "epoch": 1.2217090069284064,
      "grad_norm": 0.35918850454361245,
      "learning_rate": 8.278012851299082e-06,
      "loss": 0.6604,
      "mean_token_accuracy": 0.793212516605854,
      "num_tokens": 818822580.0,
      "step": 860
    },
    {
      "epoch": 1.2288150648427785,
      "grad_norm": 0.36931586095521574,
      "learning_rate": 8.179476668956799e-06,
      "loss": 0.6697,
      "mean_token_accuracy": 0.789932218939066,
      "num_tokens": 823577622.0,
      "step": 865
    },
    {
      "epoch": 1.2359211227571505,
      "grad_norm": 0.3618218146346168,
      "learning_rate": 8.081313994138857e-06,
      "loss": 0.6573,
      "mean_token_accuracy": 0.7943486146628856,
      "num_tokens": 828319732.0,
      "step": 870
    },
    {
      "epoch": 1.2430271806715225,
      "grad_norm": 0.3947688956702076,
      "learning_rate": 7.983537826027808e-06,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7911505416035652,
      "num_tokens": 833074724.0,
      "step": 875
    },
    {
      "epoch": 1.2501332385858945,
      "grad_norm": 0.49837585254908734,
      "learning_rate": 7.886161112623072e-06,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7948375590145588,
      "num_tokens": 837847437.0,
      "step": 880
    },
    {
      "epoch": 1.2572392965002666,
      "grad_norm": 0.3875626366429437,
      "learning_rate": 7.789196749026349e-06,
      "loss": 0.6519,
      "mean_token_accuracy": 0.7962292313575745,
      "num_tokens": 842595165.0,
      "step": 885
    },
    {
      "epoch": 1.2643453544146386,
      "grad_norm": 0.3461153977326627,
      "learning_rate": 7.692657575733928e-06,
      "loss": 0.6591,
      "mean_token_accuracy": 0.7930607885122299,
      "num_tokens": 847377566.0,
      "step": 890
    },
    {
      "epoch": 1.2714514123290104,
      "grad_norm": 0.3539458234583982,
      "learning_rate": 7.596556376936328e-06,
      "loss": 0.6585,
      "mean_token_accuracy": 0.7939456604421139,
      "num_tokens": 852138878.0,
      "step": 895
    },
    {
      "epoch": 1.2785574702433826,
      "grad_norm": 0.3776944250010228,
      "learning_rate": 7.500905878825335e-06,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7954832412302494,
      "num_tokens": 856908856.0,
      "step": 900
    },
    {
      "epoch": 1.2785574702433826,
      "eval_loss": 0.6587108373641968,
      "eval_mean_token_accuracy": 0.7879830112582759,
      "eval_num_tokens": 856908856.0,
      "eval_runtime": 150.1109,
      "eval_samples_per_second": 24.242,
      "eval_steps_per_second": 0.759,
      "step": 900
    },
    {
      "epoch": 1.2856635281577544,
      "grad_norm": 0.35417475896948203,
      "learning_rate": 7.405718747908743e-06,
      "loss": 0.6554,
      "mean_token_accuracy": 0.7936457127332688,
      "num_tokens": 861668793.0,
      "step": 905
    },
    {
      "epoch": 1.2927695860721264,
      "grad_norm": 0.3717646188257627,
      "learning_rate": 7.311007589332986e-06,
      "loss": 0.6587,
      "mean_token_accuracy": 0.7932697109878063,
      "num_tokens": 866418403.0,
      "step": 910
    },
    {
      "epoch": 1.2998756439864985,
      "grad_norm": 0.38878740217599195,
      "learning_rate": 7.216784945213913e-06,
      "loss": 0.6625,
      "mean_token_accuracy": 0.7936202257871627,
      "num_tokens": 871159945.0,
      "step": 915
    },
    {
      "epoch": 1.3069817019008705,
      "grad_norm": 0.3422730868265931,
      "learning_rate": 7.123063292975889e-06,
      "loss": 0.6525,
      "mean_token_accuracy": 0.794795686006546,
      "num_tokens": 875924929.0,
      "step": 920
    },
    {
      "epoch": 1.3140877598152425,
      "grad_norm": 0.3694651695886196,
      "learning_rate": 7.02985504369949e-06,
      "loss": 0.6547,
      "mean_token_accuracy": 0.7951326429843902,
      "num_tokens": 880666672.0,
      "step": 925
    },
    {
      "epoch": 1.3211938177296145,
      "grad_norm": 0.3621570732014425,
      "learning_rate": 6.937172540477944e-06,
      "loss": 0.6654,
      "mean_token_accuracy": 0.7919820554554462,
      "num_tokens": 885436601.0,
      "step": 930
    },
    {
      "epoch": 1.3282998756439865,
      "grad_norm": 0.7187487529688721,
      "learning_rate": 6.8450280567826074e-06,
      "loss": 0.6636,
      "mean_token_accuracy": 0.792605972290039,
      "num_tokens": 890209266.0,
      "step": 935
    },
    {
      "epoch": 1.3354059335583586,
      "grad_norm": 0.3607746297969546,
      "learning_rate": 6.753433794837663e-06,
      "loss": 0.655,
      "mean_token_accuracy": 0.7943654432892799,
      "num_tokens": 894978447.0,
      "step": 940
    },
    {
      "epoch": 1.3425119914727306,
      "grad_norm": 0.34656870317990024,
      "learning_rate": 6.662401884004226e-06,
      "loss": 0.6594,
      "mean_token_accuracy": 0.7929094567894935,
      "num_tokens": 899731953.0,
      "step": 945
    },
    {
      "epoch": 1.3496180493871024,
      "grad_norm": 0.3556403292186795,
      "learning_rate": 6.571944379174128e-06,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7939096741378308,
      "num_tokens": 904484204.0,
      "step": 950
    },
    {
      "epoch": 1.3567241073014746,
      "grad_norm": 0.3481029203946856,
      "learning_rate": 6.482073259173533e-06,
      "loss": 0.6558,
      "mean_token_accuracy": 0.795223805308342,
      "num_tokens": 909254980.0,
      "step": 955
    },
    {
      "epoch": 1.3638301652158464,
      "grad_norm": 0.36246749377385634,
      "learning_rate": 6.39280042517666e-06,
      "loss": 0.6576,
      "mean_token_accuracy": 0.794485367834568,
      "num_tokens": 914013636.0,
      "step": 960
    },
    {
      "epoch": 1.3709362231302185,
      "grad_norm": 0.35744165552118257,
      "learning_rate": 6.304137699129758e-06,
      "loss": 0.6521,
      "mean_token_accuracy": 0.7954901576042175,
      "num_tokens": 918774652.0,
      "step": 965
    },
    {
      "epoch": 1.3780422810445905,
      "grad_norm": 0.34190890162690535,
      "learning_rate": 6.216096822185591e-06,
      "loss": 0.6596,
      "mean_token_accuracy": 0.7934505857527256,
      "num_tokens": 923523836.0,
      "step": 970
    },
    {
      "epoch": 1.3851483389589625,
      "grad_norm": 0.3417362920171981,
      "learning_rate": 6.12868945314862e-06,
      "loss": 0.6647,
      "mean_token_accuracy": 0.7919908218085766,
      "num_tokens": 928304038.0,
      "step": 975
    },
    {
      "epoch": 1.3922543968733345,
      "grad_norm": 0.3435048840223176,
      "learning_rate": 6.041927166931078e-06,
      "loss": 0.6577,
      "mean_token_accuracy": 0.7943571574985981,
      "num_tokens": 933073919.0,
      "step": 980
    },
    {
      "epoch": 1.3993604547877065,
      "grad_norm": 0.40933619958354717,
      "learning_rate": 5.9558214530201784e-06,
      "loss": 0.6575,
      "mean_token_accuracy": 0.7943412482738494,
      "num_tokens": 937846004.0,
      "step": 985
    },
    {
      "epoch": 1.4064665127020786,
      "grad_norm": 0.3515117903938119,
      "learning_rate": 5.870383713956601e-06,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7938548773527145,
      "num_tokens": 942601267.0,
      "step": 990
    },
    {
      "epoch": 1.4135725706164506,
      "grad_norm": 0.38419237136120965,
      "learning_rate": 5.785625263824531e-06,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7948469713330268,
      "num_tokens": 947375335.0,
      "step": 995
    },
    {
      "epoch": 1.4206786285308226,
      "grad_norm": 0.38108767014595607,
      "learning_rate": 5.701557326753375e-06,
      "loss": 0.6504,
      "mean_token_accuracy": 0.7960710853338242,
      "num_tokens": 952105402.0,
      "step": 1000
    },
    {
      "epoch": 1.4206786285308226,
      "eval_loss": 0.6552348136901855,
      "eval_mean_token_accuracy": 0.7893575147578591,
      "eval_num_tokens": 952105402.0,
      "eval_runtime": 149.5045,
      "eval_samples_per_second": 24.34,
      "eval_steps_per_second": 0.763,
      "step": 1000
    },
    {
      "epoch": 1.4277846864451944,
      "grad_norm": 0.3429174906843955,
      "learning_rate": 5.6181910354314265e-06,
      "loss": 0.6596,
      "mean_token_accuracy": 0.7940759062767029,
      "num_tokens": 956874826.0,
      "step": 1005
    },
    {
      "epoch": 1.4348907443595666,
      "grad_norm": 0.37307240441832,
      "learning_rate": 5.5355374296316e-06,
      "loss": 0.6589,
      "mean_token_accuracy": 0.7940549589693546,
      "num_tokens": 961632193.0,
      "step": 1010
    },
    {
      "epoch": 1.4419968022739384,
      "grad_norm": 0.3463607871324184,
      "learning_rate": 5.4536074547495055e-06,
      "loss": 0.6576,
      "mean_token_accuracy": 0.7948333404958248,
      "num_tokens": 966392410.0,
      "step": 1015
    },
    {
      "epoch": 1.4491028601883105,
      "grad_norm": 0.34652042003246536,
      "learning_rate": 5.372411960353996e-06,
      "loss": 0.6636,
      "mean_token_accuracy": 0.7924063883721828,
      "num_tokens": 971170949.0,
      "step": 1020
    },
    {
      "epoch": 1.4562089181026825,
      "grad_norm": 0.33480848996072166,
      "learning_rate": 5.2919616987504205e-06,
      "loss": 0.6436,
      "mean_token_accuracy": 0.7979453206062317,
      "num_tokens": 975920452.0,
      "step": 1025
    },
    {
      "epoch": 1.4633149760170545,
      "grad_norm": 0.35444982881545845,
      "learning_rate": 5.212267323556754e-06,
      "loss": 0.6488,
      "mean_token_accuracy": 0.7975021339952946,
      "num_tokens": 980657772.0,
      "step": 1030
    },
    {
      "epoch": 1.4704210339314265,
      "grad_norm": 0.3236146641950486,
      "learning_rate": 5.1333393882927776e-06,
      "loss": 0.6656,
      "mean_token_accuracy": 0.7911154888570309,
      "num_tokens": 985424225.0,
      "step": 1035
    },
    {
      "epoch": 1.4775270918457986,
      "grad_norm": 0.36258575664825715,
      "learning_rate": 5.055188344982549e-06,
      "loss": 0.653,
      "mean_token_accuracy": 0.7950214244425297,
      "num_tokens": 990170268.0,
      "step": 1040
    },
    {
      "epoch": 1.4846331497601706,
      "grad_norm": 0.3626547592700128,
      "learning_rate": 4.977824542770279e-06,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7932340361177921,
      "num_tokens": 994933612.0,
      "step": 1045
    },
    {
      "epoch": 1.4917392076745426,
      "grad_norm": 0.3424014804959087,
      "learning_rate": 4.901258226549855e-06,
      "loss": 0.6499,
      "mean_token_accuracy": 0.7964041963219642,
      "num_tokens": 999695033.0,
      "step": 1050
    },
    {
      "epoch": 1.4988452655889146,
      "grad_norm": 0.3457300931458133,
      "learning_rate": 4.825499535608169e-06,
      "loss": 0.659,
      "mean_token_accuracy": 0.7942204736173153,
      "num_tokens": 1004453306.0,
      "step": 1055
    },
    {
      "epoch": 1.5059513235032864,
      "grad_norm": 0.407404649365606,
      "learning_rate": 4.750558502282403e-06,
      "loss": 0.6466,
      "mean_token_accuracy": 0.7969782948493958,
      "num_tokens": 1009222958.0,
      "step": 1060
    },
    {
      "epoch": 1.5130573814176587,
      "grad_norm": 0.36455466451492874,
      "learning_rate": 4.676445050631517e-06,
      "loss": 0.6669,
      "mean_token_accuracy": 0.7919491566717625,
      "num_tokens": 1013988411.0,
      "step": 1065
    },
    {
      "epoch": 1.5201634393320305,
      "grad_norm": 0.3352428252596833,
      "learning_rate": 4.603168995122048e-06,
      "loss": 0.653,
      "mean_token_accuracy": 0.7959543123841286,
      "num_tokens": 1018736541.0,
      "step": 1070
    },
    {
      "epoch": 1.5272694972464027,
      "grad_norm": 0.336146168164258,
      "learning_rate": 4.530740039328427e-06,
      "loss": 0.6527,
      "mean_token_accuracy": 0.795566051453352,
      "num_tokens": 1023492540.0,
      "step": 1075
    },
    {
      "epoch": 1.5343755551607745,
      "grad_norm": 0.35998475713849004,
      "learning_rate": 4.4591677746479935e-06,
      "loss": 0.6542,
      "mean_token_accuracy": 0.7954114884138107,
      "num_tokens": 1028251642.0,
      "step": 1080
    },
    {
      "epoch": 1.5414816130751465,
      "grad_norm": 0.4210390284353885,
      "learning_rate": 4.38846167903085e-06,
      "loss": 0.6501,
      "mean_token_accuracy": 0.7963161066174507,
      "num_tokens": 1033004523.0,
      "step": 1085
    },
    {
      "epoch": 1.5485876709895185,
      "grad_norm": 0.4960482537413715,
      "learning_rate": 4.318631115724741e-06,
      "loss": 0.6553,
      "mean_token_accuracy": 0.7946652464568615,
      "num_tokens": 1037760729.0,
      "step": 1090
    },
    {
      "epoch": 1.5556937289038906,
      "grad_norm": 0.3799332566110809,
      "learning_rate": 4.2496853320351424e-06,
      "loss": 0.6607,
      "mean_token_accuracy": 0.7947109803557396,
      "num_tokens": 1042523723.0,
      "step": 1095
    },
    {
      "epoch": 1.5627997868182626,
      "grad_norm": 0.33946558046244235,
      "learning_rate": 4.1816334581006656e-06,
      "loss": 0.6651,
      "mean_token_accuracy": 0.792590418457985,
      "num_tokens": 1047291640.0,
      "step": 1100
    },
    {
      "epoch": 1.5627997868182626,
      "eval_loss": 0.6525910496711731,
      "eval_mean_token_accuracy": 0.7899543700510996,
      "eval_num_tokens": 1047291640.0,
      "eval_runtime": 150.3086,
      "eval_samples_per_second": 24.21,
      "eval_steps_per_second": 0.758,
      "step": 1100
    },
    {
      "epoch": 1.5699058447326346,
      "grad_norm": 0.32287060668756534,
      "learning_rate": 4.114484505684019e-06,
      "loss": 0.6541,
      "mean_token_accuracy": 0.7952132284641266,
      "num_tokens": 1052042031.0,
      "step": 1105
    },
    {
      "epoch": 1.5770119026470066,
      "grad_norm": 0.3370631693518313,
      "learning_rate": 4.048247366978606e-06,
      "loss": 0.658,
      "mean_token_accuracy": 0.7935857936739922,
      "num_tokens": 1056804804.0,
      "step": 1110
    },
    {
      "epoch": 1.5841179605613784,
      "grad_norm": 0.414769500614358,
      "learning_rate": 3.9829308134309995e-06,
      "loss": 0.6475,
      "mean_token_accuracy": 0.7969807527959347,
      "num_tokens": 1061577783.0,
      "step": 1115
    },
    {
      "epoch": 1.5912240184757507,
      "grad_norm": 0.35600756916022547,
      "learning_rate": 3.9185434945793725e-06,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7951311826705932,
      "num_tokens": 1066355020.0,
      "step": 1120
    },
    {
      "epoch": 1.5983300763901225,
      "grad_norm": 0.36188029855593157,
      "learning_rate": 3.855093936908081e-06,
      "loss": 0.6664,
      "mean_token_accuracy": 0.7921065390110016,
      "num_tokens": 1071139121.0,
      "step": 1125
    },
    {
      "epoch": 1.6054361343044947,
      "grad_norm": 0.3632538405728989,
      "learning_rate": 3.7925905427185504e-06,
      "loss": 0.6569,
      "mean_token_accuracy": 0.7936886362731457,
      "num_tokens": 1075914044.0,
      "step": 1130
    },
    {
      "epoch": 1.6125421922188665,
      "grad_norm": 0.3669437696101656,
      "learning_rate": 3.7310415890166e-06,
      "loss": 0.6512,
      "mean_token_accuracy": 0.7960372731089592,
      "num_tokens": 1080682478.0,
      "step": 1135
    },
    {
      "epoch": 1.6196482501332385,
      "grad_norm": 0.3331543509156551,
      "learning_rate": 3.6704552264163695e-06,
      "loss": 0.6561,
      "mean_token_accuracy": 0.7935027062892914,
      "num_tokens": 1085456231.0,
      "step": 1140
    },
    {
      "epoch": 1.6267543080476106,
      "grad_norm": 0.3299992523767914,
      "learning_rate": 3.6108394780609513e-06,
      "loss": 0.6506,
      "mean_token_accuracy": 0.7957184061408042,
      "num_tokens": 1090215557.0,
      "step": 1145
    },
    {
      "epoch": 1.6338603659619826,
      "grad_norm": 0.3450779481067245,
      "learning_rate": 3.552202238559953e-06,
      "loss": 0.6429,
      "mean_token_accuracy": 0.798128329962492,
      "num_tokens": 1094959524.0,
      "step": 1150
    },
    {
      "epoch": 1.6409664238763546,
      "grad_norm": 0.3531321257611881,
      "learning_rate": 3.4945512729440413e-06,
      "loss": 0.6503,
      "mean_token_accuracy": 0.7954187601804733,
      "num_tokens": 1099731395.0,
      "step": 1155
    },
    {
      "epoch": 1.6480724817907266,
      "grad_norm": 0.3498741705682094,
      "learning_rate": 3.437894215636661e-06,
      "loss": 0.6578,
      "mean_token_accuracy": 0.7941137261688709,
      "num_tokens": 1104494157.0,
      "step": 1160
    },
    {
      "epoch": 1.6551785397050987,
      "grad_norm": 0.3777101584751149,
      "learning_rate": 3.382238569443045e-06,
      "loss": 0.6529,
      "mean_token_accuracy": 0.7957448020577431,
      "num_tokens": 1109252674.0,
      "step": 1165
    },
    {
      "epoch": 1.6622845976194704,
      "grad_norm": 0.3915263029819705,
      "learning_rate": 3.3275917045566596e-06,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7957381546497345,
      "num_tokens": 1114004017.0,
      "step": 1170
    },
    {
      "epoch": 1.6693906555338427,
      "grad_norm": 0.33568853526043657,
      "learning_rate": 3.2739608575832056e-06,
      "loss": 0.6412,
      "mean_token_accuracy": 0.7980836987495422,
      "num_tokens": 1118768157.0,
      "step": 1175
    },
    {
      "epoch": 1.6764967134482145,
      "grad_norm": 0.34752341789739466,
      "learning_rate": 3.2213531305823125e-06,
      "loss": 0.6613,
      "mean_token_accuracy": 0.7935202896595002,
      "num_tokens": 1123535145.0,
      "step": 1180
    },
    {
      "epoch": 1.6836027713625867,
      "grad_norm": 0.34083716495360583,
      "learning_rate": 3.1697754901270477e-06,
      "loss": 0.6507,
      "mean_token_accuracy": 0.7964440450072289,
      "num_tokens": 1128307445.0,
      "step": 1185
    },
    {
      "epoch": 1.6907088292769585,
      "grad_norm": 0.3774093707155365,
      "learning_rate": 3.1192347663813684e-06,
      "loss": 0.6547,
      "mean_token_accuracy": 0.7946882367134094,
      "num_tokens": 1133071242.0,
      "step": 1190
    },
    {
      "epoch": 1.6978148871913306,
      "grad_norm": 0.36333647453863616,
      "learning_rate": 3.0697376521956377e-06,
      "loss": 0.6526,
      "mean_token_accuracy": 0.7956908911466598,
      "num_tokens": 1137831284.0,
      "step": 1195
    },
    {
      "epoch": 1.7049209451057026,
      "grad_norm": 0.33780601240351715,
      "learning_rate": 3.021290702220331e-06,
      "loss": 0.6561,
      "mean_token_accuracy": 0.7948304824531078,
      "num_tokens": 1142587626.0,
      "step": 1200
    },
    {
      "epoch": 1.7049209451057026,
      "eval_loss": 0.6507056355476379,
      "eval_mean_token_accuracy": 0.790492679466281,
      "eval_num_tokens": 1142587626.0,
      "eval_runtime": 149.5446,
      "eval_samples_per_second": 24.334,
      "eval_steps_per_second": 0.762,
      "step": 1200
    },
    {
      "epoch": 1.7120270030200746,
      "grad_norm": 0.3537743025090112,
      "learning_rate": 2.9739003320380237e-06,
      "loss": 0.6624,
      "mean_token_accuracy": 0.793489520996809,
      "num_tokens": 1147357460.0,
      "step": 1205
    },
    {
      "epoch": 1.7191330609344466,
      "grad_norm": 0.4133446002595886,
      "learning_rate": 2.927572817313823e-06,
      "loss": 0.6585,
      "mean_token_accuracy": 0.7936319254338742,
      "num_tokens": 1152138440.0,
      "step": 1210
    },
    {
      "epoch": 1.7262391188488186,
      "grad_norm": 0.3563260686728207,
      "learning_rate": 2.8823142929643043e-06,
      "loss": 0.6426,
      "mean_token_accuracy": 0.797927625477314,
      "num_tokens": 1156890428.0,
      "step": 1215
    },
    {
      "epoch": 1.7333451767631907,
      "grad_norm": 0.3989494711298427,
      "learning_rate": 2.838130752345092e-06,
      "loss": 0.6582,
      "mean_token_accuracy": 0.7947382763028145,
      "num_tokens": 1161657895.0,
      "step": 1220
    },
    {
      "epoch": 1.7404512346775625,
      "grad_norm": 0.3432409644849126,
      "learning_rate": 2.7950280464572066e-06,
      "loss": 0.6541,
      "mean_token_accuracy": 0.7953485876321793,
      "num_tokens": 1166423043.0,
      "step": 1225
    },
    {
      "epoch": 1.7475572925919347,
      "grad_norm": 0.3368844710605464,
      "learning_rate": 2.7530118831722286e-06,
      "loss": 0.6481,
      "mean_token_accuracy": 0.796825060248375,
      "num_tokens": 1171166643.0,
      "step": 1230
    },
    {
      "epoch": 1.7546633505063065,
      "grad_norm": 0.45251140872957446,
      "learning_rate": 2.7120878264764437e-06,
      "loss": 0.6473,
      "mean_token_accuracy": 0.7977175071835518,
      "num_tokens": 1175924107.0,
      "step": 1235
    },
    {
      "epoch": 1.7617694084206788,
      "grad_norm": 0.3545121531531998,
      "learning_rate": 2.67226129573403e-06,
      "loss": 0.6512,
      "mean_token_accuracy": 0.7960038974881172,
      "num_tokens": 1180681893.0,
      "step": 1240
    },
    {
      "epoch": 1.7688754663350506,
      "grad_norm": 0.3323525365320921,
      "learning_rate": 2.633537564969398e-06,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7952632494270802,
      "num_tokens": 1185447027.0,
      "step": 1245
    },
    {
      "epoch": 1.7759815242494228,
      "grad_norm": 0.34600724842490793,
      "learning_rate": 2.5959217621687823e-06,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7938597463071346,
      "num_tokens": 1190231791.0,
      "step": 1250
    },
    {
      "epoch": 1.7830875821637946,
      "grad_norm": 0.3678379104234008,
      "learning_rate": 2.5594188686011616e-06,
      "loss": 0.6541,
      "mean_token_accuracy": 0.7947786100208759,
      "num_tokens": 1194998688.0,
      "step": 1255
    },
    {
      "epoch": 1.7901936400781666,
      "grad_norm": 0.36986419243022695,
      "learning_rate": 2.524033718158621e-06,
      "loss": 0.6492,
      "mean_token_accuracy": 0.7966626077890396,
      "num_tokens": 1199764688.0,
      "step": 1260
    },
    {
      "epoch": 1.7972996979925386,
      "grad_norm": 0.36634691155162535,
      "learning_rate": 2.489770996716227e-06,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7945116639137269,
      "num_tokens": 1204526423.0,
      "step": 1265
    },
    {
      "epoch": 1.8044057559069107,
      "grad_norm": 0.3592637763107603,
      "learning_rate": 2.456635241511491e-06,
      "loss": 0.6436,
      "mean_token_accuracy": 0.7984024800360203,
      "num_tokens": 1209280668.0,
      "step": 1270
    },
    {
      "epoch": 1.8115118138212827,
      "grad_norm": 0.3637888435360703,
      "learning_rate": 2.4246308405435314e-06,
      "loss": 0.6503,
      "mean_token_accuracy": 0.7954847238957882,
      "num_tokens": 1214048139.0,
      "step": 1275
    },
    {
      "epoch": 1.8186178717356547,
      "grad_norm": 0.3515690726134511,
      "learning_rate": 2.3937620319919966e-06,
      "loss": 0.6471,
      "mean_token_accuracy": 0.7975172877311707,
      "num_tokens": 1218805359.0,
      "step": 1280
    },
    {
      "epoch": 1.8257239296500267,
      "grad_norm": 0.37901691417441497,
      "learning_rate": 2.3640329036558167e-06,
      "loss": 0.6458,
      "mean_token_accuracy": 0.7973252393305301,
      "num_tokens": 1223580683.0,
      "step": 1285
    },
    {
      "epoch": 1.8328299875643985,
      "grad_norm": 0.48078408366926273,
      "learning_rate": 2.3354473924118843e-06,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7954902827739716,
      "num_tokens": 1228344380.0,
      "step": 1290
    },
    {
      "epoch": 1.8399360454787708,
      "grad_norm": 0.3487084891659478,
      "learning_rate": 2.3080092836937124e-06,
      "loss": 0.649,
      "mean_token_accuracy": 0.7968501009047031,
      "num_tokens": 1233124681.0,
      "step": 1295
    },
    {
      "epoch": 1.8470421033931426,
      "grad_norm": 0.34640131501737065,
      "learning_rate": 2.2817222109901442e-06,
      "loss": 0.6448,
      "mean_token_accuracy": 0.7978550389409065,
      "num_tokens": 1237873166.0,
      "step": 1300
    },
    {
      "epoch": 1.8470421033931426,
      "eval_loss": 0.6490960121154785,
      "eval_mean_token_accuracy": 0.7908800155447241,
      "eval_num_tokens": 1237873166.0,
      "eval_runtime": 149.9569,
      "eval_samples_per_second": 24.267,
      "eval_steps_per_second": 0.76,
      "step": 1300
    },
    {
      "epoch": 1.8541481613075148,
      "grad_norm": 0.3243906306128693,
      "learning_rate": 2.256589655364193e-06,
      "loss": 0.6593,
      "mean_token_accuracy": 0.7929202131927013,
      "num_tokens": 1242627340.0,
      "step": 1305
    },
    {
      "epoch": 1.8612542192218866,
      "grad_norm": 0.37597198041798413,
      "learning_rate": 2.2326149449920653e-06,
      "loss": 0.6446,
      "mean_token_accuracy": 0.797098808735609,
      "num_tokens": 1247387461.0,
      "step": 1310
    },
    {
      "epoch": 1.8683602771362586,
      "grad_norm": 0.35265594906604686,
      "learning_rate": 2.2098012547224197e-06,
      "loss": 0.6513,
      "mean_token_accuracy": 0.7950267992913723,
      "num_tokens": 1252135688.0,
      "step": 1315
    },
    {
      "epoch": 1.8754663350506307,
      "grad_norm": 0.3583812845832173,
      "learning_rate": 2.188151605655942e-06,
      "loss": 0.6521,
      "mean_token_accuracy": 0.7945805780589581,
      "num_tokens": 1256903702.0,
      "step": 1320
    },
    {
      "epoch": 1.8825723929650027,
      "grad_norm": 0.3577801661959976,
      "learning_rate": 2.1676688647452795e-06,
      "loss": 0.6437,
      "mean_token_accuracy": 0.7986263297498226,
      "num_tokens": 1261633144.0,
      "step": 1325
    },
    {
      "epoch": 1.8896784508793747,
      "grad_norm": 0.35744367217582623,
      "learning_rate": 2.1483557444153795e-06,
      "loss": 0.649,
      "mean_token_accuracy": 0.7966003373265267,
      "num_tokens": 1266390903.0,
      "step": 1330
    },
    {
      "epoch": 1.8967845087937467,
      "grad_norm": 0.39747974453689555,
      "learning_rate": 2.1302148022042993e-06,
      "loss": 0.6491,
      "mean_token_accuracy": 0.7970162339508533,
      "num_tokens": 1271162270.0,
      "step": 1335
    },
    {
      "epoch": 1.9038905667081187,
      "grad_norm": 0.3547995480225708,
      "learning_rate": 2.113248440424526e-06,
      "loss": 0.643,
      "mean_token_accuracy": 0.7987522542476654,
      "num_tokens": 1275906083.0,
      "step": 1340
    },
    {
      "epoch": 1.9109966246224905,
      "grad_norm": 0.3924659274346196,
      "learning_rate": 2.0974589058448456e-06,
      "loss": 0.6499,
      "mean_token_accuracy": 0.7970600210130214,
      "num_tokens": 1280649985.0,
      "step": 1345
    },
    {
      "epoch": 1.9181026825368628,
      "grad_norm": 0.3450930632168253,
      "learning_rate": 2.0828482893928208e-06,
      "loss": 0.6525,
      "mean_token_accuracy": 0.795515525341034,
      "num_tokens": 1285434113.0,
      "step": 1350
    },
    {
      "epoch": 1.9252087404512346,
      "grad_norm": 0.33567083461731984,
      "learning_rate": 2.069418525877897e-06,
      "loss": 0.644,
      "mean_token_accuracy": 0.798255106061697,
      "num_tokens": 1290191830.0,
      "step": 1355
    },
    {
      "epoch": 1.9323147983656068,
      "grad_norm": 0.36694627449422723,
      "learning_rate": 2.0571713937351834e-06,
      "loss": 0.6397,
      "mean_token_accuracy": 0.7977312818169594,
      "num_tokens": 1294948980.0,
      "step": 1360
    },
    {
      "epoch": 1.9394208562799786,
      "grad_norm": 0.362232508705221,
      "learning_rate": 2.0461085147899497e-06,
      "loss": 0.6457,
      "mean_token_accuracy": 0.7973731994628906,
      "num_tokens": 1299719386.0,
      "step": 1365
    },
    {
      "epoch": 1.9465269141943506,
      "grad_norm": 0.3669523670641092,
      "learning_rate": 2.0362313540428485e-06,
      "loss": 0.6472,
      "mean_token_accuracy": 0.797086289525032,
      "num_tokens": 1304487261.0,
      "step": 1370
    },
    {
      "epoch": 1.9536329721087227,
      "grad_norm": 0.37200879998568015,
      "learning_rate": 2.027541219475922e-06,
      "loss": 0.6475,
      "mean_token_accuracy": 0.7960396580398083,
      "num_tokens": 1309241194.0,
      "step": 1375
    },
    {
      "epoch": 1.9607390300230947,
      "grad_norm": 0.3688539723341265,
      "learning_rate": 2.020039261879382e-06,
      "loss": 0.6573,
      "mean_token_accuracy": 0.7950874969363213,
      "num_tokens": 1314011836.0,
      "step": 1380
    },
    {
      "epoch": 1.9678450879374667,
      "grad_norm": 0.43966459213132447,
      "learning_rate": 2.013726474699225e-06,
      "loss": 0.6505,
      "mean_token_accuracy": 0.7958736583590508,
      "num_tokens": 1318761485.0,
      "step": 1385
    },
    {
      "epoch": 1.9749511458518387,
      "grad_norm": 0.34824004564185856,
      "learning_rate": 2.008603693905673e-06,
      "loss": 0.6476,
      "mean_token_accuracy": 0.7972124963998795,
      "num_tokens": 1323527340.0,
      "step": 1390
    },
    {
      "epoch": 1.9820572037662108,
      "grad_norm": 0.37468360563296904,
      "learning_rate": 2.0046715978824663e-06,
      "loss": 0.6496,
      "mean_token_accuracy": 0.7958178780972958,
      "num_tokens": 1328302362.0,
      "step": 1395
    },
    {
      "epoch": 1.9891632616805826,
      "grad_norm": 0.3561744251531493,
      "learning_rate": 2.001930707337034e-06,
      "loss": 0.6501,
      "mean_token_accuracy": 0.7963785864412785,
      "num_tokens": 1333062144.0,
      "step": 1400
    },
    {
      "epoch": 1.9891632616805826,
      "eval_loss": 0.6482434868812561,
      "eval_mean_token_accuracy": 0.7910316936802446,
      "eval_num_tokens": 1333062144.0,
      "eval_runtime": 149.6853,
      "eval_samples_per_second": 24.311,
      "eval_steps_per_second": 0.762,
      "step": 1400
    },
    {
      "epoch": 1.9962693195949548,
      "grad_norm": 0.3552540336023921,
      "learning_rate": 2.000381385231536e-06,
      "loss": 0.656,
      "mean_token_accuracy": 0.7951462939381599,
      "num_tokens": 1337810166.0,
      "step": 1405
    },
    {
      "epoch": 2.0,
      "mean_token_accuracy": 0.7962638111341567,
      "num_tokens": 1340314150.0,
      "step": 1408,
      "total_flos": 1.0314062938243072e+16,
      "train_loss": 0.7184352108531378,
      "train_runtime": 49977.4428,
      "train_samples_per_second": 14.417,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 5,
  "max_steps": 1408,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0314062938243072e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}