{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7629947544110635,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0019074868860276585,
      "grad_norm": 23.25,
      "learning_rate": 1.9987282207808727e-05,
      "loss": 1.5688,
      "mean_token_accuracy": 0.5786232218146324,
      "step": 5
    },
    {
      "epoch": 0.003814973772055317,
      "grad_norm": 15.0625,
      "learning_rate": 1.997456441561745e-05,
      "loss": 1.4629,
      "mean_token_accuracy": 0.6007462203502655,
      "step": 10
    },
    {
      "epoch": 0.005722460658082976,
      "grad_norm": 15.0625,
      "learning_rate": 1.9961846623426175e-05,
      "loss": 1.4357,
      "mean_token_accuracy": 0.6233335413038731,
      "step": 15
    },
    {
      "epoch": 0.007629947544110634,
      "grad_norm": 17.75,
      "learning_rate": 1.9949128831234897e-05,
      "loss": 1.4754,
      "mean_token_accuracy": 0.6211968064308167,
      "step": 20
    },
    {
      "epoch": 0.009537434430138292,
      "grad_norm": 14.25,
      "learning_rate": 1.9936411039043622e-05,
      "loss": 1.3715,
      "mean_token_accuracy": 0.6358444899320602,
      "step": 25
    },
    {
      "epoch": 0.011444921316165951,
      "grad_norm": 12.875,
      "learning_rate": 1.9923693246852348e-05,
      "loss": 1.366,
      "mean_token_accuracy": 0.6299699194729328,
      "step": 30
    },
    {
      "epoch": 0.01335240820219361,
      "grad_norm": 14.25,
      "learning_rate": 1.9910975454661073e-05,
      "loss": 1.3211,
      "mean_token_accuracy": 0.6434777334332467,
      "step": 35
    },
    {
      "epoch": 0.015259895088221268,
      "grad_norm": 14.75,
      "learning_rate": 1.98982576624698e-05,
      "loss": 1.3934,
      "mean_token_accuracy": 0.6441360361874103,
      "step": 40
    },
    {
      "epoch": 0.017167381974248927,
      "grad_norm": 14.25,
      "learning_rate": 1.988553987027852e-05,
      "loss": 1.3203,
      "mean_token_accuracy": 0.6629756145179272,
      "step": 45
    },
    {
      "epoch": 0.019074868860276584,
      "grad_norm": 13.0625,
      "learning_rate": 1.9872822078087246e-05,
      "loss": 1.3518,
      "mean_token_accuracy": 0.623377176374197,
      "step": 50
    },
    {
      "epoch": 0.020982355746304245,
      "grad_norm": 12.875,
      "learning_rate": 1.986010428589597e-05,
      "loss": 1.3845,
      "mean_token_accuracy": 0.623815793544054,
      "step": 55
    },
    {
      "epoch": 0.022889842632331903,
      "grad_norm": 15.5625,
      "learning_rate": 1.9847386493704694e-05,
      "loss": 1.264,
      "mean_token_accuracy": 0.6513356983661651,
      "step": 60
    },
    {
      "epoch": 0.02479732951835956,
      "grad_norm": 15.0,
      "learning_rate": 1.983466870151342e-05,
      "loss": 1.2541,
      "mean_token_accuracy": 0.6393217265605926,
      "step": 65
    },
    {
      "epoch": 0.02670481640438722,
      "grad_norm": 14.625,
      "learning_rate": 1.982195090932214e-05,
      "loss": 1.3066,
      "mean_token_accuracy": 0.6521439477801323,
      "step": 70
    },
    {
      "epoch": 0.02861230329041488,
      "grad_norm": 14.125,
      "learning_rate": 1.9809233117130867e-05,
      "loss": 1.3278,
      "mean_token_accuracy": 0.6283318139612675,
      "step": 75
    },
    {
      "epoch": 0.030519790176442536,
      "grad_norm": 13.125,
      "learning_rate": 1.9796515324939593e-05,
      "loss": 1.3905,
      "mean_token_accuracy": 0.6241413362324237,
      "step": 80
    },
    {
      "epoch": 0.03242727706247019,
      "grad_norm": 16.375,
      "learning_rate": 1.9783797532748318e-05,
      "loss": 1.3635,
      "mean_token_accuracy": 0.6261423952877522,
      "step": 85
    },
    {
      "epoch": 0.034334763948497854,
      "grad_norm": 15.125,
      "learning_rate": 1.977107974055704e-05,
      "loss": 1.2991,
      "mean_token_accuracy": 0.6268114626407624,
      "step": 90
    },
    {
      "epoch": 0.036242250834525515,
      "grad_norm": 13.75,
      "learning_rate": 1.9758361948365766e-05,
      "loss": 1.3033,
      "mean_token_accuracy": 0.6382434226572513,
      "step": 95
    },
    {
      "epoch": 0.03814973772055317,
      "grad_norm": 14.4375,
      "learning_rate": 1.974564415617449e-05,
      "loss": 1.3264,
      "mean_token_accuracy": 0.6418077692389488,
      "step": 100
    },
    {
      "epoch": 0.04005722460658083,
      "grad_norm": 13.625,
      "learning_rate": 1.9732926363983213e-05,
      "loss": 1.3281,
      "mean_token_accuracy": 0.6261951096355916,
      "step": 105
    },
    {
      "epoch": 0.04196471149260849,
      "grad_norm": 12.4375,
      "learning_rate": 1.972020857179194e-05,
      "loss": 1.3703,
      "mean_token_accuracy": 0.6287036083638669,
      "step": 110
    },
    {
      "epoch": 0.043872198378636144,
      "grad_norm": 14.3125,
      "learning_rate": 1.970749077960066e-05,
      "loss": 1.258,
      "mean_token_accuracy": 0.6674343384802341,
      "step": 115
    },
    {
      "epoch": 0.045779685264663805,
      "grad_norm": 12.375,
      "learning_rate": 1.9694772987409387e-05,
      "loss": 1.2793,
      "mean_token_accuracy": 0.6343503654003143,
      "step": 120
    },
    {
      "epoch": 0.047687172150691466,
      "grad_norm": 13.75,
      "learning_rate": 1.9682055195218112e-05,
      "loss": 1.2583,
      "mean_token_accuracy": 0.6366156145930291,
      "step": 125
    },
    {
      "epoch": 0.04959465903671912,
      "grad_norm": 11.6875,
      "learning_rate": 1.9669337403026834e-05,
      "loss": 1.2335,
      "mean_token_accuracy": 0.6459783628582955,
      "step": 130
    },
    {
      "epoch": 0.05150214592274678,
      "grad_norm": 14.5,
      "learning_rate": 1.965661961083556e-05,
      "loss": 1.2978,
      "mean_token_accuracy": 0.6397392503917217,
      "step": 135
    },
    {
      "epoch": 0.05340963280877444,
      "grad_norm": 15.0625,
      "learning_rate": 1.9643901818644285e-05,
      "loss": 1.407,
      "mean_token_accuracy": 0.6118167527019978,
      "step": 140
    },
    {
      "epoch": 0.055317119694802096,
      "grad_norm": 12.6875,
      "learning_rate": 1.963118402645301e-05,
      "loss": 1.2377,
      "mean_token_accuracy": 0.6603214554488659,
      "step": 145
    },
    {
      "epoch": 0.05722460658082976,
      "grad_norm": 14.1875,
      "learning_rate": 1.9618466234261733e-05,
      "loss": 1.2917,
      "mean_token_accuracy": 0.6472210057079792,
      "step": 150
    },
    {
      "epoch": 0.05913209346685742,
      "grad_norm": 13.5,
      "learning_rate": 1.9605748442070458e-05,
      "loss": 1.2779,
      "mean_token_accuracy": 0.657531713694334,
      "step": 155
    },
    {
      "epoch": 0.06103958035288507,
      "grad_norm": 14.5,
      "learning_rate": 1.9593030649879184e-05,
      "loss": 1.3979,
      "mean_token_accuracy": 0.6118378482758999,
      "step": 160
    },
    {
      "epoch": 0.06294706723891273,
      "grad_norm": 13.625,
      "learning_rate": 1.9580312857687906e-05,
      "loss": 1.3048,
      "mean_token_accuracy": 0.6219266936182976,
      "step": 165
    },
    {
      "epoch": 0.06485455412494039,
      "grad_norm": 12.125,
      "learning_rate": 1.956759506549663e-05,
      "loss": 1.3086,
      "mean_token_accuracy": 0.6731891065835953,
      "step": 170
    },
    {
      "epoch": 0.06676204101096805,
      "grad_norm": 13.0,
      "learning_rate": 1.9554877273305353e-05,
      "loss": 1.292,
      "mean_token_accuracy": 0.6188046306371688,
      "step": 175
    },
    {
      "epoch": 0.06866952789699571,
      "grad_norm": 13.1875,
      "learning_rate": 1.954215948111408e-05,
      "loss": 1.2729,
      "mean_token_accuracy": 0.6499899953603745,
      "step": 180
    },
    {
      "epoch": 0.07057701478302336,
      "grad_norm": 16.0,
      "learning_rate": 1.9529441688922804e-05,
      "loss": 1.306,
      "mean_token_accuracy": 0.6558002933859826,
      "step": 185
    },
    {
      "epoch": 0.07248450166905103,
      "grad_norm": 12.125,
      "learning_rate": 1.951672389673153e-05,
      "loss": 1.3232,
      "mean_token_accuracy": 0.6388124503195286,
      "step": 190
    },
    {
      "epoch": 0.07439198855507868,
      "grad_norm": 12.4375,
      "learning_rate": 1.9504006104540255e-05,
      "loss": 1.22,
      "mean_token_accuracy": 0.6645313084125519,
      "step": 195
    },
    {
      "epoch": 0.07629947544110634,
      "grad_norm": 12.1875,
      "learning_rate": 1.9491288312348978e-05,
      "loss": 1.3315,
      "mean_token_accuracy": 0.6429102905094624,
      "step": 200
    },
    {
      "epoch": 0.078206962327134,
      "grad_norm": 11.875,
      "learning_rate": 1.9478570520157703e-05,
      "loss": 1.2854,
      "mean_token_accuracy": 0.6392446413636208,
      "step": 205
    },
    {
      "epoch": 0.08011444921316166,
      "grad_norm": 13.625,
      "learning_rate": 1.9465852727966425e-05,
      "loss": 1.3131,
      "mean_token_accuracy": 0.6423079304397106,
      "step": 210
    },
    {
      "epoch": 0.08202193609918931,
      "grad_norm": 13.5,
      "learning_rate": 1.945313493577515e-05,
      "loss": 1.343,
      "mean_token_accuracy": 0.6454595476388931,
      "step": 215
    },
    {
      "epoch": 0.08392942298521698,
      "grad_norm": 16.75,
      "learning_rate": 1.9440417143583876e-05,
      "loss": 1.2251,
      "mean_token_accuracy": 0.6403672114014626,
      "step": 220
    },
    {
      "epoch": 0.08583690987124463,
      "grad_norm": 12.75,
      "learning_rate": 1.9427699351392598e-05,
      "loss": 1.3682,
      "mean_token_accuracy": 0.6467891149222851,
      "step": 225
    },
    {
      "epoch": 0.08774439675727229,
      "grad_norm": 14.625,
      "learning_rate": 1.9414981559201324e-05,
      "loss": 1.3697,
      "mean_token_accuracy": 0.6114679872989655,
      "step": 230
    },
    {
      "epoch": 0.08965188364329996,
      "grad_norm": 14.625,
      "learning_rate": 1.940226376701005e-05,
      "loss": 1.3222,
      "mean_token_accuracy": 0.6396318718791008,
      "step": 235
    },
    {
      "epoch": 0.09155937052932761,
      "grad_norm": 12.1875,
      "learning_rate": 1.938954597481877e-05,
      "loss": 1.2226,
      "mean_token_accuracy": 0.6718489579856396,
      "step": 240
    },
    {
      "epoch": 0.09346685741535526,
      "grad_norm": 12.5,
      "learning_rate": 1.9376828182627497e-05,
      "loss": 1.3327,
      "mean_token_accuracy": 0.6280755452811718,
      "step": 245
    },
    {
      "epoch": 0.09537434430138293,
      "grad_norm": 14.0,
      "learning_rate": 1.9364110390436222e-05,
      "loss": 1.3255,
      "mean_token_accuracy": 0.6481422707438469,
      "step": 250
    },
    {
      "epoch": 0.09728183118741059,
      "grad_norm": 12.5,
      "learning_rate": 1.9351392598244948e-05,
      "loss": 1.2724,
      "mean_token_accuracy": 0.6506570190191269,
      "step": 255
    },
    {
      "epoch": 0.09918931807343824,
      "grad_norm": 12.75,
      "learning_rate": 1.933867480605367e-05,
      "loss": 1.199,
      "mean_token_accuracy": 0.6626401949673891,
      "step": 260
    },
    {
      "epoch": 0.10109680495946591,
      "grad_norm": 14.25,
      "learning_rate": 1.9325957013862396e-05,
      "loss": 1.3612,
      "mean_token_accuracy": 0.6227854460477829,
      "step": 265
    },
    {
      "epoch": 0.10300429184549356,
      "grad_norm": 13.3125,
      "learning_rate": 1.9313239221671118e-05,
      "loss": 1.3032,
      "mean_token_accuracy": 0.6405224844813346,
      "step": 270
    },
    {
      "epoch": 0.10491177873152122,
      "grad_norm": 13.5,
      "learning_rate": 1.9300521429479843e-05,
      "loss": 1.2347,
      "mean_token_accuracy": 0.6489498361945152,
      "step": 275
    },
    {
      "epoch": 0.10681926561754888,
      "grad_norm": 14.375,
      "learning_rate": 1.928780363728857e-05,
      "loss": 1.2301,
      "mean_token_accuracy": 0.6662219226360321,
      "step": 280
    },
    {
      "epoch": 0.10872675250357654,
      "grad_norm": 16.375,
      "learning_rate": 1.927508584509729e-05,
      "loss": 1.3138,
      "mean_token_accuracy": 0.6375241696834564,
      "step": 285
    },
    {
      "epoch": 0.11063423938960419,
      "grad_norm": 11.5625,
      "learning_rate": 1.9262368052906016e-05,
      "loss": 1.3117,
      "mean_token_accuracy": 0.6601852528750897,
      "step": 290
    },
    {
      "epoch": 0.11254172627563186,
      "grad_norm": 11.0,
      "learning_rate": 1.9249650260714742e-05,
      "loss": 1.2498,
      "mean_token_accuracy": 0.6576425515115261,
      "step": 295
    },
    {
      "epoch": 0.11444921316165951,
      "grad_norm": 14.625,
      "learning_rate": 1.9236932468523467e-05,
      "loss": 1.2453,
      "mean_token_accuracy": 0.6428922191262245,
      "step": 300
    },
    {
      "epoch": 0.11635670004768717,
      "grad_norm": 11.1875,
      "learning_rate": 1.9224214676332193e-05,
      "loss": 1.1744,
      "mean_token_accuracy": 0.673005498945713,
      "step": 305
    },
    {
      "epoch": 0.11826418693371483,
      "grad_norm": 13.4375,
      "learning_rate": 1.9211496884140915e-05,
      "loss": 1.2113,
      "mean_token_accuracy": 0.6825620189309121,
      "step": 310
    },
    {
      "epoch": 0.12017167381974249,
      "grad_norm": 15.0625,
      "learning_rate": 1.919877909194964e-05,
      "loss": 1.1482,
      "mean_token_accuracy": 0.6523631751537323,
      "step": 315
    },
    {
      "epoch": 0.12207916070577014,
      "grad_norm": 12.5,
      "learning_rate": 1.9186061299758362e-05,
      "loss": 1.2886,
      "mean_token_accuracy": 0.6448531322181225,
      "step": 320
    },
    {
      "epoch": 0.12398664759179781,
      "grad_norm": 10.9375,
      "learning_rate": 1.9173343507567088e-05,
      "loss": 1.273,
      "mean_token_accuracy": 0.6480912029743194,
      "step": 325
    },
    {
      "epoch": 0.12589413447782546,
      "grad_norm": 13.25,
      "learning_rate": 1.916062571537581e-05,
      "loss": 1.363,
      "mean_token_accuracy": 0.6523602977395058,
      "step": 330
    },
    {
      "epoch": 0.12780162136385312,
      "grad_norm": 16.0,
      "learning_rate": 1.9147907923184536e-05,
      "loss": 1.2477,
      "mean_token_accuracy": 0.6430979609489441,
      "step": 335
    },
    {
      "epoch": 0.12970910824988077,
      "grad_norm": 12.8125,
      "learning_rate": 1.913519013099326e-05,
      "loss": 1.1793,
      "mean_token_accuracy": 0.6552196487784385,
      "step": 340
    },
    {
      "epoch": 0.13161659513590845,
      "grad_norm": 15.25,
      "learning_rate": 1.9122472338801987e-05,
      "loss": 1.2032,
      "mean_token_accuracy": 0.6600343361496925,
      "step": 345
    },
    {
      "epoch": 0.1335240820219361,
      "grad_norm": 14.125,
      "learning_rate": 1.9109754546610712e-05,
      "loss": 1.3241,
      "mean_token_accuracy": 0.6307924754917622,
      "step": 350
    },
    {
      "epoch": 0.13543156890796376,
      "grad_norm": 18.875,
      "learning_rate": 1.9097036754419434e-05,
      "loss": 1.2946,
      "mean_token_accuracy": 0.6583775602281093,
      "step": 355
    },
    {
      "epoch": 0.13733905579399142,
      "grad_norm": 14.6875,
      "learning_rate": 1.908431896222816e-05,
      "loss": 1.3141,
      "mean_token_accuracy": 0.6407140150666237,
      "step": 360
    },
    {
      "epoch": 0.13924654268001907,
      "grad_norm": 12.75,
      "learning_rate": 1.9071601170036885e-05,
      "loss": 1.3233,
      "mean_token_accuracy": 0.6416009657084942,
      "step": 365
    },
    {
      "epoch": 0.14115402956604672,
      "grad_norm": 13.3125,
      "learning_rate": 1.9058883377845607e-05,
      "loss": 1.1646,
      "mean_token_accuracy": 0.6636185184121132,
      "step": 370
    },
    {
      "epoch": 0.1430615164520744,
      "grad_norm": 11.875,
      "learning_rate": 1.9046165585654333e-05,
      "loss": 1.3162,
      "mean_token_accuracy": 0.6456755690276623,
      "step": 375
    },
    {
      "epoch": 0.14496900333810206,
      "grad_norm": 11.5625,
      "learning_rate": 1.9033447793463055e-05,
      "loss": 1.2697,
      "mean_token_accuracy": 0.6551977261900902,
      "step": 380
    },
    {
      "epoch": 0.1468764902241297,
      "grad_norm": 14.6875,
      "learning_rate": 1.902073000127178e-05,
      "loss": 1.2712,
      "mean_token_accuracy": 0.6613594964146614,
      "step": 385
    },
    {
      "epoch": 0.14878397711015737,
      "grad_norm": 15.375,
      "learning_rate": 1.9008012209080503e-05,
      "loss": 1.1864,
      "mean_token_accuracy": 0.6685264468193054,
      "step": 390
    },
    {
      "epoch": 0.15069146399618502,
      "grad_norm": 18.375,
      "learning_rate": 1.8995294416889228e-05,
      "loss": 1.3601,
      "mean_token_accuracy": 0.6272764652967453,
      "step": 395
    },
    {
      "epoch": 0.15259895088221268,
      "grad_norm": 11.9375,
      "learning_rate": 1.8982576624697954e-05,
      "loss": 1.3435,
      "mean_token_accuracy": 0.6371615134179592,
      "step": 400
    },
    {
      "epoch": 0.15450643776824036,
      "grad_norm": 13.25,
      "learning_rate": 1.896985883250668e-05,
      "loss": 1.3172,
      "mean_token_accuracy": 0.6476671509444714,
      "step": 405
    },
    {
      "epoch": 0.156413924654268,
      "grad_norm": 11.75,
      "learning_rate": 1.8957141040315405e-05,
      "loss": 1.204,
      "mean_token_accuracy": 0.658133564144373,
      "step": 410
    },
    {
      "epoch": 0.15832141154029566,
      "grad_norm": 13.6875,
      "learning_rate": 1.8944423248124127e-05,
      "loss": 1.3395,
      "mean_token_accuracy": 0.6478057160973549,
      "step": 415
    },
    {
      "epoch": 0.16022889842632332,
      "grad_norm": 12.9375,
      "learning_rate": 1.8931705455932852e-05,
      "loss": 1.2254,
      "mean_token_accuracy": 0.655465979874134,
      "step": 420
    },
    {
      "epoch": 0.16213638531235097,
      "grad_norm": 14.25,
      "learning_rate": 1.8918987663741578e-05,
      "loss": 1.3114,
      "mean_token_accuracy": 0.6299714539200068,
      "step": 425
    },
    {
      "epoch": 0.16404387219837863,
      "grad_norm": 12.625,
      "learning_rate": 1.89062698715503e-05,
      "loss": 1.2496,
      "mean_token_accuracy": 0.6370454408228398,
      "step": 430
    },
    {
      "epoch": 0.1659513590844063,
      "grad_norm": 14.25,
      "learning_rate": 1.8893552079359025e-05,
      "loss": 1.2617,
      "mean_token_accuracy": 0.6422188803553581,
      "step": 435
    },
    {
      "epoch": 0.16785884597043396,
      "grad_norm": 13.25,
      "learning_rate": 1.8880834287167747e-05,
      "loss": 1.3337,
      "mean_token_accuracy": 0.6599226906895638,
      "step": 440
    },
    {
      "epoch": 0.16976633285646162,
      "grad_norm": 11.8125,
      "learning_rate": 1.8868116494976473e-05,
      "loss": 1.2401,
      "mean_token_accuracy": 0.6384981378912926,
      "step": 445
    },
    {
      "epoch": 0.17167381974248927,
      "grad_norm": 15.1875,
      "learning_rate": 1.88553987027852e-05,
      "loss": 1.3429,
      "mean_token_accuracy": 0.6290579028427601,
      "step": 450
    },
    {
      "epoch": 0.17358130662851692,
      "grad_norm": 14.625,
      "learning_rate": 1.8842680910593924e-05,
      "loss": 1.26,
      "mean_token_accuracy": 0.6353619039058686,
      "step": 455
    },
    {
      "epoch": 0.17548879351454458,
      "grad_norm": 11.9375,
      "learning_rate": 1.882996311840265e-05,
      "loss": 1.2258,
      "mean_token_accuracy": 0.6596131481230258,
      "step": 460
    },
    {
      "epoch": 0.17739628040057226,
      "grad_norm": 12.0,
      "learning_rate": 1.881724532621137e-05,
      "loss": 1.3572,
      "mean_token_accuracy": 0.6373824059963227,
      "step": 465
    },
    {
      "epoch": 0.1793037672865999,
      "grad_norm": 13.0625,
      "learning_rate": 1.8804527534020097e-05,
      "loss": 1.2938,
      "mean_token_accuracy": 0.6647212252020835,
      "step": 470
    },
    {
      "epoch": 0.18121125417262757,
      "grad_norm": 14.0625,
      "learning_rate": 1.879180974182882e-05,
      "loss": 1.2844,
      "mean_token_accuracy": 0.6200387306511402,
      "step": 475
    },
    {
      "epoch": 0.18311874105865522,
      "grad_norm": 12.5625,
      "learning_rate": 1.8779091949637545e-05,
      "loss": 1.2827,
      "mean_token_accuracy": 0.6431592881679535,
      "step": 480
    },
    {
      "epoch": 0.18502622794468288,
      "grad_norm": 13.1875,
      "learning_rate": 1.876637415744627e-05,
      "loss": 1.2237,
      "mean_token_accuracy": 0.667415677011013,
      "step": 485
    },
    {
      "epoch": 0.18693371483071053,
      "grad_norm": 13.5,
      "learning_rate": 1.8753656365254992e-05,
      "loss": 1.376,
      "mean_token_accuracy": 0.6330628030002117,
      "step": 490
    },
    {
      "epoch": 0.1888412017167382,
      "grad_norm": 13.5,
      "learning_rate": 1.8740938573063718e-05,
      "loss": 1.3092,
      "mean_token_accuracy": 0.6471766166388988,
      "step": 495
    },
    {
      "epoch": 0.19074868860276586,
      "grad_norm": 11.125,
      "learning_rate": 1.872822078087244e-05,
      "loss": 1.1906,
      "mean_token_accuracy": 0.6722134962677956,
      "step": 500
    },
    {
      "epoch": 0.19265617548879352,
      "grad_norm": 14.875,
      "learning_rate": 1.8715502988681165e-05,
      "loss": 1.3313,
      "mean_token_accuracy": 0.6335577562451362,
      "step": 505
    },
    {
      "epoch": 0.19456366237482117,
      "grad_norm": 13.9375,
      "learning_rate": 1.870278519648989e-05,
      "loss": 1.2115,
      "mean_token_accuracy": 0.6548780754208565,
      "step": 510
    },
    {
      "epoch": 0.19647114926084883,
      "grad_norm": 11.5,
      "learning_rate": 1.8690067404298616e-05,
      "loss": 1.3229,
      "mean_token_accuracy": 0.6398176200687885,
      "step": 515
    },
    {
      "epoch": 0.19837863614687648,
      "grad_norm": 12.875,
      "learning_rate": 1.8677349612107342e-05,
      "loss": 1.2795,
      "mean_token_accuracy": 0.6450474753975868,
      "step": 520
    },
    {
      "epoch": 0.20028612303290416,
      "grad_norm": 11.9375,
      "learning_rate": 1.8664631819916064e-05,
      "loss": 1.3635,
      "mean_token_accuracy": 0.6569343976676464,
      "step": 525
    },
    {
      "epoch": 0.20219360991893182,
      "grad_norm": 13.5,
      "learning_rate": 1.865191402772479e-05,
      "loss": 1.2394,
      "mean_token_accuracy": 0.6203682988882064,
      "step": 530
    },
    {
      "epoch": 0.20410109680495947,
      "grad_norm": 12.3125,
      "learning_rate": 1.863919623553351e-05,
      "loss": 1.2598,
      "mean_token_accuracy": 0.6244672931730747,
      "step": 535
    },
    {
      "epoch": 0.20600858369098712,
      "grad_norm": 11.75,
      "learning_rate": 1.8626478443342237e-05,
      "loss": 1.2364,
      "mean_token_accuracy": 0.6448377221822739,
      "step": 540
    },
    {
      "epoch": 0.20791607057701478,
      "grad_norm": 13.5625,
      "learning_rate": 1.8613760651150963e-05,
      "loss": 1.3184,
      "mean_token_accuracy": 0.6233433380722999,
      "step": 545
    },
    {
      "epoch": 0.20982355746304243,
      "grad_norm": 11.125,
      "learning_rate": 1.8601042858959685e-05,
      "loss": 1.1649,
      "mean_token_accuracy": 0.6621956214308738,
      "step": 550
    },
    {
      "epoch": 0.2117310443490701,
      "grad_norm": 10.8125,
      "learning_rate": 1.858832506676841e-05,
      "loss": 1.2191,
      "mean_token_accuracy": 0.6589386224746704,
      "step": 555
    },
    {
      "epoch": 0.21363853123509777,
      "grad_norm": 11.6875,
      "learning_rate": 1.8575607274577136e-05,
      "loss": 1.1812,
      "mean_token_accuracy": 0.6640269085764885,
      "step": 560
    },
    {
      "epoch": 0.21554601812112542,
      "grad_norm": 13.9375,
      "learning_rate": 1.856288948238586e-05,
      "loss": 1.3898,
      "mean_token_accuracy": 0.6331407696008682,
      "step": 565
    },
    {
      "epoch": 0.21745350500715308,
      "grad_norm": 10.875,
      "learning_rate": 1.8550171690194583e-05,
      "loss": 1.1636,
      "mean_token_accuracy": 0.6876399561762809,
      "step": 570
    },
    {
      "epoch": 0.21936099189318073,
      "grad_norm": 13.5,
      "learning_rate": 1.853745389800331e-05,
      "loss": 1.1624,
      "mean_token_accuracy": 0.683189244568348,
      "step": 575
    },
    {
      "epoch": 0.22126847877920838,
      "grad_norm": 12.625,
      "learning_rate": 1.8524736105812034e-05,
      "loss": 1.1666,
      "mean_token_accuracy": 0.6829775631427765,
      "step": 580
    },
    {
      "epoch": 0.22317596566523606,
      "grad_norm": 17.625,
      "learning_rate": 1.8512018313620756e-05,
      "loss": 1.3125,
      "mean_token_accuracy": 0.6385770931839942,
      "step": 585
    },
    {
      "epoch": 0.22508345255126372,
      "grad_norm": 12.5625,
      "learning_rate": 1.8499300521429482e-05,
      "loss": 1.2159,
      "mean_token_accuracy": 0.6778845816850663,
      "step": 590
    },
    {
      "epoch": 0.22699093943729137,
      "grad_norm": 11.3125,
      "learning_rate": 1.8486582729238204e-05,
      "loss": 1.1959,
      "mean_token_accuracy": 0.6777172073721885,
      "step": 595
    },
    {
      "epoch": 0.22889842632331903,
      "grad_norm": 13.0625,
      "learning_rate": 1.847386493704693e-05,
      "loss": 1.2577,
      "mean_token_accuracy": 0.6450985379517078,
      "step": 600
    },
    {
      "epoch": 0.23080591320934668,
      "grad_norm": 12.5,
      "learning_rate": 1.8461147144855655e-05,
      "loss": 1.1582,
      "mean_token_accuracy": 0.6681318923830986,
      "step": 605
    },
    {
      "epoch": 0.23271340009537433,
      "grad_norm": 19.75,
      "learning_rate": 1.8448429352664377e-05,
      "loss": 1.2834,
      "mean_token_accuracy": 0.6435618877410889,
      "step": 610
    },
    {
      "epoch": 0.23462088698140202,
      "grad_norm": 11.8125,
      "learning_rate": 1.8435711560473103e-05,
      "loss": 1.2699,
      "mean_token_accuracy": 0.6606324508786201,
      "step": 615
    },
    {
      "epoch": 0.23652837386742967,
      "grad_norm": 13.625,
      "learning_rate": 1.8422993768281828e-05,
      "loss": 1.2522,
      "mean_token_accuracy": 0.6440395541489125,
      "step": 620
    },
    {
      "epoch": 0.23843586075345732,
      "grad_norm": 15.4375,
      "learning_rate": 1.8410275976090554e-05,
      "loss": 1.2139,
      "mean_token_accuracy": 0.6698227688670159,
      "step": 625
    },
    {
      "epoch": 0.24034334763948498,
      "grad_norm": 13.3125,
      "learning_rate": 1.8397558183899276e-05,
      "loss": 1.2841,
      "mean_token_accuracy": 0.6581710763275623,
      "step": 630
    },
    {
      "epoch": 0.24225083452551263,
      "grad_norm": 12.0625,
      "learning_rate": 1.8384840391708e-05,
      "loss": 1.2246,
      "mean_token_accuracy": 0.6571968667209148,
      "step": 635
    },
    {
      "epoch": 0.24415832141154029,
      "grad_norm": 11.625,
      "learning_rate": 1.8372122599516727e-05,
      "loss": 1.2693,
      "mean_token_accuracy": 0.6628240890800953,
      "step": 640
    },
    {
      "epoch": 0.24606580829756797,
      "grad_norm": 12.75,
      "learning_rate": 1.835940480732545e-05,
      "loss": 1.2361,
      "mean_token_accuracy": 0.6445854142308235,
      "step": 645
    },
    {
      "epoch": 0.24797329518359562,
      "grad_norm": 12.5625,
      "learning_rate": 1.8346687015134174e-05,
      "loss": 1.2374,
      "mean_token_accuracy": 0.6404064007103443,
      "step": 650
    },
    {
      "epoch": 0.24988078206962328,
      "grad_norm": 13.1875,
      "learning_rate": 1.8333969222942896e-05,
      "loss": 1.1423,
      "mean_token_accuracy": 0.6834298320114612,
      "step": 655
    },
    {
      "epoch": 0.25178826895565093,
      "grad_norm": 12.625,
      "learning_rate": 1.8321251430751622e-05,
      "loss": 1.1964,
      "mean_token_accuracy": 0.6344361655414105,
      "step": 660
    },
    {
      "epoch": 0.2536957558416786,
      "grad_norm": 12.375,
      "learning_rate": 1.8308533638560347e-05,
      "loss": 1.183,
      "mean_token_accuracy": 0.6825651362538337,
      "step": 665
    },
    {
      "epoch": 0.25560324272770624,
      "grad_norm": 11.625,
      "learning_rate": 1.8295815846369073e-05,
      "loss": 1.1489,
      "mean_token_accuracy": 0.6593642435967922,
      "step": 670
    },
    {
      "epoch": 0.2575107296137339,
      "grad_norm": 17.0,
      "learning_rate": 1.82830980541778e-05,
      "loss": 1.3731,
      "mean_token_accuracy": 0.6474427200853825,
      "step": 675
    },
    {
      "epoch": 0.25941821649976154,
      "grad_norm": 11.0625,
      "learning_rate": 1.827038026198652e-05,
      "loss": 1.2505,
      "mean_token_accuracy": 0.6625294893980026,
      "step": 680
    },
    {
      "epoch": 0.2613257033857892,
      "grad_norm": 15.0625,
      "learning_rate": 1.8257662469795246e-05,
      "loss": 1.3258,
      "mean_token_accuracy": 0.6423907749354839,
      "step": 685
    },
    {
      "epoch": 0.2632331902718169,
      "grad_norm": 11.125,
      "learning_rate": 1.8244944677603968e-05,
      "loss": 1.1728,
      "mean_token_accuracy": 0.6944520533084869,
      "step": 690
    },
    {
      "epoch": 0.26514067715784456,
      "grad_norm": 15.6875,
      "learning_rate": 1.8232226885412694e-05,
      "loss": 1.3058,
      "mean_token_accuracy": 0.6184810683131218,
      "step": 695
    },
    {
      "epoch": 0.2670481640438722,
      "grad_norm": 13.5,
      "learning_rate": 1.821950909322142e-05,
      "loss": 1.2324,
      "mean_token_accuracy": 0.6492418631911278,
      "step": 700
    },
    {
      "epoch": 0.26895565092989987,
      "grad_norm": 13.9375,
      "learning_rate": 1.820679130103014e-05,
      "loss": 1.249,
      "mean_token_accuracy": 0.6514696642756462,
      "step": 705
    },
    {
      "epoch": 0.2708631378159275,
      "grad_norm": 10.9375,
      "learning_rate": 1.8194073508838867e-05,
      "loss": 1.1874,
      "mean_token_accuracy": 0.6706778854131699,
      "step": 710
    },
    {
      "epoch": 0.2727706247019552,
      "grad_norm": 11.875,
      "learning_rate": 1.8181355716647592e-05,
      "loss": 1.1423,
      "mean_token_accuracy": 0.6606415674090386,
      "step": 715
    },
    {
      "epoch": 0.27467811158798283,
      "grad_norm": 11.0,
      "learning_rate": 1.8168637924456314e-05,
      "loss": 1.2079,
      "mean_token_accuracy": 0.6652269288897514,
      "step": 720
    },
    {
      "epoch": 0.2765855984740105,
      "grad_norm": 12.5,
      "learning_rate": 1.815592013226504e-05,
      "loss": 1.1645,
      "mean_token_accuracy": 0.6542063765227795,
      "step": 725
    },
    {
      "epoch": 0.27849308536003814,
      "grad_norm": 11.5625,
      "learning_rate": 1.8143202340073765e-05,
      "loss": 1.1272,
      "mean_token_accuracy": 0.6796695061028004,
      "step": 730
    },
    {
      "epoch": 0.2804005722460658,
      "grad_norm": 13.625,
      "learning_rate": 1.813048454788249e-05,
      "loss": 1.2422,
      "mean_token_accuracy": 0.6447897091507911,
      "step": 735
    },
    {
      "epoch": 0.28230805913209345,
      "grad_norm": 16.375,
      "learning_rate": 1.8117766755691213e-05,
      "loss": 1.2117,
      "mean_token_accuracy": 0.6523264884948731,
      "step": 740
    },
    {
      "epoch": 0.2842155460181211,
      "grad_norm": 14.1875,
      "learning_rate": 1.810504896349994e-05,
      "loss": 1.2406,
      "mean_token_accuracy": 0.6362795114517212,
      "step": 745
    },
    {
      "epoch": 0.2861230329041488,
      "grad_norm": 11.125,
      "learning_rate": 1.809233117130866e-05,
      "loss": 1.1137,
      "mean_token_accuracy": 0.6698605574667453,
      "step": 750
    },
    {
      "epoch": 0.28803051979017646,
      "grad_norm": 12.125,
      "learning_rate": 1.8079613379117386e-05,
      "loss": 1.1863,
      "mean_token_accuracy": 0.6513051472604274,
      "step": 755
    },
    {
      "epoch": 0.2899380066762041,
      "grad_norm": 14.3125,
      "learning_rate": 1.806689558692611e-05,
      "loss": 1.2335,
      "mean_token_accuracy": 0.689585417509079,
      "step": 760
    },
    {
      "epoch": 0.2918454935622318,
      "grad_norm": 12.4375,
      "learning_rate": 1.8054177794734834e-05,
      "loss": 1.1806,
      "mean_token_accuracy": 0.6750295326113701,
      "step": 765
    },
    {
      "epoch": 0.2937529804482594,
      "grad_norm": 12.375,
      "learning_rate": 1.804146000254356e-05,
      "loss": 1.2288,
      "mean_token_accuracy": 0.6504852309823036,
      "step": 770
    },
    {
      "epoch": 0.2956604673342871,
      "grad_norm": 13.8125,
      "learning_rate": 1.8028742210352285e-05,
      "loss": 1.2207,
      "mean_token_accuracy": 0.6786825984716416,
      "step": 775
    },
    {
      "epoch": 0.29756795422031473,
      "grad_norm": 11.5625,
      "learning_rate": 1.801602441816101e-05,
      "loss": 1.2771,
      "mean_token_accuracy": 0.6656466156244278,
      "step": 780
    },
    {
      "epoch": 0.2994754411063424,
      "grad_norm": 12.5625,
      "learning_rate": 1.8003306625969736e-05,
      "loss": 1.2357,
      "mean_token_accuracy": 0.6720115698873996,
      "step": 785
    },
    {
      "epoch": 0.30138292799237004,
      "grad_norm": 11.1875,
      "learning_rate": 1.7990588833778458e-05,
      "loss": 1.2388,
      "mean_token_accuracy": 0.6488339100033045,
      "step": 790
    },
    {
      "epoch": 0.3032904148783977,
      "grad_norm": 12.6875,
      "learning_rate": 1.7977871041587183e-05,
      "loss": 1.1258,
      "mean_token_accuracy": 0.6850755415856838,
      "step": 795
    },
    {
      "epoch": 0.30519790176442535,
      "grad_norm": 12.0625,
      "learning_rate": 1.7965153249395905e-05,
      "loss": 1.2621,
      "mean_token_accuracy": 0.6553989514708519,
      "step": 800
    },
    {
      "epoch": 0.307105388650453,
      "grad_norm": 11.6875,
      "learning_rate": 1.795243545720463e-05,
      "loss": 1.084,
      "mean_token_accuracy": 0.6827043130993843,
      "step": 805
    },
    {
      "epoch": 0.3090128755364807,
      "grad_norm": 13.5625,
      "learning_rate": 1.7939717665013353e-05,
      "loss": 1.1707,
      "mean_token_accuracy": 0.6674947261810302,
      "step": 810
    },
    {
      "epoch": 0.31092036242250837,
      "grad_norm": 14.125,
      "learning_rate": 1.792699987282208e-05,
      "loss": 1.2618,
      "mean_token_accuracy": 0.6420316450297833,
      "step": 815
    },
    {
      "epoch": 0.312827849308536,
      "grad_norm": 14.375,
      "learning_rate": 1.7914282080630804e-05,
      "loss": 1.2122,
      "mean_token_accuracy": 0.6519798688590527,
      "step": 820
    },
    {
      "epoch": 0.3147353361945637,
      "grad_norm": 13.8125,
      "learning_rate": 1.790156428843953e-05,
      "loss": 1.3042,
      "mean_token_accuracy": 0.6403947852551937,
      "step": 825
    },
    {
      "epoch": 0.31664282308059133,
      "grad_norm": 11.3125,
      "learning_rate": 1.788884649624825e-05,
      "loss": 1.2524,
      "mean_token_accuracy": 0.6245795480906964,
      "step": 830
    },
    {
      "epoch": 0.318550309966619,
      "grad_norm": 14.4375,
      "learning_rate": 1.7876128704056977e-05,
      "loss": 1.2826,
      "mean_token_accuracy": 0.6210857503116131,
      "step": 835
    },
    {
      "epoch": 0.32045779685264664,
      "grad_norm": 12.0,
      "learning_rate": 1.7863410911865703e-05,
      "loss": 1.2216,
      "mean_token_accuracy": 0.664474394917488,
      "step": 840
    },
    {
      "epoch": 0.3223652837386743,
      "grad_norm": 12.5625,
      "learning_rate": 1.7850693119674428e-05,
      "loss": 1.1351,
      "mean_token_accuracy": 0.647052352130413,
      "step": 845
    },
    {
      "epoch": 0.32427277062470194,
      "grad_norm": 10.8125,
      "learning_rate": 1.783797532748315e-05,
      "loss": 1.1048,
      "mean_token_accuracy": 0.7011352255940437,
      "step": 850
    },
    {
      "epoch": 0.3261802575107296,
      "grad_norm": 15.75,
      "learning_rate": 1.7825257535291876e-05,
      "loss": 1.2485,
      "mean_token_accuracy": 0.6584809772670269,
      "step": 855
    },
    {
      "epoch": 0.32808774439675725,
      "grad_norm": 12.6875,
      "learning_rate": 1.7812539743100598e-05,
      "loss": 1.2427,
      "mean_token_accuracy": 0.6549804627895355,
      "step": 860
    },
    {
      "epoch": 0.3299952312827849,
      "grad_norm": 12.625,
      "learning_rate": 1.7799821950909323e-05,
      "loss": 1.0888,
      "mean_token_accuracy": 0.690093956142664,
      "step": 865
    },
    {
      "epoch": 0.3319027181688126,
      "grad_norm": 15.6875,
      "learning_rate": 1.7787104158718045e-05,
      "loss": 1.1655,
      "mean_token_accuracy": 0.6605223380029202,
      "step": 870
    },
    {
      "epoch": 0.33381020505484027,
      "grad_norm": 13.0625,
      "learning_rate": 1.777438636652677e-05,
      "loss": 1.1255,
      "mean_token_accuracy": 0.6727245457470417,
      "step": 875
    },
    {
      "epoch": 0.3357176919408679,
      "grad_norm": 12.4375,
      "learning_rate": 1.7761668574335496e-05,
      "loss": 1.1536,
      "mean_token_accuracy": 0.6633619382977486,
      "step": 880
    },
    {
      "epoch": 0.3376251788268956,
      "grad_norm": 11.6875,
      "learning_rate": 1.7748950782144222e-05,
      "loss": 1.1295,
      "mean_token_accuracy": 0.6744188234210015,
      "step": 885
    },
    {
      "epoch": 0.33953266571292323,
      "grad_norm": 11.875,
      "learning_rate": 1.7736232989952947e-05,
      "loss": 1.1849,
      "mean_token_accuracy": 0.6710595056414604,
      "step": 890
    },
    {
      "epoch": 0.3414401525989509,
      "grad_norm": 13.9375,
      "learning_rate": 1.772351519776167e-05,
      "loss": 1.1048,
      "mean_token_accuracy": 0.686171543598175,
      "step": 895
    },
    {
      "epoch": 0.34334763948497854,
      "grad_norm": 11.5,
      "learning_rate": 1.7710797405570395e-05,
      "loss": 1.1954,
      "mean_token_accuracy": 0.6756417036056519,
      "step": 900
    },
    {
      "epoch": 0.3452551263710062,
      "grad_norm": 13.375,
      "learning_rate": 1.7698079613379117e-05,
      "loss": 1.1613,
      "mean_token_accuracy": 0.6741260103881359,
      "step": 905
    },
    {
      "epoch": 0.34716261325703385,
      "grad_norm": 14.5625,
      "learning_rate": 1.7685361821187843e-05,
      "loss": 1.2515,
      "mean_token_accuracy": 0.6590819746255875,
      "step": 910
    },
    {
      "epoch": 0.3490701001430615,
      "grad_norm": 11.0625,
      "learning_rate": 1.7672644028996568e-05,
      "loss": 1.2335,
      "mean_token_accuracy": 0.6556867949664593,
      "step": 915
    },
    {
      "epoch": 0.35097758702908916,
      "grad_norm": 13.125,
      "learning_rate": 1.765992623680529e-05,
      "loss": 1.3305,
      "mean_token_accuracy": 0.6548508778214455,
      "step": 920
    },
    {
      "epoch": 0.3528850739151168,
      "grad_norm": 11.375,
      "learning_rate": 1.7647208444614016e-05,
      "loss": 1.1825,
      "mean_token_accuracy": 0.6824756883084774,
      "step": 925
    },
    {
      "epoch": 0.3547925608011445,
      "grad_norm": 12.6875,
      "learning_rate": 1.763449065242274e-05,
      "loss": 1.2301,
      "mean_token_accuracy": 0.6782255977392196,
      "step": 930
    },
    {
      "epoch": 0.3567000476871722,
      "grad_norm": 11.125,
      "learning_rate": 1.7621772860231467e-05,
      "loss": 1.175,
      "mean_token_accuracy": 0.6592901304364205,
      "step": 935
    },
    {
      "epoch": 0.3586075345731998,
      "grad_norm": 12.625,
      "learning_rate": 1.760905506804019e-05,
      "loss": 1.1576,
      "mean_token_accuracy": 0.6741487175226212,
      "step": 940
    },
    {
      "epoch": 0.3605150214592275,
      "grad_norm": 12.25,
      "learning_rate": 1.7596337275848914e-05,
      "loss": 1.2168,
      "mean_token_accuracy": 0.6609480068087578,
      "step": 945
    },
    {
      "epoch": 0.36242250834525513,
      "grad_norm": 11.375,
      "learning_rate": 1.758361948365764e-05,
      "loss": 1.2009,
      "mean_token_accuracy": 0.6507039844989777,
      "step": 950
    },
    {
      "epoch": 0.3643299952312828,
      "grad_norm": 11.5,
      "learning_rate": 1.7570901691466362e-05,
      "loss": 1.1247,
      "mean_token_accuracy": 0.6828011214733124,
      "step": 955
    },
    {
      "epoch": 0.36623748211731044,
      "grad_norm": 12.0625,
      "learning_rate": 1.7558183899275088e-05,
      "loss": 1.2232,
      "mean_token_accuracy": 0.6623214483261108,
      "step": 960
    },
    {
      "epoch": 0.3681449690033381,
      "grad_norm": 16.5,
      "learning_rate": 1.754546610708381e-05,
      "loss": 1.3721,
      "mean_token_accuracy": 0.6213653467595577,
      "step": 965
    },
    {
      "epoch": 0.37005245588936575,
      "grad_norm": 10.1875,
      "learning_rate": 1.7532748314892535e-05,
      "loss": 1.1464,
      "mean_token_accuracy": 0.6750129237771034,
      "step": 970
    },
    {
      "epoch": 0.3719599427753934,
      "grad_norm": 12.4375,
      "learning_rate": 1.752003052270126e-05,
      "loss": 1.1425,
      "mean_token_accuracy": 0.6775874204933643,
      "step": 975
    },
    {
      "epoch": 0.37386742966142106,
      "grad_norm": 12.375,
      "learning_rate": 1.7507312730509983e-05,
      "loss": 1.2007,
      "mean_token_accuracy": 0.6537452466785908,
      "step": 980
    },
    {
      "epoch": 0.3757749165474487,
      "grad_norm": 14.875,
      "learning_rate": 1.7494594938318708e-05,
      "loss": 1.2564,
      "mean_token_accuracy": 0.6193661518394947,
      "step": 985
    },
    {
      "epoch": 0.3776824034334764,
      "grad_norm": 10.8125,
      "learning_rate": 1.7481877146127434e-05,
      "loss": 1.1159,
      "mean_token_accuracy": 0.6820607483386993,
      "step": 990
    },
    {
      "epoch": 0.3795898903195041,
      "grad_norm": 12.5625,
      "learning_rate": 1.746915935393616e-05,
      "loss": 1.0954,
      "mean_token_accuracy": 0.681636007130146,
      "step": 995
    },
    {
      "epoch": 0.38149737720553173,
      "grad_norm": 12.625,
      "learning_rate": 1.7456441561744885e-05,
      "loss": 1.1578,
      "mean_token_accuracy": 0.6307322606444359,
      "step": 1000
    },
    {
      "epoch": 0.3834048640915594,
      "grad_norm": 12.75,
      "learning_rate": 1.7443723769553607e-05,
      "loss": 1.1466,
      "mean_token_accuracy": 0.6797019310295582,
      "step": 1005
    },
    {
      "epoch": 0.38531235097758704,
      "grad_norm": 12.9375,
      "learning_rate": 1.7431005977362332e-05,
      "loss": 1.2437,
      "mean_token_accuracy": 0.665465061366558,
      "step": 1010
    },
    {
      "epoch": 0.3872198378636147,
      "grad_norm": 11.4375,
      "learning_rate": 1.7418288185171054e-05,
      "loss": 1.1273,
      "mean_token_accuracy": 0.6660737812519073,
      "step": 1015
    },
    {
      "epoch": 0.38912732474964234,
      "grad_norm": 10.875,
      "learning_rate": 1.740557039297978e-05,
      "loss": 1.0377,
      "mean_token_accuracy": 0.6913766637444496,
      "step": 1020
    },
    {
      "epoch": 0.39103481163567,
      "grad_norm": 11.6875,
      "learning_rate": 1.7392852600788502e-05,
      "loss": 1.2057,
      "mean_token_accuracy": 0.6652260690927505,
      "step": 1025
    },
    {
      "epoch": 0.39294229852169765,
      "grad_norm": 11.75,
      "learning_rate": 1.7380134808597228e-05,
      "loss": 1.1664,
      "mean_token_accuracy": 0.6586895115673542,
      "step": 1030
    },
    {
      "epoch": 0.3948497854077253,
      "grad_norm": 13.125,
      "learning_rate": 1.7367417016405953e-05,
      "loss": 1.2026,
      "mean_token_accuracy": 0.6655789151787758,
      "step": 1035
    },
    {
      "epoch": 0.39675727229375296,
      "grad_norm": 13.75,
      "learning_rate": 1.735469922421468e-05,
      "loss": 1.1472,
      "mean_token_accuracy": 0.67775809019804,
      "step": 1040
    },
    {
      "epoch": 0.3986647591797806,
      "grad_norm": 12.0,
      "learning_rate": 1.7341981432023404e-05,
      "loss": 1.2337,
      "mean_token_accuracy": 0.6594470135867596,
      "step": 1045
    },
    {
      "epoch": 0.4005722460658083,
      "grad_norm": 11.5,
      "learning_rate": 1.7329263639832126e-05,
      "loss": 1.1844,
      "mean_token_accuracy": 0.6957078948616982,
      "step": 1050
    },
    {
      "epoch": 0.402479732951836,
      "grad_norm": 12.75,
      "learning_rate": 1.731654584764085e-05,
      "loss": 1.0933,
      "mean_token_accuracy": 0.6744157910346985,
      "step": 1055
    },
    {
      "epoch": 0.40438721983786363,
      "grad_norm": 12.625,
      "learning_rate": 1.7303828055449577e-05,
      "loss": 1.2524,
      "mean_token_accuracy": 0.6677799835801125,
      "step": 1060
    },
    {
      "epoch": 0.4062947067238913,
      "grad_norm": 11.375,
      "learning_rate": 1.72911102632583e-05,
      "loss": 1.0756,
      "mean_token_accuracy": 0.689930209517479,
      "step": 1065
    },
    {
      "epoch": 0.40820219360991894,
      "grad_norm": 12.1875,
      "learning_rate": 1.7278392471067025e-05,
      "loss": 1.1762,
      "mean_token_accuracy": 0.6775659635663033,
      "step": 1070
    },
    {
      "epoch": 0.4101096804959466,
      "grad_norm": 10.875,
      "learning_rate": 1.7265674678875747e-05,
      "loss": 1.2004,
      "mean_token_accuracy": 0.6647816874086857,
      "step": 1075
    },
    {
      "epoch": 0.41201716738197425,
      "grad_norm": 12.4375,
      "learning_rate": 1.7252956886684472e-05,
      "loss": 1.0648,
      "mean_token_accuracy": 0.7046946279704571,
      "step": 1080
    },
    {
      "epoch": 0.4139246542680019,
      "grad_norm": 14.1875,
      "learning_rate": 1.7240239094493198e-05,
      "loss": 1.2307,
      "mean_token_accuracy": 0.6585713855922222,
      "step": 1085
    },
    {
      "epoch": 0.41583214115402956,
      "grad_norm": 11.5625,
      "learning_rate": 1.722752130230192e-05,
      "loss": 1.212,
      "mean_token_accuracy": 0.6784304775297642,
      "step": 1090
    },
    {
      "epoch": 0.4177396280400572,
      "grad_norm": 11.125,
      "learning_rate": 1.7214803510110646e-05,
      "loss": 1.1185,
      "mean_token_accuracy": 0.690502467751503,
      "step": 1095
    },
    {
      "epoch": 0.41964711492608486,
      "grad_norm": 11.3125,
      "learning_rate": 1.720208571791937e-05,
      "loss": 1.2032,
      "mean_token_accuracy": 0.643844348937273,
      "step": 1100
    },
    {
      "epoch": 0.4215546018121125,
      "grad_norm": 14.8125,
      "learning_rate": 1.7189367925728097e-05,
      "loss": 1.1605,
      "mean_token_accuracy": 0.6835091434419155,
      "step": 1105
    },
    {
      "epoch": 0.4234620886981402,
      "grad_norm": 14.0625,
      "learning_rate": 1.717665013353682e-05,
      "loss": 1.3624,
      "mean_token_accuracy": 0.6280046261847019,
      "step": 1110
    },
    {
      "epoch": 0.4253695755841679,
      "grad_norm": 14.0625,
      "learning_rate": 1.7163932341345544e-05,
      "loss": 1.2297,
      "mean_token_accuracy": 0.6916924893856049,
      "step": 1115
    },
    {
      "epoch": 0.42727706247019553,
      "grad_norm": 10.875,
      "learning_rate": 1.715121454915427e-05,
      "loss": 1.1743,
      "mean_token_accuracy": 0.6540060944855213,
      "step": 1120
    },
    {
      "epoch": 0.4291845493562232,
      "grad_norm": 11.875,
      "learning_rate": 1.7138496756962992e-05,
      "loss": 1.1755,
      "mean_token_accuracy": 0.6581062689423561,
      "step": 1125
    },
    {
      "epoch": 0.43109203624225084,
      "grad_norm": 12.125,
      "learning_rate": 1.7125778964771717e-05,
      "loss": 1.2194,
      "mean_token_accuracy": 0.665621517598629,
      "step": 1130
    },
    {
      "epoch": 0.4329995231282785,
      "grad_norm": 11.8125,
      "learning_rate": 1.711306117258044e-05,
      "loss": 1.0739,
      "mean_token_accuracy": 0.709334583580494,
      "step": 1135
    },
    {
      "epoch": 0.43490701001430615,
      "grad_norm": 14.125,
      "learning_rate": 1.7100343380389165e-05,
      "loss": 1.1086,
      "mean_token_accuracy": 0.6792764872312546,
      "step": 1140
    },
    {
      "epoch": 0.4368144969003338,
      "grad_norm": 12.0625,
      "learning_rate": 1.708762558819789e-05,
      "loss": 1.263,
      "mean_token_accuracy": 0.6668927378952503,
      "step": 1145
    },
    {
      "epoch": 0.43872198378636146,
      "grad_norm": 12.125,
      "learning_rate": 1.7074907796006616e-05,
      "loss": 1.1677,
      "mean_token_accuracy": 0.6571429625153542,
      "step": 1150
    },
    {
      "epoch": 0.4406294706723891,
      "grad_norm": 10.9375,
      "learning_rate": 1.706219000381534e-05,
      "loss": 1.102,
      "mean_token_accuracy": 0.6752917438745498,
      "step": 1155
    },
    {
      "epoch": 0.44253695755841677,
      "grad_norm": 14.0625,
      "learning_rate": 1.7049472211624063e-05,
      "loss": 1.1673,
      "mean_token_accuracy": 0.6898197934031487,
      "step": 1160
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 12.0,
      "learning_rate": 1.703675441943279e-05,
      "loss": 1.1781,
      "mean_token_accuracy": 0.6933842703700066,
      "step": 1165
    },
    {
      "epoch": 0.44635193133047213,
      "grad_norm": 12.75,
      "learning_rate": 1.702403662724151e-05,
      "loss": 1.1156,
      "mean_token_accuracy": 0.6760947778820992,
      "step": 1170
    },
    {
      "epoch": 0.4482594182164998,
      "grad_norm": 14.3125,
      "learning_rate": 1.7011318835050237e-05,
      "loss": 1.1263,
      "mean_token_accuracy": 0.657148540019989,
      "step": 1175
    },
    {
      "epoch": 0.45016690510252744,
      "grad_norm": 11.875,
      "learning_rate": 1.6998601042858962e-05,
      "loss": 1.1793,
      "mean_token_accuracy": 0.6318694293498993,
      "step": 1180
    },
    {
      "epoch": 0.4520743919885551,
      "grad_norm": 11.875,
      "learning_rate": 1.6985883250667684e-05,
      "loss": 1.1116,
      "mean_token_accuracy": 0.6872119233012199,
      "step": 1185
    },
    {
      "epoch": 0.45398187887458274,
      "grad_norm": 12.9375,
      "learning_rate": 1.697316545847641e-05,
      "loss": 1.2359,
      "mean_token_accuracy": 0.6691461108624935,
      "step": 1190
    },
    {
      "epoch": 0.4558893657606104,
      "grad_norm": 11.375,
      "learning_rate": 1.6960447666285135e-05,
      "loss": 1.0583,
      "mean_token_accuracy": 0.6924574553966523,
      "step": 1195
    },
    {
      "epoch": 0.45779685264663805,
      "grad_norm": 11.5625,
      "learning_rate": 1.6947729874093857e-05,
      "loss": 1.1006,
      "mean_token_accuracy": 0.6897717788815498,
      "step": 1200
    },
    {
      "epoch": 0.4597043395326657,
      "grad_norm": 17.375,
      "learning_rate": 1.6935012081902583e-05,
      "loss": 1.2088,
      "mean_token_accuracy": 0.6763196066021919,
      "step": 1205
    },
    {
      "epoch": 0.46161182641869336,
      "grad_norm": 12.5,
      "learning_rate": 1.6922294289711308e-05,
      "loss": 1.2367,
      "mean_token_accuracy": 0.6802153252065182,
      "step": 1210
    },
    {
      "epoch": 0.463519313304721,
      "grad_norm": 12.75,
      "learning_rate": 1.6909576497520034e-05,
      "loss": 1.1484,
      "mean_token_accuracy": 0.67167152389884,
      "step": 1215
    },
    {
      "epoch": 0.46542680019074867,
      "grad_norm": 11.25,
      "learning_rate": 1.6896858705328756e-05,
      "loss": 1.0818,
      "mean_token_accuracy": 0.6954056262969971,
      "step": 1220
    },
    {
      "epoch": 0.4673342870767763,
      "grad_norm": 12.9375,
      "learning_rate": 1.688414091313748e-05,
      "loss": 1.2057,
      "mean_token_accuracy": 0.6596958786249161,
      "step": 1225
    },
    {
      "epoch": 0.46924177396280403,
      "grad_norm": 10.5,
      "learning_rate": 1.6871423120946204e-05,
      "loss": 1.121,
      "mean_token_accuracy": 0.6667362228035927,
      "step": 1230
    },
    {
      "epoch": 0.4711492608488317,
      "grad_norm": 10.75,
      "learning_rate": 1.685870532875493e-05,
      "loss": 1.2371,
      "mean_token_accuracy": 0.6504265196621418,
      "step": 1235
    },
    {
      "epoch": 0.47305674773485934,
      "grad_norm": 14.1875,
      "learning_rate": 1.6845987536563655e-05,
      "loss": 1.2297,
      "mean_token_accuracy": 0.6640058435499668,
      "step": 1240
    },
    {
      "epoch": 0.474964234620887,
      "grad_norm": 13.4375,
      "learning_rate": 1.6833269744372377e-05,
      "loss": 1.2614,
      "mean_token_accuracy": 0.6515400633215904,
      "step": 1245
    },
    {
      "epoch": 0.47687172150691465,
      "grad_norm": 12.5625,
      "learning_rate": 1.6820551952181102e-05,
      "loss": 1.1777,
      "mean_token_accuracy": 0.6837553754448891,
      "step": 1250
    },
    {
      "epoch": 0.4787792083929423,
      "grad_norm": 12.25,
      "learning_rate": 1.6807834159989828e-05,
      "loss": 1.2948,
      "mean_token_accuracy": 0.6862575516104699,
      "step": 1255
    },
    {
      "epoch": 0.48068669527896996,
      "grad_norm": 12.3125,
      "learning_rate": 1.6795116367798553e-05,
      "loss": 1.0523,
      "mean_token_accuracy": 0.707449996471405,
      "step": 1260
    },
    {
      "epoch": 0.4825941821649976,
      "grad_norm": 12.25,
      "learning_rate": 1.6782398575607275e-05,
      "loss": 1.193,
      "mean_token_accuracy": 0.6783672124147415,
      "step": 1265
    },
    {
      "epoch": 0.48450166905102526,
      "grad_norm": 11.875,
      "learning_rate": 1.6769680783416e-05,
      "loss": 1.2587,
      "mean_token_accuracy": 0.6748954504728317,
      "step": 1270
    },
    {
      "epoch": 0.4864091559370529,
      "grad_norm": 12.8125,
      "learning_rate": 1.6756962991224726e-05,
      "loss": 1.1465,
      "mean_token_accuracy": 0.6888084650039673,
      "step": 1275
    },
    {
      "epoch": 0.48831664282308057,
      "grad_norm": 11.9375,
      "learning_rate": 1.674424519903345e-05,
      "loss": 1.1009,
      "mean_token_accuracy": 0.6846926853060722,
      "step": 1280
    },
    {
      "epoch": 0.4902241297091082,
      "grad_norm": 13.125,
      "learning_rate": 1.6731527406842174e-05,
      "loss": 1.1359,
      "mean_token_accuracy": 0.6763195641338825,
      "step": 1285
    },
    {
      "epoch": 0.49213161659513593,
      "grad_norm": 15.75,
      "learning_rate": 1.6718809614650896e-05,
      "loss": 1.0937,
      "mean_token_accuracy": 0.6663194999098778,
      "step": 1290
    },
    {
      "epoch": 0.4940391034811636,
      "grad_norm": 11.8125,
      "learning_rate": 1.670609182245962e-05,
      "loss": 1.1554,
      "mean_token_accuracy": 0.6896326825022697,
      "step": 1295
    },
    {
      "epoch": 0.49594659036719124,
      "grad_norm": 13.0,
      "learning_rate": 1.6693374030268347e-05,
      "loss": 1.1328,
      "mean_token_accuracy": 0.6875351458787918,
      "step": 1300
    },
    {
      "epoch": 0.4978540772532189,
      "grad_norm": 12.5,
      "learning_rate": 1.6680656238077072e-05,
      "loss": 1.1569,
      "mean_token_accuracy": 0.6798019893467426,
      "step": 1305
    },
    {
      "epoch": 0.49976156413924655,
      "grad_norm": 12.0625,
      "learning_rate": 1.6667938445885795e-05,
      "loss": 1.2192,
      "mean_token_accuracy": 0.6738032080233097,
      "step": 1310
    },
    {
      "epoch": 0.5016690510252741,
      "grad_norm": 12.5625,
      "learning_rate": 1.665522065369452e-05,
      "loss": 1.131,
      "mean_token_accuracy": 0.6774978704750538,
      "step": 1315
    },
    {
      "epoch": 0.5035765379113019,
      "grad_norm": 12.4375,
      "learning_rate": 1.6642502861503246e-05,
      "loss": 1.0869,
      "mean_token_accuracy": 0.6816208437085152,
      "step": 1320
    },
    {
      "epoch": 0.5054840247973296,
      "grad_norm": 12.125,
      "learning_rate": 1.6629785069311968e-05,
      "loss": 1.1625,
      "mean_token_accuracy": 0.6730111822485924,
      "step": 1325
    },
    {
      "epoch": 0.5073915116833572,
      "grad_norm": 12.4375,
      "learning_rate": 1.6617067277120693e-05,
      "loss": 1.1635,
      "mean_token_accuracy": 0.6660488948225975,
      "step": 1330
    },
    {
      "epoch": 0.5092989985693849,
      "grad_norm": 12.75,
      "learning_rate": 1.660434948492942e-05,
      "loss": 1.1048,
      "mean_token_accuracy": 0.6875575929880142,
      "step": 1335
    },
    {
      "epoch": 0.5112064854554125,
      "grad_norm": 11.5,
      "learning_rate": 1.659163169273814e-05,
      "loss": 1.0917,
      "mean_token_accuracy": 0.6965513400733471,
      "step": 1340
    },
    {
      "epoch": 0.5131139723414402,
      "grad_norm": 13.8125,
      "learning_rate": 1.6578913900546866e-05,
      "loss": 1.0532,
      "mean_token_accuracy": 0.6765949100255966,
      "step": 1345
    },
    {
      "epoch": 0.5150214592274678,
      "grad_norm": 12.25,
      "learning_rate": 1.656619610835559e-05,
      "loss": 1.225,
      "mean_token_accuracy": 0.6685496076941491,
      "step": 1350
    },
    {
      "epoch": 0.5169289461134955,
      "grad_norm": 11.875,
      "learning_rate": 1.6553478316164314e-05,
      "loss": 1.2317,
      "mean_token_accuracy": 0.6922548785805702,
      "step": 1355
    },
    {
      "epoch": 0.5188364329995231,
      "grad_norm": 10.5625,
      "learning_rate": 1.654076052397304e-05,
      "loss": 1.1384,
      "mean_token_accuracy": 0.6778303541243076,
      "step": 1360
    },
    {
      "epoch": 0.5207439198855508,
      "grad_norm": 12.125,
      "learning_rate": 1.6528042731781765e-05,
      "loss": 1.0721,
      "mean_token_accuracy": 0.6984316289424897,
      "step": 1365
    },
    {
      "epoch": 0.5226514067715784,
      "grad_norm": 15.125,
      "learning_rate": 1.651532493959049e-05,
      "loss": 1.1862,
      "mean_token_accuracy": 0.6730318419635296,
      "step": 1370
    },
    {
      "epoch": 0.5245588936576061,
      "grad_norm": 11.9375,
      "learning_rate": 1.6502607147399213e-05,
      "loss": 1.0065,
      "mean_token_accuracy": 0.7321307882666588,
      "step": 1375
    },
    {
      "epoch": 0.5264663805436338,
      "grad_norm": 13.0,
      "learning_rate": 1.6489889355207938e-05,
      "loss": 1.3213,
      "mean_token_accuracy": 0.6621546871960163,
      "step": 1380
    },
    {
      "epoch": 0.5283738674296614,
      "grad_norm": 14.9375,
      "learning_rate": 1.647717156301666e-05,
      "loss": 1.219,
      "mean_token_accuracy": 0.6426276586949825,
      "step": 1385
    },
    {
      "epoch": 0.5302813543156891,
      "grad_norm": 11.375,
      "learning_rate": 1.6464453770825386e-05,
      "loss": 1.0934,
      "mean_token_accuracy": 0.6920969501137734,
      "step": 1390
    },
    {
      "epoch": 0.5321888412017167,
      "grad_norm": 12.0625,
      "learning_rate": 1.645173597863411e-05,
      "loss": 1.1581,
      "mean_token_accuracy": 0.6565943203866482,
      "step": 1395
    },
    {
      "epoch": 0.5340963280877444,
      "grad_norm": 12.375,
      "learning_rate": 1.6439018186442833e-05,
      "loss": 1.1194,
      "mean_token_accuracy": 0.6673160851001739,
      "step": 1400
    },
    {
      "epoch": 0.536003814973772,
      "grad_norm": 11.5625,
      "learning_rate": 1.642630039425156e-05,
      "loss": 1.1693,
      "mean_token_accuracy": 0.6715315200388432,
      "step": 1405
    },
    {
      "epoch": 0.5379113018597997,
      "grad_norm": 15.0,
      "learning_rate": 1.6413582602060284e-05,
      "loss": 1.1414,
      "mean_token_accuracy": 0.6831993453204632,
      "step": 1410
    },
    {
      "epoch": 0.5398187887458273,
      "grad_norm": 14.3125,
      "learning_rate": 1.640086480986901e-05,
      "loss": 1.1978,
      "mean_token_accuracy": 0.6750304147601127,
      "step": 1415
    },
    {
      "epoch": 0.541726275631855,
      "grad_norm": 10.0625,
      "learning_rate": 1.6388147017677732e-05,
      "loss": 1.1056,
      "mean_token_accuracy": 0.6595890045166015,
      "step": 1420
    },
    {
      "epoch": 0.5436337625178826,
      "grad_norm": 12.5625,
      "learning_rate": 1.6375429225486457e-05,
      "loss": 1.1298,
      "mean_token_accuracy": 0.684663999825716,
      "step": 1425
    },
    {
      "epoch": 0.5455412494039104,
      "grad_norm": 12.1875,
      "learning_rate": 1.6362711433295183e-05,
      "loss": 1.0179,
      "mean_token_accuracy": 0.7125121988356113,
      "step": 1430
    },
    {
      "epoch": 0.547448736289938,
      "grad_norm": 13.1875,
      "learning_rate": 1.6349993641103905e-05,
      "loss": 1.1698,
      "mean_token_accuracy": 0.6814395613968373,
      "step": 1435
    },
    {
      "epoch": 0.5493562231759657,
      "grad_norm": 14.25,
      "learning_rate": 1.633727584891263e-05,
      "loss": 1.2112,
      "mean_token_accuracy": 0.6521120138466359,
      "step": 1440
    },
    {
      "epoch": 0.5512637100619934,
      "grad_norm": 14.25,
      "learning_rate": 1.6324558056721353e-05,
      "loss": 1.0544,
      "mean_token_accuracy": 0.7097316659986973,
      "step": 1445
    },
    {
      "epoch": 0.553171196948021,
      "grad_norm": 10.875,
      "learning_rate": 1.6311840264530078e-05,
      "loss": 1.1943,
      "mean_token_accuracy": 0.6261555813252926,
      "step": 1450
    },
    {
      "epoch": 0.5550786838340487,
      "grad_norm": 12.375,
      "learning_rate": 1.6299122472338804e-05,
      "loss": 1.2382,
      "mean_token_accuracy": 0.6481642045080662,
      "step": 1455
    },
    {
      "epoch": 0.5569861707200763,
      "grad_norm": 11.9375,
      "learning_rate": 1.6286404680147526e-05,
      "loss": 1.1301,
      "mean_token_accuracy": 0.6791691981256008,
      "step": 1460
    },
    {
      "epoch": 0.558893657606104,
      "grad_norm": 12.125,
      "learning_rate": 1.627368688795625e-05,
      "loss": 1.1533,
      "mean_token_accuracy": 0.6759828574955463,
      "step": 1465
    },
    {
      "epoch": 0.5608011444921316,
      "grad_norm": 10.4375,
      "learning_rate": 1.6260969095764977e-05,
      "loss": 1.1125,
      "mean_token_accuracy": 0.6778176836669445,
      "step": 1470
    },
    {
      "epoch": 0.5627086313781593,
      "grad_norm": 11.5625,
      "learning_rate": 1.6248251303573702e-05,
      "loss": 1.194,
      "mean_token_accuracy": 0.6579376481473446,
      "step": 1475
    },
    {
      "epoch": 0.5646161182641869,
      "grad_norm": 13.75,
      "learning_rate": 1.6235533511382428e-05,
      "loss": 1.1965,
      "mean_token_accuracy": 0.6491686496883631,
      "step": 1480
    },
    {
      "epoch": 0.5665236051502146,
      "grad_norm": 13.4375,
      "learning_rate": 1.622281571919115e-05,
      "loss": 1.1953,
      "mean_token_accuracy": 0.6784386828541755,
      "step": 1485
    },
    {
      "epoch": 0.5684310920362422,
      "grad_norm": 11.875,
      "learning_rate": 1.6210097926999875e-05,
      "loss": 1.2534,
      "mean_token_accuracy": 0.663827870041132,
      "step": 1490
    },
    {
      "epoch": 0.5703385789222699,
      "grad_norm": 11.75,
      "learning_rate": 1.6197380134808597e-05,
      "loss": 1.163,
      "mean_token_accuracy": 0.6678325220942497,
      "step": 1495
    },
    {
      "epoch": 0.5722460658082976,
      "grad_norm": 11.125,
      "learning_rate": 1.6184662342617323e-05,
      "loss": 1.1818,
      "mean_token_accuracy": 0.6880650117993354,
      "step": 1500
    },
    {
      "epoch": 0.5741535526943252,
      "grad_norm": 11.9375,
      "learning_rate": 1.6171944550426045e-05,
      "loss": 1.1857,
      "mean_token_accuracy": 0.6938965663313865,
      "step": 1505
    },
    {
      "epoch": 0.5760610395803529,
      "grad_norm": 13.75,
      "learning_rate": 1.615922675823477e-05,
      "loss": 1.1482,
      "mean_token_accuracy": 0.696337477862835,
      "step": 1510
    },
    {
      "epoch": 0.5779685264663805,
      "grad_norm": 11.125,
      "learning_rate": 1.6146508966043496e-05,
      "loss": 1.112,
      "mean_token_accuracy": 0.6995782241225242,
      "step": 1515
    },
    {
      "epoch": 0.5798760133524082,
      "grad_norm": 12.5,
      "learning_rate": 1.613379117385222e-05,
      "loss": 1.157,
      "mean_token_accuracy": 0.6632382079958916,
      "step": 1520
    },
    {
      "epoch": 0.5817835002384358,
      "grad_norm": 14.1875,
      "learning_rate": 1.6121073381660947e-05,
      "loss": 1.1297,
      "mean_token_accuracy": 0.6936305865645409,
      "step": 1525
    },
    {
      "epoch": 0.5836909871244635,
      "grad_norm": 12.25,
      "learning_rate": 1.610835558946967e-05,
      "loss": 1.182,
      "mean_token_accuracy": 0.6679765857756138,
      "step": 1530
    },
    {
      "epoch": 0.5855984740104911,
      "grad_norm": 15.25,
      "learning_rate": 1.6095637797278395e-05,
      "loss": 1.1584,
      "mean_token_accuracy": 0.6810820989310742,
      "step": 1535
    },
    {
      "epoch": 0.5875059608965189,
      "grad_norm": 13.625,
      "learning_rate": 1.608292000508712e-05,
      "loss": 1.173,
      "mean_token_accuracy": 0.6857012517750263,
      "step": 1540
    },
    {
      "epoch": 0.5894134477825465,
      "grad_norm": 11.3125,
      "learning_rate": 1.6070202212895842e-05,
      "loss": 1.1198,
      "mean_token_accuracy": 0.6993751585483551,
      "step": 1545
    },
    {
      "epoch": 0.5913209346685742,
      "grad_norm": 11.625,
      "learning_rate": 1.6057484420704568e-05,
      "loss": 1.1397,
      "mean_token_accuracy": 0.682830485701561,
      "step": 1550
    },
    {
      "epoch": 0.5932284215546018,
      "grad_norm": 15.0,
      "learning_rate": 1.604476662851329e-05,
      "loss": 1.1759,
      "mean_token_accuracy": 0.6623894415795804,
      "step": 1555
    },
    {
      "epoch": 0.5951359084406295,
      "grad_norm": 14.1875,
      "learning_rate": 1.6032048836322015e-05,
      "loss": 1.1408,
      "mean_token_accuracy": 0.658209715038538,
      "step": 1560
    },
    {
      "epoch": 0.5970433953266572,
      "grad_norm": 12.3125,
      "learning_rate": 1.601933104413074e-05,
      "loss": 1.1133,
      "mean_token_accuracy": 0.7014020837843418,
      "step": 1565
    },
    {
      "epoch": 0.5989508822126848,
      "grad_norm": 11.0625,
      "learning_rate": 1.6006613251939463e-05,
      "loss": 1.1147,
      "mean_token_accuracy": 0.6807858303189278,
      "step": 1570
    },
    {
      "epoch": 0.6008583690987125,
      "grad_norm": 12.9375,
      "learning_rate": 1.599389545974819e-05,
      "loss": 1.1151,
      "mean_token_accuracy": 0.6718579567968845,
      "step": 1575
    },
    {
      "epoch": 0.6027658559847401,
      "grad_norm": 12.125,
      "learning_rate": 1.5981177667556914e-05,
      "loss": 1.0995,
      "mean_token_accuracy": 0.6849846243858337,
      "step": 1580
    },
    {
      "epoch": 0.6046733428707678,
      "grad_norm": 13.9375,
      "learning_rate": 1.596845987536564e-05,
      "loss": 1.2112,
      "mean_token_accuracy": 0.6757240429520607,
      "step": 1585
    },
    {
      "epoch": 0.6065808297567954,
      "grad_norm": 11.9375,
      "learning_rate": 1.595574208317436e-05,
      "loss": 1.1318,
      "mean_token_accuracy": 0.6792904518544673,
      "step": 1590
    },
    {
      "epoch": 0.6084883166428231,
      "grad_norm": 11.75,
      "learning_rate": 1.5943024290983087e-05,
      "loss": 1.1354,
      "mean_token_accuracy": 0.6702633060514926,
      "step": 1595
    },
    {
      "epoch": 0.6103958035288507,
      "grad_norm": 11.75,
      "learning_rate": 1.5930306498791813e-05,
      "loss": 1.0983,
      "mean_token_accuracy": 0.6991154357790947,
      "step": 1600
    },
    {
      "epoch": 0.6123032904148784,
      "grad_norm": 12.875,
      "learning_rate": 1.5917588706600535e-05,
      "loss": 1.0745,
      "mean_token_accuracy": 0.681308564543724,
      "step": 1605
    },
    {
      "epoch": 0.614210777300906,
      "grad_norm": 15.1875,
      "learning_rate": 1.590487091440926e-05,
      "loss": 1.0767,
      "mean_token_accuracy": 0.7030037745833397,
      "step": 1610
    },
    {
      "epoch": 0.6161182641869337,
      "grad_norm": 10.8125,
      "learning_rate": 1.5892153122217982e-05,
      "loss": 1.1064,
      "mean_token_accuracy": 0.6708254374563694,
      "step": 1615
    },
    {
      "epoch": 0.6180257510729614,
      "grad_norm": 12.875,
      "learning_rate": 1.5879435330026708e-05,
      "loss": 1.13,
      "mean_token_accuracy": 0.6800018325448036,
      "step": 1620
    },
    {
      "epoch": 0.619933237958989,
      "grad_norm": 11.0625,
      "learning_rate": 1.5866717537835433e-05,
      "loss": 1.0084,
      "mean_token_accuracy": 0.7097429156303405,
      "step": 1625
    },
    {
      "epoch": 0.6218407248450167,
      "grad_norm": 13.1875,
      "learning_rate": 1.585399974564416e-05,
      "loss": 1.062,
      "mean_token_accuracy": 0.7025215804576874,
      "step": 1630
    },
    {
      "epoch": 0.6237482117310443,
      "grad_norm": 14.75,
      "learning_rate": 1.5841281953452884e-05,
      "loss": 1.1089,
      "mean_token_accuracy": 0.6919012367725372,
      "step": 1635
    },
    {
      "epoch": 0.625655698617072,
      "grad_norm": 11.875,
      "learning_rate": 1.5828564161261606e-05,
      "loss": 1.1117,
      "mean_token_accuracy": 0.6917063646018505,
      "step": 1640
    },
    {
      "epoch": 0.6275631855030996,
      "grad_norm": 12.125,
      "learning_rate": 1.5815846369070332e-05,
      "loss": 1.1046,
      "mean_token_accuracy": 0.6739561937749385,
      "step": 1645
    },
    {
      "epoch": 0.6294706723891274,
      "grad_norm": 13.375,
      "learning_rate": 1.5803128576879054e-05,
      "loss": 1.1264,
      "mean_token_accuracy": 0.6709033444523811,
      "step": 1650
    },
    {
      "epoch": 0.631378159275155,
      "grad_norm": 13.375,
      "learning_rate": 1.579041078468778e-05,
      "loss": 1.2637,
      "mean_token_accuracy": 0.669117308408022,
      "step": 1655
    },
    {
      "epoch": 0.6332856461611827,
      "grad_norm": 14.5625,
      "learning_rate": 1.5777692992496505e-05,
      "loss": 1.1087,
      "mean_token_accuracy": 0.6606097273528576,
      "step": 1660
    },
    {
      "epoch": 0.6351931330472103,
      "grad_norm": 10.75,
      "learning_rate": 1.5764975200305227e-05,
      "loss": 1.107,
      "mean_token_accuracy": 0.676217395812273,
      "step": 1665
    },
    {
      "epoch": 0.637100619933238,
      "grad_norm": 13.4375,
      "learning_rate": 1.5752257408113953e-05,
      "loss": 1.153,
      "mean_token_accuracy": 0.6523927465081215,
      "step": 1670
    },
    {
      "epoch": 0.6390081068192656,
      "grad_norm": 11.8125,
      "learning_rate": 1.5739539615922678e-05,
      "loss": 1.2302,
      "mean_token_accuracy": 0.6501938365399837,
      "step": 1675
    },
    {
      "epoch": 0.6409155937052933,
      "grad_norm": 11.0625,
      "learning_rate": 1.57268218237314e-05,
      "loss": 1.0566,
      "mean_token_accuracy": 0.6942887641489506,
      "step": 1680
    },
    {
      "epoch": 0.642823080591321,
      "grad_norm": 14.25,
      "learning_rate": 1.5714104031540126e-05,
      "loss": 1.224,
      "mean_token_accuracy": 0.6493762314319611,
      "step": 1685
    },
    {
      "epoch": 0.6447305674773486,
      "grad_norm": 11.375,
      "learning_rate": 1.570138623934885e-05,
      "loss": 1.0819,
      "mean_token_accuracy": 0.6909515604376792,
      "step": 1690
    },
    {
      "epoch": 0.6466380543633763,
      "grad_norm": 11.75,
      "learning_rate": 1.5688668447157577e-05,
      "loss": 1.0433,
      "mean_token_accuracy": 0.6917465507984162,
      "step": 1695
    },
    {
      "epoch": 0.6485455412494039,
      "grad_norm": 13.5625,
      "learning_rate": 1.56759506549663e-05,
      "loss": 1.1417,
      "mean_token_accuracy": 0.6675561875104904,
      "step": 1700
    },
    {
      "epoch": 0.6504530281354316,
      "grad_norm": 15.875,
      "learning_rate": 1.5663232862775024e-05,
      "loss": 1.1317,
      "mean_token_accuracy": 0.6959464140236378,
      "step": 1705
    },
    {
      "epoch": 0.6523605150214592,
      "grad_norm": 14.375,
      "learning_rate": 1.5650515070583746e-05,
      "loss": 1.0547,
      "mean_token_accuracy": 0.6967534452676774,
      "step": 1710
    },
    {
      "epoch": 0.6542680019074869,
      "grad_norm": 11.75,
      "learning_rate": 1.5637797278392472e-05,
      "loss": 1.2029,
      "mean_token_accuracy": 0.6825399219989776,
      "step": 1715
    },
    {
      "epoch": 0.6561754887935145,
      "grad_norm": 12.375,
      "learning_rate": 1.5625079486201197e-05,
      "loss": 1.2765,
      "mean_token_accuracy": 0.6506742417812348,
      "step": 1720
    },
    {
      "epoch": 0.6580829756795422,
      "grad_norm": 10.875,
      "learning_rate": 1.561236169400992e-05,
      "loss": 1.1471,
      "mean_token_accuracy": 0.6708492331206799,
      "step": 1725
    },
    {
      "epoch": 0.6599904625655698,
      "grad_norm": 11.8125,
      "learning_rate": 1.5599643901818645e-05,
      "loss": 0.9993,
      "mean_token_accuracy": 0.7227230161428452,
      "step": 1730
    },
    {
      "epoch": 0.6618979494515975,
      "grad_norm": 12.5625,
      "learning_rate": 1.558692610962737e-05,
      "loss": 1.155,
      "mean_token_accuracy": 0.6793848961591721,
      "step": 1735
    },
    {
      "epoch": 0.6638054363376252,
      "grad_norm": 12.25,
      "learning_rate": 1.5574208317436096e-05,
      "loss": 1.1064,
      "mean_token_accuracy": 0.6699606157839298,
      "step": 1740
    },
    {
      "epoch": 0.6657129232236528,
      "grad_norm": 12.1875,
      "learning_rate": 1.5561490525244818e-05,
      "loss": 1.0567,
      "mean_token_accuracy": 0.704989293217659,
      "step": 1745
    },
    {
      "epoch": 0.6676204101096805,
      "grad_norm": 13.375,
      "learning_rate": 1.5548772733053544e-05,
      "loss": 1.0546,
      "mean_token_accuracy": 0.6963534206151962,
      "step": 1750
    },
    {
      "epoch": 0.6695278969957081,
      "grad_norm": 14.4375,
      "learning_rate": 1.553605494086227e-05,
      "loss": 1.1009,
      "mean_token_accuracy": 0.6913698375225067,
      "step": 1755
    },
    {
      "epoch": 0.6714353838817358,
      "grad_norm": 11.75,
      "learning_rate": 1.552333714867099e-05,
      "loss": 1.1649,
      "mean_token_accuracy": 0.6772311896085739,
      "step": 1760
    },
    {
      "epoch": 0.6733428707677634,
      "grad_norm": 10.625,
      "learning_rate": 1.5510619356479717e-05,
      "loss": 1.0516,
      "mean_token_accuracy": 0.6774869538843632,
      "step": 1765
    },
    {
      "epoch": 0.6752503576537912,
      "grad_norm": 11.0625,
      "learning_rate": 1.549790156428844e-05,
      "loss": 1.1433,
      "mean_token_accuracy": 0.6827952593564988,
      "step": 1770
    },
    {
      "epoch": 0.6771578445398188,
      "grad_norm": 12.4375,
      "learning_rate": 1.5485183772097164e-05,
      "loss": 1.2089,
      "mean_token_accuracy": 0.6664685860276223,
      "step": 1775
    },
    {
      "epoch": 0.6790653314258465,
      "grad_norm": 12.625,
      "learning_rate": 1.547246597990589e-05,
      "loss": 1.108,
      "mean_token_accuracy": 0.6785644575953483,
      "step": 1780
    },
    {
      "epoch": 0.6809728183118741,
      "grad_norm": 12.0625,
      "learning_rate": 1.5459748187714615e-05,
      "loss": 1.123,
      "mean_token_accuracy": 0.6967614680528641,
      "step": 1785
    },
    {
      "epoch": 0.6828803051979018,
      "grad_norm": 11.1875,
      "learning_rate": 1.5447030395523338e-05,
      "loss": 1.0844,
      "mean_token_accuracy": 0.6894414715468884,
      "step": 1790
    },
    {
      "epoch": 0.6847877920839294,
      "grad_norm": 11.5,
      "learning_rate": 1.5434312603332063e-05,
      "loss": 1.211,
      "mean_token_accuracy": 0.6685944899916649,
      "step": 1795
    },
    {
      "epoch": 0.6866952789699571,
      "grad_norm": 12.4375,
      "learning_rate": 1.542159481114079e-05,
      "loss": 1.1606,
      "mean_token_accuracy": 0.6708570115268231,
      "step": 1800
    },
    {
      "epoch": 0.6886027658559848,
      "grad_norm": 13.1875,
      "learning_rate": 1.540887701894951e-05,
      "loss": 1.1955,
      "mean_token_accuracy": 0.6612754940986634,
      "step": 1805
    },
    {
      "epoch": 0.6905102527420124,
      "grad_norm": 13.0625,
      "learning_rate": 1.5396159226758236e-05,
      "loss": 1.1045,
      "mean_token_accuracy": 0.7275897234678268,
      "step": 1810
    },
    {
      "epoch": 0.6924177396280401,
      "grad_norm": 12.5,
      "learning_rate": 1.538344143456696e-05,
      "loss": 1.1093,
      "mean_token_accuracy": 0.7059011451900006,
      "step": 1815
    },
    {
      "epoch": 0.6943252265140677,
      "grad_norm": 12.75,
      "learning_rate": 1.5370723642375684e-05,
      "loss": 1.0552,
      "mean_token_accuracy": 0.7078164145350456,
      "step": 1820
    },
    {
      "epoch": 0.6962327134000954,
      "grad_norm": 14.0,
      "learning_rate": 1.535800585018441e-05,
      "loss": 1.1243,
      "mean_token_accuracy": 0.6750999264419079,
      "step": 1825
    },
    {
      "epoch": 0.698140200286123,
      "grad_norm": 13.1875,
      "learning_rate": 1.534528805799313e-05,
      "loss": 1.0626,
      "mean_token_accuracy": 0.6938546419143676,
      "step": 1830
    },
    {
      "epoch": 0.7000476871721507,
      "grad_norm": 11.625,
      "learning_rate": 1.5332570265801857e-05,
      "loss": 1.0889,
      "mean_token_accuracy": 0.6872342020273209,
      "step": 1835
    },
    {
      "epoch": 0.7019551740581783,
      "grad_norm": 11.875,
      "learning_rate": 1.5319852473610582e-05,
      "loss": 1.0794,
      "mean_token_accuracy": 0.7006407782435418,
      "step": 1840
    },
    {
      "epoch": 0.703862660944206,
      "grad_norm": 12.1875,
      "learning_rate": 1.5307134681419308e-05,
      "loss": 1.1733,
      "mean_token_accuracy": 0.6737750858068466,
      "step": 1845
    },
    {
      "epoch": 0.7057701478302336,
      "grad_norm": 10.6875,
      "learning_rate": 1.5294416889228033e-05,
      "loss": 1.0611,
      "mean_token_accuracy": 0.7101491972804069,
      "step": 1850
    },
    {
      "epoch": 0.7076776347162613,
      "grad_norm": 12.625,
      "learning_rate": 1.5281699097036755e-05,
      "loss": 1.197,
      "mean_token_accuracy": 0.6738182365894317,
      "step": 1855
    },
    {
      "epoch": 0.709585121602289,
      "grad_norm": 10.625,
      "learning_rate": 1.526898130484548e-05,
      "loss": 0.9995,
      "mean_token_accuracy": 0.7093352362513542,
      "step": 1860
    },
    {
      "epoch": 0.7114926084883166,
      "grad_norm": 10.875,
      "learning_rate": 1.5256263512654203e-05,
      "loss": 1.1021,
      "mean_token_accuracy": 0.7152903437614441,
      "step": 1865
    },
    {
      "epoch": 0.7134000953743443,
      "grad_norm": 12.9375,
      "learning_rate": 1.5243545720462929e-05,
      "loss": 1.1298,
      "mean_token_accuracy": 0.6734328977763653,
      "step": 1870
    },
    {
      "epoch": 0.7153075822603719,
      "grad_norm": 13.6875,
      "learning_rate": 1.5230827928271654e-05,
      "loss": 1.0921,
      "mean_token_accuracy": 0.6812200739979744,
      "step": 1875
    },
    {
      "epoch": 0.7172150691463997,
      "grad_norm": 14.0,
      "learning_rate": 1.5218110136080378e-05,
      "loss": 1.0641,
      "mean_token_accuracy": 0.6840920761227608,
      "step": 1880
    },
    {
      "epoch": 0.7191225560324273,
      "grad_norm": 12.125,
      "learning_rate": 1.5205392343889103e-05,
      "loss": 1.1002,
      "mean_token_accuracy": 0.6968449607491494,
      "step": 1885
    },
    {
      "epoch": 0.721030042918455,
      "grad_norm": 12.8125,
      "learning_rate": 1.5192674551697825e-05,
      "loss": 1.1009,
      "mean_token_accuracy": 0.6776654615998268,
      "step": 1890
    },
    {
      "epoch": 0.7229375298044826,
      "grad_norm": 13.5625,
      "learning_rate": 1.5179956759506551e-05,
      "loss": 1.1619,
      "mean_token_accuracy": 0.6722353339195252,
      "step": 1895
    },
    {
      "epoch": 0.7248450166905103,
      "grad_norm": 14.8125,
      "learning_rate": 1.5167238967315275e-05,
      "loss": 1.0884,
      "mean_token_accuracy": 0.7083397641777992,
      "step": 1900
    },
    {
      "epoch": 0.7267525035765379,
      "grad_norm": 12.3125,
      "learning_rate": 1.5154521175124e-05,
      "loss": 1.1197,
      "mean_token_accuracy": 0.7005624413490296,
      "step": 1905
    },
    {
      "epoch": 0.7286599904625656,
      "grad_norm": 12.375,
      "learning_rate": 1.5141803382932724e-05,
      "loss": 1.1168,
      "mean_token_accuracy": 0.6706736356019973,
      "step": 1910
    },
    {
      "epoch": 0.7305674773485932,
      "grad_norm": 12.3125,
      "learning_rate": 1.5129085590741448e-05,
      "loss": 1.1169,
      "mean_token_accuracy": 0.6706511251628399,
      "step": 1915
    },
    {
      "epoch": 0.7324749642346209,
      "grad_norm": 12.9375,
      "learning_rate": 1.5116367798550173e-05,
      "loss": 1.0641,
      "mean_token_accuracy": 0.7128054112195968,
      "step": 1920
    },
    {
      "epoch": 0.7343824511206486,
      "grad_norm": 13.5625,
      "learning_rate": 1.5103650006358897e-05,
      "loss": 1.0722,
      "mean_token_accuracy": 0.6964303895831108,
      "step": 1925
    },
    {
      "epoch": 0.7362899380066762,
      "grad_norm": 12.5,
      "learning_rate": 1.5090932214167621e-05,
      "loss": 1.1099,
      "mean_token_accuracy": 0.670535609871149,
      "step": 1930
    },
    {
      "epoch": 0.7381974248927039,
      "grad_norm": 12.875,
      "learning_rate": 1.5078214421976347e-05,
      "loss": 1.1376,
      "mean_token_accuracy": 0.6716626837849617,
      "step": 1935
    },
    {
      "epoch": 0.7401049117787315,
      "grad_norm": 12.5625,
      "learning_rate": 1.506549662978507e-05,
      "loss": 1.1237,
      "mean_token_accuracy": 0.6802854061126709,
      "step": 1940
    },
    {
      "epoch": 0.7420123986647592,
      "grad_norm": 11.5,
      "learning_rate": 1.5052778837593796e-05,
      "loss": 1.1543,
      "mean_token_accuracy": 0.6751244008541107,
      "step": 1945
    },
    {
      "epoch": 0.7439198855507868,
      "grad_norm": 11.0,
      "learning_rate": 1.5040061045402518e-05,
      "loss": 1.0762,
      "mean_token_accuracy": 0.682145906984806,
      "step": 1950
    },
    {
      "epoch": 0.7458273724368145,
      "grad_norm": 11.6875,
      "learning_rate": 1.5027343253211243e-05,
      "loss": 1.1921,
      "mean_token_accuracy": 0.6897265024483203,
      "step": 1955
    },
    {
      "epoch": 0.7477348593228421,
      "grad_norm": 11.875,
      "learning_rate": 1.5014625461019967e-05,
      "loss": 1.1296,
      "mean_token_accuracy": 0.6912665694952012,
      "step": 1960
    },
    {
      "epoch": 0.7496423462088698,
      "grad_norm": 11.125,
      "learning_rate": 1.5001907668828693e-05,
      "loss": 0.997,
      "mean_token_accuracy": 0.7142476089298725,
      "step": 1965
    },
    {
      "epoch": 0.7515498330948974,
      "grad_norm": 15.375,
      "learning_rate": 1.4989189876637418e-05,
      "loss": 1.105,
      "mean_token_accuracy": 0.6805791586637497,
      "step": 1970
    },
    {
      "epoch": 0.7534573199809251,
      "grad_norm": 11.5625,
      "learning_rate": 1.497647208444614e-05,
      "loss": 1.1086,
      "mean_token_accuracy": 0.6618591591715812,
      "step": 1975
    },
    {
      "epoch": 0.7553648068669528,
      "grad_norm": 11.125,
      "learning_rate": 1.4963754292254866e-05,
      "loss": 1.2405,
      "mean_token_accuracy": 0.6381066031754017,
      "step": 1980
    },
    {
      "epoch": 0.7572722937529804,
      "grad_norm": 10.9375,
      "learning_rate": 1.495103650006359e-05,
      "loss": 1.0646,
      "mean_token_accuracy": 0.6994791410863399,
      "step": 1985
    },
    {
      "epoch": 0.7591797806390082,
      "grad_norm": 11.1875,
      "learning_rate": 1.4938318707872315e-05,
      "loss": 1.0498,
      "mean_token_accuracy": 0.721826684474945,
      "step": 1990
    },
    {
      "epoch": 0.7610872675250357,
      "grad_norm": 12.75,
      "learning_rate": 1.492560091568104e-05,
      "loss": 1.0933,
      "mean_token_accuracy": 0.677222141623497,
      "step": 1995
    },
    {
      "epoch": 0.7629947544110635,
      "grad_norm": 12.0,
      "learning_rate": 1.4912883123489763e-05,
      "loss": 1.0058,
      "mean_token_accuracy": 0.7074526056647301,
      "step": 2000
    }
  ],
  "logging_steps": 5,
  "max_steps": 7863,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.3261800470707405e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}