{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998149861239592,
  "eval_steps": 100,
  "global_step": 1351,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0037002775208140612,
      "grad_norm": 3.239915386527218,
      "learning_rate": 7.352941176470589e-07,
      "loss": 1.2082,
      "mean_token_accuracy": 0.6904086382244544,
      "step": 5
    },
    {
      "epoch": 0.0074005550416281225,
      "grad_norm": 2.8480570558845186,
      "learning_rate": 1.4705882352941177e-06,
      "loss": 1.2101,
      "mean_token_accuracy": 0.688909531616815,
      "step": 10
    },
    {
      "epoch": 0.011100832562442183,
      "grad_norm": 2.425023099484193,
      "learning_rate": 2.2058823529411767e-06,
      "loss": 1.1302,
      "mean_token_accuracy": 0.7074111736775649,
      "step": 15
    },
    {
      "epoch": 0.014801110083256245,
      "grad_norm": 1.8152419374117768,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 1.1277,
      "mean_token_accuracy": 0.704069345586691,
      "step": 20
    },
    {
      "epoch": 0.018501387604070305,
      "grad_norm": 1.5172611375815148,
      "learning_rate": 3.6764705882352946e-06,
      "loss": 1.0942,
      "mean_token_accuracy": 0.7089358281360809,
      "step": 25
    },
    {
      "epoch": 0.022201665124884366,
      "grad_norm": 1.2356119543050552,
      "learning_rate": 4.411764705882353e-06,
      "loss": 1.04,
      "mean_token_accuracy": 0.7169246015229918,
      "step": 30
    },
    {
      "epoch": 0.025901942645698426,
      "grad_norm": 1.2317231184779383,
      "learning_rate": 5.147058823529411e-06,
      "loss": 1.0126,
      "mean_token_accuracy": 0.720928252798097,
      "step": 35
    },
    {
      "epoch": 0.02960222016651249,
      "grad_norm": 1.094000980605897,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.9576,
      "mean_token_accuracy": 0.7326068845537983,
      "step": 40
    },
    {
      "epoch": 0.03330249768732655,
      "grad_norm": 0.8898258637631962,
      "learning_rate": 6.61764705882353e-06,
      "loss": 0.9722,
      "mean_token_accuracy": 0.7278379936229317,
      "step": 45
    },
    {
      "epoch": 0.03700277520814061,
      "grad_norm": 0.9622645010733354,
      "learning_rate": 7.352941176470589e-06,
      "loss": 0.9281,
      "mean_token_accuracy": 0.7367932115453535,
      "step": 50
    },
    {
      "epoch": 0.04070305272895467,
      "grad_norm": 0.9736757281340813,
      "learning_rate": 8.088235294117648e-06,
      "loss": 0.9054,
      "mean_token_accuracy": 0.7442706350426345,
      "step": 55
    },
    {
      "epoch": 0.04440333024976873,
      "grad_norm": 1.042826803009636,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.8948,
      "mean_token_accuracy": 0.7443939037516841,
      "step": 60
    },
    {
      "epoch": 0.04810360777058279,
      "grad_norm": 0.9490284166948695,
      "learning_rate": 9.558823529411766e-06,
      "loss": 0.9171,
      "mean_token_accuracy": 0.7375623924285362,
      "step": 65
    },
    {
      "epoch": 0.05180388529139685,
      "grad_norm": 0.9728295714818891,
      "learning_rate": 1.0294117647058823e-05,
      "loss": 0.8826,
      "mean_token_accuracy": 0.7468788718777591,
      "step": 70
    },
    {
      "epoch": 0.05550416281221091,
      "grad_norm": 0.8791162221139119,
      "learning_rate": 1.1029411764705885e-05,
      "loss": 0.9399,
      "mean_token_accuracy": 0.7313582120538165,
      "step": 75
    },
    {
      "epoch": 0.05920444033302498,
      "grad_norm": 0.8984533054240379,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.8946,
      "mean_token_accuracy": 0.74295032091544,
      "step": 80
    },
    {
      "epoch": 0.06290471785383904,
      "grad_norm": 1.011416137474189,
      "learning_rate": 1.25e-05,
      "loss": 0.9146,
      "mean_token_accuracy": 0.7366250851820709,
      "step": 85
    },
    {
      "epoch": 0.0666049953746531,
      "grad_norm": 1.0421339267946321,
      "learning_rate": 1.323529411764706e-05,
      "loss": 0.8674,
      "mean_token_accuracy": 0.7498001647381672,
      "step": 90
    },
    {
      "epoch": 0.07030527289546716,
      "grad_norm": 0.9085816073653793,
      "learning_rate": 1.3970588235294118e-05,
      "loss": 0.8848,
      "mean_token_accuracy": 0.7459213455908199,
      "step": 95
    },
    {
      "epoch": 0.07400555041628122,
      "grad_norm": 1.108924753563469,
      "learning_rate": 1.4705882352941179e-05,
      "loss": 0.8984,
      "mean_token_accuracy": 0.7407395864040437,
      "step": 100
    },
    {
      "epoch": 0.07400555041628122,
      "eval_loss": 0.8967365026473999,
      "eval_mean_token_accuracy": 0.7362222154420599,
      "eval_runtime": 14.9927,
      "eval_samples_per_second": 17.142,
      "eval_steps_per_second": 1.134,
      "step": 100
    },
    {
      "epoch": 0.07770582793709528,
      "grad_norm": 1.0749851910600146,
      "learning_rate": 1.5441176470588237e-05,
      "loss": 0.9179,
      "mean_token_accuracy": 0.7357399681666867,
      "step": 105
    },
    {
      "epoch": 0.08140610545790934,
      "grad_norm": 0.9047444784181967,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 0.8645,
      "mean_token_accuracy": 0.7504314694269566,
      "step": 110
    },
    {
      "epoch": 0.0851063829787234,
      "grad_norm": 0.9376349243794684,
      "learning_rate": 1.6911764705882355e-05,
      "loss": 0.8529,
      "mean_token_accuracy": 0.7516550848858883,
      "step": 115
    },
    {
      "epoch": 0.08880666049953746,
      "grad_norm": 1.0917351629769148,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.8017,
      "mean_token_accuracy": 0.7668538841405976,
      "step": 120
    },
    {
      "epoch": 0.09250693802035152,
      "grad_norm": 1.001563932314443,
      "learning_rate": 1.8382352941176472e-05,
      "loss": 0.839,
      "mean_token_accuracy": 0.7564128590771355,
      "step": 125
    },
    {
      "epoch": 0.09620721554116558,
      "grad_norm": 1.0798631661399483,
      "learning_rate": 1.911764705882353e-05,
      "loss": 0.8884,
      "mean_token_accuracy": 0.7426265352132153,
      "step": 130
    },
    {
      "epoch": 0.09990749306197964,
      "grad_norm": 0.9968532290106067,
      "learning_rate": 1.985294117647059e-05,
      "loss": 0.8807,
      "mean_token_accuracy": 0.7442322917876578,
      "step": 135
    },
    {
      "epoch": 0.1036077705827937,
      "grad_norm": 1.109473932122476,
      "learning_rate": 1.9999465148392906e-05,
      "loss": 0.8846,
      "mean_token_accuracy": 0.74246846644443,
      "step": 140
    },
    {
      "epoch": 0.10730804810360776,
      "grad_norm": 0.9648937799207011,
      "learning_rate": 1.999729241179462e-05,
      "loss": 0.8574,
      "mean_token_accuracy": 0.7496809848232011,
      "step": 145
    },
    {
      "epoch": 0.11100832562442182,
      "grad_norm": 1.1187334021697934,
      "learning_rate": 1.999344872485215e-05,
      "loss": 0.887,
      "mean_token_accuracy": 0.7429769166906245,
      "step": 150
    },
    {
      "epoch": 0.1147086031452359,
      "grad_norm": 1.0683788241365297,
      "learning_rate": 1.9987934730000457e-05,
      "loss": 0.8419,
      "mean_token_accuracy": 0.7538007970944414,
      "step": 155
    },
    {
      "epoch": 0.11840888066604996,
      "grad_norm": 1.1320511934956428,
      "learning_rate": 1.998075134885022e-05,
      "loss": 0.867,
      "mean_token_accuracy": 0.7468512909716426,
      "step": 160
    },
    {
      "epoch": 0.12210915818686402,
      "grad_norm": 0.9103029614671497,
      "learning_rate": 1.9971899782033853e-05,
      "loss": 0.8576,
      "mean_token_accuracy": 0.7499938720204854,
      "step": 165
    },
    {
      "epoch": 0.12580943570767808,
      "grad_norm": 0.981129476759838,
      "learning_rate": 1.9961381509004785e-05,
      "loss": 0.8662,
      "mean_token_accuracy": 0.7473178017703339,
      "step": 170
    },
    {
      "epoch": 0.12950971322849214,
      "grad_norm": 1.0506640779724106,
      "learning_rate": 1.9949198287790215e-05,
      "loss": 0.8448,
      "mean_token_accuracy": 0.7532863939278541,
      "step": 175
    },
    {
      "epoch": 0.1332099907493062,
      "grad_norm": 1.026757070568888,
      "learning_rate": 1.9935352154697257e-05,
      "loss": 0.8289,
      "mean_token_accuracy": 0.7571774470525083,
      "step": 180
    },
    {
      "epoch": 0.13691026827012026,
      "grad_norm": 1.024410882605988,
      "learning_rate": 1.9919845423972603e-05,
      "loss": 0.8708,
      "mean_token_accuracy": 0.7462624656471927,
      "step": 185
    },
    {
      "epoch": 0.14061054579093432,
      "grad_norm": 0.9976772238865053,
      "learning_rate": 1.9902680687415704e-05,
      "loss": 0.8516,
      "mean_token_accuracy": 0.7509371514896652,
      "step": 190
    },
    {
      "epoch": 0.14431082331174838,
      "grad_norm": 0.9523302539052055,
      "learning_rate": 1.9883860813945596e-05,
      "loss": 0.866,
      "mean_token_accuracy": 0.7468851074446029,
      "step": 195
    },
    {
      "epoch": 0.14801110083256244,
      "grad_norm": 0.9850889047221103,
      "learning_rate": 1.986338894912137e-05,
      "loss": 0.8455,
      "mean_token_accuracy": 0.7527866267647105,
      "step": 200
    },
    {
      "epoch": 0.14801110083256244,
      "eval_loss": 0.865836501121521,
      "eval_mean_token_accuracy": 0.7421542283263433,
      "eval_runtime": 14.4283,
      "eval_samples_per_second": 17.812,
      "eval_steps_per_second": 1.178,
      "step": 200
    },
    {
      "epoch": 0.1517113783533765,
      "grad_norm": 1.0229873534868676,
      "learning_rate": 1.9841268514616434e-05,
      "loss": 0.8252,
      "mean_token_accuracy": 0.7574686598573471,
      "step": 205
    },
    {
      "epoch": 0.15541165587419056,
      "grad_norm": 0.9622467160372926,
      "learning_rate": 1.9817503207646606e-05,
      "loss": 0.8554,
      "mean_token_accuracy": 0.7478246659284545,
      "step": 210
    },
    {
      "epoch": 0.15911193339500462,
      "grad_norm": 0.9131276938848428,
      "learning_rate": 1.979209700035216e-05,
      "loss": 0.8244,
      "mean_token_accuracy": 0.7584659661867365,
      "step": 215
    },
    {
      "epoch": 0.16281221091581868,
      "grad_norm": 0.9051238256817666,
      "learning_rate": 1.976505413913393e-05,
      "loss": 0.8335,
      "mean_token_accuracy": 0.7552826906170925,
      "step": 220
    },
    {
      "epoch": 0.16651248843663274,
      "grad_norm": 0.8323380442965079,
      "learning_rate": 1.9736379143943565e-05,
      "loss": 0.8206,
      "mean_token_accuracy": 0.7582460617181218,
      "step": 225
    },
    {
      "epoch": 0.1702127659574468,
      "grad_norm": 0.9191267666766435,
      "learning_rate": 1.9706076807528044e-05,
      "loss": 0.8497,
      "mean_token_accuracy": 0.7505949453078798,
      "step": 230
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 0.8750491803678914,
      "learning_rate": 1.967415219462864e-05,
      "loss": 0.8676,
      "mean_token_accuracy": 0.7436590583163561,
      "step": 235
    },
    {
      "epoch": 0.17761332099907493,
      "grad_norm": 0.9239017879909354,
      "learning_rate": 1.9640610641134383e-05,
      "loss": 0.8559,
      "mean_token_accuracy": 0.7475432119302973,
      "step": 240
    },
    {
      "epoch": 0.18131359851988899,
      "grad_norm": 0.9375160912449586,
      "learning_rate": 1.9605457753190224e-05,
      "loss": 0.835,
      "mean_token_accuracy": 0.7547475756381888,
      "step": 245
    },
    {
      "epoch": 0.18501387604070305,
      "grad_norm": 0.9492613957957171,
      "learning_rate": 1.9568699406260016e-05,
      "loss": 0.8051,
      "mean_token_accuracy": 0.7635365375511847,
      "step": 250
    },
    {
      "epoch": 0.1887141535615171,
      "grad_norm": 0.8639060336030094,
      "learning_rate": 1.953034174414449e-05,
      "loss": 0.8424,
      "mean_token_accuracy": 0.7528160265907698,
      "step": 255
    },
    {
      "epoch": 0.19241443108233117,
      "grad_norm": 0.8984792755320175,
      "learning_rate": 1.9490391177954383e-05,
      "loss": 0.8386,
      "mean_token_accuracy": 0.7550477950939954,
      "step": 260
    },
    {
      "epoch": 0.19611470860314523,
      "grad_norm": 1.122121252111232,
      "learning_rate": 1.944885438503888e-05,
      "loss": 0.8063,
      "mean_token_accuracy": 0.7605705069134607,
      "step": 265
    },
    {
      "epoch": 0.1998149861239593,
      "grad_norm": 0.9283272086566748,
      "learning_rate": 1.9405738307869565e-05,
      "loss": 0.8276,
      "mean_token_accuracy": 0.7565203062763346,
      "step": 270
    },
    {
      "epoch": 0.20351526364477335,
      "grad_norm": 0.8413884737177622,
      "learning_rate": 1.936105015288003e-05,
      "loss": 0.8282,
      "mean_token_accuracy": 0.7568203022016651,
      "step": 275
    },
    {
      "epoch": 0.2072155411655874,
      "grad_norm": 0.9409390882210004,
      "learning_rate": 1.9314797389261426e-05,
      "loss": 0.8123,
      "mean_token_accuracy": 0.7606654417233719,
      "step": 280
    },
    {
      "epoch": 0.21091581868640147,
      "grad_norm": 0.9393490463149954,
      "learning_rate": 1.9266987747714036e-05,
      "loss": 0.8573,
      "mean_token_accuracy": 0.747553958645714,
      "step": 285
    },
    {
      "epoch": 0.21461609620721553,
      "grad_norm": 0.8638492423841437,
      "learning_rate": 1.9217629219155172e-05,
      "loss": 0.8123,
      "mean_token_accuracy": 0.7590153618433214,
      "step": 290
    },
    {
      "epoch": 0.2183163737280296,
      "grad_norm": 0.846309176173066,
      "learning_rate": 1.916673005338357e-05,
      "loss": 0.8056,
      "mean_token_accuracy": 0.7618184813128182,
      "step": 295
    },
    {
      "epoch": 0.22201665124884365,
      "grad_norm": 0.872871552134233,
      "learning_rate": 1.9114298757700508e-05,
      "loss": 0.8481,
      "mean_token_accuracy": 0.7489313630762335,
      "step": 300
    },
    {
      "epoch": 0.22201665124884365,
      "eval_loss": 0.8483777642250061,
      "eval_mean_token_accuracy": 0.7457049964976622,
      "eval_runtime": 14.427,
      "eval_samples_per_second": 17.814,
      "eval_steps_per_second": 1.178,
      "step": 300
    },
    {
      "epoch": 0.22571692876965774,
      "grad_norm": 0.999826315813207,
      "learning_rate": 1.9060344095487916e-05,
      "loss": 0.8166,
      "mean_token_accuracy": 0.758320414906213,
      "step": 305
    },
    {
      "epoch": 0.2294172062904718,
      "grad_norm": 0.8969904896726869,
      "learning_rate": 1.9004875084743624e-05,
      "loss": 0.8315,
      "mean_token_accuracy": 0.7541319645798021,
      "step": 310
    },
    {
      "epoch": 0.23311748381128586,
      "grad_norm": 0.8663093018774404,
      "learning_rate": 1.8947900996574133e-05,
      "loss": 0.8213,
      "mean_token_accuracy": 0.7579362435854416,
      "step": 315
    },
    {
      "epoch": 0.23681776133209992,
      "grad_norm": 0.9755265816894861,
      "learning_rate": 1.8889431353645004e-05,
      "loss": 0.8488,
      "mean_token_accuracy": 0.7504882625355576,
      "step": 320
    },
    {
      "epoch": 0.24051803885291398,
      "grad_norm": 0.9335300912517395,
      "learning_rate": 1.8829475928589272e-05,
      "loss": 0.827,
      "mean_token_accuracy": 0.7555998423977681,
      "step": 325
    },
    {
      "epoch": 0.24421831637372804,
      "grad_norm": 0.8900082915560992,
      "learning_rate": 1.8768044742374008e-05,
      "loss": 0.8152,
      "mean_token_accuracy": 0.7585311452224557,
      "step": 330
    },
    {
      "epoch": 0.2479185938945421,
      "grad_norm": 0.9294234165727436,
      "learning_rate": 1.870514806262544e-05,
      "loss": 0.8334,
      "mean_token_accuracy": 0.7540371079296321,
      "step": 335
    },
    {
      "epoch": 0.25161887141535616,
      "grad_norm": 0.8974570795671991,
      "learning_rate": 1.8640796401912805e-05,
      "loss": 0.8387,
      "mean_token_accuracy": 0.7533950740061741,
      "step": 340
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 0.9304337164615034,
      "learning_rate": 1.8575000515991283e-05,
      "loss": 0.8373,
      "mean_token_accuracy": 0.7522675141398661,
      "step": 345
    },
    {
      "epoch": 0.2590194264569843,
      "grad_norm": 0.9033200669469308,
      "learning_rate": 1.850777140200427e-05,
      "loss": 0.8244,
      "mean_token_accuracy": 0.7568203171895187,
      "step": 350
    },
    {
      "epoch": 0.2627197039777983,
      "grad_norm": 0.8758137392511858,
      "learning_rate": 1.843912029664531e-05,
      "loss": 0.7854,
      "mean_token_accuracy": 0.7673373370996341,
      "step": 355
    },
    {
      "epoch": 0.2664199814986124,
      "grad_norm": 0.8264158075854122,
      "learning_rate": 1.8369058674280004e-05,
      "loss": 0.8315,
      "mean_token_accuracy": 0.7537363388382118,
      "step": 360
    },
    {
      "epoch": 0.27012025901942643,
      "grad_norm": 0.9670401033713641,
      "learning_rate": 1.8297598245028173e-05,
      "loss": 0.8038,
      "mean_token_accuracy": 0.7621257946600963,
      "step": 365
    },
    {
      "epoch": 0.2738205365402405,
      "grad_norm": 0.80837274332651,
      "learning_rate": 1.8224750952806626e-05,
      "loss": 0.8099,
      "mean_token_accuracy": 0.7602013870948292,
      "step": 370
    },
    {
      "epoch": 0.27752081406105455,
      "grad_norm": 0.9804858790315341,
      "learning_rate": 1.815052897333284e-05,
      "loss": 0.8053,
      "mean_token_accuracy": 0.7617221638816465,
      "step": 375
    },
    {
      "epoch": 0.28122109158186864,
      "grad_norm": 0.9022919090042508,
      "learning_rate": 1.8074944712089925e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7580943305793408,
      "step": 380
    },
    {
      "epoch": 0.2849213691026827,
      "grad_norm": 0.8906726263514316,
      "learning_rate": 1.799801080225316e-05,
      "loss": 0.7982,
      "mean_token_accuracy": 0.7643851506636017,
      "step": 385
    },
    {
      "epoch": 0.28862164662349676,
      "grad_norm": 0.9254743961939594,
      "learning_rate": 1.7919740102578482e-05,
      "loss": 0.8303,
      "mean_token_accuracy": 0.7550160319572066,
      "step": 390
    },
    {
      "epoch": 0.2923219241443108,
      "grad_norm": 0.8866451111188923,
      "learning_rate": 1.7840145695253258e-05,
      "loss": 0.8289,
      "mean_token_accuracy": 0.7549396953188148,
      "step": 395
    },
    {
      "epoch": 0.2960222016651249,
      "grad_norm": 0.9181449537488579,
      "learning_rate": 1.7759240883709745e-05,
      "loss": 0.8081,
      "mean_token_accuracy": 0.7601195554230622,
      "step": 400
    },
    {
      "epoch": 0.2960222016651249,
      "eval_loss": 0.8367598056793213,
      "eval_mean_token_accuracy": 0.7479890842555498,
      "eval_runtime": 14.4256,
      "eval_samples_per_second": 17.816,
      "eval_steps_per_second": 1.178,
      "step": 400
    },
    {
      "epoch": 0.299722479185939,
      "grad_norm": 1.020831268230977,
      "learning_rate": 1.7677039190401538e-05,
      "loss": 0.8015,
      "mean_token_accuracy": 0.762128841557159,
      "step": 405
    },
    {
      "epoch": 0.303422756706753,
      "grad_norm": 0.9003231285941122,
      "learning_rate": 1.759355435454342e-05,
      "loss": 0.8058,
      "mean_token_accuracy": 0.7601393580378268,
      "step": 410
    },
    {
      "epoch": 0.3071230342275671,
      "grad_norm": 0.8236287029577125,
      "learning_rate": 1.7508800329814993e-05,
      "loss": 0.8184,
      "mean_token_accuracy": 0.7579837604334689,
      "step": 415
    },
    {
      "epoch": 0.3108233117483811,
      "grad_norm": 0.878406946421321,
      "learning_rate": 1.7422791282028457e-05,
      "loss": 0.8244,
      "mean_token_accuracy": 0.756998624229836,
      "step": 420
    },
    {
      "epoch": 0.3145235892691952,
      "grad_norm": 0.8740565941244159,
      "learning_rate": 1.7335541586760928e-05,
      "loss": 0.8093,
      "mean_token_accuracy": 0.758314911393715,
      "step": 425
    },
    {
      "epoch": 0.31822386679000925,
      "grad_norm": 0.8246101567782078,
      "learning_rate": 1.7247065826951694e-05,
      "loss": 0.841,
      "mean_token_accuracy": 0.7503554458795133,
      "step": 430
    },
    {
      "epoch": 0.32192414431082333,
      "grad_norm": 0.9006537182294639,
      "learning_rate": 1.715737879046483e-05,
      "loss": 0.8274,
      "mean_token_accuracy": 0.7544573421624281,
      "step": 435
    },
    {
      "epoch": 0.32562442183163737,
      "grad_norm": 0.8655847169317894,
      "learning_rate": 1.7066495467617552e-05,
      "loss": 0.7768,
      "mean_token_accuracy": 0.768584957125417,
      "step": 440
    },
    {
      "epoch": 0.32932469935245146,
      "grad_norm": 0.8873250103741515,
      "learning_rate": 1.6974431048674714e-05,
      "loss": 0.8102,
      "mean_token_accuracy": 0.7594749343726344,
      "step": 445
    },
    {
      "epoch": 0.3330249768732655,
      "grad_norm": 0.9413246521565958,
      "learning_rate": 1.6881200921309914e-05,
      "loss": 0.8249,
      "mean_token_accuracy": 0.7560174557216164,
      "step": 450
    },
    {
      "epoch": 0.3367252543940796,
      "grad_norm": 0.824311990821373,
      "learning_rate": 1.6786820668033596e-05,
      "loss": 0.783,
      "mean_token_accuracy": 0.7668897253243784,
      "step": 455
    },
    {
      "epoch": 0.3404255319148936,
      "grad_norm": 0.8487771854424017,
      "learning_rate": 1.6691306063588583e-05,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7621946857360491,
      "step": 460
    },
    {
      "epoch": 0.3441258094357077,
      "grad_norm": 0.8528899682952513,
      "learning_rate": 1.6594673072313478e-05,
      "loss": 0.8288,
      "mean_token_accuracy": 0.7542124144532096,
      "step": 465
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 0.8630488475790786,
      "learning_rate": 1.6496937845474375e-05,
      "loss": 0.8226,
      "mean_token_accuracy": 0.7555700287541637,
      "step": 470
    },
    {
      "epoch": 0.3515263644773358,
      "grad_norm": 0.8605658074115097,
      "learning_rate": 1.639811671856535e-05,
      "loss": 0.8236,
      "mean_token_accuracy": 0.7557499752346134,
      "step": 475
    },
    {
      "epoch": 0.35522664199814985,
      "grad_norm": 0.7797518506572444,
      "learning_rate": 1.6298226208578127e-05,
      "loss": 0.7813,
      "mean_token_accuracy": 0.7673481576245396,
      "step": 480
    },
    {
      "epoch": 0.35892691951896394,
      "grad_norm": 0.8968525161832575,
      "learning_rate": 1.6197283011241423e-05,
      "loss": 0.8144,
      "mean_token_accuracy": 0.7588428362993352,
      "step": 485
    },
    {
      "epoch": 0.36262719703977797,
      "grad_norm": 0.8655098146119614,
      "learning_rate": 1.6095303998230432e-05,
      "loss": 0.7944,
      "mean_token_accuracy": 0.7637976691011525,
      "step": 490
    },
    {
      "epoch": 0.36632747456059206,
      "grad_norm": 0.8535889063927236,
      "learning_rate": 1.599230621434687e-05,
      "loss": 0.812,
      "mean_token_accuracy": 0.7584837868241149,
      "step": 495
    },
    {
      "epoch": 0.3700277520814061,
      "grad_norm": 0.869018976065693,
      "learning_rate": 1.5888306874670112e-05,
      "loss": 0.7997,
      "mean_token_accuracy": 0.7605666985241383,
      "step": 500
    },
    {
      "epoch": 0.3700277520814061,
      "eval_loss": 0.8272697925567627,
      "eval_mean_token_accuracy": 0.7506664118007147,
      "eval_runtime": 14.4281,
      "eval_samples_per_second": 17.813,
      "eval_steps_per_second": 1.178,
      "step": 500
    },
    {
      "epoch": 0.3737280296022202,
      "grad_norm": 0.8518524507404158,
      "learning_rate": 1.5783323361679865e-05,
      "loss": 0.8168,
      "mean_token_accuracy": 0.757339361928324,
      "step": 505
    },
    {
      "epoch": 0.3774283071230342,
      "grad_norm": 0.826955116427965,
      "learning_rate": 1.567737322235084e-05,
      "loss": 0.7977,
      "mean_token_accuracy": 0.7626890473452755,
      "step": 510
    },
    {
      "epoch": 0.3811285846438483,
      "grad_norm": 0.81749324197106,
      "learning_rate": 1.557047416521996e-05,
      "loss": 0.8004,
      "mean_token_accuracy": 0.7616827532540068,
      "step": 515
    },
    {
      "epoch": 0.38482886216466233,
      "grad_norm": 0.8922348675427028,
      "learning_rate": 1.546264405742654e-05,
      "loss": 0.8158,
      "mean_token_accuracy": 0.7574806745169147,
      "step": 520
    },
    {
      "epoch": 0.3885291396854764,
      "grad_norm": 0.8314989282626728,
      "learning_rate": 1.535390092172597e-05,
      "loss": 0.8343,
      "mean_token_accuracy": 0.7518461075825386,
      "step": 525
    },
    {
      "epoch": 0.39222941720629045,
      "grad_norm": 0.9803006136249208,
      "learning_rate": 1.5244262933477401e-05,
      "loss": 0.8125,
      "mean_token_accuracy": 0.7575044146062361,
      "step": 530
    },
    {
      "epoch": 0.39592969472710454,
      "grad_norm": 0.8240339887315228,
      "learning_rate": 1.5133748417605878e-05,
      "loss": 0.8046,
      "mean_token_accuracy": 0.7602933891525141,
      "step": 535
    },
    {
      "epoch": 0.3996299722479186,
      "grad_norm": 1.0034971319001893,
      "learning_rate": 1.5022375845539537e-05,
      "loss": 0.8428,
      "mean_token_accuracy": 0.7486714904011393,
      "step": 540
    },
    {
      "epoch": 0.40333024976873266,
      "grad_norm": 0.8668996786648001,
      "learning_rate": 1.4910163832122278e-05,
      "loss": 0.8212,
      "mean_token_accuracy": 0.7548812709692821,
      "step": 545
    },
    {
      "epoch": 0.4070305272895467,
      "grad_norm": 0.9141308972741845,
      "learning_rate": 1.4797131132502464e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.770355880311961,
      "step": 550
    },
    {
      "epoch": 0.4107308048103608,
      "grad_norm": 0.8537164385394441,
      "learning_rate": 1.4683296638998192e-05,
      "loss": 0.7992,
      "mean_token_accuracy": 0.7629726221010971,
      "step": 555
    },
    {
      "epoch": 0.4144310823311748,
      "grad_norm": 0.8326818981936996,
      "learning_rate": 1.4568679377939619e-05,
      "loss": 0.7802,
      "mean_token_accuracy": 0.7676385973158337,
      "step": 560
    },
    {
      "epoch": 0.4181313598519889,
      "grad_norm": 0.832442552040938,
      "learning_rate": 1.4453298506488896e-05,
      "loss": 0.7913,
      "mean_token_accuracy": 0.7640136926495742,
      "step": 565
    },
    {
      "epoch": 0.42183163737280294,
      "grad_norm": 0.8484820193679721,
      "learning_rate": 1.4337173309438236e-05,
      "loss": 0.7775,
      "mean_token_accuracy": 0.767823896982353,
      "step": 570
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.8808404728533856,
      "learning_rate": 1.4220323195986649e-05,
      "loss": 0.8008,
      "mean_token_accuracy": 0.7603237102804046,
      "step": 575
    },
    {
      "epoch": 0.42923219241443106,
      "grad_norm": 0.8447018318022254,
      "learning_rate": 1.4102767696495885e-05,
      "loss": 0.8307,
      "mean_token_accuracy": 0.7518857396509806,
      "step": 580
    },
    {
      "epoch": 0.43293246993524515,
      "grad_norm": 0.8085246144584103,
      "learning_rate": 1.398452645922611e-05,
      "loss": 0.7829,
      "mean_token_accuracy": 0.7674445881359748,
      "step": 585
    },
    {
      "epoch": 0.4366327474560592,
      "grad_norm": 0.8736484311681693,
      "learning_rate": 1.3865619247051916e-05,
      "loss": 0.7944,
      "mean_token_accuracy": 0.7624168954027166,
      "step": 590
    },
    {
      "epoch": 0.44033302497687327,
      "grad_norm": 0.7989835433698877,
      "learning_rate": 1.3746065934159123e-05,
      "loss": 0.7934,
      "mean_token_accuracy": 0.7634149574109332,
      "step": 595
    },
    {
      "epoch": 0.4440333024976873,
      "grad_norm": 0.8032554489053177,
      "learning_rate": 1.3625886502723008e-05,
      "loss": 0.7576,
      "mean_token_accuracy": 0.7739669137787928,
      "step": 600
    },
    {
      "epoch": 0.4440333024976873,
      "eval_loss": 0.8197986483573914,
      "eval_mean_token_accuracy": 0.7513783085827505,
      "eval_runtime": 14.4284,
      "eval_samples_per_second": 17.812,
      "eval_steps_per_second": 1.178,
      "step": 600
    },
    {
      "epoch": 0.4477335800185014,
      "grad_norm": 0.9180541639485074,
      "learning_rate": 1.3505101039568494e-05,
      "loss": 0.8295,
      "mean_token_accuracy": 0.752840844934445,
      "step": 605
    },
    {
      "epoch": 0.4514338575393155,
      "grad_norm": 0.7752356138164175,
      "learning_rate": 1.3383729732812814e-05,
      "loss": 0.7879,
      "mean_token_accuracy": 0.764613230505008,
      "step": 610
    },
    {
      "epoch": 0.4551341350601295,
      "grad_norm": 0.83145675114526,
      "learning_rate": 1.3261792868491267e-05,
      "loss": 0.7914,
      "mean_token_accuracy": 0.7635481399350937,
      "step": 615
    },
    {
      "epoch": 0.4588344125809436,
      "grad_norm": 0.9055291804893332,
      "learning_rate": 1.3139310827166613e-05,
      "loss": 0.7701,
      "mean_token_accuracy": 0.7685765923501461,
      "step": 620
    },
    {
      "epoch": 0.46253469010175763,
      "grad_norm": 0.7854598351220646,
      "learning_rate": 1.3016304080522657e-05,
      "loss": 0.7704,
      "mean_token_accuracy": 0.7688856993881648,
      "step": 625
    },
    {
      "epoch": 0.4662349676225717,
      "grad_norm": 0.8441815485432693,
      "learning_rate": 1.2892793187942588e-05,
      "loss": 0.7626,
      "mean_token_accuracy": 0.7723868739695786,
      "step": 630
    },
    {
      "epoch": 0.46993524514338575,
      "grad_norm": 0.8667439139705619,
      "learning_rate": 1.2768798793072708e-05,
      "loss": 0.7833,
      "mean_token_accuracy": 0.7663687223264415,
      "step": 635
    },
    {
      "epoch": 0.47363552266419984,
      "grad_norm": 0.8321053356387224,
      "learning_rate": 1.2644341620372025e-05,
      "loss": 0.797,
      "mean_token_accuracy": 0.7617867821279096,
      "step": 640
    },
    {
      "epoch": 0.47733580018501387,
      "grad_norm": 0.7876516967194911,
      "learning_rate": 1.2519442471648364e-05,
      "loss": 0.7935,
      "mean_token_accuracy": 0.7634835072725441,
      "step": 645
    },
    {
      "epoch": 0.48103607770582796,
      "grad_norm": 0.7906296446679114,
      "learning_rate": 1.2394122222581557e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.7648497977323175,
      "step": 650
    },
    {
      "epoch": 0.484736355226642,
      "grad_norm": 0.8411500456259221,
      "learning_rate": 1.226840181923427e-05,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7703495787160083,
      "step": 655
    },
    {
      "epoch": 0.4884366327474561,
      "grad_norm": 0.8625525401373217,
      "learning_rate": 1.214230227455106e-05,
      "loss": 0.7905,
      "mean_token_accuracy": 0.7643257328337282,
      "step": 660
    },
    {
      "epoch": 0.4921369102682701,
      "grad_norm": 0.8769883186756258,
      "learning_rate": 1.201584466484629e-05,
      "loss": 0.8025,
      "mean_token_accuracy": 0.7592148887938102,
      "step": 665
    },
    {
      "epoch": 0.4958371877890842,
      "grad_norm": 0.7838112843020255,
      "learning_rate": 1.1889050126281405e-05,
      "loss": 0.7909,
      "mean_token_accuracy": 0.7630134287220484,
      "step": 670
    },
    {
      "epoch": 0.49953746530989823,
      "grad_norm": 0.874298908965393,
      "learning_rate": 1.1761939851332241e-05,
      "loss": 0.7691,
      "mean_token_accuracy": 0.7678435427558977,
      "step": 675
    },
    {
      "epoch": 0.5032377428307123,
      "grad_norm": 0.806343521404904,
      "learning_rate": 1.1634535085246903e-05,
      "loss": 0.7892,
      "mean_token_accuracy": 0.7629181254463294,
      "step": 680
    },
    {
      "epoch": 0.5069380203515264,
      "grad_norm": 0.8419778500876326,
      "learning_rate": 1.1506857122494832e-05,
      "loss": 0.8051,
      "mean_token_accuracy": 0.7606549427357923,
      "step": 685
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 0.7877466333281109,
      "learning_rate": 1.1378927303207637e-05,
      "loss": 0.7768,
      "mean_token_accuracy": 0.7675751056593663,
      "step": 690
    },
    {
      "epoch": 0.5143385753931545,
      "grad_norm": 0.9351990005133819,
      "learning_rate": 1.12507670096123e-05,
      "loss": 0.786,
      "mean_token_accuracy": 0.762621351731412,
      "step": 695
    },
    {
      "epoch": 0.5180388529139686,
      "grad_norm": 0.854162080782847,
      "learning_rate": 1.1122397662457352e-05,
      "loss": 0.7829,
      "mean_token_accuracy": 0.7657675443087595,
      "step": 700
    },
    {
      "epoch": 0.5180388529139686,
      "eval_loss": 0.8126487135887146,
      "eval_mean_token_accuracy": 0.7547239760153944,
      "eval_runtime": 14.4295,
      "eval_samples_per_second": 17.811,
      "eval_steps_per_second": 1.178,
      "step": 700
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 0.8645040092134876,
      "learning_rate": 1.0993840717432582e-05,
      "loss": 0.7931,
      "mean_token_accuracy": 0.7640120292612093,
      "step": 705
    },
    {
      "epoch": 0.5254394079555966,
      "grad_norm": 0.8306086563251239,
      "learning_rate": 1.0865117661582958e-05,
      "loss": 0.7908,
      "mean_token_accuracy": 0.7626317172664291,
      "step": 710
    },
    {
      "epoch": 0.5291396854764108,
      "grad_norm": 0.8020466355732573,
      "learning_rate": 1.0736250009717249e-05,
      "loss": 0.7725,
      "mean_token_accuracy": 0.7694951870682798,
      "step": 715
    },
    {
      "epoch": 0.5328399629972248,
      "grad_norm": 0.8902488653913394,
      "learning_rate": 1.0607259300812047e-05,
      "loss": 0.7956,
      "mean_token_accuracy": 0.7613192935638047,
      "step": 720
    },
    {
      "epoch": 0.5365402405180388,
      "grad_norm": 0.8214216473694839,
      "learning_rate": 1.0478167094411733e-05,
      "loss": 0.7813,
      "mean_token_accuracy": 0.7666191635211226,
      "step": 725
    },
    {
      "epoch": 0.5402405180388529,
      "grad_norm": 0.7826795279308211,
      "learning_rate": 1.0348994967025012e-05,
      "loss": 0.7904,
      "mean_token_accuracy": 0.7615527950542014,
      "step": 730
    },
    {
      "epoch": 0.543940795559667,
      "grad_norm": 0.7400827688733695,
      "learning_rate": 1.0219764508518595e-05,
      "loss": 0.813,
      "mean_token_accuracy": 0.7559395006050915,
      "step": 735
    },
    {
      "epoch": 0.547641073080481,
      "grad_norm": 0.8998834617524183,
      "learning_rate": 1.0090497318508687e-05,
      "loss": 0.8011,
      "mean_token_accuracy": 0.7619206875092797,
      "step": 740
    },
    {
      "epoch": 0.5513413506012951,
      "grad_norm": 0.7948918646757527,
      "learning_rate": 9.961215002750799e-06,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7669139188304434,
      "step": 745
    },
    {
      "epoch": 0.5550416281221091,
      "grad_norm": 0.8243901741332256,
      "learning_rate": 9.831939169528565e-06,
      "loss": 0.7877,
      "mean_token_accuracy": 0.7651489953053691,
      "step": 750
    },
    {
      "epoch": 0.5587419056429233,
      "grad_norm": 0.8525260298716054,
      "learning_rate": 9.702691426042124e-06,
      "loss": 0.8082,
      "mean_token_accuracy": 0.7574811962767495,
      "step": 755
    },
    {
      "epoch": 0.5624421831637373,
      "grad_norm": 0.8537478736826423,
      "learning_rate": 9.573493374796694e-06,
      "loss": 0.7921,
      "mean_token_accuracy": 0.7631372878452614,
      "step": 760
    },
    {
      "epoch": 0.5661424606845513,
      "grad_norm": 0.8430252803933753,
      "learning_rate": 9.444366609991916e-06,
      "loss": 0.7816,
      "mean_token_accuracy": 0.7655811713537104,
      "step": 765
    },
    {
      "epoch": 0.5698427382053654,
      "grad_norm": 0.7819283016691769,
      "learning_rate": 9.315332713912593e-06,
      "loss": 0.7766,
      "mean_token_accuracy": 0.7672953120755375,
      "step": 770
    },
    {
      "epoch": 0.5735430157261795,
      "grad_norm": 0.7854217248693685,
      "learning_rate": 9.18641325332142e-06,
      "loss": 0.7791,
      "mean_token_accuracy": 0.7654781090301902,
      "step": 775
    },
    {
      "epoch": 0.5772432932469935,
      "grad_norm": 0.7797472528150745,
      "learning_rate": 9.057629775854314e-06,
      "loss": 0.786,
      "mean_token_accuracy": 0.7640452228560917,
      "step": 780
    },
    {
      "epoch": 0.5809435707678076,
      "grad_norm": 0.7467612879682155,
      "learning_rate": 8.929003806418934e-06,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7636666656288214,
      "step": 785
    },
    {
      "epoch": 0.5846438482886216,
      "grad_norm": 0.8108378163356028,
      "learning_rate": 8.800556843597002e-06,
      "loss": 0.7947,
      "mean_token_accuracy": 0.7620531323988468,
      "step": 790
    },
    {
      "epoch": 0.5883441258094357,
      "grad_norm": 0.8711772849690956,
      "learning_rate": 8.672310356051023e-06,
      "loss": 0.7978,
      "mean_token_accuracy": 0.7624323429145154,
      "step": 795
    },
    {
      "epoch": 0.5920444033302498,
      "grad_norm": 0.8137827849359022,
      "learning_rate": 8.544285778936004e-06,
      "loss": 0.8179,
      "mean_token_accuracy": 0.7553667683649044,
      "step": 800
    },
    {
      "epoch": 0.5920444033302498,
      "eval_loss": 0.807556688785553,
      "eval_mean_token_accuracy": 0.7549181587193091,
      "eval_runtime": 14.4357,
      "eval_samples_per_second": 17.803,
      "eval_steps_per_second": 1.178,
      "step": 800
    },
    {
      "epoch": 0.5957446808510638,
      "grad_norm": 0.9030717079561702,
      "learning_rate": 8.416504510316774e-06,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7819614746996253,
      "step": 805
    },
    {
      "epoch": 0.599444958371878,
      "grad_norm": 0.8188514357400608,
      "learning_rate": 8.28898790759152e-06,
      "loss": 0.7799,
      "mean_token_accuracy": 0.7673006684255306,
      "step": 810
    },
    {
      "epoch": 0.603145235892692,
      "grad_norm": 0.8057726363169286,
      "learning_rate": 8.161757283922084e-06,
      "loss": 0.7727,
      "mean_token_accuracy": 0.7686252774437621,
      "step": 815
    },
    {
      "epoch": 0.606845513413506,
      "grad_norm": 0.8421037028882805,
      "learning_rate": 8.034833904671698e-06,
      "loss": 0.786,
      "mean_token_accuracy": 0.7641876212394639,
      "step": 820
    },
    {
      "epoch": 0.61054579093432,
      "grad_norm": 0.8334908295515484,
      "learning_rate": 7.908238983850666e-06,
      "loss": 0.7932,
      "mean_token_accuracy": 0.7616740275010933,
      "step": 825
    },
    {
      "epoch": 0.6142460684551342,
      "grad_norm": 0.7732913249201244,
      "learning_rate": 7.781993680570656e-06,
      "loss": 0.8014,
      "mean_token_accuracy": 0.7606977785932187,
      "step": 830
    },
    {
      "epoch": 0.6179463459759482,
      "grad_norm": 0.8529863935049032,
      "learning_rate": 7.656119095508155e-06,
      "loss": 0.761,
      "mean_token_accuracy": 0.7720907621215842,
      "step": 835
    },
    {
      "epoch": 0.6216466234967623,
      "grad_norm": 0.7790046842614121,
      "learning_rate": 7.530636267377706e-06,
      "loss": 0.7634,
      "mean_token_accuracy": 0.7701395597098513,
      "step": 840
    },
    {
      "epoch": 0.6253469010175763,
      "grad_norm": 0.8225616793183522,
      "learning_rate": 7.405566169415481e-06,
      "loss": 0.7736,
      "mean_token_accuracy": 0.7693085459512584,
      "step": 845
    },
    {
      "epoch": 0.6290471785383904,
      "grad_norm": 0.7995615588703834,
      "learning_rate": 7.280929705873818e-06,
      "loss": 0.8007,
      "mean_token_accuracy": 0.7614276357940275,
      "step": 850
    },
    {
      "epoch": 0.6327474560592045,
      "grad_norm": 0.7568937242521389,
      "learning_rate": 7.15674770852727e-06,
      "loss": 0.7358,
      "mean_token_accuracy": 0.7783971531791127,
      "step": 855
    },
    {
      "epoch": 0.6364477335800185,
      "grad_norm": 0.7756432966734992,
      "learning_rate": 7.033040933190776e-06,
      "loss": 0.803,
      "mean_token_accuracy": 0.7595857140786891,
      "step": 860
    },
    {
      "epoch": 0.6401480111008325,
      "grad_norm": 0.8218979611135344,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.7839,
      "mean_token_accuracy": 0.7639222007393546,
      "step": 865
    },
    {
      "epoch": 0.6438482886216467,
      "grad_norm": 0.7688613456202562,
      "learning_rate": 6.787135671208126e-06,
      "loss": 0.7572,
      "mean_token_accuracy": 0.7727190284700474,
      "step": 870
    },
    {
      "epoch": 0.6475485661424607,
      "grad_norm": 0.7591858458864597,
      "learning_rate": 6.6649782852385554e-06,
      "loss": 0.7714,
      "mean_token_accuracy": 0.7696598108468179,
      "step": 875
    },
    {
      "epoch": 0.6512488436632747,
      "grad_norm": 0.7777825640170363,
      "learning_rate": 6.543378315762634e-06,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7740491454057086,
      "step": 880
    },
    {
      "epoch": 0.6549491211840888,
      "grad_norm": 0.7917932009302419,
      "learning_rate": 6.42235608703441e-06,
      "loss": 0.7971,
      "mean_token_accuracy": 0.7617020404301962,
      "step": 885
    },
    {
      "epoch": 0.6586493987049029,
      "grad_norm": 0.7691064861967819,
      "learning_rate": 6.301931826744189e-06,
      "loss": 0.7739,
      "mean_token_accuracy": 0.7681618066719192,
      "step": 890
    },
    {
      "epoch": 0.6623496762257169,
      "grad_norm": 0.7996344447354707,
      "learning_rate": 6.18212566263765e-06,
      "loss": 0.7856,
      "mean_token_accuracy": 0.7637340466330773,
      "step": 895
    },
    {
      "epoch": 0.666049953746531,
      "grad_norm": 0.8226708174097512,
      "learning_rate": 6.0629576191517035e-06,
      "loss": 0.7633,
      "mean_token_accuracy": 0.7714390934857177,
      "step": 900
    },
    {
      "epoch": 0.666049953746531,
      "eval_loss": 0.8014763593673706,
      "eval_mean_token_accuracy": 0.7567653930021836,
      "eval_runtime": 14.4293,
      "eval_samples_per_second": 17.811,
      "eval_steps_per_second": 1.178,
      "step": 900
    },
    {
      "epoch": 0.669750231267345,
      "grad_norm": 0.8005238604002097,
      "learning_rate": 5.944447614067588e-06,
      "loss": 0.7575,
      "mean_token_accuracy": 0.771954640275627,
      "step": 905
    },
    {
      "epoch": 0.6734505087881592,
      "grad_norm": 0.8084142017934829,
      "learning_rate": 5.8266154551818225e-06,
      "loss": 0.7384,
      "mean_token_accuracy": 0.777882152510502,
      "step": 910
    },
    {
      "epoch": 0.6771507863089732,
      "grad_norm": 0.7778625881879635,
      "learning_rate": 5.709480836995509e-06,
      "loss": 0.7701,
      "mean_token_accuracy": 0.7698965224005008,
      "step": 915
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.7577955469774116,
      "learning_rate": 5.593063337422595e-06,
      "loss": 0.7805,
      "mean_token_accuracy": 0.765536335028292,
      "step": 920
    },
    {
      "epoch": 0.6845513413506013,
      "grad_norm": 0.9227266710988027,
      "learning_rate": 5.477382414517625e-06,
      "loss": 0.798,
      "mean_token_accuracy": 0.7598567457034476,
      "step": 925
    },
    {
      "epoch": 0.6882516188714154,
      "grad_norm": 0.8508290504613905,
      "learning_rate": 5.362457403223495e-06,
      "loss": 0.7884,
      "mean_token_accuracy": 0.7633147515282703,
      "step": 930
    },
    {
      "epoch": 0.6919518963922294,
      "grad_norm": 0.789919298542449,
      "learning_rate": 5.248307512139818e-06,
      "loss": 0.7862,
      "mean_token_accuracy": 0.7651956223578331,
      "step": 935
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 0.775028897496873,
      "learning_rate": 5.134951820312402e-06,
      "loss": 0.7608,
      "mean_token_accuracy": 0.7710794013752429,
      "step": 940
    },
    {
      "epoch": 0.6993524514338575,
      "grad_norm": 0.8430703919544058,
      "learning_rate": 5.022409274044346e-06,
      "loss": 0.7915,
      "mean_token_accuracy": 0.7635476913138959,
      "step": 945
    },
    {
      "epoch": 0.7030527289546716,
      "grad_norm": 0.8044082255650751,
      "learning_rate": 4.910698683729371e-06,
      "loss": 0.7638,
      "mean_token_accuracy": 0.7705783943727836,
      "step": 950
    },
    {
      "epoch": 0.7067530064754857,
      "grad_norm": 0.7439402505547626,
      "learning_rate": 4.799838720707847e-06,
      "loss": 0.7835,
      "mean_token_accuracy": 0.76646702805682,
      "step": 955
    },
    {
      "epoch": 0.7104532839962997,
      "grad_norm": 0.8141310220719153,
      "learning_rate": 4.6898479141460415e-06,
      "loss": 0.7583,
      "mean_token_accuracy": 0.7721620830045813,
      "step": 960
    },
    {
      "epoch": 0.7141535615171137,
      "grad_norm": 0.7863013724531347,
      "learning_rate": 4.580744647939163e-06,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7700705090647101,
      "step": 965
    },
    {
      "epoch": 0.7178538390379279,
      "grad_norm": 0.7919610069204563,
      "learning_rate": 4.472547157638674e-06,
      "loss": 0.7756,
      "mean_token_accuracy": 0.7670807847229706,
      "step": 970
    },
    {
      "epoch": 0.7215541165587419,
      "grad_norm": 0.8241671907963101,
      "learning_rate": 4.365273527404384e-06,
      "loss": 0.7595,
      "mean_token_accuracy": 0.7707993119384532,
      "step": 975
    },
    {
      "epoch": 0.7252543940795559,
      "grad_norm": 0.7638723992566342,
      "learning_rate": 4.258941686981864e-06,
      "loss": 0.7529,
      "mean_token_accuracy": 0.7727072913006545,
      "step": 980
    },
    {
      "epoch": 0.72895467160037,
      "grad_norm": 0.7934309413562728,
      "learning_rate": 4.15356940870567e-06,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7670910625057229,
      "step": 985
    },
    {
      "epoch": 0.7326549491211841,
      "grad_norm": 0.8485940316250707,
      "learning_rate": 4.049174304528857e-06,
      "loss": 0.802,
      "mean_token_accuracy": 0.7588791241905861,
      "step": 990
    },
    {
      "epoch": 0.7363552266419982,
      "grad_norm": 0.8810207610120062,
      "learning_rate": 3.945773823079315e-06,
      "loss": 0.7595,
      "mean_token_accuracy": 0.7711046119500352,
      "step": 995
    },
    {
      "epoch": 0.7400555041628122,
      "grad_norm": 0.833110497686039,
      "learning_rate": 3.8433852467434175e-06,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7675894720358956,
      "step": 1000
    },
    {
      "epoch": 0.7400555041628122,
      "eval_loss": 0.7973992824554443,
      "eval_mean_token_accuracy": 0.757112345312609,
      "eval_runtime": 14.427,
      "eval_samples_per_second": 17.814,
      "eval_steps_per_second": 1.178,
      "step": 1000
    },
    {
      "epoch": 0.7437557816836263,
      "grad_norm": 0.8237006050954356,
      "learning_rate": 3.742025688777413e-06,
      "loss": 0.8037,
      "mean_token_accuracy": 0.7586444848081989,
      "step": 1005
    },
    {
      "epoch": 0.7474560592044404,
      "grad_norm": 0.7526823694442858,
      "learning_rate": 3.641712090447125e-06,
      "loss": 0.7716,
      "mean_token_accuracy": 0.7684351614558499,
      "step": 1010
    },
    {
      "epoch": 0.7511563367252544,
      "grad_norm": 0.7943312700833844,
      "learning_rate": 3.542461218196379e-06,
      "loss": 0.731,
      "mean_token_accuracy": 0.7802367124302477,
      "step": 1015
    },
    {
      "epoch": 0.7548566142460684,
      "grad_norm": 0.7798186842393364,
      "learning_rate": 3.444289660844665e-06,
      "loss": 0.8063,
      "mean_token_accuracy": 0.7586571868727487,
      "step": 1020
    },
    {
      "epoch": 0.7585568917668826,
      "grad_norm": 0.765497355738645,
      "learning_rate": 3.347213826814456e-06,
      "loss": 0.7583,
      "mean_token_accuracy": 0.7724547572887597,
      "step": 1025
    },
    {
      "epoch": 0.7622571692876966,
      "grad_norm": 0.7857443566387864,
      "learning_rate": 3.2512499413887255e-06,
      "loss": 0.7543,
      "mean_token_accuracy": 0.7736901779455366,
      "step": 1030
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.802530120535559,
      "learning_rate": 3.1564140439990256e-06,
      "loss": 0.7567,
      "mean_token_accuracy": 0.7718877631758259,
      "step": 1035
    },
    {
      "epoch": 0.7696577243293247,
      "grad_norm": 0.7989942251381315,
      "learning_rate": 3.0627219855446667e-06,
      "loss": 0.7661,
      "mean_token_accuracy": 0.7682660057488334,
      "step": 1040
    },
    {
      "epoch": 0.7733580018501388,
      "grad_norm": 0.840504807173006,
      "learning_rate": 2.970189425743383e-06,
      "loss": 0.7551,
      "mean_token_accuracy": 0.7720390981661918,
      "step": 1045
    },
    {
      "epoch": 0.7770582793709528,
      "grad_norm": 0.7779487830464692,
      "learning_rate": 2.8788318305139808e-06,
      "loss": 0.7655,
      "mean_token_accuracy": 0.7693840268148309,
      "step": 1050
    },
    {
      "epoch": 0.7807585568917669,
      "grad_norm": 0.7927243683319699,
      "learning_rate": 2.7886644693913333e-06,
      "loss": 0.7743,
      "mean_token_accuracy": 0.7681611530624838,
      "step": 1055
    },
    {
      "epoch": 0.7844588344125809,
      "grad_norm": 0.8373582713281018,
      "learning_rate": 2.6997024129742544e-06,
      "loss": 0.7785,
      "mean_token_accuracy": 0.766058403299143,
      "step": 1060
    },
    {
      "epoch": 0.788159111933395,
      "grad_norm": 0.8441970949667804,
      "learning_rate": 2.611960530406572e-06,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7704607514620688,
      "step": 1065
    },
    {
      "epoch": 0.7918593894542091,
      "grad_norm": 0.7819472527987796,
      "learning_rate": 2.5254534868919077e-06,
      "loss": 0.745,
      "mean_token_accuracy": 0.7760465989840439,
      "step": 1070
    },
    {
      "epoch": 0.7955596669750231,
      "grad_norm": 0.7781088934347913,
      "learning_rate": 2.4401957412425213e-06,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7700818907749126,
      "step": 1075
    },
    {
      "epoch": 0.7992599444958371,
      "grad_norm": 0.8191690736963815,
      "learning_rate": 2.3562015434626784e-06,
      "loss": 0.7427,
      "mean_token_accuracy": 0.7759877209976991,
      "step": 1080
    },
    {
      "epoch": 0.8029602220166513,
      "grad_norm": 0.7935070459901263,
      "learning_rate": 2.273484932366874e-06,
      "loss": 0.8047,
      "mean_token_accuracy": 0.7583013868934,
      "step": 1085
    },
    {
      "epoch": 0.8066604995374653,
      "grad_norm": 0.7331859606400628,
      "learning_rate": 2.192059733233408e-06,
      "loss": 0.8008,
      "mean_token_accuracy": 0.7600684615036566,
      "step": 1090
    },
    {
      "epoch": 0.8103607770582794,
      "grad_norm": 0.8074434856892805,
      "learning_rate": 2.111939555493603e-06,
      "loss": 0.7794,
      "mean_token_accuracy": 0.7657683649247553,
      "step": 1095
    },
    {
      "epoch": 0.8140610545790934,
      "grad_norm": 0.8086682573035457,
      "learning_rate": 2.0331377904571303e-06,
      "loss": 0.7586,
      "mean_token_accuracy": 0.7709054565375336,
      "step": 1100
    },
    {
      "epoch": 0.8140610545790934,
      "eval_loss": 0.7944868803024292,
      "eval_mean_token_accuracy": 0.7578580899604019,
      "eval_runtime": 14.41,
      "eval_samples_per_second": 17.835,
      "eval_steps_per_second": 1.18,
      "step": 1100
    },
    {
      "epoch": 0.8177613320999075,
      "grad_norm": 0.8343820210266487,
      "learning_rate": 1.9556676090737803e-06,
      "loss": 0.7709,
      "mean_token_accuracy": 0.7677138846863649,
      "step": 1105
    },
    {
      "epoch": 0.8214616096207216,
      "grad_norm": 0.8274108362888261,
      "learning_rate": 1.879541959732072e-06,
      "loss": 0.772,
      "mean_token_accuracy": 0.767794811632847,
      "step": 1110
    },
    {
      "epoch": 0.8251618871415356,
      "grad_norm": 0.7986915851539247,
      "learning_rate": 1.8047735660950427e-06,
      "loss": 0.7824,
      "mean_token_accuracy": 0.7641448767963437,
      "step": 1115
    },
    {
      "epoch": 0.8288621646623496,
      "grad_norm": 0.7912930392931041,
      "learning_rate": 1.7313749249736266e-06,
      "loss": 0.7722,
      "mean_token_accuracy": 0.7688348884744852,
      "step": 1120
    },
    {
      "epoch": 0.8325624421831638,
      "grad_norm": 0.7998408807770482,
      "learning_rate": 1.6593583042379192e-06,
      "loss": 0.7777,
      "mean_token_accuracy": 0.7664428074095984,
      "step": 1125
    },
    {
      "epoch": 0.8362627197039778,
      "grad_norm": 0.7582195922034395,
      "learning_rate": 1.5887357407667314e-06,
      "loss": 0.7635,
      "mean_token_accuracy": 0.7685496955885053,
      "step": 1130
    },
    {
      "epoch": 0.8399629972247918,
      "grad_norm": 0.7846551733743673,
      "learning_rate": 1.5195190384357405e-06,
      "loss": 0.7623,
      "mean_token_accuracy": 0.7700296736101973,
      "step": 1135
    },
    {
      "epoch": 0.8436632747456059,
      "grad_norm": 0.8102544575623821,
      "learning_rate": 1.4517197661445893e-06,
      "loss": 0.8151,
      "mean_token_accuracy": 0.7554465880686968,
      "step": 1140
    },
    {
      "epoch": 0.84736355226642,
      "grad_norm": 0.7915517548834473,
      "learning_rate": 1.3853492558832472e-06,
      "loss": 0.8012,
      "mean_token_accuracy": 0.759899401391139,
      "step": 1145
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 0.771064167787231,
      "learning_rate": 1.3204186008379926e-06,
      "loss": 0.7466,
      "mean_token_accuracy": 0.7737188447666707,
      "step": 1150
    },
    {
      "epoch": 0.8547641073080481,
      "grad_norm": 0.7634680191740536,
      "learning_rate": 1.2569386535372807e-06,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7721920358524959,
      "step": 1155
    },
    {
      "epoch": 0.8584643848288621,
      "grad_norm": 0.7702642560131528,
      "learning_rate": 1.1949200240378577e-06,
      "loss": 0.7681,
      "mean_token_accuracy": 0.7690720719541746,
      "step": 1160
    },
    {
      "epoch": 0.8621646623496763,
      "grad_norm": 0.7798114837698836,
      "learning_rate": 1.1343730781513896e-06,
      "loss": 0.747,
      "mean_token_accuracy": 0.774875145015143,
      "step": 1165
    },
    {
      "epoch": 0.8658649398704903,
      "grad_norm": 0.7949938733034089,
      "learning_rate": 1.0753079357119134e-06,
      "loss": 0.7849,
      "mean_token_accuracy": 0.765570759792907,
      "step": 1170
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.7698843089224456,
      "learning_rate": 1.017734468884417e-06,
      "loss": 0.7242,
      "mean_token_accuracy": 0.7807185988785081,
      "step": 1175
    },
    {
      "epoch": 0.8732654949121184,
      "grad_norm": 0.7280858038390325,
      "learning_rate": 9.616623005147952e-07,
      "loss": 0.7586,
      "mean_token_accuracy": 0.7718187615798253,
      "step": 1180
    },
    {
      "epoch": 0.8769657724329325,
      "grad_norm": 0.839720555473333,
      "learning_rate": 9.071008025214767e-07,
      "loss": 0.7946,
      "mean_token_accuracy": 0.7638036068173906,
      "step": 1185
    },
    {
      "epoch": 0.8806660499537465,
      "grad_norm": 0.7555588990753439,
      "learning_rate": 8.540590943290128e-07,
      "loss": 0.7571,
      "mean_token_accuracy": 0.7721729084932626,
      "step": 1190
    },
    {
      "epoch": 0.8843663274745606,
      "grad_norm": 0.7969942371823111,
      "learning_rate": 8.025460413438457e-07,
      "loss": 0.8112,
      "mean_token_accuracy": 0.75678643764634,
      "step": 1195
    },
    {
      "epoch": 0.8880666049953746,
      "grad_norm": 0.7364947816684886,
      "learning_rate": 7.525702534725443e-07,
      "loss": 0.7528,
      "mean_token_accuracy": 0.7727457500345997,
      "step": 1200
    },
    {
      "epoch": 0.8880666049953746,
      "eval_loss": 0.7930368185043335,
      "eval_mean_token_accuracy": 0.7581294874020488,
      "eval_runtime": 23.741,
      "eval_samples_per_second": 10.825,
      "eval_steps_per_second": 0.716,
      "step": 1200
    },
    {
      "epoch": 0.8917668825161887,
      "grad_norm": 0.8056054421008482,
      "learning_rate": 7.041400836827439e-07,
      "loss": 0.7526,
      "mean_token_accuracy": 0.7748112663916402,
      "step": 1205
    },
    {
      "epoch": 0.8954671600370028,
      "grad_norm": 0.7430784621258909,
      "learning_rate": 6.572636266070265e-07,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7691522840115936,
      "step": 1210
    },
    {
      "epoch": 0.8991674375578168,
      "grad_norm": 0.7294844027936921,
      "learning_rate": 6.119487171899807e-07,
      "loss": 0.7605,
      "mean_token_accuracy": 0.7704982028684197,
      "step": 1215
    },
    {
      "epoch": 0.902867715078631,
      "grad_norm": 0.8002640649815743,
      "learning_rate": 5.682029293786673e-07,
      "loss": 0.7926,
      "mean_token_accuracy": 0.7625801633473456,
      "step": 1220
    },
    {
      "epoch": 0.906567992599445,
      "grad_norm": 0.7836990755620872,
      "learning_rate": 5.26033574856708e-07,
      "loss": 0.7672,
      "mean_token_accuracy": 0.7700011696772932,
      "step": 1225
    },
    {
      "epoch": 0.910268270120259,
      "grad_norm": 0.8159687086578604,
      "learning_rate": 4.854477018222103e-07,
      "loss": 0.8008,
      "mean_token_accuracy": 0.7607755917340568,
      "step": 1230
    },
    {
      "epoch": 0.913968547641073,
      "grad_norm": 0.7645217245522713,
      "learning_rate": 4.464520938097294e-07,
      "loss": 0.7788,
      "mean_token_accuracy": 0.7658377209811824,
      "step": 1235
    },
    {
      "epoch": 0.9176688251618872,
      "grad_norm": 0.7673840812542029,
      "learning_rate": 4.0905326855646186e-07,
      "loss": 0.7573,
      "mean_token_accuracy": 0.7721266403991972,
      "step": 1240
    },
    {
      "epoch": 0.9213691026827012,
      "grad_norm": 0.7454299232676801,
      "learning_rate": 3.732574769128738e-07,
      "loss": 0.7659,
      "mean_token_accuracy": 0.7694314927154986,
      "step": 1245
    },
    {
      "epoch": 0.9250693802035153,
      "grad_norm": 0.7654595178316245,
      "learning_rate": 3.390707017979311e-07,
      "loss": 0.745,
      "mean_token_accuracy": 0.7750965684798319,
      "step": 1250
    },
    {
      "epoch": 0.9287696577243293,
      "grad_norm": 0.8067427671781613,
      "learning_rate": 3.06498657199108e-07,
      "loss": 0.7732,
      "mean_token_accuracy": 0.768182908414515,
      "step": 1255
    },
    {
      "epoch": 0.9324699352451434,
      "grad_norm": 0.7562039729125304,
      "learning_rate": 2.7554678721735675e-07,
      "loss": 0.7787,
      "mean_token_accuracy": 0.7642553045991999,
      "step": 1260
    },
    {
      "epoch": 0.9361702127659575,
      "grad_norm": 0.7430603594511384,
      "learning_rate": 2.4622026515717654e-07,
      "loss": 0.8077,
      "mean_token_accuracy": 0.7573527261389765,
      "step": 1265
    },
    {
      "epoch": 0.9398704902867715,
      "grad_norm": 0.7593630984334607,
      "learning_rate": 2.1852399266194312e-07,
      "loss": 0.7793,
      "mean_token_accuracy": 0.764669397435321,
      "step": 1270
    },
    {
      "epoch": 0.9435707678075855,
      "grad_norm": 0.7357935865065981,
      "learning_rate": 1.9246259889464935e-07,
      "loss": 0.7781,
      "mean_token_accuracy": 0.7665922229483206,
      "step": 1275
    },
    {
      "epoch": 0.9472710453283997,
      "grad_norm": 0.7288331937335505,
      "learning_rate": 1.6804043976418438e-07,
      "loss": 0.7704,
      "mean_token_accuracy": 0.7690436128045272,
      "step": 1280
    },
    {
      "epoch": 0.9509713228492137,
      "grad_norm": 0.7843857219453947,
      "learning_rate": 1.4526159719728595e-07,
      "loss": 0.7768,
      "mean_token_accuracy": 0.7667279502666673,
      "step": 1285
    },
    {
      "epoch": 0.9546716003700277,
      "grad_norm": 0.8205041527973095,
      "learning_rate": 1.24129878456285e-07,
      "loss": 0.808,
      "mean_token_accuracy": 0.759965865777703,
      "step": 1290
    },
    {
      "epoch": 0.9583718778908418,
      "grad_norm": 0.7786180946366419,
      "learning_rate": 1.0464881550276362e-07,
      "loss": 0.7656,
      "mean_token_accuracy": 0.7689572567793539,
      "step": 1295
    },
    {
      "epoch": 0.9620721554116559,
      "grad_norm": 0.8324732784427329,
      "learning_rate": 8.682166440721729e-08,
      "loss": 0.7554,
      "mean_token_accuracy": 0.7726877638582714,
      "step": 1300
    },
    {
      "epoch": 0.9620721554116559,
      "eval_loss": 0.7924448251724243,
      "eval_mean_token_accuracy": 0.7585140244194415,
      "eval_runtime": 14.4043,
      "eval_samples_per_second": 17.842,
      "eval_steps_per_second": 1.18,
      "step": 1300
    },
    {
      "epoch": 0.96577243293247,
      "grad_norm": 0.8003294805368201,
      "learning_rate": 7.065140480483235e-08,
      "loss": 0.7953,
      "mean_token_accuracy": 0.7616640102372079,
      "step": 1305
    },
    {
      "epoch": 0.969472710453284,
      "grad_norm": 0.8056079696464618,
      "learning_rate": 5.6140739397474445e-08,
      "loss": 0.7998,
      "mean_token_accuracy": 0.7604724691954434,
      "step": 1310
    },
    {
      "epoch": 0.973172987974098,
      "grad_norm": 0.7398525398002953,
      "learning_rate": 4.329209350195651e-08,
      "loss": 0.7687,
      "mean_token_accuracy": 0.7694833489373637,
      "step": 1315
    },
    {
      "epoch": 0.9768732654949122,
      "grad_norm": 0.767475456178718,
      "learning_rate": 3.210761464466639e-08,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7690996438734377,
      "step": 1320
    },
    {
      "epoch": 0.9805735430157262,
      "grad_norm": 0.804536573614064,
      "learning_rate": 2.2589172202635014e-08,
      "loss": 0.7773,
      "mean_token_accuracy": 0.7664188170808871,
      "step": 1325
    },
    {
      "epoch": 0.9842738205365402,
      "grad_norm": 0.8454088605208244,
      "learning_rate": 1.4738357091084177e-08,
      "loss": 0.768,
      "mean_token_accuracy": 0.7688064486234464,
      "step": 1330
    },
    {
      "epoch": 0.9879740980573543,
      "grad_norm": 0.8321469884127546,
      "learning_rate": 8.556481497521418e-09,
      "loss": 0.764,
      "mean_token_accuracy": 0.7714298871126578,
      "step": 1335
    },
    {
      "epoch": 0.9916743755781684,
      "grad_norm": 0.8508534520033296,
      "learning_rate": 4.044578662419918e-09,
      "loss": 0.7853,
      "mean_token_accuracy": 0.7642099197772693,
      "step": 1340
    },
    {
      "epoch": 0.9953746530989824,
      "grad_norm": 0.7643692450259899,
      "learning_rate": 1.203402706525525e-09,
      "loss": 0.7728,
      "mean_token_accuracy": 0.7677490371223821,
      "step": 1345
    },
    {
      "epoch": 0.9990749306197965,
      "grad_norm": 0.7766881531885754,
      "learning_rate": 3.342850480869686e-11,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7655295798263551,
      "step": 1350
    },
    {
      "epoch": 0.9998149861239592,
      "mean_token_accuracy": 0.7588118329384822,
      "step": 1351,
      "total_flos": 76959556042752.0,
      "train_loss": 0.8100585834261755,
      "train_runtime": 5700.9233,
      "train_samples_per_second": 7.585,
      "train_steps_per_second": 0.237
    }
  ],
  "logging_steps": 5,
  "max_steps": 1351,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 76959556042752.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}